Screening Violence

Screening Violence
Rutgers Depth of Field Series
Charles Affron, Mirella Jona Ajjron, Robert Lyons, Series Editors
Richard Abel, ed., Silent Film

John Belton, ed., Movies and Mass Culture
Matthew Bernstein, ed., Controlling Hollywood: Censorship and Regulation
in the Studio Era
John Thornton Caldwell, ed., Electronic Media and Technoculture
Peter Lehman, ed., Defining Cinema
James Naremore, ed., Film Adaptation
Stephen Prince, ed., Screening Violence
Valerie Smith, ed., Representing Blackness: Issues in Film and Video
Janet Staiger, ed., The Studio System
Linda Williams, ed., Viewing Positions: Ways of Seeing Film
Edited and with an introduction by
Stephen Prince
Rutgers
University
Press
New Brunswick,
New Jersey
Library of Congress Cataloging-in-Publication Data
Screening violence / edited and with an introduction by Stephen Prince,

p. cm. — (Rutgers depth of field series)
Includes bibliographical references and index.
ISBN 0-8135-2817-8 (cloth : alk. paper) — ISBN 0-8135-2818-6 (pbk : alk. paper)
1. Violence in motion pictures. I. Prince, Stephen, 1955- U. Series.
PN1995.9.V5 S395 2000
791.43'655—dc21 99-054344
British Cataloging-in-Publication Data for this book is available from the British Library
This collection copyright © 2000 by Rutgers, The State University

For copyrights to individual pieces please see first page of each essay.
All rights reserved
No part of this book may be reproduced or utilized in any form or by any means, elec¬
tronic or mechanical, or by any information storage and retrieval system, without writ¬
ten permission from the publisher. Please contact Rutgers University Press, 100 Joyce
Kilmer Avenue, Piscataway, NJ 08854-8099. The only exception to this prohibition is
"fair use" as defined by U.S. copyright law.
Manufactured in the United States of America

For My Parents and Tami
«
Contents
Graphic Violence in the Cinema: Origins,

Aesthetic Design, and Social Effects 1
Stephen Prince
The Historical Context of Ultraviolence
The Thin Red Line 47

Joseph Morgenstern
Movies to Kill People By 51

Bosley Crowther
Another Smash at Violence 54

Bosley Crowther
Crowther’s ‘Bonnie’-Brook: Rap at Violence Stirs

Brouhaha 57
Ronald Gold
Statement by Jack Valenti, MPAA President,

before the National Commission on the
Causes and Prevention of Violence 62
The Aesthetics of Ultraviolence
Bang Bang Bang Bang, Ad Naseum 79

John Bailey, ASC
Death and Its Details 86

David Thomson
Violence: The Strong and the Weak 99

Devin McKinney
The Violent Dance: A Personal Memoir

of Death in the Movies 110
Vivian C. Sobchack
vii
vm
Contents
Her Body, Himself: Gender in the Slasher Film 125

Carol J. Clover
The Aesthetic of Slow-Motion Violence

in the Films of Sam Peckinpah 175
Stephen Prince
The Effects of Ultraviolence
Some Effects of Thoughts on Anti- and

Prosocial Influences of Media Events:
A Cognitive-Neoassociation Analysis 205
Leonard Berkowitz
Mass Media Effects on Violent Behavior 237

Richard B. Felson
Contributors 267
Index 269
l
Screening Violence
Stephen Prince
Graphic Violence in the

Cinema: Origins, Aesthetic
Design, and Social Effects
Graphic violence is an inescapable and ubiquitous characteristic of con¬

temporary cinema. Severed heads and spurting arteries are plentiful on
today's screens, and filmmakers of all stripes, from auteurs to hacks,
have become proficient at staging sanguinary spectacles. Martin Scor¬
sese has shown characters with their faces shot apart (Taxi Driver
[1976]) and beaten into bloody but still living pulp [Casino [1995]). Paul
Verhoeven's giant bugs in Starship Troopers (1998) rip their victims into
ragged, bloody pieces. Exploding bodies (The Fury [1978], Videodrome
[1983]), decapitation [The Omen [1976], Wild at Heart [1990]), dis¬
memberment [Re-Animator [1985], Henry: Portrait of a Serial Killer
[1990]), and homicide committed with unconventional weapons rang¬
ing from ice pick [Basic Instinct [1992]) to power drill [Body Double
[1984])—such bloody spectacles now crowd the screen. Indeed, their
popularity demonstrates the widespread embrace of violence by con¬
temporary moviegoers and the film industry, despite ongoing social
controversies about the unhealthy effects of viewing ultraviolence.
This introduction examines the origins of ultraviolent movies,
the long-standing controversies over the effects of viewing film vio¬
lence, the evidence furnished by social science about these effects, and
the inherent characteristics of screen violence that subvert its progres¬
sive, legitimate uses (the reasons why, in other words, filmmakers can¬
not control the reactions of viewers to the graphic violence they put on
screen). Screen violence provokes -an inherently volatile set of viewer
responses. These do not include catharsis, and they should make us
Copyright © 1999 by Stephen Prince
1
2
Stephen Prince
pessimistic about the psychological health promoted in viewers by

much contemporary visual culture.
Violence in the movies is not of recent origin. Screen violence is
deeply embedded in the history and functioning of cinema. It is as old as
the medium and has arguably been of central importance for the popular
appeal of film. Edwin Porter's The Great Train Robbery (1903) shows a
beating victim (albeit transmuted into a none-too-convincing dummy)
thrown from a moving train and climaxes with a massacre of the train
robbers. D. W. Griffith's Intolerance [ 1916) shows decapitation and other
gruesome sights and climaxes, as did many of his pictures, with thrilling
scenes of physical action. The appeal of violence in cinema—for film¬
makers and viewers—is tied to the medium's inherently visceral proper¬
ties. These make the cinema especially suited to the depiction of violence.
Brian De Palma, a well-known contemporary exponent of movie violence,
has noted this fundamental connection between the plastic components
of the medium and the emotional pleasures it offers. "Motion pictures are
a kinetic art form; you're dealing with motion and sometimes that can be
violent motion. There are very few art forms that let you deal with things
in motion and that's why Westerns and chases and shoot-outs crop up in
film. They require one of the elements intrinsic to film: motion."1
Although movie violence has a long history, in contrast with
today's films, screen violence in earlier periods was more genteel and
indirect. From 1930, when it was formulated, until the 1960s, Holly¬
wood's Production Code regulated all aspects of screen content, with
an elaborate list of rules outlining what was permissible to show and
what was not. These regulations placed great constraints on filmmak¬
ers and helped to prevent the emergence of ultraviolence in American
film during these earlier periods. (With its decapitation [and its nudity],
Intolerance was a pre-Code film.) In filmic depictions of crime, for ex¬
ample, the Code stipulated that brutal killings must not be shown in
detail, murder must not be glamorized so as to inspire imitation, and
the use of firearms should be sparing. This censure against excessive
depictions of firearms stands in sharp contrast with the fetish for high-
tech firepower in today's action films, which devote long, lingering
close-ups to weapons, as cradled in the arms of Sylvester Stallone,
Arnold Schwarzennegger, and other superheroes.
From 1930 on, filmmakers had to shoot and cut their material
with the Code's provisions in mind because the major studios would
not distribute films that lacked a Code seal of approval. Thus, explicit
footage showing King Kong trampling his victims or Frankenstein's
3
Graphic Violence in the Cinema
Figure 1. In earlier decades, gunshot victims gracefully expired with minimal fuss. In Casablanca,
Major Strasser (Conrad Veidt) takes a bullet in the chest at close range and drops out of the frame
without losing his grip on a telephone.
monster tossing a little girl into a lake (she drowns) was deleted from
the final cuts of those pictures. And in countless Westerns and urban
crime dramas, shooting victims frowned and sank gracefully out of
frame, with their white shirts immaculate.
With its roots in Catholic Church doctrine, the Production
Code aimed to enforce a strict morality and healthy-mindedness in the
movies. This policy was the industry's response to a prolonged period
in the 1910s and 1920s of social agitation over movie content,2 during
which time many citizens' groups and state and local agencies tried to
censor the movies. To protect itself, the industry passed the Production
Code and created in 1934 an enforcement agency, the Production Code
Administration (PCA). For decades, the Code and the PCA held vio¬
lence on screen to a minimum. ,
Inevitably, though, some films and filmmakers pushed the
boundaries of what was deemed acceptable. The gangster film cycle of
the early 1930s, in particular, inflamed national controversy about the
unwholesome effects of the movies. Little Caesar (1930), The Public
4
Stephen Prince
Figure 2. Tony Montana (Al Pacino) expires in an ultrabloody showdown at the end ojScarface.
The ferocity of his death scene could not have been shown in earlier decades of American cinema.
Enemy (1931), and especially Scarface (1932) showed cold-blooded mur¬

der by tommy-gun, and these pictures boasted an extraordinary body
count. The violence of this cycle, and the viciousness personified in the
gangster heroes portrayed by James Cagney, Edward G. Robinson, and
Paul Muni, helped instigate the PCA, as well as the Payne Fund Stud¬
ies, the first systematic effort to use social science methodologies to
study the effects of motion pictures on their audiences. Conducted in
the early 1930s in response to controversies over allegedly indecent
film content, the studies comprised eight volumes of research on the
effects of movies on children and youth, with particular emphasis on
issues of delinquency and sexuality. I will have more to say about the
Payne Fund Studies later in this introduction.
For now, the salient point to grasp is that Hollywood's Produc¬
tion Code restricted the possibilities for graphic depictions of screen vi¬
olence, although, as the inception of the Payne Fund Studies shows, it
did not end controversy over the social effects of motion pictures. For
the most part, screen violence remained relatively discreet, and the
camera turned away from its uglier manifestations. Consider the rep¬
resentational differences at work in similar material in the gangster
genre filmed during Code and post-Code eras. At the end of White Heat
5
Gmphic Violence in the Cinema
Figure 3. In the 1930s, Hollywood gangster films aroused public outcry over their violence and
romantic depiction of criminals. In The Public Enemy (1931), Tom Powers (James Cagney) pre¬
pares to coldly execute a boyhood pal.
(1949), a government sniper repeatedly shoots Cody Jarrett (the film's

gangster hero portrayed by James Cagney), but Jarrett refuses to fall. He
takes the high-powered shots without succumbing in an episode of sus¬
tained and brutal violence. The visual treatment, though, makes the vi¬
olence implicit and surprisingly indirect. The action shows Jarrett at a
distance, in longshots that make it difficult to see that he is, in fact,
being shot repeatedly. Furthermore, in keeping with the period's filmic
norms, none of the bullet strikes on Jarrett are visualized. The pictorial
treatment glosses the scene's exceptional brutality by hiding its details.
A viewer has to look through the style to grasp the content that it has
buried. By contrast, at the conclusion of Brian De Palma's remake of
Scarf ace (1984), the gangster hero (Al Pacino) is gunned down by a rival
gang using automatic weapons. Blasted by a hail of gunfire, he is shot
many times over. The gore is detailed and inescapable. The bullet hits
peppering the character are shown in a graphic spray of blood and torn
clothing and flesh. Instead of hiding it, the style flamboyantly empha¬
sizes the physical carnage. The contrast between these scenes in White
6
Stephen Prince
Heat and Scarface, the difference between concealing and displaying

hyperviolence/ shows the disparity between the Code and post-Code
eras and is the result of complex changes in the film industry, and in
society, that helped produce the turn toward explicit violence.
The Genesis of Ultraviolence
The beginning of the shift, that point when the graphic representation
of physical violence first became a distinct stylistic possibility in the
American cinema, is easy to date. Ultraviolence emerged in the late
1960s, and movies have never been the same since. The factors that
helped produce this new violence were instigated by two watershed
events in Hollywood history: the revision in September 1966 of Holly¬
wood's thirty-six-year-old Production Code and the creation two years
later of the Code and Rating Administration with its G-M-R-X classifi¬
cation scheme. These changes were responses to the more liberal and
tolerant culture of the period, particularly the revolution in social
mores tied to the youth movement. Shackled by the Production Code,
movies were thirty years behind the times. Accordingly and led by the
Motion Picture Association of America (MPAA), the Hollywood indus¬
try mounted an aggressive campaign to make films relevant again for a
society whose attitudes and practices no longer coincided with the
morality institutionalized in the Production Code.
Thus, the 1966 revision scrapped the Code's injunctions about
specific content areas and substituted in their place a few broadly
phrased guiding principles that, as MPAA head Jack Valenti pointed
out, significantly expanded the creative license of filmmakers. Where
the old Code told filmmakers precisely how they were to approach and
show scenes of violence, the revised Code merely recommended that
filmmakers exercise discretion in showing the taking of human life.
"Discretion" is an elastic and relative concept, and as the intensifying
violence in American cinema in the late 1960s shows, it was mostly an
ineffective principle. But the scrapping of specific content injunctions
was not the most significant revision. In a development that was far
more important and influential for subsequent filmmaking, the MPAA
created a new Suggested for Mature Audiences (SMA) designation for
films that had harder, more adult content. This new designation en-
7
abled filmmakers for the first time to target an adult audience and on
that basis take sex and violence much further than in the past when the
audience mix included young viewers.
An additional revision a few years later accelerated these devel¬
opments. In 1968, the MPAA unveiled a four-way classification system—
G-M-R-X—that differentiated films by audience segment. G-rated films
were suitable for the entire family, y/hile underage viewers were prohib¬
ited from seeing X-rated films and could only view R-films if accompa¬
nied by a parent or adult guardian. The G-M-R-X scheme made even
greater freedoms available to filmmakers because films could now be
niche-marketed to adult audiences, bypassing the content restrictions
that the presence of young viewers had hitherto necessitated. These mod¬
ifications of the Production Code helped produce a wave of tougher,
harder-edged, and controversial films whose graphic violence, profanity,
and sexuality exemplified the new artistic freedoms that the MPAA had
been seeking and promoting. Films like Bonnie and Clyde (1967), The
Fox (1968), The Detective (1968), In Cold Blood (1967), Point Blank
(1967), and Barbarella (1968) would have been unthinkable a mere five
years previously, and they collectively demonstrate the emergence of the
new, adult-themed cinema that the MPAA helped inaugurate.
Indeed, the MPAA's advocacy of new filmmaking freedoms was
a remarkable feature of this revolution in American cinema. The
MPAA pushed aggressively for these freedoms despite vocal protest
from groups opposed to these changes. Testifying before the National
Commission on the Causes and Prevention of Violence on December
19, 1968, Jack Valenti proudly announced, "There is a new breed of
filmmaker. And mark you well this new filmmaker, because he's an ex¬
traordinary fellow. .. . He's reaching out for new dimensions of expres¬
sion. And he is not bound—not bound—by the conventions of a
conformist past. I happen to think that's good."3 In another context,
Valenti stressed the importance of bringing "old movie standards out of
the archaic and arcane and into current trends."4 Criticizing viewers
who protested the rising levels of violence, sexuality, and profanity in
the movies, an MPAA community relations associate cited the "new
revolutions" in taste and morality in society. She recommended that
viewers reorient themselves to these social changes and the new cin¬
ema they were helping to produce. "They [the public] fail to realize that
films have changed to reflect our changing culture."5
The MPAA's responsiveness to late 1960s culture, in this re-
8
Stephen Prince
spect, was strikingly different from the posture and policy it assumed
in the early 1980s, when a more conservative national climate produced
vehement criticism of movie violence by government officials and
social watchdog groups. Brian De Palma's lurid and bloody thriller
Dressed to Kill (1980) aroused tremendous criticism because of its vio¬
lence against women. Anticipating problems, the MPAA threatened
during post-production to slap an X-rating on the picture unless De
Palma trimmed its.violence. De Palma protested the cuts but never¬
theless made them so that the picture could be released with an R-
rating. Explaining the tough line taken by the MPAA with regard to
Dressed to Kill, Jack Valenti pointed to the political realignments of the
Reagan era and implicitly announced that the agency's late 1960s lib¬
eralism was over. He said, "The political climate in this country is
shifting to the right, and that means more conservative attitudes to¬
ward sex and violence. But a lot of creative people are still living in the
world of revolution."6
During the 1960s, however, in its period of high liberalism, the
MPAA aimed to bring films into closer accord with the youth audience
and its general questioning of Establishment values. This was a vital
demographic and one that the agency was determined to court; a 1968
MPAA audience survey showed that 16- to 24-year-olds were responsi¬
ble for 48 percent of national ticket sales.7 To capture this young audi¬
ence, the MPAA believed that films would have to become more
attuned to contemporary mores, which adherence to a thirty-six-year-
old Production Code prevented. In addition to the youth audience, how¬
ever, a multitude of other factors influenced and helped shape the new
direction of American cinema. The period's general social turmoil, its
climate of political violence, and, most especially, the war in Vietnam
convinced many filmmakers and the MPAA that movie violence paled
next to the real-life bloodshed in the nation's cities and the jungles of
Southeast Asia. The savage bloodshed of the Vietnam War established
a context whereby filmmakers felt justified in reaching for new levels
of screen violence. Moreover, the war and the political assassinations
of the 1960s fed a general cultural fascination with violence to which
the movies responded.8
Jack .Valenti remarked on the interconnections between the
era's social and media violence and did so in a manner that claimed
film's prerogative in responding to real-world violence. In 1968, he re¬
marked, "For the first time in the history of this country, people are ex¬
posed to instant coverage [on television] of a war in progress. When so
9
many movie .critics complain about violence on film, I don't think they
realize the impact of 30 minutes on the Huntley-Brinkley newscast—
and that's real violence."9 Bonnie and Clyde was the most explicitly
violent film that had yet been made, and its director, Arthur Penn,
claimed that he didn't consider it especially violent when taken in con¬
text of the Vietnam era. "Not given the times in which we were living,
because every night on the news wejsaw kids in Vietnam being airlifted
out in body bags, with blood all over the place. Why, suddenly, the cin¬
ema had to be immaculate, I'll never know."10
Resulting from the industry's efforts to connect with a young,
contemporary audience and the period in which that audience lived,
motion picture violence began its remarkable escalation in 1967.
United Artists released Sergio Leone's Dollars trilogy of Westerns—A
Fistful of Dollars, For a Few Dollars More, and The Good, the Bad, and
the Ugly—on staggered dates throughout that year to accentuate the se¬
ries nature of the films. Their release triggered a storm of protest over
Leone's more cold-blooded and brutal depiction of the West. A reviewer
for Variety, the film industry's trade journal, called Fistful a "blood¬
bath," with "sadism from start to finish, unmitigated brutality, a pil¬
ing up of bodies."11 The paper's subsequent review of The Good, the
Bad, and the Ugly objected, "One sequence in particular, a five-minute
torture session that climaxes in an attempted eye-gouging, may well
serve as the battle cry for opponents of screen violence."12
The Leone films, which had been made in Italy, aroused con¬
siderable protest from groups opposed to the new screen violence, but
it was two American films that year—The Dirty Dozen and Bonnie and
Clyde—that ignited the loudest cries. The Dirty Dozen was an uncom¬
monly cynical World War II action picture about hardened criminals re¬
cruited for a suicide mission into Nazi Germany. In the climax of the
picture, the Americans incinerate their Nazi enemies by locking them
in closed quarters, dousing them with petrol, and setting them afire.
The sadism of the sequence appalled many commentators. Among the
most vocal of these was Bosley Crowther, the prominent New York
Times critic. He condemned the film, and the new movie violence of
which it was part, for its "glorification of killing" and for blatantly ap¬
pealing to an audience's aggressive and sadistic appetites.13 Bonnie and
Clyde went even farther by explicitly portraying the details and physi-
cality of violent death, concluding with an extended slaughter sequence
that shows the outlaws raked with automatic weapons' fire, their bod¬
ies riddled by bullet strikes. Both Time and Newsweek condemned the
10
Stephen Prince
film as bloodthirsty trash in initial reviews, only to issue retractions in

an unprecedented second round of reviews that praised the picture as
trendsetting.
Indeed, Penn was the first American filmmaker to utilize the
cinematic techniques that quickly became the normative means of
filming violent gun battles. Taking his cue from Japanese director Akira
Kurosawa, who had used these techniques in Seven Samurai (1954) and
other films, Penn employed multicamera filming (i.e., filming with
more than one camera running simultaneously), slow motion, and
montage editing (i.e., building a sequence out of many, very short, brief
shots). To these techniques which rendered gun violence with greater
intensity than ever before, Penn added squibs. Probably more than any
other effects tool, squibs changed the way screen violence looked.
Squibs were condoms filled with fake blood, concealed within
an actor's clothing, and wired to detonate so as to simulate bullet
strikes and blood sprays. Squibs enabled filmmakers to graphically vi¬
sualize the impact of bullets on the human body, a detailing that is ab-
Figure 4. Squibs registering bullet strikes are an essential component of contemporary movie vi¬
olence. His face obscured with blood spray, Mr. Blonde (Michael Madsen) is gunned down by Mr.
Orange (Tim Roth) in Reservoir Dogs. The character’s white shirt provides maximum contrast
for the squibs.
11
Figure 5. In The Wild Bunch, Peckinpah innovated in the use of squibs by employing them to
represent exit wounds. When a bounty hunter shoots an army soldier, viewers see the exit wound
but not the bullet strike on the victim’s chest.
sent in film prior to 1967 and which helped give violence in these ear¬
lier periods an unreal and sanitized appearance. To be sure, certain pic¬
tures in these earlier periods anticipated the modern staging of gun
violence. In Shane (1953), for example, a gunfire victim is hurled back¬
ward by the force of a bullet's impact, and in Penn's own The Left-
Handed Gun (1958), a hapless deputy is literally blown out of his boot
by a gun blast. But despite such striking details, these scenes lacked
squib-work and the palpable physicality that it lent gun violence. Dur¬
ing and after 1967, by contrast, the savage impact of gunfire on human
flesh became an enduring feature of screen killing. Films released in the
same year as Bonnie and Clyde and which lacked squib-work—The
Professionals, El Dorado—seemed bloodless, irrevocably part of a now-
archaic era in screen violence and not at all contemporary with Penn's
film. In the film's climactic sequence, Bonnie (Faye Dunaway) and
Clyde (Warren Beatty) are ambushed by the Texas Rangers, who are
armed with machine guns. The actors were rigged with multiple squibs.
When detonated in sequence and augmented by the writhing of the ac¬
tors, these provided a horrifying visualization of the outlaws' bodies
being punctured by scores of bullets.
Penn filmed this action with four cameras running at different
speeds. Most operated at very high speeds to provide slow-motion
footage for the construction of a complexly edited sequence. Penn and
his editor Dede Allen assembled a brief but sophisticated montage of
the ambush. The editing juxtaposes differential rates of slow motion to
extend the outlaws' death agonies and to capture, as Penn put it, the
balletic and the spastic qualities of their violent deaths. Penn said,
12
Stephen Prince
"What I did do, which I think had not been done, was to vary the speeds
of the slow motion so that I could get both the spastic and the balletic
qualities at the same time."14 These are qualities in tension with one
another, yet Penn and Allen's masterful editing conjoin movements
that are equally dancelike and convulsive. The scene's alternately beau¬
tiful and horrifying qualities are the result of its choreography and
specifically of Penn's determination to yoke dance and convulsion to¬
gether in a novel combination. The results were truly seminal. The
montage of slow-motion, squib-fired death agony powerfully influ¬
enced subsequent filmmakers, especially Sam Peckinpah, who took the
design to new extremes the following year in The Wild Bunch.
The Dirty Dozen and Bonnie and Clyde sent the industry a
clear signal that would not be ignored. These controversial pictures
were extraordinarily popular. The Dirty Dozen was the biggest money¬
maker of 1967, and so extraordinary was repeat business for Bonnie and
Clyde that Variety, tracking its box-office performance, placed it in an
"impossible to project" category.15 The picture's popularity placed it on
Variety's weekly list of the top dozen box-office earners twenty-two
times, a record topped at the time only by Mary Poppins (thirty-two
times on the list in late 1964-early 1965). Time magazine noted the in¬
dustry's response to this success and its implications for the future:
"There is an almost euphoric sense in Hollywood that more such
movies can and will be made."16
The outstanding commercial success of Leone's Dollars trilogy,
The Dirty Dozen, and Bonnie and Clyde showed that the public had
hitherto unappeased appetites for screen carnage and that the industry
could make a lot of money from filming hyperviolence. Thus, in short
order, the threshold that Bonnie and Clyde had crossed in 1967, with
its images of slow-motion bloodletting, was surpassed. Director Sam
Peckinpah's The Wild Bunch (1969) offered two extended slaughter se¬
quences, opening and closing the film, that were far more gruesome,
graphic, and protracted than the gun battle in Penn's picture had been.
Nothing in the American cinema had been as remotely violent as what
Peckinpah now put on screen in The Wild Bunch.
Its gun battles had ferocity and an intensity that Bonnie and
Clyde only approximated. Peckinpah and his producer Phil Feldman
knew exactly what they were doing in crafting blood-soaked images
that were the most audacious yet conceived. While Peckinpah was in
pre-production on The Wild Bunch, Feldman wrote to congratulate him
for offending the MPAA with the savagery of the picture's proposed
13
bloodshed and urged him to view the Leone films before deciding how
far to go. “I think it might be good for you to have a comparative basis
before you finally decide just how far other people have gone in the field
of blood and gore and what the public is comparing us to."17 Further¬
more, on March 19, 1968, Warner Bros, sent a letter to Peckinpah's pro¬
duction manager confirming that a print of Bonnie and Clyde would be
shipped to Parras, Mexico, where^Peckinpah was on location for The
Wild Bunch.18 The film would arrive for screening the weekend of
March 23-24. This was just prior to the start of principal photography
on March 25. Peckinpah evidently wished to take another look at
Penn's achievement, to reinforce his understanding of that achieve¬
ment and his desire to surpass it.
Many viewers were appalled by the film's gore. An early test
screening of the film elicited such negative responses as "The movie is
nothing but mass murder" and "Nauseating, unending, offensive
bloody violence."19 But the film also had its passionate defenders, and
these included ordinary viewers as well as prominent film critics. Penn
and Peckinpah were enormously talented filmmakers, and they had
crafted the two most vivid, audacious, and ambitious films that Holly¬
wood had seen in many years. Thus, these pictures gained a significant
measure of respect and critical stature that helped legitimize the in-
your-face bloodletting that otherwise made them so notorious. Penn
and Peckinpah were both radical social critics, disturbed by the cor¬
ruption of America in its Vietnam years, and they proved that film¬
makers in the late 1960s could use graphic violence for serious
purposes. Peckinpah, for example, repeatedly remarked that he wished
to deglamorize movie violence in order to show how ugly and awful real
violence was.20
Unfortunately, and perhaps inevitably, the stylistics of graphic
violence proved to hold tremendous fascination for subsequent gener¬
ations of filmmakers who did not share Penn and Peckinpah's radical
social objectives. The explicitness of this violence quickly escalated.
Made only a few years after The Wild Bunch, for example, Taxi Driver
(1976) was far bloodier and much more graphic, with images of dis¬
memberment and a gunshot victim's brains splattered on a wall. Penn
and Peckinpah helped establish the stylistic features of ultraviolence,
while subsequent filmmakers have replicated and exaggerated them.
Squib-work, multicamera filming, and montage editing utilizing dif¬
ferential rates of slow motion—this combination of elements became
one of the two dominant aesthetic forms of ultraviolence. It is today
14
Stephen Prince
Figures 6 & 7. One of the most horrific moments of violence in recent cinema occurred in Alien,
when a baby monster bursts from the chest of its human host. Graphic imagery of physical mu¬
tilation would become a staple of horror films.
very difficult to find gun battles in movies that have not been stylized
in this fashion. (The elaborate gun battles in L.A. Confidential [1997]
are notably deviant because they lack slow motion.) Moreover, the form
has been internationalized. Hong Kong director John Woo (The Killer
[1989], Hard Boiled [1992]) is the best-known contemporary exponent
of the furiously bloody gun battles in the style elaborated by Penn and
Peckinpah. Despite the fact that Woo's work is situated in another cul¬
ture and country (at least until his current Hollywood period), the for¬
mal design bf violence in his films follows the now-familiar and
conventional parameters of the Penn-Peckinpah stylistic.
Contemporary .ultraviolence exhibits a second predominant
aesthetic form. In addition to the Penn-Peckinpah stylistic, ultravio¬
lence includes graphic imagery of bodily mutilation. This type of im-
15
agery was not part of the Penn-Peckinpah stylistic, beyond the use of
squib-work, because that style stressed the kinetic effects of montage,
making violence balletic, a dance of death. But graphic mutilation—eye
gouging, impalement, and dismemberment—surfaced in the horror
film in the late 1970s and the 1980s, as that genre abandoned the at¬
mospherics of earlier decades and offered instead stomach-churning
and gut-wrenching experiences. As Carol Clover notes,
The perfection of special effects has made it possible to show maim¬

ing and dismemberment in extraordinarily credible detail. The horror
genres are the natural repositories of such effects; what can be done is
done, and slashers, at the bottom of the category, do it most and worst.
Thus we see heads squashed and eyes popped out, faces flayed, limbs
dismembered, eyes penetrated by needles in close-up, and so on.21
The genre's remarkable resurgence in the 1980s was tied to this

investment in ultraviolence. The decade opened with a huge spike in
horror film production, which rose from 35 pictures produced in 1979
to 70 in 1980 and 93 in 1981.22 Films in this spike included Dressed to
Kill, Friday the 13th, Prom Night, and Maniac, all notoriously violent
pictures. The genre spiked again in 1986-87, hitting a peak of 105 pic¬
tures in 1987. By mid-decade, the video market was thriving, and many
of these pictures, especially the most violent and disreputable, were
low-budget, throw-away entries aimed at this ancillary market.
Developments in two areas of makeup special effects, practiced
by a new generation of makeup artists, stimulated the genre's turn to
ultraviolence. The artists included Tom Savini (Dawn of the Dead,
Maniac, Friday the 13th, Eyes of a Stranger), Rick Baker [It’s Alive,
Squirm, The Funhouse), and Rob Bottin (The Howling, Piranha, The
Thing). These artists employed prosthetic limbs and latex (as a con¬
vincing stand-in for human skin) to simulate exquisitely detailed body
mutilations. Limbs could be hacked off with a convincing show of bone
and gristle, and skin (latex) could be ripped, scored, punctured, and
peeled away, much to the roaring delight of audiences who patronized
these films and enjoyed the new sadism.
The most controversial of the new horror pictures were the
slasher films about serial killers slaughtering promiscuous teenagers.
These slaughterfests seemed to demonstrate rampant sadism in popular
culture, and they disturbed many observers inside and outside the film in¬
dustry. FJorror film bloodfests were present in high numbers at the inter¬
national film markets in 1980, the first year of the genre's big production
16
Stephen Prince
boom. Distributors and studio sales reps expressed revulsion at these

products. “All they want is blood pouring off the screen. I question the
mental balance of the people making and buying this stuff/' noted a
Carolco executive.23 The imagery of victims dismembered by spikes,
axes, chain saws, or power drills, or run through meat grinders evoked
a swift and stern backlash from critics, especially feminist scholars,
who pointed out that a basic slasher film premise was a male killer
stalking and slaughtering female victims.
Moreover, the extended subjective camerawork in these films,
showing the killer's point of view as he stalks his victims, seemed to
invite viewers to accompany the killers and to participate in their stalk¬
ing rituals. Prominent film critics Gene Siskel and Roger Ebert used
their newspaper columns and television show to attack the slasher
films. Calling them ''gruesome and despicable," they argued that the
pictures expressed a hatred for women. Ebert noted his personal dis¬
comfort watching the pictures with an audience. "Audiences are cheer¬
ing the killers on. It's a scary experience."24 The Chicago Tribune
noted, "The public would not tolerate this kind of eye-ball gouging
sadism if it were directed against animals instead of women. But in
these new films, the fact that the horror is inflicted on women appears
to be the very point."25
Not all viewers or scholars felt that the slasher films were ob¬
jectionable or were without nuance. In her book Men, Women, and
Chain Saws, for example, Carol Clover offered an interpretation of
these films that emphasized the presence of a recurring character
type—a heroic woman who eludes the killer, survives his rampage, and
eventually defeats him. Clover thus argued (see the essay included in
this volume) that a heroic female archetype repeatedly emerged from
these narratives. She also presents a persuasive case that viewers iden¬
tify with the characters in these films, killer and victims, in a fluid and
dynamic manner and particularly that male viewers identify with the
pain and anguish of the female victims. "Just as attacker and attacked
are expressions of the same self in nightmares, so they are expressions
of the same viewer in horror film."26 In contrast to most observers,
Clover emphasizes the complexity of slasher film discourse, the way it
plays on ambivalent gender messages. But, as she acknowledges, this
discourse contains sadistic elements. Whatever their manner of subli¬
mating and displacing viewer anxiety, the films appealed in part, at
least, to a prurient sadism, vividly visualized, and this is what moti¬
vated much of the critical outcry. The controversies, though/did not
17
Figure 8. In the prelude to a highly detailed killing scene, a knife-wielding and masked assailant
grabs his victim (Drew Barrymore) in the opening sequence of Scream. Because she does not die
easily or quickly, the sadism goes on at length.
stop the productions, although by the late 1990s the slasher cycle was
nearly played out, as evidenced in part by the degree of self-conscious¬
ness it attained in Wes Craven's Scream films. But ultraviolent horror
films remain very popular, as the huge collections of gory movies in
video stores demonstrate and as the Web pages devoted in cyberspace
to Jason, Freddy Kreuger, and the other killer heroes indicate.
Ultraviolent films since 1967 have utilized these two aesthetic
formats—montage-slow motion and graphic mutilation—to provide
very powerful and intense experiences for spectators. The medium's
power to agitate, horrify, and excite movie audiences has generated per¬
sistent concern about the psychological and social effects of violent
films. These concerns have appeared throughout film (and television)
history. The 1930 Production Code, for example, was premised upon
characteristics of the medium and the audience, which in combination
were deemed socially undesirable. These characteristics were formally
written into the Code as its reasons for being. They included cinema's
intense appeal to mature as well as immature viewers, the spread of cin¬
ema to all sectors and regions of society and consequent erosion of local
community standards, the suggestibility of a mass audience, and the
vividness and emotional appeal of the medium. "In general, the mobil¬
ity, popularity, accessibility, emotional appeal, vividness, [and]
straightforward presentation of fact in the film make for more intimate
contact with a larger audience and for greater emotional appeal. Hence
the larger moral responsibilities of the motion pictures."27
An empirical examination of the effects of movies on young
18
Stephen Prince
viewers, the Payne Fund Studies (1933), closely followed passage of the
Production Code. The studies were not concerned with violence per se
but with juvenile delinquency more broadly conceived in terms of
young lives given over to petty crime and sexual promiscuity, possibly
resulting from Hollywood's influence. The studies are an early mani¬
festation of a persistent fear that visual media may have harmful effects
on viewers due to the media's vivid presentation and modeling of un¬
desirable behaviors. In Movies and Conduct, a volume in the series, the
sociologist Herbert Blumer developed the concept of "emotional pos¬
session" to explain the impact of film on viewers and the potential of
the medium to exert strong effects.28 For Blumer, emotional possession
had three components. Viewers identify strongly with characters and
situations on screen (e.g., romantic lovers in a kissing scene), and this
leads to the arousal of normally repressed feelings. Heightened arousal
produces a loss of ordinary self-control as the unleashed feelings gain
or threaten to gain expression. Blumer speculated that, for some view¬
ers over time, the emotions and behavioral repertoires associated with
emotional possession may become entrenched and lead to a long-term
alteration of character.
Blumer's terminology—emotional possession—tends to sound
a bit lurid, and it certainly overstates the affective components of cin¬
ema viewing at the expense of their cognitive correlates and the
viewer's cognitive framing of the experience. Most viewers, for exam¬
ple, doubtless realize they are watching a movie and understand that
there are boundaries between movie experience and real-life experi¬
ence. But Blumer was right to consider the special vividness and emo¬
tive power of the medium as key variables affecting viewer responses.
By varying camera positions, editing rhythms, and using music and spe¬
cial effects, filmmakers stylize violence and heighten its emotional and
visceral impact. The history of cinema is, in part, a history of increas¬
ingly vivid presentational techniques. Public fears that the medium's
intense appeals might bypass traditional avenues of socialization and
reconfigure social mores were acute in the 1930s, as the Payne Fund
Studies demonstrate, but these anxieties did not receive definitive an¬
swers in the period's research. The evidence furnished by the Payne
Fund Studies was ambiguous, and it pointed toward limited media ef¬
fects. Moreover, graphic violence was not a feature of American movies
in that period. After 1967, though, when the censorship barriers came
down and explicit sex and violence filled the screen, the old fears about
the medium's impact were again relevant. The rise of ultras iolence
19
compelled a reconsideration of the medium's vividness and emotive

power, those factors that Blumer had identified as constituents of the
medium's ability to influence viewers. Generations of film and televi¬
sion viewers were being socialized to accept graphic violence as a screen
norm, and this development created a vital need for continuing research
on the effects of screen violence. Scholars in cinema studies, though,
neglected questions of social effects because the field as a whole was
hostile to empirical methodologies,- scholars from the fields of psy¬
chology and communication took up the questions.
The Empirical Evidence on the Effects of Viewing Violence
Empirical investigations of media violence have confronted and dis¬

credited a deeply entrenched popular notion about the nature of proba¬
ble effects. This is the idea of catharsis, the notion that screen violence
provides a viewer with the opportunity to purge hostile feelings in the
safe realm of art. The idea of catharsis has acquired tremendous force
in contemporary culture and has become a foundational concept for ex¬
plicating the relationship between visual media content and viewers.
Critics routinely refer to the cathartic effects of violent films—and de¬
fend them on this basis—as in Mark Crispin Miller's description of the
concluding gun battle in The Wild Bunch as a "cathartic holocaust."29
Defending the new violence of late 1960s pictures, Jack Valenti cited
their probable cathartic effects. He told the National Commission on
the Causes and Prevention of Violence, "I am personally convinced that
there is genuine validity in the belief that disturbing emotions may be
purged through the vicarious experience of aggressive acts on screen."30
The concept of catharsis derives from Aristotle's well-known
discussion in the Poetics wherein he asserted that tragedy uses lan¬
guage and acting to evoke pity (eleos) and fear [photos] in a manner that
purges such feelings from the spectator.. Aristotle's discussion is brief,
and, rather than constituting a theory, it provides only the hint of an
idea. He does not work out the body or the implications of his idea in
any detail. As the critic and philosopher of classical tragedy, Walter
Kaufmann, points out, "The celebrated doctrines of the Poetics are
for the most part peremptory dicta of a few lines, and not theories that
Aristotle tries to establish with care."31 Significantly, Aristotle does
not mention aggression, and, furthermore, he identified language and
20
Stephen Prince
acting as the vehicles effecting the cathartic purge. He described

a medium—classical tragedy—which conveyed its effects through
language and in which horrific violence occurred offstage.
Thus, his remarks about catharsis do not necessarily generalize
to a medium like cinema, whose design is so different than classical
tragedy. Cinema images are far more sensual and powerful. As I noted
previously, filmmakers use editing, camerawork, and sound to heighten
the sensory qualities of violent episodes, and these are shown in film
rather than described verbally, as in classical tragedy. Thus, as Singer
and Singer point out, "When you see a violent image on the screen in
all its graphic gore, that image intrudes with greater shock as a literal
picture, relatively uncontrolled by one's own imagination [or] val¬
ues."32 As a result of these differences, film and television images of
violence have a greater ability to excite and arouse their viewers, and
these medium-specific differences point away from the cathartic ideal.
The empirical evidence on media effects has discredited the
idea of catharsis, and it points toward a link between viewing film or
television violence and subsequent aggressive behavior. As Leonard
Berkowitz, a prominent researcher in this area, writes, "the majority of
researchers are agreed that the depiction of violence in the media in¬
creases the chances that people in the audience will act aggressively
themselves."33 A meta-analysis of 217 studies on television violence
conducted between 1957 and 1990 found a consistent and significant
correlation between viewing violent television and aggressive or anti¬
social behavior.34 In Felson's recent review of the literature (included
in this volume), he concedes that the empirical inquiries have demon¬
strated media effects on aggressive behaviors, although he adds the
qualifier that, in his assessment, these effects are likely to be weak, to
affect only a small percentage of viewers, and are unlikely to be a sig¬
nificant factor in the high crime rate in this country. Berkowitz, how¬
ever, points out that a small percentage of affected viewers, in a country
with a large population, may produce a not insignificant number of
violent acts.
Studies on the viewing of media violence show aggression-
inducing effects, rather than cathartic ones, and they have pointed in
this directiqn for some time.35 In 1972, for example, after examining the
relation between television viewing and social behavior, the Surgeon
General's Committee concluded that there was little evidence support¬
ing the catharsis hypothesis (that viewing film or television violence
would lessen an individual's propensity to behave aggressively).36 From
21
an empirical standpoint, the catharsis hypothesis is primarily associ¬

ated with the work of Seymour Feshback, who has reported some evi¬
dence offering conditional support for the cathartic effects of viewing
fantasy aggression.37 Feshback's findings, though, do not represent the
consensus of most researchers in this field, and even Feshback reported
evidence pointing away from the catharsis hypothesis.38
Based on the empirical evidence, it is now possible to specify
the program characteristics that are most implicated in the findings of
aggression inducement. These involve violence that is relatively free of
pain and suffering victims who deserve what they get, stories that pos¬
tulate scenarios of righteous, justifiable aggression, and a match be¬
tween the cue properties of situations and characters on screen and the
viewer's real-world situation. When the salient characteristics of film
victims, for example, match the characteristics of available targets in
everyday life, aggressive responses toward those real-life targets may
become more likely. Many of the most notorious and widely reported
cases of apparent media-inspired violence fall into this category. The
incineration of a Manhattan subway-token clerk following the release
of the film Money Train (1995), which depicted a similar incident,
seems to exemplify this principle, but it has also been demonstrated
experimentally.39
A huge proportion of our film and television programming pres¬
ents stories in which the hero or protagonist must behave aggressively
in order to defeat the villain, avenge some crime, or realize goals. In this
context, the depiction of the antecedents and consequences of aggres¬
sion has been shown to influence viewers' tendencies to aggress. When
aggression is rewarded in a film or television show, it tends to elicit
more imitative aggression from viewers. Albert Bandura, who has ar¬
gued persuasively in favor of a social learning theory of aggression,40
points out that the persuasive example of successful villainy in a show
or film "may outweigh the viewers' value systems."41 Summarizing his
experimental work in which he found young viewers imitating the ag¬
gressive behavior of a film character whose behavior was rewarded,
Bandura noted that children watching television "have opportunities
to observe many episodes in which anti-socially aggressive behavior
has paid off abundantly."42 In light of this, the villain's routine pun¬
ishment at the end of an episode "may have a relatively weak inhibitory
effect on the viewer."43
Bandura has also demonstrated the modeling functions of aggres¬
sive media content—that is, the tendency for observed filmic aggression
22
Stephen Prince
to shape the form of a viewer's subsequent antisocial behavior. Children's

play activities and choice of toys tended to assume forms imitative of the
aggressive activities they had seen on film. Seeing film aggression not
only seemed to heighten subsequent aggressive reactions; it offered a
model for imitative learning.44 When aggression was rewarded in a
show, the salience of this modeling increased. Similarly, when a show
or film presents aggression as a justifiable response by the hero or pro¬
tagonist, it may have a disinhibiting effect on viewers.45 When Rocky
patriotically pummels the robotic Soviet boxer Drago in Rocky IV
(1985), or the rebel forces battle the Empire in the Star Wars sagas, the
film narratives present scenarios of justifiable aggression that provide
a moral orientation for the explicit violence that ensues. Summarizing
the association between filmic aggression, presented as justified, and
antisocial responses by viewers, which they demonstrated experimen¬
tally, Berkowitz and Rawlings conclude that "Seeing the fantasy villain
'get what he deserved' may make the angered individual more inclined
to hurt the villain in his life, the person who had angered him."46 Zill-
mann has framed this process of enjoying the villain's fate as one of
'counterempathy.' "As we have morally condemned a villain for raping
and maiming, for instance, we are free to hate such a person, can joy¬
ously anticipate his execution, and openly applaud it when we finally
witness it. In fiction, we can enjoy his being bullet riddled; and outside
fiction, we can rejoice at the drop of the guillotine."47
Zillmann has suggested that viewers form affinities for screen
characters based on how such people would be viewed were their
behavior occurring in the real world (i.e., the one outside the fiction).
"Fictional characters toward whom an audience holds particular dis¬
positions will draw emotional involvement and responses similar to
nonfictional persons toward whom the same dispositions are held."48
Depictions of suffering may complicate these affinities, that is, the
viewer's disposition to feel for or against the perpetrators of screen vi¬
olence. The presence or absence of a victim's pain, and, when present,
its degree of severity, tends to affect a viewer's disposition to aggress.
Visible or audible signs of a victim's suffering, for example, tend to
depress, or inhibit, aggressive responses49 and may also diminish a
viewer's enjoyment of the film violence, depending on the amplitude of
the expressions of pain.50 One of the striking facts about much screen
violence is its relatively pain-free quality. In the action epics of Arnold
Schwarzenegger or Sylvester Stallone, the heroes massacre hordes of
bad guys with little evident feeling and with no evident agonies on the
23
part of those righteously dispatched. Of importance in this regard, view¬

ers may perceive scenes that lack expressions of suffering by victims as
being less violent than those in which the victim's pain is apparent.
Blanchard, Graczylc, and Blanchard note that viewers tended to rate the
violence level as much higher in The Deer Hunter (1978) than in either
The Bugs Bunny Movie (1979) or the James Bond film Tor Your Eyes
Only (1981), because the latter two films show relatively pain-free vio¬
lence, unlike The Deer Hunter, which presents scenes of physical and
emotional suffering.
This suggests that suffering of victims is a major constituent of per¬

ceptions of violence. This interpretation is supported by the reliably
lower ratings on violence of the Bond film, which contained relatively
little emphasis on suffering of victims although several dozen charac¬
ters appeared to be injured or killed as the result of a nonstop action
episode. It is also congruent with the very low suffering ratings for the
cartoon, which was rated much lower in violence than any of the live-
actor sequences despite portrayals of exaggerated physical violence,
including physical consequences such as heads being shot off."51
In assessing these consequences of screen violence for the view¬

ing audience, I wish to emphasize two points. First, the viewing audi¬
ence is not monolithic but is stratified by personality variables. These
latter may interact strongly with program content and genres to produce
undesirable effects. Zillmann and Weaver, for example, have shown that
male viewers, whose personality characteristics show hostile disposi¬
tions and lack of empathy, preferred violent means of conflict resolution
after watching the "superviolence" in contemporary action films and
that these effects are relatively nontransient.52 My second point is that
the reactions of viewers are not mechanistic but, instead, have strong
cognitive components, and these cognitive components may operate to
reinforce undesirable effects. Berkowitz provides (in the essay included in
this volume) a detailed explanation of the role that cognitive factors
(thoughts and ideas) play in viewer reactions to media violence. Specifi¬
cally, he examines the role of thought (induced by media programming)
in activating bodily reactions. He terms this a priming process that in¬
volves the "spreading activation" of thoughts related to media images and
content and which, under the right conditions, can dampen a viewer's in¬
hibitions against acting aggressively. "The aggressive ideas suggested by
a violent movie can prime other semantically related thoughts, height¬
ening the chances that viewers will have other aggressive ideas in this
24
Stephen Prince
period."53 Berkowitz maintains that a high level of aggression-related

thoughts can activate aggressive inclinations. This perspective is con¬
sistent with the phenomenon of desensitization, that is, the tendency
for viewers accustomed to media violence to show a lowered physio¬
logical response to the material.54 Berkowitz points out that these
lower levels of physiological arousal should be understood as signs of
lowered anxiety levels and that they can coexist with a higher disposi¬
tion to aggress. "It-is important to note that the lower arousal is pre¬
sumably an indicant of a reduced concern about aggression and not a
sign of a decreased inclination to aggression."55 Berkowitz's view that
people derive ideas from aggressive material in the media and that these
ideas can, for a time afterward, foster antisocial behavior has a great
deal of empirical support, which he cites in his discussion. His view
suggests that the aggression-inducing characteristics of media violence
have cognitive consequences for viewers, in addition to emotional and
physiological ones. (Of these latter, several studies have shown that the
arousal of excitement, in itself and independently of specific program
content, is associated with inclinations to aggress.)56
The interplay of these physiological and ideational responses is
complex and can be, at the nonstatistical level of an individual viewer,
difficult to assess. This complexity may account for some apparent con¬
tradictions in the empirical literature and in efforts to predict the likely
effects of violence in a given film or sequence. In the Blanchard et al.
study, for example, respondents rated the Russian roulette scene from
The Deer Hunter as being the least enjoyable and most violent scene of
all the material under study, yet that scene was associated with thirty-
one violent incidents involving use of a handgun in rituals of Russian
roulette between 1978 and 1982.57 Some viewers reacted aversely to the
sequence, while others were apparently stimulated to reenact it. In an¬
other context, anecdotal evidence about the effects of Bonnie and
Clyde ran counter to speculations that the carnage in that film was so
graphic as to inhibit aggressive responses. In 1969, for example, Leonard
Berkowitz spoke to the National Commission on the Causes and Pre¬
vention of Violence and speculated that the graphic violence in Bonnie
and Clyde might inhibit a viewer's aggressive responses because of the
manner in which that film showed the horrific consequences of vio¬
lence. The film, he said, may have "a good effect of dampening the like¬
lihood of the audience member acting aggressively himself, if he says
to himself, yes, it can have this effect."58 Congressman Boggs followed
up on this point during a subsequent interview before the Commission
25
with MPAA president Jack Valenti. "I was interested in the responses
about Bonnie and Clyde," Boggs said. "I might tell him that we had a
murder in my town committed by an 18-year-old boy who had come
out of Bonnie and Clyde one hour before. He killed a young man who
was running a drive-in grocery store. And it was just a senseless mur¬
der. Now, whether or not what he saw in Bonnie and Clyde had any im¬
pact on the murder, I don't know. But I know that what I say to you is
a fact—that he saw this movie which glorifies violence."59
Aestheticized Violence and Volatile Responses
As these examples indicate, viewers respond to graphic violence in

often divergent ways, and this may be one reason why the observable
effects in the empirical literature occur in a small percentage of view¬
ers, albeit a statistically significant one. The reactions of viewers to
screen violence tend to be inherently volatile, a phenomenon encour¬
aged by some of the essential features of cinema and one that has sig¬
nificant social import. In light of the empirical findings, I now examine
these issues. Prior to releasing The Wild Bunch in June 1969, Warner
Bros, tested the film before preview" audiences and sampled their reac¬
tions. Viewers wTere invited to fill out preview cards describing their re¬
sponses, and the film's producer, Phil Feldman, and studio marketing
personnel studied these responses for clues about the demographics of
the film's likely audience. The studio's analysis of 768 preview cards,
derived from screenings in Fresno and Kansas City, showed a prepon¬
derantly negative response to the film.60
Sixty percent of those viewers who turned in cards rated the
film unfavorably. Only 20 percent rated the film as excellent or out¬
standing, and these tended to be viewers in the 17-25 age group, the au¬
dience segment that Warners subsequently targeted in its release. The
written responses on the cards run a gamut of extreme reactions:61
"Truly a product of our sick society." "Nauseating, unending, offen¬
sive, bloody violence." "The kind of picture [mass murderer] Charles
Whitman would have enjoyed." "It's difficult to believe someone could
make such a movie—Let's all kill more!" "When a film like yours (and
Bonnie and Clyde) makes me cheer for the bad guys—it is not a good
picture." "Stark realism at its best." "It is the most honest and accu¬
rate film of human nature I have ever seen." "Most realistic motion
Kankakee Community CoHeg©
Learning Resource Center
26
Stephen Prince
picture yet conceived." "I hope the viciousness of the fights stays in the
film so people can really know how bad killing is." "The violence ob¬
viously was handled at a point near brilliance. The really shocking hor¬
ror of this mass killing definitely makes one think."
As these responses indicate, disturbing or outre images of vio¬
lence may stimulate some viewers to reflect upon the filmmaker's
moral point of view and probable intentions in crafting the images, al¬
though such viewers may arrive at very different conclusions about
these issues. At the same time, other viewers of The Wild Bunch clearly
exhibited less reflective and more excitatory reactions. Contemporary
reviewers commented upon the laughter that greeted the film's violent
episodes: "At all this the audience laughed (and so did I), not with mer¬
riment, exactly, but in tribute to such virtuosity of gore."62 Screen¬
writer Charles Higson recalled his first reaction to the film as a
teenager, one marked by a rush of excitement: "Once the film was over,
Figure 9. Covered in blood after a frenzy of shooting and stabbing that claims multiple victims,
Travis Bickle (Robert De Niro) taunts the police by mimicking suicide. Taxi Driver escalated the
threshold of permissible violence well past what Sam Peckinpah had shown in The Wild Bunch.
27
I was exhausted and in a state of high nervous excitement. I wanted to

go out in a blaze of glory. I wanted a Gatling gun. I wanted to be pierced
by a hundred bullets."63
Bonnie and Clyde and The Wild Bunch were not made as ex¬
ploitation pictures. Nevertheless, the gore in these films, and others,
elicited reactions from viewers that troubled the filmmakers who cre¬
ated the images. Hearing that a group of Nigerian soldiers screened The
Wild Bunch to get psyched-up before going into battle, Peckinpah re¬
marked, “1 heard that story and I vomited, to think that I had made that
film."64 Similarly, the climactic shoot-out in Taxi Driver aroused vigi¬
lante reactions in excited viewers, and Scorsese remarked that he found
these reactions disturbing. While wishing to distance himself from
these reactions, however, he nevertheless admitted that he derived
tremendous pleasure as a filmmaker in staging and filming this mate¬
rial. The audience, he said,
was reacting very strongly to the shoot-out sequence in Taxi Driver.

And I was disturbed by that. It wasn't done with that intent. You can't
stop people from taking it that way. What can you do? And you can't
stop people from getting an exhilaration from violence, because that's
human, very much the same way as you get an exhilaration from the
violence in The Wild Bunch. But the exhilaration of the violence at the
end of The Wild Bunch and the violence that's in Taxi Driver—because
it's shot a certain way, and I know how it's shot, because I shot it and
I designed it—is also in the creation of that scene in the editing, in the
camera moves, in the use of music, and the use of sound effects, and
in the movement within the frame of the characters. . . . And that's
where the exhilaration comes in.65
Viewers who whoop with approval at ultraviolence are often in¬

tuiting the filmmaker's own aesthetic pleasure in creating such scenes.
It is not simply that the design elicits the response. Rather, the viewer
grasps the filmmaker's own relationship to the materials, the sensuous
pleasures that a Penn, Peckinpah, Scorsese, or Tarantino has derived
from the audiovisual design of graphic violence and is manifesting
through those designs.
To a large extent, the cinema cannot present violence in other
than a pleasure-inducing capacity, and as the foregoing review of the em¬
pirical literature suggests, serious social consequences follow from this.
The medium inevitably aestheticizes violence. The arousal and expres¬
sion in cinema of 'negative' emotions— fear, anxiety, pain—typically
occur as part of a pleasure-inducing aesthetic experience. Admittedly,
28
Stephen Prince
there is much about this phenomenon—why viewers seek emotional

experiences in art that they would avoid in a nonfictional context—
which is little understood.66 But it seems likely that representations of
violence on screen that are unrelentingly horrifying, nauseating, or
disgusting will fail to attract viewers, in comparison with films that
provide aesthetic pleasures, even when the work in question, like Peck¬
inpah's, aims to be shocking and upsetting. McCauley, for example,
reported showing - a sample of university undergraduates three
documentary-style films depicting extreme violence.67 One film
showed head surgery on a young girl, during which the surgeons pulled
her face inside out, away from her skull. Another film showed cattle
stunned and butchered in a slaughterhouse, and the third showed a
group of diners butchering a live monkey and eating it. The students
watching the films were given the option of turning them off whenever
they chose to do so. On average, they turned the films off halfway
through, and only 10 percent of the sample watched the films until the
end. Unlike Hollywood films, the screen violence here was a record of
real events, and it lacked the normative aesthetic frame of commercial
features, that is, attributes such as music and special effects. The stu¬
dents found these films unattractive and did not wish to watch them.
McCauley suggests that "these three films were disgusting rather than
enjoyable because they were loaded with cues for reality and were lack¬
ing the frame of dramatic fiction."68
The evident contrast with commercial cinema is quite clear,
even in categories of film, like horror, that present extremely graphic
images of violence. Changing camera positions, controlled lighting,
montage editing, music, and special effects create significant aesthetic
pleasure and emotional distance for viewers, who can use these cues as
a means of insulating themselves from the depicted violence.69 The
episodes can be enjoyed because they are perceived as being 'not real'
by virtue of their elaborate design and special effects. A recent study by
the United Kingdom's Broadcasting Standards Commission examined
male viewers' responses to a variety of violent film and television pro¬
gramming. With regard to the elaborate violence in contemporary Hol¬
lywood films, viewers reported using the special effects to distance
themselves’from the otherwise shocking violence: "I think when it
goes that far it's not really disturbing." "You see the guy getting pierced
all over the place and blood spurting out, it doesn't really bother me,
but having to watch a documentary or something like that, and there
was an operation getting carried out, you just saw a simple incision and
29
stuff like thait, you'd be cringing, turning away from the telly."70 Zill-
mann has discussed a similar concentration on special effects by male
viewers of horror films.71
Even a film whose violence is as harrowing as that in Saving
Private Ryan employs an elaborately artificial audiovisual design.
Spielberg and cinematographer Janusz Kaminski used a number of un¬
usual techniques to get the striking; vivid, and unexpected visual qual¬
ities of that film's battle scenes. For some shots, they employed a
shutter set at 45 and 90 degrees, instead of the more usual 180-degree
configuration, in order to pixelate the action. They stripped the coating
off the lenses to flatten contrast and get a foggy but sharp look. They
shot with the camera shutter out of sync to produce a streaking effect
from the top to bottom of the image. They used a Clairmont Camera
Image Shaker to produce horizontal and vertical shaking of the camera,
and they flashed the film to desaturate the color and used ENR to add
contrast. (Named after Technicolor technician Ernesto N. Rico, ENR is
a bleach-bypass process that leaves silver in the print during develop¬
ing. This gives the film image richer, deeper shadows and desaturated
pastels.) These effects made the battlefield carnage more vivid, but they
also supplied the visual effects that viewers could use to create some
emotional and cognitive distance from the depicted violence. These
techniques gave the violence an elaborate and explicit aesthetic frame,
which was intensified by the picture's narrative of heroism and moral
redemption. The violence was not raw, that is, it was not 'real.' It was
staged for the cameras and filtered through the various effects and tech¬
niques employed by the filmmakers. By contrast, when screen violence
lacks this aesthetic dimension, its evocation of negative emotions can
be markedly unpleasant for viewers, as in the McCauley study.
These considerations suggest some uncomfortable and unpleas¬
ant facts about the medium of cinema. Filmmakers who wish to use
graphic violence to offer a counterviolence message—that is, to use vi¬
olence in a way that undercuts its potential for arousing excitatory re¬
sponses in viewers—may be working in the wrong medium. The
medium subverts the goal. In significant ways, the aesthetic contract
that the filmmaker must honor with viewers entails that screen vio¬
lence be made to offer sensory pleasures. These pleasures are implicated
in the aggression responses studied in the empirical literature. The prob¬
lem is especially serious due to the phenomenon of divergent audience
responses. The difficulty here can be stated very simply: filmmakers
cannot control the reactions of their viewers. Neither Peckinpah nor
30
Stephen Prince
Scorsese, both consummate filmmakers, wished to arouse aggressive,

vigilante reactions among their viewers, but the scenes that they
crafted did so. To a large extent, the formal design of the scenes is im¬
plicated in these reactions. Each plays up the aestheticizing organiza¬
tion of the violence. In addition to the squib effects common to both
scenes, Peckinpah's montage editing enhances the artifice and specta¬
cle of the violence, and in Taxi Driver Scorsese ends the shoot-out with
an elaborate series of tracking shots to emphasize the artfully arranged
bodies and Jackson Pollack-like collage of blood spatters. In these ways
the scenes play to responses the filmmakers wished to disavow.
But even when a filmmaker goes to great lengths to avoid play¬
ing violence to spectacle, strikingly divergent -viewer reactions may
ensue. In Schindler’s List, Spielberg films a number of killing scenes—
Nazis executing prisoners—with an excruciating intensity. In one par¬
ticularly disturbing scene, a Nazi officer executes a Jewish woman
employed as an engineer helping construct a new building in the death
camp. In this scene, Spielberg lets victim and executioner interact in an
extended, lengthy take. Spielberg avoids montage here in order to en¬
hance the emotional horror of the killing, in other words, to minimize
the viewer's ability to take refuge in the spectacle and visual effects that
montage offers. In addition, by playing the action in an extended take,
he captures the humanity of the victim by letting the action unfold in
the real time of the shot, as opposed to edited time. The filmic design
is carefully considered, and it demonstrates serious moral purpose. A
life is being taken, and the filmmaker tries to show in honest terms
what that means and to show horror in the lack of existential reciproc¬
ity between victim and executioner, in the disparity of their emotional
reactions (i.e., the earnest intensity of the victim in contrast to the cool
indifference of the killer).
It is safe to say that Spielberg did not intend for any of the vio¬
lence in Schindler’s List to be funny or to provoke laughter, yet that is
what this execution scene did for a group of Oakland, California high
school students.72 Promised an outing on Martin Luther King Day
when the school was closed, the students were taken to a screening of
the film with little preparation or context for the picture. Their laugh¬
ter prompted complaints from other patrons in the theater, a request
from the theater manager that all seventy-odd students leave, and a
great deal of media coverage of the incident. The teenagers denied that
they were expressing any racism or anti-Semitism and maintained that
they were simply doing what they always do at the movies, which is to
talk and express themselves. One 15-year-old girl described their col-
31
Figure 10. The-Basketball Diaries has been implicated in several real-world shootings in which
the killers cited the inspirational nature of the film sequence where a high school kid (Leonardo
Di Caprio) shoots his teacher and classmates. Note how the lighting and camera angle aestheti-
cize the violence by accentuating the drama and making it visually striking.
lective reaction: "It wasn't like people were laughing because people
were dying. The woman who got shot fell funny and people just
laughed."73 For some of the viewers, no doubt, the laughter was a dis¬
tancing response to alleviate tension, to insulate their feelings from the
depicted violence, and, in a high school environment rife with peer
pressure, to prevent peers from seeing that one is touched or sad and
therefore emotionally vulnerable. At the same time, however, such
laughter can signal a failure of empathy, an inability (or an unwilling¬
ness) to imaginatively place oneself inside the fiction and relate to the
pain or violence on an immediately personal level. In some films, like
Taxi Driver, the audiovisual structure of a scene might work to prevent
a viewer's empathic response by stressing the artifice and spectacle of
the violence, but here Spielberg evidently sought to show the violence
within an appropriately moral and empathy-inducing framework.
Viewer reactions were nevertheless somewhat unpredictable.
The Claremont High incident was not the only case of variant
responses elicited by the film. Washington Post columnist Donna Britt
encountered a similar incident when she went to see the film.74 She
reported that a group of middle-aged and elderly women were whoop¬
ing and laughing during scenes that prompted flinching, gasps, and tears
from the rest of the audience. A more extreme variant response oc¬
curred when an evidently deranged man in a San Diego theater shot the
woman sitting in front of him during a scene in the film where Nazis
32
Stephen Prince
were shooting Jewish victims in Krakow, Poland.75 The 45-year-old gun¬

man said that he wanted to “test God" and protect Jews. These are vari¬
ant reactions because they are not the responses that Spielberg intended
his viewers to have and that he, to the best of his not inconsiderable
filmmaking talent, had worked to create.
Spielberg's inability to control the responses of his viewers, de¬
spite his skill as a filmmaker and evident desire to show violence that
hurts and horrifies,- is a significant failure. It is one that all filmmakers
share, whatever their intentions in depicting ultraviolence. Coupled
with the findings in the empirical literature about the effects of view¬
ing violence and with theories which hold that aggression, in many
manifestations, is a socially learned response, the film industry's
continuing investment in violent spectacle does not leave one very san¬
guinary about the social health of contemporary culture. Viewer reac¬
tions to screen violence are volatile, and filmmakers cannot reliably
control these responses, that is, they cannot craft their scenes so as to
eliminate the variant reactions. Furthermore, to compel viewers to
watch, filmmakers routinely embed violence within an audiovisual de¬
sign that provides aesthetic pleasures. Screen violence is made attrac¬
tive, whether by dressing it up in special effects or by embedding it in
scenarios of righteous (i.e., morally justified) aggression. Without these
aesthetic pleasures, viewers are unlikely to consent to viewing grossly
disturbing violence.
The filmmaker who wishes to use graphic violence to instruct
or make a statement against violence pursues an elusive goal. The ef¬
fort is easily undermined by the medium's inherent aestheticizing of
violence and by the prevalence of deviant viewer reactions, chief among
these being the instigation of aggression as documented in the empiri¬
cal studies. Furthermore, popular film and culture, in general, exhibit
a fascination with the spectacle and pleasures of violence. In this re¬
gard, the forum and context from which a filmmaker would speak is
often counterproductive to the message. Working with graphic vio¬
lence, the filmmaker inherits the history of a stylistic whose conven¬
tionalized meanings are readily understood by contemporary viewers
and are, in significant ways, counter to an antiviolence philosophy.
These considerations suggest that a critique of violence may be best
pursued on screen in its absence, that is, by not showing—at least not
in graphic detail—the very phenomenon that a film would address.
Otherwise, filmmakers risk that their moral efforts be undone by
established characteristics of the medium in which they would work.
33
On the one hand, the policy implications of these considera¬

tions and the empirical findings are clear. As the current manifestation
toward which cinema violence has been moving, graphic ultraviolence
is not conducive to the long-term health of the social polity. However,
efforts to censor program material create their own problems, and calls
for the industry to regulate itself have been mostly ineffective. The
CARA ratings have always showed greater leniency toward violence
than toward sex, and graphic violence is now so endemic to the
medium, such a pervasive feature of contemporary style, that film¬
makers have become disconnected from the carnage on which they
turn their cameras. In an odd but understandable turn of events, many
filmmakers who purvey ultraviolence are emotionally disengaged from
it and show it in a dispassionate manner. For them, it is a special effect
and a box-office asset. This is quite evident in routine action films, but
it sometimes afflicts even fine directors. Oliver Stone's Natural Born
Killers (1994), for example, plays violence as a cartoon in a discon¬
nected, postmodern style that romanticizes the serial killers it pro¬
fesses to critique. Its attentive and flamboyant depictions of violence
make it a part of the violence-loving media it aims to target. Most of
the ultraviolence in Pulp Fiction is played as comedy, with no ground¬
ing in suffering or pain. Cinematographer John Bailey has remarked that
"the artifice of movie mayhem" has become "routine and unreal to us
as filmmakers."76
This unreality is symptomatic of the social disconnect of many
contemporary filmmakers. For them, violence is an image to be con¬
structed, a special effect to be staged, but not a social effect that is pro¬
duced. When characters die spectacularly bloody deaths in contemporary
crime and action films, they are, for the individuals who make these
films, just movie characters, without real-life correlates. In the culture
of ultraviolence that now engulfs the medium, moviemakers operate in
a kind of postmodern bubble, treating violence as an image and not as a
social process. Furthermore, the sheer pervasiveness of media violence
helps augment this sense of unreality. It has become an object for con¬
sumption, a familiar part of the social landscape as defined by movies and
television. As German director Wim Wenders has said, "Violence appears
in so many contexts where you cannot reflect on it any more, where you
cannot experience it any other way than consuming."''7
Violence has a legitimate place in art as part of the human ex¬
perience and as one of the mysteries of life that haunts and fascinates.
But viewers rarely experience screen violence in this fashion, treated in
34
Stephen Prince
a serious and provocative way that invites reflection and contemplation

