American Academy of Political and Social Science

Cultural Indicators: The Case of Violence in Television Drama

Author(s): George Gerbner
Source: Annals of the American Academy of Political and Social Science, Vol. 388, Political
Intelligence for America's Future (Mar., 1970), pp. 69-81
Published by: Sage Publications, Inc. in association with the American Academy of Political and
Social Science
Stable URL: http://www.jstor.org/stable/1038317
Accessed: 10-04-2015 19:24 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact [email protected].

Sage Publications, Inc. and American Academy of Political and Social Science are collaborating with JSTOR to digitize,
preserve and extend access to Annals of the American Academy of Political and Social Science.

http://www.jstor.org

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

Cultural Indicators: The Case of Violence in

Television Drama
By GEORGEGERBNER

The cultural transformation of our time stems

from the extension of the industrial-technologicalrevolution
into the sphere of message-production. The mass production
and rapid distributionof messages create new symbolic environments that reflect the structureand functions of the institutions
that transmit them. These institutional processes of the massproduction messages short-circuit other networks of social
communicationand superimpose their own forms of collective
consciousness-their own publics-upon other social relationships. The consequencesfor the quality of life, for the cultivation of human tendenciesand outlooks, and for the governingof
societies, are far-reaching. Informed policy-making and the
valid interpretation of social behavior require systematic indicators of the prevailing climate of the changing symbolic environment. A central aspect of cultural indicatorswould be the
periodic analysis of trends in the composition and structure of
message systems cultivating conceptions of life relevant to socialization and public policy. Findings of studies of the portrayal of violence in network television drama illustrate the
terms of such analysis, and demonstrate the need for more
comprehensive, cumulative, and comparative information on
mass-culturaltrends and configurations.
ABSTRACT:

George Gerbner, Ph.D., is Professor of Communications and Dean of the Annenberg

School of Communications, University of Pennsylvania. He has been principal investigator on international communication research projects sponsored by the United States
Department of Health, Education, and Welfare; UNESCO; the International Sociological
Association; and the National Science Foundation. Recently, he directed the study of
violence in television drama for the National Commission on Causes and Prevention of
Violence, and he is now continuing that research for the Scientific Advisory Panel of the
Surgeon General.
69

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

70

THE ANNALS OF THE AMERICAN ACADEMY

HE Nation has no comprehensive are made with as little reliable, system-

set of statistics reflecting social

progress or retrogression," begins the
recent Department of Health, Education and Welfare (HEW) report to the
President.1 The report recommendeda
procedure for the periodic stock-taking
of the social health of the nation. Steps
in that direction include pending legislation and the growing literature on social
and political accounting; the recent
White House order setting up a National Goals Research Staff, charged,
among other things, with "developing
and monitoringsocial indicators that can
reflect the present and future quality of
Americanlife, and the direction and rate
of its change"; and this issue of THE
ANNALS itself.

My purpose is to develop and illustrate a frameworkfor cultural indicators

as one aspect of social accounting.
Cultural indicators have been alluded to,
but have not yet been articulated in
policy statements, legislative proposals,
or the research literature. I would first
like to outline the case for indicators of
the mass-produced symbolic environment that I call the common culture.
Then I shall describe a central aspect of
the framework for such indicators, and
illustrate the framework with our
study of television violence 2 and related
research.
THE CASE FOR CULTURAL INDICATORS

There is no area of significant social

policy in which far-reaching decisions
1 U.S., Department
of Health, Education
and Welfare, Toward a Social Report (Washington, D.C.: Government Printing Office,
1969) p. xi.
2 George Gerbner, Martin Brouwer, Cedric
C. Clark and Klaus Krippendorff, "Dimensions of Violence in Television Drama," The
Annenberg School of Communications, University of Pennsylvania, 1969 (hereinafter referred to as the "television violence study").
However, some findings are presented in this
paper for the first time.

