EMOTION AND THE STRUCTURE

OF NARRATIVE FILM
Film as an Emotion Machine
EMOTION AND THE STRUCTURE
OF NARRATIVE FILM
Film as an Emotion Machine

Ed S. Tan
Vrije Universiteit Amsterdam
Utrecht University

Translated by Barbara Fasting

The translation of this book from Dutch into English

has been made possible by a grant from
The Netherlands Organization for Scientific Research.

R Routledge
Taylor &. Francis Group
NEW YORK AND LONDON
First Published by
Lawrence Erlbaum Associates, Inc., Publishers
10 Industrial Avenue
Mahwah, New Jersey 07430

Routledge
711 Third Avenue, New York, NY 10017
2 Park Square, Milton Park, Abingdon, Oxfordshire OX 14 4RN

Copyright © 1996 by Lawrence Erlbaum Associates, Inc. All rights

reserved. No part of this book may be reproduced in any form, by
photostat, microform, retrieval system, or any other means, with-
out the prior written permission of the publisher.

First issued in paperback 2011

Cover design by Cheryl Minden

Library of Congress Cataloging-in-Publication Data

Tan, Ed S.
Emotion and the structure of narrative film : film as an
emotion machine / Ed S. Tan.
p. cm.
Includes bibliographical references and index.

1. Motion pictures—Psychological aspects. I. Title.

PN1995.T26 1996
791.43'01'9—dc20 95-6454
CIP

Publisher's Note
The publisher has gone to great lengths to ensure the quality of this reprint
but points out that some imperfections in the original may be apparent.

ISBN 978-0-805-81409-5 (hbk)

ISBN 978-0-415-51560-3 (pbk)
ISBN 978-1-136-69497-4 (ebk)
For Anyke
 Contents

Preface ix

1 Introduction 1

2 The Psychological Functions of Film Viewing 15

3 Film and Emotion: Theoretical Background 41

4 The Structure of Interest 85

5 Thematic Structures and Interest 121

6 Character Structures, Empathy, and Interest 153

7 The Psychological Affect Structure of the Feature Film 195

8 Conclusion: The Feature Film as an Emotion Machine 225

References 253

Author Index 275

Subject Index 283

vii
 Preface

This book is meant to be a contribution to the psychology of film. Books on

the psychology of film may be rare, but psychology and film share a lot. They
have both recently celebrated their 100th birthday, and it may well be that
they both originated in the technology of the late 19th century, or at least in
the social spirit of this technology. On the one hand, film has lent service to
psychology. It soon became an important tool in the psychological laboratory,
and it is still widely in use (e.g., in the study of perception and emotion). On
the other hand, psychology has contributed to theoretical reflection on film.
The 1916 essay on the "photoplay" by Hugo Munsterberg, a Harvard psychol-
ogist, has become known as the first attempt to formulate a theory of film.
However early this attempt may be, subsequent psychological studies of the
film have remained scarce. As recently as in 1979, James Gibson observed,
"The technology of cinema and television has reached the very highest level
of applied science. The psychology of the awareness provided by a motion
picture, however, is nonexistent. . . " (p. 292). Only a few years later, Hoch-
berg's (1986) brilliant contribution to Boff, Kaufman, and Thomas' Hand-
book of Perception was published. Virtually all knowledge that existed on the
perception of motion pictures had been accumulated in the chapter. The au-
thor complained that this is not really much: "In my first draft of this chap-
ter, I included titles for 27 tables which . . . would have been valuable to de-
signers of moving pictures in film. . . . Those tables have not been given here
because their data, although obtainable and desirable, are as yet imaginary"
(p. 22-60).

ix
X PREFACE

In this book, I attempt to further our understanding of the emotional ex-

perience that films constitute for the viewer, rather than the mysteries of per-
ception of the moving image. When viewing any film, say, The Remains of
the Day, we do not only see solid bodies in motion, and understand that they
represent people, but also, and perhaps above all, we feel something for the
characters and are somehow moved by the sight of them. The perception and
comprehension of butler Stevens and Miss Kenton are part of our dear feel-
ings toward them, hopes that they will finally come together, frustration when
they do not, and resignation as we realize that the characters were children
of their time. The awareness provided by a motion picture is an emotional
one in the first place. In the face of this simple fact, the lack of psychologi-
cal research is rather astonishing. Despite the fact that film has left its first
centennial behind, and film-produced emotion is now part and parcel of
everyday life, hardly any knowledge is available about the what and how of emo-
tion in the film viewer, and hardly any data have been reported. To describe
what an emotion produced by film is, and to open strands for empirical re-
search into its determinants, is the aim of this book.
Film awareness is also a complex awareness, and this may be the reason
that psychological research has been vastly absent. How does a film viewer
extract relevant meaning from what is shown on screen? How can it be that
I am gripped by the enacted sorrow of an actress filmed at a staged set? Can
viewers have real emotions? What are "real" emotions? These are all issues
that are crucial for explaining the film experience. It will also be crucial to
have an understanding of what an emotion is. The psychology of emotion has
gone a long way since William James proposed his well-known theory, also
100 years ago. Current theories of emotion provide a basis for describing the
film experience, but they are by no means sufficient to do justice to its com-
plexity. However, in film theory, concepts have been developed that can be
put to use. Hochberg (1986) noted that perceptual processes and compre-
hension of even elementary sequences in the cinema cannot be explained
properly unless narrative structure is taken into account. We cannot under-
stand how the viewer grasps a cut between two shots, let alone a longer se-
quence of shots, without assuming that some mental structure is in opera-
tion, one that has been made active by the narration. At this point, film theory
takes over. Current film theory is concerned with the role of narrative in film
and the way narrative determines film style and film technology. Concepts of
narration in film are especially relevant for understanding emotion in the
viewer, because narrative is a way to arouse emotion. Not surprisingly, how
narrative succeeds in bringing about powerful and highly specific emotion
has been the subject of investigation for ages in the humanities. This book
tries to gather and join insights from the humanities and psychology, partic-
ularly from the theory of film narration and the theory of human emotion.
In integrating the two theoretical perspectives, I hope to show that an at first
PREFACE xi

sight abhorrently complex question, "What is film emotion?" can fruitfully

be tackled and decomposed into questions of somewhat more manageable
proportions.
Bringing together insights from two widely different fields yields useful re-
sults and the joyful discovery that what seems so different at first turns out
to be more similar than you had ever suspected. In explaining the viewer's
emotion, a number of deep correspondences between the theory of emotion
and that of film narration have emerged. The integration also has its costs, in
this case a near paradox, that time alone may solve. The study has resulted
in an all but entirely theoretical approach to a phenomenon that excels in the
immediacy of its manifestations. Nevertheless, many of the notions and ideas
have developed in field experiments, to which I have referred throughout this
book. As a most concrete film-psychological result, I propose the affect struc-
ture of the traditional film. The affect structure systematically exposes traces
of the operation of film as a machine, a technology for manufacturing emo-
tion. It is a hypothetical framework for gathering data that I have left, al-
though obtainable and desirable, as yet imaginary. I can only hope that they
will not remain so for another 100 years.

ACKNOWLEDGMENTS

I call myself lucky to have found companions in my expedition into such an

ill-charted area of psychological research. Many of the ideas in this book owe
their existence to collaboration with Nico Frijda. His erudition and keen per-
ception have been a delight to me during the years, while his support and en-
couragement have greatly inspired the work. Henry Schoenmakers intro-
duced me to the humanities; he has made possible psychological research
into film and the viewer and has contributed a lot of ideas in all stages. Emiel
van Moerkerken, a cinematographer and perception psychologist who has en-
joyed more films more intensely than anybody else, has been an unwearying
teacher over the years, until the end of his days. I hope that one of his many
lessons, that film artistry can be found in all genres, has found its way in
the book's argument. "The Club," Adri van der Wurff, Dick Ruimschotel, Jos
de Bruin, Marten den Uyl, and Wil Zeegers, had an indelible influence on
the study reported in this book. Wil also helped me greatly in preparing the
references section. Barbara Fasting has most skillfully translated the Dutch
manuscript into English, and I will remember with pleasure the profound dis-
cussions we had on some translation problems. All remaining flaws are due
to inconsistency and negligence on my part, and this goes not only for the
translation. Adri van Dijk took care of the figures in text, and assisted in
preparing the manuscript.
I am most indebted to The Netherlands Organization for Scientific Re-
xii PREFACE

search (NWO) for providing a grant that enabled translation of the book. Ad-
ditional support was received from the Faculties of Arts and Humanities of
the University of Amsterdam and Utrecht University. A number of institutes
have granted me time and other facilities to prepare the book. I owe special
thanks to the Department of Theater Studies of the University of Amsterdam,
The Institute for History and Culture of Utrecht University, the Institute for
Emotion and Motivation of the University of Amsterdam, and the Department
of Film and Television Studies of Amsterdam University.
I acknowledge Walter de Gruyter & Go. for permission to reproduce the
contents of Fig. 7.2, and Lawrence Erlbaum Associates for adaptation of the
contents of Fig. 5.1.
And finally, this book could never have been written without the ceaseless
approval of little Emiel, my son, and the endless patience and continuous sup-
port of Anyke, my wife. It is to her that I dedicate this book.
Ed S. Tan
 1 Introduction

DOES WATCHING A FILM PRODUCE EMOTION?

The story goes that during the showing of L'Arrivee d'un train en gave de la
Ciotat in 1895, people were so terrified at the sight of the oncoming loco-
motive that they tried to hide under their seats. Today's filmgoers are un-
doubtedly a good deal more hardened, but the cinema has itself evolved
considerably since the days of the Lumiere brothers. Thus even today any
cinema visitors who are in a position to observe their fellow film spectators
will see reactions that are not too different from those of the primal filmgo-
ers. People cover their faces, shrink back against their seats, and scream "Oh,
no! Not that!" The irrationality of such reactions is striking. It is, after all,
only a film. All the usual cliches present themselves: the plastic shark, the
tomato ketchup blood, the starstruck lovers played by two people who can-
not abide each another. Film is make-believe, and we know it.
If we fail to be intrigued by the apparent irrationality of these emotional
reactions, then we will at any rate be struck by their intensity. In Portrait de
Lillian Gish (1986), the aged actress recalls a showing of Birth of a Nation
(1915), and the violent emotions that seized the old men who had fought in
the Civil War: "Their sobs shook the seats." Today's cinema audiences, too,
are often surprised at the force of the emotion that grips them. Their surprise
is triggered by the fact that they know full well that what they are seeing is a
series of images projected onto a screen: in other words, an illusion.
There is another interesting side to the emotions evoked by a film, and that

1
2 CHAPTER 1

is the sheer diversity of those feelings. A cinema audience can be brought to

such a fever pitch of excitement that the people squirm in their seats, or even
call out, as did happen in the well-known film scene in Saboteur (1942). Or
they may respond in exactly the opposite way, by following breathlessly, in to-
tal silence, the events taking place on the screen. This is nicely illustrated by
a scene in 84 Charing Cross Road (1986) where Helen Hanff is sitting alone,
smoking a cigarette and watching Brief Encounter. Or film viewers may feel
a pang of guilt when they realize that they have wrongly suspected one of the
characters of treachery or betrayal. Or again, they may be grateful to an ac-
tor or actress for a fine performance or jealous of the fortune amassed by the
villain of the piece. Cinema audiences smile condescendingly, titter nervously,
or burst out laughing. One and the same film can produce tears of joy or tears
of frustration. And then there is that subtle emotion that is a combination of
contradictory feelings, like hope and fear, embarrassment and mirth, or pity
and gloating. But regardless of which particular registers of emotion are
played upon, no one can deny the universal popularity of films. Apparently
every conceivable human emotion, whether shunned or sought after in every-
day life, is capable of providing entertainment when experienced in the cin-
ema. And again we are struck by the suspicion that what filmgoers experi-
ence is a very special kind of reality. In any case, the fact that not only pleas-
ant sensations but also such normally unpleasant feelings as terror, helpless-
ness, and disgust are in a certain sense experienced as pleasurable leads us
to wonder what all the different feelings evoked by films have in common.
Obviously, filmgoers themselves are not likely to rack their brains over
such questions nor are they expected to do so. However, if a psychologist man-
aged to step out of his or her role of film spectator and revert to that of skep-
tical scientist, two fundamental questions would immediately spring to mind.
First, while viewers may themselves be convinced that what they feel is true
emotion, that feeling is not necessarily consistent with an objective definition
of the term. There is reason to question the authenticity of cinematic emo-
tion, because viewers know full well that what they are seeing is a fictional
world created by means of an artefact. According to some emotion theorists,
notably Sartre (1940), the feeling evoked by an artefact—he was referring
here to the image—can never give rise to a genuine emotion. 1 Is it possible
to be afraid of a film shark or terrified by Jack Nicholson? Sartre's view on
this point is not without its complications. First, we must ask ourselves to
what extent the reality, which is the source of any emotion, is itself subjec-
tive and dependent on the good will of the viewer.2 Thus we may choose to

^ h i s argument is supported by Peters (1989), with special reference to the cinema.

2
In his major work dealing with emotion (1939), Sartre was somewhat ambiguous about the
object to which the term reality applies; he appears to be referring to both the emotion and the
source of the emotion. He suggested that the intensity of the physical sensation accompanying
the emotion affects our assessment of the genuineness of the objects and situations involved.
INTRODUCTION 3

surrender to the film, accepting the idea that the shark is "real." A much sim-
pler criterion for assessing the authenticity of the feeling evoked by films is
the degree to which it resembles that experienced by the subject in the real
world outside the cinema. What relationship is there between my fear of the
shark and the fear I would feel if I actually found myself face to face with a
shark? What is the relationship between the specific feeling experienced by
the film viewer and "emotion" in the most general sense of the word? Is the
sensation experienced by the viewer an emotion? If so, to what extent is it a
particular kind of emotion? And in what respect does this emotion differ from
that which we experience in the reality of everyday life?
Second, the psychologist has to explain the systematics behind the feel-
ings that filmgoers experience. What mechanism produces the relatively or-
derly response to most feature films? On the basis of empirical observation,
it is safe to say that the response to films is fairly unanimous within a given
body of viewers. When the film is a comedy, most people laugh, although
some may do so more exuberantly than others. There is also something sys-
tematic about the timing of the response. A comedy, for example, produces
waves of laughter: most people laugh at roughly the same time, perhaps one
just before the others, another slightly longer. The same is true of the other
responses that we are inclined to see as emotional: apprehension, excitement,
relief, and so on. There seems to be a plan governing the course of our feel-
ing, and we know from sources other than the film itself that this is indeed
so. Films are designed to produce a particular effect and, as artefacts, they
display both a functional design and a certain consistency. That orderly struc-
ture and consistency are reflected in the systematics of the affective reactions
of the viewers, reactions that they themselves are not aware of. The second
goal of this study is to describe the essence of such reactions and their an-
tecedents and, in so doing, to take the initial steps toward explaining the ex-
perience of the film viewer.

BETWEEN PSYCHOLOGY AND FILM THEORY

The problem addressed by this study is important for both film theory and
psychology. To begin with the former, much of the material published in this
field focuses on qualities that are considered characteristic of this or that film
genre. Such research tends to localize the singularity of a particular genre in
the emotional effect that such films evoke in the viewer. The classic genre de-
scriptions, in literature and drama as well as in film, have traditionally been
rooted in the sort and the intensity of the emotion engendered. One need
only think of the dictionary definition of words like tragedy and comedy; in
fact, we all know the meaning of terms like thriller and tear-jerker without
even reaching for a dictionary. Indeed, the only publications focusing on the
4 CHAPTER 1

subject of the emotional reactions of viewers may well be found exclusively

among such genre studies. 3 Unlike these and similar studies, the aim of this
work is to contribute to a general description of the emotional potential of
films and the structural characteristics involved. In this way, it may contribute
to what, in the view of Bordwell (1989a), is seen by many as the crucial prob-
lem in cognitive film theory, namely, establishing a relationship between the
cognition of film and its emotional effect.
As a psychological study, this book aims to extend our insights into the
manner in which a film interacts with the human emotional system. If it can
be shown that films do indeed generate genuine emotions, then we may have
identified one of the major manifestations of human emotion. It may be that
the emotions evoked by films are related to those that accompany fantasies,
daydreams, and certain forms of play. It may also prove possible to shed some
light on the question of aesthetic emotion in a wider sense of the word.
It is the intention of this study to integrate film theory and psychology. If
we are to make any claims at all about what a viewer experiences, let alone
about the systematics governing that experience, we must first examine what
is already known about the structure of the stimulus, the film itself. This idea
is an elaboration of a number of more general views on the best way of gain-
ing information about complex information processing. On the basis of re-
search directed toward a description of expertise in specific areas—chess, for
instance, or solving physics problems—we know that a knowledge of the
structure of the surroundings in which psychological tasks are carried out is
often of crucial importance in understanding complex psychological
processes (Frijda, 1981; Simon, 1970). Gibson went a step further when he
formulated the principles of an ecological psychology that bridges the gap be-
tween stimulus and psychological process by viewing the two as a single sys-
tem (Gibson, 1979). The integration that will be developed in the course of
this study is characterized by the same general orientation. The rationale is
that the film, to the extent that it is seen as a narrative, systematically ma-
nipulates fictional situations and aspects of those situations in such a way that
they fulfill the requirements for the creation, maintenance, and modulation
of emotions. In short, to narrate is to produce emotion. The more the de-
scription of the process of narration and the description of the emotion felt
by the viewer intertwine, the more narration and emotion become a single
system.
The notion that film narration and viewer emotion are actually two sides
3
If we confine ourselves to the discipline of film theory, then the titles speak volumes. Films
and Feelings by Durgnat (1967) is a series of essays on the popular genre film. A work by Affron
(1982), with the somewhat more sophisticated title Cinema and Sentiment, deals with the tear-
jerker and a partially overlapping genre, the melodrama. The Philosophy of Horror (Carroll,
1990) is a milestone in the investigation of viewer affect because its import reaches far beyond
the genre.
INTRODUCTION 5

of the same coin provides us with a hypothetical answer to both parts of our
central question. First, it assumes that films do evoke emotions, and second,
it presupposes that film narration is responsible for a characteristic system-
atics in the emotion felt by the viewer. In the course of the book, this work-
ing hypothesis will be elaborated and provided with a theoretical basis.

The Theory of Film Narration

Like any theory, film theory must be based on usable observations. In his
book Making Meaning (1989b), Bordwell argued that although considerable
material has been collected on films, this material consists of a hodgepodge
of observations that are of doubtful utility if what one is interested in is a the-
ory dealing with a particular aspect of the film. Reviews, essays, polemic
pieces, analytical applications of fragments of theory, and assorted other con-
tributions to the publicistic enterprise that has formed around the film are
all useless as observations precisely because they fail to describe the most di-
rectly observable manifestations. A perusal of two studies by the methodol-
ogist Carroll (1988a, 1988b) leads to a similar conclusion concerning the
usefulness of existing film theory. In Carroll's view, the classic texts on film
theory emphasize the specific characteristics of the medium, while more or
less disregarding its use and function (Carroll, 1988a). For this reason such
contributions are not particularly helpful in explaining cinematic phenom-
ena. Modern theory, on the other hand, which according to Carroll begins
with the dawn of film semiology, suffers from an excessively broad perspec-
tive and a plethora of widely varying phenomena that, for the sake of conve-
nience, have been lumped together. Moreover, the vehicle of the metaphor,
which is commonly used to explain phenomena, is totally unsuitable for that
purpose. 4 I would, however, suggest that some caution should be exercised
with respect to the conclusion that in order to describe the main elements of
film we are obliged to start more or less (Bordwell) or even entirely (Carroll)
from scratch. I agree with Carroll that the classic film theories contain a great
many valuable elements and characteristics that are typical of the film
medium. The work of Munsterberg (1915/1980) and Arnheim (1933), as well
as the studies of the predominantly French filmologues who wrote for the Re-
vue Internationale de Filmologie immediately after the Second World War,
contain valuable observations—not least for the film psychologist—by the
very fact that they refer to the specifics of the medium. Modern film theory
is based on the notion that certain characteristics of the fiction film are
shared with other media, 5 while a number of those characteristics have long

4
Cf. "the fallacy of the 'indigent metaphor'," Carroll (1988b, p. 228).
5
For important discussions on the partial specificity of the film, see Cohen-Seat (1958) and
Metz (1971).
6 CHAPTER 1

been the object of observation well beyond the confines of the cinema. The
theoretical notions pertaining to the film as narrative have long had their par-
allels in the fields of literature and drama research. Attempts to clarify the
meaning and affects of film can benefit greatly from the observations that
have been collected in these fields.
For purposes of this introduction, I confine myself to a tentative descrip-
tion of the concept of narration as the process by which fictional events are
presented in an ordered and temporally structured manner, thereby produc-
ing a certain effect upon the listener (in our case, the viewer). The structur-
ing by means of narration goes deeper than a simple ordering of events. The
verbal narrative employs highly concrete vehicles of language, including syn-
tax, vocabulary, and figures of speech. The film as narrative similarly manipu-
lates individual and easily observable characteristics of the medium in order
to convey subtle meanings and produce highly specific effects. Film technol-
ogy, including the acting, directing, and camera work, serve to present fic-
tional events in such a way that they produce the intended effect on the
viewer.
The work of David Bordwell (1985) on the film narrative is indispensable
to this study. While it would take us too far afield to examine in detail all of
his theoretical concepts, it is important to note three aspects of Bordwell's
approach. First, the theory is comprehensive. The narrative does, of course,
represent only one—highly abstract—point of view from which the film can
be described. However, in Bordwell's approach, even the most directly ob-
servable phenomena of the surface structure, that is, technical and stylistic
features of individual films, can be understood within the framework of the
theory of film narrative. Second, Bordwell's theory comprises a body of films
that may be seen as the dominant genre, the traditional, full-length feature
film. This implies, furthermore, that the theory encompasses at least some
phenomena that are also characteristic of a much older narrative tradition,
such as Aristotelian drama and the traditional folk tale or legend, which means
that it is based on a greater wealth of observations. Third, Bordwell based his
theory on the activity of the viewer. And because he described this activity
largely in terms of existing cognitive-psychological insights concerning the
processing of information, the link with a cognitive-psychological theory of
emotion would appear to present no great difficulties, at any rate in compari-
son with other film theories.

The Psychological Theory of Emotion

The tradition of reflecting upon emotion is at least as old as the theories on
drama and literature that have come down to us through the ages. Most of
the great philosophers have had something to say about human emotions. 6

6
For a limited selection, see Calhoun & Solomon (1984).
INTRODUCTION 7

And ever since psychology attained the status of an independent scientific

discipline at the end of the last century, the theory of emotion has played an
important role in research and the formation of psychological theory One
need only think of William James; the behaviorists; the field theorists who
followed in the footsteps of Lewin; Freud, of course; and finally Schachter, all
of whom have left their mark on the most recent work within the cognitive
theory of emotion.
When emotion is studied from a cognitive-psychological perspective, as I
do here, the steering function is perhaps the most important characteristic
of emotion. Watching films is seen as a complicated process of information
processing. All processing of information by means of the human system is
governed, on the one hand, by the characteristics of the environment and,
on the other hand, by the limitations of a system formed by biological and
cultural evolution. One may expect a cognitive theory of emotion to indicate
exactly how this steering of the processing of information takes place.
In the last few years we have seen the appearance of a great many publi-
cations on emotion and the relationship to cognition, (i.e., the processing of
information). The scope of this introduction does not permit a discussion of
these studies.7 It is, however, important to note here that very few of these
publications are so general as to cover a broad spectrum of emotional phe-
nomena and their determinants. Such a spectrum must be extensive enough
to include the aesthetic emotions, more specifically, those that may present
themselves in the course of watching a film.
In this study, the theory of emotion developed by Frijda (1986) serves as
the basis for a description of the affective processes of the film viewer. In an-
ticipation of a more detailed discussion in chapter 3, it will suffice to say that
this theory will help us in two ways. First, it provides a working definition of
what constitutes a genuine emotion. This definition covers emotion in gen-
eral, as it occurs in daily life; it does not deal specifically with the situation
of the film viewer, and its plausibility is based in equal measure on the gen-
eral situation. Thus we are justified in concluding that films produce genuine
emotion if we can show that within the situation of the film viewer the crite-
ria for a genuine emotion have been met in the sense of the general theory
of emotions, which has been validated on grounds that are unrelated to our
research question.
Second, Frijda's theory describes in great detail how information is
processed, making it easier to highlight certain aspects of the complex pro-
cessing of information on the part of the viewer. The theory accounts for the
mechanisms governing experience and behavior and identifies the underly-
ing psychological mechanisms. There is no theory that covers the emotional
experiences of the subject as effectively as Frijda's, and it is this that makes
it so appropriate for our purposes.

7
References to the literature may be found in chapter 3.
8 CHAPTER 1

DELIMITING THE EMPIRICAL DOMAIN:

THE TRADITIONAL FEATURE FILM
AND ITS NATURAL VIEWER

In the light of our research question, the empirical domain of this study must
be defined by determining the range of two concepts, the film and the viewer.
The range of the third concept, feeling or emotion, will be left open here, to
be filled in at the end of our study by means of a hypothesis.

The Traditional Feature Film

We restrict our research into the nature of the affect generated by watching
feature films to a body of films known as the classical Hollywood cinema. At
the risk of over-simplification, a number of characteristics of the classical film
narrative, as presented by Bordwell in his prototypical description of this
body, are given. For a proper understanding of these characteristics, we must
keep in mind that most of them derive their significance in part from a com-
parison with other bodies of work, such as the art film or the historical-
materialist film. (See Bordwell [1985] for a description of these two genres.)
The prime characteristic of the classical film narrative is that the ambiguity
and uncertainty that it creates are subject to qualitative restrictions; the nar-
ration is as straightforward as possible. This characteristic is given concrete
shape in features such as the following, all from Bordwell (1985):
• Uncertainties or gaps in the information that the viewer is given concern-
ing the action are always temporary. In the end, the action invariably proves
to be a complete series of causally related events.
• There is usually a happy ending.
• The source of the causality lies in the main characters, who act more or
less purposefully and who are psychologically defined, albeit naively.
• Wherever possible, the narrative presents the events in chronological order.
• The viewer sees and hears only as much as is functionally necessary at a
particular time.
• It is always clear whether one is witnessing an objective or a subjective
scene.
• The film does not draw attention to itself as artefact: for one thing, the use
of technical means, such as staging, camera handling, and editing, are sub-
ordinate to the progress of events and the clarification of their causality.
• Conventions associated with a particular genre further restrict stylistic
freedom.
It is possible to see the body as a closed set of titles, all of which were made
within a certain period (roughly between 1920 and 1960), in a certain place
INTRODUCTION 9

(Hollywood), by means of a certain method (studio production). In fact, a

sample from this set formed the basis for Bordwell's generalizations con-
cerning classical film narration (Bordwell in Bordwell, Staiger, & Thompson,
1985). However, we prefer to see the classical body as a "fuzzy set," the size
of which is not precisely known; the members of this collection can only grad-
ually be established by determining the multidimensional distance from Bor-
dwell's prototype, as previously sketched. However, there is no formal
method, let alone a metric, available for this purpose. In each separate case,
it will be necessary to establish, on rational grounds, the extent to which a
particular film conforms to the classical style of film narration. 8 In this way
contemporary films might prove to be just as classical as the highly proto-
typical examples from the period in question. Indeed, there are a great many
contemporary films for which a case might be made (Bordwell, Staiger, &
Thompson, 1985 ).9
To avoid confusion, we will nevertheless consider as traditional all films
that broadly meet the criteria of the classical canon, whether or not they are
part of the classical body in a chronological sense. Where necessary, we will
be more precise and refer to classical films, meaning titles from the histori-
cal body, or classical mode of narration.

The Natural Viewer of the Traditional Feature Film

Just as there are different categories of films, there are also different cate-
gories of film viewers. If there is such a thing as a horror genre and a film
musical, then there must necessarily be lovers of the horror genre and fans
8
The obvious alternative would be to have the degree of conformity to the classical norms
determined by experts, in order to establish a derived index.
9
It may be useful to provide a somewhat more substantial impression of the position of the
traditional feature film within the entire range of films on offer. Contemporary films can be
ranked on an artistic continuum, according to the degree to which they conform to the classi-
cal canon. One end of this continuum might be formed by the popular film, which meets all the
criteria of the classical canon, with the contemporary avant-garde film at the other end. In the
ranking of films along such a continuum, differing and interrelated criteria could be of impor-
tance, such as cultural sophistication or a more specific cinematographical expertise required
to appreciate a film. The latter criterion is in turn related to the size of the natural audience of
the genre in question. Depending on the precision with which films could be ranked by means
of this method, the following genres might be found between the two extremes noted above: the
quality film which deviates only minimally from classical norms and appeals to a large audience,
though not so large as that of the popular film; the art house film, which in many respects devi-
ates from the classical norms and appeals to a more select audience; and the art film in a nar-
rower sense or avant-garde film, which deviates sharply from the classical canon in several re-
spects and is attractive to specialized audiences of the cinema clubs and film museums. In a hy-
pothetical arrangement such as the above, the traditional feature film would in any case include
the categories of popular film and quality film, as well as a portion of the art house category.
(For an arrangement that was used for more or less the same purpose, i.e., to give some idea of
a hypothetical range of films, see ANF, 1988.)
 10 CHAPTER 1

of the film musical. Categories of viewers are continually being formed through
a process of self-selection. Most film viewers have at the very least a strong
intuitive feeling about what does not appeal to them, and it is not necessary
to be a true film enthusiast to know approximately what type of film you are
looking for. The more films you see and the more informed you are about the
films on offer, the more pronounced your preferences are likely to be. Pref-
erences are determined by specific film characteristics. Bordwell (1985) used
the term norm to indicate that films are subject to historically determined
conventions with respect to the mode of narration, from plot to stylistic
choices. In our view, norms, as part of the theory of film narration, corre-
spond to attitudes, that is, affectively charged preferences on the part of an
audience, that have been formed during a learning process encompassing a
great many films. Films or film types may be seen as systems of norms, while
audiences are groups of spectators characterized by certain attitudes. Norms
and attitudes are geared to one another by means of a historical process of
selection. From the perspective of the film, we can say that each film or type
of film has its natural viewer; we will take this to mean the viewer who has a
preference for the film or type of film in question. Strictly speaking, the re-
sults of empirical research into the way a particular film is experienced can
only be generalized with reference to the population of its natural viewers.
The results must be obtained by means of a random sample from that popu-
lation, that is, all those persons who consider it conceivable that, if the op-
portunity presented itself, they would choose to see that film.
The natural audience of the traditional feature film can be further divided
into subaudiences, which differ from one another in the particular needs that
can be satisfied by attending a film. The division of audiences may corre-
spond to that of traditional genres, and it seems that there are significant lim-
itations to crossover of audiences. Nevertheless, it can be maintained that
what all the subaudiences of traditional films have in common is that they
want to be entertained, as we shall see in chapter 2. Furthermore, research
has shown that the differences between the natural viewer of the traditional
feature film and the natural viewer of the avant-garde film, such as those shown
during a program of experimental films, are far greater than the differences
between the devotees of the various traditional genres. 10
Finally, it is important to note that by the natural viewer or the natural au-
dience of the traditional feature film we do not only refer to an empirical cat-

10
There are at least two empirical studies that support this assumption. In a study involving
fans of the more artistic film (drama students) and a more general film audience, Tan, Egger-
mont, and Joosten (1989) found significant differences in the appreciation of certain charac-
teristics of the film narrative, reflecting a difference in primary motivation (see chapter 2). Like
Tan et al., Faber, O'Guinn, and Hardy (1988) found that those who preferred art films saw a
greater number of films, planned their film visits longer in advance, were less likely to go to the
cinema in the company of others, and were less likely to go to see a film for the emotional stimu-
lation it provided.
INTRODUCTION 11

egory of subjects characterized by particular interests, knowledge, prefer-

ences, and pragmatic attitudes but also to a category of subjects who share a
certain way of looking at a film, a certain set We are referring here to the
normal, voluntary visit to the cinema, where the viewer watches the film with
an open mind, that is, nonanalytically, and generally makes no effort to es-
cape the attraction of the fictional world portrayed on the screen. 11
To recapitulate, the natural viewer of a particular type of film is one whose
more or less stable film preference is for that type of film and who watches
such films with a set that is characteristic of people who watch films without
any special, say, analytical purpose in mind.
We have defined the traditional feature film and its audience not by means
of an exhaustive enumeration but by a prototypical description. It is impor-
tant to at least point the way to a form of empirical demarcation, (i.e., to in-
dicate a procedure). The fact that it is possible to identify by empirical means
the natural viewer of the traditional feature film as described previously is il-
lustrated by a questionnaire study by Tan, Eggermont, and Joosten (1989;
Tan, Eggermont, Joosten, & Spinhof, in prep.). On the basis of Bordwell's de-
scription of prototypes of the classical narrative style, preferential attitudes
were derived that had as their object characteristics of that narrative style, or
conversely, characteristics that contrast with that style. This collection of at-
titudes was presented to a group of respondents who might be expected to
be heterogeneous with regard to their preference for the traditional film or
for other genres. The attitude list displayed a high degree of consistency and
correlated well with other indices of taste in films and sophistication of be-
havior linked to film preferences, such as keeping abreast of, and making a
selection from, the films on offer. Depending on the selection aims, it should
be possible to determine a criterion score for this or a similar attitude list and
thus identify the natural viewers of the traditional feature film. A larger group
selected in this manner could then serve as a panel, with a view to establish-
ing the degree to which films may be termed traditional. Thus there appear
to be sufficient empirical grounds for the idea that there is indeed such a
thing as a category of natural viewers of the traditional feature film.

SCOPE AND LIMITS OF THE PRESENT STUDY

As noted earlier, Bordwell's theory of the film narrative plays an important

part in this study; its validity will not be questioned but rather assumed for
11
One consequence of this point of view is that most of the experimental-psychological re-
search that makes use of film as stimulus is not representative for the manner of viewing films
that is dealt with in this study. Often fragments of film material are shown, which, because they
are seen out of context, evoke different emotions, or emotions of a greater or lesser intensity
than would be the case in the natural viewing situation. The likelihood of this producing a pleas-
ant viewing experience is minimal. Where reference is made to experiments of this type, they
should be viewed with considerable reservation; wherever possible, this will be indicated.
12 CHAPTER 1

the sake of our argument. The theory is based on a long tradition of studying
narrative texts, so that we feel justified in drawing upon the insights it pro-
vides. Other theoretical viewpoints that might have been used to describe the
stimulus, the feature film, will receive limited attention. This study will not
seek association with other theories nor will there be any systematic discus-
sion of alternative theories, such as semiology, Lacanian psychoanalysis, or
ideological-critical theory, however valuable they are in themselves.
The cognitive-psychological study presented here is theoretical rather than
empirical. Wherever possible, the reasoning will be supported by references
to the empirical literature. There is an extensive body of literature dealing
with empirical psychological research into cognitive processes comparable to
those that form the focus of this study. These include processes that play a
role in understanding and remembering texts, notably stories and their com-
ponents (events, episodes, etc.). The nature of our research question is such
that it cannot easily be answered by means of a series of experimental stud-
ies. It entails a large network of concepts that must be explored and clarified.
For this reason, we decided at the outset that it would not be fruitful to seek
affiliation with that other large area of research within cognitive science, the
computer simulation of cognitive and emotional processes. 12 The latter ap-
proach uses programming techniques based on artificial intelligence, and we
considered it unlikely that the necessary accuracy in the description of the
concepts could immediately be achieved.
One of the main advantages of a theoretical approach is that it allows for
a more extensive description of the actual experience of the film, an experi-
ence that—as we shall be investigating here—may or may not prove to be
emotional. This is a major advantage over a largely experimental approach,
because it is precisely the measurement of the experience that is the great-
est obstacle, in particular where the more subtle aspects of that experience
are concerned. The simulation method is even more likely than the experi-
mental method to result in a neglect of conscious experience. The limitations
would seem to be inherent in the methodology itself. The problem of dealing
with conscious experience is recognized as one of the most important the-
oretical questions called up by the use of computer programs for modeling
mental activity.

PLAN OF THE BOOK

The question of whether films evoke genuine emotion can be answered in part
by examining to what extent watching films fulfills major psychological needs.
Watching a film may be assumed to serve a psychological need of some kind.

12
Computer models of emotion are as yet scarce. Exceptions are Colby (1975), Pfeifer (1982),
Frijda & Swagerman (1987), and Wegman (1985).
INTRODUCTION 13

Chapter 2, titled The Psychological Functions of Film Viewing, contains a

considered view of what brings people—both consciously and uncon-
sciously—to invest time, money, effort, and attention in watching films. The
chapter closes with a proposal for a classification of motives. Chapter 3, Film
and Emotion: Theoretical Background, presents a rough sketch of the expe-
rience of the viewer. Proceeding from Frijda's theory of emotion and Bord-
well's theory of film narrative, it formulates an answer to the question of the
contribution of the stimulus, the film narrative, to that experience. An initial
attempt is made here to explain why the experience of the viewer may be seen
as an emotional one. We argue that the film evokes witness emotions, and
the chapter closes with a provisional survey of the types of emotion that a
viewer undergoes while watching a traditional feature film. In chapter 4, The
Structure of Interest, we suggest that the film experience may be seen as an
emotional episode centering on a single most important emotion and ex-
tending with some degree of continuity throughout the entire film. This chap-
ter also identifies the psychological mechanisms cued directly by the film nar-
rative, which are responsible for this single most important emotion: inter-
est. In separate chapters, two determinants of emotion are discussed. Chap-
ter 5, Thematic Structures and Interest, deals with the affective function of
themes in films. The question here is whether it is possible, in general terms,
to identify the cognitive structures that are important for an understanding
of the events with which the viewer is confronted and that also result in af-
fectively charged expectations. Chapter 6, Character Structures, Empathy,
and Interest, examines how an understanding of the most important carrier
of the action, the protagonist, is accompanied by emotional experiences.
Contrary to the customary views on identification, we propose a theoretical
alternative, namely that the film systematically manipulates a certain view of
the film characters. Chapter 7, The Psychological Affect Structure of the Fea-
ture Film, addresses our second research question: How can the systematics
of film emotion best be described? In this chapter an effort is made to do jus-
tice to several characteristic qualities of the emotion and its determinants,
with special reference to the organization of both in time. And finally, in chap-
ter 8, The Feature Film as an Emotion Machine, we return to our first research
question. In the light of all our findings, we conclude that the film involves
the viewer in a series of overlapping illusions that appeal to more or less uni-
versal sensitivities and thus lead to a true emotional experience.
 The Psychological Functions
2 of Film Viewing

Filmgoers do not passively experience the stream of sounds and images

reaching them. They search out those aspects that appeal to them. In any film
there are certain event_.;; that arc more significant than others, and the aver-
age audience will probably be of one mind about which these are. On what
do we base this simple fact?
Let us compare the filmgoer with a tabula ra..,a, such as the subject of .John
Searle's Chinese Room thought experiment (1980). The subject, let us imag-
ine Searle himself, who knows no Chinese, answers a series of questions con-
tained in Chinese symbols passed to him from outside the room. Hy means
of a fixed code contained in a book, he translates these symbols into other
series of Chinese symbols, which are in turn passed to someone outside. In
this way, the subject answers the questions without attaching any meaning to
what he receives or to what he passes on. According to Searle, such a subject
cannot be said to understand-in the sense of attaching intentional meaning
to-the information that he processes. For this reason, Searle argued, a com-
puter program, which does no more than manipulate symbols, can never be
said to have the power of thought.
The subject in the Chinese room does not think. Rut more important, he
does not feel. He does not feel because he attaches no importance to the in-
formation that he is processing, which goes a bit further than saying that he
does not sec any meaning in the information. For our purposes, a general de-
finition of to attach importance to information will suffice. Let us simply say
that a feeling subject-not necessarily intentionally-selects from the in-

15
16 CHAPTER 1

coming information something that affects him, something that matters,

something that immediately and spontaneously strikes him as significant.
The situation of film viewers is quite different from that of Searle's subject,
for they are actually searching—although not always consciously—for items
that appear to them to be important. It might be more correct to say that in
the course of the film such items present themselves as such; they impress
themselves on the viewer as important and, for this reason, are more or less
automatically given preferential treatment. Both of the previously mentioned
formulations are based on the assumption that there is something within the
viewer to which the film appeals, something that is triggered, so to speak, by
the stimulus of the film, instead of passing unnoticed along the secret path
of the fixed code.
If we accept that a film audience does indeed feel, then there must be some-
thing in the viewer to which the stimulus appeals: a predisposition, sensitiv-
ity, preference, or motivation. This prompts us to ask what the spectator looks
for in a film? Because we are not referring here exclusively to motives the viewer
is aware of, this question can also be formulated in more general terms: What
are the psychological functions of the feature film?
It is important to note that the functions we are looking for are associated
directly with watching films, that is, the actual experience of film, as opposed
to going to the cinema. The reason for this restriction is that our main re-
search question deals with those experiences associated with watching a film.
The surroundings in which this takes place, such as a visit to the cinema or
an evening out with friends, are of lesser importance in seeking to under-
stand the nature of the feelings evoked by films.
There is no existing model for the description of the psychological func-
tions of the feature film. One method would be to collect all the motives fa-
miliar from the research on motivation and then to determine to what extent
watching films contributes to the realization of these motives. McGuire
(1974) opted for this method in his discussion of motives for media use, and
to a degree it has served as a model for this discussion. I decided, however,
to make use of a convergent approach, bringing together in a limited num-
ber of categories functions from various disciplines, including television re-
search, art theory, and film theory.

IS MEDIA USE MOTIVATED?

Empirical studies on the functions of the film are few and far between. There
is, however, a body of psyehoanalytically oriented theory on the subject, and
a great deal of research has focused in a more general sense on the functions
of the media, notably television.
THE PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 17

Film and Entertainment

Austin (1988) offers an overview of research into motivations for moviegoing.
A number of surveys have been carried out in which film viewers were asked
directly about their motivation (Austin, 1986; Faber, O'Guinn, & Hardy,
1988; Lichtenstein & Rosenfeld, 1983; Moller & Karppinen, 1983; Palm-
green, Cook, Harvill, & Helm, 1988; Wyatt & Badger, 1988). 13 Without ex-
ception, all these studies indicate that entertainment is the major reason for
going to the cinema. Jarvie (1970), who relied on older survey research,
came to the same conclusion. This is not surprising, considering Hollywood's
reputation as the home of the entertainment industry. Entertainment has
been found to be the major reason for filmgoing in surveys held among widely
varying audiences in the United States and in Europe, including the Soviet
Union, before 1920, and since the 1940s (Austin, 1988, pp. 51-57). There
are, however, a number of more subtle distinctions to be made here.
In a survey among college students, Austin (1988) found that entertain-
ment was the main reason for going to the film ("to forget and get away and
escape," "to enjoy a pleasant activity," "to pass the time"). He does, however,
identify other motives almost as important as certain aspects of entertain-
ment, namely "to learn information," "to relieve loneliness," "to impress or
conform to others," and "to learn about self." Palmgreen, Cook, Harvill, and
Helm (1988) concluded on the basis of their survey among college students
that, "While entertainment certainly ranks as an important motivation for
moviegoing, there are many other gratifications which are sought in varying
digress from the theatrical experience" (p. 17). They mention content, social
context, and structural and technical attributes of the medium. Moller and
Karppinen (1983) identified not only a Relaxation and Mood Change factor,
but also dimensions concerned with Interest and Information, Social Rela-
tionships, and Aesthetics and Art. Moreover, they unearthed a number of dif-
ferences between the various categories of viewers: Relaxation and Change
were more important for those watching a traditional feature film than for an
art film audience, for whom Aesthetics and Art were a prime factor. Faber,
O'Guinn, and Hardy (1988) reported a similar difference between these two
groups.
The review of the results of questionnaire research stressing the impor-
tance of entertainment provides only a partial answer to the question of what
the viewer looks for in a film. In the first place, there is a methodological
problem attached to the use of questionnaires, namely, that this type of re-
search does not bring to light any deeper motives. Even more important is a

13
For a survey of a number of older studies, see Bannerman & Lewis (1977, p. 128-129) and
Handel (1950).
 18 CHAPTER 1

conceptual shortcoming: What do we actually mean by entertainment? The

concept is a sociological one, which means that it is more at home in a de-
scription of the sociological functions of the film than in an inventarization
of its psychological functions. Entertainment appears in different classifica-
tions of the social functions of the mass media. In a review study by McQuail
(1987), for instance, entertainment is to be found alongside such functions
as information, correlation, continuity, and mobilization, all of which are de-
fined on a higher aggregate level of analysis than that of the individual. In the
third place, specific characteristics of the film are almost totally absent from
the research. This fact was not lost on Palmgreen et al. (1988), who observed,
"More noteworthy for their absence, though, are motives connected to ele-
ments such as plot, characterization, film type, theme, conflict, suspense, etc."
(p. 20).

Psychological functions

The Filmic State

Psychoanalytically oriented film theory, which is based to a large extent
on Freud's theory of the unconscious, deals with the cathartic effect of films.
All the elements necessary to describe the cathartic function of the feature
film are contained in his work on jokes (Freud, 1905/1940). The crux of this
theory is that a saving in psychological activity is a major source of pleasure.
Freud suspected that the mechanism on which jokes are based, which in-
volves such a saving, also operates in aesthetic processes in the broader sense
of the word. Like jokes, works of art may employ a technique that generates
tension, for example, by creating a contradiction and subsequently resolving
it. In addition, the technique of a work of art, like the contents of dreams and
free fantasy, may enable the subject to gain access to representations, desires,
and impulses that are normally unconscious. 14 This presupposes that the work
of art has a tendency, that is, an aggressive or sexual charge. Thus in Freud's
view, the effect of the work of art has both a formal aspect (the technique)
and a content aspect. The form may in itself give rise to pleasure, while in
other cases it makes possible an effect of the content on the unconscious,
opening a door that is usually closed. And finally, according to Freud, there
is a third aspect of the effect of jokes, one that is actually beyond the scope
of the homeostatic view of human motivation. I am referring to the tendency
of jokes to touch upon the content of consciousness shared by social groups,
such as prejudices. Laughing at a tendentious joke reinforces the bond be-
14
For Freud's well-known ideas on dreams as the veiled expression of unconscious psychic
contents, see Freud (1900). For the function of fantasizing, see Freud (1908). For the possibil-
ity that fiction gives the subject access to unconscious desires and impulses, see Freud (1912-13;
1919; 1942).
THE PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 19

tween individuals and strengthens their sense of belonging to a particular

group.
Film theorists Metz (1975a, 1975b) and Baudry (1975/1986) described the
filmic state of the viewer as one in which, psychologically speaking, the sub-
ject functions quite primitively. There is a certain regression in the filmic
state, which may in itself be pleasurable. The circumstances under which
films are shown are highly favorable to this state: the spectator is seated al-
most motionless in the dark, so that his or her perceptions are more domi-
nant with respect to other functions than is customary in most situations. The
film appeals strongly to the individual viewer. Indeed, watching a film is an
individual and isolated activity, one which in effect takes place despite the
presence of the other viewers. The state of primary functioning, 15 together
with the nature of the fictional material and its organization, contribute to ef-
fects that are comparable to dreams. According to Metz (1975b), films share
with dreams and fantasy what is known as the romanesque formula: the
structuring and organization of fictional contents are the result of precon-
scious primary processes, while the thematic content originates in the un-
conscious. Structure and organization facilitate access to unconscious con-
tents. Finally, the theory of the filmic state also includes assouvissement pul-
sionnel or drive satisfaction (Metz, 1975b), enabled by access to unconscious
material and restoring affective balance. Here we have an interesting vari-
ation on catharsis, as Metz emphasizes that there should be some drive sat-
isfaction but not too much. An excess would result in heightened defense.
In the work of Baudry (1975/1986) the reinforcement or recovery of the
affect balance is of secondary importance; it is the regressive state itself that
is dominant. In his view, the subject unconsciously wishes to return to the
primitive narcissistic phase of psychodynamic development, in which the sep-
aration of one's own body and the external world has not been fully com-
pleted. No distinction is made between representations and actual observa-
tions, and desires are satisfied in a hallucinatory manner.
In the psychoanalytical literature on the filmic state there is an unusual
dual metaphor, whereby consciousness is seen as a kind of cinema and the
cinema as a kind of consciousness. This comparison becomes bizarre when
it is no longer clear to what extent it is intended to be taken literally. It was
B. Lewin (1950) who launched the idea of a dream screen, the basis for a
kind of darkroom of consciousness onto which fantasies and dreams are pro-
jected. This psychological structure allegedly bears a symbolic relationship to
the mother's breast. Borrowing this imagery, Baudry suggested that the sub-
15
The term refers to the primary processes distinguished by Freud, in which the ego func-
tions, notably reality testing, are diminished with respect to the secondary processes. Thus one
functions in a primary fashion just before dropping off to sleep or when one is in a state of ex-
haustion. See, for a discussion of these processes, Hilgard (1962), who also includes references
to Freud's formulations.
20 CHAPTER 1

ject sees the cinema screen as his or her own dream screen, so that con-
sciousness actually encompasses the cinema. 16 In effect, one is observing
one's own imagination at work (Baudry, 1975/1986). This is the same phe-
nomenon to which Metz (1975a) alluded when h e claimed that the viewer
identifies with himself or herself. And finally, Eberwein (1984) saw the ac-
tual projection and the actual screen as a psychological prosthesis of our
dream screen. It is possible to distill from this rather vague metaphor the view
that something akin to becoming immersed in one's own perception is an es-
sential characteristic of the filmic state of the viewer.
Another psychoanalytical concept, scopophilia, which is variously known
as Schaulust and voyeurism, has also been put forward as an elementary drive
that motivates one to watch films. Metz (1975a) explained that scopophilia
implies a distinction between the viewing subject and the desired object. The
film is a perfect example of this desire, since the object seemingly present on
the screen is in reality absent. In a more specific characterization, Mulvey
(1975) described scopophilia as the libido-driven act of looking at another
person or object, involving the fulfillment of desires by means of narcissistic
identification. The body as displayed in traditional feature films represents an
ideal ego, which in Mulvey's view is generally the ideal male ego: the suc-
cessful seducer of women.
The application of psychoanalytical concepts, notably those borrowed
from Lacan, to the experience of the film viewer has met with scathing sci-
entific criticism and rightly so. The ontological status of the concepts is un-
clear, the logical consistency of ideas leaves something to be desired, and the
frugality requirement appears to have been reversed (Carroll, 1988b). Terms
such as scopophilia and voyeurism have a pathological connotation that is
quite gratuitous here. The metaphor that places the cinema inside the head
of the viewer—and vice versa—is totally unacceptable, mainly because it rests
on a homunculus theory of consciousness: someone who is in the head of the
viewers is watching the screen. But this would mean that in his head there is
another viewer at work, and so on into infinity.17 If, however, we disregard
these shortcomings for a moment, the essence of the contribution of psycho-
analysis might well be as follows: What viewers find so appealing in films is
the fact that they help them to fill a gap, a gap that originated somewhere in
their past. Watching a film makes it possible to return—albeit fleetingly and
only in one's mind's eye—to a lost paradise and to see wishes fulfilled that
otherwise must be repressed. The fact that while viewing a film one delivers
oneself over to fantasy, laying aside the oppressive rationality of everyday life,
is in itself a source of pleasure that may be highly motivating. These princi-
ples provide us with functional hypotheses concerning the etat filmique. This
16
See Carroll (1986) for a related critical discussion of the "dream screen" metaphor.
17
See Pylyshyn (1973) for a similar criticism of the homunculus theories of mental imagi-
nation.
T H E PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 21

approach will be of no great help in establishing precisely how the viewer is

gripped by a film, why it has such a direct and powerful effect, and which styl-
istic and technical mechanisms appeal to which of the viewer's perceptual
and cognitive sensitivities. These questions as to precise mechanisms have
seldom been the object of systematic research within other branches of cin-
ematic theory and have, in any case, remained largely unanswered.

Film Ideology and Participation

Psychoanalytically oriented film theorists have attempted to link the con-
dition of the viewer with the ideological effects of the film. Heath (1981), for
example, maintained that the spectator is continually being urged to fill in
gaps in the action, say, or the fictional space. This activity fulfills a deep-
rooted desire of the subject for closure. The traditional film forces the spec-
tator to call up ideologically determined knowledge, which thus becomes
even more firmly established. Others refer to the positioning of the specta-
tor by the film, by which they mean that the film is continuously allocating
to the viewer not only a physical view of the scene but also an ideological po-
sition with respect to the reality being portrayed (Browne, 1975; MacCabe,
1976). A major elaboration on this approach is that put forward in feminist
film studies; it documents the dominance in feature films of those points of
view that fit in best with the ideology and desires of masculine viewers. The
study by Mulvey mentioned previously is a case in point, and it has been fol-
lowed by a number of others, including those by Kaplan (1983), de Lauretis
(1984), and Kuhn (1982).
Certain sociological film theories, notably the earlier ones, bear some re-
semblance to psychoanalytical film theory. There, however, the filmic state is
the precondition for another effect, namely ideological manipulation. In the
mirror theory of the film developed by Kracauer (1947), film motifs reflect
desires rooted in a collective subconscious. His reasoning is that—within the
classic studio system of film fabrication, at any rate—producers must give the
audience what they want. In an ongoing competitive process, film motifs,
such as themes, givens, problems, and so on, are shaped and adjusted to con-
form to the contents of the subconscious. Hence the capacity of film to act
as a moral safety valve.
Prokop (1974) reviewed those authors who characterize aspects of the
filmic state as dreamlike or trancelike and thus primitive and emotional as
opposed to rational. He suggested that it is not this state that the audience
wants and expects, but rather that this is the way the dominant films shape
cinema experience. The film situation gives the spectator an opportunity to
transgress social norms and codes of behavior, albeit within strictly deter-
mined boundaries, in such a way that the stability of society is not endan-
gered and may even be enhanced.
Similar conclusions are reached in an earlier piece of research that has re-
22 CHAPTER 1

ceived more attention than Prokop's solid study: the socioanthropological

work Hollywood: The Dream Factory by Powdermaker. A small elite in Holly-
wood sets before the cinema viewer films that conjure up . . vicarious ex-
periences beyond his own activities, . . . portray solutions to his problems,
[and] . . . provide models for human relationships, a set of values and new
folk heroes" (Powdermaker, 1950, p. 15).
Today's traditional feature film continues to offer its natural audience this
type of informal schooling in the dominant ideology (Jowett & Linton, 1989).
Genre formulas are excellent vehicles for passing on the dominant morality,
as noted by Schatz (1981) in his work on the major Hollywood genres. The
conservative effect of the traditional feature film, notably the classical Amer-
ican film, is often stressed in studies on censorship and self-censorship. This
is an aspect of Hollywood films that Powdermaker deals with in some detail
and in the more recent works on the history of studio production has been
even more extensively documented (Bordwell, Staiger, & Thompson, 1985).
The cultural studies approach has, however, contributed important nu-
ances to the study of feature films. Acknowledging that some films preserve
dominant values, certain genres are thought to appeal not exclusively to the
insights and stereotypes shared by the cultural majority. A random example
of a recognized corroborative genre is the romantic film comedy of the thir-
ties (Neale & Krutnik, 1990). In this case, the insights and stereotypes are
concerned with female sexuality in relation to marriage and the family. An
equally random example of a genre that might be expected to foster change,
opening the way to negotiation on ideological standpoints pertaining to these
same issues, is the classical American film melodrama (Lang, 1989).
There is also room for another, somewhat more hedonistic view of the pre-
sumed consolidation of existing doctrines. People's desire for variety and new
experiences is more or less balanced by the fact that they enjoy what they
know. According to McGuire (1974), dramatic story material fulfills an almost
universal need for categorization and identification; stock characters and for-
mula plots are an ideal way of ensuring easy assimilation. Results reported in
experimental psychology likewise point to a preference for familiarity. First,
there is the mere exposure effect (Zajonc, 1968; 1980): subjects prefer stim-
uli that have been presented a number of times to those seen less often, even
when they are not aware of the fact that they have already seen them. Sec-
ond, experiments have been carried out to assess the aesthetic effects of the
typicality of stimuli, based on a categorization theory in which so-called pro-
totypes play an important part (Rosch, 1978; Rosch & Mervis, 1975). The he-
donic value of a stimulus is said to be based on the extent to which it agrees
with a prototypical instance of the class to which the stimulus belongs. Results
differ on the degree of prototypicality associated with maximum preference. 18

18
Examples of such research are to be found in Gaver and Mandler (1987), Hekkert and van
THE PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 23

According to some researchers, a need to confirm one's convictions de-

termines the choice of films. Atkin (1985) described the results of research
into selective exposure to television programs that reinforce cognitive pre-
dispositions, such as political and moral values. Gunter and Wober (1983)
found a positive correlation between a belief in a just world and a preference
for action and adventure programs on television; these results can be inter-
preted to mean that people watch a particular type of program in order to see
their convictions confirmed. Here, too, a number of studies adhering to the
cultural studies approach have shown that contemporary television offerings
are in fact quite diverse and that viewers are not interested solely in corrob-
oration. The study by Ang (1985) on the significance that television viewers
attach to the series "Dallas" is a case in point (see Fiske, 1987, for a sum-
marizing view).
Bolstering social identity could be another function of watching feature
films. The recognition of socially shared views can fulfill the individual's need
for a sense of belonging. Take, for instance, the way people laugh at a biased
joke, told at the expense of someone or something. This strengthens the bond
between the laughers, as Freud (1905/1940) points out, because they share
both the joke and the underlying view of the target. Not only sharing a laugh
but sharing a variety of other feelings serves to reinforce the realization that
one is part of a community rooted in one and the same worldview. This might
include such shared feelings as the primal anxiety that grips audiences watch-
ing a horror film. The relevant social function has been studied by Fenigstein
and Heyduk (1985) with reference to television viewers. In that connection
it is appropriate to cite McGuire, who translated a well-known view on the
ceremonial significance of the theater to the conditions of media use:19 "Me-
dia contents serve to reaffirm for the recipients their sense of participating
in the events and institutions depicted, of being part of the human drama on
a broader stage than their own personal concerns" (McGuire, 1974, p. 189).
It may be enlightening to examine more closely a social gratification that
the traditional feature film does not provide, not at any rate to the same ex-
tent as a visit to a museum, a concert, or a play: namely the considerable
prestige value attached to participation in art and culture. Theater, especially
modern theater, which is considered a form of high culture, a place to be
seen, would probably score high when it comes to satisfying motives related
to distinction (Bourdieu, 1979). The traditional film, notably the popular fea-
ture film, owes much less of its appeal to this motive. For one thing, it is of-

Wieringen (1990), Light, Hollander, and Kayra-Stuart (1981), Mandler (1982), Martindale and
Moore (1988), Whitfield (1983), and Whitfield and Slatter (1979).
19
See for views on theater as ceremony, for example, Duvignaud (1965/1973). See also the
well-known anthropological studies carried out by Malinowski and the Cassirer's philosophical
research into the function of myths. The latter is said to touch upon the expression of systems
of belief which promote the unity of a group.
24 CHAPTER 1

ten difficult to be seen in the cinema, while attaining status would seem to
be of much less importance than simply being with one's friends on an
evening out. 20 (The latter reason is also put forward by Jowett and Linton
[1989] as a major reason for going to the cinema.) According to Bourdieu,
choosing to see a popular feature film may even serve to express quite the re-
verse of a taste for high culture: an anti-Kantian attitude, in which pleasure
is foremost and participation and empathy are major goals.

Media Entertainment and the Regulation of Emotion

Within the uses-and-gratifications paradigm a great deal of research has
been done that provides insight into the psychological functions of television
viewing. I will confine myself here to the functions that have thus far been es-
tablished for television drama, as this type of program resembles most closely
the feature film. In the more recent television research, it has been assumed
that media entertainment plays a role in the regulation of affect.
According to Tannenbaum (1980,1985), television entertainment is always
characterized by an increase in emotional arousal. This increase is said to oc-
cur regardless of the genre one is watching. Zillmann (1991a) likewise referred
to research that indicated that "excitatory reactions produced by exposure to
extremely amusing films are very similar to those produced by exposure to sus-
penseful, violence-laden films." The views of Scheff on this point are in the
same vein (Scheff, 1979; Scheff & Scheele, 1980). He assumed that televi-
sion entertainment results in catharsis, defined as the experience and ex-
pression of feelings. Like Tannenbaum, he took it for granted that the emo-
tional arousal brought about is not in itself pleasurable and can be described
as a combination of such emotions as fear, anger, and embarrassment. In
addition, however, the stimulus must offer the proper degree of distancing.
If there is excessive distancing, for example, when violence is exceptionally
stylized, no unpleasant emotions are evoked, and there can be no catharsis.
The same is true when there is too little distancing, because viewers are then
so overwhelmed by emotions that they have no opportunity to release them.
The latter situation might occur during highly realistic and explicit scenes of
violence.
Zillmann and his associates have attempted to demonstrate the relation-
ship between media use and affect regulation in a more direct manner than
is possible with the customary survey research. They have shown that both
an increase and a decrease in excitation may be functional in mood man-
agement (Bryant & Zillmann, 1984; Zillmann, 1988; Zillmann & Bryant,

20
However, certain films appear to have a high distinction value for certain categories of view-
ers. See, for instance, the motive identified by Austin (1988): "To impress or to conform to oth-
ers," referred to in the section Film and Entertainment.
T H E PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 25

1985). In an ingenious experiment carried out by Bryant and Zillmann

(1984), subjects selected those drama programs that led to optimum arousal.
When experimental manipulation was used to induce in subjects a state of
boredom, they opted for programs with a high arousal potential. Other sub-
jects, who had been manipulated into a state of stress, displayed a preference
for programs with a low arousal content. The programs had been tested be-
forehand, to determine their arousal potential, and the dependent variable
was measured by means of both choice of program and duration of viewing.
An experimental study by Meadowcroft and Zillmann (1984) indicated that
both the quality and the intensity of the emotional state of the media user
can be regulated by television entertainment. Women who were in the middle
of the menstrual cycle chose to watch serious drama rather than comedies,
while those in the premenstrual or menstrual phase were more interested in
comedy programs.
Zillmann (1988) summed up the affect-regulation function of entertain-
ment (not only film, but also drama in general, music, spectator sports, etc.)
by means of a hedonistic theory, the main features of which are as follows:
1. There is a general tendency toward maximalization of gratification and
minimalization of aversion.
2. In the case of extreme under stimulation, individuals display a preference
for varied, potentially arousing stimuli.
3. In the case of extreme overstimulation, a reversal of 2 (above) is seen.
4. Individuals in an acute state of aversion seek relief in the most engaging
and absorbing types of stimulation, of a positive hedonic tone.
5. Individuals who experience gratification—and thus feel good—want to
continue this state and therefore seek less absorbing stimuli. They do not
seek hedonically positive stimuli or are less likely to do so than the indi-
viduals in (4).
It is important to note here the reservations Zillmann (1985) made with re-
spect to this operant conditioning theory of selective choice. There are limits
to the possibilities for mood regulation by means of television and film drama.
An affective state may be too strong or the stimuli may have a permanent ef-
fect that would cancel out influence of the entertainment product itself.
Christ and Medoff (1984), for example, came to the conclusion that individ-
uals with an acute problem, such as anger due to a provocation that they felt
had to be resolved immediately, temporarily refrained from the use of enter-
tainment media at all.
The operant-conditioning function of media use is not limited to television
programs. Fiction can function in the same way. Survey data presented by Nell
(1988) indicated the importance of reading as a means of mood enhance-
ment. On the basis of her study on reading satisfaction among primary and
26 CHAPTER 1

secondary school pupils, Tellegen-van Delft (1987) concluded that reading

with a view to maintaining or regaining affective balance is the main form of
reading behavior. Furthermore, there is a more than superficial parallel to be
drawn between the previously mentioned notions on media entertainment
and the sociology of leisure, in any case with respect to the excitation-en-
hancing aspects of affect regulation. Media use, notably watching feature
films, may be seen as a mimetic leisure activity. According to Elias and Dun-
ning (1986), such activities are an enclave in the midst of everyday life, where
emotional excitement can be kept within bounds, resulting in emotional
catharsis. This forms a kind of compensation for the monotony and boredom
of modern society. A feature common to television drama and films is the fact
that both the emotional arousal and the safe experience of emotion are at-
tained by means of some form of identification with the characters portrayed.

Fantasy and Emotion Regulation. An interesting effect of the cathar-

sis hypothesis that is also found in Freud is the fact that under certain con-
ditions media entertainment, especially television drama, evokes fantasies
that are capable of reducing the intensity of emotions, those of an aggressive
nature, for example. 21 The work of Feshbach and J. L. Singer focuses in part
on this hypothesis (Feshbach, 1976; Feshbach & R. Singer, 1971; J. L. Singer,
1973; J. L. Singer & D. G. Singer, 1976, 1981). Feshbach (1976) discusses
the fantasy mechanisms that can lead to emotional catharsis, in this case, a
reduction in the level of aggression. These include goal substitution and the
related (symbolic) expression of aggression, 22 which involve expressing the
aggressive response in imaginary or symbolic form. In the case of positive re-
inforcement, a fantasy, such as a wish fulfillment, can evoke a pleasurable
emotion that is incompatible with the aggressive response.
Finally, it would seem plausible that some fantasies evoked by television
drama are in themselves pleasurable. In other words, not only do they con-
tribute to a catharsis but the emotion-laden symbolic representations that ac-
company television viewing could in themselves be a motive for watching the
programs concerned. This has been noted by various researchers studying
fantasy. Klinger (1971), for example, stresses that fantasy, that is, mentation
devoid of obligation, reflects current concerns. These may involve as yet un-
realized goals, corresponding to the familiar psychoanalytical views on wish
fulfillment through daydreams, or something as simple as instrumental ac-
tivity that has been interrupted. The most famous form of fantasy involves
identification with powerful persons, the Walter Mitty effect, or the more con-

21
The restriction "under certain conditions" should be stressed because then the opposite ef-
fect may well result. The literature on the effect of the media on aggression is too extensive and
too complex to discuss here. See Van der Voort (1982).
22
The terms are Feshbach's.
THE PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 27

troversial Rambo phenomenon. 2 3 This type of wish-fulfilling fantasies may

contribute to more permanent identity effects.
Film theorists have often pointed to the potential of the film for liberating
fantasy. Sorlin (1977) distinguished the rational pleasure involved in follow-
ing the narrative from the irrational element represented by emotionalizing
fantasy. As Miinsterberg (1916/1970) wrote, "The massive outer world has
lost its weight, it has been freed from space, time and causality, and it has
been clothed in the forms of our consciousness . . ." (p. 95), and part of this
effect may linger after the performance is over. Perhaps we may go a step
further by assuming that the activity of fantasizing, regardless of the content,
is in itself enjoyable and that the media provide a powerful stimulation to
fantasize.

Media Entertainment and Learning

It is by no means inconceivable that readers actually learn from television
drama, thus fulfilling Horace's dictum utile et dulce. More important for the
purposes of our investigation into the psychological functions of entertain-
ment than any objective learning effect is the possibility that viewers prefer
this type of program precisely because they expect to learn from it (or expect
to feel as if they have learned from it). Here learning is employed as a col-
lective term for a variety of effects that are more permanent than simple di-
version or mood enhancement. In fact, the definition is broad enough to in-
clude the acquisition of new behavior and new insights, as well as the reverse
effect of reinforcing existing knowledge, as referred to in the section on ide-
ological effects.
Research into the relationship between aggression and watching television
has not only examined the hypothesis that there is a reduction in aggressive
impulses but has also put forward the possibility that certain emotional be-
havior, not least aggression, can be learned via the fantasies evoked by the
medium. This possibility has been studied in detail by Bandura (1973), who
based his research on the theory of social learning; he carefully balanced his
hypotheses against the claims of the catharsis theory. J. L. Singer and D. G.
Singer (1981) concluded that on the basis of empirical data, the social learn-
ing theory is more plausible than the catharsis theory. According to the so-
cial learning theory, television drama, which in the United States is dominant,
contributes less to the regulation of emotions than to something akin to emo-
tional learning. Action drama programs serve to reinforce certain emotion-
ally charged behavior, such as aggression, and the cognitions that accompany
these emotions. For example, they portray fantasy models for whom aggres-
sion is instrumental in attaining prosocial goals.
23
They refer to The Secret Life of Walter Mitty (1947 ) and Rambo: First Blood, Part II (1985),
respectively.
28 CHAPTER 1

A great many studies have demonstrated a positive correlation between

the aggressive disposition of viewers and their viewing preferences in terms
of violent themes. It is, of course, difficult to distinguish between cause and
effect; the earlier correlation may or may not mean that an aggressive dis-
position is necessarily exacerbated by seeing a particular program. (For sev-
eral surveys, see Atkin, 1985, pp. 77-79, and Gunter, 1985, pp. 108-109).
Moreover, a positive correlation between an affective disposition and the the-
matic content of programs does not necessarily mean that reinforcement of
this disposition—assuming that it actually takes place—is a motive for se-
lecting the programs in question. In the case of a disagreeable affect, such as
fear, we would have to assume a masochistic motive on the part of the viewer.
Catharsis, or more precisely, the anticipation of catharsis, would appear to
be a more promising assumption. On the basis of their discussion of several
studies focusing on the connection between one's fear of crime (i.e., becom-
ing the victim of crime) and a preference for television drama with a crime-
related theme, Zillmann and Wakshlag (1985) believed that expected cathar-
sis is indeed a major motive. In the majority of the television crime series,
justice triumphs. Therefore, Zillmann and Wakshlag assumed that the more
fearful viewers, who are anxious for everything to turn out all right, derive
more satisfaction from such series than the less fearful.
Experiencing intense emotion is capable of reinforcing one's self-esteem.
Exciting films, such as thrillers and, above all, horror films give viewers the
opportunity to experience a feeling of mastery or "competence" (White,
1959). Audiences viewing films of this type are actually afraid, but they never-
theless experience the satisfaction of being able to master their fear (Piet,
1987). This goes especially for men, who gain additional benefit from ex-
hibiting mastery in the company of women (Zillmann, Weaver, Mundorf, &
Aust, 1986). The same may be true of films that evoke sadness, pity, or re-
vulsion, although we have no research to back this up. According to Zillmann
(1991a), the evidence for habituation due to regular viewing of exciting tele-
vision drama is both scant and inadequate. The intensity of the response may
indeed decrease, but it has never been demonstrated that reduced recep-
tiveness to emotion can be generalized to events in the real world. Neverthe-
less, anticipation of such an effect, by which viewers gradually develop a kind
of callousness, could form a motive for certain types of viewers.
Compared to the relative wealth of results that have been reported with re-
gard to affective learning processes, only incidental empirical information is
available to illustrate the possibility of cognitive learning effects as a motiva-
tion for watching dramatic media entertainment. However, a great many well-
known claims in this direction have been made by theoreticians in various
branches of the arts. In recent psychological aesthetics, the notion that art
produces insights that go beyond structural order in the stimulus has received
somewhat less attention. This is no doubt due more to a methodological
THE PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 29

choice than to an assumption that works of art have no message. 24 In many

cases art is about something; it sets people thinking and reveals new per-
spectives. A work of art does not speak to its beholders directly, but rather
involves them in a process that leads to insight, whether it is a question of a
truth, an ideal, or a possibility.
In contrast to experimental aesthetics, the romantic notion of gaining in-
sight into a truth that is not directly available is by no means foreign to con-
temporary art theory. Works of art have layers of meaning; one of these is a
symbolic layer, which is concealed beneath, or interacts with, a more super-
ficial denotation. According to Langer (1953), an artwork evokes an aesthetic
emotion, which gives way to a well-rounded insight into both our inner life
and external reality. To take an entirely different example, the Gestalt psy-
chologist Arnheim (1971) maintained that the viewer finds order in the stim-
ulus, and that "this order [should] reflect a genuine, true, profound view of
life" (Arnheim, 1971, p. 56). However, the problem of the relationship be-
tween art and reality is so complex that it is doubtful whether it will ever be
fully understood.
Films are also about something, and they can teach us a great deal about
reality, as is clear from the reviews of feature films. 25 What feature films ac-
tually say about reality and how they do so is difficult to explain; this is a spe-
cial aspect of the problem of reality in art. The psychoanalytical viewpoint,
according to which the film provides access to an inner reality, that of re-
pressed fantasies, wishes, and fears, has already been dealt with at length.
But there are also countless interpretations of films that are based on their
symptomatic meaning—indeed, too many to examine here. (See Bordwell,
1989b, for a critical overview.) The relationship between film and reality may
well be the subject on which the classic film theoreticians differ most. The re-
alism theoreticians, among them Bazin (1958, 1959, 1961, 1962) and Kra-
cauer (1960), defend the standpoint that film should portray reality as it is.
It should, however, be noted that they entertain very different views on what
reality actually is. Bazin believes that film, because it is capable of register-
ing change, is the ultimate answer to a human being's cherished desire to cre-
ate a perfect analogon of reality. Kracauer (1960) stressed that reality should
24
In general, experimental psychology is more comfortable with processes and formal vari-
ables of mental processes than with the contents of those processes. This was noted by McClel-
land during the 1950s, in connection with his defense of the Thematic Apperception Test (see
McClelland, 1955). The cognitive revolution has not brought about any appreciable change in
this situation. While the structure of mental schemas, say, was studied, the contents was left to
researchers in other disciplines.
25
By touching upon the question of what films are actually about, I am treading on danger-
ous ground. By its very nature, the interpretation of films is open-ended. For practical reasons,
I have opted for the simple standpoint that thematic structures can be identified by empirical
means. An inspection of reviews is a first step. There are, of course, more formal methods, such
as the analysis of synopses and protocols on the basis of cognitive models of narrative structure.
30 CHAPTER 1

be revealed; the maker has the entire technical arsenal of cinematography at

his or her disposal in the struggle to expose what is mistakenly seen as reality.
Another well-known viewpoint concerning the relationship between film
and reality is that rooted in drama theory. Drama, which includes the feature
film, may be seen as a model for social reality. Dramatizing and portraying
familiar problematical situations may help viewers to face reality. Gurvitch
(1956/1973) referred to a sublimation of social situations, which makes it
possible for theater audiences to see new perspectives on social conflicts and
even to discover solutions.
All the views that attribute to art the power to teach the viewer something
about reality invariably call up the question of interpretation. Interpretation
is itself something that must be learned. It cannot be assumed that the nat-
ural viewer of the traditional film will automatically discover those hidden
meanings the critic is so adept at identifying. Insofar as they go unnoticed,
this cannot be said to be a motive for watching feature films. One may well
ask whether what art has to say to the critic or the expert is always under-
stood by the untrained viewer. The traditional views on art often imply that
there is a considerable reward in store for the beholder who is prepared to in-
vest a certain degree of effort. In return for the effort required to come to a
disinterested contemplation—a visit to a museum, say, and the perusal of cer-
tain background information—there is assumed to be some exceptional value
in the form of insights gained. On the other hand, it is argued, traditional
films, as products of the entertainment world, are able to reveal more simple
truths at a lower cost and therefore have a greater appeal for large groups of
viewers. Atkin (1985) reported the results of two surveys involving viewers of
soap operas who support this view. These subjects were convinced that the
information contained in soap operas is in some way useful in solving prob-
lems encountered in everyday life and in understanding reality and other peo-
ple. Cognitive learning effects become considerably more probable when they
do not pertain exclusively to the knowledge and information transmitted by
authorized institutions. Referring to a study focusing on the effects of radio
soap operas, McGuire (1974) affirms, "The sophisticated educator might
wonder whether what the soap operas teach is the best lesson; but from the
preoccupation with audience gratification that concerns us here, the impor-
tant point is that the audience believed to a poignant extent that they were
learning how to live and cope from these radio dramas" (1974, p. 182).
The lesson of the soap operas, or of any popular feature film need not be
new to serve as a motive: the possibility that the viewer is anxious to see ex-
isting views confirmed has already been discussed.

Enjoying Films as an Aim in Itself

Many of the motives mentioned thus far are concerned with a desirable
THE PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 31

state that is far removed from the film experience in the more restricted
sense. Such motives may be seen as secondary, in analogy to a well-known
distinction familiar from the psychology of reading (Purves & Beach, 1972).
Secondary motives can be realized comparatively easily in other ways than
by going to see a film. For example, the sense that one is part of a commu-
nity of like-minded individuals can probably be more easily acquired by seek-
ing the company of family and friends or joining in the social life offered by
clubs, societies, and cafes. Boredom can be dispelled by doing a crossword
puzzle or taking a short nap. But most secondary motives have to do with rel-
atively permanent desirable states, which, moreover, are not immediately at-
tainable. A strong ego or a well-defined social identity are not lost in the space
of a day, but then they can only be gained through years of experience.
Primary motives differ from secondary ones in that they can be immedi-
ately realized; they are often fleeting and are closely associated with the film
medium. The state of regression postulated by psychoanalytical theorists is
one example, affect regulation another. The most primary motives are by de-
finition concerned with the experience of the feature film as an aim in itself.
In this sense, they may be considered aesthetic. The experience of watching
can be a reward in itself, apart from the gratification of any need for preser-
vation and growth. Certain qualities of a work of art are in themselves ap-
pealing. To borrow from Aristotle, these range from proportion—for exam-
ple, arrangement, amplitude, unity, or plausibility—to a convincing portrayal
or imitation of people. 26 And then there is the enjoyment provided by spe-
cific filmic qualities. Some people find enjoyment in the view that films offer
of a particular reality, as stressed by the realism theorists. The viewer is in-
vited to observe reality in itself or reality as spectacle. The feature film also
offers viewers a unique opportunity to observe people in all their comings and
goings.
The so-called formative film theorists have described in detail the cine-
matic means by which the film experience is modeled, and which in their
view render that experience so pleasurable. Unlike the realism theorists, they
see the reality played out in front of the camera as no more than the raw ma-
terial that has yet to be processed, with the aid of filmic means. Arnheim
(1933) gave countless examples of the inadequacy of the film as an imitation
of the—perception of—nonfilmic reality. The degree to which these imper-
fections are visibly exploited rather than camouflaged determines the artis-
tic value of a film. Balazs (1938) saw the camera as the main resource the
filmer has available to recreate reality. In Balazs' view, traditional arts are al-
tar arts: they present the work of art as a closed microcosm that can be con-
templated from a distance. What is unique about films is the fact that the

26
The characteristics are from Aristotle (1972).
32 CHAPTER 1

viewer is given an opportunity to enter that microcosm and to identify with

the characters portrayed. Not only do we see what they see, we see how they
see it, which makes possible emotional identification.
A contemporary advocate of the formative tradition is Peters (1977, 1981,
1989), who analyzed the ability of what he calls the camera eye to transform
the reality it is filming. Camera eye is a comprehensive term that covers
not merely framing, camera positioning, and camera movement but also
the imagination created by such filmic means, an imagination that is envi-
sioned and experienced by the viewer. In the filmic imagination, the reality
depicted takes on certain qualities: aesthetic, emotive, persuasive, expres-
sive, and so on.
The views of Arnheim, Balazs, and Peters described here reflect the prac-
tice of the traditional filmmaker. In addition, the examples chosen by Arn-
heim indicate that cinematic art is by no means limited to the recognized
experimental and avant-garde genres, such as the work of Eisenstein and
Pudovkin or the form experiments of the Bauhaus filmmakers but is part and
parcel of quality movies, such as those made by popular Hollywood directors
like Chaplin, Lang, and DeMille.
As we saw during the discussion of survey research into the motivation of
film viewers, there is little or no material available from the viewers them-
selves on the appeal of specific filmic qualities. Although film scholars have
distinguished artistic or aesthetic dimensions in the motivation for film view-
ing, there are no data on the concrete qualities most appreciated by actual
film viewers. This is understandable, as natural viewers of the traditional fea-
ture film do not normally have at their disposal the necessary terminology to
describe these qualities. Research using more sophisticated experimental
techniques is needed to bring them to light.

Tension Reduction as the Main Primary Motive

There are two sources of primary satisfaction specific to the feature film: the
first is the fictional world depicted by the film; the second derives from the
technical-stylistic qualities of the medium. We examine first the motivational
effects of the fictional world.
Viewers may derive pleasure from the free and unencumbered contem-
plation of other people. We know from practical experience that the tradi-
tional feature film, regardless of the genre, gives viewers the impression that
they are seeing others without being seen themselves. Curiosity as to the
comings and goings of the characters is awakened and subsequently satisfied.
Also, watching a film evokes a filmic state that according to Metz (1975b),
can best be described as a "pseudo regression." In other words, the fact that
the members of the audience are fascinated by—or even lose themselves in—
the fictional world would appear to be one of the most important primary mo-
T H E PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 33

tives. The pleasure of observation and the pleasure of losing oneself in the
fictional world may be considered two sides of the same coin. Viewers so-
journ, in the imagination, in a fictional world where they run absolutely no
risk; their fantasy is both encouraged and directed. The cognitive basis for
this experience is the realization that one is in a fictional space.
For the moment we will refer to this compound primary motive, the de-
tails of which are covered in chapter 3, as safe involvement in the fictional
world, or simply involvement.
There is a second quality, complementary to involvement, which likewise
contributes to the appeal of films, a characteristic common to all forms of
fiction. It is important for a film to be understood not only as a representa-
tion of something else—in the case of the feature film, a fictional world—but
also as an independent construction or artefact. For the viewer, the most im-
portant elements of the film as artefact are plot and style. The plot is gener-
ally constructed in such a way that the viewer is presented with an often com-
plicated sequence of events in which the fictional action alternately pro-
gresses and stagnates. As a result, the viewer, who is constantly striving for a
more complete overview of, and a better insight into, the overall action, is
alternately frustrated and rewarded. The so-called mystery plot is the most
obvious manifestation of this construction, but all narrative films create un-
certainties that are put before the viewer at selected intervals and in selected
doses.
Viewers must grasp the surface structure of the film, those aspects that are
immediately observable, if they are to understand the plot. In some cases,
however, this surface structure is enjoyable in its own right. There may be a
special appeal in the way a certain scene is filmed, regardless of its impor-
tance for the development of the plot. Not only the specialist, such as the for-
mative theorist, but also the natural viewer may be motivated by the antici-
pation of a particular manner of filming. This last statement requires some
clarification. Unlike the specialist, the average viewer is not necessarily in-
trigued by the technical and stylistic details in themselves: the length of a
dolly shot, say, or the ingenious shift in focus within a single shot. But cer-
tain qualities of the immediate correlate to the film stimulus, the image, that
is, the impression that is the result of those technical manipulations of which
one is unaware, cannot fail to strike regular film viewers. I submit that style
and technique work, even when they are not perceived.
Nevertheless, the average viewers of traditional feature films may well be
aware of—and enjoy—certain technical aspects of films. Special effects, for
example, or the acting of a favorite star are undeniably pleasurable in them-
selves. Like Bordwell (1985), we assume that the stylistic characteristics in-
herent in the film technique make themselves felt because they form patterns
structured in time. Although the effects of these patterns generally go un-
noticed, they are occasionally recognized, as in the case of repeated or con-
34 CHAPTER 1

trasting camera positions, two-dimensional composition of the image, and the

mobility of the mise-en-scene. Such patterns of style enhance a film even
when viewers are not aware of them, and in some cases they are recognized
and enjoyed by a large portion of the audience.
The enjoyment of formal attributes is related to those motives that under-
lie games, as well as to a larger group of aesthetic experiences. The film ex-
perience is like a game in that the viewer, driven by cognitive curiosity, takes
pleasure in discovering order for its own sake and not for any use that might
be made of the knowledge thus obtained.
In psychological-aesthetical theories, finding order of some kind, abstract
patterns in systems of plot and style, such as repetition, contrast, symmetry,
parallelism, and so on, has been put forward as a possible motive. Arnheim's
(1971) theory of the structural order of works of art, to which we referred
earlier, is a case in point, as is Gombrich's (1960/1977) approach to the un-
derstanding and appreciation of representations. It is Gombrich's view that
no representation can be understood by the viewer unless some kind of ex-
isting schema is available within which elements of the work of art can be
placed.
It requires an effort to discover order. Within certain limits, this could be
a precondition for aesthetic enjoyment. Fechner (1876) defined beauty as
unity in diversity. Resolving inconsistency has often been described as the
cognitive basis of aesthetic experience (see Bever, 1986; Lasher, Carroll, &
Bever, 1983; Zusne, 1986). According to Mandler's theory of valuation, a
stimulus is valued more highly when a certain amount of effort is required to
find a schema to fit it (Gaver & Mandler, 1987; Mandler, 1982). And finally,
several of Berlyne's experiments provide additional insight into the exact re-
lationship between effort and pleasure. As complexity increased, a linear in-
crease in interest was established, as well as a reverse U-shaped appreciation
curve. The more difficult a stimulus is to process, the more effort people are
able to summon, but only up to a certain point. That effort is initially ac-
companied by pleasurable feelings, but these can become unpleasurable even
before the stimulus becomes so complex that one actually withdraws one's
attention (Berlyne, 1970, 1971, 1974).
Such formal attributes of feature films as plot and style, and their dynam-
ics and mutual interaction, cannot easily be captured in a single, unequivo-
cal term. Beauty sounds too formal for our purposes, but the term should be
mentioned here because it has often been used to describe the aesthetic emo-
tion that flows from the formal characteristics of a work of art, as opposed to
its contents. Here we will use the term appreciation of the artefact to refer
to the motive that consists in finding enjoyment in formal film characteris-
tics. Appreciation of the feature film as artefact can be measured. Our own
research has shown that film viewers can be divided along a dimension that
we call cinephilia. Thus far cinephilia has proved to be a good predictor of
T H E PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 35

preference for various film types that differ in the degree to which film style
plays a conspicuous role. The same results were obtained when other vari-
ables that influence one's taste in films, such as level of education, cultural
sophistication, and even film fan behavior, are kept constant (Tan, Egger-
mont, & Joosten, 1989; Tan, Eggermont, Joosten, & Spinhof, in prep.).
Obviously, films differ in the degree to which they satisfy each of the two
groups of motives: involvement and artefact appreciation. Some abstract film
forms present themselves almost exclusively as artefact; they do not provide
access to a fictional world in which viewers can lose themselves, at least not
in the sense of a fantasized presence in that world. As we have seen in chap-
ter 1, the traditional feature film is by definition good at realizing involve-
ment, which in general fits in with the motivation of its natural audience. In
the previously mentioned study of Tan et al. (1989), preference of the most
traditional type of feature film was accompanied by a fairly low level of
cinephilia. Nevertheless, we should not underestimate the appeal of the tra-
ditional feature film as artefact or the aesthetic motivation of its natural au-
dience, that is, the desire to enjoy the film as film.
Finally, if we were asked to reduce the various primary motivational at-
tributes of the feature film to the single most comprehensive motive, with the
smallest possible loss of information, then the answer would probably be
something like "tension reduction." By this I mean the regulation of affect on
the part of the viewer, a small-scale emotional catharsis, admittedly in the
most limited and specific sense of the word. A traditional feature film creates
a specific emotional tension but then goes on to resolve that tension. This has
been demonstrated most clearly by the work of Zillmann and his associates
discussed earlier, although in his experiments subjects were presented mainly
with television drama rather than feature films.
Each of the two clusters of primary motives discussed involvement in the
fictional world and appreciation of the artefact, makes its own contribution
to the combined effect of tension. The events in the fictional world create un-
certainty, which by the end of the film has been resolved. In general, any he-
donically negative affect has by then been reversed. At the same time, how-
ever, the initially chaotic structural organization of the artefact, in particular
its systems of plot and style, requires a mental effort on the part of the viewer
and creates a desire for order. Gradually the representation of these systems
does indeed take on a more orderly form and ultimately ends in a good
Gestalt. This most comprehensive primary motive for watching feature films
will hereafter be referred to as the creation and resolution of tension. Figure
2.1 summarizes the interplay of the various motives.
If the argument put forward thus far is sound, then the most important
motive for viewing feature films may well be of an affective nature. The most
important primary motivation lies in the expectation of undergoing a highly
specific emotional experience.
36 CHAPTER 2

EXPERIENCE FICTION

Imaginary sojourn
'In' fictional world

Safety

TENSION
CHEAT/ON
&
R£[){}(77()N

EXPERIENCE AR'JUACf
Appreciation
Comprehension of plot
of the
Stylistic features artefact

'Formal order'

FIG. 2.1. Summary of the most important primary motives for film viewing.