(and, as noted, the medium of cinema does not make this kind of de¬
piction easy to achieve). Instead, commercial films offer it as spectacle,
an easy way to get to the viewer emotionally and to solve narrative is¬
sues. And it all becomes ever more unreal, ever more stylized and dis¬
connected from a viewer's personal experience, except, as the empirical
studies suggest, for its impact on the social psychology of the culture.
There seems, at present, no way out from the blood and circus that
much of present cinema has become, especially as the culture at large
manifests such fascination (albeit horrified) with an ongoing spate of
headline-grabbing homicides. The Littleton shootings were recent in¬
stances in a lengthy spate of murders in which -the killers had feasted
on media ultraviolence. In the words of a teenager sentenced to life for
killing his mother, "The first stage you see a guy's head being blown off
and you feel compassion. The second stage you see it again, you feel
compassion, but it's not as strong as the first . . . the fifth stage, you
want to do it but it's just a thought. The last stage you do it and you
want to do it again."78 Certainly this killer's words are self-serving. But
they also speak to a worst-case, nightmare scenario about the connec¬
tions between ultraviolent movies and real violence.
Once American cinema turned the corner toward graphic vio¬
lence, there has been no going back. And the culture as a whole has ac¬
companied cinema on this journey, becoming bloodier, ever more grim,
and ever more confused about the accelerating spiral of movie-induced
carnage.
The essays collected in this volume are drawn from a variety of sources
that demonstrate the complexity of the topic and the manner in which
it has cut across different realms of public discourse. These sources in¬
clude the popular press, scholarly journal articles representing work in
cinema studies and empirical social science, and testimony elicited by
congressional committees investigating problems of violence in soci¬
ety and the media. By examining movie violence in three contexts—its
history, its aesthetic characteristics and structure, and its potential ef¬
fects upon individual and society—this volume presents the topic in its
richness and complexity. As a result, readers will develop an under¬
standing of cinema violence as a representational construct and a social
phenomenon, a function of distinct historical periods, characteristics
of film form, and ongoing intellectual and research paradigms.
The first five essays in this volume set the historical context.
35
They evoke the controversies that followed the explosion of graphic

bloodshed in cinema during and after 1967 (the pivotal year that saw
the release of The Dirty Dozen, Bonnie and Clyde, A Fistful of Dollars,
and For a Few Dollars More). These films unleashed a stormy debate
about the turn contemporary cinema had taken and the likely long¬
term effects of the new violence. The essays in this section are symp¬
toms of that period and its controversy; they acquaint the reader with
the immediate reactions of critics, reviewers, and others who were
attempting to think through this confusing and seminal period in the
history of American film and of the country itself.
After first labeling Bonnie and Clyde a trash film, the reviewer
for Newsweek took the extraordinary step of publishing a second col¬
umn on the film, this time retracting his earlier opinion and reassess¬
ing the picture and its violence in more favorable terms. In "The Thin
Red Line," the reviewer, Joseph Morgenstern, defended the film by ar¬
guing that violent movies are a consequence of violent times (alluding
to the diverse currents of late 1960s unrest). He added, though, an im¬
portant qualification that resonates more powerfully twenty-some
years later: "When artists are able to bring characters to life and keep
them alive, they should not leave death to the special effects depart¬
ment." More often than not, this is exactly what movies have done—
shown violent death as a special effect, as spectacle—and Morgenstern
suggests that this is an abdication of the artist's responsibility. Al¬
though impressed with the artistic possibilities represented by Bonnie
and Clyde and the new creative freedoms the industry was granting
filmmakers, Morgenstern remained uneasy about the film's graphic
violence, and his attitude was shared by many.
With less caution and ambivalence than Morgenstern dis¬
played, Bosley Crowther, the film critic for the New York Times, ex¬
pressed strong opposition to the increase of violence brought on by the
pictures released in 1967. In a pair of columns, Crowther attacked the
new violence and worried over its likely long-term effects on the cul¬
ture. He felt that the explicitly detailed murders depicted in these
pictures were gross, bloody, massive, and excessive. Furthermore, he
believed that they expressed an attitude of "killing is fun," and he char¬
acterized this attitude as "an antisocial venom." Crowther's intensely
negative reaction to these popular pictures quickly made him a target
for criticism, but his worries were prescient. The questions he raised
are with us still. Moreover, contrary to many critics and filmmakers
(e.g., Sam Peckinpah and Arthur Penn) who argued that violent times
36
Stephen Prince
justified the presentation of explicit movie violence, Crowther argued

that violent times made it essential for the media to show balance and
restraint. Rather than give new license to artists, periods of social un¬
rest, for Crowther, ought to make them more prudent. Writing for Va¬
riety, the industry tradepaper, Ronald Gold summarized the reactions
to Crowther's columns in the film industry and among movie review¬
ers and the widespread suggestions that Crowther was out of touch
with contemporary film and the sensibilities of the audiences for these
pictures.
The year that saw release of The Dirty Dozen and Bonnie and
Clyde, and the one that followed, were among the most violent periods
in modern American history. The Vietnam War continued to escalate,
with American troop levels at 460,000, and massive antiwar protests in
New York and Washington, D.C. Urban riots erupted in 128 cities dur¬
ing the first nine months of 1967, and in April and June of 1968, Mar¬
tin Luther King Jr. and Robert Kennedy were murdered. The nation's
turmoil led Congress to form a committee of inquiry into the sources
producing the diverse skeins of domestic unrest. The National Com¬
mission on the Causes and Prevention of Violence heard testimony
from a range of social scientists, political officials, and concerned citi¬
zens. Because of the upsurge in movie violence, the Commission ex¬
amined, as part of its broad-ranging inquiries, whether Hollywood films
might be playing a role in fomenting the contemporary unrest.
In December 1968, Jack Valenti, president of MPAA, testified
before the Commission and defended the industry and its films from
suggestions that it was peddling violence for a buck and that such vio¬
lence might have undesirable social consequences. The MPAA was (and
is) a buffer group, interceding between the industry and its public and
promoting the industry's interests. In this context, Valenti's testimony
is a masterpiece of rhetoric, treating the commissioners with great re¬
spect while defending Hollywood's right to make its pictures. Of spe¬
cial interest are the Commission's concerns that focus specifically on
Bonnie and Clyde and Valenti's defense of graphic violence and the rev¬
olutionary new films and filmmakers that had been causing such con¬
troversy. He deftly insulated the industry and its films from the threat
posed by the findings of empirical studies pointing to a connection be¬
tween media violence and aggressive reactions among viewers. Valenti
maintained that evaluating screen violence was an inevitably subjec¬
tive undertaking and one not amenable to social science methods.
Quantitative studies, he suggested, especially laboratory experiments,
37
were flawed in design and conception. Thus, he claimed, social scien¬

tific findings about media effects were of limited use and that, in any
case, the real question to worry over is that of censorship. Who decides
the standards, and who applies them? As evidence that Hollywood took
its social responsibilities seriously, he offered the industry's newly for¬
mulated ratings code. Much discussion focused on the 1966 revision of
Hollywood's Production Code, and Valenti used the occasion of his tes¬
timony to introduce the newly created G-M-R-X classification scheme.
This, he said, would for the first time prevent children from seeing films
inappropriate for them. He might also have added that it significantly
expanded the creative license of filmmakers, who could now make
harder films for the R-rated adult audience, precisely the kinds of pic¬
tures that were embroiled in controversy.
The next six essays examine the aesthetics of ultraviolence.
Hollywood cinematographer John Bailey (The Big Chill [1983], In the
Line of Fire [1993], Nobody’s Fool [1994]) explores the issues faced by
filmmakers (or, at least, those issues that the honest ones think about)
who craft screen violence. Bailey asks where the difference lies between
the exploration of violence and its exploitation, and he examines the
differences of style in two violent films, Natural Born Killers and The
Shawshank Redemption. He finds that the camerawork in Natural
Born Killers visualizes and makes exciting the sociopathic point of
view of the killers, while the imagery in Shawshank humanizes the vic¬
tims of violence and creates empathy for them. For Bailey, the distinc¬
tion between exploration and exploitation turns on these differences,
on the extent to which film style deglamorizes violence and focuses on
its human consequences. He relates his own effort to grapple with these
issues during the making of China Moon, a crime film that he directed.
Bailey closes by asking that filmmakers consider their complicity in the
images they create: "If our own sensibilities are askew, if we have no
moral compass to guide us, what point of view are we going to create?"
David Thomson examines the ease and slickness of death as
cinema portrays it. He discusses the distinctive mannerisms that stars
as diverse as Humphrey Bogart, James Cagney, and Clint Eastwood
brought to their screen encounters with death and dying, and he ex¬
plores the fascination that staging deadly violence holds for filmmak¬
ers. Through a wealth of example, Thomson shows how easy death is
in the movies—glamorized and 'slo-moed'—and how false to life. He
finds that "something essential to the medium cleaves to ... the chance
to stare at death without honoring pain or loss." By contrast with its
38
Stephen Prince
cinematic renditions, death in real life typically involves waiting and

suffering, attributes that cinema has not often depicted. Thus does cin¬
ema become habituated to its own stylistic conventions: "When killing
is so easy and death so brief, it becomes a way of life."
Devin McKinney proposes a schema for classifying movie vio¬
lence: it is either "strong" or "weak." Weak violence, he suggests, is the
most plentiful sort, and 'weakness' here is not to be understood as spec¬
ifying a relative degree of explictness. The graphic ear-cutting scene in
Reservoir Dogs is an example for McKinney of weak violence. This kind
of violence carries few emotional consequences for the filmmaker who
designs it, the characters who enact it, or the viewer who watches it. It
is easily consumed and proposes no moral challenges for viewers who
are asked, merely, to assent to it and, typically, to enjoy it. The action
epics of Arnold Schwarzenegger fit this pattern, as do the killings in
Paul Verhoeven's Basic Instinct. Indeed, McKinney finds a resurgence
of weak violence in contemporary film: "The kind of disconnected, un¬
committed movie mayhem that began with James Bond, came of age
with the sociopathic crime dramas of the early 1970s, and served the
reactionary agenda of the 1980s is once again ascendant." Strong vio¬
lence, by contrast, opens an emotional trapdoor under the viewer, up¬
sets one's moral equilibrium, and, above all, grasps the consequences
of violent acts, for the characters on screen as for the viewer who
watches. McKinney explicates this category with reference to Bad
Lieutenant, The Crying Game, and Henry: Portrait of a Serial Killer. By
emphasizing the importance of aesthetic designs that delineate conse¬
quences, McKinney's argument links with those of John Bailey and
David Thomson. All propose that morally grounded designs are those
that treat violence not simply as an image, but as a phenomenon and
an experience that connects with life. For them, strong violence, vio¬
lence responsibly rendered, designates the dilemmas of life and visual¬
izes these, rather than hermetically and self-consciously referencing
cinema and its stylistic conventions.
Vivian C. Sobchack finds that graphically violent movies per¬
form a service for the viewer—they exorcise fears of chaos and sense¬
less death. Discussing such early 1970s films as A Clockwork Orange
and Straw Dogs, she suggests that the stylization of violence, achieved
through slow motion and squib effects, elicits the viewer's real-world
fears about chance encounters with violence while simultaneously al¬
leviating those anxieties. Sobchack thus suggests that graphic screen
violence performs a cathartic effect. Writing of Bonnie and Clyde and
39
its seminal influence, she notes that "it was the first major film to allow
us the luxury of inspecting what frightened us-the senseless, the un¬
expected, the bloody. And most important, it kindly stylized death for
us; it created nobility from senselessness, it choreographed a dance out
of blood and death, it gave meaning and import to our mortal twitch-
ings." By choreographing violence in such a way as to impose an aes¬
thetic order on the experience of violent death, films of graphic
bloodshed impose this order upon the viewer's conceptual and emo¬
tional experience. Thus, Sobchack suggests that graphic violence per¬
forms a useful social function. Her essay and her reflections date from
the early 1970s. In the essay's afterword, Sobchack reassesses these
questions some twenty-five years later, in light of the intervening
decades of movie violence. The contrast between her reflections past
and present illuminates the trajectory that the nation's culture, as a
whole, has traveled in relation to cinema violence.
Carol J. Clover explicates the sexual and gendered aspects of
violence in horror films, specifically the slasher film. Since the 1980s,
these grisly slaughter-fests had taken over the horror genre, offering
viewers stalk-and-murder narratives overlaid with detailed imagery of
butchery and mutilation. While most critics were quick to condemn
these films as trash, Clover, instead, took a close view and found rich
psychological material in the format. Her illuminating analysis clari¬
fies the complex appeal of these films for viewers, amid all of the prob¬
lematic and prurient violence.
In the concluding essay in this section, I examine the origins
and aesthetic of slow-motion violence. Since Bonnie and Clyde and
The Wild Bunch, this has become the predominant way of stylizing vi¬
olent gun battles. Filmmakers insert slow-motion shots into extended
montages of death and destruction. I explore how filmmakers (chiefly,
Sam Peckinpah, who did more than any other director to popularize this
aesthetic) orchestrate these scenes and the emotional and moral impli¬
cations of these constructions.
Explorations of aesthetics inevitably open onto issues of effect.
The remaining essays examine these issues. The essays by Leonard
Berkowitz and Richard B. Felson offer somewhat contrasting portraits
of the state of research on media violence. Berkowitz provides a detailed
explanation of the role that cognitive factors (thoughts and ideas) play
in viewer reactions to media violence. Readers should not be put off by
the somewhat wordy title of this essay,- the discussion is clearly pre¬
sented and very straightforward. While arguing that the social costs of
40
Stephen Prince
violent imagery in the mass media are real and may be deemed unac¬
ceptable, Berkowitz shows the complexity of media-viewer interac¬
tions. Specifically, he explicates the role of thought (induced by media
programming) in activating bodily reactions. He terms this process the
"spreading activation" of thoughts related to media images and content
and which, under the right conditions, can dampen a viewer's inhibi¬
tions against acting aggressively. Berkowitz examines the factors that
can trigger and facilitate this "spreading activation." His account
shows that empirical research on media violence does not construct a
mechanistic portrait of media messages and passively reacting viewers.
Instead, while the bulk of the evidence clearly shows aggression-
inducing characteristics of media programming, .the processes involved
are complicated and multifactoral.
While a general consensus prevails among social scientists
about the aggression-inducing characteristics of movie and television
content, some controversy exists about the strength and duration of
these effects. In the final essay, sociologist Richard B. Felson undertakes
a thorough review of the empirical literature and offers a more tenta¬
tive assessment of media effects. He concludes that, while effects are
present, they are likely to be weak and to affect only a small percent¬
age of viewers, and are unlikely to be a significant factor in the high
crime rate in this country.
The essays by Berkowitz and Felson give the reader an exten¬
sive account of the available literature and theory on the effects of
media violence and a sampling of the controversy and range of opinion
that bear on this issue. There is as yet no definitive resolution of the
problem. The questions, though, remain urgent ones, from the stand¬
point of individual as well as social psychology. Neither Felson nor
Berkowitz suggests that the media play a clearly causative role in fos¬
tering real-world aggression. But they do implicate film and television
in the mix of social factors that incline individuals to behave violently.
In light of the empirical work on media effects, the fascination with
violence in popular culture should give us pause. It is an attribute
unlikely to be conducive to the general well-being or social health of
our society.
i
NOTES
1. Marcia Pally, "'Double' Trouble," Film Comment 20 (September-October 1984),

p. 14.
2. A thorough review of this period can be found in Garth Jowett, Film: The Demo¬
cratic Art (Boston: Little, Brown, 1976).
41
3. Mass Media Hearings, vol. 9A: A Report to the National Commission on the
Causes and Prevention of Violence (Washington, D.C.: U.S. Government Printing Office,
1969), p. 193.
4. "'Brutal Films Pale Before Televised Vietnam'—Valenti," Variety, February 21,
1968, p. 2.
5. "Theatre Operators and Public Require Updating on Social Point of View," Va¬
riety, March 6, 1968, p. 7.
6. Quoted in Peter Wood, "How a Film Changes from an 'X' to an 'R,'" New York
Times, July 20, 1980, sec. C. .*■
7. "Pix Must 'Broaden Market,'" Variety, March 20, 1968, p. 1.
8. I examine these factors in Savage Cinema: Sam Peckinpah and the Rise of
Ultraviolent Movies (Austin: University of Texas Press, 1998).
9. "Brutal Films Pale Before Televised Vietnam," p. 2.
10. Gary Crowdus and Richard Porton, "The Importance of a Singular, Guiding Vi¬
sion: An Interview with Arthur Penn," Cineaste 20:2 (Spring 1993), pp. 8-9.
11. Robert J. Landry, "It's Murder, Italian Style," Variety, February 8, 1967, p. 7.
12. Review of The Good, the Bad and the Ugly, Variety, December 27, 1967, p. 6.
13. Bosley Crowther, "Movies to Kill People By," New York Times, July 9, 1967, sec.
2, p. 10.
14. Crowdus and Porton, p. 9.
15. "'Bonnie and Clyde's' Booming Repeats," Variety, February 14, 1968, p. 3.
16. "Hollywood: The Shock of Freedom in Films," Time, December 8, 1967, p. 73.
17. Phil Feldman memo, May 13, 1969, folder no. 46, Sam Peckinpah Collection,
AMPAS Library, Los Angeles.
18. Wild Bunch—Bill Faralla, folder no. 43, letter of March 19, 1968, Sam Peckin¬
pah Collection.
19. Wild Bunch reaction cards, folder no. 65, Sam Peckinpah Collection.
20. I explore Peckinpah's efforts in this regard in Savage Cinema.
21. Carol J. Clover, Men, Women, and Chain Saws: Gender in the Modern Horror
Film (Princeton, N.J.: Princeton University Press, 1992), p. 41.
22. Variety, June 8, 1988, p. 24.
23. Roger Watkins, "'Demented Revenge' Hits World Screens," Variety, October 29,
1980, p. 3.
24. "Chi Tribune Blasts Gory X-Films in R-Rated Clothing," Variety, November 12,
1980, p. 6.
25. Ibid., p. 30.
26. Clover, p. 12.
27. Ruth A. Inglis, "Self-Regulation in Operation," in The American Film Industry,
rev. ed., ed. Tino Balio (Madison: University of Wisconsin Press, 1985), p. 378.
28. Herbert Blumer, Movies and Conduct (New York: Macmillan, 1933), p. 74.
29. Mark Crispin Miller, "In Defense of Sam Peckinpah," Film Quarterly 28:3
(Spring 1975), p. 13.
30. "Films, Like TV, Lack Research on 'Violence,'" Variety, December 25, 1968, p. 5.
31. Walter Kaufmann, Tragedy and Philosophy (New York: Anchor Books, 1969),
p. 34.
32. Dorothy L. Singer and Jerome L. Singer, "TV Violence: What's All the Fuss
About?," Television and Children 7:2 (1984), p. 30.
33. Leonard Berkowitz, "Some Effects of Thoughts on Anti- and Prosocial Influ¬
ences of Media Events: A Cognitive-Neoassociation Analysis," Psychological Bulletin
95, no. 3(1984), p. 414.
34. Haejung Paik and George Comstock, "The Effects of Television Violence on An¬
tisocial Behavior: A Meta-Analysis," Communication Research, 21, no. 4 (August 1994),
pp. 516-46. For a different view of media effects, see Felson in this volume.
42
Stephen Prince
35. Eli A. Rubinstein, "Television Violence: A Historical Perspective," in Children

and the Faces of Television, ed. Edward L. Palmer and Aimee Dorr (New York: Academic
Press, 1980).
36. Surgeon General's Scientific Advisory Committee on Television and Social Be¬
havior, Television and Growing Up: The Impact of Televised Violence (Washington, D.C.:
U.S. Government Printing Office, 1972).
37. Seymour Feshback, "The Role of Fantasy in the Response to Television," Jour¬
nal of Social Issues 32, no. 4 (1976), pp. 71-85; "Reality and Fantasy in Filmed Violence,"
in Television and Social Behavior: A Technical Report to the Surgeon General’s Scien¬
tific Advisory Committee on Television and Social Behavior, ed. John P. Murray et al.
(Washington, D.C.: U.S. Government Printing Office, 1972), pp. 318-45; "The Stimulat¬
ing Versus Cathartic Effects of a Vicarious Aggressive Activity," Journal of Abnormal and
Social Psychology 63, no. 2 (1961), pp. 381-85; "The Drive Reducing Function of Fantasy
Behavior," Journal of Abnormal and Social Psychology 50 (1955), pp. 3-11. See also Gary
A. Copeland and Dan Slater, "Television Violence and Vicarious Catharsis," Critical
Studies in Mass Communication 2 (1985), pp. 352-62; Sidney A. Manning and Dalmas
A. Taylor, "Effects of Viewed Violence and Aggression: Stimulation and Catharsis," Jour¬
nal of Personality and Social Psychology 31, no. 1 (1975), pp. 180-88; Ann Roth Pytkow-
icz, Nathaniel N. Wagner, and Irwin G. Sarason, "An Experimental Study of the
Reduction of Hostility Through Fantasy," Journal of Personality and Social Psychology
5, no. 3 (1967), pp. 295-303.
38. Feshback, "The Catharsis Hypothesis and Some Consequences of Interaction
with Aggressive and Neutral Play Objects," Journal of Personality 24 (1956), pp. 449-62.
39. Leonard Berkowitz and Russell G. Geen, "Film Violence and the Cue Properties
of Available Targets," Journal of Personality and Social Psychology 3, no. 5 (1966), pp.
525-30.
40. Albert Bandura, "Psychological Mechanisms of Aggression," in Human Ethol¬
ogy, ed. M. von Cranach et al. (New York: Cambridge University Press, 1979).
41. Albert Bandura, Dorothea Ross, and Sheila A. Ross, "Vicarious Reinforcement
and Imitative Learning," Journal of Abnormal and Social Psychology 67, no. 6 (1963), pp.
601-7.
42. Ibid., p. 606.
43. Ibid.
44. Albert Bandura, Dorothea Ross, and Sheila A. Ross, "Imitation of Film-Medi¬
ated Aggressive Models," Journal of Abnormal and Social Psychology 66, no. 1 (1963),
pp. 3-11.
45. Leonard Berkowitz and Edna Rawlings, "Effects of Film Violence on Inhibitions
against Subsequent Aggression," Journal of Abnormal and Social Psychology 66, no. 5
(1963), pp. 405-12.
46. Ibid., p. 411.
47. Dolf Zillmann, "The Psychology of the Appeal of Portrayals of Violence," in
Goldstein, Why We Watch, p. 202.
48. Ibid., p. 200.
49. Robert A. Baron, "Magnitude of Victim's Pain Cues and Level of Prior Anger
Arousal as Determinants of Adult Aggressive Behavior," Journal of Personality and So¬
cial Psychology 17, no. 3 (1971), pp. 236-43; Glenn S. Sanders and Robert Steven Baron,
"Pain Cues and Uncertainty as Determinants of Aggression in a Situation Involving Re¬
peated Instigation," Journal of Personality and Social Psychology 32, no. 3 (1975), pp.
495-502.
50. D. Caroline Blanchard, Barry Graczyk, and Robert J. Blanchard, "Differential Re¬
actions of Men and Women to Realism, Physical Damage, and Emotionality in Violent
Films," Aggressive Behavior 12 (1986), pp. 45-55.
43
51. Ibid., p. 51.

52. Dolf Zillmann and James B. Weaver III, "Psychoticism in the Effect of Prolonged
Exposure to Gratuitous Media Violence on the Acceptance of Violence as a Preferred
Means of Conflict Resolution," Personality and Individual Differences 22, no. 5 (1997),
pp. 613-27. On nontransient effects, see also Zillmann and Weaver, "Effects of Prolonged
Exposure to Gratuitous Media Violence on Provoked and Unprovoked Hostile Behavior,"
Journal of Applied Social Psychology 29, no. 1 (1999), pp. 145-65.
53. Berkowitz, p. 411.
54. Russell Green, "Behavioral and Physiological Reactions to Observed Violence:
Effects of Prior Exposure to Aggressive Stimuli," Journal of Personality and Social Psy¬
chology 40 (1981), pp. 868-75; M. Thomas, R. Horton, E. Lippincott, and R. Drabman,
"Desensitization to Portrayals of Real-Life Aggression as a Function of Exposure to Tele¬
vision Violence," Journal of Personality and Social Psychology 35 (1977), pp. 450-58; Vic¬
tor R. Cline, Roger G. Croft, and Steven Courrier, "Desensitization of Children to
Television Violence," Journal of Personality and Social Psychology 27, no. 3 (1973), pp.
360-65.
55. Berkowitz, p. 418.
56. Dolf Zillmann, "Excitation Transfer in Communication-Mediated Aggressive
Behavior," Journal of Experimental Social Psychology 1 (1971), pp. 419-34; P. H. Tan-
nenbaum and Dolf Zillmann, "Emotional Arousal in the Facilitation of Aggression
Through Communication," in Advances in Experimental Social Psychology, vol. 8, ed.
Leonard Berkowitz (New York: Academic Press, 1975), pp. 149-92.
57. Wayne Wilson and Randy Hunter, "Movie-Inspired Violence," pp. 435-41.
58. Mass Media Hearings, vol. 9A: A Report to the National Commission on the
Causes and Prevention of Violence (Washington, D.C.: U.S. Government Printing Office,
1969), p. 43.
59. Ibid., p. 207.
60. Sam Peckinpah Collection, Feldman memos, folder no. 46.
61. Sam Peckinpah Collection, Wild Bunch preview reaction cards, folder no. 65.
62. Review of The Wild Bunch in The Nation, July 14, 1969, p. 61.
63. Charles Higson, "The Shock of the Old," Sight and Sound 5, no. 8 (August 1995),
p. 36.
64. P. F. Kluge, "Director Sam Peckinpah, What Price Violence," Life (August 11,
1972), p. 53.
65. Anthony DeCurtis, "What the Streets Mean: An Interview with martin Scors¬
ese," in Plays, Movies, and Critics, ed. Jody McAuliffe (Durham, N.C.: Duke University
Press, 1993), p. 211.
66. Goldstein's Why We Watch explores this issue and concludes with a series of
unanswered questions.
67. The citation that McCauley provides for this information is unhelpful because
it does not, in fact, contain information about these films. He cites J. Haidt, R. C. ALc-
Cauley, and P. Rozin, "Individual Differences in Sensitivity to Disgust: A Scale Sampling
Seven Domains of Disgust Elicitors," Personality and Individual Differences 16 (1994),
pp. 701-13. McCauley discusses the research involving the films in "When Screen Vio¬
lence Is Not Attractive," Why We Watch, pp. 144-62.
68. McCauley, "When Screen Violence Is Not Attractive," p. 161.
69. "Men Viewing Violence," Broadcasting Standards Commission, London, 1998.
70. Ibid., p. 41.
71. Zillmann, "The Psychology of the Appeal of Portrayals of Violence," p. 198.
72. Christine Spolar, "The Kids Who Laughed Till It Hurt," Washington Post,
March 10, 1994, pp. Cl, C4.
73. Ibid., p. C4.
44
Stephen Prince
74. Donna Britt, "Lights, Camera, Sad Reaction," Washington Post, March 15, 1994,
pp. Bl, B5.
75. "Gunman Gets 6 Years for 'Schindler's List' Shooting," Los Angeles Times, Au¬
gust 10, 1994, p. B8.
76. John Bailey, "Bang Bang Bang Bang, Ad Nauseum," American Cinematographer
75:12 (December 1994), p. 26.
77. Manohla Dargis, "Sleeping with Guns," Sight and Sound 7, no. 5 (May 1997),
p. 21.
78. "'Scream' Killer Fingers Film, TV," The Hollywood Reporter, July 25, 1999, on¬
line version.
<
The Historical Context
of Ultraviolence
v
■s
Joseph Morgenstern
The Thin Red Line
Last week this magazine said that Bonnie and Clyde, a tale of two
young bank robbers in the 1930s, turns into a "squalid shoot-'em for
the moron trade" because it does not know what to make of its own
violence. I am sorry to say I consider that review grossly unfair and
regrettably inaccurate. I am sorrier to say I wrote it.
Seeing the film a second time and surrounded by an audience no
more or less moronic than I, but enjoying itself almost to the point of
rapture, I realized that Bonnie and Clyde knows perfectly well what to
make of its violence, and makes a cogent statement with it—that vio¬
lence is not necessarily perpetrated by shambling cavemen or quivering
psychopaths but may also be the casual, easy expression of only slightly
aberrated citizens, of jes' folks.
I had become so surfeited and preoccupied by violence in daily
life that my reaction was as excessive as the stimulus. There are indeed
a few moments in which the gore goes too far, becomes stock shockery
that invites standard revulsion. And yet, precisely because Bonnie and
Clyde combines these gratuitous crudities with scene after scene of
dazzling artistry, precisely because it has the power both to enthrall and
appall, it is an ideal laboratory for the study of violence, a subject in
which we are all matriculating these days.
Violent movies are an inevitable consequence of violent life.
They may also transmit the violence virus, but they do not breed it any
more than the Los Angeles television stations caused Watts to riot. Dis¬
tinctions can and must be made between violent films that pander and
violent films that enlighten, between camp, comment, and utter cyni¬
cism. And there is nothing like the movies for giving us historical per¬
spective on violence we have known, and in many cases loved.
No one but Charlie Chaplin's competitors ever deplored his
From Newsweek, August 28, 1967. © 1967, Newsweek, Inc. All rights reserved. Reprinted
by permission.
47
48
Joseph Morgenstern
early comedies although they served up staggeringly large helpings of

mayhem. Cruelty to animals, children, and adults was the crucial ingre¬
dient in W. C. Fields's social satire. No one ever accused Cagney of
excess gentility in Public Enemy, but people consoled themselves in
those days with the notion that violence was the particular province of
a particular minority, namely, the violent. They called the group The
Underworld, at least until 1939.
World War- II brought back primitivism, which had been on the
skids ever since 1918, and its popularity today has not discernibly
declined. A fair amount of contemporary movie violence is still
conventionally primitive, unadorned by anything but gangrenous nos¬
talgia—Battle of the Bulge, for instance. Some-movie violence is styl¬
ishly primitive—St. Valentine’s Day Massacre pumps slugs into lugs by
the thousands, but a few good performances lift the sleazy legend from
the sewer to the gutter. Some is pretentiously primitive—The Chase
was all awash with racial and social symbols yet seemed most pleased
with itself when Marlon Brando's battered face was awash with
ketchup. And some is ingeniously primitive—The Dirty Dozen spends
more than two hours on an outlandishly detailed setup for a half-hour
payoff in which the GI demolition squad really demolishes, the charges
explode, the Kraut machine guns chatter, and the victims (including
lots of screaming females) cry themselves a river of blood.
Such stuff as this is trash, and at least has the bad grace to give
itself away. More serious complications arise when the overlay of com¬
edy or comment is done more artfully. A Fistful of Dollars and its
sequel, For a Few Dollars More, were synthetic Westerns made on the
cheap in Italy and unmistakably brutal. But their director, Sergio Leone,
had done his homework and studied the models he wanted to copy
and/or parody. His primary motive clearly was profit, and therefore imi¬
tation, but that did not prevent him from adding a pinch of put-on and
a dollop of dubious satire.
Yet violence can serve thoroughly satiric or artistic ends, and
not only in Shakespeare, Marlowe, Bunuel, "Marat/Sade," or the com¬
ing film version of In Cold Blood. Violence was downright charming in
The Quiet Man, delightful in The Crimson Pirate, enthralling in Psy¬
cho and The Hill, relevant in Dutchman. West Side Story might well
have done with more of it to stiffen its spine, and so might Up the Down
Staircase, which only tiptoed timorously around the crazy chaos of
America's slum schools. ¥
There is nothing timorous about Bonnie and Clyde, in which

violence is at once a virtue and a vice. Director Arthur Penn and his col-
49
'The Thin Red Line
Figure 11. Using multiple cameras and slow motion, director Arthur Penn was the first American
filmmaker to capture the spastic agonies of gunshot victims. Bonnie Parker (Faye Dunaway) ex¬
pires in a hail of machine gunfire.
leagues perform poignant and intricate wonders with a Loony Toon gang
of outlaws who bumble along from one bank job to another, from one
blood bath to another, in an inchoate, uncomprehending, and fore¬
doomed attempt to fulfill their stunted selves. Both Bonnie and Clyde
and St. Valentine's Day Massacre deal with the same slice of life, yet the
characters in the latter are gun racks and the characters in Bonnie have
a rushing of blood in their veins and a torrent of thought in their minds.
From time to time, however, all artistry falls by the wayside:
when a cop trying to stop the getaway car is shot in the face, when a
grocer is bludgeoned in a close-up, when'a grenade demolishes a police
tank, when one outlaw has his skull blown open and another has her
eye shot out. These scenes are reprehensible not because they are ugly
or shocking, but because they are familiar, gross, and demeaning. When
artists are able to bring characters to life and keep them alive, they
should not leave death to the special effects department.
There is, in the depiction of violence, a thin red line between
the precisely appropriate and the imprecisely offensive. Sometimes a
few too many frames of film may mean the difference between a shot
50
Joseph Morgenstern
that makes its point concisely and one that lingers slobberingly. These
few frames or scenes in Bonnie and Clyde will hardly change the course
of human events. When we talk about movies, we are not talking about
urban renewal programs, nuclear nonproliferation treaties, or rat con¬
trol bills. Art cannot dictate to life and movies cannot transform life,
unless we want to retool the entire industry for the production of prop¬
aganda. But art can certainly reflect life, clarify and improve life; and
since most of humanity teeters on the edge of violence every day, there
is no earthly reason why art should not turn violence to its own good
ends, showing us what we do and why. The clear danger, of course, is
that violence begets violence in life and engenders confusion in art. It
is a potent weapon, but it tends to aim the marksman.
I
Bosley Crowther
Movies to Kill People By
v
/
Something is happening in the movies that has me alarmed and dis¬

turbed. Moviemakers and moviegoers are agreeing that killing is fun.
Not just old-fashioned, outright killing, either, the kind that is quickly
and cleanly done by honorable law enforcers or acceptable competitors
in crime. This is killing of a gross and bloody nature, often massive and
excessive, done by characters whose murderous motivations are mor¬
bid, degenerate, and cold. This is killing of the sort that social misfits
and sexual perverts are most likely to do. And the eerie thing is that
moviegoers are gleefully lapping it up.
Not all moviegoers, thank goodness. There are plenty who dis¬
gustedly eschew this new flock of sadistic pictures that is coming to
roost on our screens. But enough of them are eagerly swarming to see
such slaughterhouse films as Robert Aldrich's The Dirty Dozen and
Sergio Leone's For a Few Dollars More that it makes one tremble with
amazement and wonder what this interest represents.
Take this picture The Dirty Dozen, which is a brazen and bru¬
tal account of how a group of American military prisoners, condemned
for murder, rape, and other major crimes, are taken from a military
prison in England shortly before D-Day in World War II and secretly
trained as a team of commandos to mop up a chateau-full of Nazi offi¬
cers on the eve of the assault on the Normandy coast.
It is not just a standard glorification of killing Nazis by brave
American troops, which is a sad enough ethical perversion to practice and
prolong on the screen. It is glorification of killing by hardened criminals
who are willfully trained by a hard-bitten major (Lee Marvin) with an evi¬
dent sadistic streak. And although most of these corrupt commandos are
killed in the course of their slaughterhouse task, and thus are denied
the enjoyment of the commutation of their sentences that has been
From The New York Times, July 9, 1967. Copyright © 1967 by The New York Times.
Reprinted by permission.
51
52
Bosley Crowther
promised them, they are elevated as heroes by the attitude that is blaz-
ingly assumed by Mr. Aldrich's direction of Nunnally Johnson and
Lukas Heller's script.
What is more, the initial recalcitrance and arrogance of these
brutes is given a tone of social defiance and individuality. The inculca¬
tion in them of a sort of hoodlum esprit de corps is developed as a fea¬
ture of excitement that is obviously meant to stimulate glee. And the
ultimate killing of the Nazis and their shrieking, cowering concubines
is a superfluous amount of shooting and slashing and throwing of fire.
If one could find in the structure of this picture or in the way it
was angled and staged the slightest hint of intentional, sardonic com¬
ment upon the fundamental nature of war—the slightest glimmer of
revelation that all killing is essentially criminal—then the hideous bru¬
tality of it might be regarded as subtle irony, and the glorifying of its
felons as a tragic travesty.
But there are no such hints or glimmers in it. It is a blatant and
obvious appeal to the latent aggressiveness and sadism in undiscrimi¬
nating viewers. And I would guess that the vast majority of the people
who are seeing it at the Capitol and at the 34th Street East, where it is
playing, are taking it for kicks and thrills and are coming away from it
palpitating with a vicarious sense of enjoyment in war.
A similar sense of fun out of killing is being induced, I feel sure,
by Mr. Leone's excessively violent For a Few Dollars More. This is
another of those clever Italian-made Western films, shot in Spain, with
a cast of actors speaking a babel of languages, but with English dialogue
dubbed, and it follows the further adventures of the character created
by Clint Eastwood in Mr. Leone's A Fistful of Dollars.
This gray-eyed, cool-cat gunman is a bounty killer here—that
is, a self-commissioned huntsman whose prey are bandits with prices
on their heads—and he is after a gang of Mexican bank robbers who are
being stalked by another bounty killer, too. So the deadly confronta¬
tions are not only between the huntsman and the men he is out to bring
back dead, but also between the two huntsmen, who clearly have no
feeling for each other's lives.
Having found a market for bloodlust with his previous Dollar
film, Mr. Lepne is out to exploit it for all it's worth with this further
killing spree. He fills the screen with violence, men stalking other men
like animals—with blazing guns and whining bullets and bodies falling
in spasm-twitching heaps. And he makes the deadlines of Mr. East-
53
Movies to Kill People By
wood—his urge and efficiency at knocking men off—a matter of cool

bravado and glee-provoking triumph.
This film was a huge success in Europe before it opened here
last week at the Trans-Lux West and other theaters, so the passion for
this sort of thing is not exclusive to audiences in the United States.
There have been previous films of this order, and there are going to be
many more. Even though they merely purge aggressive spirits, which
some people say is all they do, they seem to me as socially decadent and
dangerous as LSD.
Bosley Crowther
Another Smash at Violence
There appears to have been some disagreement with the feeling of

alarm I expressed at the increase of violence in movies a couple of
weeks ago. One reader, who said she is a teacher—and a wife and
mother—in Syracuse, was incensed that I should have used The Dirty
Dozen as an example of excessive violence aimed, as I candidly calcu¬
lated, to provide the general audience with kicks and thrills.
How could I be so callous as to criticize a picture that shows
the "inner sensitivities" of criminals who are given a chance to redeem
themselves by performing a "productive" mission (killing Germans),
she wanted to know. My attitude is not shared by "people whose fam¬
ilies may be involved in Vietnam," she assured me.
Several others repeated this notion, as though the fact that
there is violence in the world is sufficient justification for exploding an
irrational excess of it on the screen.
Another reader slapped my wrist soundly for not appreciat¬
ing For a Few Dollars More because its cold-blooded cowboy bounty
killer is as much of "an existentialist hero as Camus' Caligula." This
literary-intellectual slotting of a fellow who guns people down with
deliberate self-serving indifference elevates the splurge of blood and
death in a viciously sadistic picture to the dignity of art, it seems.
"It is odd," continued this reader, "that one should show so
much distaste for [this hero's] killing powers in a society which reads
panegyrics to the Vietcong-killing power of the M-16 rifle."
Right there, with that extreme assumption, this reader and sev¬
eral others who wrote echoing his sweeping generalization put the case
for restraint on the screen.
It is precisely because there are vast areas of violence and blood¬
shed in our world—because there are certain elements that foment and
From The New York Times, July 30, 1967. Copyright © 1967 by The New York Times.
54
55
Another Smash at Violence
justify killing on the grounds of its ultimately leading to solutions of

political and social ills—that our media of so-called entertainment
should strive for balance and moral truth. It is precisely because there
are panegyrists of killing that the exponents of life should cry out with
distaste.
The fact that there are cold discussions of how efficiently a cer¬
tain weapon kills, weekly publications of boxscores of casualties in
Vietnam, enthusiastic glorifications of military conquerors in the Mid¬
dle East, and irresponsible boastings and backslappings for violence in
Newark and Detroit is not to say that many people—many good and
humane people in this world—do not deplore these manifestations and
yearn for their hastened cessation.
It is the fallacious idea that violent movies are playing an im¬
portant cultural role as ironic reflection and commentators on these sad
events or are offering release for anxieties and torn emotions with their
excessive fantasies that some thoughtful critics and philosophers use
to rationalize this trend.
Significantly, this idea is interestingly put in the new Peter
Watkins picture, Privilege, which came to the Sutton last week. The
calculated device of a pop singer to have himself tormented and tor¬
tured by fake police while he is agonizingly singing lurid folk songs cry¬
ing for freedom is coolly justified by his managers as providing
emotional release for the pent-up aggressions of his idolaters who ago¬
nize and moan with him. The hypocrisy of this thesis is indicated by
the evidence of the hold that this illusory symbol of protest has on his
followers and by how readily he can lead them to a religion of con¬
formity, all to serve the purposes of a developing British dictatorship.
Such is the strange intoxication that this wave of violent
movies can have for a public that is happily submissive to their wild
and gory fantasies. They can lead the halfway preconditioned public to
condone preposterous values and cruel deceits.
The fine French director, Claude Autant-Lara, observed several
years ago, in commenting on the nature of war films for a volume pub¬
lished by Robert Hughes, that "by habituating the public to brutality
in the exceptional as in the everyday war film, they (the war-mongers
of the world) bring about the general acceptance of force as the only
solution to any drastic situation—even in international problems."
By habituating the public to violence and brutality—by making
these hideous exercises into morbid and sadistic jokes, as is done in
The Dirty Dozen—these films of excessive violence only deaden their
56
Bosley Crowther
sensitivities and make slaughter seem a meaningless cliche. One reader

wrote me she sat next to a 10-year-old boy at the above cited film. She
asked him how he liked it. He looked at her blandly and shrugged.
I hate to have to be insistent and alarmist about these films. But
when I see one such as Frank Sinatra's The Naked Runner, which posits
the idea that a man might be righteously suborned by British Intelli¬
gence to assassinate a defector because he is going to bear military
secrets to the "enemy," or another such as The St. Valentine’s Day Mas¬
sacre, which seems to take a morbid delight in reenacting in great detail
the slaughter of Bugs Moran's Chicago gang in 1929, I feel again and
again the penetration of an antisocial venom into my own flesh and I
dread how widely such deliberate exploitation of the public's suscepti¬
bilities is poisoning and deadening our fiber and strength.
»
Ronald Gold
Crowther’s ‘Bonnie’-
Brook: Rap at Violence
Stirs Brouhaha
New York Times reviewer Bosley Crowther has long been the most
influential critic in America, and during the course of his career he has
been one of the industry's staunchest supporters and defenders. But
almost as often he has been at odds with the Hollywood moguls, some¬
times leading a crusade of his fellow scribes against the current tide of
crass "commercialism."
Lately, Crowther has begun another crusade, against what he
calls a "wave of violent movies." But this time he seems to have taken
on not only the filmmakers, but his fellow critics, the bulk of whom
have pointedly disagreed with his "violent" blasts at two recent pix,
Robert Aldrich's The Dirty Dozen and Arthur Penn's Bonnie and Clyde.
Indeed some of them have gone so far as to suggest that his cru¬
sading spirit has clouded his perception, that he's been unable to sepa¬
rate a film's statement about violence from the thing itself—and,
particularly in the case of Bonnie, that he has hurt the cause of serious
filmmaking in America by shooting down a work of art.
Crowther's basic thesis, as he expressed it in the second of two
Sunday pieces, ironically entitled "A Smash at Violence," is that "By
habituating the public to violence and brutality . . . films of excessive
violence only deaden their sensitivities and make slaughter seem a
meaningless cliche." He added that "such deliberate exploitation of the
public's susceptibilities is poisoning and deadening our fiber and
strength," and he rejected the view that violence in the world is an
excuse for it on the screen. "It is precisely because there are vast areas
of violence and bloodshed in our world . . . that our media of so-called
entertainment should strive for balance and moral truth," he said.
From Variety, August 30, 1967, pp. 5, 26. Reprinted by permission.
57
58
Ronald Gold
Some Don’t Agree
While nobody has come out in favor of violence for its own sake, some
of those who've been arguing with Crowther have thrown back at him
his own arguments in favor of increased on-screen liberality in the area
of sex. Indeed, Moira Walsh, critic for the Jesuit weekly America and
reviewer for the National Catholic Office for Motion Pictures, sounds
like an echo of the Times reviewer himself, when he defended such
films as The Pawnbroker, Who’s Afraid of Virginia Woolf! and Blow-
Up from the onslaughts of the bluenoses.
"When I started looking at movies in a 'quasi-official' capacity,"
Miss Walsh said, "I was thoroughly indoctrinated with the prissy
notion that 'common folk' should be protected from 'bad' and 'danger¬
ous' ideas in films. By slow and painful steps I was disabused of this out¬
look. It simply cannot be squared with any coherent view of the
realities of mid-20th-century life.. .. Without being dogmatic about it,
I tend to the opinion that the hardest thing for a film to do is to change
anyone's point of view; that films reflect the realities of the age far more
than they influence them."
Rather than approaching the subject on this abstract level, most
of Crowther's confreres who've chosen to take up his challenge (usu¬
ally without mentioning him by name) have responded negatively to
his dismissal of work by two highly respected filmmakers as nothing
but crassly commercial trash, essentially "sympathetic" to violence
and devoid of irony.
They've directly contradicted his view and, by implication,
have accused Crowther of totally ignoring the technical felicities of the
films at hand as well as the intentions of the men who made them.
In his review of The Dirty Dozen, for example, and in the Sun¬
day piece which followed it, Crowther was unable to find "the slight¬
est hint of intentional sardonic comment on the fundamental nature of
war," and his overview of the film was that it is "a raw and preposter¬
ous glorification of a group of criminal soldiers" and "a studied indul¬
gence of sadism that is morbid and disgusting beyond words."
Dirty Dozen
None of the other critics ignored the violence in Dirty Dozen, and some
of them were quite unhappy about it too, accusing director and screen-
Figures 12 & 13. For 1960s audiences, one of the most shocking scenes in Bonnie and Clyde oc¬
curs following a bungled holdup, when Clyde shoots a bank official in the face at point blank-range.
The victim hangs for a moment on the side of the car, his face pressed against the shattered glass.
60
Ronald Gold
writers of copping out in the final third of the film and spoiling the
point. But all of them were prepared to see that there was a point—one
which some of them emphasized later when taking up the discussion
of violence.
"This particular film/' said Judith Crist on the "Today" Show,
"very clearly implies that it takes killers and psychopaths to do a suc¬
cessful job in war, which is a murderous and perhaps psychotic pas¬
time." And Miss Walsh said Dozen contained "at least implicitly, the
most uncompromising attack on the 'typical military mind' and expo¬
sition of the insanity and hell of war that I have ever encountered in a
potentially popular 'mass audience' movie."
Accepting the picture on its own terms, most other critics also
avoided Crowther's preoccupation with whether or not the central sit¬
uation of Dozen (allegedly taken from an actual incident) was "a fic¬
tional supposition that is silly and irresponsible." And despite the
Times man's accusation that "a raw and unmitigated campaign of sheer
press-agentry has been trying to put across the notion that Warner
Brothers' Bonnie and Clyde is a faithful representation of the desper¬
ado careers of Clyde Barrow and Bonnie Parker," they stayed away from
the question of historical accuracy in their reviews of this pic too—
most of them suggesting that what Penn had tried to create was a "folk
ballad." (Warners claims it's never said anything different.)
While a number of the Bonnie reviews (including Variety's own)
thought Penn had been unsuccessful in welding comedy to tragedy, all
without exception treated the concept respectfully, and all seemed to
understand what was intended when, in Crowther's words, Penn went
"out of his way to splash the comedy holdups with smears of the vivid
blood."
Some, like Cue's William Wolf, called Bonnie "a major artistic
accomplishment, but even those who turned thumbs down, like New
York Post's Archer Winston, said it had "qualities that can be praised
extravagantly." And none duplicated Crowther's accusations that it had
been "assembled in a helter-skelter fashion"; that its characters were
"ridiculous, camp-tinctured travesties" or that it is "strangely antique,
sentimental claptrap."
Penn’s View
If one accepts the view that Penn had no intention of being either
"antique" or "sentimental" and that both he and Aldrich were trying
61
Crbwther’s ‘Bonnie’-Brook
to make statements against violence, rather than in favor of it, how was
Crowther able to view the films as he did?
One answer may be that he is not only concerned about, but in
a sense shares the reactions of, that portion of the public which may be
besieging the box office in search of kicks rather than illumination, and
is more concerned with what it sees than what is implied. ("But will
the picture do well because of its more probing aspects or because of
the vivid violence with which it is filled?" asks the Saturday Review's
Hollis Alpert about Bonnie and Clyde.)
If this is so, then Crowther is in a better position than most to
express concern about the effect of violent films on the public.
But what some in the trade are worried about is that these con¬
cerns over violence might spark a return to the genuinely "antique"
days of filmmaking, when every "commercial" picture had to make an
explicit statement of its point, rather than relying on the ability of its
audience to see beneath the surface. And these folks have accused
Crowther of making no effort to comprehend what the new breed of
filmmakers is attempting—not only where violence is involved, but in
other areas as well.
Offered as a recent example of what these people are talking
about is Joseph Losey's Accident, which Crowther called "just a sad lit¬
tle story of a wistful don," and virtually everybody else referred to as
"disturbing," "thought-provoking," or "a glacial dissection of human
passion."
"We don't mind if Mr. Crowther doesn't like our picture," said
one distributor this week. "What we're concerned about is that the
Times readers may come away from his reviews with no idea of what
it's supposed to be about."
Statement by Jack Valenti,
MPAA President, before
The National Commission
on the Causes and
Prevention of Violence
As you are aware and as we are, there are many disagreements today on
the subject of violence. People say there is too much violence in the
society and more than there used to be; there is disagreement about
whether there is more violence in movies today than there was.
And there is disagreement among laymen as well as social sci¬
entists about the effect that violence in the media has, particularly on
children.
The problem of violence was one of the principal matters that
occupied my attention when I first became president of this Associa¬
tion. It is only one, I might say, because it's only one part of the human
condition.
But I recognized immediately, from May 1966 when I became
president of the MPAA, that for the filmmaker the treatment of vio¬
lence in scenes and incidents of a story that he is trying to tell really
involves the whole fundamental issue of the responsibility of the artist,
the creative artist, not only toward his art but toward the society in
which he lives.
And I think that as you talk to filmmakers you know that this
issue confronts them almost in everything that they do. The question
that confronts the artist is: How much is too much?
I have said on many occasions that what is important is for the
Statement issued by Jack Valenti on December 19, 1968.
62
63
The National Commission on the Causes and Prevention oj Violence
creative man to be honest in his portrayals, to tell the story as he thinks

it ought to be told. But the question is always: When does the balance
tip from violence which is honest to portrayals which are excessive and
overweighted with violence? In short, the whole question is: Where
does one draw the line?
And I might interject here to the members of this honorable
Commission that the next questipmis: Is there a man or an assembly
or a group that is so divinely inspired that they can make those kinds
of final judgments for others?
Almost everything I have said as president of the Association
has been based on the theme that the screen must be free if it is going
to flourish. There is no way to have a flourishing creativity in this land
if you are going to put fetters on the creative man. But I have also said
that this freedom must be responsible—must be responsible—lest lib¬
erty becomes license. I have said that to creative people countless times
as I have tried to establish a rapport between what I am trying to do and
the creative man in both Hollywood and New York and all over this
country. Because this theme poses—and I think you must understand as
I do—the artist and the ethical and the moral distinction between what
a creative man must have for his art and what he must demand of him¬
self. It is a mingling of inspiration and imagination and discipline.
That sounds a little esoteric and far-fetched, but it's really true
to the responsible man of integrity who is creating motion pictures.
Because even for this man, the conscientious man, this gray line—and
that's all I can call it—the line between what is enough and what is too
much—is so extraordinarily difficult to measure. It is so shadowy and
dimly fit that it's very difficult to measure.
So that the essential point becomes not the inclusion of vio¬
lence or the quantity of it or the nature of it but really how it is treated,
how it is handled.
I don't have to tell you—I think it's almost a cliche to say—that
throughout the whole history of drama, violence is a common ingredi¬
ent. That goes without saying. The very nature of drama is conflict.
New plays and old plays, ancient chants, litanies, the epic poems, and
traditional literature of practically every country and civilization that
you can name are rooted in violencebecause man's whole existence has
been a story of conflict.
And I might add that we know, even to this very hour, that all
civilizations have been alternately horrified and fascinated by death
and violence. I didn't make it that way. That's the way it is.
64
Jack Valenti
So it's my judgment that violence should not be presented as a

way of life—not at all—but for what it truly is, one of the facets in the
complex fabric of the human condition.
Now, let me go to my second point. It is: What is the motion pic¬
ture industry doing to fulfill its obligation of responsibility to the soci¬
ety in which it lives? I'm going to trace this for you very briefly. Since
1927 we have had codes, guides for producers, voluntary guides. And in
1930 we adopted the so-called Production Code, now very famous,
which was a self-regulatory rubric through which producers, directors,
and writers in a voluntary way tried to regulate themselves. This Code
has been updated from time to time to change with the mores and the
customs of the society, because all mores and Gustoms change. And in
1966 we reaffirmed and we strengthened the Production Code.
The Production Code is operated separately but in tandem with
an Advertising Code which does the same thing.
We try to avoid what we call a cumulative overemphasis on sex
and violence, which are the two great facets of the human condition.
We recognize that sometimes incidents standing alone are quite
permissible, but once they are allowed to accumulate they become
almost intolerable. We understand this. And both in our Advertising
Code and in our Production Code we are constantly trying to deal with
this as we have been dealing with it for over thirty years.
There are objections; some say: "Well, that's very fine, Mr.
Valenti, but you're dealing with it in general terms, and you're depend¬
ent on the subjective views of those people who are managing this Code."
The answer is that's very true, because the very nature of what
we are doing is subjective. And I don't believe that you can base a deci¬
sion as to whether a particular portrayal of violence is detailed or pro¬
tracted or excessive on the number of killings or the number of blows
or how many grams of blood were spilled.
The very nature of the problem makes it absolutely imperative
that you deal with it, not numerically or quantitatively but subjectively
or qualitatively. And these are the kinds of decisions, frail decisions,
human decisions, that are made under this voluntary Code by people
who are vastly experienced, not only in the appraisal of motion pictures
but also in this very tenuous and sensitive relationship between the cre¬
ative man and his monitor. It is a very difficult relationship.
The people on the Production Code operation are literate peo¬
ple, skilled people. While I would be the first to say that their judgments
are no better or no worse than anyone else's, they are rooted in a better
kind of experience.
65
The National Commissibn on the Causes and Prevention of Violence
Now, in addition to the Code, we have taken several other steps

in dealing with the portrayal of violence. This spring, immediately fol¬
lowing the tragic murder of Senator Kennedy, Louis Nizer, general
counsel of this Association, and I traveled to Hollywood and called a
special meeting with all heads of studios, with directors, writers, and
actors, with producers. We urged upon them increased restraint and
heightened responsibility in portraying violence.
The response was very heartening. Later on, more than 350 pro¬
ducers, writers, directors, and actors signed an open pledge that they would
forgo scripts which had anything to do with aimless cruelty and senseless
brutality. It is a voluntary act, of course, but I think it does testify to the
accountability of creative people about their own responsibilities.
Now let me say a word about audiences. It has always been a
great cliche in the motion picture business, that there was a single com¬
mon denominator, a single audience, and that films were made for the
14-year-old level. You have heard that before. If this ever was true, it
certainly isn't today.
The popular media, I don't have to tell you, produce a veritable
tidal wave of products that almost drowns this country. Motion pictures
do not appeal to a single audience. There is no mass audience today. The
mass audience, in my judgment, just doesn't exist, if it ever did.
Today, we must understand the following: films explore more
deeply into the human condition than they ever did before. A substan¬
tial number of films coming into this country are foreign in origin.
There is a new breed of filmmaker. And mark you well this new film¬
maker, because he's an extraordinary fellow. He's young. He's sensitive.
He's dedicated. He's reaching out for new dimensions of expression.
And he is not bound—not bound—by the conventions of a conformist
past. I happen to think that's good. Moreover, this new style in film-
making is matched by a new audience. It is seeking new fulfillment. Its
members are better educated. . . .
Conscious of these findings and of the fact that the kind of soci¬
ety we live in today is different from the kind of society we used to live
in, for the first time in the history of the motion picture industry we
have developed a plan of rating films for audience suitability. It is a vol¬
untary film rating system developed with the active assistance of the¬
ater owners and creative people and distributors and producers.
Its dominant, preeminent, overriding concern is for children.
This is a rating system for parents and families. Films are rated not on
their excellence or lack of it, not on their excitement or lack of it, but
whether or not the content of the film is suitable for children. . . .
66
Jack Valenti
After November 1, all films released to the public will carry a

rating. The first rating is "G"—suggested for general audiences. That
rating means a parent may send his child in to that picture. There is no
objectionable material in the film. However, this doesn't necessarily
mean it is a children's film, because some of the most powerful and pro¬
foundly significant films of this generation would be G-films. One of
the classic examples of such a film that, would surely have been a "G,"
had it been rated is Man for All Seasons, the great story of Sir Thomas
More and the irreconcilable conflict of conscience between Sir Thomas
More and his king.
The second rating is "M," suggested for mature audiences,
mature young people, with parental discretion advised. What we're say¬
ing to a mother or father here is: "Look, don't take your child in to see
this picture until you know more about it. For it may be—just may be—
unsuitable for your child. There are no restrictions, but we want you to
know more about it."
The third category is "R"—meaning restricted. Here for the
first time there are restrictions on the audience. Children under 16,
unless accompanied by a parent or an adult guardian, are barred from
such pictures. These are adult films. However, there may be some adult
films that a parent would want his child to see. The parent may want
to go with his child so they can discuss it afterward together.
I can name a number of films of this nature that carry a mes¬
sage for young people but the parent ought to be there with him.
Figure 14. In 1968, the Motion Picture Association of America instituted its G-M-R-X ratings for
film content. To satisfy the MPAA and earn an R rather than an X rating, producer Phil Feldman
and editor Lou Lombardo had to abbreviate the throat-cutting scene in The Wild Bunch, delet¬
ing a side-angle view of the blood spurt.
67
The National Commission on the Causes and Prevention of Violence
The final category is "X," in which we say that this picture

should not be shown to any child under 16, regardless of who accom¬
panies him.
I think the filmmaker has to remember that discipline and
restraint are part of the definition of true artistry and that therefore he
must practice restraint.
It is my judgment that responsible filmmakers in this country,
with whom I have been in contact, intend to do just this, but it would
be dishonest to tell you that there won't be fringe operators on the
periphery who are going to try to make a buck out of this thing. Of
course there will be. But that's true in all professions and all enterprises,
and even in families.
The third question is, How well will these ratings be enforced
at the box office? I place my faith in the vast majority of responsible
theater owners in this country. I place my faith and my hope in them,
because I believe they will do it. They have told me that they will do it
and I believe they will do it.
I have personally talked to more than seventy-five of the own¬
ers of the leading exhibitor chains in this country, probably represent¬
ing four thousand to five thousand theaters, probably representing 80
to 85 percent of box office. These leaders of this industry have all told
me that they will support and implement this program.
And, finally, how cooperative are the parents of this land? If the
parents abandon their responsibility for the conduct of their child, it is
very difficult for the motion picture industry to make up for that
parental lack of responsibility.
Now for my third major point: Is the depiction of violence on
the screen probably harmful to people, to children particularly? I'm not
an expert. But what I have read tells me that the evidence is not con¬
clusive. As we examine the writings of the past forty years of social sci¬
entists and others who are experienced observers, the best that can be
said is that the opinions are ambivalent in intent. . . and contradictory
and the differences among social scientists reach imposing levels. I am
sure that you ladies and gentlemen, as you have examined the litera¬
ture know that I'm not speaking in hyperbole. If there is one conclu¬
sion that appears to be warranted, it is simply this: most authorities are
reluctant to conclude that the portrayal of violence in motion pictures
results in harmful social behavior. That's one conclusion that I think is
warranted.
68
Jack Valenti
Now, one of the things that makes it clear why experimenta¬

tion in this area is so hard to design, hard to construct, is that it is
morally unacceptable to induce delinquency experimentally in a child.
It's wrong to do it. And it's the most serious barrier to experimentation.
There are additional reservations that I have about research in
this area. I will list them very quickly.
First, fears that motion pictures may set off real-life acts of
aggression, many times are based on very little solid evidence. Most of
the time it's case histories of maladjusted people who are under treat¬
ment and this is not a valid kind of conclusion.
Second, alleged acts of aggression that happen in laboratory
experiments are said to be brought on by what scientists call artificially
induced preconditions. Therefore one begins to doubt the relationship
to a real-life reaction in a live theater.
Number 3, very little is known of the effects in long-range
behavior. In my personal judgment that is one of the key weaknesses in
that whole scheme of social research.
And fourth, most clinical opinions are too heavily dependent
on the deviant, the disturbed, the already mentally disfigured child.
You will find that the literature is filled with statements that
well-adjusted children in a well-adjusted home life can't be harmed by
anything shown on a motion picture screen.
The overwhelming evidence shows that the root causes of
behavior are developed in the early years of the child and are primarily
environmental, physical, and psychological, arising out of home and
family life. This is a truth I'd stake my being on because I believe it.
Finally, you have a right to ask, "Well, all that is very fine, but
what are your plans to be alert to the newest developments in the search
for new social knowledge?"
First, I have been in consultation with an eminent social sci¬
entist on the West Coast, and I am concluding a similar arrangement
with one on the East Coast, so that I can keep abreast of the latest devel¬
opments in this rather fuzzily defined field.
Second, I have recently appointed to the Code and Rating
Administration a woman who brings with her a very strong background
in child psychology, family relations, and the behavioral sciences.
Basically, our approach to the problem is not to wait for scien¬
tific demonstration that some lurid depiction of violence is harmful to
children. It may very well be, ladies and gentlemen, that this will never
be proven. But my common sense tells me that the depiction of extreme
69
The National Commission on the Causes and Prevention oj Violence
violence or anything in the extreme is simply offensive to normal sen¬

sibilities. I don't need a scientist to tell me that. Whether or not it causes
juvenile or adult delinquency, it's just offensive, and I am against it.
Therefore, if I had to pick a watchword, it would be "modera¬
tion." Our whole Code, all of our voluntary programs are based on that
principle. "Nothing in excess," said the sign over the Delphic Oracle,
and I think that ancient maxim is a good one here—moderation. . . .
Mr. Tone: Will you agree with me that when the members of
MPAA adopted the 1966 Code they subscribed to this view: that motion
pictures do have an effect on the moral standards and conduct of those
who watch them, especially juveniles? Is that a fair statement?
Mr. Valenti: I'm not aware of such a statement.
Mr. Tone: Well, let's see whether I inferred too much. The 1966
Code does say under the heading "Particular Applications": "Crime
shall never be presented in such a way as to throw sympathy with the
crime as against law and justice or to inspire others with a desire for
imitation."
It says: "Methods of crime shall not be explicitly presented or
detailed in a manner calculated to glamorize crime or inspire imitation."
It says, under the heading "Reasons Supporting the Code":
"The moral importance of entertainment is something which has been
universally recognized. It enters intimately into the lives of men and
women and affects them closely. It occupies their minds and affections
during leisure hours and ultimately touches the whole of their lives. A
man may be judged by his standard of entertainment as easily as by his
standard of work. So correct entertainment raises the whole standard
of the nation,- wrong entertainment lowers the whole living conditions
and moral ideals of the race."
Then it says (and this is the last passage I will read): "Hence,
the important objective must be to avoid the hardening of the arteries,
especially of those who are young and impressionable, to the thought
and fact of crime. People can become accustomed even to murder, cru¬
elty, brutality and repellant crimes if these are too frequently repeated."
Would you not say that the author of those statements believed
that motion pictures could influence the conduct of the people who
watch the pictures?
Mr. Valenti: Yes, I do. As a matter of fact, we have gone beyond
that 1966 Code, way beyond it. Because we have now instituted some¬
thing that the 1966 Code didn't have. That philosophy was related to a
public feeling that anyone could go to any picture in this country. But
70
Jack Valenti
today, for the first time, we are excluding children. We are excluding
children from pictures we think are unsuitable for them. So I would say
we have gone beyond the concept set by that statement.
Mr. Tone: I'm at the moment speaking only of the philosophy
of the 1966 Code. And I recall your statement at the time the Code was
announced in October, which I think is similar to your statement
today: "There is no valid evidence at this time that proves movies have
anything to do with antisocial behavior." Would you not say that that
statement represents some change in viewpoint from the statement of
principles in the 1966 Code? And if it does, I'm interested in how the
motion picture industry's thinking has changed on that subject. What
have the—
Mr. Valenti: Well, the principal change has been that the Asso¬
ciation has a new president. Administrations may change and points of
view—if indeed this is a change in point of view. That is a judgment.
I bear no responsibility for the 1966 Code, any more than the
incoming president who bears responsibility for our nation would bear
total responsibility for carrying forward something that his predecessor
did—as long as it doesn't violate a principle. I truly believe that I have
made the philosophy stronger by bringing it into an active program. To
me this is far more important. It is well to say words, but I believe that
to translate these words into an active program shows progress. That's
why we have taken a philosophy and hardened it into an active program
that keeps children under 16 out of certain movies. That in my judg¬
ment, Mr. Tone, is an advance. That's keeping in step with the chang¬
ing mores and customs of a society and taking that which was first
rooted in philosophy and constructing a living, breathing program
through our new film rating system.
Mr. Tone: One more question before we get to the rating system
specifically. You stated the children who are not disturbed, normal chil¬
dren who come from normal homes, are not likely to be affected by any¬
thing they see. Do you believe that motion picture producers have an
obligation with respect to disturbed children or children who because
of some problem or other could be affected by the amount of violence
they see? Or do you believe that the advantages of creative freedom out¬
weigh any obligation to such a small minority? What are your views?
Mr. Valenti: Mr. Counsel, my answer to that is quite obviously
we can't make pictures at the level of the disturbed child. As Mr. Jus¬
tice Marshall said in his 1968 decision, that would turn movies into a
wasteland. We would make what he called inane movies. And I surely
would agree with that. There is no rational person in the motion pic-
71
ture industry who would even suggest that we should make pictures
aimed at the level of the disturbed child. You simply can't do it. Any
more than you would write all books or portray all of life itself at this
level. . . .
There are some people in America who have no patience or tol¬
erance for the workings of a democratic society, and they would have
the government intrude by law to control motion pictures in the belief
that this is the way to control violence.
As a humble citizen of this land, I don't think there is any delu¬
sion so slippery or any act more perilous than the intrusion of the gov¬
ernment into making such judgments for the communications media.
Of course, I would hope that all thoughtful citizens would be opposed
to that alternative.
The question that needs to be asked is: Can censorship cure the
portrayal of violence in the media? That's the question everybody must
ask. Some people would answer yes. I would have a larger question. I
would ask: Can censorship curb violence in the society? I think it's a
truism that movies are not beacons but rather mirrors of society. They
don't lead; they follow the already established course.
That leads to another question and that is: How much is too
much violence, Senator? How do you determine that?
And, then, the really tough question to be asked is: Who would
make these judgments? Who would appoint them or who would anoint
them? And by what omniscient or divine authority would they claim
accurate judgment?
This is the crux. The more I get involved in even voluntary rat¬
ing of pictures, the more convinced I am of the lunacy and the absurd¬
ity of governmental involvement in making such cultural judgments.
I promise you, ladies and gentlemen, all you have to do is read
two leading critics and you will find one saying, "This is wonderful,"
another saying "It's pornography," and one saying, "This is the great
moral play of our time," another saying, "It's cheap violence."
Who's right? This is the thing that causes me the greatest con¬
cern—knowing we are frail humans and yet we are making these diffi¬
cult, even impossible, judgments. We do the best that we can. What's
important is that it's voluntary.
Senator, I can only state this as my answer to your question
regarding alternatives, which obviously must concern anybody looking
at the problem: what we are doing now is not quite to our liking. What
else can we do? It is a tormented question that is both attractive and
repulsive to different people.
72
Jack Valenti
Figure 15. The graphic violence of Bonnie and Clyde was tied to the social and political turmoil
of the 1960s. Director Arthur Penn, for example, included a visual reference to the death of Presi¬
dent Kennedy. When the Texas Rangers gun down Clyde, a piece of his head (like JFK’s) flies off.
I find sanction by law in this field odious beyond measure

because, as I say, just trying to rate films voluntarily emphasizes how
impossible it is to do.
Judge McFarland: Thank you very kindly. . . . Now, I'll have to
admit my ignorance in not having seen this picture, Bonnie and Clyde,
but I am told that it is replete with violence. How do you justify the
amount of violence, if my information is correct, in that picture?
Mr. Valenti: Well, Senator, you are one of the few people in this
country who seem not to have seen that picture, because it has gained
wide audience.
Judge McFarland: A lot of them have.
Mr. Valenti: However, you bring out what I think is a legitimate
question, and I will speak to that briefly. This picture, more than any
other, I think, illuminates the great dichotomy of opinion that so bedev¬
ils this whble subject of rating films.
For example, I was being confirmed by the Senate in another job
on the Corporation for Public Broadcasting, and a distinguished sena¬
tor of the United States Senate took me to task on this picture. I will
give you the same answer that I gave him, because I think it is germane,
and it is simply this:
73
The National Commissioti on the Causes and Prevention of Violence
A number of people did think Bonnie and Clyde was a picture

of extreme violence with a tendency to cause people to think kindly of
bandits and robbers and hoodlums. And, as Congressman Boggs knows,
I came from that part of the world, and as a young boy, I knew about
Bonnie and Clyde, and I must say my great hero was not Bonnie or
Clyde, but Frank Hamer, who doesn't come out too well in the picture.
But, on the other hand, may Tpoint out something to you that
you may not know? This picture, so disfigured by a number of critics,
was chosen by the National Catholic Office for Motion Pictures as the
best mature picture of 1967, they called it a "great morality play." I have
said to some of my critics when they talk about Bonnie and Clyde:
"Well, you've got to determine who you're going to follow, those peo¬
ple who criticize Bonnie and Clyde, or the Catholics who are probably
the most indefatigable monitors of the motion picture screen and
whose integrity is almost impeccable."
There is a good example of a great and prestigious group in
America, beyond personal gain, that says this is a great motion picture.
And there are others who say it's extremely violent. And I must also
add that the Catholic Church is also a great critic of senseless violence
on the screen, more than almost any other group that I know. I go along
with the National Catholic Office of Motion Pictures. I think I would
follow their judgment in this particular case.
Judge McFarland: Well, thank you very kindly. I must not take
up more time. I know my colleagues have questions. But again I want
to express my appreciation and say to you it's nice to see you and visit
you even across the table.
Mr. Valenti: Thank you, Senator.
Judge McFarland: Although you have just been in your present
position two and a half years, you certainly have the understanding of
the problems confronting the motion picture industry as though you
have been there for a much longer period.
Mr. Valenti: Thank you, Senator.
Judge Higginbotham: I would hate to think, Mr. Valenti, what
a survey would show if we took a survey on this Commission as to how
many have seen Bonnie and Clyde, [laughter.)
Congressman Boggs: Mr. Chairman, I did see Bonnie and
Clyde, and the distinguished Justice anticipated some of my questions.
I might say that I am a great admirer of the witness. I have
known him for many years and have been very closely associated with
him. I am very happy to have him in his job. I think he does a very dif¬
ficult and trying job.
74
Jack Valenti
But I was interested in the responses about Bonnie and Clyde.

