Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Christian Metz and the reality of film
(USC Thesis Other)
Christian Metz and the reality of film
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
CHRISTIAN METZ AND THE REALITY OF FILM
by
George Agis Cozyris
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Communication-— Cinema)
January 1979
C o p y rig h t © by G e o rg e A g is C o z y r is 1979
UMI Number: DP22243
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
Dissertation Publishing
UMI DP22243
Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
U N IVE R S ITY O F S O U TH E R N C A L IF O R N IA
TH E G RADUATE SCHO O L
U N IV E R S IT Y PARK
LOS A NG ELES, C A L IF O R N IA 9 0 0 0 7
C - f r i
c SSI
This dissertation, written by
GEp^E,^IS^CpZTOIS
under the direction of his.... Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Graduate
School, in partial fulfillment of requirements of
the degree of
D O C T O R O F P H I L O S O P H Y
Dean
Date...Q.&%MU...£$J.l7¥
SERTATIO
Q j A Chairman
...
ACKNOWLEDGMENT
I would like to thank the members of my dissertation
committee for their help and critical comments. In
particular, I wish to express my appreciation to Professor
Irwin Blacker for his encouragement and counsel throughout
the preparation of this study.
i±
TO
MY MOTHER AND FATHER
CONTENTS
Page
ACKNOWLEDGMENTS ...................................... ii
INTRODUCTION ........................................... 1
Statement of the Problem..................... 1
Significance of the Study ................... 4
Limits of the S t u d y ......................... 6
Review of the Literature................... . 10
Outline of the Study......................... 13
Chapter
I. THE TORTURED MIND ..................... 18
Reasons for Metz's "Intervention" ........... 18
Metz's Awareness of the Reality of Film . . . 22
Building on Vagueness, Uncertainty,
and Hesitancy......................... 27
Neologisms, Levels, and Ranks................. 32
The Impossible Distinctions ................. 3 5
The Drifting Point of View................... 43
Circles, Spirals, and Lyrical Passages .... 47
Proof via Manifesto and the Great Fear
in the Horizon......................... .. . 52
II. FROM MOVIES TO "CINE-TEXT" ..................... 58
"^Cinematic Language System .................. 68
Cinematic Codes and Sub-Codes................. 69
System........................................ 7 2
Message and Text.............................. 7 4
System/Code, Text/Message ................... 7 5
III. METZ AND THE FILM AUDIENCE..................... 80
Metz and the Limitations of the
Phenomenological-Structural
Linguistic Methods ......................... 80
Metz's Approach and the Film As A r t .......... 92
iii
Chapter Page
IV. THE SEVERED R O O T S ..............................'• 100
V. THE IMAGINARY SIGNIFIER....................... 116
General Characteristics and Features of
Metz’s Psychoanalytic Approach ........... . 116
From the Linguistic to the Linguistic-
Psychoanalytic Approach................... 128
The Linguistic-Psychoanalytic Approach and
the Reality of F i l m ....................... 133
SUMMARY AND CONCLUSIONS .............................. 155
BIBLIOGRAPHY .......................................... 17 0
iV)
INTRODUCTION
Statement of the Problem
Christian Metz is considered by many to be one of the
most important film theoreticians of our times. Even those
who have reservations about the validity of applying to the
study of film the principles of structural linguistics,
realize that Metz has significantly influenced those who
have been trying to develop this new "scientific" approach
to film theory and criticism. Certainly, Metz was the first
to give to the semiological, structuralist approach in film
an aura of seriousness and respectability.
By now it is practically impossible to find any piece
of cine-structuralist writing which does not contain some
reference to a concept, distinction, or argument which
appeared first in Metz's works.
In the manner of a Renaissance "master," Metz even
finds himself surrounded by student-disciples who either
undertake projects assigned by him, or conduct studies
bearing his influence.1
An increasing number of books which are offered as
textbooks for film courses at colleges and universities not
only contain references to Metz, but show very clearly the
1
Influence which his writings have had on their authors.z
Up to now the world of professional filmmaking has been
almost totally indifferent toward the cine-structuralist
movement. This has been an approach for studying film, not
for making films. However, since Metz claims that his brand
of cine-structuralism, the semiology of film, is an attempt
to understand how films communicate, one would expect that
such an understanding will be utilized eventually in
producing films. In that sense, his work can be of a
normative nature.
The problem which I tried to deal with in this study
was to determine to what extent Christian Metz has contrib
uted something of value toward our understanding of how
films communicate; something which goes beyond what we
already know about film communication from the writings of
the film theoreticians of the conventional, non-semiological,
approaches to film theory and criticism.
To deal with this problem I confronted the work of
Christian Metz with what I call in this study the "reality
of film."
This was done because, although a phenomenon may be
studied from a variety of viewpoints and Metz is entitled to
choose his viewpoint, one may not distort or pervert the
phenomenon to make it fit the viewpoint. This is a funda
mental principle of the scientific procedure.
If one claims to be studying how films communicate, one
2
must be studying tilms. One may not invent something which
is not film, study it, and hope to learn how films
communicate.
By "film" I mean the feature-length fiction film (what
Metz calls the "narrative film") which up to now has been
the object of Metz's investigations.
The reality of film, as perceived by me, contains the
following elements:
1. A narrative film has an author. Although film-
making is a collaborative art, the author of the film is the
director who bases his work on the inspiration provided by
the screenplay.
2. A narrative film is meant to be screened in a
motion picture theatre for the benefit of an audience.
The audience does not have physical access to the
print of the film; the audience experiences the film as
projected. This experience is usually limited to a single
Showing of the film although people are free to see a film a
number of times.
3. The relationship between a narrative film and its
audience is that between a work of art and a group of people
experiencing a work of art.
4. The content of a narrative film is an important
element in the comprehension, appreciation, and enjoyment
of the film.
5. Each individual film is a unique determination on
_______________ _3
the level of each shot, and as far as the overall organiza-
tion of its visual and auditory elements is concerned.
There are no two films which are alike.
6. A narrative film is rooted in a given society and
culture and reflects the period in which it was created.
7. A narrative film nearly always represents a
commercial venture requiring that the film recoup its
production cost and, hopefully, show a profit.
The film theoreticians, critics, and, in general,
"writers" of the conventional, non-semiological approach (all
those dismissed by Metz and his followers as naive and
sentimental "impressionists"), have been dealing with films
as an audience experiences them at a motion picture
theatre. The questions are, what has Christian Metz been
studying for the past decade and a half? and where does he
differ from those he criticizes?
It is hoped that by confronting his work with the
reality of film we could find suitable answers to the
problem.
Significance of the Study
As the review of the literature and the bibliography
indicate, there have been many studies on Metz's work.
However, to the best of my knowledge, there has never been
a study such as the one attempted here.
A considerable number of the existing studies are
concerned with "internal matters" within the framework of
4
the semiological and cine-semiological movement. They
analyze and argue the appropriateness and validity of Metz's
methods in terms of structural linguistics and semiotics.
In short, they accept his basic premises.
In those studies where concern is shown for Metz's
achievements in terms of our understanding of how films
communicate, there are only a few and scattered references
to the reality of film.
Most of the studies which express approval for Metz's
methods have been written by semiologists like himself, or
by those who have been influenced by him and adopted his
methods. These studies do not appear to be making an
effort to examine closely Metz's methods.
It is mostly in studies which are critical of Metz
that the references to the reality of film can be found.
But, these few and scattered references are made in connec
tion with very specific points raised by Metz, not with his
overall method. Where there are comments about his work as
a whole, they are not sufficiently documented.
The most serious problem with all of the studies which
were located is that they do not deal adequately with
something very fundamental: Is Metz really studying film?
This is a very important question for several reasons.
First, there is a need, an urgent need, to find out
more about how films communicate. Any activity which
honestly claims, or pretends to study film communication,
5
when in reality it does not, cannot advance or deepen our
knowledge and understanding of film as a medium of
communication.
Second, Metz's semiological approach has been accepted
by many as a major key for unlocking the doors to our
understanding of film communication. As a result, countless
students in film courses all over the world are struggling
with Metz's concepts and they are attempting to use his
methods to analyze films. Major periodicals in the area of
film scholarship and in other fields have devoted and con
tinue to devote precious space to articles on cine-semiol-
ogy. Professional meetings and conferences at times become
battlegrounds for the cine-structuralists and their
opponents.
If Metz is not really studying film, is it not possible
that all the activities which his work has generated can be,
somehow, harmful to film scholarship?
It seems to me that confronting Metz's work with the
reality of film has been long overdue. Unless, perhaps,
one can claim that a certain amount of distance, a perspec
tive, was required. Now, more than a decade and a half
after Metz started his investigations, perhaps it is the
appropriate time to conduct a study such as the one at hand.
Limits of the Study
This study has taken into consideration three works by
Christian Metz which are the primary sources. The first
6
is Film Language: A Semiotics of the Cinema„3 The second is
Language and Cinema.^ The third work by Metz is his essay "The
Imaginary Signifier."5
The choice of Film Language and Language and Cinema does not
really need to be defended. Although they do not contain
every essay Metz wrote in the 1960s and early 1970s, they
are without doubt the works which contain the essence of his
approach during that decade.
"The Imaginary Signifier" is a different case. It is
one of a number of essays in which Metz has tried to incor
porate psychoanalytic concepts and techniques to his already
established semiological approach. "The Imaginary Signi
fier" is characteristic, I believe, of Metz's current phase;
a phase which is obviously still evolving. As it often
happens in Metz's writings this essay contains references to
his earlier works which he attempts now to relate to his
current psychoanalytic approach.
Since his new approach is still in the process of
evolving, we cannot be very critical of what he is attempt
ing to do. However, there is substantial evidence, which
will be discussed in Chapter V, that Metz is in the process
of doing with the psychoanalytic approach what he did
earlier with the semiological.
In an effort not to lose sight of the goal of this
study, to confront Metz's work with the reality of film, and
in order to avoid the mistakes and weaknesses of previous
7
studies, I have tried as much as possible not to become
involved in the "internal affairs" of the cine-semiologists.
That is, I have tried to stay clear of arguments which
involve comparisons between the approach of this and that
semiologist or linguist.
I have also tried to disentangle Metz's views from
those of other authorities in linguistics and in cine-
structuralism. This was not an easy task because of the
manner in which Metz and his colleagues present their views.
In arguing over a point, there is a tendency of these
scholars to invoke one another as an authority on a subject,
thus creating a confusing circularity.
When Metz is applying a certain principle borrowed from
linguistics or another field, I hold him accountable for it
regardless of what a Roland Barthes or a Louis Hjelmslev has
to say about it. If one fails to do this, it is practically
impossible to move on and to attempt to talk about motion
pictures.
Related to the above point is the matter of the semi
ological and other jargon with which Metz's work is filled.
Other than dealing with this problem in Chapter I and
Chapter V, as part of Metz's tendency to constantly generate
new terms and to revise old ones, I have not attempted in
this study to keep track of any sort of evolution of Metz’s
terminology.
In a number of places in Metz's writings, especially in
8
Language and cinema, I encountered passages which appeared to
lie to be incomprehensible after considerable study. This
was obviously a source of great frustration. Only when I
came upon references by scholars who pointed out that
Metz's thought is characterized by great confusion and
obscurity, I consoled myself that there was nothing wrong
with my ability to grasp complex semiological ideas.6
One of the reasons given by certain members of the
editorial board of Screen for resigning their positions in
1976 was the obscurity and inaccessibility of the views of
those contributing to'this "bible" of cine-structuralism and
cine-semiology; especially in connection with the new
psychoanalytic approach. These former editors of Screen
spoke of dealing with the contents of the magazine as a
torment of endless rereadings in an effort to understand.7
Screen has published a considerable number of Metz's essays
as well as analyses of his works.
Upon further inquiry into the writing style of some of
Metz's mentors such as Michel Foucault and Jacques Lacan, I
came upon references to an intentional obscurity on their
part as a matter of strategy.8
I am now confident that such gaps in my comprehension
of certain passages in Metz's works are not serious enough
to undermine any important aspects of this study.
The secondary sources for this study were analyses and
evaluations of Metz's work appearing in a variety of
9
periodicals and books.
