I

I \\ “1“,“ Y n .
‘ H \ ‘
‘ H w
‘, ‘
m Wm , ‘le’
I

_ ,‘7fii

Ill/W /!

.l

{NFSRMATWN AS A CDNSTRUCT éN
HUMAN CGMMUMSATifiN SYSTEMS

I

 

'—I
:8:
CID-Am

Thesis for the Degree of Ph.. D.
MiCHIGAN STA‘fE fiNiVERSITY
CLYDE D. MORRES
1968

 

flllfljllllfllfllflfllfllUUIIIUHIHHIHJIIIUUIIIWI ’

10375 6825

This is to certify that the

thesis entitled

Information as a Construct in
Human Communication Systems

presented by

Clyde D. J. Morris

has been accepted towards fulfillment
of the requirements for

Ph.D. degree in Communication

 

§

Major professor

Date July 18¢_]

0-169

-’ ammo av

 

 

 

 

 

(16 v0

scie

And

th
Co

Co

ABSTRACT

INFORMATION AS A CONSTRUCT IN
HUMAN COMMUNICATION SYSTEMS

by Clyde D. J. Morris

In the past two decades, a new discipline has appeared that is
devoted to the scientific study of human communication. As with any
science, one of the discipline's main goals is theory construction.

And one of the first tasks of theory construction is construct
explication. One construct frequently used in communication research
is "information." This paper describes ways the construct has been
explicated during the last 20 years and suggests some areas for further
explication of the construct.

Bxplication of a construct, for purposes of this paper, is taken
to be establishing two linkages: a linkage to physical events (Operational
definition) and a linkage to other terms (constitutive definition).
These linkages are examined in various areas of communication research
such as applications of information theory, dissonance theory, mass
communication and diffusion of innovations.

After analyzing how the term has been used in communication
research, a matrix of usages for the term "information" is presented
that shows the locus of the usage, the function, the relationship to
uncertainty and the advantages and disadvantages to the researcher of
the particular usage. Usage number fOur is related to states of the
communication system and is part of an approach to the study of human

communication based on General Systems Theory.

Accepted by the faculty of the Department
of Communication, College of Communication Arts,
Michigan State University, in partial fulfillment of

the requirements for the Doctor of Philosophy degree.

9

67W

ctor of Thesis

Guidance Committee: M. Chairman

E ‘

 

INFORMATION AS A CONSTRUCT IN

HUMAN COMMUNICATION SYSTEMS

By

Clyde D. J. Morris

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements ,
for the degree of
DOCTOR OF PHILOSOPHY

Department of Communication

1968

 

peo

be

par

the

Ca:
the

uh,

ACKNOWLEDGMENTS

My warmest thanks go to Dr. Randall P. Harrison, my thesis
advisor; and to Dr. Hideya Kumata, Dr. R. Vincent Farace, and
Dr. John Vinsonhaler, the members of my guidance committee. My thanks
also to two men whom I greatly admire: Dr. Miles Martin and
Dr. Everett Rogers.

A special kind of thanks to Dr. David K. Berlo - a man who
understands how to create an intellectual climate in which researchers
and scholars feel free to work on problems they're interested in with
people that they like to work with.

So many colleagues have influenced my thinking that they can not
be listed. But Jeffrey Katzer, David Beatty and Linda Davis have been
particularly helpful to me in the formulation of this thesis.

And thanks to Gail Morris. Graduate programs are much harder
on wives of graduate students than they are on graduate students.

Gail was patient and understanding when she should have been and not
patient and understanding when she should not have been.

Two men who have greatly influenced my thinking should be
thanked here: Mr. J. L. Thomas and Dr. John Paul.

My sincerest thanks to Mrs. Shirley Sherman, Mrs. Joyce Flea
and Mrs. Donna Kokx for their help in preparing the several drafts of
this thesis.

And a long overdue thanks to John Morris and Agnes Morris. I
can think of several highly articulate and eloquently written books
that have said things that my parents told me, in simpler language,

when I was a boy.

ii

CE;

CHAPTER

I. INTRODUCTION: THE PROBLEM . .

II.

III.

IV. INFORMATION

TO COMMUNICATION

V. CONCLUSIONS AND IMPLICATIONS FOR RESEARCH

BIBLIOGRAPHY . . . .

iii

TABLE OF CONTENTS

SHANNON'S USE or INFORMATION (11) , , ,

- GENERAL SYSTEMS THEORY APPROACH

INFORMATION2 AND INFORMATION3(ABSOLUTE AND DISTRIBUTED)

22

28

39

#3

Table 1.

LIST OF TABLES

Four Uses of "Information" . . .

iv

LIST OF FIGURES

Page

Figure 1. Kinds of Information and Uncertainty . . l2

devot
cien
and o
Berle
funda

(Ber;

of ti
SCier

that

Ix)
*5

w.

CHAPTER I

INTRODUCTION: THE PROBLEM

In the past two decades, a new discipline has appeared that is
devoted to the scientific study of human communication. As with any
science, one of the discipline's main goals is theory construction.

And one of the first tasks of theory construction is construct explication.
Berlo notes that "the creation, care and feeding of constructs is a
fundamental task, and one which must precede most other scientific work."
(Berlo, p. 2)._

One construct frequently used in communication researCh is "in-
fbrmation." This paper will describe ways the construct "infOrmation"
has been explicated during the last 20 years and will suggest some areas -

for further explication of the construct.

The Problem of Construct Explication

For purposes of this paper, construct is used in Conant's sense
of the word - a concept that has been created or appropriated for
scientific purposes. (Conant, p. 31). A construct is a word (a symbol)
that the scientist can manipulate and for which he can provide meaning.
To be useful to the scientist, the construct must be reliable; that is,
when it is applied it must consistently include the same phenomena
andexclude other phenomena. This suggests two things:
1. Tightness or rigor at a logical, formal level.

2. Tightness in operationalization or measurement.

In Berlo's terms, there are two linkages for the construct; the
linkage of the term to physical events (Operational definition) and the
linkage of the term to other terms (the constitutive definition). (Berlo, p. 2).

These problems of definition are of practical importance when we
deal with the construct information because the term is used by both the
scientist and the layman. To the layman, "fish" is practically anything
that swims, including whales, eels, etc. But to an icthyologist, it
has a rigorous, Specific meaning. And to the layman, "electricity" is a
useful construct to describe whatever it is that makes the lights go on
but the electrical engineer needs a good deal more precision. To him,
electricity is almost a meaningless construct unless it is specified in
terms of volts, amps, resistance, etc. in the circuit. These few
sentences are, of course, only a clarification of what is involved in
construct explication for purposes of this paper. Rudner deals at some
length with the problem of construct explication in the social sciences

(Rudner, p. 19).

Information — Boundaries and Measures

As has been noted, to the layman "information" means a variety of
vague things. For example, he could mean information about new cars,
or the information explosion, or the service the phone company provides.
Sometimes the communication researcher uses it like the layman, sometimes
he uses it more precisely. When using the word information, this paper
suggests that the researcher could be using the word even more precisely.
As it has been used in research, the word seems to shift focus and
change the boundaries of what is in and what is out. Sometime the boundary

is drawn so that the "infermation" is in the external world (that is, in

the stimuli). Sometimes the boundary is drawn so the information is in
the person's head (what is informative is what will change what is already
"known.") This problem of changing focus and boundary will be dealt

with as a problem of locus.

On the measurement dimension, the operationalization of information
differs. It is often at the nominal level of measurement (that is in-
formation and this isn't); sometimes it is at the ordinal level of measure-
ment (200 words is defined to be more information than 100 words but
not twice as much). Sometimes it is difficult to measure information at
all (the amount and kind of reSponses we get when we present a question
to an infbrmation retrieval system). In this last case, the question
of what is a relevant answer and what is not relevant becomes important,
as does the problem of how much of the relevant information has been

retrieved and how much has not been retrieved.