I might tell him that we had a murder in my town committed by an
18-year-old boy who had come out of Bonnie and Clyde one hour
before. He killed a young man who was running a drive-in grocery store.
And it was just a senseless murder. Now, whether or not what he saw
in Bonnie and Clyde had any impact on the murder, I don't know. But
I know that what I say to you is a fact—that he saw this movie which
glorifies violence.-
Those Bonnie and Clyde characters lived in my State. They
were reprehensible criminals. There was nothing about them that was
commendable. They killed in cold blood, as the movie depicts.... I can
assure you that I share your fear of censorship. But oftentimes what
brings repressive measures is abuse. . . .
And I read the Code here on page 4 of your statement where you
set out five basic standards of the eleven. You say:
"The basic dignity and value of human life shall be respected
and upheld." That certainly isn't true in that movie.
"Restraint shall be exercised in portraying the taking of life."
God knows it isn't true in that one.
"Evil, sin, crime and wrongdoing shall not be justified." Well, I
don't know what that means.
"Special restraint shall be exercised in portraying criminal or
anti-social activities in which minors participate or are involved."
"Detailed and protracted acts of brutality, cruelty, physical vio¬
lence, torture and abuse shall not be presented." That's the essence of
that movie.
Now I presume that your whole emphasis is on self-policing.
What happens when the self-policing doesn't work? You used the word
"responsible operators." I understand that word perfectly. The average
responsible citizen is not a criminal. We don't pass laws to deal with
him. We finally get into the business of regulation and law passing and
law enforcement because of the irresponsible. So what do you do about
the irresponsible in your voluntary code? . . .
Valenti: There is no way to deal with the problem directly
unless one chooses to make pictures aimed only for the disturbed
youngster,«aimed for the lowest-common-denominator audience. I
don't have to tell this distinguished group what would happen to the
level of art in the community, how all art, including movies, would
cease to flourish. Art would become totally stagnant, and it would soon
disappear.
75
The attempt we have made . . . with our rating system is, at

least, to try to tell families what is in that picture vis-a-vis their
children, so they don't wander into a theater without any knowledge
whatsoever.
I think you are in a very difficult position when you try to make
films that would exclude adults.
Now, the question was brought up: What do you do about a film
that triggers a disturbed youngster? Well, anything might trigger such
a person. You have to isolate him from life, I suppose. He may see some¬
thing happen on a street corner. He may read something in a book.
Somebody may speak harshly to him and he is triggered. So I find that
not at all an argument for not making pictures that might trigger him.
To answer your question, I don't know of any way in a demo¬
cratic society in which you can segregate pictures. I think the very fact
you do have controversy about motion pictures is an indication of the
interest, the hidden interest, that we find in it now.
'■
>
.
The Aesthetics of
Ultraviolence
John Bailey, ASC
Bang Bang Bang Bang,

Ad Nauseum
In September, over the three-day Labor Day weekend, I made a serious

effort to catch up on the major summer films I had missed. I saw six
films along with about two dozen trailers for the autumn releases. Not
a single one of the trailers (not to mention the features themselves) was
devoid of considerable firepower. I'm not speaking of just action
excitement, but of a veritable litany of handgun and automatic weapons
discharges, incendiary effects, stabbings, and throat slittings. There
were also, of course, a few garrotings and numerous beatings of women.
This is studio entertainment, after all.
But before your P.C. antennae start bobbing around, let me say
this. I don't believe in censorship. It does violate basic freedoms, and it
doesn't work. I don't support any alliterative watchdog group such as
the Christian Coalition. I do support my own conscience.
We have all been deluged recently with magazine cover stories
documenting, as individual and national tragedies, the 'senseless' mur¬
ders on our streets and in our homes. We've digested heart-rending pro¬
files of preteen murderers who could barely entertain the concept of
mortality before they were swallowed up in the blitz of news that is atten¬
dant on an act as effortless as squeezing the trigger of a cheap handgun.
We're tired, angry, and frustrated by our seeming inability to
have any influence over the ever-escalating statistics of real-life mur¬
der, rather than the artifice of movie mayhem that is so routine and
unreal to us as filmmakers.
We've had plenty of finger-pointing and buck passing from Con¬
gress, the media, and the giant conglomerates that pass for studios and
From American Cinematographer, vol. 75, no. 12 (December 1994), pp. 26, 28-29.
79
80
John Bailey
which crank out our "entertainment." Nobody's to blame. We're all at

the video store.
I have been wanting to put down my thoughts on screen vio¬
lence for some time, but I'm finding the path into the subject to be
somewhat daunting, like Daedalus standing before his own creation,
the Labyrinth. As a cinematographer I have tried to be responsible, or
at least conscious, about the amount of, and approach to, violence in
the films I have photographed. I think it became a crucial issue for me
in the mid-1970s while I was still a camera operator.
I had accepted an offer to do an MOW with a cameraman who
had been a very real mentor to me, a man whom I had assisted for some
years. Accepting the job on short notice, I wasn't able to read the script
in advance. I went into the production stone-cold.
At the start of the first day's shooting, machine gun-toting 'ter¬
rorists' charged the camera, herding a group of scantily clad beauty con¬
testants into an abandoned Quonset hut, where they were soon to be
held hostage. These very bad guys had just hijacked a beauty pageant. I
swear, this really was the story line. Is it high-concept enough for you?
Between grunts and vague threats of unmentionable sexual acts, these
goons fired their guns into the tin ceiling. It was loud. It was dumb. I
was ashamed.
Two days later I told the cinematographer that I couldn't con¬
tinue the show (and I really did need the work). He felt I had left him in
the lurch and promised that he would never work with me again. And he
didn't. But, painful as it was, I think I crossed a personal Rubicon of self-
respect that day and addressed a sense of my own part in the use and mis¬
use of filmed violence. Even within the cartoon cliches of this forgettable
film, an embryonic reference emerged that serves me to this day.
An ironic footnote to this tale is that a short time later [cine¬
matographer] Nestor Almendros phoned me. I had met him in Paris in
the late 1960s when, fresh out of film school, I made a pilgrimage to
France to meet my idols, the cinematographers of the Nouvelle Vague.
Nestor had called to ask me to be his camera operator on the upcoming
Days of Heaven. Thus I ascended from the classic slough of despond to
a friendship with one of the gentlest humanists who ever looked
through a lens.
I don't mean this to be a cautionary tale about the efficacy of
doing "the right thing." But I guess I do believe in some kind of just
reward. I do believe there is strong correlation between our work, the
kind of films that we choose to work on, those we refuse to do, and the
people we become, certainly over the long haul.
81
Bang Bang Bang Bang, Ad Nauseum
It may seem I've diverged from the nominal topic of this essay,
but there is for me an emotional "throughline" that remains intact.
And that is pretty simple. As filmmakers, like any artists, we inhabit
our work. Not only do we take it home at night, but often, especially
when we are on location, our location is our home. We form and re-form
ad hoc families and residences that are as real to us as our own spouses
and children. t
Because we live and work in such a violent society, it is natu¬
ral that our films reflect and explore this violence. But often we only
explore it deep enough to wallow in its muck. This is where the ques¬
tion of a demarcation arises. What is exploration and what is exploita¬
tion? And who decides?
I've always had difficulty with this question. It gets to the heart
of the issue of censorship. Because one's intent and point of view are
key factors, it is no easy subject to investigate. But recently I saw two
films that so crystallized the matter that everything started to fall into
place, and I found for myself a clear distillate of the dilemma.
I hope you have seen both Natural Born Killers and The Shaw-
shank Redemption. They are magnificently made films, brilliantly con¬
ceived and photographed. They cast an unflinching look at the strains
of violence in Americans and in our films. At times both films make you
want to avert your eyes. Otherwise, they could not be more different.
The philosophical stand these two films take on the issue of
violence may be fundamentally different. But just as telling and more
germane to this essay is the photography of that violence.
Natural Born Killers is a manic, all-stops-out journey through
the madness and darkness of the American obsession with violence as
acte gratuit, an existential fury that defines all that is alien, angry, and
antilife in our character. Its style is hyperkinetic and frenzied, mixing
media and formats in a non-logical way that is said to have been deter¬
mined at times by the flip of a coin. Its images seem to spin out of con¬
trol, held in place only by an incantatory sound track, pivoting around
the threnodic voice of Leonard Cohen.
The Shawshank Redemption is visually very measured and
controlled, often almost processional. Its character and narrative skeins
are nuanced and defined by circumspection, a tone set by the reflective
timbre of Morgan Freeman's elegiac voice-over narration. The plot
points are neatly resolved, although the ending may seem too attenu¬
ated to some. The depiction and eruption of violence is unflinching but
distanced. None of this makes the film remote. In fact, it draws you in
it and by the end the cumulative power of empathy is heroic.
82
John Bailey
The Shawshank Redemption observes moments of awful vio¬

lence enacted upon the character played by Tim Robbins. And violence
or the threat of it permeates many of the relationships in the film. But
the point of view of the violence is external. It is an ugly fact of the char¬
acters' lives and its force is elemental and Darwinian. It is not heroic;
it is not insightful; it is not stylized, slo-moed, or eroticized. The cam¬
era records, documents, then retreats and plays out the action in long-
shot and in the shadows. The point of view is that of empathy for the
victim. In short, it is human. And this is crucial to the transforming
majesty of the film.
There is no doubt that the team of Oliver Stone and Robert
Richardson have redefined many of the parameters of mainstream the¬
atrical image-making. The raw power and immediacy of their films has
been one of the defining markers of American cinema in the past
decade. It has come at a price. The dragon has finally swallowed its own
tail. Technique has become raison d'etre. And therein lies the caution¬
ary tale.
Natural Born Killers places you inside the violence, makes you
a part of it. Its point of view is that of the killers, not because of a clear
Figure 16. Woody Harrelson as a multiple murderer in Natural Born Killers. The low-angle cam¬
era and the character’s dramatically extended arm manifest the film’s fascination with the lurid
violence it purports to satirize and critique. It is very difficult for filmmakers to resist the attrac¬
tions of aestheticizing violence.
83
critical or moral perspective but because that's where the action is. The
defining aesthetic is MTV. And if it is meant to be otherwise, if it is
meant to be a de facto critique or satire of the American and media
obsession with violence, I can only say that on this level it is, for me at
least, a complete failure. The filmmaking tools which are wielded so
artfully and with such panache distort the putative intent. The film
eroticizes violence, wallows in it, and struggles to incite the viewer.
If you are an impressionable child, an angry adolescent, or an
alienated sociopath, you will probably leave Natural Born Killers feel¬
ing jacked up. But if you are a rational adult you will feel drained, sullen,
enervated. You leave The Shawshank Redemption feeling renewed,
buoyant, sanguine. Both films spend a good part of their screen time
inside the confines of a prison. Your experience of that space could not
be more polarized by the two films.
The space and light of Natural Born Killers is that of a Sadean
stage set replete with Grand Guignol effects, the whole grotesquerie
awash in blazingly assaultive light and aggressive camera motion. The
space and light of Shawshank Redemption is, like its narrative, almost
classical. Roger Deakins and Frank Darabont brought close scrutiny
and a sense of the importance of detail to their work. You have to watch
this film closely. You have to live in its images. They don't beat you
over the head. And because your head is left intact, you experience the
film, not just watch it.
The restraint of Bresson's camera in A Man Escaped has no
inherent moral force superior to the first-reel frenzy of Truffaut's Jules
and Jim. Nor is The Shawshank Redemption with its formal pictori-
alism superior to the visual disjunction of Natural Born Killers. My
point is not to negate one style of filmmaking in favor of another.
Nor do I wish to anathematize in toto the efficacy of violence
in the arts. From the Greek tragedies, Jacobean drama, French theater
of the Revolution, German Sturm und Drang, to Antonin Artaud and
Peter Brook, the use of violence in theater has been the vehicle for prob¬
ing the complexities and dysfunctions of the human condition and the
means to catharsis and growth. Violence is as endemic to the human
soul as is love. Both emotions are interwoven with the fear of and
inevitability of death. Experiencing, violence and death in our art is a
very real way of affirming our life.
But we must know that there are dangerous shoals when we set
out to conceptualize and visualize violence, especially in film. A cam¬
era may be a mechanical recording device but the eye behind it is artful
84
John Bailey
and intentional—always. Storyboards are not comic strips. They are blue¬
prints for complex and powerful images. If "actions have consequences,"
so do images. We create images. We are conjurers. And, like it or not, we
are teachers. Our images are more haunting and more influential than any
of the words that will ever be written about them. If you can deny that
fact, you can deny the movie memories of your own childhood.
Several years ago during pre-production of China Moon I had to
deal with these issues as a director. There is a murder in the film: Made¬
line Stowe fires a handgun at close range into the head of her husband,
Charles Dance. So, in the interest of responsible research, I found
myself one afternoon on the police firing range in Lakeland, Florida,
9 mm semiautomatic pistol at the ready. I had never fired a handgun. I
felt I needed to experience the sensation. The emotional shock of the
first discharge far exceeded its ballistic "kick." The sense of power was
instant and terrifying. I understood. That same week, also for research,
I subjected myself to viewing the full autopsy of a local murder victim.
I had never seen a human body rendered into parts. The sadness of this
needless death and of my own empathic mortality haunted me for days.
I decided to film in a realistic way the scripted autopsy of the
woman found murdered in her kitchen at the beginning of the film. I
also filmed in detail Ed Harris's forensic investigation of the murder vic¬
tim, with emphasis on the "bonding" between detective and deceased.
Joe Laude, a homicide detective who served as our tech adviser, said
that even in death the murder victim often "speaks" to him. And I pho¬
tographed Charles Dance's murder as simply and as directly as I could.
The second gunshot, the one to his head, was rigged with an elaborate
prosthetic to cause a stream of blood to pulse out under pressure, just
as I'd seen in documentary footage. It was graphic and shocking.
And it is not in the finished film. The close scrutiny of the
autopsy, the even closer, humanizing look of the murder victim as Har¬
ris stares into her open eyes, and the shot in the head of Charles Dance
all made the preview audience uncomfortable and was reported to have
pulled down the survey scores. Some of the offending material quickly
went onto the editing room floor,- more fell victim to strategic com¬
promises. At first, I felt completely co-opted, but then my own anxiety
made me v,ery tentative and uncertain about the choices I had made in
being so direct—that is, so "uncinematic"—in the depiction of murder
and its aftermath.
But the ensuing time has convinced me that it was correct to
deglamorize the violence and to focus on its human consequences. In
85
the end, my own complicity in agreeing to "tone down" the darkness

of this film noir compromised the moral perspective I had intended. It
had reduced the spiritual agony of Ed Harris, a once moral man now
become accessory to murder, to the level that some in the audience
would see as a man just trying to save his own skin. This was not the
film Roy Carlson, the writer, and I had set out to make. While I remain
proud of China Moon for many reasons, it is my own cautionary tale.
Directing it, I learned much about the intersection of screen violence
and box office. The final irony was that my effort to deal with violence
and death responsibly proved not to be too violent but too much of a
"downer."
I believe we have a great responsibility to the people who see
the images we create. But we have as great a responsibility to ourselves.
If our own sensibilities are askew, if we have no moral compass to guide
us, what point of view are we going to create? Part of the responsibility
we have to ourselves is to know and respect the trust we have been
given to influence others.
Film is arguably the most influential of all of the arts,- it is the
art form of choice for young people. How we choose to show them those
images of life and death, the emotive wallop they pack, and the imprint
they make on their hearts and minds, are questions that we cannot keep
in the shadows.
David Thomson
Death and Its Details
The seven screenwriters are decent men and women. Four of their chil¬
dren have been in Non-Violence Awareness programs. They are all
devout in the faith that there are too many guns in America. Not to
mention greater Los Angeles. They have written letters to senators urg¬
ing resistance to National Rifle Association pressures. But they all have
a problem with this script.
“What do we do with Arthur?"
“Arthur's a loose end."
“He did love Dolores."
“And he was useful in our second act bridge."
“Arthur was the second act bridge. Now he's spare."
“We kill him?"
“We have to. We leave him around the audience is wondering.
How?"
“Arthur could get ill."
"Illness is a year, it's doctors."
“It's got to be quick. We're over two hours already."
“Suppose Roger shoots him? In a fit of hitherto repressed
anger?"
“Love it. It makes Roger stronger,which is good for Angie."
“And it gets Roger put away, too."
“Ground clearing."
“We do it as a sudden epiphany. Ten-second scene."
“Bang bang, Arthur."
This is the way the world ends

No^ with a bang but a whimper.
—T. S. Eliot, The Hollow Men
From Film Comment, vol. 29, no. 5, Sept.-Oct. 1993. Reprinted by permission.
86
87
Gittes pulls the car door open and Evelyn falls out. Her face is cov¬
ered with blood. She is dead, shot through the back of the head,
coming out through her left eye.
—Robert Towne, Chinatown, the screenplay
The mass of men lead lives of quiet desperation.

—Thoreau, Walden
Forget it, fake. It’s Chinatown!

—Robert Towne, Chinatown
To illustrate the empire of bang-bang, let me list the deaths in

The Godfather, the original and first part:
• Luca Brasi, one hand pinned to the bar counter with a knife, stran¬
gulation inflating his horrified dignity
• Paulie, the treacherous driver, plugged in the back of the head in a
parked car, all in serene longshot, while Clemenza takes an easeful
leak in the wheatfield
• Sollozzo, shot in the forehead so that an odd Indian castemark puts a
dent in his pasta
• Police Chief McCloskey, shot in the throat and the forehead; these
two men are left with their tipped-over table in the small Italian
restaurant like lovers' clothes thrown off in haste—there is some¬
thing orgasmic in this double murder, it's Michael getting laid
• Sonny, at the toll booth, with 612 bullets (you count them)
• Apollonia, eager to drive, blown up in the car, her persimmon breasts
tossed to either end of the garden
• Vito Corleone, while playing in the tomato plants on a hot day with
his grandson
• two Tattaglia brothers shotgunned in an elevator
• Moe Greene, having just put on his glasses, so blood can creep
through a cracked lens
• a man in a revolving door
• a man and a woman astonished in bed, naked, and then quickly dead
• Barzini and Barzini's man, Barzini only after he has run to the top of
some steps so that he may topple down again
• Tessio: we don't see Tessio's death, only his stooped figure being
taken away and Abe Vigoda's dreadful glance
88
David Thomson
• and Carlo, Connie's husband, throttled in a car, his frantic feet kicking
out the windscreen so that we can see he didn't even take care of his
shoes
That's seventeen, to say nothing of a horse's head, a fish (to let

us know where Luca Brasi sleeps), and at least three corpses in a mon¬
tage. Not to mention Kay's hopes.
Over the.years, I have taken it as an axiom that the American
18-year-old has seen twenty thousand acts of killing in movies and on
TV. Most college freshmen with whom I shared this nodded like con¬
noisseurs: few professors had taken their youth so seriously. Yet I am
no longer sure how I ever knew this fact. Having totted up one movie
(17 deaths in 171 minutes), I suspect the statistic is rather conservative.
It would be as hard as counting the bullets in Sonny as it might be to
keep tabs on the deaths in, say, Sergeant York, Kriemhild’s Revenge,
The Wild Bunch, Alan Clarke's Elephant, Spartacus, or Tom and Jerry.
Let us agree that we have witnessed a lot of killings. By my age, 52, with
more films endured than I can remember, I wouldn't doubt a hundred
thousand. Yet in what I will call the rest of my life, I have seen just two
dead bodies. From all I can gather, asking around, two is on the high
side-—enough to be thought a little morbid.
What a marvel that our bang-bang movies are so seldom chided
for morbidity. Yet, as I recorded the deaths in The Godfather, I did feel
infected or aroused by the sheer exuberance and stylistic slam of the
movie. I recalled that aside from Lolita: "You can always count on a
murderer for a fancy style." In turn that triggered the recollection of
Francis Coppola admitting, long ago:
You know, I took my kid to see a 45-minute assembly of some of the

stuff of the old Godfather and I said what parts do you like better? He
said, "I like when the guys get shot." Everyone is like that. Even when
you're shooting the film. The second you're going to do a throatcut¬
ting or something, everyone including the crew crowds round.
He's right, of course (and he's no more bloodthirsty than most

directors today). Kids—or boys—generally exult in the best balletic
death scepes: it may be as close as they ever come to sissyness, prefer¬
ring style over content. In the late 1940s and early 1950s, getting out of
cinemas on Streatham High Road, I galloped to the prairies of Tooting
Bee Common to reenact death scenes. My friends and I vied for the spec¬
tacular bits in these remakes: it was fine to be John Wayne on the bare
89
land that would become the Red River D ranch, but so much more lib¬
erating to be Don Diego's man who is tumbled from the saddle by
Wayne's shot. Or Edmond O'Brien running through the streets of
D.O.A., clutching his bright stomach, loaded with irreversible poison,
running for his death. Or Elisha Cook Jr. in Shane, lifted off his feet and
thrown back in the mud by Jack Palance's guns. Cagney in White Heat,
seeming to go up like a rocket. There was a blooming in these deaths,
a jete, reaching for beauty. We grazed our knees and dirtied our clothes
doing death falls. But we glowed in their rapture.
As time passed, I guessed some actors had the same fun and
release in extravagant death scenes. Cagney and Edward G. Robinson
were treasured for those cadenzas in which they occupied time and
space with their delirious dance of death. "Is this the end of Rico?"
Robinson asked in Little Caesar. Well, eventually. Was there anything
in movies so much in love with life as these whirling expirations? Slum
kids could at least look forward to the brilliant strut and fret of passing.
Cagney's stricken grace leaned toward the fits and fevers of
approaching rigor mortis. "Made it, Ma!" in White Heat was only a vet¬
eran's fond tribute to the death throes from Public Enemy, He Was Her
Man, Angels with Dirty Faces (one of the first "showtime" deaths, proof
Figure 17. Ever the dynamo, James Cagney dispensed death while staying in constant motion. In
White Heat, in one fluid move Cagney eats his lunch (a chicken leg) and ‘‘ventilates’’ a victim
stashed in a car trunk.
90
David Thomson
that acting can get you over that hump), and The Roaring Twenties.
Cagney yearned to throw himself about, to smell extinction, tousle his
hair, and let his eyes see oblivion. He expanded on death scenes, he grew
lithe and poetical. It was only then that he could disclose his passion
for movement and his love of the precarious. Death gave this would-be
real-life radical his greatest sense of insurrection.
If you doubt the nihilistic elan in Cagney's demises, then look
at how bitterly Bogart went to his deaths. He never liked losing control.
In such predicaments, we feel the truth in Louise Brooks's observation
that "Humphrey" was a socially correct young man who liked to keep
his cool and his distance. Bogart tensed up when he had to die: his body
often folded in on itself, like hired evening dress packed in a suitcase.
If he had to "act," in death, it was his worst acting. In The Roaring
Twenties Bogart is intimidated, whipped, and mocked by Cagney, as if
he were Liston being taught psychic danger by Cassius Clay. It's notable
that in High Sierra Bogart's Roy Earle dies in extreme longshot, with¬
out benefit of triumphant staggering against the skyline. High Sierra
was Raoul Walsh (who made White Heat), but Sierra is drugged by Bog¬
art's depressive reticence. Bogart's glory learned a lot from High Sierra:
it developed a smoldering, still fatalism, the chance for a few wry words
before conclusions he had foreseen. But Cagney's death hound was
always lit up by the surprise—the discovery!—of bullets.
We know of actors who held a nearly contractual right not to
perish in fiction. To stay the hero, Wayne, Gary Cooper, and, more
recently, Clint Eastwood walked up and down in the shadow of death,
yet kept a beacon keylight on their ever more haggard faces. As one of
those not swept away by Unforgiven, I note the greater historical and
artistic plausibility if William Munny had died on his mission, rather
than revert to that reassuring bringer of death. The film might then con¬
clude with a simple scene of the forsaken Munny children, dying in
their cabin from cholera or loneliness. But the Munny who had lost
such edge and youth did pick up quickness at the end. Then he rode
home in a grim spirit, condemning all the fates and movie conventions
that had made him be lethal.
The comfort in Munny's regaining deadly impact surely ex¬
tends to Eastwood himself. No matter how far this moviemaker has
poked and prodded old genres, no matter his candor with age and fatigue
lines, he cannot do without looking good and potent. He is—and he
knows it—the last classic star. Thus, allegedly, he spent time research¬
ing Secret Service agentry for In the Line of Fire, but still indulged the
cockamamy of a 60-year-old jogging along beside the limo and the sen-
91
timentality of a Service that keeps such maverick problem children on

the payroll. Even then, the ludicrous fun of In the Line of Fire depends
on John Malkovich's fastidious killer. He is a delicious tribute to the
bliss of murderous daydreams. The fantasy appeal of In the Line of Fire
is a balanced two-hander, but it is Malkovich who knows the whole
thing is just a game. It is so often our killers now who are blessed with
wisdom and insight. They are the only characters allowed to turn to
philosophy or talk for the sake of talking.
It would not be out of order if Malkovich crooned into the
phone, "Frank, don't be petulant, you know, and I know, we're just play¬
ing checkers for the audience, and they love to think about killing, and
you're there just to waste me at the end so they feel okay about it. I've
told you all along I'm ready to die for the picture—but, Frank, are you
really ready to stop a bullet?"
And Clint is not: its thought seems indecent and un-American.
On the other hand, Malkovich is a master of all those infinitesimal
droops and melancholies that could while away a whole film with
dying—"As I Lay Dying" seems to be the dream in Malkovich's remote
eyes. Kirk Douglas in that ebullient youth of his seemed to crave hor¬
rid execution, agonies to prompt his throbbing cry, and movies where
he could be mutilated, marred, and generally pecked at by those birds
of story who knew his needs. Lee Marvin was made for the world-weari¬
ness that is a killer-for-hire only to stay awake. Thus, in Don Siegel's
The Killers, the actor/assassin who is gradually clearing the film of life,
mortally wounded himself, can say (to Angie Dickinson), "Lady, I'm
just too tired. I haven't got the time." A career and an attitude are made
lucid. James Mason was another actor born to see the sense in his own
extinction—think of Odd Man Out, North by Northwest, and Lolita,
or even Heaven Can Wait, where he is in charge of death's best hotel.
And who can mistake the self-discovery in William Holden,
gazing down from the top of the pool in Sunset Boulevard, the hack who
finally has a drop-dead story to peddle?
Death is so slick in film, it has become tongue-in-cheek. "If his¬
tory has taught us anything," Michael Corleone will announce in The
Godfather Part II, "it says you can kill anybody." That line portends
the resolute evil with which Michael orders his older brother Fredo
dead—in a rowboat on desolate Tahoe. It inspires confidence that any
Hyman Roth can be taken out in broad daylight at an airport. And it
tickles us to think that Michael may have assented to events in Dallas
on November 22, 1963—as In the Line of Fire makes plain, killing a
president is not that difficult.
92
David Thomson
But Michael's line signals a more pervasive mastery: that movie

can kill whomever its weary, plot-crazed eye falls on. It can put Shel¬
ley Winters on the bottom of the lake in The Night of the Hunter (so
that we marvel at the coup); it can consign Janet Leigh to the swamp
after a scant forty minutes of Psycho (outraging stardom's expecta¬
tions); it can make a studious tracking of so many serial killers for con¬
noisseurs, so long as the killers are caught before the final
crawl—except that after The Silence of the Lambs these killers may
roam the earth if they are nicely spoken, have discerning tastes, and
remember to call Clarice now and then.
There's a giggle in the end of Silence of the Lambs as Lecter goes
a-roaming. There was a more smothered chuckle in the setup for Sun¬
set Boulevard, a way in which contempt for Hollywood began to turn
against the audience; there is malice and self-loathing in the camp supe¬
riority that betrays our disbelief. Long before the clamor to be on death's
side in such titles as Die Hard, Lethal Weapon, The Terminator, and
Death Wish, our movies had made deals with the glamour of death. The
timing and polish of our well-made melodramas were like the ingenu¬
ity of Malkovich's gun in In the Line of Fire. They were to die for.
Deathliness is in the mise-en-scene. Think of the killings in
The Godfather: Isn't the consciousness observing them that of a Cor-
leone? These deaths are not messy or untidy. The attitude is proud,
masterful, in love with meticulous detail. Nothing in the sensibility
disturbs the remorseless efficiency of vengeance, or departs from the
managerial pleasure in seeing intricate plans work sweetly.
There is gallows humor in the spatial tranquillity of Paulie's
death—the wind-swept location; Paulie's stupid patience,- Clemenza's
ruminative urination; the far-away bump of the shot to match the
lovely longshot: it is a kill-master's dream with Paulie the spare pin
taken out by a dab hand. When Sollozzo and McCloskey get it, we are
nearly palpitating with the urge to give it to them—one reason why
"they" can kill anyone is because we are such willing, voyeur accom¬
plices. Finally, when the total elimination of enemies is orchestrated
with the rite of baptism, we might be witnessing the adorable fit of
nuclear physics.
Is there irony in the magnificence? Is Michael being con¬
demned? Watch the sequence again, and there is no escaping our
deranged complicity in the lethal arrangement. There is such macabre
comfort in feeling a part of the Corleones. The psychic infancy that
dreads all strangers has been protected—they have got theirs, and our
supreme plan has been vindicated. Michael is patriarch and paranoid,
93
for surely, one day, he will have to eliminate every family member and
anyone who knows.
It's during the grease-quiet, digestive mechanics of films like
The Godfather that we may recall how frequently film has appealed to
fascists. Not that I mean to suggest some directors are readier for the
jackboot than others. No, the dilemma is tougher for those of us who
love film: something essential to the medium cleaves to uncorrected
powers, the magic of plot (or organization), and the chance to stare at
death without honoring pain or loss. Still, for the moment let's pull
back from that comprehensive unease and offer an intriguing dis¬
claimer—roughly, that these corpses don't smell, none of these guys
were ever "alive" anyway. They're 'toons, for crying out loud!
There is something childlike in the easy dispatch of so many
people as quick as a wipe. My 4-year-old takes as much delight in hurl¬
ing himself to many deaths, and in inflicting them with pointed finger
and inner-mouth explosions, as ever I did. But did children have this
game, or its risk, before moving pictures? Perhaps boys keep the game
alive—and most things that depend on boyishness are becoming harder
to sustain. Yet there are plenty of decent film critics, some of them
women, who seem untroubled by the extended boyhood of, say, Sam
Peckinpah and the very cinematic motto, "Kill anything that moves"—
the line that introduces Peckinpah's credit on The Wild Bunch.
The deaths are easier to face when one gets in the habit of
knowing movie deaths are akin to a bucket of dip cleaning out the
premises. Dip, you may remember, is the green fluid that brings death
or erasure to all 'toons in Who Framed Roger Rabbit. It works in the
way water did on the Wicked Witch of the West in The Wizard of Oz—
Christopher Lloyd's judge dies with the same cry of frustration, "What
a world!", that Margaret Hamilton uttered in Oz.
Lloyd's judge actually dies twice. He is first flattened out by a
steamroller: this is what reveals his secret looniness, although it is a
cunning conceit of the Zemeckis film to say some characters are flat¬
ter than others in movie's two-dimensional illusion. Then the judge
reinflates his own balloon and comes on wicked again—with gestures
from The Texas Chainsaw Massacre—before dip takes him back to
primordial ooze. He is too good a villain for one death. Zemeckis is so
wantonly inventive in Roger Rabbit (and so inspired by other films), I
could believe in the judge bouncing back as often as the cat in Tom and
Jerry cartoons.
Did that cat ever die? Or were his lives infinite? He was reduced
to fragments, blown to smithereens, electrocuted, pancaked—you
94
David Thomson
Figure 18. Martin Scorsese’s films are often full of harsh, flamboyantly rendered physical violence.
In Cape Fear, the villain Max Cady (Robert De Niro) is subjected to an extended beating
sequence, one of several mutilating ordeals the character undergoes.
think of a way to go, Hanna-Barbera did it—and always there was the
swift fade-out, fade-in, and puss was back again ready for worse. He was
a character out of Bunuel. When killing is so easy and such fun, and
death so brief, it becomes a way of life. And not just for card-carrying
boons. We know how reluctant the film business is to let its most vital
killers take retirement. At the end of Halloween, the demonstrably
deceased Michael vanishes, so that he could return for sequels. As I
tried to puzzle out why Scorsese had made Cape Fear, I noticed the
clearly posed cue for Max Cady to grab back once more from the river,
just as Carrie came out of the black earth of her own grave. The moment
passed; it was presumably just a joke about such tricks. But Cady could
have escaped—so burned, so crushed, and so drowned that he would be
the harder to recognize next time.
With death so climatic or constitutional, ghastliness becomes
a subject for movies. Our watching from the dark, intensely "with" the
images yet powerless to intervene in their progress, is a model for sto¬
ries in which those left alive may keep some kind of community with
lost ones. After all, the thrust of movies is so much more imaginary
than actual. So there have been films in which ghosts come back, or the
living make a journey to the realm of the dead. From Nosferatu and A
Matter of Life and Death to Beatty's Heaven Can Wait and Ghost,
95
movies have played with the undead (without having to hire other than
the regular actors). Some of those films have resorted to "ghostly" spe¬
cial effects, superimposition, and so on. Yet, truly, no tricks are re¬
quired. No one on screen has a real life or corporeality. The films keep
playing long after the actors die. These are 'toons reread as appealing
solids by our fond credulity.
Ghost hints at a way movies- might—on the scent of Shirley
MacLaine and Marianne Williamson—burrow into the self-help of pro¬
jection. In several recent movies, there are wishful thinkings beyond
the grave, psychic schmoozings: in Field of Dreams, Kevin Costner has
the chance to meet his dead father again, so that the load of misunder¬
standing can be tolled away. That's not what I want an art form to offer.
But that won't deter the development, and who knows if movie isn't
less an art than just one of those fun boxes the Good Guys offer. We
may not be far from a household video facility that could take all the
hours of home movie of a loved one and then put that passed-away per¬
son into computer regeneration, so that those left behind have a house
guest/ghost to chat with.
Already, TV commercials (today's pioneering) have worked this
magic with the look and sound of dead stars. And there are rumors of
Jurassic Parks in northern California where that readiness is poised for
new feature films—with Bogart and Louise Brooks together again at
last—if only the legal details can be worked out. The one interest in the
actual Jurassic Park was that such ghosts shared a frame with the very
pale humans Spielberg had time for. Most of the time, the seams didn't
show. But suppose next that we could resurrect and write dialogue for
Elvis and Marilyn? Or you could have your own home video tete-a-tete
with the star/celebrity of your choice—the star as ultimate pet.
The bullet goes in, and life goes out. Movie prefers it as an
instant, switchlike adjustment, without suffering or waiting. Yet so
much of death is in those two grim departments. How do we stand up
for the very few deaths, and corpses, we may meet in our lives? They
are unscripted, no matter the anticipation; and they are not there for
slo-mo analysis. Are there moments in films when we have a better
than bang-bang understanding of what cessation means for the passer-
on and those left behind? I can think; of film deaths that move me, or
give me a sharper sense of the precariousness of life. In every case I'm
going to list, somehow, life remains the subject:
• the deaths in The Missouri Breaks are epic and lugubrious, undigni¬
fied yet very skilled, but the best is Lee Clayton's, asleep after his
96
David Thomson
horse pissed during his love song, then awake to the snap of bracken
or sinews, his own vain efforts to breathe or stay calm at Tom Logan's
dry inquiry, "You know what woke you up? Lee, you just had your
throat cut."
• in Pat Garrett and Billy the Kid, Slim Pickens realizes he has a death
wound while James Coburn is still fighting L. Q. Jones. But the music
rises for his death, and leads him down to the river (Pickens has been
building a boat to "drift out of this damn territory"). His woman,
Katy Jurado, follows him and they gaze at each other by the water as
he holds his startled face and his belly wound up to the evening light.
And then we see that Coburn (Garrett) is watching the death and tak¬
ing it in as evidence and responsibility. For he had hired Pickens away
from the boat-building. And then there will be Garrett camped by
another river, half an hour and months later in the film, as a vagrant
family drifts by on a raft.
• in Renoir's The River, the little boy Bogey is entranced by the cobra
at the end of the garden. He is warned. But he wants to charm the
snake, and one afternoon he goes too close. We never see the cobra
strike. There is just the boy's sprawled body. The loss comes not long
after a siesta in which the several members of the household are seen
sleeping, with the camera simmering on their breath. The boy dies,
a new child is born. This easily sounds trite as a philosophy, but
Renoir's structure, his camera, and the sense of breathing transform
the quietism so that it becomes as steady and flowing as the river and
the flooding sitar music.
• the moment when Isabelle dies in Amber sons, with the Major gazing
into the fire, speaking of the sun as the source of life—Richard Ben¬
nett rambling very near his own death—and then the fateful word,
the Major jerked out of his revery, ready to die, and the rapacious
embrace that Fanny has for George. Family in a few seconds
• and the death of Tom Joslin in Silverlake Life, the documentary made
by Joslin and his lover, Mark Massi, as they both faced the destiny of
AIDS. In this case, it is the entire film, the ending of which never
enjoys doubt. It takes a movie like this to remind us how gradually
and faithfully death comes. Bodies diminish and waste, the lesions of
sarcoma spread and join; courage and tact fail. Joslin sometimes rants
out of fear and horror at what is happening to him. Months and years
come down to ninety minutes or so of film, and we see Joslin seconds
after he has died—a skull with skin, yet freed or deserted by life and
the illness. We see so few authentic deaths on film, Silverlake Life
can put you off movies.
97
I saw Silverlake Life on June 15 on PBS. Two days later I flew to Eng¬
land for my father's funeral. I was reading Philip Larkin on the plane
because, for many Englishmen, Larkin has been like a life-sustaining
illness: he had a sensibility I loathed, and a capacity with words that
was piercing. And now, eight years after his death, Larkin is being
revealed—in Selected Letters and a biography by Andrew Motion—as a
furtive, less than honest man, darker and more afraid than he could
admit. I felt there was some kind of kinship between Larkin and my
father. My father left my mother when I was being born. He lived with
another woman for over forty-five years. But he came home to us at
weekends and Christmases, and he never once said anything about the
double life. This is Larkin in Aubade (he never sold the movie rights):
I work all day, and get half-drunk at night.

Waking at four to soundless dark, I stare.
In time the curtain-edges will grow light.
Till then I see what's really always there:
Unresting death, a whole day nearer now,
Making all thought impossible but how
And where and when I shall myself die.
Arid interrogation: yet the dread
Of dying, and being dead,
Flashes afresh to hold and horrify.
Before London, I went to Dublin to see my oldest friend, Kieran

Hickey. We met on the steps of the National Film Theatre in London
thirty-three years ago. He was the first naturally eloquent person I had
ever known—smart, gruff, lyrical, and caustic. He was a filmmaker, and
a good one: he made documentaries and short fiction films in Ireland—
Faithful Departed, A Child’s Voice, Exposure, Criminal Conversation,
Attracta, The Rockingham Shoot. I went to see him because he was set
for double bypass surgery on June 28.
We had a fine weekend in Dublin, his house on the south side
full of friends and all the things he collected. We watched a tape of Bit¬
ter Victory, letter-boxed yet incomplete, but with that scene where
Richard Burton's officer remarks on his skill at killing the living and
saving the dead. We watched Stephen Frears's wonderful The Snapper.
Then I went to England for the funeral. My father was 84. He had had
a stroke from which he never regained consciousness. I went to the
funeral home to see him in his coffin before the cremation.
It was him, yet the fierceness had gone, and with it the last hope
98
David Thomson
that he and I would ever talk about our history. You see, I had never quite
been able to make him tell me; and instead there were ways in which I
had imitated him. He was cremated and I came back to America.
Kieran had his operation. It was a complete success. The news
from friends in Dublin could not have been more positive. Two weeks
after the operation he was to be moved from the hospital to a nursing
home for further convalescence. But in a matter of hours he developed
a pulmonary embolism and died. The day after he died I got a card from
him, written from the hospital —"My improvement is marked." You
have to believe me: he would have chuckled and said, "Oh, dear, yes,"
for he loved irony.
The limit to death in most of our films is that it shows what
poor attention they pay to life.
I
Devin McKinney
Violence: The Strong

and the Weak
For many of us, our earliest and most lasting moviegoing memories
involve acts of violence. Whereas an older generation was marked by the
murder of Bambi's mother, my epiphany came with seeing Taxi Driver
at the age of 12 and being disturbed nearly to the point of physical sick¬
ness by its violence. But more intriguing to me now is the nightmare I
awoke with two days later. Somewhere I realized that I had responded
so viscerally not only because the violence felt physically real but
because it was emotionally and morally complex: it brought up ambiva¬
lences and dreads that no amount of rationalization could overcome.
Time swings like a pendulum, and violence is once again an
issue. Over the past few years a new ethos of violence has been accru¬
ing in the commercial cinema; directors have been attempting to take
it further—but not necessarily deeper. More than ever, violence has
emerged as thematic matter, the true meat even of movies that claim
to be about something else. This has given a sharper edge to both
“strong" and “weak" (as I'll call them) portrayals of violence, since both
must compete in a media marketplace that is ever more vicious and in
a social context that is ever more apathetic.
Of all that films contain, violence may be the most resistant to
quantification. Only at its weakest does it yield to patterns and pre¬
dictabilities. (Robin Wood conceded as much in his theory of the "inco¬
herent text," a theory whose unvoiced admission is that many of the
most violent and interesting films are so multilayered and resistant to
singular meaning that their systematization can be built only on con¬
tradictions.) But perhaps some generalities can be discerned. Perhaps
movie violence must contain multitudinous meanings if it hopes to
Copyright © 1993 by the Regents of the University of California. Reprinted by permis¬

sion from Film Quarterly, vol. 46, no. 4, Summer, 1993, pp. 16-22.
99
100
Devin McKinney
avoid the prisons of ideology and cliche/ if it seeks to draw anything

more than a distant, formalized response, if it wants to outlast its
moment. Perhaps it must bring the heart and mind together, and aim
for the emotions as much as the viscera. And perhaps in order to enter¬
tain a discussion of "strong" violence versus "weak" violence, one has
to accept the notion that some nightmares are worth having.
In an essay on "excessive" film types (weepie melodrama,
pornography, horror), Linda Williams determined that essential to the
allure of these "body" genres is their capacity to bring up unmitigated,
unsocialized emotions—the extremes of feeling not elicited by pictures
that take the straight and narrow path. Strong violence, while it often
has the physical effect of the body genres, also acts on the mind by refus¬
ing it glib comfort and immediate resolutions. If successful socialization
depends on a neutralizing of extremes, then violence of this kind
amounts to a rent in the curtain of rationality, a glimpse of the ultimate
questions one spends a lifetime denying. And it amounts to carnage that
is haunting in the truest sense—that gives "meaning and import to our
mortal twitchings," in Vivian Sobchack's fine phrase. This is a rarity on
today's screen, but it shows up just often enough to make you feel its
larger absence. It's the sort of violence that Wild at Heart pulsed with,
that framed the good small thriller One False Move, and that Scorsese,
despite the unwelcome gloss twinkling on his late work, can still
deliver. It was the violence, not the time-saving homilies about fallen
kings, that gave JFK its sense of loss. And it carried the climactic mur¬
der in Casualties of War, so surreally angled, to the level of catharsis.
Movie violence this strong communicates intensely the sense
that a person who in one moment is fully alive has been reduced to
God's garbage (as Joseph Heller perceived his doomed gunner, Snow¬
den). It holds unspoken contingencies, and by its nature is crazed. It
need not be particularly kinetic, but it shakes everything up, re-form¬
ing the entire Active environment around itself.
Plainly the terrain mapped here is not an easy place to reach.
Among recent films there are three in particular that in their impact—
both immediate and residual—make the discrimination between
strong and weak violence solid and meaningful. In each, the bloodshed
has subtext, carries the weight of fear and mystery, and is piercing
enough to shoot past the crap violence we all drink like beer.
If Neil Jordan's The Crying Game fits the present discussion, it
isn't because its violence is plentiful or showy but because it exempli¬
fies how strong violence can reshape both the entire context of a nar¬
rative and the audience's enunciation of the experience. The deaths that
101
Violence: The Strong and the Weak
occur in this picture mark beginnings as well as endings, and open a

world of new threats and possibilities. The nasty trick of fate that ends
the life of a British soldier (Forest Whitaker) signals the beginning of a
new life for his captor, an IRA terrorist (Stephen Rea), as does the mur¬
der committed by the Irishman's lover (Jaye Davidson) later on. In each
case, violence is committed—as most movie violence is not—because
it demands commitments of those still living.
This does not square with the conventions of movie violence,
where the cataclysm of death constitutes no more than a momentary
lull ending with a cutaway to the next sequence. Although stylistically
well within a certain classical-realist tradition, The Crying Game
breaks those bounds on the narrative level by according the conse¬
quences of violence a determining role. As much as anything, it is this
grasp of consequence that distinguishes strong violence from weak.
The emotions we are enabled to feel by The Crying Game are
ultimately comforting, affirmative, even warm—in the nonironic
senses of those words. This does not, of course, detract from the film
(whose warmth, in fact, is as central as its violence), but it does point
out that strong violence is often put to the service of perceptions that
are ugly and cold, and that it often etches a horrid picture. Abel Ferrara's
savage cop movie, Bad Lieutenant, swan-dives into the waste of the
excremental city, and whatever sympathy one extends to its eponymous
antihero is repaid only after a punishing swim. Ferrara pursues the new,
riskier violence in every respect, pushing a rigorous cum-stained filthi¬
ness against the keening excess of grand opera: the result is a Times
Square men's room with the design of St. Patrick's Cathedral.
FFarvey Keitel's doping, whoring, extorting L.T. is the film's cen¬
ter, and its metaphor: he is a walking, talking consequence. Bad Lieu¬
tenant is so single-minded and so steeped in the ubiquity of brutality
that its violences come to feel organic to one another—symbolically, if
not actually, intertwined. The L.T.'s act of autoeroticism—aided by two
frightened girls in a car—echoes and counters the event that spurs his
death-trip to redemption: the rape and vaginal mutilation of a nun by
two boys. The skein of suggestions underlying these disparate acts of
violence—active and passive subjugation of women, the priming of the
male pump versus the attempted destruction of the female organ—is
tight and intricate, with no vanishing point. That such echoes are heard
at all implies that the film has deepened its grotesquerie by problema-
tizing it, making it stand for more than a scream and a crotch shot.
And as relentless as it is, Bad Lieutenant has plenty of color;
it's lined with rock 'n' roll; it moves. Other films don't. On one level,
102
Devin McKinney
John McNaughton's Henry: Portrait of a Serial Killer can be denigrated

as a slack, unmodulated piece of filmmaking. But the truth of this judg¬
ment also underlines the fact that strong violence (like the features of
Williams's "excessive" genres) is not always an experience defensible
or divisible by aesthetic means. For though it may fail as an art object,
Henry can't be dismissed.
Its tonelessness and sensual inertia apotheosize the particularly
cruel, moribund tenor of violence in the 1980s. But unlike American
Psycho, Bret Easton Ellis's portrait of a serial killer, Henry does not
commit the imitative fallacy of depicting a barren mind through the
barren accumulation of idiot minutiae, mistaking a willed emptiness
for postmodern substance. It stands at just the right distance from its
subject, never enforcing a sociologically judgmental thesis but never¬
theless entailing that the killings be as grubbily, unexplosively real as
kitchen-sink style can make them.
Perhaps the indirect success of Henry as a study in violence is
that it comes to life only when witnessing death, and even then it
comes only to a kind of life. The banality of its violence issues from a
beclouding feeling of moral lassitude. There is a grinding insistence on
murder as a mere relief of tension, a dully masturbatory act, and it
infuses even the nonviolent scenes with a glowering menace. The life
seen here is entirely of a piece with the death: there is no "real world,"
no normality to return to. What this means in practice is that although
the presentation of violence is outwardly neutral, its effect is extreme.
Unlike the common run of hermetic, low-budget bloodbaths, Henry
puts its banality to a purpose. Its very monotony induces paranoia,
hypersensitivity to what was once ordinary. Like all works of strong
violence, it leaves an audience feeling dead inside, yet, somehow, more
alive than it was two hours before.
The kind of disconnected, uncommitted movie mayhem that
began with James Bond and that came of age with the sociopathic crime
dramas of the early 1970s and served the reactionary agenda of the
1980s is once again ascendant. The increasingly visible violence of the
global society virtually demands that its art be more violent, and
already all the barricades seem to have been breached, no crimes left
undepicted. But repetition always vitiates, particularly if it's the
lowest-common-denominator variety habitually practiced by the
movie industry. The paradoxes of strong violence are rich and mazelike,
but weak violence thrives on a sterile contradiction: it reduces blood¬
shed to its barest components, then inflates them with hot, stylized air.
103
Violence: -The Strong and the Weak
Figure 19. In Goodfellas Tommy (Joe Pesci) is executed with a bullet to the head and a flam¬
boyant blood spray. Note how director Martin Scorsese overtly plays this violence to the camera.
The characters face the camera, and the composition gives the viewer the ideal vantage from which
to watch the violence.
Weak violence appears nearly everywhere in both popular and

highbrow culture, and by its nature ridicules the powerful empathies
that hard, personalized violence can make an audience feel. It's akin to
what has been called camp, and it recalls John Fraser's account (in his
touchstone work, Violence in the Arts) of camp aesthetics "draining
off" the "charge of feeling and meaning" possible in a violent artwork.
It informs not only the products of the Schwarzenegger school—which
are merely bionic mutations of earlier action styles—but also works of
some moment: Peter Greenaway's The Cook, the Thief, His Wife and
Her Lover; parts of Goodfellas-, Pedro Almodovar's postcommercial-
breakthrough films,- David Cronenberg's adaptation of Naked Lunch;
much of Alan Parker's work; the French blockbuster La Femme Nikita.
The violence of these pictures simply doesn't last; it gets left on the
floor with the candy wrappers. It's too rationalized, too articulate—
either in the limited sense of "nice" cinematic effects too well con¬
trived to have any other content, or because the outrages are stapled,
memo-like, to external signifiers that bury their very peculiar mean¬
ings. Either way, violence is used only as a device: something a crowd
pays for when it goes in, but not when it comes out.
Although weak violence is marked by its lack of pluralities,
the sensibilities that create it are various. Among the hacks of the world
it is still the most popular pipeline to grim titillation, as witness Paul
104
Devin McKinney
Verhoeven's Basic Instinct. In building a dream world whose moral

parameters are so easily drawn (straight = good, gay or bi = evil), Joe
Eszterhas's screenplay abdicates any claim on psychological insight,
and the director's angle on violence is a congruent refusal to believe that
anything might be more than it appears. Like his demonized femme
fatale, Verhoeven wields an ice pick—in the form of visual violence—
and he uses it to pin his characters to the mat. Significantly, no picto¬
rial reference is made to the accidental murders gratuitously
committed by the hero cop, while the brutalities of the central lesbian
are virtually turned into production numbers. The characters are not
revealed by their acts, only summarily defined, and violence is used not
as an entry to human depths but as a means of shutting them off.
But this sort of lacquered, discrete violence is no longer the
exclusive province of the schlockmeisters. It also serves the needs of
young, earnest filmmakers eager to showcase their formal skills, and
who therefore seem worthier of considered attention because they're
walking the art house walk. A paradigmatic case is Phil Joanou's State
of Grace, a fitfully impressive, high-voltage melodrama about the stub¬
born dregs of New York's Irish mob. The picture follows its bumpy,
overdeliberate, but generally absorbing course until a climactic mas¬
sacre that comes charging right out of the movie past, impeccably chor¬
eographed to evoke the shades of Travis Bickle and the Wild Bunch.
Joanou's marshaling of time and space is voluptuous and breathtaking,
but the scene's brilliance is an affront. One is severed cleanly from the
film's involving fiction—from a world where nearly everything seems
real and vital to one where nothing does. Unlike the fierce, flatly
observed throat-cutting earlier on, the bullets fired in this slow-motion
shootout inflict no pain on a viewer; the blood-squibs explode with all
the portent of popcorn.
Even more accomplished than State of Grace in its hollow
treatment of violence is Reservoir Dogs, the debut of writer-director
Quentin Tarantino. This high-gloss, low-budget job about the prepara¬
tion for and disastrous fallout from a botched jewel heist is aggressively,
conscientiously violent. In terms of perceptible intent, the tyro auteur
seems involved in the playful crossbreeding of pop culture references
from the pqst three decades with his own witty, state-of-the-art thriller
technique. But despite its conscious avoidance of "relevance"—all its
backward glances and retrograde finery—Reservoir Dogs is a pristine
reflection of its socioartistic climate,- this brutal exercise says much
about the way we're absorbing violence today.
105
Violence:.'The Strong and the Weak
The picture's savagery is so assiduously appointed that it

demands analysis as a thing in itself, a component of the work overar¬
ching all others—which is to say that the story usually serves the vio¬
lence rather than the reverse. In theory, this is hardly a fatal flaw, and
indeed, there is an initial passage whose strength leads one to expect
that the subsequent violence will evolve into something more con¬
summate. Speeding from the scene ohthe crime, the most psychologi¬
cally complex of the thieves (Harvey Keitel) holds the hand of his friend
(Tim Roth), who lies writhing in the back seat, slowly bleeding to death.
The sequence is brought down to the tightly gripped hands and the
strangled growls of the backseat passenger. It doesn't have many levels,
but it does have some, and above all the scene is obstinate—it sinks in;
and this at least bodes well for Tarantino's willingness to allow the vio¬
lence a lifelike rhythm and character. But when the crew of hoods con¬
gregates in a warehouse to lick its wounds, the picture relaxes into its
oddball formula and, despite its blood and curses, turns quisling, water¬
ing its frenzy with self-satisfied cool.
Figure 20. The notorious ear-cutting scene in Reservoir Dogs is a moment of gratuitous screen
violence. Failing to connect with any previous material in the film, it performs no narrative func¬
tion. But it does offer a striking visual showcase for ultraviolence, in which the bloody effect
becomes its own end and purpose.
106
Devin McKinney
The film proclaims itself to be a jaunty, hyperbolic comedy of

horrors that aspires to no particular realism or social import. It occurs
in a comic-existential dead zone, and is as nonreferential to a reality
outside itself as a (theoretically) mimetic work can be. Initially, it
requires a stretch to allow that a gang of robbers—and self-designated
"professionals" at that—would think it clever to outfit themselves in
identical Blues Brothers costumes, or to go by ostentatiously colorful
code names like "Mr. Pink" and "Mr. Orange." But Reservoir Dogs has
approximately the same relation to true crime as Mandingo had to life
in the antebellum South. This is a formalist filmmaker's logic, and it
has only one self-apparent subject: the set-piece.
The set-piece that has drawn the most attention—perhaps
because it draws the most attention to itself—is the one in which Mr.
Blonde (Michael Madsen) gives the bloody business to a trussed-up
policeman, all the while dancing the dance of the blissfully insane and
swishing away with a razor to the accompaniment of a wonderful old
Stealers Wheel tune. Although this may have seemed great on paper,
the combination of pop music, self-infatuated dance, and offhand
amputation inspires no clashing sensation in a viewer, perhaps because
the incongruous ingredients are too obviously chosen for their effect of
creative perversity. With every authorial impulse visible, the suppos¬
edly wild scene has an inapposite smoothness and objectivity, rein¬
forced by a Formica-like sheen over the colors. When Blonde cuts off
the cop's ear, it's not a determined act in any real sense but a show of
flamboyance for its own hip, photogenic sake.
Where does a scene like this locate the audience? What is a
viewer's level and kind of complicity, if any? On the surface (which is
where weak violence phenomenalistically stays), one identifies with
neither the victimizer (the infantile Blonde) nor the victim (the objec¬
tified cop), and is thus put in the position of the most passive, disinter¬
ested of observers. This is the limbo that weak violence inhabits:
empathies are not engaged, commitments are not brought to bear,
ambivalences are not acknowledged, neutrality is the currency.
Strong violence enables—and often entails—shifts in one's
moral positioning. This is a part of its power, and a great deal of its
threat. In fact, the realm of violence is one of the few aspects of film
that can be legitimately called co-constructive. Movies, despite the
penis-envious treatises of academically oriented critics, offer the
viewer no degree of imaginative co-authorship comparable to that in lit¬
erature. The image, after all, is right there before us, concrete block-like
107
in its sensual solidity, ranging in opacity from a stiff Eisenstein tableau

to Eric Rohmer's airy idylls. But when acts of strong violence are
involved, a picture can (like Garry Wills's comment on Oliver Twist)
open moral trapdoors under us, engendering an immediate and very sub¬
jective confrontation with the material. The most memorable of Hitch¬
cock's murder scenes, for instance (the Psycho shower murder and its
aftermath, the slow killing of the German agent in Torn Curtain), are at
once acts of pure calculation and spontaneous disruptions of rational
order; the calculation courts the audience's desires and the disruption
serves to foreground those (discomfiting) desires. The audience is both
acted upon and made to act by acknowledging its role in the fulfillment
of a wish it barely knew it had: it is both victimizer and victim.
The new exploiters of violence use bloodshed as a kick, but in
an agile, knowing, and self-consciously formal way. Verhoeven's kinky,
whorey thriller is tarted and rouged with Hitchcock allusions, while Mr.
Blonde's torture of the cop is meant not to horrify but to succeed as a
ghastly entertainment, a bloody song and dance. It's a clean kick, and
guilt-free, too: it obviates a viewer's self-disgust at getting off on sadism
by positing figures who are plasticized and unreal to begin with, lacking
even a certain crude B-movie immediacy. The gut-wrench one feels at
1 . X III- .--- - OO L L
ing approach to ultraviolence. Schwarzenegger’s heroes usually concluded their mayhem with a
quip that took the bloodshed in the direction of comedy and “weak” violence.
108
Devin McKinney
the parboiled savageries of a Basic Instinct bears the same relation to

real emotional empathy as the tears provoked by one of Douglas Sirk's
moist epics: a conditioned reflex. But without emotional response, film
violence trades flesh and blood for hamburger and ketchup.
If a film makes the decision to be violent, it shouldn't go about
its business timidly: no art ever came of a hedged bet. But most of the
violent pictures that cross the screen these days, however dangerous
they appear, are as conservative at heart as a Disney fable. These films
hedge their bets on the level of audience involvement by refusing a full
commitment to their own content: they want to look at horror, but they
don't want to feel it, smell it, take the chance of getting sick from it. By
insuring itself in this way, a violent film can't help but resist a viewer's
emotional investment, which, frustrated, displaces itself onto an aca¬
demic admiration of style.
This consideration of involvement, finally, is perhaps the most
important. All the other issues—consequences, multiplicity of mean¬
ing, arousal of empathy—are united by the questions that measure how
successfully a group of film artists have involved their audience in the
experience of violence. Do I care about this character? Will his or her
passing leave a gap? Will I remember what I've seen?
More and more, violence—not merely as an aspect of art but as
an everyday presence—seems to come equipped with its own escape
hatch, its own assurance that involvement can be avoided. Those who
communicate violence in its varied forms are eager to provide the
means by which the receptor can reify it into a construct, something
not messy and uncontrollable but regimented, with the workable out¬
lines of fiction. A currently popular example is the phenomenon of
"reality television," which in its seamier modes is like a Marxist's
nightmare of capitalist decadence, or the bottoming-out of what Guy
DeBord called "the society of the spectacle." Desperate for new and
exciting diversions, millions gather around images of real-life violence
edited and rechanneled as entertainment, images either caught in the
flesh by video cameras or reenacted by those whom the violence befell,
complete with stage blood and stage screams. "You'll think you're
watching a disaster movie," one such show promises, hawking its
enticing miseries, "but it's real."
As the violence of the world, the country, the city, and the street
crowd closer, threatening to collapse whatever peaceful center one has
tried to maintain, the more one is tempted to seek insulation wherever
it can be found. There is a politics to all of this—or to be precise, a non-
109
politics—and it cloaks itself in the absolving cleanliness of denial.

Movies that offer the queasy pleasures of violence half-baked con¬
tribute just as much to the process of insulation as the remote control
that shuts off the nightly news. Today's average violent movie doesn't
ask suffering from those already inclined to stand aloof from it; it exacts
none but monetary payment. Which brings me back to my childhood
nightmare. The reclamation of that same empathy and receptivity is a
project as old as human history, and when a film artist makes us cry
over spilt blood it can start to seem like the project is worth it.
Vivian C. Sobchack
The Violent Dance:

A Personal Memoir
of Death in the Movies
Violence and death do not only sell movie tickets today; they also are
the source of countless arguments, safely confined to paper, to panel
discussions, to cocktail parties. What is all that technicolor blood doing
to our youth? Are we a nation of voyeurs? Can the movies make us cal¬
lous, unfeeling? Are the films innocent, merely presenting to us some¬
thing which is already there? Is today's movie violence reflective of
some phenomenon presently existent in our society (good, "honest,"
documentary revelation like The French Connection) or is it teaching
us to regard blood and death with a blase aplomb we would not other¬
wise acquire in our own insular lifetimes? Answers to these questions
often boil down to whether one believes that cinema is reflective or
affective, mimetic or cathartic. At any rate, issues such as these have
been worried about and nosed over until one's response becomes, at
best, gentle boredom and, at worst, calculated oversight.
Violence and death have always been with us in the darkness of
the theater. They were there before I became a regular moviegoer in the
late 1940s and will be there long after I have bloodlessly expired in bed
or been juicily run down by an aggressive taxicab. They've always been
there, as familiar as the smell of popcorn, and yet the violent deaths I
remember from long ago—the deaths which have stuck to my ribs, so
to speak—are rare. It certainly was not because I was shielded from the
horrors of the screen. The first film I ever saw was Bambi and his
mother's death before the forest fire was a movie experience I'll never
From Journal of Popular Film 3 (Winter 1974), pp. 2-14. Copyright © 1974 Heldref Pub¬
lications. Reprinted with permission.
Afterword copyright © 1999 by Vivian Sobchack.
110
Ill
The Violent Dance: A Personal Memoir of Death in the Movies
forget. Little sobs, including my own, rose in the audience,- animated or

not, the violence of that fire and the death of a mommy were real. Sat¬
urday and Sunday matinees followed—triple features plus ten car¬
toons—and the cinematic fare I grew up on consisted of little else but
violence: Errol Flynn, Tyrone Power, Stewart Granger ran their enemies
through and through, themselves suffered the rack, the Spanish Inqui¬
sition, and the slings and arrows of assorted outrageous fortunes,- Joel
McCrea, Randolph Scott, Audie Murphy said they hated guns but used
them expertly, made their opponents dance to the graveyard or the scaf¬
fold in the name of decency and law and order,- John Wayne and his bat¬
talions bombed, machine-gunned, and grenaded their way to Bataan and
beyond, tearing themselves on barbed wire, dragging useless but pre¬
cious left legs after them like so many sacks of mail from home.
Was it simply that as a child I had no intimations of my own
mortality? Those deaths certainly never had the impact of the one in
Bambi. (Perhaps I unconsciously believed that only mothers could die.)
I watched those scenes of piracy and gunfighting and war with a great
deal more forbearance than I gave to the gratuitous love scenes at which
we all groaned in pubescent unison, but they didn't really touch me as
did Bambi or the violent films I see today. Was it merely childish hubris
that kept me essentially unmoved and only marginally fascinated by
the deaths I saw in what were supposed to be fairly realistic movies? I
think not. I feared and cowered before death in the horror films; I
believed in Dracula and Frankenstein and Vincent Price. Death by fang,
claw, and hot wax were very real to me while bullet holes and stab
wounds never once caused my steps to quicken on my way home from
the theater.
Perhaps the violent deaths I saw in those somewhat representa¬
tional films were too realistic to move me to an awareness of my own
fragile flesh. For one thing, they were often expertly concealed in a
Breughel-like mise-en-scene, similar to one of those puzzles we tried to
solve on rainy days. There are twenty-three rabbits hidden in this pic¬
ture. Find them! Death came swiftly, noisily, and in the midst of confu¬
sion—on the decks of pirate ships, in the circle of covered wagons, over
the teeming battlefield. As in real life, no one told us precisely where to
look and by the time we had found the locus of interest, a character or
two had already neatly keeled over. The longshot, the panoramic view,
kept death far from us and that was real. The bullet holes were too small
to see well; the sword wounds were always on the side facing away from
the camera. Sitting in the theatre, it was as realistically futile for us to
112
Vivian C. Sobchack
crane our necks to see—really see—as it was when we passed an auto¬

mobile accident in the car and my father refused to slow down in the
name of good taste. In this sense those movies (and I'll include most of
the domestic variety up through the mid-1960s) were indeed realistic.
Real violence happened far away, neatly in the straight columns of a
newspaper, safely confined in the geometric box of a television set.
As in the films I saw when I was growing up, real violence
occurred quickly, giving neither the participants nor the spectators
much time to abstract themselves from emotion, time to examine and
explore and perhaps comprehend the internal mystery of the human
body. Of course, in these films, we did have the drama of anticipating
death, a drama usually not present in reality; the long ticking seconds
before the shoot-out, the tense crawl over rough terrain. But death itself
was quick, the camera didn't extend the last bleeding seconds of the
antagonist's life nor did it linger on the carnage. Death was acknowl¬
edged in these films, but not inspected. Too few of us in the audience
(our parents included) felt threatened enough by the presence of death
and violence outside the theater to need the comfort of a microscopic
inspection of it on the screen. Although it was talked about, violence
was not yet an everyday occurrence, a civil occurrence; it was not yet
a time when we feared not only the incomprehensibility of death but
the incomprehensibility, the irrationality and senselessness of man as
well. Death in the movies may have been quick, but it was dramatic,
meaningful. Those who died did so for a reason. In those days, we did¬
n't even think in terms of assassination (something that happened to
Lincoln and livened our history books up a bit), or about junkies, mad¬
men, snipers. We could be pretty certain (who even doubted it?) that the
clean-cut kid next door was a clean-cut kid. Our relationship with vio¬
lence and death in those Saturday movies was the same relationship we
had with them in life. They happened to someone else and were mildly
titillating, mildly disturbing.
Intimations of what was to come—given the right climate, the
right circumstances and environment—were revealed in our youthful
attitudes toward blood. In another sense, those movies were unrealis¬
tic; they didn't satisfy the very human curiosity that only children
cared to vpice in those safer times. They never told us what we wanted,
albeit hesitantly, to know: the color and the texture of blood. All of us
children, superficially satisfied with the realistic limitations of movie
violence, worried over our cuts and picked our scabs and wondered at
the rich red that coursed through our insides and occasionally came to
the surface. Blood was something we were rarely given enough of on
113
the screen—it oozed rather than spurted, it was most often black or a
rusty Cine-color. Even in medium close-up (the man-to-man shoot-out
or duel, the pink traceries of a whiplashing), I can't remember a single
movie death which fired my imagination as much as the bright red
blood from my own finger. Then as now, we humans attempted to hide
from the frightening reality of our fragile innards by believing in the
strength of plastic and supermarkets. kYet we were fascinated, as we
have always been, by blood and tissue and bone. Snowden's secret
(Catch-22) is everyone's secret. In the late 1940s, however, and in the
1950s and early 1960s it was a bit easier to keep that secret; everyone
was comfortably mum.
Then, suddenly it seemed, in the mid-1960s, there was blood
everywhere. We didn't have to go to war to find it—or stretch our necks
to peer at it on the highway. Blood appeared in living color in more and
more of our living rooms. And it was there all around us in the streets,
is still there. Politicians became surprisingly more than caricatures,-
they became mortal. People who looked and lived exactly as we did shot
at us from water towers, slit our throats, went berserk, committed
murder next door. Even children died and supermarkets—meats neatly
packaged and displayed to exorcise the taint of the slaughterhouse—
were no longer havens of safety and sanity and civilization. No place,
however ordinary, was safe; blood ran in busy streets, on university
campuses, in broad daylight, everywhere. We were all threatened and
terrified, all potential victims of our not so solid flesh and some
unknown madman's whimsy.
Public figures splattered against our consciousness: John Ken¬
nedy, Martin Luther King Jr., Malcolm X, Robert Kennedy, Governor
Wallace. Little men with strange other-worldly lights in their eyes took
up arms against anonymity: Charles Whitman, Richard Speck, Richard
Hickoclc, Perry Smith, Charles Manson. Students, police, and the
National Guard fought it out on university campuses where the most
burning issue when I went to college had been whether we had the right
to wear Bermuda shorts to class. None of the blood spilt was pictur¬
esque or patriotic. Death by violence became a possibility for all of us
because it lacked sense and meaning much of the time; there was no
drama and catharsis. The blood in our lives had nothing of art or dis¬
tance about it and we all felt personally threatened. Each one of us could
die, each one of us could bleed, each one of us had to consciously
acknowledge the secret which Snowden so shyly concealed in his flak
suit—we all had pink and vulnerable guts.
One can hardly think that someone's careful calculations
114
Vivian C. Sobchack
somewhere in Hollywood read fear in all our faces and saw our hunger
for the seeming security of knowledge, our yearning to find meaning in
the senselessness of random violence. Films which had previously spo¬
ken to our unconscious desires and fears, like Topsy, "just growed."
And in the fear-ridden 1960s and 1970s, it has been no different.
There always is, of course, a first film—the film which tran¬
scends its surface intentions and burns into us some unstated message
with the intensity not of an arc lamp but of a laser. Bonnie and Clyde,
released in 1967, was just such a film. Although it was not the first film
to overtly bathe itself in blood, it was the first one to create an aesthetic,
moral, and psychological furor. Uneven in tone yet brilliantly con¬
ceived, it fired our imaginations not merely because it was a good film,
but because it was the first major film to allow us the luxury of inspect¬
ing what frightened us—the senseless, the unexpected, the bloody. And,
most important, it kindly stylized death for us; it created nobility from
senselessness, it choreographed a dance out of blood and death, it gave
meaning and import to our mortal twitchings.
There has always been violence and death in the cinema. But the
cinematic phenomenon in the films of our decade which is new and sig¬
nificant is the caressing of violence, the loving treatment of it by the
camera. The most violent of deaths today is treated with the slow-
motion lyricism of the old Clairol commercials in which two lovers
glide to embrace each other. The once abrupt drop into nonbeing has
become a balletic free fall. This, of course, is what has incensed those
who fear for the nation's children and the nation's morality. Making
death and violence "beautiful," they suggest, may seduce all of us into
a cheerful acceptance of gore, may whet our aesthetic appetites for more
and more artful and prolonged bloodletting, may even cause us to com¬
mit violence. After all, The Wild Bunch and their victims died so grace¬
fully. Sonny's body was a poem in motion when he was machine-gunned
to death in The Godfather. A Clockwork Orange kept our toes tapping
to the tunes of rape and beatings. The French Connection, while
certainly not musical, possessed a staccato tempo which was brutal
and bloody.
I suppose that what bothers the moralists most is that these
films are selling, are popular, are even critically acclaimed. The idea
that we are a nation of voyeurs is indeed disturbing. Our Wednesday
and Saturday night outings to see blood and gore while we eat buttered
popcorn is disturbing. But then being human has always been disturb¬
ing. People are no more corrupt, no more voyeuristic than they were in
the days of my childhood. But we are certainly more fearful. And it is
115
this fear, rather than peeping Tom-ism, which has caused violence to
almost literally blossom—like one of Disney's time-lapse flowers—on
today's movie screens.
Fear has made even the most squeamish of us take our hands
from our eyes. We still are afraid of violence and blood and death, but
we are more afraid of the unknown—particularly when it threatens us
personally and immediately. Even those of us who couldn't stand the
sight of blood at one time find our desire to know and understand blood
stronger than our desire not to see what frightens and sickens us. In a
time when we seemed safe, not immediately threatened, we could
ignore our fear and indulge our squeamishness. Today, this is hardly
possible.
After many viewings of the film, I still cannot watch the scene
in Dali and Bunuel's Un Chien Andalou in which a woman's eyeball is
slit by a razor. I have read about it, heard about it, but I have not seen
it. I cannot keep my eyes open; I have tried, but I physically and men¬
tally cannot do it. The content of that scene is violent and bloody
(although in black and white), and I am deeply afraid to watch it, to
know it. On the other hand, I sat motionless and wide-eyed through
Straw Dogs; my eyes refused to leave the screen for even a moment. I
can remember my husband wanting to leave in the middle (he said his
arm had fallen asleep and we had already watched The French Con¬
nection). I didn't want to go; I didn't want to miss anything. If that
sounds voyeuristic, in essence it wasn't. I got no pleasure at all out of
watching Straw Dogs. I felt extraordinarily tense, upset, sick. And yet
I could not leave the theater until the film was over, even at the expense
of a family argument. For some reason, it seemed to be a matter of life
and death—mine—that I stay.
What is the difference between my responses to the two films?
Both play upon hidden fear in the audience. Both have moments of
extreme violence. Why should I be able to watch one, but not the other?
The answer lies, I think, in the qualitative nature of the violence
involved. I don't have the pressing need to see a woman's eyeball slit by
a razor; seeing it will do nothing but disturb me. And this particular vio¬
lent action—although terrifying with or without its Freudian implica¬
tions—seems to have little to do with my life as I live it every day. I am
not afraid of someone's slitting my eyeball while I passively submit.
Watching that scene, in other words, is not going to instruct me; it is
not going to reveal to me something that is terrible, but which I need
to know. The nature of the violence in Straw Dogs is different. It may
not be treated surrealistically, but it is not totally realistic either. (One
116
Vivian C. Sobchack
Figure 22. David Sumner (Dustin Hoffman) defends his home in the climax of Sam Peckinpah’s
treatise on violence, Straw Dogs.
remembers the little tailor in the fairy tale who wore a sash that pro¬
claimed "Seven at One Blow.") Yet the violence in Straw Dogs touches
contemporary nerves in a way that Un Chien Andalou doesn't. It
involves the kind of violence that one fears now, today. Sickened, ter¬
rified, I had to watch the film. I had to learn and know what I fear and,
however painful the experience was, for the moment I found a certain
security in the fact that I had not backed away from instruction. In
short, I was doing my homework—trying to learn how to survive. David
in that movie was much like myself, the people around me. We all just
wanted to mind our own business and yet found ourselves, our homes,
our lives, threatened by people and things which plainly didn't make
sense, weren't at all rational.
Popular films have always given us—the audience—what we
want; otherwise they would not be popular. Today (just recently
twenty-seven bodies were unearthed in Texas), we want to know blood
and death. Although we retain little of the optimism which sprang from
the Age of Reason, we still believe, even if it's half-heartedly and hope¬
lessly, in knowledge, enlightenment. Knowledge is the magic which
will save ys; cataloguing will restore our crumbling sanity; inspection
will cure our anxiety. Blood and tissue, death and killing, rape and beat¬
ing don't please us, don't titillate us, can't be glibly compared to the cen¬
terfold of Playboy. Yet we have the clear and present need to know
them, to have them made significant rather than senseless, to have
them dramatized. Hence the slow motion, the lingering look at physi-
117
cal agony, the tongue-in-cheek treatment of brutality, the indecent