CONCLUSION

For a number of reasons, it is good to exercise a degree of caution in drawing

general conclusions from a discussion of different psychological functions. In
the first place, one of the major results of any investigation into functions is
that the relationship between motives and media usc is a complicated one,
which docs not lend itself to a simple and unequivocal interpretation. This is
doubly true if one is attempting to do more than establish certain correla-
tional associations. Fcnigstcin and Ileyduk (1985) pointed out that there arc
important interactions between motivation and use, notably, the vicious cir-
cle. Exposure to a particular theme in a program, such as sex, increases the
likelihood of fantasies and behavior related to that theme. Sexual fantasies
and active sexual interest then in turn increase the likelihood of exposure to
programs that may be expected to feature sexual themes. In the second place,
the perception that media users have of the gratifications obtained does not
necessarily reflect the gratifications they actually draw from that use (Davis
& Woodall, 1982; Palmgrccn, Wenner, & Rayburn, 1981). In the third place.
it is by no means certain that the findings of studies devoted to the function
of television drama are entirely applicable to the viewing of feature films. By
far the majority of the empirical data dealt with earlier were related to tele-
vision viewers, while the natural audience of the feature film only partially
overlaps the television audience. Moreover, the motivation to go to see a film
is no doubt stronger than the motivation to watch a television program. More
T H E PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 37

effort is involved, and if the film is disappointing, then the "cost" is higher. 27
While the choice of a particular title or genre is not always the result of a con-
scious choice (Ghoffray & Pas, 1980), people generally have a fairly good idea
of the kind of film they want to see or, conversely, the type of film they have
a n aversion to. There are, however, structural similarities between the two
media, especially when every effort is made to disregard the context, as we
have strived to do here.
Reviewing the body of insights and empirical studies, we may nevertheless
conclude that media use, notably television viewing but also film viewing,
may indeed be motivated. If we compare the cinema with the Chinese Room,
then the subject—the viewer—is anything but a tabula rasa. Even as they
take their place in the cinema, they have certain needs and desires that con-
fer an intentional significance on the stream of symbols passing before them.
The motives discussed include the need for entertainment, a temporary re-
treat into a less rational, dreamlike state, participation in a collective signifi-
cation, mood regulation, fantasizing, and various kinds of learning. The
viewer will in any case evaluate the incoming images according to their abil-
ity to evoke a desirable experience, namely a well-rounded emotional episode
characterized by the creation of tension, followed by a resolution of that ten-
sion. Everything that appears on the screen is assessed on the basis of this
criterion.
This concluding assumption is not based on a systematic evaluation of all
the theoretical and empirical data that are in some way related to the moti-
vation of the film viewer. On the other hand, my discussion does not differ
essentially from that of McGuire (1974). But unlike him, I have attempted not
so much to deal with all the possible psychological functions as to find an an-
swer on which to base our understanding of the emotion of the viewer. That
answer implies that people watch films because they expect to experience a
specific kind of emotion. I suspect that there are two other possible approaches
that would lead to a different concluding assumption.
The best known approach that deviates from this view is no doubt the the-
matic approach, which holds that films appeal to people because they are
about something, and that it is this that draws them to the cinema. In this
view, the film theme does not have to be exalted to qualify as such. The most
banal adventure film is about justice versus injustice, or friendship versus
treachery; the most overworked romance is about true love; while drama the-
orists have traditionally maintained that drama is always about a conflict of
some kind. However, the point stressed by the thematic approach is that a
specific content is more important than any creation and resolution of ten-
27
Fowles (1992) maintains that television viewing requires an activity that is closer to "dream
sleep" than to an active intake of information. There is also much less active attention involved
in television viewing, compared to the intense concentration that accompanies the viewing of a
cinema film.
38 CHAPTER 1

sion that may be provided. The answer to this counterproposal is that the
viewer does indeed strive to discover the theme or themes in a film, some of
which speak to the viewer directly and intuitively, through a kind of univer-
sal relevance. But the existence of such a theme in no way guarantees that a
traditional feature film will be a success in the eyes of its natural audience. A
Woody Allen film about the neuroticism of the New York in-crowd would not
be a good film—either in the eyes of the critics or those of devotees of the
traditional film—if there was no plot development. To be a success, the theme
must be embedded in a good story. 28 The reason that all feature films have
one or more themes is simply that themes make for good stories. They com-
bine a number of complications in the action with a number of possible re-
sults, each with its own wider import or morale; this makes them an emi-
nently suitable means of creating and later resolving tension. Themes serve
not only as a vehicle for the action but can themselves be manipulated in such
a way that they help to build up the tension. To mention just one example:
the famous shower murder in Psycho (1960) shows not only violence but also
punishment, thanks to an ingenious editing (Perkins, 1972). There is even
something to be said for the view that what a film is about, and the beliefs
that it propagates, are in effect uncertain and preliminary. This notion was
recently advanced by Branigan (1992), who maintains that the meaning put
forward by the fiction film is not characterized by truth or untruth but by
what he calls partial determination. Reference is established in the course
of a process of continuous improvement in which the subject takes an active
part. This does not necessarily lead to a total specificity of meanings. Brani-
gan refers explicitly to the affective capacity of the preliminary nature of the-
matic meaning: "Thus indefinite reference does not mean that we can't have
specific and intense emotional reactions to fiction; quite the contrary, indefi-
nite reference may facilitate such reactions." (Branigan, 1992, p. 195).
In unfolding themes, the feature film kills two birds with one stone: not
only is it entertaining, but it can put forward something that goes beyond the
value of a brief experience. In a film not all subjects can be dealt with in such
a way as to do justice to their complexity and relevance on the basis of some
absolute criterion. 29 But this does not mean that fiction films cannot deal with
important questions, in particular, when the importance of a theme is mea-
sured by the needs of a particular type of audience rather than by absolute

28
Even in the documentary, the genre best suited to the presentation of a theme as aim in it-
self, narrativization is almost universally employed to hold the viewer's attention.
29
A neat illustration of this was given by Dudley Andrew, who wryly noted that the so-called
leftist film is merely yet another genre of the popular film. He took the example of the work of
Costa-Gavras: "Viewers are led to identify with spectacular heroes incarnated by noted actors
such as Yves Montand. Seeking to reach the largest possible audience, Costa-Gavras necessarily
subjects the spectator to that state of childish wonder fostered by the readerly text. The assur-
ance of the form contradicts the message of alarm and outrage that these films presumably want
to transmit" (Andrew, 1984, p. 122).
T H E PSYCHOLOGICAL FUNCTIONS OF FILM VIEWING 39

criteria. (See McGuire's quotation earlier concerning the lesson of the soap
opera.) And it may well be that the more importance the viewer attaches to
a particular theme, the more effectively it can be used in a film story to cre-
ate and ultimately resolve tension. Examining a more fundamental truth, un-
earthing a secret fear, revealing some unsuspected possibility: these are some
of the powers of complex themes that are eminently suited to evoke involve-
m e n t and tension, although they admittedly make higher artistic demands
(see again the example of Perkins earlier). In general, satisfaction of the pri-
m a r y motive of tension reduction does not rule out the gratification of other
motives, ranging from affective preservation to cognitive learning. What
other psychological functions can be realized by a particular film, and to what
degree, will depend to some extent on the genre.
A second approach that appears to run counter to what we have main-
tained is a more socially relevant one. Some readers may be disappointed that
wherever possible I have dealt separately with the functions of a visit to the
cinema and the pleasures afforded by the film itself, and that the former have
not been given a great deal of weight in my considerations. However, a num-
ber of the motives we have dealt with here do have social relevance. To take
two examples, the desire to see one's norms validated and one's participation
in a collective identity reinforced are the psychological counterparts of rec-
ognized social functions: socialization and the promotion of cohesion. In the
viewpoint that I have put forward here, however, these social functions are
not linked to the action of going to see a film but rather to distinguishable el-
ements within the fictional film itself. Not all films have the capacity to grat-
ify social motives. Popular action films and comedies actively steer the bias
of the viewer, which tends to result in a uniform vision on the part of the au-
dience concerning the behavior, motives, and character of the characters por-
trayed. These are the necessary conditions for active and overt participation
in those visions: additional enjoyment by laughing—or booing—together.
Other traditional film genres, such as the psychological drama—Mankiewicz'
The Barefoot Contessa, for instance—are much less likely to give rise to such
reactions. Yet, the most important motives, involvement and tension reduc-
tion, are primarily psychological, despite their significant social implications.
The claim that these psychological motives influence the way viewers watch
all traditional fictional films is the major empirical claim put forward in this
chapter, and it will have to be tested by means of psychological research.
From the perspective of the viewer, it could be said that what all natural
viewers of the traditional feature film have in common is their desire to ex-
perience emotion as intensely and as abundantly as possible, within the safe
margins of guided fantasy and a closed episode. The degree to which other
demands are made on the fulfillment of motives that are related, on the one
hand, to the precise contents of the film and, on the other, to a possible col-
lective experience will vary according to the particular genre and the corre-
sponding subgroup of the larger audience of the traditional feature film.
 Film and Emotion:
3 Theoretical Background

THE PRAGMATIC FRAMEWORK

OF THE EMOTION EVOKED BY FILMS

As we saw in chapter 2, one of the major incentives for watching feature films
is the emotional experience they offer. However, not all emotions will be wel-
comed by all filmgoers at any arbitrary point in the film. There are certain—
largely unwritten—rules that limit the emotional effect produced by films. To-
gether they form what might be called the pragmatics of the feature film. Like
the rules that govern conversational speech acts, they are derived not so much
from conventions as from a historical development within a particular con-
text; in this sense they ultimately come to be seen as natural.
The fundamentals of linguistic pragmatics, as developed by such theorists
as Grice (1975, 1978), are valid for everyday conversation and language use
in the more common interactive situations. They are, however, not easily trans-
ferred to more complex domains, such as literary discourse (Lanser, 1981;
van Dijk, 1981) and the feature film. The main problem is that the concept
of efficient communication, which is of prime importance here, is fairly easy
to describe in purposeful interactive situations, but much more difficult to
characterize in interactions that are not of direct use. As a consequence, it is
very difficult to grasp the codes that coordinate literary or filmic communi-
cation (Fokkema, 1989). In a situation where A is explaining to B how a cer-
tain piece of machinery works, we can assume, as indeed Grice did (1975),
that there is a cooperative principle active that generates certain simple max-

41
42 CHAPTER 1

ims, such as quantity, sincerity, relevance, manner, and clarity. The common
aim of A and B is to see that the working of the machine is made clear. Yet
such maxims are in themselves unable to explain even the simplest frills of
everyday conversation. It is only with the greatest difficulty that such com-
mon phenomena as irony and jokes can be made to fit into the pragmatic
framework referred to above. (Leech, 1983). 30 And if we were to apply the
maxims to the feature film in some overall way, the result would probably be
as dry as dust. Conversely, there are pragmatic aspects of good film narra-
tives that are in direct conflict with Grice's maxims. A well-known example
is the so-called unreliable narrative, 31 where it later appears that a particular
scene created a false impression of actions crucial to the plot.
For an adequate translation of the maxims, the cooperative principle must
be formulated differently. First, in the case of the traditional feature film—
and the qualification traditional must be stressed—the result might be some-
thing like the film viewer is always right and may expect to be entertained.
Be entertaining could then replace the maxim Be informative, which com-
bines the quantity and relevance maxims. If a certain film passage (e.g., an
exposition) is slightly dull, the viewer accepts this as a necessary investment,
one that will be richly rewarded later on. Second, cooperation means that it
is the duty of the filmmaker to make the film interesting. Although the viewer
plays an active role, that role is guided by the film. This trust on the part of
the viewer is the counterpart of the original sincerity maxim. 32 Third, al-
though the genres differ somewhat in this respect, the viewer expects the film
to be interesting, without being addressed personally by the artefact. The film
creates for the viewer a safe involvement by entertaining discreetly, by pre-
senting emotionalizing events that take place in a fictional world; the viewer
is no more than a witness to things that happen to others—and fictional oth-
ers, at that. This may be seen as a politeness maxim: a film is a fantasy, guided
by a firm but discreet hand. In the following sections we look at the qualities
that the fictional world must display in order to give rise to real emotions,
while at the same time guaranteeing protection against effects that are too
extreme to be truly entertaining. In the fourth place, according to the man-

30
Pratt (1977) discussed in detail the relationship between the pragmatics of conversation
and literary discourse, also focusing on the shortcomings of speech act theory for ordinary con-
versation.
31
See Buckland (in press) and Chatman (1978, p. 235-237). According to Bordwell (1985),
the unreliable narrative is relatively uncommon in classical cinema films, at any rate compared
with the frequency with which the phenomenon is used in novels.
32
This does not mean that the attitude of the viewer is naive; he or she may be compared to
the "sophisticated reader" (Kay, 1983) who, at least to some extent, anticipates the tricks and
ruses of the filmmaker and, looking back, can appreciate them. In fact, as we shall see, this is
one of the conditions that must be met if the film is to be interesting.
FILM AND EMOTION: THEORETICAL BACKGROUND 43

ner maxim, the film may not display any ambiguity or confusion unless these
are necessary in order to heighten interest.
These maxims should be seen as limiting rather than generative. They can-
not generate good films, although departures may result in serious anomalies.
There are, however, countless ways of remaining within the pragmatic bound-
aries of traditional cinema. Both laughter and tears can be entertaining: the
power to entertain can be derived from the ability to endear, to move, or to
astonish. Thus it is possible to distinguish various types or genres of films, ac-
cording to the nature of the events in the fictional world that they portray and
the characteristic viewer's emotion that they evoke. All genres still remain
within the boundaries of the maxims.

WHAT IS AN EMOTION?
FRIJDA'S THEORY OF EMOTION

The Functionality of Emotion

The claim that watching films evokes genuine emotion presupposes the exis-
tence of a generic description of emotion. What is felt by the viewer can only
be said to be an emotion if it can be shown to be a special variant of that gen-
eral emotion. Although in recent years much of the literature has focused on
specific aspects of the subject, there are few general theories of emotion. Inte-
grative overviews are to be found in Frijda (1986), Izard (1993), Ortony,
Glore, and Collins (1988), and Strongman (1978). As pointed out in chap-
ter 1, the assumption that emotion is functional may be an appropriate cri-
terion by which to distinguish it from other states. The idea that emotions
serve some purpose, and that it is this that determines their structure and ef-
fect, was put forward by Darwin, who stressed the communicative function
of emotion. The functional viewpoint has been elaborated by Plutchik (1980,
1991), who discusses the evolutionary significance of emotion, and Lazarus
(1991), who dealt with the role of emotions in an individual's adaptation to
the environment.
An important standpoint within the functional viewpoint maintains that
the main task of emotion is the control of cognitive processes. Some emo-
tions arise out of acute situations, for example, the fear you feel at the sight
of a car that is out of control and racing toward you on the wrong side of the
road. Here the emotion is functional in that it takes control, precluding ex-
tended reasoning: a decision to take action—any action—is better than fail-
ing to act in time. But in nonurgent situations, too, the control of cognitive
processing by the emotions may be seen as functional. De Sousa compared
the emotions involved—for example, grief and envy—with "judgments, in the
44 CHAPTER 1

sense that they are what we see the world 'in terms of'" (De Sousa, 1987,
p. 196).

Relevance, Concern, and Emotion

Of the functional theories, that of Frijda (1986, 1988, 1993) is the most com-
prehensive. Characteristic of his theory is the standpoint that emotions con-
trol not only cognition, in approximately the same sense as that used by De
Sousa, but also action. According to Frijda, emotion regulates the interaction
between the individual and the environment in that it is directed toward the
realization of what is of importance to the individual, that is, his or her con-
cerns. Without concerns, there can be no emotion; conversely, emotion sig-
nifies that some concern of the individual has been affected. Frijda (1988)
referred to the law of concern. Instances of such concerns are elementary
needs, stable motives, norms, and attitudes. In a more general sense, con-
cerns may be seen as universal preferred states, such as having enough to eat
and drink. But these concerns also include preferred psychological states,
such as a sense of security, experience that is sufficiently diverse, and cog-
nitive mastery of situations with which one is confronted.
Thus the briefest description of emotion is "a concern realization or sat-
isfaction system" (Frijda, 1986, p. 478). The emotional system is geared to-
ward establishing the relevance of certain situations for the concerns of the
individual and, if such relevance exists, to enforce the priority of cognition
and action in accordance with those concerns. Emotion as process is ac-
companied by a specific way of attributing significance to the situation, as
well as control precedence for certain action tendencies. In the latter case,
this means that a tendency to do or not to do something presents itself and
persists, even if it runs counter to other behavior that has already been initi-
ated or a different type of ongoing mental activity.

Situational Meaning Structure and Emotion

Frijda's theory is one of a number of cognitive theories of emotion (Arnold,
1960; Averill, 1982; de Rivera, 1977; Ellsworth, 1991; Lazarus, 1966, 1991;
Mandler, 1984; Oatley, 1992; Ortony, Clore & Collins, 1988). The one pos-
tulate that they share is that some form or other of cognitive activity medi-
ates between the stimulus and the emotional response. This distinguishes
them from those theories that stress that the origin of emotions is to be found
in neural (Gray, 1982), motivational (Tomkins, 1962), or sensorimotor sys-
tems (James, 1890/1950). 33

33
This division is derived from Izard (1993). None of the alternatives excludes the effect of
cognitive factors.
FILM AND EMOTION: THEORETICAL BACKGROUND 45

The emotional process may be said to have an appraisal dimension and an

action dimension, linked to the stimulus and the response respectively. The
result of appraisal is an appreciation—not necessarily conscious—of the situ-
ation in terms of personal significance. The various emotions are character-
ized by different appraisal and action patterns, which are in principle unique
to that emotion. (Frijda, 1986, pp. 204-221, for the components of situa-
tional meaning structure and an overview of various structures and accom-
panying emotions.) For instance, sadness is the experience of the absence of
a valued object and the knowledge that the possibilities for regaining that ob-
ject have been exhausted (Frijda, 1986). In the action pattern, the loss of ac-
tion tendencies is paramount. Guilt is linked to the realization that one has
caused another harm; the action tendency is undoing the harm (Frijda,
1993). If it is to give rise to an emotion, the appraisal of a situation must con-
tain certain critical meaning features. This is the law of situational eaning
(Frijda, 1988). Conversely, it is possible to deduce from the emotion the per-
sonal significance that the situation has for the subject. Critical features of
situational meaning structure, known as core components (Frijda, 1986), in-
clude relevance, valence, reality, and difficulty. Relevance is related to the
involvement of concerns: the chance that one will experience sadness when
a loved one is seriously ill is considerably greater than when a stranger is ill.
Valence refers to the intrinsic value of a situation in terms of a concern. Re-
ality has to do with the encoding of the stimulus: we are not afraid when we
know for sure that someone is only pretending to threaten us. Difficulty refers
to the presence of an obstacle to the attainment of a desired situation. Frijda
distinguished not only core components but also context components, which
determine the exact quality of the emotion. These are concerned with such
things as the potential for action offered by the situation. Each situation is
classified according to the capacity for action that it offers: an impending neg-
ative event that is unavoidable does not produce the same emotion as one
that is only a possibility or that can be actively prevented. In the same way,
the emotion evoked by a negative event that was deliberately caused by some-
one differs from a negative event that occurred accidentally. Urgency, seri-
ousness, and intentionality are the products of this aspect of appraisal.
The action tendency that is part of the emotion is closely related to the ex-
tracted relevance and meaning. Survey research has shown that action pat-
terns can be predicted on the basis of appraisal patterns. Frijda (1987) and
Frijda, Kuipers, and ter Schure (1989) reported considerable multiple cor-
relations between appraisal and action readiness scores. In the case of anger,
the meaning of the situation is that an agent is held responsible for damage
to concerns; the action readiness is likewise an urge for revenge. Fear con-
sists in an urge to run away in response to a threat to one's concerns. Action
readiness may be general, as in excitement, or specific, as in fear. In other
words, the tendency to act consists of an inclination to initiate, maintain, or
46 CHAPTER 1

alter a relationship with the environment or, alternatively, a readiness to un-

dertake specific activities that satisfy a certain concern. This latter variant of
action readiness is characteristic of a special class of emotions, namely, de-
sires and enjoyments. Action readiness also has cognitive forms, such as, in
the case of interest, the tendency to devote attention. It would perhaps be
more correct to say that emotion is accompanied by a change in action readi-
ness rather than by an action tendency. Some emotions, such as surprise and
sadness, are clearly characterized by a reduction in the action tendency. Nev-
ertheless, for purposes of simplification, we use the term action tendency,
except where this might lead to confusion.

Definition of Emotion
An emotion may be defined as a change in action readiness as a result of the
subject's appraisal of the situation or event. To avoid any possible misunder-
standing, a few preliminary remarks are in order here.
First, the description of emotion may suggest that each part of the emo-
tional process can be neatly separated from the others. The term emotion, in
the sense of the process, refers to the whole body of psychological events,
from appraisal to a change in action readiness, and it is not always possible
to separate the various stages of the process. Thus the appraisal of possibili-
ties for action and the generation of an action tendency may overlap. Fur-
thermore, causal relationships between various stages may work in two di-
rections: for example, through feedback, the action tendency may reinforce
a certain appraisal.
Second, the impression may have been given that emotions are totally
blind. They are not, and regulatory processes are capable of influencing each
of the subprocesses of an emotion. Once evoked, emotions have a tendency
to take off, counter to all kinds of resistance. They are also relatively inac-
cessible to other cognitive processes. 34 But an emotion of any kind also evokes
caution, which is itself an emotion. The law of care for consequence (Frijda,
1988) holds that a tendency to change the response is called up in the wake
of any emotion, generally in the form of inhibition. The stimulus for inhibi-
tion consists of the anticipation of consequences, often associated with sig-
nals from the environment, such as disapproving glances from others. As a
result of the persistence of emotion, regulation is often indirect and incom-
plete and, in some cases, is unable to prevent negative consequences.
Third, the description of the cognitive process of appraisal must be refined
somewhat in regard to the question of the possibility to regulate emotion. Frijda
(1993) clarified the distinction between primary and secondary appraisal.
Primary appraisal is an advanced automated process whereby basic elements
of situational meaning, such as intrinsic hedonic quality, agency, and causal-

34
They are "eognitively impenetrable" (Pylyshyn, 1984).
FILM AND EMOTION: THEORETICAL BACKGROUND 47

ity are determined. Primary appraisal is immediate, often requiring no more

than a simple comparison between two basal representations, say, perceptual
characteristics on the one hand and possible actions on the other; these are
described by Gibson (1979) as immediately observable affordances. The re-
sults of primary appraisal are directly related to the stimulus and may be seen
as antecedents of the emotion. Primary appraisal triggers continued cogni-
tive elaboration; this requires more inferences, in some instances whole se-
ries of inferences. These cognitions are often available more or less in their
entirety in the form of affect schemas (Fiske, 1982). Secondary appraisal is
accompanied by emotional significance that is not directly evident from the
situation itself. An example of primary and secondary appraisal can be clearly
seen in the study of the more subtle emotions, such as guilt. Frijda (1993)
demonstrated that the feeling that one is responsible for harm done to some-
one else is not part of primary appraisal. It would appear that primary appraisal
includes only the realization that one has in some way caused that harm. It
is only in the later stage that one is conscious of a sense of responsibility.
And fourth, the earlier account of Frijda's theory does not do justice to the
role played by experience. Emotional experience is, on the one hand, the re-
alization of the significance of the situation as it has been appraised. In the
case of fear, for instance, the situation is experienced as something that
threatens a concern. On the other hand, emotional experience consists in the
action tendency, for example, the urge to take to one's heels. The emotion
cannot be distinguished from the emotional experience. Thus we can add to
the definition given earlier an experiential component: emotion is the con-
sciousness of a change in action readiness, which is experienced as motivated
or caused by situations that have been appraised in a specific manner (Frijda,
1986).
There is a distinction to be made between primary appraisal and sec-
ondary appraisal in the manner in which they are experienced. Frijda (1993)
showed that primarily appraised elements are less conscious than the more
detailed secondary elements.
We now attempt to interpret the characteristics of the traditional feature
film in the light of the general theory of emotion. In doing so, we follow
Frijda (1988) in his treatment of the essential concepts of that theory namely,
concerns, emotional stimulus, appraisal mechanism, action readiness, and
regulation.

THE LAW OF CONCERN, EMOTION,

AND FILM VIEWING

If it is true that feature films evoke genuine emotions, then they also touch
upon concerns. Traditional feature films call up characteristic emotions,
which generally fit one of the film genres. When we look at a larger group of
48 CHAPTER 1

viewers, the nature of the emotion appears to be fairly uniform: the audience
laughs at Buster Keaton and cries at the death of Jennifer in Love Story
(1970). Thus it may be argued that the traditional feature film, which is in-
tended to evoke homogeneous reactions, appeals to concerns shared by most
or even all individuals, at any rate in Western culture. If this is true, then af-
fective reactions to films are rooted not so much in idiosyncratic sensitivities,
such as the traumatic memories of a personal life event or associations with
a person close to the subject, as in their ability to appeal to more or less uni-
versal concerns. Admittedly, idiosyncratic concerns play a role, but they are
not so important as to endanger the relative homogeneity of the affect
evoked.
The concerns that may be touched on by watching a film are many and
varied. In chapter 2 we attempted to bring some kind of order out of chaos;
we referred there to motives and functions, because the exact meaning of the
concept of concerns had not yet been touched on. Frijda (1986) distin-
guished between source concerns and surface concerns. The first group is re-
lated to the preferred states of the subject, such as security and sexual grati-
fication. (For a discussion of a number of source concerns, see Frijda, 1986,
pp. 344-359.) The second group has to do with specific objects and aims. 35
The distinction between the two types of concerns is important for our dis-
cussion of film emotion. The feature film concretizes source concerns, and
in so doing activates a variety of surface concerns. When we watch a film, our
general interest in the fortunes of our own loved ones and friends takes the
form of sympathy with the fate of a particular character or characters. In the
same way, it is our general sense of justice that underlies our hope for the tri-
umph of this hero and the defeat of this villain.
The most important surface concern realized by the feature film as a
whole, namely, the creation and resolution of tension (see chap. 2), springs
from various source concerns. In the first place, the tension episode evoked
by the film fulfills viewers' needs for a sensation and an experience that they
are just able to cope with (Zuckerman, 1988) and that at the same time forms
no threat to their safety concern. In the second place, tension is related to
cognitive uncertainty: What's going to happen next? Thus cognitive assimi-
lation is another concern that plays a role here. In the third place, tension is
heightened by sympathy, a concern that has thus far received little mention
in the literature on film and television viewing. In Frijda's discussion it is seen
as an altruistic concern: sympathizing with and caring for others (Frijda,
1986). The fact that most, if not all, traditional feature films create tension
by introducing a sympathetic protagonist faced with one or more problems
suggests that this concern is in some way realized by the act of watching films.

35
Note that there is a parallel of sorts here with the distinction between primary and sec-
ondary functions, as discussed in chap. 2; the two pairs of concepts, however, are not identical.
FILM AND EMOTION: THEORETICAL BACKGROUND 49

There is reason to believe that the effect of at least some of the traditional
feature films goes beyond the cycle of tension followed by the resolution of ten-
sion. They also present a vision of reality, or a possible way of life, that would
be more difficult to portray in a verbal or discursive manner. Most of the sec-
ondary motives referred to in chapter 2 may be seen as similar to the source
concerns under discussion here. The motive of ideological reinforcement or
attitude confirmation may be seen as a satisfactory condition for Frijda's
value concern or familiarity concern. The motive of social participation and
identity satisfies Frijda's proximity and coherence concern.
All the concerns mentioned thus far are addressed by the events taking
place in the fictional world. The film narration mobilizes the surface concern
in question: concrete expectations and sensitivities with respect to the action
and the characters in the film one is watching. But there are other concerns
as well, which have to do with the film as product, such as a system of film
techniques. Although in the traditional feature film these are subordinate to
the fictional plot, the natural viewer of this type of film will have certain pref-
erences pertaining to the film as artefact. The well-known phenomenon of
movie stars, which has been extensively documented by a number of re-
searchers (Dyer, 1979) is proof enough that the viewer is interested in the
film as artefact.
To some extent, the emotions evoked by the film as artefact are rooted in
the cognitive assimilation concern, which, according to Frijda, often plays a
role in aesthetic emotions. The ability to fit the stimulus into existing sche-
mata and structures and to respond to it in a new way is a preferred state, in
other words, a source concern. The source concerns proximity and coher-
ence are also satisfied. The active anticipation of structural developments and
the synthesis or unconscious imitation of mobility, camera movement, say,
produce a pleasant involvement with the artefact, a kind of identification with
the film that is experienced in place of or alongside the viewer's identifica-
tion with the fictional characters. Moreover, it is possible to attain compe-
tence (White, 1959), whereby the viewer anticipates a particular develop-
ment in the technical means (sees it coming, as it were). This might be the
first bars of the theme music or the switch to a different camera angle. We
know from research by Lynch (1972) that even viewers with no background
or training in film analysis have a certain intuition in this respect that is
largely correct.
Finally, almost any film characteristic is capable of becoming a surface
concern. Films are made to appeal to audiences who over the years have fol-
lowed the development of certain features and who welcome something that
is new but nonetheless rooted in the familiar. A particular film may be a suc-
cess because it takes the genre just that little bit further than its predecessor
in spectacular special effects, say, the staging of dance numbers or the cut-
ting rate during action scenes. The cinematic taste of viewer groups is molded
50 CHAPTER 1

by ongoing developments in their preference for particular film characteris-

tics. This type of surface concern has everything to do with the differences
among various categories of film viewers. It has been noted that having a par-
ticular emotion is itself a surface concern and is part of the film taste. The
type of emotion that one prefers—or abhors—and the intensity of that emo-
tion are guided at least in part by personal preferences. Some viewers enjoy
the feeling of chills running down their spine, others prefer commiserating
with the fate of a helpless victim. Younger audiences tend to prefer stronger
emotions, whether they have to do with tension, humor, or romantic feelings.
These variations in viewer preference are due in part to the fact that one and
the same source concern, the evocation of emotion through the action of fic-
tional characters, affects different people in different ways.
Even within the context of a single film, sensitivities are awakened, and
recognition and appreciation triggered, if formative film theory has any rele-
vance for what appeals to ordinary film viewers (see chapter 2). Feature films
are chock full of motives, elements that capture the audience's attention and
appeal to viewers by simple repetition. 36 Not only the props and behavioral
quirks of a character—the great comics are a case in point—but also unob-
trusive stylistic devices, such as a camera movement, a still life shot in be-
tween two action scenes, or a special montage sequence, are aspects that
viewers look forward to or that afford them unexpected pleasure.

THE FEATURE FILM AS

AN EMOTIONAL STIMULUS

Subject Matter
The emotional stimuli of feature films can best be characterized by compar-
ing them with an imaginary cross-section of episodes from the mundane re-
ality of the everyday life of everyday people. While situations of high drama
do occasionally occur, in the more prosperous countries of our world it is not
often that the essential concerns of average individuals are directly addressed.
Although there is no quantitative measure by which to assess such a state-
ment, it would seem reasonable to say that, relatively speaking, feature films
contain an unusually high concentration of emotional stimuli.
To a large extent, film events that address concerns evoke emotions by
means of characteristics that lend a certain affective potential to comparable
events in daily life. In the feature film these characteristics are both concen-
trated and magnified, not only with respect to the technical realization but

36
For a broad definition of this notion, see Bordwell & Thompson (1986, p. 37).
FILM AND EMOTION: THEORETICAL BACKGROUND 51

also the fictionalization that precedes it. If the reality of everyday life is the
measure of the events taking place on the screen, then it could be said that
the emotion evoked by something of which we have n o experience at all in
our daily lives, but that is just conceivable, represents not so much the com-
parative as the superlative degree of the comparison. No one has ever met a
unicorn or a dragon, and this explains why we are overawed by them. Flies
and human beings are familiar to everyone, but the idea of a man turning into
a fly verges on the incredible. Fantastic film events tread a thin line between
the familiar and natural and the unfamiliar and supernatural. 37 In the feature
film, things we have only heard of are acted out before our very eyes; a sud-
den rise to riches, prominence, or genius is something you hear about but
never experience. Even less spectacular developments are blown up until they
reach excessive proportions. In the same way, occurrences that one ordinar-
ily encounters only once or twice in a lifetime, such as the fatal illness of a
loved one, suicide, a sudden and drastic reversal of fortune, serious crime, a
fire or flood, a passionate love affair, and a meeting with an exceptional hu-
man being, in the film world are more or less a daily occurrence and within
a time frame of an hour and a half. Most events on the screen may be seen
as condensed or exaggerated versions of reality, in the sense that the likeli-
hood of their actually taking place in the real world, simultaneously or in se-
quence, in the presence of one and the same observer, ranges from zero to
minimal, even if they are among the best-known life themes. 38 This form of
statistical improbability is what we mean when we say that film portrayals are
larger than life.
The relative rarity of the situations depicted in films does not extend to
the domain of fantasy. In daily life people fantasize about precisely those
themes that are portrayed in films. While the fantasies people already enter-
tain may be unusual and statistically unlikely, they do appear to display a cer-
tain degree of reality. For example, the fear of crime, especially serious and
violent crime, is exaggerated if one considers the actual chance of becoming
the victim of such a crime. But anyone who thinks it is easy to talk a person
out of such a fear is quite mistaken (van der Wurff, 1992).

37
Carroll (1990), following Todorov, refers to "fantastic hesitation." He elaborates on this no-
tion, using an example from Cat People (1942), in which the character Irena is able to turn into
a cat. Although this is what the audience is led to believe, it is not actually confirmed until close
to the end of the film. All that time the supernatural interpretation is at odds with the natural
interpretation.
38
According to this assumption, the elderly are more likely than young people to recognize
events that they themselves have experienced at some time in their lives. There are indications
that this could indeed be the case. Older adults tend to find fictional television programs more
realistic than young people do. According to Potter (1988), the empirical evidence on this point
is contained in at least six studies.
 52 CHAPTER 4

The Diegetic Effect

In principle, the control that the film narrative has over the stimulus is com-
plete, far more complete than anyone seeking emotions in real situations
could ever hope to attain and generally more complete than in fantasy. In a
manner of speaking, the control of the stimulus goes as far as the psycho-
logical researcher could wish for in an ideal emotion experiment. (In fact,
films are often used in emotion experiments.) The feature film has an enor-
mous potential for manipulating the meaning of events taking place. There-
fore, emotions may be said to be controlled as to strength and quality, for
each emotion has its own unique situational meaning structure (Frijda, 1986;
see also Frijda, 1988 for the Law of Situational Meaning). The effectiveness
of the stimulus derives, first, from the fact that much of the processing by the
subject has been made largely unambiguous, easily comprehensible, or even
superfluous; the stimulus is preprocessed in accordance with certain emo-
tions. And second, a certain development of the stimulus over time deter-
mines or optimalizes emotions according to their nature and intensity.
The perceptual and cognitive basis for the inevitability of situational mean-
ing is that the feature film creates the illusion of being present in the fictional
world. In film theory this is known as the diegetic effect. Burch (1979) re-
ferred to the "experience of the fictional world as the environment." 39 "[The
history of the ] basic mode of representation in the cinema of the West points
to an all but universal tendency within the dominant cinema (as well as the-
ater and literature) to maximize and generalize the diegetic effect" (p. 19).
The scope of this tendency is considerable: it is fully applicable to the classi-
cal Hollywood film narrative, and thus to what we refer to as the traditional
feature film, although there are differences in the degree to which this is true
of the various genres. Burch believes that the effect is equally strong in the
films of the Italian neorealists and art films such as the later work of Fellini.
In his view, it is still "intermittently operative for many spectators of Michael
Snow's (1967) Wavelength or Jean-Luc Godard's (1966) Deux ou Trois
Choses Que Je Sais d'Elle" (Burch, 1979, p. 19). The diegetic effect is sup-
ported by a more general one, observed in studies of the perception of pic-
tures. It has been noted that monocular perspective in pictures, such as paint-
ings and photographs, draws the beholder in a position that is defined in
relation to an imaginary space behind the window formed by the picture
plane and the frame. (See Panofsky, 1927/1991. The point has recently been
made in a very convincing manner by Kubovy, 1986). The perspective illu-
sion is very robust. If the beholder's vantage point does not correspond to the

39
In this context I disregard the distinction between fiction and diegesis, because it would not
help to clarify the point at issue here. Burch's (1979) exact words are "diegetic effect, whereby
spectators experience the diegetic world as environment" (p. 19).
FILM AND EMOTION: THEORETICAL BACKGROUND 53

center of projection (located in the real space before the picture plane), he
or she does still experience depth in the representation, and the layout of the
space beyond the picture plane is perceived as if the beholder is looking from
the center of projection. The awareness of the picture as a two-dimensional
object, or even as an artefact, may recede to the background to a consider-
able degree, in favor of the illusion of looking into the virtual space.
In terms of emotion theory, the diegetic effect implies that the situational
meaning structure in film viewing is related primarily to the situation in the
fictional world; the fact that one is watching an artefact is of lesser impor-
tance. In the traditional feature film, the film as artefact is hidden from view.
The illusion of a magic window through which one observes another space is
particularly strong in the cinema. 40 The classical narrative is above all covert,
whereby the diegetic world remains as intact and solid as can be (Bordwell,
1985; Bordwell, in Bordwell, Staiger, & Thompson, 1985).
The contemplation of the artefact is almost entirely replaced by the sen-
sation that one is not only faced with, but also literally—indeed, physically—
present in the fictional world and is witnessing the action taking place around
one. Gibson (1979) formulates this point as follows: "We are onlookers in the
situation, to be sure, not participants, but we are in it, we are oriented to it,
and we can adopt points of observations in its space" (p. 298).
The primary basis for the diegetic illusion is the fact that the camera is mov-
ing within the scene. The cinema adds motion to the perspective illusion, and
motion is capable of reducing the distance between the viewer and the ob-
jects in the imaginary space. Among the classical film theorists, Balazs above
all has stressed that motion distinguishes cinematography from the other
arts, which tend to keep the spectator at a distance from their altar (Balazs,
1938). Bordwell (1985) makes the same point, using slightly different word-
ing: "Films—above all, classical films—allow the spectator a privileged posi-
tion within the fictional world" (p. 161). "The camera not only becomes the
storyteller but the viewer as well; the absent narrator is replaced by the 'ideal
observer'" (Bordwell, in Bordwell, Staiger, & Thompson, 1985, p. 37). 41 View-
ers experience the fictional events as if they were happening all around them;
the events appear to be real, concrete, and taking place in the here and now.
This physical immediacy forms the basis for the situational meaning.
40
The human visual system appears to be extremely tolerant for deviations between the cen-
ter of projection and the seating position vis-a-vis the screen. Cutting (1987) observes that in the
cinema no viewer can be at the center of projection, because the projector is at that location.
Of course the problem of deviation is most severe for those seated in the front row, side aisle. It
would seem for instance that when seated in this position, we should perceive rotating objects
as deforming. This is hardly so. Cutting's experiments suggest that extensive mental computa-
tion by the viewer (e.g. using some kind of Euclidean rectification to restore the true screen
slant), is unnecessary in order to preserve rigidity of objects in motion.
41
He uses the term privileged with certain restrictions; these have to do with the other nar-
rative principles, such as the planned dosage of information.
54 CHAPTER 4

Another significant quality of the situational meaning structure that flows

directly from the diegetic effect is the fact that one personally witnesses what
is taking place in the fictional world. We saw in chapter 1 that the film viewer,
unlike the theater spectator, say, is described as an individual, even an iso-
lated, viewer. In phenomenological film theory the specific impression that
the film image makes on viewers, namely, that it is an observation made ex-
clusively by the viewers themselves, has been singled out as essential. As Metz
(1975a) said, the viewer identifies with himself or herself. Morin (1956) de-
scribed that same impression when he argued that the camera imitates the
stages of perception, the separate, saccadic eye movements of the observer.
He noted the sensation, reported by viewers, of their eyes darting about the
film scene, as if divorced from the body. And if we disregard for a moment
the additional symbolic charge that psychoanalytic theory bestows on its no-
tions, then we could say that it refers to this same phenomenon when it deals
with the allocation of a subject position to the viewer by the film. The most
cogent formulation I know of is that of the phenomenologist Sesonke: "In
contrast [to the theater audience], there is a sense in which we see a film
alone rather than as part of an audience; my relation to the space and the
events of the film is felt as my relation, not as ours. When the camera moves
it is I and not we who move with it through the world of the film" (Sesonke,
1973, pp. 408-409).
As regards the appraisal of the situation, one result of the diegetic effect
is that the objectivity of the situation is exceptionally high. This crucial as-
pect of the situational meaning may be seen as indispensable to the evoca-
tion of an emotion. The situation in the fictional world is appraised as "spon-
taneous"; the subject experiences himself or herself as "affected and passive."
He or she does not "confer meaning" (Frijda, 1986, p. 205). In film viewing,
this is due primarily to the fact that the viewer relinquishes control over the
stimulus, allowing it to pass to the film, that is, the narrative process, but also
to the nature of the fictional world. That world exists and is independent of
the viewer, or so it appears. It was there before the viewers made their en-
trance, and it will continue to exist when the film is over. What does end is
the viewer's opportunity to play out his or her own role of witness to the
events of that world. What happens there is governed by a kind of logic that
cannot be influenced, although it can to a considerable extent be understood.
Events do not befall viewers physically, nor even as participants in the fic-
tional action. As a viewer I am an observer of that fictional world, and as such
my fate is not determined by the events I observe. Many—perhaps even the
majority—of the emotions evoked by feature films may be characterized as
object fate emotions (Frijda, 1986). (See also the related notion fortunes-of-
others emotions employed by Or tony, Glore, & Collins, 1988). Not one's own
welfare but that of others is the source concern addressed here. Just as in the
traditional novel and traditional drama, the events in the fictional world are
FILM AND EMOTION: THEORETICAL BACKGROUND 55

determined by characters who act in a more or less purposeful fashion and

whose mental life is more or less comprehensible to the viewer. The theoreti-
cal implications of this aspect of the diegetic effect are legion. We postpone
our discussion of those implications until chapter 6, where we examine the
concepts of identification and the empathetic emotions. We can, however,
ask ourselves for whom the situational meaning is intended, the character or
the viewer? The answer is that the situation addresses viewers in their imag-
inary role as witnesses to the events of the fictional world. And it is to this sit-
uational meaning structure that the components are related. Thus urgency
signifies that in the eyes of the viewer it is high time that something is done
by or on behalf of the protagonist with whom the viewer sympathizes, re-
gardless of whether the protagonist himself or herself shares that urge. But
the component controllability is always minimal: the viewer can do nothing.
It is this condition that guarantees safe involvement.
In the traditional feature film, the essential characteristics of film tech-
nique are placed entirely at the service of the diegetic effect. As a privileged
witness, viewers are allowed to see precisely what they need to see, at pre-
cisely the moment they need to see it. Let me illustrate this point with a num-
ber of examples. The mobility of the camera consists largely in reframings,
movements that are necessary in order to keep the action in the center of the
screen (Bordwell & Thompson, 1986). The demarcation of the foreground
and background of the visible action, which is done by means of framing, fo-
cus, and the exact staging, will depend on what it is important for the viewer
to see at that particular point in the action. Of all the sounds that are part of
the scene, what viewers hear are those considered important for them at that
moment. Even the music, which clearly comes from outside the fictional
world, is governed by the fictional events taking place. It fades away during
dialogue, and forward and back reference functions are used only sporadi-
cally. Music serves primarily to highlight those moments that are crucial to
the plot. It can illustrate these moments, signal them, or strengthen any
moods already inherent to them. Similar considerations, that is ones having
to do with centrality of events to the plot, dictate the omission of superflu-
ous visual information. Even within a continuous scene, the action is com-
pressed almost invisibly, so that only that which is dramatically important ap-
pears on the screen (de Wied, 1991, 1995; Reisz & Millar, 1968). And what
is dramatically important is determined by the situational meaning structure,
that is, the meaning of the story event that is related to a certain emotion.
Thus in guiding the appraisal the plot makes use of all conceivable cine-
matic means. The elements of the primary appraisal are provided directly by
the film; those of the secondary appraisal are at the very least supported by
it. The major elements of the situational meaning structure are made con-
spicuous and redundant by such technical means as acting, staging, framing,
montage, camera movement, use of props, and selective amplification of the
56 CHAPTER 4

sound. The viewer is furious with the villain because he is deliberately stand-
ing in the hero's way. It is quite obvious that he is refusing to let the hero pass,
and it requires no great insight to see that he is acting according to some plan
or other. The villain is portrayed in such a way that he is objectively evil, and
there is no room for doubt on that point (Frijda, 1986). The intensity of the
emotion is also determined by the parameters of the situational meaning,
which the cinematic technique allows to attain a maximum value: these in-
clude not only objectivity but also such aspects as closedness (Frijda, 1986)
and proximity. The viewer fears increasingly for the life of the heroes when
seeing that they are unable to untie the rope that binds their hands, while the
water—or fire—is getting closer and closer. There are countless other exam-
ples of the affordances that are directly observable as part of the primary ap-
praisal steered by the film technique.
While the plot controls the situational meaning structure, this does not
necessarily mean that the viewer is allowed to see and hear everything. Con-
trol may mean that certain inferences or representations are forced upon the
viewer. For example, a few minor but highly expressive consequences of an
accident or act of violence may be shown (traces of blood, the twisted frame
of a child's bicycle, etc.). The very inescapability of the conclusion renders
it more tangible. A familiar variation on this theme is the case where the
viewer hears only a series of noises, and yet is capable of conjuring up a de-
tailed visual representation (as in the case of the off-stage sounds of blows
striking a body). Nor does control of situational meaning automatically mean
that the viewer is conscious of the relevant characteristics of the situation, at
least in the case of primary appraisal. This is a general feature of human emo-
tion. "They [meaning structures] do their work, whether one knows it or not"
(Frijda, 1988, p. 351). It is indeed conceivable that too blatant a presenta-
tion of relevant characteristics leads to a certain irritation.

THE LAW OF CHANGE, EMOTION,

AND TEMPORAL STRUCTURES IN FILM

The effect of the temporal dimension is determined by the fact that emotions
are rooted not so much in the presence of positive or negative conditions, as
in changes in the stimulus, the so-called Law of Change (Frijda, 1988). A posi-
tive or negative turn of events gives rise to an emotion. The intensity of the
emotion is related to the magnitude of the change. In addition, Carver and
Scheier (1990) argued convincingly that the speed of the change, for better
or worse, determines the intensity of the emotion. They claimed that this fac-
tor can be further differentiated according to the abruptness of the change.
When a desirable state approaches more rapidly than expected, this is
greeted with delight; when the approach of an undesirable state is acceler-
FILM AND EMOTION: THEORETICAL BACKGROUND 57

ated, this is cause for disappointment. An abrupt decrease in the distance to

the preferred state results in a superlative degree of joy, or exhilaration, while
an abrupt increase in that distance causes a response that can best be defined
as a sinking feeling.
Changes are appraised on the basis of a criterion. This may be a future
preferred state, or a halfway point on the way to that state. But there are many
other contrasts between an existing state and other situations, past or imag-
inary, which help to determine the nature and intensity of affect. This is what
Frijda (1988) calls the Law of Comparative Feeling. Thus poverty is harder
to endure when one has known wealth, when the neighbors are rich, or when
wealth that was within one's grasp has been lost.

Characteristic Changes in Situational Meaning:

Canonic Course of the Film Narrative
A feature film is a story, which means that knowledge is conveyed. And yet
the film audience is barely aware of the process by which this takes place.
Only rarely is there an overt narrator, and even then he or she often recedes
into the background. The narrative process is concealed behind the diegetic
effect, or more precisely, behind the apparent objectivity of events.42 Events
are discrete elements that constitute the continual changes in the situation
in the fictional world.
Narration is only possible by virtue of disparity of knowledge (Branigan,
1992), that is, knowledge that is concerned with the fictional world. At the
end of the story the viewer is aware of all relevant aspects of a complete
episode that has taken place in that fictional world. The knowledge that is
conveyed, and the manner in which it is conveyed, are characteristic of narra-
tives. Bordwell (1985) described the most important parameters of the nar-
rative for the classical feature film. These include communicativeness, the
measure in which the narrative divulges knowledge, and the depth of the
knowledge divulged, in other words, the degree to which the viewer is in-
formed about the inner life of the characters. The narrative determines both
the nature and the intensity of emotions through a careful dosage of infor-
mation by means of these parameters. But emotion is played upon first and
foremost through the presentation of ongoing change.
The description of this function of film narration, the presentation of on-
going change in the fictional situation, can be approached in various ways. In
principle, all the theories of plot structure are important to such a descrip-
tion. For our purposes, however, it will be sufficient to highlight the major
facets of the changes in the situation in the fictional world, without claiming
to be exhaustive.
42
According to Branigan, the viewer's understanding of the story is guided by the "illusion of
occurrence" (Branigan, 1992, p. 115).
58 CHAPTER 4

The changes that take place in the course of a traditional film narrative
can be broadly summarized. 43 According to Bordwell, the story structure of
the classical film follows a conventional schema revealed by story researchers
(Bordwell, 1985). More recently, the same notion has been developed by
Branigan (1992). According to many story theorists, the most elementary
formula to which a story, that is, a series of events in a fictional world, can
be reduced is a characteristic course of events. This involves a systematic
change brought about by a process of cause and effect. The change is enacted
as follows: A balance is disturbed, and then restored. (See Todorov, 1971; for
the film, Heath, 1981, Neale, 1980; Thompson, in Bordwell, Staiger, & Thomp-
son, 1985). In the more refined cognitive story grammars (Mandler, 1984;
Rumelhart, 1977; Thorndyke, 1977; van Dijk, 1978), the story also has a rec-
ognizable middle, a Complication that may consist of a recursion or linked
series of episodes in which the struggle to attain one or more subgoals takes
place.
The overall organization of the story is reflected in the characteristic emo-
tion of the viewer, likewise conceived as an overall process. The viewer under-
goes an emotion episode. According to Frijda, Mesquita, Sonnemans, and van
Goozen (1991, p. 201), an emotion episode is "a continuous emotion se-
quence resulting from the more or less continuous impact of one given event
or series of events." Watching a film sets in motion a continuous witness-
emotion episode, by virtue of the fact that the film story consists of a causal
sequence of events with a certain cohesion. Branigan (1992) aptly employed
the term focused causal chain. The cohesion may be attributed to the fol-
lowing characteristics of the classical narrative style: (a) a limited number of
characters are involved in separate and independent actions, (b) if there is
more than one plot line, which is generally the case, these lines are inte-
grated; thus the action line and the romance line of the story always converge,
and (c) in the end, causality is based on psychological features of protago-
nists acting in a rational manner (Bordwell, in Bordwell, Staiger, & Thomp-
son, 1985).
In this light, the Disturbance of Balance is crucial, in the sense that it must
bring about an irreversible involvement. That is, one or more concerns of the
viewer must be addressed for an extended period. We may assume that the
event in question leads to cognitive curiosity, a drive toward knowledge (How
can this problem be solved?), the activation of the viewer's sympathy, and an

43
This summary does not presume to represent anything resembling a "story grammar," with
a broad domain of validity. That would be decidedly too ambitious. More than one attempt to
present a universally valid schema has come to grief due to the complexity of existing narratives.
See, Chatman (1978, pp. 9 2 - 9 5 ) and the well-known discussion on psychological "story gram-
mars" (Wilensky, 1983b). However, in our view, the characterization of the typical plot structure
given here reflects the consensus that exists concerning the structure of simple, traditional
stories.
FILM AND EMOTION: THEORETICAL BACKGROUND 59

involvement with the problem as it presents itself to t h e protagonist. Sympa-

thy also means that the audience takes the side of the protagonist, so that as
far as possible the protagonist's concerns become the viewer's concerns. The
episode ends when the satisfaction conditions of the curiosity and sympathy
concerns have been met, that is, when the Disturbance has been resolved.
The Complication phase may be seen as a series of minor, embedded emo-
tion episodes. (For a more detailed discussion of the processing of a story as
emotion episode, see Tan, 1994).
Change is measured both by what the viewer has already been told is go-
ing to happen and what can be suspected. Thus changes in emotion are spe-
cific, the result of systematic changes in the situational meaning structure.
We will examine briefly a number of changes in important components of the
situational meaning structure that can be derived f r o m the three phases of
the canonic story schema. (For an account of the components of situational
meaning presented here, see Frijda, 1986, pp. 2 0 4 - 2 1 4 . )
Initially there is a situation of balance, in which everything is going well for the
protagonist. In this phase, the Balance phase, the intensity of emotion may be
fairly low. As regards the core components of the situational meaning structure,
difficulty and urgency are low. Especially when the situation of balance is shown
at the very beginning of the film—which is not always the case—both clarity
and demand character will be low. Valence is positive, at any rate in compari-
son with that of the situation after the disturbance. With respect to the context
components, modifiability may be high: there may be signs indicating that
things are about to become less calm and pleasant. This is not, however, always
the case.
In the following phase, the Complication phase, the balance is disturbed—
often quite unexpectedly—by an event which has a highly negative valence for
the protagonist (a murder, a holdup, financial reverse, loss of the loved one, out-
break of a feud, etc.), which is contrasted with the valence of the previous bal-
ance. Assuming that the fate of the characters is related to a sympathetic sur-
face concern of the viewer, then such negative events will have a negative va-
lence for the viewer.44 This event, the Disturbance of the Balance, produces
such emotions as sadness, pity, and anger, and affects directed toward the fu-
ture, such as fear and a desire for improvement in the situation. The Distur-
bance introduces a constant change in the valence of the situation. There is a
stream of Complications, ups and downs in the restoration of the balance. At
this point of the narrative several criteria against which change is measured are

44
The fact that the disturbance implies a deterioration may have to do with the fact that this
is less likely to induce habituation: although we can get used to anything, we find it more diffi-
cult to accustom ourselves to unpleasant circumstances than to pleasant ones, as Frijda argues
by means of his Law of Hedonic Assymmetry (Frijda, 1988, p. 353). A change for the better
loses its emotional potential more quickly. If the objective of the film is to entertain by evoking
emotion, then it may be expected that much of the film will be taken up by misery and misfor-
tune and the struggle of the characters to surmount them.
60 CHAPTER 4

simultaneously operative: 1) the expectation that in the end things will get bet-
ter; 2) the memory of the original situation; 3) the subsequent deterioration at
the beginning of the Complication phase; and 4) the immediately preceding sit-
uation within the Complication episode. To illustrate the fourth criterion: A sud-
den reversal in the fortunes of the characters is not only pleasant or unpleasant
in terms of the expected objective, for example, victory, but also because things
had been going so well. The emotions which present themselves in the Com-
plication phase of the story, as a result of the changing valences, include hope,
fear, uncertainty, and relief. Desires are also characteristic of this phase: the
viewer longs for righting of the wrong, the safe return of the protagonist, the
reconciliation of the lovers. As regards the core components of situational mean-
ing, this phase is accompanied not only by changes in valence, but also by con-
siderable variations in the intensity of urgency, difficulty, and seriousness. As a
rule, the value of these core components will increase during the Complication
phase, together with those of demand character and clarity, as the more im-
portant objectives and obstacles become clearer; in the course of the action it
is these which will remain and converge, while those less relevant will disap-
pear. The value of a number of context components also change during the
Complication phase. To name a few examples: Certainty increases, because as
a rule the number of possible outcomes tends to decline. Closedness increases,
as obstacles appear on the road leading to the final goal.
The Disturbance is ultimately undone by events, which obviously have a pos-
itive valence (the culprit is arrested, the hero regains his fortune or his beloved),
resulting in such emotions as relief, joy, and triumph. This phase is referred to
as the Restoration of Balance, the decisive event(s) leading to the restoration
are sometimes referred to as the Resolution. The change involved in the Reso-
lution is compared with two criteria. Not only the previous disturbance plays a
role in the comparison, but also the original situation. Thus the final situation
may eliminate the problems, and at the same time constitute an improvement
over the situation in the beginning (in the end there is not only happiness but
happiness that lasts forever). With respect to the change in situational mean-
ing, brought about by presentation of the Restoration, in the traditional feature
film the final situation is unequivocal. Of all the situations that were presented
as possible final results (for example, A, B, or C gets D, or perhaps none of
them), only one prevails as a final outcome (for example, A gets D). Even in the
case of an open end, only one result is implied, one that is clearly defined, even
if dynamically: A gets D, but not for long; the game will soon start all over again,
and the chances of all three will again be equal. This means that clarity and cer-
tainty concerning the outcome are maximal, while modifiability and demand
character are minimal. Difficulty and urgency decline, as do seriousness and fo-
cality. As regards the latter aspect, this is illustrated by the fact that the famil-
iar formula "And they lived happily ever after" is remarkably devoid of details,
at least as compared to the specific nature of the previous difficulty. The va-
lence is, in general, positive: two thirds of the films made during the classical
Hollywood period have a happy ending (Bordwell, 1985). Thus presence is high:
FILM AND EMOTION: THEORETICAL BACKGROUND 61

the sympathetic protagonist gets what he or she wants. Desire gives way to en-
joyment, hope and fear to relief, and pity to joy. (Frijda, 1986, pp. 204-214)
There are countless possible variations on the above pattern of change in
t h e stimulus and the accompanying emotions. The three phases that serve as
a framework for the description of change may be closely intertwined in in-
dividual films. To give an example, it is very common for a film to start with
t h e final episode of the Complication, then go on to tell the Balance and its
Disturbance by means of flashbacks, and finally return to the here and now
of the Restoration of the Balance. 45 The film narrative often makes use of ac-
celeration or deceleration or the withholding or foreshadowing of informa-
tion. Elements that clarify the significance of a situation can be presented at
almost any subsequent moment in the film. Conversely, a particular event—
articulated to a greater or lesser extent—such as a final accident, may cast
its shadow far ahead. 4 6

Change and Situational Irony

It will be clear that a film may evoke a wide variety of emotions at one and
t h e same time, and that the blend of these emotions is constantly subject to
change. One type of emotional blend is characteristic of the experience
evoked by films. A particular situation may have a multiple valence: one and
t h e same situation may involve several different changes with respect to var-
ious previous situations. Thus the end situation of the film story is not always
a n improvement over all aspects of the disturbed equilibrium or the initial sit-
uation. Perhaps the protagonist ends up sadder but wiser or even rich and
still sadder but wiser. The heroine becomes a great star but loses her lover in
t h e process; chastened, she cherishes his memory (A Star is Born, 1954).
There are several criteria for change: the protagonist, once a poor, unknown
singer, is now a celebrated actress; once they were happy together, now she
misses him; he helped her on her way to success but is not there to share it
with her; he contributed to her rise to stardom but has himself sunk to the
depths of despair and suicide. As a consequence, the feelings of the audience
are as mixed as those of the screen characters.
This constellation of multiple references to previous situations is known as
situational irony. The term is most commonly used to refer to the ambigu-
ity of meaning in comic situations, such as the conversation that Jack Lem-
m o n and Tony Curtis (dressed as women) carry on with their "suitors" ( S o m e
Like It Hot, 1959). But it is clear from the example of A Star is Born that

45
The Remains of the Day (1993) is a typical example.
46
For instance, Accident (1967); A Wedding (1978); Rafelson's The Postman Always Rings
Twice (1981), and its predecessors by Visconti and Garnett.
62 CHAPTER 4

ironic relationships between changing elements of the situational meaning

structure and the criterion do not always give rise to humorous effects. Situ-
ational irony is employed in the plots of melodrama and tragedy as well and
also features in other genres. (For an overview of characteristic antitheses,
paradoxes, and parallels in popular film narrative, see Durgnat, 1967. Lu-
cariello, 1994, classified a large number of situational ironies that are easily
recognized and even produced by subjects lacking any special training.)

Specific Changes in Situational Meaning

as Opposed to Aspecific Changes in the Response
Up to now we have used the term emotion episode to refer to the total emo-
tional process evoked by a feature film because it is based on an ongoing
change. Various subepisodes may, however, have an emotional tone all their
own. Exciting scenes, humorous scenes, and romantic scenes may be found
in one and the same film. Contrast of emotional tone across scenes is such a
frequent phenomenon that it elicits speculation as to its functions. One at-
tractive assumption is that a strategic placement of various scenes, in which
the type and intensity of the emotions evoked is alternated, leads to an opti-
mum evocation of emotion during each of those scenes. Any given scene may
affect another, as a result of excitation transfer (Zillmann & Bryant, 1974)
and emotional response compatibility (Baron, 1977). The excitation trans-
fer effect involves the reinforcement of an emotional situation by previous but
unrelated experiences that have a high activation value. The study by Baron
and Bell (1977) shows that negative emotions, such as aggression, can be re-
duced by calling up unrelated positive emotions, for example, through mildly
erotic stimuli. However, the conditions under which excitation transfer and
response compatibility effects operate, namely, the inability of the subject to
attribute remaining activation or hedonic value to the previous stimulus, does
not hold true for the film viewer. A feature film is not an arbitrary series of
mutually independent events capable of evoking emotion, but rather a con-
tinuing series of causally related changes.47
What at first glance appears to consist of response effects such as activa-
tion transfer, or a contrast effect of hedonic tone, may on closer examination
reveal itself as an underlying change in situational meaning. There is always
something to which the change responds, and this is less likely to be the va-
lence of an emotion, or the intensity of the activation, than one or more fea-
tures of a past or future situation. 48
47
Strangely enough, the excitation transfer effect has also been demonstrated by showing frag-
ments of totally different films (Cantor, Zillmann, & Bryant, 1975; Ramirez, Bryant, & Zillmann,
1982; Zillmann, Mody, & Cantor, 1974).
48
For the same reason, it may be assumed that the so-called "mood congruency" effects are
limited, that is, the effects of our knowledge of previous events on our knowledge of later events
FILM AND EMOTION: THEORETICAL BACKGROUND 63

It should be noted here that a special sequence of scenes with a specific

intensity and hedonic value undoubtedly influence the emotional effect of the
individual scenes. The manner in which director Maurice Pialat alternates the
description of Van Gogh's depressions with happy and carefree scenes (an
outdoor celebration, a dinner where people are laughing and smiling) is only
one example of the masterly manipulation of viewer emotion (Van Gogh,
1992). In such effects, however, the ongoing story line still plays a role; it is
this line that is responsible for the compatibility of, and the contrast between,
the successive scenes, rather than the sequencing of emotions per se.

Emotion and the Scenic Structure of the Film Narrative

According to current film theory, the scenic structure of the film narrative is
one of the most constant features of the classical feature film (Bellour,
1976/1979; Bordwell, 1985; Metz, 1968, 1972). The scenic structure allows
the narrative to select from the total history of action a particular segment
and to weave it through that history more or less at will. The rest of the com-
plex of events that comprise the story then form the background. Thus view-
ers are not obliged to have at their fingertips each and every aspect of the
story known to them thus far. Because each scene in the traditional film is
accompanied by a brief introduction, the activation of relevant prior knowl-
edge is ensured. As regards the situational meaning structure, this means that
the focality of the events taking place, and the most important aspects of
those events, is high. (Frijda, 1986). As a result, the feature film first evokes
emotions, and only then moods. The latter are the result of nonfocal aspects
of the action, such as the atmosphere, or the general tenor of the action,
which in certain scenes may take over from the action itself (den Uyl & Frijda,
1984; Frijda, Mesquita, Sonnemans, & van Goozen, 1991).
The best known quality of the scene is its unity of action. In the traditional
feature film, the action within a single scene is continuous. 49 As a result, the
situational meaning is based on continuous time in the fictional world, the

(Bower, 1978, 1981; Gilligan & Bower, 1984). They undoubtedly occur during the viewing of fea-
ture films but are overshadowed by more specific emotional influences of cognitive elements.
Mood congruency may also be an overall effect of one's mood at the beginning of the film. In
chap. 2 we discussed the possibility that viewers select films that echo or contrast with their own
mood. The revenge of the unhappy protagonist is doubly sweet when one's own mood is some-
what aggressive (Straw Dogs, 1971; Taxi Driver, 1976).
49
The alternating viewpoints within the continuity of a single scene that contribute to the
diegetic effect, in the sense that as a spectator one is generally allowed a privileged view of the
essential details of the action, are fairly predictable and logical (Hochberg, 1986; Frith & Rob-
son, 1975; Lynch, 1972; Messaris, 1994). The logic of the continuity employed by the makers of
the traditional feature film has been well documented. (See, above all, Reisz & Millar, 1968). Re-
search into eye movements by d'Ydewalle and Vanderbeeken (1990) showed that cuts that fol-
low the traditional rules of continuity are less likely to be observed than those that do not.
64 CHAPTER 4

experience of which—in a prototypical manner—corresponds to the tempo-

ral experience of real life. The viewer's impression of time, as in the reality
of everyday life, is one of an advancing now.
Scenic structure serves to demarcate emotions in time. The scenes in clas-
sical feature films display a structure comparable to that of the story as a
whole (Bordwell, 1985). They contain equivalents to the introduction, a prob-
lem, and a solution. Thus a certain scene may evoke limited-emotion episodes
that are embedded in the larger emotion episode provided by the film story
as a whole. The hierarchy of emotion episodes corresponds to that of the ac-
tion structure of the film story, which also contains embedded episodes, each
consisting of an attempt to realize a subgoal (Rumelhart, 1977; Thorndyke,
1977).
The relationship between scenic structure and the course of emotions is
such a close one that we are probably justified in assuming that the viewers'
perception of the course of their own emotions is just as valid an indication
for the segmentation of the narrative as the action itself or the conventional
segmentation signals (filmic punctuation), such as fade and dissolve.50

FICTIONAL WORLD VERSUS ARTEFACT

Thus far we have proceeded on the assumption that in traditional cinema the
film as artefact is subordinate to the fictional world to which the attention of
the audience is directed. This does not mean that the artefact plays no role
in emotion. For one thing, there may be deviations from the customary prag-
matic principles; these result in emotions related first to the film as artefact
and second to the makers. If a film is not entertaining, for example, because
there is too little structure in it or because it departs from the acceptable do-
main of fictional worlds, the natural viewer of the traditional feature film will
have negative feelings. This is a relatively rare occurrence, but there are more
subtle gradations conceivable in the violation of pragmatic principles, as well
as individual differences between viewers in the appraisal of such violations.
Some film viewers may feel that a gory scene or a bout of torrid lovemaking
has gone on too long or is too explicit to be truly entertaining, and it is not
unknown for dissatisfied viewers to leave the theater. 51
Even when a film remains well within the bounds of the cooperation prin-
ciple, the artefact does occasionally emerge. Perhaps the audience is startled
by an attack on the hero that comes as a total surprise. They may admire the

50
Forgas (1982, pp. 84 ff) likewise pointed to the possibility that emotions serve as "mark-
ers" of episodes in daily life.
51
Survey research by Palmgreen, Cook, Harvill, and Helm (1988) showed that for some peo-
ple violence and sex in feature films is a reason for not going to the cinema.
FILM AND EMOTION: THEORETICAL BACKGROUND 65

narrative technique, the ingenuity, and wiliness of the filmmaker. Or during

a particularly gruesome scene the viewers may find themselves wondering
how the event was staged or indeed if it was staged (e.g., where they have just
seen a dog shot dead).
Sudden twists in the plot can also cause the viewer t o become aware of the
telling of the story. Although the diegetic effect makes the manipulation of
the events invisible, most story lines will contain at least a few unexpected
twists and turns. But the element of surprise is bounded by a certain logic.
The viewer is continually being misled, and expectations may often prove
wrong. In this way, the traditional feature film makes it clear to the audience
that there is an editorial intelligence at work (Bordwell, 1985), who has care-
fully ordered the events that they are viewing.52 As soon as they are aware,
no matter how fleetingly, of the operation of that intelligence, they are in that
instant aware of the film as artefact. In a general sense, it may be that the
more intense the emotion, the greater the likelihood the viewer will realize
that this is a special experience and be aware of what he or she is seeing is
indeed an artefact. It is interesting that, in that case, the emotion evoked by
the situation in the fictional world is one element of the situational meaning,
which results in an emotion that has the artefact as its object.
As the final manifestation of the film artefact, we have the tours de force:
the spectacular special effects, a sample of superb acting talent, and/or im-
pressive photography or staging. Bordwell points to the origins of the Holly-
wood film, such "spectacle-centered entertainments" as vaudeville melo-
drama and music hall, which are reflected in the virtuoso and energetic dis-
play of "showmanship" (Bordwell, in Bordwell, Staiger, & Thompson, 1985,
p. 21). However, it is not unlikely that in the case of the traditional feature
film this display is not distributed arbitrarily but occurs primarily in combi-
nation with fictional action that is itself calculated to bring forth intense emo-
tion. As Bordwell says, the artistic motivation of stylistic choices follows, or
is concealed within, narrative and causal motivation.
It should be noted that awareness of artistry and manipulation by the film-
maker may depend on a competence that is lacking in most natural viewers.
Messaris (1994) convincingly argued that production experience rather than
viewing or analytic training is needed in order to detect certain aspects of
film style. However, other aspects of the artefact, such as acting, may strike
almost any viewer. More research on viewer competence is clearly needed.
From now on we will refer to emotions that arise from concerns related to
the artefact, as well as stimulus characteristics based on those concerns, as
artefact emotions or A emotions. Those emotions rooted in the fictional
world and the concerns addressed by that world will be referred to as F emo-

52
It should be noted here that Bordwell uses this term in the course of an argument in which
he stresses the fact that in the case of the classical cinema narration is relatively hidden.
66 CHAPTER 4

tions. The latter term is linked to the experience of being present in the fic-
tional world.

THE LAW OF APPARENT REALITY

AND FILM VIEWING

Coding and Reality

In the previous sections dealing with the emotional stimulus, we have tried
to make it plausible that the traditional feature film prepares the way for much
of the viewer's work. With respect to the fictional world, the cognitive coding
of the stimulus requires little more than a knowledge of the world, which al-
most everyone who is part of Western society has at his or her disposal.53
The coding processes referred to here are cognitive-constructive and are
based on cognitive schemata. More direct perceptions based on ecological
principles (Gibson, 1979) also play a role, not only at the most elementary
level of the perception of events (Johansson, 1950) but also in the recogni-
tion of more complicated interactions between characters (Knowles & Smith,
1979; McArthur & Baron, 1983). Thus the appraisals that structure the emo-
tional meaning of situations portrayed in feature films are determined largely
by quite common perceptual and cognitive processes. At first glance, the is-
sues that complicate the effects of coding and appraisal mechanisms (Frijda,
1986, 1993) would seem to be of lesser importance for the viewing of a fea-
ture film. However, there is a catch here. One of the most important charac-
teristics of the feature film stimulus, which for the sake of convenience we
will call "realism," is the result of coding and appraisal processes that are ex-
traordinarily difficult to clarify.
The reason for dealing with this subject here, rather than in the section on
the emotional stimulus, is the fact that the viewer makes a major contribu-
tion to this characteristic of the stimulus. This is presumably one of the rea-
sons that the problem of realism has for so long been a topic of discussion
within the theory of the arts. As many theater, television, and film theorists
argue, the viewers know that they are watching a construction but neverthe-
less succeed in seeing it as "real," to some degree. If fiction is to have any ef-
fect, there must be some kind of multiple coding by the viewer.54 The theory

53
It includes knowledge of elementary actions, situations, events, goals, causal action
schemas, plans, and themes (Johnson-Laird, 1983; Schank & Abelson, 1977), persons (see chap.
6), and canonical story structures (Thorndyke, 1977).
54
See also, among others, Allen (1993), Ang (1985), Bazin (1958-1962), Bonitzer (1982),
Carroll (1988b), Ellis (1982), Fiske (1988), Metz (1975a), Michotte (1948), Schoenmakers
(1988), and van Vliet (1991). All these authors discuss some kind of ambiguity in the observa-
tion of the reality in the film that boils down to the fact that the viewer distinguishes two or more
FILM AND EMOTION: THEORETICAL BACKGROUND 67

of emotion is related to the observation that awareness of reality is not a mat-

ter of all or nothing at all. The law of apparent reality (Frijda, 1988) states
that emotions are evoked exclusively by events that are appraised as real and
that their intensity corresponds to the degree to which this is the case. A com-
parable relationship has been postulated by Ortony, Glore, and Collins (1988).
The law of apparent reality makes no pronouncement on reality in any ob-
jective sense, but rather on the impression of reality created in the subject.
Frijda (1989) likewise referred to the assumption of multiple coding in clari-
fying the aesthetic emotions, including those based on the fictional action
portrayed in feature films.
The assumption that the apparent reality of the fictional action is not one-
dimensional is developed in the structuralist theory of literature in the no-
tions concerned with apparent reality or truth. Culler (1975) integrated the
relevant ideas of Todorov, Genette, and, in particular, Barthes in a survey of
the levels of meaning on which the reader, guided by the text, attempts to
construct an apparent reality or verisimilitude. A literary text confronts read-
ers with things that are strange, which they then render natural with the aid
of certain literary conventions. At the most elementary level the reader rec-
ognizes the reality that is familiar from simple observation. At the next level,
a socially and culturally determined reality is evoked. On the next three lev-
els, purely literary conventions are involved in the formation of a reality,
namely, genre models, the act of narration, and the attitude of the narrator
toward the realities created on the previous levels.
Empirical research has produced dimensions of subject reality that corre-
spond in part to this literary analysis. A review of the empirical literature by
Potter (1988), focusing on the perceived reality of television fiction, is of par-
ticular interest here. On the basis of the results of a relatively large number
of studies, he distinguished three dimensions that met certain simple crite-
ria: empirically distinguishable, theoretically significant, and independent of
one another. Magic-window reality is the most important of these and may
be described as a belief in the literal reality of the television message. This
dimension has a syntactic style component, and a semantic content compo-
nent. The second dimension is known as perceived utility and refers to the
applicability to their own life of what viewers observe. The last dimension,
identity reality, consists of a feeling of closeness to the characters who ap-
pear in a particular program.

separate levels of representation, which differ in the degree or type of "reality" attributed to
them. The differences between the various views could easily lead to an extensive discussion cen-
tering on the classic problem of realism in aesthetics. To take only one example, the problem is
complicated by the various views as to what actually constitutes an illusion. It would not be wise
to embark upon such a discussion at this time. It will be taken up in the final chapter of this
book.
68 CHAPTER 4

Magic-window Reality
As its name implies, this kind of apparent reality has the most intimate
links with the diegetic effect. The syntactic component of magic-window re-
ality is the result of the highly elementary coding of the stimulus information.
Unconditioned affective associations are important in the appraisal of the re-
ality of the situation, as are the sensory associations linked to them by con-
ditioning (Frijda, 1986, 1988). For example, the images and sounds of in-
nately frightening stimuli or stimuli that produce revulsion or sexual arousal
in a highly direct manner are quite common in feature films. In chapter 6 we
look at other examples of characteristic feature-film stimuli that operate on
this level of coding.
A more far-reaching assumption is that a diffuse emotionalizing effect is
brought about by the photographic realism that is characteristic of the film
medium. The most important reflection of this is the diegetic effect referred
to earlier. Even though a filmed scene is recognized as having been staged,
viewers still have the impression that they are physically present in front of
or in the scene. If we ignore the element of movement in film for a moment,
there is already a strong effect of depth in the projected image. That is, it
seems as if the viewer is looking into a three-dimensional scene extending be-
hind the screen, from a specific viewing point. Generally, distance to the
screen is as large as to make the surface of the screen as a surface hardly per-
ceptible. Due also to the darkness of the surroundings, the impression of ap-
parent depth and the perspective illusion created by pictures are strongly
facilitated under these conditions. (See Kubovy, 1986, for a thorough discus-
sion of the illusion of depth in pictures.) In addition, the particular brightness
and resolution of photographic projection in the cinema ensures maximal
conservation of depth cues that were present in the scene, such as texture
gradients, at least relative to other pictures, such as printed photographs,
drawings, and paintings. Brightness also makes colors, as well as nuances in
black and white, more true to life. In short, cinematic photography con-
tributes reality to the picture by way of such elements as depth and vividness.
Now movement within the picture, the most specific aspect of the cinema,
contributes in various ways to the experience of reality, and this has not es-
caped the attention of film theorists (Bazin, 1962; Metz, 1968; Morin, 1956).
According to Michotte (1948/1991), the perceptual system cannot distin-
guish between apparent (stroboscopic) movement and true movement; re-
cent research in human perception appears to confirm this claim. Phenome-
nally speaking, movement is always real, a view put forward also in an essay
by Metz (1968). Moreover, according to Michotte, certain movements, re-
gardless of their origin, give rise to immediate and specific impressions (e.g.,
of causality). Thanks to their kinetic structure, complex movements can
evoke immediate impressions of such actions as impact, approach, separa-
FILM AND EMOTION: THEORETICAL BACKGROUND 69

tion, flight, and so on. Michotte demonstrates this by means of a series of in-
genious experiments involving a special kind of projector (Kiewiet & Tan, in
prep.; Michotte, 1946). Gibson (1979) viewed such impressions as directly
observable affordances, qualities that have to do with action potential. These
movements and complexes of movement lend reality to the fictional world be-
cause they automatically call up impressions that are familiar from the real
world. The authenticity of the movement, in turn, lends reality to moving ob-
jects. Camera movement makes the direct perception of ego motion in the
scene almost inevitable, thanks to the operation of (differential) motion par-
allax. In a recent study, Warren and Kurtz (1992) experimentally demon-
strated how powerful the effects of simulated camera movement on apparent
self-motion may be. Subjects exposed to animated sequences exhibit overt
body movements corresponding to the impression of self-motion. The strong
illusion of self-motion reinforces the impression that one is surrounded by a
reality in which one is moving. That impression also owes something to the
fact that the screen functions as a kind of artificial boundary: objects that are
cut off at the edge of the screen appear to continue beyond the field visible
to the viewer. This impression is the basis for the illusion of an off-screen
space that extends in all directions around the viewer. Allen (1993), who dis-
cussed the diegetic effect as the "projective illusion" adds two more obser-
vations. First, the impression of a continuous space is intensified by the fact
that objects regularly move across the boundary between the screen and the
off-screen space behind the camera; moreover, the camera itself also moves.
Second, stereophonic sound contributes to the illusion of a diegetic space
that is larger than that viewed through the magic window.
For that matter, both Michotte (1949) and Allen (1993) stressed that the
impression of reality created by movement in the frame, while strong, is not
imperative, especially since the viewer is free to look away from the screen
and to again become aware of the other space, that of the theater. Seeing depth
and continuous action in conventional motion pictures is immediate in the
sense that it does not require any learning and is not very culture specific.
Messaris (1994) argued convincingly that no "visual literacy" of any kind is
needed to understand pictures—and even line drawings—based on linear per-
spective, and that continuity editing in conventional film appeals to percep-
tual and cognitive mechanisms that are well developed in all adults due to
normal experience in the real world.
One explanation for the specific relationship between photographic real-
ism and emotion comes from Lang (1984) and his associates. Their experi-
ments show that dramatic presentations are more effective in producing af-
fect than spoken text and that experimental subjects with a well-developed
imagination display stronger emotions than poor imagers. Their explanation
is based on an associative network model of emotion (Bower & Cohen, 1982).
This means that the associative network in which the various aspects of the
70 CHAPTER 4

emotion process are represented can best be activated by a maximum num-

ber of propositions corresponding to parts of the network. For the most part
these propositions have been established in the association with real stimuli
outside the film. Due to its photographic and real-time qualities, the feature
film stimulus, like the dramatic presentation and vivid mental imagery, is rich
in propositions, far richer than texts, for example. By chance alone, these
propositions undoubtedly include many that are emotionally effective (i.e,
relevant for the situational meaning). Alternatively, as Lang and his associ-
ates mentioned in passing, it is possible that the transformations necessary
to turn stimulus information into effective propositions are smaller in num-
ber and less complex than in the case of texts.
Diffuse completeness effects of the perceptual realism of films must be
carefully distinguished from the power of the film narrative to highlight cer-
tain components of the situational meaning structure. In terms of the asso-
ciative network theory of emotions, only those propositions that occupy a
central place in the network are selectively activated. The feature film—and
perhaps other art forms as well—evokes a powerful suggestion of a particular
reality by presenting the essentials of that reality. Omission and condensa-
tion are among the most important means of obtaining this effect. (It should
be noted that this theory is based on the assumption that some complete and
true-to-life portrayal is at all conceivable, from which it is possible to deviate.)
At various levels of coding, the film shows the essence of situations, events,
actions, and characters, as well as social groupings and their way of life. Re-
duction and transformation make it possible to put forward aspects of things
we have a general sensory or conceptual notion of that could not have been
formulated as concisely by other, say, discursive means. In theory, it is pos-
sible to make a distinction between reduction and transformation. In the case
of reduction, emotion is directed through the depiction of certain funda-
mental aspects. As an alternative for a detailed overall representation of the
situation, it seems possible that overall reading strategy is determined to no
small degree by the reader's curiosity about important plot developments that
evoke affect ("story-driven reading," Vipond & Hunt, 1984). Elsewhere it has
been postulated that what readers pick up from the text are special elements
crucial to the emotional meaning structure (Tan, 1994). If we apply this to
the viewing of feature films, then all sorts of details that do not contribute to
the representation of the emotional meaning of a situation may be lost on the
viewer or may even result in confusion. Thus it is possible that in this way the
abundantly present practice of restricting detailed information conforms to
the strategy of the viewer, which is itself the result of the film narrative, at
any rate the more global aspects of that narrative.
The coding of transformed situations (again, in relation to some imaginary
true-to-life representation) would seem to require a greater cognitive effort
than the coding of situations that have simply been reduced. An initial as-
FILM AND EMOTION: THEORETICAL BACKGROUND 71

sumption that might explain why the viewer accepts transformation, or rather
stylization, is that the manner in which it is done is in itself entertaining, while
the result is an expressive portrayal of events. Occasionally a film may resort
to caricature, but if it is a good caricature, a possible reality will be there to
see. Another explanation might be that stylization actually makes it easier for
the viewer to recognize familiar objects drawn from reality. It may be useful
to look for a moment at the psychological principles of the caricature, which
we have drawn mainly from Hochberg (1972). Caricatures of objects may be
easier to recognize than photographically realistic representations of those
same objects because the distinctive features of their canonic or prototypical
form are emphasized. 55 By prototype we refer to a cognitive representation
rather than some Platonic category or other metareality. A number of stud-
ies on categorization processes have shown that it is difficult for people to de-
cide whether or not an object belongs to a given category. For example, does
a lamp belong to the category furniture. As a rule, people do not agree on
category boundaries. In contrast, they do agree on best examples of a cate-
gory. The best example or prototype of a category is unanimously recognized
as a member of that category and is also recognized more quickly than a less
typical instance (see Mervis, 1980, pp. 283-287 and Rosch & Mervis, 1975,
for a discussion of related research). Prototypes may share attributes with
most members of the category, in which case they count as best averages, or
alternatively, represent some ideal, for example, in terms of functionality. Em-
phasis on prototypical attributes is at the expense of the representation of
those features that are part of the photographic but not the prototypical rep-
resentation. All sorts of important elements of the fictional world are simi-
larly typified in feature films. At an elementary coding level, all we need do
is show the most typical features of an object, a chair, say, as regards both
components and visual angle. Characters become real people, that is, indi-
viduals, if on the basis of common-sense psychology they meet the criteria
for the prototypical person, notably intentionality. Some prototypes are based
on objective, that is, ecological, features of stimuli. In addition to such nat-
ural categories (Rosch, 1978), social and cultural stereotypes may also func-
tion as the canonic form. Structuralist poetics speaks of cultural vraisem-
blance (Culler, 1975). The traditional feature film conforms to the social and
cultural stereotypes of its natural audience. These include not only such as-
pects as the role of the sexes, family life, and ethnic identity but also proto-
typical views on justice, utility, good and evil, and so on, which we deal with
in chapters 5 and 6.
55
Experimental evidence for this assumption has been provided by Ryan & Schwartz (1956).
In later experiments, which differed in some respects, Goldman and Hagen (1978), Hagen and
Perkins (1983), and Tversky and Baratz (1985) recorded a number of results that conflict with
those of Ryan & Schwartz as well as with one another. Thus the empirical status of the assump-
tion is still somewhat unclear.
72 CHAPTER 4

The experience of the magic window realism of android monsters, men

from Mars, an elevator with human features ( T h e Lift, 1983), and all those
other creations that deviate from everyday reality can be explained by the
fact that fictional elements in films can take on many different forms, as long
as the prototype is not relinquished. In this connection, it is important to keep
in mind the accepted key characteristics of the traditional feature film, that
is, that the action is determined by causality and that ultimately this can be
traced to the intentionality of individuals (Bordwell, 1985; see also Chap. 1).
These same features characterize prototypical action schemas (Schank,
1975). In the traditional cinema the impression of causality is often reinforced
in a manner that is both subtle and efficient. The closer the agreement be-
tween the demands of causality based on intentions, the more freedom there
is to omit, replace, or introduce irrelevant characteristics. This also explains
why far-reaching stylization of the action is almost never experienced as un-
desirable. This is nicely illustrated by the well-known experiment by Heider
and Simmel (1944). Subjects were shown a short film in which geometric fig-
ures moved across a two-dimensional plane. Although not instructed to do so,
35 of the 36 subjects saw these figures as actors. Their actions were sponta-
neously attributed to motives and needs, which were in turn seen as flowing
from personality traits: "The plucky little male triangle fights off the big 'bad'
male triangle and rescues the 'helpless' female circle" would be an apt sum-
marization of subject protocols.
Heider and Simmel's experiment was recently replicated by Berry, Kean,
Misovich, and Baron (1991). Their study demonstrates that the information
on which the perception of acts as intentional is based is to be found mainly
in movement. Although the disruption of structural information in the Hei-
der and Simmel film did not result in a decrease in the anthropomorphic de-
scriptions, this was the case when movement was omitted from the film. This
replication supports an important concept put forward by Michotte (1950),
namely, that structures of motion contain expressive qualities that are im-
mediately picked up by the viewer, regardless of the nature of the moving ob-
jects, 56 thus producing the experience of reality. In addition, according to Mi-
chotte, although the viewer is not the object of the movements, he or she im-
mediately recognizes the emotion from which they spring, and this can in
turn produce emotion in the viewer. Thus in the case of films, styling intended
to have a maximum emotional effect on the viewer might usefully focus on
movements, that is, the movements in the mise-en-scene, and, above all, the
acting.

56
Kiewiet & Tan (in preparation) reported an experiment in which it was shown that shape
of two geometrical figures A and B does not affect judgment of such expressive qualities as ag-
gressiveness and strength of A. In contrast, speed of movement of A, and duration of contact
with B did affect judgment.
FILM AND EMOTION: THEORETICAL BACKGROUND 73

Furthermore, it will be clear that those elements that are of lesser impor-
tance for the action and the emotion of the viewer, do not have to be char-
acterized in detail and can be drawn along broad lines. The gaudily painted
cardboard backdrops of Mamie (1964) and The Man Who Knew Too Much
(1956) will be disregarded by viewers as long as their attention is not drawn
to them.

Utility Reality
Like the caricature, the fictional action of the feature film displays a key
quality of a more general nature. The perception of the reality of a work of
art as a whole is derived not only from the objective existence of the work of
art as artefact but also from the reality perspectives revealed by the fictional
situation or action contained within it. Love, acceptance following rejection,
and a happy ending after intense suffering are demonstrated to exist as a pos-
sibility by the sentimental film. Although the action in the film may be seen
as fiction, a more abstract significance of that action, such as a theme or
moral, tends to transcend fiction. Significance is not bound to the special ex-
istents (Ghatman, 1978) of the story, such as the fictional situation, place,
time, and characters; these may serve only to provide access to pronounce-
ments and visions of a far wider scope. "Possibilities come down in the shape
of the actual, in the object of art" (Frijda, 1989, p. 1547).
This form of reality perception—insight into a truth or possibility—is a spe-
cial example of Potter's utility dimension. The prototypical knowledge to
which the coding process appeals consists of commonsense truths, which
strangely enough may conflict with one another (Ruimschotel, 1987). The in-
sight that intense suffering may be followed by a happy ending contrasts with
the acceptance of the common mortal's fate, which is less fortunate, as mean-
ingful ("the tragic structure of feeling," Ang, 1985; Brooks, 1976) to which
the melodrama appeals; it contrasts with the idea that in a topsy-turvy world
like ours suffering must either be taken with a pinch of salt, as in comedy, or
has to be borne in order to achieve a form of purification, as in tragedy. The
essence depicted may be an absurd exaggeration, often with a subversive so-
cial message (see Stam, 1989, on the films of Bunuel, Godard, and Monty
Python), or a combination of the distressing and the absurd, as in various
types of comedy (Milner Davis, 1977). In either case the audience invariably
succumbs to this essence.
Both magic-window reality and utility reality are based on socially shared
knowledge, that is, on stereotypes and cultural beliefs. There may be two-way
traffic between cultural beliefs and individually held ones. Because of self-
selection mechanisms, we would not expect major shifts of worldviews in the
natural audience as a result of watching a movie. Natural viewers tend to be
attracted by the kind of film that fits their beliefs. However, traditional fea-
74 CHAPTER 4

ture films can contribute to the fine tuning of social representations in their
natural audience. This may involve minute adjustments of social stereotypes
held in particular groups, but also some reduction of individual differences
among viewers. Minor deviations of the film's representations from social rep-
resentations, particularly where less crucial subjects are concerned, may lead
the individual viewer to adjust some attitudes, without rejecting them. Ger-
rig and Prentice (1987) provided experimental evidence that written stories
may indeed have this effect.
Differences between viewer groups with respect to stereotypes and beliefs
may be responsible for genre preferences, with a mediating role reserved for
emotion. What one viewer sees as reality may strike another as artificial and
insincere. And if an individual who does not belong to the natural audience
of the traditional feature film were to see that same film, it would probably
be experienced as less real, and the emotional response would be less intense.
Many people hate sentimental films, because they are unwilling—for even
one moment—to believe in the solutions presented.
The power of social stereotypes becomes almost palpable in the case of
films from other cultures or older films from one's own culture. These older
films are often less realistic because the social representations to which they
appeal have long since changed. The contents of these representations are
no longer taken for granted. Zeegers (1988) showed that it is precisely the
self-evident aspects of social representations that are constantly subject to
change. As a result, they lack emotional impact, or evoke other emotions than
the original ones (e.g., exhilaration instead of pity or fear).

Identity Reality
With respect to Potter's identity dimension, we can be brief. In his dis-
cussion of the role of empathy in understanding a work of art, Gombrich
(1970) noted that the experience that comes from inside may be the basis
for the illusion of authenticity, in the sense of likeness to reality. Feature films
create reality by presenting events as seen through the eyes of one or more
protagonists, one effect of which is to make them seem more real to us. There
is, however, an aspect of the relationship between identification and sense of
reality that might complicate the discussion, namely, the fact that a strong
sense of reality also intensifies one's identification with the protagonist. In
chapter 6 we return to this complication.

THE LAW OF CLOSURE AND EMOTIONAL

ACTION TENDENCIES IN FILM VIEWING

One of the core propositions of the functional theory of emotion is that emo-
tions tend to persist. All emotions are characterized by an urge to act in one
FILM AND EMOTION: THEORETICAL BACKGROUND 75

way or another, counter to any other cognitions or influences from outside.