A Matrix for Examining the Multiple Usages of Information

 

Section V of this paper presents a table that illustrates four
uses for the word "infermation," the locus of the usage, the function of
the usage, how the usage relates to uncertainty, and some advantages
and disadvantages of the particular usage of the construct when studying
the process of human communication. The four usages are labeled I1, 12,
13, and Iu°

11 is statistical information. The word is used here in Shannon's

sense (Shannon, P- 32). 12 may be looked at as "pieces" of information.
("Pieces" is used to avoid confusion with Shannon's "bits"). The pieces
are known collectively as the world's knowledge. Another way of'con-

ceptualizing this usage is to borrow Teilhard de Chardin's word nooSphere -

 

auord he
intellect‘
earth, mu
way of co
as oppose
Absolute

is more a
a musici.
Creative
infomat
fomatio
dictiona
formed,

an answe
concem
interpe
EESture
funCtic
SYSIEms
t0 the

0f inf

for fu

a word he used to describe the "thinking layer" or the total Sphere of
intellectual activity which he predicted would eventually cover the

e arth, much as the atmoSphere does. (Teilhard, p. 180). Still another

Way Of conceptualizing 12 is in Brillouin's term "absolute information"
as opposed to "distributed information," (Brillouin, 1950, p. 595).
Absolute infermation exists as soon as at least one person has it. There
is more absolute information as a result of a new scientific discovery,

a musician composing a symphony, an author writing a book - anything
creative or imaginative which adds to the noosphere. When the absolute
information spreads to more than one person, we have distributed in-

formation which is labeled I This usage is close to the more normal

30
dictionary sense of the word. This usage is an informing or being in-
formed, a telling or being told of something, news, acquisition of facts,
an answer to a question, etc. The fourth usage, In: is proposed as

concerning the state of the communication system. Much of In at the

interpersonal level is received via nonverbal channels (smile, frown,
gestures, etc.) which makes it difficult to articulate its meaning. Its
functions are largely to: 1) define the relationship of the living
systems to one another, 2) define the communication system's relationship
to the environment, and 3) define the relationship of the actions of
the system to the goals of the system.

The next feur sections of this paper will deal with the four usages
of information. Section V presents some conclusions and implications

for further research.

Sh;
Specified
a certair
are recef
is produ
measnm
much unc

maSUpe

Which s
the nu:
altem;
in the
“ceiv

exampl

If
m

We hat

CHAPTER II

SHANNON'S USE OF INFORMATION ( Il )

Shannon's concern is with the transmission of symbols within a
Specified system. As the symbols are sent from a source to a receiver,
a certain amount of uncertainty of the receiver is reduced as the symbols
are received. Shannon was looking for a measure of how much information
is produced as the receiver receives the symbols. His measure is a
measure of the freedom of choice in selection, which is related to how
much uncertainty there is to begin with as the message is received. The

measure he derived is:

H““ Pi 1°8Pi
1

which states that the freedom of choice or uncertainty is a function of
the number of alternatives and the probability of occurrence of these
alternatives (Shannon, p. 50). Shannon was not concerned with meaning
in the sense of what human value might be attached to the symbols being
received. The symbols are not necessarily letters, as the next few

examples attempt to illustrate.

Information as a Choice between "on" and "off."

 

A stimulus from the environment can be a source of information if

we have a_ priori knowledge that the stimulus can either be on or off.

If we have
ringing).
assign ou:
symbols it
of: ringi
is at the
the door"

It
continuox
the bell
The latt.
beam tel
115 some
Signs uh

E
and we 1
(anothe:
by answq
for him
2? km
w2’11 r
3? diff
Stimuli
”rosin
Host oh

an Oper

If we have a doorbell, it can be in either of two states (ringing or not
ringing). The meaning of the ring is arbitrary — something which we
assign ourselves. In Shannon's sense of information, there are two
symbols in the repertoire that we as receivers have §_priori knowledge
of: ringing and no ringing. In general, we assign the meaning "someone
is at the door" to the state of ringing and the meaning "no one is at
the door" to the state of not ringing.

It could easily be arranged the other way. The bell could ring
continuously when no one was there and, when someone pushed the button,
the bell would stop ringing which would tell us that someone was there.
The latter case is the way we rig electronic eyes. The continuous
beam tells us no one has crossed it. The cessation of the beam tells
us someone is there. In each of the instances described, we have
signs which are capable of being in more than one state (on or off).

Suppose that we get tired of answering the doorbell ourselves
and we hire a butler. We'll call him Harry. If someone knocks
(another on - off system) on the door or rings the hell, he responds
by answering the door. But on Harry's day off his brother stands in
for him. Alfred won't open the door until the caller rings the bell
and knocks on the door. Our friends will ask us "What's with those two?"
We'll reply with a shrug and say "I don't know, they're just different."
By different, what we're saying is Harry and Alfred seem to process
stimuli from the environment differently. All "objective" observers
perceive two stimuli - a knock on the door and a ring of the bell.

Most observers agree that the expected response for either stimuli is

an opening of the door. Yet, Alfred doesn't open the door until he hears

both. 01‘
curve dro
nhhahe
Tr
we percei
forwhe
someone .
to them
thaxone
or hi be
no touc]
wrmo
we take

to us.

(in m
set
be pac

Could

Betty

"0n" 5
the s<

Sigma.

op So

both. Or maybe he doesn't "hear" both. Maybe his hearing response

curve drops sharply at the frequency range of the ring of the bell. Or,

maybe he requires a ring and_a knock as conditions for opening the door.
This discussion underscores that we know what we know by what

we perceive through our senses. But we may not all have the same meaning

f<>r what we perceive. And even if we perceive the same stimuli as

someone else, maybe we process incoming stimuli differently and respond

to them differently. A stimulus isn't a stimulus unless there is more

than one alternative on the incoming channel. It must be bell-no bell,

or bi bellelow bell, light-no light or green light-red light, touch-

no touch or hot touch-cold touch, smell of roses or smell of beer. If

our receptors (our senses) perceive some change in the environment and

we take these changes into account, then these changes become information

to 118.

Information as Freedom of Choice (Degrees of Freedom)

' Information in this sense is a measure of the freedom of choice
(in the statistical sense) when Selecting a "message" from an available
set. Many more symbols to which more meanings could be assigned could
be packed into our doorbell code. A bachelor with five girl friends
could agree in advance with each one on a unique ring. Mary has one,
Betty has two, etc. Now when the bell rings, he knows which of the
"on" states he hears which tells him: 1. someone is there, and 2. who
the someone is. The patterned input has a referent; the pattern now
signals a unique young lady.

‘But a problem arises. Any stranger is likely to use one or two

or so rings. And our friend may wish to know whether a friend or a stranger

is at the
second be
hears a i
i which

I
"teaming
unique r
acquaint
find him
friends1
assignec
with feet
he may a
Closed .
more in

(mm,er
is the

The man
Shamnon
tramSrni
“use
ShamIon

(I

Hf
w

h
.e Uses

is at the door. So he puts in a conVeniently located but secret

second bell which has a different sound from the first. Now when he
hears a hell he knows: l. someone is there, 2. friend or stranger, and
3. which of the five girls it is.

If our friend has a desire for even more certainty as to the
"meaning" of the bell when he heard it, he could assign every friend a
unique ring. He may even want to discriminate between close friends and
acquaintances. He could use more bells and more codes. But then he may
find himself in the position, if he is a popular young man with many
friends, of waiting for several minutes as his friends ring out their
assigned code. He could reduce his waiting time by putting in more bells
with fewer codes per hell, but bell installation can become costly.

He may ask himself if it wouldn't be cheaper and faster to install a
closed circuit TV system. He has figured intuitively that he can get
more information per unit of time per unit of money on a TV channel than
on many Spearate wire channels.

This comparison of the relative information carrying capacity
(number of alternatives) of different channels of information transmission
is the one faced for many years by engineers at Bell Telephone Laboratories.
The man who devised a statistic to compare different channels was Claude
Shannon. He didn't care what the "meaning" of the information being
transmitted was. That would be left to the discretion of the person
using the channel.