amounts of blood. Our films are trying to make us feel secure about vio¬
lence and death as much as it is possible; they are allowing us to purge
our fear, to find safety in what appears to be knowledge of the unknown.
To know violence is to be temporarily safe from the fear of it.
Of course, our belief in our sophistication, our comprehension
of violence lasts about as long as it tajtes to walk from the theater to
the subway or to the car—and sometimes not even that long. We
quickly realize that the orgies of blood on the screen have told us noth¬
ing really useful. Our fear returns, and makes us return to the theatre
to see more, hoping against hope that we will finally understand.
Our world has changed from the world of the 1940s, and 1950s,
and early 1960s. And a world is changed by people. Yet the individual
moviegoer today is not really much different than he was a decade ago.
When I was a child, it was considered in bad taste to look at blood and
talk about death. It wasn't "nice." Adults politely and tactfully avoided
the issue because it was possible to avoid it; only we children, lacking
manners, refused to hide our fascination with mortality. Today man¬
ners and niceness are too small a social device to cover up the fear and
actuality of imminent violence.
At the drive-in, I can see my 3-and-a-half-year-old pajama-clad
son become disturbed at the bloody and violent coming attractions, dis¬
turbed in a strange but very human way. He wants to know why and
how. He stops eating his popcorn and wanting to drive the car. He leans
forward from the back seat and carefully watches the screen. He
becomes even more inquisitive than usual, running a high fever of fas¬
cination and questions. He is too young to be personally threatened by
the particulars of the violence he sees (the death of Bambi's mommy
would make him howl), but the blood on the screen seems to sing out
to his blood. He may not know it, but he does have intimations of mor¬
tality. We, his parents, share his uneasiness, but don't tell him to hush.
We don't tell him his curiosity is not nice. We sit there in the car, a con¬
temporary family, united with all those other adults and children in
their cars by a desire to know.
In the past we also wanted to know blood and violence but were
afraid to ask, found it inconvenient, unnecessary for our day-to-day sur¬
vival. Today, we are afraid not to ask. It seems our very lives depend upon
the answers we get. The movies today merely reflect our search for
meaning and significance—for order—in the essentially senseless.
Drama can give us that meaning and order through form, through style.
What frightens us in our daily lives is not only the possibility of personal
118
Vivian C. Sobchack
violence, but the omnipresence of the chaos which surrounds contem¬

porary violence, the fact that we may die for no other reason than that
we were there to be killed.
The films today which stylize death and blood (and I can't think
of many which don't) paradoxically reflect our fear of chaos but also cre¬
ate order. The senselessness and purposelessness, the randomness of
real violence is certainly upheld in the content of current movies. In
The French Connection, for example, a young mother walking her baby
in the park is accidentally shot, commuters on a subway train find
themselves the victims of something they cannot comprehend. Current
films reflect our fear for our lives, our fear that no matter how unin¬
volved we may be, we are all potential victims of accident, that our
deaths will horribly have no meaning at all. Yet the very presence of
random and motiveless violence on the screen elevates it, creates some
kind of order and meaning from it; accident becomes Fate. If we should
get shot minding our own business, if we should be mowed down acci¬
dentally, at least we have seen it on the movie screen, have felt—along
with a theatre full of people—for the hapless victim. And we also have
realized that there was some reason even if the poor bystander who fell
to the pavement didn't know it: the cop had to catch his crook. Even in
films which are not lyrical in their presentation of violence and do not
use the mannerisms of a Peckinpah, death—when it comes to Every¬
man—is given a nobility simply by its presence on the screen, its
acknowledgment by the camera. These films reflect our fears, but also
allay them.
All the current movies, then, which deal with violence say
something important, if not particularly helpful. If we are to die for no
apparent purpose, they seem to say, at least we will die with style, with
recognition. We will create our own purpose and reason as we die; style
will give our senseless death some sort of significance and meaning. The
form of death in the movies today—the way the camera treats it—allows
us to find some brief respite from our fears. The moment of death can
be prolonged cinematically (through editing, slow motion, extreme
close-ups, etc.) so that we are made to see form and order where none
seems to exist in real life. The movement of the human body toward
nonbeing,is underlined, emphasized, dramatized and we all become
Olympic participants of Olympian grace. We can also see ourselves on
the fringes of the frame, falling by the wayside, but falling in the movies.
We may be horrified by the senselessness of violence and death,
but we are also lulled and soothed by the possibility that there is—for
all of us—a moment of truth, a moment of drama. We can believe, if
119
only briefly, that even the most senseless and violent and horrible of
deaths has at least a form, an internal order, and therefore a meaning.
Those bloody and brutal films which appear on our theatre screens
today perform, for us all, a kindness.
Afterword: The Postmorbid Condition
It is exceedingly strange to revisit something one has written twenty-

five years ago. My son is now 29 and has his own car—and I no longer
feel I have some desperate need or wish to see violence on the movie
screen as I did when he was 3-and-a-half and the country had just begun
to recognize explicitly that it was irreparably and irrevocably altered in
the aftermath of "the sixties," Vietnam, and civil unrest. Perhaps this
is because Pm a quarter of a century older and, facing my own fragile
flesh and mortality, find violence on the screen so much "play-acting":
that is, only a trivial figure of the larger annihilation that awaits each
of us and looms larger on my existential horizon that it did when I was
in my early 30s. Or, perhaps, I avoid violence in the movies because
now, after various and intense experiences of physical pain, it affects
me more strongly than it did before, writing itself on my body as it
writes itself on the screen. Perhaps, however, my growing avoidance of
screen violence (I would be dishonest if I called it "disinterest") also has
much to do with changes in our culture's relationship to violence over
the past quarter of a century and, correlatively and reciprocally, in the
representation of violence in the movies themselves.
In the essay above, written so many years ago, I argued that screen
violence in American, films of the late 1960s and early 1970s was new and
formally different from earlier "classical" Hollywood representations of
violence. This new interest in violence and its new formal treatment not
only literally satisfied an intensified cultural desire for "close-up" knowl¬
edge about the material fragility of bodies,- but also—and more impor¬
tant—made increasingly senseless violence in the "civil" sphere sensible
and meaningful by stylizing and aestheticizing it, thus bringing intelligi¬
bility and order to both the individual and social body's increasingly ran¬
dom and chaotic destruction. Indeed, I argued that random and senseless
violence was elevated to meaning in these then "new" movies, its "tran¬
scendence" achieved not only by being up there on the screen, but also
though long lingering gazes at carnage and ballets of slow motion that
conferred on violence a benediction and the grace of a cinematic "caress."
120
Vivian C. Sobchack
Today, most American films have more interest in the presence

of violence than in its meaning. There are very few attempts to confer
order or perform a benediction upon the random and senseless death,
the body riddled with bullets, the laying waste of human flesh. (The
application of such order, benediction, and transcendental purpose is,
perhaps, one of the explicit achievements of Steven Spielberg's high-
tech but emotionally anachronistic Saving Private Ryan, and it is no
accident that its context is a morally intelligible World War II.) Indeed,
in today's films (and whatever happened started happening sometime
in the 1980s), there is no transcendence of "senseless" violence: it just
is. Thus, the camera no longer caresses it or transforms it into some¬
thing with more significance than its given instance. Instead of caress¬
ing violence, the cinema has become increasingly careless about it:
either merely nonchalant or deeply lacking in care. Unlike medical
melodramas, those films that describe violent bodily destruction evoke
no tears in the face of mortality and evidence no concern for the fragility
of flesh. Samuel L. Jackson's violent role and religious monologues in
Quentin Tarantino's Pulp Fiction notwithstanding, we see no grace or
benediction attached to violence. Indeed, its very intensity seems
diminished: we need noise and constant stimulation and quantity to
make up for a lack of significant meaning.
Perhaps this change in attitude and treatment of violence is a
function of our increasingly technologized view of the body and flesh.
We see this view dramatized outside the theater in the practices and
fantasies of "maintenance" and "repair" represented by the "fitness
center" and cosmetic surgery. Inside the theater, we see it dramatized
in the "special effects" allowed by new technological developments and
in an increasingly hyperbolic and quantified treatment of violence and
bodily damage that is as much about "more" as it is about violence. It
seems to me that this quantitative move to "more" in relation to vio¬
lence—more blood, more gore, more characters (they're really not peo¬
ple) blown up or blown away—began with the contemporary horror
film, with "slasher" and "splatter" films that hyperbolized violence
and its victims in terms of quantity rather than through exaggerations
of form. Furthermore, unlike in the "New Hollywood" films of the late
1960s and 1970s (here one thinks of Peckinpah or Penn), excessive vio¬
lence in these "low" genre films, while eliciting screams also elicited
laughter, too much becoming, indeed, "too much": incredible, a "gross-
out," so "outrageous" and "over the top" that ironic reflexivity set in
(for both films and audiences) and the mounting gore and dead bodies
121
Figure 23. In Quentin Tarantino’s Reservoir Dogs, Mr. Orange (Tim Roth) spends the film bleed¬
ing out from a gunshot wound.
became expected—and funny. (Here Scream and its sequel are recent
examples.)
This heightened sense of reflexivity and irony that emerges
from quantities of violence, from "more," is not necessarily progressive
nor does it lead to a "moral" agenda or a critique of violence. (By virtue
of its excesses and its emphasis on quantity and despite his intentions,
Olive Stone's Natural Born Killers is quite ambiguous in this regard.)
Indeed, in its present moment, this heightened reflexivity and irony
merely leads to a heightened sense of representation: that is, care for
the film as experience and text, perhaps, but a lack of any real concern
for the bodies blown away (or up) upon the'screen. In recent "splatter"
films, in Tarantino films like Reservoir Dogs and Pulp Fiction, and in
quite a number of action thrillers, bodies are more carelessly squan¬
dered than carefully stylized. Except, of course, insofar as excess, as
hyperbole, itself constitutes stylization. Thus, most of the violence we
see on the screen today suggests Grand Guignol rather than Jacobean
tragedy. However, in our current cultural moment, tiredly described as
"postmodern" but filled with new forms of violence like "road rage,"
122
Vivian C. Sobchack
the exaggeration and escalating quantification of violence and gore are a

great deal less transgressive than they were—and a great deal more absurd.
Thus, Tarantino has said on various occasions that he doesn't take vio¬
lence "very seriously" and describes it as "funny" and "outrageous."
This hyperbolic escalation and quantification of violence also
has become quite common to the action picture and thriller, where the
body count only exceeds the number of explosions and neither matters
very much to anyone: here violence and the laying waste of bodies
seems more "naturalized": that is, it regularly functions to fill up
screen space and time in lieu of narrative complexity, and to make the
central character look good by "virtue" of his mere survival (see, for
example, Payback). Again, there seems no moral agenda or critique of
violence here—only wisecracks and devalution uttered out of the sides
of a Bruce Willis-type mouth. Indeed, here is the careless violence and
laconic commentary of comic books (where the panels crackle with
zaps and bullets and explosions and the body count is all that counts).
On a more progressive note, I suppose it is possible to see this
new excessive and careless treatment of violence on screen as a satiric
form of what Russian literary theorist Mikhail Bakhtin has called
"grotesque realism." That is, excessive representations of the body and
its messier aspects might be read as containing critical and liberatory
potential-—this, not only because certain social taboos are broken, but
also because these excessive representations of the grotesquerie of
being embodied are less "allegorical" and fantastic than they are exag¬
gerations of concrete conditions in the culture of which they are a part.
In this regard, and particularly relevant to "indie" crime dramas and the
action thriller (a good deal of it science-fictional), much has been writ¬
ten recently about the "crisis of the body" and a related "crisis of mas¬
culinity." Both of these crises are no longer of the Bonnie and Clyde or
Wild Bunch variety: they are far too much inflected and informed by
technological concerns and confusions and a new sense of the body as
a technology, altered by technology, enabled by technology, and dis¬
abled by technology. Indeed, along with the Fordist assembly line and
its increasing production of bodies consumed as they are violently
"wasted" on the screen, comes the production of bodies as both tech¬
nological* subjects and subjected to technology: enhanced and ex¬
tended, but also extinguished by Ouzis, bombs, whatever the latest in
firepower. Thus, we might argue, the excessive violence we see on the
screen, the carelessness and devaluation of mere human flesh, is both
a recognition of the high-tech, powerful, and uncontrollable subjects
we (men, mostly) have become through technology—and an expression
123
of the increasing frustration and rage at what seems a lack of agency

and effectiveness as we have become increasingly controlled by and
subject to technology.
This new quantification of and carelessness toward violence on
the screen also points to other aspects of our contemporary cultural con¬
text. We have come both a long way and not so far from the assassins,
serial killers, and madmen who made their mass presence visibly felt in
the late 1960s and early 1970s. They, like the bodies wasted on the screen,
have proliferated at an increasingly faster and decreasingly surprising rate.
They and the violence that accompanies them are now a common,
omnipresent phenomenon of daily life—so much so that, to an unprece¬
dented degree, we are resigned to living with them in what has become
an increasingly uncivil society. "Senseless" and "random" violence per¬
vades our lives and is barely remarkable or specific any longer—and while
"road rage" and little children killed by the stray bullets of gang bangers
do elicit a moral frisson, for the most part we live in and suspect the
absence of a moral context in this decade of extreme relativism. Violence,
Figure 24. John Travolta and Samuel L. Jackson are buddies and professional killers in Pulp Fic¬
tion, which takes a cartoonish approach to graphic violence, using a character’s exploding head
as the basis for an extended comic sketch.
124
Vivian C. Sobchack
like "shit," happens—worth merely a bumper sticker nod that recon¬

ciles it with a general sense of helplessness (rather than despair).
No longer elevated through balletic treatment or narrative pur¬
pose, violence on the screen is sensed—indeed, appreciated—as sense¬
less. But then so is life under the extremity of such technologized and
uncivil conditions. Indeed, what has been called the "postmodern con¬
dition" might be more accurately thought of as the "postmortem con¬
dition." There's a kind of meta-sensibility at work here: life, death, and
the movies are a "joke" or an "illusion" and everyone's in on it. Vio¬
lence on the screen and in the culture is not related to a moral context,
but to a proliferation of images, texts, and spectacle. And, given that we
cannot contain or stop this careless proliferation, violence and death
both on the street and in Pulp Fiction become reduced to the practical—
and solvable—problem of cleanup.
Pain, too, drops out of the picture. The spasmodic twitching
that ends Bonnie and Clyde has become truly lifeless. The bodies now
subjected to violence are just "dummies": multiple surfaces devoid of
subjectivity and gravity, "straw men," if you will. "Wasting" them
doesn't mean much. Hence, the power (both appealing and off-putting)
of those few films that remind us that bodily damage hurts, that vio¬
lently wasting lives has grave consequences. Hence, the immense pop¬
ularity of Saving Private Ryan, a movie in which the massive quantity
of graphic physical damage and the violent "squandering" of bodies and
lives is "redeemed" to social purpose and meaning, its senselessness
made sensible by its (re)insertion in a clearly defined (and clearly past)
moral context. Hence, also, the popular neglect of Beloved or Affliction,
movies in which violence is represented "close up" as singularly felt:
graphically linked to bodily pain and its destruction of subjectivity. In
these films, violence is not dramatized quantitatively or technologi¬
cally and thus becomes extremely difficult to watch: that is, even
though an image, understood by one's own flesh as real.
I am not sure how to end this particular postmortem on my
original essay. I still can't watch the eyeball being slit in Un Chien
Andalou. But, as with Straw Dogs and The French Connection, I could
and did watch all the violence in Pulp Fiction. Nonetheless, there's
been a qualitative change as well as a quantitative one: while I watched
those earlier violent films compulsively, with some real need to know
what they showed me, I watch the excesses of the current ones casu¬
ally, aware they won't show me anything real that I don't already know.
Carol J. Clover
Her Body, Himself:

Gender in the Slasher Film
On the high side of horror lie the classics: F. W. Murnau's Nosferatu

(1922), King Kong (1933), Dracula (1931), Frankenstein (1931), and
various works by Alfred Hitchcock, Carl Theodor Dreyer, and a few
others—films that by virtue of age, literary ancestry, or fame of direc¬
tor have achieved reputability within the context of disreputability.
Farther down the scale fall the productions of Brian De Palma, some of
the glossier satanic films (Rosemary’s Baby [1968], The Omen [1976],
The Exorcist [1973]), certain science-fiction hybrids (Alien [1979],
Aliens [1986], Blade Runner [1982]), some vampire and werewolf films
[Wolfen [1981], An American Werewolf in London [1981]), and an
assortment of other highly produced films, often with stars (Whatever
Happened to Baby Jane1 [1962], The Shining [1980]). At the very bot¬
tom, down in the cinematic underbrush, lies—horror of horrors—the
slasher (or splatter or shocker) film: the immensely generative story of
a psycho killer who slashes to death a string of mostly female victims,
one by one, until he is himself subdued or killed, usually by the one girl
who has survived.
Drenched in taboo and encroaching vigorously on the porno¬
graphic, the slasher film lies by and large beyond the purview of the
respectable (middle-aged, middle-class) audience. It has also lain by and
large beyond the purview of respectable criticism. Staples of drive-ins
and exploitation houses, where they ''rub shoulders with sex pictures
and macho action flicks," these are films that are "never ever written
up."1 Books on horror film mostly concentrate on the classics, touch
on the middle categories in passing, and either pass over the slasher in
silence or bemoan it as a degenerate aberration.2 The one full book on
Copyright © 1987 by the Regents of the University of California. Reprinted by permis¬

sion from Representations, no. 20, Fall 1987, pp. 187-228.
125
126
Carol J. Clover
the category, William Schoell's Stay Out of the Shower, is immacu¬

lately unintelligent.3 Film magazine articles on the genre rarely get past
technique, special effects, and profits. The Sunday San Francisco Exam¬
iner relegates reviews of slashers to the syndicated "Joe Bob Briggs,
Drive-In Movie Critic of Grapevine, Texas," whose lowbrow, campy
tone ("We're talking two breasts, four quarts of blood, five dead bodies
. . . Joe Bob says check it out") establishes what the paper and others
like it deem thejiecessary distance between their readership and that
sort of film.4 There are of course the exceptional cases: critics or social
observers who have seen at least some of these films and tried to come
to grips with their ethics or aesthetics or both. Just how troubled is their
task can be seen from its divergent results. For one critic, The Texas
Chainsaw Massacre (1974) is "the Gone with the Wind, of meat
movies."5 For another it is a "vile little piece of sick crap . . . nothing
but a hysterically paced, slapdash, imbecile concoction of cannibalism,
voodoo, astrology, sundry hippieesque cults, and unrelenting sadistic
violence as extreme and hideous as a complete lack of imagination can
possibly make it."6 Writes a third: "[Director Tobe] Fiooper's cinematic
intelligence becomes more apparent in every viewing, as one gets over
the initial traumatizing impact and learns to respect the pervasive felic¬
ities of camera placement and movement."7 The Museum of Modern
Art bought the film in the same year that at least one country, Sweden,
banned it.
Robin Wood's tack is less aesthetic than anthropological:
"However one may shrink from systematic exposure to them [slasher
films], however one may deplore the social phenomena and ideological
mutations they reflect, their popularity . . . suggests that even if they
were uniformly execrable they shouldn't be ignored."8 We may go a step
farther and suggest that the qualities that locate the slasher film out¬
side the usual aesthetic system—that indeed render it, along with
pornography and low horror in general, the film category "most likely
to be betrayed by artistic treatment and lavish production values"9—
are the very qualities that make it such a transparent source for
(sub(cultural attitudes toward sex and gender in particular. Unmediated
by otherworldly fantasy, cover plot, bestial transformations, or civi¬
lized routine, slasher films present us in startlingly direct terms with a
world in which male and female are at desperate odds but in which, at
the same time, masculinity and femininity are more states of mind
than body. The premise of this essay, then, is that the slasher film, not
despite but exactly because of its crudity and compulsive repetitive¬
ness, gives us a clearer picture of current sexual attitudes, at least
127
Her Body, Himself: Gender in the Slasher Film
among the segment of the population that forms its erstwhile audience,
than do the legitimate products of the better studios.
Before we turn to the generic particulars, however, let us review
some of the critical and cinematic issues that attend the study of the
sensation genres in general and horror in particular. We take as our
point of departure not a slasher film but Brian de Palma's art-horror film
Body Double (1984). The plot—a man witnesses and after much strug¬
gle solves the mysterious murder of a woman with whom he has
become voyeuristically involved—concerns us less than the three
career levels through which the hero, an actor named Jake, first ascends
and then descends. He aspires initially to legitimate roles (Shake¬
speare), but it becomes clear during the course of a method-acting class
that his range of emotional expression is impaired by an unresolved
childhood fear. For the moment he has taken a job as vampire in a "low-
budget, independent horror film," but even that job is threatened when,
during a scene in which he is to be closed in a coffin and buried, he suf¬
fers an attack of claustrophobia and must leave the set. A plot twist
leads him to the underworld of pornography, where he takes on yet
another role, this time in a skin flick. Here, in the realm of the flesh
with a queen of porn, the sexual roots of Jake's paralysis—fear of the
(female) cavern—are exposed and finally resolved. A new man, he
returns to A Vampire’s Kiss to master the burial scene, and we are to
understand that Shakespeare is the next stop.
The three cinematic categories are thus ranked by degree of
sublimation. On the civilized side of the continuum lie the legitimate
genres,- at the other end, hard on the unconscious, lie the sensation or
"body" genres, horror and pornography, in that order. For De Palma, the
violence of horror reduces to and enacts archaic sexual feelings.
Beneath Jake's emotional paralysis (which emerges in the "high" genre)
lies a death anxiety (which is exposed in the burying-alive of horror),
and beneath that anxiety lies a primitive sexual response (which
emerges, and is resolved, in pornography). The layers of Jake's experi¬
ence accord strikingly, and perhaps not Coincidentally, with Freud's
archaeology of "uncanny" feelings. "To some people," Freud wrote,
"the idea of being buried alive by mistake is the most uncanny thing of
all. And yet psychoanalysis has taught us that this terrifying phantasy
is only a transformation of another phantasy which originally had noth¬
ing terrifying about it at all, but was qualified by a certain lascivious¬
ness—the phantasy, I mean, of intrauterine existence [der Phantasie
vom Leben im Mutterleib}."10 Pornography thus engages directly (in
pleasurable terms) what horror explores at one remove (in painful
128
Carol J. Clover
terms) and legitimate film at two or more. Beneath the "legitimate"

plot of The Graduate (1967), in which Ben must give up his relation¬
ship with a friend's mother in order to marry and take his proper social
place, lies the plot of Psycho (1960), in which Norman's unnatural
attachment to his own mother drives him to murder women to whom
he is attracted; and beneath that plot lies the plot of the porn film Taboo
(1980), in which the son simply has sex with his mother ("Mom, am I
better than Dad?"). Pornography, in short, has to do with sex (the act)
and horror with gender.
It is a rare Hollywood film that does not devote a passage or
two—a car chase, a sex scene—to the emotional/physical excitement
of the audience. But horror and pornography, are the only two genres
specifically devoted to the arousal of bodily sensation. They exist solely
to horrify and stimulate, not always respectively, and their ability to do
so is the sole measure of their success: they "prove themselves upon
our pulses."11 Thus in horror-film circles, "good" means scary, specif¬
ically in a bodily way (ads promise shivers, chills, shudders, tingling of
the spine,- Lloyds of London insured audiences of Macabre [1958]
against death by fright);12 and Hustler’s Erotic Film Guide ranks porno¬
graphic films according to the degree of erection they produce (one film
is ranked a "pecker popper," another "limp"). The target is in both cases
Figure 25. joe Spinell plays a serial killer in Maniac, which devotes lingering and graphic atten¬
tion to the savagery he inflicts on his victims. The special effects makeup by Tom Savini details
gunshot wounds to the head, garrotings, and scalping. No manner of death was now off-limits for
the commercial cinema.
129
the body, our witnessing body. But what we witness is also the body,
another's body, in experience: the body in sex and the body in threat.
The terms flesh film (skin flicks) and meat movies are remarkably apt.
Cinema, it is claimed, owes its particular success in the sensa¬
tion genres (witness the early and swift rise of vampire films) to its
unprecedented ability to manipulate point of view. What written nar¬
rative must announce, film can accomplish silently and instanta¬
neously through cutting. Within the space of seconds, the vampire's
first-person perspective is displaced by third-person or documentary
observation. To these simple shifts can be added the variables of dis¬
tance (from the panorama of the battlefield to the close-up of an eye¬
ball), angle, frame tilt, lighting effects, unsteadiness of image, and so
on—again, all subject to sudden and unannounced manipulation.13 Fri¬
day the 13th (1980) locates the I-camera with the killer in pursuit of a
victim; the camera is hand- held, producing a jerky image, and the frame
includes in-and-out-of-focus foreground objects (trees, bushes, window
frames) behind which the killer (I-camera), is lurking—all accompanied
by the sound of heartbeats and heavy breathing. "The camera moves in
on the screaming, pleading victim, 'looks down' at the knife, and then
plunges it into the chest, ear, or eyeball. Now that's sick."14
Lagging behind practice is a theoretical understanding of effect.
The processes by which a certain image (but not another) filmed in a
certain way (but not another) causes one person's (but not another's)
pulse to race finally remains a mystery—not only to critics and theo¬
rists but even, to judge from interviews and the trial-and-error (and
baldly imitative) quality of the films themselves, by the people who
make the product. The process of suture is sensed to be centrally impor¬
tant in effecting audience identification, though just how and why is
unclear.15 Nor is identification the straightforward notion some critics
take it to be.16 Where commentators by and large agree is in the impor¬
tance of the "play of pronoun function."17 If the fantastic depends for
its effect on an uncertainty of vision, a profusion of perspectives, and a
confusion of subjective and objective, then cinema is preeminently
suited to the fantastic. Indeed, to the extent that film can present
"unreal" combinations of objects and events as "real" through the cam¬
era eye, the "cinematic process itself anight be called fantastic."18 The
"cinefantastic" in any case succeeds, far more efficiently and effec¬
tively and on a far greater scale than its ancestral media, in the pro¬
duction of sensation.
The fact that the cinematic conventions of horror are so easily
130
Carol J. Clover
and so often parodied would seem to suggest that, individual variation

notwithstanding, its basic structures of apperception are fixed and fun¬
damental. The same is true of the stories they tell. Students of folklore
or early literature recognize in the slasher film the hallmarks of oral
story: the free exchange of themes and motifs, the archetypal charac¬
ters and situations, the accumulation of sequels, remakes, imitations.
This is a field in which there is in some sense no original, no real or
right text, but only variants,- a world in which, therefore, the meaning
of the individual example lies outside itself. The "art" of the horror
film, like the "art" of pornography, is to a very large extent the art of
rendition, and it is understood as such by the competent audience.19 A
particular example may have original features, but its quality as a hor¬
ror film lies in the ways it delivers the cliche. James B. Twitchell rightly
recommends an "ethnological approach, in which the various stories
are analyzed as if no one individual telling really mattered. . . . You
search for what is stable and repeated; you neglect what is 'artistic' and
'original.' This is why, for me, auteur criticism is quite beside the point
in explaining horror. . . . The critic's first job in explaining the fascina¬
tion of horror is not to fix the images at their every appearance but,
instead, to trace their migrations to the audience and, only then, try to
understand why they have been crucial enough to pass along."20 That
auteur criticism is at least partly beside the point is clear from inter¬
views with such figures as John Carpenter (Halloween [1978], The Fog
[1980])—interviews that would seem to suggest that, like the purvey¬
ors of folklore, the makers of film operate more on instinct and formula
than conscious understanding. So bewildered was Hitchcock by the
unprecedented success of Psycho that he approached the Stanford
Research Institute about doing a study of the phenomenon.21
What makes horror "crucial enough to pass along" is, for critics
since Freud, what has made ghost stories and fairy tales crucial enough
to pass along: its engagement of repressed fears and desires and its reen¬
actment of the residual conflict surrounding those feelings. Horror films
thus respond to interpretation, as Robin Wood puts it, as "at once the
personal dreams of their makers and the collective dreams of their audi¬
ences—the fusion made possible by the shared structures of a common
ideology."22 And just as attacker and attacked are expressions of the
same self in nightmares, so they are expressions of the same viewer in
horror film. Our primary and acknowledged identification may be with
the victim, the adumbration of our infantile fears and desires, our mem-
131
ory sense of ourselves as tiny and vulnerable in the face of the enormous
Other; but the Other is also finally another part of ourselves, the pro¬
jection of our repressed infantile rage and desire (our blind drive to anni¬
hilate those toward whom we feel anger, to force satisfaction from those
who stimulate us, to wrench food for ourselves if only by actually
devouring those who feed us) that we have had in the name of civiliza¬
tion to repudiate. We are both Red"Riding Hood and the Wolf; the force
of the experience, the horror, comes from "knowing" both sides of the
story—from giving ourselves over to the cinematic play of pronoun
functions. It is no surprise that the first film to which viewers were not
admitted once the theater darkened was Psycho. Whether Hitchcock
actually meant with this measure to intensify the "sleep" experience is
unclear, but the effect both in the short run, in establishing Psycho as
the ultimate thriller, and in the long run, in altering the cinema-going
habits of the nation, is indisputable. In the current understanding, hor¬
ror is the least interruptable of all film genres. That uninterruptability
itself bears witness to the compulsive nature of the stories it tells.
Whatever else it may be, the slasher film is clearly "crucial
enough to pass along." Profits and sequels tell much of the story. Hal¬
loween cost $320,000 to make and within six years had grossed over
$75 million,- even a highly produced film like The Shining has repaid
itself tenfold.23 The Hills Have Eyes (1977), The Texas Chainsaw Mas¬
sacre, and Alien (a science-fiction/slasher hybrid) are [as of 1987] at Part
Two. Psycho and A Nightmare on Elm Street are at Part Three. Hal¬
loween is at Part Four, and Friday the 13th is at Part Six. These are bet¬
ter taken as remakes than sequels; although the subsequent part
purports to take up where the earlier part left off, it in most cases sim¬
ply duplicates with only slight variation the plot and circumstances—
the formula—of its predecessor. Nor do different titles indicate
different plots,- Friday the 13th is set at summer camp and Halloween
in town, but the story is much the same, compulsively repeated in
those ten films and in dozens like them under different names. The
audience for that story is by all accounts largely young and largely
male—most conspicuously groups of boys who cheer the killer on as he
assaults the victims, then reverse their sympathies to cheer the sur¬
vivor on as she assaults the killef.24 Our question, then, has to do with
that particular audience's stake in that particular nightmare, with what
in the story is "crucial" enough to warrant the price of admission and
what the implications are for the current discussion of women and film.
132
Carol J. Clover
The Slasher Film
The immediate ancestor of the slasher film is Hitchcock's Psycho

(1960). Its elements are familiar: the killer is the psychotic product of a
sick family but still recognizably human; the victim is a beautiful, sex¬
ually active woman; the location is not-home, at a Terrible Place,- the
weapon is something other than a gun; the attack is registered from the
victim's point of view and comes with shocking suddenness. None of
these features is original, but the unprecedented success of Hitchcock's
particular formulation, above all the sexualization of both motive and
action, prompted a flood of imitations and variations. In 1974, a film
emerged that revised the Psycho template to a degree and in such a way
as to mark a new phase: The Texas Chainsaw Massacre, directed by
Tobe Hooper. Together with Halloween, it engendered a new spate of
variations and imitations.
The plot of The Texas Chainsaw Massacre is simple enough:
five young people are driving through Texas in a van,- they stop off at an
abandoned house and are murdered one by one by the psychotic sons of
a degenerate local family; the sole survivor is a woman. The horror, of
course, lies in the elaboration. Early in the film the group picks up a
hitchhiker, but when he starts a fire and slashes Franklin's arm (having
already slit open his own hand), they kick him out. The abandoned
house they subsequently visit, once the home of Sally's and Franklin's
Figure 26. Suddenly looming into view, Leatherface claims his first victim in The Texas Chain¬
saw Massacre.
133
grandparents, turns out to be right next door to the house of the hitch¬
hiker and his family: his brother Leatherface; their father; an aged and
only marginally alive grandfather; and their dead grandmother and her
dog, whose mummified corpses are ceremonially included in the fam¬
ily gatherings. Three generations of slaughterhouse workers, once
proud of their craft but now displaced by machines, have taken up
killing and cannibalism as a way of life. Their house is grotesquely dec¬
orated with human and animal remains—bones, feathers, hair, skins.
The young people drift apart in their exploration of the abandoned
house and grounds and are picked off one by one by Leatherface and
Hitchhiker. Last is Sally. The others are attacked and killed with dis¬
patch, but Sally must fight for her life, enduring all manner of horrors
through the night. At dawn she manages to escape to the highway,
where she is picked up by a passing trucker.
Likewise, the nutshell plot of Halloween: a psychotic killer
(Michael) stalks a small town on Halloween and kills a string of teenage
friends, one by one; only Laurie survives. The twist here is that Michael
has escaped from the asylum in which he has been incarcerated since
the age of 6, when he killed his sister minutes after she and her
boyfriend parted following an illicit interlude in her parents' bed. That
murder, in flashback, opens the film. It is related entirely in the killer's
first person (I-camera) and only after the fact is the identity of the per¬
petrator revealed. Fifteen years later, Michael escapes his prison and
returns to kill Laurie, whom he construes as another version of his sis¬
ter (a sequel clarifies that she is in fact his younger sister, adopted by
another family at the time of the earlier tragedy). But before Michael
gets to Laurie, he picks off her high school friends: Annie, in a car on
her way to her boyfriend's; Bob, going to the kitchen for a beer after sex
with Lynda; Lynda, talking on the phone with Laurie and waiting for
Bob to come back with the beer. At last only Laurie remains. When she
hears Lynda squeal and then go silent on the phone, she leaves her own
babysitting house to go to Lynda's. Here she discovers the three bodies
and flees, the killer in pursuit. The remainder of the film is devoted to
the back-and-forth struggle between Laurie and Michael. Again and
again he bears down on her, and again and again she either eludes him
(by running, hiding, breaking through windows to escape, locking her¬
self in) or strikes back (once with a knitting needle, once with a hanger).
In the end, Dr. Loomis (Michael's psychiatrist in the asylum) rushes in
and shoots the killer (although not so fatally as to prevent his return in
the sequels).
134
Carol J. Clover
Before we turn to an inventory of generic components, let us

add a third, more recent example: The Texas Chainsaw Massacre, Part
2 (1986). The slaughterhouse family (now named the Sawyers) is the
same, although older and, owing to their unprecedented success in the
sausage business, richer.25 When Mr. Sawyer begins to suspect from her
broadcasts that a disk jockey named Stretch knows more than she
should about one of their recent crimes, he dispatches his sons Leather-
face and Chop Top (Hitchhiker in the earlier film) to the radio station
late at night. There they seize the technician and corner Stretch. At the
crucial moment, however, power fails Leatherface's chainsaw. As
Stretch cowers before him, he presses the now still blade up along her
thigh and against her crotch, where he holds it unsteadily as he jerks
and shudders in what we understand to be orgasm. After that the sons
leave. The intrepid Stretch, later joined by a Texas Ranger (Dennis Hop¬
per), tracks them to their underground lair outside town. Tumbling
down the Texas equivalent of a rabbit hole, Stretch finds herself in the
subterranean chambers of the Sawyer operation. Here, amid all the
slaughterhouse paraphernalia, the Sawyers live and work. The walls
drip with blood. Like the decrepit mansion of the first film, the resi¬
dential parts of the establishment are quaintly decorated with human
and animal remains. After a long ordeal at the hands of the Sawyers,
Stretch manages to scramble up through a culvert and beyond that up
onto a nearby pinnacle, where she finds a chainsaw and wards off her
final assailant. The Texas Ranger evidently perishes in a grenade explo¬
sion underground, leaving Stretch the sole survivor.
The spiritual debt of all the post-1974 slasher films to Psycho
is clear, and it is a rare example that does not pay a visual tribute, how¬
ever brief, to the ancestor—if not in a shower stabbing, then in a purl¬
ing drain or the shadow of a knife-wielding hand. No less clear,
however, is the fact that the post-1974 examples have, in the usual way
of folklore, contemporized not only Hitchcock's terms but also, over
time, their own. We have, in short, a cinematic formula with a twenty-
six-year history, of which the first phase, from 1960 to 1974, is domi¬
nated by a film clearly rooted in the sensibility of the 1950s, while the
second phase, bracketed by the two Texas Chainsaw films of 1974 and
1986, responds to the values of the late 1960s and early 1970s. That the
formula in its most recent guise may be in decline is suggested by the
campy, self-parodying quality of The Texas Chainsaw Massacre, Part 2,
as well as the emergence, in legitimate theater, of the slasher satire
Buckets of Blood (1991). Between 1974 and 1986, however, the formula
135
evolved and flourished in ways of some interest to observers of popular

culture, above all those concerned with the representation of women
in film. To apprehend in specific terms the nature of that mutation,
let us, with Psycho as the benchmark, survey the genre by component
category: killer, locale, weapons, victims, and shock effects.
Killer
The psychiatrist at the end of Psycho explains what we had already
guessed from the action: that Norman Bates had introjected his mother,
in life a "clinging, demanding woman," so completely that she consti¬
tuted his other, controlling self. Not Norman but "the mother half of
his mind" killed Marion—had to kill Marion—when he (the Norman
half) found himself aroused by her. The notion of a killer propelled by
psychosexual fury, more particularly a male in gender distress, has
proved a durable one, and the progeny of Norman Bates stalk the genre
up to the present day. Just as Norman wears his mother's clothes dur¬
ing his acts of violence and is thought, by the screen characters and also,
for a while, by the film's spectators to be his mother, so the murderer
in the Psycho imitation Dressed to Kill (1980), a transvestite psychia¬
trist, seems until his unveiling to be a woman,- like Norman, he must
kill women who arouse him sexually. Likewise, in muted form, Hitch¬
hiker/Chop Top and Leatherface in the Texas Chainsaw films: neither
brother shows overt signs of gender confusion, but their cathexis to the
sick family—in which the mother is conspicuously absent but the pre¬
served corpse of the grandmother (answering the treated body of Mrs.
Bates in Psycho) is conspicuously present—has palpably arrested their
development. Both are in their 20s (30s, in the later film), but Hitch¬
hiker/Chop Top seems a gangly kid and Leatherface jiggles in baby fat
behind his butcher's apron. Like Norman Bates, whose bedroom dis¬
plays his childhood toys, Hitchhiker/Chop Top and Leatherface are per¬
manently locked in childhood. Only when Leatherface "discovers" sex
in The Texas Chainsaw Massacre, Part 2 does he lose his appetite for
murder. In Motel Hell (1980), a send-up of modern horror with special
reference to Psycho and The Texas Chainsaw Massacre, we are repeat¬
edly confronted with a portrait of the dead mother, silently presiding
over all manner of cannibalistic and incestuous doings on the part of
her adult children.
No less in the grip of boyhood is the killer in The Eyes of Laura
Mars (1978). The son of a hooker, a hysterical woman gone for days at
136
Carol J. Clover
a time, the killer has up to now put his boyish anger to good use in
police work—the film makes much of the irony—but the sight of
Laura's violent photographs causes it to be unleashed in full force. The
killer in Hell Night (1981) is the sole member of his family to survive,
as a child, a murderous rampage on the part of his father; the experience
condemned him to an afterlife as a murderer himself. In Halloween the
killer is a child, at least in the first instance: Michael, at the age of 6, is
so enraged at his sister (evidently for her sexual relations with her
boyfriend) that he stabs her to death with a kitchen knife. The remain¬
der of the film details his return rampage at the age of 21, and Dr.
Loomis, who has overseen the case in the interim, explains that
although Michael's body has attained maturity, his mind remains
frozen in infantile fury. In It’s Alive (1974), the killer is literally an
infant, evidently made monstrous through intrauterine apprehension
of its parents' ambivalence (early in the pregnancy they considered an
abortion).
Even killers whose childhood is not immediately at issue and
who display no overt gender confusion are often sexually disturbed. The
murderer in A Nightmare on Elm Street is an undead child molester.
The killer in Slumber Party Massacre (1982) says to a young woman he
is about to assault with a power drill: "Pretty. All of you are very pretty.
I love you. Takes a lot of love for a person to do this. You know you want
it. You want it. Yes." When she grasps the psychodynamics of the situ¬
ation in the infamous crotch episode of Texas Chainsaw, Part 2, Stretch
tries a desperate gambit: "You're really good, you really are good," she
repeats; indeed, immediately after ejaculation Leatherface becomes pal¬
pably less interested in his saw. The parodic Motel Hell spells it out.
"His pecker don't work; you see when he takes off his overalls—it's like
a shriveled prune," Bruce says of his killer-brother Vincent when he
learns of Terry's plans to marry him. Terry never does see, for on her
wedding night he attempts (needless to say) not sex but murder. Actual
rape is practically nonexistent in the slasher film, evidently on the
premise— as the crotch episode suggests—that violence and sex are not
concomitants but alternatives, the one as much a substitute for and a
prelude to the other as the teenage horror film is a substitute for and a
prelude to the "adult" film (or the meat movie a substitute for and prel¬
ude to the skin flick).26 When Sally under torture (The Texas Chainsaw
Massacre) cries out, "I'll do anything you want," clearly with sexual
intention, her assailants respond only by mimicking her in gross terms;
she has profoundly misunderstood the psychology.
137
>
Female killers are few and their reasons for killing significantly
different from men's. With the possible exception of the murderous
mother in Friday the 13th, they show no gender confusion. Nor is their
motive overtly psychosexual; their anger derives in most cases not from
childhood experience but from specific moments in their adult lives in
which they have been abandoned or cheated on by men (Strait-jacket
[1964], Play Misty for Me [1971], Attack of the 50-Foot Woman [1958]).
(Films like Mother’s Day [1980], Ms. 45 [1980], and I Spit on Your Grave
[1978] belong to the rape-revenge category.) Friday the 13th is some¬
thing of an anomaly. The killer is revealed as a middle-aged woman
whose son, Jason, drowned years earlier as a consequence of negligence
on the part of the camp counselors. The anomaly is not sustained in the
sequels (Parts Two to Six), however. Here the killer is Jason himself, not
dead after all but living in a forest hut. The pattern is a familiar one; his
motive is vengeance for the death of his mother, his excessive attach¬
ment toward whom is manifested in his enshrining of her severed head.
Like Stretch in the crotch episode of Texas Chainsaw, Part 2, the girl
who does final combat with Jason in Friday the 13th Part 2 (1981) sees
the shrine, grasps its significance (she's a psych major), and saves her¬
self by repeating in a commanding tone, "I am your mother, Jason; put
down the knife." Jason, for his part, begins to see his mother in the girl
(I-camera) and obeys her.
In films of the Psycho type (Dressed to Kill, The Eyes of Laura
Mars), the killer is an insider, a man who functions normally in the
action until, at the end, his other self is revealed. The Texas Chainsaw
Massacre and Halloween introduced another sort of killer: one whose
only role is that of killer and one whose identity as such is clear from
the outset. Norman may have a normal half, but these killers have
none. They are emphatic misfits and emphatic outsiders. Michael is an
escapee from a distant asylum,- Jason subsists in the forest; the Sawyer
sons live a bloody subterranean existence outside town. Nor are they
clearly seen. We catch sight of them only in glimpses—few and far
between in the beginning, more frequent toward the end. They are usu¬
ally large, sometimes overweight, and often masked. In short, they may
be recognizably human, but only marginally so, just as they are only
marginally visible—to their victims and to us, the spectators. In one
key aspect, however, the killers are superhuman: their virtual inde¬
structibility. Just as Michael (in Halloween) repeatedly rises from blows
that would stop a lesser man, so Jason (in the Friday the 13th films) sur¬
vives assault after assault to return in sequel after sequel. Chop Top in
138
Carol J. Clover
Texas Chainsaw, Part 2, is so called because of a metal plate implanted

in his skull in repair of a head wound sustained in the truck accident
in the earlier film. It is worth noting that the killers are normally the
fixed elements and the victims the changeable ones in any given series.
Terrible Place
The Terrible Place, most often a house or tunnel in which the victims
sooner or later find themselves, is a venerable element of horror. The
Bates mansion is just one in a long list of such places—a list that con¬
tinues, in the modern slasher, with the decaying mansion of The Texas
Chainsaw Massacre, the abandoned and haunted mansion of Hell
Night, the house for sale but unsellable in Halloween (also a point of
departure for such films as Rosemary’s Baby and The Amityville Hor¬
ror [1979]), and so on. What makes these houses terrible is not just their
Victorian decrepitude but the terrible families—murderous, incestu¬
ous, cannibalistic—that occupy them. So the Bates mansion enfolds the
history of a mother and son locked in a sick attachment, and so the
mansion/labyrinth of the Texas Chainsaw movies shelters a lawless
brood presided over by the decaying corpse of the grandmother. Jason's
forest hut (in the Friday the 13th sequels) is no mansion, but it houses
another mummified mother (or at least her head), with all the usual
candles and dreadful paraphernalia. The terrors of the Hell Night man¬
sion stem, we learn, from an early owner's massacre of his children. Into
such houses unwitting victims wander in film after film, and it is the
conventional task of the genre to register in close detail those victims'
dawning understanding, as they survey the visible evidence, of the
human crimes and perversions that have transpired there. That per¬
ception leads directly to the perception of their own immediate peril.
In The Texas Chainsaw Massacre, Part 2, house and tunnel
elide in a residential labyrinth underground, connected to the world
above by channels and culverts. The family is intact, indeed it thrives,
but for reasons evidently having to do with the nature of their sausage
business has moved residence and slaughterhouse underground. For
Stretch, trying desperately to find a way out, it is a ghastly place: dark,
full of blind alleys, walls wet with blood. Likewise the second basement
of the haunted mansion in Hell Night: strewn with decaying bodies and
skeletons, lighted with masses of candles. Other tunnels are less famil¬
ial: the one in Body Double that prompts Jack's claustrophobic faint,
and the horror-house tunnel in He Knows You’re Alone (1980) in which
the killer lurks. The morgue episode in the latter film, certain of the
139
hospital scenes in Halloween II (1981), and the bottom-cellar scenes

from various films may be counted as Terrible Tunnels: dark,
labyrinthine, exitless, usually underground and palpably damp, and
laced with heating ducts and plumbing pipes. In Hell Night, as in Texas
Chainsaw, Part 2, Terrible House (the abandoned mansion) and Terri¬
ble Tunnel (the second basement) elide.
The house or tunnel may at first seem a safe haven, but the
same walls that promise to keep the killer out quickly become, once
the killer penetrates them, the walls that hold the victim in. A phe¬
nomenally popular moment in post-1974 slashers is the scene in which
the victim locks herself in (a house, room, closet, car) and waits with
pounding heart as the killer slashes, hacks, or drills his way in. The
action is inevitably seen from the victim's point of view,- we stare at the
door (wall, car roof) and watch the surface break with first the tip and
then the shaft of the weapon. In Hitchcock's The Birds (1963), it is the
birds' beaks we see penetrating the door. The penetration scene is com¬
monly the film's pivot moment; if the victim has up to now simply fled,
she has at this point no choice but to fight back.
Weapons
In the hands of the killer, at least, guns have no place in slasher films.
Victims sometimes avail themselves of firearms, but like telephones,
fire alarms, elevators, doorbells, and car engines, guns fail in the
squeeze. In some basic sense, the emotional terrain of the slasher film
is pretechnological. The preferred weapons of the killer are knives,
hammers, axes, ice picks, hypodermic needles, red-hot pokers, pitch-
forks, and the like. Such implements serve well a plot predicated on
stealth, the unawareness of later victims that the bodies of their friends
are accumulating just yards away. But the use of noisy chainsaws and
power drills and the nonuse of such relatively silent means as bow and
arrow, spear, catapult, and even swords would seem to suggest that
closeness and tactility are also at issue.27 The sense is clearer if we
include marginal examples like Jaws (1975) and The Birds, as well as
related werewolf and vampire genres. Knives and needles, like teeth,
beaks, fangs, and claws, are personal, extensions of the body that bring
attacker and attacked into animalistic embrace.28 In I Spit on Your
Grave, the heroine forces a rapist at gunpoint to drop his pants, evi¬
dently meaning to shoot him in genitals. But she changes her mind,
invites him home for what he all too readily supposes will be a volun¬
tary follow-up of the earlier gang rape. Then, as they sit together in a
140
Carol J. Clover
bubble bath, she castrates him with a knife. If we wondered why she
threw away the pistol, now we know: all phallic symbols are not equal,
and a hands-on knifing answers a hands-on rape in a way that a shoot¬
ing, even a shooting preceded by a humiliation, does not.29
Beyond that, the slasher evinces a fascination with flesh or
meat itself as that which is hidden from view. When the hitchhiker in
The Texas Chainsaw Massacre slits open his hand for the thrill, the
onlookers recoil in horror-all but Franklin, who seems fascinated by the
realization that all that lies between the visible, knowable outside of
the body and its secret insides is one thin membrane, protected only by
a collective taboo against its violation. It is no surprise that the rise of
the slasher film is concomitant with the development of special effects
that let us see with our own eyes the "opened" body.
Victims
Where once there was one victim, Marion Crane, there are now many:
five in The Texas Chainsaw Massacre, four in Halloween, fourteen in
Friday the 13th Part III (1982), and so on. (As Schoell puts it, "Other
filmmakers figured that the only thing better than one beautiful
woman being gruesomely murdered was a whole series of beautiful
women being gruesomely murdered.")30 Where once the victim was an
adult, now she is typically in her teens (hence the term teenie-kill pic).
Where once she was female, now she is both girl and boy, although most
often and most conspicuously girl. For all this, her essential quality
remains the same. Marion is first and foremost a sexual transgressor.
The first scenes show her in a hotel room dressing at the end of a lunch
hour, asking her lover to marry her. It is, of course, her wish to be made
an honest woman that leads her to abscond with $40,000, an act that
leads her to the Bates motel in Fairvale. FFere, just as we watched her
dress in the opening sequences, we now watch her undress. Moments
later, nude in the shower, she dies. A classic publicity poster for Psy¬
cho shows Janet Leigh with a slightly uncomprehending look on her
face sitting on the bed, dressed in a bra and half-slip, looking backward
in such a way as to outline her breasts. If it is the task of promotional
materials to state in one image the essence of a film, those breasts are
what Psycho is all about.
In the slasher film, sexual transgressors of both sexes are sched¬
uled for early destruction. The genre is studded with couples trying to
find a place beyond purview of parents and employers where they can
have sex, and immediately afterward (or during) being killed. The
141
theme enters the tradition with the Lynda-Bob subplot of Halloween.

Finding themselves alone in a neighborhood house, Lynda and Bob
make hasty use of the master bedroom. Afterward, Bob goes downstairs
for a beer. In the kitchen he is silently dispatched by the killer, Michael,
who then covers himself with a sheet (it's Halloween), dons Bob's
glasses, and goes upstairs. Supposing the bespectacled ghost in the door¬
way to be Bob, Lynda jokes, bares her breasts provocatively, and finally,
in irritation at "Bob's" stony silence, dials Laurie on the phone. Now
the killer advances, strangling her with the telephone cord, so that what
Laurie hears on the other end are squeals she takes to be orgasmic. Hal¬
loween II takes the scene a step farther. Here the victims are a nurse
and orderly who have sneaked off for sex in the hospital therapy pool.
The watching killer, Michael again, turns up the thermostat and, when
the orderly goes to check it, kills him. Michael then approaches the
nurse from behind (she thinks it's the orderly) and strokes her neck.
Only when he moves his hand toward her bare breast and she turns
around and sees him does he kill her.
Other directors are less fond than John Carpenter of the
mistaken-identity twist. Denise, the English vamp in Hell Night, is sim¬
ply stabbed to death in bed during Seth's postcoital trip to the bathroom.
In He Knows You’re Alone, the student having the affair with her pro¬
fessor is stabbed to death in bed while the professor is downstairs chang¬
ing a fuse,- the professor himself is stabbed when he returns and discovers
the body. The postcoital death scene is a staple of the Friday the 13th
series. Part Three offers a particularly horrible variant. Invigorated by
sex, the boy is struck by a gymnastic impulse and begins walking on
his hands,- the killer slices down on his crotch with a machete. Unaware
of the fate of her boyfriend, the girl crawls into a hammock after her
shower; the killer impales her from below.31 Brian De Palma's Dressed
to Kill presents the infamous example of the sexually desperate wife,
first seen masturbating in her morning shower during the credit
sequence, who lets herself be picked up later that day in a museum by
a man with whom she has sex first in a taxi and later in his apartment.
On leaving his place in the evening, she is suddenly attacked and killed
in the elevator. The cause-and-effect relationship between (illicit) sex
and death could hardly be more clearly drawn. All of the killings in
Cruising (1980) occur during (homo)sexual encounters; the difference
here is that the killer is one of the participants, not a third party.
Killing those who seek or engage in unauthorized sex amounts
to a generic imperative of the slasher film. It is an imperative that
crosses gender lines, affecting males as well as females. The numbers
142
Carol J. Clover
are not equal, and the scenes not equally charged; but the fact remains
that in most slasher films after 1978 (following Halloween), men and
boys who go after "wrong" sex also die. This is not the only way males
die,- they also die incidentally, as girls do, when they get in the killer's
way or try to stop him, or when they stray into proscribed territory. The
victims in Hell Night and in the Texas Chainsaw Massacre and Friday
the 13th films are, respectively, those who trespass in Garth Manor,
those who stumble into the environs of the slaughterhouse family, and
those who become counselors at a cursed camp, all without regard to
sex. Boys die, in short, not because they are boys but because they make
mistakes.
Some girls die for the same mistakes. Others, however, and
always the main one, die—plot after plot develops the motive—because
they are female. Just as Norman Bates's oedipal psychosis is such that
only female victims will do, so Michael's sexual anger toward his sis¬
ter (in the Halloween series) drives him to kill her—and after her a
string of sister surrogates. In much the same way, the transsexual psy¬
chiatrist in Dressed to Kill is driven to murder only those women who
arouse him and remind him of his hated maleness. In The Eyes of Laura
Mars, the killer's hatred of his mother drives him to prey on women
specifically—and, significantly, one gay male. He Knows You’re Alone
features a killer who in consequence of an earlier jilting preys exclu¬
sively on brides-to-be. But even in films in which males and females are
killed in roughly even numbers, the lingering images are inevitably
female. The death of a male is always swift; even if the victim grasps
what is happening to him, he has no time to react or register terror. He
is dispatched and the camera moves on. The death of a male is more¬
over more likely than the death of a female to be viewed from a dis¬
tance, or viewed only dimly (because of darkness or fog, for example),
or indeed to happen offscreen and not be viewed at all. The murders of
women, on the other hand, are filmed at closer range, in more graphic
detail, and at greater length.
The pair of murders at the therapy pool in Halloween II illus¬
trates the standard iconography. We see the orderly killed in two shots:
the first at close range in the control room, just before the stabbing, and
the second as he is being stabbed, through the vapors in a medium-long-
shot; the orderly never even sees his assailant. The nurse's death, on the
other hand, is shot entirely in medium close-up. The camera studies her
face as it registers first her unwitting complicity (as the killer strokes
her neck and shoulders from behind), then apprehension, and then, as
143
she faces him, terror; we see the knife plunge into her repeatedly, hear
her cries, and watch her blood fill the therapy pool. This cinematic stan¬
dard has a venerable history, and it remains intact in the slasher film.
Indeed, "tits and a scream" are all that is required of actresses audi¬
tioning for the role of victim in Co-Ed Frenzy, the Active slasher film
whose making constitutes the frame story of Blow Out (1981). It is
worth noting that none of the actresses auditioning has both in the
desired amount and that the director must resort to the use of doubles:
one for the tits, one for the screams.
Final Girl
The image of the distressed female most likely to linger in memory is
the image of the one who did not die: the survivor, or Final Girl. She is
the one who encounters the mutilated bodies of her friends and per¬
ceives the full extent of the preceding horror and of her own peril; who
is chased, cornered, wounded; whom we see scream, stagger, fall, rise,
and scream again. She is abject terror personified. If her friends knew
they were about to die only seconds before the event, the Final Girl lives
with the knowledge for long minutes or hours. She alone looks death
in the face,- but she alone also finds the strength either to stay the killer
long enough to be rescued (ending A) or to kill him herself (ending B).
She is inevitably female. In Schoell's words: "The vast majority of con¬
temporary shockers, whether in the sexist mold or not, feature cli¬
maxes in which the women fight back against their attackers—the
wandering, humorless psychos who populate these films. They often
show more courage and levelheadedness than their cringing male coun¬
terparts."32 Her scene occupies the last ten to twenty minutes (thirty
in the case of The Texas Chainsaw Massacre) and constitutes the film's
emphatic climax.
The sequence first appears in full-blown form (ending A) in The
Texas Chainsaw Massacre with Sally's spirited self-defense and eventual
rescue. Her brother and companions' were dispatched suddenly and
uncomprehendingly, one by one, but Sally survives the ninth round—
long enough to see what has become of her fellows and what is in store
for her, long enough to meet and- even dine with the whole slaughter¬
house family, long enough to undergo all manner of torture (including
the ancient grandfather's effort to strike a fatal hammer blow on the tem¬
ple as they bend her over a washtub), and long enough to bolt and rebolt,
be caught and recaught, plead and replead for her life, and eventually
144
Carol J. Clover
escape to the highway. For nearly thirty minutes of screen time—a third
of the film—we watch her shriek, run, flinch, jump through windows,
and sustain injury and mutilation. Her will to survive is astonishing; in
the end, bloody and staggering, she finds the highway, Leatherface and
Hitchhiker in pursuit. Just as they bear down on her, a truck comes up
and crushes Hitchhiker. Minutes later a pickup driver plucks Sally up
and saves her from Leatherface. The final shots show us Leatherface
from her point of view (the bed of the pickup): standing on the highway,
wounded (having gashed open his abdomen during the truck episode)
but upright, waving the chainsaw crazily over his head.
Halloween’s Final Girl is Laurie. Her desperate defense is
shorter in duration than Sally's but no less fraught with horror. Limp¬
ing from a knife wound in the leg, she flees to a garden room and breaks
in through the window with a rake. Neighbors hear her scream for help
but suspect a Halloween prank and shut the blinds. She gets into her
own babysitting house—by throwing a potted plant at a second-story
window to rouse the children—just as the killer descends. Minutes
later he comes through the window and they grapple,- she manages to
fell him with a knitting needle and grabs his butcher knife—but drops
it when he seems dead. As she goes upstairs to the children, the killer
rises, takes the knife, and goes after her. She takes refuge in a closet,
lashing the two doorknobs together from the inside. As the killer
slashes and stabs at the closet door—we see this from her inside per¬
spective—she bends a hanger into a weapon and, when he breaks the
door down, stabs him in the eye. Again thinking him vanquished, she
sends the children to the police and sinks down in pain and exhaustion.
The killer rises again, but just as he is about to stab her, Dr. Loomis,
alerted by the children, rushes in and shoots the killer.
Given the drift in just the four years between The Texas Chain¬
saw Massacre and Halloween—from passive to active defense—it is no
surprise that the films following Halloween present Final Girls who not
only fight back but do so with ferocity and even kill the killer on their
own, without help from the outside.33 Valerie in Slumber Party Mas¬
sacre (a film directed by Amy Jones and scripted by Rita Mae Brown)
takes a machete-like weapon to the killer, striking off the bit from his
drill, severing1 his hand, and finally impaling him. Alice assaults and
decapitates the killer of Friday the 13th. Pursued by the killer in Hell
Night, Marti pries the gate key from the stiff fingers of a corpse to let
herself out of the mansion grounds to safety; when the car won't start,
she repairs it on the spot; when the car gets stuck in the roadway, she
inside and killer on top, she releases it in such a way as to cast the killer
145
Her Body,'Himself: Gender in the Slasher Film
Figure 27. Freddy Kreuger stalked and slashed his way through the popular Nightmare on Elm Street
films. Unlike other killers in the slasher cycle, Kreuger leavened his sadism with an urbane wit.
on the gate's upper spikes. The grittiest of the Final Girls is Nancy of A
Nightmare on Elm Street. Aware in advance that the killer will be pay¬
ing her a visit, she plans an elaborate defense. When he enters the
house, she dares him to come at her, then runs at him in direct attack.
146
Carol J. Clover
As they struggle, he springs the contraptions she has prepared; he is

stunned by a swinging sledgehammer, jolted and half incinerated by an
electrical charge, and so on. When he rises yet again, she chases him
around the house, bashing him with a chair.34
In The Texas Chainsaw Massacre, Part 2, the Final Girl se¬
quence takes mythic measure. Trapped in the underground slaughter¬
house, Stretch repeatedly flees, hides, is caught, tortured (at one point
forced to don the flayed face of her murdered technician companion),
and nearly killed. She escapes with her life chiefly because Leatherface,
having developed an affection for her after the crotch episode, is reluc¬
tant to ply his chainsaw as the tyrannical Mr. Sawyer commands.
Finally Stretch finds her way out, leaving the Texas Ranger to face
certain death below, and clambers up a nearby pinnacle, Chop Top in
pursuit. At the summit she finds the mummified grandmother, cere¬
moniously enthroned in an open-air chamber, and next to her a func¬
tional chainsaw. She turns the saw on Chop Top, gashing open his
abdomen and tossing him off the precipice. The final scene shows her
in extreme longshot, in brilliant sunshine, waving the buzzing chain¬
saw triumphantly overhead. (It is a scene we are invited to compare to
the final scene of the first film, in which the wounded Leatherface is
shown in longshot at dawn, staggering after the pickup on the highway
waving his chainsaw crazily over his head.) In the 1974 film the Final
Girl, for all her survivor pluck, is, like Red Riding Hood, saved through
male agency. In the 1986 sequel, however, there is no male agency; the
figure so designated, the Texas Ranger, proves so utterly ineffectual that
he cannot save himself, and much less the girl. The comic ineptitude
and failure of would-be "woodsmen" is a repeated theme in the later
slasher films. In Slumber Party Massacre, the role is played by a
woman—although a butch one (the girls' basketball coach). She comes
to the slumber party's rescue only to fall victim to the drill herself. But
to focus on just who brings the killer down, the Final Girl or a male res¬
cuer, is—as the easy alternation between the two patterns would seem
to suggest—to miss the point. The last moment of the Final Girl
sequence is finally a footnote to what went before, to the quality of the
Final Girl's fight, and more generally to the qualities of character that
enable her, of all the characters, to survive what has come to seem
unsurvivable.
The Final Girl sequence too is prefigured, if only rudimentar-
ily, in Psycho's final scenes, in which Lila (Marion's sister) is caught
reconnoitering in the Bates mansion and nearly killed. Sam (Marion's
147
boyfriend) detains Norman at the motel while Lila snoops about (tak¬
ing note of Norman's toys). When she perceives Norman's approach,
she flees to the basement. Here she encounters the treated corpse of
Mrs. Bates and begins screaming in horror. Norman bursts in and is
about to strike when Sam enters and grabs him from behind. Like her
generic sisters, then, Lila is the spunky inquirer into the Terrible
Place—the one who first grasps, however dimly, the past and present
danger, the one who looks death in the face, and the one who survives
the murderer's last stab. There the correspondences end, however. The
Psycho scene turns, after all, on the revelation of Norman's psychotic
identity, not on Lila as a character—she enters the film midway and is
sketchily drawn—and still less on her self-defense.
The Final Girl of the slasher film is presented from the outset
as the main character. The practiced viewer distinguishes her from her
friends minutes into the film. She is the girl scout, the bookworm, the
mechanic. Unlike her girlfriends (and Marion Crane), she is not sexu¬
ally active. Laurie (Halloween) is teased because of her fears about dat¬
ing, and Marti (Hell Night) explains to the boy with whom she finds
herself sharing a room that they will have separate beds. Although
Stretch [Texas Chainsaw, Part 2) is hardly virginal, she is not available,
either,- early in the film she pointedly turns down a date, and we are
given to understand that she is, for the present, unattached and even
lonely. So too Stevie of Carpenter's The Fog, like Stretch a disk jockey.
Divorced mother and a newcomer in town, she is unattached and lonely
but declines male attention. The Final Girl is also watchful to the point
of paranoia; small signs of danger that her friends ignore she takes in
and turns over. Above all, she is intelligent and resourceful in extreme
situations. Thus Laurie even at her most desperate, cornered in a closet,
has the wit to grab a hanger from the rack and bend it into a weapon;
Marti can hot-wire her getaway car, the killer in pursuit; and the psych
major of Friday the 13th Part 2, on seeing the enshrined head of Mrs.
Voorhees, can stop Jason in his tracks by assuming a stridently mater¬
nal voice. Finally, although she is always smaller and weaker than the
killer, she grapples with him energetically and convincingly.
The Final Girl is boyish, in a word. Just as the killer is not fully
masculine, she is not fully feminine—not, in any case, feminine in the
ways of her friends. Her smartness, gravity, competence in mechanical
and other practical matters, and sexual reluctance set her apart from
the other girls and ally her, ironically, with the very boys she fears or
rejects, not to speak of the killer himself. Lest we miss the point, it is
148
Figures 28 & 29. The shower murder of Marion Crane in Alfred Hitchcock’s Psycho exerted a
seminal influence over the subsequent slasher horror cycle.
spelled out in her name: Stevie, Marti, Terry, Laurie, Stretch, Will. Not
only the conception of the hero in Alien and Aliens but also her name,
Ripley, owes a clear debt to slasher tradition.
With the introduction of the Final Girl, then, the Psycho for¬
mula is radically altered. It is not merely a question of enlarging the fig¬
ure of Lila but of absorbing into her role, in varying degrees, the
functions of Arbogast (investigator) and Sam (rescuer) and restructuring
the narrative action from beginning to end around her progress in rela¬
tion to the killer. In other words, Psycho's detective plot, revolving
around a revelation, yields in the modern slasher film to a hero plot,
149
v
revolving around the main character's struggle with and eventual tri¬
umph over evil. But for the femaleness, however qualified, of that main
character, the story is a standard one of tale and epic.
Shock
One reason that the shower sequence in Psycho has "evoked more
study, elicited more comment, and generated more shot-for-shot analy¬
sis from a technical viewpoint than any other in the history of cinema"
is that it suggests so much but shows so little.35 Of the forty-odd shots
in as many seconds that figure the murder, only a single fleeting one
actually shows the body being stabbed. The others present us with a
rapid-fire concatenation of images of the knife-wielding hand, parts of
Marion, parts of the shower, and finally the bloody water as it swirls
down the drain. The horror resides less in the actual images than in
their summary implication.
Although Hitchcock is hardly the first director to prefer the
oblique rendition of physical violence, he may, to judge from current
examples, be one of the last. For better or worse, the perfection of special
effects has made it possible to show maiming and dismemberment in
extraordinarily credible detail. The horror genres are the natural reposi¬
tories of such effects,- what can be done is done, and slashers, at the bot¬
tom of the category, do it most and worst. Thus we see a head being
stepped on so that the eyes pop out, a face being flayed, a decapitation, a
hypodermic needle penetrating an eyeball in close-up, and so on.
With this new explicitness also comes a new tone. If the horror
of Psycho was taken seriously, the "horror" of the slasher films is of a
rather more complicated sort. Audiences express uproarious disgust
("Gross!") as often as they express fear, and it is clear that the makers
of slasher films pursue the combination. More particularly, spectators
fall silent while the victim is being stalked, scream out at the first stab,
and make loud noises of revulsion at the sight of the bloody stump. The
rapid alternation between registers—between something like "real"
horror on one hand and campy, self-parodying Horror on the other—is
by now one of the most conspicuous characteristics of the tradition. In
its cultivation of intentionally outrageous excess, the slasher film inter¬
sects with the cult film, a genre devoted to such effects. Just what this
self-ironizing relation to taboo signifies, beyond a remarkably compe¬
tent audience, is unclear—it is yet another aspect of the phenomenon
that has lain beyond criticism—but for the time being it stands as a
defining characteristic of the lower genres of popular culture.
150
Carol J. Clover
The Body
On the face of it, the relation between the sexes in slasher films could
hardly be clearer. The killer is with few exceptions recognizably human
and distinctly male; his fury is unmistakably sexual in both roots and
expression; his victims are mostly women, often sexually free and
always young and beautiful ones. Just how essential this victim is to
horror is suggested by her historical durability. If the killer has over
time been variously figured as shark, fog, gorilla, birds, and slime, the
victim is eternally and prototypically the damsel. Cinema hardly
invented the pattern. It has simply given visual expression to the abid¬
ing proposition that, in Poe's famous formulation, the death of a beau¬
tiful woman is the "most poetical topic in the world."36 As slasher
director Dario Argento puts it, "I like women, especially beautiful ones.
If they have a good face and figure, I would much prefer to watch them
being murdered than an ugly girl or a man."37 Brian De Palma elabo¬
rates: "Women in peril work better in the suspense genre. It all goes
back to the Perils of Pauline. ... If you have a haunted house and you
have a woman walking around with a candelabrum, you fear more for
her than you would for a husky man."38 Or Hitchcock, during the film¬
ing of The Birds: "I always believe in following the advice of the play¬
wright Sardou. He said 'Torture the women!' The trouble today is that
we don't torture women enough."39
What the directors do not say, but show, is that "Pauline" is at
her very most effective in a state of undress, borne down upon by a bla¬
tantly phallic murderer, even gurgling orgasmically as she dies. The case
could be made that the slasher films available at a given neighborhood
video rental outlet recommend themselves to censorship under the
Dworkin-MacKinnon guidelines at least as readily as the hard-core
films the next section over, at which that legislation is aimed; for if some
victims are men, the argument goes, most are women, and the women
are brutalized in ways that come too close to real life for comfort. But
what this line of reasoning does not take into account is the figure of the
Final Girl. Because slashers lie for all practical purposes beyond the
purview of legitimate criticism and, to the extent that they have been
reviewed at all, have been reviewed on an individual basis, the phe¬
nomenon of the female victim-hero has scarcely been acknowledged.
It is, of course, "on the face of it" that most of the public dis¬
cussion of film takes place—from the Dworkin-MacKinnon legislation
to Sislcel and Ebert's reviews to our own talks with friends on leaving
the movie house. Underlying that discussion is the assumption that the
151
sexes are what they seem; that screen males represent the Male and
screen females the Female,- that this identification along gender lines
authorizes impulses toward sexual violence in males and encourages
impulses toward victimization in females. In part because of the mas¬
sive authority cinema by nature accords the image, even academic film
criticism has been slow—slower than literary criticism—to get beyond
appearances. Film may not appropriate the mind's eye, but it certainly
encroaches on it; the gender characteristics of a screen figure are a vis¬
ible and audible given for the duration of the film. To the extent that
the possibility of cross-gender identification has been entertained, it
has been in the direction female-with-male. Thus some critics have
wondered whether the female viewer, faced with the screen image of a
masochistic/narcissistic female, might not rather elect to "betray her
sex and identify with the masculine point of view."40 The reverse ques¬
tion—whether men might not also, on occasion, elect to betray their
sex and identify with screen females—has scarcely been asked, pre¬
sumably on the assumption that men's interests are well served by the
traditional patterns of cinematic representation. Then too there is the
matter of the "male gaze." As E. Ann Kaplan sums it up: "Within the
film text itself, men gaze at women, who become objects of the gaze;
the spectator, in turn, is made to identify with this male gaze, and to
objectify the women on the screen; and the camera's original 'gaze'
comes into play in the very act of filming."41 But if it is so that all of
us, male and female alike, are by these processes "made to" identify
with men and "against" women, how are we then to explain the appeal
to a largely male audience of a film genre that features a female victim-
hero? The slasher film brings us squarely up against a fundamental
question of film analysis: Where does the literal end and the figurative
begin,- how do the two levels interact and what is the significance of the
particular interaction,- and to which, in arriving at a political judgment
(as we are inclined to do in the case of low horror and pornography), do
we assign priority?
A figurative or functional analysis of the slasher begins with the
processes of point of view and identification. The male viewer seeking
a male character, even a vicious one, with whom to identify in a sus¬
tained way has little to hang on to in the standard example. On the good
side, the only viable candidates are the schoolmates or friends of the
girls. They are for the most part marginal, undeveloped characters;
more to the point, they tend to die early in the film. If the traditional
horror film gave the male spectator a last-minute hero with whom to
identify, thereby "indulging his vanity as protector of the helpless
152
Carol J. Clover
female/'42 the slasher eliminates or attenuates that role beyond any