Once the emotion has arisen, the action tendency will have control prece-
dence over other activities until the emotion episode is closed due to a change
in the situation. This is Frijda's (1988) law of closure. What action tenden-
cies are typical of emotion in the film viewer?
There are few film theorists who have not pointed out the great paradox
of the cinema: the fact that motion, the very essence of film, is accompanied
by the no less essential motionlessness of the spectator. It is not difficult to
associate this motionlessness of the viewer with the illusion that one is a wit-
ness to the events taking place in the fictional world. The viewer imagines
himself or herself to be an onlooker, a witness who is not only invisible but is
also incapable of any relational action. When the protagonist is unknowingly
under threat, the viewer cannot grab the gun or even shout a warning to the
potential victim. It must be stressed that the realization of one's impotence
is not based on the fact that one is watching the projection of a staged world.
It is rather the result of the role of spectator that the film imposes upon view-
ers, comparable to that of a spectator sitting in a real stadium and watching
a real football game, or the train passenger who witnesses a murder being
committed in a passing train. There is nothing one can do except watch or
turn away. In the case of film viewing the choices are even more limited. All
we can see is what is on the screen; our gazes are directed automatically to
the spot where the action is.
Thus, in effect, the action repertoire of the viewer in his or her capacity of
spectator to the fictional events is empty. Many, but not all, F emotions are
what Frijda calls feelings (1986), that is, emotions that evoke a virtual action
tendency, one that does not compete for precedence with other actions, ei-
ther planned or in progress. This action tendency differs from other emotions
only with respect to this latter aspect; the appraisal of the situation in rela-
tion to concerns is no different from that which gives rise to complete emo-
tions. There may also be a nonvirtual action tendency (e.g., to help a victim),
but this is transformed into a virtual action, an imagination, say, or a desire,
so that it does not actually require action. Nevertheless, a feeling is not nec-
essarily experienced any less intensely than the corresponding emotion with
a genuine action tendency. We are still dealing here with true emotions: there
is a surface concern involved, and the core components of the situational
meaning are present. The action tendency can be virtual and still contribute
to the intensity of the emotional experience. For example, in Bad Day at
Black Rock (1955) the feeling of anger and the intention to teach the ene-
mies of the one-armed sheriff a lesson are intense, more so because one can-
not act. It would be more precise to paraphrase the feeling as "When are they
finally going to get their comeuppance" or "I'd like to get my hands on them,"
rather than "If they don't stop, I'm going to go up there and wallop them
myself."
76 CHAPTER 4

In a sense the traditional feature film offers desires and longings in place
of genuine action tendencies: "I wish Spencer Tracy would go out there and
get 'em." The film narrative fulfills such desires, "This is what we've been wait-
ing for!" Furthermore, while the viewer cannot undertake any action leading
to contact or interaction, his or her readiness to continue watching is subject
to change. Ensuing changes in viewer interest are dealt with in the following
chapter.
The limited repertoire of possibilities for action that is inherent in the role
of film viewer should not be seen as a shortcoming in the medium. As we have
seen, the experiencing of emotion from a safe distance is one of the prime
pragmatic characteristics of film viewing. A terrifying situation is entertain-
ing precisely because you can do no more than watch; if you were in a posi-
tion to intervene, in order to protect yourself and others, then you would feel
responsible and would no longer be able to enjoy the fictional events on the
screen. But there is a more important, if less obvious, point to be made here:
not only are viewers not capable of taking action, they are not called upon to
do so. The invisible witness is not addressed, indeed, not even ignored: for
the characters of the fictional world the witness simply does not exist. This is
clearly a happy circumstance when it comes to witnessing misery, but even
the contemplation of pleasant events might become less enjoyable if the
viewer were one of the actors or even a visible witness. For instance, in the
latter case, the witness would be less likely to approve of the revenge meted
out to the villain. More generally, many morally dubious practices on the part
of the heroes taking place in the fictional world would no longer be given a
natural benefit of the doubt were the invisible witness addressable by the
action.
The inability to take action is an established and recognized component
of the situational meaning structure and thus of the emotion evoked by fea-
ture films. It creates a certain measure of distance without detracting from
the intensity of the feelings experienced.
In the case of the action tendencies made possible by the feature film as
artefact, it is a different—and simpler—story. The artefact, the projection on
the screen, does not rule out actual relational actions. Film viewers can watch
the screen or turn away; they can look around the theater or think about
something else. They can even get up and walk away. But as a rule changes
in action readiness related to the film as artefact are still the result of the situ-
ational meaning related to the fictional events on the screen. In the little town
featured in High Noon (1952) the situation is becoming more tense by the
minute, as the sheriff, newly married, faces a major confrontation. That is
why we stare spellbound at the screen, ignoring the rustling of a candy-bar
wrapper next to us.
FILM AND EMOTION: THEORETICAL BACKGROUND 77

THE LAW OF CARE FOR CONSEQUENCE:

REGULATION OF EMOTION IN FILM VIEWING

Frijda's (1988) Law of Care for Consequence states that in every emotion
there is a tendency toward regulation. Any display of strong emotion has its
risks. For instance, blind rage brings along loss of esteem of others and self.
The subject has a concern for safety. Although experiencing emotion is one
of the very reasons for going to the cinema, as we have already seen, if some
major film produced emotions of extreme intensities, say, fear or disgust, it
probably would be avoided by most film viewers.
The most obvious way of regulating emotion in a theater is by challenging
the reality of the events taking place on the screen. We know from our dis-
cussion of the coding of film fiction as more or less real that viewers do have
some say in that coding. No matter how tempting it is for them to believe in
the film reality, they can still cling to the knowledge that everything they are
seeing on the screen is fake. Of course, few viewers are likely to take such a
radical stance; they may, however, from time to time take a step backward,
distancing themselves from the reality of the fiction being played out before
them. Fiske (1988) spoke of a kind of shuttling back and forth between "im-
plication in" and "extrication from" that fiction.57 It is interesting to note that
viewers make use of this strategy in order to enhance their viewing enjoy-
ment. According to Fiske, they are able to adapt the degree of implication to
the sympathy they feel for one of the characters. More importantly, the strat-
egy implies that there are gradations in our perception of reality that are to
some extent subject to conscious control.
It is, however, reasonable to assume that there are limits to the conscious
control that viewers of feature films can exercise over the perception of real-
ity. Lang (1984) recorded considerable emotional effects of dramatic pre-
sentations, despite the fact that the subjects were informed in advance that
the situations were not real. We know from the well-known research on stress
and coping carried out by Lazarus and his coworkers (Koriat, Melkman, Aver-
ill & Lazarus, 1972; Lazarus, Speisman, Mordkoff, & Davison, 1962; Nomi-
kos, Opton, Averill, & Lazarus, 1968) that the strategy that consists of de-
nying the reality of the accidents depicted in the film reduces but does not
eliminate the viewer's experience of that reality. The results recorded by
Horowitz (1976) also deserve a mention here; he found that the use of intel-
lectualization as a defense in viewing a stressful film did reduce the degree
of emotion but that the feelings of fear later recurred in full intensity.58 Ap-
57
Fiske is referring here to television drama. However, the difference between television and
film viewing is not relevant here.
58
A comparable effect has been recorded for sexually stimulating films. Fisher and Byrne
(1978) concluded that subjects who consider an erotic film to be "pornographic," and for this
78 CHAPTER 4

parently then, in actual viewing, the experience of reality has not entirely
been switched off.
Yet we would perhaps be justified in asking whether the viewer is at all in-
terested in either denying or weakening reality in the sense described earlier.
An important answer to the question of the freedom of the viewer is that only
a very limited potential for spontaneous initiatives may be desirable for him
or her. As we saw at the end of chapter 2 and the beginning of the present
chapter, there is reason to believe that the viewer is eager for various affec-
tive gratifications built into the film performance, as in a game. Understand-
ably, as viewers we are not out to spoil the game; on the contrary, as a ratio-
nal consumer we will try to get the maximum emotional potential out of the
film. Coding of fictional events promotes that strategy.
Thus it appears that the viewer's perception of reality is not unrelated to
concerns. It is no accident that at first glance the three dimensions of reality
identified by Potter display a similarity to the concerns served by watching
feature films. The magic window satisfies the concern that consists in watch-
ing events as a spectacle, especially the comings and goings of other people
from a safe distance. Utility reality confirms the social identity one shares
with others, the sense of belonging, and the discovery or rediscovery of truths
and possibilities, whereas identity reality satisfies the need to lose oneself in
something greater than oneself.
Finally, we must not forget that the film itself may well be the most im-
portant source of emotion regulation. As we have seen, a fixed component of
the situational meaning structure is the distance created by the absence of
both the reason and the opportunity for action. The narrative evokes emo-
tion, while at the same time steering the ultimate fate of that emotion.
Broadly speaking, everything we have said about the evocation and mainte-
nance of emotion can also be employed to tone down emotion. It is also true
to say that the realization of emotion A in the viewer is always accompanied
by regulation, in the sense of a reduced probability of emotion B. Appropri-
ate emotions, those intended by the filmmaker, exclude less appropriate
ones. The narrative controls the development of any emotion in time. It would
be interesting to examine exactly how films both evoke and regulate emotion.
On the basis of our present level of knowledge, we can say that an important
point of departure here is secondary appraisal (Frijda, 1993). Fear, once
evoked, is made specific, as it were, by the further development of the scene.
Its focus may narrow; the pace of threatening events may be increased or al-
lowed to slow down; the protagonist may accept a loss more easily than ex-
pected, or a surprise happening may lend a twist to events, something that is

reason attempt to diminish its impact, display a greater increase in sexual activity in the two days
following on the experiment than the subjects who had not seen the film as particularly porno-
graphic.
FILM AND EMOTION: THEORETICAL BACKGROUND 79

often used to provide comic relief. There are any number of possibilities, all
of which merit further study. An important form of regulation lies in the fact
that the narrative is organized into episodes: the close of each episode rounds
off the emotion that it has evoked. At the end of chapter 2 we postulated that
the primary aim of the viewer is the experience of a certain emotion episode.
Regulation is inherent in such an episode, in which tension is followed by a
relaxation of tension. A negative emotion is allowed to escalate precisely be-
cause it will later be resolved. And, given the prospect of a resolution, the
more intense it is, the better. Finally, as part of the larger appraisal, there is
also the knowledge that even the larger episode of the film itself must ulti-
mately come to an end and with it all emotion.

Maximization of Cinema Emotion

The conditions created by a good cinematic performance, quite apart from
the film narrative, help to ensure that it is difficult to distance oneself from
the fiction. We now examine briefly a number of those qualities that, because
they assist viewers in their activity, may shed some light on strategies for
heightening viewing pleasure.
First, we may assume that an overall effect is brought about by the pres-
ence of an audience around us: emotions tend to be contagious. On the ba-
sis of cinema experience, we would be justified in saying that the greater the
number of spectators, the easier it is to react in an emotional way. This effect
has been confirmed by at least one field study (Levy & Fenley, 1979), which
found a positive correlation between the size of the audience and the pro-
portion of laughers recorded by observers. The mechanism of contagion, on
the other hand, has not yet been satisfactorily explained. It may be that in
the beginning overt behavior is imitated, which subsequently reinforces the
emotional experience as a whole.59
Second, the cinema is characterized by specific physical conditions that
may reinforce one another. Rabbie and Visser (1984) studied the individual
contribution of darkness, anonymity, and physical nearness of other specta-
tors to the expression of emotion. The aim of the research was to establish
whether manifestations of collective impulsive behavior are attributable to a
situation of deindividuation (Zimbardo, 1969) or to special social norms that
take shape. Deindividuation is part of the filmic state (chap. 2), and is char-
acterized by an increase in impulsiveness and a decrease in self-control; the
subject is particularly susceptible to contagion by the behavior of a collective.
Rabbie and Visser's results show that acquaintance versus anonymity is

59
The most important element in such a peripheral influence on emotion might be "facial
feedback," in other words, the awareness of one's own facial motor functions (Duclos et al., 1989;
Ekman, Levenson, & Friesen, 1983; Levenson, Ekman, & Friesen, 1990; Leventhal, 1979).
80 CHAPTER 4

the prime factor in determining the degree to which one feels free to express
emotion. People who know each other feel freer and more comfortable and
are less likely to be inhibited by one another. The results suggest that people
who do not know each other would be well advised to sit further apart if they
wish to feel free to express emotion, while friends or acquaintances should
sit closer together. Among people who know each other, it is apparently not
so much a question of deindividuation as it is a kind of norm that encourages
the expression of emotion.
In the pattern of these experimental results there are interesting parallels
with the situation in the theater during the showing of a traditional feature
film, which attracts larger groups of filmgoers than, say, the art genres (Faber,
O'Guinn & Hardy, 1988; Palmgreen, Cook, Harvill & Helm, 1988). Broadly
speaking, acquaintances sit closer together, thus reducing each other's inhi-
bitions, while strangers, again broadly speaking, sit further away from one an-
other, so that the inhibiting effect of the presence of others, already reduced
by the darkness, is less marked. As Rabbie and Visser indicated, the norm ex-
planation would appear to be more plausible than the deindividuation hy-
pothesis. 60 In their view, the free expression of emotion will manifest itself
most forcefully when there is already a norm present. This is often the case
in the theater. Certain theaters have a reputation for providing a particular
type of entertainment and behavior, while many traditional feature films an-
ticipate an emotional response even before they are released: the members
of the audience come to the theater prepared to laugh, cry, be frightened,
and so forth. The younger viewers, in particular, go to see films that may be
expected to evoke strong emotions; these emotions are even more pleasur-
able because they are shared with friends and with the rest of the audience.
All things considered, the theater situation appears to meet the individ-
ual's desire to freely experience emotions, guided by the qualities of the film
in which those emotions are rooted. What the facilities provided by the the-
ater offer viewers is an opportunity to get the maximum amount of emotion
out of a film. The process of shuttling back and forth between implication and
extrication described by Fiske (1988) would appear to be both easier and more
appropriate where the initiative of the viewer is concerned when one is sit-
ting at home in front of the TV than in a film theater.

The Freedom of the Viewer

It is now time to replace the question about the regulation of emotion put for-
ward at the beginning of this section by an improved version: What reason do

60
It should, however, be added that the previously mentioned experiment by Levy & Fenley
(1979) provided indications—albeit indirect ones—for the assumption that social facilitation has
a greater influence on audience-size effect than conformity.
FILM AND EMOTION: THEORETICAL BACKGROUND 81

people watching a feature film have to direct their attention toward the arti-
ficiality of the fiction? The question is more or less rhetorical, as there is ac-
tually only one logical answer: because the essence of what is being depicted
on the screen is not to their liking. This is fairly unlikely in the case of the
natural viewer of the traditional feature film: self-selection ensures that the
average viewer ends up watching the film best suited to him or her.
Should regulation in the cinema be deemed necessary, it would be more
correct to say that viewers can maximize their pleasure by not resisting the
emotion, simply doing what the film expects of them. All the qualities of a
good cinematic performance are designed to help them do so. In incidental
cases, when imagination threatens to overpower them, viewers can always
remind themselves that what they are watching is fiction. But we may assume
that the reverse is just as easy: when doubt or skepticism rears its ugly head
viewers are free to suspend disbelief, so as not to interfere with their own
pleasure.
The Laws of the Lightest Load and the Greatest Gain (Frijda, 1988) are
applicable to the film situation as well as to daily life, although the coding of
the stimulus is somewhat more complicated. In everyday life the subject
strives to code the situation in such a way that there is as little occasion as
possible for negative (i.e., unpleasant) emotions (the Lightest Load). At the
same time the coding is guided in the direction of the hedonically positive
emotions (the Greatest Gain). During the viewing of a feature film, coding of
the stimulus ignores the rule that the negative charge of such emotions as
fear and sadness must be kept to a minimum. For the film viewer it may be
shivers and tears that provide the greatest gain by contributing to the build-
up and release of tension.

AN OVERVIEW OF THE TYPES OF EMOTION

EXPERIENCED BY THE FILM VIEWER

In concluding this chapter, it may be useful to review the types of emotion

evoked in the viewer by the traditional feature film. These are summarized in
Fig. 3.1.
The first classification of film emotions is related to the aspect of the stim-
ulus on which the situational meaning is based, namely, either the fictional
action or the film as artefact. In one case we are dealing with F emotions, and
in the other with A emotions. In the traditional feature film, F emotions are
dominant; the diegetic effect dictates that the viewer is interested primarily
in the events taking place in the fictional world, seen from a safe vantage
point, whereas artefact characteristics are generally hidden. In the light of
the match between traditional feature films and their natural audience, A
emotions are generally positive, since the pragmatic rules are respected. A
82 CHAPTER 4

EMPATHY NO EMPATHY

Feelings Feelings

FICTION Emotions Emotions

EMOTIONS proper proper Interest

Enjoyments Enjoyments
and desires and desires

Emotions Emotions
proper proper
ARTEFACT
Interest
EMOTIONS Enjoyments Enjoyments
and desires and desires

FIG. 3.1. Classification of the types of emotion experienced by the film viewer.

emotions may follow in the wake of intense F emotions, as filmic surface con-
cerns that make up traditional film taste are gratified.
Second, emotions can be characterized according to the criterion of em-
pathy. It is not difficult to imagine what is meant by empathetic F emotions:
an obvious example is pity for one of the characters. Empathetic F emotion
is based on an understanding of or "feeling into" fictional characters. But
there are also empathetic A emotions, which have to do with synthetic pro-
prioceptive activity, such as mirroring a certain type of movement on the
screen. One example might be the viewer's delight at a whirling camera move-
ment in combination with a lyrical sound track. The major A emotions are,
however, nonempathetic. These include enjoyment, desire (for example, one
hopes for the return of an element that one particularly enjoyed), admiration,
and astonishment.
Thanks to the diegetic effect, all F emotions are witness emotions, compar-
able to affect evoked by the sight of nonfictional emotional events in real life.
Most F emotions are empathetic because the action in the traditional feature
film narration is realized by protagonists who display human traits and whose
goals and fate are of interest to the film viewer. These include such emotions
as hope and fear, anxiety, sympathy, pity, relief, gratitude, admiration, shame,
anger, terror, joy, and sorrow. These are discussed in detail in chapter 6.
Yet nonempathetic F emotions are quite common. It is possible—analyti-
cally, at any rate—to distinguish between, on the one hand, fear of witness-
ing certain fictional events or a desire to watch other events and, on the other
FILM AND EMOTION: THEORETICAL BACKGROUND 83

hand, anxiety about the fate of a protagonist or a desire for the resolution fa-
vorable to him or her. Disgust is an emotion that is nearly always nonempa-
thetie: watching a torture scene fills the viewer not only with empathetic pity
for the individual being tortured and loathing for the torturer but also revul-
sion for the action and all its visual details as such. Spectacle in film is, as the
term implies, appealing, simply because it is largely divorced from the fate of
the protagonists. And there is a great deal of spectacle in films because the
medium itself is spectacular. Few film plots are set in a totally empty space,
and there is always some aspect of the background to enjoy, from a breath-
taking landscape to indoor spaces that most people have never been privi-
leged to enter (the editorial offices of a daily newspaper, an air force base,
the headquarters of the German army, or a jungle camp). And through the
invisible witness effect, even the most soberly executed Kammerspiel offers
the viewer an opportunity to observe the outward appearance and the actions
of people quite separate from the significance that a particular situation has
for them. This is in fact one of the most important primary motives for watch-
ing feature films.
The passive role of the spectator in the fiction that the diegetic effect im-
poses on the viewer means that many F emotions may be characterized as feel-
ings, that is, the associated action tendency is virtually or totally absent. This
does not, however, hold true for the complex of emotions that are examined
in detail in chapter 4, namely, interest The action tendency in the case of emo-
tional interest consists of an increase in the readiness to follow the fictional
happenings. This is based on the anticipation of coming events and their ac-
companying emotions. Interest is a tonic emotion. The narrative structure of
traditional feature films is such that expectations are constantly being intro-
duced about what is about to happen. Some of these expectations, including
the more important ones, are projected far into the future and have a long
survival time, so that changes in the intensity of viewer interest take place
relatively slowly. This is in contrast to most of the other F emotions, which
are evoked by more rapid changes in the situation; these become habitual
much more quickly and thus may be characterized as phasic emotions.
The object of interest is both the fictional world and the artefact. The
events of the fictional world hold promises of things to come; they encourage
the viewer to follow attentively the action taking place in order to learn more
about that world. But the artefact, too, encourages the activity of the viewer,
through its development and composition, as in the area of plot and style.
Both interest fostered by the fiction and aesthetic interest ultimately direct
the viewer's attention to the artefact and nothing but the artefact. Because
my main concern is whether or not Gary Grant and Eva Marie Saint are about
to plunge to their death in the closing scenes of North by Northwest (1959),
my attention is captured first and foremost by the film itself.
The survey of emotions evoked by feature films is far from complete. It
84 CHAPTER 4

does not entirely do justice to the dynamics of emotion over time. In chapter
7 the survey is refined with regard to this aspect; we will also be better able
to flesh out the various emotions.
 The Structure of Interest
4

INTEREST AS EMOTION

The act of watching any feature film is accompanied by interest. The nature
and intensity of that interest are such that one is tempted to speculate on the
mechanisms behind the phenomenon. Films hold their audiences spell-
bound, bewitched, fascinated, mesmerized, and captivated. These are the
terms used not only by the advertising copywriters but also by professional
film critics—who are by nature more reserved in their judgments—and film
theorists. The regular filmgoer would have no quarrel with the terminology.
He or she would freely admit to becoming wrapped up in the plot and find-
ing it difficult, if not impossible, to detach himself or herself from the spec-
tacle on the screen. The germ of an explanation may well lie in the character-
ization of this phenomenon as emotional: the intense interest in the action
on the screen is the result of or even part of an emotion, and this very fact
makes it difficult to suppress.
In various theories of emotion, interest is seen as a basic emotion, that is,
one that cannot be reduced to one or more other emotions. (Frijda, 1986;
Izard, 1977, 1992; Panksepp, 1982; Tomkins, 1984). The status of interest as
an emotion is somewhat controversial.61 The major criticism of this view is
61
The trouble begins with the very concept of "basic emotion." Doubt has been expressed
about the possibility of establishing criteria for defining such a concept (Ortony, Glore, & Collins,
1988; Ortony & Turner, 1990; Turner & Ortony, 1992). But this doubt has also been repeatedly
contradicted (Ekman, 1992; Izard, 1992; Panksepp, 1992).

85
86 CHAPTER 4

that expressed by Ortony and Turner (1990), who do not consider interest a
valenced reaction. They suggest that interest belongs to the cognitive rather
than the affective states. One of the purposes of this chapter is to show—with
special reference to film theory—that interest, like any other emotion, is
based on concerns, that the film stimulus occasions a specific situational
meaning structure, and that together these two things give rise to a tendency
toward action. The concerns that are relevant for the creation of interest
would appear to be fairly obvious: the satisfaction of curiosity, the assimila-
tion of cognitive structures, and the need for tension and diversity. As we shall
see, other concerns may also influence interest evoked by the act of watch-
ing a feature film, notably sympathy and value concerns. The situational
meaning structure, which is based on the illusion of presence in the fictional
world, involves a promise, that is, the situation gives rise to strong expecta-
tions, no matter how diffuse and unarticulated, with respect to other, subse-
quent situations within the film. These aniticpated situations appear to be
highly emotional and, moreover, offer at least some prospect of a satisfying
denouement in the fictional world.
The action tendency may lie in the realm of approach and exploration,
however bounded it is by the role of spectator or witness imposed by the
diegetic effect. As a witness, one cannot take any part in the action. But the
witness is able to follow intently the action of the film, speculating about com-
ing developments by forming expectations and searching for confirmation of
those expectations. The action tendency distinguishes interest from the un-
emotional, cold cognitive states of attention described by such theorists as
Ortony and Turner (1990). Like all action tendencies, it attempts to drive
other conduct, and with it contradictory cognitions and perceptions, into the
background. In other words, interest as an emotion displays a strong control
precedence. Subjectively, this emotion is variously referred to as fascination,
enthrallment, tension, and absorption, in preference to the rather bland term
interest.
In this chapter we will examine the concept of interest as emotion and
attempt to make it more plausible. We make use of the following working
definition:
Definition 4.1. By interest we mean the inclination to call on re-
sources from a limited capacity, and to employ them for the elaboration
of a stimulus, under the influence of the promises which are inherent
in the present situation with respect to expected situations.
Interest during the act of watching films is of a hedonically positive tone
because the promises imply the realization of the concerns involved, even
thought this may only take place at a later stage. In chapters 5 and 6 we ex-
amine in detail the stimulus characteristics of the feature film that are rele-
vant for the core of the situational meaning structure for the promise of the
T H E STRUCTURE OF INTEREST 87

situation as appraised. They include aspects of the action themes and the per-
ceived traits of the various characters. This chapter centers on the principles
that the mechanisms of interest in film viewing may satisfy.

Interest and attention

It is necessary to distinguish interest from such unconscious responses as
startle and orientation reflexes, which is not to say that these are of no im-
portance for the intensity of interest. However, unlike interest, these re-
sponses do not involve an inclination toward specific elaborations. Given the
formulation of the question on the role of affect in film viewing, however, the
relationship with explorative behavior is of importance; in our working defi-
nition of interest the freeing of resources at the expense of other activity is
stressed. Interest has more chance of arising "to the extent that one is free
of survival needs and negative emotions" (Izard, 1977, p. 212; see for com-
parable views such authors as Apter, 1984 and Nunnally, 1981). The fact that
it is not associated with primary needs does not mean that interest is a lux-
ury, a kind of extra we do not get around to until all the other necessities of
life have been provided for. Izard, for example, believes that interest may be
seen as the "most prevalent motivational condition for the day-to-day func-
tioning of human beings."(Izard, 1977, p. 211). And from the standpoint of
environmental psychology, Kaplan and Kaplan (1982) postulated that inter-
est is necessary in one's normal contact with the environment, namely, in
order to acquire familiarity with the environment.
In theoretical propositions concerning interest, there is always some kind
of link between interest and attention. The two concepts are almost synony-
mous where interest is seen as a cognitive state (Ortony, Clore, & Foss, 1987)
or a perceptual state (Hochberg, 1970). Attention does not in itself presup-
pose a tendency to act, as interest does. In Frijda's theory, interest is the ten-
dency to acquire knowledge or certainty (Frijda, 1986) in a situation that is
characterized by strangeness or unfamiliarity, and clear points of attention
(focality) and is related to an event rather than an object. The fact that in-
terest in the sense of willingness to bestow attention differs from the actual
bestowal of attention will be clear from the unusual case where a film frag-
ment fails to evoke the interest of viewers. Even when the level of interest is
low, it is possible to attract the attention of the viewer by mechanical means:
a rapid succession of shot changes (Hochberg & Brooks, 1978), say, or un-
predictable movements in the peripheral field of vision.
Interest involves not only intensity but also selectivity of attention. As we
have seen, feature films steer attention to an extraordinary degree, by ren-
dering salient those elements of the fictional situation that constitute the sit-
uational meaning structure appropriate to a particular emotion. Reverse ef-
fects may also occur. The nature of the emotion steers the selectivity of at-
88 CHAPTER 4

tention, a familiar phenomenon to which emotion owes its reputation for ir-
rationality. This point was examined in chapter 3. Various emotion theories
have as one of their major postulates the control of cognition, including at-
tention. These include, alongside Frijda's theory, the evolutionary-functional
emotion theory of Klinger (1977) and the differential emotion theory of Izard
(1989, 1992). However, according to Ritchie (1986), there has been very
little empirical research focusing on the way attention is influenced by emo-
tion. Ritchie himself demonstrated the effect of emotion on visual attention.
Subjects were asked to look at a monitor portraying a series of faces with hats
on. The faces all displayed schematic representations of emotions. One of the
hats was always different from the others, and the subjects had to give the
name of the hat that was different (i.e., the target). By means of bonuses in
one condition, and fines in the other, a momentary affect state was created.
The results showed that the visual search time was shorter when the target
lay in an area that contained schematic information that corresponded to the
momentary emotion of the subject.
This same emotionally steered selective attention may form the basis of
mood-congruent learning and recognition (Bower, 1981; Bower & Cohen,
1982; Isen, 1984; Isen, Shalker, Clark, & Karp, 1978). Feature films often
make use of this phenomenon. For instance, it would appear plausible that
threatening stimuli that are important to the narrative, such as a knife, can
be shown more casually in a scene that the viewer already finds scary. As we
saw in chapter 3, however, the cognitive structure of the action and the result-
ing specific expectations and anticipations are far more important in deter-
mining the direction of attention than a diffuse emotional state. In addition,
the filmmaker has at his or her disposal a wealth of mechanical techniques
to control attention, such as inserting a sudden close-up of a detail.

Behavioral consequences
As we have seen, one of the features considered characteristic of the state of
the film viewer is a kind of freeze, the result of a total absorption in the per-
ception. Arnheim (1958/1983) describes how interest can be suggested in a
silent film. In Chicago (1927), 62 for example, we see a group of factory girls
watching a trial. They are shown sitting in a row, chewing gum. All of a sud-
den, as the tension in the courtroom reaches its peak, they all stop chewing.
In the same way, the viewer may freeze when the higher levels of interest are
reached in following the fictional action. Conversely, in the late 19th century,
Francis Galton proposed that the amount of fidgeting or wiggling on the part

62
In the book referred to, Arnheim incorrectly attributes the film to Cecil B. DeMille. The di-
rector was Frank Urson; DeMille supervised the production. I am indebted to Ruud Bishoff who
found out the correct filmography of Chicago.
THE STRUCTURE OF INTEREST 89

of the audience should be used to measure boredom. The experienced per-

former is quick to catch the slightest shifting or rustling in the audience, sig-
nifying that they are in danger of becoming bored. Experiments by Kretsinger
(1952) have shown that an interesting radio program is accompanied by less
"gross bodily movement" than the tape of a dry lecture.
It is a generally accepted fact in the psychobiology of attention that alert-
ness inhibits motor activity. (See, Izard, 1992; Posner, 1975; and, in particu-
lar, Obrist, 1981). Pribram and McGuiness described the neural systems that
are responsible for the control of perceptual attention and motor activity
(McGuiness & Pribram, 1980; Pribram & McGuiness, 1975) and found that
there are clear similarities to emotional aspects of behavior. There is, how-
ever, a complicating factor: the fact that when behavioral activity brings the
desired object closer, interest, or, at any rate, anticipatory eagerness, may be
accompanied by excitement and excessive movement.
The definition of interest as the inclination to bestow on the stimulus a
great deal of attention drawn from a limited capacity is supported by an ex-
periment carried out by Anthony and Graham (1985). Interesting foreground
stimuli, such as music and color slides of human faces, led to a greater re-
duction in cross-modal probe response than dull stimuli, such as isolated pure
tones and transparent slides. When the foreground stimulus was an interest-
ing color slide, the subject responded with a less pronounced blinking reflex
to a sudden loud noise in the background than when a somewhat boring slide
was being shown. In a comparable experiment (Simons & Zelson, 1985),
stimuli were used that might be expected to produce stronger emotional in-
terest. Subjects were shown a stimulus, and, at the same time, unpredictable
auditory stimuli were presented. It appeared that the presentation of color
slides of male and female nudes was accompanied by a less pronounced
startle-blink reflex in response to the probes than the presentation of slides
showing straw baskets. And finally, there is the experiment by Neisser and
Becklen (1975) that illustrates the fact that anticipatory elaboration of a sim-
ple film can make considerable demands on a subject's capacity for attention.
They showed their subjects two short films projected one on top of the other
and asked them to follow one of the two stimuli. The subjects were unable to
give any details of the film to which their attention was not directed. The re-
searchers concluded that cognitive operations involved in following a very
simple action, structured in such a way as to evoke anticipations—in one case
a ball game and in the other a game played with the hands—took up the en-
tire attention capacity of the subjects. This was not due to the operation of
some form of filter mechanism, but rather to the subject's anticipatory ac-
tivity triggered by the target stimulus.
We may conclude that the present level of informal observation and psy-
chological theory with respect to the characteristic behavior of the involved
film viewer in no way conflicts with the view that interest is an emotion. Con-
90 CHAPTER 4

trary to the views of Ortony and Turner (1990), it is quite possible that in-
terest is a valenced reaction, as witnessed by the resistance to distraction. In
Frijda's theory of emotion, the action readiness is a primary feature of true
emotion. And during the act of watching a film, interest is marked by the in-
clination to devote one's full attention to the stimulus, at the cost of all other
matters, including the completion of motor programs.

The Experience of Interest

The description of interest during the act of watching a film would be in-
complete without a discussion of certain important aspects of the character-
istic experiential phenomena. If we look at interest in a somewhat broader
sense, then contributions to psychology, literature theory, and drama theory
may offer useful points of departure. We now examine the following features:
the flow aspect, the problem-solving aspect, and the sensation that one is ca-
pable of meeting the cognitive and affective challenge represented by the film
narrative.

Involvement as Flow
The involvement that is characteristic of interest in feature films has been
described by researchers such as Child (1978) as a necessary element of aes-
thetic activity, one that is also seen in other explorative and recreational ac-
tivities. Csikszentmihalyi (1975; Csikszentmihalyi & Csikszentmihalyi, 1988)
spoke of "flow," a state in which the subject appears to merge with his or her
environment; this is sometimes seen when exceptionally high levels of
achievement are reached, as in chess or composing. In this state, it is almost
as if the plan for some efficient form of action presents itself almost effort-
lessly. It is conceivable that the film narrative produces in the viewer an im-
itation of this state, or even the true state. The witness does not invent the
screen developments in the same way the chess player devises moves and the
composer the end of a musical sequence. As we saw in chapter 3, the tradi-
tional film narrative runs extremely smoothly. More important, it allows the
fantasy of the viewer a certain degree of freedom, albeit within a previously
determined plan. The viewer carries out a highly specific mental activity de-
signed to fill in the gaps (Bordwell, 1985), which in the end prove to be sur-
mountable.

THE AESTHETIC ACTIVITY OF THE VIEWER

What we refer to in our working definition of interest as elaboration—a men-

tal activity that lies somewhere between free association or fantasy on the one
hand and a strictly prescribed process of inference controlled by the text, on
THE STRUCTURE OF INTEREST 91

the other—is in the view of some researchers akin to problem solving or work-
ing out a puzzle. Even if we confine ourselves to the more recent psychologi-
cal studies, we see that a great many descriptions of aesthetic activity and
motivation are based on some kind of incongruity between mental represen-
tations that are formed under the influence of the work of art itself. The sub-
ject solves the puzzle, which results in aesthetic pleasure. The incongruity
may be accompanied by tension or arousal, or any other motivational state
that is not directly pleasurable, while the solution always has some positive
emotion attached to it, such as the "aha experience" of the problem solvers
described by Duncker (1945, p. 2). 63
Along these same lines, the structuralist theory of literature stresses the
resolution of inconsistency by the reader: smoothing away strangeness (Cul-
ler, 1975), resolving tension (Preminger, 1965), solving incoherences (Ker-
mode, 1967), filling in gaps (Sternberg, 1978), and completion of qualities
such as abstract forms, promised qualities, and expectations based on con-
ventions (Booth, 1961). Finally, in the field of drama theory, there is a simi-
lar conception of what tension is, namely, "partielle Informiertheit" (Pfister,
1977; Piitz, 1970).
The idea that the effort required to create order out of complexity is the
source of aesthetic pleasure is also in line with the hypothesis that cognitive
interest consists in the solution of not too complex incongruities between in-
formation and knowledge based on schemata or between two competing
schemata. This is a popular hypothesis in current research into cognitive pro-
cessing of stories (Hidi & Baird, 1986; Kintsch, 1980). It is most clearly ar-
ticulated by Graesser (1981 and, in particular, Graesser & Clark, 1985).
Fairly simple stories are ultimately understood through the instantiation of a
number of generic knowledge structures that only roughly match the input.
These later converge in a smaller number of general knowledge structures
that overlap and are selectively retained. Oatley associates the pleasure af-
forded by reading stories with the Aristotelian idea that happiness is to be
found in complete surrender to an activity (Oatley, 1995). According to Oat-
ley, the activity consists in the act of assimilating into schemata and the as-
similation of schemata.

63
For general formulations, see such authors as Berlyne (1974), Bever (1986), Carver and
Scheier (1990), Child (1969), Gaver and Mandler (1987), Hochberg (1978), Humphrey (1973),
Kaplan and Kaplan (1982), Kreitler and Kreitler (1972), Lasher, Carroll, and Bever (1983), Man-
dler (1982, 1984), Meyer (1956), Miall (1989), Nunnally (1981), Purcell (1984). Certain cog-
nitive theories of metaphor might also be mentioned in this connection. The relation between
the "vehicle" and the "target" of the metaphor may also be accompanied by tension (Ortony
1978, 1979, 1980). So-called incongruity theories of humor (Godkewitsch, 1974; McGhee,
1971, 1972; Morreal, 1983; Suls, 1972) also fall into this category. According to these theories,
a joke creates a kind of tension between two incongruent elements, which is released by the
punch line.
92 CHAPTER 4

The adequate challenge

The term problem solving may seem a trifle pretentious when used to charac-
terize the experience of the filmgoer watching My Fair Lady (1964) or Brief
Encounter (1945). Yet if we exclude the inappropriate connotations, this
should leave us with some useful starting points. There are two main restric-
tions to this description. First, when a viewer is watching a feature film the
danger of failure is almost totally absent, and second, both the solution and
the path leading to that solution have an important surplus value, over and
above their value as elements of the puzzle. To begin with the latter consid-
eration; viewers do not have the feeling that they are solving a problem by
processing arbitrary abstract elements and relations. 64 It would not be cor-
rect to say that the viewer just keeps on working away at an abstract plot. The
elements of the plot are significant within the fictional world and, moreover,
the puzzle-solver is not indifferent to them. A character is someone, from
whom the viewer expects something and whose fate engenders in him both
hope and fear. The characters, the events, and the outcomes touch upon con-
cerns; the viewer invests concern in what is being played out on the screen.
This fact sets film viewing far apart from the typical problem-solving situa-
tion. The cognitive component of the interest is totally overshadowed by the
affective component. Moreover, it is often difficult to distinguish between the
cognitive and affective sources of uncertainty created by the film, while ex-
perientially they are inextricable. A question such as "When did the action
take place?" is not only relevant when it comes to unravelling the structure
of the plot; it may also determine the fate of a protagonist. When the body of
a woman is discovered in the apartment of a highly sympathetic character
and it is not clear when she was killed, we are not just inclined to figure out
that the killing took place after the man left the apartment and that he is
therefore innocent (The 39 Steps, 1935), indeed, we hope that this is the
case. When the protagonist seems about to be murdered by his lover, not only
does cognitive uncertainty give rise to the question of whether she is playing
a double role but we also fervently hope that this is so. In this way, the wish—
rather than just the expectation—that she does truly care for him can be
maintained (North by Northwest, 1959).
The second inconsistency with respect to the usual connotations of the
term problem solving is concerned with the cognitive level of difficulty. The
natural viewer of the traditional feature film is capable of filling in the gaps
in the film narrative. The pragmatic rules of the traditional feature film make
this possible, and he or she knows that in the end the solutions will always

64
There is the famous example of Newell (Newell & Simon, 1972) in which it is given that
DONALD + GERALD = ROBERT. The letters must be replaced by numbers in such a way that
the sum is correct.
THE STRUCTURE OF INTEREST 93

present themselves. This lends to the experience of the film an aspect of se-
curity complementary to that referred to in chapter 3.
There is somewhat more to be said on the subject of the relative simplic-
ity of the traditional film narrative if we compare it with other forms of fic-
tion. As we have seen, the narrative of the traditional feature film follows that
of the conventional story. There is evidence that the latter genre is easy to
understand and remember, especially in comparison with other forms of
prose (Britton, Graesser, Glynn, Hamilton, & Penland, 1983). Graesser (1981)
found, in addition, that there are far fewer individual differences between
readers with regard to their understanding of a simple story than between
readers of expository texts and other genres. Graesser and Clark (1985) have
shown that, as we postulated earlier, the final representation of a story is de-
termined exclusively by prototypical knowledge. Their analyses of subject
protocols demonstrate that there is very little evidence of new inferences dur-
ing the processing of the narrative and that, with few exceptions, these are
not retained in the ultimate cognitive story structure. It is generally agreed
that in the majority of traditional films—the popular cinema supply—narra-
tive structure is subliterary: plot dominates theme and style, and there is no
striving for form experiments or alienation effects. This means that the viewer
has no difficulty in understanding the average film story. It is precisely the
well-timed sequence of challenges that the viewer is just barely capable of
meeting that constitutes flow. Gsikszentmihalyi (Gsikszentmihalyi & Csik-
szentmihalyi, 1988) maintains that a precondition for flow is a certain bal-
ance between challenge and skill. For those in the creative professions, this
balance is a part of daily life. For the average subject, however, most situations
in everyday life provide either too many challenges or—more often—too few,
represented by those routine things you could do with your eyes shut. Where
it is possible to optimalize these two elements, ordinary individuals can also
attain a state of flow. The feature film clearly meets these conditions, just as
the ritual events, sports, games, and artistic performances specifically men-
tioned by Gsikszentmihalyi do.
We are assuming here that the secret of the Challenge that Everyone can
Meet lies in creating problems that the viewer can attempt to solve but that
are ultimately always solved by the film narrative itself, with no detriment to
the viewer's ego. Thus the ordeal that the problem solver must undergo con-
sists essentially in a delay, one that can be enriched by a game of cat-and-
mouse: expectations concerning the ultimate solution are created that can
subsequently be manipulated in a variety of ways.
Thus the notion of aesthetic interest as a problem-solving activity is useful
only up to a certain point. The fact that a solution has been found is not the
only desirable result of following a film from beginning to end. There is also
the significance and value of that solution in the fictional world. Moreover,
the significance and value of the final situation outside the fictional world may
94 CHAPTER 4

be among the major results, notably when the film presents some suggestion
or possibility of a more general nature, such as some worldview (chaps. 2 and
3). In short, in the feature film, interest depends largely on the prospect of
knowledge outcomes that are charged with concerns.

PRINCIPLES OF INTEREST

Principle 1: Temporal Dynamics and

Real-time Nature of Interest
The structure of the interest experienced by viewers of traditional feature
films, here loosely referred to as the inclination to continue to follow the film,
can be expressed in a number of principles.
We must begin by saying that the formulation of principles serves a heuris-
tic purpose, namely, to take intuitions collected from various sources and to
order and arrange them in such a way that they contribute to the develop-
ment of a theory. This theory, which involves only a relatively small number
of determinants, can then be used to explain the course of interest through-
out the presentation of as many different traditional feature films as possible.
Efforts to find a formal and experimental explanation for interest or the
hedonic appreciation of aesthetic stimuli are not new. After some scattered
attempts dating back to the work of Wundt, the "New experimental aesthet-
ics" (Berlyne, 1974) and the "Informationsasthetik" (Dorner & Vehrs, 1975)
flowered particularly throughout the sixties and seventies. (For a survey of
formal and experimental aesthetics, see Berlyne, 1978; and Hochberg, 1978.)
Existing formulas for hedonic appreciation explain preference for certain arti-
ficial stimuli, in particular geometrical figures, in terms of their cognitive
complexity or arousal value.65 In some cases (Berlyne, 1974), a distinction is
made between appreciation and interest, whereby interest is operationalized
as the amount of attention spent on the stimulus, as measured by inspection
time in the case of visual stimuli.
In light of the discussion of the nature of interest, however, there are cer-
tain shortcomings attached to the existing psychological-aesthetical formu-
las. The first is the fact that, although the response has a certain affective
aspect, the processing of aesthetic stimuli, to the extent that it has been
described, is exclusively cognitive and bears a strong resemblance to simple
problem solving. This cannot be the whole truth. For one thing, purely cog-
nitive accounts fail in telling us what makes the viewer embark on a fairly ex-
65
One example is the model devised by Birkhoff (1933), who is often referred to as the
founder of formal aesthetics. The aesthetic value of a stimulus M is equal to the ratio of O to C,
in which O stands for the number of rewarding order elements in the stimulus, and C for the ef-
fort involved, reflecting the complexitity of the stimulus.
THE STRUCTURE OF INTEREST 95

acting cognitive exercise. The second shortcoming of existing accounts is that

appreciation of the stimulus is assessed on the basis of the full and complete
process of reception. However, in the case of stimuli like the film, which are
presented over time, the dynamics of interest throughout the entire course
of the presentation must be distinguished from the appeal of that same stim-
ulus after it has been fully processed. If there is any functional relationship
between the two at all, that relationship may be far from simple. In addition,
there is a logical objection to equating intermediate interest with final ap-
preciation. The simple postulation that interest registered at a certain point
in the processing of the stimulus is a function of the final result presupposes
a foreknowledge of results not actually available at that point. This means that
the explanation is not psychologically realistic. 66 Thus a musical composition
that, taken as a whole, is assessed as enjoyable or beautiful may well have in-
cluded noticeable dissonances that at the time of appearance were not liked
very much and were only resolved in the course of the piece. Conversely, a
film may create enormous promise of exciting plot developments and a satis-
fying and just ending, but in the final analysis fail to deliver the goods.
A film is a narrative and as such it is characterized by a development over
time. Unfortunately, the disadvantage of theories of literary narrative struc-
ture is that they are not really suited to describing the dynamics of stories.
Martin concluded a discussion of structuralist theories by pointing out their
major limitations: "One such limitation is that they tend to reduce narratives
to a static, atemporal deep structure and thus cannot account for the tensions
and reversals of situations that makes us want to find out what happens next"
(Martin, 1986, p. 105). Martin points to recent efforts to fill this gap. One of
the theories involved describes how dynamic the process of following a story
with interest can be: "each new situation is construed as a revised configu-
ration of the entire sequence up to that point. . . leading to revised expecta-
tions concerning the outcome." (p. 106).
The distinction between final judgment and interest displayed from one
moment to the next is expressed in the following principle.
Principle 1. The appreciation of a film after it has been completely pro-
cessed must be distinguished from momentary interest.
I now attempt to contribute to a psychologically realistic explanation of
the real-time, momentary values of interest. These must be influenced by the
stimulus elements presented earlier in the film but only to the extent that
66
Information-theoretical proposals are available for the reduction in complexity or uncer-
tainty achieved by the "problem-solving subject" (e.g., Giinzenhaiiser, 1965). But these are in
effect quasi-realistic models, since they deal only with the effect of the total reduction in uncer-
tainty, leaving aside the distribution over time and the resulting affective dynamics. In the case
of static stimuli, this argument is of lesser importance than for films and other stimuli that un-
fold over time.
96 CHAPTER 4

these are available in the memory of the subject. Similarly, these values are
influenced by expectations about what is to come to the extent that the ex-
pectations are based on elements already received and are actually active in
the subject's memory.

Principle 2: Interest is Determined

by the Prospect of Return

Preferred Final Situation, Investment and Return

From the very beginning of a film, the viewer is aware of the fact that the
film narrative is leading somewhere, and expectations with respect to the fi-
nal situation gradually become more specific. As noted earlier, the viewer is
not indifferent to the nature of this final situation. A particular final situation
is satisfying to the degree that it satisfies certain cognitive and affective con-
cerns.
The cognitive concern corresponds broadly to the assimilation of the ac-
tion observed, into a canonic narrative structure. That structure is determined
largely by typical causal relationships between successive parts of the action.
Completion of an unambiguous canonic narrative structure satisfies cogni-
tive curiosity. It is completely clear who did what, when and why, and what
the consequences were. This concern is also satisfied by a wealth of different
relationships other than causal ones within the ultimate cognitive represen-
tation of the narrative. The cognitive rewards concerned include symmetry,
ironic parallels and contrasts, and peculiarities of structure with regard to
prototypes, such as genre categories (richly represented in the traditional
cinema).
The affective concern is an end state of the fictional world that corre-
sponds to (a) the sympathies of viewers and (b) their values. The sympathetic
protagonist gains the person or thing she wants and has been striving for,
while the unsympathetic character gets nothing or is punished. A happy end-
ing corresponds to prototypical representations of justice. (For "a just world,"
see Lerner, 1980).67 Bordwell reported that over 60% of classical films had a
happy ending with respect to the romantic story line, (i.e., boy gets girl)
(Bordwell, 1985). This does not mean that a satisfying final situation always
meets all the mentioned, somewhat schematic, conditions, which appear to
be characteristic of certain genres, above all the popular action film. There
67
The extent to which the interaction of values and sympathies, as referred to earlier, deter-
mines the affective return of a narrative has been examined by Jose and Brewer (1984). These
researchers had children of various ages read a number of exciting stores. Using path analyses
of questionnaire data, they found that, regardless of age, children prefer stories about "good"
protagonists. Moreover, from about the age of 10 on, they show a preference for stories in which
the final situation conforms to the view of "a just world."
T H E STRUCTURE OF INTEREST 97

are other qualities of the final state of affairs that are appreciated by the nat-
ural audience of the traditional feature film. For example, the open ending
produces cognitive closure since the story, as canonic episodic structure, is
completed. And if justice does not triumph, as in the case of an unhappy end-
ing, then some value concern may still be satisfied, for instance poetic jus-
tice as opposed to a strict moral justice. In contrast, the happy ending may
appear contrived and childish, so that it sows doubt about the views of jus-
tice expressed or sets the viewer thinking about the ideological determina-
tion of the customary happy ending (Bordwell, 1985; Bordwell, in Bordwell,
Staiger, & Thompson, 1985). The film narrative as a whole—of which the end-
ing is only one part—may present a certain vision of the fictional world that
is in itself valuable as a pronouncement on everyday reality. The contrast be-
tween good and evil may play a secondary role, or no role at all, and the sym-
pathies of the viewer may be distributed equally over all the parties. As noted
above, the vision presented by the film may include a solution that is seldom
seen in everyday life, or even one that is not entirely plausible, as in roman-
tic films, say, where the ideal of unconditional love comes to life. In a com-
edy, there are no real good guys, and all the characters are portrayed with a
healthy portion of irony. In other cases, the vision may rely heavily on hope
or trust, as in the open-ended film, where one may at least hope that the lovers
will one day be reunited. Alternatively, the tragic aspects of the narrative can
be magnified, lending significance to mundane problems, as in melodrama
(Ang, 1985; Brook, 1973, 1976). The final situation is unfortunate for the pro-
tagonist, but seen from a different or more distant standpoint, it serves to il-
lustrate the necessity and the value of acceptance.
Thus it is not only sympathy and a somewhat oversimplified form of jus-
tice that determine the appeal of the film representation. The thematic
charge of a film, the moral, and the special attitude that the narrative displays
with respect to the fictional world all contribute to its attractiveness. This con-
tribution differs significantly from one film to another. Popular genres depend
to a greater degree on sympathy and typical justice concerns of the viewer,
while quality films usually focus on themes that are more complex.
Of course, a match of concerns with the actual state of the story and the
fictional world is not postponed until the end of the film narrative. The tra-
ditional film narrative activates those concerns almost from the beginning.
And from the beginning the viewer has a preference for a certain develop-
ment; that preference is determined by the desire for cognitive closure, one
that does justice to the viewer's sympathies and values. Some evidence for
the truth of this hypothesis is to be found in a series of experiments by Albrit-
ton and Gerrig (1991). They succeeded in showing that readers who are pro-
cessing a story generate participatory responses, such as hope and fear,
which are related to the preferred final situations of a story. The researchers
manipulated the preferred final situation so as to be a priori either positive
98 CHAPTER 4

or negative. This resulted in the subjects' formulation of mental preference

and participatory responses. When the actual outcome differed from the pref-
erences, verification of the former was impaired, as witnessed by longer
latencies.
As the film narrative proceeds, these preferences become stronger, and
the viewer forms a more and more detailed image of what the final situation
might be. In this way the final situation gradually acquires a higher value.
However, there are factors at work, simultaneously or consecutively, that are
favorable to the preferred final situation, and others that render it impossi-
ble. Ups and downs alternate in quick succession. The effect of the narrative
is situated primarily in the Complication phase (see chap. 3). The narrative
attempts to find a perfect balance, one that will maximize the viewer's inter-
est, between frustration and reward. Frustration results from the prevention
or postponement of the preferred final situation, whereas reward is associ-
ated with the enhancement and promotion of the desired final situation. This
assumption of balancing two antagonistic effects is to be found in the small
body of literature on interest in relation to the processing of discourse (Brit-
ton, Glynn, Meyer, & Penland, 1982; Hidi & Baird, 1986; Kintsch, 1980;
Moynihan & Mehrabian, 1981).
The antagonistic effect of the narrative has a counterpart in the process-
ing carried out by the film viewer. The narrative maintains the interaction be-
tween the viewer's investment and return, which is what determines the
viewer's interest. The return that the viewers get on their activity is deter-
mined by the progress made in the direction of closure, or more precisely,
the preferred final situation of the process of reception. The preferred final
situation refers both to the situation in the fictional world and to the cogni-
tive representation of the story structure. The investment involved in follow-
ing the traditional film narrative lies in the effort it takes to participate in the
construction of the events of the fictional world, such as decoding causal re-
lationships and enduring the uncertainty that exists concerning the final
structure of the action. This uncertainty means that the viewer must keep in
mind a number of possible structures, complete some of them by means of
inference, and continually test their probability. Questions must be kept on
hold until they are answered. In addition to this cognitive investment, there
is also the affective investment that arises because the fate of the protago-
nists is not in accordance with the sympathies and value concerns of the
viewer, i.e., what the viewer would wish for the protagonist. Viewers must be
patient, biding their time in hope, fear, and unfulfilled desire, all emotions
that are not in themselves pleasant.
For the most part, these cognitive and affective investments take the form
of anticipations. In describing the activity of the viewer, Bordwell (1985)
showed that this consists of a certain measure of anticipation with regard to
changes in the present state of the action structure and the fate of the pro-
THE STRUCTURE OF INTEREST 99

tagonists. Some anticipations are extremely pronounced with respect to these

changes and the moment at which they take place, as in the case of suspense:
either the hero is going to die or he is not, and the presentation of one out-
come or the other will take place within a few seconds, (de Wied, 1991; Tan
& Diteweg, in press). In such anticipations, hope and fear are affective ele-
ments. There are other anticipations whose content is specific but that are
unspecified with respect to the moment of redemption. One example would
be the question of who committed the murder at the beginning of a detective
film or murder mystery, where the affective component is challenge. Other
anticipations are much more diffuse: something is about to happen, but what
and where is not clear. Such anticipations take the form of a fairly open ques-
tion; the accompanying emotion might be simple curiosity. Carroll (1988b,
1990) used the term erotetic narration to refer to the coherence of popular
stories: "The constraints the narrative questions place on what will happen
next are the source of the story's coherence. Audience expectations, then,
are not a matter of the audience knowing what will happen next—in the sense
of my expecting to go to work tomorrow—but expectations about the likely
range of what can happen next" (Carroll, 1990, p. 133). It might be added
here that the constraints on what is likely to happen are often visible only in
retrospect: expectations are frequently no more than constraints on coming
surprises; in other words, and conversely, surprises are postdictable, as a re-
sult of the previously planted information. In retrospect, surprising outcomes
were not expected, but somehow they make sense, because the set of likely
outcomes was restricted by implicit information that was given previously.
(The term postdictable was coined by Kintsch, 1980.)
Progress leading to the preferred final situation takes the form of the ful-
fillment of expectations, some of them quite specific, others less so. If a specific
expectation is met, then the return is what we usually call an outcome. The
expectation at the beginning of the final shoot-out in Winchester '73 (1950)
is that one of the two characters will triumph (James Stewart being the fa-
vorite) or that both will get killed. In other instances, the return may consist
of an event that generates expectations that are more pronounced than the
present ones or one that renders impossible, unlikely, or irrelevant expecta-
tions that until now had appeared possible. Think, for instance, of the arrival
of a stranger in a quiet little Western town. This leads to diffuse anticipations:
something is going to happen, but what? The characterization of the man, his
past history and intentions, his relationships with friend and foe—all these
make it possible for the viewer to form more specific expectations concern-
ing future events. It may be clear that the stranger is not a baddy, but a goody,
and a very good shot; he has come to town for a special purpose, perhaps to
settle an old score. Who could his target be? These expectations represent
important steps in the direction of the preferred final situation, although no
specific outcomes have yet been realized.
100 CHAPTER 4

In the light of the previously discussed definitions, it will not be difficult

to see that at any given point in the film interest is determined by the rela-
tionship between investment and return. This relationship may be expressed,
without any further quantitative specification, as the difference between re-
turn and investment: the higher the return in relation to the investment, the
higher the interest. Meanwhile, it is assumed that the two terms can be ex-
pressed in the same order of magnitude, we will refer to this as the net return
or NR of following a certain feature film. The higher the NR, the more return
exceeds investment. The NR is, by definition, a measure of the maximum re-
turn that can be expected. Investment and net return are variables that at any
point during the viewing time (which corresponds to the presentation time
of the film) may be said to have a specific value.

The Effect of the Return Gained in the Past

and That Expected in the Future

Due to the dynamic nature of interest, expectations concerning the fur-

ther course of events and the results already obtained are both important for
the level of interest at any given moment during the film. The effort required
of the viewer in order to comply with the action tendency, namely, examin-
ing the stimulus for clues to possible consequences for the progress of the
narrative, is directed toward the future, which means that the relationship be-
tween the future costs and the future benefits is of importance for interest.
But the ratio of past investment to past return also plays a role here: my will-
ingness to continue to follow the narrative depends to a considerable extent
on whether or not I am satisfied with what I have seen up to now.
The dynamics of interest formulated in Principle 1 also means that un-
certainty is a major determinant of the value of interest. The natural viewer
has no foreknowledge of the final situation, and his or her expectations about
what is to come are constantly changing. Thus the NR is actually a subjective
and provisional estimate of the actual, final net return. At any given moment
in the film, the more benefits viewers see in the offing, the stronger the in-
terest is, even if their estimation later proves totally wrong.
Furthermore, it is of the greatest importance that the future NR is related
to the most attractive possibility, where there is a degree of uncertainty. At
any given moment, the viewer has some notion of the maximum possible NR,
if the best-case scenario is realized. For example, when the protagonist is at-
tacked by a plane spraying pesticides, the preferred outcome is that he man-
ages to escape. This possibility, no matter how uncertain or even improbable,
determines the maximum NR and the degree of interest. This does not alter
the fact that the value of this anticipated outcome is also determined by the
alternative, the possibility that the hero will be on the losing end. What is im-
portant here is that the value of the most favorable conceivable final situa-
tion determines the future NR.
T H E STRUCTURE OF INTEREST 101

The earlier considerations can be summarized in a second principle. The

principle conforms to a general feature of emotion, namely, that it is a re-
sponse to change (see chap. 3). Interest tends to grow with increases in
expected future NR, while at the same time the expected NR is seen in
comparison with that already gained. Signs of progress (i.e., a favorable de-
velopment of future expectations as compared to the past) serve to heighten
interest.
Principle 2. There is a positive relation between interest at any given
moment and the anticipated increase in the future Net Return expected
at that moment.

Empirical Research Related to Principle 2

While little or no empirical research has been done on the determinants
of interest during the act of watching a film, several experiments have been
carried out on the determinants of suspense. These results do not prove Prin-
ciple 2, but they do appear to be compatible with it. Principle 2 can be in-
terpreted to mean that higher levels of interest are reached when there is a
promise or indication of important events. By important we mean that they
are of value for the clarification of the final situation of the narrative and for
the protagonist in particular. Suspense, which is a special case of this general
rule, may be characterized as follows: an important outcome, often the result
of a highly dangerous situation that is decisive for the ultimate fate of the pro-
tagonist, is announced by the narrative (Brewer & Lichtenstein, 1982). The
outcome is highly probable but not certain (Zillmann, 1991c). Suspense is a
quality of the stimulus; it is a narrative procedure, the result of which is an
increase in interest in the viewer. As we can predict on the basis of Principle
2, the more or less explicit announcement of such events is calculated to
heighten interest. There is a some experimental evidence for the validity of
this prediction. Cantor, Ziemke, and Sparks (1984) report an increase in the
emotional response to a horror film when the subjects received a warning just
before particularly frightening scenes. Brewer and Lichtenstein (1982), on
the other hand, found no effect of warnings that were integrated into a sus-
penseful narrative. They attribute this result to the difficulty of formulating
a good foreshadowing, namely, the fact that the information must not be too
specific. Some degree of uncertainty must remain. They do report that in the
case of a mystery, the curiosity response of subjects increased when the rid-
dles at the beginning of the plot were presented in an explicit manner. In
terms of our principle, this manipulation means that the contrast between
the NR that has been gained and that which is expected has increased. In an
unpublished experiment, de Wied found that suspense increased when the
presentation of an exciting excerpt from C'era una volta il West/Once Upon
a Time in the West (1969), the slaughter of the McBain family, was accom-
panied by a suggestion—in written form—of impending evil. Using a short
102 CHAPTER 4

suspense film, Gomisky and Bryant (1982) varied the subjective chances that
the protagonist would escape. They found that the level of suspense was low-
est when there was absolute certainty concerning the fate of the protagonist
(in a positive or negative sense) and the highest when there was only the
slightest chance that the protagonist would survive, in other words, when the
contrast between the NR gained and the NR expected was the greatest. 68
Moreover, the more sympathy there was for the hero, the greater the degree
of suspense. In terms of our principle, sympathy increases the value of the
intermediate outcome for the preferred final situation. And a number of ex-
periments point to a more or less direct positive relationship between a sym-
pathetic disposition toward a protagonist and the level of suspense. This may
also be seen as support for Principle 2 (Brewer & Lichtenstein, 1981;
Gomisky & Bryant, 1982; Jose & Brewer, 1984; Tannenbaum & Gaer, 1965;
Zillmann, 1983; Zillmann, Bryant, & Sapolsky, 1989; Zillmann & Cantor,
1976; and Zillmann, Hay, & Bryant, 1975). The expected return is higher, the
more sympathetic one's feelings are for the struggling hero, while the invest-
ment already made likewise increases with one's sympathy for the protago-
nist. As a result, the discrepancy between the NR already gained and the fu-
ture NR may increase still further. Finally, it has recently been demonstrated
by de Wied, Zillmann, and Ordman (1995) that empathetic distress experi-
enced during a tragedy film is related to the magnitude of hedonic reversal,
that is the enjoyment that is felt following exposure to the entire film. High
empathizers were shown to feel both more empathetic distress and enjoyment
of the film as a whole than low empathizers. Tan (in prep.) has obtained com-
parable findings.

Principle 3: Priority for Direct Prospects of Return

Foreground and Background Return

Various net returns are important for interest; these vary according to the
period in the film presentation to which they refer. At any given moment i,
the viewer has some idea of the NR gained in the past—for short, NR (past)—
calculated from the beginning of the film—time of presentation O—up until
the moment 1
Similarly, at any given moment there is an expectation concerning the fu-
ture NR—for short NR (future)—calculated from how, that is, moment i, un-
til the end of the film—time of presentation n. The latter return consists of a
preferential value attached to the best conceivable result and the certainty of
that result. Obviously, past and future refer to the time taken up by the act

68
As we know, the future return is based on the maximum that is possible under the most fa-
vorable circumstances.
THE STRUCTURE OF INTEREST 103

EARLIER CURRENT SCENE LATER

PAST FUTURE

past future
(PP) (p) (0 (FF)

presentation time

FIG. 4.1. Subjective temporal structure of the traditional feature film

of watching the film—which is the same as the presentation time—and not

to the fictional time in the diegetic world.
We now examine an important characteristic of the film narrative, one that
necessitates a more precise definition of the past and future referred to here.
As we saw in chapter 3, the traditional feature film is made up of a sequence
of scenes. Each scene consists of a continuous action that takes place during
an uninterrupted period within the fictional narrative. The impression the au-
dience viewing the scene has of the fictional time is one of advancing now,
bounded on either side by an immediate past and an immediate future (see
Fig. 4.1). The consciousness of the viewer contains recollections of events in
the immediate past and expectations of events in the immediate future. 69 For
the most part, investments such as anticipation about what is going to hap-
pen and the consequences that this will have for the characters are expected
to be rewarded in the near future of the extended present formed by the
scene. To the extent that there are any expectations of a return on the in-
vestment, that is, progress along the road to closure, these are related to that
immediately surveyable future. The viewer wants to know what is about to
happen and is eager for closure of the present episode. Of course, it could be
said that the viewer does not know exactly how long the current scene is go-
ing to last. However, for purposes of the earlier hypothesis, it is sufficient that

69
The term consciousness is used here in order to preclude the necessity for technical de-
tails related to the architecture of the information-processing system of the viewer. However, one
can think here of a buffer or working memory with a limited capacity, which corresponds to De-
finition 4.1.
104 CHAPTER 4

he or she realizes that the actual action will be concluded within the fore-
seeable future.
Just as the future ends where the current action is concluded, the past over
which the Net Return is calculated extends up to the onset of the current se-
quence of actions. The fact that in the traditional feature film the continuity
of the intra-scene actions calls up and maintains certain expectations that are
clearly to be fulfilled within the foreseeable future contributes in no small way
to the experience of flow, which, as we have seen, greatly contributes to the
fascination of the feature film.
Figure 4.1 shows the complete subjective temporal structure of the feature
film. During each scene, expectations are created that the viewer expects to
see fulfilled in the present course of the action, as well as expectations with
regard to events that the viewer believes will take place much later, in all prob-
ability outside the present scene. In the traditional feature film, scenes in-
variably provide an answer to one or more questions, while at the same time
opening up new ones or leaving unanswered at least one other question (Bord-
well, 1985; Bordwell, in Bordwell, Staiger, & Thompson, 1985). Because cer-
tain actions between scenes may be omitted—such deletions ranging from
minimal to substantial—and because chronologically successive series of
events may appear in a completely different order, as in flashbacks, viewers
often have only the vaguest idea of when or indeed whether a certain long-
lived anticipation will be fulfilled, until the introduction of a scene tells them
that the expected event may well occur in that very scene. 70 Slumbering ex-
pectations become active, and investments are linked to the expected maxi-
mum return. It is not until that moment that emotionally charged anticipa-
tions acquire control precedence (Frijda, 1988), at the expense of those that,
no matter how strong and specific, look as if they will not lead to an appro-
priate outcome until later.71

70
In the classical case, however, the beginning of a scene always provides certainty concern-
ing a complication introduced in the previous scene, while a new problem remains unsolved at
the end of the scene. Thus toward the end of the scene the discrepancy between the NR gained
and the expected NR increases, after having initially declined. This does not affect the limited
potential for looking back into the past or forward into the future. On the contrary, it even con-
tributes to limitations, due to the fact that great leaps forward or backward are relatively rare.
71
The influence of background investment and return on local interest illustrates Frijda's Law
of the Conservation of Momentum (Frijda, 1988). An emotional anticipation retains its force,
even when unrelated events appear to force it into the background. The account of the vicissi-
tudes of character A may temporarily recede in favor of those of character B. Even when the ac-
tion has nothing to do with A, many expectations that center on A continue to exist, and will
again become topical when the appropriate scene change takes place. Atkinson's concept of "in-
ertial tendency" is also relevant here. He advocates research into the changes and constants in
motivation during the transition from one "goal-directed episode" to another (Atkinson, 1969,
p. 106). Here the goal is twofold: cognitive (the solution of the plot structure) and affective (see-
ing how the protagonist reverses the Disturbance of the Balance).
 THE STRUCTURE OF INTEREST 105

Bordwell described in detail how the narrative within the classical scene
regulates investment and return:
limited kinds of resolution occur early in the scene, as old lines of action get
closed off. And the dangling cause often leaves the scene unresolved, open, and
leading to the next. The classical scene progresses steadily toward a climax and
then switches the resolution of that line of action to another, later scene. From
the standpoint of reception, this pattern enhances the viewer's confidence in
understanding the story action . . . short-term resolutions also promise a final
resolution as well. (Bordwell, in Bordwell, Staiger, & Thompson, 1985, p. 65)

The finding that viewers look forward only to a limited degree has been
confirmed by research into inference mechanisms during the reading of nar-
rative texts. Readers only form a more detailed image of the situation de-
scribed to the extent that the necessary information is directly available. They
barely look beyond the boundaries of the present situation. So-called prim-
ing experiments by McKoon and Ratcliff provide no proof that expectations
based on causality are developed with respect to the outcomes of the narra-
tive at a later stage. In their view, previous research that ran counter to these
findings is not convincing. However, they do not exclude the possibility that
inferences that go beyond strictly local coherence are only partially encoded
(McKoon & Ratcliff, 1990). It is also possible that the term inferences has
c o m e to be understood as too specific a knowledge structure. An unarticu-
lated intuition concerning the further course of events, such as Carroll's
(1990) erotetic question previously discussed, may be impossible to demon-
strate within the experimental paradigm of McKoon and Ratcliff. 72 Or the fu-
ture-directed inference may be of a more affective nature. Elsewhere I have
maintained that the emotion of the reader controls the inference (Tan, 1994).
However, inferences that correspond to emotion-laden expectations, such as
hope and fear, are made by the reader, as shown by Albritton and Gerrig
(1991). 7 3
The two Net Returns introduced earlier can now be extended to include
the following details: (a) the Net Return gained in the past, NR(past), con-
sists of NR(p), gained in the immediate past of the current scene and NR(PP),
gained in the scenes that preceded the present one, and (b) The future ex-
pected Net Return, NR(future) consists of NR(f), the immediately expected
n e t return in the present scene, and NR(FF), expected later on in the future.
The present NRs, NR(p) and NR(f), operate in the foreground of conscious-
ness. All the investments that are not expected to produce returns in the pre-

72
It should be noted that the experiments of McKoon & Ratcliff have not remained unchal-
lenged. See chap. 5.
73
It may well be the case that quality films that lean heavily toward the literary similarly in-
vite viewers to make different kinds of inferences, for example, concerning the intentions of the
filmmaker. (See Oatley, 1995; Vipond & Hunt, 1984; Zwaan, 1993.)
106 CHAPTER 4

sent action sequence are relegated to the background because the traditional
scene is restricted to showing a single, continuous action. Knowledge that is
not relevant and affective outcomes that have already been registered are al-
located the secondary status of background return. As regards the determi-
nation of interest during a scene, NR(PP) and NR(FF) occupy a background
position.

The Relationship Between Foreground

and Background Returns
The fact that the viewer cannot see very far back into the past or forward
into the future has certain consequences for the situational meaning struc-
ture. The past and future NRs, which are calculated for limited periods, NR(p)
and NR(f), are in the foreground in the situational meaning structure; thus
they are more important in determining the interest at any given moment than
those NRs that are not current, NR(PP) and NR(FF). The urgency component
of the situational meaning (Frijda, 1986; see also chap. 3) is determined
largely by the events in the scenic present. In other words, the relationship
between foreground and background return is such that the background re-
turn, the difference between NR(FF) and NR(PP), determines the relatively
stable baseline value of the interest, while the foreground return, the differ-
ence between NR(f) and NR(p), causes the more rapid fluctuations around
this value.
Principle 2, which expresses the fact that interest is dependent on the fu-
ture expected return, may be sharpened as follows:
Principle 3. Interest at any given moment throughout the film pre-
sentation time is (a) positively related to the prospect of the greatest
possible background net return and (b) higher, the more favorably the
greatest possible foreground net return of the present action sequence
compares with the foreground net return already gained, the fore-
ground return always weighing more heavily than the background
return.
The relationship between foreground and background NRs and interest
can, in certain characteristic cases, be summarized as in Table 4.1. Of course,
the four cases represented in this figure are not all equally probable at any
given moment. The pattern of background and foreground NRs is determined
in part by the stage in the film narrative. At the beginning of a film there are
few specific expectations, so that all the NRs have a kind of indefinite value.
As soon as these values can be determined, cases 3 and 4 of Table 4.1 are
more probable than the other two, because there are not too many long-term
investments that can have been made, and the maximum return that can be
gained in the future outside the scene is not entirely clear. This renders
THE STRUCTURE OF INTEREST 107

TABLE 4.1
The Relation Between Foreground and Background NRs, on the One Hand,
and Interest at an Arbitrary Moment, on the Other Hand

Case (NR(FF) - NR(PP)}i (NR(f) - NR(p)}i Interest at moment i

1 + + + +
2 +
3 - + +
4

Note. Pluses and minuses in the different columns can stand for different values.

NR(FF) highly negative, while NR(PP) is of course still close to zero. The pos-
sibility of a good plot becomes clear only gradually, and protagonists must
first endear themselves to the audience before viewers can begin to care
about their ultimate fate.
Case 4 of Table 4.1 will probably occur only briefly at the very beginning
of a traditional feature film or at the odd weak point in the action. If it con-
tinues, then the film is guaranteed to flop. All the pragmatic principles and
motivations that are associated with an evening out will then be needed to
keep the audience in their seats. But even these do not continue to operate
indefinitely.
If all goes well, Case 1 will be dominant. The present action occupies the
viewer, while the prospects associated with actions that have yet to be shown
are also promising. Case 2 occurs in intermezzo or neutral scenes, which we
will be looking at shortly.

Distribution of Foreground and Background Return

over the Presentation Time
In the presentation of the four combinations of foreground and back-
ground NRs (Table 4.1), the interval between the NR already gained and the
expected NR has been disregarded. This is of course a simplification. The
principles of interest are based on a careful balance of frustration and reward.
Too much frustration or excessive or lengthy uncertainty is calculated to
make the viewer lose interest. The next question is that of the timing of an
increase in NR in relation to the size of that increase. How long can you ex-
pect a dog to keep racing after a sausage, and how much does the flavor of
the sausage matter? The answer may be that there is a trade-off between, on
the one hand, the temporal proximity or uncertainty of the return, and, on
the other hand, the size or weight of that same increase in return. A particu-
lar interest value can be achieved when the degree of subjective proximity of
the increased return is low and the value of that increase is high. Conversely,
 108 CHAPTER 4

that same interest value can be achieved when the degree of subjective prox-
imity is low and the value of the expected increase in return is high.
The idea of such a trade-off is based on the observation that there are dif-
ferences between various genres. In the mystery or action thriller the prox-
imity of a significant increase in the return, at any rate that of the background
NR, is relatively small throughout, whereas the size is quite substantial. 74 In
a flash, the outcome presents the viewers with the preferred final situation:
they now know who did it, the culprit has been arrested, and the protagonist
has lived to tell the tale. This is what the viewer has been looking forward to
for the entire length of the film. In the case of those comfortable, slow-moving
family films with a weak macroplot, such as Mary Poppins (1964), Those Mag-
nificent Men in Their Flying Machines (1956), or E.T. the Extra-Terrestrial
(1982), the reverse is true. On the whole, the proximity of the fulfillment of
anticipations is high and the maximum expected return comparatively low.
Zillmann (1991c) pointed to the possibility that uncertainty over the lines of
the macroplot may lead to such high levels of suspense that many viewers no
longer find the film enjoyable.
The thriller and the family film are used as examples here because these
are familiar genres. Admittedly, they are described somewhat stereotypically;
thrillers do, of course, contain exciting and suspenseful episodes, in which
important issues are addressed, long before the denouement, while family
films do develop some measure of background NR in the course of the plot.
But these two genres neatly illustrate two different types of film, those with a
strong plot and those with a weak plot. We may assume that these types also
differ in the degree to which they satisfy the sensation-seeking motive (chap.
2). As in gambling, where the greatest stakes result in the highest winnings,
the subjective risk—and the accompanying "kick"—is high for the viewer of
the thriller. The subjective risk for the family film audience is low, compara-
ble to the risk of the cautious gambler, who regularly bets small sums and is
rewarded with modest winnings.
Both the similarities and the differences between the high-stake gambler
and the film viewer are illustrative of the interest in watching a film. The prag-
matic principles of the traditional feature film make it extremely unlikely that
a lengthy and intense period of frustration will not be rewarded. In the thriller
and the horror film, the high stakes are only simulated; in the end we simply
step out of the nightmare. The question of the degree to which these princi-
ples determine the expectations of future return beyond the background is
deserving of further study. Judging by the fact that many people consider the

74
In fact we are referring here to several genres. Thrillers differ considerably among them-
selves. In order to get an idea one may think of a few arbitrary samples: The Day of the Jackal
(1973), Les diaboloques (1954), Dial M for Murder (1954), Don't Look Now (1973), Dirty
Harry (1971), and Klute (1971).
T H E STRUCTURE OF INTEREST 109

investment involved in watching such films too high, the "safety belt" func-
tion of these principles is limited.
In chapter 7 we continue the examination of the characteristics related to
Principle 3 that are associated with the distribution of interest across pre-
sentation time.

Further Remarks on the Relationship Between

Foreground and Background Return
The exact connection between foreground and background NRs is fairly
complex. Awaiting research aiming at the further formalization of Principle
3 and its parameters, it will be sketched only briefly here. Although the role
of the two factors in the affect process differs, they are not unrelated and on
occasion may even communicate. At each scene change, and on the basis of
the introduction to that scene, a number of background investments and
background returns are activated, while at the same time foreground invest-
ments and foreground returns recede into the background. In this way the
interest potential of one scene is placed temporarily on hold before being
transferred to a subsequent scene in the presentation time, one that may be
separated from the original scene by one or more other scenes. This process
will become clearer if we take a concrete example. Suppose that the plot of
our film has a thriller line and a romance line. In scene P the detective is on
the track of the suspect. In Q and R meetings take place between the detec-
tive and two different girlfriends. In scene S the detective goes to the home
of the suspect. At the beginning of scene Q investments and returns related
to the thriller line are relegated to the background, while the investments and
returns related to the romance line, which during P had been dormant, are
activated. At the beginning of scene S the reverse process takes place: in-
vestments and returns from R, which include those from Q, disappear from
the foreground to make way for those retained from P and its predecessors.
A scene transition is the ideal moment in the film narrative to adjust fore-
ground and background NRs. This follows from the scenic structure of the
narrative. But investments and returns from the background may also be re-
trieved during the scene itself. For example, in the thriller-romance the sus-
pect might suddenly break in on the rendezvous in scene R.
The overall interest level of a scene depends on the background NR in a
number of different ways. This is illustrated by a well-known phenomenon
whereby the background NR restricts the interest to a minimum when the
contribution of local factors is low. In many films there are intermezzo or neu-
tral scenes, during which very little happens. In such scenes the difference
between NR(f) and NR(p) is not very large. In musicals, song and dance num-
bers take up whole scenes, during which the action has in effect come to a
standstill. Gained and expected returns are in balance and are distributed
 110 CHAPTER 4

evenly over the scene. 75 Here background NR keeps interest above a required
minimal level. The viewer knows that in a moment the action will continue.
In general, however, we can say that the relationship between background
return and foreground return is as follows. First, the overall level of the fore-
ground NR is determined by the level of the background NR: (NR(f) -
NR(p)}i is proportional to (NR(FF) - NR(PP)}i. At a given point in a scene,
the discrepancy between the return already gained and the one related to im-
portant events that are about to take place is limited by what the narrative as
a whole may be expected to produce in the best case by the end of the film.
In the present scene not much more can be expected than what the film as
a whole appears to produce. More formally, (NR(FF) - NR(PP)}i > (NR(f) -
NR(p)}i. A knowledge of the entire situation is worth more than a single step
on the way. Of course, as the narrative advances, the viewer gradually will
have acquired a greater return on investment. But as a rule there remains a
considerable discrepancy, and it is only as the lights go on that foreground
and background NR coincide.
Second, in most scenes there will be a relation between the levels of fore-
ground and background NRs. If the development of the plot has raised a num-
ber of important questions that have remained largely unanswered, then the
likelihood of one or more of these questions being addressed in a particular
scene is greater then when they have already been answered. In other words,
with i increasing, (NR(f) - NR(p)}i on average approaches to (NR(FF) -
NR(PP)}i. As we shall see in chapter 7, in many traditional film narratives the
background NR gradually increases as the film progresses and with it the level
of interest per scene.
And third, having noted some points of association between the two Net
Returns, we must also look at a number of significant differences. These are
due in part to the selection of plot lines. Certain scenes focus on secondary
plot lines, and because these are often of little significance for the preferred
final situation, the maximum discrepancy between NR(p) and NR(f) is rela-
tively small, at any rate in comparison with that of scenes in which the events
of the major plot line are presented. Such is the case in the above-mentioned
intermezzo scenes, for example. In these scenes, thanks to the background
NR, the interest declines only to a certain minimum value. (Here again we
must keep in mind that the subjective estimate of this NR—and the fore-
ground NR, for that matter—is based on the maximum conceivable value for
the most favorable final situation.)
By varying the foreground NR from one scene to the next—so that it is
closer to or further away from the background NR -it may be possible to maxi-

75
This is only the case when we confine ourselves to the expectations and results of the plot.
There is another play of promise and fulfillment, in the realm of the music and the choreogra-
phy. However, the interest generated by this play will not be taken into consideration here.
THE STRUCTURE OF INTEREST 111

mize interest throughout the film as a whole, (e.g., as measured by the grand
mean across the total presentation time). Habituation and fatigue as a result
of continuous high levels of interest can be reduced by introducing scenes in
which the discrepancy between the NR already gained and the NR expected
within the short term is smaller. (See also the general remarks in chapter 3
on the role of change in the stimulus.)

Principle 4: Interest as a Self-enhancing Process

The action tendency associated with interest is the inclination to devote more
attention to the stimulus and a willingness to make a further investment in
order to follow that stimulus. In the case of films, it consists in actively an-
ticipating information pertinent to the action of the film as a whole and af-
fectively charged outcomes. The higher interest, the greater the active par-
ticipation of the viewer in the reception, which results in an increase in the
previously-mentioned cognitive and emotional processes. This in turn im-
proves the conditions for a high maximum future NR, since cognitive invest-
ment contributes to a total identification of the promises of the narrative for
the final situation. The narrative provides the clues for the future structure
of the story. The better the clues are utilized, the more valuable the antici-
pated final situation. The more I do my best to follow the structure of the
story, to anticipate new information, to link up various facts, and to watch for
irony and hidden meanings, the greater will be my long-term reward in terms
of structural order. The same is true for the affective investment: the more I
sympathize with the protagonist in all her trials, try to imagine what she is
feeling, and give her my unconditional support, the greater my enjoyment
will be of the ultimate triumph, as well as of each little step that brings that
triumph closer.
In other words, any investment, which is the result of the action tendency
inherent in interest, influences the various subjective NRs. It heightens in-
vestment in the direct past, thus contributing to lowering NR(p), and it in-
creases the maximum future NR, as the anticipated final representation gains
in value. And as we know from Principle 2, it is precisely these conditions
that heighten interest. Thus, interest enhances itself.
Principle 4: The action tendency inherent in interest raises the in-
vestment and this increase, in turn, has a positive effect on interest by
increasing the contrast between NR already gained and the maximum
future return that can be expected.
The self-enhancement of interest is due mainly to the fact that it is a
prospect-based emotion, which obeys the law of change. The term prospect-
based emotion was introduced by Ortony, Glore, and Collins (1988) to refer
to the fact that some emotions owe their character and intensity to a future
112 CHAPTER 4

prospect. Thus, hope is the prospect of improvement. Other emotions cov-

ered by this term are determined by a backward look: it might be more cor-
rect to refer to these as retrospect-based emotions. Relief would be a good ex-
ample here. What the two types have in common is the fact that they con-
form to Frijda's (1988) Laws of Change and Comparative Feeling (chap. 3),
that is, they are both determined by a comparison between the present and
another moment. Interest looks both forward and backward and is enhanced
by both the future and the past. Looking back, we can say that the more ef-
fort that has been put into the return that has been gained, the more plea-
sure it affords. Looking forward, the investment enhances the appeal of the
anticipated return, which, in turn, increases the willingness to invest, and so
on, and so on.

Empirical Support for Principle 4

The somewhat paradoxical idea that an investment of effort increases the
value of an aesthetic stimulus is not new. In studies of motivation, for exam-
ple, there are several different formulations of this idea. Inertial tendency
models of motivation, for example, are characterized by the same paradox.
And yet the results of empirical research do provide some support for the
plausibility of such models. (Weiner, 1972). The investment of effort in fol-
lowing a feature film can also be seen as a state of cognitive dissonance (Fes-
tinger, 1957), which, according to the results of a large number of experi-
ments, may have a motivating effect (see Brehm & Cohen, 1962, for an over-
view). The theory of the effect of insufficient reward, in particular, an offshoot
of the theory of cognitive dissonance, contains parallels with the situation in
which film viewers find themselves. Festinger (1961) put forward evidence
from experiments with animals to support the idea that subjects persist in be-
havior that either elicits insufficient reward or—and this would appear rele-
vant to the situation of the film viewer—is rewarded only after a considerable
delay. Lewis (1964a) reported findings of three animal experiments that show
that the positive value of a reward increases along with the effort needed to
obtain it. Effort was varied by having rats pull various weights. The rats that
pulled hardest were those assigned the heaviest weights. Comparable results
for human subjects were obtained by Aronson (1961) and Lewis (1964b),
whereas the findings of Child (1946) and Wright (1937) in other types of stud-
ies can also be interpreted in this same light. The last two researchers found
that under certain conditions children opt for the most difficult roads lead-
ing to the goal, even though the easier roads were clearly offered.
The core of Festinger's explanation for such results is the occurrence of
dissonance. This is an unpleasant state, but the unpleasantness is reduced
when the reward value associated with accomplishing such an activity, or the
activity itself, is heightened. "I sometimes like to summarize all this by say-
THEMATICSTRUCTURESANDINTEREST 113

ing that rats and people come to love things for which they have suffered"
(Festinger, 1961, p. 10). This mechanism is applicable to the action of watch-
ing a narrative film, which is characterized by an increasing dissonance be-
tween the return that has already been gained and the return that is expected.
I am assuming here that dissonance is reduced when (1) the expected reward
or return in the eyes of the viewer increases in value and (2) a more intrin-
sic pleasure is provided by the actual action of following the film. In contrast
to the reward in the experiments just referred to, the appeal of the anticipated
final situation also increases for other, if you will, more objective reasons. The
promise of a satisfying final situation in part grows and develops almost au-
tomatically as the film narrative proceeds. The phenomenon whereby a stim-
ulus becomes attractive as a result of the effort needed to understand it is
also familiar from psychological aesthetics. 76 According to Berlyne (1960,
1962, 1974) contradictory symbolic responses, which include the cognitive
and affective uncertainty characteristic of the incomplete film plot, lead to
physiological arousal that is accompanied by epistemic curiosity. A similar
line of thinking is reflected in the theory of Mandler (1982, 1984; Gaver &
Mandler, 1987) on valuation emotions and the role played by the appropri-
ateness of schemata. In Mandler's view, valuation is dependent on the fit of
an object to a schema. In general, fit is valued positively. However, the in-
tensity of the affect is greater when there is effort involved in obtaining fit.
The more effort involved, the stronger the affect. The accommodation of a
schema requires more effort than making an alternative schema fit, which,
in turn, is more difficult than straightforward assimilation of the object.
The results of a number of experiments by Gerrig (1989) are more rele-
vant to the act of watching a film. Subjects were asked to read stories that re-
counted well-known historical events, such as the inauguration of George
Washington or the bombing of Japan. Then obstacles were placed on the road
leading to those events that produced suspense and uncertainty about the fi-
nal situation. In the suspense conditions it took the subjects longer to verify
the sentences containing the well-known historical events, such as "George
Washington became president of the United States." It appeared that in these
conditions the subjects did not resort to the well-known historical facts in or-
der to reduce the tension. Gerrig concluded that "the present experiments
suggest exactly how readily readers become immersed in stories. Even when
attempts are made to force them to access long-term memory . . . responses
to verification statements still show the effect of deep involvement" (Gerrig,

76
Eysenck's (1942, 1968) formula for the appeal of a pen and ink drawing is as follows: "good-
ness" [of the Gestalt] = O x C, where O stands for order and C for complexity. The experimen-
tal work of Dorner & Vehrs (1975), in which configurations of colored chips are used as stimu-
lus, forms an operationalization of the same insight, which, according to Boselie (1982, p. 15)
and Berlyne (1971, p. 125), goes back to the ideas of Descartes.
114 CHAPTER 4

NR

Start t Resolution End

Presentation time

FIG. 4.2. Typical course oi interest across presentation time. Interest is a function of cxpcetcd
minus gained returns.

19R<J, p. 646). In the now familiar terms, the willingness to invest kept them
from stepping out of the fictional tale and into reality.