atic, cumulative, and comparative information about the actual state of affairs
as in the sphere of the mass production
of the common culture. Confused by
our own rhetoric of some automatic
mechanisms at work in some mythical
marketplace of ideas, we are only
vaguely aware of the fact that decisive
policy-making is going on, and that
cultural politics is as much a part of the
fabric of modern life as economic, welfare, or military politics.3 Debates
about "censorship"obscure the realities
of direction, constraints, and controls in
the mass production of messages. Application of formal aesthetic categories
derived from other times and places
ignore functions, resources,and power at
the heart of the cultural process.
We know next to nothing about trends
in the composition and structure of
mass-produced message systems that
govern men's lives and inform men's
minds in urbanized societies. We know
little more about the institutional processes that compose and structure those
message systems. Consequently, much
of our high-powered research on how
people respond and behave in specific
situations is unenlightened by insight
into the common cultural context in
which and to which they respond.
Historically, we are dealing with a still
galloping industrial revolution in methods of producing and distributing messages. The rise of mass communication
is a profound change in the management
of information, and in the creation of
the common symbolic environment that
gives public direction and meaning to
3Governments, Presidents, and, more recently, a Vice-President, usually call attention
to this fact when a deep split in economicmilitary-communications policy-making threatens their ability to cultivate mass support, or
at least acquiescence. Any campaign to mobilize a "silent majority" of the "forgotten man"
is an attempt to force the media to publicize
views that such a campaign expects to elicit.

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

CULTURAL

INDICATORS:

VIOLENCE

human activity. The purpose of a

scheme of cultural indicators is to monitor those aspects of our system of generating bodies of broadly shared messages that are most amenable-and
most relevant-to public policy decisions, and to take the pulse and measure
the tempo of their transformations.
Selective habits of participation limit
each of us to risky and usually faulty
extrapolation about the cultural experience of different or heterogeneous communities. The reliable observation of
regularities in large message systems is
a specialized enterprise that requires
not only methodological sophistication,
but also a clear conception of dimensions
of analysis and of relevance to investigative purpose. What I have called
elsewhere the institutional approach to
mass communications research4 is the
basis of such an enterprise. It is the
study of technologically mediated message systems and processes as historically new ways of looking at life, as new
forms of institutionalized public acculturation, and the broadest common
bases of social interaction and policyformation in modern societies. Such
study revolves around problems of message system theory and analysis, institutional process analysis, and the investigation of relationships between message
systems, corporate forms and functions,
collective image-formation, and public
policy. It asks these questions:
What perspectives and relationships
are expressed in message systems produced for large and diverse communities? How do these systems vary over
time, across cultures, and in different
societies? How do media compose and
structure these message systems? How
is the mass-production and distribution
4George Gerbner, "An Institutional Approach to Mass Communications Research," in
Lee Thayer ed., Communication: Theory and
Research (Springfield, Ill.: Charles C. Thomas,
1966).

IN TELEVISION

DRAMA

71

of messages organized, controlled, and

managed? What institutional and technological functions and what organizational decision-makingprocesses govern
the production and distribution of these
message systems? What common assumptions do message systems cultivate
over and above those apparent in single
or selected messages or individual and
selective responses? And, finally, how
does the cultivation of these collective
assumptions shape the conduct of public
affairs (and, of course, vice versa)?
The questions designate three areas of
analysis. Study of the composition and
structure of large bodies of mass-mediated messages is the analysis of message
systems. Study of the organizational
forms, functions, and decision-making
that composeand structurethese systems
is what I called institutional process
analysis in mass communications. And
study of the relationships between institutional processes, message systems, and
the public assumptions, images, and
policies that they cultivate is what we
may call cultivation analysis.
Social research into the "behavioral
effects" of communications might be
seen as having concentrated on the last
area of studies. Yet, the area of cultivation analysis is perhaps the least developed. The reason is that most "effects"
research stemmed from theoretical perspectives that did not consider relevance
to the mass-cultural process a principal
criterion. From the point of view of
cultural indicators, therefore, such research will be inadequate to the task until institutional process and messagesystem analysis can provide the necessary framework of common terms and
the relevant dimensions to be investigated.
Institutional process-analysis has a
researchbase in organizationtheory and
studies of managementpolicy-formation.
It is a more focused and limited area of
study. But research on the policy proc.