such function; indeed, would-be rescuers are not infrequently blown
away for their efforts, leaving the girl to fight her own fight. Policemen,
fathers, and sheriffs appear only long enough to demonstrate risible
incomprehension and incompetence. On the bad side, there is the
killer. The killer is often unseen, or barely glimpsed, during the first
part of the film, and what we do see, when we finally get a good look,
hardly invites immediate or conscious empathy. He is commonly
masked, fat, deformed, or dressed as a woman. Or "he" is a woman: woe
to the viewer of Friday the 13th who identifies with the male killer only
to discover, in the film's final sequences, that he was not a man at all
but a middle-aged woman. In either case, the killer is himself eventu¬
ally killed or otherwise evacuated from the narrative. No male charac¬
ter of any stature lives to tell the tale.
The one character of stature who does live to tell the tale is of
course female. The Final Girl is introduced at the beginning and is the
only character to be developed in any psychological detail. We under¬
stand immediately from the attention paid it that hers is the main story
line. She is intelligent, watchful, level-headed; the first character to
sense something amiss and the only one to deduce from the accumu¬
lating evidence the patterns and extent of the threat; the only one, in
other words, whose perspective approaches our own privileged under¬
standing of the situation. We register her horror as she stumbles on the
corpses of her friends; her paralysis in the face of death duplicates those
moments of the universal nightmare experience on which horror
frankly trades. When she downs the killer, we are triumphant. She is
by any measure the slasher film's hero. This is not to say that our
attachment to her is exclusive and unremitting, only that it adds up and
that in the closing sequence it is very close to absolute.
An analysis of the camerawork bears this out. Much is made of
the use of the I-camera to represent the killer's point of view. In these
passages—they are usually few and brief, but powerful—we see through
his eyes and (on the sound track) hear his breathing and heartbeat. His
and our vision is partly obscured by bushes or window blinds in the
foreground. By such means we are forced, the argument goes, to iden¬
tify with the killer. In fact, however, the relation between camera point
of view and the processes of viewer identification are poorly under¬
stood; the fact that Steven Spielberg can stage an attack in Jaws from
the shark's point of view (underwater, rushing upward toward the
swimmer's flailing legs) or Hitchcock an attack in The Birds from the
153
bird's-eye perspective (from the sky, as they gather to swoop down on

the streets of Bodega Bay) would seem to suggest either that the viewer's
identificatory powers are unbelievably elastic or that point-of-view
shots can sometimes be pro forma.43
But let us for the moment accept the equation point of view =
identification. We are linked, in this way, with the killer in the early
part of the film, usually before we have seen him directly and before we
have come to know the Final Girl in any detail. Our closeness to him
wanes as our closeness to the Final Girl waxes—a shift underwritten by
story line as well as camera position. By the end, point of view is hers:
we are in the closet with her, watching with her eyes the knife blade
stab through the door; in the room with her as the killer breaks through
the window and grabs at her,- in the car with her as the killer stabs
through the convertible top; and so on. With her, we become if not the
killer of the killer then the agent of his expulsion from the narrative
vision. If, during the film's course, we shifted our sympathies back and
forth and dealt them out to other characters along the way, we belong
in the end to the Final Girl; there is no alternative. When Stretch evis¬
cerates Chop Top at the end of Texas Chainsaw, Part 2, she is literally
the only character left alive, on either side.
Audience response ratifies this design. Observers unanimously
stress the readiness of the "live" audience to switch sympathies in mid¬
stream, siding now with the killer and now, and finally, with the Final
Girl. As Schoell, whose book on shocker films wrestles with its own
monster, "the feminists," puts it:
Social critics make much of the fact that male audience members cheer
on the misogynous misfits in these movies as they rape, plunder, and
murder their screaming, writhing female victims. Since these same
critics walk out of the moviehouse in disgust long before the movie is
over, they don't realize that these same men cheer on (with renewed
enthusiasm, in fact) the heroines, who are often as strong, sexy, and
independent as the [earlier] victims, as they blow away the killer with
a shotgun or get him between the eyes with a machete. All of these men
are said to be identifying with the maniac, but they enjoy his death
throes the most of all, and applaud the heroine with admiration.44
What filmmakers seem to know better than film critics is that gender
is less a wall than a permeable membrane.45
No one who has read "Red Riding Hood" to a small boy or par¬
ticipated in a viewing of, say, Deliverance ([1972], an all-male story that
154
Carol J. Clover
women find as gripping as men) or, more recently, Alien and Aliens,
with whose spaceage female Rambo, herself a Final Girl, male viewers
seem to engage with ease, can doubt the phenomenon of cross-gender
identification.46 This fluidity of engaged perspective is in keeping with
the universal claims of the psychoanalytic model: the threat function
and the victim function coexist in the same unconscious, regardless of
anatomical sex. But why, if viewers can identify across gender lines and
if the root experience of horror is sex blind, are the screen sexes not
interchangeable? Why not more and better female killers, and why (in
light of the maleness of the majority audience) not Pauls as well as
Paulines? The fact that horror film so stubbornly genders the killer
male and the principal victim female would seem to suggest that rep¬
resentation itself is at issue—that the sensation of bodily fright derives
not exclusively from repressed content, as Freud insisted, but also from
the bodily manifestations of that content.
Nor is the gender of the principals as straightforward as it first
seems. The killer's phallic purpose, as he thrusts his drill or knife into
the trembling bodies of young women, is unmistakable. At the same
time, however, his masculinity is severely qualified: he ranges from the
virginal or sexually inert to the transvestite or transsexual, is spiritually
divided ("the mother half of his mind"), or even equipped with vulva and
vagina. Although the killer of God Told Me To (a.k.a. Demon [1979]) is
represented and taken as a male in the film text, he is revealed, by the
doctor who delivered him, to have been sexually ambiguous from birth:
"I truly could not tell whether that child was male or female,- it was as
if the sexual gender had not been determined... as if it were being devel¬
oped."47 In this respect, slasher killers have much in common with the
monsters of classic horror—monsters who, in Linda Williams's formu¬
lation, represent not just "an eruption of the normally repressed animal
sexual energy of the civilized male" but also the "power and potency of
a non-phallic sexuality." To the extent that the monster is constructed
as feminine, the horror film thus expresses female desire only to show
how monstrous it is.48 The intention is manifest in Aliens, in which
the Final Girl, Ripley, is pitted in the climactic scene against the most
terrifying "alien" of all: an egg-laying Mother.
Nor,can we help noticing the "intrauterine" quality of the Ter¬
rible Place, dark and often damp, in which the killer lives or lurks and
whence he stages his most terrifying attacks. "It often happens," Freud
wrote, "that neurotic men declare that they feel there is something
uncanny about the female genital organs. This unheimlich place, how¬
ever, is an entrance to the former Heim [home] of all human beings, to
155
the place where each one of us lived once upon a time and in the begin¬
ning. ... In this case too then, the unheimlich is what once was
heimisch, familiar,- the prefix 'uri ['un-'] is the token repression."49 It
is the exceptional film that does not mark as significant the moment
that the killer leaps out of the dark recesses of a corridor or cavern at
the trespassing victim, usually the Final Girl. Long after the other par¬
ticulars have faded, the viewer \yill remember the images of Amy
assaulted from the dark halls of a morgue [He Knows You’re Alone),
Sally or Stretch facing dismemberment in the ghastly dining room or
underground labyrinth of the slaughterhouse family (the Texas Chain¬
saw films), or Melanie trapped in the attic as the savage birds close in
(The Birds). In such scenes of convergence the Other is at its bisexual
mightiest, the victim at her tiniest, and the component of sado¬
masochism at its most blatant.
The gender of the Final Girl is likewise compromised from the
outset by her masculine interests, her inevitable sexual reluctance (pen¬
etration, it seems, constructs the female), her apartness from other
girls, sometimes her name. At the level of the cinematic apparatus, her
unfemininity is signaled clearly by her exercise of the "active investi¬
gating gaze" normally reserved for males and hideously punished in
females when they assume it themselves. Tentatively at first and then
aggressively, the Final Girl looks for the killer, even tracking him to his
forest hut or his underground labyrinth, and then at him, therewith
bringing him, often for the first time, into our vision as well.50 When,
in the final scene, she stops screaming, looks at the killer, and reaches
for the knife (sledgehammer, scalpel, gun, machete, hanger, knitting
needle, chainsaw), she addresses the killer on his own terms. To the
critics' objection that Halloween in effect punished female sexuality,
director John Carpenter responded: "They [the critics] completely
missed the boat there, I think. Because if you turn it around, the one
girl who is the most sexually uptight just keeps stabbing this guy with
a long knife. She's the most sexually frustrated. She's the one that killed
him. Not because she's a virgin, but because all that repressed energy
starts coming out. She uses all those phallic symbols on the guy-She
and the killer have a certain link: sexual repression."51
For all its perversity, Carpenter's remark does not underscore
the sense of affinity, even recognition, that attends the final encounter.
But the "certain link" that puts killer and Final Girl on terms, at least
briefly, is more than "sexual repression." It is also a shared masculinity,
materialized in "all those phallic symbols"—and it is also a shared fem¬
ininity, materialized in what comes next (and what Carpenter, perhaps
156
Carol ]. Clover
significantly, fails to mention): the castration, literal or symbolic, of the

killer at her hands. His eyes may be put out, his hand severed, his body
impaled or shot, his belly gashed, or his genitals sliced away or bitten
off. The Final Girl has not just manned herself; she specifically unmans
an oppressor whose masculinity was in question to begin with. By the
time the drama has played itself out, darkness yields to light (often as
day breaks) and the close quarters of the barn (closet, elevator, attic,
basement) give way to the open expanse of the yard (field, road,
lakescape, cliff). With the Final Girl's appropriation of "all those phallic
symbols" comes the queuing, the dispelling, of the "uterine" threat as
well. Consider again the paradigmatic ending of Texas Chainsaw, Part
2. From the underground labyrinth, murky and. bloody, in which she
faced saw, knife, and hammer, Stretch escapes through a culvert into the
open air. She clambers up the jutting rock and with a chainsaw takes her
stand. When her last assailant comes at her, she slashes open his lower
abdomen—the sexual symbolism is all too clear—and flings him off the
cliff. Again, the final scene shows her in extreme longshot, standing on
the pinnacle, drenched in sunlight, buzzing chainsaw held overhead.
The tale would indeed seem to be one of sex and parents. The
patently erotic threat is easily seen as the materialized projection of the
dreamer's (viewer's) own incestuous fears and desires. It is this disabling
cathexis to one's parents that must be killed and rekilled in the service
of sexual autonomy. When the Final Girl stands at last in the light of
day with the knife in her hand, she has delivered herself into the adult
world. Carpenter's equation of the Final Girl with the killer has more
than a grain of truth. The killers of Psycho, The Eyes of Laura Mars, Fri¬
day the 13th, Parts Two through Six, and Cruising, among others, are
explicitly figured as sons in the psychosexual grip of their mothers (or
fathers, in the case of Cruising). The difference is between past and pres¬
ent and between failure and success. The Final Girl enacts in the pres¬
ent, and successfully, the parenticidal struggle that the killer himself
enacted unsuccessfully in his own past—a past that constitutes the
film's backstory. She is what the killer once was,- he is what she could
become should she fail in her battle for sexual selfhood. "You got a
choice, boy," says the tyrannical father of Leatherface in Texas Chain¬
saw, Part 2, ^'sex or the saw; you never know about sex, but the saw—
the saw is the family."
But the tale is no less one of maleness. If the early experience
of the oedipal drama can be—is perhaps ideally—enacted in the female
form, the achievement of full adulthood requires the assumption and,
157
apparently, brutal employment of the phallus. The helpless child is gen¬

dered feminine,- the autonomous adult or subject is gendered mascu¬
line; the passage from childhood to adulthood entails a shift from
feminine to masculine. It is the male killer's tragedy that his incipient
femininity is not reversed but completed (castration) and the Final
Girl's victory that her incipient masculinity is not thwarted but real¬
ized (phallicization). When De Palma says that female frailty is a pred¬
icate of the suspense genre, he proposes, in effect, that the lack of the
phallus, for Lacan the privileged signifier of the symbolic order of cul¬
ture, is itself simply horrifying, at least in the mind of the male
observer. Whereas pornography (the argument goes) resolves that lack
through a process of fetishization that allows a breast or leg or whole
body to stand in for the missing member, the slasher film resolves it
either through eliminating the woman (earlier victims) or reconstitut¬
ing her as masculine (Final Girl). The moment at which the Final Girl
is effectively phallicized is the moment that the plot halts and horror
ceases. Day breaks, and the community returns to its normal order.
Casting psychoanalytic verities in female form has a venerable
cinematic history. Ingmar Bergman has made a career of it, and Woody
Allen shows signs of following his lead. One immediate and practical
advantage, by now presumably unconscious on the part of makers as
well as viewers, has to do with a preestablished cinematic "language"
for capturing the moves and moods of the female body and face. The
cinematic gaze, we are told, is male, and just as that gaze "knows" how
to fetishize the female form in pornography (in a way that it does not
"know" how to fetishize the male form),52 so it "knows," in horror,
how to track a woman ascending a staircase in a scary house and how
to study her face from an angle above as she first hears the killer's foot¬
fall. A set of conventions we now take for granted simply "sees" males
and females differently.
To this cinematic habit may be added the broader range of emo¬
tional expression traditionally allowed women. Angry displays of force
may belong to the male, but crying, cowering, screaming, fainting,
trembling, begging for mercy belong to the female. Abject terror, in
short, is gendered feminine, and the more concerned a given film with
that condition—and it is the essence of modern horror—the more likely
the femaleness of the victim. It is no accident that male victims in
slasher films are killed swiftly or offscreen, and that prolonged strug¬
gles, in which the victim has time to contemplate her imminent
destruction, inevitably figure females. Only when one encounters the
158
Carol ]. Clover
rare expression of abject terror on the part of a male (as in I Spit on Your
Grave) does one apprehend the full extent of the cinematic double stan¬
dard in such matters.53
It is also the case that gender displacement can provide a kind
of identificatory buffer, an emotional remove, that permits the major¬
ity audience to explore taboo subjects in the relative safety of vicari¬
ousness. Just as Bergman came to realize that he could explore
castration anxiety more freely via depictions of hurt female bodies (wit¬
ness the genital mutilation of Karin in Cries and Whispers [1972]), so
the makers of slasher films seem to know that sadomasochistic incest
fantasies sit more easily with the male viewer when the visible player
is female. It is one thing for that viewer to hear the psychiatrist intone
at the end of Psycho that Norman as a boy (in the backstory) was abnor¬
mally attached to his mother; it would be quite another to see that
attachment dramatized in the present, to experience in nightmare form
the elaboration of Norman's (the viewer's own) fears and desires. If the
former is playable in male form, the latter, it seems, is not.
The Final Girl is, on reflection, a congenial double for the ado¬
lescent male. She is feminine enough to act out in a gratifying way, a
way unapproved for adult males, the terrors and masochistic pleasures
of the underlying fantasy, but not so feminine as to disturb the struc¬
tures of male competence and sexuality. Her sexual inactivity, in this
reading, becomes all but inevitable,- the male viewer may be willing to
enter into the vicarious experience of defending himself from the pos¬
sibility of symbolic penetration on the part of the killer, but real vagi¬
nal penetration on the diegetic level is evidently more femaleness than
he can bear. The question then arises whether the Final Girls of slasher
films —Stretch, Stevie, Marti, Will, Terry, Laurie, and Ripley—are not
boyish for the same reason that the female "victims" in Victorian flag¬
ellation literature—"Georgy," "Willy"—are boyish: because they are
transformed males. The transformation, Steven Marcus writes, "is
itself both a defense against and a disavowal of the fantasy it is simul¬
taneously expressing"—namely, that a "little boy is being beaten ... by
another man."54 What is represented as male-on-female violence, in
short, is figuratively speaking male-on-male sex. For Marcus, the liter¬
ary picture of flagellation, in which girls are beaten, is utterly belied by
the descriptions (in My Secret Life) of real-life episodes in which the
persons being beaten are not girls at all but "gentlemen" dressed in
women's clothes ("He had a woman's dress on tucked up to his waist,
showing his naked rump and thighs.... On his head was a woman's cap
tied carefully round his face to hide whiskers") and whipped by prosti-
159
Her Body, Himself: Gender in the Slasher Film.
tutes. Reality, Marcus writes, "puts the literature of flagellation out of

the running... by showing how that literature is a completely distorted
and idealized version of what actually happens."55 Applied to the
slasher film, this logic reads the femaleness of the Final Girl (at least
up to the point of her transformation) and indeed of the women victims
in general as only apparent, the artifact of heterosexual deflection. It
may be through the female body that the body of the audience is sen¬
sationalized, but the sensation is an entirely male affair.
At least one director, Hitchcock, explicitly located thrill in the
equation victim = audience. So we judge from his marginal jottings in
the shooting instructions for the shower scene in Psycho: "The slash¬
ing. An impression of a knife slashing, as if tearing at the very screen,
ripping the film."56 Not just the body of Marion is to be ruptured, but
also the body on the other side of the film and screen: our witnessing
body. As Marion is to Norman, the audience of Psycho is to Hitchcock;
as the audiences of horror film in general are to the directors of those
films, female is to male. Hitchcock's "torture the women" then means,
simply, torture the audience. De Palma's remarks about female frailty
likewise contemplate a male-on-"female" relationship between director
and viewer. Cinefantastic horror, in short, succeeds in the production of
sensation to more or less the degree that it succeeds in incorporating its
spectators as "feminine" and then violating that body—which recoils,
shudders, cries out collectively—in ways otherwise imaginable, for
males, only in nightmare. The equation is nowhere more plainly put
than in David Cronenberg's Videodrome (1983). Here the threat is a
mind-destroying video signal and the victims television viewers.
Despite the (male) hero's efforts to defend his mental (and physical)
integrity, a deep, vaginalike gash appears on his lower abdomen. Says the
media conspirator as he thrusts a videocassette into the victim's gaping
wound, "You must open yourself completely to this."
If the slasher film is "on the face of it" a genre with at least a
strong female presence, it is in these figurative readings a thoroughly
strong male exercise, one that finally has very little to do with female¬
ness and very much to do with phallocentrism. Figuratively seen, the
Final Girl is a male surrogate in things oedipal, a homoerotic stand-in,
the audience incorporate; to the extent she "means" girl at all, it is only
for purposes of signifying phallic lack, and even that meaning is nulli¬
fied in the final scenes. Our initial question—how to square a female
victim-hero with a largely male audience—is not so much answered as
it is obviated in these readings. The Final Girl is (apparently) female not
despite the maleness of the audience, but precisely because of it. The
160
Carol J. Clover
discourse is wholly masculine, and females figure in it only insofar as

they "read" some aspect of male experience. To applaud the Final Girl
as a feminist development, as some reviews of Aliens have done with
Ripley, is, in light of her figurative meaning, a particularly grotesque
expression of wishful thinking.57 She is simply an agreed-upon fiction,
and the male viewer's use of her as a vehicle for his own sado¬
masochistic fantasies an act of perhaps timely dishonesty.
For all their immediate appeal, these figurative readings loosen
as many ends as they tie together. The audience, we have said, is pre¬
dominantly male; but what about the women in it? Do we dismiss them
as male-identified and account for their experience as an "immascu-
lated" act of collusion with the oppressor?58 This is a strong judgment
to apply to large numbers of women; for while it may be that the audi¬
ence for slasher films is mainly male, that does not mean that there are
not also many female viewers who actively like such films, and of course
there are also women, however few, who script, direct, and produce
them. These facts alone oblige us at least to consider the possibility that
female fans find a meaning in the text and image of these films that is
less inimical to their own interests than the figurative analysis would
have us believe. Or should we conclude that males and females read
these films differently in some fundamental sense? Do females respond
to the text (the literal) and males the subtext (the figurative)?59
Some such notion of differential understanding underlies the
homoerotic reading. The silent presupposition of that reading is that
male identification with the female as female cannot be, and that the
male viewer/reader who adjoins feminine experience does so only by
homosexual conversion. But does female identification with male expe¬
rience then similarly indicate a lesbian conversion? Or are the
processes of patriarchy so one-way that the female can identify with the
male directly, but the male can identify with the female only by trans-
sexualizing her? Does the Final Girl mean "girl" to her female viewers
and "boy" to her male viewers? If her masculine features qualify her as
a transformed boy, do not the feminine features of the killer qualify him
as a transformed woman (in which case the homoerotic reading can be
maintained only by defining that "woman" as phallic and retransfom-
ing her into a*male)? Striking though it is, the analogy between the Vic¬
torian flagellation story's Georgy and the slasher film's Stretch falters
at the moment that Stretch turns on her assailant and unmans him. Are
we to suppose that a homoerotic beating fantasy suddenly yields to
what folklorists call a "lack-liquidated" fantasy? Further, is it simple
coincidence that this combination tale—trials, then triumph—bears
161
such a striking resemblance to the classic (male) hero story? Does the
standard hero story featuring an anatomical female "mean" differently
from one featuring an anatomical male?
As Marcus perceived, the relationship between the Georgy sto¬
ries of flagellation literature and the real-life anecdote of the Victorian
gentleman is a marvelously telling one. In his view, the maleness of the
latter must prove the essential or‘functional maleness of the former.
What his analysis does not come to full grips with, however, is the cloth¬
ing the gentleman wears—not that of a child, as Marcus's "childish"
reading of the scene contemplates, but explicitly that of a woman.60
These women's clothes can of course be understood, within the terms
of the homoerotic interpretation, as a last-ditch effort on the part of the
gentleman to dissociate himself from the (incestuous) homosexuality
implicit in his favored sexual practice. But can they not just as well, and
far more economically, be explained as part and parcel of a fantasy of
literal femaleness? By the same token, cannot the femaleness of the
gentleman's literary representatives—the girls of the flagellation sto¬
ries—be understood as the obvious, even necessary, extension of the
man's dress and cap? The same dress and cap, I suggest, haunt the mar¬
gins of the slasher film. This is not to deny the deflective convenience,
for the male spectator (and filmmaker), of a female victim-hero in a con¬
text so fraught with taboo; it is only to suggest that the femaleness of
that character is also conditioned by a kind of imaginative curiosity
about the feminine in and of itself.
So too the psychoanalytic case. These films do indeed seem to
pit the child in a struggle, at once terrifying and attractive, with the
parental Other, and it is a rare example that does not directly thema-
tize parent-child relations. But if Freud stressed the maternal source of
the unheimlich, the Other of our films is decidedly androgynous:
female/feminine in aspects of character and place (the "intrauterine"
locale) but male in anatomy. Conventional logic may interpret the
killer as the phallic mother of the transformed boy (the Final Girl), but
the text itself does not compel such a reading. On the contrary, the text
at every level presents us with hermaphroditic constructions—con¬
structions that draw attention to themselves and demand to be taken
on their own terms.
For if we define the Final Girl as nothing more than a figurative
male, what do we then make of the context of the spectacular gender play
in which she is emphatically situated? In his essay on the uncanny, Freud
rejected out of hand Jentsch's theory that the experience of horror pro¬
ceeds from intellectual uncertainty (curiosity?)—feelings of confusion,
162
Carol J. Clover
induced by an author or a coincidence, about who, what, and where one

is.61 One wonders, however, whether Freud would have been quite so
dismissive if, instead of the mixed materials he used as evidence, he
were presented with a coherent story corpus—forty slashers, say—in
which the themes of incest and separation were relentlessly played out
by a female character, and further in which gender identity was repeat¬
edly thematized as an issue in and of itself. For although the factors we
have considered thus far—the conventions of the male gaze, the femi¬
nine constitution of abject terror, the value for the male viewer of emo¬
tional distance from the taboos in question, the special horror that may
inhere, for the male audience, in phallic lack, the homoerotic deflec¬
tion—go a long way in explaining why it is we have Pauline rather than
Paul as our victim-hero, they do not finally account for our strong sense
that gender is simply being played with, and that part of the thrill lies
precisely in the resulting "intellectual uncertainty" of sexual identity.
The "play of pronoun function" that underlies and defines the
cinefantastic is nowhere more richly manifested than in the slasher; if
the genre has an aesthetic base, it is exactly that of a visual identity
game. Consider, for example, the by now standard habit of letting us
view the action in the first person long before revealing who or what
the first person is. In the opening sequence of Halloween, "we" are
belatedly revealed to ourselves, after committing a murder in the cine¬
matic first person, as a 6-year-old boy. The surprise is often within gen¬
der, but it is also, in a striking number of cases, across gender. Again,
Friday the 13th, in which "we" stalk and kill a number of teenagers
over the course of an hour of screen time without even knowing who
"we" are; we are invited, by conventional expectation and by glimpses
of "our" own bodily parts—a heavily booted foot, a roughly gloved
hand—to suppose that "we" are male, but "we" are revealed, at film's
end, as a woman. If this is the most dramatic case of pulling out the gen¬
der rug, it is by no means the only one. In Dressed to Kill, we are led to
believe, again by means of glimpses, that "we" are female—only to dis¬
cover, in the denouement, that "we" are a male in drag. In Psycho, the
dame we glimpse holding the knife with a "visible virility quite ob¬
scene in an old lady" is later revealed, after additional gender teasing,
to be Normair in his mother's clothes.62 Psycho II (1983) plays much
the same game. Cruising (in which, not accidentally, transvestites play
a prominent role) adjusts the terms along heterosexual/homosexual
lines. The tease here is whether the originally straight detective
assigned to the string of murders in a gay community does or does not
163
succumb to his assumed homosexual identity; the camerawork leaves

us increasingly uncertain as to his (our) sexual inclinations, not to
speak of his (our) complicity in the crimes. Even at film's end we are
not sure who "we" were during several of the first-person sequences.63
The gender-identity game, in short, is too patterned and too per¬
vasive in the slasher film to be dismissed as supervenient. It would
seem instead to be an integral element of the particular brand of bodily
sensation in which the genre trades. Nor is it exclusive to horror. It is
directly thematized in comic terms in the recent "gender benders"
Tootsie ([1982], in which a man passes himself off as a woman) and All
of Me ([1984], in which a woman is literally introjected into a man and
affects his speech, movement, and thought). It is also directly thema¬
tized, in the form of bisexual and androgynous figures and relations, in
such cult films as Pink Flamingos (1973) and The Rocky Honor Picture
Show (1975). (Some version of it is indeed enacted every few minutes
on MTV.) It is further thematized (predictably enough, given their bod¬
ily concerns) in such pornographic films as Every Woman Has a Fan¬
tasy (1984), in which a man, in order to gain access to a women's group
in which sexual fantasies are discussed, dresses and presents himself as
a woman. (The degree to which "male" pornography in general relies
for its effect on cross-gender identification remains an open question,-
the proposition makes a certain sense of the obligatory lesbian se¬
quences and the phenomenal success of Behind the Green Door [1972],
to pick just two examples.)64
All of these films, and others like them, seem to be asking some
version of the question: What would it be like to be, or to seem to be,
if only temporarily, a woman? Taking exception to the reception of
Tootsie as a feminist film, Elaine Showalter argues that the success of
"Dorothy Michaels" (the Dustin Hoffman character), as far as both plot
and audience are concerned, lies in the veiling of masculine power in
feminine costume. Tootsie’s cross-dressing, she writes,
is a way of promoting the notion of masculine power while masking

it. In psychoanalytic theory, the male transvestite is not a powerless
man,- according to the psychiatrist Robert Stoller, in Sex and Gender,
he is a "phallic woman" who qan tell himself that "he is, or with prac¬
tice will become, a better woman than a biological female if he chooses
to do so." When it is safe or necessary, the transvestite "gets great
pleasure in revealing that he is a male-woman. . . . The pleasure in
tricking the unsuspecting into thinking he is a woman, and then
164
Carol J. Clover
revealing his maleness (e.g., by suddenly dropping his voice) is not so

much erotic as it is proof that there is such a thing as a woman with a
penis." Dorothy's effectiveness is the literal equivalent of speaking
softly and carrying a big stick.65
By the same literalistic token, then, Stretch's success must lie in the
fact that in the end, at least, she "speaks loudly" even though she car¬
ries no "stick." Just as "Dorothy's" voice slips serve to remind us that
her character really is male, so the Final Girl's "tits and scream" serve
more or less continuously to remind us that she really is female—even
as, and despite the fact that, she in the end acquits herself "like a
man."66 Her chainsaw is thus what "Dorothy's",skirt is: a figuration of
what she does and what she seems, as opposed to—and the films turn
on the opposition—what she is. The idea that appearance and behavior
do not necessarily indicate sex—indeed, can misindicate sex—is predi¬
cated on the understanding that sex is one thing and gender another,- in
practice, that sex is life, a less-than-interesting given, but that gender
is theater. Whatever else it may be, Stretch's waving of the chainsaw is
a moment of high drag. Its purpose is not to make us forget that she is
a girl but to thrust that fact on us. The moment, it is probably fair to
say, is also one that openly mocks the literary/cinematic conventions
of symbolic representation.
It may be just this theatricalization of gender that makes pos¬
sible the willingness of the male viewer to submit himself to a brand of
spectator experience that Hitchcock designated as "feminine" in 1960
and that has become only more so since then. In classic horror, the
"feminization" of the audience is intermittent and ceases early. Our
relationship with Marion's body in Psycho halts abruptly at the
moment of its greatest intensity (slashing, ripping, tearing). The con¬
siderable remainder of the film distributes our bruised sympathies
among several lesser figures, male and female, in such a way and at such
length as to ameliorate the Marion experience and leave us, in the end,
more or less recuperated in our (presumed) masculinity. Like Marion,
the Final Girl is the designated victim, the incorporation of the audi¬
ence, the slashing, ripping, and tearing of whose body will cause us to
flinch and scream out in our seats. But unlike Marion, she does not die.
If Psycho, like other classic horror films, solves the femininity problem
by obliterating the female and replacing her with representatives of the
masculine order (mostly but not inevitably males), the modern slasher
solves it by regendering the woman. We are, as an audience, in the end
"masculinized" by and through the very figure by and through whom
165
we were earlier "feminized." The same body does for both, and that
body is female.
The last point is the crucial one: the same female body does for
both. The Final Girl (1) undergoes agonizing trials and (2) virtually or
actually destroys the antagonist and saves herself. By the lights of folk
tradition, she is not a heroine, for whom phase 1 consists in being saved
by someone else, but a hero, who rises to the occasion and defeats the
adversary with his own wit and hands. Phase 1 of the story sits well on
the female; it is the heart of heroine stories in general (Red Riding Hood,
Pauline), and in some figurative sense, in ways we have elaborated in
some detail, it is gendered feminine even when played by a male.
Odysseus's position, trapped in the cave of the Cyclops, is after all not
so different from Pauline's position tied to the tracks or Sally's trapped
in the dining room of the slaughterhouse family. The decisive moment,
as far as the fixing of gender is concerned, lies in what happens next:
those who save themselves are male, and those who are saved by oth¬
ers are female. No matter how "feminine" his experience in phase 1,
the traditional hero, if he rises against his adversary and saves himself
in phase 2, will be male.
What is remarkable about the slasher film is that it comes close
to reversing the priorities. Presumably for the various functional or fig¬
urative reasons we have considered in this essay, phase 1 wants a female:
on that point all slashers from Psycho on are agreed. Abject fear is still
gendered feminine, and the taboo anxieties in which slashers trade are
still explored more easily via Pauline than Paul. The slippage comes in
phase 2. As if in mute deference to a cultural imperative, slasher films
from the 1970s bring in a last-minute male, even when he is rendered
supernumerary by the Final Girl's sturdy defense. By 1980, however, the
male rescuer is either dismissably marginal or dispensed with alto¬
gether,- not a few films have him rush to the rescue only to be hacked to
bits, leaving the Final Girl to save herself after all. At the moment that
the Final Girl becomes her own savior, she becomes a hero; and the
moment that she becomes a hero is the 'moment that the male viewer
gives up the last pretense of male identification. Abject terror may still
be gendered feminine, but the willingness of one immensely popular
current genre to rerepresent the hero as an anatomical female would
seem to suggest that at least one of the traditional marks of heroism, tri¬
umphant self-rescue, is no longer strictly gendered masculine.
So too the cinematic apparatus. The classic split between "spec¬
tacle and narrative," which "supposes the man's role as the active one
of forwarding the story, making things happen," is at least unsettled in
166
Carol J. Clover
the slasher film.67 When the Final Girl (in films like Hell Night, Texas
Chainsaw, Part 2, and even Splatter University [1985]) assumes the
"active investigating gaze/' she exactly reverses the look, making a
spectacle of the killer and a spectator of herself. Again, it is through the
killer's eyes (I-camera) that we saw the Final Girl at the beginning of
the film, and through the Final Girl's eyes that we see the killer, often
for the first time with any clarity, toward the end. The gaze becomes,
at least for a while, female. More to the point, the female exercise of
scopic control results not in her annihilation, in the manner of classic
cinema, but in her triumph; indeed, her triumph depends on her
assumption of the gaze. It is no surprise, in light of these developments,
that the Final Girl should show signs of boyishness. Fler symbolic phal-
licization, in the last scenes, may or may not proceed at root from the
horror of lack on the part of audience and maker. But it certainly pro¬
ceeds from the need to bring her in line with the epic laws of Western
narrative tradition—the very unanimity of which bears witness to the
historical importance, in popular culture, of the literal representation
of heroism in male form—and it proceeds no less from the need to ren¬
der the reallocated gaze intelligible to an audience conditioned by the
dominant cinematic apparatus.
It is worth noting that the higher genres of horror have for the
most part resisted such developments. The idea of a female who out¬
smarts, much less outfights—or outgazes—her assailant is unthinkable
in the films of De Palma and Hitchcock. Although the slasher film's
victims may be sexual teases, they are not in addition simple-minded,
scheming, physically incompetent, and morally deficient in the man¬
ner of these filmmakers' female victims. And however revolting their
special effects and sexualized their violence, few slasher murders
approach the level of voluptuous sadism that attends the destruction of
women in De Palma's films. For reasons on which we can only specu¬
late, femininity is more conventionally elaborated and inexorably pun¬
ished, and in an emphatically masculine environment, in the higher
forms— the forms that are written up, and not by Joe Bob Briggs.
That the slasher film speaks deeply and obsessively to male
anxieties and desires seems clear—if nothing else from the maleness of
the majority audience. And yet these are texts in which the categories
masculine and feminine, traditionally embodied in male and female,
are collapsed into one and the same character—a character who is
anatomically female and one whose point of view the spectator is
unambiguously invited, by the usual set of literary-structural and cin¬
ematic conventions, to share. The willingness and even eagerness (so
167
we judge from these films' enormous popularity) of the male viewer to

throw in his emotional lot, if only temporarily, with not only a woman
but a woman in fear and pain, at least in the first instance, would seem
to suggest that he has a vicarious stake in that fear and pain. If the act
of horror spectatorship is itself registered as a "feminine" experience—
that the shock effects induce bodily sensations in the viewer answer¬
ing the fear and pain of the screen victim—the charge of masochism is
underlined. This is not to say that the male viewer does not also have
a stake in the sadistic side; narrative structure, cinematic procedures,
and audience response all indicate that he shifts back and forth with
ease. It is only to suggest that in the Final Girl sequence his empathy
with what the films define as the female posture is fully engaged, and
further, because this sequence is inevitably the central one in any given
film, that the viewing experience hinges on the emotional assumption
of the feminine posture. Kaja Silverman takes it a step farther: "I will
hazard the generalization that it is always the victim—the figure who
occupies the passive position—who is really the focus of attention, and
whose subjugation the subject (whether male or female) experiences as
a pleasurable repetition from his/her own story," she writes. "Indeed, I
would go so far as to say that the fascination of the sadistic point of view
is merely that it provides the best vantage point from which to watch
the masochistic story unfold."68
The slasher is hardly the first genre in the literary and visual arts
to invite identification with the female,- one cannot help wondering
more generally whether the historical maintenance of images of women
in fear and pain does not have more to do with male vicarism than is
commonly acknowledged. What distinguishes the slasher, however, is
the absence or untenability of alternative perspectives and hence the
exposed quality of the invitation. As a survey of the tradition shows, this
has not always been the case. The stages of the Final Girl's evolution—
her piecemeal absorption of functions previously represented in males—
can be located in the years following 1978. The fact that the typical
patrons of these films are the sons of marriages contracted in the 1960s
or even early 1970s leads us to speculate that the dire claims of that era—
that the women's movement, the entry of women into the workplace,
and the rise of divorce and woman-Jieaded families would yield massive
gender confusion in the next generation—were not entirely wrong. We
may prefer, in the 1980s, to speak of the cult of androgyny, but the point
is roughly the same. The fact that we have in the killer a feminine male
and in the main character a masculine female—parent and Everyteen,
respectively—would seem, especially in the latter case, to suggest a
168
Carol J. Clover
loosening of the categories, or at least of the equation sex = gender. It is

not that these films show us gender and sex in free variation,- it is that
they fix on the irregular combinations, of which the combination mas¬
culine female repeatedly prevails over the combination feminine male.
The fact that masculine males (boyfriends, fathers, would-be rescuers)
are regularly dismissed through ridicule or death or both would seem
to suggest that it is not masculinity per se that is being privileged, but
masculinity in conjunction with a female body—indeed, as the term
"victim-hero" contemplates, masculinity in conjunction with femi¬
ninity. For if "masculine" describes the Final Girl some of the time and
in some of her more theatrical moments, it does not do justice to the
sense of her character as a whole. She alternates between registers from
the outset; before her final struggle she endures the deepest throes of
"femininity"; and even during that final struggle she is now weak and
now strong, now flees the killer and now charges him, now stabs and is
stabbed, now cries out in fear and now shouts in anger. She is a physi¬
cal female and a characterological androgyne: like her name, not mas¬
culine but either/or, both, ambiguous.69
Robin Wood speaks of the sense that horror, for him the by¬
product of cultural crisis and disintegration, is "currently the most
important of all American [film] genres and perhaps the most progres¬
sive, even in its overt nihilism."70 Likewise Vale and Juno say of the
"incredibly strange films," mostly low-budget horror, that their vol¬
ume surveys: "They often present unpopular—even radical—views
addressing the social, political, racial, or sexual inequities, hypocrisy in
religion or government."71 And Tania Modleski rests her case against
the standard critique of mass culture (stemming from the Frankfurt
School) squarely on the evidence of the slasher, which does not propose
a spurious harmony; does not promote the "specious good" (but indeed
often exposes and attacks it); does not ply the mechanisms of identifi¬
cation, narrative continuity, and closure to provide the sort of narrative
pleasure constitutive of the dominant ideology.72 One is deeply reluc¬
tant to make progressive claims for a body of cinema as spectacularly
nasty toward women as the slasher film is, but the fact is that the
slasher does, in its own perverse way and for better or worse, constitute
a visible adjustment in the terms of gender representation. That it is an
adjustment largely on the male side, appearing at the farthest possible
remove from the quarters of theory and showing signs of trickling
upward, is of no small interest.
169
NOTES
I owe a special debt of gratitude to James Cunniff and Lynn Hunt for criticism and encour¬
agement. Particular thanks to James (not Lynn) for sitting with me through not a few of
these movies.
1. Morris Dickstein, "The Aesthetics of Fright/' American Film 5, no. 10 (Sep¬
tember 1980): 34.
2. "Will Rogers never met a man he didn't like, and I can truly say the same about
the cinema," Harvey R. Greenberg says in his paean to horror, The Movies on Your Mind
(New York: Dutton/Saturday Review Press, 1975); yet his claim does not extend to the
"plethora of execrable imitations [of Psycho] that debased cinema" (p. 137).
3. William Schoell, Stay Out of the Shower: Twenty-Five Years of Shocker Films
Beginning with Psycho (New York: Dembner, 1985).
4. "Joe Bob Briggs" was evidently invented as a solution to the Dallas Time Her¬
ald's problem of "how to cover trashy movies." See Calvin Trillin's "American Chroni¬
cles: The Life and Times of Joe Bob Briggs, So Far," New Yorker, December 22, 1986, pp.
73-88.
5. Lew Brighton, "Saturn in Retrograde,- or, The Texas Jump Cut," Film Journal 2,
no. 4 (1975): 25.
6. Stephen Koch, "Fashions in Pornography: Murder as Cinematic Chic," Harper’s,
November 1976, pp. 108-9.
7. Robin Wood, "Return of the Repressed," Film Comment 14, no. 4 (July-August
1978): 30.
8. Robin Wood, "Beauty Bests the Beast," American Film 8, no. 10 (September
1983): 63.
9. Dickstein, "Aesthetics of Fright," p. 34.
10. Sigmund Freud, "The Uncanny," in Standard Edition of the Complete Psycho¬
logical Works of Sigmund Freud, ed. and trans. James Strachey (London: Hogarth Press,
1964), 17:244.
11. Steven Marcus, The Other Victorians: A Study of Sexuality and Pornography in
Mid-Nineteenth-Century England (New York: Basic Books, 1964), p. 278.
12. William Castle, Step Right Up! I’m Gonna Scare the Pants Off America (New
York: Pharos Books, 1992), p. 262.
13. Given the number of permutations, it is no surprise that new strategies keep
emerging. Only a few years ago, a director hit upon the idea of rendering the point of view
of an infant through the use of an I-camera at floor level with a double-vision image (Larry
Cohen, It's Alive). Nearly a century after technology provided a radically different means
of telling a story, filmmakers are still uncovering the possibilities.
14. Mick Martin and Marsha Porter, Video Movie Guide, 1987 (New York: Ballan-
tine, 1987), p. 690. Wood, "Beauty," p. 65, notes that the first-person camera also serves
to preserve the secret of the killer's identity for a final surprise—crucial to many films—
but adds: "The sense of indeterminate, unidentified, possibly supernatural or superhu¬
man Menace feeds the spectator's fantasy of power, facilitating a direct spectator-camera
identification by keeping the intermediary character, while signified to be present, as
vaguely defined as possible." Brian De Palma's Blow Out opens with a parody of just this
cinematic habit.
15. On this widely discussed topic, see Kaja Silverman, The Subject of Semiotics
(New York: Oxford University Press, 1983), pp. 194-236; and Lesley Stern, "Point of View:
The Blind Spot," Film Reader, no. 4 (1979): 214-36.
16. In this essay I have used the term identification vaguely and generally to refer
to both primary and secondary processes. See Mary Ann Doane, "Misrecognition and
170
Carol J. Clover
Identity," Cine-Tracts, no. 11 (1980): 25-32; and Christian Metz, "The Imaginary Signi-
fier," The Imaginary Signifier: Psychoanalysis and the Cinema (Bloomington: Indiana
University Press, 1982).
17. Mark Nash, "Vampyr and the Fantastic," Screen 17, no. 3 (Autumn 1976): 37.
Nash coins the term cinefantastic to refer to this play.
18. Rosemary Jackson, Fantasy: The Literature of Subversion (London and New
York: Methuen, 1981), p. 31.
19. As Dickstein puts it, "The 'art' of horror film is a ludicrous notion since horror,
even at its most commercially exploitative, is genuinely subculture like the wild child
that can never be tamed, or the half-human mutant who appeals to our secret fascination
with deformity and the grotesque." "Aesthetics," p. 34.
20. James B. Twitchell, Dreadful Pleasures: An Anatomy of Modern Horror (New
York: Oxford University Press, 1985), p. 84.
21. Donald Spoto, The Dark Side of Genius: The Life of Alfred Hitchcock (Boston
and Toronto: Little, Brown, 1983), p. 421.
22. Wood, "Return of the Repressed," p. 26. In Wes Craven's A Nightmare on Elm
Street (1985), it is the nightmare itself, shared by the teenagers who live on Elm Street,
that is fatal. One by one they are killed by the murderer of their collective dream. The
one girl who survives does so by first refusing to sleep and then, at the same time that
she acknowledges her parents' inadequacies, by conquering the feelings that prompt the
deadly nightmare. See, on the topic of dream/horror, Dennis L. White, "The Poetics of
Horror," Cinema Journal 10, no. 2 (Spring 1971): 1-18. Reprinted in Film Genre: Theory
and Criticism, ed. Barry K. Grant (Metuchen, N.J.: Scarecrow Press, 1977), pp. 124—44.
23. It is not just the profit margin that fuels the production of low horror. It is also
the fact that, thanks to the irrelevance of production values, the initial stake is within
the means of a small group of investors. Low horror is thus for all practical purposes the
only way an independent filmmaker can break into the market. Add to this the film¬
maker's unusual degree of control over the product, and one begins to understand why it
is that low horror engages the talents of such people as Stephanie Rothman, George
Romero, Wes Craven, and Larry Cohen. As V. Vale and Andrea Juno put it, "The value of
low-budget films is: they can be transcendent expressions of a single person's individual
vision and quirky originality. When a corporation decides to invest $20 million in a film,
a chain of command regulates each step, and no person is allowed free rein. Meetings with
lawyers, accountants, and corporate boards are what films in Hollywood are all about."
Incredibly Strange Films, ed. V. Vale and Andrea Juno (San Francisco: ReSearch #10,
1986), p. 5.
24. Despite the film industry's interest in demographics, there is no in-depth study
of the composition of the slasher-film audience. Twitchell, Dreadful Pleasures, pp. 69-72
and 306-7, relies on personal observation and the report of critics, which are remarkably
consistent over time and from place to place; my own observations concur. The audience
is mostly between the ages of 12 and 20, and disproportionately male. Some critics
remark on a contingent of older men who sit separately and who, in Twitchell's view, are
there "not to be frightened, but to participate" specifically in the "stab-at-female"
episodes. Roger Ebert and Gene Siskel corroborate the observation.
25. The development of the human-sausage theme is typical of the back-and-forth
borrowing in low horror. The Texas Chainsaw Massacre hints at it; Motel Hell turns it
into an industry ("Farmer Vincent's Smoked Meats: This is it!" proclaims a local bill¬
board); and The Texas Chainsaw Massacre, Part 2 expands it to a statewide chili-tasting
contest.
26. The release of sexuality in the horror film is always presented as perverted, mon¬
strous, and excessive, both the perversion and the excess being the logical outcome of
repressing. Nowhere is this carried further than in The Texas Chainsaw Massacre. Here
171
sexuality is totally perverted from its functions, into sadism, violence, and cannibalism.
It is striking that there is no suggestion anywhere that Sally is the object of an overtly
sexual threat; she is to be tormented, killed, dismembered, and eaten, but not raped."
Wood, "Return of the Repressed," p. 31.
27■ With some exceptions—for example, the speargun used in the sixth killing in
Friday the 13th, Part III.
28. Stuart Kaminsky, American Film Genres: Approaches to a Critical Theory of
Popular Film (Dayton, Ohio: Pflaum, 1974), p. 107.
29. The shower sequence in Psycho is ptobably the most echoed scene in all of film
history. The bathtub scene in I Spit on Your Grave (not properly speaking a slasher,
though with a number of generic affinities) is to my knowledge the only effort to reverse
the terms.
30. Schoell, Stay Out of the Shower, p. 35. It may be argued that Blood Feast (1963),
in which a lame Egyptian caterer slaughters one woman after another for their body parts
(all in the service of Ishtar), provides the serial-murder model.
31. This theme too is spoofed in Motel Hell. Farmer Vincent's victims are two hook¬
ers, a kinky couple looking for same (he puts them in room 1 of the motel), and Terry and
her boyfriend Bo, out for kicks on a motorcycle. When Terry (allowed to survive) won¬
ders aloud why someone would try to kill them, Farmer Vincent answers her by asking
pointedly whether they were married. "No," she says, in a tone of resignation, as if
accepting the logic.
32. "Scenes in which women whimper helplessly and do nothing to defend them¬
selves are ridiculed by the audience, who find it hard to believe that anyone—male or
female—would simply allow someone to kill them with nary a protest." Schoell, Stay
Out of the Shower, pp. 55-56.
33. Splatter University (1984) is a disturbing exception. Professor Julie Parker is
clearly established as a Final Girl from the outset and then killed just after the beginning
of what we are led to believe will be the Final Girl sequence (she kicks the killer, a psy¬
chotic priest-scholar who keeps his knife sheathed in a crucifix, in the groin, runs for the
elevator—and then is trapped and stabbed to death). So meticulously are the conventions
observed, and then so grossly violated, that we can only assume sadistic intentionality.
This is a film in which (with the exception of an asylum orderly in the preface) only
females are killed and in highly sexual circumstances.
34. This film is complicated by the fact that the action is envisaged as a living
dream. Nancy finally kills the killer by killing her part of the collective nightmare. See
note 22 above.
35. Spoto, Dark Side of Genius, p. 454. See also William Rothman, Hitchcock: The
Murderous Gaze (Cambridge, Mass.: Harvard University Press, 1982), pp. 246-341.
36. Edgar Allan Poe, "The Philosophy of Composition," in Selected Prose, Poetry
and Eureka (San Francisco: Rhinehart Press, 1950), p. 425.
37. Quoted in Schoell, Stay Out of the Shower, p. 56.
38. Quoted in ibid., p. 41.
39. Spoto, Dark Side of Genius, p. 483.
40. Silvia Bovenschen, "Is There a Feminine Aesthetic?" New German Critique 10
(Winter 1977): 114. See also Doane, "Misrecognition and Identity."
41. E. Ann Kaplan, Women and Film: Both Sides of the Camera (New York and Lon¬
don: Methuen, 1983), p. 15. The discussion of the gendered "gaze" is lively and extensive.
See, above all, Laura Mulvey, "Visual Pleasure and Narrative Cinema," Screen 16, no. 3
(Autumn 1975): 6-18; also see Christine Gledhill, "Recent Developments in Feminist
Criticism," Quarterly Review of Film Studies 3, no. 4 (Fall 1978): 458-93.
42. Wood, "Beauty Bests the Beast," p. 64.
43. The locus classicus in this connection is the view-from-the-coffin shot in Carl
172
Carol J. Clover
Dreyer's Vampyr [ 1932), in which the I-camera sees through the eyes of a dead man. See
Nash, "Vampyr and the Fantastic," esp. pp. 32-33. The 1986 remake of The Little Shop
of Horrors (originally a low-budget horror film, made the same year as Psycho in two days)
lets us see the dentist from the proximate point of view of the patient's tonsils.
44. Two points in this paragraph deserve emending. One is the suggestion that rape
is common in these films; it is in fact virtually absent, by definition (see note 26 above).
The other is the characterization of the Final Girl as "sexy." She may be attractive
(although typically less so than her friends), but she is with few exceptions sexually
inactive. For a detailed analysis of point-of-view manipulation, together with a psycho¬
analytic interpretation of the dynamic, see Steve Neale, "Halloween: Suspense, Aggres¬
sion, and the Look," Framework 14 (1981): 25-29; reprinted in Planks of Reason: Essays
on the Horror Film, ed. Barry Keith Grant (Metuchen, N.J.: Scarecrow Press, 1984), pp.
331-45.
45. Wood is struck by the willingness of the teenage audience to identify "against"
itself, with the forces of the enemy of youth: "Watching [The Texas Chainsaw Massacre]
recently with a large, half-stoned youth audience, who cheered and applauded every one
of Leatherface's outrages against their representatives on the screen, was a terrifying
experience" ("Return of the Repressed," p. 32).
46. "I really appreciate the way audiences respond," Gail Anne FFurd, producer of
Aliens, is reported to have said in the San Francisco Examiner Datebook, August 10,
1986, p. 19. "They buy it. We don't get people, even rednecks, leaving the theatre saying,
'That was stupid. No woman would do that.' You don't have to be a liberal ERA supporter
to root for Ripley." Time, July 28, 1986, p. 56, suggests that Ripley's maternal impulses
(she squares off against the worst aliens of all in her quest to save a little girl) give the
audience "a much stronger rooting interest in Ripley, and that gives the picture reso¬
nances unusual in a popcorn epic."
47. "When she [the mother] referred to the infant as a male, I just went along with
it. Wonder how that child turned out—male, female, or something else entirely?" The
birth is understood to be parthenogenetic, and the bisexual child, literally equipped with
both sets of genitals, is figured as the reborn Christ.
48. Linda Williams, "When the Woman Looks," in Re-Vision: Essays in Feminist
Film Criticism, ed. Mary Ann Doane, Patricia Mellencamp, and Linda Williams (Freder¬
ick, Md.: University Publications of America/American Film Institute, 1984), p. 90.
Williams's emphasis on the phallic leads her to dismiss slasher killers as a "non-specific
male killing force" and hence a degeneration in the tradition. "In these films the recog¬
nition and affinity between woman and monster of classic horror film gives way to pure
identity: she is the monster, her mutilated body is the only visible horror" (p. 96). This
analysis does justice neither to the obvious bisexuality of slasher killers, nor to the new
strength of the female victim. The slasher film may not, in balance, be more subversive
than traditional horror, but it is certainly not less so.
49. Freud, "The Uncanny," p. 245. See also Neale, "Halloween," esp. pp. 28-29.
50. "The woman's exercise of an active investigating gaze can only be simultane¬
ous with her own victimization. The place of her specularization is transformed into the
locus of a process of seeing designed to unveil an aggression against itself." Mary Ann
Doane, "The Woman's Film: Possession and Address," in Re-Vision, p. 72.
51. John Carpenter, interviewed by Todd McCarthy, "Trick or Treat," Film Com¬
ment 16, no. 1 (January-February 1980): 23-24.
52. This is not so in traditional film or in heterosexual pornography, in any case.
Gay male pornography films female bodies.
53. Compare the visual treatment of the (male) rape in Deliverance (1972) with the
(female) rapes in Hitchcock's Frenzy (1972), Craven's Last House on the Left (1972), or
Bergman's The Virgin Spring (1959). The latter films study the victims' faces at length
173
and in close-up during the act; the former looks at the act intermittently and in longshot,
focusing less on the actual victim than on the victim's friend who must look on.
54. Marcus, The Other Victorians, pp. 260-61. Marcus distinguishes two phases in
the development of flagellation literature: one in which the figure being beaten is a boy,
and a second in which the figure is a girl. The very shift indicates, at some level, the irrel¬
evance of apparent sex. "The sexual identity of the figure being beaten is remarkably
labile. Sometimes he is represented as a boy, sometimes as a girl, sometimes as a combi¬
nation of the two—a boy dressed as a girl, or the reverse." The girls often have sexually
ambiguous names as well. The beater is a female but, in Marcus's reading, a phallic one—
muscular, possessed of body hair—representing the father.
55. Ibid., pp. 125-27.
56. "Suspense is like a woman. The more left to the imagination, the more the
excitement.... The perfect 'woman of mystery' is one who is blonde, subtle, and Nordic.
... Movie titles, like women, should be easy to remember without being familiar, intrigu¬
ing but never obvious, warm yet refreshing, suggest action, not impassiveness, and finally
give a clue without revealing the plot. Although I do not profess to be an authority on
women, I fear that the perfect title, like the perfect woman, is difficult to find." Quoted
in Spoto, Dark Side of Genius, p. 431.
57. This would seem to be the point of the final sequence of De Palma's Blow Out,
in which we see the boyfriend of the victim-hero stab the killer to death but later hear
the television announce that the woman herself vanquished the killer. The frame plot of
the film has to do with the making of a slasher film (Co-Ed Frenzy), and it seems clear
that De Palma means his ending to stand as a comment on the Final Girl formula of the
genre. De Palma's (and indirectly Hitchcock's) insistence that only men can kill men, or
protect women from men, deserves a separate essay.
58. The term is Judith Fetterly's. See her The Resisting Reader, A Feminist
Approach to American Fiction (Bloomington: Indiana University Press, 1978).
59. On the possible variety of responses to a single film, see Norman N. Holland,
"Ting Film," Critical Inquiry 12 (Summer 1986): 654-71.
60. Marcus, The Other Victorians, p. 127. Marcus contents himself with noting that
the scene demonstrates a "confusion of sexual identity." In the literature of flagellation,
he adds, "this confused identity is also present, but it is concealed and unacknowledged."
But it is precisely the femaleness of the beaten figures that does acknowledge it.
61. Freud, "The Uncanny," esp. pp. 219-21 and 226-27.
62. Paymond Durgnat, Films and Feelings (London: Faber and Faber, 1967), p. 216.
63. Not a few critics have argued that the ambiguity is the unintentional result of
bad filmmaking.
64. So argues Susan Barrowclough: the "male spectator takes the part not of the
male but of the female. Contrary to the assumption that the male uses pornography to
confirm and celebrate his gender's sexual activity and dominance, is the possibility of his
pleasure in identifying with a 'feminine' passivity or subordination." "Not a Love Story,"
Screen 23, no. 5 (November-December 1982): 35-36. Alan Soble seconds the proposal in
his Pornography: Marxism, Feminism, and the Future of Sexuality (New Haven, Conn.:
Yale University Press, 1986), p. 93. Porn/sexploitation filmmaker Joe Sarno: "My point
of view is more or less always from the woman's point of view; the fairy tales that my
films are based on are from the woman's point of view; I stress the efficacy of women for
themselves. In general, I focus on the female orgasm as much as I can." Quoted in Vale
and Juno, eds., Incredibly Strange Films, p. 94. "Male identification with women," Kaja
Silverman writes, "has not received the same amount of critical attention [as sublima¬
tion into professional 'showing off' and reversal into scopophilia], although it would seem
the most potentially destabilizing, at least as far as gender is concerned." See her discus¬
sion of the "Great Male Renunciation" in "Fragments of a Fashionable Discourse," in
174
Carol J. Clover
Studies in Entertainment: Critical Approaches to Mass Culture, ed. Tania Modleski

(Bloomington: Indiana University Press, 1986), p. 141.
65. Elaine Showalter, "Critical Cross Dressing: Male Feminists and the Woman of
the Year," Raritan 3 (Fall 1983): 138.
66. Whatever its other functions, the scene that reveals the Final Girl in a degree of
undress serves to underscore her femaleness. One reviewer of Aliens remarks that she
couldn't help wondering why in the last scene, just as in Alien, "we have Ripley wan¬
dering around clad only in her underwear. A little reminder of her gender, lest we lose
sight of it behind all that firepower?" Christine Schoefer, East Bay Express, September 5,
1986, p. 37.
67. Mulvey, "Visual Pleasure and Narrative Cinema," p. 12.
68. Kaja Silverman, "Masochism and Subjectivity," Framework 12 (1979): 5. Need¬
less to say, this is not the explanation for the girl hero offered by the industry. Time, July
28, 1986, p. 44, on Aliens: "As director Cameron says, the endless 'remulching' of the
masculine hero by the 'male-dominated industry' is, if nothing else, commercially short¬
sighted. 'They choose to ignore that 50% of the audience is female. And I've been told
that it has been proved demographically that 80% of the time it's women who decide
which film to see.'" It is of course not Cameron who established the female hero of the
series but Ridley Scott (in Alien), and it is fair to assume, from his careful manipulation
of the formula, that Scott got her from the slasher film, where she has flourished for some
time among audiences that are heavily male. Cameron's analysis is thus both self-serv¬
ing and beside the point.
69. If this analysis is correct, we may expect horror films of the future to feature
Final Boys as well as Final Girls. Two recent figures may be incipient examples: Jesse, the
pretty boy in A Nightmare on Elm Street, Part 2: Freddy’s Revenge (1985), and Ashley,
the character who dies last in The Evil Dead (1983). Neither quite plays the role, but their
names, and in the case of Jesse the characterization, seem to play on the tradition.
70. For the opposite view (based on classic horror in both literary and cinematic
manifestations), see Franco Moretti, "The Dialectic of Fear," New Left Review, no. 136
(1982): 67-85.
71. Vale and Juno, eds., Incredibly Strange Films, p. 5.
72. Tania Modleski, "The Terror of Pleasures: The Contemporary Horror Film and
Postmodern Theory," in Studies in Entertainment, pp. 155-66. (Like Modleski, I stress
that my comments are based on many slashers, not all of them.) This important essay
(and volume) appeared too late for me to take it into full account here.
»
Stephen Prince
The Aesthetic of Slow-Motion

Violence in the Films
of Sam Peckinpah
Violence assumes a myriad of forms in contemporary cinema, but the

aesthetic of violence has evolved to emphasize a cluster of predominant
and recurring characteristics. In scenes involving gunplay, editors
assemble footage filmed with multiple cameras into complex mon¬
tages, intercutting normal-speed footage with slow-motion imagery.
This style—multicamera montage with slow motion—has become the
predominant aesthetic form for rendering gun battles in modern cin¬
ema. In their work on The Wild Bunch (1969), director Sam Peckinpah
and editor Lou Lombardo gave this stylistic the emphasis and elabora¬
tion that transformed it into the essential template that it has now
become, and Peckinpah has remained the most inventive and influen¬
tial director to use this aesthetic. Accordingly, after examining the
roots of this aesthetic, I examine Peckinpah's elaboration and trans¬
formation of it, his objectives and the controversies his work has
provoked, and the implications of that work for later filmmakers.
Peckinpah used the montage aesthetic to break with realism in
order to substitute a stylized rendition of violence. This point is most
important, given the common critical (and wrongheaded) view that
Peckinpah's signal contribution in the late 1960s was to bring to Amer¬
ican cinema a more realistic depiction of violence whereby the blood¬
less deaths portrayed in previous decades of film gave way to a more
forthright, and truer, presentation of gore. This view has tended to
obscure the important point that Peckinpah aimed to stylize his mate¬
rials and that this stylization proceeded from his conviction that it was
From Savage Cinema: Sam Peckinpah and the Rise of Ultiaviolent Movies by Stephen
Prince, copyright © 1998. Courtesy of the University of Texas Press.
175
176
Stephen Prince
the only way to wake people up to violence in a culture—late 1960s

America—whose convulsions, he believed, had anesthetized them to
bloody death. Peckinpah felt that people had become inured to violence
through the medium of television, which domesticated the violence of
the Vietnam War and, by sandwiching it between commercials, insin¬
uated it into the daily routines of consumer life. In a culture and dur¬
ing a period so heavily saturated with killing, people had become oddly
desensitized, he believed, and he felt that by heightening violence
through the artifice of style he could break the cycle of consumption in
which the era's disturbing social violence was embedded. In this regard,
he was a filmmaker with a didactic social agenda, and he aimed to place
camera style at the service of that agenda:
We watch our wars and see men die, really die, every day on television,
but it doesn't seem real. We don't believe those are real people dying
on that screen. We've been anesthetized by the media. What I do is
show people what it's really like—not by showing it as it is so much
as by heightening it, stylizing it. . . . The only way I can do that is by
not letting them gloss over the looks of it, as if it were the seven
o'clock news from the DMZ. When people complain about the way I
handle violence, what they're really saying is, "Please don't show me;
I don't want to know; and get me another beer out of the icebox."1
The montage aesthetic served these didactic intentions because

it was a decisive break with previous screen traditions of representing
violence and with the unremarkable visual presence of daily TV vio¬
lence. Peckinpah insisted that his visual approach was a reaction
against the existing movie traditions, which he considered to be mis¬
leading and grossly out of step with the times. He believed that "vio¬
lence in motion pictures is usually treated like fun and games."2 In a
letter to Paul Staniford, a lawyer and friend of the Peckinpah family (and
whose name Peckinpah gave to a character in Ride the High Country),
Peckinpah declared, "I personally feel it's time Hollywood quit glam¬
orizing violence and let people see how brutalizing and horrible it really
is. This is what I tried to do [in The Wild Bunch]."3 By breaking the
established representational conventions, Peckinpah hoped to convey
the horror of violence to viewers he believed had been rendered com¬
placent by decades of painless, bloodless movie killings. In late 1960s
America, the traumatic impact of real social violence was misaligned
with the tradition of sanitized movie violence.
For Peckinpah, conventional movie violence and television
news performed a narcotizing function, insulating people from the
177
The Aesthetic of Slow-Motion Violence in the Films of Sam Peckinpah
Figure 30. Sam Peckinpah, the modern poet of screen violence, popularized the essential film style
for depicting gun violence.
events around them. His belief in this narcotizing function was consis¬
tent with the radical critique of popular culture in that period (as repre¬
sented in the writings of Susan Sontag and Herbert Marcuse).4 By using
graphic imagery of bloodletting and the montage aesthetic, Peckinpah
aimed to bring the era's violence inside the movie theater, which would
no longer function as a place of refuge by shielding viewers from horrific
images. His work would place the filmic representation of violence into
proper synchronicity with the era, whose convulsions engulfed the sen¬
sibilities of filmmaker and audience alike. His object in doing this was
to create a socially beneficial effect. "To negate violence it must be
178
Stephen Prince
shown for what it is," he argued, "a horrifying, brutalizing, destructive,

ingrained part of humanity."5 If the narcotizing functions of the media
were broken, Peckinpah believed, people would see violence for what
it is and thereby stand a chance of gaining more control over it and its
destructive effects on a nation in turmoil. After receiving a letter from
a viewer critical of The Wild Bunch, Peckinpah wrote back and asked
that the viewer consider that the graphic screen carnage had these larger
objectives. "I am sorry you did not enjoy The Wild Bunch. Perhaps
some of its vulgarity and violence will remain with the people who will
see it and they will understand better the nature of this continuing
plague that infects our country."6
With this description of Peckinpah's objectives in mind, we
should now examine the origins of the aesthetic and the distinctive
structure of the montage editing in his films, its operation and effects.
What are the characteristic manipulations of time and space that this
aesthetic makes possible? And what are their effects on the viewer?
Does the montage aesthetic enable a filmmaker to develop a critical
perspective on the represented violence, or does it tend, instead, toward
spectacle and the incitement of excitatory responses?
Antecedents of Peckinpah’s Montage Style
The essential influences on his montage aesthetic are easily identified.