Shannon's Mathematical Theory of Communication
( Information Theory)
Shannon, in his classic paper, points out that information, as

he uses the term, is significant in that the actual message is one

selected from a set of possible messages. His problem, as a communi-
cation engineer, is to design a system to operate for each possible
selection, not just for the one which will actually be chosen since
this is unknown at the time of designing the system.

Information theory is not a theory as much as it is a measurement.
It is a useful statistical technique for quantifying amount of informa-
tion.

Information in Shannon's sense as we have noted is a function of

l. the number of alternatives.

2. the probability of occurrence of these alternatives.

To clarify this point, let us consider an example. Imagine a
dime on one square of a checkerboard. A checkerboard has 6n squares
counting red and black squares. Pretend you have your'back to the
board and a friend is standing there who will answer yes or no to any
question you ask about the location of the dime. With one question,
you could know for certain which half (left or right) of the board the
(lime is on. If you say "Is it on the left half," and it is on the left
half, he will answer "yes." If it is not on the left half, he will
answer "no." In either case, you know which half for sure with just
one question.

With another question, you could find out whether it is in the
top or the bottom of the half. If you try this yourself with a dime
and a board, you will see that with six questions, you can locate the
dime precisely. In other words, each yes or no answer reduces your
uncertainty (remember you are the receiver) by one half. You start with

en possibilities and, with one appropriately asked question, reduce

these al'
to 16 an
to one -
infomat
tematix

statistf

whether
50y or
imagine
altema
0.999 a
0f alte
b0y in

case sj

eqUallj
binary
half, 1
the al-
inesu.
aVai la]
decima
that in
People

at PPE:

10

these alternatives to 32. The next question reduces the alternatives

to 16 and so on down to the last question which reduces the alternatives
to one - the one with the dime. This example is intended to illustrate
information as a function of the number of alternatives; the more a1-
ternatives to begin with, the more information, i.e., the bigger the
statistics. Now to the probabilities.

Imagine a school with 500 boys and 500 girls. You are to guess
whether or not the first person you see coming through the door is a
boy or a girl. You have a 50-50 chance or a probability of 0.5. Now,
imagine a school with 999 boys and one girl. There are still two
alternatives (boy or girl) but the probability of seeing a boy is
0.999 and the probability of seeing a girl is 0.001. While the number
of alternatives remains the same (boy or girl), the appearance of a
boy in the first case is more "informative" than a boy in the second
case since it is less probable (.50 vs. .999).

In general then, the more alternatives we have and the more
equally probable these alternatives are, the more "information." Each
binary decision (yes or no) reduces the receiver's uncertainty by one
half, given equal probabilities. It is the different probability of
the alternatives that makes the game 20 Questions possible. Each

question has a binary (yes or no) reSponse. The total infermation

available 18 220 alternatives or 1,098,576 if we express it in the
decimal system. If we restricted the game to just people, this means
that we can eliminate as possible answers just a little over a million
people with 20 questions. But there are over 3 billion peOple on earth

at present and more than that if we count those who have lived and those

11

who will live. The only reason we can, with 20 guesses, reduce the
choices to Liz Taylor or Julius Caesar is because they are more probable
selections than the third name from the top on page 33 of the New York
telephone directory or the centurion who was in charge of the palace
guard on the night Caesar was killed. Shannon's theory then gives us

a measurement for "information" Specifically based on the number of al-

ternatives and probabilities for the alternatives.

Information and Uncertainty - Some Distinctions

 

This paper attempts to illustrate different kinds of information.
These different kinds of information each reduce some uncertainty. But,
just as there seems to be different kinds of information, there seems
to be different kinds of uncertainty, which might be useful to distinguish.

For some kinds of uncertainty, there are methods for having
a priori knowledge of the range of uncertainty. If we roll a die, for
example, we have some knowledge that the outcome will be one of six
possibilities. In reading a message written in English, we have a priori
knowledge of the letters that will be used and the words that will be
used. But we have no way of calculating the probability that the nth
person reading this paper will drink a glass of orange juice while
he or she reads it. There appears to be some limits for the different
kinds of uncertainty.

Brillouin (1962a) presents a discussion of uncertainty in terms
of scientific experiments which illustrates some of the different kinds
of uncertainty. If we plot two variables against one another, for

example x and y, and let a be the full range for x and b be the full

range fo:
fall wit]
a certai'
range of

Range of

 

3‘19 note
1. has

2. is c
With,ut

and Val

is the

that we

12

range fer y, then after a few measurements we conclude that x and y always
fall within a certain shaded stripe (See Figure 1). This accounts fer

a certain empirical law and certain limits of error. P0 = ab the full

range 0f variation 0f X and Y while P1 is the area of the shaded region.

Range of error is el to 32,

 

 

 

 

Y
---.-po
91
------- b-Pl
82
—---Experimental Law
x

a

Figure 1. Kinds of Information and Uncertainty

He notes that a scientific law:
1. has a limited field of application
2. is correct "within possible errors"
Without specifying both 1 and 2 the statement of the law is incomplete
and valueless.

In summary, then, there are several levels of uncertainty. There
is the uncertainty of measurement error but this falls within a range

that we can specify. There is the uncertainty of where the relationship

13

will lie when measured but, before measuring, we can predict that it
will fall in the range ab. Finally, there is the uncertainty of all
other variables not covered by the area ab. If the system we are
examining is a complex one, we may not know if ab is orthogonal or what
the relationship of a and b are to the rest of the variables we could

use if we considered them separately.

Information as Uncertainty Reduction

 

We may now examine the point of view in the previous examples.
It is the receiver’s uncertainty which is being reduced, and the
uncertainty is within Specified bounds. The sender encodes a message
using a set of symbols which is agreed upon as a code by the sender
and the receiver. The receiver must have prior knowledge of the code
in order to decode the symbols. The meaning of the symbols is
arbitrary. Even in the game, 20 Questions, the sender already knows the
answer. It is the receiver who is reducing the uncertainty that he,

the receiver, has.

A Note on Shannon and Weaver

 

As this chapter progresses, the examples shift from what has been
labeled Il to what we have labeled 13- This is, in part, due to a shift
in emphasis from the writing of Shannon to the writing of Weaver. Shannon

deals primarily with what has been labeled I Weaver deals with in-

lo
formation, but suggests an extension of the work into the semantic level
and the behavioral level. These levels treat meaning - something

Shannon does not do in his paper. Weaver is careful to make the distinction

in his paper but, in the application of Shannon's work by other researchers,

 

infomatic

his conf'

Ma
the logon
due to th
periment;
He descm'
we take.
sense of

Ga
function
Stimuli:
fomatio
to a PEG
1°80?! cc
This mee
by the F

8th Pee

1n

information (in Shannon's sense) and meaning are sometimes confused.

This confusion will be discussed later in this section.

USes of "Information" Similar to Shannon's

MacKay distinguishes two kinds of psychological information -
the logon and the metron. He defines logon as fa_priori information"
due to the logical structure of a set of alternatives or of an ex-
periment; this is closely related to the concept of degrees of freedom.
He describes metron as information which is obtained from measurements
we take. It is related to the concept of precision in the statistical
sense of the word.

Garner uses MacKay's notions in an example of reaction time as a
function of the modality of the stimulus. Let's say we have three
stimuli: visual, auditory and tactual. This limits the amount of in-
:fbrmation we can obtain. Three stimuli have more potential infermation
‘to a receiver than one stimulus. This aspect of the problem provides
.logon content. We then measure reaction time for each of these stimuli.
'This measurement also provides information, but its amount is limited
lay the precision of the measurement; and we ordinarily determine several
asuch reaction times in order to obtain some estimate of the precision
<Df the measurement. This aSpect of the problem provides metron content
(Garner, p. 10).