INTEREST AND COURSE OF TIME

One of the most important consequences of the operation of Principle 4 is

that during the act of watching a feature film interest tends to steadily in-
crease. As \VC have seen, this increase is due to the fact that the narrative
postpones closure. Investments increase faster than the Net Return already
gained, whereas the prospect of future return also increases, because the ul-
timate representation gains in value in the course of the presentation time.
1-ioreover, as the film progresses it seems more and more likely that the an-
swer to one of the primary questions is going to be presented any minute now.
Apart from the effect of the narrative, the interest response tends to inten-
sify itself, which also results in an increase with time. Figure 4.2 illustrates
the characteristic course of interest throughout the presentation time of a
traditional feature film. Although we have dealt with that course only in broad
terms, it can in fact be differentiated according to the type of film. The dy-
namics within separate scenes is discussed in chapter 7, which focuses on the
affect structure of the feature film.
THEMATICSTRUCTURESANDINTEREST 115

Interest is subject to inertia. At some point in the film, interest reaches an

intensity value that makes it improbable, if not impossible, for it to drop back
to zero. This goes for the course of the background value as well as the fluc-
tuations within a separate scene. As the discrepancy between the return that
has been obtained and the return still to be gained increases, so does the loss
that viewers would suffer if they were to decide to "drop out." The investment
already made is high, while the profit—the return that has already been
gained—is minimal. Figure 4.2 shows the loss that the viewer would suffer if
at a certain point t he or she decided to stop watching the film. If we neglect
the distinction between foreground and background NR for a moment, we
can say that typically, expected returns increase faster than gained ones. This
causes inertia; the viewer is unlikely to drop out. Immediately after the Res-
olution (i.e., after the Restoration of the Balance and attainment of the pre-
ferred final situation of the narrative), the discrepancy between expected and
gained diminishes sharply and drops to zero. Somewhat later the viewer is
only receiving returns without investing. Interest drops and is joined by sat-
isfaction and enjoyment.
Before the Resolution, the loss at some time t is equal to the integral ex-
pected NR minus the integral gained NR, summated over time from 0 ("Start")
to time t, plus the integral gained NR summated from t to time n ("End"). If
all goes well, the idea of stopping is hardly likely to occur to the viewer from
some point in time. It is difficult to say exactly where this point of no return
lies. Research into its determinants present in the film narrative is urgently
needed. It is logical to assume that what we already know about the problem
of exposition in narratives (Sternberg, 1978) will play a key role in finding an
answer to this question, and that the first signs of the Complication play a de-
cisive role in creating an interest that cannot easily be reversed.
This same inertia explains the familiar observation that even after the last
image has faded from the screen, the viewer is still inside the film, as it were,
and continues to fantasize about the fate of the characters and the structure
of the action.
Another important question that requires further study is how long the re-
turn on investment can hold off. Expected return cannot be postponed in-
definitely. As we know from the previously-mentioned experiments on insuf-
ficient reward, behavior that is not rewarded will eventually die out. The same
is true of the viewer's willingness to invest. We have seen that as the viewer
follows a particular action, possibilities of looking backward or forward are
only limited. Although closure continues to come closer and closer, the
viewer still has no more than a suspicion concerning the likelihood that it will
materialize in the present scene, or at some later point. As a rule the viewer
is not able to form any more precise expectations concerning the moment at
which closure will take place. One exception is the narrative procedure that
depends on some kind of deadline. An important event in the fictional world
116 CHAPTER 4

is set for a certain day or hour, as in Around the World in 80 Days (1956).
This is a narrative device that is very common in classical cinema (Bordwell,
1985). And if the narrative regulates the relationship between the time in the
fictional world and the presentation time in a homologous manner, then the
viewer will be able to provide expectations with a temporal index. The clas-
sic example is, of course, High Noon (1952), where the fictional time runs al-
most parallel to the presentation time: for over an hour we count the seconds
that separate the sheriff from the decisive confrontation with the villain. The
relationship between temporal expectancies and suspense has recently been
worked out by de Wied (1995). She proposes that interest during a suspense
episode is highest when (1) expectations are time indexed and (2) some de-
lay of the outcome is introduced.

INTEREST AND TENSION

In a sense, interest is comparable to what we commonly refer to as tension.

However, interest is a far more precise term than tension, which tends to call
up inappropriate connotations with hydraulic metaphors. Moreover, the con-
cept of tension is often used to designate a quality of the stimulus, such as a
film or a drama text. (Compare dramatic tension in Freytag, 1911; Pfister,
1977.) As we saw in chapter 2, the act of watching a film may owe its appeal
to the fact that it enables the viewer to experience a moment of excitement,
of tension, followed by resolution of that tension. The principles that interest
fulfills correspond to the familiar idea that the value of resolution is propor-
tionate to the tension previously built up. This was recently elaborated by Zill-
mann in his theory of suspense (Zillmann, 1991c). In his view, the enjoyment
of suspense depends on the strength of the euphoria that accompanies the
resolution of the suspense. And that euphoria in turn depends on the nega-
tive affect that precedes it, namely, the situation where the protagonist is in
danger. Zillmann's negative affect corresponds to what we referred to earlier
as investments. Recent research into the enjoyment of tragic film can be in-
terpreted in a similar fashion, de Wied, Zillmann, and Ordman (1994) found
that the degree of empathetic distress during the Complication phase of a sad
film (Steel Magnolias, 1989) was positively related to feelings of enjoyment
after the film was over. In addition, high empathizers showed more distress
and higher enjoyment than low empathizers.
Several theorists have pointed out that drama consists of a series of prob-
lem episodes, related to the goals of the characters, which produce in the
viewers episodes of tension (Beckerman, 1970). Laughing and crying are
thought to be typical reactions caused by the resolution of tension, the result
of an abrupt cessation of opposition to a difficulty. Efran and Spangler
(1979), who provide several references to relevant studies, make it plausible
T H E M A T I CSTRUCTURESANDINTEREST 117

that the whole audience cries at those points in The Miracle Worker (1962)
where a barrier is removed that had previously been created. Analogously,
the force of the laugh, according to Suls (1972) is proportional to the tension
that was created on the way to the denouement.
Interest is also related to tension in the sense that uncertainty and a he-
donically negative element are part of the situational meaning structure. We
referred to the determinants of interest and their relationship to Festinger's
dissonance and the motivational state created by cognitive incongruity. In ad-
dition to the positive element, the consideration of expected future return,
there is also the investment that has been made but for which no return has yet
been gained. Interest may be accompanied by a kind of approach-avoidance
conflict (Miller, 1944), whereby the approach is a function of the expected
NR, and the avoidance of the NR already gained: the lower the latter is, the
more likely one is to pull out. For the film viewer, however, the approach of
the goal is realized with the passing of time. Interest increases over time, and
the loss is greater when the viewer prematurely takes leave. The discrepancy
between the NR gained and the expected NR increases, and with it the con-
flict. In the case of very high interest values—extreme suspense, for example
— the return already gained can be so low and the urge to drop out so strong
that the viewer is torn between the two. This is a sensation that not everyone
experiences as pleasurable, especially if it goes on too long. In general, how-
ever, the net promise is such that the inclination to drop out is negligible.

CONCLUSION: INTEREST AS PERMANENT,

EPISODIC EMOTION

Traditional cinema initiates two episodes, the first of which leads to the sec-
ond. The first is that of the film narrative, in which a film story develops along
the structured lines of Balance, a Complication, and the Restoration of the
Balance. The second is an emotion episode on the part of the viewer, in which
interest dominates. This emotion, which initially has a low value, reaches an
irreversible intensity when the Complication appears but ultimately returns
to base level as soon as Balance has been restored. According to Frijda,
Mesquita, Sonnemans, and van Goozen (1991, p. 201) an emotion episode is
"a continuous emotion sequence resulting from the more or less continuous
impact of one given event or series of events." In every emotion episode, a
problem presents itself, which the subject copes with until it is solved. When
it takes the form of an episode, the emotion can reach such a level of inten-
sity that it is actually higher than in the alternative—acute—case (Sonne-
mans, 1991). The reason for this may be that the situational meaning of the
episodic emotion also involves the element of change, in this case, change
with respect to a multiple criterion, (i.e., both prospective and retrospective).
118 CHAPTER 4

This means that interest is governed both by the return already gained and
by the expected gain (see Tan, 1994).
Both the film narrative and the viewer emotion may be seen as recursive
episodic structures. The narration consists of scenes that are embedded in
the narrative as a whole. The interest covers an episode that spans the nar-
rative as a whole; at any given point in the film this episode can serve as the
background for a shorter episode, corresponding to the present scene. In this
way, the interest level of an interest episode of a high order is transferred to
an episode of a lower order, where it functions as the background value.
In the case of interest, the action tendency is real, unlike so many other
affects evoked by films. As we saw in chapter 3, the action tendency created
by the act of watching a film is largely virtual. When the viewer feels com-
passion, this is not the same thing as the willingness to actually do something;
rather, the viewer wants the poor protagonist on the screen to be helped. It
is for this reason that pity, anger, and fear felt by the viewer are more like
strong feelings than emotions. Interest, on the other hand, is an emotion. It
makes possible relational action, even if the action is only cognitive. It makes
the viewer willing to make a true effort: to devote one's attention to the film,
to go along with the narrative, to form an idea of a story that is often re-
counted in a highly fragmented manner. The behavioral characteristics of the
filmic state of the viewers, in particular their motionless fascination, are first
and foremost expressions of an emotion: viewers shut themselves off from the
environment in order to concentrate on the events being enacted on the
screen. The fact that it is so difficult to escape the pull of a good feature film
is due mainly to the self-enhancing nature of interest. This quality, in combi-
nation with the capacity of all emotions, once they have been evoked, to com-
pletely "take off." (cf. the law of closure, Frijda, 1988, pp. 3 5 4 - 3 5 5 ) makes
it difficult to escape the interest, or rather the fascination, evoked by the film.
Not until the ultimate closure are viewers able to step outside the continuous
current of guided fantasy. Until then, they become more and more wrapped
up in their own exercise of imagination as the film proceeds.
The characterization of interest as a self-reinforcing process in no way di-
minishes the role of the film. Somewhat paradoxically, in fact, you could say
that without film, there is no self-reinforcement: the film reinforces the op-
portunity to invest and regulates both the beckoning prospect of reward and—
in due time—the actual attainment of that reward. The film also keeps the
discrepancy between the return already gained and expected return within
bounds, so that the challenge represented by the film does not prove too
much for the viewer.
As the ever-present and self-enhancing emotion, interest dominates the af-
fect structure of the feature film. It is not only a tonic but also a permanent
emotion. The discussion of that episodic structure as a whole is addressed in
chapter 7. There we examine the typical course that interest takes, the best
THEMATICSTRUCTURESANDINTEREST 119

known of which is determined by three narrative processes: suspense, sur-

prise, and mystery But even now it is not difficult to see that watching fea-
ture films involves the experience of quite differing emotions at one and the
same time. No matter what the combination, however, interest is present.
Viewers may be frightened, indignant, or moved, they may be laughing or
feeling extreme disgust, but there is always the need to know what happens
next and what the ultimate fate of the protagonists will be. If it is correct to
refer to the film as an emotion machine, then it is above all an interest engine.
 Thematic Structures
5 and Interest

TWO DETERMINANTS OF INTEREST

In chapter 4 it was maintained that interest is the major emotion experienced

during the watching of a traditional feature film. What then are the determi-
nants of interest in the stimulus, that is, the fictional events? Existing insights
into the structure of traditional narratives would appear to offer more than
enough points of departure for a description of those elements that deter-
mine costs and benefits—what we refer to here as investment and return—
and are responsible for interest. There are two such elements, namely, the-
matic structures and character structures. The former regulates expectations
with respect to the course of the action, while the latter governs expectations
regarding the characters and the viewer's involvement in their fortunes. This
latter element, the cognitive character structures, will be discussed in chap-
ter 6. The literature of drama theory recounts a number of efforts to define
the dramatic situations that form the core of existing dramas or might fulfill
such a role in future dramatic works. Polti (1921/1977) undertook to char-
acterize 36 such situations, including Fatal Imprudence, Self-sacrifice for
Kindred, Erroneous Judgment, and Mistaken Jealousy. Souriau (1950) em-
ployed a more rigid system, distinguishing six separate dramaturgical func-
tions, including the Aimed Dramatic Force (often a passion, such as desire
or fear), the Representative of Value, and the person who obtains the Value,
and the Adversary. In Souriau's view, various functions can be united in a sin-
gle character; he also takes into account the different viewpoints according

121
122 CHAPTER 4

to which a situation can be defined. Thus, by means of permutation and com-

bination, he arrives at over 200,000 different situations. These, in turn, form
thematic clusters, such as Temptation and Betrayal. A system closely associ-
ated with that of Souriau is the structuralist "actant" model devised by
Greimas (1966) for the interpretation of mythical stories. Whereas Souriau
divides drama into situations, Greimas sees a story in terms of sequences.
These are characterized by a particular constellation of actants, such as the
Sujet-Heros, the Objet Valeur, the Destinataire, and the Opposant.

THE INTRINSIC INTEREST VALUE OF THEMES

It is clear from the work of Souriau in particular that while in our culture it
is possible to describe such familiar fiction themes as Temptation, Jealousy,
and Revenge by means of a limited number of semantic primitives, these form
a much smaller number of clusters than is theoretically possible. This sug-
gests that a smaller number of universal or culturally and historically rooted
subjects are at work here. Such subjects have an intrinsic appeal, one that
cannot be traced to other factors; all one can say is that they come to be ap-
preciated through social learning mediated by a dominant supply in oral sto-
ries, news narratives, fiction books, television drama, and feature films. The
best-known view on the intrinsic interest value of themes states that sex and
violence, or such variations on that theme as the three S's—Sex, Sensation,
and Sadism—are a guarantee of success, as witness the flourishing existence
of the boulevard press. Empirical findings also point in this direction.
Graesser and his coworkers found that when their subjects (college students)
were free to choose, they almost always read news items concerned with a
subject that was familiar to them and which broadly coincided with these
"SSS" themes (Graesser, Higginbotham, Robertson, & Smith, 1978).77 In one
of the conditions of the experiment, the subjects were given the impression
that they were waiting for the start of an experiment; in the room a number
of newspapers were lying within easy reach. Some 98% of the subjects spon-
taneously picked up one of the tabloids, namely, the National Enquirer; while
periodicals such as Time magazine, the Wall Street Journal, and even Play-
boy were much less in demand.
Within the cognitive theory of story comprehension the conception of in-
trinsically interesting structures is put forward by Schank (1979). A theme is
a cognitive structure containing a hierarchy of aims, plans and events, which
may be compared to Souriau's functions: Death, Danger, Power, Large Sums
of Money.78 The different themes have various learned and intrinsic interest

77
Additional information on this experiment is given in Graesser (1981, pp. 63-67).
78
For more formal definitions of themes, see Schank and Abelson (1977).
THEMATIC STRUCTURES AND INTEREST 123

values, which determine the number of meaningful inferences the reading

subject will make at any particular point in the processing of a narrative. Ac-
cording to Schank, themes optimize the number of inferences made. The
presence of themes leads to the generation of inferences, while the intrinsic
value of interest determines the number of inferences: the more interesting
the narrative, the more inferences are made.
However, it may be objected that inferences must be optimized in terms
of content as well as number if they are to play a role in interest. The ten-
dency to invest is not determined by the sheer number of expected anticipa-
tions. In fact, high numbers of actual or promised inferences may actually
create confusion and thus turn people off. In theory, each narrative could trig-
ger an almost unlimited range of qualitative inferences; however, in practice
this never happens, as not all inferences are equally relevant. 79 If it is to func-
tion as an investment, an expectation must have some connection with the
anticipation of the preferred outcome. It must offer some prospect of the ful-
fillment of cognitive and affective concerns.

THEMES AND THE

PARADOXICAL COMPLICATION

The Prospect of Complications, Followed by a Solution

The proposal put forward by Schank (1979) on the cognitive effect of themes
requires some adjustment with respect to the nature of the inferences that
they generate. Thematic structures are only capable of explaining the mech-
anism of interest when they promise not only complications but also satis-
factory outcomes. Let us first examine the complications. In an exciting story,
quick and easy revenge is a much weaker given than revenge that is achieved
only after countless difficulties and complications have been overcome. The
same may be said of a romance that proceeds without any noteworthy prob-
lems, a treasure that is found within minutes, or an immediately successful ad-
justment to stereotypical norms of civilization or of physical or mental health.
This notion of complication and resistance appears in the dramatic and nar-
rative theories of Souriau and Greimas discussed earlier, notably in the pres-
ence of an adversary category in the constellation of elements that together
provide the action. In fact, you could say that in traditional drama theory the
essence of drama can be equated with conflict (Barry, 1970, in Van den Berg,
79
There is a joke in Abel (1986) that illustrates the theoretical possibility of irrelevant infer-
ences. Certain inferences are not made by the natural viewer of the traditional feature film. If
put forward, they appear somewhat foolish: Abel's father is against his son watching westerns on
television "because they never tell you who the parents of a cowboy are, where he comes from,
and what kind of education he's had."
124 CHAPTER 4

1979, p. 83; Pfister, 1977). This may be a conflict between forces (Souriau,
1950; van der Kun, 1938) or strivings (Beckerman, 1970; Brunetiere, 1894/
1914).
The fact that a minimal number of complications are required to make a
story interesting is also illustrated by the presence of a Problem, a Distur-
bance of the Balance, or some similar element in the canonic narrative struc-
ture proposed by various story grammars. This characteristic may be implicit,
as when an episode consists of an Attempt to attain a goal which may be suc-
cessful or unsuccessful, or is related to an Outcome or Resolution (Mandler,
1984; Thorndyke, 1977). Rumelhart (1980) explicitly limited the definition
of a story to a series of goals and actions by one or more protagonists directed
toward the solution of a Problem.
A second, related, amendment of Schank's original views on themes holds
that the thematic structure must offer at least a minimal, real possibility that
solutions will be found for the complications. The viewer is more inclined to
invest if he or she feels at an early stage that ultimately all the entanglements
will be satisfactorily resolved.
For the audience of traditional feature films, at any rate, a satisfactory out-
come of the fictional action could be one that is in accordance with stereo-
typical views of justice or a kind of general commonsense truth. It is not, how-
ever, necessary for the outcome to be satisfactory from the perspective of a
particular system of values. A solution simply means that complications have
been smoothed over, contrasts reconciled, and unpleasantness rendered
bearable. In short, the solution disposes of the entanglements once and for
all, thanks to the power of the paradoxical cliche. Often the outcome is be-
lievable in quite a direct manner, due to some ironic twist in the action or the
striving of the characters. We learn from the misunderstanding, or the solu-
tion, or both. The mottoes and adages resulting from the complication and
its solution, while often on the level of the calendar quotation, do appear to
have their merits. These old bromides include such pearls of wisdom as "All's
well that ends well," "Honesty's the best policy," "He who pays the piper calls
the tune," "The chickens have come home to roost," "Every cloud has a sil-
ver lining," and so on.
The point that can be distilled from the complications and the solution and
crystallized in an adage often reflects a kind of situational irony, here de-
scribed broadly as a contrast between the intention of the protagonists and
the outcome of their actions: the means employed to solve the problem are
worse than the problem itself; two misunderstandings combine to provide a
solution; the character unintentionally damages his or her own case; a coun-
termove results in defeat. Durgnat (1967) gave us a telling description of such
points. He saw the multiple paradox and the ensuing surprise as the most im-
portant thematic characteristic of the popular film:
THEMATIC STRUCTURES A N D INTEREST 125

Many a film is built on the crescendo of an original surprise, on a series of pat-

terns like the inversion of a musical phrase in a symphony. The bitter bit, sud-
den reversals of fortune, sins coming home to rest, getting more than you bar-
gained for, being beaten at your own game, losing what you love too much—all
these and many other dramatic ironies or "reversals of expectation" form the
substructures of most tales. (Durgnat, 1967, pp. 182-183)

Durgnat is referring here to the popular film rather t h a n the traditional fea-
ture film. But the attribute of the surprising and enlightening paradox is also
valid to a greater or lesser extent for the traditional feature film. There is a
kind of continuum with, at one end, the schematic tale, in which the paradox
is far more prominent than the classical realistic psychology of the charac-
ters; comedy is one example, the lighter action film another. At the other end
of the continuum we find the psychological drama, which is actually closer
to the psychological novel. But even here there is room for irony in the given,
alongside the in-depth exploration of character. We n e e d only think of the
tragic flaw, the character fault that in some paradoxical manner stands in the
way of a favorable outcome for a sympathetic, well-intentioned character. In
Interiors (1978) by Woody Allen, Eve cannot help molding her husband and
children into perfect human beings, just as she transforms rooms into perfect
interiors. The result is that they only talk about her, not to her, and in the
end she loses them. She asks herself frantically why everyone is against her
and what has happened to her family; ultimately she chooses death. A well-
known melodramatic elaboration of tragic flaws can be seen in Douglas Sirk's
Written on the Wind (1956), where a brother and a sister cannot but ruin
themselves, due to their neuroses.

Themes and Concerns

Thematic structures are linked to the concerns of the viewer. When a theme
is broached, concerns are activated and the viewer becomes aware of the
promise of realization, which forms the emotional stimulus of interest. The
closure of a thematic structure serves a cognitive assimilation concern, where-
by complications are resolved in a satisfactory manner. Moreover, the cogni-
tive structure obtained is capable of reconciling complexity and unity. It
appeals to a need for beauty because it can encompass all manner of regu-
larities. Value and sympathy concerns are similarly served, since the solution
is in accordance with accepted ideas or with the sympathy of the viewer.
Even when only partially activated by the fictional events, themes hold out
to viewers the promise of emotions, indeed, precisely the emotions they are
waiting for: those generated by the complications and their resolution. The
condition that themes promise to realize affective concerns is met, since a
t h e m e contains the elements of intentional actors and their goals. The
126 CHAPTER 4

themes that are activated by the traditional feature film are also character-
ized by the fact that the goals they encompass are universal. A Disturbance
of the Balance affects the concerns of the protagonist that he or she shares
with everyone else and leads to the definition of recognizable goals, such as
the restoration of intimate relations, ownership, health, security, acceptance
and recognition, freedom, autonomy, and justice. The fact that they refer to
shared concerns means that the subjects designated by Schank (1979) as in-
trinsically interesting, such as Danger, Death, Large Sums of Money, and so
on are eminently suitable for eliciting interest from the viewer. It should be
noted here that recognition is itself emotionally charged, precisely because
concerns are involved. The film viewer, like any observer in the real world, is
interested when someone is rescued from mortal danger for the simple rea-
son that he or she, too, has a strong will to stay alive. Not only are we all fa-
miliar with the theme, it is of vital importance to all of us!
Themes contain complications and outcomes. The recognition of con-
cerns implies the identification of both preferred and undesirable outcomes.
Negative outcomes would appear to be more frequently associated with
themes than positive ones. Many of the most common themes, such as Be-
trayal, Self-sacrifice, and Deceit, activate the expectation that a negative out-
come is in store for the protagonist. This is inherent in the story of the tra-
ditional feature film, with its inevitable complications. The actual outcome is
less likely to be negative, especially as regards the final outcome of the story.
And even when it is, this is still in keeping with concerns, because we have
seen earlier that one characteristic of the final outcome is that it may offer
the promise of a moral. We can learn from complications and their resolu-
tion, but it is also possible that a philosophy of life or attitude is confirmed.
Truths and adages implied by the outcome may have significance to the
viewer's life. Andrews (1989) distinguished a number of different views of re-
ality that may be characteristic of different personalities. Lazarus (1991) sug-
gested that these views fulfill an important adaptive function for the species
and the individual. This would justify seeing them as concerns in the sense
of Frijda's theory of emotion. The visions of life delineated by Andrews would
appear to coincide with the general thrust of certain fictional genres and their
dominant themes, namely, the romantic, the ironic, the tragic, and the comic
vision of reality. Film themes reinforce these views, perhaps through their
complications and outcomes. To take an example, it is possible that individ-
uals with a tragic view of life are attracted by the given of an impossible love
or some other melodramatic theme. (See Ang, 1985, for an intriguing an-
alysis of melodramatic television drama in relation to this type of viewer's
concern.)
If it is indeed so that the self-involvement of the viewer stems at least in
part from the recognition of the concerns of the characters in the fictional
world, then it is possible to formulate a general proposition on the relation-
THEMATIC STRUCTURES AND INTEREST 127

ship between themes and concerns, namely, that each theme contains the
representation of one or more concerns of one or more characters. The lack
of concerns could be behind Souriau's finding that not just any arbitrary
combination of semantic primitives of the dramatic situation leads to the for-
mation of a theme: concerns limit themes. The extent to which there is a cor-
relation between psychologically realistic concerns and their fictional repre-
sentation is an interesting empirical question that is worthy of further study.
If the similarity is better than average, then it is not inconceivable that there
is a more or less automatic mechanism of recognition operating here. A
theme then enhances the potential inclination of the viewer to invest inter-
est because there is an automatically recognized concern that is shared by
the viewer in the real world. This may be one way in which the fictional world
evokes genuine feeling: the emotional chords of the viewer's concerns are
played upon by the film narrative. We return to this possibility in chapter 6.

THE OPERATION OF THEMES

A S COGNITIVE STRUCTURES

General Characteristics
Let us summarize, from a cognitive standpoint, the characteristics that until
now have served to identify themes. A theme is a generic cognitive structure
that is activated by the action in the fictional world.80 This structure contains
one or more actors with concerns and intentions from which a hierarchy of
goals and plans can be derived. Furthermore, each theme contains a com-
plication or failure scenario consisting of a more or less elaborated sequence
of events and a number of mutually exclusive possible outcomes. Some of
these must be satisfactory, in the sense defined in the previous section or,
better yet, both satisfactory and valuable. The qualification valuable refers to
the possibility that they appeal to visions of life and so answer more distal,
secondary motives such as cognitive and affective preservation (see chap. 2).
From a more technical point of view, the activation of a theme is followed
by expectations concerning the action elements that have not yet been in-
stantiated. The solution, the compromise, or the characteristic failure that is
the result of a sequence of action covered by the theme, together with the ac-
companying past history can be summarized as a valued stereotypical truth.
Arriving at a cognitive representation that meets this condition of succinct
expression is a criterion for comprehension. Cognitive processing ceases and
any remaining expectations, which are now irrelevant, are neutralized. But

80
By this I mean that it is shared by most people in present-day Western society, and thus by
the group that consists of the natural viewers of the traditional feature film.
128 CHAPTER 4

before this happens, an active theme is always accompanied by uncertainty,

because it fosters the expectation of mutually incompatible outcomes. A
theme places a maximum and a minimum limit on the uncertainty concern-
ing the progress of the action. The uncertainty about what is going to hap-
pen is limited because it is not difficult to derive relevant inferences and ex-
pectations from generic knowledge. This restriction of the uncertainty about
future action, not least that which is expected in the near future, is an im-
portant characteristic of the traditional feature film, which distinguishes it
from, say, the art film (Bordwell, 1985). But certainty is similarly restricted.
Interest is determined precisely by the dosage of partial uncertainty, which,
as we have seen, must be neither too large nor too small (see the "adequate
challenge," chap. 4). What is important is that the thematic structure defines
the space available for potential outcomes, while within that space a larger
number of possible outcomes must compete with each other for the invest-
ment of the viewer.
The recognition of a theme generally precedes the outcome. Activation of
a theme calls up the need for a suitable closure of the episode governed by
the theme, within the limited scope available for outcomes. Restriction of this
scope makes it both worthwhile and pleasurable to guess at possible out-
comes. This is almost always accompanied by surprise and other phasic emo-
tions that depend on the thwarting of an expectation: amazement, dismay,
disappointment, and so on.81

Cognitive Models of Thematic Structure

TOPs, TAUs, Points and Plot Units

Cognitive science has developed models for the comprehension of narra-
tives that take as their starting point the work of Schank and that meet the
general characteristics described earlier (Dyer, 1983; Lehnert, 1982; Wilen-
sky, 1983a). In these models the comprehension of a narrative develops
through the construction of a cognitive representation from the bottom up.
Thematic structures of various dimensions regulate the representation of the
action on different levels of detail. In a very general sense, they specify the
goals of one or more protagonists, obvious plans for attaining those goals, in-
teractions between goals, plans and actions of the various protagonists, and
the possible outcomes of the actions. They also contain characteristic plan-
ning errors.
On the most global level of the narrative, that of the story as a whole,
themes may be defined as Thematic Organization Packages, or TOPs (Schank,

81
That is, if surprise is an emotion. See Ortony, Glore, and Collins (1988). According to Frijda
(1986) and Izard (1977), this is indeed the case.
THEMATIC STRUCTURES AND INTEREST 129

1982). These serve as a source of anticipations of complications and out-

comes on the highest level of abstraction. Examples of TOPs are Competition
Goal, which is found in the representation of all the stories in which the good
guys battle it out with the bad guys, and the general Possession Goal, which
speaks for itself. On a lower level of abstraction, thematic structures operate
as postulated and implemented by Dyer (1983) in BORIS, a simulation pro-
gram that is capable of understanding stories. A Thematic Abstraction Unit
(TAU) does not govern a complete story, but only an episode, an action se-
quence that in a film would cover one or two scenes. TAUs contain infor-
mation about planning errors and how to correct them, which can be sum-
marized by means of mottoes. Thus they meet the criterion that the action
contains complications and that the final outcome of the sequence of actions
is a satisfactory one. A simulation program—Plan Applyer Mechanism (PAM)
—has been developed by Wilensky (1982, 1983a), which is capable of under-
standing stories by identifying "Points." Using these Points, Wilensky strives
to identify exactly what it is that makes stories interesting. Points represent
"human dramatic situations"; in general this means that a character is faced
with problems because of the interaction of various goals. In a somewhat sim-
plified form, one could say that in Wilensky's model Problem points and So-
lution points can be identified. The thematic structures distinguished by
Wilensky call up expectations, "demons," the activation, change, and deacti-
vation of which are recognizable in the output of the computer program. For
example, in the model a Problem produces the expectation of a Solution. A
Problem demon is activated that searches through all the input for informa-
tion that conforms to the pattern of a solution for the problem in question.
The model developed by Lehnert (1981, 1982; Lehnert & Vine, 1987) for
the cognitive processing of stories represents most clearly the requirement
that thematic structures must be capable of being constructed in a bottom-
up manner. In this model identification of the simplest thematic units, known
as simple plot units, is followed by the composition of more complex themes,
known as complex plot units, consisting of combinations of the simple units.
In statu nascendi they generate expectations concerning complications and
the ensuing solutions and failures. Examples of a simple plot unit are a Suc-
cess and a Loss. More complex units include Retaliation, Hidden Blessing,
and Fortuitous Resolution. Ironies and other striking forms of regularity or
contrast between elements of the action are clearly expressed in Lehnert's
model. This model lends itself to the processing of thematic structures in or-
der to illustrate interest.

Example 1: "The Gift of the Magi"

An instructive example of a complex plot unit is provided by the thematic
analysis in Lehnert (1982) of "The Gift of the Magi," a short story by O.
130 CHAPTER 4

TABLE 5.1
Input sentences in "The Gift of the Magi"

1. A young couple wants to buy each other Christmas presents.

2. They are both very poor.
3. Delia has long beautiful hair.
4. Jim has a prized pocket watch.
5. To get money for their presents, Delia sells her hair and Jim sells his pocket watch.
6. She buys him a gold chain for his watch.
7. He buys her an expensive ornament for her hair.
8. When they find out what they have done, they are consoled by the love behind each
other's sacrifices.

Note: From Lehnert (1982). Reprinted by permission.

Henry, which has been filmed as one of the episodes in O. Henry's Full House
(1952). The story is about a young couple, Delia and Jim. They are very poor,
but each wants to buy the other a Christmas present. To get the money for a
present, she sells her beautiful long hair and he sells his pocket watch. He
buys her an expensive ornament to wear in her hair, and she buys him a gold
chain for his watch. When each discovers what the other has done, they are
consoled by the thought behind the presents (see Table 5.1).
Figure 5.1 shows the thematic structure according to Lehnert (1982). In
terms of the characters' goals and values, the states represented may be
pleasant or positive (+), unpleasant or negative ( - ) , or neutral (M). Those des-
ignated as + or - are events or, to be more precise, event outcomes, while M
reflects a neutral mental state or mental act, such as (having in mind) a goal,
a plan, or a preferred state that has yet to be realized. One affect state may
relate causally to another in various ways. A mental state is motivated (m-
link) by another mental state or an event, or it is the equivalent of (e-link)
another mental state, or it is terminated (t-link) by another mental state or
an event. An event is an actualization (a-link) of a mental state, or an equiv-
alent of another event, or it is terminated by another event. Any event or men-
tal state may have more than one relation with one or more other events or
mental states.
Lehnert proposed a bottom-up process of story comprehension. Affect
state patterns can combine to form primitive plot units, such as a Problem
( — m - > M), a Success (M -a-> +; see wants money-gets money configura-
tion in Fig. 5.1) or a Hidden Blessing ( e—> +; see regrets [present]-ap-
preciation in Fig. 5.1). Primitive plot units, in turn, form complex ones. Sac-
rifice, for example, combines Success with Loss. Complex plot units may in-
volve multiple characters. The configuration in Fig. 5.1 shows the primitive
plot units Nested Subgoals, Nested Success, two Losses (one of valuable ob-
ject, the other of pleasure in giving), and a Hidden Blessing. There is also
the complex plot unit Regrettable Mistake, a shared negative event caused by
THEMATIC STRUCTURES AND INTEREST 131

o n e of two characters, and a Sacrifice. The representation accounts for the

reader's intuition of symmetry. All plot units mentioned are doubled a n d / o r
shared by Jim and Delia.
Lehnert's model accounts for summarization behavior. If it is assumed that
t h e complete complex affect unit structure is available in m e m o r y after read-
ing the story properly, that is, even without having to produce a summary, we
c a n conceive of this structure as the story's final representation. In order to
grasp how promise develops, it may be instructive to conjecture about how
t h e final representation evolves. It should be stressed that the analysis is com-
pletely hypothetical. It rests on the assumption that subjects work toward a
final representation by processing a sequence of input sentences one by one,
using generic knowledge structures, such as plot units, to construct a final
representation in a bottom-up fashion. A simplified protocol of input-by-input
processing might look like the one shown in Table 5.2. I have left out many
propositions that later prove to be false or superfluous. The numerals in the
left-hand margin correspond to input sentences in Table 5.1. In describing
t h e on-line processing of narrative, I employ Graesser's terminology (Graes-
ser, 1981; Graesser & Clark, 1985). He distinguishes the following opera-
tions: (1) matching of new with old information, (2) bridging of conceptual
gaps, (3) pruning of links between propositions that appear to be obsolete,
a n d (4) projection of expectations to be matched by future inputs.

Mr
w a n t s to give gift / M M w a n t s t o give gift
A
m\ \ Jzn
wants gift M M w a n t s gift
\ )m

"I \m
wants money M M wants money
\

J
w a n t s to sell hair M w a n t s t o sell watch

( J J
sells hair +
( \ sells w a t c h

gets money + gets m o n e y

gets chain t + ^ / + t gets o r n a m e n t
gives chain
V + ^ + / gives o r n a m e n t
gets ornament
V __ -
r gets chain
regrets chain

appreciation
r "

v. + —
< ^ -

+
)e
regrets o r n a m e n t

appreciation
FIG. 5.1. Affect unit representation of "The gift of the Magi,' according to Lehnert (1982):
M = a plan; m = motivates; a = is actualizes; e = equivalent; t = terminates; + = positive affect;
- = negative affect. (From Lehnert and Ringle (1982). Adapted by permission.
132 CHAPTER 4

TABLE 5.2
Overview of simplified processing of "The Gift of the Magi"

1. Goals (M) are established for two protagonists (Fig. 5.1). Young couple may activate world-
view: Love will overcome. Bridging: Man and woman are in love. Projection: goals, gift to be
specified; subgoals to be specified.
2. Sets a circumstance, logically preceding input sentence 1, not accounted for in Fig. 5.1.a
Enables sympathy to be established. Onset of Complication phase; Problem is clear. Projec-
tion of expectations: mental states involving subgoals to get money, further complications
and resolution.
3. Introduces first protagonist. Increases liking. Bridging, using world and story knowledge:
hair is valued by Delia. Projection from story knowledge: something may happen to her
hair. Sympathy: poor girl has only one valuable possession.
4. Introduces second protagonist. Bridging, using world and story knowledge: watch is valued
by Jim. Projection: something may happen to the watch. Sympathy: poor boy is proud of his
only possession. Matching with sentence 3: first awareness of some symmetry in plot struc-
ture. Projection: symmetry to be expanded.
5. First Resolution: they both get money. It is a Sacrifice to both, that is a Success in terms of
a subgoal (wants money) and a Loss. Matching: further awareness of symmetry of protago-
nists' actions. Projection: they are going to buy something. Sympathy: they both sacrificed
something valuable.
6. First half of second Resolution, one gift has been obtained. The nature of the Resolution is
a surprise. It is a Regrettable Mistake (Lehnert's name for the complex plot unit). This con-
stitutes the point of the story. Projection: Jim will be disappointed when he receives the
chain. Sympathy: she is unaware of this. Projection: he may choose a present that has the
same effect on her.
7. Second half of second Resolution. Point is complete. Matching and projection: ornament is
Regrettable Mistake, symmetrical to chain. Projection: a Problem for both is expected, and
a Resolution. Sympathy: they are unaware of the problem.
8. Final Resolution: The presents are exchanged, which is a Success in terms of the first goal
(input sentence 1); it terminates earlier Successes, but at the same time something more
valuable is revealed, resulting in two Hidden Blessings.

Note: See also Table 5.1 and Fig. 5.1. Processing step number corresponds to input sen-
tence number in Table 5.1.
a
This may be a general difficulty in working with Lehnert's plot units (see Tan, 1986a).

On the basis of this example, it can be shown how a thematic structure ac-
tivates concerns and calls up promises:
1. Cognitive closure is called up by step 1 (see Table 5.2). Because a spe-
cific expectation is possible, this concern immediately leads to a cognitive in-
vestment. With each new input sentence a new projection becomes possible,
which to a certain degree specifies the further expansion of the plot-unit
structure. Each projection proceeds in the direction of the completion of a
primitive plot unit, while gradually expectations of more complex units and
symmetry links begin to play a role. Projections count as investments,
matches as return. It is important for the course of the cognitive component
of interest that the value of both increases with the input sentences, but that
THEMATIC STRUCTURES A N D INTEREST 133

the increase in return lags behind that of the investment. The main reason
for this is that the possibility of a more complex structure suddenly emerges
in sentences 5 and 6, and that it is not until sentence 8 that the most com-
plex plot unit is revealed, that is, a double Mixed Blessing.
2. Sympathy may be called up directly, by the words "young couple" in
sentence 1. The reader recognizes the motives of Jim and Delia, a universal
love concern. Reinforcement follows in sentence 2, where the Problem es-
tablishes the superiority of the reader. Due to the lack of action possibilities,
sympathy takes the form of hope and fear. Next, input sentences 3 and 4 lead
to emphasis on the protagonists' lack of possibilities, which enhances the
sympathy. Hope and fear are confirmed by the events in sentences 5 and 6
and are renewed when a new problem crops up. Sympathy is reinforced by
what Durgnat calls the "reversal of fortune." Sympathy functions as invest-
ment, since the reader must wait to hear what the fate of the protagonists
will be.
3. View of the world is a second affective concern that might be evoked at
an early stage, namely, in processing step 1. The view inherent in the final
outcome becomes more articulated, from "Love will conquer" to something
like "True love is reflected in the intention." The view of the world contained
within the story gradually becomes more explicit, and this is accompanied by
an increase in the value of the preferred final representation. Moreover, the
moral of the theme lends the final outcome its definitive character, because
it immediately raises the cohesion of the structure to its maximum value.
Thus the Hidden Blessings and the moral mutually reinforce one another,
whereas each of the Hidden Blessings is directly or indirectly linked to each
of the simple affect states that together make up the complete configuration.
In the case of the affective component of interest, return also lags behind
investment across the board. Gradually more and more sympathy is invested.
Each plan (M) can evoke hope and fear in connection with the uncertainty
of the outcome. The outcomes are slow in coming. Each plus in the repre-
sentation stands for pleasure on the part of the subject, that is, return. Figure
5.1 shows that return based on sympathy cannot be brought about before in-
put sentences 5 and 6, although these sentences also call up a new and affec-
tively more heavily charged investment. This investment is not rewarded un-
til the very last sentence. The most valuable vision of the world also presents
itself quite late and does not become completely clear until the last sentence.

Thematic Structures as Constraints on Surprise

A theme can take form gradually, with each advance contributing to interest.
Once a theme has been wholly or partially activated by the input, it has the
status of a kind of summary of one or more hypotheses; in this way it con-
134 CHAPTER 4

tributes to the evocation and steering of responsible investments, efforts of

the viewer that promise to be productive. Directly or through inference, the
viewer distills from the input information concrete instances that meet the
requirements of the theme. Thus the theme "Christmas presents" generates
the expectation that something of value will be purchased. "Chain" and "or-
nament" comply with the pattern and produce a match. However, expecta-
tions need not be particularly specific in order to support interest. The viewer
does not always search the input to find a single concrete instance that is
highly restricted by the structure is already known. Various possible thematic
structures may have presented themselves already on the basis of earlier in-
put information. In that case, the anticipation consists in searching for new
input patterns that differentiate between the alternatives.
Anticipations may even have so little articulation that viewers can only
look forward to unspecified Points, without any idea of what they are looking
for. In the case of minimal projection, thematic structures may yet function
in such a way as to make surprises "postdictable" (see p. 99, this volume). A
postdictable surprise is caused by an input that, while its nature is totally un-
expected, can be integrated into the existing representation. To make this
possible the existing representation of all the input up to now may have to be
revised. Without thematic structure, such surprises would be totally illogical.
For instance, in "Gift of the Magi," it is conceivable that the author might de-
cide to follow the vicissitudes of only one of the protagonists, say, Delia, so
that it is only in the last events of the story that both the final outcome and
the entire previous history are revealed. The surprise is then postdictable if
the sacrifice she has made and her secrecy toward Jim have been perfectly
clear throughout.
All things considered, it may generally be preferable to see expectations
and hypotheses more as schematic structures defining the limits of possi-
bilities and limiting the significance of coming surprises, rather than as far-
reaching and detailed propositions that need only be compared with new
input. Some readers will remember Noel Carroll's description "erotetic nar-
ration"—quoted in chapter 4 to stress the nature of open questions—that may
be attributed to the expectations evoked by the film narrative. The viewer's
ability to see into the future is limited. Recent research questions whether
readers are able to form detailed impressions of the fictional world that go
further than the here-and-now sketched by the story itself. Research by Mc-
Koon and Ratcliff (1990, 1992) has already been discussed (Principle 3 of
interest, chap. 4). The minimalist position advocated by these researchers,
whereby few if any inferences are made beyond what is explicitly given in the
text and what one's knowledge of the world provides in the way of strictly
local coherence is quite extreme. Other experiments show that people do
make predictive inferences during reading (Morrow, Bower, & Greenspan,
1989; Suh & Trabasso, 1993; Wilson, Rinck, McNamara, Bower, & Morrow,
THEMATIC STRUCTURES AND INTEREST 135

1993). But even then we may assume that readers do not make any more in-
ferences than necessary. Inferences linked to causal relationships that are rel-
evant in the light of the present situation take precedence over those that are
not necessary in order to understand the action now taking place. Let us look
at two examples of research findings consistent with this claim. First, if a story
contains a hierarchy of goals, the present action is interpreted in the light of
the closest subgoal (Suh & Trabasso, 1993). Second, Wilson et al. (1993) re-
ported a steering influence of task characteristics: a detailed model of the sit-
uation, specifically the exact location of the protagonist, was not constructed
until attention was actually focused on the protagonist.
Conventional narratives, including the traditional feature film, steer the
reader (in our case the natural film viewer) in the direction of "story-driven"
processing. This term was devised by Vipond and Hunt (1984), in contrast to
"point-driven reading," in which the reader is interested mainly in discover-
ing the point of the author. (Here, "point" is a kind of global linguistic act,
which should not be confused with Wilensky's "point.") This is the aim of lit-
erary texts. The reader can make use of the surface structure of the text (the
discourse) in trying to understand the author's point. In contrast, story-
driven processing largely disregards the surface structure. As noted earlier,
the traditional feature film is designed to maintain the diegetic illusion at the
expense of the viewer's awareness of the artefact. The intentions of the maker
are not taken into consideration. Viewers of the traditional feature film may
be seen as trusting readers (Kay, 1983). They do not look any further forward
than they can see, as a rule no further than the current action. And even
within the action being played out, the view forward is limited. There is little
use in speculating, when the road to the (happy) ending is paved with rever-
sals of expectation.

Recursiveness of Thematic Structures

All things considered, and despite the fact that they differ in degree of com-
prehensiveness, the cognitive models described earlier illustrate how difficult
it is to differentiate between thematic structures by means of size. In many
cases they do not allow for a distinction, in any absolute sense, on the basis
of the size of the action unit they govern. This has to do with the fact that the
sequences of goals and plans in a story display a recursive structure: a higher
goal is attained through the realization of subgoals.
The theme Allied Deception (Dyer, 1983), for example, may integrate
goals and plans of the protagonist on various levels. A narrative may in its en-
tirety recount the perpetration of fraud through alliance, as in Rosemary's
Baby (1968) where Mia Farrow becomes the victim of a satanic society,
which, as gradually becomes clear, includes all those whom she trusts. This
TAU, Allied Deception, governs virtually the entire plot of the story, since it
136 CHAPTER 4

encompasses one of the main goals and the associated plans of the antago-
nist. However, the same thematic structure may also serve to integrate the
information on a more elementary level of the action. To realize a goal that
is higher up in the hierarchy, namely, to maintain the level of their sales, the
protagonists of Tin Men (1987), two hucksters of aluminum siding, in one
episode conspire to defraud an unsuspecting couple. In other episodes of this
film, the Allied Deception goal is attained by means of other schemes. One
theme that operates on various levels of the goals and action sequence then
predicts different typical complications and solutions for each separate level.
The recursiveness of thematic structures is related to Principle 3 of inter-
est. The immediate prospect of return is determined by progress in the di-
rection of the most deeply embedded structure, such as achieving a Success
in solving a smaller Problem. This Problem is embedded in a larger thematic
structure, for example, via a relationship subgoal/main goal. The state of af-
fairs with respect to the construction of the larger structure determines the
background return.

Example 2: Punishment 82

Background of the Analysis

In the following analysis an effort is made to provide insight into the man-
ner in which the anticipations of the preferred final state are related to the
various NRs that together determine interest. Because these variables cannot
be quantified precisely, it is not possible to demonstrate the activity of the-
matic structures in the form of a simulation program. The analysis, on the
other hand, is the qualitative representation of a hypothetical process, a
model protocol, which highlights the joint effect of the principles of interest
dealt with in chapter 4 and the cognitive thematic structures covered in this
chapter.

Materials. For purposes of the analysis a Dutch feature film, Straf (Pun-
ishment, 1974) by Olga Madsen, that was shown as a short film in traditional
cinemas, was selected. The film is only 11 minutes 30 seconds long, and thus
eminently suited to be analyzed in its entirety. It would appear that general-
ization in the direction of the processing of regular-length feature films is not
inconceivable, since within its short space, the film reflects the full canonic
range of events characteristic of the traditional narrative. The following is an
informal summary of the story on which the film is based:
82
This section is based on a previously published study (Tan, 1986a). I would like to thank
Minet de Wied and Dorien van Otterloo for their part in the standardization study, as well as the
10 subjects who participated. My thanks also goes to the four subjects who carried out the time-
consuming clustering work.
THEMATIC STRUCTURES AND INTEREST 137

Marjan, a little girl of about eight, smashes her father's violin. She goes down-
stairs to join her mother and younger brother in the living room. After a while
her father arrives home from work. He appears to be a first-class tyrant. The
family sits down to the evening meal. After dinner the father goes up to his study
and discovers the remains of his violin. To punish Marjan, he cuts up her dolls
and hangs them up on a string.

Structural Analysis. Five steps can be distinguished in the structural

analysis:
1. Identification of elementary events,
2. Specification of the hypothetical thematic structure, that is, the identifi-
cation of affect states and primitive and complex plot units,
3. Empirical verification of the hypothetical thematic structure obtained
in 2,
4. Determination of progress in the direction of the ultimate thematic struc-
ture per elementary event, and
5. The determination of foreground and background NRs and the intensity
of interest at successive moments of presentation.
Step 1, the identification of elementary events, was carried out by empirical
m e a n s (Tan & de Wied, 1986). Four subjects then reduced the 181 elemen-
t a r y events identified in that study to 24 clusters, designated by the re-
searcher as shown in Table 5.3.

TABLE 5.3
List of Events in the Film Punishment

1. Marjan goes into her father's room and 14. Father completes his ritual reading of the
destroys his violin (with a pair of scissors). newspaper.
2. Marjan is called downstairs by her 15. Father tells Marjan how she ought to set
mother. the table.
3. Marjan goes downstairs and into the liv- 16. Marjan sets the table.
ing room. 17. Father forbids Robbie to mash his food.
4. Marjan and Robbie sit down to play the 18. Father corrects all the members of the
piano. family.
5. Mother sits watching them. 19. Robbie tries to recount an experience but
6. Mother watches for Father. is interrupted by Father.
7. Mother announces that Father is coming. 20. Marjan is belittled by Father.
8. Marjan, Robbie, and Mother straighten 21. Everyone gets up from the table and Fa-
up the living room. ther goes to his study.
9. Father comes in. 22. Marjan waits in suspense in the kitchen.
10. Marjan makes no response. 23. Marjan is called upstairs by Father and
11. Father carries out his homecoming ritual. sees the slashed dolls.
12. Mother urges Marjan to go set the table. 24. Marjan and Father exchange glances; pen-
13. Marjan goes off to set the table. sively, Marjan reaches for the scissors.
138 CHAPTER 4

Hypothetical Thematic Structure. The central complex plot unit is a

Retaliation, the first element of which, the positive affect state, corresponds
to the destruction of the violin (Event l). 8 3 The positive affect state, part of
a Success on the part of Marjan, forms a Mixed Event, as it signifies a Loss
for Father. The Mixed Event connects Event 1 with Events 22 and 23, in which
Father devises a Loss equivalent to his own and shows Marjan the result (Suc-
cess Born of Adversity). The affect state of Events 2 through 20 would appear
to have a lower value than that of Events 1, 22, and 23. They are also more
difficult to relate to the complex and more primitive plot units already men-
tioned. In accordance with an accepted procedure, (Gee & Grosjean, 1984),
a function of Exposition or Setting was attributed to these events. They help
to clarify the relationships within the family and, above all, contribute to the
characterization of the father. The Setting category is lacking in Lehnert's
plot units, but they must often be presumed when a story is being analyzed
(Tan, 1986a; Tan & Diteweg, in press). The ultimate hypothetical thematic
structure of the film narrative is shown in Fig. 5.2.
The Setting clearly suggests a hate relationship between Marjan and Fa-
ther. This means that the destruction of the violin and Father's answer were
preceded by mental acts, namely the desire to deal the other a definitive blow.
Thus Retaliation overlaps two Malicious Acts. These consist not only of a
Mixed Event, the harm done to the other person was planned in advance and
counts as a Success.
The last event corresponds to a mental act that is the result of the desire
of both to settle the conflict once and for all. Father's look seems to be say-
ing that Marjan had better not try anything again. But the way she thought-
fully fingers the scissors could mean that she is planning a new Malicious Act.
The most comprehensive plot unit then presents itself, namely, Starting Over.

Informal Empirical Testing. The centrality of the Retaliation theme in

the cognitive representation of the action of the film was verified informally
by questioning subjects who had seen the film in its entirety. Some 24 sub-
jects, first- and second-year arts students, were asked to name the two most
important events in the film. All 24 mentioned the destruction of the violin,
while 18 subjects also mentioned Marjan's confrontation with the slashed
dolls. When asked about the reason for the second event, 16 of the 18 sub-
jects gave some equivalent or other of Retaliation related to the smashing of
the violin: revenge (6), punishment (4), retaliation (3), and hitting back (3).
However, by no means did all the subjects recognize the most comprehensive
plot unit: Starting Over. In answer to the question: "What is the significance
of the fact that at the end of the film the scissors are produced again?" only
10 of the 24 subjects answered that this could be seen as the prelude to a new
Malicious Act episode.

83
All the names of plot units are taken from Lehnert (1981) and Lehnert and Vine (1987).
THEMATIC STRUCTURES AND INTEREST 139

FIG. 5.2. Detailed plot unit configuration of the film Straf/Punishment, including Setting in-
formation. Mml-Mm4: mental acts Marjan 1 - 4 ; Mfl-Mf3: mental acts Father 1 - 3 ; + m l - + m 3 :
positive affect states Marjan 1; - m l : negative affect state Marjan 1; + f l - + f 2 : positive affect state
Father; - f l : negative affect state Father 1; 1, 23 and 24: events corresponding to affect states
(see Table 5.3).

On the basis of a summarizing study involving 31 new subjects, it was clear

that Events 2 through 20 did indeed contribute to the Setting category (Tan,
1986a).
The present research is obviously not suitable for testing the more detailed
thematic structure of Fig. 5.2, let alone for the empirical testing of Lehnert's
proposal. This would require a more sophisticated analysis of the summaries,
140 CHAPTER 4

one that could bring to light clear traces of the constituent primitive plot units.
Furthermore, an empirical measure for the strength of the association be-
tween various plot unit configurations would be needed, in order to subject
the theoretical structure as a whole to test. 84

Development of the Thematic Structure during Processing, and Pro-

jections Generated. Now that we have an impression of the thematic struc-
ture established following the processing of the film as a whole, we can trace
the development of the constituent plot units involved and their relationships
during processing. In doing so, however, we must exclude prior knowledge of
the final outcome, at any rate to the extent that this result could not be pre-
dicted at that particular moment. (Principle of interest 1, chap. 4). I confine
myself to a description of the formation of the most complex plot units, those
that appear in capitals in Fig. 5.2. It is possible to indicate for each elemen-
tary event the extent to which there is an increase in likelihood that a com-
plex plot unit will be instantiated as a theme by the subjects.
1. A Success is instantiated: the intention of smashing the violin (Mm3) is
realized. There is a possibility of the start of Retaliation. It is probably
clear that the destruction of the violin is an intention that has succeeded
(+ml) and that it constitutes a Mixed Event, that is, one that will bring
harm to someone (who?). In addition, a reaction of that someone may
be expected: Mm3 -a-> + m l — ? -t—> +? and - ? -m—> M?. It is still un-
clear how negative the reaction will be for Marjan. The question arises of
what drove Marjan to her deed. Search for the Mental act Mm2 and any
Setting preceding it.
2. Very weak Setting information, in itself no occasion for anything resem-
bling a Malicious Act.
3. Idem.
4. Idem.
5. Idem.
6. Idem.
7. Idem.
8. Idem. In this sequence, 3 through 8, the foundation is laid for the Set-
ting; the family is clearly under some kind of strain. What is it? Sneak-
ing suspicion that it has something to do with the father.
9. Filmed in such a way that the father is seen as a threatening figure. The
surprise element represented by the unusual father figure is constrained
by the previous Setting. Retaliation will by now have suggested itself to
84
To the best of our knowledge, Gee and Grosjean (1984) are the only researchers who have
undertaken efforts in this direction. They have successfully used the length of the pauses be-
tween sentences and phrases during the recounting of a story as a measure for the association
of theoretically distinguished plot unit clusters.
THEMATIC STRUCTURES AND INTEREST 141

many of the subjects, especially those who suspect that Father is the
owner of the violin. Also contributes to the formation of the Setting. It is
indeed the father from whom the threat emanates, and this could feed
the suspicion of serious negative affect for Marjan in the Retaliation.
10. Confirms suspicion that the violin is indeed Father's, and reinforces Re-
taliation. However, it also becomes clear that there is no immediate
prospect of a serious negative affect state for Marjan, because Father
does not go upstairs.
11. Strong Setting information that could introduce t h e beginning of Mali-
cious Act by Marjan, in other words, her intention to smash the violin in
order to deal Father a blow. A crucial element here is the fact that the
characterization of Father as a tyrant takes place is from the point of view
of Marjan.
12. Weaker Setting information.
13. Idem.
14. Strong Setting information, negative affect Retaliation and Malicious Act
by Marjan becomes more likely.
15. Idem
16. Weaker Setting information. A rapid answer to the question of the con-
sequences of the destruction of the violin, the rounding off of Retaliation,
and the seriousness of the accompanying negative affect for Marjan is
now even less likely.
17. Strong Setting information, negative affect Retaliation more likely, Mali-
cious Act Marjan more likely.
18. Idem.
19. Idem.
20. Idem. Malicious Act by Marjan will probably be rounded off, in other
words, Mm2 and +m2.
21. Father will now discover the violin. Expectation that Retaliation is about
to be closed and that an answer will soon be forthcoming concerning the
nature and seriousness of the cluster Mf3 -a—• +fl — m l .
22. Does not alter 21.
23. Closure of complex Retaliation plot unit. Nature of the punishment is a
postdictable surprise. This was not expected, but given the nature of the
father, it is understandable. Malicious Act of Father is immediately sus-
pected, as a result of the inference of Mf2 -a—> +f2, in which Mf2 indi-
cates that Father intends to mete out something much nastier than a ped-
agogical punishment. 85 Father's Malicious Act could have been picked up

85
Following Graesser and Clark (1985) it would be preferable to speak of "bridging" rather
than inference, since in the inferences there is little or no projection of a pattern for which match-
ing with a future input is sought.
 142 CHAPTER 4

from the Setting even before 23, but the indication was so weak that up
to now it has not been mentioned.
24. The fingering of the scissors is a postdictable surprise. The thematic
structure of the entire film up to now can easily be adjusted: a relation
of equivalence is established between Mm4 and the now inferred Mml:
Starting Over. The Hate theme, which as Setting provides the back-
ground for Mml and Mfl, can be rounded off. If the Malicious Act by
Marjan were not already assured, it would now be a certainty.
The analysis illustrates the functioning of plot units as a thematic struc-
ture that generates projections. The process corresponds to an on-line model
for the comprehension of narrative prose devised by Graesser and Clark
(1985). Both the primitive and the complex plot units may be seen as generic
knowledge structures, which according to these researchers are instantiated
on the basis of the activation of a partial pattern by the stimulus. Thus a Mixed
Event carried out deliberately calls up a Retaliation plot unit, because the
former is a major component of the latter. The portions of this complex plot
unit that have not yet been matched by incoming events remain active in the
foreground, as long as there is a clear prospect of matching relevant events,
at any rate, in the near future. There need be no specific expectation con-
cerning the exact contents of the matching event. The subjects will have very
little idea of the nature and severity of the punishment until these are re-
vealed. They will tend to search for an event that meets the broad constraints
of the complex plot unit: a deed planned by Father with negative consequence
for Marjan.
In some cases the expectation or projection, to which we have repeatedly
referred, can best be paraphrased by an open question: "what motive would
Marjan have for destroying the violin?" or "who would be harmed by the
smashing of the violin?" In other cases the projection is more specific: "Fa-
ther is now going to punish Marjan in a terrible manner, by doing this or that."
Graesser and Clark adopted a neutral tone, referring to Projections that are
cast forward by the GKSs. In any case it is always a smaller segment of a more
complex plot unit that functions as a demon for which a matching input pat-
tern is sought. 86
It will be clear from the example that available knowledge of the world only
serves as the basis for an understanding of the film narrative. Complex plot
units such as Retaliation are permanently available to viewers as schemas:
they understand that there is a reason for revenge, that there is a plan and
that plan must be executed. But when it comes to filling in the actual values
86
The term was introduced by Selfridge, in relationship to his well-known computer model of
pattern recognition. See Selfridge and Neisser (1960); for applications to the comprehension of
narrative texts, see Charniak (1972) and Dyer (1983); for computational formalisms, see Win-
ston and Horn (1981).
THEMATIC STRUCTURES AND INTEREST 143

of the attributes of this schema (where, when, and precisely how) in order to
understand the film, there is a kind of planning required that goes beyond
ready knowledge in schematic form. In this connection, Barsalou (1991) de-
scribes how planning is the result of a combination of, on the one hand, sta-
tic knowledge of the world, that is, taxonomic categories, and, on the other
hand, more dynamic goal-derived categories. "Vacation," for example, is a
taxonomic category. In planning a vacation there are goals that lead to re-
stricted vacations, such as vacation fun in combination with privacy and
peace and quiet. Goal-derived categories considerably narrow the choice of
possible candidates: not only do they impose certain constraints, but they
also generate ideal instances. In our vacation example, all the top destina-
tions are out, while—depending on the entertainment preferences of the
planner—ideal places are selected. Returning to our film, this means that as
soon as the taxonomic category Retaliation is active, from about input propo-
sition 10, the viewer begins to search for a more precise specification. What
is in any case typical for an understanding of this film is that not only the
choice of possible instances is shrouded in uncertainty. This choice is logi-
cally preceded by the determination of constraints and ideals, and here, too,
there is uncertainty. It gradually becomes clear from the growing body of Set-
ting information that the worst "ideal" is not inconceivable: a physical pun-
ishment, not even excluding murder. One constraint could be the extent to
which the genre permits certain denouements. But even the most macabre
outcome imaginable still does not appear impossible. An ideal that is specific
for fiction might be that the punishment must fit the crime, not only juridi-
cally, but also poetically. I am assuming here that such constraints and ideals
are considered, even though viewers are not necessarily trying to arrive at a
choice or decision, as they might do when planning an action that they them-
selves intend to carry out.
In rudimentary planning, unavoidable surprises are in any case post-
dictable, that is, they fall within certain boundaries. Nevertheless, the con-
sideration of constraints and ideals and the incidental odd guess may be seen
as a substantial cognitive investment. This certainly holds true for the toler-
ance of uncertainty and the anticipation of the precise outcome.
In this example the affective concern lies mainly in the fate of Marjan. The
seriousness and nature of - m l , the punishment, is what viewers are most anx-
ious to know. The affective value of that outcome is heightened by the fact
that the film narrative retains the perspective of Marjan, and by means of the
Setting generates expectations concerning the outcome that evokes such
fears. In addition, the main themes, in their ultimate form, could honor var-
ious views of life a la Andrews (1989). Starting Over and Retaliation border
on the tragic and the ironic visions. The romantic vision, on the other hand,
comes off rather badly, so that highly romantic personalities might be less
likely to appreciate the film.
144 CHAPTER 4

Hypothetical Course of Net Return and Interest. The hypothetical

course of interest and its determinants is shown in Table 5.4. In order to fol-
low the changes in interest intensity, we must compare them with the devel-
opment of the projections discussed in the previous section. The course of
the foreground Net Return, abbreviated to fgNR will be described in some de-
tail by following, from one event to the next, the difference between the NR
expected in the present action episode and the NR already obtained in that
same episode, in other words (NR(f) - NR(p)}. The development of the back-
ground NR, abbreviated to bgNR, on the other hand, will be represented in
less detail, whereby the component terms NR(FF) and NR(PP) will be omit-
ted. The absolute values used in this analysis, including the initial value of
NRs and Interest, are arbitrary. What is important is the relationship between
attained and expected return and the changes that take place as a function
of the input propositions. I have opted to determine the strength of Interest
on the basis of the simplest comparison, as follows:
I = { sum of the number of minuses Attained NR(p) +
sum of the number of pluses Expected NR (f) }.
Event 1, the smashing of the violin produces a foreground return attained
in relation to an important Success, important in the sense that in some way
or other it will surely be part of the final cognitive representation of the fic-
tional action. However, far more questions are called up than answers, and
the beginning of an affective investment gradually becomes possible: this is
not pleasant for the owner and thus will lead to something that is not pleas-
ant for Marjan. We may assume that the investment surpasses, to a consid-
erable degree, the return attained, and this is expressed by means of five mi-
nuses for the value of the fgNR attained. Because very little else is known and
it is somewhat unusual for a film to present the viewer with such riddles at
the very beginning, it is reasonable to expect an answer to these questions in
the short term; however, the possible emergence of new riddles is also taken
into account.
Thus the fgNR expected is positive, but to a lesser degree than the fgNR
attained is negative: three pluses, say. The bgNR that the viewer expects for
the film as a whole, (i.e., including the period that cannot be overseen), is
not particularly high. And as we have noted, answers are to be expected in
the short term. There are as yet few investments kept in store in the back-
ground; we set the bgNR at two pluses, see Table 5.4. The possibility of a Re-
taliation might count in favor of the bgNR, but it is still only faintly present
in the background.
Because the bgNR as a piece of background information is of lesser im-
portance for interest (Principle of interest 3), we can set the value of the in-
terest at, say, 8, the difference between expected return and return already
attained, which reflects a full determination by foreground investments and
returns.
THEMATIC STRUCTURES AND INTEREST 145

TABLE 5.4
Global representation of the hypothetical course of foreground and background NR
and resulting interest among viewers of the film Punishment

Event fgrNR attained fgrNR attained background NR Interest

1 +++ ++ 8
2 — ++ +++ 4
3 — ++ +++ 4
4 -- + +++ 3
5 -- - +++ -1(+1)
6 -- - +++ - 1 (+ 1)
7 + +++ 3
8 + +++ 4
9 ++ +++ 6
10 ++ +++ 6
11 ++ ++++ 6
12 ++ ++++ 6
13 ++ ++++ 6
14 +++ ++++ 6
15 +++ ++++ 6
16 - - ++ ++++ 4
17 - - +++ ++++ 5
18 - - +++ ++++ 5
19 - - +++ ++++ 5
2 0 ' - - +++ +++++ 5
21 +++++ ++ 10
22 ++++++ + 12
23 - ++ + 3
24 -- + (++) 3

Event 2 implies that for the time being no resolution can be expected. This
means that the demons, for which no match can be expected in the coming
input, are allocated background status. 87 For example, we subtract three mi-
nuses from the fgNR attained and one plus from the expected fgNR. Interest
then declines.
The demons derived from the consequences of Marjan's deed, the begin-
ning of the path to completion of Retaliation, (+ml — ? -m-> M? -a->?), will
now become background investment, while the expected return associated
with it will also disappear into the background. Sleeping demons increase the
bgNR, as this is based on an estimate of the value of the maximum preferred
final outcome. A complex plot unit such as Retaliation meets the requirement
that, if completed, it must close the action of an episode in a satisfactory man-

87
In the model devised by Graesser and Clark, GKSs can be deactivated for precisely the
same reason. They also refer to experimental evidence concerning various activation statuses of
knowledge structures in the comprehension of narrative prose. See Graesser and Clark (1985,
pp. 194, 196).
146 CHAPTER 4

ner, because the course of the entire action can b e summarized in the form
of a "bromide," e.g., "Where vice is, vengeance follows." The plot structure
is neatly symmetrical and closed. In addition, the road leading to completion
of this plot unit promises a considerable dose of emotion.
Thus the bgNR increases somewhat, and it is clear that the film is not con-
tinuing as it began, that is, the major questions will be answered later.
Events 3 through 8, the whole episode in the living room up until Father
enters, bring little alteration in the NRs. There is very little return, but the
active investment is not high either. In contrast to the surprise beginning,
which made possible a great deal, there is now very little return to be ex-
pected in the short term. Because nothing is happening, the fgNR may drop
to a minimum. Interest is then at a low. The bgNR for the film as a whole de-
creases less markedly, because (a) the nonactive background investments
and returns, the unanswered questions, are passed on, (b) the sympathy for
Marjan may now slowly increase, thanks to increasing comprehension of the
Setting and the fact that the events are being narrated from her perspective,
and (c) because an expectation of the bgNR based on pragmatic principles
plays a role here, namely that films ultimately give what they promise, and
often more. This background return could serve here as a safety net for in-
terest. There is still a certain willingness to elaborate the input, because in
the long term the prospects are still reasonably good. At this point of mini-
mum discrepancy, the background return could correct the interest value: 1
point higher, say, following Events 5 and 6. In 7 and 8, however, there is light
at the end of the tunnel: the announcement of the arrival of the father and
the excitement that this generates indicate that there may be more return to
be attained here: the viewer now looks forward to the entrance of the father.
Event 9, Father enters, projections are again being made. Retaliation be-
comes stronger and might even be realized in the short term. Worry about
Marjan's fate is active investment, worth, say, a minus and a plus; not much
more, since there are no indications that the father will immediately discover
the damage to the violin.
Event 10, Marjan makes no response, again relegates a portion of the Re-
taliation investment to the background, since Father does not go straight up-
stairs. But, in the light of Marjan's reaction, it also makes it clear that the
owner of the violin is now in the house. On balance, no change.
Events 11 through 20, Father's ritual and the entire evening meal push the
investments associated with the Retaliation theme further into the back-
ground, since when the family sits down at the table it is clear that the match-
ing of discovery ( — f l -t-> +) and its consequences will not immediately be
forthcoming. Here a decisive role is reserved for generic knowledge of the
world, a kind of "evening meal" script, in the time indexing of the expecta-
tions involved. There is a slight increase in the attained and expected fgNR,
as the characterization of the father will probably be seen as amusing, espe-
THEMATIC STRUCTURES AND INTEREST 147

cially to the extent that it is further developed. The return attained also in-
creases slightly when the reason for Marjan's deed, Mm2 -m—• Mm3, is added.
All in all, the interest could decline very slightly, as shown in Table 5.4. The
bgNR, however, increases because the expectation of an unusual punishment
beyond the present episode is greater, the more the Setting, in particular the
characterization of the father, is worked out in detail
It is quite conceivable that the expectation regarding the end of the film,
and thus the expected total return, increases here, because the Setting en-
riches both cognitive and affective aspects of the preferred final outcome.
There is more to the story than the spiteful act of a spoiled child that triggers
the father's response. A Malicious Act is poetic justice, a suitable closure of
a history of oppression. A plus is added from 11 on and another one some-
where between 11 and 20.
Event 21, Father goes to his room, brings the matching of the missing units
of Retaliation very close indeed. Almost all the background investments set
in motion by Event 1 and reinforced and supplemented by the Setting and af-
fect, slightly in favor of Marjan and against Father, are now activated in the
foreground, expressed by, say, three extra minuses. Fulfillment appears close,
but it may yet be delayed. For this reason the expected fgNR increases some-
what, but less than that already attained decreases: from three to five pluses,
say. The bgNR, of course, decreases.
Event 22, Marjan waits, is pure delay. The fgNR attained thus decreases;
for example, one minus is added. The expected fgNR increases, because as
time goes by, the definitive answer becomes inevitable: say, one extra plus.
The discrepancy between attained and expected fgNR reaches maximum, and
interest rises toward a climax.
At the end of Event 23, which shows the outcome: Marjan is confronted
with the slashed dolls, a big return is attained. Almost all projections that
arose in the past have now been fulfilled. The point, that the maliciousness
of the punishment exceeds that of the deed, is unexpected. Retaliation is con-
cluded with the adage "He who laughs last, . . ." and this leads to a final in-
crease in the foreground return attained. For some viewers the film is now
over, which causes the fgNR and the bgNR to drop sharply. But one might
also wonder what Marjan is going to do now. Admittedly, this is a weak in-
vestment, because we already have an entirely acceptable final outcome, both
cognitively and affectively. Accordingly, interest declines after Event 23. This
also explains why so many subjects do not pick up the hint of a sequel in the
following Event; the tendency toward elaboration (see def. 4.1, chap. 4) has
dropped to an absolute minimum. On the other hand, it is conceivable that
some subjects will continue to put forward anticipations, if only the fleeting
thought: "Is it really over now?" due perhaps to the self-reinforcing effect,
Principle of Interest 4. We can represent this by not immediately allowing
both the attained and expected fgNR and bgNR to drop to zero.
148 CHAPTER 4

Following the reception of Event 24, Marjan pensively reaches for the scis-
sors, attentive subjects will no doubt recognize the complex plot unit Start-
ing Over, which calls up the notion of what Marjan's next move might be, and
the weak—and idle—hope that they might be allowed to see what happens
next. The completion of the thematic structure in this way does briefly evoke
the fleeting interest that is characteristic of the effect of the open end. The
filmic realization of this final event contains signs that the film is now over.
The musical theme fades, the film comes to a halt, and the credits begin. On
the other hand, it is quite possible that certain viewers, forgetting that they
are watching a short film, will continue to expect a new scene or episode,
right up to the point where these signs become evident. Their interest de-
clines but not as far as that of the viewers to whom it is clear that the film is
over. They await the introduction of a new scene, saving the total return, in
order to use it as a first criterion in determining the expected NR in the fol-
lowing scene. Their bgNR increases somewhat, which we indicate by two
pluses in parentheses.

Concluding Remarks in Connection with Example 2

Elsewhere results will be reported of empirical research that broadly re-
flects the hypothetical course of interest given in Table 5.4 (Tan, in prep.).
The analysis presented here shows how the principles of interest sketched in
chapter 4 might operate in the processing of a short feature film. The terms
foreground net return "attained in the past" and "to be expected in the fu-
ture" have been given concrete form. An illustration has also been given of
the limiting role that background return can play at various points in the film.
In addition, the role of thematic cognitive structures has been highlighted.
We have also seen how a theme that has been evoked more or less automat-
ically generates projections of input patterns. Matching between input events
and such projections definitely establishes the complex theme as component
of the preferred final outcome and prepares the way for the instantiation of
more complex themes.
However, the example is quite simple, and care should be taken in gener-
alizing in the direction of longer feature films. The major problem is perhaps
the fact that no attention could be given to the way originally instantiated
themes are replaced by more suitable ones, because the narrative of Pun-
ishment invariably postpones the fulfillment of expectations, while keeping
future results vague. At the same time, there is little or no temporary mis-
leading of the viewer, in the sense of creating an expectation that later proves
to be false. In a comparable but longer film, for example, a temporary but re-
peated suggestion might have been put forward that Marjan may yet escape
her punishment or a specific punishment might initially be suggested, which
is later disproved. Furthermore, in a longer film there would be more room
THEMATIC STRUCTURES AND INTEREST 149

for repeated suggestions that discovery and punishment are imminent. The
mechanism of pruning, which consists in doing away with erroneous or ir-
relevant information and expectations has been clearly described by Graesser
and Clark (1985) in their on-line model of the comprehension of narratives,
so that anyone interested in extending the analysis to more complex films is
referred to this model. Despite these reservations, the short feature film may
be seen as a miniature of the traditional full-length feature film, just as the
short story is a miniature of the adventure novel. The psychological explo-
ration of the characters is, of course, minimal, although the difference be-
tween short and full-length feature films in this respect should not be exag-
gerated. The longer film also makes use of somewhat sketchy characteriza-
tion, as we will see in chapter 6. The "tale" quality of this type of film, whose
plot is based on "point" and paradoxical surprise, appears in concentrated
form in the short film and eminently so in the film analyzed here.
In the case of a brief and simple film narrative, such as that of Punish-
ment, the complexity of the processing act is such that the limits of the pos-
sibilities afforded by a "pencil-and-paper" simulation, for which the above
analysis provides a basis, have been reached. 88 It is just possible to keep track
of the changes and the possible interactions of various mechanisms. For ex-
ample, it is assumed that the increase in attained and expected foreground
return, as a result of the characterization of the father in Events 11 through
20, together with the decrease in investment, could on balance lead to a de-
cline in interest. However, this interaction cannot be further specified, and
there is an implicit subjective weighing in this statement. Further research
might focus on computer simulation of the process by which expectations are
evoked by themes. In addition, simulation would require further formaliza-
tion of the processes involved in generating certain patterns by means of
themes and the matching and integration of input patterns, for example, by
means of production systems.
The greatest problem is no doubt the implementation of the difference be-
tween foreground and background expectations. How does the system know
that, for the time being, a certain inference will not be matched? In the sam-
ple analysis this knowledge has been kept implicit. The fact that the viewer
knows that the discovery of the smashed violin is only possible if the father
goes into his study is of influence here. If the system keeps track of infor-
mation about locations, and if its knowledge of the world includes such in-
formation as the fact that the evening meal is taken in the living room, then
it knows that as soon as the family sits down at the table, the danger of dis-
covery is almost nil, at least for the time being. However, there is other knowl-
edge, of a different sort, on which the viewer bases the intuition about
88
de Groot (1965) had already some time ago pointed to the usefulness of the "paper simu-
lation."
 150 CHAPTER 4

whether or not something is about to happen. Some action sequences, such

as a meal, are long, according to the simple, scriptlike, knowledge of the
world. The deadlines, which the classic film narrative regularly employs to
provide viewer expectations with a limiting time index (Bordwell & Thomp-
son, 1986), address the duration and the natural end of events such as rep-
resented in schematic knowledge structures that the viewer effortlessly has
available. However, the narrative may handle such a long action in a frag-
mented or elliptical manner. It may well be pragmatic knowledge that ex-
plains why the viewer, having watched this action for a time, begins to sus-
pect that the meal will be shown at length.
Furthermore, the tempo in which the events have succeeded one another
up to the present moment, as seen over a larger or smaller portion of the film,
may be of influence in assessing the future presentation of the point at which
matching of an outstanding projection will follow. If the narrative has been
fairly slow up to now, then the viewer will expect resolution to occur later
when events have proceeded at high speed. And finally, the actual duration
of the projections involved may also be a cue for resolution. The longer an in-
ference survives the input of new information, the greater the chance that it
will be matched. The research carried out by de Wied (1991, 1995) is rele-
vant for the study of the last two factors. It suggests that for the creation of
suspense, there may be an optimum duration for expectations, given a cer-
tain tempo of events.
And finally, future research aiming at development of a simulation pro-
gram for "on-line computation" of interest, would require considerably fur-
ther parameterization of the principles of interest.

CONCLUSION

All things considered, it would seem to be possible to arrive at a description

of the on-line process involved in the comprehension of feature films by a nat-
ural audience, a description based on the operation of generic knowledge
structures in relation to the course of the action and capable of explaining
the audience's interest. The generic knowledge structures or themes involved
contain information on the action that may include larger or smaller chains
of action and interaction. They are hierarchically ordered, and thus make pos-
sible a bottom-up explanation of the process of comprehending the fictional
action.
These themes, after total or partial instantiation, have the power to relate
each input event to the existing representation of the action up to now. This
may be a direct manner of relating, notably "role binding," or what Graesser
(1981; Graesser & Clark, 1985) referred to as the "bridging" of minor link-
age problems, by means of inference. This capacity is an important initial
THEMATIC STRUCTURES AND INTEREST 151

source of return. Second, there is the quality that has been highlighted in the
present chapter, namely, the fact that themes can generate anticipations con-
cerning the future course of the fictional action. The anticipations consist of
more or less specific events and states that are not specified, or not fully spec-
ified in the existing thematic structure. And third, themes—in particular the
more complex ones—represent generic knowledge of a preferred final out-
come that, provisionally anticipated and surrounded by inevitable uncer-
tainty, can be derived from the fictional action encoded thus far.
In these various—and not entirely unrelated—ways, the thematic struc-
tures should serve as generators of investments and returns, and thus as deter-
minants of interest. This theoretical proposition supplements the explanation
for the interest potential of thematic structures that is based exclusively on
the assumption that there is an intrinsic element of interest in the subject
matter of certain themes. The cognitive thematic structures distinguished
here, which were originally developed in an effort to shed some light on the
comprehension of narratives, were shown capable of producing and sustain-
ing affect, that is interest. Themes are rounded portions of a story that have
been torn apart and are being reintegrated in the course of the narration, cre-
ating the conditions of conflict, complication, and retardation, followed by
cognitive and affective closure.
 Character Structures,
6 Empathy, and Interest

INTRODUCTION OF TERMS

There are various contexts in which the viewer has the sensation of being "in
the film," a sensation characteristically experienced by the natural audience
of the traditional feature film. As a viewer I do not only entertain the illusion
that I am present in the scene—the diegetic effect—I may even feel that to a
greater or lesser degree the adventures of the protagonists are actually hap-
pening to me. This experience can take many different forms, which makes
it somewhat difficult to describe. In La Peau Douce (1964) we fill in Nicole's
thoughts after Lachenay has turned his back on her in the street in order to
keep their love secret. We have a different experience, when Reuven is struck
full in the face by a baseball in The Chosen (1981); we almost feel as if we
ourselves are the victim. And watching King Kong (1933), one sympathizes
with both the girl and the huge gorilla as the animal majestically undergoes
his fate, amid a rain of bullets fired by stupid, insensitive human beings. In
describing such experiences, people often speak of identification or empa-
thy. Although based on the diegetic effect, these phenomena are clearly dis-
tinct from it. It is quite conceivable, for example, that the illusion of being
present in the fictional world is absolute without the viewer experiencing any
appreciable involvement with the events taking place on the screen. The ex-
act meaning of the terms identification and empathy differs according to the
context and the field of research; the result is an assortment of widely dif-
fering concepts commonly referred to by the same name. (For recent surveys

153
154 CHAPTER 4

undertaken from a broader perspective, see Schoenmakers, 1988; van Vliet,

1991; and Zillmann, 1991b.)
We are fortunate in being able to examine the phenomena of identification
and empathy within quite a limited framework, focusing on viewer emotion
and, more particularly, interest. Before doing so, however, we must mention
one special manner of participating in a feature film. We all know that it is
possible to feel drawn into or borne along by the movement of the camera or
the objects in the scene, or by the stream of images created by montage and
reinforced by the music; numerous lyrical passages in feature films could be
cited as examples. By analogy with the use of music and the visual arts in psy-
chology, it is possible to speak of empathy. Lipps (1906) referred to the total
immersion ("projection") of the self in an object as characteristic of total em-
pathy. Michotte (1953/1991) used the term "motor empathy" for the imita-
tion of the rhythmic movements observed on the screen. (For a general
overview, see also Kreitler & Kreitler, 1972, pp. 2 6 4 - 2 8 1 ) . Following on the
terms introduced in chapter 3, A—artefact—empathy may be seen as the op-
posite of F—fiction—empathy. We will not be examining A empathy here, as
this would involve a separate study of not inconsiderable proportions. More-
over, according to our model, the processes of F identification/empathy in-
volved in the viewing of a traditional feature film in themselves make up a
large proportion of the determinants of interest.
As we know from the Principles laid down in chapter 4, interest is deter-
mined by the dynamics of anticipations and outcomes, called up and re-
inforced by the fictional events. Anticipations and resolution may be seen as
a form of costs and benefits. In chapter 5 we saw that one of the sources of
expectations and anticipations consists in thematic structures, and we ex-
amined their role in the cognitive processing of feature films. A second source
of investment and return, (i.e., another determinant of interest), is to be
found in the expectations and anticipations that are rooted in the viewer's
own knowledge of, and feelings toward, the characters of a film.
Our description of these determinants is based on two simple assumptions.
First, we assume that the watching of a feature film is accompanied by ho-
mogeneous experiences, (i.e., that the experiences of various viewers are
comparable), the reason being that the effects intended by the maker are op-
erative in all viewers. The sadness expressed by a particular viewer when a
sympathetic protagonist contracts a fatal disease is less relevant for our pur-
poses where that sadness is caused by the fact that a close relative has been
similarly struck down. While this fact is extremely important for the viewer
in question and for the manner in which he or she experiences the film, ex-
planations based on life experiences and personality factors can at most help
to account for the preference that certain natural viewers of the traditional
feature film display for a particular genre. Thus when we speak of identifica-
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 155

tion/empathy, we will be referring to processes that in principle operate in

all viewers in response to a certain film. As regards o u r second assumption,
during the discussion of interest in chapter 4 we emphasized the activity of
t h e viewer and the fact that active participation that respects the rules of the
game is rewarded. This point of departure clashes with certain views on iden-
tification that have attained some popularity within contemporary film the-
ory. I would like to begin with a few remarks on this subject.

IDENTIFICATION IN CONTEMPORARY
FILM THEORY: THE TRAGIC VIEWER

Under influence of the psychoanalytical orientation in French film studies,

which has been evident in the last two decades, identification has become a
prominent subject of theorizing (see also the discussion of the etat filmique
in chap. 2). Metz (1975a) introduced a distinction between primary and sec-
ondary identification: the former refers to the identification of the viewer
with him or herself, the latter to the identification with the film character.
Identification with oneself has to do with the fact that t h e viewer is forced to
share the perspective of events presented by the camera. You identify with a
certain view, which ultimately becomes your own. The basis of the expres-
sion "identifying with one's own view" is not entirely clear. We assume that
Metz is referring here to the cornerstone of the diegetic effect, the illusion
that the viewer is actually present in the fictional world as an (invisible) spec-
tator. A similar reading of Metz was put forward by Aumont, Bergala, Marie,
a n d Vernet (1983). According to these authors, primary identification is iden-
tification with the camera. On logical grounds, I have chosen to distinguish
t h e diegetic effect from identification/ empathy.
In its more general sense, the secondary identification referred to by Metz
is an identification with characters. It is possible that what primary and sec-
ondary identification have in common is the reflection of an imaginary, ab-
sent object, based on a lack of some kind, a shortcoming that is at once com-
pensated for and constantly recalled. Film technique denies the existence of
this lack, creating the illusion of presence; in this sense it is a fetish, a sur-
rogate, which is enjoyed not only by the cinephile, b u t also by the average
filmgoer. This tragic representation of affairs is most clearly defined in the
formulation of primary identification by Aumont, Bergala, Marie, and Vernet
(1983). The viewer is prompted by a desire the precise nature of which

. . . est certainement a chercher du cote d'un etat d'abandon, de solitude, de

manque: le spectateur de cinema est toujours plus ou moins un refugie pour
qu'il s'agit de reparer quelque perte irreparable, serait-ce au prix d'une regres-
156 CHAPTER 4

sion passagere, socialement reglee, le temps d'une projection. (Aumont et al.,

1983, p. 172)
. . . is assuredly to seek a state of abandon, solitude, want. The cinema spec-
tator is always something of a fugitive, whose concern is to redress some ir-
reparable loss, even at the cost of a passing regression, socially controlled, last-
ing only as long as a projection, [author's translation].
For Metz, establishing a paradox between what viewers want and what they
ultimately get is the crux of the argument, which is in itself an amusing point.
In the view of Aumont et al., this obscures the issue of the operations that
constitute identification and the circumstances u n d e r which they are carried
out. Aumont et al. propose several such operations that are quite useful, as
we will see. First, there is the operation that consists in following the narra-
tive in search of the Restoration of Balance (chap. 3). In chapter 5 we saw
that something resembling empathy is involved in understanding the various
events, which are themselves governed by themes. This was abundantly clear
from the analysis of the process of understanding the film Punishment Sec-
ond, Aumont et al. mention identification with the character through the op-
eration of a type; third identification with a character due to the structure of
the situation; fourth identification at the level of t h e surface structure of the
film (the decoupage); and fifth identification with the narrative. It is not en-
tirely clear whether some of the factors are identical to or merely influenced
by identification. I hope that the following exposition will help to clarify this
point. One fundamental idea put forward by Aumont et al. is deserving of our
attention, namely, the fact that identification requires no in-depth psycho-
logical explanation, because, given the diegetic effect, it flows directly from
the viewer's attention to, and comprehension of, the film. It may be added
that the same goes for other affective attitudes toward film characters as well.
Let us begin with the viewer's comprehension of characters.

UNDERSTANDING C H A R A C T E R S

Because viewers have the feeling that they are spectators in the fictional
world, a world that has at least some similarity with the real world, it is con-
ceivable that many well-known mechanisms of social cognition are applicable
to their experience of characters in a film. Thus we are justified in seeing the
comprehension of characters as a guided impression formation that extends
to the entire film narrative. There is reason to believe that our comprehension
of fictional characters takes place in the same way as our comprehension of
people in the everyday world or in that of the psychological laboratory
(Hoffner & Cantor, 1991). We examine first a number of insights related to
fictional characters, and second, impression formation by actual individuals,
in an attempt to establish possible similarities.
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 157

Literary Theories on Characterization:

Type Versus Person
In narrative theory there have been a number of attempts to arrive at a sys-
tem for categorizing the characters that appear in traditional feature films.
The proposed systems are in line with recent social psychological views on
impression formation. No doubt the best known distinction that has been put
forward in the field of literary analysis is that between flat and round char-
acters, as launched by E. M. Forster (1927). This distinction also appears in
drama theory, for example, in the contrast between one-dimensional and mul-
tidimensional characters (Pfister, 1977). According to Forster flat characters
or types appear in strong plots, such as those employed by Dickens and Wells.
The strength of the plot precludes the possibility of the reader seeing through
the shallowness of the characterization. On the other hand, one will search
in vain for types in a tragedy. Forster contrasts the type with the round char-
acter, who carries within him or her the unpredictability of life itself. Char-
acters like this predominate in the novels of Tolstoy and Dostoevsky.
In general one would do well to exercise a certain restraint in applying this
distinction to whole genres. However, the usefulness of such a dimension, say,
from type to person to individual (Pfister, 1977), is, in the light of our objec-
tives, undeniable, if only because it implies a kind of trade-off, in terms of
complexity, between the course of the narrative events on the one hand and
the characters on the other. In certain genres, such as the tale or the novel,
the complexity of the narrative is concerned mainly with the way the plot de-
velops. The themes that feed the plot contain a multitude of possibilities for
reversals of expectation, countless shifts and turns, and an abundance of ironic
twists (see chap. 5). In other genres, such as novels or tragedies, the complex-
ity, that is, the postponement of the major outcomes of the narrative, is rooted
in the development of the characters and their interrelationships.
A similar dichotomy in genres is to be found in Barthes (1966). He pro-
posed a distinction that recognizes, on the one hand, functional narratives,
which rely heavily on linear sequences of action, guided by narrative func-
tions, such as those distinguished by Propp and Bremond. The popular fairy
tale is a case in point. On the other hand, there are the indicial narratives,
such as the psychological novel, in which the emphasis is on atmosphere and
characterization. 89 This distinction may overlap that between the popular fea-
ture film on the one hand, and the quality film on the other.
89
Obviously, there are a great many conceivable transitional forms. Thus it is quite possible
that what we designated earlier (chap. 5) as a quality film, notably the so-called psychologically
realistic drama, meets Auerbach's (1946) description of the great realist literary work dating
from before the 20th century, which combines a tragic involvement with the individual, regard-
less of his or her particular place in the hierarchy, and a comic interest in society, all of which
results in an apt representation of the actual conditions. Examples include Host sonaten/Autumn
158 CHAPTER 4

In chapter 5 we said that in the traditional feature film the major focus is
on the thematic action, while less attention is given to an insightful exami-
nation of character, especially in comparison with the so-called modern art
cinema, with its subjective realism (Bordwell, 1985). Pfister (1977) pointed
out that the limited psychological treatment of the characters has a long cul-
tural and historical tradition, while the opposite approach did not make its
appearance until the Sturm und Drang period. This does not mean that the
viewer of the traditional feature film in general, and the popular action film
in particular, is deprived of in-depth psychological examinations of character.
For one thing, even types are to some extent individualized. As Scholes and
Kellogg (1966) noted, Mr. Pickwick and Sam Weller have quite striking traits
of their own, despite their admitted resemblance to other famous literary
duos, such as Quixote/Panza and Holmes/Watson. And although it would be
misleading to speak of an exploration of character in the strict sense of the
word, the viewer must at the very least go along with a characterization that
often transcends routine typification. Harvey (1965) maintained that in
everyday life "the bore bores us" and we are repelled by the hypocrite, and
yet when they appear in the fictional world we cannot get enough of them. In
the second place, the interaction between the various types makes possible
a greater degree of depth. According to Harvey (1965), the protagonist ac-
quires a semblance of depth precisely because there are other characters
playing out their roles in the background. Let us look now at an example of
relationships between different types. Cards are relatively static and pre-
dictable characters, although not necessarily simple, who are often at once
comical and pathetic. (For example, the faithful Cheyenne in Cera una volta
il West/Once Upon a Time in the West (1969), who is invariably turning up
at the side of the hero, the "Harmonica Man," and in the end dies a heroic
death, or the faithful Lieutenant Bates in The Third Man (1949), who catches
Harry Lime's first shot, intended for the hero Martins.) Harvey referred to
Cards as "chemically pure," whereas drama theorist Beckerman (1970) char-
acterized them as "narrow," in a reference to the extent of their potential for
development throughout the drama as a whole. A second example is the Fi-
celle, the representative of the reader/viewer in the fictional world. He or she
is contrasted with the protagonist by virtue of the common sense that the hu-
man, driven, or spontaneous protagonist lacks or is contrasted with the pro-
tagonist in the matter of social background. In 'Round Midnight (1986)
Frangois, who associates with the "tenor hero" Dexter Gordon, could be a Fi-
celle, like Scott Fitzgerald's young alter ego in The Great Gatsby (1974) and
Thompson, the journalist in Citizen Kane (1941).
If no great attention is bestowed upon character in film theory, film criti-

Sonata (1978), Dersu uzala (1975), Under the Volcano (1984), The Dresser (1983), The Dead
(1987), Smultronstallet/Wild Strawberries (1957), and A Woman Under the Influence (1974).
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 159

cism does highlight the significance of characters. The bridge between liter-
ary theories concerned with character reading and psychological theories re-
lated to the comprehension of actual people is formed by Bordwell's account
of the interpretation of film characters by professional film critics (Bordwell,
1989b). However, the schemas devised by Bordwell, on the basis of which crit-
ics interpret characters, are of such a general nature that it may be assumed
that they are also employed by the natural viewers of the traditional feature
film. 90 The comprehension of characters is a major strategy for lending co-
herence to the film text. According to Bordwell, it consists of two compo-
nents. First, characters are conceived of as persons in a folk psychological
sense. They are linked to a body, they display perceptual activity such as self-
awareness, they have thoughts and feelings, they are characterized by per-
sistent dispositions, and they are capable of self-generated actions. All these
qualities are included in the term "intentionality." In the second place, the
image that one has of characters guides one's understanding of all the other
elements of the film narrative and the artefact. Thus there are various con-
nections between character and cinematography. For instance, a threatening
villain may be filmed from a lower point of view, or irregularities of physiog-
nomy may be emphasized by lighting.