The most important influence is the work of Akira Kurosawa because
it was Kurosawa who first showed filmmakers how to intercut slow-
motion and normal-speed footage in scenes of violence. As an initial
transposition of Kurosawa's work and the immediate stimulus for The
Wild Bunch, there is Arthur Penn's demonstration of slow-motion and
multicamera filming in Bonnie and Clyde. In a distant context, of
course, there is the montage editing of Sergei Eisenstein. (Other prece¬
dents, more prosaic, perhaps, were also important. While working in
the early 1950s as a stagehand at KLAC-TV in Los Angeles, Peckinpah
watched an experimental film made by another station employee. It
included a ^low-motion shot of a falling lightbulb that intrigued Peck¬
inpah. Also, prior to cutting The Wild Bunch, Lou Lombardo showed
Peckinpah an episode of the TV show "Felony Squad" that he had edited
that included some slow-motion work during a gunfight.)7
Kurosawa, whose Rashomon (1950) Peckinpah always cited as
179
a favorite film, doubtless because its theme of the relativity of truth

deeply appealed to his sense of irony, exerted a decisive stylistic influ¬
ence on Peckinpah's work in several ways. Beginning with Seven Samu¬
rai (1954), Kurosawa customarily used from three to five cameras
running simultaneously to film his scenes. This approach gave him
much better coverage of that film's complexly choreographed fight
scenes and also helped to elicit better performances from the actors by
extending the length of each take. Peckinpah obviously appreciated the
strategic advantages that multicamera filming afforded the shooting of
action scenes. For the scene in Bring Me the Head of Alfredo Garcia,
where the professional killers Quill and Sappensly massacre a Mexican
family, Peckinpah used five cameras, two of which were running at high
speed to produce slow motion.8 (Slow motion is produced by filming at
very high speeds—this captures the action on a much greater number
of frames that, when projected at the industry standard of twenty-four
frames per second, will appear as slow motion.) On Junior Bonner, for
the critical sequence in which the bulldozers wreck Ace's house, Peck¬
inpah employed four cameras: two high-speed cameras with zoom
lenses at opposite angles on the house, an Arriflex wide angle for a long-
shot on the front of the house, and an Arriflex telephoto for a low angle
on the right of the house.9 Covering the action simultaneously from so
many angles and with different camera speeds amplifies the material
available for montage editing, an obvious strategic advantage given
Peckinpah's desires to transform the normative conventions of Ameri¬
can cinema.
Kurosawa's multicameras, though, described a fixed and unique
geometry of space. They were often set at right angles to one another so
that the cuts shift the viewer's axis of vision by 90 degrees. While space
in Kurosawa's cinema is extremely angular, the disjunctiveness of his
cutting is softened somewhat by the recurrent regularities of these 90-
degree perspective realignments. The angularity of Peckinpah's cutting
exhibits none of the rectilinear "normality" of Kurosawa's 90-degree-
angle shifts. Peckinpah's angularity is totally acute or oblique, always
off-center, and, as a result, it imposes a much higher degree of fragmen¬
tation upon the space that it carves up. Peckinpah learned from Kuro¬
sawa's disjunctive editing of space and carried its implications much
further, as the cutting throughout Straw Dogs clearly demonstrates.
Kurosawa's cinema also taught Peckinpah about the perspec¬
tive-distorting effects of telephoto lenses, a signature Kurosawa ele¬
ment that became a signature Peckinpah element, because the
180
Stephen Prince
telephoto lens works extremely well in conjunction with multicamera

filming. By equipping multicameras with long-focal-length lenses, the
cameras can be positioned more easily about the periphery of the set.
Since the focal length of the lenses will produce a narrow field of view
that can be used to prevent the cameras from seeing each other, tele¬
photo lenses facilitate the blocking of multicamera positions. Peckin¬
pah quickly grasped the implications of this advantage. Ride the High
Country, a nonmon.tage-based film, does not conspicuously utilize tele¬
photo lenses, while The Wild Bunch, a montage and multicamera film,
clearly does, as do many of his later films.
In addition to the multicamera filming, disjunctively angular
cutting, and reliance on telephoto lenses that Peckinpah found in Kuro¬
sawa's work, the most explicit area of influence from Kurosawa to Peck¬
inpah is, of course, Kurosawa's exploration of slow motion within
scenes of violent death. (Kurosawa's use of slow motion was also an
important influence on Walon Green in his thinking about the script
for The Wild Bunch, although he did not write out these ideas. He told
an interviewer, "The violence in slow motion is very expressly in the
script. I put the slow motion in because when I wrote it, I had just seen
The Seven Samurai, which had the first use of slow motion in an action
scene that I'd ever seen." Peckinpah scholar Paul Seydor points out that
Green told him this claim was an error: that he did not, in fact, write
slow motion into the script, but was thinking about it while working
on the script because of the Kurosawa film, which had tremendously
impressed him.)10
This interest appeared as early as Kurosawa's first film, San-
shiro Sugata (1943), but it was Seven Samurai (1954), widely seen and
admired in the West, with its all-male band of heroes, adventure narra¬
tive, and martial values, that explicitly demonstrated the stylistic pat¬
terning that the intercutting of footage shot at different camera speeds
could bring to the dramatic and temporal rhythms of a scene. In this
respect, Seven Samurai furnished the stylistic template utilized and
popularized by Peckinpah and by Arthur Penn and used by every action
scene choreographer ever since. Kurosawa's work occasioned Peckin¬
pah's famous remark, following completion of Ride the High Country,
"I'd like to ,be able to make a Western like Kurosawa makes West¬
erns."11 Weddle's biography claims that Kurosawa's use of slow motion
was primitive compared to the complexities introduced by Peckinpah.
"Editorially it was static. The weaving of slow motion into the very fab¬
ric of a sequence . . . had still to be achieved."12 While it is true that
Peckinpah used slow motion far more extensively than Kurosawa (and
181
The Aesthetic of Slow-Mgtion Violence in the Films of Sam Peckinpah
that he began using it well before The Wild Bunch, as Weddle and Sey-
dor have pointed out), who seemed to have only an occasional interest
in the device, the essentials of Peckinpah's usage are clearly already
contained in Seven Samurai.
Contrary to Weddle's claims, Kurosawa does cut in and out of
the slow-motion footage in a dynamic manner. During the scene where
the leader of the samurai, Kambei (Takashi Shimura), rescues a kid¬
napped child from a crazed thief and kills the thief with a short sword,
Kurosawa dynamically intercuts footage filmed at normal speed with
slow-motion footage so that the rhythms of the scene oscillate between
these two different temporal modes. The mortally wounded thief
crashes through the hut's doorway to the village square outside, where
amazed onlookers witness his dying. Kurosawa intercuts three slow-
motion shots of the thief crashing through the door, running a few steps
forward, and rising up on tiptoe with three normal-speed shots of the
onlookers' reactions. Since movement also occurs in these shots, the
scene builds an internal tension between these differing rhythms. After
Figure 31. In Seven Samurai, Kurosawa intercuts slow-motion shots of the dying thief (fore¬
ground) with normal-speed shots of onlookers. This scene provided the essential template for mod¬
ern movie violence, showing other filmmakers how to integrate footage shot at differing camera
speeds to aestheticize violence.
182
Stephen Prince
these six shots, Kurosawa shows the thief fall to the ground in slow
motion but without the sound of an impact. This sound has been with¬
drawn from the scene, setting up a dynamic visual-acoustic conflict
that accompanies the temporal conflicts.
But the visual-acoustic conflicts in the film are more subtle
still and hold yet greater relevance for Peckinpah's work and for the
slow-motion aesthetic. During the scene's slow-motion shots, Kuro¬
sawa includes amplified sound effects—the baby's cry, the mother's
scream, howling wind—which involve no temporal distortion. Ampli¬
fied, but temporally unmodified, sound accompanies the slow-motion
images, and this disjunction emphasizes the heightened artifice of these
images, their uniquely expressive power. What we have yet to appreci¬
ate is not only how fundamentally Kurosawa's brief exploration of
slow-motion effects influenced Peckinpah's work, but also how Kuro¬
sawa had already made the essential discovery that a temporal non¬
synchrony of image and sound accentuates the contrast of footage shot
at differing camera speeds. The normal-speed sound emphasizes the
otherness of the slow-motion image. This visual-acoustic principle is
basic to the expressive power of such sequences.
Following Kurosawa, Peckinpah manipulates sound during the
slow-motion shots by amplifying selected effects. The cries of the baby
and mother and the howling wind are selectively amplified during the
slow-motion shots in Seven Samurai to accentuate the temporal mis¬
match between the audiovisual tracks. In the opening shoot-out in The
Wild Bunch, when an outlaw crashes his horse through a glass window
in slow motion, the sound of shattering glass heard at normal tempo is
mixed above the general battle sounds to accentuate the temporal mis¬
matches, and when Dutch (Ernest Borgnine) falls off his horse into a
wooden structure in slow motion, the amplified crunch of wood is the
dominant sound. In Peckinpah's work, the slow-motion image is care¬
fully contrasted with amplified sound effects to create an intermodal,
cross-sensory montage, and, as we have established, the expressive
power of these combinations had been explicitly demonstrated by
Kurosawa.
Kurosawa's disjunctive editing and audiovisual combinations
are indebte4 to the montage tradition exemplified by Eisenstein and so,
too, therefore, is Peckinpah's work. Care needs to be exercised in
extending these comparisons, however. Eisenstein's montage princi¬
ples belong to a rich and specific ideological and social context that
informed his filmmaking, a context that does not translate to either
183
The Aesthetic of Slow-Mption Violence in the Films of Sam Peckinpah
Kurosawa or Peckinpah, who both employed a much more limited use

of montage. The Odessa Steps massacre in Potemkin, for example, fea¬
tures a complex orchestration of graphic, volumetric, kinetic, and tem¬
poral elements in a design whose intricacy surpasses the narrower range
of manipulations Peckinpah carried out in his montages. This is why
Eisenstein looms as a more distant and general example for Peckinpah's
cinema rather than as a direct and immediate influence. Despite this
caveat, it is worth noting that Peckinpah's artistic relationship with
Kurosawa and Eisenstein achieved a degree of self-consciousness dur¬
ing the production of Peckinpah's films. On Cross of Iron, production
designer Ted Haworth told Peckinpah that a particular effect would be
"Kurosawa Peckinpah at his best."13 Discussing with Peckinpah the
story structure as scripted for The Wild Bunch, producer Phil Feldman
noted that an improperly placed scene with the bounty hunters "would
interrupt your Eisenstein structure."14
While Kurosawa's work had impressed Peckinpah, it was Arthur
Penn's demonstration in Bonnie and Clyde of Kurosawa-style montage
that fired Peckinpah with the determination to outdo the level of blood¬
shed—and its aesthetic rendering—that Penn had achieved. Penn's
work is historically important in this context because of its timing—
Bonnie and Clyde appeared two years before The Wild Bunch. Fur¬
thermore, as with Peckinpah, Penn was a stylistic disciple of Kurosawa,
whose editing in Seven Samurai had furnished the essential template.
Penn graciously acknowledged Kurosawa's importance for the multi¬
speed montage that caps Bonnie and Clyde. Discussing his conceptu¬
alization of that scene, he remarked, "Having seen enough Kurosawa
by that point, I knew how to do it."15 Throughout his career, Penn
remained a vivid stylist of violence, but he never again filmed anything
like the slow-motion, multicamera carnage that caps Bonnie and
Clyde. Thus, although that is a seminal film, Penn remains an inter¬
mediary figure—not a primary one—in the history of modern screen
violence, standing between Kurosawa (as the stylistic mentor) and
Peckinpah (as the filmmaker who became-the exponent of slow-motion
violence). In this regard, it is important to understand the contribution
of Bonnie and Clyde to Peckinpah's work, but equally important to rec¬
ognize that, while Penn moved on to make films on other topics, Peck¬
inpah remained preoccupied by the issue of violence in human life and
as a subject for film.
Peckinpah claimed to have seen Bonnie and Clyde only after
finishing The Wild Bunch and claimed ownership of Penn's slow-
184
Stephen Prince
motion work, remarking "they did all my shtick."16 Despite this claim,
however, wardrobe supervisor Gordon Dawson recollected that Peck¬
inpah wanted to surpass Bonnie and Clyde's violence and stylistics
while in production on The Wild Bunch. Furthermore, in a letter of
March 19, 1968, Warner Bros, confirmed with the film's production
manager that a print of Bonnie and Clyde would be shipped to Peckin¬
pah's Mexico location for a screening the next weekend (March 23-24).
This was immediately prior to the start of principal photography on the
25th. Peckinpah was studying Penn's film and wanted to see it before
commencing work on his own. He knew exactly what he was doing on
The Wild Bunch and how it related to Penn's achievements in Bonnie
and Clyde.17
If Peckinpah did not discover anything new about intercutting
slow-motion shots into a montage sequence, he undeniably extended
and built upon the principles informing Kurosawa's editing of multi¬
camera footage. Like Penn's use of slow motion within the bloody mon¬
tage that concludes Bonnie and Clyde, which Penn said conveyed
"both the spastic and the balletic" qualities of the gangsters' death ago¬
nies,18 Peckinpah's editing emphasized the brutality of physical vio¬
lence while also giving it a graceful beauty. This contradiction between
the aesthetic beauty of the visual spectacle and the emotional and phys¬
ical pain that Peckinpah also dramatized as part of his screen violence
is a complex and important one, and Peckinpah used slow motion and
montage to stylize screen violence in ways that far surpassed what Penn
had done in Bonnie and Clyde. An comparison with Penn's films can
help illuminate why Peckinpah's slow-motion inserts are more strik¬
ing and achieve a more heightened stylistic intensity than do Penn's.
Penn began to explore slow motion in The Left-Handed Gun
(1958), about the legend of Billy the Kid, during the scene where Billy
(Paul Newman) kills Deputy Ollinger (Denver Pyle), a scene that Peck¬
inpah closely re-creates in Pat Garrett and Billy the Kid. It is important
to note here that Penn's film contains numerous images and bits of
business that Peckinpah borrowed for his own film, including the
Christ-pose Billy adopts when Garrett arrests him following the Stink¬
ing Springs shoot-out. Both films also reference the fascinated reaction
of children, to violent death, a major Peckinpah preoccupation. In
Penn's film, a little girl runs out and laughs at the dead Ollinger, who
has been blown out of his boot, while Peckinpah's film shows kids play¬
ing and laughing on the gallows that has been erected for Billy. And,
again, much of this can be traced back to Kurosawa, who in Seven
185
Samurai had shown children playing and climbing on the fortress walls
built to protect the village from the bandits and across which much
killing would occur.
When Billy shoots Ollinger, Penn cuts from a slow-motion shot
of Ollinger waving his arms and starting to fall backward to a fast-
motion shot of his body hitting the ground. The transition from slow
to fast motion is abrupt, and as a stylistic design it is clearly an exper¬
imental effort. "I was just playing with the medium/' says Penn.19 The
experimentation doesn't work very well because the action rendered in
slow motion—Ollinger flailing his arms—is not effectively suited for
the temporal manipulation. It is neither balletic nor spastic. By contrast,
in Bonnie and Clyde, Penn more shrewdly incorporates slow motion by
intercutting it at multiple points with the jerky convulsing of the gang¬
sters as they are riddled with bullets. By alternating between slow and
apparently accelerated tempos (the apparent acceleration produced at
normal film speed by virtue of the Texas Rangers' fast rate of fire), Penn
successfully brings out the balletic and spastic qualities of the scene.
Furthermore, Penn switches to slow motion at a more judicious
moment than in The Left-Handed Gun. After a quick series of glances
between Bonnie and Clyde that conveys their awareness of what is about
to happen, Clyde runs toward Bonnie, at which point they are raked
with machine-gun fire. As Clyde starts to fall, Penn switches to slow
motion for the first time so that the arc of Clyde's dying fall is poeti¬
cally extended.
The imagery is extremely vivid, and it discloses a fundamental
principle that Peckinpah would observe in his own films: slow motion
is especially powerful when it correlates with a character's loss of phys¬
ical volition. Clyde's dying arc; the trajectories of falling, dying men shot
from the rooftops of San Rafael or the army personnel blasted off the flat¬
cars of the train in The Wild Bunch; the Gorch brothers dead on their
feet but kept up convulsively by the impact of bullets fired by Mapache's
men.; Holly's mortal fall, blasted backward across the saloon by the force
of Pat Garrett's shot; Lt. Triebig's grotesque writhing under Steiner's
machine gun fire in Cross of Iron—all of these slow-motion images
derive their poetic force from the metaphysical paradox of the body's
continued animate reactions during.a moment of diminished or extin¬
guished consciousness. Slow motion intensifies this paradox by pro¬
longing it. It is not just the moment of violent death which is extended,
but the mysteries inherent in that twilit zone between consciousness
and autonomic impulse, that awful moment when a personality ceases
186
Stephen Prince
to inhabit a body that is still in motion. Peckinpah, and Penn, intensi¬

fied the trauma of violent death by visualizing this loss of human voli¬
tion in a tangle of rioting flesh and nerve. To achieve maximum
intensity on this point, it was necessary to employ extremely fast cam¬
era speeds, for only by slowing down the action could the metaphysical
poetry of these scenes be elicited. This is why the slow-motion insert in
The Left-Handed Gun does not work very well. The slow speed is not
slow enough, and Ollinger has not yet lost control of his body.
As we can now see, ample precedent existed in the films of Kuro¬
sawa and Penn for the stylistic inflections that Peckinpah would
explore. However, with characteristic solipsism, Peckinpah claimed to
have gained insight into the cinematic usage of slow motion through
personal experience. He got into the habit of telling interviewers that
during his military service in China in 1945 he realized how slow
motion might apply to such scenes after seeing a Chinese passenger shot
while riding on a train. Peckinpah called it one of the longest split sec¬
onds of his life.20 On other occasions when he told the story of learning
about slow motion, it was he who had been shot: "I was shot once and
I remembered falling down and it was so long ... I noticed that time
slowed down and so I started making pictures where I slowed down
time, because that's the way it is."21 We should be very skeptical of these
claims, because they sound like retrospective attempts to justify a sty¬
listic inflection in the face of hostile critical reception (criticism that
Peckinpah's slow motion was self-indulgent) by attributing to the style
an empirical and phenomenological foundation in personal experience.
If, as he claims, Peckinpah's slow motion has its basis in real
perceptual experience, then—and this is the implied message to his
critics—he is no exploiter and glorifier of screen violence but merely an
observer of the psychological reality of living through a violent experi¬
ence. But, phenomenologically, there seems no necessity for equating
the vividness of a brief traumatic episode with a subjective sense of
extended duration. This may occur, and perhaps it did for Peckinpah,
but it does not seem to be a necessity. It seems more likely that Peck¬
inpah was struck by the stylistic manipulations of Kurosawa and Penn,
began trying them himself, and subsequently projected his World War
II memory tinto the results. Through their montage structures, Peck¬
inpah's films effect a formal transformation of violence, not an imita¬
tion of its psychological contours. These montages may incorporate
psychological dimensions of meaning, but they function as aesthetic
translations of the idea of violence, not as mimetic constructions that
seek to imitate faithfully the contours of an experience. The complex-
187
The Aesthetic of Slow-Mbtion Violence in the Films of Sam Peckinpah
ity of these formal transformations constitutes one of Peckinpah's

claims to being a great filmmaker, not his dubious assertions about the
psychological basis of his slow-motion editing.
Slow-Motion Inserts
Peckinpah's aesthetic transformation of violence through montage led

him toward three principal types of montage construction: the relatively
simple, slow-motion insert crosscut into the body of a normal-tempo
sequence; the synthetic superimposition of multiple lines of action with
radical time-space distortions in a montage set-piece,- and montages
approaching Eisenstein's notion of intellectual editing, wherein the
viewer is moved to cognitively grasp psychological or social truths. I
examine here the first two categories of montage. Peckinpah used the
third category—the psychological and poetic montages—to probe and
visualize the subjective responses of characters to emotional and phys¬
ical trauma. In this regard, he did not use it primarily to stylize gun vio¬
lence, and since it falls outside the strict focus of this essay, I omit it
from discussion here. (Readers interested in this category of montage
should consult my extended discussion in Savage Cinema.)22
Slow-motion inserts crosscut into the body of a normal-tempo
sequence may be found in all of Peckinpah's post-Dundee films. Even
The Ballad of Cable Hogue, distinguished by its use of fast-motion
footage and general absence of montage-based violence, opens with a
Mexican beaded lizard (subbing for an iguana) exploding from gunfire
in a slow-motion shot (followed by a three-frame 'subliminal' flash)
that is inserted (but not crosscut) into the body of an otherwise normal-
tempo sequence. It is easy enough to locate similar moments in the
other films. As previously noted, Holly's backward lurch in Pat Garrett
and Billy the Kid, after Garrett shoots him, describes a beautiful, slow-
motion arc across the saloon floor and is cut into this scene, which is
otherwise free of such temporal distortions. The Wild Bunch's train
heist is edited, for the most part, without temporal distortions, but
when Pike throws the engineer and a crewman off the locomotive, the
editing crosscuts their falling bodies, in slow motion, with the dying
falls of the two soldiers Lyle Gorch shoots off the front of the train. In
Straw Dogs, when the thuggish Tom Hedden (Peter Vaughan) blasts
Major Scott (T. P. McKenna) with his shotgun, three shots crosscut with
other action in the scene show Scott's misshapen body flying backward
188
Stephen Prince
with slow-motion grace. In The Killer Elite, when professional killer

George Hansen (Robert Duvall) executes Vorodny (Helmut Dantine),
three slow-motion close-ups of Dantine falling onto the couch are
crosscut with normal-tempo close-ups of Hansen watching this action.
In the next scene, when Hansen cripples his friend Mike Locken (James
Caan) by shooting him in the elbow and knee, Locken convulses in nor¬
mal time but rolls off the stool onto the floor in slow motion.
I could continue to multiply examples, but the essential point
should be clear. One of Peckinpah's basic montage structures involves
the sudden intrusion of one or more slow-motion details inserted or
crosscut into the body of a sequence whose temporal rhythms are oth¬
erwise normal. The perceptual shock of such intrusions comes from the
sudden disruption of ordinary time through the influx of an alternate
mode of time. In most cases, when squib-work is involved, the explo¬
sion of blood is not the main focus of the slow-motion insert. While the
detonating squibs (electrical firing devices used to simulate bullet hits)
were certainly shocking for audiences in 1969 when Peckinpah
unleashed them, the bulk of the visual attention in the slow-motion
inserts is devoted, as previously noted, to the body's loss of volitional
control over its actions. The exploding squib behind Vorodny's head
when Hansen shoots him is one of Peckinpah's most elaborate and
graphically bloody, but it is only a few frames long, so that it appears as
a flash cut despite occurring in slow motion. The aftereffects—
Vorodny's slow fall onto the couch—take up much more screen time.
Paul Seydor points out that Peckinpah purposely kept his slow-motion
shots brief: "It is the build-up and the release that he wanted to capture,
because perception and feeling, violence as psychological effect, are
what chiefly interested him."23
While this is certainly true for such scenes as Vorodny's killing
or, more remarkably, Garrett's killing of the Kid, the presence of
extended, violent set-pieces in The Wild Bunch, Straw Dogs, Bring Me
the Head of Alfredo Garcia, Cross of Iron, The Killer Elite, and The
Osterman Weekend demonstrate that the act of violence, in itself,
exerted tremendous fascination for Peckinpah. While I argue in Savage
Cinema that Peckinpah's work is distinguished by the emotional and
self-reflective frameworks it builds around the violence that it depicts,
Peckinpah was also obsessive about this concern for violence and
enthralled by the possibilities that cinema offers for visualizing it. His
interests included the build-up and release of tension, the psychologi¬
cal effects of violent action, as well as the action and act of killing itself,
only part of which is visualized in the squib-work.
189
The Aesthetic of Slow-Mdtion Violence in the Films of Sam Peckinpah
The simplest of these cinematic possibilities lies in the

momentary disruption of time by the brief, slow-motion insert placed
to accentuate the lyrical appearance of the human body acted upon by
violent physical forces that have extinguished its ability to respond in
an intentional manner. It seems most probable that Peckinpah kept his
slow-motion imagery brief not because he was interested exclusively
in the psychological effects of violence, but rather because it worked
best that way from a visual standpoint. Brevity accentuates the poetic
effects of slow motion. Too much slow motion, or for too extended a
period, would rob the scenes of their kinetic charge and their physical
edge by making the action seem like it is occurring underwater or in a
strange condition of weightlessness. By quickly (i.e., briefly) punctur¬
ing normal time and space with the slow- motion imagery, Peckinpah
could stress the balletic beauty that, as a filmmaker, he discovered he
could create within a maelstrom of death, and he could retain the sharp
edge of physicality that was essential to his didactic intention.
This physicality is communicated by the normal-tempo
images, not the slow-motion inserts, and by the sound effects that sen¬
suously detail the thud of bullets into flesh, the violent exhalation of
breath, shattering pottery, or crashing glass. When one of the outlaws
is shot from his horse in the opening massacre of The Wild Bunch, we
see rider and horse fall in beautiful slow motion. Because the pair's falls
are so extended and the rate at which man and animal strike the ground
is so gradual, the spill, as an image, lacks a strong physical dimension.
But on the sound track as the horse goes down we hear a loud cracking
sound like a bone breaking, and this gives the image a concreteness that
the slow motion has removed from it. Peckinpah used the expressive
poetry of slow motion to elicit balletic effects and to visualize that
moment when death or grievous wounding robs or threatens to rob the
body of its spirit or personality. He was striking a delicate balance
between the slow-motion inserts and the normal-tempo continuum of
the sequence proper. Too much slow motion would become ludicrous
because it would bog down the violent outburst and remove all sense
of its physical consequences. Slow motion, therefore, had to exist in a
state of tension with the normal-tempo sound track and body proper of
the sequence. Extended slow-motion imagery would not create this req¬
uisite tension. Slow motion had to constitute a brief interlude, dis¬
rupting the texture of the scene to offer a privileged glimpse at the
metaphysical mysteries of violent death. Too long a glimpse and its
effects would be vitiated.
Peckinpah rarely employed extended slow-motion imagery. In
190
Stephen Prince
the climax of Straw Dogs, when David Sumner (Dustin Hoffman) grap¬
ples with Charlie Venner following the shooting of Norman Scutt (both
Venner and Scutt are members of the gang that has invaded the Sum¬
ners' house), Peckinpah presented their struggle in a lengthy series of
slow-motion shots. It is an interesting usage, but it softens the hard
edge the film's violence has heretofore had. The viewer feels that little
harm can come to either David or Charlie while they slog slowly
around as if underwater. Significantly, when David brings the mantrap
down on Charlie's head, the film reverts to normal speed. Furthermore,
because the slow-motion shots have lasted an uncommonly long time,
the transition back out, to normal tempo, feels abrupt and harsh. The
paradoxical thing about the brief slow-motion inserts that typify Peck¬
inpah's work is that they mesh so well with the ordinary temporal con¬
tinuum. When the insert is brief, the editor can slip into and out of the
decelerated moment in a highly fluid manner. Despite the temporal dis¬
ruption, strong continuity prevails, unlike the just-described scene
from Straw Dogs, where the return to normal time occasions a percep¬
tible loss of continuity.
Slow-motion images are not of themselves dynamic. Their ten¬
dency is toward inertia, a deceleration not only of represented time but
of the internal rhythms and pacing of the sequence in which they
appear. They become dynamic with reference to their surrounding con¬
text—the normal-tempo actions against which they play as stylistic
opposites. By maximizing this opposition, Peckinpah and his editors
could give the slow- motion inserts a dynamism which they do not in
themselves possess. Intercutting slow motion with normal speed
became an essential and highly effective way of achieving this. When
two thugs in Bring Me the Head of Alfredo Garcia lose control of their
station wagon, it skids off the road, churning up a huge spray of dirt.
This is rendered in two slow-motion shots that are crosscut with the
startled reactions of passengers on a passing bus. The decelerated action
of the skid is slowed down so much that the resulting images seem
robbed of nearly all movement, which heightens the dynamic contrast
with the surrounding normal-tempo imagery. The viewer experiences
a perceptual shock because of this radical misalignment between the
alternate tempos. Intercutting the two accentuates the misalignment
and the dynamic contrast and works against the tendency for slow
motion to create inertia and a brake on the action.
Later in that film, when the protagonist, Bennie (Warren Oates),
ambushes these thugs and shoots them, Peckinpah had his editors
crosscut three slow-motion shots of one thug's dying fall with normal-
191
tempo images of Bennie exchanging fire with the other. Peckinpah's

most effective uses of slow motion almost always occurred with this
kind of intercutting. As we have noted, when Weddle discusses Peck¬
inpah's editing, he refers generally to his "weaving of slow motion into
the very fabric of a sequence."24 Although Weddle does not discuss how
this occurs, we can now see that it occurs through a brief but sustained
contrasting or crosscut series of shots that accentuate the different
modes of time. The dynamic qualities of the technique lie in the accen¬
tuation of these differences. To achieve this accentuation, it is impera¬
tive that the represented actions—a skidding car, a shooting victim
flung backward—be ones that explicitly occur with, and denote, speed
or force. This, in conjunction with the crosscutting, sets up two types
of opposition within the editing. The slow motion is set into a relation
of striking opposition with the normal tempo of the surrounding
imagery, and, internally, the slow-motion shots by themselves contra¬
dict the viewer's narrative understanding of the speed at which these
events are actually occurring.
When Peckinpah's slow-motion inserts fail to observe either of
these principles, and when there is an insufficient narrative context
supporting the device, the dynamic force of the technique is dimin¬
ished. Near the end of Alfredo Garcia, when Bennie guns down a pair
of corrupt executives and their hired guns in the El Camino Real hotel
suite, he steps in front of a mirror to drill his last opponent. Peckinpah
showed this man flung backward against a table and chair and crashing
to the floor in a single, rather extended slow-motion shot. Because this
action is not crosscut with normal-tempo shots of Bennie's reactions or
any other ongoing activity, and because it lasts too long, the decelera¬
tion of time here becomes what it rarely ever does in Peckinpah's
films—a simple slowing down. It is nondynamic because it has mini¬
mal structural relationship with the surrounding material. A much bet¬
ter set of isolated slow-motion inserts occurs in Straw Dogs when
Major Scott is shotgunned by Tom Hedden. The inserts acquire con¬
siderable force by virtue of the narrative context. Scott is the narrative's
chief authority figure, and with him dead, the viewer knows that all
hell is about to break loose as the gang of thugs converges on the
Sumners' house.
Peckinpah incorporated the brief slow-motion interlude into
his more complex montage sequences because the dynamic oscillation
between normal and decelerated time demands a continuing perceptual
reorientation from viewers. He apparently hoped the stylistic artifice
would alternately immerse viewers in the spectacle on screen and then
192
Stephen Prince
realign their perspective through the nonrealistic slow-motion inser¬

tions. These perceptual realignments he hoped would establish a new,
less complacent, and passive relationship between viewers and the
screen spectacle, would, as he put it, wake viewers up to what violence
is really all about. This functional intent was in addition to the other
uses to which he put slow motion, which we have just reviewed: to cre¬
ate a temporal dialectic across the body of a scene,- to interrupt the con¬
crete physicality of violence with more abstract contemplations of its
balletic and metaphysical aspects,- and to shuttle between these con¬
crete and abstract dimensions in a way that would superimpose them
on top of each other.
Extended Montage Set-Pieces
Before evaluating the extent to which his montage editing of different

film speeds successfully establishes a new viewing position for film
spectators, we need to explore how Peckinpah incorporated slow
motion into more elaborate montages that work as extended violent
set-pieces. These extended spectacles of death and destruction appear,
of course, in The Wild Bunch and Straw Dogs and, to a lesser extent, in
Alfredo Garcia, The Killer Elite, Cross of Iron, and The Osterman
Weekend.
As we have seen, the editing of the opening and closing battles
in The Wild Bunch has been compared to the Odessa Steps scene in
Eisenstein's Potemkin.25 Yet despite their more exhilarating quality,
they lack the enormous structural variety and richness that character¬
ize the Eisenstein sequence. Like that sequence, however, these scenes
in The Wild Bunch take what exists in the narrative as linear, separate
lines of action (e.g., each member of the Bunch separately trying to
escape the ambush, the bounty hunters picking their targets, and the
panicked reactions of the pedestrians caught in the cross fire) and inte¬
grate them as a synthesized collage of activity. The film's editor, Lou
Lombardo, remarked that, following Peckinpah's advice, he
intercut all the separate lines of action. I might start with this guy
being hit, then cut to that guy being hit, cut to this guy falling, that
guy still falling, then cut to somebody else over there getting hit, to a
horse spinning over there, somebody going through a window there,
and then back to the first guy just landing on the ground. I meshed it.
I took every piece of action and intercut it with another.26
193
Indeed, the shot lists that survive among Peckinpah's papers, corre¬
sponding to the opening and closing shoot-outs in San Rafael and Agua
Verde, present a linear and chronological list of images.
As Lombardo's description indicates, and as a careful viewing
of the film demonstrates, the basic device used to mesh the lines of
action is crosscutting. However, instead of a simple cutting back and
forth, the lines of action are interrupted for extended periods by cut¬
aways to other things before they resume. In its elaborate patterns of
crosscutting, Peckinpah's editing performs four distinct functions. It
slows down, interrupts, parallels, and returns to ongoing lines of action.
These functions collectively establish the collage-like structure of the
montage set-pieces, and they are clearly illustrated by a scene within
the San Rafael massacre that opens The Wild Bunch, one that Lom¬
bardo alludes to in his description of how he meshed the lines of action.
The Bunch have just robbed the depot and are being fired upon by the
rooftop snipers. They return fire, hitting two victims on the rooftop. An
early, written visualization of this scene differs substantially from the
final montage and describes mainly a linear progression of images with
little intercutting.
23. Bounty hunter Shannon rises up to fire and is hit and wounded,
starts to pitch forward over the parapit [sic],
24. Another bounty hunter (who?) grabs for him, exposes himself and
is hit and killed, slumps back releasing Shannon who rolls down
the roof and pitches to the street.
25. Burt sees the body fall beside him and jumps on his horse racing out
firing steadily until he is caught in the bounty hunter fire, causing
both he [sic] and his horse to plunge through the window of the
dress shop.
26. Wild Bunch:—Pike stays in the street with Dutch as the others die
or mount and ride out.
27. Then following, Frank who has been wounded while mounting, is
shot while riding out and goes into a drag still clutching the bags of
silver from the pay station.
28. As Frank is dragged down the street, the horse is shot down.27
The editing in the finished film parallels the falls of both rooftop vic¬
tims by crosscutting between them, but it also interrupts these lines of
action by cutting away to other, ongoing lines of action before return¬
ing to the two victims. As victim one ('Shannon' in the early shot con¬
tinuity) topples forward, off the roof, the first cutaway occurs to a shot
194
Stephen Prince
Figures 32 & 33. Peckinpah uses slow motion and montage editing to mesh together multiple
lines of action in the violent gun battles that open and close The Wild Bunch, as in this inter¬
cutting of the trajectories of two victims shot off the rooftops of the town of Starbuck.
of Pike running out of the depot office. The next shot returns to a con¬
tinuation of the previous action as victim one falls below the bottom
of the frame line. The next three shots are all cutaways. Angel, another
of the Bunch, dashes out of the depot office, returning fire as he runs.
A low-angle longshot of the rooftop snipers is followed by a high-angle
longshot of several of the Bunch shooting toward the snipers from
inside the depot. After these three cutaways, the montage returns to
victim one, and slow motion is introduced to retard the rate of his fall.
The composition is very dramatic. The victim arcs against the blue sky,
the slow motion suspending him weightlessly in space.
The next three shots are more cutaways, but they also intro¬
duce the fate of the second rooftop victim. In the first cutaway, a
medium shot shows the Bunch firing from inside the depot. Then, in
the next shot, the second rooftop victim is struck and falls forward,
195
toward the camera. Next, a medium shot shows Angel running from
the depot, firing, continuing an action introduced six shots previously.
The next shot, in slow motion, returns to the first victim, continuing
his lethargic descent to earth. Next, the montage cuts to the second vic¬
tim still toppling forward, but not off the roof. His fall will be broken
by the rooftop ledge. A medium close-up shows Deke Thornton firing
from the roof. Then, in the next two ^shots, both victims land simulta¬
neously, in a matched cut, victim one in the street and victim two on
the rooftop ledge.
The editing reconfigures, stylistically transforms, the deaths of
the two rooftop victims. The action cuts away from victim one's fall
four times, initially for one shot, then for three shots, again for three
shots, and finally for two shots. Two cutaways interrupt the fall of the
second victim, for two shots each time. The cutaways and the use of
slow motion impose a marked distortion upon the time and space of
the represented action. The editing creates a false parallel between the
two victims. Victim one falls off the roof in slow motion, and victim
two, hit later in the sequence, falls at normal speed, yet they strike
ledge and street at precisely the same moment, as the matched cut that
closes off these events indicates. The editing imposes a false parallel
between normal time and decelerated time. The simultaneous impact
of the two victims represents an impossible time-space relationship
within the sequence, yet Peckinpah and Lombardo convincingly inter¬
cut normal speed and slow motion to extend this discontinuity.
This is certainly what Peckinpah meant when he rejected alle¬
gations that his work represented a greater realism, stressing instead
that it stylized violence, heightened it through artistic transformation.
The elaborate montages of The Wild Bunch effect such an artistic trans¬
formation of space, time, and perception by rendering space and time
as totally plastic and unstable entities. Time slows, stretches, folds
around on itself and becomes the fourth dimension of a spatial field in
which the ordinary laws of physics do not apply. When the outlaw try¬
ing to flee the rooftop sniper ("Burt" in the early visualization) crashes
his horse through a storefront window, three slow-motion shots of this
action are intercut with three normal-speed shots of another rider
crashing to the street after his horse is shot out from under him. The
cutting creates a false spatial parallel. The amount of space traversed
by each rider who takes a spill is roughly the same—from saddle to
ground—but the different rates of time indicate different spaces, since
time and space alike are part of a four-dimensional continuum. By
196
Stephen Prince
intercutting the falls of these riders in slow motion and normal speed,
Peckinpah and Lombardo reconfigured space as well as time. The slow
motion implies an alternate spatial field in which the reconstituted
dynamics play out.
Peckinpah and Lombardo also employed the pattern of inter¬
rupting ongoing lines of action with cutaways to other events in a way
that is analogous with their use of slow motion—to extend the dura¬
tion of the represented events and retard their completion. In the con¬
cluding shoot-out at Mapache's headquarters, the film's crowning
montage set-piece, Pike bursts into a room where a woman stands,
shoots into a mirrored door to kill a soldier hiding there, and is then
shot by the woman on whom he had turned his hack. This simple series
of events is broken up and extended with a very long set of cutaways.
After Pike shoots the soldier, the action cuts away from him for seven¬
teen shots, which show, primarily, Dutch using a woman as a shield as
he exchanges gunfire with Mapache's men. Then, returning to Pike, a
single close-up shows him glancing outside the woman's room. A sec¬
ond lengthy cutaway, twenty-eight shots long, shows Lyle Gorch
behind the machine gun, howling like a demon. Then the action returns
to Pike, still in the room, at which point he is shot. The elaborate cut¬
aways extend the duration of the action and thereby expand time.
As this description indicates, Peckinpah and Lombardo did not
employ slow motion in any simple capacity. The slow-motion inserts
are placed within a complex montage that crosscuts multiple lines of
action so that the slow motion functions in concert with the extended
cutaways to reconfigure time. Furthermore, the temporal manipula¬
tions include not just deceleration but also parallelism, disruption, and
resumption. The resulting stylistic transcends a naturalistic presenta¬
tion of violence, and it is notable that critics have discussed Peckin¬
pah's work as if its use of bloody squibs and slow motion was more
realistic than previous generations of Flollywood gunfights. It certainly
is bloodier, but Peckinpah's is far from a realist's aesthetic. Peckinpah's
montage set-pieces in The Wild Bunch work primarily on the level of
form rather than by their representational content. They work as
exquisitely crafted artifacts that emphasize physical spectacle. Their
design foregrounds the hyperkinetic spectacle so that it becomes a
detachable part of the film. Although, within the narrative, complex
issues of character and theme lead up to and into the film's climactic
slaughter at Mapache's headquarters, in terms of its montage design
this scene is complete, self-contained, and utterly sufficient unto itself.
For the filmmakers involved, the scene must have been a lot of fun to
197
The Aesthetic of Slow-Mvtion Violence in the Films of Sam Peckinpah
craft and edit: Herein lies a significant problem for the didactic uses to
which Peckinpah wanted to put his screen violence. This problem has
persisted, and it affects the work of virtually all filmmakers who have
chosen to stylize graphic violence by using this aesthetic approach.
Peckinpah himself noted that audience reactions were at vari¬
ance with the responses he intended viewers to have. Writing to Paul
Stamford, who had condemned the brutality of The Wild Bunch, Peck¬
inpah observed that "better than 50% of the people who saw the pic¬
ture felt as you did. However, better than 30% of the people thought it
was an outstanding and much needed statement against violence."28
Elsewhere, he admitted, "unfortunately most people come to see it [vio¬
lence] because they dig it, which is a study of human nature, and which
makes me a little sick."29
As Peckinpah recognized, viewer reactions to The Wild Bunch
included outrage and shock over the scale and explicitness of the vio¬
lence but also excitement and exhilaration. By aestheticizing it, Peck¬
inpah's montages made violence pleasurable and beautiful and turned
it into an exciting spectacle. The very stylization that Peckinpah
thought would wake people up to the horror of violence instead excited
and gratified many. This pleasure is an inevitable result of the aes¬
theticizing functions of Peckinpah's montage editing and its balletic
incorporation of slow motion, and it is a virtually inescapable effect of
the montage aesthetic that characterizes the work of all filmmakers
who have employed it. The director occasionally acknowledged his
own culpability in eliciting a dualistic response. "In The Wild Bunch I
wanted to show that violence could be at the same time repulsive and
fascinating."30
Many critics noted these contradictions in 1969 when The Wild
Bunch was released. The reviewer for The Nation confessed his own
complicity in sharing the audience's gleeful reaction to all the carnage.
"At all this the audience laughed (and so did I), not with merriment,
exactly, but in tribute to such virtuosity of gore."31 The reviewer for
The Christian Century noted that "even while gasping at Peckinpah's
bloodbath, the people seated near me in the theater continued to cram
their mouths with popcorn."32
Several perceptive reviewers pointed to the contradiction be¬
tween Peckinpah's stated aims and his montage style: the style is so rit-
ualistically elaborated in the opening and closing scenes of The Wild
Bunch that it turns violence into a pleasurable spectacle. Furthermore,
the excessive elaboration of these montages indicates that the filmmaker
took an overwhelming interest and delight in the mechanics of crafting
198
Stephen Prince
them. Arthur Knight in Saturday Review pointed out that the stylistics
of the film tend to displace the intended moral commentary. "But when
the movies attempt to show violent killing in detail, the mind turns
against the fact of death and toward the mechanics that produced it. And,
curiously, one comes away convinced that the director was also more
concerned with the mechanics than with the fact."33 Joseph Morgen-
stern in Newsweek stressed the apparent falsity of Peckinpah's premise
that the repetition of stylized violence could become an artistic device
commenting on itself.34 These reviewers are wrong in their dismissal of
Peckinpah's moral involvement with his material and in their objections
that the film is without ethical content. But they are dead right in argu¬
ing that the tendency of Peckinpah's montage -set-pieces in The Wild
Bunch is to spectacularize violence in ways that will incite the aggres¬
sive fantasies of many viewers. Recalling his first encounter with The
Wild Bunch when he was a teenager, screenwriter Charles Higson
stressed that the film's violence stimulated an "orgiastic" release of
energy in him. "Once the film was over, I was exhausted and in a state
of high nervous excitement. I wanted to go out in a blaze of glory. I
wanted a Gatling gun. I wanted to be pierced by a hundred bullets."35
Peckinpah did not overtly intend to elicit this kind of reaction in
his viewers, any more than Martin Scorsese wanted to arouse vigilante
responses in viewers of his bloody Taxi Driver (1976), but both film¬
makers crafted sequences that have done so. If we ask why, the answer
lies in the way that Peckinpah, and Scorsese, too, was seduced by the
artistic excitement of putting those violent montages together, of
manipulating time and space with images of bodies flying this way and
that. The sheer pleasure of crafting these montage scenes would have
been exhilarating for Peckinpah. That pleasure is plainly evident in the
the San Rafael massacre and the flamboyant Agua Verde shoot-out, vio¬
lent set-pieces that open and close the film. Through their dynamic
energies, these montages convey the excitement and thrill of a film¬
maker no longer in moral control of his material to the viewer, who
reacts accordingly. Discussing Taxi Driver, which climaxes with an
astoundingly bloody shoot-out, Scorsese exhibits confusion about his
intellectual intent and the seductive excitement he derives from craft¬
ing screen Violence. The audience, he says, was
reacting very strongly to the shoot-out sequence in Taxi Driver. And I

was disturbed by that. It wasn't done with that intent. You can't stop
people from taking it that way. What can you do? And you can't stop
people from getting an exhilaration from violence, because that's
199
human, very much the same way as you get an exhilaration from the
violence in The Wild Bunch. But the exhilaration of the violence at the
end of The Wild Bunch and the violence that's in Taxi Driver—because
it s shot a certain way, and I know how it's shot, because I shot it and
I designed it—is also in the creation of that scene in the editing, in the
camera moves, in the use of music, and the use of sound effects, and
in the movement within thefr^me of the characters. . . . And that's
where the exhilaration comes in.36
As did Peckinpah, Scorsese wanted to disassociate himself from

the aggressive reactions of his viewers, but, like Peckinpah, he was
keenly responsive to the physical and artistic pleasures of crafting
screen violence. This contradiction accounts for the filmmakers' dis¬
may at the responses of viewers to the scenes that they had so lovingly
crafted and choreographed. Neither Scorsese nor Peckinpah wished to
evoke violent fantasies in their viewers. When asked if that was their
intention, both passionately denied it. But they could not disengage
themselves, as artists, from the sensuous gratifications of assembling
spectacularized violence. Wfiiile one should not doubt the sincerity of
their belief in their own stated intentions, one may still be amazed at
their blindness to their own artistic complicity in stimulating the
aggressive reactions of their viewers.
Recognizing this contradiction between Peckinpah's laudable
moral intention of shocking his viewers into confronting the horror of
violence and his own fascination with the montage spectacle in The Wild
Bunch brings us to an important point. If the montage-based representa¬
tion of violence were Peckinpah's only artistic contribution to late 1960s
cinema and to the dilemmas of social violence wracking American soci¬
ety in those years, he should be condemned as an aesthete of violence,
an inciter to aggression, a director whose films reinforced and added to
the violence of those years. If the montage aesthetic were the only frame
of analysis deployed on the issue of violence in Peckinpah's films, then
he would be everything his detractors have claimed him to be: a glamor-
izer and glorifier of violence. But this was not the only frame through
which he approached the violence issue, as I have discussed elsewhere.37
To the extent, though, that the montage aesthetic has become the endur¬
ing template for filmic presentations of graphic violence, cinema has
turned death into a mechanized spectacle. The style has taken film¬
makers down a gigantic artistic dead end. The montage aesthetic chore¬
ographs the outward manifestations of violence, not its inwardly
spiritual and emotional components and consequences. These elude the
200
Stephen Prince
template. Thus this aesthetic, when it becomes the chief means for rep¬
resenting violent death, as it all too often has been, is an insufficient
means for probing the meaning and consequences of violence, should
those be a filmmaker's intentions. The montage aesthetic conveys
a filmmaker's delight to the audience, who is encouraged to respond
in kind. The aesthetic evokes vigilante responses. Thus does death be¬
come pleasurable and its protracted physical agonies become emptied
of human meaning.
NOTES
1. "Playboy Interview: Sam Peckinpah," Playboy 19, no. 8 (August 1972), p. 68.
2. Sam Peckinpah Collection, Margaret Herrick Library, Wild Bunch—response
letters, statement on violence, folder no. 92.
3. SPC, Wild Bunch—response letters, May 13, 1969, folder no. 92.
4. See Stephen Prince, Savage Cinema: Sam Peckinpah and the Rise of Ultravio-
lent Movies (Austin: University of Texas Press, 1998), pp. 27^4-5.
5. SPC, Wild Bunch—response letters, statement on violence, folder no. 92.
6. SPC, Wild Bunch—response letters, September 3, 1969, folder no. 92.
7. David Weddle, If They Move . . . Kill ‘Em: The Life and Times of Sam Peckin¬
pah (New York: Grove Press, 1994), pp. 99, 333.
8. SPC, Alfredo Garcia—daily log, folder no. 45.
9. SPC, funior Bonner—daily shots, folder no. 19.
10. Nat Segaloff, "Walon Green: Fate Will Get You," Backstory 3: Interviews with
Screenwriters of the 1960s ed. Pat McGilligan (Berkeley: University of California Press,
1997), pp. 143,135-56; and author's telephone conversation with Paul Seydor, July 22,1997.
11. Ernest Callenbach, "A Conversation with Sam Peckinpah," Film Quarterly 17,
no. 2 (Winter 1963-64), p. 10.
12. Weddle, p. 271.
13. SPC, Cross of Iron—script notes, letter of February 5, 1976, folder no. 26.
14. SPC, Feldman memos, October 29, 1967, folder no. 45.
15. Gary Crowdus and Richard Porton, "The Importance of a Singular, Guiding
Vision," Cineaste 20, no. 2 (Spring 1993), p. 9.
16. Stephen Farber, "Peckinpah's Return," Film Quarterly 23, no. 1 (Fall 1969), p. 11.
17. SPC, The Wild Bunch—correspondence, letter of March 19, 1968, folder no. 43;
and Weddle, p. 331.
18. Ibid.
19. Ibid., p. 5.
20. Garner Simmons, Peckinpah: A Portrait in Montage (Austin: University of
Texas Press, 1982), p. 19.
21. Bryson, "Wild Bunch in New York," p. 28.
22. Savage Cinema, pp. 72-97.
23. Seydor, p. 20.
24. Weddle, p. 271.
25. See, for example, Weddle, p. 356.
26. Weddle, p. 355.
27. SPC, Wild Bunch—editing, folder no. 40.
28. SPC, Wild Bunch—response letters, May 13, 1969, folder no. 92.
29. John Bryson, "Wild Bunch in New York," New York (August 19, 1974), p. 28.
201
The Aesthetic of Slow-Mdtion Violence in the Films of Sam Peckinpah
30. SPC, D&Devoir interview, interviews, folder no. 96.

31. Review of The Wild Bunch in The Nation, vol. 209, July 14, 1969, p. 61.
32. Sherwood Ross, "Blood and Circuses," The Christian Century (August 20,
1969), p. 1095.
33. Arthur Knight, "Violence Flares Anew," Saturday Review (July 5, 1969), p. 21.
34. Joseph Morgenstern, "The Bloody Bunch," Newsweek (July 14, 1969), p. 85.
35. Charles Higson, "The Shock of the Old," Sight and Sound 5, no. 8 (August 1995),
p. 36.
36. Anthony DeCurtis, "What the Streets Mean: An Interview with Martin Scors¬
ese," in Plays, Movies and Critics, ed. Jody McAuliffe (Durham, N.C.: Duke University
Press, 1993), p. 211.
37. Prince, Savage Cinema.
- < '
f*f" ? s ...
-
■
The Effects
of Ultraviolence
'
Leonard Berkowitz
Some Effects of
Thoughts on Anti- and
Prosocial Influences of Media
Events: A Cognitive-
Neoassociation Analysis
Well before the onset of the twentieth century, the French sociologist
Gabriel Tarde discussed what he termed suggesto-imitative assaults.
He noted how news of a sensational violent crime often seemed to
prompt similar incidents elsewhere, and pointed to the aftereffects of
the Jack the Ripper murders as an example. According to Tarde, the
great national attention given these brutal crimes led to eight
absolutely identical crimes in London as well as "a repetition of these
same deeds outside of the capital (and abroad)" (Tarde, 1912). "Epi¬
demics of crime," he concluded in a once famous phrase, "follow the
line of the telegraph."
Regardless of how well known Tarde's principle was several
generations ago, social scientists have only recently obtained system¬
atic quantitative support for his thesis. Berkowitz and Macaulay (1971)
reported a significant jump in violent, but not property, crimes above
what would be expected from the prevailing trend after several spec¬
tacular murders in the early and mid-1960s, including those after Pres¬
ident Kennedy's assassination. Phillips published much more
impressive evidence in a series of studies of the consequences of widely
publicized suicides. He found that when the media paid considerable
attention to a famous person's suicide, there was an increase in the
number of people who also took their lives. According to Phillips (1974)
From Psychological Bulletin, vol. 95, no. 3, 1984, pp. 410-427. Copyright © 1984 by the
American Psychological Association. Reprinted with permission.
205
206
Leonard Berkowitz
the more publicity given to the suicide story, the greater the apparent
impact. In the same vein, he later demonstrated that the number of
automobile accidents increased several days after a widely publicized
suicide; many of the accidents were single-vehicle incidents (Phillips,
1979). Interestingly, in these single-vehicle cases, the driver tended to
be more like the suicide victim than were any passengers in the car.
Phillips (1980) then cast his empirical net even further and showed that
news stories of suicide-murders were followed by an increase in non¬
commercial airplane crashes. There is evidence that even the portrayal
of suicides in television soap operas led to an increase in the number of
white Americans who end their own lives (Phillips, 1982).
The implications of this research seem fairly clear: real and fic¬
tional depictions by the media of violence—killings, shootings, or
suicides—can prompt audience members to act aggressively toward
others or themselves. The present essay also argues that some of the
now familiar effects of violent scenes on television and movie screens
are similar to those producing the suggesto-imitative assaults discussed
by Tarde and documented in the research into the contagion of violence
just mentioned. In addition to these negative consequences, the mass
media can also promote prosocial behavior (Liebert, Sprafkin, & David¬
son, 1982; Rushton, 1979). A comprehensive account of media effects
should deal with both positive and negative influences.
In attempting such an analysis, this essay considers a number
of psychological processes that can contribute to these anti- and proso¬
cial effects. In particular, it focuses on the role of the audience's
thoughts. I suggest that the ideas activated by the witnessed or reported
event tend to evoke temporarily other, semantically related thoughts
in the audience that, under some circumstances, justify conduct simi¬
lar to that portrayed by the communication and even lead the viewers
to anticipate the benefits they might derive from this behavior. These
ideas also tend to evoke associated expresive-motor responses or per¬
haps even behavioral reactions that might intensify the viewer's exist¬
ing behavioral tendencies.
Priming Effects and Associative Networks
Several years ago I suggested that conditioning notions could profitably

be used to account for many of the media effects considered here
207
Some Effects of Thoughts on'Anti- and Prosocial Influences of Media Events
(Berkowitz, 1973, 1974). However, because of the overriding impor¬

tance of the media communications' meaning for the audience and the
processes of memory and recall, I now believe that the concepts of
cognitive-neoassociationism (Anderson St Bower, 1973; Landman St
Manis,19S3) provide a better framework for the analysis of these phe¬
nomena. How people react to the message they read, hear, or see
depends considerably on their interpretations of the message, the ideas
they bring with them to the communication, and the thoughts that are
activated by it. It is therefore advisable to study media effects in a way
that gives explicit attention to these matters; we can find such a useful
formulation in cognitive-neoassociationism.
Only a brief outline of this conceptual scheme is necessary. Nor
need we be concerned with the intricacies of this reasoning and the con¬
troversies regarding a number of the details (cf. Klatzky, 1980; Landman
St Manis, 1983; Wyer & Hartwick, 1980; Wyer St Srull, 1981, for a more
complete discussion). What is important is that memory is regarded as
a collection of networks, with each network consisting of units or
nodes (representing substantive elements of thought, feelings, and so
on) that are interconnected through associative pathways. The strength
of these pathways is presumably a function of a variety of factors,
including contiguity, similarity, and semantic relatedness.
Priming Effects and Spreading Activation
Along with other similar analyses, the present formulation adds the
concept of spreading activation advanced by Collins and Loftus (1975)
and others to these structural notions. This argument maintains that
when a thought element is brought into focal awareness, the activation
radiates out from, this particular node along the associative pathways
to other nodes. This process can lead to a priming effect. Although the¬
orists are not in agreement as to why the priming takes place, they gen¬
erally maintain that for some time after a concept has been activated,
there is an increased likelihood that it and associated thought elements
will come to mind again. It is as though some residual excitation has
remained at the activated node for a while, making it easier for this and
other related thoughts and feelings to be activated. Thus, the aggessive
ideas suggested by a violent movie can prime other semantically related
thoughts, heightening the chances that viewers will have other aggres¬
sive ideas in this period. All of this can happen automatically and with¬
out much thinking. Cognitive theorists now differentiate between
208
Leonard Berkowitz
automatic and controlled processing (e.g., Shiffrin & Schneider, 1977),

and it is often suggested that the memory effects just mentioned take
place passively and involuntarily. Controlled processing, in contrast,
requires conscious attention to a much greater extent.
Several social-psychological investigations (Wyer & Hartwick,
1980; Wyer St Srull, 1981) of the impact of recently activated concepts
on impression formation illustrate this important priming effect. In one
the subjects first had to construct sentences, each time using three of
the four words supplied to them. Some of these sets contained words
that had aggressive connotations, whereas other sets had neutral words.
At a specified interval after this initial priming phase (immediately
after, one hour later, or a day later), the participants were required to
indicate their impression of a target person on the basis of a brief
description. As the researchers predicted, the more aggression-related
words in the sentence construction task and the more recent this task,
the more unfavorable the subjects' evaluations of the target person. The
use of the aggression-related words apparently activated other aggression-
related thoughts that colored the participants' judgments of the target
for a time afterward. Bargh and Pietromonaco (1982) showed how this
process can operate automatically and without awareness. In their ex¬
periment the subjects were unknowingly exposed to single words,- some
of the words were semantically related to hostility. The subjects then
had to rate a target person on the basis of a description provided to them.
Although the subjects had not been consciously aware of the priming
words, the greater the proportion of hostility-related words presented
to them, the more negative their judgment of the target person.
Links with Emotion Components
The present conception does not stop with the individual's thoughts
and memories. Making use of recent theoretical developments in cog¬
nitive psychology, it holds that externally presented ideas can activate
particular feelings and even specific action tendencies as well as certain
kinds of thoughts and recollections. This formulation builds on the net¬
work analyses of emotion, advanced by Bower (1981), Clark and Isen
(1982), Lang (1979), and Leventhal (1980), which postulate associative
connections among thoughts, memories, and feelings. As an example,
Bower (1981) posited a memory-emotion network in which feelings or
emotion nodes are associatively (and, especially, semantically) related
to particular concepts and propositions and also to the autonomic
209
responses and expressive behaviors exhibited on these occasions. For

Bowers and the other investigators cited, the activation, of any of the
components in a given emotion network tends to evoke the other units
to which it is tied. Isen and her associates demonstrated such a con¬
nection when they showed that the experience of a positive mood tends
to facilitate the recall of information having a positive meaning (cf.
Clark St Isen, 1982). *
An experiment by Vaughan and Lanzetta (1981) also provides
supporting evidence. When their subjects watched a person receive
electric shocks, those observers who were asked to grimace at the time
of the shock exhibited other indications of emotion arousal relatively
strongly. The expressive facial movements apparently helped to acti¬
vate other components of the emotional state. Lang (1979) emphasized
the role of psychophysiological reactions in his conception. He noted
that the imagery thoughts are often accompanied by specific patterns
of efferent outflow. Also consider Velten's (1968) mood-induction pro¬
cedure. As the subjects read a series of effectively toned statements
(either positive or negative), they recalled conceptually similar inci¬
dents in their past that had the same affective tone. Consequently, they
tended to have expressive-motor reactions and feelings similar to those
experienced in the past situations. The activated components sum-
mated and reverberated as the subjects read from one statement to the
next, and they then consciously experienced (to some degree) the feel¬
ings associated with the statements they read.
Because of the links among thought elements, feelings, and
expressive-motor reactions, externally activated ideas can prompt auto¬
matic (involuntary) actions or even some controlled behaviors that are
semantically associated with them. Lang's (1979) theory of emotional
imagery is consistent with this suggestion in that it postulates seman¬
tic connections between "the cognitive structure of the image" and
"specific patterns of physiological and behavioral responding" (p. 506).
There is also some support for my position in Greenwald's (1970) dis¬
cussion of the ideo-motor mechanism in performance control. On
reviewing the available evidence, he concluded, along with William
James (1890), that the perceptual image or idea of an action can initiate
the performance of this behavior. More recent research by Anderson
(1983) also yielded findings in keeping with the present analysis. Peo¬
ple who imagined themselves carrying a particular type of behavior
later reported having a greater intention to engage in that action than
did others who imagined someone else performing the behavior. The
210
Leonard Berkowitz
thought of themselves carrying out the action evidently activated a

readiness to perform the behavior. Later a number of conditions will be
considered that might influence the degree to which media-induced
thoughts are translated into open action.
Applications to Research Findings
Priming Effects in Prosocial Influences
Experiments involving children and young adults indicate that people

seeing a display of helpfulness or generosity are often inclined to be
helpful or generous themselves (Hearold, 1979; Liebert et alv 1982;
Rushton, 1979). It is not likely that the communication's influence
comes about in these instances through the reduction of the audience's
restraints. In addition, the media do not necessarily produce a persist¬
ent learning because their effects are often relatively short-lived
(Liebert et al., 1982; Rushton, 1979). Rather, the viewers appear to have
been spurred temporarily into action. The first step in such a process
might involve a priming effect in which there is an activation of
thoughts semantically related to the depicted occurrence.
Hornstein and his colleagues (Blackman St Hornstein, 1977;
Holloway, Tucker, St Hornstein, 1977; Hornstein, LaKind, Frankel, St
Manne, 1975) provided evidence of such a priming effect in a series of
experiments concerned with the impact of good news. Supposedly
while waiting for the experiment to begin, the subjects in these studies
listened to a news program on the radio playing in the laboratory, and
then engaged in bargaining with a peer. In general, those who heard a
report of prosocial behavior were especially inclined to be cooperative
in their bargaining, particularly if the radio story was of someone who
had intentionally given help (Holloway et al., 1977).
Other findings indicate that the overheard story did more than
convey the specific message that the listeners ought to be generous in
their dealings with others. Ratings made at the end of the session
showed that the participants' conceptions of human nature and their
social world had been affected by the news report. After listening to the
prosocial news, the subjects had a stronger expectation that their bar¬
gaining partner would be cooperative toward them and thought that
there were more people who led clean, decent lives and who were basi¬
cally honest. Putting their results together with comparable observa¬
tions published by other investigators, Hornstein and his associates
211
Some Effects of Thoughts on 'Anti- and Prosocial Influences of Media Events
concluded that "people function as social actuaries, relying upon the

information they receive from the mass media and other sources . . .to
make generalized inferences about human nature" (Blackman & Horn-
stein, 1977, p. 303).
These findings can readily be interpreted from a cognitive per¬
spective. First, note the generalized effects. In the Hornstein research,
as in other experiments (e.g., Aderman & Berkowitz, 1970; Collins St
Getz, 1976; Sprafkin, Liebert, & Poulos, 1975), the portrayed prosocial
behavior was quite different from the acts the subjects could carry out,
but yet there was a carryover. The subjects apparently were responding
to semantically related concepts (e.g., helpfulness, generosity, coopera¬
tion) that had been activated in their minds by the priming communi¬
cation; they were not simply and narrowly imitating the reported
conduct. In several studies the activation spread some distance along
the associated pathways to other prosocial concepts such as honesty
(Blackman St Hornstein, 1977) or good work and rule obedience
(Friedrich St Stein, 1973). The present formulation holds that the mass
communication's influence stems largely from the activation of con¬
cepts and propositions semantically related to the portrayed event.
Those who lack these semantically related ideas will not be affected in
the same way. In keeping with this possibility, Silverman and Sprafkin
(1980) found that prosocial scenes in the children's television program
Sesame Street had little effect on the behavior of young children.
Although other interpretations are possible, it may be that the children
did not possess the concepts and propositions that could be related to
the depicted incidents.
In line with our analysis of the evidence, the thought activation
often was only temporary, and the communication's impact typically
declined with the passage of time. Priming effects usually subside as
the initiating stimulus recedes into the past. In addition, rather than
merely bringing related ideas to mind, the depicted action affected the
observers' estimates of the prevalence of prosocial behavior in their
social world, as Hornstein noted. The availability heuristic can account
for this finding. Readily recalling the vivid instance of altruism reported
in the mass media, the participants automatically assumed this con¬
duct was relatively frequent and likely.
Research on the Priming Effects of Television and Movie Violence
Introductory remarks. Because a good deal of this analysis relies on

investigations of the results of observed aggression, a few introductory
212
Leonard Berkowitz
comments about these studies are warranted. It is not necessary to sum¬

marize the hundreds of investigations in this area because a number of
fine reviews now exist (e.g., Andison, 1977; Comstock, 1975; Com¬
stock, Chaffee, Katzman, McCombs, & Roberts, 1978; Geen, 1976;
Goranson, 1970; Liebert, Sprafkin, & Davidson, 1982; Murray St Kip-
pax, 1979; Palmer & Dorr, 1980). However, it should be noted that the
majority of researchers are agreed that the depiction of violence in the
media increases the chances that people in the audience will act aggres¬
sively themselves. Meta-analyses of the published studies support such
a conclusion (Andison, 1977; Hearold, 1979). There is controversy, how¬
ever, concerning the magnitude of this effect (e.g., see Comstock, 1975;
Cook, Kendzierski, &. Thomas, 1983), and no claim is made here as to
how important media depictions are relative to the other sources of
antisocial conduct. Furthermore, because the greatest dispute in this
research area deals with the results of frequent exposure to violent
scenes (see Eron, 1982; Eron, Huesmann, Lefkowitz, St Walder, 1972;
Milavsky, Kessler, Stipp, St Rubens, 1982), most of our attention will
be given to experiments in which the participants see only one or a few
aggressive movies. The consequences of a high level of television or
movie viewing will receive much less emphasis here, and findings from
studies bearing on this issue will be discussed only if they are pertinent
to a particular theoretical point.
Some shortcomings of traditional interpretations. Before considering

the priming effects of violent communications, a few of the difficulties
in the traditional analyses of media aggression should be examined. In
my opinion the priming effect analysis overcomes these problems.
The most popular interpretations of the effects of media vio¬
lence emphasize the observers' learning. The depicted (or reported)
action serves as an example, teaching people in the audience how to
behave, what type of conduct is appropriate in a given situation, and
especially whether this action is likely to bring the rewards they desire
(e.g., Bandura, 1965, 1971, 1973). I do not want to deny or even mini¬
mize the importance of this learning. Nevertheless, it is apparent that
media influences do not operate through observational learning only, if
this concept is understood to refer to a relatively long-lasting acquisi¬
tion of new knowledge or the adoption of a novel form of behavior.
Some media effects are fairly transient (e.g., Buvinic & Berkowitz, 1976;
Doob & Climie, 1972; Mann, Berkowitz, Sidman, Starr, St West, 1974),
as if the observed event had activated reactions or thoughts only for a
213
Some Effects of Thoughts on Anti- and Prosocial Influences of Media Events
relatively brief, period. Phillips's previously mentioned studies of the

impact of widely publicized suicide stories also obtained evidence of a
time decay. In a recent investigation, for example, Bollen and Phillips
(1982) noted that the publicity given to suicide stories on television
evening news programs did not produce a heightened probability of
other suicides more than ten days after the initial report. It is my con¬
tention that at least part of this time decay is due to the diminution in
a priming effect that often occurs with the passage of time (cf. Wyer &.
Srull, 1981).
Furthermore, I believe that the often demonstrated generality
of the media influence is troublesome for those observational learning
interpretations that are couched in terms of imitation. Imitation gen¬
erally implies that the reproduced action is physically similar to the
portrayed behavior, but yet, most of the experiments in this area use
aggression measures that are physically different from the depicted con¬
duct (e.g., Parke, Berkowitz, Leyens, West, & Sebastian, 1977). Phillips
(1983) reported a startling demonstration of such a generalization.
Again making use of demographic data, Phillips found that widely pub¬
licized heavyweight championship prize fights between 1973 and 1978
were typically followed by an increase in homicides throughout the
United States. Rather than explaining these findings in terms of imita¬
tion (or even modeling), it may be better to say that the observers
responded to the meaning of the media event and exhibited behavior
having the same general meaning. Further, the depicted occurrence
should be considered as producing a priming effect that activates
semantically related ideas and behavioral tendencies.
For many researchers the knowledge transmission interpreta¬
tion of media influences maintains that these influences arise largely
through a permission-giving (or disinhibitory) process. People in the
audience are supposedly disposed to engage in some antisocial behav¬
ior but are reluctant to do so until the media tells them, directly or indi¬
rectly, that the behavior is permissible or even profitable. Wheeler's
(1966) discussion of behavioral contagion essentially follows this rea¬
soning, and other investigators have offered a similar analysis. Com¬
stock and his associates (Comstock, 1980; Comstock et ah, 1978)
emphasized the disinhibitory consequences of media violence, main¬
taining that the portrayed action lessens the audience's restraints
against such conduct.
I do not doubt that media events can lower observers' inhibitions
by indicating that the behaviors they are tempted to carry out may ben¬
efit them (Bandura, 1965) or might be appropriate in the given situation.
214
Leonard Berkowitz
Indeed, my colleagues and I have published a good number of experi¬

ments demonstrating a permission-giving phenomenon (e.g., Berko¬
witz St Geen, 1967; Berkowitz St Rawlings, 1963). My contention here
is that the relatively short-lived aggression-enhancing influences are
not due to a disinhibitory process only.
Priming Effects and Antisocial Influences
I have already implied that the contagion of violence reported by Tarde,