MacKay's distinction between logon and metron is important to
IPSychology because humans apparently make use of both kinds of information.
Sometimes one or the other is more important for human processing of in-

formation. Broadbent (p. 113) has said:

Neither ‘
occurred
to be im
illustra
he has d
If the "
feel bad
the comm
(
says he
with She
gained ¢

15

1. SR psychology uses occurrence of stimuli past or present.

2. Gestaltists consider patterns of stimuli present at any one
time.

Neither talks about effect at any moment of the stimuli which might have
occurred but did not. Broadbent considers what might have happened

to be important in psychology. An example in human communication
illustrates this. A person may do a good job and turn to the person

he has done the job fer expecting to hear a well deserved "thanks."

If the "thanks" is not said, the person who has done the good job may
feel bad. What has not occurred, in this case, becomes important in
the communication system involving two peOple.

Garner in his Uncertainty and Structure as Psychological Concepts
says he uses "uncertainty" rather than "information" to avoid confusion
‘with Shannon's work. His "uncertainty" is reduced by "information"
gained during a communicative act.

Some of the more important assumptions he makes are:

l. Uncertainty is a prerequisite to structure. Structure is
related uncertainty, not the lack of it, and to have
structure is to have uncertainty. Figure 1 illustrates
this point.

2. The search for structure is inherent in human behavior.
Peeple in any situation will search for meaningful relations
between the variables existing in the situation.
Psychologists who run experiments where minimal or
ambiguous cue differences exist find that subjects will
find any cue at all to relate their reSponses to even if

the cue is irrelevant to the task.

in a hl
radar .
if We .
(1131313.
Dumber

Steps

the Si
higher
we Use
shows .
S'eps 1

discI‘ir

l6

3. Structure can exist meaningfully only within Specified
systems. The subject who sees a Single stimulus as structure
does so only because he generates an implied set of
stimuli against which the particular stimulus can be con-
trasted or within which the particular stimulus can be
subsumed.

u. Structure as such is a property of systems. Structure is
an entire set of relations between variables and its amount
and form can be specified without statement of the particular
significations which operate because of the structure. The
structure can be perceived even without the ability to
verbalize the significations.

One application of Garner's notions of information measurement

in a human communication situation is when we build disPlayS such as a
radar display for an aircraft controller. We can ask the question -
if we use any perceptual or sensory continuum as a method of coding and
diSplaying any information continuum, is it better to use a small
inumber of discrete steps on the continuum, a large number of small
steps or even to use the complete continuum with its infinity of steps?

For example, how could we Show the altitude of an aircraft by

the size of the symbol used to represent the aircraft? Assuming the
higher the aircraft, the larger it appears on the radar screen, Should
'we use 2, 3, S, 10, 20, or no steps? A simple "errors made" experiment
shows that perfOrmance deteriorates with an increase in the number of
steps beyond some minimum value. But, we may want to trade off increased

discrimination to offset the number of errors. Five categories may

17

give us no errors and ten categories may give us just one error. We

get twice as many steps at a modest increase in the number of errors.
In summary, with writers like MacKay, Broadbent and Garner, we

find "information" measures being applied to problems of psychological

uncertainty.

Information and Meaning -- A Controversy

A controversy in the literature concerns Shannon's use of
information and the use of information by others to mean "meaning."
Bar-Hillel for example, says there's no connection between the amount
of semantic information conveyed by a statement and the measure of
rarity of kinds of symbol sequences. He says "Apples are red" carries
semantic information regardless of how frequently the statement is trans-
mitted.

Weaver's discussion of three levels of communication problems
helps reduce some of the confusion (Shannon and Weaver, p. u). He
refers to the following levels:

Level A. How accurately can the symbols of communication be
‘transmitted? (The technical problem).

Level B. How precisely can the transmitted symbols convey
the desired meaning? (The semantic problem).

Level C. How effectively does the received meaning affect

Conduct in the desired way? (The effectiveness problem).

.lnformation and Behavior
While Weaver's distinction of levels helps reduce the confusion,

it also illustrates some of the confusion. His levels B and C imply that

meaning:
enough .
meaning
coming

This pl
between

living

organis

adapt .

which ‘
descrf
can di
By thi
with e
intens
label

envipc
CODdit
Obtain

manne r

of C03]

18

meanings can somehow be transmitted if the sender is only skillful
enough in his selection of symbols. It may be more useful if we view
meaning as an attribute of the person and not of the message. The in-
coming stimuli can elicit certain meanings in the person receiving them.
This places the fecus of communication research on the linkage system
between two persons or, as will be suggested later, between any two
living systems. Some questions we may wish to determine are:

1. What Signals from the environment are detected by the
organism we are studying? (What infermation is produced?),

2. How subtle are the discriminations of these signals?

3. How are these incoming signals used by the organism to
adapt to its environment?

For example, we use the word "blind" to describe those organisms
which are incapable of sensing visual stimuli. And George Miller (1956)
describes experiments which suggest that even those who see normally
can discriminate a limited number of changes on a given channel.
By this we mean that given a range of light intensities, for example,
with each intensity being given a label, above seven or eight different
intensities most people make errors in responding with the correct
label when presented with a given intensity. Finally, stimuli from the
environment take on "meanings" for a particular organism. Rats being
conditioned, fer example, will associate the ringing of a bell with
obtaining food and respond tO‘a bell in a manner greatly similar to the
manner in which they originally responded to food.

One thing we might consider when talking of the behavioral aspects

of communication (Weaver's Level C) is whose point of view are we taking?

Ackoff's
particula
purposivu
make and
His deci:
If we no.
goals he
his esti:
that it
the mess
temativ:
message
A
envimnm
Periment
SUbject
knowledg
Past exp

°Perate

19

Ackoff's model of communication illustrates this point. Ackoff has a
particular use for information. First, he assumes a person is in a
purposive state. By this is meant the individual has a decision to
make and he has at least two alternative methods of achieving a goal.
His decision consists of selecting an alternative to achieve his goal.
If we modify the person's value of the goal (in relation to the set of
goals he might have) the message is one of "motivation." If we change
his estimate of the relative efficiency of a given alternative such
that it changes the probability of his selecting that alternative,

the message is one of "instruction." If we provide him with an al-
ternative which he was not aware of before he received our message, the
”message is "information."

Ackoff's labeling is from the point of view of someone in the
"environment watching the person who is making the decision. An ex—
perimenter’has some knowledge of what messages he is sending to the
subject from the eXperimenter'S point of view but he has no certain
Iknowledge of what the subject perceives and he has no knowledge of what
jpast experience or present needs the subject is bringing to bear to

Operate on the stimuli from his environment.

Mg As An Information Processor

James Miller, Norbert Wiener, Colin Cherry, Stafford Beer, and
<Ithers have suggested that we look at man as an information processor.
‘iiener has said the world may be viewed as a myriad of "To whom it may
concern" messages. Man gets data from his environment in the form of
lights, colors, sounds, smells, symbols, pictures, touch, taste, etc.

If we view man as an infermation processor, we might ask what kinds

 

20

of operations he can perform on raw sensory data coming in from the
environment. Eventually we would like to be able to say under what
circumstances a given person will perform a given operation. Some of
the possible operations are:

1. He has the ability to filter and smooth data, to

reject extraneous detail and to fill in missing elements.
2. He has the ability to integrate, or combine, an array

of coded data into a composite picture of a physical

process or event.

3. He can store and relate large quantities of data far

beyond the capacity of any known computing device.

u. He can extrapolate from diverse data to establish

future trends and interpolate between available data
to predict missing data or determine the need to seek out
additional or new data.

The human also has remarkable adaptability at times. He will often
reSpond to overload by reducing the number of input sources he attends
to; or change the pattern of scan across the inputs. He will often
reduce the number and kind of his outputs to adapt to emergency situation.
An example here is the random crawl of a baby heading toward a swimming
pool and the variation in sampling frequency the attending parent, who
.is also reading a book, will devote to watching the baby as a function
<3f distance to the pool. In this instance, the data sought from
the environment is based on observation of the environment plus ex-
trapolation of the baby's Speed and direction. If the baby approaches

the point of no return (point at which present heading and Speed will

21

carry him into the pool before the parent can overtake him) the parent
b ehaves by dropping the book, jumping up and picking up the baby.