The majority of these sources presented rather polar
ized views on the value of Metz's works. The man is either
admired or dismissed for wasting our time. For this reason
I was very cautious about making use of such secondary
sources. In most cases, secondary sources merely provided
me with clues as to which aspects of Metz's work needed to
be studied more closely in terms of the confrontation with
the reality of film.
Since the works by Metz which I decided to use as my
primary sources were available in English, I saw no reason
to study them in the original French. The translator's
notes indicated no unusual problems with the translations
despite Metz's style and the subject.
Review of the Literature
The specific works consulted for this study are listed
in the Bibliography. Basically, in this review of the
literature I want to point out the way I approached the task
of acquiring a background in the fields of scholarship on
which Christian Metz bases his work. This was a necessary
task because without an understanding of the analytic tools
which Metz borrows from fields outside cinema it is not
possible to deal with his work.
After consulting a number of general works on linguis
tics, I tried to acquire a background in structural linguis
tics which is related to semiology in a number of ways.
10
In the areas of linguistics and structural linguistics
I paid particular attention to the writers, such as Louis
Hjelmslev, who exerted the greatest influence on Metz.
Although no attempt was made in my study to place Metz
in any kind of a perspective within the overall field of
literary analysis and criticism, I found it necessary to
look into fields such as Russian Formalism, Structuralism,
and French New Criticism.
Some of the key objections raised about the validity of
certain aspects of structural analysis of literature,
especially the most "mechanical" ones, are very similar to J:
the objections which have been raised about Metz's approach
to film.
Some of the most useful works which deal with problems
in structural analysis of literary works are those by
Jameson,9 Scholes,10 Segre,11 and Wetherill.12
Wetherill's book provided some interesting insights
into the problems involved in connection with Metz's claim
that his is a scientific approach to the study of film. The
extensive bibliography in Wetherill led me to other inter
esting works which deal with the subject more thoroughly.
Since there are quite a few phenomenological aspects to
Metz's approach, I made an effort to get acquainted with the
basic principles of phenomenology and its relationship to
art and literary analysis. Interesting works in this area
are those of Ihde13 and Wolf.14
. 11
Metz's references to the work of Claude Levi-Strauss
and his structural analysis of myths forced me to inquire
into Levi-Strauss and his work. In the extensive literature
on Levi-Strauss a number of works were useful to me because
certain features of Strauss' structuralism present problems
which are similar to the ones Metz faces. These are the
works of Hayes15 and Rossi.16
The dialectical aspects of Metz's method made it neces
sary for me to look into the subject of dialectics and, to
some extent, Marxist aesthetics. In this area the most
useful works are those of DeGeorge,17 Dembo,18 and Jame
son.19 An exceptionally useful and extensive bibliography
on Marxist aesthetics is that of Baxandall.20
To deal with Metz's recent efforts to introduce to his
semiology the psychology of Sigmund Freud and Jacques Lacan,
I consulted a number of works on psychology and psychiatry.
Particular emphasis was placed on analyses of those aspects
of the work of Freud and Lacan which are utilized by Metz in
"The Imaginary Signifier."
Since there are references to Melanie Klein in Metz's
essay, a number of her works and several analyses of her
writings were consulted.
The most useful of the studies on Lacan, a writer with
a particularly difficult style, as those of Miel,21 Rifflet-
Lemaire,22 Shands,23 and Wilden.21*
Repeated references by Metz to Andre Bazin and to other
12
key theoreticians of the cinema made it necessary for me to
read again certain basic texts of the cinematic literature
and to relate them to the points which Metz was making. I
paid particular attention to those aspects of Bazin's
writings which deal with the impression of reality in the
cinema.
Outline of the Study
The material in this study is organized in five
chapters, a Summary and Conclusions, and a Bibliography.
Chapter I deals with all three of Metz's works. Chap
ters II, III, and IV deal only with Film Language and Language
and cinema.. Chapter V deals with "The Imaginary Signifier."
The first chapter, entitled "The Tortured Mind," is an
examination of Metz's way of approaching his subject; it
deals with his methods and his style.
This chapter refers to all three of his works and
points out the state of confusion, frustration, tension, and
great hesitancy which characterizes his thought.
An attempt is made to describe the circularity of his
thought, the ad infinitum fragmentation of every dimension
and every element of every dimension he touches upon, the
seemingly endless neologisms, and the lack of documentation
and evidence which plague his work despite his claim to
being "scientific."
This chapter also points out that Metz's approach
resembles a game with an infinite number of rules which are
______________________________________________________________13
modified at any time to suit the purposes of the player.
I felt that such a chapter was necessary because the
manner in which Metz thinks and writes is very much related
to his inability to deal with the reality of film.
The second chapter, entitled "From Movies to Cine-
Text," examines how Metz methodically reduces film to a
detached, authorless, and lifeless object. It is this
object, and not film, which he proceeds to make the subject
of his study.
This chapter points out that Metz ignores the content
of films and attempts to deal only with codes, sub-codes,
and structures; a process which allows him to engage in very
abstract theorizing but which does not permit him to deal
with the reality of film.
The third chapter, "Metz and the Film Audience," deals
with the place of the film audience in Metz's scheme of
things. It attempts to show that the manner in which Metz
studies film is almost totally unrelated to those aspects of
the reality of film which involve the audience. One of the
things which I try to point out in this chapter is that all
the mental activity which Metz describes in connection with
his codes, sub-codes, and structures, does not describe the
experience of an audience viewing a motion _picture film, and
it is of no value in letting us know how films communicate.
The fourth chapter, "The Severed Roots," looks into the
rootless and barren nature of Metz's cine-text when it comes
14
to the social and cultural aspects of the reality of film.
The fifth chapter, "The Imaginary Signifier," deals
with the essay which establishes Metz's new psychoanalytic
direction. In this chapter, rafter a brief presentation of
the general characteristics and features of this new
approach, Metz's attempt to combine structural linguistics
with psychoanalysis is examined in terms of the key dimen
sions of the reality of film.
The Summary and Conclusions is followed by the
Bibliography.
Due to the very frequent references to the primary
sources, the following abbreviations are used in this study:
FL = Film Language, LC = Language and Cinema, IS = "The Imaginary
Signifier."
15
FOOTNOTES
xSee Sylvain du Pasquier's "Buster Keaton's Gags,"
Journal of Modern Literature , 3, No. 2 (April, 1973) , 269-291.
2A good example is James Monaco's How to Read a Film (New
York: Oxford University Press, 1977) .
3Christian Metz, Film Language: A Semiotics of the Cinema
New York: Oxford University Press, 1974). A collection of
essays originally published in French as Essais sur la
signification au cinema. On the occasion of publishing the
collection in English, Metz made a number of revisions and
added footnotes in which he qualified and modified some of
his views. However, this book still represents Metz's
point of view of the early and mid-1960s.
1 1 Christian Metz, Language and Cinema (The Hague: Mouton,
1974) . A translation into English of his Langage et cinema,
originally published in French in 1971. It is representa
tive of Metz's thought in the late 1960s and early 1970s.
5Christian Metz, "The Imaginary Signifier," screen, 16,
No. 2 (Summer, 1975), 14-76.
6 See Brian Henderson, "Metz: 'Essais I' and Film
Theory," Film Quarterly, 28, No. 3 (Spring, 1975), 12-33.
Also see Alfred Guzzetti, "Christian Metz and the Semiology
of the Cinema," Journal of Modern Literature, 3,. No. 2 (April,
1973), 272-308.
7Edward Buscombe, Christine Gledhill, Alan Lovell, and
Christopher Williams, "Psychoanalysis and Film," Screen, 16,
No. 4 (Winter, 1975/76), 119-130. Also see "Why We Have
Resigned from the Board of screen," Screen, 17, No. 2 (Sum
mer, 1976), 106-109.
8Edward Buscombe, et al., "Psychoanalysis and Film,"
120.
9Fredric Jameson, The Prison-House of Language: A Critical
Account of Structuralism and Russian Formalism (Princeton, New
Jersey: Princeton University Press, 1972).
1 0 Robert Scholes, Structuralism in Literature (New Haven:
Yale University Press, 1974).
1 1 Cesare Segre, Semiotics and Literary Criticism (The Hague:
Mouton, 1973).
16
12P. M. Wetherill, The Literary Text: An Examination of
Critical Methods (Oxford: Basil Blackwell, 1974) .
1 3Don Ihde, Hermeneutic Phenomenology (Evanston: North
western University Press, 1971).
1 * * Janet Wolff, Hermeneutic Philosophy and the Sociology of Art
(London: Routledge & Paul, 1975) .
15E. Nelson Hayes and Tanya Hayes, eds., Claude Levi-
Strauss: The Anthropologist as Hero (Cambridge: The Massachu
setts Institute of Technology Press, 1970).
1 6 Ino Rossi, ed. , The Unconscious in Culture: The Structural
ism of Claude Levi-Strauss in Perspective (New York: E. P. Dutton
& Co., 1974) .
17Richard T. DeGeorge and M. Fernande, eds., The Struc
turalists from Marx to Levi-Strauss (Garden City, N. Y.: Anchor
Books, 1972) .
18L. S. Dembo, ed., Criticism: Speculative and Analytical
Essays (Madison, Milwaukee: The University of Wisconsin
Press, 1967).
1 9Fredric Jameson, Marxism and Form: Twentieth-Century
Dialectical Theories of Criticism (Princeton: Princeton Univer
sity Press, 1972).
20Lee Bauxandall, comp., Marxism and Aesthetics::'. ' A Selective
Annotated Bibliography; Books and Articles in the English Language (New
York: Humanities Press, 1968).
21 Jan Miel, Jacques Lacan and the Structure of the Unconscious,
Yale French Studies, Nos. 36-37 (New Haven: Yale University
Press, 1966), pp. 104-111.
2 2A. Rifflet-Lemaire, Jacques Lacan (Brussels: Dessart,
1970) .
23Harley C. Shands, "Semiotic Approaches to Psychiatry,"
Approaches to Semiotics, ed. T. A. Sebeok (The Hague: Mouton,
1970), pp. 170-187.
2 * * Anthony Wilden, "Freud, Signorelli, and Lacan: The
Repression of the Signifier," American imago, 23, No. 4
(Winter, 1966), 332-366.
17
CHAPTER I
THE TORTURED MIND
This chapter deals with certain aspects and features of
Metz's methodology and style which serve as useful back
ground information in the investigation of how his work is
related to the reality of film.
The Reasons for Metz's "Intervention"
As one of the pioneers and main spokesmen for the cine-
structuralist movement, Christian Metz has been quite
critical of the conventional, non-structuralist approach to
film theory and criticism.
Film Language, Language and Cinema, "The Imaginary Signi
fier," and practically everything else he has written,
contain numerous references to the inadequacy of the
conventional approach. His criticism usually takes the form
of dismissing in a condescending manner most of the writings
of film theoreticians appearing before his "intervention" in
the early 1960s. When he does give credit to contributions
he considers useful, he places them within a historical
perspective; they were contributions which had some value
during the era in which they were made.1
18
However, Metz has been kind enough to express his will
ingness to "forgive the cinema of the past for its excesses,
because it has given us Eisenstein and a few others. But
one always excuses genius" (FL, p. 56).
What Metz finds wrong with the conventional approach is
what all the cine-structuralists object to: film has not
been studied systematically and scientifically.
Until Metz and his fellow "scientists" came to the
rescue, it is said, film theory and criticism were basically
an impressionistic account of a critic's response to films.
There was no critical method with a deductive analytic and
evaluative apparatus. Dialogue between theoreticians and
critics was an exchange of rival assertions. There was a
failure to distinguish between evaluation and description.
A subjectivism, humanism, and moralism undermined most
theories and approaches to the cinema. Classical film
aesthetics posited a unity and a coherence to every work
which required that a search be made through the content of
films to discover a "message." Marxist cine-structuralists
even suggested that conventional film theory is part of the
"conspiracy" of the bourgeois capitalist society to force
the oppressed masses to accept the "reality" of an unjust
social system.2
To all of the above reasons for rejecting the conven
tional approach, Metz added one which according to him was
quite important: the time had come for a new era.
19
Obviously echoing the manner and style of Ferdinand de
Saussure, who in the early 1900s announced that the new
science of semiotics had the right to exist,3 Metz in the
196 0s heralded in Film Language that "The time has come for a
semiotics of the cinema" (FL, p. 91).
