The following is something I found from October 7, 2001:
I am haunted by hyperspheres of signals and mystically encoded
metasignals--all made out of words--made out of bits. From every thought I
have, I can decode the metasignal of Wallace Steven's "Men Made Out Of
Words" if I just write it down and factor out its implicit propositions.
What should we be without the sexual myth,
The human reverie or poem of death?
Castratos of moon-mash - Life consists
Of propositions about life. The human
Reverie is a solitude in which
We compose these propositions, torn by dreams,
By the terrible incantations of defeats
And by the fear that defeats and dreams are one.
The whole race is a poet that writes down
The eccentric propositions of its fate.
Shannon when info theory was still on the drawing boards, ignored Donald
MacKay and his contextual metacommunication. He took human information and
divorced it from meaning. In separating information from meaning Shannon
gave us a hypersphere:
E=x1 ssquared+ x2 squared + x3 squared...
Shannon took a signal, broke it into samples which defined his
hypershpere--the object which information theory and the information society
are founded on. To do so however, Shannon had to build a generator of
English text or at least a statistical approximation of one. Shannon
generated approximations of English text through the consultation of tables
of monogram and digram letter frequencies. When these tables proved
insufficient (as they were for higher-order simulations), he consulted books
from his library. His simulations included zero-order, first-order,
second-order, and third-order by letter. He also conducted first-order and
second-order simulations by word. Due to the labor involved, he stopped at
this point. Here are his 0-3rd order approximations:
XFOML RXHRJFFJUJ ZLPWCFWKCYJ FFJEYVKCQSGXYD QPAAMKBZAACIBZL
OCRO HLI RGWR NMIELWIS EU LL NBNESEBYA TH EEI ALHENHTT OOBTTVA NAH BRL
ON IE ANTSOUTINYS ARE T INCTORE ST BE S DEAMY ACHIN DILONASIVE TUCOOWE AT
TEASONARE FUSO TIZIN ANDY TOBE SEACE CTISBE
IN NO IST LAT WHEY CRATICT FROURE BIRS GROCID PONDENON OF DEMONSTURES OF THE
REPTAGIN IS REGOACTIONA OF CRE
Now notice that the output of these Shannon text generators is a signal...
which also the hypersphere E. We can recognize the humanness even in the
lowest order approximation. Yes, it is just noise. But it is human
noise--statistically different than every other kind of noise. As the order
increases, the text becomes very obviously human, but discoherent--like a
whole bunch of human sentences averaged toegether--like unfocused thought.
Tought without connection to human meaning. This is what bugged MacKay.
Information theory obviously works, but it factors out the context of the
human and is meaningless independent of us, with our portable human
contexts. MacKay said there are two kinds of information--selective and
structural. Selective information--Shannon's information--is information
calculated by selecting eelementsfrom a set. Sturctural infromation,
communicates how the selected infromation is to be understood--it is
metainformation.
Okay, I agree.
That's what wrong with Shannon's approximations--they aren't focused because
they are lacking metainformation about how to structure the signal they are
generating to be understood by people. So, let's extend Shannon's
hypersphere:
E=x1 suqared + x2 squared + x3 squared...
and embed it into another hypersphere--a hypersphere of metainformation by
adding another set of terms:
y1 suqared + y2 squared + y3 squared...
Yielding:
E=x1 squared + x2 squared + x3 squared... + y1 squared + y2 squared + y3
squared...
Now E is no longer appropriate because we are no longer talking about just
energy--just a signal. Now we are faces with energy, plus the entire human
context--or psychology. Thus:
E Psi=x1 squared + x2 squared + x3 squared... + y1 squared + y2 squared + y3
squared...
The greek letter Psi (which is tatooed on the back of my left hand) is the
symbol for psychology. The only other use of the letter in math that I am
aware of is in Schrodingers wave function.
What does this haunting hypersphere buy us?
Nothing unless we identify some candidates for Psi terms.
I propose as a first approximation : C P A D
Coherence
Pleasure
Arousal
Dominance
The theoretical rationale and experimental foundations for the Coherence
Semantic Model of Minimum Intelligent Signasls have been detailed by
McKinstry (1995). The Model consists of one dimension that is used to
describe and measure semantic resonance (how true or false a given signal
feels.)
The Specific term describing meaning can be visualized as a point in a
one-dimensional Coherence-Discoherence space. Alternatively, when the scale
scores are standardized, each signal can be described succinctly in terms of
its values on a Coherence-discoherence axis. The following sample ratings
illustrate the coherence or ability for a human to understand a signal when
the scores on the scale range from -1 to +1: Most people hate "Why did the
chicken cross the road?" jokes. (.81) "Why did the chicken cross the road?
To get to the other side." is funny (.45) (These are real examples!)
The theoretical rationale and experimental foundations for the PAD Emotional
State Model have been detailed by Mehrabian (1980, 1995b, 1997). The Model
consists of three nearly independent dimensions that are used to describe
and measure emotional states (or feelings, affective conditions): pleasure-
displeasure, arousal-nonarousal, and dominance-submissiveness.
"Pleasure-displeasure" distinguishes the positive-negative affective quality
of emotional states, "arousal-nonarousal" refers to a combination of
physical activity and mental alertness, and "dominance-submissiveness" is
defined in terms of control versus lack of control.
Specific terms describing emotions can be visualized as points in a
three-dimensional PAD emotion space. Alternatively, when the PAD scale
scores are standardized, each emotion term can be described succinctly in
terms of its values on the pleasure-displeasure, arousal-nonarousal, and
dominance- submissiveness axes. The following sample ratings illustrate
definitions of various emotion terms when scores on each PAD scale range
from -1 to +1: angry (-.51, .59, .25), bored (-.65, -.62, -.33), curious
(.22, .62, -.01), dignified (.55, .22, .61), elated (.50, .42, .23), hungry
(-.44, .14, -.21), inhibited (-.54, -.04, -.41), loved (.87, .54, -.18),
puzzled (-.41, .48, -.33), sleepy (.20, -.70, -.44), unconcerned
(-.13, -.41, .08), violent (-.50, .62, .38).
Thus, according to ratings given for "angry," it is a highly unpleasant,
highly aroused, and moderately dominant emotional state. "Sleepy" consists
of a moderately pleasant, extremely unaroused, and moderately submissive
state, whereas "bored" is composed of highly unpleasant, highly unaroused,
and moderately submissive components.
Combining C, P A and D (the metasignal) to Shannon's hypersphere (the
signal) gives us something that looks like this:
Signal: (Most people hate "Why did the chicken cross the road?" jokes.)
Metasignal: (.81, -.6, -.6, -.3)
It's not important that the signal is written as text. I could have
converted it to numbers. What is important is that we have a new object--a
hypersphere of meaning and emotion on the surface of which we can plot any
coherent human thought. Thought space (N+4 Dimensions. Where N is the length
of the signal in bits)
Just a note: the metasignal is structured internally from public to private.
The Coherence bit is the most public because everyone can agree on it. The
most private bit would be the subcognitive modification of the signal that
the encoder is unaware of (subconscious influence)... see French.
The next step is to build a statistical model of an intelligent signal
generator the same way Shannon built his human noise generators. That's what
mindpixel is about. It collects data so that we can build a statistical
model that will allow a machine to test any signal for coherence by
comparing it to very large amount of coherent and discoherent examples. If
we can teach a machine to recognize an meaningful signal, then just by
looking at a large enough number of random signals, it will discover and
thus generate meaningful signals. Which is a pretty remarkable idea.
The whole race is a poet that writes down
The eccentric propositions of its fate.