It would appear that the kinds of data the human seeks from the
environment depends sometimes on what activities he is engaged in, what
his goals are (e.g., save baby from falling in pool) and the kinds of
decisions he must make.

This section has attempted to illustrate how changes in a given
stimulus from the environment are perceived by man and how these
changes take on meaning as a function of the man and the environment
in which the stimulus appears.

Shannon's use of information is a measure of the freedom of
choice one has when selecting a message from an available set. In-
:formation, in this sense, is related to the number of alternatives and
the probability of occurrence distributed across these alternatives.
Researchers such as Garner, Berlyne, Attneave, Broadbent, and MacKay
have adapted Shannon's notion and have looked at characteristics of the
stimuli which can be varied in a manner that seems to affect human
behavior.

Although Shannon's work has some use in the study of human communi-
cation, it deals with an optimal theory of coding of sets of arbitrary
symbols. Since any stimulus in the environment can be considered as
a.set of symbols, the theory has provided some insights into the mechanical
limitations of man's sensory system. The question of why man does as

he does when he senses data from the environment is another matter.

CHAPTER III

INFORMATION2 AND INFORMATION3 (ABSOLUTE AND DISTRIBUTED)

It may be useful to begin this section by considering Brillouin's
notions of absolute and distributed infermation. He suggests that
active, creative thinking requires invention, methodological imagination,
unexpected comparisons; and it results in a new idea, representing a
completely new type of absolute infermation. He says that there is
more absolute information when we have a scientific discovery, a new
scientific law, an author writing a book, a musician composing a
symphony, an artist painting a canvas, and other examples of creative
work. Brillouin (1962a) defines absolute infermation and distributed
information as:

l. absolute information, exists as soon as one person has it, and

 

should be counted as the same given amount of information, whether
it is known to one man or to millions.

2. distributed infermation, defined as the product of the amount of

 

absolute information and the number of peeple who share that
infermation. In the process of broadcasting the information,
from the point where it originates to all the persons who finally
receive it,.a certain amount of information may get lost and the
total amount of distributed information will be reduced by a

certain percentage.

22

23

These two types of information suggest different types of re-
search. The area of creativity research seems to be devoted to finding
out how some people devise new answers to questions; to how they devise
new questions to ask; and to how they devise new ways of looking at some
phenomena. A creative researcher is a man who can ask the significant
question.

Distributed information, on the other hand, suggests a focus on
the linkage system between persons. This linkage system is found in
teaching, in therapy, in mass communication, in diffusion of innovations -

in any area of communication research involving transfermation of 12 to

13. The next few pages attempt to present some boundaries of what is

and what is not information in the I2 and I3 usage of the word.

Past Operationalizations of Informationa

 

Informationa has been the fecus of many areas of communication
research. It is often operationalized as a measure of some kind of
information gain or recall. It is sometimes operationalized as time
spent attending to a communication. Dervin (p. 5) notes that in-
‘formation gain, information level, learning, memory, retention and
recall are used interchangeably. Some of the definitions used are:

1. Information level or support: the knowledge an individual

 

possesses which is relevant to his attitude, the amount of
information an individual is capable of bringing to bear in

'appraising the topic of an attitude. (Smith, Bruner’and White)

21+

2. Recognition: information level measured with true/false or multiple

 

choice items which require a discrimination of what one has
learned in the past. (Mednick)

3. Recall: free response by the individual of information learned in
the past, a method of measuring information gain and retention in
which individuals must freely verbalize and produce reSponses
without structured guidelines. (Mednick)

Examples of information operationalized according to one of the
definitions above are abundant in the communication literature.
Recognition is used by mass communication researchers in studying the
spread of a news story (Tannenbaum and Greenberg), recognition or
recall is used in the study of the diffusion of innovations (Rogers),
recognition is used in the study of rumor (Shibutani), recall and "time
Spent listening to a communication" are used in dissonance research
(Brehm and Cohen). In a sense, much communication research concen-
trates 0D informationa if we consider some message being presented
to some group of persons. It may be more useful to consider messages
as raw data in the environment and that information3 depends on the

message, the person and the context for more explicit specification.

Informationa Viewed as an Interaction Process

Let us consider, for a moment, the number "80" as a potential
element Of infermationa, If the question is posed by a stockbroker
"What is General Motors selling at today?" the answer "80" becomes
infermationa to him. But if a physician asks his nurse "What is that

patient's temperature?" The answer "80" is not within the range of

2S

acceptable answers. The physician's experience with body temperature

of live patients causes him to "believe" temperatures from perhaps

95 to 105 degrees. The number "80" is a data point in the environment.

In other words, as raw data from the environment becomes transfermed

into a cognition, the data become information3 to the person. The

informationa can be correct or incorrect depending on some belief
system. In this view, then, information3 can be taken as an inter—
action between the person and the environment or, more generally, an
interaction between a living system and its environment.

In this view, there appear to be a number of elements:

1. Some kind of energy pulse which is: a) above sensory thresholds and
b) similar to some energy pulse which has had Sign or symbolic
relevance for the organism previously. (That is, it's recognized
as "80" not as the nurse clearning her throat - it is seen as a
signal rather than noise).

2. Some aSpect of the environment for which the energy pulse acts as a
symbol (that is, 80 refers to temperature, not weight, or dollar
value, or number of patients).

3. A "program" inside the receiver which tells him how to handle
incoming inputs. ("I ask the temperature because it is related to
x,y, and z; if it is above point t, I take steps a, b, and c; if
it is below point t, I take actions d,e, and f; if it is outside
the range 95 to 105, I recheck...)

More generally, information (in any of the usages) is a meaning-
less construct unless it is Specified in terms of its constituents -
some set of symbols interacting with some receiver with information

being produced as a result.

26

For example, let us consider ID again. Weaver (p. 33) notes
that information may be a function of time f(t) as in radio or
telephony,0r it may be more complex. Black and white TV, for example,
may be thought of as a function of two Space coordinates and time; the
light intensity at point (x,y) at time t on a pickup tube plate.

A color TV provides more information and is even more complex. It is
several functions of several variables f(x,y,t), g(x,y,t), and
h(x,y,t). For us to Specify the message of a combined color reception
and associated sound at any time t, requires the specification of the
several functions noted for color TV plus the specification of the
associated sound message f(t).

Weaver's comments have their locus in the channel and the code.
But information in the 13 sense becomes information in an analogous
manner. The distributed infermation is a function of some living
system in some context at some point in time. (For example, the stock-
broker in the context of his office receiving the answer "80" to his
question.)

In a similar manner, the next section examines I,4 (infermation
about the state of the communication system). In the case of the
physician, he had a "theory" of what the answer should have been when
he asked his nurse for the patient's temperature. An answer that fell
outside the range of answers predicted by his theories was not accepted
or believed. The informationa in this instance was erroneous in the
cunntext of his question but it may have caused the physician to think
about the state of the communication system between him and his nurse

(for example, she didn't understand my question, She's incompetent,

27

she must be kidding me, I didn't hear her correctly). The "80"
becomes information” - infOrmation about the state of the communication
system.
We can use an example given earlier in this paper to illustrate

In. In an earlier section of this paper, the game 20 Questions is used
to illustrate the gradual reduction of uncertainty by receiving "yes"
or "no" answers to questions. This example can be used in another
sense to illustrate In - information about the state of the system.
If'a person were playing the game 20 Questions and were told that
"yes" and "no" represented the complete repertoire of symbols, how
would he react if he got the reSponse "maybe," to a question he posed?
If he were a Simple receiver (as a radio or TV is), the "maybe" would
be perceived as noise. But the human receiver (fer example, the
physician) is more complex than that. An answer like "maybe" would
be a cue that the rules of the game were being changed. It would be
,perceived as a communication about the game because it is not one of

' the symbols permitted within the game.