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

72

THE ANNALS OF THE AMERICAN ACADEMY

ess in mass communications is scarce.5

More importantly, such research cannot
serve the purposes of cultural indicators
until media policies can be related not
only to theories of organization and
decision-making, but also to specific
characteristics of message systems that
they shape.
The analysis of message systems is,
then, the starting point of researchleading to cultural indicators. It is the step
that must be taken before the study of
the institutional-policy and public-cultivation processes can proceed on coherent
terms. The central dimensions of cultural analysis stem, not from intentions
or policies or individual cognitions, but
from the actually shared messages that
mediate public perspectives and provide
such common bases for social interaction (both dissent and consensus) as
shape the course of public events.
MESSAGESYSTEMANALYSIS

The analysis of message systems rests

on the conception of the role of communication in human life. Communication is interaction through messages.
Messages are specialized events (or aspects of events) that signify other things
in enormously varied and creative ways
unique to human culture.6 Social inter5 Some illustrations and the beginnings of a
scheme for process analysis are contained in
George Gerbner, "Institutional Pressures upon
Mass Communicators," in Paul Halmos, ed.,
The Sociology of Mass-Media Communicators,
Sociological Review Monographs, no. 13 (University of Keele, England, 1969), pp. 205-248.
6An earlier development of this definition
and its implications may be found in George
Gerbner, "On Content Analysis and Critical
Research in Mass Communication," 6 AV
Communication Review (Spring 1958), pp. 85108, reprinted in Lewis A. Dexter and David
M. White, People, Society and Mass Communications (New York: Free Press, 1969),
pp. 476-501. The present discussion of message system analysis is a further development
of ideas presented in George Gerbner, "Toward
'Cultural Indicators': The Analysis of Mass-

action through such symbols is the "humanizing" process of our species. That
process creates the symbolic environment from which behavior derives its
distinctively human significance. It also
cultivates man's notions of the facts and
potentials of existence, his orders of priorities and ranges of values, and the
clusters of associations among all these
dimensions of imagery and imagination.
The terms of our analysis stem from
and relate to the dimensions of common
consciousness that mass-produced message systems cultivate in large and
heterogeneouspublics. We have identified these dimensions as message-mediated assumptions about existence, priorities, values, and relationships. Table
1 summarizes the questions, terms, and
measures of analysis relevant to each
dimension.
The dimension of assumptions about
existence deals with the question "What
is?," that is, what is available (referred to) in public message systems at
all, how frequently, and in what proportions. The availability of shared messages defines the scope of public attention. The measure of attention, therefore, indicates the presence, prevalence,
rate, complexity, and varying distributions of items, topics, themes, and the
like, representedin message systems.
The dimension of priorities raises the
question "What is important?" We may
use measures of emphasis to study the
context of relative prominence and the
order or degrees of centrality or importance. Measures of attention and emphasis may be combined to indicate, not
only the allocation, but also the channeling attention in a message system.
The dimensions of values inquires
Mediated Public Message Systems," in George
Gerbner, Ole R. Holsti, Klaus Krippendorff,
William J. Paisley and Philip J. Stone, The
Analysis of Communication Content (New
York: Wiley & Sons, 1969), pp. 122-132.

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

CULTURALINDICATORS:
VIOLENCE IN TELEVISIONDRAMA

73

TABLE I-DIMENSIONS, QUESTIONS,

TERMS,AND MEASURESOF
MESSAGE

DIMENsloNS

EXISTENCE

SYSTEM ANALYSIS

PRTORITIES

VALUES

RELATIONSHIPS

Assumptions
about:

WHAT IS?

WHAT IS
IMPORTANT?

Questions:

What is available
for public attention? How much
and how frequently?

In what context or In what light,

order of importfrom what point
ance?
of view, with
what associated
judgments?

In what over-all
proximal, logical, or
causal structure?

Terms and
measures of
analysis:

ATTENTION
Prevalence, rate,
complexity, variations

EMPHASIS
Ordering, ranking,
scaling for prominence, centrality,
or intensity

STRUCTURE
Correlations, clustering; structure of
action

into assumptions about right and wrong,

good and bad, and other associated qualities. It asks about the point of view
from which things are presented, about
the characteristics, traits, or connotations attached to different items of reference and emphasis. Measures of tendency are used to assess the direction
and intensity of value judgments observed in messages.
The dimensionof relationshipsfocuses
on the more complex associations within
and among all measures. When we deal
with patterns of attention, emphasis, or
tendency, instead of only simple distributions, or when we relate the clustering
of measures to one another, we illuminate the underlyingstructure of assumptions about existence, priorities, and values representedin message systems.
The four dimensions, then, yield measures of attention, emphasis, tendency,
and structure. I shall illustrate some of
the terms of each measure with findings
of research on mass media violence, and
suggest how such research might be developed to serve the tasks of indicators
of the role of cultural production in
image-cultivationand policy-formation.