However, for something as important as the end of an
era and the dawning of a new one featuring a new science,
Metz provided no reason other than his feeling that it had
to be so. The "old" film theory was declared to have
contributed whatever it had to contribute and it had to go.h
Half a decade after that triumphant announcement, and
after exhaustive investigations on his part, in the Conclu
sion to his Language and Cinema, Metz made another proclama
tion: ". . . the semiotics of the cinema does not yet
sxist" (LC, p. 287). It is also interesting to note that
soon after that proclamation Metz turned toward a new hope
for the proper study of film: psychoanalysis.
The structural linguistic approach which Metz applies
to film theory has as its purpose to accomplish in the field
of cinema "... the great Saussurian dream of studying the
mechanisms by which human significations are transmitted in
human society" (FL, p. 91).
Because Metz feels that the fiction film, the "narra
tive film" as he calls it, has assumed a social superiority
over other types of film, he applies his structural linguis
tic approach to this type of film only.
20
Til Language and Cinema, Metz points out that
the semiotician finds before him the already realized film. He
thus has nothing to say about how it ought to be made . . . He is
concerned with seeing how the film is constructed . . . The
semiotician . . . would also like to be able to understand how
the film is understood. (LC, pp. 73-74)
Halfway through Language and Cinema , he points Out that
the task, of the semiotics of what he calls the "filmic fact"
is "to analyze film texts in order to discover either
textual systems, cinematic codes, or sub-codes" (LC,
p. 150) . He hopes that these elements within the "filmic
fact" will enable him to understand how films communicate.
In "The Imaginary Signifier," Metz describes in this
manner the new psychoanalytic approach to the study of film:
To sum up, what I have analyzed, or attempted to analyse, in my
first two Freudian inspired studies . . . turns out, without
having precisely intended it, to be already established on one
of the flanks of the ridge-line, that of the imaginary: cine
matic fiction as a semioneiric instance, in one of these
articles, and in the other, the spectator-screen relationship as
a mirror identification. That is why I should now like to
approach my subject from the symbolic flank, or rather along the
ridge-line itself. My dream today is to speak of the cinematic
dream in terms of a code: of the code of this dream. (IS,
p. 18)
His introduction of psychoanalysis in the study of film
is justified in "The Imaginary Signifier" by saying that
"linguistics and psychology are the two main ’sources' of
semiology, the only disciplines that are semiotic through
and through" (IS, p. 28). As it happened earlier with his
heralding of a new era of the semiotics of the cinema, in
this essay a new hope is expressed and a new dawn is
predicted. He now says that the semiotics of the cinema
2JJ
will become possible through a combination of structural
linguistics and psychoanalysis.
Metz's Awareness of the Reality of Film
It is important to point out that Metz appears to be
very much aware of the reality- of film. This is clearly
evident in the primary sources used for this study and I
want to quote Metz rather extensively.
In Film Language, Metz indicates his awareness of the
reality of film in a variety of ways. On the impression of
reality upon the spectator he says:
Films give us the feeling that we are witnessing an almost real
spectacle. . . . Films release a mechanism of affective and
perceptual participation in the spectator . . . they speak to
us with the accents of true evidence, using the argument that
"it is so." (FL, p. 4)
The reason why cinema can bridge the gap between true art and
the general public ... is that films have the appeal of a
presence and of a proximity that strikes the masses and fills
the movie theatres. This phenomenon, which is related to the
impression of reality, is naturally of great aesthetic
significance. (FL, p. 5)
He identifies movement and the dimension of time as key
elements for the vivid impression of reality:
. . . why the impression of reality is so much more vivid in a
film than it is in a photograph . . . ? An answer immediately
suggests itself: It is movement . . . that produces the strong
impression of reality. (FL, p. 7)
Before the cinema, there was photography . . . But, accurate
as it was, this means was still not sufficiently lifelike: It
lacked the dimension of time; it could not render volume
acceptably. (FL, p. 14)
Through repeated references to the theatrical and
narrative nature of the fiction film he admits that story-
22
telling is an important element of the reality: of film:
The basic formula, which has never changed, is the one that
consists in making a large continuous unit that tells a story
and calling it a "movie." . . . Going to the movies is going
to see this type of story. . . . The rule of the "story" is
so powerful that the image, which is said to be the major
constituent of film, vanishes behind the plot it has woven.
(FL, p. 45)
All one retains of a film is its plot and a few images. Daily
experience confirms this. (FL, p. 46)
It is difficult to see how the cinema can ever become truly
"non-theatrical." (FL, p. 191)
Remove "drama," and there is no fiction, no diegesis, and
therefore no film. (FL, p. 194)
To exclude the dimension of the scenario from the modern cinema,
or to belittle it, is like saying that the only scenarios are
those that are like the scripts of Aurenche and Bost. (FL,
p. 203)
Metz admits that film is an art and that it started
developing along the lines of an art very early,:
From the very beginning, threatened with extinction, it became
an art. (FL, p. 58)
As an art, film possesses- the qualities of uniqueness
and one-way communication:
Every image is a hapax [a unique determination]. (FL, p. 69)
Now, like all the arts, and because it is itself an art, the
cinema is one-way communication. (FL, p. 75)
The poetic and "language of art" dimensions of film are
mentioned by Metz a number of times:
Resembling true languages as it does, film, with its superior
instancy, is of necessity projected "upward" into the sphere
of art— where it reverts to a specific language. The film
total can only be a language if it is already an art. (FL,
58)
In point of fact, the cinema is not a language but a language
of art. (FL, p. 64)
2.3J
One will have to recognize that the cinematographic enterprise,
whether successful or not, is initially poetic. (FL, p. 204)
Metz knows that, being an art, film can only give the
impression that it renders the external world as it really
is. He says:
A film is never objective. (FL, p. 195)
When it comes to the sociocultural roots of the cinema,
Metz demonstrates a very clear understanding of what is
involved in terms of this aspect of the reality of film:
There remains, of course, a problem, and that is that cinemato
graphic language itself, in as it is a body of orderings, must
certainly be influenced by various sociocultural codifications
. . . the types of filmic ordering must in one way or another
refer to given patterns of intelligibility within society.
(FL, p. 215)
Important elements of the sociocultural influences
evident in film are, according to Metz, various kinds of
"censorships";
The mutilation of the content of films is frequently the pure
and simple result of political censorship . . . More frequently
still, it is the result of commercial censorship . . . a true
economic censorship . . . so that in the cinema the problems of
content are linked to external permission much more than in the
other arts. (FL, p. 236)
An interesting kind of "censorship" which Metz calls
"The Plausible," is related to the kind of stories and other
expressive elements in films which are present or absent as
a result of pressures brought upon the filmmakers by the
social conditioning of the audiences:
The Plausible, I said, is cultural and arbitrary: I mean that
the line of division between the possibilities it excludes and
those it retains (and even promotes socially) varies considerably
according to the country, the period, the art, and the genre.
(FL, p. 244)
24
The admission that, as cultural objects, films reflect
one another, is another way by which Metz indicates his
awareness of the reality of film:
Since the film-maker shoots films, to some degree he often shoots
the films of other people, believing that he is shooting his own.
To tear oneself away, even partially, from an attraction so
profoundly rooted in the very fact of culture, in the shape of
fields, requires unusual strength of mind. Books reflect each
other; so do paintings, and so do films. (FL, p. 245)
In Language and cinema there are numerous indications that
Metz is well aware of the reality of film.
He admits that film is an art:
It is equally clear that the study of film is of interest to
aesthetics: the film is always a "work of art." (LC, p. 15)
The complexity of films is also the result of the fact that the
cinema is what we call an art. (LC, p. 37)
Film is an art because of its social status and
function:
We will maintain only that the film— and even the ugliest,
dullest, and most obscure one— is always a work of art by virtue
of its social status. (LC, p. 38)
The film is a work of art by its intention. . . . It is also a
work of art by its consumption . . . The cinema is an art because
it functions socially as such. (LC, p. 38)
As works of art, films are unique systems of expressive
elements:
Each film has its own structure, which is an organized whole, a
fabric in which everything fits together; in short a system.
But this system is valid only for one film. ... To the extent
that films are considered as unique totalities, each contains
within itself a system which is as unique as the film itself.
(LC, p. 63)
The sociocultural dimension of film is admitted by Metz
a number of times:
25
What is referred to globally as "cinema" (and to a lesser degree
as "film") is, in reality, a vast and complex socio-cultural
phenomenon, a sort of total social fact . . . which includes, as
is well known, important economic and financial elements. (LC,
p. 9)
But today the cinema, although a recent development as noted
above, has become an established cultural fact. (LC, p. 10)
A film, in other words, is not only an example of cinema, but
also a culture. (LC, p. 72)
As cultural objects films contain reflections of the
society in which they are produced and "consumed":
It is common occurrence that films, consciously or not, reflect
various systems of political thought. (LC, p. 99)
The film reflects social behavior, but may also remodel it to a
certain extent. (LC, p. 116)
In "The Imaginary Signifier" there are quite a few
indications that Metz is aware of the reality of film. Most
of them reflect Metz's marxist orientation, but they are
still indications of an awareness of this reality:
It has very often, and right, been said that the cinema is a
technique of the imaginary. A technique, on the other hand,
which is peculiar to a historical epoch (that of capitalism) and
a state of society, so-called industrial civilization. (IS,
p. 15)
Let me insist once again, the cinematic institution is not just
the cinema industry . . . it is also the mental machinery— another
industry— which spectators "accustomed to the cinema" have
internalized historically and which has adapted them to the
consumption of films. (IS, pp. 18-19)
Metz knows well that films are not abstract linguistic
entities but popular entertainment presented for the enjoy
ment of a fee-paying audience:
The cinema is attended out of desire, not reluctance, in the hope
that the film will please, not that it will displease. (IS,
P- 19)
2f>
The "desire to go to the cinema" is a kind of reflection shaped
by the film industry, but it is also a real link in the chain of
the overall mechanism of that industry. It occupies one of the
essential positions in the circulation of money, the turnover of
capital without which films could no longer be made. (IS, p. 19)
If I am concerned to define the cinematic institution as a wider
instance than the cinema industry ... it is because of this dual
kinship . . . between the psychology of the spectator . . . and
the final mechanisms of the cinema. (IS, p. 20)
While discussing the traditions from which the cinema
has drawn its expressive means, Metz admits again that film
is an art deeply rooted in western culture:
Since its birth at the end of-the nineteenth century the cinema
has, as it were, been snapped up by the Western, Aristotelian
tradition of the fictional and representational arts, of diegesis
and mimesis, for which its spectators were prepared— prepared in
spirit, but also instinctively— by their experience of the novel,
of theatre, of figurative painting, and which was thus the most
profitable tradition for the film industry. (IS, p. 44)
With such an awareness of the reality of film, one
cannot help but wonder why and how could Metz fail to deal
with this reality in his effort to study how films communi
cate.
Building on Vagueness, Uncertainty,
and Hesitancy
For someone setting out to create a new science, the
semiotics of the cinema, Metz has the right to be cautious.
In Film Language he takes great pains to develop very slowly
his concepts; especially that of cinematic narrativity.
Very large sections of Language and Cinema are devoted to a
painstaking clearing of the ground for the establishment of
what he calls the "cinematic codes" and the "distinctive
units." In "The Imaginary Signifier" it takes Metz fourteen
27
pages (out of the sixty-three page essay) to state why he is
interested in a psychoanalytic approach to film, and then
another eighteen pages for him to indicate what the subject
of this essay is going to be.
However, there is a difference between a slow, delib
erate, and methodical approach, which is based on solid
ground, and an approach which is based on methodical vague
ness, uncertainty, and hesitancy.
In Film Language, where he is setting the foundations
of his cine-semiology, Metz declares that "I am persuaded on
the contrary that the 'filmolinguistic' venture is entirely
justified, and that it must be fully 'linguistic'— that is
to say, solidly based on linguistics itself" (FL, p. 60).
He then proceeds in a very efficient manner to discover
not the ways in which the cinema is like a language which
can be studied through linguistics, but to discover the many
ways in which the cinema is not like a language.