Berkowitz and Macaulay, and Phillips might be partly due to priming
effects generated by news stories of spectacular violent crimes. Indeed,
the rise in homicide rates often following a war, in the victorious as well
as defeated nations (Archer St Gartner, 1976), conceivably might also
be affected by aggression-related ideas activated by media reports of the
conflict and kept alive by other influences.
Activation of aggression-related ideas. Both direct and indirect evi¬

dence indicates that the observation of aggression evokes aggression-
related thoughts and ideas in the viewers. In the latter category,
Berkowitz, Parker, and West (cited in Berkowitz, 1973) offered infor¬
mation suggesting such a possibility. In their experiment, schoolchil¬
dren were asked to complete sentences by selecting one of two words
supplied to them. The children were more likely to choose words hav¬
ing aggressive connotations if they had just read a war comic book
rather than a neutral comic book. Exposure to the violent scenes in the
former condition apparently had brought aggression-related ideas to
mind so that words with aggressive meanings seemed more appropri¬
ate in constructing the sentences.
Turning to more direct evidence, the research by Wyer and Srull
(1981) and Bargh and Pietromonaco (1982) already has been cited as a
demonstration of how the priming effect can influence one's reaction
to others. In these studies, presenting the participants with words hav¬
ing hostile connotations led them to make hostile evaluations of an
ambiguous target person. Carver, Ganellen, Froming, and Chambers
(1983) extended the research paradigm that Wyer and Srull and Bargh
and Pietromonaco used and showed that people who had watched a
brief film depicting a hostile interaction between a businessman and
his secretary subsequently perceived more hostility in an ambiguous
stimulus person. A study of the consequences of aggressive humor
(Berkowitz, 1970) is conceptually similar to these investigations. In this
215
experiment, young women who listened to a tape recording of a hostile

comic routine were subsequently harsher in their ratings of a job appli¬
cant than were other women who heard a nonaggressive comic routine.
This happened even when the subjects had not been provoked by the
person they were judging.
In these instances the initial communication had evidently
primed aggression-related ideas which then affected the subjects' inter¬
pretation of an ambiguous stimulus person. These activated ideas
might also change at least temporarily the viewers' impression of the
desirability of aggression. Other studies have shown that people seeing
violent encounters are at times inclined to favor the use of violence in
interpersonal conflict. Leifer and Roberts (1972) and Drabman and
Thomas (1974) reported such an effect in experiments with school-
children, and Malamuth and Check (1981) found that movies portray¬
ing sexual violence against women increased the male viewer's belief
that this type of violence was sometimes acceptable. Even the mere
presence of weapons might also produce a heightened acceptance of vio¬
lence at times. The participants in an experiment by Leyens and Parke
(1975) who were shown slides of weapons were more willing to punish
an available target severely than were the participants who had seen
only neutral slides.
Conceptions of others: The cultivation thesis. A number of students

of the mass media have gone even further with the line of reasoning just
introduced. George Gerbner and his associates (e.g., Gerbner, Gross,
Morgan, St Signorielli, 1980; Gerbner, Gross, Signorielli, Morgan, St
Jackson-Beeck, 1979) argued that television violence does more than
produce a relatively transient acceptance of aggression. It may also cul¬
tivate a long-lasting conception of the social environment as wicked
and dangerous. Television programs often portray a world of violence,
danger, and evil. Consequently, it is maintained that those who watch
television frequently develop a perception of social reality that is con¬
sistent with this usual depiction of life and society. In a word, the view¬
ers supposedly acquire a paranoid conception of their environment.
Evidence suggests that the Gerbner argument is valid in some respects
but may be overstated. The best support for this position comes from
research in which subjects are exposed to only a single communication.
Thus, in an experiment by Thomas and Drabman (1977), children
shown a television detective scene were subsequently inclined to
believe that other children would prefer to act aggressively in their
216
Leonard Berkowitz
interpersonal encounters. Thinking this, the subjects would be apt to

conclude that aggression is relatively frequent in their social world. The
previously cited findings of Hornstein and his colleagues (e.g., Black¬
man & Hornstein, 1977) are also in accord with the Gerbner formula¬
tion, even though these studies involve the perception of prosocial
behavior. Recall that in the Hornstein experiments, the subjects who
heard a radio report of helpful behavior tended to guess that helpfulness
was relatively prevalent in the world.
The research evidence is much less clear, however, when we
turn to studies of people repeatedly exposed to scenes of violence. In
the investigation by Gerbner and his associates (Gerbner, Gross, Sig-
norielli, Morgan, &. Jackson-Beek, 1979), adolescents who watched a
great deal of television were especially likely to overestimate the
amount of violence in society and believe that the social world was dan¬
gerous. However, several investigators have published opposing obser¬
vations. Wober (1978) failed to obtain indications of a paranoid
conception of life in a national sample of British television viewers, and
Doob and MacDonald (1979) showed that the relationship between tel¬
evision viewing and fear of crime was eliminated in a Toronto, Ontario
sample when the actual incidence of crime in the viewers' areas was
controlled for statistically. Gerbner et al. (1980) produced further data
from a national sample of adults and a statistical analysis indicating
that heavy television viewing can lead to later perceptions of danger.
Cook et al.'s (1983) position lies somewhere between that of Gerbner
and his critics. Upon examining the published data, Cook et al. con¬
cluded that television viewing can influence general beliefs about the
world, but suggested that this cultivation effect was extremely modest.
One way to reconcile the opposing contentions is to refer to two
possibilities. The first possibility is that the priming effect is relatively
short-lived. As noted earlier, media reports apparently activate seman¬
tically related thoughts for only a brief period. It is during this time that
the viewers, guided by the availability heuristic, are apt to exaggerate
the prevalence of antisocial behavior in the world. However, other
influences may be necessary if the activated ideas are to be learned so
that there is an abiding conception of the surrounding environment.
These influences are not always present and television's portrait of
social reality is not necessarily always accepted as true by the people in
the audience.
The second possibility has to do with a difference between esti¬
mates of the prevalence of antisocial behavior and judgments of being
personally in danger. Hawkins and Pingree (1980) suggested that a
217
Some Effects of Thoughts on 'Anti- and Prosocial Influences of Media Events
heavy diet of television increases the perceived frequency of antisocial

behavior in the world, but does not necessarily create the belief that the
viewers themselves are likely to be victimized. Some support for this
suggestion can be found in path analyses conducted by Tyler (1980).
Making use of samples of respondents in cities on the East and West
Coasts, Tyler demonstrated that mass media reports of crimes affected
respondents' estimates of the crime rate in their neighborhoods, but did
not influence their sense of being personally vulnerable to crime. This
latter perception was influenced, however, by their own or another's
personal experience with crime.
All in all, theory and empirical research, and Gerbner's (Gerb-
ner et al., 1979; Gerbner et al., 1980) cultivation thesis, propose that
vivid portrayals of antisocial behavior on television and radio can lead,
at least temporarily, to the belief that this kind of behavior is relatively
frequent in the social environment. Nevertheless, additional factors
apparently have to operate if these activated judgments are to be turned
into a persistent perception of the world or are to affect the sense of
being personally in danger.
Inciting an action tendency. The research reviewed to this point indi¬

cates that the media can activate particular thoughts and ideas. The
question now is whether specific action tendencies are also incited. In
my summary of the network analyses of emotion, I argued for such an
effect when I suggested that activation can spread from the concepts
and propositions that come to mind to feelings, expressive movements,
and even action tendencies. Is there any evidence that media-activated
thoughts can generate such behavioral inclinations?
The aggressive humor experiment mentioned before (Berko-
witz, 1970) points to such a possibility. The hostile comic routine led
the subjects to treat the job applicant harshly. Better evidence is pro¬
vided in the second study conducted by Carver, Ganellen, Froming, and
Chambers (1983). Making use of the Wyer and Srull (1981) scrambled
sentence task, the investigators showed that male undergraduates who
had been primed to have aggressive thoughts subsequently adminis¬
tered the most intense electric shocks to a fellow student whenever
that person made a mistake. Their aggressive ideas apparently led to
aggressive acts.
Other relevant findings were obtained in experiments on the
effects of aggression-associated cues in the surrounding environment.
This research is discussed in more detail later, and only one such study
218
Leonard Berkowitz
need be mentioned here. Josephson (1982) conducted an experiment in

which deliberately frustrated schoolboys were shown an excerpt from
a popular television program that was either violent or nonviolent in
nature and then played a game of floor hockey in which they could
exhibit naturalistic aggression against their opponents. The villains in
the violent television program used walkie-talkie radios in the course
of their misdeeds, so Josephson assumed that the presence of a walkie-
talkie would serve as a retrieval cue: it would remind the subjects in
the violent movie condition of the aggression they had seen and might
also activate the thoughts and feelings that the film had evoked. Thus,
the adult referees supervising the boys7 game sometimes carried
walkie-talkies. The researcher found that the children were most
aggressive if they had previously watched the violent program and the
referees carried the aggression-associated walkie-talkie radios. My pro¬
posal here is that this cue strengthened the activation generated by the
aggressive movie.
An experiment by Worchel (1972) also suggests that violent
scenes can incite aggressive inclinations. Young boys and girls were
shown a brief fight scene or an interesting movie about boats or they
were not shown any film. They then were told they could choose
between engaging in a pie fight (an aggressive activity) or riding on a
raft (nonaggressive activity). The children's choices of activities tended
to be consistent with the movie they had seen, as if this film had evoked
an appetite for that type of behavior. Another experimenter then
entered the room and put pressure on the children to select one of the
activities, the activity that was either consistent or inconsistent with
the movie they had watched. In keeping with Worchel's expectation
(based on reactance theory), when the subjects were again asked to state
their preferences, they overwhelmingly reacted against the pressure if
it had been inconsistent with the film. Both films had evidently gener¬
ated a desire for a particular type of activity, and the subjects countered
the opposing pressure with an even stronger insistence on this activity.
Disinhibition and indifference. The media-induced activation of

aggression-felated thoughts can account for other consequences of
media depictions of violence. I have suggested so far that these thoughts
can (1) influence the audience's interpretation of an ambiguous stimu¬
lus person, (2) heighten the estimated frequency of aggressive and other
antisocial acts in the social world, (3) incite inclinations to aggressive
behavior, and (4) strengthen the belief that aggression is desirable or
219
Figures 34 & 35. Director Paul Verhoeven (Robocop, Total Recall, Starship Troopers) has cre¬
ated some of the most extreme and outlandish violence in mainstream films. In Robocop, a police
robot machine-guns a corporate executive, whose body explodes in a shower of blood and torn
clothing. The MPAA threatened the film with an X rating unless Verhoeven trimmed its violence.
acceptable at times. All of these possible reactions, and especially the

last one, can lead to a lowering of restraints against aggression and even
lessen the observers' anxieties about aggression.
Some investigators have labeled this process a trivialization of
aggression,- violence becomes a small matter not worthy of special note
as it becomes more acceptable. I have already mentioned one indication
220
Leonard Berkowitz
of this in the experiment by Malamuth and Check (1981), but much

more direct evidence has been provided by Zillmann and Bryant (1982).
In accord with Gerbner's cultivation thesis, the men and women in this
experiment who were exposed to a massive amount of filmed sex scenes
gave the highest estimate of the prevalence of unusual sexual practices
in the general population, and they were also least likely to object to
the open display of pornography. More important, when all of the sub¬
jects were asked to recommend a jail sentence for a convicted rapist,
those men and women who had been exposed to a large number of
unusual sex scenes favored the shortest terms of imprisonment. Rape
had evidently become a more trivial matter for them. Can it be that the
activation of sexual-aggressive ideas in these people and the resulting
increased acceptance of such behavior also led to a diminution in
sexual-aggressive concerns?
This conjectured reduction in aggression anxiety implies that
observers will tend to become physiologically less aroused with
repeated exposure to aggressive scenes. Thomas, Horton, Lippincott,
and Drabman (1977) conducted two experiments—one involving
schoolchildren and the other involving college students—in which sub¬
jects first saw either a fictional portrayal of violence or a nonaggressive
movie and then watched a film clip showing realistic aggression. In
both studies the viewers exhibited less emotionality (inferred from
changes in skin resistance) in reaction to the realistic aggression if they
had previously been exposed to the aggressive incident. Geen (1981)
also reported that prior exposure to a violent scene reduced arousal to
a later aggressive depiction. He suggested that the earlier observation
of aggression was particularly likely to lessen the viewers' sensitivity
to cues that normally restrain their aggression.
It is important to note that the lower arousal is presumably an
indicant of a reduced concern about aggression and not a sign of a
decreased inclination to aggression. Thomas (1982) produced some sup¬
porting evidence. In her experiment angry men exposed to a fifteen-
minute aggressive television program subsequently tended to have
a lower heart rate than their angered counterparts who were shown a
neutral movie. Nevertheless, in spite of their lower arousal level, the
former also tended to give their antagonist more electric shocks
soon afterward. Apparently less anxious about aggression because of
the aggression-approving ideas generated by the violent program, the
former group became calm, but also punished their tormentor
more severely.
221
%
Conditions Facilitating the Occurrence of Overt Aggression
There is a substantial gap between the media-activated ideas, feelings,

and action tendencies and the open display of behavior. Other factors can
intervene to affect the chances that people in the audience will be overtly
assaultive themselves. It is important to consider these conditions.
)p
Meaning of the Communication

Defining the witnessed action as aggression. The present essay repeat¬
edly emphasizes the importance of the media's meaning for the audi¬
ence. Aggression is in the mind of the beholder, and a movie will not
activate aggression-associated thoughts unless the viewer regards what
is seen as aggression. This term, of course, is used in a variety of ways
in our everyday language, but most people seem to think of aggression
as the deliberate injury of another. Berkowitz and Alioto (1973) assumed
this was the common meaning of the term when they argued that con¬
tact sports are most likely to stimulate aggressive inclinations in the
onlookers when they believe the players are trying to hurt each other.
The angered university men in this study watched a film of either a prize
fight or a football game and were induced to interpret the contest either
as aggressive (the opponents supposedly were trying to injure each other)
or as nonaggressive (the contestants were professionals unemotionally
engaged in their business). When the subjects were given an opportunity
to shock their tormentor at the end of the movie, those who had seen
the encounter described as aggressive were most punitive. The viewers
had to define what they saw as aggression if the event was to activate
strong aggression-related ideas and responses in them.
Donnerstein and Berkowitz (1983) obtained findings consistent
with this theory. The intensity of their male subjects' punishment of a
woman who had provoked them earlier was significantly correlated
with the rated aggressiveness of the movie they had seen before they
could deliver the shocks. The observers' definition of the scene as
aggressive had presumably activated aggression-related thoughts and
other aggression-facilitating reactions in them which strengthened the
punishment they gave to the provocateur.
Other aggressive ideas. Besides imparting an aggressive meaning to the

witnessed event, when they watch this occurrence the observers might
also have other aggressive thoughts that can intensify the activated
222
Leonard Berkowitz
aggressive inclinations. Turner and Berkowitz (1972) showed their pro¬

voked male subjects a movie of a prizefight and asked most of the sub¬
jects to think of themselves either as the winner or the referee. The
others were asked simply to watch the film. Half of the people in each
condition were led to have aggressive ideas as they viewed the scene by
requiring them to think "hit" each time the victor landed a blow. The
two experimental variations interacted to affect the number of shocks
the subjects subsequently gave to their tormentor. In general, the men
attacked this person most severely if they had imagined themselves as
the winner and thought "hit" with each punch they saw.
The "imagine-self " finding in this experiment is reminiscent
of the results reported more recently by Anderson (1983). The college
students in this later study expressed the strongest intention to carry
out a certain kind of behavior if they had previously imagined them¬
selves engaging in that action.
The "hit" ideas conceivably might have served as aggression
retrieval cues, especially for those who thought of themselves as the
victor. As the subjects imagined themselves hitting their opponent, the
word hit that they periodically uttered to themselves might have
evoked memories of other occasions in which they had fought an
antagonist. Thoughts, feelings, and expressive movements related to
the aggression they exhibited might have been activated by these words
so that their present aggression was intensified.
Turner (Turner & Layton, 1976) was the first to recognize how
the words used in thoughts can function as aggression retrieval cues to
affect subsequent behavior. On the basis of other research on the psy¬
chology of memory, he predicted that words having an aggressive mean¬
ing would be especially strong activators of aggressive inclinations to
the degree that they could easily evoke images. These high-imagery
aggression words presumably would better retrieve prior aggressive
episodes from memory, and thus should activate other aggression-
related thoughts and inclinations. To test this prediction, Turner and
Layton (1976) required their male subjects to learn lists of words vary¬
ing in aggression connotation (aggressive or neutral) and imagery value
(high or low), and then gave them an opportunity to shock a partner. All
of the subjects were physiologically aroused by exposure to white noise
when they were delivering these shocks. In keeping with Turner's pre¬
diction, the aroused subjects were most punitive if they had been pre¬
viously exposed to the high-imagery aggressive words.
223
Identification and imagination. Several students of mass communi¬

cations (e.g., Dorr, 1981; Tannenbaum St Gaer, 1965) have sugested that
the observers' identification with the media characters influences the
extent to which they are affected by the witnessed occurrence. From
the present perspective, onlookers who identify with the observed
actors are vividly imagining themselves as these characters and are
thinking of themselves as carrying, out the actions they see. The view¬
ers identifying with a movie aggressor should therefore be providing
themselves with aggressive thoughts having a relatively high imagery
value, and thus should be priming themselves aggressively. The previ¬
ously cited experiment by Turner and Berlcowitz (1972) supports this
analysis, as does a later study by Leyens and Picus (1973). In this inves¬
tigation the angry subjects who had been asked to imagine themselves
as the victor of the filmed fight shown to them were later most aggres¬
sive toward the person who had insulted them. Their high degree of
aggression was presumably because of their high level of aggression-
related thoughts as they watched the movie.
Several field studies have also highlighted the importance of the
media viewers' identification with the observed aggressor. When
McLeod, Atkin, and Chaffee (1972) questioned high school students in
Maryland and Wisconsin about their television viewing and social
behavior, they found, as other investigators had, that exposure to tele¬
vision violence was positively associated with self-reported aggression.
Most pertinent to the present issue, however, was that those students
indicating a high degree of identification with violent television char¬
acters tended to be highly aggressive themselves. Huesmann, Eron,
Klein, Brice, and Fischer (1983) also reported that children who watched
violent television programs frequently and who identified with the
aggressive characters tended to be highly aggressive in their dealings
with peers. Aggressive fantasies generally might have a similar effect.
According to Rosenfeld et al. (1978), young boys who often fan¬
tasize about aggression tend to be highly aggressive. The data from
these studies indicate that a highly aggressive imagination might keep
the influence of the media exposure alive. Bandura (1971) maintained
that the observers' imaginal and verbal representations of a witnessed
incident can perpetuate the effects of that event. Huesmann (in
National Institute of Mental [NIMH], 1982) maintained that children's
hostile daydreams or aggressive make-believe play increase the chances
that they will display overt aggression after seeing violence on the
224
Leonard Berkowitz
screen. In keeping with these concepts, continuing to think about the

filmed violence and one's own aggression in the depicted situation
might prolong the media-generated activation of aggressively related
ideas and expressive-motor reactions. Some of the research results
obtained by Singer and Singer (1981) are also consistent with this rea¬
soning. They noted that children who engaged in imaginative play with
a prosocial theme generally exhibited little aggression. In this case the
children's prosocial thoughts could have activated ideas and feelings
that were incompatible with aggression, thereby lessening their aggres¬
sive tendencies.
The meaning of the observed aggression. Viewers can impart different

kinds of meaning to the reported behavior in addition to defining it as
aggressive or not. They might regard the depicted action as justified or
unjustified, as warranted revenge for an earlier insult or an improper
assault on an innocent victim as a profitable act yielding benefits for the
aggressor, or as costly conduct leading to negative consequences. In
many instances, perhaps because of a spreading activation of related con¬
cepts, these meanings apparently then generalize to affect the observers'
interpretations of any aggression they themselves might display.
There is little doubt that the outcome of the model's aggression
often influences the audience's willingness to engage in aggression
themselves (Bandura, 1965, 1971). The viewers appear to draw a lesson
from what they see: what happens on the screen (or is reported in the
media) might also happen to them if they exhibited the same behavior
(Bandura, 1971; Comstock, 1980; Comstock et al., 1978; Huesmann, in
NIMH, 1982). As was suggested earlier, this phenomenon is readily
interpreted in cognitive terms. Seeing the result of the model's action,
the observers are reminded of other occasions in which there was a sim¬
ilar outcome. With this consequence prominently in mind, the fre¬
quency and probability of such a result are apt to be overestimated (as
the availability heuristic suggests). Consequently, showing the actor
punished for a given behavior tends to dampen their willingness to act
in the same way, whereas a report that the behavior had favorable con¬
sequences often increases their inclination to follow suit.
The audience can be affected by the outcome of the aggression
they watch even if the movie aggressor is not directly affected. Goran-
son (1969) found that his angry subjects attacked their tormentor only
weakly after learning that the movie aggression victim died of the
225
injuries received. In this case the thought of the possible unfortunate

consequences of aggression apparently was uppermost in the observers'
minds, and these ideas evidently led the subjects to restrain their
assault on their insulter. It is important to note that Goranson's sub¬
jects had no ill feelings toward the movie aggression victim. Nor did
they associate that movie person with another individual whom they
might have wanted to hurt. In suchf a case, if the victim in the film had
suffered rather than died, the angry observers conceivably might have
regarded the victim's pain as gratifying for them. The observed aggres¬
sion could then have been seen as successful aggression leading to a
desired consequence—revenge—and the end result might have been a
disinhibition (Berkowitz, 1974; Donnerstein St Berkowitz, 1981).
This matter of the observed aggression's outcome cannot be
ignored if an adequate understanding of the effects of movie violence is
to be gained. The old-fashioned Western movies might have had an
adverse impact on their viewers because they never showed any tragic
consequences of aggression. Characters fell down when they were shot
but no one was seen suffering. There is an interesting and suggestive
difference between U.S. and Japanese television programming in this
regard. According to Iwao, de Sola Pool, and Hagiwara (1981), Japanese
and American TV films have approximately the same amount of vio¬
lence, but the Japanese movies are much more likely to show the vic¬
tim suffering and are especially more apt to depict the protagonists as
victims. This emphasis on the protagonists' pain could lessen the audi¬
ence's willingness to act aggressively.
Other related phenomena are less obvious but may also affect the
viewers' inhibitions against carrying out the reported behavior. A fairly
substantial body of research has demonstrated that the audience mem¬
bers' attitude toward the observed aggressor and the victim influences
their willingness to attack their own tormentors. In these investigations,
initiated by Berkowitz and Rawlings (1963) and replicated by Berkowitz
and his colleagues (Berkowitz, 1965; Berkowitz, Corwin, St Heironimus,
1963; Berkowitz St Geen, 1967; Berkowitz, Parke, Leyens, St West, 1974;
Berkowitz St Powers, 1979) and others (e.g., Geen St Stonner, 1973; Hoyt,
1970; Meyer, 1972), the angry male subjects were given a brief introduc¬
tion to the violent movie they were about to see that portrayed the vic¬
tim of the observed aggression in a certain manner. In some instances
this movie character was described as an immoral and unpleasant per¬
son, whereas in other cases the character was depicted favorably. At the
conclusion of the scene, all of the men had an opportunity to punish the
226
Leonard Berkowitz
person who had insulted them earlier. The experiments all found that the
subjects attacked their tormentor more severely after seeing the "bad
guy" being assaulted than after watching the "good man" being hurt.
Berkowitz interpreted the results of his studies as being due to
differences in the perceived justification of the witnessed aggression.
That is, the subjects presumably regarded the villain's punishment as
justified aggression, whereas the "good guy's" beating was viewed as
morally unwarranted. The observers then interpreted their own possi¬
ble attacks on their tormentor in the same way. In essence, the thought
of the movie violence as good, justified aggression presumably activated
other ideas of morally proper aggression. However, other researchers
extended the original findings in a way that suggests a slightly differ¬
ent interpretation. Beginning with Hoyt (1970), investigators (Geen fit
Stonner, 1973, 1974; Meyer, 1972) showed that provoked observers
were most punitive toward their insulter if they had been led to think
of the witnessed violence as vengeance for an earlier injustice. Believ¬
ing themselves to have been ill-treated, the subjects evidently applied
the movie-activated thoughts of revenge to their own situation so that
their inhibitions against aggression were lessened. Those who had not
been angered, however, might have regarded these ideas as inappropri¬
ate or even wrong, which might have produced increased restraints
(Geen St Stonner, 1973). Geen (1976) concluded that the viewers com¬
pared their own motivational states with those of the actors in the vio¬
lent incident and used cues from this incident "as information
concerning [what is] appropriate behavior" (p. 212). Whatever the exact
meaning of the observed aggression, the context in which it occurs
clearly can affect the viewers' inhibitions against aggression.
Nature of available target. The probability of overt aggression after

seeing violence depends on the nature of the potential target as well as
the other factors which have been discussed. The viewers are unlikely
to attack someone who might punish them, but there are also more sub¬
tle and interesting effects that should be noted. Most important, a series
of experiments has demonstrated that after watching an aggressive
scene, angry observers tend to direct the strongest aggression toward
those around them who are associated with the victim of the witnessed
violence (Berkowitz fit Geen, 1967; Donnerstein fit Berkowitz, 1981;
Geen fit Berkowitz, 1966, 1967). In most of these studies deliberately
provoked subjects were shown a filmed fight and then were given an
227
opportunity to punish their tormentor. The results indicated that the

participants punished their antagonist more severely if the antagonist
had the same name as the loser of the observed fight as opposed to hav¬
ing the winner's name or a name not used in the movie. A more recent
experiment by Donnerstein and Berkowitz (1983) extends the general¬
ity of this phenomenon. The male subjects in this study were first pro¬
voked by a woman. Some of the subjects then saw a film clip in which
a woman was assaulted by two men. In accord with the earlier findings,
when these subjects were able to punish the insulting woman, their
aggression was more intense if she had the same name as the female vic¬
tim in the movie than if she had a different name. The portrayed assault
might have activated thoughts in the angry viewers concerning the sat¬
isfaction of attacking a woman (probably because they had just been
provoked by a woman). These thoughts could have been strengthened
when the subjects faced the female antagonist having the same name as
the film victim, perhaps because her name functioned as a retrieval cue.
The result was the strong aggression toward her. In general, then, poten¬
tial targets who remind viewers of the victim of violence they had
recently seen might be especially likely to draw hostility toward them¬
selves, particularly if the observed aggression was gratifying.
Reality of the observed incident. All of the factors just mentioned,

such as the meaning of the witnessed event, will affect the viewers only
to the extent that they attend to this occurrence. Moreover, they are
especially inclined to be influenced if they are highly involved in the
observed scene, thinking of themselves as carrying out the reported
behaviors. The scene's reality can help determine how involved they
will be in what they see.
A number of experiments suggest that viewers are more likely
to be aggressively incited by a violent occurrence if this event is realis¬
tic rather than fictional. Noble (1973) reported that the constructiveness
of children's play was more adversely affected by realistic rather than
stylistic depictions of aggression. Feshbach (1972) has more direct evi¬
dence in a better controlled investigation. All of the children in his study
saw a movie of a campus riot. Half of them were told this was a ficti¬
tious film,- the others were informed they were watching a newsreel of
an actual incident. When the children were able to punish a person soon
afterward, those who had viewed the supposedly real violence were
more aggressive. Berkowitz and Alioto (1973) extended this finding to
228
Leonard Berkowitz
young adults. The angry college men in their experiment were more
punitive toward their provocateur after watching a war scene that was
described as realistic rather than a staged depiction.
Research by Geen and his colleagues provides important addi¬
tional information. An experiment by Geen and Rakosky (1973) sug¬
gests that the observers' definition of the witnessed aggression as only
fictitious seemed to distance them from the event psychologically so
that it had less of an impact on them. Subjects who were reminded that
the fight they were about to see was fictional were less physiologically
aroused by it than were those subjects who were not given this
reminder. Consistent with this finding, in a later study Geen (1975)
reported that angry men shown a real fight instead of a staged one were
more likely to maintain their high level of physiological arousal during
the movie and to give their tormentor strong punishment soon after¬
ward. The subjects in these investigations were less likely to have
aggression-activating thoughts as they watched the violent incident if
they regarded the event as only fictional.
Another experiment indicates that the observer's focus on the
artistic nature of the scene can have a similar result. Leyens, Cisneros,
and Hossay (1976) asked some of their subjects to attend to the aesthetic
qualities of the aggression-related pictures shown to them. These sub¬
jects were less aggressive toward the person who had insulted them ear¬
lier than were those subjects who were also exposed to aggressive
pictures but who were not given this "decentration" instruction. In
other words, the subjects seeing the aggressive pictures without the
aesthetic focus were more aggressive than controls seeing aggressive
pictures with the aesthetic focus. The researchers noted that the atten¬
tion to the aesthetic qualities had not altered the aggressive meaning
of the pictures, so it could only be that it led to a weaker activation of
aggression-facilitating ideas.
Leyens and his associates pointed out an intriguing implication
of their findings. It may be possible to lessen the aggression-inciting
capacity of violent communications without imposing censorship by
teaching viewers to focus on the aesthetic or physical aspects of the
movies they watch. A research team led by Huesmann and Eron (Hues-
mann et al.’ 1983) followed a comparable line of reasoning, emphasiz¬
ing the reality versus fictional dimension. As part of their research
project, schoolchildren in the experimental condition were required to
write essays on why televised violence was unreal and bad.. This pro¬
cedure, based on social-psychological studies of role playing, was
intended to inculcate an unfavorable attitude toward violent programs
229
as well as to strengthen the children's awareness of the fictional nature

of these shows. The training procedure seemed to affect their subse¬
quent behavior. Not only were the experimental subjects less aggres¬
sive than their controls (as rated by peers), but there was no relationship
between television violence viewing and peer-rated aggression in the
experimental group, although such a relationship did exist in the con¬
trol condition. The children in thd experimental group might have
learned to think about televised violence in a way that dampened the
aggression-activating capacity of these aggressive movies.
Conclusion
Criticisms of the heavy diet of violence in American movies and tele¬

vision programs typically are based on two principal fears: (1) the fear
that children will learn to favor and use aggression as a way of solving
their interpersonal problems, and (2) the fear that many persons will
become indifferent to the suffering caused by violence. It certainly is
not wrong to be concerned about these matters, but the present essay
emphasizes another type of danger: people can get ideas from the com¬
munications reporting violent incidents and, for a short time afterward
at least, these thoughts can help foster antisocial behavior.
The media can be a force for good as well as bad, and I have sug¬
gested that appropriate media-activated thoughts and memories can
facilitate helpfulness and prosocial conduct generally. The outcome
depends on the action that is communicated. Newspapers, radio, and
television make news as well as report it. They can promote socially
desirable behavior by publicizing instances of such conduct, or they can
increase the likelihood of aggression by depicting violent incidents in
either realistic or fictitious form. My guess is that realistic portrayals
of violence are more inclined to have these adverse consequences
(assuming that they are highly vivid and capture the audience's full
attention).
Of course, it could be argued that the present analysis exagger¬
ates the impact of the mass media. Even if the reasoning spelled out here
is upheld by further research, some might say that the consequences are
due to only one or a few exposures to media communications. Frequent
viewing could conceivably lead to an even smaller effect so that the audi¬
ences become inured to the scenes or stories reported by the media. The
trivialization of aggression or desensitization to aggression discussed
230
Leonard Berkowitz
earlier might be interpreted in such a manner: As the observers repeat¬

edly encounter violent scenes, they might become habituated to events
of this type so that there is a decreasing priming effect and a decline in
the activation of aggression-related ideas and inclinations. The desen¬
sitization, then, presumably reflects this reduced activation.
This essay has implicitly rejected this habituation interpreta¬
tion of desensitization. I have suggested that the signs of desensitization
to aggression with repeated exposure to aggressive scenes are basically
due to an increased acceptance of aggressive behavior. Aggressive ideas
theoretically are more likely to be activated with this repeated exposure
and, as a consequence, the observers are more apt to think (at least for a
short time) that aggression is proper or worthwhile. Because of this
increased acceptance of aggression, their concerns and anxiety about
aggression presumably diminish. Thus, I have proposed that the lowered
level of physiological arousal resulting from several viewings of aggres¬
sive movies might actually reflect a reduced internal conflict or a
decreased aggression anxiety, but not a decline in the observer's incli¬
nation to aggression. The previously discussed experiment by Thomas
(1982) has yielded findings consistent with this thesis.
Even if habituation is involved in the desensitization to aggres¬
sion, this does not necessarily mean that there is a lower probability of
open aggression with frequent exposure to aggressive scenes. The habit¬
uation could create a diminution of two different reaction tendencies:
(1) ideas and inclinations facilitative of aggression, and (2) ideas and
other responses that tend to inhibit overt aggression. Available evi¬
dence suggests, although only tentatively, that any habituation might
result in a faster decrease in the inhibitory as compared with the facil¬
itative tendencies. Again, overt aggression is then more likely.
This is not to say that antisocial and prosocial actions are
inevitable when antisocial or prosocial incidents are reported. The
mass communications only affect the probability of this behavior, and
many other factors can intervene to affect the chances that the media-
activated thoughts will be translated into overt acts. This report has
examined a number of these moderating influences, but even when
they are taken into account, there undoubtedly is only a relatively
small likelihood that any one person will openly display the type of
behavior that was portrayed. However, this low probability must be
considered in the context of the audience size. For example, the chance
that only one individual in one hundred thousand will exhibit overt
aggression as a result of the depicted violence means that a hundred
231
more violent acts will occur in an audience of 10 million. This partic¬

ular example may have underestimated the probability of overt aggres¬
sion and certainly minimized the size of American audiences.
Whatever the exact numbers, our society has to decide whether the ben¬
efits of portrayed aggression outweigh the cost.
REFERENCES
Aderman, D., St Berkowitz, L. (1970). Observational set, empathy, and helping. Journal
of Personality and Social Psychology, 14, 141-148.
Anderson, C. A.(1983). Imagination and expectation:The effect of imagining behavioral
scripts on personal intentions. Journal of Personality and Social Psychology, 45,
293-305.
Anderson, J., St Bower, G. (1973). Human associative memory. Washington, DC: V. H.
Winston.
Andison, F. (1977). TV violence and viewer aggression: A cumulation of study results
1956-1976. Public Opinion Quarterly, 41, 314—331.
Archer, D., St Gartner, R. (1976). Violent acts and violent times: A comparative approach
to postwar homicide rates. American Sociological Review, 41, 937-963.
Bandura, A. (1965). Influence of models' reinforcement contingencies on the acquisition
of imitative responses. Journal of Personality and Social Psychology, 1, 589-595.
Bandura, A. (1971). Social learning theory. New York: General Learning Press.
Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ:
Prentice-Hall.
Bandura, A., Ross, D., St Ross, S. (1963). Imitation of film-mediated aggressive models.
Journal of Abnormal and Social Psychology, 66, 3-11.
Bargh, J., & Pietromonaco, P. (1982). Automatic information processing and social per¬
ception: The influence of trait information presented outside of conscious aware¬
ness on impression formation. Journal of Personality and Social Psychology, 43,
437^149. Berkowitz, L. (1965). Some aspects of observed aggression. Journal of Per¬
sonality and Social Psychology, 2, 359-369.
Berkowitz, L. (1970). Aggressive humor as a stimulus to aggressive responses. Journal of
Personality and Social Psychology, 16, 710-717.
Berkowitz, L. (1973). Words and symbols as stimuli to aggressive responses. In J. Knutson
(Ed.), Control of aggression: Implications from basic research (pp. 113-143).
Chicago: Aldine.
Berkowitz, L. (1974). Some determinants of impulsive aggression: Role of mediated asso¬
ciations with reinforcements for aggression. Psychological Review, 81, 165-176.
Berkowitz, L., St Alioto, J. (1973). The meaning of an observed event as a determinant of
its aggressive consequences. Journal of Personality and Social Psychology, 28,
206-217.
Berkowitz, L., Corwin, R., St Heironimus, M. (1963). Film violence and subsequent
aggressive tendencies. Public Opinion Quarterly, 27, 217-229.
Berkowitz, L., St Geen, R. (1967). Stimulus qualities of the target of aggression: A further
study. Journal of Personality and Social Psychology, 5, 364-368.
Berkowitz, L., St Macaulay, J. (1971). The contagion of criminal violence. Sociometry, 34,
238-260.
Berkowitz, L., Parke, R., Leyens, J., St West, S. (1974). The effects of justified and unjus¬
tified movie violence on aggression in juvenile delinquents. Journal of Research in
Crime and Delinquency, 11, 16-24.
232
Leonard Berkowitz
Berkowitz, L., & Powers, P. (1979). Effects of timing and justification of witnessed aggres¬
sion on the observers' punitiveness. Journal of Research in Personality, 13, 71-80.
Berkowitz, L., & Rawlings, E. (1963). Effects of film violence on inhibitions against sub¬
sequent aggression. Journal of Abnormal and Social Psychology, 66, 405-412.
Blackman, J., St Hornstein, H. (1977). Newscasts and the social actuary. Public Opinion
Quarterly, 41, 295-313.
Bollen, K., St Phillips, D. (1982). Imitative suicides: A national study of the effects of tel¬
evision news stories. American Sociological Review, 47, 802-809.
Bower, G. (1981). Mood and memory. American Psychologist, 36, 129-148.
Buvinic, M., St Berkotyitz, L. (1976). Delayed effects of practiced versus unpracticed
responses after observation of movie violence. Journal of Experimental Social Psy¬
chology, 12, 283-293.
Carver, C., Ganellen, R., Froming, W., St Chambers, W. (1983). Modeling: An analysis in
terms of category accessibility. Journal of Experimental Social Psychology, 19,
403-421.
Clark, M., St Isen, A. (1982). Toward understanding the relationship between feeling
states and social behavior. In A. Hastorf St A. Isen (Eds.), Cognitive social psychol-
ogy (pp. 73-108). New York: Elsevier.
Collins, A., St Loftus, E. (1975). A spreading-activation theory of semantic memory. Psy¬
chological Review, 82, 407-428.
Collins, W., St Getz, S. (1976). Children's social responses following modeled reactions to
provocation: Prosocial effects of a TV drama. Journal of Personality, 44, 488-500.
Comstock, G. (1975). Television and human behavior: The key studies. Santa Monica,
CA: Rand.
Comstock, G. (1980). New emphases in research on the effects of television and film vio¬
lence. In E. Palmer St A. Dorr (Eds.), Children and the faces of television (pp.
129-148). New York: Academic Press.
Comstock, G., Chaffee, S., Katzman, N., McCombs, M., St Roberts, D. (1978). Television
and human behavior. New York: Columbia University Press.
Cook, T. D., Kendzierski, D. A., St Thomas, S. V. (1983). The implicit assumptions of tel¬
evision research: An analysis of the 1982 NIMH report on television and behavior.
Public Opinion Quarterly, 47, 161-201.
Donnerstein, E., St Berkowitz, L. (1981). Victim reactions in aggressive erotic films as a
factor in violence against women. Journal of Personality and Social Psychology, 41,
710-724.
Donnerstein, E., St Berkowitz, L. (1983). Effects of film content and victim association
on aggressive behavior and attitudes. Unpublished manuscript, University of Wis-
consin-Madison.
Doob, A., St Climie, R. (1972). Delay of measurement and the effects of film violence.
Journal of Experimental Social Psychology, 8, 136-142.
Doob, A., St MacDonald, G. (1979). Television viewing and fear of victimization: Is the
relationship causal? Journal of Personality and Social Psychology, 37, 170-179.
Dorr, A. (1981). Television and affective development and functioning: Maybe this
decade. Journal of Broadcasting, 25, 335-345.
Drabman, R., St Thomas, M. (1974). Does media violence increase children's toleration
of real-life aggression? Developmental Psychology, 10, 418-421.
Eron, L. (1982)! Parent-child interaction, television violence, and aggression of children.
American Psychologist, 37, 197-211.
Eron, L., Huesmann, L., Lefkowitz, M., St Walder, L. (1972). Does television violence
cause aggression? American Psychologist, 27, 253-263.
Feshbach, S. (1972). Reality and fantasy in filmed violence. In J. Murray, E. Rubinstein,
St G. Comstock (Eds.), Television and social behavior (vol. 2, pp. 318-345). Wash¬
ington, DC: Department of Health, Education, and Welfare.
233
Friedrich, L., St Stein, A. (1973). Aggressive and prosocial television programs and the nat¬
ural behavior of preschool children. Monographs of the Society for Research in Child
Development, 38 (4, Serial No. 151).
Geen, R. (1975). The meaning of observed violence: Real versus fictional violence and
effects of aggression and emotional arousal. Journal of Research in Personality, 9,
270-281.
Geen, R. (1976). Observing violence in the mass media: Implications of basic research. In
R. Geen St E. O'Neal (Eds.), Perspectives on aggression (pp. 193-234). New York:
Academic Press. t *
Geen, R. (1981). Behavioral and physiological reactions to observed violence: Effects of
prior exposure to aggressive stimuli. Journal of Personality and Social Psychology,
40, 868-875.
Geen, R., St Berkowitz, L. (1966). Name-mediated aggressive cue properties. Journal of
Personality, 34, 456-465.
Geen, R., St Berkowitz, L. (1967). Some conditions facilitating the occurrence of aggres¬
sion after the observation of violence. Journal of Personality, 35, 666-676.
Geen, R., St Rakosky, J. (1973). Interpretations of observed violence and their effects of
GSR. Journal of Experimental Research in Personality, 6, 289-292.
Geen, R., St Stonner, D. (1973). Context effects in observed violence. Journal of Person¬
ality and Social Psychology, 25, 145-150.
Geen, R., St Stonner, D. (1974). The meaning of observed violence: Effects on arousal and
aggressive behavior. Journal of Research in Personality, 8, 55-63.
Gerbner, G., Gross, L., Morgan, M., St Signorielli, N. (1980). The "mainstreaming" of
America: Violence profile No. 11. Journal of Communication, 30, 10-29.
Gerbner, G., Gross, L., Signorielli, N., Morgan, M., St Jackson-Beeck, M. (1979). The demon¬
stration of power. Violence profile No. 10. Journal of Communication, 29, 177-195.
Goranson, R. (1969). Observed violence and aggressive behavior: The effects of negative
outcomes to the observed violence. Unpublished doctoral dissertation. University
of Wisconsin.
Goranson, R. (1970). Media violence and aggressive behavior: A review of experimental
research. In L. Berkowitz (Ed.), Advances in experimental social psychology (vol. 5,
pp. 1-31). New York: Academic Press.
Greenwald, A. G. (1970). Sensory feedback mechanisms in performance control: With
special reference to the ideo-motor mechanism. Psychological Review, 77, 73-99.
Hawkins, R., St Pingree, S. (1980). Some processes in the cultivation effect. Communi¬
cation Research, 7, 193-226.
Hearold, S. (1979). Meta-analysis of the effects of television on social behavior. Unpub¬
lished doctoral dissertation, University of Colorado.
Holloway, S., Tucker, L., St Hornstein, H. (1977). The effects of social and nonsocial infor¬
mation on interpersonal behavior of males: The news makes news. Journal of Per¬
Hornstein, H. A., LaKind, E., Frankel, G., St Manne, S. (1975). The effects of knowledge
about remote social events on prosocial behavior, social conception, and mood.
Journal of Personality and Social Psychology, 32, 1038-1046.
Hoyt, f. (1970). Effect of media violence "justification" on aggression. Journal of Broad¬
casting, 16, 455-464.
Huesmann, L., Eron, L., Klein, L., Brice, P., St^Fischer, P. (1983). Mitigating the imitation
of aggressive behaviors by changing children's attitudes about media violence. Jour¬
nal of Personality and Social Psychology, 44, 899-910.
Iwao, S., de Sola Pool, I., St Hagiwara, S. (1981). Japanese and U.S. media: Some cross-
cultural insights in TV violence. Journal of Communication, 31, 28-36.
James, W. (1890). Principles of psychology. New York: Holt.
Josephson, W. (1982). The effects of violent television upon children’s aggression: Elici-
234
Leonard Berkowitz
tation, disinhibition, or catharsis} Unpublished doctoral dissertation. University of

Manitoba.
Klatzky, R. (1980). Human memory (2nd ed.). San Francisco: Freeman.
Landman, J., St Manis, M. (1983). Social cognition: Some historical and theoretical per¬
spectives. In L. Berkowitz (Ed.), Advances in experimental social psychology (vol.
16, pp. 49-123). San Francisco: Academic Press.
Lang, P. J. (1979). A bio-informational theory of emotional imagery. Psychophysiology,
16, 495-512.
Leifer, A., St Roberts, D. (1972). Children's response to television violence. In J. Murray,
E. Rubinstein, St G. Comstock (Eds.), Television and social behavior (vol. 2). Wash¬
ington, DC: U.S. Government Printing Office.
Leventhal, H. (1980). Toward a comprehensive theory of emotion. In L. Berkowitz (Ed.),
Advances in experimental social psychology (vol. 13, pp. 140-207). New York: Aca¬
demic Press.
Leyens, J., Cisneros, T., St Hossay, J. (1976). Decentration as a means for reducing aggres¬
sion after exposure to violent stimuli. European Journal of Social Psychology, 6,
459-473.
Leyens, J., St Parke, R. (1975). Aggressive slides can induce a weapons effect. European
Journal of Social Psychology, 5, 229-236.
Leyens, J., St Picus, S. (1973). Identification with the winner of a fight and name media¬
tion: Their differential effects upon subsequent aggressive behavior. British Journal
of Social and Clinical Psychology, 12, 374-377.
Liebert, R., Sprafkin, J., St Davidson, E. (1982). The early window: Effects of television on
children and youth (2nd ed.). New York: Pergamon Press.
Malamuth, N., St Check, f. (1981). The effects of mass media exposure on acceptance of
violence against women: A field experiment. Journal of Research in Personality, 15,
436-446.
Mann, J., Berkowitz, L., Sidman, f., Starr, S., St West, S. (1974). Satiation of the transient
stimulating effects of erotic films. Journal of Personality and Social Psychology, 30,
729-735.
McLeod, J., Atkin, C., St Chaffee, S. (1972). Adolescents, parents and television use: Ado¬
lescent self-report measures from Maryland and Wisconsin samples. In G. Comstock
St E. Rubinstein (Eds.), Television and social behavior (vol. 3, pp. 173-238). Wash¬
ington, DC: U.S. Government Printing Office.
Myer, T (1972). Effects of viewing justified and unjustified real film violence on aggres¬
sive behavior. Journal of Personality and Social Psychology, 23, 21-29.
Milavsky, J., Kessler, R., Stipp, H., St Rubens, W. (1982). Television and aggression: The
results of a panel study. New York: Academic Press.
Murray, J. (1980). Television and youth: 25 years of research and controversy. Boys Town,
NE: Boys Town Center for Study of Youth Development.
Murray, J., St Kippax, S. (1979). From the early window to the late night show: Interna¬
tional trends in the study of television's impact on children and adults. In L.
Berkowitz (Ed.), Advances in experimental social psychology (vol. 12, pp. 254-320).
New York: Academic Press.
National Institute of Mental Health. (1982). Television and behavior: Ten years of scien¬
tific progress. Vol 1. Summary report. Washington, DC: U.S. Government Printing
Office.
Noble, G. (1973). Effects of different forms of filmed aggression on children's construc¬
tive and destructive play. Journal of Personality and Social Psychology, 26, 54-59.
Palmer, E., St Dorr, A. (Eds.). (1980). Children and the faces of television: Teaching, vio¬
lence, selling. New York: Academic Press.
Parke, R., Berkowitz, L., Leyens, J., West, S., St Sebastian, R. (1977). Some effects of vio¬
lent and nonviolent movies on the behavior of juvenile delinquents. In L. Berkowitz
235
(Ed.), Advances in experimental social psychology (vol. 10, pp. 136-172). New York:
Academic Press.
Phillips, D. (1974). The influence of suggestion on suicide: Substantive and theoretical
implications of the Werther effect. American Sociological Review, 39, 340-354.
Phillips, D. (1979). Suicide, motor vehicle fatalities, and the mass media: Evidence toward
a theory of suggestion. American Journal of Sociology, 84, 1150-1174.
Phillips, D. (1980). Airplane accidents, murder, and the mass media: Towards a theory of
imitation and suggestion. Social Forces, 58, 1001-1024.
Phillips, D. (1982). The impact of fictional television stories on U.S. adult fatalities: New
evidence on the effect of mass media on violence. American Journal of Sociology,
87, 1340-1359.
Phillips, D. (1983). The impact of mass media violence on U.S. homicides. American Soci¬
ological Review, 48, 560-568.
Rosenfeld, E., Maloney, S., Huesmann, R., Eron, L., Fischer, P, Musonis, V., St Washing¬
ton, A. (1978). The effect of fantasy behaviors and fantasy-reality discriminations
upon the observational learning of aggression. Paper presented at the annual meet¬
ing of the International Society for Research on Aggression, Washington, DC.
Rushton, f. (1979). Effects of prosocial television and film material on the behavior of
viewers. In L. Berkowitz (Ed.), Advances in experimental social psychology (vol. 12,
pp. 322-351). New York: Academic Press.
Shiffrin, R., St Schneider, W. (1977). Controlled and automatic human information pro¬
cessing: II. Perceptual learning, automatic attending, and a general theory. Psycho¬
logical Review, 84, 127-190.
Silverman, L., St Sprafkin, J. (1980). The effects of Sesame Street's prosocial spots on coop¬
erative-play between young children. Journal of Broadcasting, 24, 135-147.
Singer, J., St Singer, D. (1981J. Television, imagination and aggression. Hillsdale, NJ:
Erlbaum.
Sprafkin, J., Liebert, R., St Poulos, R. (1975). Effects of a prosocial televised example on
children's helping. Journal of Experimental Child, 20, 119-126.
Tannenbaum, P, St Gaer, E. P. (1965). Mood changes as a function of stress of protagonist
and degree of identification in a film-viewing situation. Journal of Personality and
Social Psychology, 2, 612-616.
Tarde, G. (1912). Penal philosophy. Boston: Little, Brown.
Thomas, M. (1982). Physiological arousal, exposure to a relatively lengthy aggressive
film, and aggressive behavior. Journal of Research in Personality, 16, 72-81.
Thomas, M., St Drabman, R. (1977). Effects of television violence on expectations of oth¬
ers’ aggression. Paper presented at the annual meeting of the American Psychologi¬
cal Association, San Francisco, CA.
Thomas, M., Horton, R., Lippincott, E., St Drabman, R. (1977). Desensitization to por¬
trayals of real-life aggression as a function of exposure to television violence. Jour¬
Turner, C., St Berkowitz, L. (1972). Identification with film aggressor (covert role taking)
and reactions to film violence. Journal of Personality and Social Psychology, 21,
256-264.
Turner, C., St Layton, J. (1976). Verbal imagery and connotation as memory induced medi¬
ators of aggressive behavior. Journal of Personality and Social Psychology, 33, 755-763.
Tyler, I. R. (1980). Impact of directly and indirectly experienced events: The origin of
crime-related judgments and behaviors. Journal of Personality and Social Psychol¬
ogy, 39, 13-28.
Vaughan, K., St Lanzetta, J. (1981). The effect of modification of expressive displays on
vicarious emotional arousal. Journal of Experimental Social Psychology, 17, 16-30.
Velten, E. (1968). A laboratory task for the induction of mood states. Behavior Research
and Therapy, 6, 473-482.
236
Leonard Berkowitz
Wheeler, L. (1966). Toward a theory of behavioral contagion. Psychological Review, 73,

179-182.
Wober, J. (1978). Televised violence and paranoid perception: The view from Great
Britain. Public Opinion Quarterly, 42, 315-321.
Worchel, S. (1972). The effect of films on the importance of behavioral freedom. Journal
of Personality, 40, 417-435.
Wyer, R., Jr., St Hartwick, J. (1980). The role of information retrieval and conditional infer¬
ence processes in belief formation and change. In L. Berkowitz (Ed.), Advances in
experimental social psychology (vol. 13, pp. 243-284). New York: Academic Press.
Wyer, R., Jr., St Srull, T. (1981). Category accessibility: Some theoretical and empirical
issues concerning the processing of information. In E. Higgins, C. Herman, St M.
Zanna (Eds.), Social cognition (vol. 1, pp. 161-197). Hillsdale, NJ: Erlbaum.
Zillmann, D., St Bryant, J. (1982). Pornography, sexual callousness, and the trivialization
of rape. Journal of Communication, 32, 10-21.
Zillmann, D., Bryant, J., Comisky, R, St Medoff, N. (1981). Excitation and hedonic valence
in the effect of erotica on motivated intermale aggression. European Journal of
i
Richard B. Felson
Mass Media Effects

on Violent Behavior It
Watching violence is a popular form of entertainment. A crowd of

onlookers enjoys a street fight just as the Romans enjoyed the gladia¬
tors. Wrestling is a popular spectator sport not only in the United
States, but in many countries in the Middle East. People enjoy combat
between animals—cock fights in Indonesia, bull fights in Spain, and dog
fights in rural areas of this country. Violence is frequently depicted in
folklore, fairy tales, and other literature. Local news shows provide
extensive coverage of violent crimes in order to increase their ratings.
Technological advances have dramatically increased the avail¬
ability of violent entertainment. The introduction of television was
critical, particularly in making violent entertainment more available to
children. More recently, cable systems, videocassette recorders, and
video games have increased exposure. Hand-held cameras and video
monitors now permit filming of actual crimes in progress. Economic
competition for viewers, particularly young viewers, has placed a pre¬
mium on media depictions of violence.
Not long after the introduction of television in American
households, there occurred a dramatic increase in violent crime (Cen-
terwall, 1989). Some scholars and commentators see a causal connec¬
tion. The most common argument is that children imitate the violence
they see on television. The process of imitation is emphasized by social
learning theory—a well-established approach in social psychology (Ban¬
dura, 1983). For both practical and theoretical reasons, then, an inter¬
est developed in examining whether exposure to violence in the media
affects the incidence of violence.
Violence usually refers to physical aggression. Aggression is
With permission from the Annual Review of Sociology, vol. 22, 1996, pp. 103-128,
© 1996, by Annual Reviews.
237
238
Richard B. Felson
Figure 36. Harvey Keitel takes aim in Quentin Tarantino’s Reservoir Dogs, a film awash in
blood and mayhem.
usually defined as any behavior involving an intent to harm another

person. Some studies of media effects, however, examine behaviors that
do not involve an intent to harm. For example, a common procedure is
to see whether children will hit a "Bobo" doll after observing an adult
model do so or after being exposed to media violence. It seems unlikely
that hitting a Bobo doll involves an intent to do harm (Tedeschi et al.,
1974). Other studies include measures of nonviolent criminal behavior,
most of which do not involve an intent to do harm. Of course, it
depends on what is meant by intent, a term most researchers do not
define. Tedeschi and Felson (1994) define an intent to do harm as a
behavior in which the actor expects the target will be harmed and val¬
ues that harm.1 Offenders who commit larceny and other nonviolent
crimes know that the victim will be harmed, but in most cases they do
not value that harm,- harm is not their goal.
In the first section of this essay, I discuss the empirical evidence
regarding Whether media violence has a causal effect on the aggressive
behavior of viewers. I review the classic studies, the meta-analyses, and
some more recent research. In the second section I examine the theo¬
retical processes that might explain short-term effects, should they
exist, and discuss relevant evidence. I do the same for long-term effects
in the third section.2
239
i
Mass Media Effects on Violent Behavior
Empirical Evidence Regarding Media Effects on Aggression
The relationship between exposure to media violence and aggression

has been examined using laboratory experiments, field experiments,
natural experiments, and longitudinal analyses based on correlational
data. I review some of the key research in each of these domains.
Laboratory Experiments
Laboratory experiments examine short-term effects of media violence.
Most studies show that subjects in laboratory experiments who observe
media violence tend to behave more aggressively than do subjects in
control groups. A meta-analysis of these studies reveals consistent and
substantial media effects (Andison, 1977). However, research is incon¬
sistent in showing whether it is necessary to provoke subjects before
showing violence to get an effect (Freedman, 1984). Thus, it is not clear
whether media exposure acts as an instigator of aggression in the labo¬
ratory or merely as a facilitator.
Researchers have raised questions about the external validity of
laboratory experiments in this area (Freedman, 1984; Cook et al., 1983).
They point out that the laboratory situation is very different from
situations leading to violence outside the laboratory (e.g., Tedeschi &
Felson, 1994). For subjects to engage in aggressive behavior in the lab¬
oratory, the behavior must be legitimated. Subjects are told, for exam¬
ple, that the delivery of shocks is a teaching method or part of a game.
Subjects are then subjected to an attack by a confederate and given a
chance to retaliate. Unlike aggressive behavior outside the laboratory,
there is no possibility that this will be punished by third parties or that
this will subject them to retaliation from the target. It is unknown to
what extent these differences limit the generalizability of experimen¬
tal studies. Evidence suggests that aggression measures in many labo¬
ratory studies do involve an intent to harm (Rerkowitz &. Donnerstein,
1982). Experimental subjects may not be so different from those who
engage in violence outside the laboratory, who see their behavior as
legitimate, and who do not consider its costs.3
The demand cues in these studies are probably a more significant
problem. Demand cues are instructions or other stimuli that indicate to
subjects how the experimenter expects them to behave.4 Experimenters
who show violent films are likely to communicate a message about their
attitudes toward aggression. A violent film may imply to subjects that
the experimenter is a permissive adult or someone not particularly
240
Richard B. Felson
offended by violence. Just a few subjects aware of the demand and com¬
pliant could account for the mean differences in aggression found
between experimental conditions.
The laboratory is a setting that exaggerates the effects of con¬
formity and social influence (see Gottfredson & FFirschi, 1993). The
extent of compliance in laboratory settings is dramatically demon¬
strated in Milgram's (1974) well-known research on obedient aggres¬
sion. Subjects' behavior is easily influenced for at least three reasons:
(1) The standards for behavior are unclear and the situation is novel
(Nemeth, 1970); (2) subjects are influenced by the prestige of the ex¬
perimenter and the scientific enterprise,- (3) subjects want to avoid
being perceived as psychologically maladjusted by the psychologist-
experimenter (Rosenberg, 1969).
Field Experiments
Concerns about external validity have stimulated researchers to em¬

ploy field experiments. Field experiments retain the advantages of
experimental design but avoid the problem of demand cues since sub¬
jects do not usually know they are being studied. A number of such
studies have been carried out in institutionalized settings (Feshbach &.
Singer, 1971; Leyens et al., 1975; Parke et al., 1977). In these studies,
boys are exposed to either violent or nonviolent programming, and their
aggressive behavior is observed in the following days or weeks. Each of
the studies has some important methodological limitations (see Freed¬
man, 1984). For example, although the boys in each treatment lived
together, the studies used statistical procedures that assumed that each
boy's behavior was independent. Even if one overlooks the limitations,
the results from these studies are inconsistent. In fact, one of the stud¬
ies found that the boys who watched violent television programs were
less aggressive than the boys who viewed nonviolent shows (Feshbach
& Singer, 1971).
The results of field experiments have been examined in at least
three meta-analyses. Flearold's (1986) meta-analysis of a broad range of
experimental studies revealed an effect for laboratory experiments but
no effect fof field experiments. A meta-analysis that included more
recent studies, however, did find an effect for field experiments (Paik &.
Comstock, 1994). Finally, Wood et al.'s meta-analysis (1991) was
restricted to field studies of media violence on unconstrained social
interaction.5 In all of these studies children or adolescents were ob¬
served unobtrusively after being exposed to an aggressive or nonag-
241
gressive film. In sixteen studies subjects engaged in more aggression fol¬

lowing exposure to violent films, while in seven studies subjects in the
control group engaged in more aggression. In five of the studies there
was no difference between control and experimental groups.
Natural Experiments: The Introduction, of Television
These studies take advantage of the fact that television was introduced
at different times in different locations. They assume that people who
are exposed to television will also be exposed to a high dose of televi¬
sion violence. This is probably a reasonable assumption given the
extremely high correlation between television viewing and exposure to
television violence (Milavsky et al., 1982).
Hennigan et al. (1982) compared crime rates in American cities
that already had television with those that did not. No effect of the
presence or absence of television was found on violent crime rates in a
comparison of the two kinds of cities. Furthermore, when cities with¬
out television obtained it, there was no increase in violent crime. There
was an increase in the incidence of larceny, which the authors attrib¬
uted to relative deprivation suffered by viewers observing affluent
people on television.6
Joy et al. (1986) examined changes in the aggressive behavior of
children after television was introduced into an isolated Canadian town
in the 1970s. The town was compared to two supposedly comparable
towns that already had television. Forty-five children in the three
towns were observed on the school playground in first and second grade
and then again two years later. The frequency of both verbal and phys¬
ical aggression increased in all three communities, but the increase was
significantly greater in the community in which television was intro¬
duced during the study. Some of the results were not consistent with a
television effect, however. In the first phase of the study, the children
in the community without television were just as aggressive as the chil¬
dren in the communities that already had television. Without televi¬
sion they should have been less aggressive. The children in the
community where television was introduced then became more aggres¬
sive than the children in the other communities in the second phase,
when all three communities had television. At this point, the level of
aggressive behavior in the three communities should have been simi¬
lar. To accept the findings, one must assume that the community with¬
out television at the beginning of the study had more aggressive
children than the other communities for other reasons, but that this
242
Richard B. Felson
effect was counteracted in the first phase by the fact that they were not
exposed to television. That assumption implies that there are other dif¬
ferences between the communities and thus casts doubt on the findings
of the study.
Centerwall (1989) examined the relationship between homi¬
cide rates and the introduction of television in three countries: South
Africa, Canada, and the United States. Television was introduced in
South Africa in 1975, about twenty-five years after Canada and the
United States. The white homicide rate increased dramatically in the
United States and Canada about fifteen years after the introduction of
television, when the first generation of children who had access to tel¬
evision were entering adulthood. The white homicide rate declined
slightly in South Africa during this time period. While Centerwall ruled
out some confounding factors (e.g., differences in economic develop¬
ment), causal inference is difficult, given the many differences between
the countries involved. In addition, Centerwall could not determine at
the time he wrote whether the level of violence had increased fifteen
years after the introduction of television in South Africa,- thus an impor¬
tant piece of evidence was missing.
Centerwall also examined the effect of the introduction of tel¬
evision in the United States. He found that urban areas acquired tele¬
vision before rural areas, and their homicide rates increased earlier.
However, social changes in general are likely to occur in urban areas
before they occur in rural areas. He also found that households of
whites acquired television sets before households of blacks, and their
homicide rates increased earlier as well. It is difficult to imagine an
alternative explanation of this effect.
Still, the methodological limitations of these studies make it
difficult to have confidence in a causal inference about media effects.
The substantial differences between the comparison groups increase
the risk that the relationship between the introduction of television
and increases in aggression is spurious.
Natural Experiments: Publicized Violence
The effects of highly publicized violent events on fluctuations in homi¬

cide and suicide rates over time have been examined in a series of stud¬
ies (see Phillips, 1986 for a review). Phillips (1983) found an increase in
the number of homicides after highly publicized heavyweight champi¬
onship fights. Modeling effects were only observed when the losing
243
fighter and the crime victims were similar in race and sex. The loss of
prize fights by white fighters was followed by increases in deaths
through homicide of white males on days 2 and 8. The loss of prize
fights by blacks was followed by an increase in homicide deaths for
black males on days 4 and 5. The rise in the homicide rate was not can¬
celed out by a subsequent drop, suggesting that the prize fights affected
the incidence and not just the timing of homicides.
Baron and Reiss (1985) attribute these effects to the fact that
prize fights tend to occur during the week and homicides are more
likely to occur on weekends. They were able to replicate Phillips's find¬
ings by selecting weeks without prizefights and pretending that they
had occurred. In response to this critique, Phillips and Bollen (1985)
selected different weeks and showed that the weekend effect could not
account for all of the findings. Miller et al. (1991) replicated some of
Phillips's results, but found that the effect only occurred on Saturdays
following highly publicized fights.
Freedman (1984) has criticized Phillips's research on other
methodological grounds, and Phillips (1986) has addressed these criti¬
cisms. There are still unresolved questions, such as why effects tended
to occur on different days for different races. In addition, experimental
results suggest that watching boxing films does not affect the viewer's
aggressive behavior. Geen (1978) found that, when provoked, college
students were more aggressive after viewing vengeful aggression but
not after viewing a boxing match (see also Hoyt, 1970).
Longitudinal Surveys
Survey research demonstrates that the correlation between the amount

of exposure to television violence and frequency of aggressive behavior
generally varies between 0.10 and 0.20 (Freedman, 1984; see Paik St
Comstock, 1994 for slightly higher estimates). There are good reasons
to think the relationship is at least partly spurious. For example, chil¬
dren with favorable attitudes toward violence may be more likely to
engage in violence and also more likely to find violence entertaining to
watch. Also, children who are more closely supervised may be less
likely to engage in violence and less likely to watch television. Intelli¬
gence, need for excitement, level of fear, and commitment to school are
other possible confounding variables. Wiegman et al. (1992) found that
intelligence was negatively associated with both exposure to violence
and aggressive behavior.
244
Richard B. Felson
Longitudinal data has been used to examine whether viewing

television violence produces changes in aggressive behavior. These
studies statistically control for aggression at T1 in order to isolate
causal effects on aggression at T2. Spuriousness is still possible if some
third variable is associated with exposure to media violence and
changes in aggressive behavior over time.
The main longitudinal evidence for a causal link between view¬
ing violence and aggressive behavior has been provided by Eron, Hues-
mann, and their associates (Eron et al., 1972; Huesmann St Eron, 1986).
In the first study, they examined the effect of children's exposure to tel¬
evision violence at age 8 on aggressive behavior at age 18. A measure of
viewing television violence at Time 1 was obtained by asking parents
the names of their children's favorite television shows. These shows
were coded for the level of violence depicted. Aggressive behavior at
Time 2 was measured by ratings of aggressiveness by peers, self-reports,
and the aggression subscale on the MMPI.7 Effects of television violence
were found only for boys and only on the peer nomination measure.
In addition to the inconsistent results, there are some meas¬
urement problems in this study (see Surgeon General's Report on Tele¬
vision Violence, 1972; Freedman, 1984). First, the aggression measure
included items referring to antisocial behavior that do not involve
aggression. Second, the measure of television exposure is based on par¬
ents' beliefs about the favorite programs of their children. Later
research found that parental reports of their children's favorite pro¬
grams are not strongly correlated to children's self-reports of total expo¬
sure (Milavsky et al., 1982).
Three-year longitudinal studies of primary schoolchildren were
later carried out in five countries: Australia, Israel, Poland, Finland, and
the United States (Huesmann & Eron, 1986). Aggression was measured
by the same peer nomination measure as the one used in the earlier
research. The children were asked to name one or two of their favorite
programs and to indicate how often they watched them. Complex and
inconsistent results were obtained. In the United States, television vio¬
lence had a significant effect on the later aggressiveness of females but
not males, a reversal of the effect found in their first study (Huesmann
&. Eron, 1986). An effect of the violence of favorite programs on later
aggression was found only for boys who rated themselves as similar to
violent and nonviolent television characters. A similar conditional
effect was found for males in Finland, but there was no effect of view¬
ing television violence on later aggressiveness of females (Lagerspetz St
245
Viemero, 1986). In Poland a direct effect of violence in favorite programs

was found on later aggressiveness for both males and females (Fraczek,
1986). No effect of early viewing of television violence was found on
subsequent aggressiveness for either males or females in Australia
(Sheehan, 1986), or among children living in a kibbutz in Israel (Bach-
rach, 1986). A television effect was found for city children in Israel
when the measure of aggression was a single item asking "who never
fights." But the effect did not occur on the same peer nomination
measure that had been used in the other cross-national studies.
Negative evidence was obtained in a large-scale, methodologi¬
cally sophisticated, longitudinal study carried out by Milavsky et al.
(1982). Their study was based on data collected from 3,200 students in
elementary and junior high schools in Fort Worth and Minneapolis.
Students identified the programs they had watched in the past four
weeks and indicated how many times they had watched them; these
were coded for violent content.8 The authors refined the peer nomina¬
tion measure of aggression used by Eron et al. to include intentional
acts of harm-doing, but not general misbehavior.
There was no evidence that any of the measures of exposure to
television violence produced changes in aggressive behavior over time.
The authors corrected for measurement error and used a variety of time
lags, subsamples, and measures of exposure to television violence and
aggressive behavior. In spite of a thorough exploration of the data, they
found no evidence that exposure to violence on television affected the
aggressive behavior of children. While the coefficients in most of the
analyses were positive, they were all close to zero and statistically
insignificant. The abundance of positive correlations led some critics
to reject Milavsky et al.'s conclusion of no effect (e.g., Friedrich-Cofer
& Huston, 1986).
A more recent longitudinal study in the Netherlands also failed
to find a media effect (Wiegman et al., 1992). The children were sur¬
veyed in either the second or fourth grade and then again two years
later. Peer nominations were used as a measure of aggressive behavior.
The lagged effect of exposure on aggressive behavior was small and
statistically insignificant.
It is difficult to reach a conclusion on the long-term effects of
viewing television violence from these longitudinal studies. The stud¬
ies that used better measurement failed to find an effect. In the studies
where an effect was found, the relationship was between favorite show
violence and subsequent aggression, rather than the amount of exposure
246
Richard B. Felson
to television violence, and Milavsky et al. did not replicate that effect.
The findings reported in the cross-national studies were inconsistent
and had as many negative findings as positive ones. Therefore one must
conclude that longitudinal studies have not demonstrated a relation¬
ship between the amount of violence viewed on television and subse¬
quent aggressive behavior.9
Theoretical Explanations of Situational Effects
The experimental results described above show that exposure to media

violence can have at least a short-term effect on aggressive behavior. In
this section, I consider theoretical reasons for expecting situational
effects. I also review some of the evidence regarding these theoretical
mechanisms.
Cognitive Priming
According to a cognitive priming approach, the aggressive ideas in vio¬

lent films can activate other aggressive thoughts in viewers through
their association in memory pathways (Berkowitz, 1984). When one
thought is activated, other thoughts that are strongly connected are also
activated. Immediately after a violent film, the viewer is primed to
respond aggressively because a network of memories involving aggres¬
sion is retrieved. Evidence indicates that media violence does elicit
thoughts and emotional responses related to aggression (Bushman &
Geen, 1990).
Huesmann (1982) makes a similar argument. He suggests that
children learn problem-solving scripts in part from their observations
of others' behavior. These scripts are cognitive expectations about a
sequence of behaviors that may be performed in particular situations.
Frequent exposure to scenes of violence may lead children to store
scripts for aggressive behavior in their memories, and these may be
recalled in a later situation if any aspect of the original situation—even
a superficial one—is present.
The classic studies of these effects involve the exposure of sub¬
jects to the fight scene from a film, The Champion, starring Kirk Doug¬
las. In one of these studies subjects were either shocked frequently or
infrequently by a confederate, witnessed the fight scene, or viewed a
neutral film, and then had an opportunity to shock the confederate,
247
whose name was either Bob or Kirk (Berkowitz St Geen, 1966). Subjects
gave the confederate the most shocks in the condition when they had
been provoked, had viewed the violent film, and the confederate had
the same name as the film's star.
Tedeschi and Norman (1985) attribute the results from these
studies to demand cues (see also Tedeschi &. Felson, 1994). They point
out that experimenters mention the1'fact that the confederate's first
name is the same as Kirk Douglas's in their instructions, and that they
justify to subjects the beating that Kirk Douglas received. A series of
studies have shown that it is necessary to provide this justification to
get a violent film effect (Geen &. Berkowitz, 1967; Berkowitz, 1965;
Berkowitz et al., 1962; Berkowitz St Rawlings, 1963; Meyer, 1972b).
Josephson (1987) examined the combined effects of exposure to
a violent film and retrieval cues in a field experiment with second- and
third-grade boys. The boys were exposed to either a violent film (in
which a walkie-talkie was used) or a nonviolent film. The boys were also
frustrated either before or after the film. Later they were interviewed by
someone holding either a walkie-talkie or a microphone. After the inter¬
view, the boys played a game of field hockey and their aggressive behav¬
ior was recorded. It was predicted that boys who were exposed to both
violent television and a walkie-talkie would be most aggressive in the
game, since the walkie-talkie would lead them to retrieve scripts asso¬
ciated with the violent film. The hypothesis was confirmed for boys who
were, according to teacher ratings, aggressive. Boys who were identified
as nonaggressive inhibited their aggression when exposed to the walkie-
talkie and the film. Josephson suggested that for these nonaggressive
boys, aggression may be strongly associated with negative emotions
such as guilt and fear which, when primed, may inhibit aggression. If we
accept this post hoc interpretation, it suggests that media violence may
increase or inhibit the violent behavior of viewers, depending on their
initial predisposition. Such effects are likely to be short-term, and they
may have no effect on the overall rate of violence.
Arousal from Pornography
According to Bandura (1973), emotional arousal facilitates and intensi¬

fies aggressive behavior. The facilitating effect of emotional arousal
occurs only when the individual is already prone to act aggressively. If
the individual is predisposed to behave in some other way, then emo¬
tional arousal will facilitate that behavior. Arousal energizes any
behavior that is dominant in the situation.
248
Richard B. Felson
Zillmann (1983) explains the facilitative effects of arousal in

terms of excitation transfer. He has proposed that arousal from two dif¬
ferent sources may combine with one another and be attributable to the
same source. When the combined arousal is attributed to anger, the
individual is likely to be more aggressive than would have been the case
if only the anger-producing cue has been present.
Some research has examined whether the arousal produced by
pornography facilitates aggressive behavior. A series of experiments
have been carried out in which subjects are exposed to sexual stimuli
and then allowed to aggress against another person, who may or may not
have provoked them. The prediction is that arousal produced by pornog¬
raphy should increase aggression when a subject has been provoked. The
message communicated by pornography and the gender of actor and
target should not matter unless they affect the level of arousal.
Experiments that have examined the effects of arousal from
pornography have produced mixed results. Some studies have found
that erotic films increased the aggressiveness of subjects who had been
provoked by the victim, while others have shown that pornography has
an inhibitory effect (Zillman, 1971; Meyer, 1972a; Zillmann et al.,
1974; Baron St Bell, 1973, 1977; Donnerstein et al., 1975).
Researchers have developed hypotheses to provide explana¬
tions for the conditions under which opposite effects are obtained
(Baron, 1974; White, 1979). Zillmann et al. (1981) explained the con¬
tradictory findings using an arousal-affect hypothesis. They proposed
that arousal has both an excitation component and an affective com¬
ponent. If arousal is accompanied by negative affect, it should add to
the arousal produced by anger and increase the level of aggression. If
arousal is accompanied by positive affect, it should subtract from the
arousal produced by anger and decrease the level of aggression. The
findings from research on the arousal-affect hypothesis are inconclu¬
sive (see Sapolski, 1984; Tedeschi St Felson, 1994 for reviews).
Even if these results are real, their significance for pornography
effects outside the experimental lab seems trivial. They suggest, for
example, that a man enjoying a pornographic film is less dangerous
when provoked, while a man who dislikes the film, but is still aroused
by it, is mo’re likely to retaliate for a provocation. Perhaps the findings
have more implications for the effects of arousal from other sources. For
example, it is possible that arousal from the car chase in the Rodney
King incident contributed to the violent behavior of the police.
It is difficult to manipulate arousal in the laboratory without
249
also affecting the meanings subjects give to those manipulations (Neiss,