WHAT IS
RIGHT OR
WRONG, GOOD
OR BAD, ETC.?

TENDENCY
Measures of critical and differential
tendency; qualities,
traits

WHAT IS
RELATED TO
WHAT, AND
HOW?

Measures of attention
A mass-produced message system is
the result of institutional processes selecting some things to be brought to public attention and ignoring or rejecting
others. Measures of attention indicate
the presence and distribution of subjects, topics, themes, and the like, selected to compose the system. Knowing
something about the distribution of attention over time and across cultures is
an elementary measure of the most commonly available fund of raw materials
out of which each age and place weaves
its own patterns of public imagination
and imagery.
Focusing on an issue such as violence,
we can ask how prevalent its representation is; what rate per natural contextunit, for example, story or play, it occurs; and how its frequency varies by
different categories of analysis within
systems and across systems.
The evidence, of course, is scattered,
fragmentary,and rarely comparable. In
one of the first studies, Edgar Dale
found that "crime" prevailed in 84 percent of the movies of the early 1930's,

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

74

THE

ANNALS

OF THE

(an average of 3.9 per picture), and

violent death in 39 percent.' Our own
researchnoted violence in 66 percent of
some three thousand Hollywood movies
produced between 1950 and 1961.8
Studies of television drama in the
early 1950's found that "in over threequarters of the plays, acts of violence,
crime, or aggression occurred,"'9 that
violent acts "predominated"in 56 percent or 3.7 per play,10 and that the
hourly rate doubled between 1952 and
1954.11

Our television violence study analyzed

plays, cartoons, and feature films telecast nationally during study periods representative of 1967 and 1968 network
programming. The study found violence
portrayed in eight out of every ten plays
for both years. The rate of violent episodes was 5.0 per play in 1967 and 4.5
in 1968. An average dramatic program
hour had seven, and a cartoon hour,
twenty-two, violent episodes. Half of all
leading characters committed, and six
out of ten suffered, some violence. One
in ten became a killer. One in twenty
was killed.
Variations by other categories begin
to form patterns of life manifest in the
world of television drama. The following highlights from the findings of our
1967-1968 television violence study
come from measures of attention across
such variables as personal and demographic aspects, social environment,
place, and time.
7 Edgar Dale, The Content of Motion Pictures (New York: The Macmillan Company,
1935), pp. 133-134.
8 Unpublished research data.
9 Sidney W. Head, "Television and Social
Norms: An Analysis of the Social Content
of Television Drama" (PhD. dissertation, New
York University, 1953), p. 2.
10oSidney W. Head, "Content Analysis of
Television Programs," 9 Quarterly of Film,
Radio, and Television, 1954, pp. 184-185.
11 H. H. Remmers, Four Years of New York
Television (Urbana, Ill.: National Association
of Educational Broadcasters, 1954), pp. 37-38.

AMERICAN

ACADEMY

* Violence stuns, maims, and kills

without much visible "hurt."
Suffering was difficult to detect,
making violence appear painless,
despite the "body count" of
about 400 casualties per week or
an average of five per violent
play.
* When witnesses appear on a
scene of violence, they are passive
in seven out of ten such episodes.
When they are not passive, they
are as likely to assist or encourage as to try to prevent violence.
* The "generationgap" looms with
a vengeance. One young adult
gets killed for every five young
killers. One middle-aged character gets killed for every two
middle-aged killers. Old people
rarely appear in television plays
(6 percent of all characters), and
even more seldom kill (2 percent
of all killers); but two old men
were killed for one who was a
killer. To look at only killers
and their fate; one young killer
out of fourteen is himself the
fatal victim of violence; one middle-aged killer out of five gets
killed; the old man who kills is
killed in return.
* Social class, too, makes a difference. Half of all upper- and
middle-class characters, but
three-quartersof lower-classcharacters, commit violence. Six out
of ten upper- and middle-class
characters, but nine out of ten
lower-class characters, fall victim
to violence. Almost one-third of
upper- and middle-class characters escape both violence and
victimization; none of the lowerclass characters does. When it
comes to fatal violence, upperclass killers number 11 percent,
middle-class killers, 12 percent,
and lower-classkillers, 16 percent

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

CULTURAL

INDICATORS:

VIOLENCE

of all characters in their class.