Information, in any of the usages, seems to be a process of

lbecoming aware of, or of ordering, or comprehending, or accepting raw

<1 ata from the environment. It may be convenient to view information
:18 a process - an interaction between a person (or any living system)

and data in the environment.

CHAPTER IV

INFORMATIONu - GENERAL SYSTEMS THEORY APPROACH TO COMMUNICATION

Since the early 19503, there has been increasing interest in
General Systems Theory and its application to the study of various
areas such as management (Beer), formal organizations (Katz and
Kahn), economics (Boulding), living systems (Miller) and other areas
of scientific inquiry (See General Systems Yearbooks beginning in 1956).
There has been some attempt to apply this approach to the study of human
communication (Watzlawick, Beavin and Jackson). This section explores
some of the implications of the General Systems approach with regard
to the discussion of information. To begin, we might consider
ancther approach that has been used as a model for communication - the
S-R model from psychology.

The S-R psychologist, as we know, goes to great lengths to
<2ontrol as many variables as possible; he then manipulates one variable
aand.measures some reSponse variable. This Stimulus-Response model
has had great success in psychology and, through it, some important
'theories in psychology have been constructed.

It is tempting to think of messages and humans in terms of the
S-R model. A message is presented to a person and he reacts in some
manner which is predictable and measurable. And, in the case of

learning machines, the S-R model may be quite useful. The learner is

28

29

put into a controlled environment in which competing stimuli are
eliminated; then, the learner is presented with a stimulus or aseries
of stimuli. His "correct" responses are rewarded in a manner that
Shapes his behavior. The end result is that his reSponse to a given
set of stimuli is changed and the change is relatively enduring - in a
word, he has "learned."

But in many communication situations, the S-R model doesn't

seem to help theorists construct a theory of communication. In the
everyday world, many stimuli (cues, messages) compete for the attention
of the individual. Furthermore, a communication system involves more
than one person. A patient talks to a therapist, a student talks with

a teacher, a husband talks to his wife or, on a higher systems level,

cne company deals with another or one nation conducts diplomatic relations
saith another. These communication systems are more complex than the S-R
situation in that the stimuli available for selection are more numerous
and the reSponseS to the stimulus are also more numerous.

There is, of course, a theoretic issue involved here. Many S-R
theorists argue that there is nothing intrinsically limiting about the
S-R approach. They would say that complex S-R models - taking into
account hierarchies, etc. - will eventually handle these problems
that have been raised. They would say that the criticism offered is
just a criticism of research to date, not a criticism of the heart of
the S-R approach. Others would argue that S-R is intrinsically limited
and could never get to these problems. This paper views the General
Systems approach as appearing to be more fruitful for examining complex

communication systems.

30

As an example of a system, let us consider a boat that is rowed
by eight men with a ninth man guiding the boat by use of a steering car.
The boat plus the nine men may be considered a system, with eaCh man
being considered a subsystem. The kinds of statements we' might wish
to make about the system (boat plus men) would refer to Speed and
direction. Each man who is rowing is participating in the overall goal
of the system - getting from point a to point b. If one of the eight
men stops rowing, the velocity is affected somewhat but the direction
is not affected as much. But if the man at the helm stops steering,
the heading is affected more than the velocity.

It becomes easy to see the effect of varying the number of sub-
systems. If the boat were a large one - rowed by 32 men (sixteen on
each Side), then the effect of one oarsman (whether he rows or does not
row) becomes less important to the outcome of the overall system (boat
plus men). In this case, it is doubtful if the single oarsman could
change the heading or velocity no matter how hard or fast he rowed.

On the other hand, if there are only two oarsmen, the effect of one
oarsman is great on the heading and velocity of the overall system
(men plus boat).

In this example, it is also possible to illustrate levels,
differentiation of function of subsystems, and relative power over
systemgoals. (That is, the oarsmen can be taken as one subsystem -
with each man a component in this subsystem - while the helmsman
comprises another subsystem. The first is a power subsystem while the

second is a guidance subsystem).

31

In summary, the subsystems (men in the boat) COOperate in Sharing
a common goal or value. Each of them contributes his share to the goal
of the overall system - getting from point a to point b. In order for
the overall system to achieve the goal, the subsystems must each
contribute their share. If the overall system is large (many subsystems),
the effect of one particular subsystem may be small. If the overall
system is small (two subsystems), the effect of one of the subsystems
on the goal is likely to be quite large. And each subsystem is affected
by the others.

So far, it has been suggested that a communication system, in
many cases, may be more complex than suggested by an S-R learning
model; that a communication system must consider the total number of
living systems within it and that the minimum number seems to be two;
and that these systems plus the environment should be examined. The
manner in which information (in all its usages) is processed by the
living systems may be a useful focus for the study of communication.

If we view communication as a process involving systems, we can
look to General Systems Theory for guidance in an analysis of a communi-
cation system. One writer in General Systems Theory, Kenneth Boulding,
notes:

Almost every discipline studies the interaction of some

kind of "individual" with its environment. The individual

may be an electron, atom, molecule, crystal, virus, cell,

plant, animal, man, tribe, state, church, firm, corporation,

university, and so on. Each of these individuals exhibits

"behavior" action or change and this behavior is considered

to be related in some way to the environment of the indivi-

dual - that is, with other individuals with which it comes
into contact or into some relationship. (Boulding, p. 199)

32

Communication may be viewed as an interaction process between
at least two living systems. In order to study communication, it is
suggested, the system must be specified, and the environment in which
the system exists must be specified. The information (all usages)
produced is a function of the living system in the context of some
environment. Going back to the example of the number "80", the living

system, in one case, was a stockbroker and the environment is his

Au

caffice. In that context, the reply "80" becomes information3 to him.

Boulding posits nine levels of systems — each of which may have
some use in the study of communication systems. Each level incorporates
the characteristics of previous levels but has added characteristics
of its own. The sixth, seventh and eighth levels may be of particular
use in studying communication systems. The sixth is the animal level
in which we have the development of specialized sense receptors
(eyes, ears, etc.) and an enormous increase in the intake of information.
This information is channeled through a nervous system and is apparently
organized by the brain into what Boulding calls a knowledge structure
or "image." "Once the image structure is well established," says
Boulding, "most information received produces very little change in
the image."

At the seventh or human level, the system possesses self con-
sciousness. By this Boulding means that not only does man know but
he knows that he knows. At the eighth level - the level of social
organizations - the person in the organization is more of a "role"
than a person. Boulding would view any social system as a set of roles

tied together with channels of communication. At this level, he would

33

concern himself with content and meaning of messages, nature and
dimensions of value systems, etc.
Boulding suggests that living systems have agreed upon rules
for deciding what information to process and how to process this in-

formation. Constructs such as "role" and "value" suggest preferred

 

states of the system and of systems within the system. Boulding says
behavior is described in terms of restoration of these preferred

states when they are disturbed by changes in the environment. These

jinn. Arm-$.13; '

preferred states of the system may be considered Information”, Several

 

writers have suggested that we look at the relationship aspect of
communication as well as the content aspect. In is infbrmation about
states of the system or the relationship aSpect of communication.

Thayer suggests that information from the environment permits
the living system to adapt to the environment. The living system's
ability to take into account features and events of the environment
serves the purpose of establishing, maintaining, exploiting, or al-
tering some relationship between the living system and the environ-
ment. (Thayer, p. 33). Another living system is part of the environ-
ment to any given living system. When one living system behaves
(uses languages, gestures, etc.), the originator's product is nothing
but sense data to the other living system. The functional message the
second living system gets is an internal, subjective creation of the
second system.

Thayer suggests the only reality upon which an individual can
base his behavior is his own reality - his conceptual structuring of

the world, his model or image of it, his expectations about the future.

3‘4

The world we see and hear and comprehend depends on how we are
individually organized to convert raw event-data into information (all
usages), and the sort of conceptual systems we individually have to make
messages out of that information.