1988). Experimenters who show pornographic films communicate
information about their values and expectations and thus create
demand cues. I discuss this issue in the next section.
Sponsor Effects
/
Demand cues provide a general explanation of short-term media effects

in the experimental laboratory. Wood et al. (1991) suggest that demand
cues may be a type of "sponsor effect" that occurs outside the experi¬
mental laboratory as well:
Viewers are likely to believe that the violent presentation is condoned

by the media sponsor, whether it be an experimenter, one's family,
the television networks or movie studios, or society in general. . . .
Sponsor effects are not artifacts of laboratory procedures; they also
occur in field settings (Wood et al., 1991, 373).
Wood et al.'s (1991) concept of sponsor effects appears to in¬

clude both social learning and situational conformity. Social learning
involves socialization and enduring effects on the viewer. Viewers may
be more likely to internalize a media message if they think it is spon¬
sored by someone they respect. A sponsor effect would enhance what¬
ever message is being conveyed.
Field and laboratory experiments seem more likely to produce
sponsor effects involving situational conformity. By showing a violent
film, sponsors may communicate that they are not very strict or that
they have a permissive attitude toward aggressive behavior. Young peo¬
ple, who are normally inhibited in front of adults, may engage in aggres¬
sive behavior if they think that they can get away with it. For example,
students often misbehave when they encounter less experienced sub¬
stitute teachers. According to this line of thinking, young people who
are exposed to media violence should feel disinhibited and should be
more likely to misbehave in a variety of ways, at least while adults are
present. When the sponsors of the film are no longer present, the effects
should disappear.
Meta-analyses show that exposure to violence is related to
nonaggressive forms of antisocial behavior. Hearold (1986) performed a
meta-analysis of experiments that included studies of effects of expo¬
sure to media violence on antisocial behavior generally. The effects of
250
Richard B. Felson
media violence on antisocial behavior were just as strong as the effects

of media violence on violent behavior. A more recent meta-analysis
that focused on all types of studies yielded similar results (Paik & Corn-
stock, 1994).
A study performed by Friedrich and Stein (1973) provides an
example of an experiment showing general effects of exposure to media
violence on antisocial behaviors. They found that nursery school chil¬
dren exposed to violent cartoons displayed more aggression during free
play than children exposed to neutral films. However, they also found
that children exposed to violence had lower tolerance for minor delays
and lower task persistence, and displayed less spontaneous obedience
in regard to school rules. These behaviors clearly do not involve an
intent to harm.
Additional evidence for a sponsor effect comes from a study by
Leyens et al. (1975). They found that subjects delivered more shock to
another person when they anticipated that the experimenter would
show them violent films; it was not necessary for them to actually see
the films. The investigators attributed this effect to priming, based on
the assumption that the mere mention of violent films primes aggres¬
sive thoughts. It seems just as likely that sponsor effects were involved:
an experimenter who is willing to show a violent film is perceived as
more permissive or more tolerant of aggression.
The effects of exposure to television violence on antisocial
behavior generally cast doubt on many of the theoretical explanations
usually used to explain media effects on violence. Explanations involv¬
ing cognitive priming or arousal cannot explain why those who view vio¬
lence should engage in deviant behavior generally. Explanations that
stress modeling (to be discussed) cannot explain this pattern of effects
either. It is possible, however, that viewers imitate the low self-control
behaviors of the characters they observe in television and films, rather
than violence specifically. Children model the self-control behavior of
adults in experimental situations (Bandura & Walters, 1963), but it is not
clear whether socialization or short-term situational effects are involved.
Sponsor effects may also explain the results of experimental
studies involving exposure to pornography. Paik and Comstock's (1994)
meta-analysis shows effects of both pornography and violent pornogra¬
phy on antisocial behavior in general. Experimenters who show pornog¬
raphy, especially violent pornography, may imply that they condone or
at least are tolerant of taboo behavior (Reiss, 1986). Subjects may be dis-
inhibited in this permissive atmosphere and engage in more antisocial
behavior.
251
%
In sum, these studies suggest that subjects may assume a more

permissive atmosphere when they are shown a violent film, and that
their inhibitions about misbehavior generally are reduced. It is not yet
clear whether their behavior reflects short-term conformity or longer-
term socialization. Research is needed to determine whether subjects
who view violent films in experiments engage in more aggression and
other misbehavior in the absence ofisponsors.
Television Viewing as a Routine Activity
According to the routine activity approach, crime should be less fre¬

quent when the routine activities of potential offenders and victims
reduce their opportunities for contact (e.g., M. Felson, 1986). Any activ¬
ity that separates those who are prone to violence from each other, or
from potential victims, is likely to decrease the incidence of violence.
Messner uses this approach to argue that watching television
can decrease the incidence of violence in society (Messner, 1986; Mess¬
ner &. Blau, 1987). Since people watch television at home, the opportu¬
nities for violence, at least with people outside the family, are probably
reduced. When people watch television, they may also interact less
often with other family members, so the opportunities for domestic
violence may also be reduced. Messner found that cities with high lev¬
els of television viewing have lower rates of both violent and nonvio¬
lent crime (Messner, 1986; Messner St Blau, 1987). However, in an
aggregate analysis of this type, one cannot determine the specific view¬
ing habits of offenders or victims of criminal violence.10
The routine activities of young adult males are particularly
important since they are most prone to use violence. Young adult males
do not spend as much time as other groups watching television (Dim-
mick et al., 1979). According to the routine activity approach, their
level of violence would be lower if they did.
Theoretical Explanations Involving Socialization
It is widely believed that people are more violent because they learn to be
violent from their parents, their peers, and the mass media. These social¬
ization effects tend to endure since they involve changes in the individ¬
ual. The evidence on the versatility of criminal offenders casts doubt on
the importance of this socialization process. Considerable evidence
252
Richard B. Felson
suggests that those who commit violent crime tend to commit nonvio¬
lent crime and other deviant acts as well. Studies of arrest histories based
on both official records and self-reports show a low level of specializa¬
tion in violent crime. For example, West and Farrington (1977) found
that 80 percent of adults convicted of violence also had convictions for
crimes involving dishonesty. Violent acts were also related to non¬
criminal forms of deviant behavior, such as sexual promiscuity, smok¬
ing, heavy drinking, gambling, and having an unstable job history.
The evidence that most offenders are versatile challenges the
notion that violent offenders are more violent because of a special pro¬
clivity to engage in violence, due to exposure to media violence or any
other factor. Individual differences in the propensity to engage in crim¬
inal violence reflect for the most part individual differences in antiso¬
cial behavior generally. Variations in the socialization of self-control
and other inhibitory factors are probably important causal factors
(Gottfredson & Hirschi, 1990). Theories that emphasize specific social¬
ization to violence are likely to be limited in their utility, since most
violent offenders are generalists.
The versatility argument should not be overstated. Some peo¬
ple do specialize in violence, and exposure to media violence may play
a role in their socialization. There are a variety of reasons one might
expect viewers to learn aggressive behavior from the media. First,
media depictions of violence may suggest novel behaviors to viewers
that they otherwise might not have considered. Second, vicarious rein¬
forcements and legitimation of violent actions may increase the ten¬
dency to model media violence. Third, viewers become desensitized
about violence after a heavy diet of it on television. Finally, people may
get a false idea of reality from observing a great deal of violence on tel¬
evision and develop unrealistic fears. I now examine each of the
processes more closely.
Learning Novel Forms of Behavior
Bandura (1983) has argued that television can shape the forms that
aggressive behavior takes. Television can teach skills that may be use¬
ful for committing acts of violence, and it can direct viewers' attention
to behaviors that they may not have considered. For example, young
people may mimic karate and judo moves, or they may learn effective
tactics for committing violent crime. This information may give direc¬
tion to those who are already motivated to engage in aggression. Such a
253
modeling process could lead to more severe forms of aggression. It could

increase the frequency of violence if people who are motivated to harm
someone choose a violent method they have observed on television.
There is anecdotal evidence that bizarre violent events have fol¬
lowed soon after their depiction on television, suggesting a form of copy¬
cat behavior. In one widely reported case in Boston, six young men set
fire to a woman after forcing her to douse herself with fuel. The scene
had been depicted on television two nights before. In another instance,
four teenagers raped a 9-year-old girl with a beer bottle, enacting a scene
similar to one in the made-for-TV movie Bom Innocent. Such incidents
may be coincidental, but they suggest the possibility that unusual and
dramatic behaviors on television are imitated by viewers who might
never otherwise have imagined engaging in such behaviors.
Modeling can also be used to explain contagion effects observed
for highly publicized violence, such as airline hijackings, civil disor¬
ders, bombings, and political kidnapping. The tendency for such events
to occur in waves suggests that at least some viewers imitate real events
that are reported on television. However, the central argument about
the relationship of viewing violence on television and viewers' aggres¬
sive behavior focuses on fictional events.
Vicarious Reinforcement and Legitimations
Bandura (1983) also suggested that television may inform viewers of the
positive and negative consequences of violent behavior. Audiences can
be expected to imitate violent behavior that is successful in gaining the
model's objectives in fictional or nonfictional programs. When violence
is justified or left unpunished on television, the viewer's guilt or con¬
cern about consequences is reduced. Thus Paik and Comstock's (1994)
meta-analysis found that the magnitude of media effects on antisocial
behavior was greater when the violent actor was rewarded or the behav¬
ior was legitimated.
It is not at all clear what message is learned from viewing vio¬
lence on television. In most plots, the protagonist uses violence for
legitimate ends while the villain engages in illegitimate violence. The
protagonist usually uses violence in self-defense or to mete out an
appropriate level of punishment to a dangerous or threatening criminal.
Television conveys the message that while some forms of violence are
necessary and legitimate, criminal violence is evil.
The consequences of the illegitimate violence portrayed in fic-
254
Richard B. Felson
tional television and film are more negative than the consequences of ille¬
gitimate violence in real life. In real life violent people often evade pun¬
ishment, while in television, the villain is almost always punished. Thus,
one could argue that television violence might reduce the incidence of
criminal violence, since crime doesn't pay for TV criminals. Another
difference is in the appeal of those who engage in illegitimate violence.
In fictional television, those who engage in illegitimate violence tend
to lack any attractive qualities that would lead to sympathy or identi¬
fication. In real life, illegitimate violence may be committed by loved
ones or others who are perceived to have desirable qualities.
Other factors may limit the effects of any message about the
legitimacy, or the rewards and costs of violence. First, the lessons
learned from the media about violence may be similar or redundant to
the lessons learned about the use of violence conveyed by other sources.
In fact, most viewers probably approve of the violent behavior of the
protagonists. The influence of television on viewers who already agree
with its message would be weak at best. Second, the audience may not
take the message from fictional plots seriously. Modeling is more likely
to occur after viewing nonfiction than after viewing fiction (Feshbach,
1972; Berkowitz &. Alioto, 1973).11 Third, the violent contexts and
provocations observed on television are likely to be very different from
the contexts and provocations people experience in their own lives. Evi¬
dence suggests that viewers take context and intentions into account
before they model aggressive behavior (Geen, 1978; Floyt, 1970). Straus
(Baron &. Straus, 1987), on the other hand, suggests that people are
likely to be influenced by the violence they observe regardless of its
context, message, or legitimacy. According to cultural spillover theory,
violence in one sphere of life leads to violence in other spheres.
Finally, some young children may miss the more subtle aspects
of television messages, focusing on overt acts rather than on the inten¬
tions or contexts in which such acts occur. Collins et al. (1984) found
that kindergarten and second-grade children were relatively unaffected
by an aggressor's motives in their understanding of a violent program.
They focused more on the aggressiveness of the behavior and its ulti¬
mate consequences. However, even if young children imitate the vio¬
lence of models, it is not at all clear that they will continue to exhibit
violence as they get older. When they are older, and they pay attention
to the intentions and context in violent television, their behavior is
more likely to reflect the messages they learn. It is also at these later
ages that violent behavior, if it should occur, is likely to be dangerous.
255
%
Creating Unrealistic Fear

Bandura (1983) claims that television distorts knowledge about the dan¬
gers and threats present in the real world. The notion that television
viewing fosters a distrust of others and a misconception of the world as
dangerous has been referred to as the "cultivation effect" (Gerbner St
Gross, 1976). Research shows that heavy television viewers are more
distrustful of others and overestimate their chances of being criminally
victimized (see Ogles, 1987 and Gunter, 1994 for reviews).12 The
assumption is that these fears will lead viewers to perceive threats that
do not exist and to respond aggressively. It is just as plausible that such
fears would lead viewers to avoid aggressive behavior against others, if
they feel it is dangerous and might lead to victimization. Persons who
fear crime may also be less likely to go out at night or go to places where
they may be victimized. If viewing television violence increases fear, it
might decrease the level of violence.
Desensitization
Frequent viewing of television violence may cause viewers to be less
anxious and sensitive about violence. Someone who becomes desensi¬
tized to violence may be more likely to engage in violence. This argu¬
ment assumes that anxiety about violence inhibits its use.
Desensitization has been examined indirectly using measures
of arousal. Research shows that subjects who view violent films are less
aroused by violence later on (Thomas et al., 1977; see Rule &. Ferguson,
1986 for a review). In addition, heavy viewers of television violence
tend to respond less emotionally to violence than do light viewers.
There is no evidence that desensitization produces lower levels of
violent behavior.13 Nor is it clear what effect should occur. Studies of desen¬
sitization measure arousal not anxiety) and arousal can facilitate violent
behavior, according to the literature cited earlier (e.g., Zillmann, 1983). If
viewers are exposed to a heavy diet of television violence, one might argue
that they will be less aroused by violence and therefore less likely to engage
in violence. In addition, if viewers become desensitized to violent behavior
on television, they may become indifferent to its message. Desensitization
could thereby weaken the effect of a heavy diet of television violence.
Messages from Pornography

The discussion of situational effects of pornography on aggression
focused on arousal as a mediating variable. Feminists have argued that
256
Richard B. Felson
pornography has special effects on violence against women because of

the message it communicates (Dworkin, 1981; MacKinnon, 1984).
Exposure to pornography supposedly leads to negative attitudes toward
women which, in turn, affects the likelihood of rape and other forms of
violence against women. It is argued, for example, that pornography
leads male viewers to think of women as sex objects or as promiscuous
(Linz &. Malamuth, 1993). Furthermore, some erotica portrays scenes
of rape and sadomasochism. In such fictional forms the female victim
may express pleasure during and after being raped, suggesting that
women enjoy such treatment. Males who view such films may be
induced to believe that forceful sexual acts are desired by women. In
addition, unlike illegitimate violence not associated with sex, violence
in pornographic films rarely has negative consequences for the actor
(Palys, 1986; Smith, 1976).
Evidence does not support the hypothesis that exposure to non¬
violent pornography leads to violence toward women. Most experimen¬
tal studies show no difference in aggression toward women between
subjects exposed to pornographic films and control groups (for reviews,
see Donnerstein, 1984; Linz St Malamuth, 1993). Research outside the
laboratory has not demonstrated that exposure to pornography and vio¬
lence toward women are even correlated, much less causally related.
There is evidence that rapists report less exposure to pornography than
controls, not more (see Linz St Malamuth, 1993 for a review). Studies
of the relationship between exposure to pornography and use of sexual
coercion among college students yield mixed results (Demare et al.,
1993; Boeringer, 1994).
Research using aggregate data has also failed to demonstrate a
relationship between exposure to pornography and violence against
women. Studies of the effect of changes in restrictions on pornography
on rape rates show inconsistent results. States in which sex-oriented
magazines are popular tend to have high rape rates (Baron & Straus
1987). However, it is questionable whether the state is a meaningful
unit of analysis, given the heterogeneity within states. Gentry (1991)
found no relationship between rape rates and circulation of sexually ori¬
ented magazines across metropolitan areas.
Effects of violent pornography have been reported in laboratory
experiments, at least under certain conditions (see Linz &. Malamuth,
1993 for a review). Some studies show that an effect is obtained only if
the sexual assault has positive consequences. In this case, subjects are
told that the woman became a willing participant in the coercive sex¬
ual activities, and she is shown smiling and on friendly terms with the
257
man afterward (Donnerstein, 1980). However, in a more recent study,

exposure to a rape scene with positive consequences did not increase
subjects' aggression toward women (Fisher St Grenier, 1994).
The effects of exposure to violence with positive consequences
have been examined in a field experiment. College students were
exposed either to two films that showed women responding positively
to men who had attacked them or tohwo neutral films (Malamuth St
Check, 1981). Subjects completed a survey that they thought was unre¬
lated to the films several days later. Males who had viewed the violent
films showed greater acceptance of violence against women. Note that
these films did not involve pornography. Pornographic films in which
the victim of sexual aggression is perceived as experiencing a positive
outcome are quite rare (Garcia St Milano, 1990).
The experimental evidence is mixed concerning whether porn¬
ography or violent pornography affects male attitudes toward women,
according to Linz's (1989) review of the literature. Evidence that men
who have negative attitudes toward women are more likely to engage
in violence against women is also inconsistent. Some studies find that
men who engage in sexual coercion have different attitudes toward
women and rape than do other men, while other studies do not (Kanin,
1969; Malamuth, 1986; Ageton, 1983; Rapapport St Burkhart, 1984). It
may be that sexually aggressive men are more likely to have antisocial
attitudes generally. Thus, convicted rapists are similar to males con¬
victed of other offenses in their attitudes toward women and women's
rights (e.g., Howells St Wright, 1978) and in their belief in rape myths
(Hall et al., 1986).
The literature on violence and attitudes toward women is
plagued by conceptual and measurement problems. Measures of belief
in rape myths are problematic (Tedeschi St Felson, 1994). In addition,
traditional attitudes about gender roles do not necessarily involve neg¬
ative attitudes toward women and may be negatively associated with
violence toward women and exposure to pornography. Thus, rape rates
are twice as high at private colleges and major universities than at reli¬
giously affiliated institutions (Koss et al., 1987). Males who report
greater exposure to pornography have more (not less) liberal attitudes
toward gender roles (Reiss, 1986). Finally, even if a correlation between
certain attitudes regarding women and violence could be established,
the causal interpretation would be unclear. For example, it may be that
men express certain beliefs to justify coercive behavior already per¬
formed (Koss et al., 1985).
One limitation on the impact of pornography or any media
258
Richard B. Felson
effect is selective exposure (McGuire, 1986). Media effects are likely to

be limited to the extent that viewers choose programming that already
reflects their values and interests. The argument in regard to media vio¬
lence is that violence is so pervasive on television that all viewers,
including impressionable children, are exposed. In the case of pornog¬
raphy, particularly violent pornography, there is much more selective
exposure, since those interested in viewing this material must make
a special effort to'do so. In addition, the viewers of pornography are
usually adults, not children.
Pornography provides fantasies for masturbation. Viewers may
select material depicting activities that they already fantasize about.
When they substitute commercially produced'fantasies for their own
fantasies, the content is not necessarily more violent. Palys (1986)
found that less than 10 percent of scenes in pornography videos in¬
volved some form of aggression. A study of college students revealed
that approximately 39 percent of men and women reported that they
had fantasized about forced sex (Loren & Weeks, 1986).
The versatility evidence is also relevant to the literature on
pornography and rape. Most rapists do not specialize in rape or in vio¬
lent crime (Alder, 1984; Kruttschnitt, 1989). Therefore, theories that
emphasize socialization of rape-supportive attitudes, whether learned
from the media or elsewhere, are going to have limited utility for under¬
standing individual differences in the proclivity to rape.
In summary, some experimental research suggests that violent
pornography that depicts women enjoying the event can lead male sub¬
jects to engage in violence against women in the laboratory. The effect
of these films appears to be similar to the effects of violent films with¬
out a sexual theme. Demand cues provide an alternative explanation of
these results as well (see Reiss, 1986). The external validity of these
studies is questionable given the rarity of these themes in pornography,
and given selective exposure.
Summary and Conclusions
The inconsistencies of the findings make it difficult to draw firm con¬

clusions about the effects of exposure to media violence on aggressive
behavior. Most scholars who have reviewed research in the area believe
that there is an effect (Friedrich-Cofer St Fluston, 1986; Centerwall,
259
1989). Other scholars have concluded that the causal effects of expo¬
sure to television have not been demonstrated (Freedman, 1984;
McGuire, 1989).
Given the pervasiveness of media violence, it would be sur¬
prising if it had no effect on viewers. I agree with those scholars who
think that exposure to television violence probably does have a small
effect on violent behavior (Cook et al., 1983). The reason that media
effects are not consistently observed is probably because they are weak
and affect only a small percentage of viewers. These weak effects may
still have practical importance since, in a large population, they would
produce some death and injuries. However, it seems unlikely that
media violence is a significant factor in high crime rates in this coun¬
try. Changes in violent crimes mirror changes in crime rates generally.
In addition, the people who engage in criminal violence also commit
other types of crime. An explanation that attributes violent behavior to
socialization that encourages violence cannot easily explain the versa¬
tility of most violent criminals.
It seems likely that some people would be more susceptible to
media influence than others. Therefore it is puzzling that research has
not shown any consistent statistical interactions involving individual
difference factors and media exposure. The failure to find individual dif¬
ference factors that condition the effects of media exposure on aggres¬
sive behavior contributes to skepticism about media effects.
It seems reasonable to believe that the media directs viewers'
attention to novel forms of violent behavior they might not otherwise
consider. The anecdotal evidence is convincing in this area. There
appear to be documented cases in which bizarre events on television
are followed by similar events in the real world; the similarities seem
too great to be coincidental. In addition, hijackings and political vio¬
lence tend to occur in waves. Many parents have observed their chil¬
dren mimicking behaviors they've observed in films. Whether this
process leads to a greater frequency of violence is unclear.
There is some evidence that the effects observed in laboratory
experiments, and less consistently in field experiments, are due to spon¬
sor effects. The fact that children who are exposed to violence tend to
misbehave generally casts doubt on tiiost of the other theoretical expla¬
nations of media effects. The issue has particular significance for labora¬
tory research, where subjects know they are being studied and may be
responding to demand cues. Research is needed in which sponsor effects
are isolated and controlled. A field experiment in which subjects imitate
260
Richard B. Felson
violent behavior they have observed in the absence of the sponsor, but
do not misbehave otherwise, would be convincing. Alternatively, there
may need to be further development of the theoretical argument that
self-control behavior is modeled.
It is not clear what lesson the media teaches about the legiti¬
macy of violence, or the likelihood of punishment. To some extent that
message is redundant with lessons learned from other sources of influ¬
ence. The message is probably ambiguous and is likely to have different
effects on different viewers. Young children may imitate illegitimate
violence if they do not understand the message, but their imitative
behavior may have trivial consequences. Out of millions of viewers,
there must be some with highly idiosyncratic interpretations of televi¬
sion content who intertwine the fantasy with their own lives, and as a
result have an increased probability of engaging in violent behavior.
NOTES
1. An alternative definition is that intentional harm involves deliberate harm or

expected harm. However, teachers sometimes give low grades with the expectation that
it will make their students unhappy, but their behavior should not be defined as aggres¬
sive, unless they also value that harm. Tedeschi and Felson (1994) substitute the term
coercion for aggression and include coercive actions in which the actor values compli¬
ance as well as harm.
2. This chapter borrows from Tedeschi and Felson (1994).
3. According to Freedman (1984), effects outside the laboratory are likely to be
weaker than laboratory effects because violent programs are mixed with other types of
programs. Friedrich-Cofer and Huston (1986) dispute this point, arguing that experimen¬
tal research underestimates media effects. They claim that the stimuli used in experi¬
mental research are brief and often less violent than typical television programs and that
the presence of experimenters inhibits subjects from engaging in aggressive behavior in
laboratory settings.
4. Any cue that indicates which direction the experimenter prefers would be a
demand cue. In their strongest form demand cues give away the experimenter's hypoth¬
esis to subjects, who then compliantly act to confirm the hypothesis. In their weaker
form, demand cues simply guide behavior without creating awareness of the hypothesis.
5. Some of the studies were in laboratory settings, but subjects did not know that
their aggressive behavior was being observed as part of the study.
6. The hypothesis that consumerism, promoted by advertising and the depiction of
wealth on television, leads to more financially motivated crime has never been tested, to
my knowledge.
7. An important requirement of such studies is that they control for the aggres¬
siveness of the viewer at the earlier time period, when looking at the effect of earlier expo¬
sure on later aggression. Eron and Huesmann do so in later reanalyses of their data.
8. Also included were parental reports of a child's favorite programs, and self-reports
of children of their favorite programs. These measures of exposure to television violence
were poor indicators of overall exposure.
9. Valkenburg et al. (1992) found that violent programming increased the level of
261
aggressive-heroic'fantasies found in a longitudinal analyses among Dutch children. How¬

ever, nonviolent dramatic programming had the same effect.
10. Viewing violent television and viewing television are so highly correlated across
cites that it does not matter which measure is used in analysis. The notion of catharsis
provides an alternative explanation, but it cannot explain the negative relationship
between exposure to television violence and the incidence of nonviolent crime.
11. In Paik and Comstock's (1994) meta-analyses the strongest effects were observed
for cartoon programs. However, the subjects in these studies were children, and children
may be more easily influenced. / *’
12. There is some evidence that the relationship is spurious,- see Gunter's (1994)
review.
13. Emergency room personnel may become desensitized to the consequences of
violent behavior, but there is no evidence that they are more violent than other groups
of people.
REFERENCES
Ageton, S. (1983). Sexual assault among adolescents. Lexington, MA: Lexington.

Alder, C. (1984). The convicted rapist: A sexual or a violent offender? Criminal Justice
Behavior 11, 157-177.
Andison, F. S. (1977). TV violence and viewer aggression: A cumulation of study results:
1956-1976. Public Opinion Quarterly 41, 314-31.
Bachrach, R. S. (1986). The differential effect of observation of violence on kibbutz and
city children in Israel. In L. R. Huesmann & L. D. Eron (Eds.), Television and the
aggressive child: A cross-national comparison, (pp. 201-238. Hillsdale, NJ: Erlbaum.
Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ:
Prentice-Hall.
Bandura, A. (1983). Psychological mechanisms of aggression. In R. Geen & E. I. Donner-
stein (Eds.), Aggression: Theoretical and empirical reviews (vol. 1, pp. 1-40). New
York: Academic.
Bandura, A., Walters, R. H. (1963). Social learning and personality development. New
York: Holt, Rinehart & Winston.
Baron, J. N., &. Reiss, P. C. (1985). Same time next year: Aggregate analyses of the mass
media and violent behavior. American Sociological Review, 50, 347-363.
Baron, L., St Straus, M. A. (1987). Four theories of rape: A macrosocial analysis. Social
Problems, 34, 467-489.
Baron, R. A. (1974). The aggression-inhibiting influence of heightened sexual arousal.
Baron, R. A., & Bell, P. A. (1973). Effects of heightened sexual arousal on physical aggres¬
sion. Proceedings of the 81st Annual Convention of the American Psychological
Association, 8, 171-172.
Baron, R. A., & Bell, P. A. (1977). Sexual arousal and aggression by males: Effects of type
of erotic stimuli and prior provocation. Journal of Personality and Social Psychol¬
ogy, 35, 79-87.
Berkowitz, L. (1965). Some aspects of observed aggression. Journal of Personality and
Berkowitz, L. (1984). Some effects of thought on anti- and pro-social influences of media
effects. Psychological Bulletin, 95, 410-427.
Berkowitz, L., St Alioto, J. T. (1973). The meaning of an observed event as a determinant
of its aggressive consequences. Journal of Personality and Social Psychology, 28,
206-217.
262
Richard B. Felson
Berkowitz, L., Corwin, R.; St Heironimus, M. (1962). Film violence and subsequent
aggressive tendencies. Public Opinion Quarterly, 27, 217-229.
Berkowitz, L., fit Donnerstein, E. (1982). External validity is more than skin deep: Some
answers to criticism of laboratory experiments. American Psychologist, 37, 245-257.
Berkowitz, L., & Geen, R. G. (1966). Film violence and the cue properties of available tar¬
gets. Journal of Personality and Social Psychology, 3, 525-530.
Berkowitz, L.; fit Rawlings, E. (1963). Effects of film violence: An inhibition against sub¬
sequent aggression. Journal of Abnormal Social Psychology, 66, 405-412.
Boeringer, S. (1994). Pornography and sexual aggression: Associations of violent and non¬
violent depictions with rape and rape proclivity. Deviant Behavior, 15, 289-304.
Bushman, B. J., & Geen, R. G. (1990). Role of cognitive-emotional mediators and indi¬
vidual differences in the effects of media violence on aggression. Journal of Person¬
ality and Social Psycholology, 58, 156-163.
Centerwall, B. S. (1989). Exposure to television as a cause of violence. In G. Comstock
(Ed.), Public communication and behavior (vol. 2, pp. 1-58. Orlando: Academic.
Collins, W. A., Berndt, T. J., fit Hess, V. L. (1984). Observational learning of motives and
consequences for television aggression: A developmental study. Child Develop¬
ment, 45, 799-802.
Cook, T. D., Kendzierski, D. A., fit Thomas, S. V. (1983). The implicit assumptions of tel¬
evision: An analysis of the 1982 NIMH Report on Television and Behavior. Public
Opinion Quarterly, 47, 161-201.
Demare, D., Lips, H. M., St Briere, J. (1993). Sexually violent pornography, anti-women
attitudes, and sexual aggression: A structural equation model. Journal of Research
in Personality, 27, 285-300.
Dimmick, J. W., McCain, T. A., & Bolton, W. T. (1979). Media use and the life span. Amer¬
ican Behavioral Scientist, 23, 7-31.
Donnerstein, E. (1980). Aggressive erotica and violence against women. Journal of Per¬
Donnerstein, E. (1984). Pornography: Its effect on violence against women. In N. M. Mala-
muth St E. Donnerstein (Eds.), Pornography and sexual aggression (pp. 53-81). New
York: Academic.
Donnerstein, E., Donnerstein, M., St Evans, R. (1975). Erotic stimuli and aggression: Facil¬
itation or inhibition. Journal of Personality and Social Psychology, 32, 237-244.
Dworkin, A. (1981). Pornography: Men possessing women. New York: G. P. Putnam's Sons.
Eron. L. D., Huesmann, L. R., Lefkowitz, M. M., St Walder, L. O. (1972). Does television
violence cause aggression? American Psychologist, 27, 253-263.
Felson, M. (1986). Routine activities, social controls, rational decisions and criminal out¬
comes. In D. Cornish St R. Clarke (Eds.), The reasoning criminal: Rational choice
perspectives on offending. New York: Springer-Verlag.
Feshbach, S. (1972). Reality and fantasy in filmed violence. In J. P. Murray, E. Rubinstein,
St G. A. Comstock (Eds.), Television and social behavior (vol. 2, pp. 318-345). Wash¬
ington DC: U.S. Government Printing Office.
Feshbach, S., fit Singer, R. (1971). Television and aggression. San Francisco: Jossey Bass.
Fisher, W. A., St Grenier, G. (1994). Violent pornography, antiwoman thoughts, and anti¬
woman acts: In search of reliable effects. Journal of Sex Research, 31. 23-38.
Fraczek, A. (19^6). Socio-cultural environment, television viewing, and the development
of aggression among children in Poland. In L. R. Huesmann St L. D. Eron (Eds.), Tele¬
vision and the aggressive child: A cross-national comparison (pp. 119-160. Hills¬
dale, NJ: Erlbaum.
Freedman, J. L. (1984). Effects of television violence on aggressiveness. Psychological Bul¬
letin, 96, 227-246.
Friedrich, L. K., St Stein, A. H. (1973). Aggressive and prosocial television programs and
263
the natural behavior of preschool children. Monograph of the Society for Research
in Child Development, 38 (4, Serial No. 151).
Friedrich-Cofer, L., St Huston, A. C. (1986). Television violence and aggression: The
debate continues. Psychological Bulletin, 100, 364-371.
Garcia, L. T., St Milano, L. (1990). A content analysis of erotic videos. Journal of Law Psy¬
chiatry, 14, 47-64.
Geen, R. G. (1978). Some effects of observing violence upon the behavior of the observer.
In B. Maher (Ed.), Progress in experimental personality research (vol. 8). New York:
Academic. ,
Geen, R. G. (1983). Aggression and television violence. In R. G. Geen St E. I. Donnerstein
(Eds.), Aggression: Theoretical and empirical reviews (vol. 2, pp. 103-125). New
York: Academic.
Geen, R. G, St Berkowitz, L. (1967). Some conditions facilitating the occurrence of aggres¬
sion after the observation of violence. Journal of Personality, 35, 666-676.
Gentry, C. S. (1991). Pornography and rape: An empirical analysis. Deviant Behavior, 12,
277-288.
Gerbner, G., St Gross, L. (1976). Living with television: The violence profile. Journal of
Communication, 26, 173-199.
Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford: Stanford Uni¬
versity Press.
Gottfredson, M., St Hirschi, T. (1993). A control theory interpretation of psychological
research on aggression. In R. B. Felson St J. T. Tedeschi (Eds.), Aggression and vio¬
lence: Social interactionist perspectives (pp. 47-68). Washington, DC: American
Psychological Association.
Gunter, B. (1994). The question of media violence. In J. Bryant St D. Zillman (Eds.), Media
effects: Advances in theory and research (pp. 163-212). Hillsdale, Nf: Erlbaum.
Hall, E. R., Howard, J. A., St Boezio, S. L. (1986). Tolerance of rape: A sexist or antisocial
attitude. Psychology of Women Quarterly, 10, 101-118.
Hearold, S. (1986). A synthesis of 1043 effects of television on social behavior. In G. Com¬
stock (Ed.), Public communication and behavior (vol. 1, pp. 65-133). San Diego, CA:
Academic.
Hennigan, K. M., Del Rosario, M. L., Heath, L., Cook, T. D., Wharton, J. D., St Calder, B.
f. (1982). The impact of the introduction of television on crime in the United States.
Howells, K., St Wright, E. (1978). The sexual attitudes of aggressive sexual offenders.
British Journal of Crimnology, 18, 170-173.
Hoyt, J. L. (1970). Effect of media violence 'justification' on aggression. Journal of Broad¬
casting, 14, 455^-64.
Huesmann, L. R. (1982). Television violence and aggressive behavior. In D. Pearl, L.
Bouthilet, St f. Lazar (Eds.), Television and behavior: Ten years of scientific progress
and implications for the eighties (vol. 2, pp. 220-256). Washington, DC: National
Institute of Mental Health.
Huesmann, L. R., St Eron, L. D. (1986). The development of aggression in American chil¬
dren as a consequence of television violence viewing. In L. R. Huesmann St L. D.
Eron (Eds.), Television and the aggressive child: A cross-national comparison (pp.
45-80). Hillsdale, NJ: Erlbaum.
fosephson, W. L. (1987). Television violence and children's aggression: Testing the prim¬
ing, social script, and disinhibition predictions. Journal of Personality and Social
Psychology, 53, 882890.
Joy, L. A., Kimball, M. M., St Zaback, M. L. (1986). Television and children's aggressive
behavior. In T. M. Williams (Ed.), The impact of television: A natural experiment in
three communities (pp. 303-360). New York: Academic.
264
Richard B. Felson
Kanin, E. J. (1969). Selected dyadic aspect of male sex aggression. Journal of Sex Research,
5, 12-28.
Koss, M. P., Gidycz, C. A., St Wisniewski, N. (1987). The scope of rape: Incidence and
prevalence of sexual aggression and victimization in a national sample of students
in higher education. Journal of Consultation in Clinical Psychology, 55, 162-170.
Koss, M. P., Leonard, K. E., Beezley, D. A., St Oros, C. J. (1985). Non-stranger sexual aggres¬
sion: a discriminate analysis classification. Sex Roles, 12, 981-992.
Kruttschnitt. C. (1989). A sociological, offender-based, study of rape. Sociological Quar¬
terly, 30, 305-329.
Lagerspetz, K., St Viemero, V. (1986). Television and aggressive behavior among Finnish
children. In L. R. Huesmann St L. D. Eron (Eds.), Television and the aggressive child:
A cross-national comparison (pp. 81-118). Hillsdale, NJ: Erlbaum.
Leyens, J. P., Camino, L., Parke, R. D.; St Berkowitz, L. (1975). Effects of movie violence
on aggression in a field setting as a function of group dominance and cohesion. Jour¬
Linz, D. (1989). Exposure to sexually explicit materials and attitudes toward rape: A com¬
parison of study results. Journal of Sex Research, 26, 50-84.
Linz, D., St Malamuth, N. (1993). Pornography. Newbury Park: Sage.
Loren, R. E. A., St Weeks, G. (1986). Sexual fantasies of undergraduates and their percep¬
tions of the sexual fantasies of the opposite sex. Journal of Sex Education Therapy,
12, 31-36.
MacKinnon, C. (1984). Not a moral issue. Yale Law Policy Review, 2, 321-345.
Malamuth, N. M. (1986). Predictors of naturalistic sexual aggression. Journal of Person¬
ality and Social Psychology, 50, 953-962.
Malamuth, N. M., St Check, J. V. P. (1981). The effects of mass media exposure on accept¬
ance of violence against women: A field experiment. Journal of Research in Person¬
ality, 15, 436-446.
McGuire, W. J. (1986). The myth of massive media impact: Savagings and salvagings. In
G. Comstock (Ed.), Public communication and behavior (vol. 1, pp. 175-257).
Orlando: Academic.
Messner, S. F. (1986). Television violence and violent crime: An aggregate analysis. Social
Problems, 33, 218-235.
Messner, S. F., St Blau, J. R. (1987). Routine leisure activities and rates of crime: A macro¬
level analysis. Social Forces, 65, 1035-1052.
Meyer, T. P. (1972a). The effects of sexually arousing and violent films on aggressive
behavior. Journal of Sex Research, 8, 324-331.
Meyer, T. P. (1972b). Effects of viewing justified and unjustified real film violence on
aggressive behavior. Journal of Personality and Social Psychology, 23, 21-29.
Milavsky, J. R., Stipp, H. H., Kessler, R. C., St Rubens, W. S. (1982). Television and aggres¬
sion: A panel study. New York: Academic.
Milgram, S. (1974). Obedience to authority: An experimental view. New York: Harper St
Row.
Miller, T. Q., Heath, L., Molcan, f. R., St Dugoni, B. L. (1991). Imitative violence in the
real world: A reanalysis of homicide rates following championship prize fights.
Aggressive Behavior, 17, 121-134.
Neiss, R. (1988f). Reconceptualizing arousal: Psychobiological states in motor perform¬
ance. Psychological Bulletin, 103, 345-366.
Nemeth, C. (1970). Bargaining and reciprocity. Psychological Bulletin, 74, 297-308.
Ogles, R. M. (1987). Cultivation analysis: Theory, methodology and current research on
television-influenced constructions of social reality. Mass Communication Review,
14, 43-53.
Padget, V. R., Brislin-Slutz, J., St Neal, J. A. (1989). Pornography, erotica, and attitudes
265
toward women: The effects of repeated exposure. Journal of Sex Research, 26,
479-491.
Paik, H., Si Comstock, G. (1994). The effects of television violence on antisocial behav¬
ior: A meta-analysis. Communication Research, 21, 516-545.
Palys, T. S. (1986). Testing the common wisdom: The social content of video pornogra¬
phy. Canadian Psychology, 27, 22-35.
Parke, R. D., Berkowitz, L., Leyens, J. P., West, S., St Sebastian, R. J. (1977). Some effects
of violent and nonviolent movies on the behavior of juvenile delinquents. In L.
Berkowitz (Ed.), Advances in experimental social psychology (vol. 10, pp. 135-172).
New York: Academic.
Phillips, D. P. (1983). The impact of mass media violence on U.S. homicides. American
Sociological Review, 48, 560-568.
Phillips, D. P. (1986). The found experiment: A new technique for assessing the impact
of mass media violence on real-world aggressive behavior. In G. Comstock (Ed.), Pub¬
lic communication and behavior (vol. 1, pp. 259-307). San Diego, CA: Academic.
Phillips, D. P., St Bollen, K. A. (1985). Same time last year: Selective data dredging for
unreliable findings. American Sociological Review, 50, 364-371.
Rapaport, K., St Burkhart, B. R. (1984). Personality and attitudinal characteristics of sex¬
ually coercive college males. Journal of Abnormal Psychology, 93, 216-221.
Reiss, I. L. (1986). Journey into sexuality: An exploratory voyage. Englewood Cliffs, NJ:
Prentice-Hall.
Rosenberg, M. f. (1969). The conditions and consequences of evaluation apprehension. In R.
Rosenthal Si R. Rosnow (Eds.), Artifacts in behavioral research. New York: Academic.
Rule, B. G., St Ferguson, T. J. (1986). The effects of media violence on attitudes, emotions,
and cognitions. Journal of Social Issues, 42, 29-50.
Sapolsky, B. S. (1984). Arousal, affect, and the aggression-moderating effect of erotica. In
N. M. Malamuth St E. Donnerstein (Eds.), Pornography and sexual aggression (pp.
83-115). New York: Academic.
Sheehan, P. W. (1986). Television viewing and its relation to aggression among children
in Australia. In L. R. Huesmann St L. D. Eron (Eds.), Television and the aggressive
child: A cross-national comparison (pp. 161-200). Hillsdale, NJ: Erlbaum.
Smith, D. D. (1976). The social content of pornography. Journal of Communication, 29,
16-24.
Straus, N. (1991). Discipline and divorce: Physical punishment of children and violence
and other crime in adulthood. Social Problems, 38, 133-154.
Surgeon General's Scientific Advisory Committee on Television and Social Behavior.
(1972). Television and growing up: The impact of televised violence. Report to the
Surgeon General, U.S. Public Health Service. HEP Publ. No. HSM 72-9090.
Rockville, MD: National Institute of Mental Health, U.S. Government Printing
Office.
Sykes, G., St Matza, D. (1961). Juvenile delinquency and subterranean values. American
Sociological Review, 26, 712-719.
Tedeschi. J. T., St Felson, R. B. (1994). Violence, aggression, and coercive actions. Wash¬
ington, DC: American Psychological Association.
Tedeschi, J. T., St Norman, N. (1985). Social mechanisms of displaced aggression. In E. J.
Lawler (Ed.), Advances in group processes: Theory and research (vol. 2). Greenwich,
CT: JAI.
Tedeschi, J. T., Smith, R. B. HI, Si Brown, R. C. Jr. (1974). A reinterpretation of research
on aggression. Psychological Bulletin, 89, 540-563.
Thomas, M. H., Horton, R. W., Lippincott, E. Q, St Drabman, R. S. (1977). Desensitiza¬
tion to portrayals of real-life aggression as a function of exposure to television vio¬
lence. Journal of Personality and Social Psychology, 35, 450-458.
266
Richard B. Felson
Valkenburg, P. M., Vooijs, M. W., & Van der Voort, T. H. (1992). The influence of televi¬
sion on children's fantasy styles: A secondary analysis. Imagination, Cognition, Per¬
sonality, 12, 55-67.
West, D. J., & Farrington, D. P. (1977). The delinquent way of life. London: Heinemann.
White, L. A. (1979). Erotica and aggression: The influence of sexual arousal, positive
affect, and negative affect on aggressive behavior. Journal of Personality and Social
Psychology, 37, 591-601.
Wiegman, O., Kuttschreuter, M., & Baarda, B. (1992). A longitudinal study of the effects
of television viewing on aggressive and antisocial behaviors. British Journal of Social
Psychology, 31, 147-164.
Wood, W., Wong. F. Y., & Chachere, J. G. (1991). Effects of media violence on viewers'
aggression in unconstrained social interaction. Psychological Bulletin, 109, 371-383.
Zillman, D. (1971). Excitation transfer in communication-mediated aggressive behavior.
Journal of Experimental and Social Psychology, 7, 419-434.
Zillman, D. (1983). Arousal and aggression. In R. G. Geen & E. I. Donnerstein (Eds.),
Aggression: Theoretical and empirical reviews (vol. 1, pp. 75-101). New York:
Academic.
Zillmann, D., Bryant, J., Comisky, P. W., & Medoff, N. J. (1981). Excitation and hedonic
valence in the effect of erotica on motivated intermale aggression. European Jour¬
nal of Social Psychology, 11, 233-252.
Zillmann, D., Floyt, J. L., & Day, K. D. (1974). Strength and duration of the effect of aggres¬
sive, violent, and erotic communications on subsequent aggressive behavior. Com¬
munication Research, 1, 286-306.
Contributors
John Bailey is a Hollywood cinematographer whose work includes

The Out-of-Towners, As Good As It Gets, and In the Line of Fire, and,
as director, China Moon.
Leonard Berkowitz is Professor Emeritus of Psychology at the Uni¬

versity of Wisconsin-Madison and has published extensively on mass
media effects.
Carol J. Clover is Professor of Rhetoric and Scandinavian at the Uni¬

versity of California-Berkeley. She is author of Men, Women and
Chainsaws.
Bosley Crowther was the film critic for The New York Times.
Richard B. Felson is Professor of Sociology at University of Albany

(SUNY), where his research has focused on the social psychology of
violence.
Ronald Gold covered the Hollywood industry for Variety, the trade
paper for the motion picture business.
Devin McKinney is an independent writer-researcher whose work on

film has appeared in Film Quarterly.
Joseph Morgenstern was the film critic for Newsweek.
Stephen Prince is Associate Professor of Communication Studies at

Virginia Tech. His books include A New Pot of Gold: Hollywood Under
the Electronic Rainbow, 1980-1989, Savage Cinema: Sam Peckinpah
and the Rise of Ultraviolent Movies, and The Warrior’s Camera: The
Cinema of Akira Kurosawa.
267
268
Contributors
Vivian C. Sobchack is Professor and Associate Dean of Film and Tele¬

vision at the University of California-Los Angeles. Her books include
Meta-Morphing: Visual Transformation and the Culture of Quick-
Change, The Persistence of History: Cinema, Television and the Modern
Effect, and Address of the Eye: A Phenomenology of Film Experience.
David Thomson is the author of Beneath Mulholland: Thoughts on

Hollywood and Its Ghosts, Rosebud: The Story of Orson Welles, and A
Biographical Dictionary of Film.
Jack Valenti is President of the Motion Picture Association of America.

Index
Accident, 61 Battle of the Bulge, 48

aestheticized violence, 27-34 Beatty, Warren, 11, 94
Affliction, 124 Behind the Green Door, 163
Aldrich, Robert, 52, 57, 60 Beloved, 124
Alien, 125, 131, 148, 154 Bergman, Ingmar, 157, 158
Aliens, 148, 154, 159 Berkowitz, Leonard, 20; cognitive
All of Me, 163 approach of, 23-24, 39-40,
Allen, Dede, 11 206-210
Allen, Woody, 157 Big Chill, The, 37
Almendros, Nestor, 80 Birds, The, 139, 152, 155
Almodovar, Pedro, 103 Bitter Victory, 97
America, 58 Blade Runner, 125
American Psycho, 102 Blow Out, 143
American Werewolf in London, An, Blow-Up, 58
125 Blurner, Herbert, 18; "emotional
Amityville Horror, The, 138 possession," concept of, 18
Angels with Dirty Faces, 89 bodily mutilation, 14-17
Argento, Dario, 150 Body Double, 1, 127, 138
Aristotle, 19-20 body genres, 127-129
Artaud, Antonin, 83 Bogart, Humphrey, 37, 90
Attack of the 50-Foot Woman, 137 Bonnie and Clyde, 7, 9, 11, 13, 35,
Attracta, 97 36, 38, 57, 114, 122, 124, 178;
Aubade, 97 critical ambivalence toward,
Autant-Lara, Claude, 55 47-50, 60-61; editing of, 11-12,
39, 183-184,185; viewer reac¬
Bad Lieutenant, 38, 101 tions to, 24-25, 27, 72-74
Bailey, John, 33, 37, 38 Borgnine, Ernest, 182
Baker, Rick, 15 Bottin, Rob, 15
Bakhtin, Mikhail, 122 Bresson, Robert, 83
Ballad of Cable Hogue, The, 187 Briggs, Joe Bob, 126, 166
Bambi, 110, 111 - Bring Me the Head of Alfredo
Bandura, Albert, 21-22, 247, 252; Garcia, 179, 188, 190-191, 192
social learning theory of, 21, Britt, Donna, 31
212, 237, 252 Broadcasting Standards Commission
Barbarella, 7 (United Kingdom), 28
Basic Instinct, 1, 38, 104, 107-108 Brook, Peter, 83
269
Brown, Rita Mae, 144 Cruising, 141, 156, 162
Buckets of Blood, 134 Crying Game, The, 38, 100-101
Bugs Bunny Movie, The, 23
Bunuel, Luis, 94, 115 D.O.A., 89
Burton, Richard, 97 Dali, Salvador, 115
Dance, Charles, 84
Caan, James, 188 Dantine, Helmut, 188
Cagney, James, 4, 5, 37, 48, 89-90 Darabont, Frank, 83
Cape Fear, 94 Dawn of the Dead, 15
Carlson, Roy, 85 Dawson, Gordon, 184
Carpenter, John, 130, 141, 147, 155, Days of Heaven, 80
156 De Palma, Brian, 2, 5, 8, 125, 127,
Casino, 1 141, 150, 157, 159, 166
Casualties of War, 100 Deakins, Roger, 83
Catch-22, 113 Death Wish, 92
catharsis, concept of, 19-20; work of DeBord, Guy, 108
Seymour Feshback and, 21 Deer Hunter, The, 23; viewer
Champion, The, 246-247 responses to, 24
Chaplin, Charles, 47 Deliverance, 153
Chase, The, 48 Demon, 154
Child’s Voice, A, 97 Detective, The, 7
China Moon, 37, 84-85 Dickinson, Angie, 91
Chinatown, 87 Die Hard, 92
Christ, Judith, 60 Dirty Dozen, The, 9, 12, 35, 36, 48,
Christian Century, The, 197 57; critical ambivalence toward,
Clarke, Alan, 88 51-53, 54, 55-56, 58-60
Classification and Rating Adminis¬ Douglas, Kirk, 91, 246, 247
tration (CARA), 6, 33 Dracula, 125
Clockwork Orange, A, 38, 114 Dressed to Kill, 8, 135, 137, 141, 142,
Clover, Carol, 15, 16, 39 162
Coburn, James, 96 Dreyer, Carl Theodor, 125
Co-Ed Frenzy, 143 Dunaway, Faye, 11
Cook, Jr., Elisha, 89 Dutchman, 48
Cook, the Thief, His Wife and Her Duvall, Robert, 188
Lover, The, 103
Cooper, Gary, 90 Eastwood, Clint, 37, 52, 53, 90-91
Coppola, Francis Ford, 88 Ebert, Roger, 16
Costner, Kevin, 95 effects on viewers. See empirical
Craven, Wes, 17 research on media effects
Cries and Whispers, 158 Eisenstein, Sergei, 107, 178, 182-183,
Criminal Conversation, 97 187
Crimson Pirate, The, 48 El Dorado, 11
Cronenberg, David, 103, 159 Elephant, 88
Cross of Iron, 183, 185, 188, 192 Ellis, Bret Easton, 102
Crowther, Bosley, 9, 35-36; criticisms empirical research on media effects,
of movie violence by, 51-56; reac¬ 19-29; consequences of depicted
tions to Crowther, 57-61 aggression, 224-226; cultivation
271
Index
thesis, 215-217; demand cues, Frears, Stephen, 97

239-240; desensitization, 220, French Connection, The, 110, 114,
230, 255; discrediting of catharsis, 115, 118, 124
19-20; disinhibition, 218-220; Freud, Sigmund, 127, 130, 154, 161,
field experiments, 240-241; iden¬ 162
tification with aggressor, Friday the 13th, 15, 129, 137, 142,
223-224; inciting action, 144, 152, 156, 162
217-218; influence of thoughts Friday the 13th Part 2, 137, 147
and ideas on aggressive behavior, Friday the 13th Part III, 140
207-236, 246-247; laboratory Funhouse, The, 15
experiments, 239-240; longitudi¬ Fury, The, 1
nal surveys, 243-246; media-
inspired violence, 21, 24, 31-32, Gerbner, George, 215-217
34, 72-74, 205-206, 242-243; Ghost, 94, 95
natural experiments, 214-243, God Told Me To, 154
nature of available target, Godfather, The, 87-88, 92-93, 114
226- 227; pornography, 247-249, Godfather Part II, The, 91
255-258; priming effects, Gold, Ronald, 36
206-220; prosocial influences, Gone With the Wind, 126
210-211; reality vs. fiction, Good, the Bad, and the Ugly, The, 9
227- 229; sponsor effects, Goodfellas, 103
249-251; vicarious reinforcement, Granger, Stewart, 111
253-254 Great Train Robbery, The, 2
ENR, 29 Greenaway, Peter, 103
Eszterhas, Joe, 104 Griffith, D. W., 2
Every Woman Has a Fantasy, 163
Exorcist, The, 125 Halloween, 94, 130, 131, 132, 136,
Exposure, 97 140, 141, 142, 162; Final Girl in,
Eyes of a Stranger, 15 144, 147; John Carpenter on,
Eyes of Laura Mars, The, 135-136, 155; killer in, 137; plot of, 133
137, 142, 156 Halloween II, 139, 141, 142
Hard Boiled, 14
Faithful Departed, 97 Harris, Ed, 84, 85
Feldman, Phil, 12, 25, 183 Haworth, Ted, 183
Felson, Richard B., 20, 39, 40 He Knows You’re Alone, 138, 141,
Ferrara, Abel, 101 142, 155
Feshback, Seymour, 21 He Was Her Man, 89
Field of Dreams, 95 Heaven Can Wait, 94
Fistful of Dollars, A, 9, 35, 48, 52, Hell Night, 136, 138, 139, 141, 142,
Flynn, Errol, 111 144, 166; Final Girl in, 147
Fog, The, 130, 147 Heller, Joseph, 100
For a Few Dollars More, 9, 35, 48, 51, Henry: Portrait of a Serial Killer, 1,
52, 54 38, 102
For Your Eyes Only, 23 Hickey, Kieran, 97, 98
Fox, The, 1 Hickock, Richard, 113
Frankenstein, 125 High Sierra, 90
Fraser, John, 103 Higson, Charles, 26-27, 198
Hill, The, 48 Kurosawa, Akira, 10; montage-slow
Hills Have Eyes, The, 131 motion techniques of, 10-12,
Hitchcock, Alfred, 107, 125, 130, 178-183, 184-185, 186
131, 134, 139, 148, 150, 152, 159,
164, 166 L.A. Confidential, 14
Hoffman, Dustin, 163 La Femme Nikita, 103
Holden, William, 91 Larkin, Philip, 97
Hooper, Tobe, 126 Left-handed Gun, The, 11, 184, 185,
horror films: controversial nature of, 186
15-17; make-up effects in, 15. Leigh, Janet, 140
See also slasher films Leone, Sergio, 9, 52; Dollars trilogy
Howling, The, 15 of, 9, 12
Hughes, Robert, 55 Lethal Weapon, 92
Hustler’s Erotic Film Guide, 128 Little Caesar, 3, 89
Lloyd, Christopher, 93
I Spit on Your Grave, 137, 139-140, Lolita, 88, 91
158 Lombardo, Lou, 175, 178; editing of
In Cold Blood, 7, 48 Wild Bunch, 192, 193, 198
In the Line of Fire, 37, 90-91, 92 Losey, Joseph, 61
Intolerance, 2
It’s Alive, 15, 136 Macabre, 128
Madsen, Michael, 106
Jackson, Samuel L., 120 Magnificent Amber sons, The, 96
Jaws, 139, 152 Major Dundee, 187
JFK, 100 Malkovich, John, 91, 92
Joanou, Phil, 104 Man Escaped, A, 83
Jones, Amy, 144 Mandingo, 105
Jones, L. Q., 96 Maniac, 15,
Jordan, Neil, 100 Manson, Charles, 113
Joslin, Tom, 96 Marcus, Steven, 159-160, 161
Jules and Jim, 83 Marcuse, Herbert, 177
Junior Bonner, 179 Marvin, Lee, 51, 91
Juno, Andrea, 168 Mary Poppins, 12
Jurassic Park, 95 Mason, James, 91
Massi, Mark, 96
Kaminski, Janusz, 29 Matter of Life and Death, A, 94
Kaplan, E. Ann, 151 McCauley, R. C., 28, 29
Keitel, Harvey, 101, 105 McCrea, Joel, 111
Kennedy, John, 113 McKenna, T. P., 187
Kennedy, Robert, 113 McKinney, Devin, 38
Killer Elite, The, 188, 192 McNaughton, John, 101
Killer, The, 14, media effects. See empirical research
Killers, The, 91 on media effects
King Kong, 125 media-inspired violence, 21, 24,
King, Jr., Martin Luther, i 13 31-32, 34, 72-74, 205-206,
Knight, Arthur, 198 242-243. See also empirical
Kriemhild’s Revenge, 88 research on media effects
273
Index
Men, Women and Chainsaws, 16

Miller, Mark Crispin, 19 Pacino, Al, 5
Missouri Breaks, The, 95 Palance, Jack, 89
Modleski, Tania, 168 Parker, Alan, 103
Money Train, 21 Pat Garrett and Billy the Kid, 96,
montage-slow motion aesthetic, 39 184, 187
Morgenstern, Joseph, 35, 198 Pawnbroker, The, 58
Motel Hell, 135, 136 Payback, 122
Mother’s Day, 137 Payne Fund Studies, 4, 18
Motion Picture Association of Peckinpah, Sam, 12, 27, 28, 29, 35,
America (MPAA), 6, 12, 36, 62; 39, 93, 120, 175-200 passing-
advocacy of new freedoms, 7-9; antecedents of his style,
G-M-R-X rating system, 7, 37, 178-187; reasons for using
66-67 graphic violence, 176-178;
Movies and Conduct, 18; "emotional on use of slow motion, 186;
possession," concept of, 18 slow motion inserts, 187-192;
Ms. 45, 137 montages with slow motion,
Muni, Paul, 4 192-196
Murnau, F. W., 125 Penn, Arthur, 8, 10-12, 35, 48, 57,
Murphy, Audie, 111 60, 120, 178, 180; use of mon¬
Museum of Modern Art, The, 126 tage-slow motion, 183-185
My Secret Life, 158 Penn-Peckinpah stylistic, the, 10-14;
antecedents of, 178-183
Naked Lunch, 103 Pickens, Slim, 96
Naked Runner, The, 56 Pink Flamingos, 163
Nation, The, 197 Piranha, 15
National Catholic Office for Motion Play Misty for Me, 137
Pictures, 58, 73 Playboy, 116
National Commission on the Causes Poe, Edgar Allen, 150
and Prevention of Violence, 7, Poetics, The, 19
19, 24, 36 Point Blank, 7
Natural Born Killers, 33, 37, 81-83, pornography, 127, 128, 129, 150, 157,
121 159-160, 163, 247-249, 255-258
Newman, Paul, 184 Porter, Edwin S., 2
Newsweek, 198 Potemkin, 183
Night of the Hunter, The, 92 Power, Tyrone, 111
Nightmare on Elm Street, A, 131, Price, Vincent, 111
136; Final Girl in, 145-146 Privilege, 55
Nobody’s Fool, 37 Production Code Administration, 3
North by Northwest, 91 Production Code, 2-6, 17; revision
Nosferatu, 94 of, 6-8, 37, 64-67, 69-70; Sug¬
gested for Mature Audiences
O'Brien, Edmond, 89 (SMA) designation, 6-7
Odd Man Out, 91 Professionals, The, 11
Omen, The, 1, 125 Psycho, 48, 92, 107, 130, 131, 156,
One False Move, 100 158, 159, 162, 164; Final Girl in;
Osterman Weekend, The, 188, 192 146-147; influence on slasher
Psycho (continued) Seydor, Paul, 181
films, 132, 134-135, 137, Shane, 11, 89
147-148, 165 Shawshank Redemption, The, 37,
Psycho II, 162 81-83
Public Enemy, The, 3-4, 48, 89 Shimura, Takashi, 181
Pulp Fiction, 33, 120, 121, 124 Shining, The, 125, 131
Pyle, Denver, 184 Showalter, Elaine, 163
Siegel, Don, 91
Quiet Man, The, 48 - Silence of the Lambs, The, 92
Silverlake Life, 96-97
Rashomon, 178 Silverman, Raja, 167
Re-Animator, 1 Sinatra, Frank, 56
Renoir, Jean, 96 Sirk, Douglas, 108
Reservoir Dogs, 38, 104-106, 121 Siskel, Gene', 16
Richardson, Robert, 82 slasher films: Final Girl in, 143-149;
Ride the High Country, 176, 180 gender roles in, 150-168; plots
River, The, 96 of, 132-135; killers in, 135-138;
Roaring Twenties, The, 90 locale of, 138-139; point of view
Robbins, Tim, 82 of, 150-168; shock effects of,
Robinson, Edward G., 4, 89 149; weapons in, 139-140; vic¬
Rocky Horror Picture Show, The, tims in, 140-143
163 slow motion, 184-196
Rocky IV, 22 Slumber Party Massacre, 136, 144;
Rosemary’s Baby, 125, 138 Final Girl in, 146
Roth, Tim, 105 Smith, Perry, 113
Snapper, The, 97
Sanshiro Sugata, 180 Sobchack, Vivian C., 38-39, 100
Saturday Review, 198 social learning theory, 21, 212, 237,
Savage Cinema, 187, 188 252
Saving Private Ryan, 29, 120, 124; Sontag, Susan, 177
aestheticization of violence in, Spartacus, 88
29 Speck, Richard, 113
Savini, Tom, 15 Spielberg, Steven, 29, 30, 32, 95, 120,
Scarf ace (1932), 4 152
Scarf ace (1984), 5-6 squibs, 10-11
Schindler’s List, viewer reactions to, Squirm, 15
30-31 St. Valentine’s Day Massacre, 48, 49,
Schoell, William, 126, 140, 153 56
Scorsese, Martin, 1, 27, 29, 94; on Stanford Research Institute, 130
audience reaction to Taxi Driver, Stamford, Paul, 176, 197
198-199 Star Wars, 22
Scott, Randolph, 111 Starship Troopers, 1
Scream, 17, 121 State of Grace, 104
Sergeant York, 88 Stay Out of the Shower, 126
Seven Samurai, 10, 179, 180-82, Stoller, Robert, 163
184-185 Stone, Oliver, 33, 82, 121 •
Sex and Gender, 163 Strait-jacket, 137
275
Index
Straw Dogs, 38, 115-116, 124, 179, Wallace, George, 113

187, 188, 190, 191, 192 Walsh, Moira, 58, 60
Sunset Boulevard, 91, 92 Walsh, Raoul, 90
Watkins, Peter, 55
Tarantino, Quentin, 104, 105, 121 Wayne, John, 88, 90, 111
Taxi Driver, 1, 13, 29, 31, 99; viewer Weaver, James, 23
reactions to, 27, 198-199 Weddle, David, 180-181, 191
Terminator, The, 92 Wenders, Wim, 33
Texas Chainsaw Massacre, The, 93, West Side Story, 48
126, 131, 135,136, 140, 142, 155; Whatever Happened to Baby Jane,
Final Girl in, 143-144; killer in, 125
137; plot of, 132-133; locale in, White Heat, 4-6, 89, 90
138 Whitman, Charles, 113
Texas Chainsaw Massacre, Part 2, Who Framed Roger Rabbit, 93
The, 134, 135, 136, 137, 138, Who’s Afraid of Virginia Woolf1, 58
153, 156, 166; Final Girl in, 146, Wild at Heart, 1, 100
147; locale in, 138, 139 Wild Bunch, The, 12, 19, 39, 88, 93,
The Rockingham Shoot, 91 114, 122, 175, 176, 180, 181, 182,
Thing, The, 15 185, 187, 188, 189; Bonnie and
Thomson, David, 37-38 Clyde influence on, 183-185;
Tom and Jerry, 88 editing of, 192-198; Kurosawa
Tootsie, 163-164 influence on, 178-183; montage
Torn Curtain, 107 in, 192-198, 199; viewer reac¬
Towne, Robert, 87 tions to, 13, 25-27, 178, 197-198
Truffaut, Francois, 83 Williams, Linda, 100, 102, 154
Winston, Archer, 60
ultraviolence, genesis of, 6-19 Winters, Shelly, 92
Un Chien Andalou, 115, 116, 124 Wizard of Oz, The, 93
Unforgiven, 90 Wolf, William, 60
Up the Down Staircase, 48 Wolfen, 125
Woo, John, 14
Vale, V., 168 Wood, Robin, 99, 126, 130, 168
Valenti, Jack, 6, 7, 8-9, 19, 25;
defense of Hollywood industry, X, Malcolm, 113
36-37, 62-75
Vaughn, Peter, 187 Zemeckis, Robert, 93
Verhoeven, Paul, 1, 38, 103-104, 107 Zillmann, Dolf, 22, 23, 248; "counter¬
Videodrome, 1, 159 empathy," concept of, 22
Vietnam war, 8, 36, 119, 176
Violence in the Arts, 103

Screening Violence

Uploaded by

Copyright:

Available Formats

Screening Violence

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Screening Violence

Uploaded by

Copyright:

Available Formats

Screening Violence

Rutgers Depth of Field Series

Charles Affron, Mirella Jona Ajjron, Robert Lyons, Series Editors

Richard Abel, ed., Silent Film

Screening violence / edited and with an introduction by Stephen Prince,

This collection copyright © 2000 by Rutgers, The State University

Manufactured in the United States of America

Graphic Violence in the Cinema: Origins,

The Historical Context of Ultraviolence

The Thin Red Line 47

Movies to Kill People By 51

Another Smash at Violence 54

Crowther’s ‘Bonnie’-Brook: Rap at Violence Stirs

Statement by Jack Valenti, MPAA President,

The Aesthetics of Ultraviolence

Bang Bang Bang Bang, Ad Naseum 79

Death and Its Details 86

Violence: The Strong and the Weak 99

The Violent Dance: A Personal Memoir

Her Body, Himself: Gender in the Slasher Film 125

The Aesthetic of Slow-Motion Violence

The Effects of Ultraviolence

Some Effects of Thoughts on Anti- and

Mass Media Effects on Violent Behavior 237

Graphic Violence in the

Graphic violence is an inescapable and ubiquitous characteristic of con¬

Copyright © 1999 by Stephen Prince

pessimistic about the psychological health promoted in viewers by

Enemy (1931), and especially Scarface (1932) showed cold-blooded mur¬

(1949), a government sniper repeatedly shoots Cody Jarrett (the film's

Heat and Scarface, the difference between concealing and displaying

The Genesis of Ultraviolence

film as bloodthirsty trash in initial reviews, only to issue retractions in

The perfection of special effects has made it possible to show maim¬

The genre's remarkable resurgence in the 1980s was tied to this

boom. Distributors and studio sales reps expressed revulsion at these

compelled a reconsideration of the medium's vividness and emotive

The Empirical Evidence on the Effects of Viewing Violence

Empirical investigations of media violence have confronted and dis¬

acting as the vehicles effecting the cathartic purge. He described

an empirical standpoint, the catharsis hypothesis is primarily associ¬

to shape the form of a viewer's subsequent antisocial behavior. Children's

part of those righteously dispatched. Of importance in this regard, view¬

This suggests that suffering of victims is a major constituent of per¬

In assessing these consequences of screen violence for the view¬

period."53 Berkowitz maintains that a high level of aggression-related

Aestheticized Violence and Volatile Responses

As these examples indicate, viewers respond to graphic violence in

I was exhausted and in a state of high nervous excitement. I wanted to

was reacting very strongly to the shoot-out sequence in Taxi Driver.

Viewers who whoop with approval at ultraviolence are often in¬

there is much about this phenomenon—why viewers seek emotional

Scorsese, both consummate filmmakers, wished to arouse aggressive,

were shooting Jewish victims in Krakow, Poland.75 The 45-year-old gun¬

On the one hand, the policy implications of these considera¬

a serious and provocative way that invites reflection and contemplation

They evoke the controversies that followed the explosion of graphic

justified the presentation of explicit movie violence, Crowther argued

Graphic Violence in the Cinema

were flawed in design and conception. Thus, he claimed, social scien¬