Middle-class killers outnumber
middle-class killed 3 to 1, but for
every upper- and lower-class killer, there is an upper- and lowerclass character killed.
* Ethnic and race distinctions are
related and equally striking.
The violents comprise half of all
white American characters, six
out of every ten white foreigners,
and two-thirds of all nonwhites.
The same groups suffer from violence in the same order, with
nearly six out of ten whites, but
eight out of ten nonwhites falling
victim to some violence. Both
inflicting and suffering violence
themselves were 39 percent of
white Americans, 46 percent of
white non-Americans,and 60 percent of nonwhites. Escaping
both violence and victimization
were 35 percent of white Americans, 32 percent of white nonAmericans, and only 13 percent
of nonwhites. However, the
proportion of killers was one out
of every five white foreigners,one
out of every eight white Americans, and one out of every fifteen
nonwhites. There seems to be
something like "violent efficiency" at work here, too: onethird of all violent foreign whites
and nearly one-fourth of all violent white Americans, but only
one-tenth of all violent nonwhites,
succeed in killing an opponent.
The pattern of fatal victimization
also shows that while white
American killers outnumbered
killed 4 to 1, and white foreigners
3 to 2, for every nonwhite killer
there was a nonwhite killed.
* Time and place also affect the
patterns of violence and justice
implicit in the way things work
out in the world of television.

IN TELEVISION

DRAMA

75

Violence was involved in threefourths of all plays in a contemporary or domestic setting. But
it was featured in 98 percent of
plays set in the past, every single
play set in the future, and 92
percent of all plays depicting foreign lands or people. To look at
it another way, the past was nine
times as likely to be depicted
when there was some violence
than when there was none. The
future was always violent. The
outside world was three times as
likely to be violent as nonviolent. And, as we have seen, the
violence extracts a higher pricea tooth for a tooth-from its
predominantly violent strangers
than it does from its violent native whites.
Crude as they are, these patterns begin to lay bare some assumptions cultivated in these message systems. They
also begin to give substance to the contention that without a more specific
knowledge of these assumptions about
the role of violence, researchon "effects"
may be shallow and misdirected.
Measures of emphasis
"Emphasis" is that aspect of the
composition of message systems which
establishes a context of priorities of importance or relevance. The distribution
of emphases sets up a field of differential
appeal in which certain things stand out.
Emphasis orders the agenda of public
conceptions and discourse cultivated in
message systems. Measures of emphasis may be based on indications of size,
intensity, or stress, or on the featuring
of certain topics or themes as the major
points of stories, and are usually expressed in ranks or ratios.
Measures of emphasis are most useful
when it is necessary to examine orders
of priority channelingor even command-

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

76

THE ANNALSOF THE AMERICAN

ACADEMY

ing attention independently of terms of

prevalence and rate of messages available for attention. For example, in an
international study,12 we found that the
ratios of films in which violence was
"essential to the plot" to those in which
violence was "secondary" was higher in
Polish films (2.5) than in United States
films (1.3), despite the fact that the
prevalence of violent portrayals was
higher in the United States films. Yugoslav films with a very high, and Czechoslovak films with a very low, prevalence
of violence both yielded the lowest emphasis ratios of six countries (0.8).
Some evidence of trends in emphasis
over time comes from television studies.
The 1952 television research found violent acts "predominating"in 56 percent
of plays analyzed.1" Our 1967-1968
study found violence integral to the plot
(essential for a brief plot description)
in 66 percent of the plays in 1967 and
56 percent in 1968. However, when
television drama portrayed violence, it
emphasized it as an essential plot element in from seven to eight out of
every ten plays in both the early 1950's
and the late 1960's.