Thayer notes every individual must develop his own model of the
world either:
1. In and through his own individual sensory eXperiences of it, or
2. In and through his accumulation of standardized interpretations of

it - via education, socialization, observation, and imitation, etc.

 

The price of membership in a culture or a subculture is the abdication
of certain degrees of freedom by the individual (wittingly or not).
(Thayer, p. #8).

Other writers have suggested looking at man as a creature who
views the world in terms of theories he might have about the way
things are. Kelly's notion is that a person's processes are
psychologically channelized by the way in which he anticipates events.

Miller, Gallanter and Pribram title their book Plans and the Structure

 

of Behavior to suggest that man has expectations before he acts, then
he acts, then he checks to see if events conform to his plans.

A recent book echoes these thoughts even though the terminology
used is slightly different. Watzlawick, Beavin and Jackson see a
(untent and a relationship aSpect for any inter-communication between
individuals. Psychiatric symptoms are viewed as behavior appropriate
to an ongoing interaction. Such symptoms grow out of an interaction

with others rather than being caused by one's "insanity." In a similar

35

manner, a "leader" would also be a relationship growing out of a
particular communication situation.

Content refers to the information I1 and 13 derived from a
message. Relationship refers to the kind of message it is in terms
of relationship (In) between communicants. Watzlawick et al., say
communication is virtually interchangeable with behavior. All
behavior in an interactional situation has message value. It is
impossible to not communicate. An attempt to not communicate
communicates just that: "I do not wish to communicate." "Cause"
and "effect" begin to become meaningless in an interaction situation.
It is meaningless to describe any given behavior as the "cause."

An interactional system is one involving two or more communi-
cants seeking to define their relationship with each other. This
includes nOtions such as rejection ("I'm busy, leave me alone")
acceptance ("I'd love to go with you" or "that idea of yours is a good
one") disqualification ("close one's eyes and hope that he goes away",
evade the issue, etc.).

In terms of uncertainty reduction, I1+ reduces the uncertainty
that the living system might have concerning his relationship with
other living systems in the communication system. The "meaning" of
I“ is difficult to articulate and, hence, difficult for the
researcher to work with. For example, North Americans are often
exasperated with persons from other cultures when doing business with
them. The persons from South American countries, fOr example, do not

"get to the point" when conducting a business transaction. But, may

 

36

we not view this as a cultural difference in the relative emphasis on
the content and relationship aspects of communication? Perhaps the
main goal of the conversation in a South American country is to "define
the relationship." What a North American would call "getting to the
point" would be bad manners and cause social disapproval. This social
disapproval is an example of In which would be perceived by both persons
in the conversation as a defining of the relationship.

Out of the Hovland research we find that "social approval" or
"social disapproval" is a major factor in Opinion change or attitude
change. A person may conform (behave) in some manner because of the
threat of detection and punishment. It might be useful to keep
Hovland's "social approval" factor in mind and, viewing man as a living
system, look at communication, in part, as an attempt at two or more
living systems to define their relationship to one another. The "defining
of the relationship" may be one of the most important goals of the
attempt to communicate. Man may be viewed as a living system attempting
to relate to other living systems.

An area of communication research that is not devoted to the study
of language is currently labeled "nonverbal communication." It is
probably through nonverbal cues (nods, smiles, frowns and other behavior)
that we obtain most "social approval" or "social disapproval" messages.
If "approval/disapproval" cues are important in defining the relation-
ship and if "defining the relationship" is an important goal of communi-
cation, the study of nonverbal communication becomes as important as
the study of language. Nonverbal communication may be considered

as information“ about the state of the system.

 

37

Informationu and Learning

 

It may well be that In will be most difficult to work with in
communication research. Once a relationship has been established, it
seems to be difficult to change. Bateson suggests that learning to
define a new relationship may be the most difficult kind of learning.
He posits four levels of learning (Bateson, p. 183):

1. Simple learning or what we have called the ability to take something
into account.

2. When a person changes his ability to receive the yes or no answer
to a question (e.g. when a dog learns that a buzzer means future
meat powder).

3. Acquisition of "test wisdom or set learning". A person learns to
look out for sequences of a certain sort in his universe, which
includes external events and his own behaviors.

u. Problem of changes at the third level. To make changes at the
third level is painful and difficult.

Bateson feels that fourth level learning (or changes at the
third level) is a necessary part of human life. It obtains in courtship,
in psychotherapy, in any reconstruction of relationship. He feels we
don't know much about the phenomena.

To illustrate the problem, he tells of experiments by Alex Bavelas
in which subjects are given a set of buttons to push and told that when
they push them the correct way, a bell will ring. Actually, the bell
rings after 50 pushes, then after 45 pushes, then after 40 pushes and

so on until the number of pushes before the ring of the bell is 20.

 

38

Then, the subject is told that there is no relation between what he
is doing and the ringing of the bell - that the bell is geared to a
hypothetical learning curve. Subjects usually cannot be convinced of
this but will reassert their theory of the relationship between the
buttons and the ringing.

Several examples from other areas of communication research T
illustrate the usefulness of examining the relationship aspect of

communication. Cross cultural researchers, such as Rogers, find that

 

technical innovations are resisted in most cases until persons in the i'
target population perceive that their referent group begins to adopt
the new idea. As another example, we might consider T-Groups (training
groups). One function of these groups is to get participants to re-
examine their relationship with others (Bradford, Gibb and Benne, p. l).
The communication strategy involved is to change the environment - to
unfreeze the perceptions. In this manner the same raw data from the
environment can be reexamined for new interpretations. In Bateson's
terms, the purpose is to make changes at level 3.; to develop new
theories of viewing familiar phenomena. This area (In) deals with
learning to define relationships and learning to redefine relationships.
The relationships being defined may be viewed as:

l. Subsystems to one another

2. Communication system to the environment

3. Actions of the communication system to the goals of the system.

The next section discusses some areas of possible research in defining

'these relationships.

CHAPTER V

CONCLUSIONS AND IMPLICATIONS FOR RESEARCH

Table 1 shows the four usages of information that have been dis-
cussed in this paper. The matrix contains comments on each in terms
of locus, uncertainty, function, advantages of using information in the
particular sense and disadvantages of the particular usage. One
immediate suggestion is the level of analysis suggested by In - Information
about the state of the system. This suggests that constructs such as
attitude, opinion, belief and value be examined at a system level
consisting of two or more living systems rather than using them as
attributes of individuals.

Part of the difficulty with a discussion of this sort is the
many labels we use for "ignorance." If two football teams are evenly
matched on a number of criteria, considered to be predictors of winning
teams, but one team defeats the other soundly and we don't know why,
we say that one team was "motivated" more than the other. The word
"motivated" is a convenient label we use for an intervening variable
which we feel is there but which we don't know how to measure or
demonstrate in any publicly observable manner. It seems this may be
the case with constructs like "attitudes," "beliefS" and "values."
These are often labels for intervening variables which we posit to

explain why people behave as they do.

39

’LJ"

—‘-~«. a MC

 

40

In contrast, the systems approach is more like the sportscaster
watching the game. He watches the process - the two teams struggling
to define their relationship to one another. Then, he attempts to
identify the factors that contributed most to the outcome. This is
not to say that posited constructs are not used. A system is presumed
to have a goal (preferred state) such as the goal of getting from point a
to point b by the men in the boat. But the posited construct will be
in terms of the communication system rather than an attribute used to
eXplain why an individual does as he does. Study of communication, in
this sense, is a move away from the S-R model to the more complex systems
model.

An example was used earlier of the North American wishing
to "get to the point" whereas persons from other cultures may wish to
Spend more time defining the relationship. Communication systems may
be viewed more conveniently if the relative amounts of I3 and In
in them are Specified. A communication system consisting of a

patient and a therapist may be heavily oriented to In, for example,

whereas the communication system between a physics instructor and a
lecture section of 500 students would be oriented to 13' It may be
convenient to develop typologies of communication systems for purposes
of analysis.