By his own admission, the cinema differs from language
in the following fundamental ways:
There is nothing in the cinema that corresponds, even metaphor
ically, to the second articulation. (FL, p. 61)
The cinema has no phonemes; nor does it, whatever one may say,
have words. Except on occasion, and more or less by chance, it
is not subject to the first articulation. (FL, p. 65)
There is no precise syntax in the cinema in the proper linguistic
sense of the term "syntax." (FL, pp. 67-68)
There are many characteristics to the filmic image that distin
guish it from the preferred form of signs— which is arbitrary,
conventional, and codified. (FL, p. 76)
.............................. 2J8
The paradigmatic category in film is practically non-existent.
(FL, pp. 68-69, 101)
The cinema is not, a language system, because it contradicts three
important characteristics of the linguistic fact: a language is
a system of signs used for inter-communication . . . Now, like
all the arts, and because it is itself an art, the cinema is one
way communication. (FL, p. 75)
In the cinema connotation is nothing other than a form of
denotation. (FL, p. 118)
I would say, on the contrary, that the cinema has never had a
grammar. (FL, p. 209)
While he is discovering that cinema is not like a
language, Metz begins to attribute to this entity the char
acteristics of a language so that he can proceed anyway and
study it using an approach which is solidly based on
linguistics itself.
In this procedure, the man who intervened in the
course of film theory to do away with "impressionism," makes
statements such as these:
[Cinema] "resembles true languages." (FL, p. 58)
iThe image] "corresponds to one or more sencences." (FL, p. 65)
There is a syntax of the cinema, but it remains to be made. (FL,
p. 67)
The image is a form of speech. (FL, p. 67)
Thus a kind of filmic articulation appears. (FL, p. 99)
The cinema is certainly not a language system (langue). It can,
however, be considered as a language. (FL, p. 105)
Although it lacks distinctive signifying units, film proceeds by
whole "blocks of reality," which are actualized with their total
meaning of discourse. (FL, p. 115)
These "blocks of reality" bring about "a kind of
articulation."
251
All that can be affirmed, therefore, is that a shot is less
unlike a statement than a word but it does not necessarily
resemble a statement. (FL, p. 116)
In the middle of this process of attempting somehow to
relate film to language, Metz apparently senses that he can
no longer sustain his earlier dogmatism, and he declares
that "the concepts of linguistics can be applied to the
semiotics of the cinema only with the greatest caution"
(PL, p. 107).
But the problems created by the effort to treat film as
if it were a language, when he knows very well that film
differs fundamentally from a language, overwhelm him and a
general and pervading uncertainty, timidity, and hesitancy
undermines practically everything he is trying to say and
to establish in Film Language. Metz's writing contains every
indication that he does not have the situation under
control. Here is a sample of the manner in which he
expresses his thoughts:
"It is reasonable to think" (FL, p. 7) . "One might add" (IS,
p. 10). "Let us say, therefore--perhaps a little cavalierly— "
(FL, p. 17). "One can perhaps explain" (FL, p. 19). "We all
assume" (FL, p. 19). "Nevertheless— aside from the fact that
the distinction I am suggesting here is, in more than one case,
extremely difficult to maintain ... it seems to me" (FL, p. 27).
"Let us indeed suppose that in certain circumstances a shot can
appear to be equivalent to several sentences" (FL, p. 66). "One
can of course conclude that the cinema is not a language" (FL,
p 89). "It is not possible here to give this table in its com
plete form . . . let us content ourselves, then, with the almost
unpolished 'result'" (FL, p. 119). "On a level roughly corre
sponding to that of the 'sequence'" (FL, p. 120). "We are told
nothing, yet we are informed that a great deal could be told us"
(FL, p. 132). These three reasons might eventually compel us to
revise the status (FL, p. 133). It is impossible to locate pre
cisely the threshold separating elements we call "large" from
those we call "small" . . . and yet, however uncertain its
30
location, the existence of the threshold . . . is beyond doubt
(FL, pp. 140-141). A truth that is extremely difficult to
define, but that, somehow, one places instinctively (FL, p. 197).
The filmic laws are most probably located far beyond the place
one usually expects to find them— that is to say, on a much
deeper level, a level in some ways prior to the differentiation of
verbal language . . . from other human semiotic systems (FL,
p. 209) .
In Film Language, Metz's inability to make any statement
without undermining it by repeated and lengthy qualifica
tions, admissions of inconsistencies, and references to
hopes for improvement in the future, extends even to his
footnotes which are, at times, much longer and more complex
than his text.
The same uncertainty and hesitancy is present in
Language and Cinema and it is expressed through figures of
speech very similar to those used in Film Language. In
Language and cinema one of the best examples of his hesitancy
and confusion appears when he is discussing the "code of
montage." After he presents a fairly elaborate case for the
existence of such a code, he concludes: "Of course,
strictly speaking there is no code of montage" (LC, p. 140).
Five lines of text later he reverses himself: "But in
another sense there is a code of montage." Four lines later
he says: "In the extreme and perhaps by forcing things a
little, one could conceive of a particular type of code"
(LC, p. 140).
In Language and Cinema, the position that the semiotics
of the cinema can be based solidly on linguistics is aban
doned. Now Metz believes that ". . .a semiotic analysis
31
inevitably encounters sociology, cultural history, aesthet
ics, psychoanalysis, etc." (LC, p. 16). As it has been
mentioned already, Language and cinema ends with the conclusion
that the semiotics of the cinema does not yet exist.
Between the statement on page sixteen and the disturbing
conclusion, Metz does not deal with cine-semiology in terms
of sociology, cultural history, aesthetics, or psychoanaly
sis. He continues in the vein of Film Language but in a much
more technical and detailed fashion. This highly technical
approach is characterized by what appears to be a certain
degree of self-confidence. However, this is not real self-
confidence but the freedom which he has granted to himself
to do and to say whatever he pleases. Barely under the
surface is the same lack of certainty which characterizes
his work in Film Language.
Neologisms, Levels, and Ranks
One of the most disturbing qualities of Metz's approach
and style is his practice of coining new terms every time he
discovers that he cannot deal adequately with those which
are already available. Film Language and Language and Cinema
are filled with Metz's neologisms. Some examples of the
terms created by him are the following:
In Film Language: "Straightforward Spatiotemporal
Break." This is what has been known in film as a "Fade."
"Spatiotemporal Break with an Underlying Transitive Link."
32
Cn the language of the discredited nimpressionistic" fTIrri
theory this is known as a "Dissolve." "Alternate Syntagma,"
"Alternating Syntagma," "Frequentative Syntagma," "Full
frequentative Syntagma," "Bracket Frequentative Syntagma,"
"Semi-Frequentative Syntagma," "Nondiegetic Insert,"
"Displaced Diegetic Insert," "Extracinematographic," etc.
In Language and Cinema : "Cinematic Non-Filmic," "Abso
lute Film," "General Cinematic Codes," "Partial Cinematic
Codes," "The Cinema-Fact," "Singular System," "Cinematiza-
tion," "Filmization," "Opposing Irredentism," "Global Text,"
"Syntagi®-Token, " "Remote Co-presence," "Paradigm-Token,"
"Mode of Cinematicity," "Suprasegmental Distinctive Units,"
"Plurifilmic Textual System," "Heterogenous Simultaneous
Syntagm," "Homogenous Temporal Syntagm," "Intermixed
Specificities," "Metaparadigms," "Meta-Syntagms," "Oblique
Syntagms."
In "The Imaginary' Signifier"': . . "Sahcipioning Construc
tion," "Juxtastructure," "Circumstantial Denotation,"
"Significatory Pressure," "Manifest Filmic Material,"
"Proximate Signification," "Great Global Signified," "Super
Genres," "Great Cinematic Regimes," "Referential Illusion,"
"Cinematic Scopic Regime," "Cinematic Voyeurism," "Cinematic
Scopophilia," "Cinematic Fetishism."
Newly coined terminology can be troublesome. What
makes Metz's creations particularly disturbing is that most
of these new terms identify either relationships between
entities which Metz does not describe clearly, or they are
-------------------- _33
labels for items which belong to classification systems
which, with one exception, are never developed or explained
fully.
As far as classification systems are concerned, most
of the time Metz creates rung after rung for ladders which
he never constructs. The one classification system which
he describes in some detail is in Film Language and he labels
it "The General Table of the Large Syntagmatic Category of
the Image-Track."
But even this Grande Syntagmatique, supposedly the crown
ing achievement of his Film Language, is nothing more than a
presentation in linguistic jargon of types of shots and of
relationships between shots which have been discussed in
the works of the "conventional" theoreticians of the cinema
decades before Metz's "intervention."5
For example, a "displaced diegetic insert" is, by
Metz's own example, nothing more than a single shot of the
pursued in a sequence of shots showing the pursuers (FL,
p. 128).
And his "parallel syntagma" is none other than the
familiar parallel montage sequence.
The value of the Grande Syntagmatique is further under
mined by Metz himself when, attempting to praise Jean-Luc
Goddard, he discovers that Goddard is using in one of his
films a novel kind of syntagma (Metz calls it a "potential
sequence") which the Grande Syntagmatique cannot account for
34
(FL, pp. 217-220).
The Impossible Distinctions
Related to the tendency to coin new terms and to
create category after category of items which fit in a
model of the whole which is never presented,6 is Metz's
habit of creating a great number of distinctions so that
there is hardly a paragraph or a page in Film Language and in
Language and cinema that does not contain some attempted
distinction. The problem with these distinctions is not so
much their number and the fact that they appear with such
frequency that there is no way for someone to reflect
adequately on what is being said; what is really disturbing
is that Metz is unable to establish that there is a differ
ence between the items he attempts to differentiate.
Some examples of these attempted distinctions, with
brief comments, are the following:
In Film Language
Image - -J5 imp 1 e Desc r i P t i o n — Nar r a t i ve (FL, p. 18)'. Metz
gives examples to illustrate the difference between these
three items: IMAGE: "A motionless and isolated shot of a
stretch of desert . . . (space-significate— space-signi-
fier)." SIMPLE DESCRIPTION: "Several partial and success
ive shots of the desert waste . . . (space-significate--
time-signifier)." NARRATIVE: "Several successive shots of
a caravan moving across the desert . . . (time-significate—
time-signifier)."
35
Distinctions of this kind generate a lot of very
disturbing questions: What does he mean by a "motionless
shot?" A static-camera shot? A freeze-frame? A "still
nature" shot without a human element in it? Can a motion
picture shot exist which does not contain the "time
signifier" element? What is the important difference
between the shots making up a "simple description" and a
"narrative?" The moving caravan? In that case, is the
human element what distinguishes "description" from "narra
tive?" Why is not Metz's "narrative" a "description?" Why
is not his "description" a "narrative?" Why is the element
of space-significate absent from what he defines as a
"narrative?"
Semiotics of Connotation— Semiotics of Denotation (FL,
p. 96). Metz attempts to distinguish between these two
types of approaches, and he indicates that he will study
the semiotics of denotation. However, he is unable to make
a clear distinction between these two because of that aspect
of the reality of film which involves the impression of
reality captured by the image.
In a motion picture shot (in the image) connotation
and denotation fuse and Metz admits this a number of times.
Finally, he is forced to say that connotation and denotation
have "a common root in the actual perception of the
spectators" (FL, p. 114). Then he has to invent new
distinctions in order to proceed:
16
This is why, provisionally, I use the term iconography to
designate the prefilmic connotations of objects, in order
to distinguish them from— and at the same time draw them close
to— the iconology (likewise prefilmic) that organizes the
denotation of these same objects. (FL, pi 114)
A few pages later, Metz admits very openly that "in
the cinema, even more than elsewhere, connotation is nothing
other than a form of denotation" (FL, p. 118).
His inability here to distinguish between connotation
and denotation is a very serious matter because he proceeds
anyway to take an approach which favors denotation, and
this prevents him from focusing properly the subject of his
investigation. Since, by his own admission, it is the
reality of film which prevents a distinction from being made
between connotation and denotation, Metz cannot be blamed
for failing to make the distinction. His error is that he
attempts to go against the reality of film and, as a
result, he has to create a great number of other unreal
distinctions in order to proceed with his task.
Visual— Photographic Elements (FL, 106). Metz attempts
to distinguish between "visual" and "photographic" elements
in film. By way of example he indicates that fades and
dissolves are visual but not photographic.
How can a valid distinction be made between these two?