The broadest over-all dimension of

judgment is a summaryevaluation of the
goodness or badness, rightness or wrongness, of things. A measure of the favorable-unfavorable associations expressed
in message systems may be called "critical tendency." It is based primarily on
whether a subject or topic appears in a
supportiveor critical context. But value
judgmentis, of course, multidimensional.
Measures of "differentialtendency" can
be used which indicate directionality of
judgment in many differentdimensions.
In fiction and drama, tendency may
be implicit in justifications for action,
consequences of action, and the characterization of persons assigned different
roles and fates. The international film
study cited above found that principal
charactersin all countries' films resorted
to violence for defense, protection, and
other sanctioned or legal ends from over
two to more than four times as frequently as for illegal or immoral ends.
But illegal and immoral reasons for violence were portrayed twice or more as
frequently in the West as in Eastern
Europe. The ratio of the tendency to
commit "socially sanctioned" violence
versus antisocial violence was 2.5 in
Measures of tendency
French films, 2.6 in United States films,
The position of an institution (or an 3.5 in Italian films, 3.8 in Czechoslovak
individual) in time and space, and in films, 4.1 in Yugoslav films, and 4.4 in
the over-all structure of social rela- Polish films. (The higher the ratio, the
tions, enters into the approach, point of lower the proportion of antisocial vioview, or direction from which that insti- lence.)
tution or individual presents aspects of
The 1967-1968 television study traced
existence. The investigation of "ten- the tendency to present violence as seridency" deals with the explicit or con- ous, or as slapstick or "just for fun."
textual value judgments and other quali- Four-fifths of all violent episodes (inties that such vantage points impart to cluding comedy) were judged "serious."
The proportionof "serious"episodes demessage systems.
clined from 87 percent of all violent
12 Based on research data for George
scenes in 1967 to 74 percent in 1968.
Gerbner,"The Film Hero: A Cross-Cultural The more
complex judgments inherent
Study," University of Minnesota Journalism
in the world of dramatic violence can
Monographs, no. 13 (Minneapolis: University
be illuminated by combining dimensions
of Minnesota, 1969).
:s Head, "Television and Social Norms."
again to weave fuller patterns of life.

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

CULTURAL INDICATORS: VIOLENCE IN TELEVISION DRAMA

Measures of structure
Systematic measurement of patterns
of relationships requires the analysis of
correlationsor clusters among measures.
Such methods depend on the development of sophisticated and flexible computer programsthat can reveal complex
interrelations of different types of data.
I am limited to the few relatively simple
examplesavailable to illustrate some possibilities. These examples, from the
1967-1968 television violence study,
involve certain aspects of purpose, characterization, role, and fate.
* Most violence is interpersonal
and at close range, but relatively
impersonal. Strangers assault
each other for reasons of private
gain, power, or duty. In a world
of specialized relationships, violence is one more specialty that
rarely involves intimates and seldom stems from great emotion
or from fighting for a noble cause.
* Happy are the good guys and
unhappy the bad (at least in the
end). Good guys initiate as
much violence as bad guys, but
hurt less and kill less. Good guys
suffer more from violence, but
heroes never die. Bad guys get
hurt less than good guys, but, of
course, they lose out in the end.
* Half of all killers are good guys
who reach a happy end in the
stories.
If virtue suffers more than evil, the
ultimately happy hero must be more
decisive and efficient to triumph in the
end. "Personality-differential" scales
used to measure the intensity of selected
character traits support that inference,
and add more dimensions to the structure of judgments.
Figure 1 charts the mean scores of all
violent characters, all killers, and all

77

those who commit no violence, on scales

found to be reliably used by analysts
on the television violence study. Aside
from the sex differencewhich results, in
part, from the larger proportion of
women among nonviolents, the traits
that distinguish nonviolents from violents and killers are "usualness," efficiency, attractiveness, emotionality, and
logic. Nonviolents are more usual, less
efficient, more emotional, and less logical
than violents. Killers are the most efficient and the least emotional. (We shall
come to "attractiveness"later.)
Now let us see how a happy fate-practically identical with virtue- affects
the portrayals. Figure 2 compares violents and nonviolents by fate. The
greatest difference, aside from violence
itself, is that the "happies" are attractive whether violent or not. The "unhappies" are repulsivewhen violent.
Violence does not mar, nor nonviolence improve, the attraction of the
happy hero. But violence does appear
to have logic on its side. And the key
to violent happiness is efficiency. Let
us take a closer look.
Attractive characters reach a happy
end, (or vice versa) whether violent or
not. Attractive, happy, and violent
characters are also the most logical, but
the repulsive, unhappy, violent characters are nearly equally logical. Thus, all
violents are more logical than all nonviolents. Nonviolents-whether happy
or unhappy-are more emotionaland intuitive than violents. Unhappy nonviolents are also more irrational. Cool
logic is all on the side of violence.
Violence with a happy ending, furthermore, is a matter of efficiency. Superior efficiency separates happy violents from all others. Efficient, cool
logic is the unbeatable combination that
makes for the happy violent hero.
Characters can remain happy and lack
efficiency only if nonviolent. Violent