Psychological Theories of Impression Formation:

Type Versus Person
Very little psychological research has thus far been devoted to the compre-
hension of fictional characters, in comparison with research focusing on the
processing of narrative action structures and thematic structures. There is,
however, an excellent review study by Hoffner and Cantor (1991), where fre-
quent mention is made of research in which real people rather than fictional
characters serve as stimuli. The comparison between the formation of im-
pressions of media characters and that of real people would appear to be a
valid one, according to Hoffner and Cantor (1991). Somewhat ironically, the
most recent studies on social cognition have for practical reasons employed
mainly imaginary individuals as stimuli, which also blurs the distinction be-
tween the two types of research.
Reviewing the literature on the formation of impressions, we see that the
earlier mentioned distinction between typological and individual-based cog-
nitive structures is the subject of several of the more recent contributions.
The picture emerging from the literature involved in what it is to understand
other people shows a surprising similarity to current views within the field of
literary analysis. However, a major difference is the fact that psychological

90
As I noted elsewhere, it is regrettable that Bordwell did not strive for a more general se-
mantics of characters (Tan, 1990).
160 CHAPTER 4

models strive to capture processes instead of looking only at the final results.
This means that not only various levels of detail are distinguished but also dif-
ferent steps in the categorization of stimulus persons. Let us now take a closer
look at these processes.
The categorization of characters is presumably based on our habit of try-
ing to organize the rich array of sensory impressions confronting us by link-
ing all the different characteristics of the perceived other. It has often been
demonstrated that there is a substantial correlation between observed char-
acteristics, such as intelligence and dominance, while that between the same
objective or measured characteristics may be almost zero and that between
observed and measured characters is likewise negligible (Brunswik, 1956). 91
If subjects are asked to mentally assess others on the basis of certain char-
acteristics, the pattern of correlations between traits is quite different from
that reported on the basis of direct observation of others (Pryor, 1986;
Shweder & D'Andrade, 1980). In categorizing others, people apparently
make use of person schemas that describe the connections between traits. It
would take very little steering of the categorization process by filmic means
to produce a significant halo-effect (Hamilton & Rose, 1980).

Automatic Categorization
At the most elementary level of the encoding of the stimulus, persons are
automatically recognized by means of a process that Bruner (1957) called
"primitive categorization." It may be assumed that even at this early stage of
processing "intentionality" is attributed to the stimulus, thus introducing a
distinction between living and inanimate objects. Such basic characteristics
as sex and age are also identified. It is more difficult, however, to establish to
what extent traits that are important for affect are registered during this
stage. But it is logical to assume that the total appeal that an individual has
for the subject is not realized in a single instant. In feature films characters
are often "introduced," in the sense that a certain amount of time is devoted
to presenting them to the audience. The course of the plot also plays a vital
role in determining the appeal of a particular character: how she reacts to
events determines whether she arouses sympathy or antipathy among viewers.
At a very elementary level of processing, however, the visual image of a
film character is capable of producing direct appeal. This may be the case if
an unlearned emotional stimulus (Frijda, 1986) or innate releaser is pre-
sented. Facial and bodily characteristics that are adaptive for the survival of
the species or the attainment of a goal by the individual may be emphasized
or exaggerated. We are thinking here of such things as signs of sexual readi-
ness, dependence, or approachability (Berry & McArthur, 1986) or a variant
91
A recent survey of research into "illusory correlation" is to be found in Fiske and Taylor
(1991).
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 161

of approachability, such as cuteness, as studied by Brooks and Hochberg

(1960). These signs call up desire and protective action tendencies respec-
tively. The ecological approach to social perception (Baron, 1980) is of par-
ticular importance in explaining the elementary process of impression for-
mation. Here affordances are sought, which the individual observed offers for
a certain behavior and which are directly observable, without any significant
intervention of cognitive processes. Frijda (1993) discussed the importance
of the direct observation of affordances for the primary appraisal of situations
in the emotion process. An example of a social affordance is the impression
of infantile helplessness, which makes possible protective behavior and which
is immediately observable in such things as, say, the shape of the head (Al-
ley, 1983, 1986). According to Baron (1991), certain emotions, such as anger
and fear, are immediately evident in the movements of the model. He sug-
gests that in general the observation of personality dispositions is the same
thing as the direct observation of social affordances. This would mean that
the usual accounts of impression formation assume the existence of more—
and more complicated—cognitive operations than necessary.
According to Berscheid (1985), a great many studies attest to the role that
physical beauty plays in the attraction people feel for one another. The results
of these studies "are such that the physically attractive—across age, sex, race,
and all socioeconomic stations—receive numerous preferential social treat-
ments" (pp. 453-454). According to Berscheid, this may be due to the fact
that the observer tends to assume the presence of various other attractive
qualities behind the appealing outward appearance. Hoffner and Cantor
(1991) found support for this assumption during research that focused on
children and television. It is, however, probable that other dimensions also
play a role in determining the appeal of media personalities.
The physical attractiveness of a face is probably related to the degree to
which it reflects the average of a particular population. Light, Hollander, and
Kayra-Stuart (1981) found that faces described as attractive were also judged
to be more common and typical, whereas attractive and typical faces were
thought to look more like each other than like other faces. In a later study
Langlois and Roggman (1990) actually established that average faces are more
attractive. Using computer techniques, the researchers averaged photographs
of faces. Stimulus faces that were obtained by averaging a larger number of
photographs were judged to be more attractive than those obtained by aver-
aging a smaller number of photographs. Langlois and Roggman believe that
this is due to a closer association of the first group of stimuli with prototypi-
cal representations. They do, however, note that a random sample of portraits
of film stars might well be considered more attractive, even though they were
not necessarily prototypical. All in all, it is unclear to what extent the appeal
of sympathetic protagonists, portrayed in feature films by attractive young
men and women, is based on innate releasers, as opposed to familiarity or
162 CHAPTER 4

cultural stereotypes. 92 But there are other possibilities as well. Carello,

Grosofsky, Shaw, Pittenger, and Mark (1989) suggested that the assessment
of the physical attractiveness of a person's face should not be based on the
average of some reference group or other, but rather on the ideal for that par-
ticular individual. That ideal is related to a biological function, the chewing
function, for example. This hypothesis was confirmed by experiments in
which subjects were presented with illustrations of faces in which the shape
of the jaw deviated to a large extent, slightly, or not at all, from a functional
archetype.
It is interesting to note the complications that can be deliberately attached
to the display of signs, such as innate releasers and determinants of affor-
dances. The example of veiled or partially revealed sexual releasers has be-
come something of a cliche. There are other examples, such as the smallest
of the robots in Return of the Jedi (1983). It is cute as far as its shape is con-
cerned, but it appears to be made of a smooth, hard material. This lends it a
minimal degree of cuddliness, which is in sharp contrast to the appeal of its
shape. A somewhat different example is to be found in Le gendarme de Saint-
Tropez (1964), when Louis de Funes plants a passionate kiss on the white
helmet of a motorcycle policeman.
From the viewpoint of the film narrative, it is conceivable that innate re-
leasers are employed when there is a need to strengthen the typing of char-
acters. "Cards" are endearing by virtue of their shortcomings; "Ficelles" are
perhaps nondescript because their appearance and behavior lack innate re-
leasers. Protagonists are sexually attractive, while antagonists, in particular
bad guys, may be characterized by means of innate releasers of aversion and
fear. These include such things as a slight deformity, a rasping voice, a per-
petual expression of anger, or—less commonly, perhaps—a remote physical
resemblance to animals that generally call up a reaction of fear, such as rats,
snakes, and scorpions.
In short, we may assume that on an elementary level of impression for-
mation by the viewer, characterization makes use of innate releasers or stim-
uli that, through processes of conditioning, are related to innate releasers.
Such stimuli can be realized both by typecasting and by staging. In view of
the fact that on the elementary level of impression formation, perceptual and
cognitive processing is not susceptible to conscious steering, and proceeds
more or less automatically, the effects of such processes of characterization
are largely unconscious and often unavoidable.
92
We know that at the moment they are introduced, protagonists tend to be considered at-
tractive, although there need not be complete agreement among viewers on this point. But there
is another important question that is deserving of further study, namely, the extent to which the
appeal of famous actors for a natural audience at their first appearance in a particular film is
founded on their appearances in previous films (or even outside the film world), quite apart from
or in spite of any innate releasers.
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 163

There are many interesting avenues of research yet to be explored in this

area. Do films display innate releasers that do not immediately present them-
selves to the unsuspecting eye of the viewer? There may be dynamic charac-
teristics—even some of a higher order—in the physical movements, facial ex-
pressions, voice parameters, and paraverbal expression that have a more or
less automatic effect. Another interesting possibility is that feature films per-
manently control the viewer's classical learning processes in such a way that
the control by innate releasers is shifted to substitute stimuli (Hearst & Jenk-
ins, 1974; Suboski, 1990). For instance, a character with an unconditioned
scary face always hums a certain tune or repeats the same fragment of text.
This tune or text—heard separately from any image—then becomes scary,
even though in itself it is a pleasing or appealing sound, such as a nursery
rhyme. In general, we can say that the stylization of characters does not nec-
essarily lower their affect potential; the loss of direct releasers of affect is of-
ten compensated for by learning processes.
Readers interested in the research into the relationship between the cate-
gorization of characters and other, immediately observable, sources of infor-
mation, such as voice characteristics, dress, and behavior, are referred to the
survey by Hoffner and Cantor (1991). The previously mentioned hypotheses
on the categorization of characters on the elementary level show that it is
quite possible to derive affective investment and return from a process that
is focused primarily on comprehension of the narrative. The only thing nec-
essary for the operation of the innate releasers, for example, is that the viewer
wants to know what is going on in the fictional world. The viewer's efforts to
understand the action taking place in the film will lead almost automatically
to awareness of the details that are made salient by the narrative, details that
often have a considerable affect potential.

Categorization on Less Elementary Levels of Processing

Processes of impression formation, which follow logically upon automatic
encoding, lead to a cognitive representation of a character that is of varying
complexity and degree of integration. Several models, all offering various lev-
els of detail, have recently been proposed for impression formation of other
persons (Andersen & Klatzky, 1987; Brewer, 1988; Fiske & Neuberg, 1990;
Wyer & Srull, 1986). Fiske and Neuberg (1990) suggested that in the cate-
gorization of individuals there is a whole range of processes, ranging from
rough categorization to "piecemeal integration." If the results on a given level
prove unsatisfactory, recategorization takes place on the following, more de-
tailed, level. This pragmatic quality of the categorization process (i.e., pro-
gressive refinement) is in accordance with the possibilities offered by the
characterization of the protagonists in the film narrative.
The dual-process model proposed by Brewer (1988) is more suited to the
164 CHAPTER 4

typologies familiar from literary analysis. Moreover, it is based on a large body

of empirical research and fits the conclusions of an extensive survey of the
experimental literature carried out by Wyer and Gordon (1984). According
to Brewer, the coarse categorization of individuals runs "top-down," that is,
it is category-driven and leads to the instantiation of a type representation.
The latter is pictoliteral, that is, it is coded in an analogous, visual format.
Type categories are hierarchical combinations of roles, features, and behav-
ioral characteristics: "serious professionals with I-dare-you-to-challenge-my-
opinion attitudes," as one subject put it, or "Barbara Walters types, gossipers,
nosey, yet sly and slightly sluggish" (Brewer, 1988, p. 17). In this sense, they
correspond to some extent to the flat characters of Forster (1927). Through
individuation, subtypes then arise in which the information deviates some-
what from the instanced type.
In the type of feature films where the action plays a fairly significant role
in the narrative, as compared to character psychology, the cognitive invest-
ment related to character categorizations may consist mainly of trying out
and discarding increasingly specific individuation hypotheses. This occurs af-
ter a particular character has been introduced and a rapid categorization
made on the basis of pattern recognition and innate releasers. This first cate-
gorization displays a strong primacy effect (Fiske & Taylor, 1991). The return
of ongoing categorization is formed by improvements in the fit and a more
complex representation of the character. One of the characteristics of the pro-
cess of impression formation may well be that in the processing of the new
features—derived from the actions of the characters—which are continually
being offered by the film narrative, the main objective is confirmation. Minor
inconsistencies in the various behaviors go unnoticed, as long as the impres-
sion formation is driven by the anticipation of more global features, as is de-
cidedly the case during the watching of feature films. Major inconsistencies,
however, do give rise to revisions in the direction of a subtype.
Another method of processing is what Brewer referred to as personaliza-
tion. This runs "bottom-up," in other words, it is steered by the input and re-
sults in a propositional representation focusing on a certain individual, with
roles and features subordinated. Specific behavior is first stored as concrete
information, and only later integrated—by means of combination and infer-
ence—to form features. Here the cognitive investment consists in keeping
available a large number of particular instances and the relatively laborious
process of integration, given the lack of support afforded by a prototypical
category. The return would then consist of a rich and complex structure that
may be relatively new.
In current film theory there is consensus over the high degree of typing of
the protagonists in traditional feature films. The characters who appear in
classical cinema are not actually individuals, but can best be described as a
collection of traits that are required to realize the prototypical causality of the
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 165

action (Bordwell & Thompson, 1986; Ghatman, 1978). Because there are no
comparative studies dealing with the agreement between the categorization
of individuals, on the one hand, and fictional characters, on the other hand,
we can only surmise—by way of provisional hypothesis—that the categoriza-
tion found in traditional feature films depends largely on types, although
these are to some extent individualized. The role of personalization is prob-
ably greatest in the more sophisticated psychological-realistic genres.
A supplement to the dual-process model is to be found in Andersen and
Klatzky (1987). These researchers make an interesting differentiation in the
concept that Brewer describes as "typing." Trait typing makes use of cate-
gories such as outgoing, socially skilled, friendly, nutty, power-loving, self-con-
fident, knowledgeable, and intelligent. Social stereotyping, on the other
hand, is more vivid and concrete. The categories are also richer in attributes;
according to Andersen and Klatzky, the most important of these extra at-
tributes of social stereotypes are typical behavior and reactions to certain
events, characteristic intentions, and goals of the type. It is conceivable that
they fulfill exactly the same functions as the thematic action structures de-
scribed in chapter 5. In other words, they function as a rich source of expecta-
tions and predictions concerning events and actions. Social stereotypes meet
a more specific collection of constraints than trait types. Once the action has
gotten under way, detailed schemas of characters must be generated that co-
incide with the results that have already become clear from the action of the
plot. A social stereotype then functions as a goal-derived ad hoc category,
which makes more specific the original rough taxonomic category, as postu-
lated by Barsalou (1991; see also chap. 5). The examples suggested by An-
dersen and Klatzky include mafioso, clown/comedian, politician/diplomat,
bully/gang member, brain/genius, depressed/suicidal, wise man/guru, Ron-
ald Reagan, Woody Allen, and Ghandi. These are all characterizations that,
when activated, are capable of evoking quite specific expectations concern-
ing the further course of events. They do so on their own, and most certainly
in combination with thematic action structures. The mafioso is about to pull
a fast one on someone, Woody Allen will have a hard time coping, the brain/
genius will undoubtedly invent some ingenious device and become involved
in a comical scene highlighting his relationship with women.
The amendment suggested by Andersen and Klatzky is important to an un-
derstanding of how we form our impressions of film characters. In the first
place, it helps to explain the wealth of possible predictions offered by the film
narrative. The film stimulus unfolds quite gradually, and initially the viewer
may tend toward trait characterization and social stereotyping. Global pre-
dictions are then succeeded by more precise ones. In the second place, trait
characterization—in particular, stereotyping—may go a long way toward ex-
plaining why one character is seen as real and the other is not. It will be clear
from the examples that stereotypes can refer to both fictional characters and
166 CHAPTER 4

real persons; a combination of the two is also possible, as in the case of Woody
Allen the actor and the film character. The fact that stereotypes make no clear
distinction between fictional and real-life characters—a quality that might be
referred to as "archetyping"—means that they can render a fictional element,
in this case a character, believable or, conversely, contribute to the fictional-
ization of a real-life stimulus. In the latter case, for instance, an intimidating
male person can be dubbed a "Boris Karloff," in an effort to make it easier
to deal with him; in the former, one might try to see grotesque fictional fig-
ures in a sympathetic, compassionate, or even tragic light. Repulsive figures
are often seen as "sheep in wolf's clothing": The Hunchback of Notre Dame
(1939), The Elephant Man (1980), The Phantom of the Opera (1925), and
Beauty and the Beast (1991). The more specific the expectations created by
such a characterization, especially if they are borne out, the more natural the
fictional figure becomes.
The models of impression formation described above all have in common
that the mode or level of categorization of the stimulus person depends on
the aim of the observer. In other words, there are pragmatic limitations to
person perception. It is safe to assume that as a rule the observer will do no
more than is absolutely necessary—Taylor (1981) referred to "the cognitive
miser"—and we are assuming that revision of an established representation
will be avoided. Nevertheless, during the watching of feature films the "pro-
cessing depth" of characters is dictated to a considerable extent by the nar-
rative, 93 including in some cases the revision of established impressions. The
course of the story determines how detailed the categorization will be and in-
volves working out expectations concerning certain traits of the character.
The narrative ensures that the necessity to revise a seemingly completed cate-
gorization comes to some degree as a surprise. There are countless, largely
sentimental, films in which the plot revolves around a metamorphosis of the
protagonist or the relatively late revelation of the true nature of a character.
A few arbitrary examples will illustrate this point. The horrid boarding-school
headmaster Blanchard (whose nickname was Merleusse or "hake") proves to
be a kind and generous man who distributes Christmas presents to the boys
who have to stay at school over the holidays (Merleusse, 1935). At the be-
ginning of Ceiling Zero (1935) James Cagney, as the test pilot Dizzy David-
son, is portrayed as an unscrupulous ladies' man who connives to get another
pilot sent on a dangerous test flight; in the end, he gives up his girlfriend to
the pilot who really loves her and takes over the other's dangerous flight, dur-
ing which he is killed. A less sentimental variant is the gradual revelation of
unsuspected aspects of the character of Hannibal the Cannibal, the beast in

93
The term depth of processing is borrowed from the work of Craik (Craik & Loekhart, 1972;
Craik & Tulving, 1975) and Anderson (Anderson, 1976, 1983), concerning encoding strategies
and their effect on memory. Stimuli can be subjected to a shallow or a deeper elaboration.
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 167

The Silence of the Lambs (1991). However, it is not only the depth of the cat-
egorization that is determined by the film narrative. The story places all sorts
of restrictions on the fleshing out of traits and types. For example, we know
that there is a certain relation between the role that a fictional character plays
in relation to other characters and the former's psychological characteriza-
tion. How could a Ficelle be a dominant personality? Gould we ever visualize
the adversary of our hero as a Ghandi rather than the usual mafioso or bully
type? And the fact that as the narrative progresses Hannibal the Cannibal is
shown to possess a strange kind of sensitivity is bound up with the realization
that his opponents—the police and the governor—are being increasingly por-
trayed as stupid and corrupt.

Impression Formation, Sympathy, and

the Recognition of Concerns
The theory of impression formation not only makes it easier to describe how
viewers comprehend protagonists but also helps to explain the affective in-
vestment and return that are involved in the watching of a feature film. As
Aumont et al. (1983) say, characterization automatically leads to identifica-
tion. Stereotypical categories are not, in the end, affect-neutral. In the first
place, impression formation may be determined in part by self-categorization
needs. The theory of social identity (Tajfel, 1982), for example, holds that
stereotyping of others is a response to the need of individuals to derive their
own identity from the group to which they belong (in terms of sex, profession,
religion, etc.), and to see one's own group (the in-group) as more favorable
than another (the out-group). Older studies have shown that a certain kind
of stereotype—prejudice—springs from people's desire to see themselves in
a favorable light by contrasting themselves with one or more groups (Adorno,
Frenkel-Brunswik, Levinson, & Sanford, 1950). These self-categorization
needs make it possible to arrange psychological stereotypes on a positive-
negative affect or evaluation dimension, on the basis of in-groups and out-
groups. Such groups are culturally determined; complicated social dynamics
ordain that, in the eyes of the dominant group or the majority of other groups,
a particular group is out. It need not surprise us that research into the im-
pression formation of actual people shows that this process converges in quite
an early stage in a general assessment as either good or bad, around which
other traits are then arranged (Srull & Wyer, 1989; Wyer & Gordon, 1984).
According to Srull and Wyer (1989) this general assessment is made as soon
as sufficient behavioral information has been collected to make possible an
affective categorization. Any revision of a categorization is usually prompted
by new information that conflicts with the established likeable-dislikeable as-
sessment, rather than by descriptive information.
D'Andrade (1984) showed that the terminology that is culturally estab-
168 CHAPTER 4

lished and commonly used by lay persons—as opposed to professional psy-

chologists, at least hopefully—to describe character includes a highly evalu-
ative component. This led him to state:
It should be remembered that the natural language of character terms was cre-
ated by and for ordinary people, and that one of the major needs of ordinary
people is to determine how to react to what other people do—to approach or
avoid, to reward or punish, to imitate or correct. The natural language system
of character terms seems well suited to these purposes. What is interesting is
that we have developed so many ways of being good and bad. (D'Andrade, 1984,
p. 332)

In a more general sense, social schemas are capable of evoking affect di-
rectly. When someone fits into an existing person schema, then not only the
meanings that are inherent in the schema are invoked but also the affect that
has been attached to the schema by previous experience. Schema-triggered
affect has been demonstrated using a variety of person schemas, such as the
members of a certain profession (doctor, hotel maid, loan shark), or individ-
uals who are culturally stigmatized, such as schizophrenics (Fiske & Taylor,
1991). The fact that characters are so easy to typify and that types, in par-
ticular stereotypes, are in or out lends characters an affective charge. The
more the narrative gives rise to the instantiation of types, especially stereo-
types, the more the appeal of characters can be directly traced to the process
of impression formation. Typical intentions (good or bad) and the ensuing
character goals and plans are among the attributes of stereotypical social cat-
egories, and these trigger affect in the viewer.
The appeal of stereotypical characters is presumably related to the inten-
tions ascribed to them: good guys are sympathetic, bad guys are not (Leifer
& Robers, 1972, in Hoffner & Cantor, 1991), and deliberate bad behavior on
the part of a character results in a negative evaluation (Berndt & Berndt,
1975, in Hoffner & Cantor, 1975). Motives or intentions would appear to
carry more weight than consequences (Hoffner & Cantor, 1991). When we
say that good guys have good intentions, we mean t h a t the concerns that they
have with respect to the state of affairs in the fictional world corresponds to
more or less universal source concerns shared by t h e natural audience of the
traditional feature film. And, as we know, perceived similarity leads to at-
traction. Similarity between one's own attitudes and views on the important
issues of life—in other words, value concerns (Frijda, 1986)—and the views
of others more in particular creates a bond between people. There is consid-
erable empirical support for the positive relation between observed attitude
similarities and attraction. 94 Above all, correspondence between the charac-

94
See, for example, Seeord, Baekman, and Eaehus (1964). For surveys of research into the
relation between attraction and observed similarity of attitudes, see also Berscheid (1985, pp.
455-457). One explanation for the powerful influence of this effect may be the operation of a
reverse influence: attraction leads to a heightened awareness of shared attitudes (see Granberg
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 169

ter's views on justice to those of the viewer presumably determines the lat-
ter's sympathy for that character. The categorization of characters in terms
of good and bad is by no means restricted to popular and stereotypical films,
as one might expect. Among readers and viewers of contemporary drama
(Bordewijk-Knotter, 1988) and political drama intended for a culturally elite
audience (Tan & Schoenmakers, 1984), there are striking examples of judg-
ments based on those very same moral categories: good and evil intentions.
The categorization of characters on the basis of recognized concerns has
a number of consequences. First, there is the heightened relevance for the
viewer of the vicissitudes of the characters. Both good guys and bad guys har-
bor intentions and aims that touch the concerns of the viewer. An obvious ex-
ample is defense against threats to law and order or to sympathetic charac-
ters. This relevance of the character's fate is in itself an important condition
for the emergence of empathetic emotions. And second, viewers are inclined
to subscribe to the goals of the good guy and to reject those of the bad guy.
(Albritton & Gerrig, 1991; Zillmann & Bryant, 1975a; Zillmann & Cantor,
1977). Because intentions and goals are so crucial to the process of catego-
rization, there is presumably a pronounced halo effect attached to them; even
impression formation consists to a not inconsiderable degree in establishing
consistency between the good or evil intentions observed and other, more su-
perficial traits. When Humphrey Bogart suddenly develops a tic and begins
to pull on his ear lobe (Casablanca, 1942), this is charming and the repeti-
tion of this mannerism only makes him more attractive. The same might be
said of the chest-beating of King Kong. The fact that this Beast has the best
interests of his Beauty at heart, and even those of all people of good will,
serves to lend a human and pleasant quality to his chest beating and other
animal features, notably in the long run (King Kong, 1933). Conversely, once
we are aware of the intentions of the villain, his (more often than her) refined
manners suddenly become decadent, and there is an ominous significance to
the way he sits there, stroking the cat on his lap.
The revelation of underlying intentions of characters is a component of
the narrative process and the previously mentioned halo effects can be ma-
nipulated by the filmmaker in a variety of ways. In general, we may assume
that the Complication, which theoretically is embedded in the thematic ac-
tion of all traditional feature films, is balanced by the development of a char-
acter. In fact, the development of the character may even replace the the-
matic action structure in the sense that it creates expectations and shapes
them (chap. 5). 95
An early and firmly planted primacy effect consisting mainly of negative

& King, 1980; Schoedel, Frederickson, & Knight, 1975). Berscheid (1985, p. 456) warned
against an overly simplistic interpretation of the links between observed attitude similarities and
attraction.
95
Development is also seen by scriptwriters as a desirable trait of dramatic characters. The
seeds of growth must be planted at an early stage of the drama (Egri, 1960).
 170 CHAPTER 4

affect brought about by the instantiation of a negative type—through innate

releasers, say—can be broken off suddenly or in gradual stages, preceded by
subtle clues. Something like this occurs in A Star is Born (1954), where the
somewhat mysterious figure of Norman Maine is introduced, a disagreeable
lush with a creepy manner. Individuation—and perhaps a degree of person-
alization as well—leads to the integration of a number of traits that cohere
around an ultimately wholly good intention. Then there is the Elephant Man,
whose hideous appearance is only revealed to us after several carefully con-
structed scenes in which we are only allowed a view of a mysterious creature,
whose face is heavily shrouded. The viewer is then gradually encouraged to
conquer aversion and to discover, with Dr. Treves, the intentionality, sensi-
tivity, and intelligence of the monster (The Elephant Man, 1980). One comes
to feel compassion and ultimately admiration for a man who has endured so
much at the hands of people who look normal but behave like monsters. A
comparable mechanism operates in many whodunnits of the Le Corbeau
(1943), Death on the Nile (1978) and Murder on the Orient Express (1974)
genre; type categorizations display a dynamics that continually shifts suspi-
cion from one character to the other: Here's the kind of person who must
have done it. 96 In retrospect, of course, there must be at least a few features
discernible in the introduction of the character that are compatible with the
ultimate categorization. Examples of a total reversal of affect are rare.
Kunda, Miller, and Claire (1990) recently demonstrated that presenting
fictional characters with conflicting social stereotypical characteristics auto-
matically leads to a great deal of causal reasoning and the construction of
causal narratives. This, in turn, gives rise to the formation of new attributes,
that is, traits that were not part of the original stimulus. Examples of con-
flicting combination stimuli include the blind marathon runner, the leftist
businessman, and the carpenter graduated from Harvard. The more unusual
the combination of characteristics, the more likely they are to trigger such
reasoning. Traditional film narratives, too, often present combinations of
more or less conflicting clues to the nature of a particular character. This
leads to the construction of causal assumptions that can be used to prefigure
the expected revelations.
Other investments are rooted in the conflict between characters and
stereotypes. Cultural sex stereotypes portray women as virtuous and docile.
Pussy Galore, assistant to Goldfinger (Goldfinger; 1964), is the very antithesis
of that image. She is on the side of the baddies, presides over a squadron of
lady pilots in true military fashion, and fails to evince any interest in James
Bond, whom she floors with sheer physical force. The viewer hopes she will

96
There are, of course, interesting variations on this process. In Le Corbeau, for example, the
ultimate conclusion is that anyone could have done it, while in Murder on the Orient Express,
all suspects have done it.
CHARACTER STRUCTURES, EMPATHY, A N D INTEREST 171

side with 007, and ultimately she does; the point is that initially almost the
only grounds for such hope are provided by the stereotypical view of women.
In the same way, deep down Ninotchka (1939) can never be the real com-
munist woman who refuses to allow herself to be seduced by a man. Again,
the investment is a form of hope, based on a stereotype under pressure.
If it is empirically realistic to apply impression formation theory to the cat-
egorization of film personages—and this would have to be established experi-
mentally—then it is easy to see how empathetic emotions originate. Catego-
rization and individuation of characters clearly affect the emotional meaning
for the viewer of situations involving that character. For one thing, the rele-
vance component of the situational meaning structure is heightened, because
the fate of the characters in the categories "hero" and "villain" calls upon the
sympathy and value concerns of the viewer. The first result of categorization
is that the viewer recognizes the most important objectives and concerns of
the characters and either endorses or rejects them. Subsequently, events be-
falling the characters evoke emotion in the viewer because they are mean-
ingful for these characters in view of their concerns. The misfortunes of the
villain are welcome, those of the hero unwelcome, in that they are either in
accordance with or run counter to the wishes of the characters. The events
and outcomes of the plot may be seen as more or less desirable from the
standpoint of both the character and the viewer, to borrow a term from
Ortony, Glore, and Collins (1988).
A second result of the categorization of characters is enhanced reality. The
more depth there is to the individuation of a character, that is, the sharper
the differentiation in terms of subtypes, the more real the character is and
the higher the reality parameter of the situational meaning structure.

EMPATHY

All F emotions have as their object an element of the fictional world, as we

saw in chapter 3. An F emotion is the reaction of viewers who imagine them-
selves to be present in that fictional world, as invisible observers of the events
taking place there. Some F emotions are determined by a fictional event that
is an emotional stimulus in itself. Others—the majority—are determined by
the significance of the event in question for a particular character. The first
type of emotion is nonempathetic, the second empathetic.
An empathetic emotion is characterized by the valence of the events in re-
lation to the concerns of the protagonist. In the first place, misfortune for the
character produces a negative emotion in the viewer, and good fortune a pos-
itive emotion.
In principle, however, not only the valence of events in terms of character
concerns, but all components of the significance of the situation for the char-
172 CHAPTER 4

acter help to shape the situational meaning structure of that situation for the
viewer. This understanding of the significance of the situation for a charac-
ter, in its broadest sense, is known as empathy.97
Definition 6.1. By empathy we mean all the cognitive operations on
the part of the viewer that lead to a more complete understanding of
the situational meaning for the character.
An understanding of conventional narratives such as that of the traditional
film requires empathy in the sense of definition 6.1. Such narratives consist
of situations that are of significance for the protagonist, and a significance
that is determined by his or her intentions and psychological traits. There is
some experimental evidence for the idea that the processing of narratives is
accompanied by a representation associated with the emotion the protago-
nist is understood to have. As a rule, the emotions of the protagonist are quite
clear to the reader (Gernsbacher, Goldsmith, & Robertson, 1992). Compre-
hension of a narrative includes a representation of the aims of the protago-
nist (Long, Golding, Graesser, & Clark, 1990), as well as a presentation of
spaces and objects that are of importance for those aims (Morrow, Bower, &
Greenspan, 1989; Morrow, Greenspan, & Bower, 1987). Moreover, in the
study by Morrow et al. (1987) it was demonstrated that the location that was
part of the situational meaning structure for the protagonist was more promi-
nent in the eyes of readers than the actual location in which the protagonist
found himself or herself. And experiments by Suh and Trabassso (1993)
indicated that the as yet unrealized goals of the protagonist are being repre-
sented in the memory of readers. As we surmised in chapter 4, such an un-
realized goal may remain latently present in the background, only to be actu-
alized as expectation in a later scene (Huitema, Dopkins, Klin, & Myers,
1993). 98
Research results such as these are not, in fact, particularly surprising. The
traditional (film) narrative is centered on the actions of a human protagonist
to whom intentionality is ascribed, whose concerns are recognized, and who
displays emotions in response to events, as Bordwell (1985; 1989b) has aptly
pointed out. Situations are seen by viewers as relevant to the concerns of the
protagonist: they comprise outcomes with a valence in terms of character
concerns or desirabilities and as yet unrealized outcomes." There are several

97
This deviates from what Aumont et al. say about the situation. They do not see it as a rep-
resentation of the significance for a character.
98
It must be said that inferences that go further than what is actually given in the narrative
are somewhat controversial (McKoon & Ratcliff, 1992).
"Intentionality is not reserved exclusively for human characters. Anthropomorphic robots
(HAL in 2001: A Space Odyssey), worlds (the planets in Solaris, 1972, and of course animals
(The Birds, 1963) have all been seen as beings with concerns that color the viewer's under-
standing of situations and determine the action of the narrative.
CHARACTER STRUCTURES, EMPATHY, A N D INTEREST 173

such examples in the analysis of Punishment in chapter 5. All the pluses in

the plot unit configuration (Fig. 5.2) represent valences either for Marjan or
for Father; in the description of the processing of the story, mention was
made of several opportunities for the viewer to project desirabilities. Oatley
(1992, 1995) convincingly argued that identification and reader emotion
arise from understanding the characters' plans and their interactions. The
reader adopts the protagonist's goals, and the narrative steers planning-and-
understanding systems that are used in planning routines in the reality of
daily life as well. According to Oatley, narrative may also induce more com-
plex emotions, in which the readers are invited to examine their own plan-
ning and the thread that joins an individual's plans, the self.
Insight into the significance of the action for a particular character de-
pends in part on the manner in which the character is being represented and
understood. Broadly outlined typing does not allow for a great deal of detail
in the situational meaning structure for the character that the viewer con-
structs; more refined categorizations enable the viewer to draw up a differ-
entiated assessment of the significance of the situation. Through sub typing
and individuation, increasingly far-reaching and specific aspects of the situ-
ation acquire significance for the character. The characters in quality films
are round, in the sense that they display more emotion, and that emotion is
more developed than that of the heroes in the popular genres. As a result, the
emotion of the viewer is more empathetic, displaying more gradation and quali-
tative nuances. We now examine three examples of increasing individuation,
in this case, of piano-playing characters. In Forman's film Amadeus (1984)
Mozart is portrayed as an adolescent imp. His feats at the piano that astound
his lords and masters, mean triumphs to him, and this in itself is a source of
simple pleasure to the audience. Madame Sousatzka (1988) is a somewhat
more complex case. She is enchanted by the brilliant playing of her pupil,
and this, too, pleases the audience. At the same time, however, it confronts
her with her own failure as a concert pianist, which fills us with sadness. An-
other dimension of the situation is the fact that she is a gifted teacher. Our
admiration is empathetic, as it depends on our understanding of her pride
and pain. The musical achievements of Ada, the heroine of The Piano (1993),
mean far more to her than a display of prowess. Her music is a substitute for
speech, and thus expression; it means liberation from a claustrophobic life
with a man she does not love and imprisonment on an island. It is also a di-
rect expression of her longing for the man she does love and a great many
other things as well. The viewer is aware of all these various meanings. Empa-
thetic emotions in response are pity and admiration, and the viewer's own ob-
servations will be added to all that Ada is experiencing; the perspective is by
no means limited to that of the character Ada. She does not play the coquette,
and the viewer finds this appealing; the latter observation is empathetic, for
feelings of sympathy on the part of the viewer cannot be aroused unless it is
 174 CHAPTER 4

understood that she is not trying to get into anyone's good graces. The view-
ers may also wonder about a particular significance that her playing has for
her, one that apparently completely escapes them. It seems that there is more
to Ada than we can grasp: a mystery, leading to—nonempathetic—amaze-
ment. Some viewers may be fascinated by her elusiveness, which is often
wounding to other characters. She baffles others, who do not understand
what has taken hold of her; given the period in which the story is set, they in-
evitably suspect some form of diabolic intervention.

Empathetic Emotion
The viewer's understanding of the emotional meaning of a situation for a par-
ticular character is not always complete. Nor does it in all cases constitute the
complete situational meaning. Perhaps we could say that an emotion is em-
pathetic to the extent that the significance that an event has for the charac-
ter is part of the situational meaning structure for the viewer.
Definition 6.2. By an empathetic emotion we mean an emotion which
is characterized by the fact that the situational meaning structure of the
situation for a character is part of the meaning for the viewer.
It should be noted that according to this definition the emotion that a situ-
ation evokes in a character does not necessarily coincide with the empathetic
emotion of the viewer. Such agreement is often posited by researchers as a
precondition for the use of the term empathy. Zillmann (1991b), for exam-
ple, sees hedonic compatibility as necessary for empathy. In the definition
presented here, however, a central role is reserved for the understanding of
the meaning, whatever it may be. The situations portrayed in feature films
are significant not only for the protagonist but also for other characters. Like
the protagonist, they are understood to have certain concerns that are specifi-
cally addressed by events. Because the viewer cannot share the concerns of
all the characters—by the very nature of the narrative, these are mutually con-
tradictory—the empathetic emotions felt by the viewer often do not corre-
spond to the feelings of any particular character. And yet these emotions may
be seen as empathetic, because their quality is determined by the viewer's
understanding of the situational meaning for the character.
The emotion supposedly felt by the antagonist contrasts most clearly with
the empathetic emotion of the viewer. When the antagonist coolly deals the
hero a blow, this calls up anger on the part of the viewer—empathetic anger—
because it is clear that the blow was deliberate and that the antagonist is en-
joying her success. Conversely, schadenfreude is evoked when the bad guy
has to take his knocks. Both the negative evaluation of the situation by the
antagonist and the resulting negative emotion are part of the situational
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 175

meaning structure for the viewer. In the literature the term contrast empa-
thy is often used to refer to the opposite of empathy (Stotland, Mathews, Sher-
man, Hansson, & Richardson, 1978). However, this pair of terms suggests
that a comparison of the meaning of the situation for character and viewer
necessarily leads to either agreement or contrast. This is not so, for all the
various roles that other characters play in the course of events have their own
situational meaning, a meaning that is in a constantly changing relationship
to the concerns of the protagonist.
Definition 6.2 calls up the question of whether it is even possible for viewer
emotions to be nonempathetic. Although non-empathetic emotion plays no
major role in the traditional cinema, it is specific to the medium of film and
does regularly occur. Traditional films offer not only a dramatic plot and a
psychological development of character but also spectacle. In chapter 2 we
examined the motivation theories that stress the simple pleasure of watching
as an independent motive for viewing films. There are also traditional genres
in which watching fictional events is the prime attraction, almost divorced
from their significance for the characters. These are known as spectacle films
and include such genres as science fiction, historical extravaganzas with nu-
merous crowd scenes a la Spartacus, (1960), disaster films, costume dramas,
and fantasy films. In the same way, the major scenes in erotic and horror films
are more than anything else a feast for the eyes, although perhaps not to
everyone's taste. Such films are clearly extremely popular, but even in gen-
res that enjoy higher prestige the spectacle of fictional events may be more
important than the significance and emotional value of those events for the
protagonist. In song-and-dance films, plot and character development are per-
haps too often only an excuse for the presentation of spectacular revue num-
bers. In quality films where the dramatic events are situated in an exotic lo-
cation—such as Out of Africa (1985), for example—the invisible witness is
often detained longer than necessary or treated to a detailed view of the
sights in landscapes or city scenes where nothing very much is going on that
affects the concerns of the protagonists. Thus emotions such as enjoyment,
excitement, horror, fear, and longing are sometimes nonempathetic, to the
degree that the focus of the situational meaning is limited to the event itself,
as a scene. Nonempathetic interest might best be described as fascination:
one is caught up in the spectacle; here the promise is represented by the con-
tinuing or intensified enjoyment of the spectacle, rather than the prospect of
increased understanding or any improvement in the chances of the protago-
nist; the action tendency is an urge to go on watching. Nonempathetic emo-
tions are not associated with the implications for the protagonist, but rather
with the actual witnessing of the event. The nonempathetic fear occasioned
by The Texas Chain Saw Massacre (1974) is fear evoked by the anticipation
of the sight of horrific scenes; the nonempathetic enjoyment evoked by
176 CHAPTER 4

Down-by-Law (1986) is rooted in the sight of magnificent street scenes;

while the nonempathetic longing in Vertigo (1958) is the (male) desire to see
Judy again.100
Nonempathetic emotion are still F emotions. The fictional world is the stim-
ulus, not the artefact. A distinction can be made between the spectacle of the
dinosaurs in Jurassic Park (1993)—what we see of them in the fictional
world of the park—and the staging of that spectacle. The former evokes wit-
ness emotions, at the very least fascination and possibly surprise, whereas
the latter, when noticed by the viewer, calls up aesthetic A emotions, such as
wonder—how on earth did they manage to make the brachiosaurs so lifelike—
and admiration. In the same way, the viewer's desire to see Judy may be dis-
tinguished from his desire to see Kim Novak.
And yet the events depicted by the traditional feature film always have a
certain significance for the protagonist, which is why empathetic emotion in
its absolute sense is dominant. The effect that events have on an intentional
protagonist are the source of the causality that lends coherence to the tradi-
tional cinema plot. The examples mentioned earlier—films in which specta-
cle plays an important role—are all limited in the sense that the spectacle has
at least some minimal significance for a protagonist. To take a cliche exam-
ple, hard porno films show explicit images of sexual intercourse. Everything
revolves around these images, and the entire development of the film is gov-
erned by the dosage of shots that include intimate details of couples in the
act of copulation. The fact that displayed bodies belong to characters with
their own thoughts and feelings is of secondary importance. In many cases,
there is not even a suggestion of characterization. Still there is always the
skeleton of a narrative that lends these scenes a modicum of significance for
the characters. Perhaps they are enjoying the secrecy of their affair or have
been forced to wait an inordinately long time for the meeting. In short, there
is in every event—even the explicit portrayal of sex—a complication that ac-
tivates concerns and strivings and thus lends some significance to events for
the characters. In other genres, by contrast, the explicitation of the meaning
of events for the characters enjoys primacy, as in the case of the melodrama
and the more sophisticated psychological drama. As a rule, however, specta-
cle and psychological developments are combined with a defter hand and in
a more balanced fashion. The canonic structure of the narrative of the tradi-

100
The case of Vertigo is interesting because it is an example of empathetic and nonempa-
thetic emotion going hand in hand. Empathetic emotion is evoked in the viewer by Scottie Fer-
guson's desire to find Judy. At the same time, a nonempathetic desire to see her is evoked, es-
pecially in male viewers. Crucial to this twofold desire is the shot where Scottie first catches sight
of Judy in a restaurant. She is clearly visible, and shown at her most beautiful. It is almost im-
possible for the male viewer not to immediately fall in love with her, and he has no trouble un-
derstanding why the same thing has happened to Scottie. The almost subjective perspective of
the shot in question would appear to promote both desires.
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 177

tional cinema as a whole (chap. 3) implies that the protagonist is always char-
acterized by recognizable concerns and that each event is of importance to
the protagonist. At the same time, the film fulfills a need to examine issues—
and above all, people—at one's leisure.

The Major Empathetic Emotions

Given the fact that the film narrative is about overcoming Complications, sym-
pathy is, alongside interest, the most important sensation evoked by the tra-
ditional feature film. There are, however, two forms that can be distinguished.
Until now we have talked about sympathy or attraction, which is seen as a dis-
position with an affective aspect, based on a sympathy source concern. It is
possible to have a soft spot for the sympathetic protagonist quite apart from
the actual action of the film. It is for this reason that we speak of disposition.
This will arise somewhere near the beginning of the film and gradually in-
tensify, as the good intentions of sympathetic characters are always sus-
tained, in spite of all the obstacles placed in their way.101 The disposition is,
in other words, tonic. On the other hand, sympathy as emotion—during the
watching of feature films, but probably at other times as well—is phasic.
There are moments when the otherwise strong protagonist makes special sac-
rifices in a good cause, acknowledges vulnerability, and yet carries on, even
though no one is forcing him or her to do so. The meaning of the situation
for the protagonist is that knowingly and willingly a risk is taken on behalf of
the Good Cause or another sympathetic character. At moments such as these,
sympathy functions as a phasic emotion. One stereotypical example, which
appears to be endowed with eternal life, is the manly hero who, with barely
suppressed irritation, is obliged to accept as a companion on his rugged ex-
pedition a stereotypical female figure, helpless, bothersome, but totally de-
pendent on his help and protection (The African Queen, 1951; Indiana
Jones and the Temple of Doom, 1984; Torn Curtain, 1966; and others). A
crisis arises, during which he has an opportunity or a good reason to leave
the woman behind. He does not do so, which occasions a wave of sympathy.
Feature films are full of moments when even a minor incident triggers a simi-
101
An exception is no doubt the story in which the protagonist repents, after starting out as
the bad guy. This is a theme that has been featured in many a film. It is activated early on in the
film, creating the expectation and desire for the protagonist to turn over a new leaf. The Com-
plication in this case consists in the struggle between good and evil within the protagonist him-
self or herself, whereby both "sides" find helpers among the characters around him or her (A
Place in the Sun, 1951, The Lost Weekend, 1945, Woman of the Year, 1942). The effect of a
theme like this and the influence on the sympathy of the viewer are far more subtle and prob-
lematic than I have sketched here. For example, there are melodramatic films in which the re-
pentance of the protagonist is accompanied by a tragic loss. The viewer hopes, with him, that
what is seen as socially unacceptable—the evil—can somehow be made acceptable. But in vain.
(See the study by de Kuyper, 1984, on impossible love.)
178 CHAPTER 4

lar surge of sympathy among the audience. The—again stereotypical—manly

protagonist is somewhat awkward and diffident toward his lady love, or he
stumbles and makes her laugh; at such moments the character becomes
charming or appealing. A sympathetic disposition is also a precondition for
empathetic enjoyment. When the protagonist experiences the satisfaction of
achieving his goal, the viewer who sympathizes with her will feel an empa-
thetic pleasure. In similar fashion, the disappointments of the protagonist
lead to empathetic feelings of disappointment among viewers.
The action tendency of sympathy may be an inclination to seek proximity
and intimacy, a sharing of thoughts and feelings, and a sense of cherishing
and being cherished. 102 Only proximity and intimacy can be realized.
Sympathy is a positive emotion, a pleasurable experience, as when we en-
tertain warm feelings for one of the characters. Such feelings are considered
one of the most important aspects of our relationships with family, friends,
and partners. They provide the satisfaction that flows from these psycholog-
ically quite elementary relationships. Feelings of sympathy can be measured
by asking subjects to rate their warm feeling on a numerical scale. Accord-
ing to validation studies, this warmth is a positive and fleeting—or phasic—
emotion, which is accompanied by physiological arousal and intensified by
the real or imaginary experience of a romantic, friendly, or family relation-
ship (Aaker, Stayman, & Hagerty, 1986). In a study by Aaker et al., empathetic
warm feelings were evoked by means of narrative television commercials. 103
Sympathy is characterized by equality and reciprocity. In sympathy the
subject tends to give and is ready to receive. As viewers we cannot really share
feelings with a protagonist, since there is no direct address, but very often
protagonists or major characters disclose to the viewer intimate feelings in
an oblique fashion. They communicate their deepest feelings to another
character, and at the same time the viewer shares these, due to the latter's
position of invisible witness in the fictional world. This effect no doubt
strengthens feelings of sympathy.
In a certain sense, sympathy could be situated between two other empa-

102
The hypotheses on the structure of the empathetic emotions are based in part on the in-
sights into pity and admiration in Frijda (1986) and research by Rombouts (1992) into friend-
ship and being in love. They are in accordance with the points made by Schoenmakers (1988)
in his discussion of types of identification derived from Jauss.
103
There are also experimental research results that suggest that feelings may be quite liter-
ally warm. According to Zajonc, Murphy, and Inglehart, "The metaphors hothead, boiling mad,
hot under the collar and cool as a cucumber are not altogether accidental, and the present study
is not the first to measure head temperature change in response to an emotional episode" (1989,
p. 409). It should be noted that these researchers are referring to emotions with a negative va-
lence, such as anger and aggression, which could be related to brain temperature: hotheaded,
for instance. In contrast, warm feelings, such as those associated with sympathy and attraction,
are said to be accompanied by a lower brain temperature. This does not mean that the temper-
ature in other parts of the body is not higher. Perhaps in the chest area?
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 179

TABLE 6.1
Similarities and Differences Among Sympathy, Compassion, and Admiration

Empathetic Comparison of Situational Meaning Action Tendency

Emotion for Character and Viewer

Compassion C: weak Seek intimacy, giving, protection

V: greater than C
Sympathy C: strong and weak Seek intimacy, giving and receiving,
V: equal to C sharing
Admiration C: strong Seek proximity, receiving and giving
V: less than C

thetic emotions, namely, compassion and admiration. Table 6.1 offers a sche-
matic representation of the similarities and differences.
Admiration has features in common with sympathy, but there are also dif-
ferences. In the first place, superiority, rather than equality, is an element of
the imaginary situational meaning structure for the character; she has rea-
son to be proud because she has overcome extraordinary setbacks in an ex-
traordinary fashion. Here the viewer is the lesser god.
Experienced superiority may elicit action tendencies involving some de-
gree of distance, as compared to the tendencies inherent in sympathy. The
viewer may wish to be near the admired character, to be a follower rather
than to seek intimacy and sharing of feelings. In the reality of everyday life,
persons might want to offer something to show their respect. They may be
eager to curry favor with the admired other and to receive something, in the
sense of appropriating thoughts or qualities of the admired person. Because
of the element of distance and superiority, giving and receiving are made dif-
ficult. This is why admiration may be accompanied by symbolic activity, such
as imitation. In the fictional world actual giving and receiving are impossible
rather than difficult. Only proximity (rather than intimacy) can be realized,
and this is easier than in the reality of daily life. The impossibility of receiv-
ing and giving may stimulate even more symbolic activity than in the reality
of daily life. We can think of an enhanced formation of fantasies about pos-
sessing those sterling qualities that form the basis for admiration, which is,
after all, a form of appropriation.
In the case of compassion, the situational meaning for the character con-
tains an element of superior force, in that the setbacks suffered by the char-
acter are so overwhelming that he or she is in danger of succumbing alto-
gether. The character is inferior to the viewer. The action tendency involves
an inclination to protect, to help, and to console. The viewer is prepared to
give.104 As in admiration intimacy can actually be realized—think of the close-
104
This is why I elected to use the term compassion rather than pity. Lazarus (1991) ex-
plained that pity may involve a degree of condescension, whereas compassion is characterized
by a tendency toward caring and providing help.
180 CHAPTER 4

ups in melodrama—and, as a form of giving, a willingness to go more deeply

into the unpleasant significance of the situation, to seek hope and grounds
for defense or resignation.
As we have seen, there are limits to the realizability of empathetic action
tendencies inherent to film viewing. Intimacy can be realized, but by the film,
not by the viewer; it is the film that allows the witness some proximity to the
protagonist. Given the viewer's imaginary position as a spectator to the ac-
tion, the action tendency is largely virtual, taking the form of approval, an in-
clination to encouragement and moral support, a kind of mental thumbs up.
And like the witness in the everyday world, who can take no action but is in
an excellent position to observe the action taken, the viewer is inclined to say
to the victim "They ought to take care of you," "I wouldn't have let this hap-
pen to you," to the hero "I would like to be a little like you," and to a sym-
pathetic character "It would be nice to see more of you." The impossibility
of realizing empathetic action tendencies may also provoke feelings of frus-
tration and helplessness.
The virtual nature of the action tendency encourages fantasies. Rombouts
(1992) describes falling in love as emotion; in many ways her description sug-
gests the profile of sympathy but in a highly intense form. According to Rom-
bouts, fantasizing is an important element in the behavior of people in love.
The film evokes powerful feelings of sympathy, admiration, and compassion,
while offering only the most limited opportunity to transform those feelings
into action. All these conditions are conducive to fantasizing. But films not
only encourage fantasy, they also guide and steer the fantasy that accompa-
nies empathetic emotions.
The three major empathetic emotions have in common some inclination
to give; in addition both sympathy and admiration are characterized by a readi-
ness to receive. A not inconsiderable difficulty presents itself here, namely
the fact that giving and receiving are not the same for all viewers. The tradi-
tional male hero in the stereotypical case is of course generous, though ap-
parently not interested in being cherished himself, indeed, he actively dis-
courages attempts to do so, because independence is more important to him
than anything else. The traditional male viewer will likewise have few incli-
nations in that direction: while he may provide moral support, and does not
begrudge the hero a certain amount of sympathy, this clearly does not mean
that he cherishes him. Nor is he cherished himself, not even in imagine. What
he does receive is an imaginary example of a situation which, for a man, has
a certain appeal, namely, the opportunity to protect a defenseless woman. Ac-
cording to feminist theory, the traditional female viewer should identify with
the male character, and may indeed do so, as the dominant roles in fiction
and in everyday life are played by men. But the female viewer may also iden-
tify with the helpless woman, whereby she is cherished in her imagination,
(e.g., see also de Lauretis, 1984; Mulvey, 1981; for a discussion of an alter-
native cinema for women: Brunsdon, 1986).
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 181

Sympathy, admiration, and compassion appear in a variety of combina-

tions during the watching of feature films. The poignant closing scene of Casa-
blanca (1942) is a case in point. Bogart allows his old flame Bergman to leave,
together with Victor Laszlo, the man she loves, who is being sought by the
Nazis. This arouses our admiration, not least because he has made every ef-
fort to hide from her the enormity of his loss. For the second time he is los-
ing the love of his life, this time for good. This evokes compassion. Moreover,
h e will have to pay dearly for resorting to intrigue and violence to ensure the
safe departure of the couple. It may even cost him his life, which intensifies
our feelings of compassion. Of course, this example—like any other we might
have chosen—has its shortcomings when it comes to generalization. In many,
less melodramatic, cases, the admiration of the viewer, especially toward the
end of the film, is accompanied by the triumph of the character. Not so here,
due to the Mixed Blessing nature of the farewell. Moreover, the sympathy we
feel is based on a Great Good Deed, which is not always the case.
To the three major empathetic emotions may be added others, such as
gratitude, anger, envy, contempt, and embarrassment. These often comple-
m e n t the first three, forming a response to attendant features of the situation
that evoke sympathy, compassion, or admiration. The fate of the protagonist
is, as we know, bound up with the actions of the other characters. When the
protagonist reaches his goal, this delights the viewers; when that goal is at-
tained thanks to someone else, then they feel gratitude as well. A situation
that arouses compassion for the hero may at the same time evoke anger di-
rected toward the villain or envy at his success. But it is also possible to ex-
perience a combination of compassion and anger, for instance, when the pro-
tagonist suffers a reverse that is in some way his own fault. Admiration for
the hero can go hand in hand with contempt for the antagonist. Empathetic
embarrassment often accompanies failures of the protagonist whereby he or
she loses face (Miller, 1987). In short, empathetic emotions occur in regular
configurations that reflect the mutual relationships—and in particular the
conflict—between various characters.
Empathetic emotions may be tonic as well as phasic, as we noted in the fi-
nal sections of chapters 3 and 4. The situational meaning structure for the
character involves expectations concerning favorable or unfavorable devel-
opments, or a combination of the two. The viewer hopes and fears along with
the protagonist. And finally, the empathetic emotions based on expectation
include desire and longing.
Although there are admittedly a great many genre differences, it would be
safe to say that the average film makes possible a deeper and more complete
understanding of the inner life of the protagonist, and his or her assessment
of the situation, than it does in the case of the other characters. This has to
do with the fact that the action is driven by the protagonist. Insight into the
protagonist's concerns, aims, plans, and feelings is necessary for an under-
standing of the plot, along with the significance of events and the responses
182 CHAPTER 4

that they call up. There is a two-way relationship between sympathy and an
understanding of the feelings of protagonists. Sympathy for protagonists is
fostered by providing the viewer with a better insight into their inner life. This
sympathy—both the emotion and the disposition—also increases the viewer's
inclination to delve into the psyche of the protagonist. It follows that there
will be much less understanding of the other characters, and the emotions
directed toward them will be of secondary importance.

OBSERVATIONAL ATTITUDE AND EMOTION

The F emotions are all witness emotions, as we have seen. The diegetic effect
places the viewer at center stage, and the film narrative determines the exact
manner in which the viewer is a witness to the fictional events. The traditional
feature film imposes upon the viewer a certain observational attitude.
The traditional film revolves around characters, and the events portrayed
are almost identical to the fortunes of those characters. It may therefore be
useful to draw a comparison between watching a feature film and observing
people. In Frijda's study on the interpretation of facial expressions (1956),
three observational attitudes are distinguished, each of which lends to the ex-
pression a particular significance. These are the attitudes vis-a-vis, en-pro-
fil, and en-face.105 For our purposes these attitudes can be expanded to in-
clude not only interpreting the faces of characters but also their behavior and
all other forms of their expression, seen in a situational context.
The attitude vis-a-vis, whereby the observer interacts with the observed
person by means of a direct confrontation, does not occur in the traditional
feature film. Indeed, the fact that direct address is taboo is a major defining
characteristic of the traditional film, as we saw in chapter 3. The viewer is a
hidden observer and never forms the object of action on the part of the char-
acter. The agent behind the narrative, the filmmaker, is likewise hidden from
view and can never be the object of frontal observation on the part of the
viewer.
In the case of the attitude en-profil, the observer interprets the behavior
of the model as a signal that the situation addresses certain concerns. The
viewer tries to reconstruct the situational meaning for the character by asso-
ciating expression and situational context. This observational attitude, which
might be described simply as watching, is often imposed on the viewer by the
feature film. If the character is startled, viewers immediately look for the
source of the reaction and the consequences, because they apparently affect
the fate of a character in whom viewers take an interest. Much of the empa-

105
In fact, Frijda (1956) distinguished a fourth attitude, the interpretive attitude. However,
this attitude is irrelevant for our purposes.
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 183

thetic emotion in the traditional cinema appears to be based on this mecha-

nism: on the basis of the character's reaction, one constructs the situational
meaning for the character. It may be assumed that this attitude dominates in
the traditional cinema. All traditional feature films entice their viewers into
throwing in their lot with the character and sharing the latter's concerns; as
a result, they are anxious to know what is going to happen next. This attitude
is accompanied by a comparatively broad type categorization of the charac-
ter. There is heightened attention for the events, but empathy is definitely not
excluded, as attention is directed primarily toward aspects that are of impor-
tance to the character. 106
The attitude en-face is likewise of great importance for the traditional cin-
ema. The observer follows attentively the comings and goings of the model,
for the sake of those comings and goings as well as the character himself or
herself. This attitude is accompanied by an individualized categorization.
Many films provide an opportunity to study the emotions of a character in
greater depth. In some films the narrative examines in great detail the char-
acter's reaction to events.107 The melodrama is an excellent case in point. A
large number of extended close-up shots of the character's reactions help the
viewer to understand the significance of a particular event for the character
and to empathize with her. Accordingly, the action tendency of sympathy and
compassion heightens one's attention to that character. A typical example is
to be found in Brief Encounter (1945), where lengthy close-ups showing
Laura lost in reverie are accompanied by an inner monologue. 108 The attitude
en-face also lends itself to the ultimate viewer activity: watching characters
simply for the sake of watching (as psychoanalytically oriented film theory
will [chap. 2], or for the sheer pleasure—whether sexual, aggressive, aes-
thetic, or otherwise—aroused by the sight of the character or the actor or ac-
tress). Sympathy is often accompanied by gazing lovingly at a character. In
everyday life, seeking intimacy and a shared identity leads almost automati-
cally to an attempt to establish eye contact; if the person for whom one feels
sympathy—a lover or friend—is not aware that such contact is being sought,
then he or she can be watched in a tender, loving way. In the absence of the
loved one, the lover sees the beloved in the mind's eye. Rombouts (1992)
found in the study cited earlier that the fantasies of couples in love consist
mainly of images of the outward appearance of the loved one, without any
connection with a particular situation, such as an erotic scene. The opposite

106
There may also be a measure of self-interest involved in this attitude. In his discussion of
compassion, and later in dealing with the emotions evoked by film and drama, Lazarus (1991)
made mention of the general fascination with accidents and natural disasters, which has every-
thing to do with the thought that "something like that could happen to me."
107
Some canonic narrative grammars even reserve a separate category for the "mental" Re-
action to an Event. See, for example, Rumelhart (1977).
108
For a detailed description, see Silver and Ursini (1974).
184 CHAPTER 4

of sympathy, an antipathetic disposition, does not lend itself to the attitude

en-face. In everyday life most people tend to avoid eye contact with people
they find antipathetic, preferring to turn away. Almost by definition, antipa-
thy results in less in-depth understanding. It is, however, possible to derive
pleasure from malicious observation, for example, by searching for a defor-
mity, a repulsive feature, or a base personality trait, or to study "the crook
you love to hate." The cinema provides an opportunity to do so in perfect
safety. It has often been noted that the villain is more often shown in extreme
close-up than any other character. This may have a threatening effect, exactly
because it counters a tendency in the viewer to turn away. It also emphasizes
a permanent evil intent of the villain.
The fourth attitude described by Frijda, the interpretative, plays no role
in the case of the natural viewer of the traditional feature film. The attribu-
tion of significance is totally divorced from the situational context in which
the expression is seen. The situational meaning structure for the character is
in no way part of the situational meaning for the observer.
A major difference between the two attitudes imposed on the viewer by the
traditional feature film is in the relationship between the situational meaning
structure for the viewer and the significance that the character appears to at-
tach to the fictional situation. In the case of the attitude en-face, the signifi-
cance of the situation for the character is totally shared by the viewer. More-
over, the situational meaning for the viewer is no more than the significance
that the situation has for the character. An event that gives rise to one of the
empathetic emotions has the same valence for character and viewer and is
encoded in the same way. We can take as an example the same scene from
Brief Encounter.109 The viewer is a witness to the mental processes unfolding
inside Laura. The viewer sees how her memory of the meeting with Alec is
set off against her perception of her husband, the decent but dull pater fami-
lias Fred, a contrast that is expressed as Laura herself experiences it. Of the
empathetic emotions, sympathy above all is fostered by the attitude en-face.
The similarity between the significance of the situation for the character and
for the viewer may be seen by the latter as intimacy on equal terms. As noted
earlier, we are proceeding on the assumption that one of the action tenden-
cies associated with sympathy is the sharing of thoughts and feelings; this ten-
dency is facilitated by the attitude en-face. The attitude en-profil, by contrast,
allows more scope for differences between the situational meaning structure
for the viewer and the situational meaning for the character. For instance,
characters may be aware of only a part of the total situation seen by the
viewer, so that they do not fully understand the context; conversely, they may
know more than the viewer. This attitude promotes the major empathetic
emotions, especially passion and admiration. Compassion is reinforced by the

109
See once again Silver and Ursini (1974).
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 185

superior knowledge of the viewer, who understands that the characters have
suffered a more serious reverse than they themselves realize, or that their loss
is more definitive than they believe it to be. In the same way, admiration is
evoked or reinforced when the effectiveness of the action undertaken by the
character demonstrates that he or she actually had a different and better un-
derstanding of the situation than the viewer.
In a sense, the difference between the two attitudes coincides with what is
referred to in the literature as "imagine-self empathy" as opposed to "imagine-
h i m / h e r empathy" (Davis, Hull, Young, & Warren, 1987; Hoffman, 1982,
1984; Stotland, 1969; Stotland, Mathews, Sherman, Hanson, & Richardson,
1978). The same research likewise proposes a third attitude, the "watch-him/
her" attitude; this leads to nonempathetic emotion, which is not relevant here.
Imagine-self empathy is called up by instructing the observer to imagine the
situation as if it were happening to him or her, while imagine him/her em-
pathy is evoked by imagining how the model feels. The attitude en-profil is
more compatible with imagine h i m / h e r empathy, as the character is shown
in a broader context, while the attitude en-face offers ample opportunity to
imagine what is going on in the mind of the character. Davis et al. (1987)
used dramatic productions as stimuli, and by means of an analysis of ques-
tionnaires and protocols, established that the observers who imagined them-
selves in the place of the actor, registered and experienced the actor's nega-
tive feelings, which is a precondition for sympathy.110 The subjects who dis-
played mainly imagine h i m / h e r empathy, by contrast, reported more other-
oriented positive feelings, such as compassion. On the basis of our distinc-
tion between attitudes en-profile and en-face, the same results could have
been predicted.
The attitude en-profil makes possible a contrast between the situational
meaning structure for the viewer and that for the character, which, as we have
seen, may intensify compassion and admiration. However, there are many
other empathetic emotions that thrive on a contrast between situational mean-
ing for character and viewer. Empathetic fear is intensified when the charac-
ter unaware of impending danger while the viewer knows what is about to
happen, as is often the case in suspenseful scenes. The contrast also evokes
strong virtual action tendencies: the viewer wishes there were some way to
warn the character. When maximum contrast exists, because the character
sees nothing and knows nothing, the effect may be a comical one. Buster
Keaton is walking down the street. A house collapses just as he is passing by,
but he does not even notice. A wall topples, threatening to crush him, but at
precisely the point where he is standing, there is an open window in the wall.

110
They made use of stimulus films that concentrated on the protagonists's negative experi-
ences: the psychological drama Who's Afraid of Virginia WoolfP (Nichols, 1966) and the melo-
drama Brian's Song (B. Kulik, 1970). This may have biased their results.
186 CHAPTER 4

He walks on as if nothing has happened. Monsieur Hulot emerges unscathed

from the most precarious situations, thanks to a series of coincidences of
which he is totally unaware.
The meaning of the situation for the character can even be eliminated en-
tirely. In that case, the events that befall the protagonist are often more vio-
lent and the antagonists more cruel, hostile, and asocial, precisely because
the consequences for the victim are seen exclusively form outside. (See, for
an elaboration of this theory, Neale & Krutnik, 1990, pp. 8 6 - 9 4 and Palmer,
1987.) We need only think of the various fates suffered by Donald Duck and
Popeye: being steamrollered, burned to a crisp, electrocuted, and blown up.
We find this amusing, not only because these two characters appear to have
been gifted with eternal life, reappearing in the very best of health, or be-
cause they are "only" comic strip characters, but above all because they are
perceived as more or less devoid of feeling.
Where the character is antipathetic, schadenfreude can be intensified by
the contrast between the two situational meaning structures. The bad Nazi of-
ficer Erhard, head of the Wehrmacht in Warsaw, believes that his superiors
in Berlin are full of admiration for his work, but he is being misled by a group
of Polish actors (To Be or Not To Be, 1942). Surprise and embarrassment re-
sult when the viewer realizes that he or she had been entertaining an incor-
rect situational meaning for the character, having been misled by the narra-
tive. At first Johnny appears to be a playboy who sends his best friend to his
death for the sake of his own gain, but later on it becomes clear that he had
noble motives for doing so (Suspicion, 1941).
Contrasts between the situational meaning structure for the viewer and
that for the character arise out of narrative processes that have been well doc-
umented in film theory Aumont et al. (1983) referred to identification with
the narrative, whereby the viewer joins the narrator in a position "above" the
parties. Bordwell (1985) used the term omniscient narration, a perspective
that disposes over more information than any of the characters. Ironic con-
trasts are frequently found in highly communicative narratives, where the
narrator has more information available than the characters and makes this
fact clear to the viewer. In the literature on drama theory (Pfister, 1977) such
knowledge discrepancies are referred to as dramatic irony, information lead,
or information gap. We have looked briefly at this lead or gap in information,
as measured by the knowledge that the viewer has of the situation in com-
parison with the character. We have also examined an example of dramatic
irony, "The Gift of the Magi," in chapter 5: we alone know that the actions of
Jim and Delia will lead not to joy but to disappointment. The symmetry of their
actions was also stressed, an aspect that is similarly lost on the characters
themselves. Dramatic irony is an important component of drama and occurs
frequently in both tragedy and comedy. The irony becomes more complex
when it also includes relationships between the viewer and the ideas, feelings,
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 187

and person of the maker or author. The battle scenes in war films such as All
Quiet on the Western Front (1930), Gallipoli (1981), and Full Metal Jacket
(1987) evoke more than empathetic fear. It is quite possible that a particular
view of war entertained by the filmmakers is part of the situational meaning
structure for the viewer, thus complementing the emotion felt. There are nu-
merous examples in fiction that lead to the conclusion that such an associ-
ation must be part of any description of the emotional effect of the products
involved. It is not clear whether all these examples can be translated into situ-
ational meaning structure without losing the meaning they have within the
literature devoted to various ironies. What is important here is that empa-
thetic emotion experienced in film viewing is not always based exclusively on
a comparison between the valences of an event in the situational meaning for
oneself and those in the situational meaning for the character. Other ele-
ments, such as the degree to which the character is aware of the exact nature
of the situation and the effect of his or her actions, make the situation tragic,
comic, embarrassing, awkward, and so on, depending on the genre, among
other things.
Such ironic effects may have in common the fact that the spectator or wit-
ness role reserved for the viewer is stressed in the situational meaning struc-
ture. The viewer wants to warn the hero or heroine of impending danger but
is incapable of action. Or, conversely, the viewer delights in the sensation that
while the danger that looms for the protagonist is coming closer and closer,
there is no need to warn her. Ironic effects are compatible with empathetic
emotions because both require knowledge of the plans and objectives of the
characters, as well as their perception of the situation. Nevertheless, they in-
troduce into the experience a certain remoteness to the object that is lack-
ing in the pure emotions (i.e., those uncolored by irony) such as sympathy,
compassion, admiration, and vicarious embarrassment.
Observational attitudes are also marked by a certain way of seeing and cat-
egorizing characters. Where impression formation leads to typing, as is gen-
erally the case, the empathy is of an imagine him/her nature, rather than the
imagine-self kind. In those relatively rare cases in which the traditional film
leads to personalization, the link between the traits of the individual and the
action is often much less pronounced, and the understanding of the charac-
ter is more of a goal in itself. Here the representation of character traits is or-
ganized around the individual. The core that is thus formed integrates not
only all the elements, such as roles and traits, but also situations and the re-
lationships between the various characters. And yet the resulting cognitive
structure is similarly loose, producing a representation that is complex, ac-
cessible in various ways, and capable of different applications. Because the
representation of the self displays a similarly complex and flexible organiza-
tion (Greenwald, 1981; Markus, 1977; Markus & Nevius, 1987), in these cases
the film encourages a kind of imagine-self empathy. In fact it might be more
 188 CHAPTER 4

correct to speak of imposition rather than encouragement, because in this

case there are so few schemas available that support the elaboration and or-
ganization of the information provided by the film.111

Observational Attitude and Film Narrative

The above analysis of observational attitudes in terms of situational meaning
structure is intended as an introduction to the formation of a theory. I do not
know of any empirical studies focusing on observational attitudes among the
viewers of feature films, despite the urgent need for such research. Among
the important questions to which an answer must be found are the following.
Can observational attitudes be distinguished on the basis of categorization
processes? And can they be distinguished on the basis of the involvement of
the self? On the stimulus side, we have yet to determine exactly how the film
narrative imposes a particular attitude on the viewer or how the narrative
steers the process of appraisal. That it does so was the opening assumption
of our discussion on observational attitudes. Aumont et al. discuss—albeit in
a cursory manner—the structures responsible for this steering process, for
which they employ the term surface structure of a film. A major step forward
was taken by Branigan (1992), who saw a narrative as a process by means of
which an initial difference in level of knowledge between the narrator and the
recipient is abolished. Branigan distinguished eight levels of narration, ac-
cording to the different types of knowledge (or, to take Branigan's term, data)
that are communicated to the viewer. On the highest level, the knowledge is
extremely broad and unfocused; here an empirical maker communicates an
artefact to empirical viewers. On the lowest level the knowledge imparted has
the narrowest focus; here private thoughts are transmitted by a character to
a viewer who identifies with that character. In between these two extremes
we find, with narrowing focus, types of knowledge such as the events that a
narrator in the fictional world passes on to a character and spoken text that
one character directs to another. This view is related to empathetic emotion
as follows. Branigan's highest level of narration gives rise to an aesthetic at-
titude. The viewer concentrates on the narrative and undergoes A emotions.
The seven lower levels correspond to attitudes that give rise to F emotions,
whereby the two lowest attitudes—internal focalization (surface) and inter-
nal focalization (depth)—force the viewer to adopt an attitude en-face. The
information that is transmitted, consisting of the perceptions or thoughts of
the character, is concerned exclusively with the situational meaning for the
m
Miall (1989) demonstrated the shortcomings of prototype schemas in explaining how a lit-
erary narrative such as E. M. Forster's A Passage to India is understood. In Miall's alternative a
representation of the reader's self plays a key role in establishing coherence in the processing
of a narrative. The representation of the self by the reader is capable of linking such widely dif-
fering types of information as character behavior and descriptions of atmosphere.
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 189

character. The levels situated between the upper two and the lower two trans-
mit information with an external focus, which imposes on the viewer an atti-
tude en-profil.
Future research might concentrate on Branigan's analysis, in an effort to
discover how the dynamics of observational attitudes can be predicted. On
the basis of an examination of the opening scenes of Hangover Square (1944),
Branigan demonstrates how the various narrative levels, or focuses, can be
realized by technical and stylistic means. For example, he indicates the pre-
cise point at which the narrative switches from external to internal focaliza-
tion, by means of an initial point of view shot, and then returns to an exter-
nal focus. In our terminology the sequence not only introduces a character
but also the situational meaning structure for that character: George Harvey
Bone has knowingly and willingly committed a murder, and must face the
consequences.

Empathy and Identification

While empathetic emotions may be complex (as in The Piano), they do not
always require the observer to delve into the innermost feelings of the char-
acters. If one is justified in referring to the image of the internal life of a charac-
ter in the traditional feature film as simplistic, because it is based on a stereo-
typical cognitive structure, then empathetic emotions provoked by traditional
film in its viewers are similarly lacking in depth. This view is based on the fact
that empathizing with a character often consists simply in seeing the situation
as he or she sees it. This is particularly true in the case of films: all we have
to do as viewers is to recognize what the situation means to the protagonist.
And most films paint a fairly unambiguous picture.
A second restriction on empathy and empathetic emotion could be the fact
that the observer does not entirely share the—constructed—situational
meaning structure for the observed person. The film viewer is an observer
with a particular mind set or attitude, one that is rooted in the pragmatic prin-
ciples of watching feature films and in the qualities of the film itself, that is,
of the film structure. This mind set or attitude determines which portion of
the total conceivable situational meaning structure for the film character is
reflected in the situational meaning that the same fictional situation has for
viewers, and the exact role that it plays for within their situational meaning
structure.
The position adopted here with respect to empathetic emotion entails a
sharp delineation of what is often termed identification. Schoenmakers (1988)
presented a detailed explication of the concepts of identification and empa-
thy, with reference to the activities of spectators in a theater setting. He notes
that in the literature these concepts are mistakenly seen as synonymous. In
our own terminology there is a difference between the constructed situational
190 CHAPTER 4

meaning for the character and that for the viewer, for the simple reason that
the viewer is not the character. The events of the narrative befall the charac-
ter, not the viewer. Nor are viewers under the impression that these events
touch them; the diegetic effect creates the illusion that one is present in the
fictional world, but as a spectator or witness. In short, the situational mean-
ing for the viewer always contains, in addition to the situational meaning for
the protagonist, an element of spectacle. Of course, viewers can actively ex-
plore what it is like to be the protagonist, for example, when the film forces
them to assess the situation entirely from the perspective of the protagonist;
something that does not often occur (see, for instance, Mitry's (1965) reserve
against the assumed currency of subjective images). Imagine-self empathy
does not, however, mean that one is under the impression that one is the
other. The literature referred to above deals with research in which subjects
did their best to imagine that they were feeling the same emotion as the
model. However, it is improbable that subjects were at any time unaware of
their efforts to imagine having the emotion. Total identification, in a literal
sense, would end in delusion. Experiencing the exact same emotion, rather
than some representation of that emotion, means that one not only recog-
nizes or is able to form an impression of the protagonist's concerns but that
one actually has those concerns oneself. As the law of concern states, I can-
not have my emotion, where my concerns are not at risk.
If one is determined to retain the term identification at all, then it can
best be reserved for those cases of empathy in which the viewer imagines
what it is like to be a particular character. But because, as we have seen, it is
not possible to take on completely the situational meaning for the character
in question and because this is not necessary in order to experience empa-
thetic emotion, I consider the term identification misleading.

CONCLUSION: EMPATHY,
INVESTMENT, AND RETURN

In this chapter we have examined the role of the cognitive process that leads
to the representation of characters. We have seen that this process interacts
with an understanding of the action that was the subject of the previous chap-
ter. We have maintained that the empathetic emotion experienced by the film
viewer is far from an in-depth understanding of the innermost feelings of the
film character. It is sufficient for the viewer to simply follow the film narra-
tive. I believe that this is what Aumont, Bergala, Marie, and Vernet (1983)
meant when they say that the spectator identifies with the narrative. How-
ever, in their view, the desire to see the story end in a manner that is favor-
able to the protagonist, and the endurance of postponements and threats to
that happy outcome, satisfies a fundamental oedipal need. This tragic view is
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 191

unnecessary; as we saw in chapter 4, interest—the ubiquitous emotion—is

fed by promise. There is always some measure of return to be gained and, if
all goes well, at least some hope of more in the offing. The term identifica-
tion with the narrative is unsatisfactory; it suggests an awareness of the arte-
fact that plays no role in empathy. Empathy is concerned quite expressly with
characters, who, in the reading of Aumont et al., appear to have disappeared
from view. The film narrative forces the viewer to call up a vision of charac-
ters, their objectives and their actions, and shows the relationship between
their strivings and the actual situation. The step from comprehension of film
characters to empathetic emotion is a small one. But a more refined com-
prehension and a closer examination of character emotions make possible
deeper empathetic emotion, although not every film will give rise to such
emotion at any given moment.
Empathy and empathetic emotions are a source of both investment and re-
turn. Let us look first at return, which arises from the gratification of the mo-
tives for watching films, which we discussed in chapter 2. To begin with, un-
derstanding a character is in itself pleasurable, not least when the complex-
ity of the film narrative is rooted in the development of the character, as in
psychological film drama. But characterization of even the simplest kind is
often a source of pleasure and not exclusively in a genre like the farce. There
is no great need for the viewer to examine the feelings of supporting players
and antipathetic characters; it is enough to enjoy the apt and telling charac-
terization. Furthermore, the major empathetic emotions have a rewarding as-
pect. Sympathy and admiration contain an element of obtaining, deriving, or
receiving. And, as in compassion, there is also an element of giving involved;
empathetic emotions exhibit a cost aspect, which is important for invest-
ments. However, this aspect is limited, since here giving is accompanied by
feelings of warmth and thus need not be seen exclusively as a loss. Giving
sympathy is a source concern. The fact that people are capable of experi-
encing solidarity with others and caring about other people, which in turn
gives rise to positive feelings in themselves, cannot be traced to deeper-seated
mechanisms. Although there is every reason to do so—the question of whether
altruistic motivation actually exists still occasions spirited debate—any at-
tempt to substantiate in detail the existence of such motivation would lead us
too far afield.112 It is rooted in culture, and the history of altruistic ideologies
is a long one. I am proceeding on the assumption that stimuli that are rec-
ognized as intentional activate satisfaction conditions for sympathy con-
cerns. 113 It may well be that the same limitation—inherent in the spectator
role—whereby the action tendencies associated with empathetic emotions are
largely virtual has a favorable influence on the hedonic nature of emotions

112
See Dawkins (1978, p. 215), Batson (1987), Dovidio, Schroeder, and Allen (1990).
113
For the concept of "intentionality," see Dennett (1987).
192 CHAPTER 4

such as sympathy, empathetic sorrow, and empathetic fear, because the costs
involved are low. Viewers can indulge feelings of sympathy for a character in
distress, without ever being in danger themselves. In a more general sense,
the costs of sympathy during the watching of a film are extremely low. In
everyday life, it requires an effort to help family and friends when they are in
trouble. Often it is difficult to pay regular visits to a friend in the hospital; it
takes time, and in the more serious cases it is not easy to shrug off the painful
impressions and intense experiences with which one may be confronted dur-
ing such visits. It is quite possible that films offer not only sensation, spec-
tacular action sequences, and likeable characters whom one gets to know
intimately but also the opportunity to be altruistic at a far lower cost than re-
quired in everyday life. For next to nothing, the viewers have given their sym-
pathy; they have shown themselves benefactors simply by sympathizing with
the film characters. And if the protagonist fares badly, viewers can still find
satisfaction in a comparison with their own fate. At least for a moment, they
are in an extremely favorable position.
And finally, affectively charged beliefs are confirmed. Showing sympathy
is favorable to one's self-image and in accordance with culturally shared
norms. One may also feel a sense of solidarity with one's fellow spectators,
or a wider audience, as a result of one's sympathy and support for a just cause.
Let us now look at the investment. Empathy calls up expectations about
what a particular character is going to do and how he or she will react to
events. These expectations are cognitive investments. Affective investments
consist mainly of feelings of sympathy for the protagonist. Because the film
narrative places obstacles along the path of the protagonist, this sympathetic
disposition may be seen as a source of investment. The viewer hopes for the
best and bears up under the inevitable setbacks. Thus sympathy is a source
of investment as well as return. The investment aspect of empathetic emo-
tions with a largely negative charge, such as empathetic fear and empathetic
disappointment, is abundantly clear. As we have seen, they are nevertheless
capable of calling up anticipations of long-term benefits. In the traditional
cinema affective investments are concerned primarily with the postponement
of reward. Of course, the viewer does not remain in a state of dejection and
unhappiness right to the very end of the film. With the addition of new com-
plications, the viewer's sympathy for the characters grows, as does the
promise, that is, the size of the reward attached to a final outcome that is fa-
vorable to the protagonist. In the traditional feature film the promise of an
improvement in the situation of the protagonist never drops to absolute zero.
And thanks to that promise, every empathetic emotion is accompanied by in-
terest, the urge to know what happens next.
In art cinema, which is presented here by way of contrast, in order to elu-
cidate a particular aspect of the traditional cinema, the negative empathetic
emotions are often more intense and more painful. Imagine-self empathy is
CHARACTER STRUCTURES, EMPATHY, AND INTEREST 193

less likely to result in feelings of sympathy and an altruistic, sharing type of

action tendency. Nor does the promise of a satisfactory solution increase to
the same degree. The fact that some art cinema films do not find favor with
the natural audience of the traditional feature film (Tan, Eggermont, &
Joosten, 1989) may have to do with the fact that the imagine-self empathy
evoked by these films leads to an excess of uncomfortable empathetic emo-
tions. The futility of human existence and the shortcomings of humanity are
themes that feature in what some critics refer to as the merciless character-
ization displayed in modern novels and art cinema films.114
In the traditional cinema, at any rate, understanding characters and be-
coming engrossed in what makes them tick has very little to do with coping
with an irreparable loss, as the tragic representation of identification would
have us believe. It is a hedonic affair, which offers both short-term and long-
term rewards.