Specifying the goal of a system is of particular importance
when building a system, which is the case when building information

retrieval systems. It would seem that a distinction could be drawn

 

'41

between libraries (devoted mainly to storing I2) and information
retrieval systems (perhaps more conveniently viewed as oriented to 13),
Rather than build an IR system by storing documents first, it may be

more useful to identify a pOpulation of users, determine the kind and

 

E
frequency of questions they ask, and build a system of answers to the ;
questions. This means that questions of relevance of the answers,
acceptability of the answers and other measures of system effectiveness
might be worked out before the system is built. '

In the classroom situation, we may wish to have teachers (persons)

when the emphasis is on In. Some content areas might be well enough
Specified so that we may consider them primarily 13 and relegate them
to more automated teaching techniques. This presupposes that those
building the teacher-student communication system have some means of
specifying beforehand the relative emphasis on I3 and In for different
subjects. It also allows us, as Bruner suggests, to think about ways
and means of developing in students the ability to create I2 - in both

the artistic and the scientific sense.

u2

.oHSmmoE

0» pasowmmwo
ma pa .msnu
.mdoswwnam

soumo mm :H

osofianoop
wanSmmos
coo: ooonmm oz

onwamsoflumnooo
ou pasowmmwa

meowpmSwa sowumo
qusesoo more
now OH A3 “8.5

you moooo

 

mowmucw>ommwo

cofiumu
quseeoo
mo nowpnoo
owanOwuma
no.» masons.

coHumowcseeoo
mo cprnoa
unoucoo mesons

mH gnomes
ow onmoemum
mm mm:

mamzdmcm moon
oaaonemm
now @006

 

mommucm>o<

 

Eoumzm

mo mamom

ow mEoumhm
mo meowuo< .m

peoeconw>ao
Op Eopm>m .N

nosuonm oco
ow manpmzmnsm .H
“one
awesome meson
wowsmcowumaom

pcoeconw>co
was op power 09
conned made:

omno>qu
was camaoxo on
posouum m.:mz

maonezm
weapon now
moans menosoo

acmuocsm

.ouo .oumum cannon
none .oouoooxo mm
was: no soumhm on“
mo mopmpm ou movmaom

Awsfiumou some o>wpfiewoov
venomous on non usmueoo

canonsmmoe
hawomon poz

mo>wpmcnouam
was so spaflwaasopa ace
wm>fiumcnopam mo cowuonsm

 

hucfimvnooe:

AmEopmmmnsm
shoe no
oz: 53 ohm

Aswumhm
maa>nav
cornea

=80£Q80G=

.maomnoo
ow Hmcnouxm

oooo .Hoeemno

 

mecca

 

:cofiumenowcHz mo mow: snow

.H manna

BIBLIOGRAPHY

Ackoff, R. L., "Toward a Behavioral Theory of Communication," Management
Science, 1958, 3’ 218-23“.

 

Attneave, F.,.Applications of Information Theory to Psychology) New York:
Holt, Rinehart and Winston, 1959.

 

 

Bar-Hillel, Y., "An Examination of Information Theory," Philosophy of ’3
Science, 1955, 23, 86-105. -

Bateson, 6., "Exchange of Information about Patterns of Human Behavior,"
in W. 5. Fields and W. Abbott, (Eds.), Information Stgrgge and
Neural Control, Springfield, 111.: C. C. Thomas, 1963, pp. 173-186.

 

 

 

Beer, S., Cybernetics and Management, New York: Wiley, 1959. i

 

Berlo, D. K., "The Process of Social Research," unpublished manuscript.
Department of Communication, Michigan State University,_1967.

Berlyne, D. E., Conflict, Arousal and Curiosity, New York: McGraw-Hill,
1960.

 

Boulding, K. E., "General Systems Theory - The Skeleton of Science,"
Management Science, 1956, 33 197-208.

 

Bradford, L. P., Gibb, J. R., and Benne, K. D., T-Group Theory and
Laboratory Method, New York: Wiley, 196u.

 

Brehm, J. W., and Cohen, A. R., Explorations in Cognitive Dissonance,
New York: Wiley, 1962.

 

Brillouin, L., "Emperical Laws and Physical Theories: The Respective
Roles of Information and Imagination," in M. C. Yovits,
G. T. Jacdbi and G. D. Goldstein, (Eds.), Self-Organizing
Systems 1962, Washington, D.C.: Spartan Books, 1962a, pp. 231-2u2.

 

Brillouin, L., Science and Information Theory, New York: Academic
Press, 1962b.

Brillouin, L. , "Thermodynamics and Information Theory," American
Scientist, 1950, _3_8, sen-599. _——'

Broadbent, D. E. , "Information Theory and Older Approaches in
Psychology," Proc. 15th International Congress in Psychologz
Brussels, 1957, 111—115.

Bruner, J. S., The Process of Education, New York: Random House, 1969.

 

43

1414

Cherry, C., On Human Communication, New York: Wiley, 1957.

 

Conant, J. , Science and Common Sense, New Haven: Yale University Press,
1951.

 

Dervin, B., "A Critical Review of Research Relating Attitudes and
Attitude Change to Prior Information, Information Gain, and
Retention," unpublished paper. Department of Communication,
Michigan State University, 1967.

Garner, W. R., Uncertainty and Structure as Psychological Concepts,
New York: Wiley, 1962.

 

‘ ”“237
.i' in!

Hovland, C. I., Janis, I. L., and Kelley, H. H., Communication and
Persuasion, New Haven: Yale University Press, 1953.

 

 

Katz, D., and Kahn, R. L., The Social Psychology_of Organizations, E
New York: Wiley, 1966.

Kelly, G. A., The Psychology of Personal Constructs, New York:
Norton, 1955.

MacKay, D. M., "Quantal Aspects of Scientific Information,"
Philosophy Magazine, 1950, 3;, 289-311.

 

Mednick, S., Learning, Englewood Cliffs, N. J.: Prentice-Hall,
1964.

Miller, G. A., "The Magical Number'Seven Plus or Minus Two,"
Psychological Review, 1956, 63, 81-97.

 

Miller, G. A., Galanter, E., and Pribram, K. H., Plans and the
Structure of Behavior, New York: Holt, Rinehart and Winston,
1960.

 

 

Miller, J. G., "Living Systems: Basic Concepts," Behavioral Science,
1965a, 32, 193-237.

 

Miller, J. G., "Living Systems: Cross-Level Hypotheses," Behavioral
Science, 1965b, 12, 380-”11.

Miller, J. G., "Living Systems: Structure and Process," Behavioral
Science, 1965c, 39, 337-379.

 

Rogers, E. M., Diffusion of Innovations, New York: The Free Press, 1962.

 

Rudner, R. S., Philosophy of Social Science, Englewood Cliffs. N.J.:
Prentice-Hall, 1966.

 

Shannon, C. E., and Weaver, W., The Mathemgtical Theory of Communication,
Urbana: The University of Illinois Press, 19u9.

1+5
Shibutani, T., Improvised News: A Sociological Study of Rumor,
New York: BobbS-Merrill, 1966.

Smith, M. B., Bruner, J. S., and White, R. W., Opinions and Personality,
New York: Wiley, 1956.

 

Tannenbaum, P. H., and Greenberg, B. S., "Mass Communication,"
Annual Review of Psychology, 1968, 223 351-386.

Teilhard de Chardin, P. , The Phenomenon of Man, New York: Harper, 1961.

.3]

gen-

Thayer, L. , Communication and Communication Systems, Homewood, Ill.:
Irwin, 1968.

-___
-

Watzlawick, P., Beavin, J. H., and Jackson, D. D., Pragmatics of
Human Communication, New York: Norton, 1967.

 

 

Weaver, W. , (See Shannon and Weaver). . b

Wiener, N., gybernetics, Cambridge, Mass.: The M.I.T. Press, 191:8.