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

78

THE ANNALS OF THE AMERICAN ACADEMY

heroes must be efficient to win, while

nonviolent heroes may, and all villains
(a little older than the rest), must bungle the job.
Further confirmation comes from a
comparisonof happy and unhappy killers, and their, naturally, unhappy victims. (See Figure 3.) Unhappy killers
and killed present similar profiles because their numbers overlap; half of the
unhappy killers are themselves killed.
The happy killers stand out in manliness and attractiveness; they are heroes.
The happy killers are "first among
equals" in rational and unemotional
logic, for they, too, are killers. But the
happy killers are the most distinguished
FIGURE

1-PERSONALITY

in the quality they must possess to win,

even more than all violent heroes: efficiency.
Unhappy killers and all victims present, by comparison, an image of repulsive bunglers. The victims of fatal violence, half of them killers but some of
them innocents themselves, are also the
oldest and least masculine. Otherwise
they are almost as repulsive and bungling as are those of their killers who
come to an unhappy end.
Cool efficiency,and, to a lesser extent,
manliness and youth, appear to be the
chief correlates of success and virtue in
a fairly impersonal, self-seeking, and
specialized structure of violent action.

PROFILES OF CHARACTERS BY VIOLENCE-RELATED

IN TELEVISION DRAMA
3

7
_

OLD

ROLES

YOUNG

FEMININE

MASCULINE
..

..

UNUSUAL

USUAL
... ...

._

E
EFFICIENT

BUNGLING
.

ATTRACTIVE

REPULSIVE

EMOTIONAL

-lotUNEMOTIONAL

LOGICAL

INTUITIVE

RATIONAL

IRRATIONAL

---

VIOLENT

----

NONVIOLENT

S VIOLENTS
(N -240)
KILLERS(N-53)

I"

,,

ma

,,,

NONVIOLENTS(N
-215)

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

CULTURAL INDICATORS: VIOLENCE IN TELEVISION DRAMA

FIGURE

2-PERSONALITY
1

PROFILES

OLD

OF CHARACTERS BY FINAL
4

79

OUTCOME

7
YOUNG

__
.

MASCULINE

FEMININE

USUAL

UNUSUAL

BUNGLING

.EFFICIENT

_____-

REPULSIVE

ATTRACTIVE

EMOTIONAL

UNEMOTIONAL

\'
-

INTUITIVE

IRRATIONAL

VIOLENT

LOGICAL

RATIONAL

___

____

NONVIOLENT

_"_

- FINAL.OUTCOME:
VIOLENTS
HAPPY(N-126)
-VIOLENTS-FINAL
OUTCOME:
UNHAPPY
(N-61)
i
,,nB,,-,NONVIOLENTS
- - - --

- FINAL OUTCOME:
HAPPY(N-140)

UNHAPPY(N -28)
-NONVIOLENTS-FINALOUTCOME:

There is a need for more sensitive and

comprehensive indicators of the structure of assumptions cultivated in public-message systems than we have yet
been able to develop. What should they
encompass? What will they show? I
would like to conclude with some suggestions about the potentials and limitations of such development.
POTENTIALS

AND LIMITATIONS

Understanding the structure of cultivation is especially necessary when the

focus of attention is a complex issue of
life or policy. Violence, for example, is

not a single or simple dimension of behavior. Much of it is not a problem of

behavior at all, in the sense of violent
motion. It is equally a question of public assumptions about the role of force
and the distribution of justice; it is also
a problem of shared expectations of the
kinds and effects of violence that "we"
expect from different types of "others."
Violence involves resistance to or acquiescence in its private and public uses;
support for or opposition to policies related to its use; and other responses to
its distant, as well as to its nearby,
manifestations.
These are culturally learned assump-