114
For a discussion of "merciless characterization" in the modern novel, see Auerbach (1946).
Bordwell (1979; 1985, pp. 207-208) discussed "modernist" characterization of the protagonist
in the art cinema.
 The Psychological Affect Structure
7 of the Feature Film

AN OVERVIEW OF TilE AFFECT STRUCTURE

At the beginning of this book we put forward two questions. Our examination
of the first, which focused on the adequacy of the term film emotion, will be
postponed for a while. But we are now able to formulate an answer to the sec-
ond question, namely, how to explain the systematics of the emotion evoked
by films. The following is a proposal for the description of that systematics,
based on the insights developed in the previous chapters. It consists of a ten-
tative model representing the course of the emotion aroused by the tradi-
tional feature film, together with the determinants of that emotion. As we have
seen, films not only trigger a multitude of emotions, they also regulate their
intensity from one moment to the next. The model is intended to account for
the composite nature of those emotions, as well as their dynamics.
The model is tentative and some of the components are still programmatic,
because I have been unable to arrive at a satisfactory solution for all the the-
oretical problems touched upon in the previous chapters. However, the pro-
posed model does clearly indicate which questions arc most urgently in need
of solution.
Figure 7.la is an overview of the most important con1poncnts of the affect
structure, whereas Fig. 7.lh fills in the details of that structure. In Fig. 7.1a
we see that there are three substructures that develop over time, parallel to
one another. The time axis at the top of the figure covers the period between
the projection of the first and the last frame of the film. Thus at any given

195
 VIEWING TIME ~ ~- VIEWING 77ME-~-7
....
-c
"'
Narrational!
'' EPISODES
I NARRATIONAL TEXTBASE I Textbase j EVENTS

1 [fiJ
ory~

SITUATION MODEL :--coi.IPONE:NTS_________ _

- EMOTIONAL I valence
MEANING Emotional!
STRUCfURE prospects
l Meaning ]
retrospects
Structure i difficulty
'
' other

-------------------------
1
TOMC EMOTIONS
I EMOTIONAL I Emotional Interest
RESPONSE
Response
S}'mp8thy
J I d, I J
/JesU.S '·
(o)
Hopes and '
Feers '
+f
•.~
\~
FIG. 7.1. Th~ affect strudure of the feature PHASIC EMOTIONS ·\
...
film: (a) Overview of the emotion process; 'Co
'
(b) Model of Generic Mfect Structure (for ex- '
'------------------------------------------------------------------------------------'
"'
planation, sec text).
(b)
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 197

moment during the viewing of the film there is a text base, a situation model,
and an emotional response.
As is customary in theories of the cognition of discourse, we are assuming
that the viewer's comprehension of the film narrative begins with the forma-
tion of the text base, a propositional representation of the discourse (van Dijk
& Kintsch, 1983). This text base is the first result of following with under-
standing the filmic action, which is relatively close to the directly observable
surface structure of the film.
The second substructure, a situation model, also a customary ingredient
of theories of discourse comprehension, is derived from the text base in the
process of formation (van Dijk & Kintsch, 1983; Johnson-Laird, 1983). This
contains a representation of relevant details of the fictional action taking
place and is based on what has actually been shown on the screen plus vari-
ous inferences. The situation model also contains expectations concerning
coming events (see chap. 5). Not all of the situation model is of importance
for the emotion of the viewer. The part that is significant for that emotion is
what we refer to as the emotional meaning structure of the situation. For the
sake of clarity, it should be noted that up to now we have used the term situ-
ational meaning structure. To avoid any confusion with the more compre-
hensive situation model, in this chapter we will use the term emotional mean-
ing structure (of the situation in the fictional world). The emotional mean-
ing structure evokes the emotional response, that is, the action tendency and
the emotional experience, which together form the third substructure.
As Fig. 7.1b shows, the text base consists of a series of events that are suc-
cessively registered by the viewer during the viewing time of the film. The in-
troduction of a new event has certain consequences for the emotional mean-
ing structure and the emotional response. This is represented by the lines
that start at the events and continue on down to the bottom of the schema.
Let us look first at the text base, the representation of the fictional action. In
previous chapters we referred in a fairly general manner to the narration of
the film story. The text base is a structure that, unlike the narration, resides
with the viewer. The parsing of events by the viewer is guided by their func-
tion within the various episodes, as described by Rumelhart (1977). An (ac-
tion-)episode is a problem cycle consisting of a minimum of one initiating
event and one outcome. In Fig. 7.1b, there are two events per episode. As a
rule, however, there are also intermediate events. Let us say, for example, that
in one episode a protagonist tries to realize a goal or subgoal; each separate
step, consisting of an attempt plus an outcome, counts as one event. The epi-
sodes themselves are embedded in the larger episode, which consists of the
story as a whole, where they serve a particular function. There is at least one
episode in which there is a Balance, another in which the balance is disturbed
(a Complication) and one in which the balance is restored (the Restoration
of Balance due to a Resolution) (see chap. 3). Episodes coincide roughly with
198 CHAPTER 4

scenes. However, unlike the latter, they can be arranged not only linearly but
also hierarchically, that is, one episode can be embedded in another. Thus an
attempt to attain a particular goal can be divided into various efforts to reach
mutually dependent subgoals.
The emotional meaning structure is a representation of the components
of situational meaning and their values. In chapter 3 we discussed the com-
ponents as represented in Frijda's (1986) theory. In dealing with the affect
structure, it is important to distinguish between constant and variable com-
ponents. Some components maintain a more or less constantly high value.
Gore components such as objectivity and reality make possible the emotion
and are hardly subject to change due to the fact that the diegetic effect and
the illusion of being a witness to fictional events are constantly in operation.
The impossibility of active intervention is also important for the film emo-
tion, but then as a continuous and unvarying given. However, in the case of
the affect structure, which is a dynamic structure, other components are of
importance, namely, those that are variable. The variable components that
are always part of the emotional meaning structure include the valence of the
situation in terms of viewer concerns (including the fate of the sympathetic
protagonist; see chap. 6), prospects and retrospects, difficulty, and relevance.
Each event has a valence, calls up expectations, and displays a link with pre-
vious events, as illustrated in the analysis of Straf/Punishment (1974) in
chapter 5. Prospects and retrospects are concerned primarily with the pres-
ent episode and only secondarily with events situated further away in the past
or future (see Principle 3 of interest, chap. 4). The degree to which the pros-
pects and the change are favorable as compared to an event in the past, in
other words promise, form the basis for investment and return. During the
greater part of the film narrative there is also a difficulty. On the one hand,
the difficulty is related to the problem of the protagonist with whom the
viewer sympathizes. On the other hand, the difficulty for the viewer is the
lack of cognitive and affective closure. And finally, there is always a greater
or lesser degree of relevance. Some events are less relevant to the concerns
of the viewer, primarily the cognitive concern of closure and the affective con-
cern of an outcome that is satisfactory for the protagonist and compatible
with the values of the viewer. Thus relevance is not the same thing as promise,
although they are related. High relevance may be accompanied by either high
or low values of promise.
There are several other variable components of emotional meaning, but
these are not all applicable to every filmic situation. These include certainty,
urgency, familiarity, closedness, and intentionality, qualities that only occur
in certain situations.
The emotional response consists in an action tendency and the emotional
experience. The emotional experience is the awareness of the situation as
specified in the emotional meaning structure and the awareness of the ac-
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 199

companying action tendency. However, the subject is generally able to give a

succinct characterization of the emotional experience that coincides with
one of the familiar emotion words. For this reason, in Fig. 7.1b the emotional
responses have been designated as emotion labels, rather than as action ten-
dencies. We previously distinguished between phasic and tonic emotions
(chap. 3). Phasic emotions are relatively brief responses to a particular event.
Each event has a valence, and thus evokes enjoyment or, conversely, dis-
pleasure. In general, however, events evoke more than one emotion. An event
with a positive valence leads to enjoyment; where the event is important and
unexpected, it also calls up surprise; when the surprise consists of a moder-
ate incongruity, exhilaration may ensue (Ruch, 1993). Intentionality behind
an event with a negative valence arouses not only pain but also anger. The ir-
reparable loss of a situation with a positive valence leads to displeasure as well
as sadness. When the loss takes the place of an expected return, this produces
disappointment. If a sympathetic protagonist is the victim of such a loss, and
there is little that he or she can do about it, then the sadness and disap-
pointment are mixed with pity. If, however, the loss is the protagonist's own
fault, pity gives way to regret. In this way there are numerous variations pos-
sible in the mix of emotions accompanying a particular event, depending on
the complexity of the narrative and the observational attitude imposed at a
particular moment. Hence for every emotion (i.l or i.2 in Fig. 7.1b), there is
an emotion or a mixture of various emotions.
The major phasic emotion is enjoyment. While enjoyments can be of longer
duration, this is hardly ever the case. In general, such emotions occur only
during the Complication stage. The paradisal state of the original Balance is
rarely enjoyed; it is only after it is over that a longing is felt for its restoration.
The first joy and enjoyment are not experienced until obstacles are beginning
to be overcome. The persistence of any enjoyment is, however, highly pre-
carious. Satisfaction soon becomes habitual and thus loses its positive char-
acter (cf. Frijda's [1986] Law of Hedonic Asymmetry). Enjoyments arise
when one finally attains a desired object or situation. Other phasic emotions
may be seen as emotions proper (Frijda, 1986). They respond to changes on
the way to the desired situation, in films usually the appearance —or re-
moval—of disturbances and obstacles. Emotions proper include, on the one
hand, anger, frustration, disappointment, sadness and anxiety, and, on the
other hand, relief and joy.
Tonic emotions, which are based on prospects and retrospects, have a
longer life span. The major tonic emotions include desire, fear, and hope. It
may be useful here to give a brief description of each, based on Frijda (1986).
Desire is characterized by the absence of an event with a positive value and
a high focality, that is, an event that can be easily specified. The action ten-
dency is an urge to transform absence into presence. Fear is the response to
the presence of an event with an uncertain negative outcome in a context of
200 CHAPTER 4

closedness, that is, a threat which one can do very little about. The action
tendency is the inclination to flee. Hope is the response to the presence of
an event with an uncertain positive outcome. The accompanying action ten-
dency is an inclination to approach.
As we know, interest is the most important tonic emotion (see Fig. 7.1b).
In fact, in general, it is the most important emotion involved in the viewing
of feature films. All expectations and their resolution, all prospects and ret-
rospects, are not only of importance for other tonic emotions, they also in-
fluence the promise of the final outcome preferred by the viewer and are in
this way linked to interest.
Any event can initiate, turn into, or terminate a tonic emotion. Hopes arise
or are dashed. Desire can become fear when the loss threatens to become
permanent. Interest is strengthened or weakened by a particular event. An
event that influences a tonic emotion in some way or another always brings
with it a phasic emotion as well, for the simple reason that such an event intro-
duces an important change, and that in itself produces a short emotional re-
sponse. The affect structure also reflects other relationships between phasic
and tonic emotions. Some emotions have both a phasic and a tonic aspect.
Pity may be a brief response to short-term suffering but it may be a more last-
ing emotion, which continues as long as the difficulties that the protagonist
faces. Sympathy, as an emotional response, appears not only in a fleeting
form but also as a disposition that persists throughout every feature film, as
we saw in chapter 6. Furthermore, it is important to note that phasic emo-
tions are also influenced by expectations. Without doubt, one of the most im-
portant phasic emotions triggered by watching feature films is surprise. Any
event that is not in keeping with expectations produces a sensation of sur-
prise. In the light of the fact that many expectations are highly nonspecific
(chap. 5), it will be obvious that many, if not most events lead to at least sev-
eral small surprises. Disappointment and fears confirmed are other phasic
emotions that arise when positive expectations are thwarted or negative ones
confirmed. The intensity of many phasic emotions depends on the preceding
tonic emotions. Relief and satisfaction, for example, are brief responses that
are more intense the stronger the preceding emotions of fear or frustration.
(Note that it is not so much the expectation itself that shapes and intensifies
the phasic emotions, but rather the tonic emotion, i.e., the response to the
expectation.)
Let us return now to Fig. 7.1a, for a remark on the mutual cohesion be-
tween the text base, the emotional meaning structure, and the emotional re-
sponse. Broadly speaking, the text base is a precondition for the emotional
meaning structure, while the emotional meaning structure is a precondition
for the emotional response. This is shown by the arrows in Fig. 7.1a. However,
there are forms of feedback among the three structures. Although the emo-
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 201

tional meaning structure is fed by the text base, it is possible for the former
to evoke in the viewer an attentional set that promotes certain propositions
in preference to others. Similarly, the emotional response may influence the
salience of components of the emotional meaning structure. To take an ex-
ample, it is conceivable that empathetic fear produces a virtual flight ten-
dency, that is, the intense desire to see the protagonist flee. This tendency
heightens the salience of the closedness of the situation: there is nowhere for
the hero to go. A comparable effect is reported by Keltner and Ellsworth
(Ellsworth, 1991). Subjects were asked to read either a tragic story or one
calculated to make them angry. They were then presented with an embar-
rassing situation, in which they were asked to imagine themselves as the pro-
tagonist. Those subjects who were sad were inclined to see situational causes
that no one could do anything about, while those who were angry focused on
the evil intent of several of the characters involved.
To round off this overview, a few comments on the theoretical status of the
affect structure are in order.
1. The affect structure of the feature film is a psychological structure, por-
traying as it does processes that viewers undergo. Unlike semiotic structures,
for example, it is related to the viewing subject.
2. The affect structure is a "real-time" representation of the course of the
emotional response and its determinants. At any given moment during the
film, the viewer has the text base available, complete up to that point in the
action and stored in memory. When event i.l is presented (Fig. 7.1b), event
i.2 is not yet represented in the text base; moreover, event 1.2, say, may have
been forgotten. The situation model is related to present events, and the pros-
pects and retrospects that it contains are limited in scope. Moreover, the
prospects are uncertain.
3. An affect structure may have a theoretical or an empirical status. The
overview presented previously is a theoretical structure. By means of analy-
sis, it is possible to predict for any given film how the formation of a text base,
an emotional meaning structure, and the emotional response will proceed.
We refer to this body of mutually related predictions as the theoretical affect
structure of the film. In a sense, every film—even one that has not been seen
by anybody—has such an affect structure. By that we mean that a film has a
potential supply of viewer reactions, ranging from the comprehension of the
text base right down to the emotional response. A theoretical affect structure
can be verified by collecting measurements of the parameters of the three
substructures. The empirical affect structure is a representation of the results
of measurement followed by interpretation. Results must have been inter-
preted because predicted courses in time of the various parameters and re-
lationships among these have to be tested in the usual way.
202 CHAPTER 4

THE GENERIC AFFECT STRUCTURE

In Fig. 7.1 every effort has been made to exclude the individual particulari-
ties of feature films. The aim was to set down the generic affect structure of
the traditional feature film, the average or otherwise representative reflection
of the temporal course of the emotions and their determinants, which is valid
for all feature films. The affect structure can, however, be made specific for
each individual feature film; in this way, the specific affect structure of a par-
ticular film is exposed. In between these two extremes, affect structures of
certain corpora, such as genres, or a certain type of film dating from a par-
ticular period, can be identified.
An affect structure of any kind is defined not only according to the domain
of films to which it belongs but also according to the domain of viewers. This
is equally true of the generic affect structure. We have confined our research
into the emotion of the film viewer to the traditional feature film and its nat-
ural audience, that is, the viewer who by means of a process of self-selection,
freely chooses to watch a traditional feature film and does so in a manner that
is customary for cinema viewers (see also chap. 1).

Episodicity and Unity

What then are the core characteristics of the generic affect structure, which,
for the sake of convenience, we will shorten to "affect structure?" The most
important comprehensive characteristic is that it is episodic. As we saw at the
close of chapter 4, the affect structure may be seen as a threefold recursive
episode. The text base, the film narrative, consists of an episode that in turn
contains various subepisodes. In the emotional meaning structure, this
episodicity is reflected in the fact that each event signifies a change in rela-
tion to two criteria, namely, the cause of the problem and the anticipated im-
provement of the situation. It is primarily the change that the viewer experi-
ences. The action moves in the direction of closure of the present episode,
toward an outcome of the present attempt at improvement. In the back-
ground the closure of larger, hierarchically higher episodes also play a role.
The promise of a cognitively and affectively valuable final representation gov-
erns the emotional meaning structure. Even when the threat to the character
becomes almost unbearable, there are at least minimal indications of coming
improvement, as regards both the fate of the favorite character and closure
of the cognitive structure. The emotional response is likewise an emotion epi-
sode, whereby interest lends unity to the response. Interest also dominates
the other emotional responses, forming the backbone of the affect structure.
Alongside virtual action tendencies to protect the leading character, to flinch
when he or she does, or to turn away in disgust, there is always the inclina-
tion—rooted in the promise of improvement—to go on watching in order to
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 203

find out how it all ends, to search for closure and completion of the image
that one has of the action. The typical emotional responses peculiar to the
various genres, such as tears, laughter, and horror, all have their own indi-
vidual place in interest episodes, perhaps more commonly known as tension
spans. Within the larger interest episode of the film as a whole, smaller in-
terest episodes can be distinguished. The initiation of an "attempt" encour-
ages investment, and the outcome leads to closure of the interest episode.
Interest is reinforced by the other emotions and, in turn, reinforces them.
The phasic emotions are responses to the value of a significant event. They
underscore and heighten the significance of events and, in this way, influence
the promise of cognitive and affective closure. Because in the development
of the text base the Complication phase lasts the longest, negative events (i.e.,
reverses) will inevitably predominate. The emotions accompanying these
negative events heighten the value of the final outcome. Intermediate frus-
tration and disappointment actually serve to intensify the value of the ulti-
mate reward. Here, as in our discussion of the intensification of phasic emo-
tion, we see that the emotion tends to carry more weight than the event in
itself. Only in the emotion does the event address the self of the viewer, and
only emotions (i.e., negative ones) count as investment. Thus, on balance,
the other emotions widen the discrepancy between attained and expected re-
turn, and increase interest.
The action tendency of interest is an emotional inclination to see what hap-
pens next, and to formulate active expectations; this reinforces feelings of sym-
pathy, desire, hope, and fear. In other words, the investment readiness of in-
terest enhances the tonic emotions and, indirectly, the phasic emotions. More-
over, viewers' anticipations are sharpened, and they encode with greater atten-
tiveness the incoming events, which directly stimulates the phasic emotions.
The affect structure is characterized by unity. The viewer is aware of the
unity of the emotional experience. In a word: films produce emotion, and the
rich brew of emotions has a harmony of its own. What the viewer feels is not
pity alongside fear, but rather an integral combination of the two. Even where
the emotions are contradictory, this is experienced as an appropriate dis-
crepancy. We need only think of hope and fear or empathetic embarrassment
combined with exhilaration. The unity of the emotional experience is due in
the first place to the dominance of interest. But the other emotions are not
held together by interest alone; as we have seen, there is also a mutual cor-
relation among them. Phasic emotions are in part a response to tonic emo-
tions. Empathetic admiration for the victorious hero is accompanied by ma-
licious delight at the downfall of the villain. The generic affect structure is
also characterized by an important system of correlations over time. Thus the
greater the anxiety of the viewer, the stronger the feelings of euphoria when
events proclaim the coming Restoration of Balance. The recursiveness of
emotion episodes is of great importance here. The resistance of the viewer is
204 CHAPTER 4

built up in a stepwise manner. Anxiety grows with each successive episode;

the background investment increases and, at a stroke, is transformed into re-
turn (see chap. 4).
A final contribution to the unity of the emotional response is to be found
in the phenomenon that films create moods. The term mood is often incor-
rectly used to describe more or less protracted emotions of fairly low inten-
sity. However, Frijda, Mesquita, Sonnemans, and van Goozen (1991) noted
that moods are not necessarily either protracted or of low intensity. Accord-
ing to these researchers, a more important characteristic is the absence of
an object and the highly global nature of the appraisal. When you feel de-
pressed, the whole world looks miserable. Moods are not episodic and may
last longer than the film, as any cinema viewer knows. Indeed, getting into a
good mood is often a motive for going to the cinema (chap. 2). Mood is a dis-
position that encourages certain emotions and inhibits others. It is plausible
that the consistent predominance of a certain emotion, such as anxiety, will
in the end produce a mood, a disposition toward appraising emotional mean-
ing structures and a readiness to respond in a certain manner, for example,
with panic. One might even say that the moods evoked by films are the result
of rapid learning processes. They are part of the self-reinforcing mechanism
that is inherent in the emotion experienced by the film viewer.

The Course of Emotions

The Course of Interest

The generic affect structure is characterized by a limited number of dif-
ferent possible courses of interest. With respect to the course of interest,
there is a certain regularity in both the large emotion episode, which is built
upon the complete text base, and the constituent episodes. For one thing, in-
terest increases in the course of the film as a whole. This is not surprising,
for otherwise the viewer's attention would diminish through sheer habitu-
ation. This intuition is in line with studies dealing with the course of tension
in drama that go back to the early years of this century; the norm for a good
composition was deemed to be a gradual increase in tension throughout the
play as a whole with, in some cases, a resolution about two thirds of the way
through the performance or later (Brunetiere, 1914; Freytag, 1911; Mat-
thews, 1910). A similar course has also been established empirically by
Lazier, Zahn, and Bellinghiere (1974) among members of a theater audience
and Tan and van den Boom (1992) and Tan (1986b; in prep.) among the view-
ers of various films. However, an overall increase is also to be expected, in
the light of the theoretical principles of interest (chap. 4). As the text base
becomes fuller and more complete, the final outcome in the fictional world
gains in clarity and value. At the same time, however, the attainment of that
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 205

final outcome is postponed, for the rule is that throughout a considerable por-
tion of the film, the Restoration of Balance does not take place in the present
scene. The background interest gradually increases, which gives rise to an
overall increase in interest (see Principle 3 of Interest, chap. 4). In addition,
interest is self-reinforcing (Principle 4), that is, the action tendency of inter-
est coincides with investment readiness. This likewise promotes an overall in-
crease in interest. It is not until the end of the film that interest again de-
creases, when the final results of the last episode are made known and the
main problem of the Complication inherent in the narrative has been solved.
Viewers know that the final representation has been reached. They are satis-
fied because the traditional feature film is characterized by a complete nar-
rative structure where there is a definitive end or an end that may be open
but, in any case, immediately recognizable as such. As we noted earlier, there
is generally a happy end, one that is affectively satisfactory. The hero triumphs
and the lovers are joined together for all eternity (see chaps. 1, 3, and 5).
From a psychological point of view, it is quite understandable that films al-
low promise to increase toward the end and that the greatest reward for the
viewer is postponed until the latest possible moment. At first glance, however,
it might seem more logical for viewers to prefer as rapid a solution as possi-
ble. Is it rational to postpone reward? Research in choice behavior shows that
when isolated outcomes are presented, immediate reward is preferred; given
a sequence of outcomes, more evenly spread reward is preferred. A recent
overview is provided by Loewenstein and Prelec (1993). They referred to a
number of studies that point to a preference for series of solutions that dis-
play an improvement over time. These researchers discuss in detail an ex-
periment by Ross and Simonson (in Loewenstein & Prelec, 1993), which re-
vealed a preference for a happy ending. Subjects were presented with a choice
between sequences that ended with a loss (e.g., profit $85-loss $15) and oth-
ers that ended with a profit (e.g., loss $15-profit $85). The vast majority of
the subjects opted for the series that ended with a profit. The authors' own
research shows that when subjects themselves are asked to compose series,
the majority consistently produce sequences characterized by improvement
of outcomes toward the end. There are a number of complementary expla-
nations for this phenomenon, such as loss aversion, adaptation, and the de-
sire to get undesirable outcomes over with quickly. Moreover, research by
Loewenstein and Prelec (1993) showed that people tend to distribute favor-
able and unfavorable outcomes throughout the course of a series, so that a
preference for a certain sequence of results may be due to two interactive ten-
dencies, namely, a preference for improvement and a desire for a distribution
throughout the entire interval. In our view, a hierarchical episode structure
would do justice to both tendencies, in that most but not all improvement
may be concentrated in the last episode or episodes.
When we speak of an overall rise in interest, we are referring to an aver-
206 CHAPTER 4

age increase in the course of the film as a whole. The level of interest, how-
ever, may vary considerably from one episode to the next. All feature films
display not only moments of high tension but also scenes of repose, inter-
mezzi that give the viewers a chance to catch their breath (see also chap. 4).
In scenes of this type, the action is concerned with secondary plot lines and
often promises no significant improvement in the situation with respect to
the fate of the protagonist or the closure of the narrative structure. And in
between these two extremes—scenes of heightened tension and intermezzi—
all manner of intermediate levels are possible.
One might expect to find the same course of interest within individual
episodes. The narrative structure of the episode is, after all, the same as that
of the narrative in its entirety, but then on a smaller scale: an attempt to deal
with the problem results in an outcome. It is also characteristic of traditional
film episodes that while one problem is being brought closer to solution, a
new problem is rearing its ugly head or an existing one is being held over
(Bordwell, in Bordwell, Staiger, & Thompson, 1985). This would lead us to
expect a gradual increase in interest over the episode. However, the conven-
tional scene often begins with some form of introduction. As a result, the be-
ginning of a scene is accompanied by a fall in interest (for the time being, not
much can be expected), after which interest rises and then falls again (due
to presentation of an outcome), though not reaching its initial value (due to
a new complication). The characteristic course of interest within the scene
would then take the form of an n, whereby the vertical on the right is higher
than the one on the left. This pattern has been established in at least one
study. Figure 7.2 shows the data recorded by Tan and van den Boom (1992).
Subjects in this study viewed Opname/In for Treatment (1979), a tragic
drama film in which a terminal cancer patient becomes resigned to his fate.
They were asked to indicate continuously, by means of a slide, the level of in-
terest. There was a significant linear increase in interest over the film as a
whole. Furthermore, it was overly clear that with the exception of number 5,
every scene began with a decrease in interest. In all the scenes except 21 this
was followed by a substantial increase in interest and, except for 3, 5, and 17,
a subsequent decrease.

Three Narrative Procedures

Structuralist theory of literature distinguishes three ways of varying the
level of interest throughout the narrative as a whole or within a single episode,
namely, suspense, curiosity, and surprise. All three methods manipulate an-
ticipations and expectations, although they differ in the way fictional events
are ordered by means of the narrative (Sternberg, 1978). These three pro-
cedures also play an important role in Bordwell's account of the mode of nar-
ration employed in the classical Hollywood film (Bordwell, 1985). These same
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FIL~I 207

I
7

6~

5~

JN ~'A
4 1-

3 tl
vM ~~
;.. ~
2~

tJ
0
I 2 3 56 7
4
I 9 10 II
8 12
13 14 1s 16 11 \ 19 20 I
18 21
22 23 24

FIG. 7.2. Course of interest throughout the film Opname/Infor Treatmellt (Tan & van den
Boom, 1992). Means per segment on a scale ranging from 0 tn 7, N =' 17. From Nardm:chio
(1992). Reprinted hy permission.

qualities have also been studied from a psychological perspective by Zillmann

(1991c). In describing the procedures, we make use of the definitions of
Hrewer and Lichtenstein (1982), who hark back to the work of Sternberg
(1978). By means of a fe\v simple experiments, Brewer and Lichtenstein
demonstrate that the three procedures are seen by readers of narratives as
characteristic of a good narrative. \Vhen a base narrative was varied accord-
ing to these three procedures, by presenting a suspense version, a surprise
version, a.nd a mystery version. the course of affect registered during the
reading of each version agreed quite closely with predictions made on the
basis of each particular procedure.
According to Brewer and Lichtenstein, a su._<.,pense procedure is one in
which an initiating event (IE) is introduced in an early stage, an event that
causes the viewer to become concerned about the possible results for a sym-
pathetic character. The IE coincides with the event that introduces the at-
tempt, which will he familiar from our earlier discussions. However, there the
attempt referred to the actions of the protagonist, while the agent of the IE
may be either the antagonist or the protagonist.
The IE results in a high level of investment (e.g., a bomb has been placed
 208 CHAPTER 4

in the heroine's car), whereas the narrative also includes the outcome of the
IE (e.g., the bomb explodes just as the heroine is entering her house). Be-
tween the IE and the outcome, supplementary material is included, which,
while it must not lead to resolution, in any case does not result in boredom.
This keeps the net return already attained low, while retaining and even en-
hancing the prospect of an outcome with a maximum value (see Fig. 7.3a).
In our example of the film Straf/Punishment, t h e setting information pre-
sented after the IE—the smashing of the violin—produced an increase in the
investment and the expected return (chap. 5).
A surprise procedure is characterized by the sudden presentation of an
unexpected outcome, whereby the IE is recounted later, or even omitted, so
that it is only through inference that it is added to the text base. The outcome
follows upon the attempt. It is important here that it is not experienced as
disturbing that the IE is not presented until after the outcome. One way this
is done is to present incomplete or misleading information. Following the sur-
prising outcome, the true IE becomes clear, either immediately or gradually.
Let us suppose that we have seen someone rummaging around the heroine's
car; sometime later the bomb explodes, long after she has reached the safety
of his house. It will be clear from this example that the procedure will only
be effective if, in retrospective, the seeds—however tiny—of an IE have been
planted. If it is to be effective, surprise must never be total: there must be a

a. S u s p e n s e b. Surprise c. Mystery

IE 0 0 IE 0 IE

presentation time presentation time presentation time

Expected NR

Attained NR

FIG. 7.3. Three classical narrative procedures and the course of interest. (Interest at moment
i = NRi Expected - NRi attained).
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 209

minimal anticipation. In other words, a successful surprise is a postdictable

surprise (Kintsch, 1980). Until the IE, both investment and return are low.
The Net Return attained is close to zero, and not a great deal is happening.
Following the presentation of the IE, the return attained suddenly increases,
and, very briefly, there is a high investment and an extremely high expected
return. Viewers realize that they have been surprised but do not yet know how
until they arrive at the right postdiction (see Fig. 7.3b).
A curiosity or mystery procedure is characterized by the fact that an out-
come is launched early on in the narrative, while the absence of an IE is con-
spicuous. The investment generated takes the form of a larger number of who-
what-why questions. The maximum expected return is determined by the
prospect of an appropriate and/or surprising outcome, which reveals the IE
(Fig. 7.3c). For example, the outcome is presented in the form of a flashback
of the bomb explosion at the beginning of the narrative. The mystery-cum-
detective story is based on the procedure.
The three classical procedures designed to heighten tension differ from
one another in the manner in which the net return attained and the net re-
turn expected—and thus the interest—are distributed over time, while the av-
erage interest level may be the same. In the case of surprise and curiosity, in-
vestments increase only gradually. If we look at suspense, however, we see that
investments related to the danger that threatens the protagonist rise sharply,
while returns are concentrated around the moment when the outcome is re-
vealed. Thus the suspense procedure is characterized by a steep increase in
interest over time, while the curiosity procedure leads to only a modest rise
over the initially high level.
It is understandable that these procedures often appear in combined form,
or alternate with one another. Straf/Punishment, for example, begins with a
surprise, and this generates a mysterious effect (Why did Marjan destroy the
violin?), while the narrative slowly dissolves into a suspense procedure. "The
Gift of the Magi" opens with a mystery procedure (What are they going to
do?), followed by the surprise (the fatal sale and purchase) and the suspense
created by the viewer's knowledge advantage (How will they react when con-
fronted with the gifts?). The overall mystery-cum-detective narrative often de-
velops from mystery (Who did it?) to surprise (X did it!), after which one or
more suspense episodes follow in which the culprit is run to ground or
launches a final offensive that places the protagonist in a tight corner. 115

115
The mystery procedure is not restricted to the detective film genre. Any plot with a
"thriller" aspect, even one that is "richer" from a psychological standpoint, may display such a
development. To take an arbitrary example, in The Third Man (1949), the second-rate writer
Martins is forced to choose between love and friendship on the one hand and justice on the other.
Following his arrival in Vienna, he goes in search of his friend Harry Lime, about whom strange
rumors have been circulating, which Martins finds difficult to believe (mystery). After having
caught fleeting glimpses of Lime, he is almost convinced that Lime is indeed an unscrupulous
210 CHAPTER 4

The role played by surprises should not be underestimated. The example

of a bomb that explodes unexpectedly may be misleading to the extent that
here the IE has been all but deleted. This is not always the case, particularly
in the run-up to the quite common smaller surprise. It is not so that only a
specific type of narrative, such as the Roald Dahl story or the plots of O.
Henry (see "The Gift of the Magi," chap. 5) epitomizes this procedure. In
such narratives the surprise is complete and occurs at a fairly late stage, typ-
ically in the end, and as a consequence there is no expectation to which the
outcome corresponds. (An example would be Delia's selling her hair). As
Sternberg (1978) stressed, it is actually quite unusual in any narrative for ex-
isting expectations concerning an outcome to be fully realized. In most cases
it is not so much unexpected that some kind of resolution is presented at this
point, but rather that the nature of the event representing the outcome dif-
fers to a greater or lesser degree from what one expected. Moreover, as we
have already noted in chapter 5, expectations are as a rule not very specific,
often taking the form of diffuse anticipations, what Bordwell (1985) called a
"wait-and-see-strategy." Assuming that this is true, then our initial observa-
tion might be that surprise plays an important role in the other two proce-
dures, suspense and mystery. In the case of suspense one expects the out-
come—the heroine does not fall to her death—but it is difficult to guess how
she will get out of the tight spot in which she finds herself. The same goes for
the outcome of the mystery: an effective buildup of interest means that the
details can never be guessed at before it has been revealed. The outcome
(who the culprit is, say) is at least slightly surprising (X did it, after all) and
is liable to be extremely surprising (They all did it together). A second ob-
servation is that films are actually full of surprises. Not one single event proves
to be precisely what we expected, and each successive surprise serves to re-
inforce interest. (For an empirical demonstration of the role of surprise in a
feature film see Diteweg & Tan, in prep.) While we may not be able to pre-
dict the outcome of a particular episode, what we are given is decidedly
worthwhile. It is postdictable, that is, looking back, it is appropriate, it can
be explained, and leads to an improvement in the prospects of the protago-
nist. Looking forward, we must accept that the details, at any rate, will re-
main hidden. This is true a fortiori for the outcome of the narrative as a
whole: each intermediate surprise increases the promise of an elegant and
satisfying final representation.
Surprise is a fixed component of the pragmatic contract between viewer

dealer in counterfeit medicines. At this point Lime takes the initiative and arranges a meeting,
which is meant to be a final warning to Martins that he had better keep his nose out of Lime's
affairs. At this meeting Lime had an opportunity to kill Martins but did not do so (surprise). In
the end, Martins decides to betray Lime and his girlfriend, Anna, whom he has come to love.
The outcome is revealed after an exciting chase through the sewers of Vienna: Martins shoots
Lime, after the latter has again had an opportunity to kill him.
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 211

and filmmaker. The viewer acquiesces in the fact that many developments
cannot be guessed at in advance, although a really good film will entice the
audience into trying to do just that. Each ingenious twist turns the viewer into
a good loser, who does not mind admitting that he or she appreciates the ul-
timate result.
As we have seen, the course of interest forms the backbone of the affect
structure and determines to a large extent the other emotions. Thus the three
narrative procedures are of influence on the remaining emotions. Let us look
first at tonic emotion. The buildup of suspense is accompanied by specific ex-
pectations that are concerned with an important short-term outcome. These
expectations are affectively charged, taking the form of hope and fear. This
is because the IE touches the concern of the protagonist, a concern that is
shared by the viewer. A high level of interest, and thus a marked tendency to-
ward affective investment, goes hand-in-hand with prospect-based emotions.
In the case of the mystery procedure, anticipations are more diffuse, and the
viewer has at least a suspicion that in due course all will be revealed. The
viewer's expectations may be reflected in a desire to discover the missing
links in the final outcome of the narrative (Who did it?). The surprise pro-
cedure does not involve explicit expectations and, hence, is not accompanied
by any special tonic emotion.
What is the relation between narrative procedures and phasic emotions?
The IE in the suspense episode always causes a negative emotion, such as dis-
pleasure and—when accompanied by surprise—fright. 116 The IE creates ex-
pectations but is itself capable of fulfilling or altering expectations. Thus the
IE occasions disappointment and "fears confirmed." The outcome of the sus-
pense episode is generally greeted with relief, often in combination with dis-
appointment, where the solution creates a new problem. The phasic emotions
at the end of the suspense episode are intense, because the change in the fic-
tional world is measured against the IE and what the outcome might have
been.
Surprise, above all, reinforces the main emotion, the nature of which is de-
termined by the outcome. 117 Thus an unexpected setback clearly leads to a
greater sense of disappointment than a negative outcome that was more or
less expected. The surprising outcome is also a reason for enjoyment. Unex-
pectedness per se is rewarding. Major surprises, or ones following pronounced
expectations that prove to be wrong, may also result in an awareness of the
narration, and hence, in A emotions. The remarkable lack of an IE in the mys-
116
The negative character of the IE flows from the current definition of suspense: it is initi-
ated by an event that provokes worry or moral sympathy with a protagonist. (See Brewer & Licht-
enstein, 1981, 1982; Carroll, 1988b; Zillmann, 1991c.)
117
See also the discussion of the role of surprise by Ortony, Clore, and Collins (1988). These
researchers see surprise as one of the major intensity factors in emotion. In their view, the ex-
act nature of the emotion is determined by other factors.
212 CHAPTER 4

tery procedure naturally evokes wonder, while the outcome itself calls up
some form of emotion. The corpse inspires disgust, and the murder we have
just witnessed fear. When the presentation of the outcome takes place quite
early on in the film, the viewer is most likely to be gripped by highly non-
empathetic emotions such as horror and loathing. The element of mystery
creates a desire for knowledge, that is, curiosity, while resolution brings with
it satisfaction. In general, this desire for knowledge on the part of the viewer
will coincide with the aim of the protagonist, thus merging with empathetic
emotion. Bordwell (1985) described the similarity between the knowledge of
the viewer and that of the detective, as well as the limits to that similarity.
The obstacles that the protagonist encounters in the search for truth, and the
manner in which she triumphs over them, not only strengthen the viewers'
desire for knowledge but also their sympathy for the protagonist. A mixture
of empathetic and nonempathetic frustration and satisfaction is characteris-
tic of the phasic emotion evoked by this procedure.
As we conclude this examination of narrative procedures, it should be
noted that there may well be other ways of structuring the course of emotion
within larger and smaller interest episodes. The aim of the previous discus-
sion of the three recognized procedures was simply to demonstrate that they
fit neatly within the framework of the episodic affect structure. The proce-
dures—each in its own way—seize the opportunities that the episode offers
to intensify emotion to the maximum (see also chap. 4). This is supported by
Zillmann's explanation of at least two of the procedures (1991c; see also Zill-
mann, 1980). He sees both mystery and suspense as a means of bringing
about excitation transfer, thus increasing the intensity of the emotion. The
higher the excitement during suspense or mystery, the stronger the eupho-
ria occasioned by the outcome.
Like other emotion episodes, the affect structure of the feature film agrees
with the combined operation of the Laws of Change and Closure (Frijda,
1986, 1988). The three narrative procedures carefully mete out the changes,
whereby different criteria are used. During a suspenseful episode the viewer
is aware of a change with respect to what has gone before, the IE, and what
is probably about to happen, the outcome. In the case of mystery, the out-
come is a given, and this is the criterion against which the change is mea-
sured, while in the case of surprise, the criterion is what might have hap-
pened. However, as the narrative moves toward the end and completeness,
the viewer is also eager for closure of the present episode. As the law of clo-
sure emphasizes, the interest response, that is, the urge to go on watching, is
almost unstoppable until the viewer has gained a complete picture of the final
outcome.
The succession of different episodes also offers additional opportunities to
increase interest throughout the film as a whole. The hierarchical nesting of
episodes makes it possible for interest to increase along with the number of
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 213

levels nested. Between the attempt and the outcome of the highest episode,
a second attempt-outcome pair can be inserted, which forms the next to high-
est level of the hierarchy; this procedure can be repeated a number of times.
Each following nested attempt-outcome pair represents a source of invest-
ment and postponement of return, which causes the level of interest to rise.
For example, a failure to attain one of the subgoals results in an increase in
frustration. When episodes are nested, the speed of change, which is impor-
tant for the intensity of the emotion, can be given in measured doses. To take
an example, a series of subgoals may suddenly be realized at one fell swoop,
as it were, which leads to euphoria; conversely, a setback can undo the reali-
zation of a series of subgoals, which results in a sinking feeling (Carver &
Scheier, 1990). Or in a series of favorable outcomes the magnitude of the im-
provement may increase, resulting in even more satisfaction (Hsee & Abel-
son, in Loewenstein & Prelec, 1993).

Research into the Affect Structure

Development of the Text Base

If we look at our overview of the affect structure, we see that there are still
considerable lacunas in the knowledge required to map out the entire affect
structure of the feature film, or even substantial portions of that structure.
Let us examine again the three substructures text base, emotional meaning
structure, and emotional response, but now with the focus on urgent research
issues. There has been little if any research into the comprehension of the
surface structure of the feature film. In the analysis of Straf/Punishment
(chap. 5) use was made of the roundabout method consisting of the post hoc
construction of input propositions. Important work has been done in the area
of the comprehension of discrete transitions between shots (e.g., Hochberg,
1986; Tan, in press). But to the best of my knowledge no one has succeeded
in explaining how on-line propositions are regularly extracted from incoming
perceptual data.118 Existing film theory, which is oriented toward the de-
scription of larger units, offers little help here. There are several intuitions in
circulation concerning emotionally effective film style. For a start, there are
several smaller experiments conceivable in which a cinematographic tech-
nique that is easy to isolate can be varied. For example, questions like the fol-
lowing seem to merit an experimental approach: Does a certain scene pro-
duce the same emotional response when close-ups are replaced by long shots,
surveying the entire action. There have been a few limited studies devoted to

118
For example, there has been no research into eye movements on the basis of which claims
can be made about what the viewer sees at any given moment. We simply do not know how the
viewer views.
214 CHAPTER 4

this question, but not enough research with a sound methodological basis is
available to warrant discussion here.

Emotional Meaning Structure

Great strides have, however, been made in research into situation models
formed by readers of written narratives. Many of the results would appear to
be applicable to the comprehension of narrative film. However, no studies
have been devoted to the question of how an emotional meaning structure is
derived from the situation model and to what extent the situation model ac-
tually goes beyond the emotional meaning structure (Tan, 1994). I believe
that cognitive research into such questions is most urgently needed. The
problem of emotional expectations deserves special mention here. In the lit-
erature on the cognitive processing of stories, there is some disagreement
concerning the degree to which the subject striving to comprehend a narra-
tive constructs anticipations about coming developments. Simple top-down
processing with the aid of "story grammars" has long since proved untenable
(de Beaugrande, 1982; Wilensky, 1983b). As noted in chapter 5, the present
alternatives range from a strictly local understanding, whereby the situation
model does not extend beyond what is explicitly provided in the text base,
and a situation model that looks further ahead. We have also seen that our
own quite limited analyses give rise to the assumption that anticipations are
often tentative and diffuse, in the sense that they do no more than constrain
coming surprises. Another example of this can be found in Tan and Diteweg
(in press). There is an urgent need for research that can provide more solid
grounding for this assumption.
A more or less intuitive analysis of dramatic structures, which makes use
of existing drama theory, may at least lead to hypotheses concerning the most
important components of the emotional meaning structure of a particular
film. On the basis of such an analysis, Tan and van den Boom (1992) assumed
that in the film Opname/In for Treatment (1979) a pronounced develop-
ment of the situational meaning takes place. The film first creates uncertainty
over the fate of the protagonist. Then the viewer is told that he is going to
die, a fact the protagonist himself is unaware of. This is the situational mean-
ing of pity. In the last phase of the film the viewer's superiority is reversed,
when the protagonist takes the initiative, choosing to actively take leave of
his beloved. The result is admiration in the viewer. Questionnaire data ob-
tained in Tan and van den Boom's study provided support for this presumed
development of emotional meaning. The course of pity and admiration were
also in accord with expectations; special mention should be made of the fact
that admiration, which initially did not even approach the level of pity, far sur-
passed it in the closing scenes of the film. An intuitive approach such as this
means a major step forward in the prediction of emotional meaning, but still
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 215

leaves many questions unanswered. Which aspects of the situation are ex-
plicitly represented? Which inferences concerning future events are made?
How are the components of situational meaning and their values derived from
the elements represented? 119
Studies into the structure of situational meaning and its relationship to the
emotional response carried out by researchers of general emotion has paved
the way for understanding emotion in the film viewer (Frijda, 1993; Frijda,
Kuipers, & ter Schure, 1989; Lazarus, 1991). What is still lacking is an ex-
planation of how emotional meaning arises in the cinema. What is the mech-
anism by which the emotional meaning structure is derived from the surface
structure of the film? In particular, it might be useful to examine the manner
in which the viewing attitudes sketched in chapter 6 are reflected in the emo-
tional meaning structure. A tentative step in this direction has been taken in
the research into imagine-him versus imagine-self attitudes referred to in
chapter 6, and in the research into coping strategies in relation to stress (Folk-
man & Lazarus, 1988). Some of the strategies consist precisely in adopting
some special observer attitude toward aversive events. However, it is quite
conceivable that the general psychological theory of emotion requires sup-
plementation when it comes to the role of affect in the observations of indi-
viduals. It is difficult to translate into current psychological concepts the nu-
ances that the film viewer applies when assessing what is going on in the mind
of a character. Quite a successful preliminary effort is the study by Zillmann
and Bryant (1991), who translate the essence of humor in comedy into psy-
chological terms. They demonstrate that in a nonserious context, aspects
such as a certain view of people and a tendency to belittle characters cate-
gorized as bad, are essential in producing the response of mirth. However,
not all exhilaration (rather than mirth) is a response to disparagement hu-
mor. Some degree of incongruity in the stimulus may be just as important
(Ruch, 1993). Finally, there has been no empirical research into the manner
in which the situational meaning for characters is embedded in the emotional
meaning structure developed by the viewer.
As regards the emotional meaning structure, the free interview would ap-
pear to be the most suitable method of data collection (Lazarus, Speisman,
Mordkoff, & Davidson, 1962; Zillmann & Cantor, 1977). There has been some
research dealing with the measurement of empathy among film viewers by
means of questionnaires (de Jonge, 1994; Koriat, Melkman, Aver ill, & Laza-
rus, 1972; Stotland & Mathews, 1978; van Vliet, 1991; Tan, in prep.).
119
Genre studies may in like fashion contain hypotheses about the typical emotional mean-
ing structure of a given body of films. Examples are additional studies of comedy (e.g., Eaton,
1981; Neale & Krutnik, 1990; Palmer, 1987), and studies dealing with horror (e.g., Carroll,
1990), melodrama (e.g., Lang, 1989; Petro, 1989), film noir (e.g., Borde & Chaumeton, 1975;
Porfirio, 1976), westerns (e.g., Nachbar, 1974; Wright, 1975), and gangster movies (e.g., Parish
& Pitts, 1976).
216 CHAPTER 4

Emotional Response
A great deal of research has focused on the emotional response to films,
albeit with other aims than ours. In psychological research, films are some-
times used to evoke more or less persistent emotions in the laboratory. We
look first at these more long-lasting emotions. In an extended survey article,
Martin (1990) described how different moods are evoked with the aid of
films. In chapter 2 we referred to studies by Zillmann and his associates that
show that watching films can lead to an improvement in the mood of a sub-
ject. More recently, Zillmann, Rockwell, Schweitzer, and Sundar (1992) dem-
onstrated that widely differing drama genres—two types of comedy and one
tragedy—enhance the ability to withstand pain. A blood pressure cuff was at-
tached to the arm of the subjects, and they were asked to indicate at what
point the pressure became uncomfortable. Subjects who had seen the genres
mentioned above were able to endure more pain occasioned by the inflation
of the blood pressure cuff than control subjects. The fact that the genres in-
volved displayed maximum differences suggests that it is primarily interest
that is responsible for the improvement in the mood of subjects. The re-
searchers themselves conclude that it is not amusement, but rather the film's
potential for absorption that is responsible for the increased tolerance for
pain, which points in a similar direction.
Interest, as defined in this book, has not been studied to any extent. There
are, however, operationalizations that come close to the concept of interest.
In various studies, interest (in the sense intended here) has been measured
by means of a complex of rating scales that made use of such labels as ex-
cited, enthralled, and involved (Tan, 1986b; Tan, in prep.; Tan & van den
Boom, 1992). In these studies, the use of a questionnaire appears to differ-
entiate well between different scenes. As noted, Tan and van den Boom
(1992) also measured interest continuously and simultaneously with the aid
of a response slide.120 The correlations between the latter measurement and
the questionnaire indices were more than satisfactory. In another study (Tan,
in prep.), a questionnaire measuring interest in three extremely divergent
groups of respondents during the showing of Opname/In for Treatment dis-
played an almost identical course of interest over the 24 consecutive scenes
of the film. In contrast, the average intensity per scene was systematically
higher in one group than in the other two. It was also found that when a ques-
tionnaire covering all the scenes was filled out after the subjects had seen the
entire film, the results differed only slightly from those obtained by having

i20jn 1951 Meier made mention of experiments with an "audience response machine." Spec-
tators at a theater performance were asked to move a slide back and forth, and these movements
were transmitted to a recorder. As the account does not indicate what instructions the subjects
were given, there is no way of knowing what was being measured.
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 217

the subjects answer questions covering various series of scenes during short
breaks of the film show. The tentative conclusion is that the questionnaire
measure of interest is stable across conditions of use and provides good dif-
ferentiation. At present, no conclusions can be drawn with respect to the con-
tent validity, notably the relationship between the questionnaire measure, on
the one hand, and perceived promise and readiness to invest, on the other.
The investment of interest is accompanied by a selective allocation of at-
tention for the film stimulus. Accordingly, we may expect that (a) interesting
segments will seem shorter than less interesting ones (Gupchik & Gebotys,
1988; Ornstein, 1969) and (b) during interesting segments viewers will be
more inclined to sit still than during less interesting segments. There did in-
deed appear to be a significant negative correlation between interest as mea-
sured by two questionnaire items and the subjective duration of segments.
There was also a negative correlation between interest intensity and number
of subjects in a test audience who displayed movement; however, this corre-
lation was extremely weak (Tan, in prep.). Further experiments involving de-
tection or verification tasks are needed in order to obtain more information
on the allocation of interest and the maintenance of affectively charged ex-
pectations. Alternatively, selection and effort paradigms might be used for
this purpose. Selection behavior can be measured by establishing which of
several video channels the subject chooses (Christ & Medoff, 1984). Interest
in a target movie might be measured by recording how often the subject
zapped to another channel. Bryant and Zillmann (1977) measured the ab-
sorption potential of programs. In our terms, this is the degree to which a
television program forces the viewer to invest. According to Bryant and Zill-
mann, two measures proved useful here, namely, the number of mistakes
made when subjects were asked to perform a simple secondary task during
the watching of the film and their recall of material presented before the film.
Such ingenious experiments may shed some light on the action tendency
of interest, with a view to validating the simpler measures, such as question-
naires. In order to examine interest as a component of the affect structure,
however, continuous means of registration are needed. Not only the response
slide, but also nonverbal behavior may be useful, preferably behavior that ex-
hibits the subject qua natural viewer of a feature film. From our own research,
it appears that during some films very little nonverbal behavior is displayed.
For the most part, subjects sit quite still and the few movements they do make
are fairly random and differ widely in frequency from one person to the next.
Facial expression may well be a more suitable index of interest. Davidson,
Ekman, Saron, Senulis, and Friesen (1990) recently reported the occurrence
of a particular type of relaxed smile, the "Duchenne smile," which is associ-
ated with mild amusement. They demonstrated this by means of EEG indices
of positive affect accompanying this smile. It is conceivable that interest is
connected with this response. The Duchenne smile was registered among sub-
218 CHAPTER 4

jects watching a film.121 The fact that interest is theoretically an ever-present

emotion makes it less likely that it is accompanied by highly conspicuous
signs. Still, it may be that smaller, less expressive responses, such as eye
movements and alterations in the diameter of the pupil provide a more
promising starting point for the measurement of interest. Further research is
needed to determine to what extent such responses reflect only a peripheral
aspect of interest, namely, the intensity of attention. Interest itself refers to
investment readiness in a broader sense, as we have seen.
As regards the other tonic emotions, most of the research focusing on emo-
tionally charged expectations, such as fear, has been done by Lazarus and his
associates (Lazarus & Alfert, 1964; Lazarus, Speisman, Mordkoff, & David-
son, 1962; Nomikos, Opton, Aver ill, & Lazarus, 1968). This research made
use of films as stimuli. When the psychophysiological response was measured
over the entire film, the dynamics of the response proved to be related to the
consecutive events in the text base. Physiological excitement or arousal as a
response to suspense during the watching of a fiction film has been measured
by Zillmann, Hay, & Bryant (1975). Heart rate proved to be a reliable indi-
cator of arousal: the greater the arousal, the higher the heart rate. However,
Zillmann (1991a) believed that the reliability of this measure is reduced by
the fact that it is negatively influenced by a second mechanism, the orienta-
tion reflex. The concept of arousal is, in any case, ambiguous and for this rea-
son it has not been used in the present account of film-induced emotions.
(Frijda, 1986; Zillmann, 1991a.) In some studies where arousal is treated as
equivalent to tension it coincides with the tonic emotion linked to expecta-
tions that is under discussion here. All in all, it appears that there are a num-
ber of problems attached to research into long-lasting tonic emotion during
the watching of feature films.122 In my view, this is due largely to the fact that
the expectations connected with the emotional response are often diffuse and
difficult to describe in detail.
The registration of phasic responses may be expected to present fewer

121
Zillmann, Hay, and Bryant (1975) previously had difficulty in determining the subject's ap-
preciation for a film by means of facial expression. The interobserver consistency in their ex-
periment was lower than that of interviews. It also proved difficult to distinguish between liking
and relief.
122
The quantity of information may itself present problems when it comes to registering the
emotional response as a component of the affect structure of a feature film. Continuous mea-
surement, as in the case of psychophysiological recording, produces a considerable amount of
data when carried out during an entire feature film, generally about an hour and a half. A prac-
tical solution is that put forward by Uchiyama, Hanari, Ito, Takahashi, Okuda, Goto, & Tsuji
(1990). They had subjects watch a number of abridged feature films. These still provided a com-
plete picture of the course of emotion throughout the film as a whole, since the researchers had
shortened each individual scene. Research by de Wied (1991) also demonstrates that the dy-
namic characteristics of the response—in this case, those that accompany tension—are retained
when a film is abridged in such a way that the main action within each scene is preserved.
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 219

problems. Precisely because they are accompanied by rapid changes in ex-

perience and behavior, they are easier to observe; moreover, because they
are linked to important events, they are easier for the viewer to recall and to
name. Thus not only is the overt action tendency visible, the emotional experi-
ence and meaning are also accessible. Quite usable results can be obtained
using simple questionnaires. The same is true of the intense forms of tonic
emotion, such as terror, as long as these are related to a clearly distinguish-
able series of events. The literature contains many examples of the registration
of such overt reactions as relief, crying, laughter, and erotic reactions in re-
sponse to cinema or television films. Shock and fear also are well-documented
emotional responses. 123
Research by Philippot (1993) is more in line with the concept of the affect
structure of feature films. He has demonstrated that the emotional response
that is to be expected in the light of the genre of a particular film, that is, the
dominant emotional response, can actually be measured by means of self-
report methods. He presented his subjects with a number of widely differing
films and employed a variety of self-report formats.
Finally, we must say something about the further quantitative modeling of
the affect structure, whether generic or specific. Of course, it is possible to
set up experiments in which one element of a film, say, the text base, is ma-
nipulated in order to determine the effect on another variable of the affect
structure. There are several good examples in Zillmann's research that have
already been discussed (Zillmann & Bryant, 1975a; Zillmann & Cantor, 1977;
Zillmann, Hay, & Bryant, 1975). Elsewhere I pointed out that it is all but im-
possible to manipulate film elements, while logically excluding the possibil-
ity of confounding independent variables (Tan, 1991). The alteration of a sin-
gle element leads to changes in, or an altered perception of, other elements.
Any opportunity to carry out a worthwhile experiment should, of course, be
welcome. However, so-called passive observation is essential where the aim is
to determine the affect structure, for this can only be validated on the basis
of the response of the natural viewers of one or more existing feature films.
Generally, the low cost of passive observation contrasts with the high cost of
statistical control, and in the case of the affect structure, these costs are high
indeed. As we have seen, the empirical affect structure reflects not only the

123
Most of the studies focusing on the emotional response to cinema and television films are
contained in the standard work by Zillmann & Bryant (1991), aptly entitled Responding to the
Screen. Studies dealing with the registration of emotions in response to film include Zillmann &
Bryant (1975b) and Zillmann, Hay, and Bryant (1975) on relief; Efran and Spangler (1979) on
crying; Cogan, R., Cogan, D., Waltz, and McCue, (1987), Levy and Fenley (1979), and Schachter
and Wheeler (1962) on laughing. For erotic responses, see the survey by Weaver (1991) and the
comprehensive collection focusing on pornography by Zillmann and Bryant (1989). Cantor
(1991) provided a most useful survey of studies dealing with fright responses. Tamborini (1991)
discussed the response to horror films and the remarkable role played by individual differences.
220 CHAPTER 4

course of the emotional response but also that of its determinants, the emo-
tional meaning and the text basis, and their interrelationship. In principle,
the relations among the three substructures can be described by means of
correlations. However, such descriptions quickly become extremely complex.
Without going into detail here, a few remarks may be appropriate. If one vari-
able from each substructure is registered, this results in three correlated time
series. Autocorrelation is of influence in each of these series, as well as one
or more overall trends and seasonal effects. Some of the overall trends and
seasonal effects are interesting, in the sense that they are theoretically pre-
dictable. The effect of a linear increase in the intensity of interest and other
emotions can be explained, as can seasonal effects attributed to various
episodes or scenes. In addition, there are random fluctuations and trends that
may be regarded as background noise. The complexity is further increased
by the necessity to include the factor "film" in the equations. To determine
the generic affect structure, we must generalize about individual films and
their viewers. To do justice to the complexity of all this dynamic covariation,
the researcher will have to make use of sophisticated statistical methods, such
as dynamic structural equation modeling. Even when employing techniques
like these, it will be difficult both to detect relevant patterns of correlation
and distinguish these from various artefacts. On the other hand, it would ap-
pear that mere specification of the various effects, which is a necessary part
of complex modeling, is in itself an important addition to our present knowl-
edge of the generic affect structure.

CONCLUSION

In this chapter we proposed a model for the systematic description of the

emotion evoked by the traditional feature film in members of its natural au-
dience. This description makes it possible to do justice to certain qualities of
viewer emotion that were discussed in the previous chapters. With the aid of
the model, the emotional experience evoked by a great number of different
feature films can be charted. On the basis of the regularity emerging from
these charts, the principles set out in chapter 6 can be tested and further
worked out.
The descriptive model provides the details of a number of substructures and
their interrelationship. Empirical research into the psychological affect struc-
ture of films may also represent an improvement in relation to attempts to
study isolated responses. Both differences between types of films and differ-
ences between viewer groups gain in significance when affect structures are
studied, rather than isolated responses. We need only think of the research
into viewer differences, such as the role of gender in emotional awareness. If
we confine ourselves to the study of emotional response, the mechanism of
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 221

the film emotion will remain undisclosed. Thus we might measure crying and
discover that men do not cry as often as women. Does this mean that the
emotional meaning structures of the two groups are the same but that female
viewers cry more easily than men, or is their appraisal of the situation dif-
ferent? If the emotional meaning structure is not the same, then does the situ-
ation model also differ, that is, do men and women see films differently? More
insight into such group differences can be obtained by comparing affect
structures instead of isolated emotional responses or film plots.
An attempt was made to initiate a description of the generic affect struc-
ture of the traditional feature film. Many more empirical affect structures
must be collected if this description is to be developed and validated. In ad-
dition to the nomothetic use of the model for general theory formation, it can
also serve ideographic goals. The special quality of a particular film can be
brought to light by means of its affect structure, while comparison with other
films remains an option.
Finally, an important function of the model of the affect structure is that
it enables us to examine certain notions familiar to anyone who not only
watches films but also talks or perhaps writes about them.
The affect structure should be capable of bringing to light a number of im-
portant viewer intuitions. The first of these is the dominance of certain char-
acteristic emotions. Well-known film genres are characterized by the type of
affect that they evoke in the first place. This is almost inherent in the desig-
nation of the various genres that are so close to the hearts of the natural au-
dience of the feature film. Thriller, horror movie, comedy, and tearjerker are
descriptions that, in their simplicity, tell us a great deal about the features of
a particular film. Thus the affect structure of a film characterized as a thriller
should correspond to that designation. We expect a particular emotion, say,
fear, to be extremely intense, forming, alongside interest, the primary emo-
tion throughout large portions of the film and to be more closely intertwined
with interest than the other emotions. Naturally, not all the known genre defi-
nitions display such a close and simple relationship with a single dominant
emotional response. And yet some genres, such as melodrama, tragedy, com-
edy, and farce, are characterized by specific affect structures, since the
course of the emotion, while perhaps more difficult to characterize, is largely
the same. One example of the affect structure of tragedy appears above; it
may be characteristic of at least a certain class of tragedies that when it
comes to phasic emotions, pity and admiration dominate the affect structure,
with pity preceding admiration, while in the text base a corresponding point
of reversal in the action can be identified. And perhaps more interestingly, it
is conceivable that there are certain affect structures that regularly occur, but
that are characteristic of films that have not yet been designated as a specific
genre.
The second intuition of the viewer that should be identifiable in the affect
222 CHAPTER 4

structure is the depth of emotion. Not only does a feature film evoke emo-
tion, it is also about something. These two features are not unrelated. The
theme of a film lends depth to the emotion; it draws the viewer into the fic-
tional world and imposes a certain vision of that world. In the affect struc-
ture, this depth may be revealed in the type of emotions, distinguished on the
basis of the content of the emotional meaning structure. For instance, depth
can be associated with psychological depth, the degree to which the emo-
tional significance of the fictional events for the fictional characters is high-
lighted. By contrast, the more objective significance of a feature film, one that
extends beyond the significance of the events for the fictional characters, may
give rise to less common emotions, emotions that accompany the discovery
of a new insight, some previously unthought of perspective, or a hitherto im-
possible vision of reality including the viewer's self. But depth may also have
to do with the exceptional intensity of emotions: the scariest horror film—no
matter how flat the theme—tends to be thought provoking. An unusual com-
bination of emotions, such as embarrassment and pleasure, can also give rise
to a sensation of depth.
A third intuition is prospectivity. Some themes are substantial, others less
so. In many traditional films, the theme is little more than an excuse for show-
ing spectacles. Even if a film is less about something, it does have a certain
direction: the action is moving toward some end state, the film narrative con-
verges with time, the text base moves toward closure of the narrative struc-
ture. The affect structure of each film or group of films shows how the vari-
ous emotions are interwoven with the course of time, how they develop and,
last but not least, how they are embedded in interest. The affect structure also
makes it clear how the episodic narrative structure is responsible for the de-
velopment and completion of interrelated emotions.
A fourth intuition that can be unearthed by the affect structure is that
every film is based on a design. In effect, this intuition encompasses the three
previous intuitions, dominance, depth, and prospectivity. No matter how the
illusion of the diegetic effect forces the viewer to see through the artefact,
one of the most important experiences of films is that the viewer is led by the
hand through the fictional world. In chapter 3, we noted that the diegetic il-
lusion in its optimum form can evoke the realization that one is the victim of
premeditated deception, or at any rate, forceful steering. For instance, sus-
pense brings with it a perceptible postponement of closure, and each sur-
prising revelation the realization that one has been misled. As we maintained
in chapter 4, interest may be distributed more or less evenly over the entire
film, or rise sharply at certain moments. In some films exciting episodes are
alternated with calmer intermezzi, while in others the viewers are barely
given the chance to catch their breath from beginning to end. Similarly, in
one film the emotional tone, that is, the nature of the dominant emotion or
mood, may fluctuate wildly during consecutive scenes, while in another it is
PSYCHOLOGICAL AFFECT STRUCTURE OF FEATURE FILM 223

distributed fairly uniformly. What is more, the intensity, nature, and dosage
of each individual emotion over time and in relation to other emotions has
been preprogrammed by the filmmaker, even though such preprogramming
is not always based on conscious reflection. There is, in any case, a plan be-
hind the film narrative, and each microproposition of the text base has a cer-
tain role to play within this plan. As a viewer one will occasionally be aware
of the inescapability of this masterplan concealed behind the series of emo-
tions that one is undergoing, although this sensation may not be as strong as
in the case of high suspense or a violent shock. As we have seen, this intu-
ition on the part of the viewer is at the root of A emotions. For instance, the
amazement and admiration elicited by the composition of the film's plot have
the artefact as their object, including both the making and the maker of the
film.
Further research will be needed to establish to what extent the filmmaker,
during preparation, shooting, and editing of a film, envisions the entire body
of emotional experience that the viewer is destined to undergo. Perhaps we
could say that the affect structure is something like the score of the psycho-
logical reactions that are timed, meted out, and orchestrated by the film-
maker. The unity of the emotional experience and the near unanimity of the
reactions that can be observed in any cinema are traceable to the design of
the film that is contained within its affect structure.
 Conclusion: The Feature Film
8 as an Emotion Machine

AI\' ASSUMPTION IN CLOSING

It would he somewhat naive to attempt to conclude a study on a subject like

the present one- the structure and determinants of the emotion experienced
by viewers of the traditional feature film-in the most direct manner, namely,
by formulating a definitive theory.
If there is one thing that has become clear in the course of the preceding
chapters, it is that the sensation occasioned by a film is a highly complex phe-
nomenon. \Vhile it might be possible to gain some understanding of the
phenomenon with the aid of available theories dealing with emotion and film
narrative, it would be a difficult task, and that understanding would be far
from complete.
In a sense, we should count ourselves lucky simply to be able to formulate
certain cohesive conclusions, in the form of a more or less comprehensive
assumption. That assumption is the tentative answer to the question we posed
at the beginning of this study: Docs someone witnessing a traditional fea-
ture film experience authentic emotion and, if so, what is the nature of that
emotion?
In the preceding chapters we have tended to tacitly assume that this ques-
tion should be answered in the affirmative, while at the same time striving to
demonstrate the plausibility of that assumption. Now, however, we are in a po-
sition to evaluate some arguments for and against.
This chapter will not deal with any facts, assumptions or issues that have

225
226 CHAPTER 4

not already been touched upon and is intended primarily as a summary of

the line of reasoning set out thus far.

On the Reality of Emotions Evoked by Films.

Do Films Evoke Genuine Emotion?
The question of whether works of art are capable of producing authentic emo-
tion has a long and complicated history. In principle, everything that has been
written about the relationship between art and reality by philosophers, art
theorists, and artists down through the ages has a bearing on the answer to
this question. Fortunately, certain restrictions can immediately be made, the
most important of which is the following. The question of the authenticity of
the emotion whose object is the fictional world is more pressing than that of
the emotion occasioned by the film as artefact. There are two reasons for this.
First, we have argued that in the case of the traditional feature film the arte-
fact as stimulus is of secondary importance. Moreover, the F emotion is prob-
lematic, because its authenticity presupposes that a certain reality is attrib-
uted to the fictional world, as we saw in chapter 3. The film as artefact, on
the other hand, is itself a reality, and as a reality it can be the object of true
emotions. For example, if I find a film disgusting, then that film is a concrete
object of my emotion. It is, of course, no easy task to clarify the active ele-
ments in the A emotion. Given the complexity of works of art, there are often
no bounds to be put on the elaboration of such aspects as the composition
and import of a literary work, a painting, or a film. Nevertheless, it is possi-
ble, at least in principle, to identify those elements, or their correlates, in the
work of art.
The illusion of the diegetic effect apparently turns things upside down: the
fictional world conveys the impression of reality, while the artefact appears
to be a far-fetched construction. When struck by the economy of a narrative,
I often find that a great many words are needed to identify the elements re-
sponsible for that feeling, although they are all right there in the art work it-
self. If I am watching Do the Right Thing (1989) and I am frightened, truly
frightened that a terrible explosion of violence is imminent, then I believe—
for the moment, at least—in the reality of the threat, in the existence of the
characters and their conflict. I fear that one false move, a word, a look, will
ignite a veritable powder keg of hate. It is as if I can reach out and touch pizze-
ria boss Sal his deliveryman Mooky, Radio Raheem, and the Lord Mayor far
more easily, in fact, than filmic elements of the artefact, such as actors. But
this is not so. The actors have an indisputable existence, whereas Sal and the
other characters are fictional representations, and I know this. Is it still pos-
sible that what I feel is actually an emotion? Or am I pretending? Do I believe
that what I am experiencing is a true emotion, even though there are good
reasons why a researcher would deny this?
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 227

The need to establish to what extent the viewer experiences the fictional
world as real, and how authentic the F emotion is, takes on added urgency
when we realize that these questions are vital elements of the explanation of
A emotions. The issue of whether an art work is felt to refer to a reality out-
side itself, how that reference has been shaped, and whether the feelings as-
sociated with reality are false or not, determines to a considerable extent the
A emotion of the viewer.
In attempting to answer our question, we will start by examining a num-
ber of recent studies dealing with the problem of the authenticity of film
emotion.

Recent Answers
Some of the most recent contributions to the discussion on the illusion of
reality called up by feature films and the accompanying emotion are by Allen
(1993), Carroll (1988b, 1990), Peters (1989), and Smith (1995). For a some-
what broader, but also recent, overview of theory and research dealing with
emotion and works of art, see Schram and Frijda (1995).
Allen (1993) distinguishes between a reproductive and a projective vari-
ant of the illusion that is evoked by films. In the case of reproductive illusion,
the viewer mistakenly sees a staged event as real. This illusion is seldom
evoked by the watching of feature films, which invariably contain enough
cues in order for them not to be taken for documentaries. In the projective
illusion viewers lose the awareness that they are watching a film: "Rather than
look through the image 'from the outside' at a photographic reproduction of
something staged in this world, you perceive the events of the film directly
or 'from within'" (Allen, 1993, p. 40). The projective illusion coincides with
what—following Burch—we have referred to here as the diegetic effect (Ch.
3). The awareness of the photographic basis of the image is swept away by a
combination of movement, sound, and the projection of a fictional space in
which we appear to find ourselves. Allen argues that this illusion is not always
experienced; viewers need not necessarily go along with the projection if they
do not choose to. In a typical case, the viewer will shift back and forth be-
tween the projective illusion and medium awareness. 124 These two modes,
however, cannot occur simultaneously.
Peters (1989) takes as his starting point Sartre's ideas on the relationship
between mental imagination and emotion, which were touched upon briefly
in chapter 1. According to Sartre, an imaginary image can never bring about
true emotion. It may in some cases reinforce an existing emotion, but the ab-
sence or irreality of the object portrayed prevents any real feelings from aris-
124
Within the projective illusion, Allen also distinguishes a "character-centered" and a "spectator-
centered" mode. These coincide roughly with what we have referred to here as empathetic and
nonempathetic viewer attitudes.
228 CHAPTER 4

ing. In one's imagination, the most one can do is to imitate the emotion that
would be evoked by the object if it were actually present. According to Peters,
the material image, such as the film image, works in the same way as the
imaginary one. The viewer reacts strongly to the film image to the extent that
the feeling is already there, as a constituent layer. In viewing feature films we
play-act emotions, especially those we share with certain characters. Peters
sees this play-acting of emotions as one of the major sources of entertain-
ment. Acting as if one is experiencing an emotion, in a pretense play con-
trolled by the film, also explains why viewers find sad or gruesome scenes so
satisfying: these scenes make one act the emotion of fear.
Carroll (1988b) stressed that we only speak of an illusion when viewers
truly believe that what they are seeing is reality. That is never actually the
case. But viewers can be struck by a remarkable similarity between x in the
film and y in reality, for instance, between the flat, make-believe houses on
the film set and real houses. If we speak of an illusion here, then we are not
using the word in its literal sense, which implies that the viewer is the victim
of a deception and actually believes in the reality of the image. Carroll (1990)
examined the relationship between the illusion of reality and the emotion felt
by the viewer. Three theoretical positions are sketched, each of which pro-
vides a clarification of the authenticity of the emotion evoked by feature
films, indeed, by fiction in general. These are referred to as the illusion the-
ory, the pretend theory, and the thought theory. As the name suggests, the
illusion theory of fiction assumes that the viewer sees the film scene as real-
ity.125 This is unlikely, because the institute of fiction is so firmly and mas-
sively established in our culture that total deception is all but impossible.
Everyone is familiar with the signs and the conventions of fiction, as well as
the framework within which it is presented. In addition, Carroll marshaled a
number of objections to Coleridge's well-known variant of the "willing sus-
pension of disbelief." The one that I consider the most important is of a psy-
chological nature, namely, the fact that one cannot resolve to believe or dis-
believe something. Belief is either there, or it is not. Moreover, as Carroll said,
it is difficult to explain how the appropriate reaction to fiction comes to be
displayed, for example, a gasp or shudder at the height of a horror movie,
without at least some measure of belief on the part of the viewer. Carroll went
on to discuss what he called the pretend theory of fictional response. It pro-
poses that viewers do not believe in the reality of the fiction, and that, more-
over, their reaction is not a genuine emotion. Viewers may, of course, think
that their reaction is an emotion, in which case there is obviously no real dif-
ference between what they feel and a real emotion. On the other hand, if view-
ers do not believe that the sensation they feel is an authentic emotion, then

125
Note that the illusion theory supposes a belief in the reality of the scene. It goes beyond a
mere "magic window effect," discussed in chap. 3.
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 229

the pretend theory corresponds to the imitation theory put forward by Pe-
ters. Peters believed that viewers cannot be under the impression that they
are experiencing a real emotion, because he finds the imitation emotion
agreeable, and it is difficult to conceive that a commonly induced sensation
like disgust, the authenticity of which is unquestioned, can be considered
pleasant. Carroll described in some detail how a make-believe game could
lead to quasi-emotion. His objection to the pretend theory is that it is not in
keeping with the phenomenology of the emotion evoked by fiction. The emo-
tional response is unequivocally registered by the subject and can easily be
observed by others. The third possibility, the thought theory, is presented by
Carroll as the solution to the dilemma. According to this theory, not only a
physical reality is capable of giving rise to true emotion. A mental represen-
tation, an insight, or an idea—in short, a thought—can also trigger a com-
plete and genuine emotion. The viewer need not believe that Dracula actually
exists or that the monster portrayed on the screen is physically present in the
same room: the thought alone is sufficient to produce an emotion. The fea-
ture film imposes thoughts that in turn, evoke emotions. The thought theory
combines a number of plausible features: first, viewers are not completely
fooled by the illusion of reality presented by films; second, authentic emo-
tions are actually experienced, not least from the standpoint of the viewer;
and third, emotions can spring from imagination.
The arguments of Smith (1995) extend Carroll's analysis. He examines in
detail Brecht's view that traditional theater and film drama arouses in the
viewer a seamless illusion of reality, whereby the latter is able to lose himself
in the F emotion. According to Smith, one of Brecht's premises is that the
spectator sees the theatrical representation of reality as reality itself. This is
clearly impossible, says Smith, in the first place because an appropriate re-
sponse to art requires us to be constantly aware that what we see before us
is a representation of reality, rather than reality itself. The response in ques-
tion is action-inhibited, which, by the way, is one of Carroll's arguments
against the pretend theory. This argument corresponds to Carroll's objection
to the illusion theory, which says that fiction is an established institute. In the
second place, a work of art can strike the viewer as realistic only if it is simul-
taneously recognized as a representation. Smith then discusses Radford's in-
fluential proposition, which says in essence that emotional reactions to art
are quite simply incoherent, unintelligible, and absurd, because if these re-
actions are truly emotional, they require a belief in the reality of the repre-
sentation, while at the same time the viewer is conscious that the opposite is
true. Smith, on the other hand, believes that there is a solution to the
dilemma, the same one advocated by Carroll with his thought theory. In the
case of art, Smith says that we imagine in our mind's eye that the object ex-
ists, and it is this that opens the way for real emotion. The awareness of the
artefact is included in the theory by virtue of the fact that the work of art en-
230 CHAPTER 4

ables the viewer to form a representation, while that process takes place by
conventional means, which are recognized as such by the viewer. The mime-
sis that is realized by the film calls upon both real world knowledge and the
historically determined conventions of the narrative. The viewer envisions a
more or less realistic world, which extends beyond what is shown at any given
moment. Here we have the diegetic effect again, but in this case it is bounded
by a narrational agency, which is experienced as such by the viewer. Viewers
make their own contribution to the simulation of reality.126
If we examine the various approaches to the problem of the authenticity
of F emotions, the first thing we can say is that the possibility that viewers
truly believe that they are seeing reality, rather than a staged representation,
can be rejected altogether. Viewers believe no more that they are seeing
Napoleon himself in the reality of everyday life, than Abel Gance's actors be-
lieved that they were actually Napoleon.127 Film creates an illusion, not a delu-
sion. The term illusion is used in its "epistemologically benign meaning"
(Carroll, 1988b): viewers know full well that they are watching a staged and
projected representation.
Second, there is considerable disagreement among the various authors con-
cerning the nature of emotion, and in this respect they are perhaps represen-
tative of the larger forum of film scholars. Most researchers tend to consider
viewer emotion as authentic, just as the majority of the natural audience of
the feature film. However, it is intriguing that the assumed authenticity of the
emotion is not based on an analysis of film characteristics that make it inevi-
table that one should experience a true emotion. Ultimately, the researchers,
like the viewers, are forced to fall back upon an introspective assessment of
their own experiences, supplemented perhaps by observation of their own or
other people's behavior interpreted as emotional. Carroll rejected the pre-
tend theory because it is consistent neither with his own experience nor, he
assumed, with that of any other viewer. In this connection, he recalled his
own experience during The Exorcist (1973): "But I, at least, recall being gen-
uinely horrified by the film. I don't think I was pretending; and the degree to

126
A similar view of mimesis in literature is also to be found in Oatley (1995). The reader in-
tegrates and constructs a reality and is able to either enter into the world of the text or take up
a position opposite the text.
127
An interesting point in this connection is the parallel between the representation of real-
ity and the nature of the emotion felt by the actor and the spectator. Konijn (1994) recently
demonstrated that, contrary to the dictum of the Stanislavsky method, the emotion displayed
by actors is almost never the emotion that one would expect the character to have. Instead, ac-
tors apparently experience primarily the emotions involved in the execution of a demanding
task, namely, the credible portrayal of their character for the benefit of an audience. The dom-
inant emotions were concentration, challenge, and "warm, generous" emotions, that is, the emo-
tions of an actor at work. An obvious comparison presents itself with the dominant role of inter-
est among viewers. The feelings of the viewers are based not only on what befalls the character
in the fictional world but also on what they expect as witnesses of events in that world.
CONCLUSION: FEATURE FILM A S AN EMOTION MACHINE 231

which I was shaken by the film was visibly apparent to the person with whom
I saw the film" (Carroll, 1990, p. 74).
The authenticity of the emotion evoked by films can be demonstrated by
means of a number of simple introspective observations with which most
viewers will concur. While films are not the exclusive domain of humor, there
are perhaps few situations in daily life more likely to give rise to hilarity than
those in a film comedy, unless it is the situation where one person deliber-
ately tries to get another to laugh (an example that is not that far removed
from the film comedy). This is true not only of comic elements but also of
those that are alarming or titillating: presented in filmic form, they give rise
to emotion. Now not all emotions are equally popular with all film viewers.
This means that introspection may be insufficient, since most people delib-
erately avoid certain types of films, precisely because these films are judged
likely to evoke some unwanted emotion. There are undoubtedly horror films
that are studiously avoided by a large part of the population because they be-
lieve—and rightly so—that they are too scary, scarier than any reality en-
countered in this prosperous, well-run, secular welfare state of ours. And al-
though sex films may be regarded as an inferior substitute for the real thing
or a threat to women, and therefore avoided, this is decidedly not because
they fail to titillate. In general then, it is quite possible that certain films are
shunned by some people and devoured by others for precisely the same rea-
son, namely, that emotionally speaking, they are "larger than life."
Survey research into the emotional response to films that focuses on the
subjects' assessment of their own emotion (referred to in chap. 7) also indi-
cates that films are just as capable of serving as emotional stimuli as situa-
tions in everyday life. The fact that films are capable of arousing genuine emo-
tion can be demonstrated even more simply by observing the reactions of a
cinema audience, to which we referred in the first chapter of this book and
which are obvious not only to Carroll's companion. And finally, the physi-
ological response, which has been measured in a number of studies (Ch. 7)
and is therefore not dependent on introspection, leads to the same conclu-
sion: films evoke emotion and that emotion can be observed externally as well
as internally.
Both introspection—systematized informally in questionnaire form or as
verbal protocol research—and observation serve as evidence for the authen-
ticity of film emotion by virtue of the fact that we have a definition of au-
thentic emotion: a conscious, cognitive experience (such as fear, excite-
ment), combined with a particular behavior (laughing, crying, the shivers),
and certain physical reactions (such as galvanic skin responses, altered heart
rate, and pupil diameter changes). Such a definition is sufficient for several
reasons. To begin with, there is considerable agreement among emotion re-
searchers concerning these three characteristics. Moreover, they are proba-
bly shared by most representatives of both European and American culture.
232 CHAPTER 4

Therefore, this description of emotion as a combination of conscious experi-

ence, emotional behavior, and physical reactions can be regarded as a kind
of common-sense definition of emotion. On the basis of this definition, it is
possible to maintain that feature films do indeed evoke true emotions. But
there is more to be said on the subject.
First, the definition is circular. It implies that film viewers experience real
emotion, because they say that they experience emotion, an assessment
based implicitly on our common-sense definition. This is an argument that
both Carroll and Smith make use of. In their view, arguments that deny the
authenticity of film emotion are not in accordance with the phenomenology
of the experience. But this phenomenology rests on a definition, rendering
the argument circular.
In the second place, we would be more justified in calling the emotion
evoked by films real emotion if it were more in keeping with a condition that
is referred to as emotional in settings quite different from that of watching a
film in a movie theater. In other words, what we are looking for is an explicit
description that is not arbitrary, and that in a highly general sense distin-
guishes emotion from non-emotional states. The functional perspective would
appear to make possible such a definition.

Functional Aspects as Characteristics

of an Authentic Emotion
In our discussion of the emotion experienced by the film viewer, we opted
for a functional view of the emotions (chap. 3). Current general theory is
based on the idea that emotion developed in an evolutionary manner, from
rigid sensorimotor reflexes into a complex of flexible responses to adaption
requirements from the environment (Lazarus, 1991). Although cognition and
learning processes have become major determining factors in emotion, the
function of emotion has remained the steering of action and the processing
of information. According to de Sousa (1987), emotion can overcome the lim-
itations of reason by restricting—if only temporarily—the supply of informa-
tion on which the organism relies. In effect, the emotion system imitates the
perceptual apparatus. The perceptual apparatus consists of modules that are
informationally encapsulated, 128 in other words, they are only receptive to a
certain type of perceptual information, remaining inaccessible to other in-
formation. Their operation is automatic and impenetrable for other cognitive
operations. 129 We observe color and depth, whether we want to or not, and
regardless of cultural and ideological influences. The process is always com-

128
De Sousa referred to the well-known work done on the modularity of the mind by Fodor
(1983).
129
The term is borrowed from the expression "cognitive impenetrability" introduced by
Pylyshyn (1984).
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 233

pleted, from beginning to end: it is a question of all or nothing at all. In De

Sousa's view, the rationality of emotion lies in its ability to guide reason. (This
view does not imply that emotions always lead to the right insights; quite often
these are counterproductive, as most of us are all too aware.)
An emotion as state is a manifestation of the emotion as function and once
evoked is characterized to a greater or lesser degree by the qualities of steer-
ing, automatism, and cognitive impenetrability. Frijda's theory of emotions is
a concrete reflection of this functional perspective, as we saw in chapter 3, be-
cause it represents emotion as a concern realization system (Frijda, 1986;
1988). We recall here in brief the main functional characteristics of emotion:

1. Control precedence. A genuine emotion displays control precedence over

other processes already taking place.
2. Dependence on concerns. Where there is an emotion, there is a concern.
There is no emotion without an identifiable concern. (Law of Concern)
3. Stimulus specificity. Each emotion has a specific situational meaning
structure that reflects the emotionally critical characteristics of the stim-
ulus. (Law of Situational Meaning)
4. Response to reality. The stimulus must represent some reality or other, if
we are to speak of an emotion. (Law of Apparent Reality)
5. Response to change. Emotion responds above all to changes in the situ-
ation. (Law of Change)
6. Closure. An emotion has a strong tendency toward complete realization
of its appraisal and action tendency. It is relatively immune to outside in-
fluences, such as conscious control. (Law of Closure)
Psychological states that meet these criteria can now be characterized as
real emotions. The functional criteria differentiate better between the vari-
ous sensations than the commonsense definition. In the first place, and in
contrast to the commonsense notion, conscious experience is not itself suf-
ficient proof of an authentic emotion. If the emotional stimulus or the con-
cern is lacking, then the individual may report an emotion that the researcher
cannot confirm. Conversely, individuals may not report any emotion, in spite
of the fact that their state meets all the functional criteria. Emotions may
catch us unaware, they "can come as thieves in the night." (Frijda, 1986,
p. 337). In some cases awareness lags behind the emotional appraisal and the
response. Emotional behavior can take many different forms. As we know,
the action tendency consists of a strong inclination to make some change in
the object of the emotion. The details of the plan of action depend on the cir-
cumstances; only the so-called basic emotions (fear, anger, sorrow, etc.) are
characterized by well-defined action tendencies. But films also evoke other
emotions, combinations of basic emotions and more subtle feelings. A num-
ber of well-known anecdotes notwithstanding, viewers are more likely to watch
234 CHAPTER 4

the scenes being played out on the screen without any sign of overt action
than they are to intervene or display some other form of specific action. In
this connection, it is interesting to note that Gunning (1989, in Allen, 1993)
refuted the myth according to which the audience of L'arrivee d'un train en
gare de La Ciotat (1895) flinched and ducked behind their seats, the anec-
dote with which we opened this book! Bodily processes—actual physiological
changes—may also arise earlier or later than other components of the emo-
tional process. Moreover, important individual differences have been re-
ported in the physiology of emotion, and it is by no means certain that each
emotion has its own unique pattern of physiological changes. Thus the total
absence of physiological reaction, or of certain physiological reactions, does
not necessarily mean that there is no real emotion there. Conversely, a physi-
ological reaction, say, a heart rate deceleration, is not in itself sufficient proof
of an emotion. This may be related to other factors, such as the level of at-
tention, that is, a purely cold cognitive state. As we have seen, the emotional
core components of the situation must be evident, and there must be a con-
cern at stake. In sum, the functional definition outperforms the common-
sense notion of emotion in distinguishing emotional from nonemotional
states.
What then are the consequences of a functional definition of emotion for
the various theoretical positions on the authenticity of film emotion? The il-
lusion theory holds that the viewer has genuine emotions. The epistemolog-
ically malign version of illusion, (i.e., deception), maintains that fiction is
seen as reality, and according to the law of apparent reality—with the em-
phasis on apparent—this means that the viewer does indeed experience emo-
tion. But, as we have seen, there is no one who would venture to say that the
viewer sees the fictional world as totally real, without any reservations. The
principle of control precedence predicts that in the case of a delusion, view-
ers will be gripped by emotion: if that emotion is fear, they will jump up and
run out of the cinema, if it is anger, then they will go charging off in the di-
rection of the projection of, say, pizza boss Sal. Such reactions may or may
not have occurred in 1895, but today they are few and far between. And a
good thing, too, since otherwise we would have been saddled with Radford's
desperate paradox. Although the illusion theory can be written off without
too much trouble, it must be stressed that the fact that in an objective sense
the film fiction consists of cardboard houses, liberal dousings of ketchup, and
actors hopping around in velociraptor suits does not necessarily affect the
emotion as seen from a psychological standpoint. If the fiction has any kind
of reality at all, then emotion is there (Frijda, 1988, 1989). A specification of
the concept of illusion brings the first position concerning the phenomenon
close to the third, thought theory.
The pretend theory holds that the film viewer imitates emotion. Thus from
a functional standpoint, and in keeping with the name of the theory, imita-
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 235

tions can never be authentic emotions. There is no real and pressing concern
at stake for the subject himself. (See also the discussion of the concept of
"identification" in chap. 6.) Although I may act as if I am petrified by the
prospect of an ultimate revolt in Do the Right Thing, where everyone will be
battling everyone else, I am only pretending that I am in the situation where
the irreversible escalation is impending. In reality, I know that I am sitting in
a movie theater far away from the melee and that my safety is in no way un-
der threat. But the pretend theory cannot be disposed of so easily. As the imi-
tation meets more conditions of apparent reality, it approaches more closely
the thought theory. As Carroll rightly maintained, thoughts are capable of
producing emotion, and thus—if they have some reality for the subject, we
would add—a thought representation of the fictional world can in some way
or other actually evoke genuine emotion. The image of the big city as a pow-
der keg of ethnic contradictions may well challenge my concern that all is
safe within my own little world. The mimetic fantasy of a world that I view as
possible evokes a genuine emotion, say, alarm. The simple fact that I find it
difficult to shake off this alarming possibility testifies to the genuineness of
emotion. The law of closure and the principle of control precedence appar-
ently come into play as soon as I start believing in a particular possibility.
This brings us to the thought theory, which is totally in accord with the
functional standpoint: thoughts can have some kind of apparent reality, and
they can be related to concerns. Frijda (1989) underscored the fact that the
simple perception of a possibility represents a form of apparent reality. In
chapter 5 we saw that people cherish certain possible representations, say,
the romantic view of life, as a form of truth that in everyday life is hidden
from sight. The film enables us to see these representations come true. Fol-
lowing De Sousa (1987) and Carroll (1990), we can also point to the evolu-
tionary value of a link between fantasy and affect. It is advantageous to ex-
perience the terrifying or pleasurable side of certain events and to undergo
the fear or longing before these events actually take place. Moreover, from a
somewhat more Freudian standpoint, it is understandable that fantasy follows
concerns. 130 When the world temporarily makes no urgent demands upon us,
it may be quite pleasurable and advantageous when the free play of our imagi-
nation is related to ever-present but latent needs and desires. Plans may
emerge or connections become apparent that are prevented from occurring
in the course of goal-directed action. In the light of the functional nature of
emotion, it is logical that concerns can be touched upon, and thus activated,
by free thoughts. Concerns may be dormant, but they are never completely
eliminated. For if they were, how would we know that a situation requires us
to act or to process information in a particular way?

130
For a more recent discussion of relevant theory and research, see especially Klinger
(1977).
236 CHAPTER 4

In our representation of matters, concerns are the chords upon which fea-
ture films play, through the action of guided fantasy. Frijda's theory implies
that concerns are always there, no matter what the situation. Watching fea-
ture films is not a disinterested pastime, as emphasized by Kantian theories
on the experience of beauty. People watch feature films because of the op-
portunity they offer to realize concerns, although viewers are not necessarily
aware of this reason.
Thus the fantasy called up by films gives rise to emotion. If such is the
case, then not only apparent reality and relevance for concerns, but also the
other functional features must be present in the emotion of the viewer. We
now describe what the film does to the viewer, whereby the main points dis-
cussed in the previous chapters will be recalled in brief.

The Film Stimulus as Source of Illusions

The feature film may be seen as a conductor of fantasy, which—in keep-
ing with the thought theory—produces emotions. How exactly does this take
place? From a psychological point of view, the action of watching a feature
film is characterized above all by the fact that the viewer cherishes a number
of illusions, in Carroll's "epistemologically benign" sense. The viewer knows
that what he or she is seeing is an artefact, one that is related to reality in a
more or less complex manner. My term illusion encompasses not only un-
avoidable effects but also those from which it is possible, albeit with great ef-
fort, to escape. Furthermore, the fact that we are speaking here of illusions
does not mean that the viewer plays a passive role in the film experience. On
the contrary, the previous chapters have demonstrated that active involve-
ment in the film narrative is essential for an adequate, that is, entertaining
experience. However, as we will see, the contribution of the viewer to the
guided fantasy of film viewing is led by illusions along highly structured paths.

The Illusion of Motion. The first illusion, the apparent motion that we
see in what are actually a series of static frames, is inevitable. This illusion is
part of Allen's reproductive illusion, but is somewhat more subtle. We know
that we are watching a film. We see a projection on a flat screen, the defini-
tion and perspective of which are fixed. In spite of these limitations, we see
motion exactly as it is seen in the real world; moreover, this impression is un-
avoidable. The precise explanation of this purely perceptual illusion has still
not been found, although it was discovered over a century ago (Hochberg,
1986). One thing is certain: the elementary aspect of the film stimulus in-
volved here is the result of a module, which may consist of various compo-
nents. Movshon (1990; Sekuler, Anstis, Braddick, Brandt, Movshon, & Orban,
1990) provided an overview of the evidence for a neuroanatomically defined
motion pathway in the cerebral cortex. Crucial components of the illusion-
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 237

ary observation of motion appear to be just as cognitively impenetrable as

those responsible for the perception of depth and color. Seen in isolation, the
contribution of this illusion to emotion is quite modest. Although it con-
tributes to a major component of the emotional situational meaning, namely,
the observed reality of the situations and events portrayed in the feature film,
in this respect it is not unique.