Under what kind of logic can a dissolve or a fade in motion
pictures be visual and not photographic?
In Metz's writings such obscure distinctions open up
new territories of problems, which require more obscure
_____________________ 37
distinctions which create new problems, ad inifiniturn.
Minimum Segment— Autonomous Segment (FL, p. 124). Metz
has barely presented this distinction when he points out
that "This, as we will see shortly, does not prevent a
minimum segment from being occasionally autonomous" (FL,
p. 124).
There is pattern in Metz's writings, described in more
detail later on in this chapter under "Circularity, " where he
works very diligently to make a distinction or to establish
the validity of something, and then he proceeds to undermine
it by saying that it is not valid.
Location of the Signifier— Location of the Significate
(FL, p. 127). Metz points out that "The Screen is the
location of the signifier, and the diegesis is the location
of the significate" (FL, p. 127).
In what sense is the screen the location of the
signifier or anything else? Is the screen really the
location or the mind of the spectator? Metz gives the
impression that he is referring literally to the projection
surface, the screen, as the location of the signifier. Why
is not the diegesis "located" on the screen since what is
"diegesized" appears on the screen? How is it possible to
distinguish between the location of the signifier and the
location of the signified when it has been impossible to
distinguish between the signifier and the signified?
Ordinary— Episodic Sequence (FL, p. 130). Inside what
Metz calls the "sequence proper," further described by him
as a "single, discontinuous temporal order," he distin
guishes between an "ordinary sequence" and an "episodic
sequence" (FL, p. 130).
For him the "ordinary sequence" is one where the
temporal discontinuity is unorganized. The "episodic
sequence" is one where the temporal discontinuity is
organized.
What is, may we ask, an unorganized temporal discon
tinuity, or an organized temporal discontinuity? And why
call the first one "ordinary" and the second "episodic?"
In Language and Cinema.
This book, is' also filled-with tvery. troublesome
distinctions. This is a brief sample of these distinctions:
Cinematic Fact— Filmic Fact. The distinction between
what is cinematic and what is filmic is a major one, and
it is related to Metz's effort to limit his study. It is
also a very important distinction because, as we shall see,
it becomes the source of very serious problems in terms
of the reality of film.
In his usual style, Metz defines things gradually. He
starts by pointing out that:
Film is only a small part of the cinema, for the later represents
a vast ensemble of phenomena some of which intervene before the
film . . . [some] after the film . . . [ some] during the film
. . . [some] aside from and outside of it.
_____________________________ L______ 39
The importance of making this distinction between the cinematic
and the filmic fact lies in the fact that it allows us to
restrict the meaning of the term "film" to a more manageable,
specifiable signifying discourse, in contrast with "cinema"
which, as defined here, constitutes a larger complex (at whose
center, however, three predominant dimensions may be distin
guished: the technological, the economic, and the sociological).
(LC, p. 12)
Since Metz's interest lies with the semiotics of the
cinema, he observes that
It is clear that the so-called semiotics of the cinema is
primarily concerned with the "filmic fact." In spite of
inevitable areas of overlap, the semiotics of the cinema cannot
usefully contribute to the understanding of the "cinematic fact,"
at least not in any direct manner, and cannot at the present
stage of research. Semiotics, whether of the film or anything
else, is the study of discourses and texts.
Suffice to say that what we shall call "film," except where
otherwise indicated, is film as a signifying discourse (text),
or as a linguistic object: Cohen-Seat's filmic fact . . . the
film functioning as an object perceived by the audience for the
duration of the projection. It is the "filmophanic" film, and
it alone, that we shall call "film." (LC, p. 13)
In terms of the reality of film the above comments
about the "filmophanic" film are very important.7 Their
importance begins to emerge more clearly as we follow what
else Metz has to say in this section of Language and Cinema.
His text is being compressed to leave out his usual
complicated digressions and references to hopes for
improvement of his technique in the future:
1. The notion of filmic fact . . .is, however, still too vast
to attribute to it the principle of analytic distinctiveness
suitable to a semiotics of the film, for the film itself is
a multi-dimensional phenomenon. (LC, 14).
2. It would be tempting to say . . . that within the filmic fact
itself, two or three types of phenomena may be isolated— for
example, the psychological, sociological, and aesthetic— which
40
are not of direct concern to semiotics, and which should
therefore be restricted to the study of film considered
as a language system.
To this extent, the semiotics of film is inextricably tied
to "psychological" considerations . . . utilized in a dif
ferent perspective.
In this way, a semiotic analysis inevitably encounters
sociology, cultural history, aesthetics, psychoanalysis,
etc. (LC, p. 16)
3. . . . film . . . is a mode of expression in which language
and art maintain a quasi-consubstantial relation to one
another and where language itself is a product or an aspect
of artistic invention.
A semiotic analysis is thus closely associated with the
aesthetics of film. (LC, pp. 16-17)
Metz has barely completed his comments about the filmic
fact and the differences between the filmic and the cine
matic when he makes the following observations:
"Cinema," in fact, is not always (as it is for Cohen-Seat) the
sum of that which is related to the film but external to it;
even at the heart of a filmic analysis cinema continues to be
a notion which intrudes itself at every turn, and which is
hard to imagine doing without.
And it is clear, although paradoxical, that the cinema thus
conceived is situated within what Cohen-Seat calls the filmic
fact. (LC, p. 22)
Later on in Language and Cinema, Metz writes:
There is something which nevertheless should be called to the
reader's attention: the distinction between the cinematic and
the filmic, which everyday language maintains for better or
worse (obstinately and confusedly at the same time) and which
we would like to make explicit here, does not prevent, in
certain propositions, cinema and film (cinematic and filmic)
from becoming interchangeable and from being spontaneously
felt as such in writing. (LC, p. 55)
As if this were not enough of an admission that he has
failed to make the distinction between "cinema" and "film,"
41
he repeats:
The contexts in which "cinema" and "film" (or "cinematic" and
"filmic") become interchangeable are quite numerous. (LC, p. 56)
However, in the remaining nine chapters of Language and
Cinema, Metz acts as if the distinction between "film" and
"cinema" has been made successfully, and he develops his
analytic procedures so that he leaves out the "cinematic."
Also, no matter what Metz says about aesthetic, psycholog
ical and socio-cultural dimensions of the "filmic," in the
remaining nine chapters he does not deal at all with these
dimensions.
Code— Message. All the distinctions and all the
analytic elements which are associated with the concept of
the "code" are of great importance to Metz because of the
central position which the concept of the "code" occupies in
his work. For this reason it is quite disturbing to find a
distinction such as this:
Code always differs from message in that one is code and the other
message, and— even if the complete list of what one finds in the
message were identical with the total list of what one introduces
into the code ... it would still be true that these traits
should be understood in the future as being associated with one
another throughout a given discourse, and thus linked by the
coherence of a logic which is always tacit. (LC, p. 59)
However, he has never attempted to give examples from
actual motion picture films to illustrate the difference
between message and code.
General— Particular Cinematic Codes. Metz makes a
determined effort to distinguish between "general" and
42
"particular" cinematic codes. As this effort reaches its
conclusion he declares:
The distinction between "general" and "particular" cinematic codes
is somewhat crude, and fails to account for different degrees of
generality which a cinematic code may represent. (LC, p. 79)
On the next page Metz continues:
It should be understood, then, that the so-called "particular"
codes are codes which are more or less particular— and, from
this, that the dual division of general and particular codes has
as its only effect (and also its only goal) to distinguish clearly
from all others those cinematic systems which present the maximum
degree of generality. (LC, p. 80)
The Drifting Point of View
As a "scientist" Metz has the right to shift his point
of view as he proceeds with his investigation. Viewing
one's subject from a number of viewpoints can be used as a
safeguard against errors in observation and experimentation.
However, Metz does not merely change his point of view
from time to time and according to some plan which becomes
apparent to us. His is a continuously shifting, drifting
point of view following a pattern which he seems to be
making up as he goes along. This disturbing situation is
evident in all his writings and it is especially obvious in
Language and Cinema .
Since this practice is a matter of Metz's overall style
and approach, it is not possible here to quote all the
passages where it occurs. Instead, an attempt is made here
to isolate what causes him to do this and to point out the
consequences of this practice.
There seem to be the following reasons for Metz's
43
drifting point of view: '
1. Metz assigns to himself an impossible task which
prevents him from focusing properly on the subject of his
study. He wants to assume a scientific approach although he
is dealing with cinema which, by his own admission, is an
art.
He wants to base his approach solidly on linguis
tics although he realizes that the cinema is not a language.
He attempts to deal with cinema in terms of semiot
ics although in Film Language he admits that the semiotics of
the cinema do not exist yet but merely have the right to
exist, and in Language and cinema he admits that the semiotics
of the cinema do not exist yet.
As a result Metz is forced to take an extremely
abstract and theoretical approach containing difficulties
which he cannot overcome. Good examples of this are his
inability to overcome the problems created by the reality of
the image, which causes a fusion between signifier and
signified, and the difficulties he encounters by the lack
of a distinction between denotation and connotation.
2. Metz implies that the nature of structural analysis
is such that the shifting point of view is a methodological
requirement accounting for the dynamic interaction of the
elements found in structures.
An example of this appears in Film Language where he
talks about the necessity of accepting, as a methodological
44
requirement, the principle of a "perpetual see-saw" (FL,
pp. 143-144).
However, the "perpetual see-saw" is occurring
between so many elements and simultaneously on so many
planes and levels that confusion results. Coupled with the
great obscurity of many of Metz's distinctions and the other
shortcomings of his approach which are described in this
chapter, the "perpetual see-saw" brings about a sense of
disorientation.
An additional troublesome quality of this "see
sawing" is that Metz often describes in such a manner the
activity occurring between structural elements that he gives
the impression that he is talking about something which is
alive. Of course, all the perceived activity is in the mind
of the analyst, but Metz talks about this activity as if
these elements within the structures were interacting with
one another independently of the will and the consciousness
of the analyst.
This practice cannot be dismissed merely as an
element of the style of the author. Together with Metz's
very frequent use of the term "object" in connection with
things (such as "texts") which are not objects, this prac
tice constitutes a deception. It is deceptive because it
suggests that there is something tangible and verifiable
about the structures and the activity occurring within them,
when in reality we are not dealing with physical objects in
45
the material world and therefore nothing really can be
verified.
3. Metz utilizes a dialectical approach, or, at least,
an approach freely inspired from dialectics, which advances
on the basis of continuous and often overlapping transforma
tions and polarities.
This aspect of Metz's approach permits him to do
and to say anything he pleases, since he can assume any
position he wants to assume and, at the same time, embrace
the exact opposite position, or anything between these two
extremes.
As a result of this constant shifting Metz becomes
very elusive, and it is practically impossible to hold him
accountable for anything. He covers all bases while
occupying none.
4. Metz claims that his work is exploratory in nature
and he is performing a task long overdue in the history of
film theory. For this reason, he claims, it is proper for
him to probe in different directions through successive
questioning and not to be concerned with the fact that
there may be inconsistencies in his approach.
Time after time in Film Language and in Language and
Cinema, he expresses the hope that, somehow, the day will
come when the problems he uncovers (or creates) will be
solved through an increased efficiency of the cine-semiotic
techniques.
46
This faith in the future is also expressed in "The
Imaginary Signifier," but this time not in terms of linguis
tics but of psychoanalysis. In "The Imaginary Signifier,"
however, Metz apparently has began to suspect that this
constant shifting and his fragmenting of every dimension and
every element he touches upon may never come to an end. He
speaks of a "perpetual possibility of a finer... structura
tion," and of the fact that "the analyst will never complete
his exploration of it [he means the system of a film] and
should not seek any 'end'" (IS, pp. 3 5-36).
Circles, Spirals, and Lyrical Passages
The difficult situation in which Metz puts himself is
also reflected in the circularity of his thought and in his
very frequent excursions into a very lyrical style of writ
ing during moments when he ought to be very precise.
Some examples of this situation and of the problems it
creates are the following:
In Film Language, Metz says that "The specific nature of
film is defined by the presence of a langue tending toward
art, within an art that tends toward language" (FL, p. 59).
The above definition, cryptic and spiraling as it is,
may appear to make some sense until one reads, six pages
later, that "it seems appropriate to look at cinema as a
language without a system" (FL, p. 65), and then ten pages
later that "The cinema is not a language system" (FL,
p. 75) .