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

80

THE ANNALS OF THE AMERICAN ACADEMY

tions and expectations. Indicators of

public-message systems give us a basis
for judging the role of cultural production in their cultivation. Such indicators
can also help place cultural policy in
perspective. "How much violence
should there be on television?" is a
shallow question, useful mainly for administrative purposes. If used alone, it
can only lead to a new rating game in
which networks compete for the lowest
score, regardless of its meaning.
"What kinds of violence, what for, and
in what contexts?" are the key issues.
Violence is a symptom of irreconciled
conflict, destruction, hurt, and waste.
It is, in a sense, the opposite of communication. It negates the most
FIGURE

3-PERSONALITY

uniquely human capacity of our species,

the capacity to interact and even collide
creatively through symbols and messages. Symbolic representation of violence is, therefore, a vital function of
informationand art in their illumination
of its real-life manifestations and consequences.
Institutional compulsions to present
life in salable packages exploit existing
assumptions about violence. Our television study found that portrayals of
violence mirror, rather than illuminate,
our society's prejudices. Cultural policies can expose the role of violence
or can use its images as instruments of
that very negation of humanity that it
represents. Indicators of the structure
OF "KILLERS"

PROFILES

AND "KILLED"

7
.-YOUNG

OLD
..

MASCULINE

FEMININE

UNUSUAL

USUAL.....---

EFF CIEfT

BUNGLING

REPULSIVE
. ..

--

ATTRACTIVE
UNEMOTIONAL

EMOTIONAL

INTUITIVE

LOGICAL

RATIONAL

IRRATIONAL

VIOLENT

in

NONVIOLENT

FINALOUTCOME:
KILLERSHAPPY(N- 25)
UNHAPPY(N-lb)
KILLERS-FINAL
OUTCOME:

e""'""m""""""""""
KILLED- FINAL OUTCOME:UNHAPPY (N-25)

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions

CULTURAL

INDICATORS:

VIOLENCE

of assumptionsimbedded in our message

systems can help make these distinctions.
They can also reveal trends in attention, emphasis, tendency, and structure,
and their relationship to cultural policy.
I have used the issue of violence to
illustrate aspects of the case, and of a
scheme, for cultural indicators. But
representationsof any aspect of life can
best be seen in the context of others,
across conventional distinctions of media
or modes of representation, and against
the background of changes in time and
place. Sources of information should
range from existing indices and guides
(often compiled for reference but usable
as trend-indicators) to regular monitoring of the massive flow of messages and
images. The accounting should begin
with the message systems most broadly
shared by the most heterogeneous publics. Cutting across all these sources,
media, modes, and forms should be the
common terms and categories of analysis. These should include categories as
comparableas possible to those studied
before and to those in other cultures, but
which are also sensitive to new and
changing issues of public policy. They
should survey the history, geography,
and demographyof the symbolic worlds
produced for common vicarious experience and learning. Interpersonal and
group relationships portrayed in these
message systems should be studied.
Themes of nature, science, politics, law,

IN TELEVISION

DRAMA

81

crime, business, education, art, illness

and health, peace and war, and sex,
love, and friendship, as well as conflict
and violence, should be analyzed. Roles,
traits, goals, values, and fates of characters engaged in dramatic action should
be related to the symbolic worlds in
which they act and the issues with which
they grapple.
The analysis of message systems can
provide a framework in which comprehensive, coherent, cumulative, and comparative information can be systematically assembled and periodically reported. Indicators relevant to specific
problems or policies can then be seen in
the context of the entire structure of
assumptions cultivated at a particular
time and place.
These indicators will not necessarily
tell us what people think or do. But
they will tell us what most people think
or do something about and in common,
and suggest reasons why. They will tell
us much about the shared representations of life, the issues, the prevailing
points of view that capture public attention, occupy people's time, and animate
their imagination. They will help to
understand, judge, and shape more intelligently the changingsymbolic climate
that affects all we think and do. We
can then inquire into who thinks and
does what, how, and why in sharper
awareness of the currents that tug and
pull us all.

This content downloaded from 132.248.9.8 on Fri, 10 Apr 2015 19:24:47 UTC
All use subject to JSTOR Terms and Conditions