The Illusion of the Diegetic Effect The next illusion to be discussed is

what we have referred to as the diegetic effect131 and what Allen calls the pro-
jective illusion. The viewer is present "in imagine " in the fictional world be-
ing portrayed. This second illusion is of greater importance for the emotion-
producing effect of the feature film than the first and is difficult if not im-
possible to resist, despite Allen's claim to the contrary. Only a trained film
analyst would be consistently capable of seeing the various shots as shots and
the changes in perspective as so many camera displacements. And even then
it would be impossible to suppress the illusion of motion in space by telling
ourselves that what we are seeing is actually motion across a flat surface. 132
To clinch our argument, we need only point out that the previously men-
tioned method of coding the stimulus is taken from the theory of film analy-
sis and is totally foreign to the natural audience of the feature film.
As we saw in chapter 3, the diegetic effect may be regarded as a compre-
hensive illusion. It can encompass all the effects of the feature film as narra-
tive. By narrative we mean the process that mediates between, on the one
hand, the fictional world as a reality, which apparently exists a priori and, on
the other hand, the imaginary spectator, the representative of the viewer in
that fictional world. Not only is the impression of actually being present in
the fictional world impossible to resist, but—and here we have the second
component of the diegetic effect—that world is itself experienced as an ob-
jective reality. Film theory provides the illuminating phenomenological de-
scription of the diegetic world as one that is entirely separate from the nar-
rative operations that make it accessible to the viewer. The film narrative is
a frame or window that provides access to a self-contained world, a world that
existed before one gained access to it.
Occasionally, however, the viewer is shaken out of the illusion by the fact

131
A terminological note is appropriate here. The word diegesis is traditionally used for a form
of verbal narration. There the opposite of diegesis is mimesis, the portrayal or simulation of a re-
ality by means of images or staging. From now on, we will use mimesis in this latter sense, so that
it is actually more closely related to the diegetic than is generally the case. This goes back to the
use of the term diegetic effect by Burch, who is referring here to the quasi-sensory experience.
132
There are, of course, exceptions that prove the rule. As a result of mistakes (rare) or artis-
tic intention (less rare), there may be a lack of depth cues, so that movement is indeed perceived
as taking place against a flat surface. This reverse illusion is described by Arnheim (1933).
238 CHAPTER 4

that the artefact itself comes to the fore. Unlike Allen (1993), I maintain that
the viewer does not move freely back and forth between diegetic effect and
medium awareness, but rather that this motion is regulated by the film. For
instance, during scene transitions the diegetic effect may make way for an
awareness of the medium, while it also tends to b e somewhat weaker at the
beginning of the film or in the epilogue. At such moments, the narrative gains
in self-consciousness (Bordwell, 1985). In accord with the observations of
Smith, the viewer is aware of the artefact, but it must be added that this self-
consciousness does not arise at arbitrary points in the film. Furthermore, the
conventionality of what I see as medium interruptions limits their influence
on the illusion.
Within the diegetic effect, the artefact may give rise to all sorts of enjoyable
effects. Repetition and variation, such as patterns of staging and framing, are
enjoyed by the viewer without any necessity to step outside the fictional
world. I prefer to locate these effects of the artefact within the diegetic effect,
as they present themselves primarily as patterns of events in the fictional
world, events that can be traced to the irreducible objective logic that reigns
there. Shots of the front of Sal's pizza shop in Do the Right Thing before and
after the fire display both similarities and differences that are ingeniously
constructed, thus revealing the artefact. But the formal and artistic relation-
ship between the shots is concealed by a causal one. In these shots, repeti-
tion and variation in camera position and framing are encoded as views or
ways of seeing on the part of the invisible witness. Narrative motivation takes
precedence over artistic motivation, and where the viewer is concerned, it is
the diegetic effect that takes precedence.
Frijda's description of objectivity as a core component of the emotional sit-
uational meaning corresponds to the previously mentioned essential charac-
teristic of the fictional world. "The situation is 'spontaneous'; the subject ex-
periences himself as affected and passive in this regard: He does not confer
meaning. The sense of being overcome by the event as well as by one's own
response is the reflexive counterpart of the situation's apparent objectivity"
(Frijda, 1986, p. 205). The experience of objectivity and reality is made pos-
sible by the fact that the representation of each of three main elements of
everyday reality—physical, psychological, and social features—displays a
striking consistency. An additional factor is the link with prototypical and
stereotypical cognitive structures or some naive form of physics, psychology,
and sociology that the natural audience of the traditional feature film may be
expected to possess (see chaps 3 and 6). In my view this link between the
narrative and the naive science over which the viewer disposes is an impor-
tant addition to the thought theory of Carroll. While thoughts and fantasies
can give rise to emotions, not every thought or every fantasy—given that it
addresses a concern—is equally capable of evoking emotion. In fiction, a pos-
sibility is created, but the degree to which it is convincing is for the most part
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 239

related to its link with a more familiar reality.133 Naive laws of cause and ef-
fect in each of the three areas are at least respected and, in some cases, given
extra prominence. Naive physics, for example, is justified by the apparent con-
tinuity of the action, with respect to place and time, as well as the actual ac-
tion. Within one and the same scene, in any case, this continuity creates the
illusion of a continuous time, while what the viewer is actually watching is a
montage of discrete shots, all taken at different times and in different loca-
tions. The medium of film is particularly strong in the simulation of situations
in which the continuity of time and space plays an important role. In sus-
pense scenes the emotion is often maximized by the fact that there is a one-
to-one relationship between the proximity of a fatal threat and an advancing
physical process: the seconds ticking away in a detonator, the fuse slowly
burning down, or the inching up of water in an enclosed space. The mimesis
of magic-window reality is a precondition for emotion. There is a direct rela-
tionship not only between medium and reality but also between core compo-
nents of the situational meaning structure and the cinematic representation.
In this respect, the description in words ("The fire came closer and
closer . . .") may well be a less effective emotional stimulus. Naive psychol-
ogy, which is based on the visibility of good and evil dispositions, leads to a
considerable degree of experienced objectivity with respect to character
traits (See chap. 6.) The farmers in Bad Day at Black Rock (1955) are ob-
jectively experienced as baddies because it suddenly becomes abundantly
clear that they are out to bully to death the one-armed sheriff, played by
Spencer Tracy. Naive sociology, finally, also contributes to the self-evidence
of the fiction (see chap. 6). The moderate African Americans in Do the Right
Thing, such as the Lord Mayor and his old flame, are portrayed as old and
wise, but more radical viewers may be inclined to see them as Uncle Tom fig-
ures, a categorization that is also sustained by the narrative. The filmic mime-
sis ensures that what happens there in that world, physically, or inside and
between people, appears to conform to logic. A logic that cannot be influ-
enced but is easy to understand and accept. This logic is determined by the
narrative and what is seen as its natural control over the action. The logic of
and the control over the action are experienced as originating in the fictional
world, rather than as resulting from the operation of some narrative agency.

The Illusion of the Controlled Witness. The special position of the

viewer within the fictional world and the action taking place in that world may
133
In this connection, it should be noted that the experience of reality by the viewer is not a
question of all-or-nothing-at-all. The largely philosophical view of the genuineness of emotion put
forward by Carroll and Smith means that our discussion maintains an excessively categorical dis-
tinction between appearance and reality, van Vliet (1991) discussed from a psychological stand-
point the possibility of different degrees of modes of reality, measured against a "paramount re-
ality," which corresponds to what I refer to here as sensory experience of the real world.
240 CHAPTER 4

be seen as the second subillusion of the diegetic effect. Viewers have the im-
pression not only that they are present in the fictional world, but also that
they are present in the guise of an invisible witness or spectator. There is an
inevitability about this illusion, because the images all but exclude all other
interpretations. The events do not befall the spectator in the flesh nor do they
affect him or her as they would one of the actors in the fictional narrative.
The apparent presence of the viewer in the fictional world has—or so it ap-
pears—no consequences for that world. It follows logically from this illusion
that one feels compelled to adopt the attitude of an observer with respect to
events in the fictional world (see chap. 6).
It may be helpful to draw a comparison with situations in everyday life that
impose upon the individual the role of spectator. We might think here of cer-
emonies or court sessions. The definition of the situation that invites one to
take on the role of witness also extends to less formal examples, such as an
accident where sufficient assistance has already reached the scene or a fist-
fight where the two opponents have already been separated by others. Or sup-
pose that I am in the middle of a rather boring meeting and I suddenly find
myself observing the other participants. What are they really like? And what
are they really after? And then there are the situations where one is physi-
cally incapable of intervention and thus is doomed to stand by and watch. To
take a fictional example, there is Miss McGillicuddy, a friend of Miss Marple's
in Murder She Said (1962). While sitting in a train, she sees another train
approaching from behind along a parallel track. Suddenly, as the two trains
pass at close quarters, a shade in front of a window in the other train shoots
up, apparently by accident, and she looks straight into a compartment, not
three feet away from her. Is someone actually being murdered? 134 And there
are more authentic examples: much of what we see on television, from the
latest news from the Middle East to the proverbial quiz, imposes on us the
role of spectator or witness. In all these cases, we are drawn into a state of
what Metz (1975b) referred to as overreception: it is precisely because there
is nothing we can do that we are so eminently able to actively follow, that is,
observe, the events taking place on the screen.
The inability to take action is an inevitable part of the illusion of the con-
trolled spectator. Active observation and anticipation, on the other hand, re-
quires the cooperation of the viewer, what we referred to earlier as invest-
ments. The invitation to actively participate is one that cannot easily be re-
fused, as I hope was made clear in our discussion of the principles of inter-
est (Gh. 4). The fact that one is unable to actually do anything is of crucial
importance to the emotion evoked by feature films. In chapter 6 we stressed
that the empathetic emotions—for example, sympathy, antipathy, pity, and
admiration—that accompany the observational stance of the viewer are ac-

134
This example was also used by Frijda (1989) to illustrate a characteristic of aesthetic
emotion.
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 241

companied by a virtual action tendency. And the experience of the action ten-
dency determines to a large extent the emotional experience as a whole. In
Le boucher (1969), you wish that you were able to look around the corner—
with the emphasis on able to—to where the psychopathic killer is lying in wait
in the darkened school for Stephane Audran to arrive.
The illusion of the position of invisible spectator places maximum and min-
imum limits on the emotion. On the one hand, it protects viewers from an ex-
cess of emotion, which is in the service of their concern for safety. Freud
(1942)—and many before and after him—noted that spectators at dramatic
productions are always aware that they are not themselves the victim, that it
is someone else doing the suffering. This realization is thought to be essen-
tial to the dramatic effect. Because one can do nothing, and one is not ad-
dressed by the actors, it is possible to witness suffering without experiencing
that conflict with one's sense of responsibility that overcomes someone who
is standing on the shore and suddenly realizes that a person is drowning.
On the other hand, the precise nature of the diegetic effect also defines
the minimum intensity of the emotion. There is clearly another side to view-
ers' illusion that they are doomed to dolce far niente. They are at the mercy
of the narrative and must undergo each and every event that occurs, pleas-
ant or otherwise, expected or unexpected. While in the everyday world one
is at liberty to take a detour in order to avoid a street riot or an ambulance
with shrieking siren, in the feature film the imaginary witness heads straight
for them, to the alarm of the viewer. The narrative creates complications, that
is, cognitive and affective problems, bringing difficulty into the situation of
the viewer, Frijda's term for this component of the situational meaning struc-
ture. Viewers are drawn into this difficulty, in spite of their position as spec-
tator. In other words, that position is not gratuitous.
The viewer is represented in the fictional world by a sensitive and intelli-
gent observer, who nevertheless lacks command of any effectors. It is above
all the lack of normal control over the motor system, which is crucial to the
film experience. We can elucidate the illusion by means of a slightly grisly
metaphor: 135 the subject in fiction is actually a head without a body, which is
placed in a cart by an obliging assistant and wheeled—or even flown—around
through the time and space of the fictional world. The assistant has been de-
scribed earlier as the "editorial intelligence" behind the narration (see chap.
3). The assistant knows exactly what the viewer should see, if the latter is to
continue to understand the plot. The assistant knows the temporal and spa-
tial ins and outs of the fictional world. As a result—and because the spectator
is so limited in his or her possibilities—the assistant is in a position to ma-
nipulate the viewer. He—let us assume that he is a male—can provide infor-
135
The metaphor is an extension of an idea launched by Gibson, who sees the camera in the
fictional world as a substitute for the head of the viewing subject in the real world. Our metaphor
also bears similarity with Kubovy's (1986) "disembodied eye," the viewer's mobile mind's eye in
mental imagery.
242 CHAPTER 4

mation in careful doses and in various sequences; he can select and stress
certain events. He can highlight the importance of situations for a certain
character; he can even fade into the background or take on the role of medi-
ator between the fictional world and the spectator.
The stream of thoughts, which for a full ninety minutes engulfs the film
viewer, is to some extent imposed on him. However, the illusion of the con-
trolled spectator does not necessarily mean that one is controlled from the
outside. Although the assistant mediates between the invisible spectator and
the fictional world, for the viewer it is as if the assistant, like the invisible spec-
tator himself, is part of the fictional world. On this point, I may possibly dif-
fer with Smith, for whom a "sense of narration guiding our attention and set-
ting up flexible 'epistemological boundaries' is part of our experience of all
types of representational texts—including 'classic realist' ones, with their pur-
ported 'transparency' of style" (Smith, p. 43, 1995). If I understand this pas-
sage correctly, Smith placed the narrative—what I refer to here as "the as-
sistant"—outside the fictional world, given his reference to the transparency
of style. I suspect that viewers experience the manipulation by the narrative
as something that happens to the witness in the fictional world—which they
believe themselves to be. It is only in the second place that the viewer regards
that manipulation as something emanating from outside the fictional world,
as an effect of the artefact. Somewhat paradoxically, we have seen that this
may also be true of intense F emotions. For example, although I am gripped
with fright by some unexpected event, this fright may be followed by a sense
of admiration for the ingenuity of the filmmaker, an emotion that has the arte-
fact—in a wider sense, that is, including the maker—as its object.
I believe that only a combination of fine-grained film analysis and viewer
studies can lead to empirical conclusions on this point. The degree to which
the viewer experiences this control from outside and the filmic factors that
influence this control are major considerations here. An interesting example
of such a study may be the following. Smith discusses a scene from The Road
Warrior (1982), in which the leader of the rival gang steers his truck straight
at the truck of the protagonist Max and his helper. The distance between the
two vehicles is rapidly closing, and the viewer expects that in a fraction of a
second they will crash into each other. However, the crash is postponed for a
full 8 seconds. Smith concluded, "we realize that the norms of physical be-
havior which have governed this world throughout the film still hold, and that
a narrational agency is manipulating the way we perceive the temporal di-
mensions of it" (Smith, 1995, p. 43). But, again, I suspect that the narrational
agency is not experienced as an external force, but rather as one operating
within the fictional world. And if this is so, one can be fairly sure that the post-
ponement of the crash will find favor with the viewer.136 However, if the sus-

136
As time passes and no changes occur in the observed proximity of closure, promise in-
creases and, with it, interest (see chap. 4).
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 243

pense is strung out too long, this will lead to a decline in interest. Up to this
crucial point, the postponement can, without too much deliberation, be con-
sidered consistent with the role of the assistant in the fictional world. In ex-
periments by de Wied that we have mentioned earlier, suspenseful sequences,
assumed to provoke strong anticipatory emotion, have been shortened or
lengthened by reediting the original, preserving the structure of the action.
The results show that tension in viewers does indeed increase as the suspense
is prolonged, that is, and then decreases (de Wied, 1991, 1995). One expla-
nation for these results is that the postponement of closure in suspenseful
scenes obeys conventions. If the narrational agency were experienced as an
external force—that of the artefact—then this would b e in accordance with
conventions rather than with the logic of the fiction, which is derived from
the logic of the real world. A somewhat different, more precise, explanation
is that the convention itself is in accordance with temporal—and other—laws
of the real world, or rather with the general psychological laws of temporal
perception. In certain genres any extreme deviation from, or even violation
of, these laws is experienced as disagreeable. There conventions are conven-
tional, but convention is not entirely arbitrary. Cultural history might well
have led to other customs* but certain customs were more likely to be formed
than others. (See, for a more general treatment of the relationship between
filmic convention and elementary psychological functions, Tan, in press.)
At the beginning of chapter 3 we spoke of a pragmatic contract between
the viewer and the maker, based on a common concern. The maker of the
traditional feature film is out to entertain the audience, while the viewers are
there precisely because they want to be entertained. Maker and viewer, each
is prepared to do something, and each receives something in return. Ac-
cordingly, in our view of interest, the dominant emotion involved in watching
films, a major role is reserved for the willingness of the viewer to invest. This
willingness to invest also has a bearing upon the confidence that the viewer
has in the assistant. 137 Not only are often minimum indications of a long-term
improvement sufficient to entice viewers to continue to invest attention in the

137
It may be instructive here to consider the contrast between the film viewer and the televi-
sion viewer. It has become increasingly clear that television viewers have an entirely different
pragmatic contract with the maker. They negotiate at great length with the assistant, often re-
sponding to his suggestions in unpredictable ways. Fiske (1987), for example, described various
ways in which viewers arrive at quite different "readings" of a "television text." He suggested that
television dramas anticipate this attitude through their "openness" and "polysemy." Thus tele-
vision viewers sometimes adopt readings in which the artefact occupies a central place. These
characteristics—openness, multiplicity of meanings, and the invitation to a consideration of the
artefact—are diametrically opposed to the forceful steering of the observational attitude by the
film narrative and the resulting homogeneous experience on the part of viewers. The explana-
tion for this is probably that there is a better match between the average feature film and its nat-
ural audience than between the average television program and its audience, due to the low cost
of television viewing. The natural audience of a (popular) feature film may be more homoge-
neous than that of the average (popular) television program.
244 CHAPTER 4

film, they also go along with the assistant's directions as to how the film
should be seen. Viewers are apparently perfectly willing to accept differences
between the logic of the fictional world and that of everyday reality; there are
even indications that as a rule they are also willing to lay aside the knowledge
of the world they have. They appear to be doing something that is described
in the pretend theory: for the sake of the "game" of guided fantasy, they sus-
pend disbelief, or even deny knowledge. This latter phenomenon has been
demonstrated in a nice experiment by Gerrig (1989), in which each of the
subjects was asked to read a story. 138 This experiment showed that subjects
have the greatest difficulty in indicating the correctness or incorrectness of
historical facts when these are kept up in the air for a short time through the
creation of suspense. The facts in question were extremely well known, for
example, "George Washington was the first president of America" or "the
United States dropped an atom bomb on Japan." Gerrig indicated various
ways in which this effect comes about (Gerrig, 1989). What this shows and,
moreover, what is of importance for our discussion is that in principle the
subject accepts the suggestions of the assistant, while at the same time re-
maining under the spell of the story. Assuming that this is equally true of the
film viewer, there would appear to be considerable willingness to cooperate
with the assistant and to lend credence to the fiction, in anticipation of fu-
ture gratification. Contrary to the view of Carroll and Smith, some form of
faith may play a role here. Not all knowledge can be forgotten, and we can-
not believe something simply because we want to; however, it is possible to
put knowledge aside for a short time, and why shouldn't we be capable of be-

138
Gerrig presented to the subjects, students at Yale University, a number of stories dealing
with well-known historical facts, for instance, the fact that George Washington became presi-
dent. In one condition, suspense was built up by creating uncertainty about the outcome, some-
thing like the following: George Washington was asked to stand, but he refused because he did
not feel physically up to it. Attention then turned to John Adams. In the meantime, his friends
tried to convince him to change his mind. In another condition, there was no suspense; no ob-
stacles were placed in his path and the historical protagonist headed straight for his goal. After
suspense had been built up, the subjects were asked to verify a target sentence that contained
the well-known historical fact, for example "George Washington was elected president." In the
nonsuspense condition, the target sentence was presented in the same position in the narrative
as in the suspense condition. Verification latency times were higher in the suspense condition
than in the nonsuspense condition. When a warning sentence in which the historical fact was
referred to was inserted at the very beginning of the story, this did not reduce the experimental
effect: it was still more difficult for the subjects in the suspense condition to confirm the histor-
ical fact. In a second and third experiment, the effect continued until the suspense was resolved;
even when subjects were asked to carry out a completely different task lasting five or ten min-
utes, the effect remained intact as long as no outcome was provided. Other recent research shows
that attempts not to believe what one reads result in paradoxical effects, at least under high men-
tal load (Gilbert, Tofarodi, & Malone, 1993; Houston & Wegner, 1993, in Wegner, 1994). Per-
haps at certain points narratives create conditions of high mental load, thus facilitating belief.
Suspense and other high-interest fragments may be cases in point.
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 245

lieving in something temporarily, for the sake of the pleasures of the fiction?
Still this does not mean that viewers confuse the fiction with the real world,
as the illusion theory maintains. But they attribute a reality or potential of
some kind to what they are seeing, whereby the stimulus must meet certain
conditions. Thus the emotion that flows from the illusion is real, in contrast
to what the pretend theory holds, but the reality that one sees is a product
of, on the one hand, realistic mimesis and, on the other, a certain pretending
on the part of the viewer. Here, pretending consists in going along with the
representation of a possible reality for the sake of one's own entertainment. 139
In other words, it is not impossible that under certain conditions viewers
employ an emotion-first strategy in their comprehension of a feature film. In
addition to an active belief that something is possible, this strategy also in-
volves the viewers' willingness to allow liberties to be taken with their imag-
inary representatives. These liberties always remain within the bounds of the
pleasurable, because the viewer is always given just enough hints to be able
to anticipate, to a greater or lesser degree, coming events. One result of this
is the sense of flow that the viewer experiences. Effortlessly, the events of the
fictional world call up expectations of things to come, and these expectations
are automatically met. The work of the viewer proceeds smoothly, and, in a
sense, unpleasant emotions, as well as surprises and the postponement of
clarity and resolution, are acceptable and even desirable.
The illusion of the controlled spectator is the most direct cause of the emo-
tion that, in our view, is at the heart of the viewers' experience of a feature
film, namely, the fascination with which they follow the fictional action. In
chapters 4, 5, and 6, which dealt with the emotion of interest, we saw how
this fascination on the part of the viewer flows from the control that the film
narrative exercises over the disclosure of information pertaining to the events
and characters of the fictional world. This control functions in such a way that
the attention of the viewer is constantly steered in one direction or another.
The assistant provides incomplete information on the ultimate destination of
the journey. On the one hand, the difficulties and complications put forward
by the narrative are such that they regularly threaten to exceed the capacity
of the information-processing system. On the other hand, these are always
countered by a promise of some kind. At any given moment during the pre-
139
Kubovy (1986) in his extremely interesting book on the psychological nature and artistic
use of perspective in pictures and paintings, distinguished a number of variants of illusion. He
distinguished collusion from illusion as a perceptual error, that is, seeing what is not really there.
In collusion, the beholder willingly suspends disbelief. In order for this to be the case, one has
to be aware of a perceptual error, and at the same time one has chosen to go along with the artist
and look at the picture in a specific way, giving rise to the "erroneous" effect, which then fully
materializes. The elegant term collusion may be adequate in referring to the diegetic effect,
which is also based in part on a pragmatic cooperation between the film viewer and the film-
maker. The diegetic effect, like Kubovy's collusion, is compelling, once the viewer has chosen to
go along with the fiction.
246 CHAPTER 4

sentation of the film, the assistant not only keeps anticipations active con-
cerning what is going to happen in the near future but also dangles before
the viewer's eyes a substantial return with respect to some norm in the past.
This return may be in the form of spectacle, knowledge, or preferred out-
comes. At the same time, however, it remains to be seen what one will actu-
ally be shown. The assistant is toying with the Law of Change.
The prospective orientation directly addresses a concern, namely, the
viewer's need for complete knowledge. It produces an immediate response,
the emotional urge to create anticipations concerning coming developments,
the intentions of individual characters, and the relations between them. This
urge is triggered by the simple action of following the film, and it is difficult
to suppress until the entire emotional episode of the film has been concluded.

The Illusion of the Observational Attitude. Given the illusion of the

spectator's position as observer, the film also imposes, in a fairly compelling
manner, a particular manner of observation. This is the final component of
the diegetic effect that we will be examining here. By virtue of this illusion,
one is forced to look in a particular way at the events to which one is a silent
witness; indeed, it is made extremely difficult to look at them in any other
way. We said earlier that in the reality of daily life people often have a choice
about whether or not to watch an emotional scene, whereas in the case of
films, that choice has already been made by virtue of the fact that they have
taken their place in the movie theater. Moreover, when they witness an inter-
esting incident in daily life, people can generally choose to see it one of vari-
ous ways. The feature film imposes perspectives on the viewer. The observa-
tional attitude varies according to several film characteristics, including
genre of the film, the particular phase in the narrative, and the point of view
(e.g., whether or not it runs more or less parallel to that of a character), and
it ranges from warm understanding and passionate desire, through cool and
clinical contemplation, to malicious delight. It is difficult to escape this illu-
sion. For example, the average viewer would have trouble summoning com-
passion for a group of characters he or she hardly knows, portrayed objec-
tively as a bunch of good-for-nothing blockheads who, through their own
stupidity, are picked off, one by one. This scene from Posse (1975) is followed
by viewers with a malicious pleasure not often encountered in real life. Simi-
larly, it is almost impossible not to feel some measure of regret when a sym-
pathetic young gangster is killed in front of his girlfriend by an unfeeling
strongarm, as in Thieves Like Us (1974).
Both the empathetic, compassionate attitude and the remote or malevo-
lent attitude are imposed by the film. Together, the alternating imposition of
various observational attitudes and the prospective orientation of the control
exercised by the assistant form the basis of the affect structure of the feature
film (see chap. 7). Through the operation of these two factors, the entire body
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 247

of experiences and emotions is endowed with an orientation and cohesion

that find favor with audiences. Viewers realize that their experience has been
given a consistent form; they trust that the programming by the assistant has
been designed with a view to achieving the optimum emotional effect, and
ideally even makes every effort to cooperate in the realization of the entire
body of emotions.
As we know, the adoption of a certain observational attitude leads to char-
acteristic emotions. And yet such attitudes can also produce states that
closely resemble genuine emotions. When the film forces me to look upon a
situation in the same way as the protagonist does, I may have an empathetic
emotion, a genuine emotion that is the result of my realization of the signif-
icance of the situation for the sympathetic protagonist. But I may also make
a conscious effort to imagine what it feels like to be the protagonist. In chap-
ter 6 we maintained that this almost never means that some kind of ego fu-
sion is taking place. That would be a delusion, comparable, although not iden-
tical, to the equally improbable belief that one is watching events in the real
world. For the subject, however, it may be difficult to distinguish this state
from authentic emotion, even in spite of an awareness that a certain amount
of effort has gone into the achievement of this state. From the standpoint of
the researcher, the emotion is not real; the simulation of other people's emo-
tion will never be more than a simulation, and, most importantly, the viewer
does not share the concerns of the protagonist. I believe that there is some-
thing to be said for the pretend theory within the complex body of experi-
ences evoked by the feature film. In the midst of the genuine emotions and
the clear cases of nonemotional states, there are also states that border
closely on genuine emotion and can only with difficulty be distinguished from
it. It is even possible that the existence of near-emotions renders a service
to real emotion! To return to our example, it is conceivable that the near-
emotion I am conscious of when I try to imagine the feelings of the protago-
nist in difficulty greatly promotes the genuine emotion of pity, an emotion
that is based on my own real sympathy concern.
The imposition of an observational attitude is one of the most important
ways in which the feature film makes use of the Law of Concern. "Try to look
at it from this angle," suggests the assistant compellingly; I am shown the suf-
fering of a victim in close-up, so that I am almost forced to see it, and my
sympathy concern is addressed. In the same way, as a result of a particular
observational attitude my value concerns are addressed ("This must not be
allowed to happen") or my innermost desires for intimacy, bonding, and se-
curity ("I hope they will get back together again"). Although we may share
the view of Peters (1989) that there is a constituent layer upon which the fea-
ture film acts, such a layer is formed not by the emotion but by the concerns.
The observational attitude also acts upon the law of situational meaning.
It is important to note that at first glance the observational attitudes seem to
248 CHAPTER 4

be almost unlimited in number. In this way they make possible a wide range
of film emotions. These include far more than just the basic emotions or the
familiar ones connected with the more common genres, such as laughter, cry-
ing, "the creeps," and so on.
The feature film's ability to orient the disposition of the viewer is one of
its charms. And the fact that this is seen as a charm is yet another indication
of an emotion-first strategy on the part of the viewer. The observational atti-
tudes referred to above also occur in everyday life, but there they are liable
to exact a higher toll. The Law of Care of Consequence makes it understand-
able that people do not generally concern themselves with what happens
around them. A malicious delight in the misfortunes of others may afford one
a degree of pleasure, but it also leads to feelings of guilt. Given the fact that
people in real need are often not only denied help but are also totally ignored,
it would appear that attention, sympathy, and pity, while socially desirable,
are not automatically part of our attitude toward other people. It should be
realized that the problem we are addressing here is a complicated one. Per-
haps we want to sympathize with others but are unable to do so, for instance,
because there are no possibilities to act upon our urge to help, and the con-
flict between urge and possibility has become unbearable. Be that as it may,
feature films give rise to only virtual sympathy. And is virtual sympathy not
inexpensive in the end?
In the case of the traditional feature film, the emotions experienced by
viewers offer an extra satisfaction, namely, the sharing of feelings. In the dark-
ened movie theater, we are all united by a common experience. We all know
that the overt emotional response is contagious, but even when each and
every viewer, whether happy or sad, is sitting quietly in his or her seat with-
out any perceptible signs of emotion, there is a realization in the theater of a
collective vision of the fictional world. Each viewer imagines himself or her-
self present in the fictional world. Each viewer is individually convinced,
through the sheer force of the situational meaning, so convinced, in fact, that
he or she is positive that the others are faced with the same situation in ex-
actly the same way. As a result, all the members of the audience sense that
they are united with one another, as part of a community of meaning. Laugh-
ing or jeering together is an externalization of this solidarity, but this is not
in itself necessary for the sharing of meaning, the consummation of an ideo-
logical solidarity.

CONCLUSION: THE FEATURE FILM

AS AN EMOTION MACHINE

As we come to the end of our discussion of illusions, we note once more that
illusion is used here in the broad sense of the term. The effects of the feature
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 249

film that we are talking about are not all equally inevitable. The sequence in
which they appear—apparent movement, apparent presence in the fictional
world, apparent objectivity of the fictional world, the illusion of the controlled
spectator, and the illusion of the observational attitude—indicates a sequence
of inevitability. This means that the illusions are also logically related to one
another. If one effect works, then the operation of the following effect be-
comes more probable, that is, more difficult to resist. Apparent movement is
necessary for the diegetic effect in the narrow sense that is required for the
imagination of the role of spectator, and this in turn is necessary if one is to
adopt the observational attitude dictated by the film. The illusions are
arranged as a series. As a rule, viewers do not pause to consider the fact that
there are a number of illusions at work here. They have an undivided expe-
rience of the fictional events. Together, the previously mentioned effects,
analogous to the experience of the viewer, can best be seen by the researcher
as a single, many-sided illusion.
The assumption presented at the beginning of this chapter can now be cast
in a more definitive form, although it remains an assumption. With a mini-
m u m of cooperation on the part of the natural viewer, the traditional feature
film evokes a multiple illusion. We have seen that it is only with great diffi-
culty that the temptations of the illusions offered by the feature film can be
resisted. But the idea of resistance is not one that is likely to appeal to the
natural audience. Why would anyone want to resist such a skillful and en-
gaging opponent? Indeed, why would anyone want to see the film as an op-
ponent in the first place? Resisting the diegetic effect means depriving one-
self of the gratification of all the concerns that can be realized by films. The
sense of solidarity with characters, the satisfaction of cognitive curiosity, the
excitement, the sense of competence, the enjoyment of spectacles, and all
the pleasures associated with the filmic representation of the fictional world—
all these would be wasted if viewers were not prepared to do their bit. A more
rational strategy would be for viewers to get all the illusions out of the film
that they can because of the pleasure this provides and, not least, for the emo-
tions themselves. How to get everything possible out of a film? By following
the action, surrendering to the gentle persuasion of the assistant, giving ex-
pectations a chance to materialize, working them out, and seeking confir-
mation in the action that presents itself. By seeking explanations for the un-
expected, by making an affective investment in the fate of the protagonists,
by taking sides, sympathizing with the protagonist when he or she deserves
it, and displaying malicious delight when the narrative promises schaden-
freude or irony. The inclination to do all this, which can be quite strong, is it-
self an emotion, namely, interest. If we assume that the viewer is free to regu-
late something in the film emotion, then it would be more correct to speak
of maximizing rather than, say, tempering or transforming an emotion. The
viewer only has to make one tiny move in order for apparent motion to turn
250 CHAPTER 4

into actual emotion, into an almost uninterrupted emotion episode. That one
move need be no more than a simple willingness to follow intelligently the fic-
tional action and the suggestions of the assistant embedded in that action.
The rest follows as a matter of course, because the emotional response of the
viewer has the capacity to reinforce itself. The modicum of good will on the
part of the viewer strengthens the illusion of the diegetic effect, as a result of
which the situational meaning structure becomes more powerful and the re-
sponse even more difficult to resist. The emotional response and, above all,
the interest then reinforce themselves. In this way the intensity of the emo-
tional experience can easily reach a point of no return. Reason is no longer
capable of destroying the illusions, and the activity of the viewer is guided al-
most exclusively by the rationality of the film emotion.
The feelings produced by the film in this way possess the most important
functional characteristic of what we regard as a genuine emotion, namely, the
realization of concerns despite the pressure of reason, and, above all, gaining
and retaining control precedence. The emotional response drives into tem-
porary oblivion all responses other than the experience of situational mean-
ing structures and action tendencies, until the end of the film does away with
most of the action tendencies aimed at the fictional action, and thus the will-
ingness to continue watching.
It must be admitted that some emotions are less intense that those that
would have been evoked by a similar event in the real world.140 This phe-
nomenon is related to the fact that during film viewing such emotions are al-
ways mingled with another emotion, interest, and with a fascination with the
film as a continuing story. It is conceivable that, together, interest and the
other emotions jointly equal the intensity of emotions experienced outside
the movie theater, in otherwise comparable situations. The pleasures need
not be any less enjoyable either, because of the relative security of the expe-
rience and, perhaps more important, the appeal of interest, which occupies
such a central position.
In conclusion, our argument with respect to the role of the film stimulus
can be summarized as follows: in general, narration may be seen as the sys-
tematic evocation of emotion in an audience, according to a preconceived
plan. Narration by means of film is one way of doing this. From a psychologi-
cal standpoint, the specific contribution of the moving picture, that is, the
feature film, to the emotion of the viewer is as follows. By presenting to the
viewer a complex and continually variable illusory stimulus, it plays upon the
most universal concerns, the weaknesses—or should we say strengths—that
are inherent in the psychological makeup of the viewer. The film generates,

140
This has been demonstrated in empirical comparative research by van Vliet (1991), who
showed that empathy is stronger in an actual situation than in theater performances of a com-
parable situation.
CONCLUSION: FEATURE FILM AS AN EMOTION MACHINE 251

as it were, like a moving belt, for the entire 90 minutes, a continuous emo-
tional situational meaning, programmed as a part of its specific affect struc-
ture, that results in an ongoing, genuine emotional response. The way the
emotional response dovetails with the kind of emotion that the viewer has
come for is a miracle of precision.
In this sense, and more than anything else, the traditional feature film is
a genuine emotion machine.
 References

Aaker, D. A., Stayman, D. M., & Hagerty, M. R. (1986). Warmth in advertising: Measurement,
impact, and sequence effects. Journal of Consumer Research, 12, 365-381.
Adorno, T. W., Frenkel-Bruswik, E., Levinson, D. J., & Sanford, R. N. (1950). The authoritarian
personality. New York: Harper & Row.
Affron, G. (1982). Cinema and sentiment. Chicago: University of Chicago Press.
Albritton, D. W., & Gerrig, R. J. (1991). Participatory responses in text understanding. Journal
of Memory and Language, 30, 603-626.
Allen, R. C. (1993). Representation, illusion, and the cinema. Cinema Journal, 32(2), 21-49.
Alley, T. R. (1983). Growth-produced changes in body shape and size as determinants of per-
ceived age and adult caregiving. Child Development, 54, 241-248.
Alley, T. R. (1986). An ecological analysis of the protection of primate infants. In V. McCabe &
G. J. Balzano (Eds.), Event cognition (pp. 239-258). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Andersen, S. M., & Klatzky, R. C. (1987). Traits and social stereotypes: Levels of categorization
in person perception. Journal of Personality and Social Psychology, 53, 235-246.
Anderson, J. R. (1976). Language, memory and thought Hillsdale, NJ: Lawrence Erlbaum
Associates.
Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press.
Andrew, D. (1984). Concepts in film theory. Oxford: Oxford University Press.
Andrews, J. D. W. (1989). Integrating visions of reality. Interpersonal diagnosis and the exis-
tential vision. American Psychologist, 44, 803-817.
ANF (1988). Naar een toekomst voor de kunstzinnige film, 1988-1992 [Toward a future for the
art film: From 1988 to 1992]. Amsterdam, The Netherlands: Associatie van Nederlandse
Filmtheaters.
Ang, I. (1985). Watching Dallas. London and New York: Methuen.
Anthony, B. J., & Graham, F. J. (1985). Blink reflex modification by selective attention: Evidence
for the modulation of "automatic" processing. Biological Psychology, 21, 43-59.

253
254 REFERENCES

Apter, M. J. (1984). Reversal theory, cognitive synergy and the arts. In W. R. Crozier & A. J.
Chapman (Eds.), Cognitive processes in the perception of art (pp. 411-426). Amsterdam,
The Netherlands: North-Holland.
Aristotle (1972). Poetics. (Trans. M. E. Hubbard). In D. A. Russell & M. Winterbottom (Eds.),
Ancient literary criticism (pp. 85-112). Oxford, England: Clarendon Press.
Arnheim, R. (1933). Film. London: Faber.
Arnheim, R. (1958/1983). Film as art London: Faber & Faber.
Arnheim, R. (1971). Entropy and art. An essay on disorder and order. Berkeley: University of
California Press.
Arnold, M. B. (1960). Emotion and personality (Vols. 1 - 2 ) . New York: Columbia University Press.
Aronson, E. (1961). The effect of effort on the attractiveness of rewarded and unrewarded stim-
uli. Journal of Abnormal and Social Psychology, 63, 375-380.
Atkin, C. K. (1985). Informational utility and selective exposure to entertainment media. In
D. Zillmann & J. Bryant (Eds.), Selective exposure to communication (pp. 63-92). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Atkinson, J. W. (1969). Change of activity: A new focus for the theory of motivation. In T. Mi-
schel (Ed.), Human action (pp. 106-133). New York: Academic Press.
Auerbach, E. (1946). Mimesis: Dargestellte Wirklichkeit in der abendlandischen Literatur
[Mimesis: Represented reality in western literature]. Bern, Switzerland: Francke.
Aumont, J., Bergala, A., Marie, M., & Vernet, M. (1983). Esthetique du film [Esthetics of film].
Paris: Nathan.
Austin, B. (1986, Spring). Motivations for movie audiences. Communication Quarterly, 34,
115-126.
Austin, B. (1988). Immediate seating. A look at movie audiences. Belmont: Wadsworth.
Averill, J. R. (1982). Anger and aggression. An essay on emotion. New York: Springer.
Balazs, B. (1938). Zur Kunstphilosophie des Films [On the philosophy of film as art]. Das Wort,
3, 104-119.
Bandura, A. (1973). Aggression: A social learning analysis. Englewood Cliffs, NJ: Prentice Hall.
Bannerman, J., & Lewis, J. M. (1977). College students' attitudes toward movies. Journal of
Popular Film, 6, 126-139.
Baron, R. A. (1977). Human aggression. New York: Plenum Press.
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 Psychology, 35, 79-87.
Baron, R. M. (1980). Contrasting approaches to social knowing: An ecological perspective. Per-
sonality and Social Psychology Bulletin, 6, 590-600.
Baron, R. M. (1991, August). The implications of treating personality dispositions as social
affordances. Paper presented at the 6th International Conference on Event Perception and
Action, Amsterdam, The Netherlands.
Barsalou, L. W. (1991). Deriving categories to achieve goals. In G. H. Bower (Ed.), The psy-
chology of learning and motivation: Advances in research and theory (Vol. 27, pp. 1 - 6 4 ) .
New York: Academic Press.
Barthes, R. (1966). Introduction a l'analyse structurale des recits [Introduction to the structural
analysis of stories]. Communications, 8, 1-27.
Batson, C. D. (1987). Prosocial motivation: Is it ever truly altruistic? In L. Berkowitz (Ed.), Ad-
vances in Experimental Social Psychology (Vol. 20, pp. 65-122). New York: Academic
Press.
Baudry, J.-L. (1986). Ideological effects of the basic cinematic apparatus. In P. Rosen (Ed.), Nar-
rative, apparatus, ideology (pp. 286-298). New York: Columbia University Press. (Original
work published 1975)
Bazin, A. (1958). Qu'est-ce que le cinema? Tome I. Ontologie et langage. Paris, France: Cerf.
REFERENCES 255

Bazin, A. (1959). Qu'est-ce que le cinema? Tome II. Le cinema et les autres arts. Paris, France:
Cerf.
Bazin, A. (1961). Qu'est-ce que le cinema? Tome III. Cinema et sociologie. Paris, France: Cerf.
Bazin, A. (1962). Qu'est-ce que le cinema? Tome IV. Une esthetique de la realite. Paris, France:
Cerf.
Beckerman, B. (1970). Dynamics of drama. Theory and method of analysis. New York: Knopf.
Bellour, R. (1973). L'evidence et le code [The obvious and the code]. In D. Noguez (Ed.), Cin-
ema: Theories, lectures, (pp. 219-227). Paris: Klincksieck.
Bellour, R. (1979). Segmenter, analyser. In R. Bellour, Uanalyse du film (pp. 247-270). Paris,
France: Albatros. (Original work published 1976)
Bellour, R. (1979). Sur l'espace cinematographique. In R. Bellour, Uanalyse du film (pp. 6 4 -
72). Paris, France: Albatros.
Berlyne, D. E. (1960). Conflict, arousal and curiosity. New York: McGraw-Hill.
Berlyne, D. E. (1962). Uncertainly and epistemic curiosity. British Journal of Psychology, 53,
27-34.
Berlyne, D. E. (1970). Novelty, complexity, and hedonic value. Perception and Psychophysics,
8, 279-286.
Berlyne, D. E. (1971). Aesthetics and psychobiology. New York: Appleton-Century-Crofts.
Berlyne, D. E. (1974). Studies in the new experimental aesthetics: Steps toward an objective
psychology of aesthetic appreciation. New York: Wiley.
Berlyne, D. E. (1978). Curiosity and learning. Motivation and Emotion, 2, 97-175.
Berry, D. S., Kean, K. J., Misovich, S. J., & Baron, R. M. (1991). Quantized displays of human
movement: A methodological alternative to the point-light display. Journal of Nonverbal Be-
havior, 15, 81-97.
Berry, D. S., &McArthur, L. Z. (1986). Perceiving character in faces: Impact of age-related crano-
facial facial changes on social perception. Psychological Bulletin, 100, 3-18.
Berscheid, E. (1985). Interpersonal attraction. In G. Lindzey & E. Aronson (Eds.), Handbook
of Social Psychology (Vol. 2, 2nd ed., pp. 413-484). New York: Random House.
Bever, T. G. (1986). The aesthetic basis for cognitive structures. In M. Brand & R. M. Harnish
(Eds.), The representation of knowledge and belief (pp. 314-356). Tucson: University of Ari-
zona Press.
Birkhoff, G. D. (1933). Aesthetic measure. Cambridge, MA: Harvard University Press.
Bonitzer, P. (1982). Le champ aveugle [The blind field]. Essais sur le cinema. Paris: Gallimard.
Booth, W. C. (1961). The rhetoric of fiction. Chicago and London: University of Chicago Press.
Borde, R., & Chaumeton, E. (1975). Panorama du film noir americain: 1941-1953 [Panorama
of the American film noir: 1941-1953]. Paris, France: Editions d'Aujourd'hui.
Bordewijk-Knotter, J. M. (1988). Pinter appeal. A comparative study of responses to "The home
comingUnpublished doctoral dissertation. Amsterdam, The Netherlands: University of Am-
sterdam.
Bordwell, D. (1979). The art cinema as a mode of film practice. Film Criticism, 4, 56-64.
Bordwell, D. (1985). Narration in the fiction film. Madison, WI: University of Wisconsin Press.
Bordwell, D. (1989a). A case for cognitivism. Iris, 9, 11-40.
Bordwell, D. (1989b). Making meaning. Inference and rhetoric in the interpretation of cinema.
Cambridge, MA & London: Harvard University Press.
Bordwell, D., Staiger, J., & Thompson, K. (1985). The classical Hollywood cinema: Film style
and mode of production. New York: Columbia University Press.
Bordwell, D., & Thompson, K. (1986). Film art: An introduction (2nd ed.). New York: Knopf.
Boselie, F. (1982). Over visuele schoonheidservaring [On the visual experience of beauty]. Un-
published doctoral dissertation, Catholic University Nijmegen, The Netherlands.
Bourdieu, P. (1979). La distinction. Critique sociale dujugement [Distinction. A critique of so-
cial judgment]. Paris, France: Minuit.
256 REFERENCES

Bower, G. H. (1978). Experiments on story comprehension and recall. Discourse Processes, 1,

211-231.
Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129-148.
Bower, G. H., & Cohen, P. R. (1982). Emotional influences in memory and thinking: Data and
theory. In M. S. Clark & S. T. Fiske (Eds.), Affect and Cognition (pp. 291-331). Hillsdale,
NJ: Lawrence Erlbaum Associates.
Branigan, E. (1992). Narrative comprehension and film. London & New York: Routledge
Brehm, J. W., & Cohen, A. R (1962). Explorations in cognitive dissonance. New York: Wiley
Brewer, M. B. (1988). A dual process model of impression formation. In T. K. Srull & R. S. Wyer
(Eds.), Advances in Social Cognition (Vol. I, pp. 1-36). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Brewer, W. F., & Lichtenstein, E. W. (1981). Event schemas, story schemas, and story grammars.
In J. Long & A. Baddeley (Eds.), Attention and performance (Vol. 9, pp. 363-379). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Brewer, W. F., & Lichtenstein, E. H. (1982). Stories are to entertain: A structural-affect theory
of stories. Journal of Pragmatics, 6, 473-486.
Britton, B. K., Glynn, S. M., Meyer, B. J. F., & Penland, M. J.(1982). Effects of text structure on
use of cognitive capacity during reading. Journal of Educational Psychology, 74, 51-61.
Britton, B. K., Graesser, R. C., Glynn, S. M., Hamilton, T., & Penland, M. J., (1983). Use of cog-
nitive capacity in reading: Effects of variations in surface features of text. Journal of Exper-
imental Psychology: Learning, Memory and Cognition, 5, 318-331.
Brook, P. (1976). The melodramatic imagination. New Haven, CT: Yale University Press.
Brooks, V., & Hochberg, J. (1960). A psychophysical study of cuteness. Perceptual and Motor
Skills, 11, 205.
Browne, N. (1975). The spectator-in-the-text: The rhetoric of stagecoach. Film Quarterly, 29,
26-38.
Bruner, J. S. (1957). Perceptual readiness. Psychological Review, 64, 12-52.
Brunetiere, F. (1914). The law of the drama. New York: Dramatic Museum of Columbia Uni-
versity. (Original work published 1894)
Brunsdon, C. (Ed.).(1986). Films for women. London: British Film Institute.
Brunswik, E. (1956). Perception and the representative design of psychological experiments
(2nd ed.). Berkeley: University of California Press.
Bryant, J., & Zillmann, D. (1977). The mediating effect of the intervention potential of comu-
nications on displaced agressiveness and retaliatory behavior. In B. D. Ruben (Ed.), Com-
munication Yearbook (Vol. I, pp. 291-306). New Brunswick, NJ: Transaction.
Bryant, J., & Zillmann, D. (1984). Using television to alleviate boredom and stress: Selective ex-
posure as a function of induced excitational states. Journal of Broadcasting, 28, 1 - 2 0 .
Burch, N. (1979). To the distant observer. Berkeley: University of California Press.
Buckland, W. (in press). Relevance and cognition: Towards a pragmatics of unreliable filmic nar-
ration. In J. Miiller (Ed.), Proceedings of the Conference on Film Pragmatics. Miinster, Ger-
many: Nodus.
Calhoun, C., & Solomon, R. C. (Eds.). (1984). What is an emotion? Classic readings in philo-
sophical psychology. New York: Oxford University Press.
Cantor, J. (1991). Fright responses to mass media productions. In J. Bryant & D. Zillmann
(Eds.), Responding to the screen. Reception and reaction processes (pp. 169-197). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Cantor, J., Ziemke, D., & Sparks, G. G. (1984). Effect of forewarning on emotional response to
a horror film. Journal of Broadcasting, 28, 21-31.
Cantor, J. R., Zillman, D., & Bryant, J. (1975). Enhancement of experienced sexual arousal in
response to erotic stimuli through misattribution of unrelated residual excitation. Journal of
Personality and Social Psychology, 32, 69-75.
REFERENCES 257

Carello, C., Grosofsky, A., Shaw, R. E., Pittenger, J. B., & Mark, L. S. (1989). Attractiveness of
facial profiles is a function of distance from archetype. Ecological Psychology, 1, 227-251.
Carroll, N. (1988a). Philosophical problems of classical film theory. Princeton, NJ: Princeton
University Press.
Carroll, N. (1988b). Mystifying movies: Fads and fallacies in contemporary film theory. New
York: Columbia University Press.
Carroll, N. (1990). The philosophy of horror. Or paradoxes of the heart. New York: Routledge.
Carver, C. S., & Scheier, M. F. (1990). Origin and functions of positive and negative affect: Con-
trol-proces view. Psychological Review, 97, 19-35.
Charniak, E. (1972). Toward a model of children's story comprehension (Report No. Al-TR-
266). Cambridge, MA: MIT, Art. Int. Lab.
Chatman, S. (1978). Story and discourse. Narrative structure in fiction and film. Ithaca, NY:
Cornell University Press.
Child, I. L. (1946). Children's preference for goals easy or difficult to obtain. Psychological
Monographs, 60 (4, Whole No. 280).
Child, I. L. (1969). Esthetics. In G. Lindzey & E. Aronson (Eds.), Handbook of Social Psychol-
ogy (Vol. Ill, pp. 853-916). Reading, MA: Addison-Wesley.
Child, I. L. (1978). Aesthetic theories. In E. C. Carterette & M. P. Friedman (Eds.), Handbook
of Perception (Vol. 10, pp. 11-131). New York: Academic Press.
Choffray, J. M., & Pas, B. (1980). Determinants of market success for commercial movies. Der
Markt, 73, 3-13.
Christ, W. G., & Medoff, N. J. (1984). Affective state and selective exposure to and use of tele-
vision. Journal of Broadcasting, 28, 1 - 2 0 .
Cogan, R., Cogan, D., Waltz, W., & McCue, M. (1987). Effects of laughter and relaxation on dis-
comfort thresholds. Journal of Behavioral Medicine, 10(2), 139-144.
Cohen-Seat, G. (1958). Essais sur les principes d'une philosophie du cinema: Notions fonda-
mentales et vocabulaire de filmologie [ Essays on the principles of a philosophy of the cin-
ema. Fundamental notions and a vocabulary for the study of film]. Paris, France: PUF.
Colby, K. M. (1975). Artificial paranoia. New York: Pergamon.
Comisky, P, & Bryant, J. (1982). Factors involved in generating suspense. Human Communi-
cation Research, 9, 49-58.
Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: A framework for memory re-
search. Journal of Verbal Learning and Verbal Behavior, 11, 671-684.
Craik, F. I. M., & Tulving, E. (1975). Depth of processing and the retention of words in episodic
memory. Journal of Experimental Psychology: Human Learning and Human Memory, 104,
268-294.
Csikszentmihalyi, M. (1975). Beyond boredom and curiosity. San Francisco: Jossey Bass.
Csikszentmihalyi, M., & Csikszentmihalyi, I. S. (1988). Optimal experience. Psychological stud-
ies of flow in consciousness. Cambridge, England: Cambridge University Press.
Culler, J. (1975). Structuralist poetics: Structuralism, linguistics and the study of literature.
Ithaca: Cornell University Press.
Cupchik, G. C., & Gebotys, R. J. (1988). The experience of time, pleasure and interest during
aesthetic episodes. Empirical Studies of the Arts, 6(1), 1-12.
Cutting, J. E. (1987). Rigidity in cinema seen from the front row, side aisle. Journal of Experi-
mental Psychology, Human Perception and Performance, 13, 323-334.
DAndrade, R. G. (1984). Character terms and cultural models. In J. W. D. Dougherty (Ed.), Di-
rections in cognitive anthropology (pp. 321-343). Urbana and Chicago: University of Ilinois
Press.
Davidson, R. J., Ekman, P., Saron, C. D., Senulis, J. A., & Friesen, W. F. (1990). Approach-with-
drawal and cerebral asymmetry: Emotional expression and brain physiology I. Journal of Per-
sonality and Social Psychology, 58, 330-341.
258 REFERENCES

Davis, M. H., Hull, J. G., Young, R. D., & Warren, G. G. (1987). Emotional reactions to dramatic
film stimuli: The influence of cognitive & emotional empathy. Journal of Personality and
Social Psychology, 52, 126-133.
Davis, D. K., & Woodall, W. G. (1982). Uses of TV news: Gratification or edification? Paper pre-
sented at the annual meetings of the International Communication Association Convention,
Boston, MA.
Dawkins, R. (1978). The selfish gene. London: Granada.
de Beaugrande, R. (1982). The story of grammars and the grammar of stories. Journal of Prag-
matics, 6, 383-422.
de Groot, A. D. (1965). Methodologie [Methodology]. The Hague, The Netherlands: Mouton.
de Jonge, P. (1994). De beleving van theatrale produkten: Over empathie en identificatie [Ex-
periencing theatrical products: On empathy and identification]. Unpublished technical re-
port, Department of Psychology, University of Amsterdam, The Netherlands,
de Kuyper, E. (1984). Filmische hartstochten [Filmic passions]. Weesp, The Netherlands/An-
twerpen, Belgium: Het Wereldvenster/Standaarduitgeverij.
de Lauretis, T. (1984). Alice doesn't: Feminism, semiotics, cinema. Bloomington: Indiana Uni-
versity Press.
Dennett, D. C. (1987). The intentional stance. Cambridge, MA: MIT Press,
den Uyl, M. J., & Frijda, N. H. (1984). Mood, emotion and action: A concern realization system.
In Proceedings of the 6th Cognitive Science Conference (pp. 137-141). Hillsdale, NJ: Law-
rence Erlbaum Associates,
de Rivera, J. (1977). A structural theory of the emotions. Psychological Issues, 10, (Whole Vol.).
New York: International Universities Press.
De Sousa, R. (1987). The rationality of emotion. Cambridge & London: MIT Press,
de Wied, M. A. (1991). The role of temporal structures in the production of suspense and du-
ration experience. Unpublished doctoral dissertation. Amsterdam, The Netherlands: Uni-
versity of Amsterdam.
de Wied, M. A. (1995). The role of temporal expectancies in the production of film suspense.
Poetics, 23, (1-2), 107-123.
de Wied, M. A. (in press). The role of temporal expectancies in the production of film suspense.
In P. Vorderer, H. J. Wulff, & M. Friedrichsen (Eds.), Analysis of suspense. Hillsdale, NJ:
Lawrence Erlbaum Associates,
de Wied, M. A., Zillmann, D., & Ordman, V. (1995). The role of empathie distress in the enjoy-
ment of cinematic tragedy. Poetics, 23(1-2), 107-124.
Dorner, D., & Vehrs, W. (1975). Asthetische Befriedigung und Unbestimmtheitsreduktion [Es-
thetic pleasingness and uncertainty reduction]. Psychological Research, 37, 321-334.
Dovidio, J. F., Schroeder, D. A., & Allen, J. L. (1990). Specificity of empathy induced helping:
Evidence for altruistic motivation. Journal of Personality and Social Psychology, 59, 2 4 9 -
260.
Duclos, S. E., Laird, J. D., Schneider, E., Sexter, M., Stern, L., & van Lighten, O. (1989).
Emotion-specific effects of facial expressions and postures of emotional experience. Journal
of Personality and Social Psychology, 57, 100-108.
Duncker, K. (1945). On problem solving. Psychological Monographs, 58 (Whole No. 5).
Durgnat, R. (1967). Films and feeling. London/Cambridge, MA: Faber & Faber/MIT-Press.
Duvignaud J. (1965/1973) Sociologie du theatre. [Sociology of literature and drama]. Paris:
PUF. (Original work published 1973)
d'Ydewalle, G., & Vanderbeeken, M. (1990). Perceptual and cognitive processing of editing rules
in film. In R. Groner, G. d'Ydewalle, & R. Parnham (Eds.), From eye to mind: Information
acquisition in perception, search and reading (pp. 129-139). Amsterdam, The Netherlands:
Elsevier/North-Holland
Dyer, M. G. (1983). The role of affect in narratives. Cognitive Science, 7, 211-242.
REFERENCES 259

Dyer, R. (1979). Stars. London: BFI.

Eaton, M. (1981). Laughter in the dark. Screen, 22(2), 21-43.
Eberwein, R. T. (1984). Film and the dream screen. Princeton, NJ: Princeton University Press.
Efran, J. S., & Spangler, T. J. (1979). Why grown-ups cry. A two-factor theory and evidence from
"The Miracle Worker." Motivation & Emotion, 3, 63-71.
Egri, L. (1960). The art of dramatic writing. New York: Simon & Schuster. (Original work pub-
lished 1946)
Ekman, P. (1992). Are there basic emotions? Psychological Review, 99, 550-553.
Ekman, P., Levenson, R. W., & Friesen, W. V. (1983). Autonomous nervous system activity dis-
tinguishes among emotions. Science, 221 (4616), 1208-1210.
Elias, N., & Dunning, E. (1986). The quest for excitement in leisure. In N. Elias & E. Dunning
(Eds.), Quest for excitement: Sport and leisure in the civilizing process (pp. 63-90). Ox-
ford: Basil Blackwell.
Ellis, J. (1982). Visible fictions. London: Routledge.
Ellsworth, P. C. (1991). Some implications of cognitive appraisal theories of emotion. In K. T.
Strongman (Ed.), International Review of Studies in Emotion (Vol. 1, pp. 143-162). Chich-
ester: Wiley.
Eysenck, H. J. (1942). The experimental study of the "Good Gestalt." A new approach. Psy-
chological Review, 49, 344-364.
Eysenck, H. J. (1968). An experimental study of aesthetic preference for polygonal figures. Jour-
nal of General Psychology, 79, 3-17.
Faber, R. J., O'Guinn, T. C., & Hardy, A. P. (1988). Art films in the suburbs. In B. A. Austin (Ed.),
Current research in film audiences, economics and law (pp. 4 4 - 5 3 ) . Norwood, NJ: Ablex.
Fechner, G. T. (1876). Vorschule der Asthetik [Preliminaries of esthetics]. Leipzig, Germany:
Breitkopf und Hartel.
Fenigstein, A., & Heyduk, R. G. (1985). Thought and action as determinants of media exposure.
In D. Zillmann & J. Bryant (Eds.), Selective exposure to communication (pp. 113-139).
Hillsdale, NJ: Lawrence Erlbaum Associates.
Feshbach, S. (1976). The role of fantasy in the response to television. Journal of Social Issues,
32, 71-84.
Feshbach, S., & Singer, R. (1971). Television and aggression. San Francisco: Jossey-Bass.
Festinger, L. (1957). A theory of cognitive dissonance. Evanston, IL: Row & Petersson.
Festinger, L. (1961). The psychological effect of insufficient reward. American Psychologist, 16,
1-11.
Fisher, W. A., & Byrne, D. (1978). Individual differences in affective, evaluative and behavioral
responses to an erotic film. Journal of Applied Social Psychology, 8, 355-365.
Fiske, J. (1987). Television culture. London & New York: Methuen.
Fiske, S. T. (1982). Schema triggered affect: Application to social perception. In M. S. Clark & S. T.
Fiske (Eds.), Affect and Cognition (pp. 55-78). Hillsdale, NJ: Lawrence Erlbaum Associates.
Fiske, S. T. (1988). Compare and contrast: Brewer's dual proces model and Fiske et al.'s con-
tinuum model. In T. K. Srull & R. S. Wyer (Eds.), Advances in Social Cognition: Vol. I. A
dual process model of impression formation (pp. 65-76). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Fiske, S. T., & Neuberg, S. L. (1990). A continuum model of impression formation from cate-
gory-based to individuality responses: Influences of informatica and motivation on attention
and interpretation. In M. P. Zanna (Ed.), Advances in Experimental Social Psychology (Vol.
2, pp. 1-79). New York: Academic Press.
Fiske, S. T., & Taylor, S. E. (1991). Social cognition (2nd ed.). New York: McGraw-Hill.
Fodor, J. A. (1983). The modularity of mind. Cambridge: MIT Press.
Fokkema, D. W. (1989). The concept of convention in literary theory and empirical research. In
260 REFERENCES

D'Haen, T., Griibel, R., & Lethen, H. (Eds.), Convention and innovation in literature
(pp. 89-101). Amsterdam: John Benjamins.
Folkman, S., & Lazarus, R. S. (1988). Coping as a mediator of emotion. Journal of Personality
and Social Psychology, 54, 466-475.
Forgas, J. P. (1982). Episode cognition: Internal representations of interaction routines. In
L. Berkowitz (Ed.), Advances in Experimental Social Psychology, Vol 15 (pp. 59-101). New
York: Academic Press.
Forster, E. M. (1927). Aspects of the novel. London: Arnold.
Fowles, J. (1992). Why viewers watch (rev. ed.). Newbury Park: Sage.
Freud, S. (1900). Die Traumdeutung [The interpretation of dreams]. Leipzig, Germany:
Deuticke.
Freud, S. (1908). Der Dichter und das Phantasieren [The poet and phantasy]. Neue Revue, 1,
716-724.
Freud, S. (1913). Totem und Tabu. Leipzig, Germany: Heller. (Original work published 1912)
Freud, S. (1919). Das Unheimliche [The uncanny]. Imago, 5, 297-324.
Freud, S. (1940). Der Witz und seine Beziehung zum Unbewussten [The joke and its relations
to the unconscious]. In Gesammelte Werke, Vol. 6. London: Imago. (Original work published
1905)
Freud, S. (1942). Psychopathic characters on the stage. Psychoanalytic Quarterly, 11, 4 5 9 -
464.
Freytag, G. (1911). Die Technik des Dramas. Leipzig, Germany: Hirzel.
Frijda, N. H. (1956). Het begrijpen van gelaatsexpressies [Understanding facial expressions].
Amsterdam, The Netherlands: Van Oorschot.
Frijda, N. H. (1981). Menselijke natuur en informatie als determinanten van gedrag [Human
nature and information as determinants of behavior]. Nederlands Tijdschrift voor de Psy-
chologie, 36, 135-150.
Frijda, N. H. (1986). The emotions. Cambridge, MA: Cambridge University Press.
Frijda, N. H. (1987). Emotions, cognitive structure and action tendency. Cognition an Emotion,
1, 115-144.
Frijda, N. H. (1988). The laws of emotion. American Psychologist, 43, 349-358.
Frijda, N. H. (1989). Aesthetic emotions and reality. American Psychologist, 44, 1546-1547.
Frijda, N. H. (1993). Appraisal and beyond: The issue of cognitive determinants of emotion [Spe-
cial issue]. Cognition and Emotion, 7, 357-387.
Frijda, N. H., Kuipers, P., & ter Schure, E. (1989). Relations among emotion, appraisal, and emo-
tional action readiness. Journal of Personality and Social Psychology, 57, 212-228.
Frijda, N. H., Mesquita, B., Sonnemans, J., & van Goozen, S. (1991). The duration of affective
phenomena or emotions, sentiments and passions. In K. T. Strongman (Ed.), International
Review of Studies on Emotion (Vol. 1, pp. 187-226). Chichester: Wiley.
Frijda, N. H., & Swagerman, J. (1987). Can computers feel? Cognition and Emotion, 1, 2 3 5 -
258.
Frith, U., & Robson, J. E. (1975). Perceiving the language of films. Perception, 4, 97-103.
Gaver, W. W., & Mandler, G. (1987). Play it again Sam: On liking music. Cognition and Emo-
tion, 1, 259-282.
Gee, J. P., & Grosjean, F. (1984). Empirical evidence for narrative structure. Cognitive Science,
8, 59-85.
Gernsbacher, M. A., Goldsmith, H. H., & Robertson, R. R. W. (1992). Do readers mentally rep-
resent characters' emotional states? Cognition and Emotion, 6, 89-111.
Gerrig, R. J. (1989). Suspense in the absence of uncertainty. Journal of Memory and Language,
28, 633-648.
Gerrig, R. J., & Prentice, D. A. (1987, August). Persuasion by fiction. Paper presented at the
95th Annual Convention of the American Psychological Association, New York City.
REFERENCES 261

Gibson, J. J. (1979). The ecological approach to visual perception. Boston: Houghton Mifflin.
Gilbert, D. T., Tofarodi, R. W., & Malone, P. S. (1993). You can't not believe everything you read.
Journal of Personality and Social Psychology, 65, 221-234.
Gilligan, S. G., & Bower, G. H. (1984). Cognitive consequences of emotional arousal. In C. E.
Izard, J. Kagan, & R. B. Zajonc (Eds.), Emotion, cognition & behavior (pp. 547-588). Cam-
bridge, MA: Cambridge University Press.
Godkewitsch, M. (1974). Correlates of humor: Verbal and nonverbal aesthetic reactions as func-
tions of semantic distance with adjective-noun pairs. In D. Berlyne (Ed.), Studies in the new
experimental aesthetics: Steps toward an objective psychology of aesthetic appreciation
(pp. 279-304). New York: Wiley.
Goldman, M., & Hagen, M. A. (1978). The forms of caricature: Physiognomy and political bias.
Studies on the Anthropology of Visual Communication, 2, 3 0 - 3 6 .
Gombrich, E. H. (1977). Art and illusion (5th ed.). London: Phaidon Press. (Original work pub-
lished 1960)
Gombrich, E. H. (1970). The mask and the face: The perception of physiognomic likeness in life
and art. In E. H. Gombrich, J. Hochberg, & M. Black (Eds.), Art, perception and reality (pp.
1-46). Baltimore: Johns Hopkins University Press.
Graesser, A. C. (1981). Prose comprehension beyond the word. New York: Springer.
Graesser, A. C., & Clark, L. F. (1985). Structures and procedures of implicit knowledge. Nor-
wood, NJ: Ablex.
Graesser, A. C., Higginbotham, M. W., Robertson, S. P., & Smith, W. R. (1978). A natural inquiry
into the National Enquirer: Self-induced versus task-induced reading comprehension. Dis-
course Processes, 1, 355-372.
Granberg, D., & King, M. (1980). Cross-lagged panel analysis of the relation between attraction
and perceived similarity. Journal of Experimental Social Psychology, 16, 573-581.
Gray, J. A. (1982). The neuropsychology of anxiety. Cambridge, England: Cambridge Univer-
sity Press.
Greenwald, A. G. (1981). Self and memory. In G. H. Bower (Ed.), The psychology of learning
and motivation (Vol. 15, pp. 201-237). New York: Academic Press.
Greimas, A. J. (1966). Elements pour une theorie de 1'interpretation du recit mythique [Ele-
ments of a theory of the interpretation of mythical stories]. Communications, 8, 28-59.
Grice, P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and seman-
tics: Vol 3. Speech acts (pp. 41-58). New York: Academic Press.
Grice, P. (1978). Further notes on logic and conversation. In P. Cole (Ed.), Syntax and seman-
tics: Vol. 9. Pragmatics (pp. 113-128). New York: Academic Press.
Gunning, T. (1989). An aesthetic of astonishment. Early film and the (In-)credulous spectator.
Art & Text, 34, 34-45.
Gunter, B. (1985). Determinants of television viewing preferences. In D. Zillmann & J. Bryant
(Eds.), Selective exposure to communication (pp. 108-109). Hillsdale, NJ: Lawrence Erl-
baum Associates.
Gunter, B., & Wober, M. (1983). Television viewing and public trust. British Journal of Social
Psychology, 22, 174-176.
Giinzenhaiiser, R. (1965). Das asthetische Mass Birkhoffs in informationsasthetischer Sicht
[Birkhoff's esthetic measure from the point of view of information esthetics]. In H. Ronge
(Ed.), Kunst und Kybernetik (pp. 37-55). Koln, Germany: Dumont.
Gurvitch, A. (1973). The sociology of the theatre. In E. Burns & T. Burns (Eds.), Sociology of
literature and drama (pp. 71-81). Harmondsworth: Penguin. (Original work published
1956)
Hagen, M. A., & Perkins, D. (1983). A refutation of the hypothesis of the superfidelity of cari-
catures relative to photographs. Perception, 12, 55-61.
 262 REFERENCES

Hamilton, O. C., & Rose, T. L. (1980). Illusory correlation and the maintainance of stereotype
beliefs. Journal of Personality and Social Psychology, 38, 382-845.
Handel, L. M. (1950). Hollywood looks at its audience: A report of film audience research. Ur-
bana: University of Illinois Press.
Harvey, W. J. (1965). Character and the novel. New York: Oxford University Press.
Hearst, E., & Jenkins, H. M. (1974). Sign-tracking: The stimulus-reinforcer relation and di-
rected action. Austin, TX: The Psychonomic Society.
Heath, S. (1981). Questions of cinema. London: MacMillan Press.
Heider, F., & Simmel, M. (1944). An experimental study of apparent behavior. American Jour-
nal of Psychology, 57, 243-259.
Hekkert, P., & van Wieringen, P. C. W. (1990). Complexity and prototypicality as determinants
of the appraisal of cubist paintings. British Journal of Psychology, 80, 483-496.
Hidi, S., & Baird, W. (1986). Interestingness: A neglected variable in discourse processing. Cog-
nitive Science, 10, 179-194.
Hilgard, E. R. (1962). Impulsive versus realistic thinking: An examination of the distinction be-
tween primary and secondary processes. Psychological Bulletin, 59, 477-488.
Hilgard, E. R. (1986). Divided consciousness (2nd ed.). New York: Wiley.
Hochberg, J. (1970). Attention, organization and consciousness. In D. I. Mostofsky (Ed.), At-
tention: Contemporary theory and analysis (pp. 99-124). New York: Apple ton-Century-
Crofts.
Hochberg, J. (1972). The representation of things and people. In E. H. Gombrich, J. Hochberg,
& M. Black (Eds.), Art, perception and reality (pp. 4 3 - 4 4 ) . Baltimore: Johns Hopkins Uni-
versity Press.
Hochberg, J. (1978). Perception (2nd ed.). Englewood Cliffs, NJ: Prentice-Hall.
Hochberg, J. (1986). Representation of motion and space in video and cinematic displays. In
K. R. Boff, L. Kaufman, & J. P. Thomas (Eds.), Handbook of Perception and Human Per-
formance: Vol. I. Sensory processes and perception (pp. 2 2 / 1 - 2 2 / 6 4 ) . New York: Wiley.
Hochberg, J., & Brooks, V. (1978). Film cutting and visual momentum. In J. W. Senders, D. F.
Fischer, & R. A. Monty (Eds.), Eye movements and the higher psychological functions (pp.
293-313). Hillsdale, NJ: Lawrence Erlbaum Associates.
Hoffman, M. L. (1982). Development of prosocial behavior: Empathy and guilt. In N. Eisenberg
(Ed.), The development of prosocial behavior (pp. 281-313). New York: Academic Press.
Hoffman, M. L. (1984). Interaction of affect and cognition in empathy. In C. E. Izard, J. Kagan,
& R. B. Zajonc (Eds.), Emotions, cognition, and behavior (pp. 103-131). Cambridge: Cam-
bridge University Press.
Hoffner, C., & Cantor, J. (1991). Perceiving and responding to mass media characters. In
J. Bryant & D. Zillmann (Eds.), Responding to the screen. Reception and reaction processes
(pp. 135-168). Hillsdale, NJ: Lawrence Erlbaum Associates.
Horowitz, M. J. (1976) Stress response syndromes. New York: Jason Aronson.
Houston, C. E., & Wegner, D. M. (1993). Believe it or not: Ironic processes in belief formation.
Paper presented at the annual meeting of the Eastern Psychological Association, Arlington,
VA.
Huitema, J. S., Dopkins, S., Klin, C. M., & Myers, J. L. (1993). Connecting goals and actions
during reading. Journal of Experimental Psychology: Learning, Memory and Cognition, 19,
1053-1060.
Humphrey, N. K. (1973). The illusion of beauty. Perception, 2, 420-439.
Isen, A. M. (1984). Toward understanding the role of affect in cognition. In R. Wyer & T. Srull
(Eds.), Handbook of Social Cognition (pp. 179-236). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Isen, A. M., Shalker, T. E., Clark, M., & Karp, L. (1978). Affect, accessibility of material in mem-
ory and behavior: A cognitive loop. Journal of Personality and Social Psychology, 36, 1-12.
REFERENCES 263

Izard, C. E. (1977). Human emotions. New York: Plenum Press.

Izard, C. E. (1989). The structure and function of emotions: Implications for cognition, motiva-
tion, and personality. In I. S. Cohen (Ed.), The G. Stanley Hall Lecture Series (Vol. 9, pp.
35-73). Washington DC: American Psychological Association.
Izard, C. E. (1992). Basic emotions, relations among emotions, and emotion-cognition relations.
Psychological Review, 99, 561-565.
Izard, C. E. (1993). Four systems for emotion activation: Cognitive and noncognitive processes.
Psychological Review, 100, 68-90.
James, W. (1950). The principles of psychology: Vol. 2. New York: Dover. (Original work pub-
lished 1890)
Jarvie, I. C. (1970). Movies and reality. New York: Basic Books.
Johansson, G. (1950). Configurations in event perception. Uppsala, Sweden: Almquist &
Wiksell.
Johnson-Laird, P. N. (1983). Mental models. Cambridge: Cambridge University Press.
Jose, P. E., & Brewer, W. F. (1984). Development of story liking: Character identification, sus-
pense and outcome resolution. Developmental Psychology, 20, 911-924.
Jowett, G., & Linton, M. J. (1989). Movies as mass communications. Newbury Park, CA: Sage.
Kaplan, E. A. (1983). Women and film: Both sides of the camera. New York: Methuen.
Kaplan, S., & Kaplan, R. (1982). Cognition and environment: Functioning in an uncertain
world. NY: Praeger.
Kay, P. (1983). Three properties of the ideal reader (Tech. Rep. No. 7). Berkeley CA: Berkeley
Cognitive Science Program.
Kermode, F. (1967). The sense of an ending: Stories in the theory of fiction. New York: Oxford
University Press.
Kiewiet, R., & Tan, E. (in preparation). Effects of velocity, temporal contingency and shape on
perceived causality and judgment of expressive qualities.
Kintsch, W. (1980). Learning from text, levels of comprehension, or: Why anyone would read a
story anyway. Poetics, 9, 87-98.
Klinger, E. (1971). Structure and function of fantasy. New York: Wiley.
Klinger, E. (1977). Meaning and void: Inner experience and the incentives in people's lives.
Minneapolis: University of Minnesota Press.
Knowles, P. L., & Smith, D. L. (1979). The ecological perspective applied to social perception:
Revision of a working paper. Journal of the Theory of Social Behaviour, 12, 53-78.
Konijn, E. (1994). Acteurs spelen emoties [Actors acting emotions]. Amsterdam, The Nether-
lands: Boom.
Koriat, A., Melkman, R., Averill, J. R., & Lazarus, R. S., (1972). The self-control of emotional re-
actions to a stressful film. Journal of Personality, 40, 601-619.
Kracauer, S. (1947). From Caligari to Hitler. A psychological history of the German film.
Princeton, NJ: Princeton University Press.
Kracauer, S. (1960). Theory offilm. The redemption of physical reality. New York: Oxford Uni-
versity Press.
Kreitler, H., & Kreitler, S. (1972). Psychology of the Arts. Durham, NC: Duke University Press.
Kretsinger, E. A. (1952). Gross bodily movement as an index of audience interest. Speech Mono-
graphs, 19, 244-348.
Kubovy, M. (1986). The psychology of perspective and Renaissance art. Cambridge: Cambridge
University Press.
Kuhn, A. (1982). Women's pictures: Feminism and cinema. London: Routledge & Kegan Paul.
Kunda, Z., Miller, D. T., & Claire, T. (1990). Combining social concepts: The role of causal rea-
soning. Cognitive Science, 14, 551-578.
Lang, P. J. (1984). Cognition in emotion: Concept and action. In C. E. Izard, J. Kagan, & R. B.
264 REFERENCES

Zajonc (Eds.), Emotions, cognition and behavior (pp. 192-226). Cambridge: Cambridge
University Press.
Lang, R. (1989). American film melodrama. Princeton, NJ: Princeton University Press.
Langer, S. K. (1953). Feeling and form. New York: Scribner.
Langlois, J. H., & Roggman, L. A. (1990). Attractive faces are only average. Psychological Sci-
ence, 1(2), 115-121.
Lanser, S. S. (1981). The narrative act. Point of view in prose fiction. Princeton, NJ: Prince-
ton University Press.
Lasher, M. D., Carroll, J. M., & Bever, T. G. (1983). The cognitive basis of aesthetic experience.
Leonardo, 16, 196-199.
Lazarus, R. S. (1966). Psychological stress and the coping process. New York: McGraw-Hill.
Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford University Press.
Lazarus, R. S., & Alfert, E. (1964). Short-circuiting of threat by experimentally altering cogni-
tive appraisal. Journal of Abnormal and Social Psychology, 69, 195-205.
Lazarus, R. S., Speisman, J. C., Mordkoff, A. M., & Davidson, L. A. (1962). A laboratory study
of psychological stress produced by a motion picture film. Psychological Monographs: Gen-
eral and Applied, 76(34, Whole No. 553).
Lazier, G., Zahn D., & Bellinghiere, J. (1974). Empirical analysis of dramatic structure. Speech
Monographs, 41, 381-390.
Leech, G. N. (1983). Principles of Pragmatics. London: Longman.
Lehnert, W. G. (1981). Plot units and narrative summarization. Cognitive Science, 4, 293-331.
Lehnert, W. G. (1982). Plot Units: A narrative summarization strategy. In W. G. Lehnert & M. H.
Ringle (Eds.), Strategies for natural language processsing (pp. 375-414). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Lehnert, W. G., & Vine, E. W. (1987). The role of affect in narrative structure. Cognition and
Emotion, 1, 299-322.
Lerner, M. J. (1980). The belief in a just world: A fundamental delusion. New York: Plenum.
Levenson, R. W., Ekman, P. E., & Friesen, W. V. (1990). Voluntary facial action generates emo-
tion-specific autonomic nervous system activity. Psychophysiology, 27, 363-384.
Leventhal, H. (1979). A perceptual-motor processing model of emotion. In P. Pliner, K. Blanken-
stein, & I. M. Spiegel (Eds.), Perception of emotion in self and others (Vol 5, pp. 1 - 4 6 ) . New
York: Plenum.
Levy, S. G., & Fenley, W. F. (1979). Audience size and likelihood and intensity of response dur-
ing a humorous movie. Bulletin of the Psychonomic Society, 13, 409-412.
Lewin, B. (1950). Inferences from the dream screen. The Yearbook of Psychoanalysis, 6, 104-
117.
Lewis, M. (1964a). Some non-decremental effects of effort. Journal of Comparative and Physi-
ological Psychology, 57, 367-372.
Lewis, M. (1964b). The effect of effort on value: An exploratory study of children. Child Devel-
opment, 35b, 1337-1342.
Lichtenstein, A., & Rosenfeld, L. B. (1983). Uses and misuses of gratifications research: An ex-
plication of media functions. Communication Research, 10, 97-109.
Light, L. L., Hollander, S., & Kayra-Stuart, F. (1981). Why attractive people are harder to re-
member. Personality and Social Psychology Bulletin, 7, 269-276.
Lipps, T. (1906). Leitfaden der Psychologie. Leipzig, Germany: Engelmann.
Loewenstein, G. F., & Prelec, D. (1993). Preferences for sequences of outcomes. Psychological
Review, 100, 91-108.
Long, D. L., Golding, J. M., Graesser, A. C., & Clark, L. F. (1990). Goal, event, and state infer-
ences: An investigation of inference generaton during story comprhension. In A. C. Graesser
& G. H. Bower (Eds.), Inferences and text comprehension (pp. 89-107). San Diego: Aca-
demic Press.
REFERENCES 265

Lucariello, J. (1994). Situational irony: A concept of events gone awry. Journal of Experimen-
tal Psychology: General, 124, 129-145.
Lynch, F. D. (1972). Clozentropy: A technique for studying audience response to films. New
York: Arno Press.
MacCabe, G. (1976). Theory and film: Principles of realism and pleasure. Screen, 17, 3.
Mandler, G. (1982). The structure of value: Accounting for taste. In M. S. Clark & S. T. Fiske
(Eds.) Affect and Cognition (pp. 3-36). Hillsdale, NJ: Lawrence Erlbaum Associates.
Mandler, G. (1984). Mind and body. New York: Norton.
Mandler, J. M. (1982). Some uses and abuses of a story grammar. Discourse Processes, 5, 3 0 5 -
318.
Mandler, J. M. (1984). Stories, scripts, and scenes: Aspects of schema theory. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Markus, H. (1977). Self-schemata and processing information about the self. Journal of Per-
sonality and Social Psychology, 35, 63-78.
Markus, H., & Nevius, P. (1987). Possible selves: The interface between motivation and the self-
concept. In K. Yardley & T. Honess (Eds.), Self & identity. Psychosocial perspectives (pp.
157-172). Chichester: Wiley.
Martin, M. (1990). On the induction of mood. Clinical Psychology Review, 10, 669-697.
Martin, W. (1986). Recent theories of narrative. Ithaca, NY: Cornell University Press.
Martindale, C., & Moore, K. (1988). Priming, prototypicality, and preference. Journal of Experi-
mental Psychology: Human Perception and Performance, 14, 661-670.
Matthews, B. (1910). A study of the drama. London: Longman & Green.
McArthur, L. Z., & Baron, R. M. (1983). Toward an ecological theory of social perception. Psy-
chological Review, 90, 215-238.
McClelland, D. C. (1955). The psychology of mental content reconsidered. Psychological Re-
view, 62, 297-302.
McGhee, P. E. (1971). Development of the humor response: A review of the literature. Psycho-
logical Bulletin, 76, 328-348.
McGhee, P. E. (1972). On the cognitive origins of incongruity humor: Fantasy assimilation ver-
sus reality assimilation. In J. H. Goldstein & P. E. McGhee (Eds.), The psychology of humor
(pp. 61-80). London: Academic Press.
McGuiness, D., & Pribram, K. H. (1980). The neuropsychology of attention: Emotional and mo-
tivational contrast. In M. C. Witrock (Ed.), The brain and psychology (pp. 95-139). New
York: Academic Press.
McGuire, W. J. (1974). Psychological motives and communication gratification. In J. G. Blum-
ler & E. Katz (Eds.), The uses of mass communications. Current perspectives on gratifi-
cations research (pp. 167-196). Beverly Hills: Sage.
McKoon, G., & Ratcliff, R. (1990). Dimensions of inference. In A. C. Graesser & G. H. Bower
(Eds.), Inferences and text comprehension (pp. 313-328). San Diego, CA: Academic Press.
McKoon, G., & Ratcliff, R. (1992). Inference during reading. Psychological Review, 99, 4 4 0 -
466
McQuail, D. (1987). Mass communication theory (2nd ed.). London: Sage.
Meadowcroft, J., & Zillmann, D. (1984). The influences of hormonal fluctuations on women's
selection and enjoyment of television programs. Paper presented at the meeting of the Asso-
ciation for Education in Journalism and Mass Communication, Gainesville, FL.
Meier, N. C. (1950). The Meier audience response machine. American Journal of Psychology,
63, 87-89.
Mervis, C. B. (1980). Category structure and the development of categorization. In R. J. Spiro,
B. C. Bruce, & W. F. Brewer (Eds.), Theoretical issues in reading comprehension (pp. 279-
303). Hillsdale, NJ: Lawrence Erlbaum Associates.
Messaris, P. (1994). Visual literacy: Image, mind and reality. Boulder: Westview Press.
266 REFERENCES

Metz, C. (1968). Essays sur la signification au cinema. Paris: Klincksieck.

Metz, C. (1971). Langage et cinema. Paris: Larousse.
Metz, C. (1972) Essais sur la signification au cinema (Vol. 2). Paris, France: Klincksieck.
Metz, C. (1975a). Le signifiant imaginaire. Communications, 23, Psychanalyse et Cinema, 3 -
55.
Metz, C. (1975b). Le film de fiction et son spectateur. Communications, 23, Psychanalyse et
Cinema, 108-135.
Meyer, L. B. (1956). Emotion and meaning in music. Chicago: University of Chicago Press.
Miall, D. S. (1989). Beyond the schema given: Affective comprehension of literary narratives.
Cognition and Emotion, 3, 55-78.
Michotte, A. (1946). La perception de la causalite. Louvain, Belgium: Presses Universitaire de
Louvain.
Michotte, A. (1950). The emotions regarded as functional connections. In M. L. Reymert (Ed.),
Feelings and emotions: The Mooseheart Symposion (pp. 114-125). New York: McGraw-Hill.
Michotte, A. (1991). The character of "reality" of cinematographic projections (A. P. Costall,
Trans.). In G. Thines, A. Costall, & G. Butterworth (Eds.), Michotte's experimental phe-
nomenology of pereception (pp. 197-209). Hillsdale, NJ: Lawrence Erlbaum Associates.
(Original work published 1948)
Michotte, A. (1991). The emotional involvemnt of the spectator in the action represented in a
film: Toward a theory (A. P. Costall, Trans.). In G. Thines, A. Costall, & G. Butterworth
(Eds.), Michotte's experimental phenomenology of perception (pp. 209-217). Hillsdale, NJ:
Lawrence Erlbaum Associates. Original work published (1953)
Miller, N. E. (1944). Experimental studies of conflict. In J. McV. Hunt (Ed.), Personality and the
behavioral disorders. (Vol. 1, pp. 431-465). New York: Ronald Press.
Miller, R. S. (1987). Empathic embarrassment: Situational and personal determinants of reac-
tions to the embarrassment of another. Journal of Personality and Social Psychology, 53,
1061-1069.
Milner Davis, J. R. (1977). A structured approach to humour and farce. In A. J. Chapman &
H. C. Foot (Eds.), It's a funny thing, humour (pp. 391-394). Oxford: Pergamon Press.
Mitry, J. (1965). Esthetique et psychologie du cinema. Tome 2: Les formes. Paris, France: Edi-
tions Universitaires.
Moller, K. E. K., & Karppinen, P. (1983). Consumer motion picture choice. Journal of Economic
Psychology, 4, 239-262.
Morin, E. (1956). Le cinema, ou Vhomme imaginaire [The cinema, or imaginary man]. Paris,
France: Minuit.
Morreal, J. (1983). Humor and emotion. American Philosophical Quarterly, 20, 297-305.
Morrow, D. G., Bower, G. H., & Greenspan, S. L. (1989). Updating situation models during nar-
rative comprehension. Journal of Memory and Learning, 28, 292-312.
Morrow, D. G., Greenspan, S. L., & Bower, G. H. (1987) Accessibility and situation models in
narrative understanding. Journal of Memory and Language, 26, 165-187.
Movshon, A. (1990). Visual processing of moving images. In H. Barlow, C. Blakemore, & M. Wes-
ton-Smith (Eds.), Images and understanding (pp. 122-137). New York: Cambridge Univer-
sity Press.
Moynihan, C., & Mehrabian, A. (1981). The psychological aesthetics of narrative forms. In H. I.
Day (Ed.), Advances in intrinsic motivation and aesthetics (pp. 323-340). New York:
Plenum.
Mulvey, L. (1975). Visual pleasure and narrative cinema. Screen, 16(3), 6-18.
Mulvey, L. (1981, Summer). On 'Duel in the sun'. Framework (15/16/17), 12-15.
Munsterberg, H. (1970). The film, a psychological study. New York: Dover. (Original work pub-
lished 1916)
Nachbar, J. (1974). Focus on the Western. Englewoods Cliffs, NJ: Prentice-Hall.
REFERENCES 267

Neale, S. (1980). Genre. London: BFI.