_____________________________________________________ ___________________ . ..... 43
His definition (on page 59) of the "specific nature of
film" also becomes an invitation to a very tortuous journey
because the concepts of language system (langue) and language
(langage) are drawn by Metz through the territory of art. In
this section Metz becomes involved in his impossible
distinctions and the matter of the specific nature of film
remains unresolved.
The largest in number and the most troublesome of
Metz's circles and spirals occur in Language and cinema.
Important parts of Language and Cinema where there is a great
concentration of very disturbing circularity are Chapter I
("Within the Cinema: The Filmic Fact"), Chapter II ("Within
the Filmic Fact: The Cinema"), and Chapter III ("Film in an
Absolute Sense").
In these three chapters the "filmic" and the "cine
matic" are drawn apart, forced to intersect, overlap, and
fuse into one another. After proceeding in this manner for
forty-five pages, he says:
There is, however, one trait— and one alone— that is shared by
all those things called "cinematic," in the loose sense of the
word: they are all phenomena which are immediately discoverable
in films, phenomena which are "found" in films, which the
investigator can "attest" in films, which have films as their
place of manifestation.
In addition to being dangerous, as we have said, it is somewhat
paradoxical to call these phenomena cinematic, since another
adjective exists which is so-to-speak ready-made to refer to them
and which would appear to be naturally (or rather linguistically)
predestined to the task: the adjective filmic. In fact, what is
filmic, if not the sum of what appears in films? (LC, p. 46)
Then, Metz tries to excuse his circularity by claiming
48
that this is a feature which appears in other fields:
The quite considerable extent of this zone of overlap should not
be surprising, and the overlap itself is no anomaly to be "done
away with." This phenomenon of contextual neutralization is not
restricted to the domain of the cinema (or to the film!); its
equivalent may be found in other semiotic studies, even in
linguistics. (LC, p. 57)
Finally, Metz blames this circularity on the lack of
"rigor" in film research:
We can understand, then that if the distinction between cinema
and film is sometimes a source of difficulties, it is not because
the two notions are inherently ill-defined, or that their opposi
tion is complex and ephemeral, but rather than habits of rigor
are less well established in research on the cinema than in
other disciplines. (LC, p. 57)
Other examples of Metz's circularity and the overall
obscurity present in his work are the following:
Every code is a system, and every message is thus a text. But
the inverse is not true, and certain systems are not codes but
singular systems (despite their systemic nature, they involve a
single text); and certain texts are not messages but singular
texts: they constitute the single manifestation of a system, not
one of the multiple manifestations of a code. (LC, p. 75)
Just as all that is codical in texts is general, everything which
is general in texts is codical. (LC, p. 154)
. . . in this code, syntagmatics and paradigmatics constantly
refer back and forth to one another and are, properly speaking,
inseparable . . . A code is both a paradigmatic and a syntagmatic
mechanism. (LC, p. 170)
All that which is specifically filmic is filmic, but all that
which is filmic is not specifically filmic. (LC, p. 40)
In Chapter X ("Specific/Non-Specific: Relativity of
the Classification Used"), after attempting another major
distinction which ends in a circularity from which he cannot
escape, Metz declares:
The two circles are "concentric" (an approximate metaphor, which
should not be taken too literally). (LC, p. 229)
____________________________, ______________________________________________________ A S
The impossible situation in which Metz places himself"^
and the general confusion which results from this, forces
him to become quite lyrical and "impressionistic" during
crucial moments in the development of his ideas when the
"specificity" he has talked about ought to be present. A
good example of this lyrical quality in Film Language involves
his treatment of the concept of the "current of significa
tion," which he borrows from Bela Balazs, and which he
renames the "current of induction" (FL, pp. 46-47).
According to Metz, this "current" accounts for the
connections which are made in the spectator's mind when two
images are juxtaposed. It is the force which "refuses not :
to flow whenever two poles are brought sufficiently close
together, and occasionally even when they are quite far
apart" (FL, p. 47). Considering that Metz places great
emphasis on the syntagmatic dimension, this mysterious and
almost mystical circulating "current" is a very disturbing
impressionistic element in his work.
Language and Cinema, which is supposed to be one of
Metz's major contributions to the new "scientific" litera
ture of the cinema, contains many lyrical elements. Here is
a small sample of these lyrical elements:
1. Metz indicates that "each film is built upon the
destruction of its own codes" (LC, p. 102).
How can a film "destroy" its own codes? Since he
states very often that the codes are creations of the
5jO J
analyst, is this "destruction" performed by him? How?
Under what circumstances? For what purpose?
2. Metz describes activities between the film codes in
terms of harmony and peacefulness and then he adds:
The system of the text is the process which displaces codes,
contaminating some by means of others, meanwhile replacing one
by another, and finally— as a temporarily "arrested" result of
this general displacement— placing each code in a particular
position in regard to the overall structure, a displacement
which is itself destined to be displaced by another text. (LC,
p. 103)
How does a system "displace" codes? Is there a
process of "contamination?" What does it consist of?
"Destined" to be replaced" What force determines this
"destiny?"
3. In his attempt to give examples and refer to actual
films that supposedly illustrate a point he is trying to
make, a rare occurrence in his writings, Metz talks in terms
of "luminocity," "existential fracture," "phenomenological
schizophrenia," "perceptual sequential derailments,"
"tragicomic sense of a constant and derisive reversibility
of possibilities," "aggressive-arbitrary variegation of the
image," "abundant, egotistical, playful exuberance of the
picture track," and "tortured cult of the female idol" (LC,
p. 113).
4. Describing the alleged activity between textual
systems in a film Metz points out that:
. . . the plurality of textual systems in a film can never be
resolved into a neutral co-habitation of this sort, for each of
them is obliged to lay claim to the textual territory in its
totality, such that this totality is perpetually torn apart by
its opposing irredentisms. (LC, p. 120)
51
For some reason Metz makes a number of references in
his works to "irredentism" and the "irredencent" quality of
cinematic elements. He has never explained what is the
exact nature of this "irredentism."
Proof Via Manifesto and the
Great Fear in the Horizon
Obviously Metz has a good understanding of the fields
of linguistics, structuralism, and psychology. Most of the
references to the methodological tools he borrows from these
fields are supported by invoking recognized authorities in
these fields. There is no doubt that he is well-versed in
film theory and, as indicated earlier, his awareness of the
reality of film is evident in all his writings.
Under these circumstances one would expect him to be
able to document the claims he makes about the manner in
which signification is achieved in motion pictures. This,
however, does not occur in his works. Metz always advances
from point to point via manifesto. Things are what he says
they are because he says so. He provides no evidence that
his conclusions are based on testing of any kind. Examples
in terms of actual films are very rarely given. When
examples are given they fail to add anything of real
significance to our understanding of the specific films
involved, or to our understanding of how films communicate.
His authoritanism and the lack of real scientific rigor
on his part are somewhat concealed by the fact that he is
52
almost constantly inviting the reader to go along with what
he is proposing for the sake of whatever argument he is
presenting. The impression is given that the reader is
invited to participate in an intellectual game which requires
agreement with statements which are extremely vague, circu
lar, and unsupported by documentation and evidence.
In Film Language he devotes three out of his ten chap
ters to analysis of specific films in terms of his theoret
ical points. Chapters Six and Seven are an attempt to
analyze certain segments of Jacques Rozier's film Adieu
Philippine. Chapter Ten discusses the "mirror construction"
of Fellini's 8k. In the remaining seven chapters there are
very few references to actual films and these references
are not related to any of the major points which Metz is
trying to make. Some hypothetical film scenes which Metz
creates and attempts to use as illustrations are too sketchy
and of little value to him and to the reader.
His "analysis" of Adieu. Philippineis not really an
analysis but a process of choosing certain shots and scenes
from the film and affixing to them labels taken from his
incomplete classification system of the Grande Syntagmatique.
For example, he points to a series of shots of a boat which
are photographed and edited so as to capture the exhilara
tion of sailing, and he informs us that this constitutes a
"descriptive syntagma" (FL, p. 173). Beyond this act of
labelling segments of Adieu Philippine there is nothing in
________________ 53
what Metz says which in any way indicates how this film
communicates its ideas to us.
In Chapter Nine, a 7-page section hardly qualifying as
a "chapter" in a book of over 250 pages, there is nothing in
what Metz says about Fellini's 8 k which in any way supports
the theoretical points made in the rest of the text of Film
Language. In fact, this section is nothing more than a long
digression from the main task at hand and it is related to
cine-semiotics only in a very indirect and vague manner.
Metz's inability to provide evidence to support his
claims is particularly disturbing in Language and Cinema. The
very few attempts which are made in this book to relate to
actual films what he is talking about are very disappoint
ing. For example, his discussion of Griffith's intolerance
does not illuminate in any significant way his theoretical
points. Certainly, Metz's semiotics does not seem to add
anything new to our understanding of intolerance through
comments such as these:
The cinematic technique of the film founded on parallelism quite
obviously shapes the notion of intolerance, such that in the end
it becomes disengaged from the text, and floats toward a timeless
horizon. (LC, p. Ill)
In "The Imaginary Signifier," where the analyst and
the analysand fuse into one and Metz ends up psychoanalyzing
himself, the proof via manifesto technique is given a cloak
of respectability by an outpouring of psychoanalytic jargon
and the frequent invocation of the names of Sigmund Freud,
Jacques Lacan, and Melanie Klein.
5. 4
The fact that Metz appears to be operating under a
great deal of stress has been mentioned already. What needs
to be pointed out here is what I believe to be his greatest
source of apprehension which always looms in the horizon:
the reality of film.
Metz does not identify this source of fear as the
reality of film. He talks only in terms of "culture."
However, both his description of what is contained in
"culture" and the context in which he uses the term indicate
that he is referring to what I have called the reality of
film.
In Film Language this fear surfaces a number of times.
The most obvious manifestation of it occurs in his attempt
to determine what is the minimum unit of signification in
film. After preparing the reader for this observation by
a number of references to what lies beyond the film, he
says:
Thus, when it reaches the level of the "small" elements, the
semiology of the cinema encounters its limits, and its competence
is no longer certain. Whether one has desired it or not, one
suddenly finds oneself referred to the myriad winds of culture,
the confused murmurings of a thousand other utterances: the
symbolism of the human body, the language of objects, the system
of colors . . . or the voices of chiaroscuro . . . the sense of
clothing and dress, the eloquence of landscape. (FL, 142)
In Chapter Nine, he returns to the matter of the
minimal units of signification which in Film Language he saw
containing the "myriad winds of culture." After declaring
that the shot is not the minimal unit, he argues that the
identification of minimal units is not a "prerequisite for
_______ 5 . 5
the entire domain of cinematic semiotics" (LC, p. 193). A
few pages later he attempts to shut the door to the myriad
winds of culture which the shot would usher in by stating
that "the minimal unit is not a given in a text; it is a
tool of analysis. There are many types of minimal units as
there are types of analysis" (LC, p. 194).
Although in Film Language and in Language and Cinema
Metz does not consider the content of films and he deals
only with structural elements, in "The Imaginary Signifier"
he seems to open the door which the image of the film
represents. Standing at the threshold and confronting
Metz are the myriad winds of culture and the sciences
which study them: sociology, anthropology, history, etc.
However, as we shall see in Chapter V, the opening of the
door is deceptive. Metz is simply unable or unwilling to
face the reality of film no matter what approach he takes
in his study of how films communicate.
5B
FOOTNOTES
1 There are frequent references in Metz's works to the
inadequacy of the conventional approaches to film theory and
criticism. In Film Language see particularly the third
chapter: "The Cinema: Language or Language System?" (FL,
pp. 31-91) . In Language and Cinema see pages 9-15, and pages
26, 34, 81-83, 115, 124, 266-267, 271-273. In "The Imag
inary Signifier," see pages 21-25, 32-37, 39-40.
2See the works of Alan Lovell, Peter Wollen, Sam
Rohdie, Geoffrey Nowell-Smith, and James Roy MacBean cited
in the Bibliography.
3Saussure said in 1906 that "a science that studies
the life of signs within society is conceivable . . . I
shall call it semiology . . . since the science does not yet
exist,...no one can say what it would be; but it has a right
to existence, a place staked out in advance." See Ferdinand
de Saussure, Course in General Linguistics (London: Peter Owen,
1960), p. 16.