Neale, S., & Krutnik, F. (1990). Popular film and television comedy. London & New York:
Routledge.
Neisser, U., & Beeklen, R. (1975). Selective looking: Attending to visually specified events. Cog-
nitive Psychology, 7, 480-494.
Nell, V. (1988). Lost in a book. New Haven, CT: Yale University Press.
Newell, A., & Simon, H., (1972). Human problem solving. Englewood Cliffs, NJ: Prentice-Hall.
Nomikos, M. S., Opton, E., Averill, J. R., & Lazarus, R. S. (1968). Surprise versus suspense in
the production of stress reaction. Journal of Personality and Social Psychology, 8, 204-208.
Nunnally, J. C. (1981). Explorations of exploration. In H. I. Day (Ed.), Advances in intrinsic mo-
tivation and aesthetics (pp. 87-129). New York: Plenum.
Oatley, K. (1992). Best laid schemes: The psychology of emotion. Cambridge, England: Cam-
bridge University Press.
Oatley, K. (1995). A taxonomy and theory of the emotions of literary response and a theory of
identification in fictional narrative. Poetics, 23(1-2), 53-74.
Obrist, P. A. (1981). Cardiovascular psychophysiology: A perspective. New York: Plenum.
Ornstein, R. (1969). On the experience of time. Harmondsworth: Penguin.
Ortony, A. (1978). Remembering, understanding and representation. Cognitive Science, 2, 5 3 -
69.
Ortony, A. (1979). Beyond literal similarity. Psychological Review, 86, 161-180.
Ortony, A. (1980). Metaphor. In R. J. Spiro, B. C. Bruce, & W. F. Brewer (Eds.), Theoretical issues
in reading comprehension (pp. 349-366). Hillsdale, NJ: Lawrence Erlbaum Associates.
Ortony, A., Clore, G. L., & Collins, A. M. (1988). The cognitive structure of emotion. Cambridge:
Cambridge University Press.
Ortony, A., Clore, G. L., & Foss, M. A. (1987). The referential structure of the affective lexicon.
Cognitive Science, 11, 361-384.
Ortony, A. & Turner, T. J. (1990). What's basic about basic emotions? Psychological Review,
97, 315-331.
Palmer, J. (1987). The logic of the absurd. London: BFI.
Palmgreen, P., Cook, P. L., Harvill J. G., & Helm, D. M. (1988). The motivational framework of
moviegoing: Uses and avoidances of theatrical films. In B. A. Austin (Ed.), Current research
in film: Audiences, economics and law (pp. 1 - 2 3 ) . Norwood, NJ: Ablex.
Palmgreen, P, Wenner, L. A., & Rayburn, J. D. (1981). Gratification discrepancies and news pro-
gram choice. Communication Research, 8, 451-478.
Panksepp, J. (1982). Towards a general psychological theory of emotions. The Behavioral and
Brain Sciences, 5, 407-467.
Panksepp, J. (1992). A critical role for affective neuroscience in resolving what is basic about
basic emotions. Psychological Review, 99, 554-560.
Panofsky, E. (1991). Perspective as symbolic form. New York: Zone Books. (Original work pub-
lished 1927)
Parish, J. R., & Pitts, M. R. (1976). The great gangster pictures. Metuchen, NJ: Scarecrow Press.
Perkins, V. F. (1972). Film as film. Harmondsworth, Middlesex: Penguin.
Peters, J. M. (1977). Kijken naar beelden; psychologie van het mechanische beeld [Watching
images: The psychology of the mechanic image]. Leuven, Belgium: Centrum voor Communi-
catiewetenschappen.
Peters, J. M. (1981). Pictorial signs and the language of film. Amsterdam, The Netherlands:
Rodopi.
Peters, J. M. (1989). Het filmische denken [Filmic thinking]. Leuven, Belgium: Acco.
Petro, P. (1989). Joyless streets. Women and melodramatic representation in Weimar Ger-
many. Princeton, NJ: Princeton University Press.
 268 REFERENCES

Pfeifer, R. (1982). Cognition and emotion: An information processing approach. (CIP Work-
ing Paper No. 436). Pittsburgh: Carnegie Mellon University.
Pfister, M. (1977). Das Drama. Miinchen, Germany: Fink Verlag.
Philippot, P. (1993). Inducing and assessing differentiated emotion-feeling states in the labora-
tory. Cognition and Emotion, 7, 171-193.
Piet, S. (1987). What motivates stuntmen? Motivation and Emotion, 11, 195-213.
Plutchik, R. (1980). Emotion: A psychoevolutionary synthesis. New York: Harper & Row.
Plutchik, R. (1991). Emotions and evolution. In K. T. Strongman (Ed.), International review of
studies on emotion (Vol. 1, pp. 37-58). Chichester: Wiley.
Polti, G. (1977). The thirty-six dramatic situations. Boston: The Writer, Inc. (Original work pub-
lished 1921)
Porfirio, R. G. (1976). No way out: Existential motifs in the film noir. Sight and Sound, 45(4),
212-217.
Posner, M. I. (1975). Psychobiology of attention. In M. S. Gazzaniga & C. Blakemore (Eds.),
Handbook of psychobiology (pp. 441-480). New York: Academic Press.
Potter, W. J. (1988). Perceived reality in television effects research. Journal of Broadcasting &
Electronic Media, 32(1), 23-41.
Powdermaker, H. (1950). Hollywood: The dream factory. Boston: Little, Brown & Co.
Pratt, M. L. (1977). Toward a speech act theory of literary discourse. Bloomington: Indiana
University Press.
Preminger, A. (Ed.). (1965). Encyclopedia of poetry and poetics. Princeton: Princeton Univer-
sity Press.
Pribram, K. H., & McGuiness, D. (1975). Arousal, activation, and the effort in the control of at-
tention. Psychological Review, 82, 116-149.
Prokop, D. (1974). Soziologie des Films [Sociology of film]. Darmstadt/Neuwied, Germany:
Luchterhand Verlag.
Pryor, J. B. (1986). The influence of different encoding upon the formation of illusory correla-
tions and group processes. Personality and Social Psychology Bulletin, 12, 216-226.
Purcell, A. T. (1984). The aesthetic experience and mundane reality. In W. R. Crozier & A. J.
Chapman (Eds.), Cognitive processes in the perception of art (pp. 189-210). Amsterdam:
Elsevier.
Purves, A. C., & Beach, R. (1972). Literature and the reader: Research in response to litera-
ture, reading interest and the teaching of literature. Urbana, IL: National Council of Teach-
ers of English.
Piitz, P (1970). Die Zeit im Drama. Zur Technik Dramatischer Spannung [Time in drama: On
the technique of dramatic tension]. Gottingen, Germany: Vandenhoeck und Ruprecht.
Pylyshyn, Z. W. (1973). What the mind's eye tells the mind's brain: A critique of mental imagery.
Psychological Bulletin, 80, 1-24.
Pylyshyn, Z. W. (1984). Computation and cognition. Cambridge, MA: MIT Press.
Rabbie, J. M., & Visser, L. (1984). De'individuatie en de ontwikkeling van normen in groepen
toeschouwers, een veldexperiment [ Deindividuation and the development of norms in groups
of spectators: A field experiment]. In R. van der Vlist (Ed.), Sociale psychologie in Neder-
land: Vol. 4. Samenleving en individuen (pp. 102-127). Deventer, The Netherlands: Van
Loghum Slaterus.
Ramirez, J., Bryant, J., & Zillmann, D. (1982). Effects of erotica on retaliatory behavior as a func-
tion of level of prior provocation. Journal of Personality and Social Psychology, 43, 971-978.
Reisz, K., & Millar, G. (1968). The technique of film editing. London: Focal Press.
Ritchie, R. G. (1986). Momentary affect and attention allocation. Motivation and Emotion, 10,
387-395.
Rombouts, H. (1992). Echt verliefd [Really in love]. Meppel, The Netherlands: Boom.
REFERENCES 269

Rosch, E. (1978). Principles of categorization. In E. Rosch & B. L. Lloyd (Eds.), Cognition and
categorization (pp. 27-71). Hillsdale, NJ: Lawrence Erlbaum Associates.
Rosch, E., & Mervis, C. B. (1975). Family resemblances: Studies in the internal structure of cate-
gories. Cognitive Psychology, 7, 573-605.
Ruch, W. (1993). Exhilaration and humor. In M. Lewis & J. N. Haviland (Eds.), Handbook of
emotion (pp. 605-616). New York: The Guilford Press.
Ruimschotel, D. (1987). Explanation, causation and psychological theories. A methodological
study illustrated by an analysis of Festinger's theory of cognitive dissonance and Newell &
Simon's theory of human problem solving. Amsterdam, The Netherlands: Swets & Zeitlinger.
Rumelhart, D. E. (1977). Understanding and summarizing brief stories. In D. LaBerge & S. J.
Samuels (Eds.), Basic processes in reading: Perception and comprehension (pp. 265-303).
Hillsdale, NJ: Lawrence Erlbaum Associates.
Rumelhart, D. E. (1980). On evaluating story grammar. Cognitive Science, 4, 313-316.
Ryan, T., & Schwartz, C. (1956). Speed of perception as a mode of representation. American
Journal of Psychology, 69, 60-69.
Sartre, J.-P (1939). Esquisse d'une theorie des emotions. [Outline of a theory of the emotions].
Paris: Gallimard.
Sartre, J.-P. (1940). L'Imaginaire. Psychologie phenomenologique de Vimagination. [The imagi-
nary: Phemomenological psychology of the imagination]. Paris: Gallimard.
Schachter, S., & Wheeler, L. (1962). Epinephrine, chlorpromazine and amusement. Journal of
Abnormal and Social Psychology, 65, 121-128.
Schank, R. C. (1975). Conceptual information processing. Amsterdam, The Netherlands: North
Holland.
Schank, R. C. (1979). Interestingness: Controlling inferences. Artificial Intelligence, 12, 2 7 3 -
297.
Schank, R. C. (1982). Dynamic memory. A theory of reminding and learning in computers
and people. Cambridge, MA: Cambridge University Press.
Schank, R. C., & Abelson, R. P. (1977). Scripts, plans, goals & understanding. Hillsdale, NJ:
Lawrence Erlbaum Associates.
Schatz, T. (1981). Hollywood genres: Formulas, filmmaking and the studio system. Philadel-
phia: Temple University Press.
Scherer, K. (1988). Facets of emotion. Recent research. Hillsdale, NJ: Lawrence Erlbaum
Associates.
Scheff, T. J., & Scheele, S. C. (1980). Humor and catharsis: The effect of comedy on audiences.
In P. H. Tannenbaum (Ed.), The entertainment function of television (pp. 165-182). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Scheff, T. J. (1979). Catharsis in healing, ritual and drama. Berkeley: University of California
Press.
Schoedel, J., Frederickson, W. A., & Knight, J. M. (1975). An extrapolation of the physical at-
tractiveness and sex variables within the Byrne attraction paradigm. Memory & Cognition,
3, 527-530.
Schoenmakers, H. (1988). To be, wanting to be, forced to be: Identification processes in the-
atrical situations. In W. Sauter (Ed.), New directions in audience research: Advances in Re-
ception and Audience Research (Vol. 2, pp. 138-163). Utrecht, The Netherlands: Instituut
voor Theaterwetenschap.
Scholes, R., & Kellogg, R. (1966). The nature of narrative. Oxford: Oxford University Press.
Schram, D., & Frijda, N. (Eds.).(1995). Emotion and cultural products [Special issue]. Poetics,
23, 1 - 2 .
Searle, J. R. (1980). Minds, brains and programs. Behavioral and Brain Sciences, 3, 417-458.
Secord, P., Backman, C. W., & Eachus, H. T. (1964). Effects of imbalance in the self-concept on
perception of persons. Journal of Abnormal and Social Psychology, 68, 442-446.
270 REFERENCES

Sekuler, R., Anstis, S., Braddick, O. J., Brandt, T., Movshon, J. A., & Orban, G. (1990). The per-
ception of motion. In L. Spillman & J. S. Werner (Eds.), Visual perception: The neurophysi-
ological foundations (pp. 205-230). San Diego: Academic Press.
Selfridge, O. G., & Neisser, U. (1960). Pattern recognition by machine. Scientific American,
203(3), 60-68.
Sesonke, A. (1973). Cinema space. In D. Carr & E. S. Casey (Eds.), Explorations in phenome-
nology (pp. 399-409). The Hague, The Netherlands: Nijhoff.
Shweder, R. A., & D'Andrade, R. G. (1980). The systematic distortion hypothesis. In R. A. Shwe-
der & D. W. Fiske (Eds.), New directions for methodology of behavioral science (pp. 3 7 -
58). San Francisco: Jossey Bass.
Silver, A., & Ursini, J. (1974). Brief Encounter (1945). In David Lean and his films (pp. 3 9 -
51). London: Frewin.
Silver, A., & Ursini, J. (1993). The vampire film (2nd ed.). New York: Limelight Editions.
Simon, H. A. (1970). The sciences of the artificial Cambridge, MA: MIT Press.
Simons, R. F., & Zelson M. F. (1985). Engaging visual stimuli and reflex blink modification. Psy-
chophysiology, 22, 44-49.
Singer, J. L. (1973). The child's world of make-believe: Experimental studies of imaginative
play. New York: Academic Press.
Singer, J. L., & Singer, D. G. (1976). Family television viewing habits and the spontaneous play
of preschool children. American Journal of Orthopsychiatry, 46, 496-502.
Singer, J. L., & Singer, D. G. (1981). Television, imagination and aggression: A study of
preschoolers. Hillsdale, NJ: Lawrence Erlbaum Associates.
Smith, M. (1995). Character engagement: Fiction, emotion, and the cinema. Oxford, UK:
Clarendon Pres.
Solomon, R. L. (1980). The opponent process theory of acquired motivation. American Psy-
chologist, 35, 691-712.
Sonnemans, J. (1991). Structure and determinants of emotional intensity. Unpublished doc-
toral dissertation. Amsterdam, The Netherlands: University of Amsterdam.
Sorlin, P. (1977). Sociologie du cinema [Sociology of the cinema]. Paris: Autier Montaigne.
Souriau, E. (1950). Les deux cent milles situations dramatiques [200,000 dramatic situations].
Paris, France: Flammarion.
Srull, T. K., & Wyer, R. S., Jr. (1989). Person memory and judgment. Psychological Review, 96,
58-83.
Stam, R. (1989). Subversive pleasures: Bakhtin, cultural criticism, and film. Baltimore: Johns
Hopkins University Press.
Sternberg, M. (1978). Expositional modes and temporal ordering in fiction. Baltimore: Johns
Hopkins University Press.
Stotland, E. (1969). Exploratory studies of empathy. In Berkowitz, L. (Ed.), Advances in Ex-
perimental Social Psychology (Vol. 4, pp. 271-314). New York: Academic Press.
Stotland, E., Mathews, K. E., Sherman, S. E., Hansson, O. R., & Richardson, B. Z. (1978). Em-
pathy, fantasy and helping. Beverly Hills: Sage.
Stotland, E., & Mathews, K. E. (1978). Mrs. Reynolds needs a nurse. In E. Stotland, K. E. Math-
ews, S. E. Sherman, R. O. Hansson, & B. Z. Richardson (Eds.), Empathy, fantasy and help-
ing (pp. 51-61). Beverly Hills & London: Sage.
Strongman, K. T. (1978). The psychology of emotion (2nd ed.). Chichester: Wiley.
Suboski, M. D. (1990). Releaser-induced recognition learning. Psychological Review, 97, 271-
284.
Suh, S., & Trabasso, T. (1993). Inferences during reading: Converging evidence from discourse
analysis, talk-aloud protocols, and recognition priming. Journal of Memory and Language,
32, 279-300.
Suls, J. M. (1972). A two-step model for the appreciation of jokes and cartoons: An information
REFERENCES 271

processing analysis. In J. H. Goldstein & P. E. McGhee (Eds.), The psychology of humor (pp.
81-100). New York: Academic Press.
Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology, 33,
1-39.
Tamborini, R. (1991). Responding to horror: Determinants of exposure and appeal. In J. Bryant
& D. Zillmann (Eds.), Responding to the screen. Reception and reaction processes (pp. 3 0 5 -
328). Hillsdale, NJ: Lawrence Erlbaum Associates.
Tan, E. (1985). Beoordeling en beleving van "de Grens" [Appreciation and experiencing of "The
Frontier"] (Technical report on the reception of theatre and the media, 11). Amsterdamt,
The Netherlands: Instituten voor Theaterwetenschap.
Tan, E. (1986a). Interest when nothing happens. A note on narrative retardation. Proceedings
of the Eighth Annual Conference of the Cognitive Science Society (pp. 832-842). Hillsdale,
NJ: Lawrence Erlbaum Associates.
Tan, E. (1986b). What's in a line? Tijdschrift voor Theaterwetenschap, 4(16-17), 151-185.
Tan, E. (1990). Bespreking van D. Bordwell, Making meaning: Inference and rhetoric in the in-
terpretation of cinema. Skrien, 173, 64.
Tan, E. (1991). Filmpsychologie. In P. Bosma (Red.), Filmkunde. Een inleiding (pp. 267-280).
Nijmegen, The Netherlands: SUN.
Tan, E. (1994). Story processing as an emotion episode. In H. van Oostendorp & R. Zwaan (Eds.)
Naturalistic text processing (pp. 165-188). Norwood, NJ: Ablex.
Tan, E. (1995). Film induced affect as a witness emotion [Special issue]. Poetics, 23, (1-2), 7 -
32.
Tan, E. (in press). Constraint and convention in psychological film aesthetics. In J. Miiller (Ed.),
Proceedings of the Conference on Film Pragmaticy. Miinster, Germany: Nodus.
Tan, E. (in preparation). Exposing the affect structure of narrative film: Some empirical studies.
Tan, E., & de Wied, M. A. (1986). Empirische vaststelling van handelingseenheden in de film
"Straf" [Empirical assessment of action units in the film 'Punishment'] (Technical report on
the reception of theatre and the media, 18). Amsterdam, The Netherlands: Instituten voor
Theaterwetenschap.
Tan, E., & van den Boom, I. (1992). Explorations in the psychological affect structure of narra-
tive film. In E. Nardocchio (Ed.), Reader response to literature. The empirical dimension
(pp. 57-94). Berlin, Germany & New York: De Gruyter-Mouton.
Tan, E., & Diteweg, G. (in press). Expectation, suspense and other emotion in film viewing. In
P. Vorderer, H.J. Wulff, & M. Friedrichsen (Eds.), Analysis of suspense. Hillsdale (NJ): Law-
rence Erlbaum Associates.
Tan, E., Eggermont, L., & Joosten, P. (1989). On becoming a cinephile. A theoretical and em-
pirical analysis. Tijdschrift voor Theaterwetenschap, 7(27/28), 207-233.
Tan, E., Eggermont, L. J. M., Joosten, P. D. M., & Spinhof, H. (in preparation). Emotion in the
film viewer as a function of cinephilia and film genre.
Tan, E., & Schoenmakers, H. (1984). "Good guy and bad guy" effects in political theatre. In
H. Schmidt & A. van Kesteren (Eds.), Semiotics of drama and theatre (pp. 467-508). Am-
sterdam, The Netherlands: John Benjamins.
Tannenbaum, P. H. (1980). Entertainment as vicarious emotional experience. In P. H. Tannen-
baum (Ed.), The entertainment function of television (pp. 107-131). Hillsdale, NJ: Law-
rence Erlbaum Associates.
Tannenbaum, P. H. (1985). 'Play it again Sam.' Repeated exposure to television programs. In
D. Zillmann & J. Bryant (Eds.), Selective exposure to communication (pp. 225-241). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Tannenbaum, P. H., & Gaer, E. P. (1965). Mood change as a function of stress of protagonist and
degree of identication in a film viewing situation. Journal of Personality and Social Psy-
chology, 2, 612-616.
272 REFERENCES

Taylor, S. E. (1981). The interface of cognitive and social psychology. In J. Harvey (Ed.), Cog-
nition, social behavior, and the environment (pp. 189-211). Hillsdale (NJ): Lawrence Erl-
baum Associates.
Tellegen-van Delft, S. (1987). Waarom zouje lezen? [Why would you read?]. Groningen, The
Netherlands: Wolters-Noordhoff.
Thorndyke, R W. (1977). Cognitive structures in comprehension and memory of narrative dis-
course. Cognitive Psychology, 9, 77-110.
Todorov, T. (1971, Fall). The two priciples of narrative. Diacritics, 3 9 - 5 3 .
Tomkins, S. S. (1962). Affect, imagery, consciousness: Vol. 1. The positive affects. New York:
Springer.
Tomkins, S. S. (1984). Affect theory. In K. R. Scherer & P. Ekman (Eds.), Approaches to emo-
tion (pp. 163-195). Hillsdale, NJ: Lawrence Erlbaum Associates.
Turner, T. J., & Ortony, A. (1992). Basic emotions: Can conflicting criteria converge? Psycho-
logical Review, 99, 566-571.
Tversky, B., & Baratz, D. (1985). Memory for faces: Are caricatures better than photographs?
Memory & Cognition, 13, 45-49.
Uchiyama, F., Hanari, T., Ito, T., Takahashi, T., Okuda, K., Goto, T., & Tsuji, K. (1990). Patterns
of psychological and physiological responses for commom affects elicited by films. Psy-
chologia-An International Journal of Psychology in the Orient, 33, 36-41.
van den Berg, H. (1979). Teksten voor toeschouwers [Texts for spectators]. Muiderberg, The
Netherlands: Coutinho.
van der Run, J. I. M. (1970). Handelingsaspecten in het drama [Aspects of action in drama].
Amsterdam, The Netherlands: Polak & van Gennep. (Original work published 1938)
Van der Voort, T. H. A. (1982). Kinderen en TV-geweld; waarneming en beleving [Children and
TV violence: perception and experience]. Lisse, The Netherlands: Swets & Zeitlinger.
van der Wurff, A. (1992). Aard en achtergronden van onveiligheidsgevoelens in de woonom-
geving [The nature and background of feelings of insecurity in the residence]. Unpublished
doctoral dissertation. Amsterdam, The Netherlands: University of Amsterdam.
van Dijk, T. A. (1978). Recalling and summarizing complex discourse. In K. Holker & W. Burg-
hardt (Eds.), Text processing (pp. 49-118). Berlin, Germany & New York: De Gruyter.
van Dijk, T. A. (1981). Studies in the pragmatics of discourse. The Hague, The Netherlands:
Mouton.
van Dijk, T. A., & Kintsch, W. (1983). Strategies of discourse comprehension. San Diego, CA:
Academic Press.
van Vliet, H. (1991). De schone schijn [The beautiful apearance]. Unpublished doctoral disser-
tation. Utrecht, The Netherlands: University of Utrecht.
Vipond, D., & Hunt, R. A. (1984). Point driven understanding: pragmatic and cognitive dimen-
sions of literary reading. Poetics, 13, 261-277.
Warren, W. H., & Kurtz, K. J. (1992). The role of central and peripheral vision in perceiving the
direction of self-motion. Perception and Psychophysics, 51, 443-453.
Weaver, J. (1991). Responding to erotica: Perceptual processes and dispositional implications.
In J. Bryant & D. Zillmann (Eds.), Responding to the screen. Reception and reaction
processes (pp. 329-354). Hillsdale, NJ: Lawrence Erlbaum Associates.
Wegman, C. (1985). Psychoanalysis and cognitive psychology. London: Academic Press.
Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101, 34-52.
Weiner, B. (1972). Theories of motivation. Chigaco: Rand McNally.
White, R. W. (1959). Motivation reconsidered: The concept of competence. Psychological Re-
view, 66, 297-333.
Whitfield, T. W. A. (1983). Predicting preference for familiar, everyday objects: An experimen-
tal confrontation between two theories of aesthetic behaviour. Journal of Environmental
Psychology, 3, 221-237.
REFERENCES 273

Whitfield, T. W. A., & Slatter, R E. (1979). The effect of categorisation and prototypicality on
aesthetic choice in a furniture selection task. British Journal of Psychology, 70, 65-76.
Wilensky, R. (1982). Points: A theory of the structure of stories in memory. In W. G. Lehnert &
M. H. Ringle (Eds.), Strategies for natural language processing (pp. 345-374). Hillsdale,
NJ: Lawrence Erlbaum Associates.
Wilensky, R. (1983a). Planning and understanding. A computational approach to human rea-
soning. Reading, MA: Addison-Wesley.
Wilensky, R. (1983b). Story grammars versus story points. Behavioral and Brain Sciences, 6,
579-623.
Wilson, S. G., Rinck, M., McNamara, T. P., Bower, G. H., & Morrow, D. G. (1993) Mental mod-
els and narrative comprehension. Journal of Memory and Language, 32, 141-154.
Winston, P. H., & Horn, B. K. P. (1981). LISP. Reading, MA: Addison-Wesley.
Wollen, P. (1972). Signs and meaning in the cinema. London: Seeker & Warburg.
Woods, D. J., Beecher, G. P., & Ris, M. D. (1978). The effects of stressful arousal on conjugate
lateral eye movement. Motivation and Emotion, 2, 345-353.
Wright, H. F. (1937). The influence of barriers on strength of motivation. Contributions to Psy-
chological Theory, 1 (Whole No. 3).
Wright, W. (1975). Sixguns and society. Berkeley: University of California Press.
Wyatt, R. 0., & Badger, D. P. (1988). What newspaper critics value in film and film criticism: a
national survey. In B. A. Austin (Ed.), Current research in film (Vol 4, pp. 54-71). Norwood,
NJ: Ablex.
Wyer, R. S., & Gordon, S. E. (1984). The cognitive representation of social information. In R. S.
Wyer & T. K. Srull (Eds.), Handbook of social cognition (Vol. 2, pp. 73-149). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Wyer, R. S., & Srull, T. K. (1986). Human cognition in the social context. Psychological Review,
53, 322-359.
Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social
Psychology, 9, 1-27.
Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychol-
ogist, 39, 117-123.
Zajonc, R. B., Murphy, S. T., & Inglehart, M. (1989). Feeling and facial efference: Implications
of the vascular theory of emotion. Psychological Review, 96, 395-416.
Zeegers, W. (1988). Andere tijden, andere mensen. De sociale representatie van identiteit
[Changing times, changing people]. Amsterdam, The Netherlands: Bakker.
Zillmann, D. (1979). Hostility and agression. Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D. (1980). Anatomy of suspense. In P. H. Tannenbaum (Ed.), The entertainment func-
tions of television (pp. 133-163). Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D. (1983). Disparagement humor. In P. E. McGhee & J. H. Goldstein (Eds.), Hand-
book of humor research: Vol. 1. Basic issues (pp. 85-107). New York: Springer.
Zillmann, D. (1985). The experimental exploration of gratifications from media entertainment.
In K. E. Rosengren, L. A. Wenner, & P. Palmgreen (Eds.), Media gratifications research.
Current perspectives (pp. 225-239). Beverly Hills: Sage.
Zillmann, D. (1988). Mood management: Using entertainment to full advantage. In L. Donohew,
H. E. Sypher, & E. Tory Higgins (Eds.), Communication, social cognition and affect (pp.
147-171). Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D. (1991a). Television viewing and physiological arousal. In J. Bryant & D. Zillmann
(Eds.), Responding to the screen. Reception and reaction processes (pp. 103-133). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Zillmann, D. (1991b). Empathy: Affect from bearing witness to the emotions of others. In
J. Bryant & D. Zillmann (Eds.), Responding to the screen. Reception and reaction processes
(pp. 135-168). Hillsdale, NJ: Lawrence Erlbaum Associates.
274 REFERENCES

Zillmann, D. (1991c). The logic of suspense and mystery. In J. Bryant & D. Zillmann (Eds.), Re-
sponding to the screen. Reception and reaction processes (pp. 281-303). Hillsdale, NJ: Law-
rence Erlbaum Associates.
Zillmann, D., & Bryant, J. (1974). Effect of residual excitation on the emotional response to
provocation and delayed aggressive behavior. Journal of Personality and Social Psychology,
30, 782-791.
Zillmann, D., & Bryant, J. (1975a). Viewers' moral sanction of retribution in the appreciation of
dramatic presentations. Journal of Experimental Social Psychology, 11, 572-582.
Zillmann D., & Bryant, J. (1975b). Effect of residual excitation on the emotional response to
provocation and delayed aggressive behavior. Journal of Personality and Social Psychology,
30, 782-791.
Zillmann, D., & Bryant, J. (1985). Affect, mood and emotion as determinants of selective expo-
sure. In D. Zillmann & J Bryant (Eds.), Selective exposure to communication (pp. 157-190).
Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D., & Bryant, J. (Eds.). (1989). Pornography: Research advances and policy con-
siderations. Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D., & Bryant, J. (1991). Responding to comedy: The sense and nonsense in humor. In
J. Bryant & D. Zillmann (Eds.), Responding to the screen. Reception and reaction processes
(pp. 261-279). Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D., Bryant, J., & Sapolsky, B. S. (1989). Enjoyments from sport spectatorship. In J. H.
Goldstein (Ed.), Sport, games and play: Social and psychological viewpoints (2nd ed., pp.
241-278). Hillsdale, NJ: Lawrence Erlbaum Associates.
Zillmann, D., & Cantor J. R. (1976). A disposition theory of humour and mirth. In A. J. Chap-
man & H. C. Foot (Eds.), Humour and laughter: Theory, research and applications (pp.
93-115). London: Wiley.
Zillmann, D., & Cantor, J. R. (1977). Affective response to the emotions of a protagonist. Jour-
nal of Experimental Social Psychology, 13, 155-165.
Zillmann, D., Hay, T. A., & Bryant, J. (1975). The effect of suspense and its resolution in the ap-
preciation of dramatic presentations. Journal of Research in Personality, 9, 307-323.
Zillmann, D., Mody, B., & Cantor, J. R. (1974). Empathetic perception of emotional displays in
films as a function of hedonic excitatory state to prior exposure. Journal of Research in Per-
sonality, 8, 335-349.
Zillmann, D., Rockwell, S., Schweitzer, K., & Sundar, S. S. (1992). Does humor enhance dis-
comfort tolerance? Paper presented at the ICA, Miami, FL.
Zillmann, D., & Wakshlag, J. (1985). Fear of victimization and the appeal of crime drama. In
D. Zillmann & J. Bryant (Eds.), Selective exposure to communication (pp. 141-156). Hills-
dale, NJ: Lawrence Erlbaum Associates.
Zillmann, D., Weaver, J. B., Mundorf, N., & Aust, C. F. (1986). Effects of an opposite-gender com-
panion's affect to horror on distress, delight, and attraction. Journal of Personality and
Social Psychology, 51, 586-594.
Zimbardo, P. G. (1969). The human choice: individuation, reason and order versus deindividu-
ation, impulse, chaos. In W. J. Arnold & D. Levine (Eds.), Nebraska Symposion on Motiva-
tion. Lincoln: University of Nebraska Press.
Zuckerman, M. (1988). Behavior and biology: Research on sensation seeking and reactions to
the media. In. L. Donohew, H. E. Sypher, & E. Tory Higgins (Eds.), Communication, social
cognition and affect (pp. 173-194). Hillsdale, NJ: Lawrence Erlbaum Associates.
Zusne, L. (1986). Cognition in consonance. Perceptual & Motor Skills, 62, 531-539.
Zwaan, R. A. (1993). Aspects of literary comprehension: A cognitive approach. Philadelphia:
John Benjamins.
 Author Index

A Atkinson, J. W., 104

Auerbach, E., 157 n89, 193 n l l 4
Aaker, D. A., 178 Aumont, J., 155, 156, 167, 172, 186, 188,
Abelson, R. P., 66 n53, 122 n78, 213 190, 191
Adorno, T. W., 167 Aust, G. F., 28
Affron, G. 4 n3 Austin, B., 17, 24 n20
Albritton, D. W., 97, 105, 169 Aver ill, J. R., 44, 77, 215, 218
Alfert, E., 218
Allen, J. L. 191 n l l 2
Allen, R. G., 38, 66 n54, 69, 125, 165, 166, B
227, 227 nl24, 234, 236-238
Alley, T. R., 161 Backman, G. W., 168 n94
Andersen, S. M., 163, 165 Badger, D. P., 17
Anderson, J. R., 166 n93 Baird, W., 91, 98
Andrew, D., 38 n29 Balazs, B., 31, 32, 53
Andrews, J. D. W., 126, 143 Bandura, A., 27
ANF, 9 n9 Bannerman, J., 17 n l 3
Ang, I., 23, 66 n54, 73, 97, 126 Baratz, D., 71 n55
Anstis, S., 236 Baron, R. A., 62
Anthony, B. J., 89 Baron, R. M. 66, 72, 161
Apter, M. A., 87 Barsalou, L. W., 143, 165
Aristotle, 31, 31 n26 Barthes, R., 67, 157
Arnheim, R., 5, 29, 31, 32, 34, 88, 237 Batson, G. D., 191 n l l 2
nl31 Baudry, J. L., 19, 20
Arnold, M. B., 44 Bazin, A., 29, 66 n54, 68
Aronson, E., 112 Beach, R., 31
Atkin, G. K., 23, 28, 30 Beckerman, B., 116, 124, 158

275
 276 AUTHOR INDEX

Becklen, R., 89 Carello, C., 162

Bell, R A. 62 Carroll, J. M. 34, 91
Bellinghiere, J., 204 Carroll, N., 4 n3, 5 n4, 20 nl6, 51 n37, 66
Bellour, R., 63 n54, 99, 134, 211 n l l 6 , 215 n l l 9 ,
Bergala, A. 155, 190 227-232, 235, 236, 238, 239 nl33,
Berlyne, D. E., 34, 91 n63, 94, 113 n76 244
Berndt, 168 Carver, C. S., 56, 91 n63, 213
Berry, D. S., 72, 160 Charniak, E. 142 n86
Berseheid, E., 161, 168 n94 Chatman, S., 42 n 31, 58 n43, 73, 165
Bever, T. G., 34, 91 n63 Chaumeton, E., 215 n l l 9
Birkhoff, G. D., 94 n65 Child, I. L., 56, 90, 91 n63, 112, 147
Bonitzer, R, 66 n54 Choffray, J. M., 37
Booth, W. C., 91 Christ, W. G., 25, 217
Borde, R., 215 n l 9 Claire, T., 170
Bordewijk-Knotter, J. M., 169
Clark, L. F., 91, 93, 131, 141 n85, 142, 145
Bordwell, D., 4-6, 8-11, 13, 22, 29, 33, 42
n87, 149, 150, 172
n31, 50 n36, 53, 55, 57, 58, 61, 63-
Clark, M. 8 8
65, 65 n52, 72, 90, 96-98, 104, 105,
Clore, G. L., 43, 44, 54, 67, 85 n61, 87, 111,
116, 128, 150, 158, 159, 159 n90,
128 n81, 171, 211 n l l 7
165, 172, 186, 193 n l l 4 , 206, 210,
Cogan, D. 219 n l 2 3
212, 238
Cogan, R., 219 n l 2 3
Boselie, F., 113 n76
Cohen, A. R. 112
Bourdieu, P., 23, 24
Cohen, P. R. 69, 88
Bower, G. H., 63 n48, 69, 88, 134, 172
Braddick, O. J., 236 Cohen-Seat, G. 5 n5
Brandt, T., 236 Colby, K. M., 12 n l 2
Collins, A. M., 43, 44, 54, 67, 85 n61, 111,
Branigan, E., 38, 57, 57 n42, 58, 188, 189
128 n81, 171, 211 n l l 7
Brehm, J. W., 112
Comisky, P., 102
Brewer, M. B. 163-165
Cook, P L., 17, 64 n51, 80
Brewer, W. F. 17, 96 n67, 101, 102, 207, 211
nll6 Craik, F. I. M. 166 n93
Britton, B. K., 93, 98 Csikszentmihalyi, I. S., 90, 93
Brook, P 97 Csikszentmihalyi, M., 90, 93
Brooks, V., 73, 87, 161 Culler, J., 67, 71, 91
Browne, N., 21 Cupchik, G. C., 17
Bruner. J. S., 160 Cutting, J. E., 53 n40
Brunetiere, F., 124, 204
Brunsdon, C., 180
Brunswik, E., 160, 167 D
Bryant, J., 24, 25, 62, 62 n47, 102, 169,
215, 217-219, 218 nl21, 219 n l 2 3 D'Andrade, R. G., 160, 167, 168
Buekland, W., 42 n31 Davidson, L. A. 215, 218
Bureh, N., 52, 52 n39, 227, 237 n l 3 1 Davidson, R. J., 79, 217
Byrne, D., 77 n58 Davis, D. K., 36
Davis, M. H., 73, 185
Dawkins, R., 191 n l l 2
c de Beaugrande, R., 214
de Groot, A. D., 149 n88
Calhoun, C., 6 n6 de Jonge, P., 215
Cantor, J. R., 62 n47, 101, 102, 156, 159, de Kuyper, E., 177 nlOl
161, 163, 168, 169, 215, 219 n l 2 3 de Lauretis, T., 21, 180
AUTHOR INDEX 277

De Sousa, R., 43, 44, 232, 232 n l 2 8 , 233, Forgas, J. P., 64 n 5 0

235 Forster, E. M., 157, 164, 188 n l l l
d e Wied, M. A., 55, 99, 101, 102, 116, 136, Foss, M. A. 267
137, 150, 218 nl22, 243 Fowles, J., 37 n27
den Uyl, M. J., 63 Frederickson, W. A. 168 n94
Dennett, D. C., 191 n l l 3 Frenkel-Brunswik, E., 167
d e Rivera, J., 44 Freud, S., 7, 18, 18 nl4, 19 n l 5 , 23, 26, 241
Diteweg, G., 99, 210 Freytag, G., 116, 204
Dopkins, S., 172 Friesen, W. F., 79 n59, 217
Dorner,D., 94, 113 n76 Frijda, N. H., 4, 7, 12 n l 2 , 13, 43-49, 52,
Dovidio, J. F., 191 n l l 2 54, 56-60, 59 n 44, 63, 66-68, 73,
Duelos, S. E., 79 n59 75, 77, 78, 81, 85, 87, 88, 90, 104,
Duneker, K., 91 104 n70, 106, 112, 117, 118, 126,
Dunning, E., 26 128 n81, 160, 161, 168, 178 nl02,
Durgnat, R., 4 n3, 62, 124, 125, 133, 135 182, 182 n l 0 5 , 184, 198, 199, 204,
Duvignaud, J., 23 n l 9 212, 215, 218, 227, 233, 234-236,
d'Ydewalle, G., 63 n49 234, 238, 240 nl34, 241
Dyer, M. G., 128, 129, 135, 142 n86 Frith, U., 63 n 49
Dyer, R. 49

G
E
Gaer, E. P., 102
Eachus, H. T., 168 n94 Gaver, W. W., 22 n l 8 , 34, 91 n63, 113
Eaton, M., 215 n l l 9 Gebotys, R. J., 217
Eberwein, R. T., 20 Gee, J. P., 138, 140 n84
Efran, J. S., 116, 219 n l 2 3 Gernsbacher, M. A., 172
Eggermont, L., 10 nlO, 11, 35, 193 Gerrig, R. J., 74, 97, 105, 113, 169, 244,
Egri, L., 169 n95 244 138
Ekman, P., 79 n59, 85 n61, 217 Gibson, J. J., 4, 47, 53, 66, 69, 241 n l 3 5
Elias, N., 26 Gilbert, D. T., 244 n l 3 8
Ellis, J., 66 n54 Gilligan, S. G., 63 n48
Ellsworth, P. G., 44, 201 Glynn, S. M., 93
Eysenck, H. J., 113 n76 Godkewitseh, M., 91 n63
Golding, J. M., 172
Goldman, M., 71 n55
F Goldsmith, H. H., 172
Gombrich, E. H., 34, 74
Faber, R. J., 10 nlO, 17, 80 Gordon, S. E., 158, 164, 167
Fechner, G. T., 34 Goto, T., 218 n l 2 2
Fenigstein, A., 23, 36 Graesser, A. G., 91, 93, 122, 122 n77, 131,
Fenley, W. F., 79, 80 n60, 219 n l 2 3 141 n85, 142, 145 n85, 149, 150,
Feshbach, S., 26 172
Festinger, L., 112, 113, 117 Graham, F. J., 89
Fisher, W. A., 77 n58 Granberg, D., 168 n984
Fiske, J., 23, 47, 66 n54, 77, 77 n57, 80, Gray, J. A., 44
243 nl37 Greenspan, S. L., 134, 172
Fiske, S. T. 160 n91, 163, 164, 168 Greenwald, A. G., 187
Fodor, J. A., 232 n l 2 8 Greimas, A. J., 122, 123
Fokkema, D. W., 41 Griee, P., 41, 42
Folkman, S., 215 Grosjean, F., 138, 140 n84
278 AUTHOR INDEX

Grosofsky, A., 162 J

Gunning, T., 234
Gunter, B., 23, 28 James, W., 7, 44, 99, 166, 170
Gunzenhaiiser, R., 95 n66 Jarvie, I. C., 17
Gurvitch, A., 30 Jenkins, H. M., 163
Johansson, G., 66
Johnson-Laird, P N., 66 n53, 197
Joosten, P., 10 nlO, 11, 35, 193
H
Jose, P E., 96 n67, 102
Jowett, G., 22, 24
Hagen, M. A., 71 n55
Hagerty, M. R., 178
Hamilton, O. C., 93, 160 K
Hanari, F., 218 n l 2 2
Handel, L. M., 17 n l 3 Kaplan, E. A., 21
Hardy, A. P., 10 nlO, 17, 80 Kaplan, R., 87, 91 n63
Harvey, W. J., 158, 189 Kaplan, S. 87, 91 n63
Harvill, J. G., 17, 64 n51, 80 Karppinen, P., 17
Hay, T. A., 102, 218, 218 nl21, 219, 219 Kay, P., 42 n 32, 135
nl23 Kayra-Stuart, F., 23 nl8, 161
Hearst, E., 163 Kean, K. J., 72
Heath, S., 21, 58 Kellogg, R., 158
Heider, F., 72 Kermode, F., 91
Hekkert, P., 22 n l 8 Kiewiet, R., 69, 72 n56
Helm, D. M., 17, 64 n51, 80 King, M., 168 n94
Heyduk, R. G., 23, 36 Kintseh, W., 91, 98, 99, 134, 197, 209
Hidi, S., 91, 98 Klatzky, R. G., 163, 165
Higginbotham, 122 Klin, G. M., 172
Hilgard, E. R., 19 n l 5 Klinger, E., 26, 88, 235 n l 3 0
Hoehberg, J., 63 n49, 71, 87, 91 n63, 94, Knight, J. M., 168 n94
161, 213, 236 Knowles, P. L., 66
Hoffman, M. L., 185 Konijn, E., 230 n l 2 7
Hoffner, C., 156, 159, 161, 163, 168 Koriat, A., 77, 215
Hollander, S., 23 nl8, 161 Kraeauer, S., 21, 29
Horn, B. K. P., 142 n86 Kreitler, H., 91, 154
Horowitz, M. J., 77 Kreitler, S., 91, 154
Houston, G. E., 244 n l 3 8 Kretsinger, E. A., 89
Huitema, J., 172 Krutnik, F., 22, 186, 215 n l l 9
Hull, J. G., 185 Kubovy, M., 52, 68, 241 n l 3 5 , 245 n l 3 9
Humphrey, N. K., 91 n63, 169 Kuhn, A., 21
Hunt, R. A., 70, 105 n73, 135 Kuipers, P., 45, 215
Kunda, Z., 170
Kurtz, K. J., 69

I
L
Inglehart, M., 178 n l 0 3
Isen, A. M., 88 Laird, J. D., 79
I to, T., 218 n l 2 2 Lang, P J. 69, 70, 77, 215 n l l 9
Izard, G. E., 43, 44 n33, 85, 85 n61, 87-89, Lang, R. 22
128 n81 Langer, S. K., 29
AUTHOR INDEX 279

Langlois, J. H., 161 Mehrabian, A., 9 8

Lanser, S. S., 41 Meier, N. C. 216 n l 2 0
Lasher, M. D., 34, 91 n63 Melkman, R., 77, 215
Lazarus, 43, 44, 77, 126, 179 nl04, 183 Mervis, C. B., 22, 71
nl06, 215, 218, 232 Mesquita, B., 58, 63, 117, 204
Lazier, G., 204 Messaris, P., 63 n 49, 65, 69
Leech, G. N., 42 Metz, C., 5 n5, 19, 20, 32, 54, 63, 66 n54,
Lehnert, W. G., 128-133, 138 n83, 139 68, 155, 156, 240
Lerner, M. J., 96 Meyer, L. B., 91 6 3
Levenson, R. W., 79 n59 Miall, D. S., 91 n63, 188 n 111
Levinson, D. J., 167 Michotte, A., 66 n54, 68, 69, 72, 154
Leventhal, H. 79 n59 Millar, G., 55, 63 n49
Levy, S. G., 79, 80 n60, 219 n l 2 3 Miller, D. T., 170
Lewin, B., 7, 19 Miller, R. S. 181
Lewis, J. M., 17 n l 3 Miller, N. E. 117
Lewis, M. 112 Milner Davis, J. R., 73
Lichtenstein, A., 17 Misovich, S. J., 72
Lichtenstein, E. H. 101, 102, 207, 211 n l l 6 Mitry, J., 190
Light, L. L. 23 nl8, 161 Mody, B., 62 n47
Linton, M. J., 22, 24 Moller, K. E. K., 17
Lipps, T., 154 Moore, K., 23 n l 8
Lockhart,, R. S. 166 n93 Mordkoff, A. M., 215, 218
Loewenstein, G. F., 205, 213 Morin, E.„ 54, 68
Long, D. L., 172 Morreal, J., 91 n 6 3
Lynch, F. D., 49, 63 n49 Morrow, D. G., 134, 172
Movshon, A., 236
Moynihan, C., 98
M Mulvey, L., 20, 21, 180
Mundorf, N., 28
MacGabe, G., 21 Miinsterberg, H., 5, 27
Malone, P. S., 244 n l 3 8 Murphy, S. T., 178 n l 0 3
Mandler, G., 22, 34, 44, 58, 91 n63, 113, Myers, J. L., 172
124
Marie, M., 83, 155, 190
Markus, H., 187 N
Martin, M., 209, 216
Martin, W. 95 Nachbar, J., 215 n l l 9
Martindale, C., 23 n l 8 Neale, S., 22, 58, 186, 215 n l l 9
Mathews, K. E., 175, 185, 215 Neisser, U., 89, 142 n86
Matthews, 204 Nell, V., 25
McArthur, L. Z., 66, 160 Neuberg, S. L., 163
McClelland, D. C., 29 n24 Nevius, P., 187
McCue, M., 219 n l 2 3 Newell, A., 92 n64
McGhee, P. E., 91 n63 Nomikos, M. S., 77, 218
McGuiness, D., 89 Nunnally, J. C., 87, 91 n63
McGuire, W. J., 16, 22, 23, 30, 37, 39
McKoon, G., 105, 105 n72, 134, 172 n98
McNamara, T. P., 134 0
McQuail, D., 18
Meadowcroft, J., 25 O'Guinn, T. C., 10 nlO, 17, 80
Medoff, N. J., 25 Oatley, K., 44, 91, 105 n73, 230 n l 2 6
 280 AUTHOR INDEX

Obrist, P. A., 89 Ratcliff, R., 105, 105 n72, 134, 172 n98
Okuda, K., 218 n l 2 2 Rayburn, J. D., 36
Opton, E., 77, 218 Reisz, K., 55, 63 n49
Orban, G., 236 Rinck, M., 134
Ordman, V. 102, 116 Ritchie, R. G., 88
Ornstein, R., 217 Robertson, R. R. W., 172
Ortony, A., 43, 44, 54, 67, 85-87, 85 n61, Robertson, S. P. 122
90, 91 n63, 111, 128 n81, 171, 211 Robson, J. E., 63 n 49
nll7 Rockwell, S., 216
Roggman, L. A., 161
Rombouts, H., 178 nl02, 180, 183
P Rosch, E., 22, 71
Rose, T. L., 160
Palmer, J., 186, 215 n l l 9
Rosenfeld, L. B., 17
Palmgreen, P., 17, 18, 36, 64 n51, 80
Ruch, W., 199, 215
Panksepp, J., 85, 85 n61
Ruimschotel, D., 73
Panofsky, E., 52
Rumelhart, D. E., 58, 64, 124, 183 nl07,
Parish, J. R., 215 n l l 9
197
Pas, B., 37
Ryan, T., 71 n55
Penland, M. J., 93
Perkins, D., 71
Perkins, V. F., 38, 39
Peters, J. M., 2 n l , 32, 227-229, 247
s
Petro, P, 215 n l l 9
Sanford , R. N. 167
Pfeifer, R , 12
Sapolsky, B. S., 102
Pfister, M., 91, 116, 124, 157, 158, 186
Saron, G. D., 217
Philippot, P, 219
Sartre, J.-P., 2, 2 n l , 227
Piet, S., 28
Schachter, S., 7, 219 n l 2 3
Pittenger, A., 162
Schank, R. C., 66 n53, 72, 122, 122 n78,
Pitts, M. R., 215 n l l 9
123, 126, 128
Plutehik, R., 43
Schatz, T., 22
Polti, S., 121
Scheele, S. C., 24
Porfirio, R. G., 215 n l l 9
Scheff, T. J., 24
Posner, M. I., 89
Scheier, M. F., 56, 91 n63, 213
Potter, W. J., 51 n38, 67, 73, 74, 78
Schneider, E., 79
Powdermaker, H., 22
Schoedel, J., 168 n94
Pratt, M. L., 42 n30
Schoenmakers, H., 66 n54, 154, 169, 178
Prelec, D., 205
nl02, 189
Preminger, A., 91
Scholes, R., 158
Prentice, D. A., 74
Schram, D., 227
Pribram, K. H., 89
Schroeder, D. A., 191 n l l 2
Prokop, D., 21, 22
Schwartz, G., 71 n55
Pryor, J. B., 160
Schweitzer, K., 216
Purcell, A. T., 91 n63
Searle, J. R. 15, 16
Purves, A. C., 31
Secord, P., 168 n94
Piitz, P., 91
Sekuler, R., 236
Pylyshyn, Z. W., 20 nl7, 46 n34, 232 n l 2 9
Selfridge, O. G., 142 n86
Senulis, J. A. 217
R Sesonke, A., 54
Sexter, M. 79
Rabbie, J. M., 79, 80 Shalker, T. E., 88
Ramirez, J., 62 n47 Shaw, A., 162
AUTHOR INDEX 281

Shweder, R. A., 160 Thorndyke, P. W., 58, 64, 66 n53, 124

Silver, A., 124, 183, 184 nl09, 215 Todorov, T., 51, 58, 67
Simmel, M., 72 Tofarodi, R. W., 244 n l 3 8
Simon, H., 4, 92 n64 Tomkins, S. S., 44, 85
Simons, R. F., 89 Trabasso, T., 134, 135
Singer, D. G., 26, 27 Tsuji, K., 218 n l 2 2
Singer, J. L. 26, 27 Tulving, E., 166 n 9 3
Singer, R. 26 Turner, T. J., 85 n61, 86, 90
Slatter, R E., 23 n l 8 Tversky, B., 71 n55
Smith, M., 227, 229, 232, 238, 239 nl33,
242, 244
Smith, D. L., 66 u, v
Smith, W. R. 122
Solomon, R. C., 6 n6 Uchiyama, F., 218 n 122
Sonnemans, J., 58, 63, 117, 204 Ursini, J., 183, 184 nl09, 215
Sorlin, P., 27 van den Berg, H., 123
Souriau, E., 121-124, 127 van den Boom, I., 204, 206, 214, 216
Spangler, T. J. 116, 219 n l 2 3 van der Run, J. I. M., 124
Sparks, G. G., 101 van der Voort, T. H. A., 26 n21
Speisman, J. C., 77, 215, 218 van der Wurff, A., 51
Spinhof, H., 11, 35 van Dijk, T. A., 41, 58, 197
Srull, T. K., 163, 167 van Goozen, S., 58, 63, 117, 204
Staiger, J., 9, 22, 53, 58, 65, 97, 104, 105, van Lighten, O., 79
206 van Vliet, H., 66 n54, 154, 215, 239 nl33,
Stam, R., 73 250 n l 4 0
Stayman, D. M., 178 van Wieringen, P G. W., 22 n l 8
Stern, L., 79 Vanderbeeken, M., 63 n49
Sternberg, M., 91, 115, 206, 207, 210 Vehrs, W., 94, 113 n76
Stotland, E., 175, 185, 215 Vernet, M., 155, 190
Strongman, K. T., 43 Vine, E. W., 129, 138 n83
Suboski, M. D., 163 Vipond, D., 70, 105 n73, 135
Suh, S., 134, 135, 172 Visser, L., 79, 80
Suls, J. M., 91 n63, 117
Sundar, S. S., 216
Swagerman, J., 12 n l 2 W

Wakshlag, J., 28
T Waltz, W., 219 n l 2 3
Warren, G. G., 185
Tajfel, H., 167 Warren, W. H., 69
Takahashi, T., 218 Weaver, J. B., 28, 219 n l 2 3
Tamborini, R., 219 n l 2 3 Wegman, C., 12 n l 2
Tan, E. S., 10, 11, 35, 59, 69, 70, 72 n56, Wegner, D. M., 244 n l 3 8
99, 102, 105, 118, 132, 136-139, 136 Weiner, B., 112
n82, 148, 159 n90, 169, 193, 204, Wenner, P, 36
206, 210, 213-217, 219, 243 Wheeler, L., 219 n l 2 3
Tannenbaum, P. H., 24, 102 White, R. W., 28, 49, 68, 162
Taylor, S. E., 160 n91, 164, 166, 168 Whitfield, T. W. A., 23 n l 8
Tellegen-van Delft, S., 26 Wilensky, R., 58 n 43, 128, 129, 135, 214
ter Sehure, E., 45, 215 Wilson, S. G., 134, 135
Thompson, K., 9, 22, 50 n36, 53, 55, 58, Winston, P. H., 142 n86
65, 97, 104, 105, 150, 158, 165, 206 Wober, M., 23
282 AUTHOR INDEX

Woodall, W. G., 36 Zajonc, R. B., 22, 178 n l 0 3

Wright, H. F., 112 Zeegers, W., 74
Wright, W., 215 n l l 9 Zelson, M. F., 89
Wyatt, R. O, 17 Ziemke, D., 101
Wyer, R. S., 163, 164, 167 Zillmann, D., 24, 25, 28, 35, 62, 101, 102,
108, 116, 154, 169, 174, 207, 211
n l l 6 , 212, 215-219, 218 n 121, 219
Y, Z nl23
Zimbardo, B., 79
Young, 51, 130, 132, 133, 158, 161, 185, Zuckerman, M., 48
246 Zusne, L., 3 4
Zahn, D., 204 Zwaan, R. A., 105 n73
 Subject Index

A as a dynamic structure, 198-201

empirical, 201
Abel (A. van Warmerdam, 1986), 123, 230 generic, 202-223
Absorption, 88 and episodicity, 202-204
potential, 217 and unity, 202-204
Acceleration, 61 and course of emotions, 204-213
Accident (J. Losey, 1967) 78 ideographic use of, 221
Actant model of mythical story structure, intuitive properties of
122 depth of emotion, 222
Action design, 222-223
possibilities for, 76, 233-234 dominance of characteristic emo-
readiness, 45, 46 tions, 221
thriller, 28, 108, 108 n74 prospectivity, 222
Action tendency, 45, 74, 75, 76 unity, 223, 203
readiness to give as, 180 nomothetic use of, 221
readiness to receive, 180 overview of, 195-202
virtual, 75, 180, 240 as psychological structure, 201
limits to, 180 quantitative modeling of, 219-220
Admiration, 82, 170, 173, 178 nl02, 214 as real time representation, 201
in relation to sympathy and compassion, specific a. of particular film, 202
179, 181 substructures of, 196-199
and reward, 191 emotional meaning structure as a,
Affect, 151 197
regulation, see Motive of viewer research into, 214-215
reversal of, 170 emotional response as a, 197
state, 129, 137 research into, 216-219, 219 n l 2 3
Affect structure of the feature film, 195-223 relations among, 200-201

283
284 SUBJECT 11\'0EX

situation model as a, 197 negative empathetic emotion in, 192

text base as a, 197 Artefact
research into, 213-214 awareness of, 135, 211, 238
as system of correlations over time, 203, emotion, 65, 81, 226
220 empathetic A emotion, 82, 83
theoretical, 201 aesthetic A emotion. 176
theoretical status of, 201 non-empathetic A emotion, 8.1
AUordancc, 47, 56, 69 film as, .13, 49, 53, 64-65, 76
social. 161, 162 Assistant to viewer, 241, 243, 250
The African Queen (J. Huston, 1951), 177 Astonishment, 82
Agency, 46 Attention, R7
Aggression, 26-28, 26 n21, 178 nlOJ selective, 87
catharsis and, 26-27 Attitude, see aL~o Observational attitude
symbolic expression of, 26 en-face, 18.1
and viewing preferences, 28 en profil, 182
Aha experience, 91 vis-a-vis, 182
Allied Deception 135, 136
Audience
All Quiet on the Western Front (L. ~lilestone,
natural, 9-11. 74
1930), 187
and homogeneity of experience, 154
Amadeus (M. Forman, 1984), 17.1
and self-selection. 10, 202
Amazement. 128
demarcation of, 11
Amusement. 146, 217
specialized, 9n9
Anger, 24, 45, 174, 181, 197
size, 79, 80 n60
Antagonist
suhaudience and genre, 10
emotion of, 174
of traditional feature film, 9-11, 124,
Anticipation, 89, 98-99, 104, 123, 134,
19.1, 202. 243 nD7
143, 147, 151. 203, 206
Aversion, 162, 170
diffuse, 210-211, 214-215
Awareness of film style and technology, JJ
Antipathy, 1R4
Apparent motion, 68, 236
Apparent reality, 66-76
Apparent Reality, Law of, 66, 67, 2.1.1, 2J4
B
Appraisal, 45, 46, 54, 66
and mood, 204 Bad Day at Black Rock (J. Sturges, 1955),
primary, 46, 55, 75, 239
of character, 161 Balance phase of narrative, 59, 117, 197
secondary, 47, 55, 78 Balance of frustration and reward, 98, 107
Appreciation. 94 The Barefoot Contessa (J. Mankiewicz,
Approach, 200 1951 ), .19
Approach-avoidance conflict, 117 Beauty, .14
Approachability, 161 Beauty and tlw Reust (G. Trousdale & K.
Around the lVOrld in 80 lJays (!vi. Ander- \Vise, 1991), 166
son, 1956), 116 Belonging, sense of, 2.1
Arousal. 24, 25, 91, 218 Birth of a ;Vation (D. W. Griffith, 11)15 ), 1
optimum, 25 BORIS, 129
potential, 24 Brian's Son;; (B. Kulik, 1970), 185
L'ArrivCe d'un Train en Gare de la Ciotat / The Birds (A. Hitchcock, 196.1), 172
A..rri·val of a Train at the Station of Brief Encounter (D. Lean, 1945), 2, 92,
l.C. (A. & L. Lumiere, 1895), 1, 234 18.1, 184
Art cinema, 158 Le Boucher/ The Butcher, 241
imagine-self empathy, 193 Bridging, 131, 141 n85, 150
SUBJECT INDEX 285

C concerns of, 127

recognition and sharing of, 167, 168
Camera eye, 32 exploration of, 158
Card, 158 flat vs. round, 157, 173
Care for Consequence, Law of, 46, 77, 248 interpretation, 159
Caricature, 71 intentions, 168
Casablanca (M. Curtiz, 1942), 169, 181 structures, 121
Categories pictoliteral, 164
natural, 71 trait, 158, 160, 167
goal-derived, 143 typing, 165
taxonomic, 143 constraints posed by narrative, 167
Categorization, 71 understanding of, 156-182
Catharsis, 18, 24, 27 and coherence, 159
Cat People (J. Tourneur, 1942), 51 visual image of, 160, see also Character,
Causality (situational meaning) 46 pictoliteral
Causality (narrative structure), see Narra- Characterization, 147, 149, 157-159
tion, causality in and stylization, 162
Ceiling Zero (H. Hawks, 1935), 166 and cultural stereotypes, 162
Censorship, 22 as source of pleasure, 191
Cera Una Volta il West/ Once Upon a Chicago (F. Urson, 1927), 88
Time in the West (S. Leone, 1969), Chinese Room thought experiment, 15
101 The Chosen (J.P. Kagan, 1981), 153
Certainty/uncertainty, 60, 92, 98, 100, 108, Cinema, 79
117, 128, 143, 198 anonimity in, 79
Challenge, 92-94 physical conditions of, 79
Change in situational meaning, 56-64, 101, popular, 93
117, 146 Cinematic
criteria of, 57, 59, 60, 61, 117, 202 attitudes, 10
and situational irony, 61 taste, 35, 49
Change, Law of, 56-57, 112, 212, 233 Cinephilia, 34
Character Citizen Kane (O. Welles, 1941), 158
in art cinema, 158 Clarity, 60
apparent reality, 165-166, 171 Closure, Law of, 74, 75, 118, 212, 233
appraisal of situation, see situational Closedness, 56, 60, 198, 199
meaning/significance for the Coding, 66, 77, 81
character elementary, 71
assessment as good or bad, 168-171 multiple, 67
and empathetic emotion, 169, 171 reduction in, 70
attractiveness/appeal, 160-163 transformation in, 70
face and, 161-162 Cognition, 46
perceived similarity and, 168, control of by emotion, 43, 88
168 n94 Cognitive
categorization of, 159-171, see also Im- closure, 97, 132, 198
pression formation coding, see Coding
primacy effect in, 164 curiosity, 58, 212
primitive, 160 dissonance, 112-113
unconscious, 162 effort, 70, 112-113
and conditioning, 162-163 generic structure, see Generic knowl-
cognitive representation of, 163 edge structure
pictoliteral, 164 impenetrability, 232
complexity of, 163, 173-174 miser, 166
286 SUBJECT INDEX

processes in coding, 66 Contempt. 181

representation, 128 Continuity, 63, 239
of reality, 70-73 Controllability, 55
theories of discourse processing, 197 Control prcccdcncc, 44, 75, 86, 104, 233,
Coherence 234
local, 134 Convention, 41, 67, 243
Collective of fiction, 228
identity, .19 and norm, 10
suheonsdous, 21 Cooperative principle, 41-43, 64, 249
Comic, 61 Le Cnrbeau /The Raven, 170
Communication, 41 Crying, 116-117
Communicativeness, 57 Cultural studies approach, 22, 2.1
Comparative Feeling, Law of, 57, 112 Curiosity, 206, 209
Compassion, 170 Cuteness, 161
as related to sympathy and adminuion,
179, 181
as related to pity, 179 n104 D
and reward. 191
Competence, 28, 49 The Da_y of the Jackal (F. Zinneman, 1973),
Complication, 124, 126 108
paradoxicaL 123 The Dead (.L Huston, 1987), 157
Complication phase of narrative, 59, 60, 98. Deadline, 115
117, 197, 203, 205 Death on the Nile (J. Guillermin, 1978), 170
Concentration, 50 Deceleration. 61
Condensation, 51 Deindividuation hypothesis. 79
Condescension, 179 Demand character, 60
Concern(s) 44, 47, 78, 125 Demons. 129, 142, 142 n86
cognitive. 198 Dependence, 160
affective, 124, 143, 198 Depth of knowledge/information, 57
altruistic motivation as a, 191 Depth of processing, 166, 166 n9J
beauty, 125 Der~u Uzala (.A. Kurosawa, 1975), 157
caring for other people, 191 Desire, 46, 59, ()1, 7(), 176 n100, 181, 199,
~:ognitive assimilation, 48, 49. 125 200,203
emotion as a, 126 Deux ou Trois Chases que Je Sais d'elle I
familiarity, 49 Two or Three Things I Knou• About
idiosyncratic, 48 Her(J.-L. Godard, 1966), 52
and motivation for viewing film, 48 nJS, Dial A! fur Munler (A. Hitchcock. 1954),
49 108
proximity and coherence, 49 Les Diaboloques (G.-H. Clouzot, 1955). 108
recognition of, 127, 167-171 Diegetic effect, 52-56, 52 n39, 237, 237
source, 48 n131, see also Illusion, of diegetic ef-
surface, 4H, 59 fect
safety, 48, 76 perceptual and cognitive basis, 52
sympathy, 48, 125, 171 and empathy, 153
tension creation and resolution, 48 Difficulty, 45, 59, 60, 116, 198, 241, 200,
value, 49, 171, see also Values 241
Concern, Law of, 47-50, 190, 233, 247 Dirty Harry (D. Siegel, 1971), 108
Coniliet, 123, 124 Disappointment, 128, 197
Consciousness, 19, 47, 56 Disgust, 83, 212
Conservation of Emotional :tv[omcntum, Law Dismay, 128
of, 104n71 Distance, 24, 76, 77
SUBJECT INDEX 287

Distress, 102 and fantasy, 51, 180, 235

Disturbance of the Balance in narrative, 58, film-induced, 248-249
59, 124, 126 first strategy, 248, 249
Do the Right Thing (S. Lee, 1989), 226, functional definition of, 232-234
235, 239 functionality of, 4 3 - 4 4
Don't Look Now (N. Roeg, 1973), 108 genuine, 2, 7, 43, 127, 226-232, 2 3 4 -
Down by Law (J. Jarmusch, 1986), 176 248
Drama, 123 hedonic tone of, 81
theory, 30, 123, 157, 214 illusion theory of, 228
Dramatic situation, 121 imitation theory of, 229
Dramatic structure, 214 intensity, 1, 50, 57, 75, 77, 202-205, 250
Dramaturgical functions, 121 introspection and, 230-232
The Dresser (R Yates, 1983), 157 labels, 199
Drive satisfaction, 19 machine, film as, 119, 252
Duchenne smile, 217 maximization of, 79, 249-251
Duration, 150 mixture of, see Emotional, blend
subjective, 217 near-e., 247
non-empathetic, 171, 174
in traditional cinema, 174
E nuances in, 173
object fate, 54
Ecological psychology, 4 off-screen space of, 69
Editorial intelligence, 65, 241 persistence of, 74, see also Law of Clo-
84 Charing Cross Road (D. Jones, 1986), 2 sure, Control precedence
The Elephant Man (D. Lynch, 1980), 166, phasic, 83, 199
170 measurement of, 218-219
Embarrassment, 24, 181 narrative procedure and, 211-212
Emotion, 46 phasic vs. tonic, 83, 128, 177, 199-200
aesthetic, 49 and expectation, 200
associative network model of, 69 problems in measurement, 218
authenticity of, see Emotions, genuine proper, 199
basic, 85-86, 233 prospect-based, 111, 198
and character categorization, 173 pretend theory of, 228, 247
cognitive theory of, 7, 44 psychological mechanisms in, 7
complexity of, 173-174 rationality of, 1, 88
and contagion, 79, 248 regulation of, 24, 46, 77-81, 249
control of, see regulation and media use, 24-27
course in time of, 204-206 self-enhancement of, 204, 249-251
definition of, 44, 46, 231, 233 scenic structure and, 6 3 - 6 4
circularity of, 231-232 sharing of, 80, 175, 248
common sense, 231-232 stimulus in film, 50-76, 81
depth of, 222 and stereotype, 169-171
empathetic, 82, 174-182 , 240 subject matter of film and, 50-51
action tendencies of, 180 systematics of, 3, 195
definition of, 174 thought theory of, 229
and feelings of characters, 174, 182 theory, 6-7, 43-47, 233
going hand in hand with non- unpleasant, 2, 24
empathetic, 176 nlOO variety of, 2
overview of, 179 witness, 82, 240
as source of investment and return, Emotional
191 blend, 2, 61, 199, 250
288 SUBJECT INDEX

meaning structure, 197, 214-215 Excitement, 45, 116, 175

and genre, 215 n, 119 The Exorcist (W. Friedkin, 1973), 230
experience, 47, 198-199, 220-221 Expectation, 83, 92, 99, 103, 104, 105, 121,
unity of, 223 123, 127, 128, 129, 134, 147, 181,
response, 216-219, see also Action ten- 197, 199, 200, 206
dency, virtual, readiness; Freeze; In- reversals of, 125, 135, 157
terest sources of, 165
dominant, 219 Experience of time, 63
and genre definition, 219, 221 Explorative behavior, 87, 90
feedback of on appraisal, 201 Extrication, see Implication
as an episode, 202-203 Euphoria, 116, 213
measurement of, 218-219, 218 nl22,
219 n l 2 3
stimulus, 50 F
as compared to verbal stimuli, 239
in film, 236-248, see also Apparent Face
reality, Affordance, Appraisal, as cue in character appraisal, 161
Coding, Diegetic effect, Movement Facial feedback, 79 n59
Emotions Facial expression as index of appreciation,
classification of, 81-84, 199-200 218 n l 2 1
Empathy Fallacy of the "indigent metaphor," 5 n4
and affective investment, 192 Familiarity, 22, 87, 198
artefact vs. fiction, 65, 82, 176 Fantasy, 26
contrast, 174 models, 27
definition of, 172 Fascination, 175, 176
and identification, 153-156, 189-190 Fatigue, 111
measurement of, 215 Fear, 24, 59, 61, 82, 99, 132, 162, 175, 199,
motor, 154 203, 212, 226
and individual viewer's characteristics, of crime, 28
154-155 fears confirmed, 200
and investment/re turn, 191 empathetic, 192
Enjoyment, 46, 61, 102, 115, 175, 199, 211 psychophysiological response in, 218
as major phasic emotion in film viewing, Feeling, 75
199 sinking, 213
non-empathetic, 175, 238 warm, 178 n l 0 3
Entertainment, 18, 24, 41, 42, 76 Ficelle, 158
as major motive for cinema visit, 17 Fiction, 227-232
Episode, 202 Fictional world vs. artefact, 64
closure, 128, 202, 212 Fictional action emotion, 65, 81, 82, 83,
narrative, 79, 117, 143, 197, 202-204 226, 227
emotion, 58-59, 62, 79, 202-204, 116- empathetic F emotion, 83
119, 202-204 non-empathetic F emotion, 83, 175
hierarchy of, 64, 197, 212-213 authenticity of, 230-232
recursive, 118, 202 Film
Equality, 179 artistic contimuum of, 9 n9
Erotetic narration, 99, 105, 134 art film, 9 n9, 158
E.T. the Extra-Terrestrial (S. Spielberg, avant-garde, 9 n9
1982), 108 characterization in, 157-158
Exaggeration, 51 classical Hollywood, 9, 53
Exhilaration, 199, 215 comedy, 3, 73, 125, 221
Excitation transfer, 62 n47, 212 detective, 209 n l l 5
SUBJECT INDEX 289

natural disaster, 175 Flow, 90, 93

and dream, 19 Focality, 63, 87
erotic, 175, 215 n l 2 3 Focused causal chain, 58
and entertainment, 17-18 Foreknowledge, 9 4
family, 108 Freedom of viewer, 80-81, 248-251, see Il-
fantastic, 51 lusion, inevitability of, and Emotion,
form, abstract, 35 pretend theory of
genre, 3-4, 4 n3, 108, 157, 175, 219, 221 Freeze, 88
horror, 28, 175 Friendship, 178
hard-porno, 176 Fright, 215 n l 2 3
melodrama, 22, 73, 125, 180, 221 Frustration, 98, 199, 212
mystery, 209 Full Metal Jacket (S. Kubrick, 1987), 187
perception, 52-53, 68-69, 213 Functions of film viewing, psychological,
motion, 53-54, 236-237 15-39
popular action, 96 evening out as, 24
popular vs. quality, 9 n9, 157, 173 and gender, 20, 21
popularity of, 175 of TV viewing, 2 4 - 2 5
psychological drama, 39, 125, 157 Functioning, primary and secondary, 19, 19
quality, 32 nl5
romantic, 22
short feature, 149
science fiction, 175 G
song-and-dance, 175
sound, 55 Gallipoli (P. Weir, 1981), 187
surface structure of, 6 Game, 78, 244
style, 33-34 Le Gendarme de Saint-Tropez (J. Girault,
theme, see Theme 1964), 162
tragedy, 73, 214 Generic knowledge structure (GKS) 127,
and characterization, 157 131, 142, 145 n87, 150
thriller, 28, 209 n l l 5 Goal, 126, 127, 135
traditional feature, 8 - 9 , 158 Goldfinger (G. Hamilton, 1964), 170
viewing as opposed to cinema visit, 16 Gratitude, 181
Filmic state, 18-21, 33 The Great Gatshy (J. Clayton, 1974), 158
and pseudo-regression, 32 Guilt, 45, 248
Film narrative
canonic course of, 57-61, 124
classical Hollywood, 8 - 9 , 52, 58 H
control by, 78
story structure of, 58, 93 Habituation, 111, 204
traditional, 93, 136 due to regular viewing, 28
Film style Hangover Square (J. Brahm, 1944), 189
classical, 48 Happy ending, 61, 205
as a source of viewer satisfaction, 32 Hate, 142
Film theory, 5, 213 Hedonic Assymmetry, Law of, 59 n44
formative, 31 Helplesness, 180
classic, 5 High Noon (F. Zinneman, 1952), 76, 116
modern, 5 Hope, 61, 92, 99, 112, 132, 171, 200, 203
phenomenological, 55 Horror, 175
psychoanalytical, 18-21 Host Sonaten / Autumn Sonata
sociological, 21 (I. Bergman, 1978), 157
Flashback, 61, 104 Humor, 215
290 SUBJECT INDEX

The Hunchback of Notre Dame (W. Dieterle, integration in, 164

1939), 166 models of, 163-166
personalization in, 165
revision in, 166
I top-down process of, 164
of types vs. persons, 159-160
Ideal observer, 53 Improvement, 202, 205
Identification, 49, 74, 190 Incongruity theories of humor, 91 n63
and camera, 31 Indiana Jones and the Temple of Doom (S.
and character categorization Spielberg, 1984), 177
and empathy, 153-156, 189-190 Inertia, 115
and gender, 180 Inertial tendency models of motivation, 112
as misleading term, 190 Inference, 105, 123, 134, 135, 172 n98
primary and secondary, 155-156 irrelevant, 123
and psychoanalytical film theory, 155- predictive, 134
156 Informationsasthetik, 94
tragic view of, 155-156, 190 Initiating event (IE), 207, 208, 209, 211
with camera, 155 Innate releaser, 160, 162
with character, 156 Intentionality, 45, 71, 72, 159, 160, 172
with narrative, 156, 190 n99, 198
with oneself, 155 Intentional meaning, 15
and surface structure of film, 156 Intentions
Ideological of film maker, 105 n73
manipulation, 21-23 of character, 127
reinforcement, 49 Interest, 83, 121, 85-119, 129, 134, 146,
Ideology, 21, 248 147, 148, 151, 200, 249-250
IE, see Initiating event action tendency in, 86, 118, 2 0 2 - 2 0 3
Illusion, 227-232, 236-248 and attention, 87
vs. collusion, 245 nl39, as cognitive state, 87
of controlled witness, 2 3 9 - 2 3 6 and course of time, 114, 144-148, 2 0 4 -
vs. delusion, 247, 190, 247 206
of depth, 53, 68 as backbone of affect structure, 211
of diegetic effect, 237-239 and elaboration, 90, 146
inevitability of, 249 as emotion, 85-87, 117-119
of motion, 68, 236 aesthetic, 83
of observational attitude, 246-248 affective, 92, 96
multiple, 249 cognitive, 92, 96, 132, see also curiosity
perspective, 52 course in time of, 204
projective, 227, 237 within individual episode, 206
character vs. spectator centered, 227 concerns in, 86
n, 124 definition of, 86
reproductive, 227, 236 as dominant emotion, 203
Implication/extrication, 77, 80 dynamics of, 94, 2 0 4 - 2 0 6
Impression formation, 159-174 as an emotion, 8 5 - 8 6 , 117-119
confirmation seeking in, 164 episode, 203
dual-process model, 163-164 experience of, 90
of characters vs. real world persons, 156, increase over film, 204
159, 166 intensity of, 137, 144-148
individuation in, 164-165, 166, 170- and nesting of episodes, 212-213
171, 183 as major emotion of affect structure, 200
examples of, 173-174 measurement of, 216-218
SUBJECT INDEX 291

principles of, 148 Love Story (A. Hiller, 1970), 48

Principle 1, 95 Loving gaze, 183
Principle 2, 101
Principle 3, 106, 136, 144
Principle 4, 111 M
and relation with other emotions, 203
and tension, 116-117, 203 Madame Sousatzha (J. Schlesinger, 1988),
self-enhancement of, 111, 118, 249-251 173
as tonic emotion, 200 Malicious Act, 138, 140, 141, 142
Interiors (W. Allen, 1978), 125 The Man Who Knew Too Much (A. Hitch-
Interpretation, 30 cock, 1956), 73
Intimacy, 178, 179 Mamie (A. Hitchcock, 1964), 73
Intermezzo, 109, 206 Mary Poppins (R. Stevenson, 1964), 108
Investment, 98, 100, 109, 112, 123, 143, Matching, 131
151, 197 Mental
cognitive, 98, 111 act, 129
affective, 98, 111, 171 state, 129
and narrative procedure, 208, 209 Mere exposure effect, 22
Involvement, 89, 90 Merleusse (M. Pagnol, 1935), 166
with artefact, 49 Metaphor, 91 n63
Ironic Mimesis, 2 3 8 - 2 3 9
paradox, 62 Mimetic leisure activity, 26
twist, 124, 157 The Miracle Worker (A. Penn, 1962), 117
Irony, 129 Mirth response, 215
situational, 61, 124 Misfortunes of others, 248
Mixed Event, 138, 140
Modifiability, 59, 60
J, K Mood, 204, 216
congruency, 62, 63
Jokes, mechanisms in, 18 Motion, 53
Joy, 60, 199 Motionless spectator, 19, 75, 115
Jurassic Park (S. Spielberg, 1993), 176 Motivation, 18
Justice, 96, 97 homeostatic view of, 18
Just world theory, 23, 96, 96 n67 and vicious circle, 36
King Kong (E.B. Schoedsack, 1933), 153, Motive (in film) 50
169 Motive (viewer)/satifaction
Klute (A. Pakula, 1971), 108 affect regulation as, 24-27
and fantasy
affective preservation as, 39
L appreciation of film as artefact as, 34
bolster social identity as, 23
Laughter, 116-117, 219 n l 2 3 confirm beliefs as, 23, 30
The Lift (D. Maas, 1983), 72 disinterested contemplation as, 30
Lightest Load and Greatest Gain, Laws of, enjoyment of filmic qualities, 31-32
81 enjoyment of film style and technology
Literary theory as, 33
and characterization, 157 enjoyment of watching as, 31
Location, 175 of watching people, 32
Longing, 175, 181 finding structural order as, 34
The Lost Weekend (B. Wilder, 1945), 177 gaining insight as, 29
Love, 178, 183 identification with characters as, 26, 32
 292 SUBJECT INDEX

involvement in fictional world, safe as, North by Northwest (A. Hitchcock, 1959),
33, 42 83, 9 2
learning as, 27 Novel, 157-158
cognitive, 28
mood management as, 24
primary, 31, 83 0
secondary, 31, 127
share feelings as, 23 Objectivity, 54, 238
specific emotional experience, 35 Observational attitude, 182, 214, 246, 248
social, 39 and narrative, 182, 188-190
social distinction as, 23-24 and emotion, 182-190
tension reduction, 3 2 - 3 9 O. Henry's Full House (H. Hathaway,
Motor activity, 88 H. Hawks, H. King, H. Koster, J.
Movement, 68 Negulesco, 1952), 130
audience, 217 On-line comprehension, 142
camera, 69 Opname / In for Treatment (E. van Zuylen,
eye, 213 n l l 8 1979), 206, 214, 216
Movie stars, 49 Outcome, 126
Murder on the Orient Express (S. Lumet, affective, 143
1974), 170 final of story, 126
Murder, She Said (G. Pollock, 1962), 240 of IE, 208
Music, 55 satisfactory, 127, 198
My Fair Lady (G. Cukor, 1964), 92 valuable, 127
Mystery Out of Africa (S. Pollack, 1985), 175
thriller, 108

P
N
PAM, 129
Narrative, 98, 111 Partial determination of meaning, 38
complexity in various genres, 157-158 Participation, 49
procedure, 101, 206-213, see also Cu- Passivity, 83
riosity; Surprise; Suspense La Peau Douce / The Soft Skin (F. Truffaut,
combined, 209 1964), 153
exposition in, 115 Perspective, 143, 246
functional vs. indicial, 157 of protagonist, 190
Narration, 6, 250 subjective, 176 nlOO
theory of in film, 5 - 6 , 8 - 9 The Phantom of the Opera (R. Julian,
causality in, 58, 72 1925), 166
scenic structure of, 6 3 - 6 4 Pity, 28, 59, 61, 83, 173, 178 nl02, 199,
structure of, see Story structure 214, 247, see also Compassion
temporal structures in, 56-64, 95, 104 The Piano (J. Campion, 1993), 173, 189
Narrational instance, see Narrator, Assistant A Place in the Sun (G. Stevens, 1951), 177
to viewer, Editorial intelligence Plans, 127
Narrative procedures, 206-213 Planning, 143
Narrator/narrational agency, 230, 242 viewer's understanding of by character,
overt, 57 173
New experimental aesthetics, 94 Pleasure, 24, 79, 183, 248, see also Enjoy-
Ninotchka, 171 ment
Nonverbal behavior, 217 aesthetic p. 91
Norm hypothesis, 80 Plot
SUBJECT INDEX 293

line, 110 magic-window, 53, 6 8 - 7 3

complex units, 129, 131, 137, 140 perceived utility, 73-74
simple units, 129, 137 representing essentials of, 7 0 - 7 3
Point-driven vs. story-driven reading, 135 Reframing, 55
Points, 129, 134 Regret, 246
Portrait de Lillian Gish (Jeanne Moreau, Relevance, 44, 45, 198
1986), 1 Relief, 60, 61, 112, 199, 200, 219 n l 2 3
Posse (K. Douglas, 1975), 246 The Remains of the Day (J. Ivory, 1993), 61
The Postman Always Rings Twice (B. Rafel- Resistance, 123
son, 1981), 61 Restoration of the Balance phase in narra-
Pragmatic contract between film maker and tive, 60, 117, 197, 205
viewer, 41-43, 64, 107, 146, 210, 243 Resolution, 60, 145, 197
Preferred final situation/outcome, 96, 98, of inconsistency, 91
99, 110, 136, 148, 151 Respect, 179
Prejudice, 167 Response
Presence, 61 emotional compatibility, 62
Problem, 124 participatory, 97
Problem solving, 91, 94 Retaliation, 138, 140, 141, 145, 147
Projection, 131, 134, 140, 142, 148, 150 Return, 98, 100, 102, 109, 112, 147, 151,
Promise, 86, 107, 113, 123, 125, 175, 198, 197
205, 245 foreground/background return, 106,
Prospect, 106 109-110, 136, 143, 148
Protagonist, 58, 158, 161 net return (NR) 100, 103, 105, 143
attractiveness of, 162 n92 expected/gained/attained net return,
metamorphosis of, 166 101, 102, 143
emotions of, 172 future net return, 100
represented aims of, 172 net return past/future, 102
situational meaning for, 172 NR(f), NR(FF), NR(p), NR(PP) 105
Prototypes, 22, 71, 72 Return of the Jedi (R. Marquand, 1983), 162
Proximity, 56, 107, 179, 180, 242 n l 3 6 Revulsion, 28
Pruning, 131, 149 Reward
Psycho (A. Hitchcock, 1960), 38 insufficient, 112
Psychological affect structure, see Affect postponement of, 192
structure of the feature film rationality of, 205
The Road Warrior / Mad Max II (G. Miller,
1981), 242
R Romanesque formula, 19
Rosemary's Baby (R. Polanski, 1968), 135
Rambo: First Blood, Part II (G.P. Cosmatos, 'Round Midnight (B. Tavernier, 1986), 158
1985), 27
Realism
problem of, 67 n54 S
photographic, 68, 71
psychological, 157 n89, 165 Saboteur (A. Hitchcock, 1942), 2
subjective, 158 Sacrifice, 130
Reality, 2 nl, 29, 45 Sadness, 28, 45, 46, 59, 197
awareness of, 67, 68, 78 Satisfaction, 115, 200
categorization and, 70-71 Scene, 103, 104,197
coding and, 66-67, 7 0 - 7 1 classical, 105
denial of, 77 change, 109
identity, 74 interest within, 206
 294 SUBJECT INDEX

Schadenfreude, 174 Sorrow, 192

Schema, 91, 113, 143, 150 Spartacus (S. Kubrick, 1960), 175
person, 160, 168 Spectacle, 175, 176, 190
social, 168 and minimal significance for protagonist,
-triggered affect, 168 176
The Secret Life of Walter Mitty (N. Z. A Star is Born (G. Cukor, 1954), 61, 125,
McLeod, 1947), 27 170
Security, 93 Starting Over, 138, 143
Self-categorization needs, 167 Steel Magnolias (H. Ross, 1989), 116
Self-selection mechanisms, 81 Stereotypes
Sensation seeking, 48 cultural, 22, 71
Seriousness, 45, 60 social, 71, 73 , 74, 165
Setting, 138, 140, 141, 142, 144, 147 and thematic structures, 123-124,
Sexual releaser, 162 127
Shot transitions, 213 gender, 170-171, 177
Significance Story, 38
abstract, 73 bottom-up processing of, 130, 150
for character, see Situtional meaning/sig- cognitive processing/comprehension of,
nificance for the character 91, 122, 123, 129, 145 n87, 213
The Silence of the Lambs (J. Demme, and anticipation, 214
1991), 167 conventional, 93
Simulation final representation of, 131
computer, 136, 149 good, 3 8
pencil-and-paper, 149 grammar, 124, 214
Situation model, 135, 197 structure, 58
Situational meaning structure, 44, 55, 63 Straf / Punishment (O. Madsen, 1974), 198,
components of, 59, 198 208, 209, 213
constant vs. variable, 198 Straw Dogs (S. Peckinpah, 1971), 62
context components of, 45, 59 Structural equation modeling, 220
core components of, 45, 60, 197 Structuralist theory of literature
of interest, 86 and reality impression, 67
vs. emotional meaning structure, 197 and viewer's activity, 91
Situational Meaning, Law of, 45, 233 and characterization, 157
Situational meaning/significance for the Superiority, 179
character, 172, 173, 182 Suspense, 101-102, 108, 113, 116, 117, 150
as part of viewer's situational meaning determinants of, 101
structure, 174 Suspension of disbelief, 228
Smultronstallet / Wild Strawberries (I. Surprise, 124, 128, 146, 149, 176, 200, 208,
Bergman, 1957), 157 210-211, 211 n l l 7
Social postdictable, 99, 134, 141, 209
facilitation, 80 n60 Suspicion (A. Hitchcock, 1941), 186
identity and categorization of character, Symmetry
167 intuition of, 131
stereotyping, 165 Sympathy, 58, 96, 97, 133, 146, 153, 160,
in-group and out-group, 167 173, 179, 200, 203, 212, 248
representation, 74 action tendency of, 178
Social cognition, 156, 159 costs of, 191
Solaris (A. Tarkovsky, 1968), 172 as disposition, 177
Solidarity with others, 191 and investment, 192
Solution, 124 disposition as condition for empathetic
Some Like it Hot (B. Wilder, 1959), 61 emotion, 178
SUBJECT INDEX 295

as emotion, 177-179 Trusting reader, 135

in relation to admiration and compas- Truth, 29, 38, 73, 126
sion, 177-181 2001: A Space Odyssey (S. Kubrick, 1968),
moral and suspense, 211 n l l 6 172
and reward, 191 Type and typing, 158, 183
Symptomatic meaning, 29 and innate releasers, 162
type vs. person, 159, see also Impression
formation, individuation in
T Typecasting, 162

TAUs, 129
Taxi Driver (M. Scorsese, 1976), 62 u
Television, 77, 78, 243 nl37
drama, 24 Unconscious impulses, 18-19
preference for and aggressive disposi- Under the Volcano (J. Huston, 1984), 157
tion, 28 Urgency, 45, 55, 59, 60, 198
entertainment, 24
selective exposure to, 23
and operant learning, 25 V
soap opera, 30
viewing vs. film, 36, 37 n27 Valence, 45, 59, 198
Tempo, 150 multiple, 61,
Tenderness, 183 Values, 22, 96 n67, 124
Tension, 79, 91, 91 n63, 116-117 Van Gogh (M. Pialat, 1992), 63
dramatic, 114 Vertigo (A. Hitchcock, 1958), 176
The Texas Chain Saw Massacre (T. Viewer
Hooper, 1974), 175 analytic training of, 65
Theater equal with character, 179
as ceremony, 23 n l 9 female vs. male, 180, see also Stereo-
Thematic structure, 121, 128, 148 type, gender
cognitive models of, 128-129 freedom of, 78
completion of, 148 intuitions, 221-223
as constraints on surprise, 133-135 as invisible witness, 76, 240-242
recursiveness of, 136 natural of traditional film, 9-11, 81, 249
Theme, 21, 29 n24, 37-38, 122-128 definition, 11
fiction, 122 empirical demarcation of, 11
intrinsic interest value of, 122-123 life experiences of, 154
as cognitive structure, 29, 127-150 personality of, 126, 143, 154
and concern, 125 preference, 50
The Third Man (C. Reed, 1949), 158, 209 self, 154
The 39 Steps (A. Hitchcock, 1935), 92 and plan understanding, 173
Thieves Like Us (R. Altman, 1974), 246 rationality of, 249-250
Those Magnificent Men in Their Flying Ma- regression, 32
chines (K. Annakin, 1956), 108 set, 11
Threat, 202 superior over character, 179
Tin Men (B. Levinson, 1987), 136 tragic, 155-156, 190-191
To Be or Not to Be (E. Lubitsch, 1942), 186 as witness/observer, 75, 86, 180, 190,
TOPs, 128 239
Torn Curtain (A. Hitchcock, 1966), 177 Viewing situation
Triumph, 60 natural, 10-11, 11 n i l
Tragic structure of feeling, 73 Visions of life/world views, 126, 133
296 .SUBJECT 11\'DEX

comic, 126 \V(lvelengtf! (M. Snow, 1967). 52

ironic, 126. 14.1 .A Wedding (R. Altman, 1978), 61
romantic. 126. 14J Whodunnit, 170
tragic, 126 Who's Afraid of Virginia Woolj? (M.
V1suallitcracy, 69 Xichols, 1966 ). 185
Voyeurism, pathological connotation of, 20 Winchester '73 (A. Mann. 1950), 99
Wish fulfillment, 26
Woman of the Year (G. Stevens, 1942 ). 177
w A lVoman Under the Influence (J. Cas-
savetcs, 1974). 157
Wait-and-see strategy, 210 Worry, 146