‘ 'Film Language, p . ’ 91’ . ■
5For an analysis of the serious shortcomings of Metz's
Grande Syntagmatique, :see Jack Daniel, "Metz's Grande Syntag-
matique: Summary and Critique," Film Form, 1, No. 1 (Spring,
1976), 78-90.
6For a good analysis of Metz's work see Brian Hender
son, "Metz: 'Essais I1 and Film Theory," pp. 12-33.
7The "filmophanic film" is, according to Metz, the
film as experienced by the spectator during projection.
5Jl
CHAPTER II
FROM MOVIES TO "CINE-TEXT"
The purpose of Metz's semiology is to create a model
which will enable him to explain how a film embodies meaning
and how this meaning is transmitted, signified, to an
audience.
According to the approach taken by Metz, the raw
material of the cinema out of which meaning arises is not
reality itself. The raw material is the various channels
of information which make up a film: image, graphics,
speech, music, and sound effects.
But even this raw material is not of interest to Metz.
His semiology is concerned with the logical structures
which operate on this material and which account for the
meaning transmitted to an audience.
A major problem with Metz's work is that, apparently,
he has not been able to decide exactly what is the nature
of the "film" which he makes the subject of his study. As
indicated in the previous chapter, in Language and Cinema
Metz claims that he is studying the "filmophanic film" as
perceived by the audience for the duration of its projec
tion.
_ .......... - 5 _ 8 .
The problem is that the film as a signifying discourse
(text), or as a linguistic object, and the filmophanic film
perceived by an audience during projection are not the same
thing.
Although Metz is well aware of the reality of film and
he identifies correctly the channels of information which
make up a film, in creating his analytic tools he does not
deal with the filmophanic film. Everything he has to say
about the semiotics of film in Film Language and in Language
and cinema is related to film as a linguistic object.
In "The Imaginary Signifier" Metz treats film as a
dream and he attempts to study it with the tools of a
vaguely described linguistic and psychoanalytic
"inspiration."
The purpose of this chapter is to point out that in
Film Language and Language and Cinema Metz regards films as
"cine-texts," and he proposes analytic tools which prevent
us from achieving contact with the reality of film.
This "reduction" of films to "cine-texts" is brought
about by the banishment of the author,1 by failing to take
into consideration the content of films, and by dealing only
with codes, sub-codes, structures, and systems which exist
only in Metz's mind as analytic tools of his own creation
unrelated to the reality of film.
According to the concept of the reality of film
utilized in this study, the author of a film is the
______________________________ 59
director who bases his work on the inspiration provided by
the screenplay.
The matter of film "authorship" is a fairly complicated
and controversial one. Something needs to be said about
it here before comments can be made about Metz's problems
with the banishment of the author in his works.
The concept of the director as the "author" of a film
is European (basically French) in origin. This concept
has become almost hopelessly entangled with the so-called
"auteur theory" which is not a theory but, at best, a
"policy of authors" attributed to some directors by certain
critics.
This "theory" has not found wide acceptance in the
United States and the title of "author" has not been
applied to American directors by the majority of American
critics because of the realities of film production in the
United States. With a few, very few, exceptions such as
that of an Orson Welles directing citizen Kane, American
directors working in this country did not have and continue
not to have an opportunity to assume the full measure of
control over the creation of their films to qualify as
"authors."
When European critics accord the title of "author"
( auteur) to certain American directors, they are doing so
apparently without concern over the fairness of singling
out these directors for such an honor while denying it to
__________________________________________________________________ j6_Q
others.
Judging from the various lists of "author"-directors
published in European film journals, it seems that the
criterion used in granting the title of "author" to a
director is that of liking his work. To be called an
"author" is almost equivalent to being regarded as a
"genius" by an admiring critic.
However, regardless of the measure of control which a
director has over the creation of a film, the fact is that
the director of a narrative film, at the very least, is the
leader of the film crew and the coordinator of the activi
ties of the creative people who make a film.
Actually, in most cases even in the United States, the
director contributes enough to the making of the film to
qualify as the person with the greatest amount of responsi
bility for interpreting the screenplay and making a film
based on it. So it is relatively safe to say that the
director is the "author" of. a narrative film even if one
does not give to the term "author" the meaning given to it
by the European writers on cinema.
The screenplay, of course, is the blueprint for the
making of a feature-length narrative film. One could
hardly name a film of this kind which was not based on a
screenplay, or, at least, on some kind of written outline.
The screenplay indicates the specific shots in which
the plot is advanced, as well as the characters appearing
in the film and the lines they speak. The screenplay also
contains numerous pieces of information which indicate how
the filmmaker should guide the audience's thoughts and
feelings.
Taking all this into consideration, it seems that the
study of how films communicate cannot exclude consideration
for the screenplay and the director who interprets this
screenplay in making a film.
Film theory and criticism of the conventional, non
structuralist persuasion has always dealt with the filmo
phanic film, and in doing so has always taken into consid
eration the director and the screenplay.
Although he claims that he is considering the
filmophanic film, Metz disregards both the screenplay and
the director.
This is so despite the fact that in Film Language there
are a number of references to film directors and, as
already indicated, despite the fact that he attempts to
analyze one film ( Adieu Philippine) , and make some comments
about another (Fellini's 8k). In both of these cases Metz
makes no comments about the director and the script which
in any way illuminate his cine-semiotics. Also, he makes
no comments in terms of cine-semiotics which illuminated
either the director's work or the role which the screenplay
plays in our understanding of how these two films
communicate.
62
In Language and cinema, a few references are made to film
iirectors. However, none of these references are related
in any vital way to the development of Metz1s theories
about the codes or the other linguistic aspects of his
analysis. The same applies to consideration for the screen
play as the blueprint for the making of a film. Meaningful
references to the screenplay are missing even from the last
part of Language and Cinema which deals with "cinematic
writing."
In eliminating the author from the scene Metz is going
along with the theory of authorial irrelevance which has
been considered beneficial to literary criticism because it
shifts the focus of attention away from the author and
toward something more tangible— the work itself.2
With the author of the film out of the way, Metz
makes an effort to convince us that the "filmic text" is
somehow a real object which has qualities which the analyst
can discover and study. This is something which Metz
claims throughout Film Language and Language and Cinema.
However, in his attempt to be "scientific," Metz does
not merely eliminate the author. He eliminates the content
of a film in terms of theme, plot, characters, philosophical
underpinnings, and social commentary. In Film Language and in.
Language and cinema Metz refuses to consider any of the above
elements of the content of films.
In Film Language a key element in this banishment of the
fill
content of films was his decision not to deal with the
"myriad winds of culture" which the image represents.
With the author and the content of films eliminated,
what is left for Metz to study? The answer is that he
invents something which he claims will help us understand
how films communicate, and then he proceeds to study it—
a product of his mind and not anything related to the '
reality of film.
The procedure of inventing the object of his study is
introduced in Film Language and perfected in Language and Cinema.
In Film Language, after talking in a condescending manner
about previous efforts in the history of film scholarship
to deal with what he calls the "natural object" (the film in
terms of the reality of film), he announces:
The grand moment, which one has been waiting for and thinking
about since the beginning, is the syntagmatic moment. One
reassembles a duplicate of the original object, a duplicate
which is perfectly grasped by the mind, since it is a pure
product of the mind. It is the intelligibility of the object
that is itself made into an object.
And one never takes into account that the natural object has
been used as a model. On the contrary, the assembled object
is taken as the model object— and let the natural object
keep still I (FL, p. 36)3
Least we did not appreciate the difference between the
natural object and the reconstructed duplicate of it in the
mind of the analyst, Metz continues:
The goal of the reconstruction, as Roland Barthes emphasizes,
is not to reproduce reality: the reconstruction is not a
reproduction, it does not attempt to imitate the concrete aspect
of the original object; it is neither poiesis nor pseudo-physis,
but a simulation, a product of techne. That is to say:‘ the
result of a manipulation. As the structural skeleton of the
____________________________________________ . ■ 64
object made into a second object, it remains a kind of
prosthesis. (FL, p. 36)
Operating under the impression that the natural object
can be silenced, Metz devotes the rest of Film Language to
dealing with the product of his techne by inventing for it
a "content" which he can manipulate and study.
In order to invent this content, and in an effort to
locate it somewhere, Metz attempts to make certain distinc
tions. As already indicated, he attempts to distinguish
between denotation and connotation. He also attempts to
distinguish between the syntagmatic and paradigmatic
dimensions. Both of these efforts, as already indicated,
eventually fail because of the circularity involved and
because of his inability to overcome this circularity.
But^ while denotation and connotation, as well as the
syntagmatic and paradigmatic dimensions, are being kept
distinct artificially for the sake of the argument he
presents, Metz weaves through these dimensions the concept
of diegesis. Since he is using a term/concept which in
ancient Greek thought has an established meaning which he
cannot completely disregard, he is forced to admit that
diegesis means "narration." Then, he admits that the term
was introduced to the field of cinema by Etienne Souriau
and that it designates
. . . the sum of film's denotation: the narration itself, but
also the fictional space and time dimensions implied in and by
the narrative, and consequently the characters, the landscapes,
the events, and other narrative elements, in so far as they
are considered in their denoted aspect. (FL, p. 98)
______________________________ 6_5
Then, Metz proceeds to use the concept of diegesis in
connection with the product of his techne which has a content
which is unrelated to narration, characters, landscapes,
events, or anything else even remotely related to such
elements. What Metz's diegesis is called upon to deal with
are the structural relationships of elements which he calls
the "syntagmatic components of films" (FL, p. 121) and which
he describes in his Grande Syntagmatique.
In other words, Metz's diegesis does not refer to
anything related to the reality of film. His diegesisis also
a product of his techne. It is the organizing dimension
accounting for a content which consists of the relationships
between elements contained in the mental duplicate of the
natural object! h
In Language and Cinema the journey into the phantom world
of the cine-text is made in terms of elements of "narrativ-
ity" which Metz calls "cinematic language system," "code,"
"sub-code," "system," "message," "text," and "textual
system."
A detailed examination of these elements inevitably
will lead us to an analysis of Metz's work in terms of its
place in the field of linguistics. Of course, this is not
the purpose of this study. For this reason, only the
abstract quality of these concepts and the manner in which
they lead us away from the reality of film will be discussed
here. Since Metz's thought is plagued by great circularity,
66
a considerable amount of overlapping will be noted between
these elements.
Before dealing with these elements it is appropriate
to return to the role of the cine-semiotician as Metz sees
it. In the beginning of Language and cinema, he claims that:
A semiotic analysis is thus closely associated with the aesthetics
of film. (LC, p. 17)
However, nothing that Metz does in Language and Cinema is
related to the aesthetics of film as we know the aesthetics
of film. As was the case with his claim that he is studying
the filmophanic film when he is not, Metz again indicates
that he will do one thing and then proceeds to do something
else.
This something else follows the pattern of what he
discussed in Film Language. It is an examination of the
duplicate of the natural object which (duplicate) is
"perfectly" grasped in Metz's mind. In Language and Cinema
the role of the semiotician is seen as follows:
The semiotician follows a path which leads in the opposite
direction from that of the cineast. The cineast starts with
diverse . . . systems in order to arrive at a demonstrable:.text.",
The semiotician focuses on the text in order to reconstitute (and
always explicitly) the systems which are implied by it, which are
invisible in it, and which are discoverable in it alone. What the
cineast constructs is the text, while the analyst constructs the
system. (LC, p. 74)
Halfway through Language and Cinema he says:
We would thus summarize the tasks of the semiotics of the filmic
fact as follows: to analyze film texts in order to discover
either textual systems, cinematic codes, or sub-codes. (LC,
p. 150)
A few pages later he admits that the search for these
_ 62
elements in the filmic fact is notrelated to the reality of
film:
At any particular moment, the analysis has nothing specifically
to do with the film: behind it, in any case, are found only
"films," and after it, thanks to it, "the film" begins to
exist; not as a fourth term, but as part of the third (codes).
When one claims to "study the film," what is meant is that, in
a movement of reflection oriented toward cinematic codes and
sub-codes, one forces oneself to organize a certain number of
traits (which are themselves more or less general, and lacking
a separate existence outside of the semiotic discourse) which
present, in addition, the remarkable peculiarity of rejecting
any immediate attribution to the code itself. (LC, p. 157)
Why this kind of analysis has nothing specifically to
do with the film hopefully becomes apparent as we examine
the elements of analysis which Metz brings to his enter
prise .
Cinematic Language System
As Metz indicates in the Conclusion to Language and
Cinema, the two great tasks involved in the study of the
cinema is the analysis of the cinematic language system and
the analysis of what he calls "cinematic writing" (LC,
p. 286) . The task of Language and Cinema is, essentially, to
study the cinematic language system which he defines as "the
combination of all the particular and general cinematic
codes . . ." (LC, p. 69).
After Metz has an opportunity to reveal certain fea
tures of the codes, he returns for a refinement of his
definition of the system. It is at this point that, as
mentioned in Chapter I, he becomes very lyrical and he talks
of the system being a process of displacing, deforming, and
- _6-8
contaminating codes within the text (LC, p. 103).
The interesting thing about this cinematic language
system turns out to be that it does not exist yet! The
reason for this is that Metz does not produce the codes
which must be contained in this system.
In the Conclusion to Language and Cinema, after he has
spent approximately two hundred (20 0) pages describing in
minute detail the intricate choreography of the activities
of these codes, he states:
. . . the reader will perhaps be surprised at not having found
here an explicit enumeration of specific codes. This omission
was intentional. First, because to study the status of a
phenomenon (to define it intentionally) and to deploy its
entire content (to define it extentionally) are two distinct
steps and that, when the "phenomenon" is rather a constructed
notion (as is the case for the cinematic language system),
the detailed exposition of distinctiveness is what should take
pride of place. Next, because cinematic studies are not yet
developed enough; one is not able to seriously advance an
explicit list of all the codes and sub-codes. It is, of
course, possible, even desirable, to proceed already to a
preliminary listing, to propose a beginning of an enumeration,
even if incomplete and still approximate. But even this is a
task which, in order to be useful, demands specifications which
would require a separate book. (LC, p. 286)
Cinematic Codes and Sub-Codes
Although he is not able to produce a list of codes and
sub-codes, Metz seems to know a lot about their characteris
tics and their behavior. This is not very surprising when
one considers Metz's definition for the code and sub-code,
his explanation of how they come to exist, and the function
which he claims they have in his cine-semiotics.
In defining what a "code" is, Metz returns one more
69
time to a description of the nature of semiotic analysis:
The semiotic analysis does not create the film, which it finds
already made by the cineast. On the other hand, we can say
that, in a certain manner, the analysis "creates" the codes of
the cinema; it should elucidate them, make them explicit,
establish them as objects, while in nature they remain buried
in films, which alone are objects which exist prior to the
analysis. It should, if not invent them, at least discover
them (in the full sense of the term). It should "construct"
them, which is in one sense to create them.
The codes of the cinema are not things which one can immediately
discern somewhere, for the cinematic does not exist independently;
only an analysis can separate it out. It thus consists only of
what the analyst puts in it. The analyst's task, in sum, is to
uncover certain filmic facts and to construct the cinematic codes
by means of these facts. (LC, p. 49)
Although Metz states repeatedly that the codes are
analytic tools which are the product of the analyst and they
do not directly express empirical reality (LC, pp. 54, 78,
7 9, 141, 154), the suggestion that these codes are created
by material provided by "certain cinematic facts" (LC,
p. 49) and by "materials furnished by the message" (LC,
p. 54), makes one hope that these codes, somehow, are
related to the reality of film.
However, nowhere in Language and Cinema does Metz explain
what "materials" the film offers to the analyst so that he
can devise from them the codes and sub-codes. Regardless of
the contacts with the reality of film that Metz alludes to,
his codes and sub-codes"(which are codes appearing only in
certain types of films) remain very abstract notions which
exist only in the mind of the analyst.
Of course, this means that all of the activities of the
70
codes which Metz describes in endless detail are also in the
mind of the analyst and not in the film.
This does not prevent Metz from often describing these
activities as if they were occurring within the film
independently of the consciousness of the analyst. For
example, in talking about the "textual systems," Metz
claims that
It is not a question, then, at present, of examining individually
each of the codes that the film contains in order to see if it is
or is not of cinematic origin . . . What we are considering now
are not the partial systems integrated by the film, but the
activity of integration (or disintegration)— the process of
composition or "writing"— by which the film, relying on all of
these codes, modifies them, combines them, plays them one against
the other, eventually arriving at its own individual system, its
ultimate (or first?) principle of unification and intelligibility.
(LC, p. 100)
Metz resorts so often to this anthropomorphism of the
film, and he engages in so many practices which create the
impression that we are dealing with something concrete and
"real" when actually we are not, that one eventually comes
to suspect that Metz is simply cheating.
The suggestion that Metz may be cheating has been made
by at least one writer in connection with the concept of the
code. Professor Gilbert Harman points out that Metz (and
fellow semiologist Peter Wollen) cheats when he uses the
term "code" interchangeably to mean "standard" or "cipher."5
The only hope raised by Metz that there can be some
codes which are related to the reality of film comes in the
form of what he calls "extra-cinematic codes.” These are
codes which appear in films as well as in other arts. They
1U
consist of what Metz calls "filmed-objects-of-a-fixed-
psychoanalytic-value," and of "themes" such as those found
in the Western.
In discussing briefly these "extra-cinematic" codes,
Metz is forced to admit that they reflect society, culture,
and human values. However, he shows very clearly that he
has no intention of allowing these contacts with the reality
of film to interfere with his work. He excludes these
"extra-cinematic" codes from his study by associating them
with the "filmic" and not with the "cinematic" which is the
focus of his study. It should be remembered that Metz has
not been able to make a clear and convincing distinction
between the "filmic" and the "cinematic."
Metz justifies the exclusion of the "extra-cinematic"
codes from his study out of fear of descending to the level
of "cinematic journalism":
As for the extra-cinematic codes, we have mentioned them only in
order to recall that they play an important role in films. But
a study of these codes could not be made the goal of the analyst
of the cinema, nor of the analyst of films. Neither is it a
question of a unified study, i.e., a "discipline." The extra-
cinematic material found in films is as extensive and varied as
social life itself (from which it directly stems), and its
analysis relies upon quite diverse skills and a large number of
pre-existing disciplines. In the division of labor customary
today, one always has the tendency to confound the cinematic and
the filmic, and to expect from the analyst of the cinema a science
which covers all aspects of all films. It is not understood that
this would be an almost universal undertaking, because films may
be about anything. The immoderation of the expectation only
encourages cinematic journalism. (LC, p. 150)
System
Another element in Metz's reduction of films into
221
linguistic objects is the concept of the "system." He says:
Nevertheless, one should not forget . . . that what the analysis
is trying to bring to light--and which is no longer a code— is
still a system. The goal toward which all descriptive work
strives is not the film as a real discourse (a series of images,
sounds, and words arranged in a certain order, an object that
may be attested) . . . What a description hopes to establish is,
rather, the system which organizes this realization: the
structure of the text, and not the text itself. The system is
nowhere clearly visible in the actual unwinding of the film: a
system, as such, is never directly attested. (LC, p. 73)
A few pages later he adds:
What characterizes the systemic (the non-textual) is its nature
as a residual object constructed by the analyst. The system has
no physical existence; it is nothing more than a logic, a prin
ciple of coherence. It is the intelligibility of the text, that
must be presupposed if the text is to be comprehensible. (LC,
pp. 75-76)
With expressions such as "trying to bring to light,"
and "the intelligibility of the text that must be presup
posed," Metz hints that the system may really exist in a
film and merely waits to be discovered by the analyst.
While hinting that the system may exist independently of the
analyst, he indicates that it is an analytic tool created by
the analyst. Then, when it suits him to do so, he comes
back to the idea that the system is real and merely awaits
to be brought to the surface by the analyst:
It is characteristic of all cultural facts that they function
according to systems but are not felt or experienced as such.
It is distinctive of such systems that they remain unfixed,
purely implicit, submerged in history as well as individual
variations, etc.— but it is also distinctive that they appear
more and more clearly as systems as our analysis of them
gradually advances. (LC, p. 86)
At one point Metz attempts a distinction between a
system and a "singular system" and he gives the impression
73
that the "singular system" could be^related to a certain
dimension of the reality of film. However, he immediately
takes steps to prevent this from happening. He says:
However, we should be cautious about saying that singular systems
are "real," for a system is never real (only a text is). If the
singular systems seem to be real, it is because they are singular,
and thus located in a unique and "concrete" place. But this place
is concrete only to the extent that it is a text. The corre
sponding system, for its part, is nowhere made explicit, even in
this place . . . Thus the system is not "real," which is why it
is a system (a fabrication of the analyst-like codes). (LC,
pp. 78-79)
Message and Text
"Message" and "text" are two more elements in the
reduction of films to "cine-texts." In Language and Cinema
Metz says:
If the film— the message— is a "concrete" object, it is because
its borders coincide with those of a discourse which has been
effectively sustained, a unit which precedes the intervention
of the analyst. (LC, p. 24)
However, Metz does not regard the "film" and the
"message" as being one and the same thing. Later on he says
that "message" is merely a part of a "terminological pair"
(text/message) which stands opposite to the terminological
pair of "system/code."
Then, he shifts his point of view and he talks in terms
of the "message of the code" (LC, p. 7 5).
The matter becomes further complicated when he indi
cates that a "message" is actually the "text" of a "singular
system." Then he shifts again his point of view and points
out that
2M
Films . . . are not messages but texts, for each of them
contains several codes and many messages. (LC, p. 89)
Finally, he admits that:
Since a message is the text of a single system and since in a
domain with a single semiotic dimension all texts are studied
in relation to but one of their systems, message and text may
in practice (and provisionally) become synonymous. (LC, p. 145)
Earlier, in Language and Cinema, he had indicated about
the term "text" that:
It is evident that this term, for us, does not apply solely to the
verbal element of the film . . . But we will take it here in
Hjelmslev’s sense, that is to name any semiotic expression . . .
whether it is linguistic, non-linguistic, or a combination of
both ... A series of images is also a text, as is a symphony, a
sequence of sound effects, or a series including images, sound
effects, and music, etc., together. This point will be further
discussed, in regard to the film, in Chapter 8.5. (LC, p. 87)
However, neither in Chapter 8.5, or anywhere else, does
Metz deal with the concept of the "text" in a way which
indicates that it contains any aspect of the reality of film
such as the image or the soundtrack of films.
As for the term/concept "message," since he has used
the term "film" in so many different ways ranging from the
filmophanic film to a linguistic object, one becomes dis
oriented and does not know what Metz means by "message."
System/Code, Text/Message
The fact that Metz relates the concepts of system,
code, text, and message has already been indicated. For the
sake of appreciating the degree of abstraction which Metz
utilizes and the distance from the reality of film from
which he operates, I wish to quote him as he "plays" these
concept/terms against one another. Compressing his text
_____________________________________ . ___________ 7-5
to eliminate extraneous material emphasizes what the reader
is confronted with as he tries to comprehend what Metz is
saying:
Code always differs from message in that one is code and the
other message. (LC, p. 59)
Thus the text, as text, is distinct from any system, and even
from the unique system of which it is only text. And the
system, even if unique, is distinct from any text, including
its own. (LC, p. 73)
A code is a system which is valid for several texts (and these
texts thus become messages); a message is a text which is not
the only one to manifest a given system (and this system thus
becomes a code). The system which is not a code (a singular
system) has only one text; the text which is not a message (a
singular text) is the only one to manifest its system. (LC,
p. 76)
The problem of minimal units of signification in the
cinema is reintroduced by Metz while he is discussing these
concepts and the "play" which he fabricates for them.
In a move which seems to isolate his approach even
further from the reality of film, Metz denies that there are
minimal units of signification in the cinema which are not
related to the code, and he denies that there is such a
thing as a cinematic sign.
Concerning the minimal units he says:
No minimal unit (for specific systems of articulation) exists
in the cinema; such a unit exists only in each cinematic code.
(LC, p. 185)
The minimal units contained within the code have never
been revealed by Metz. Therefore, his is a semiology which
has never dealt with the problem of minimal units of signif
ication.
The other unusual feature of Metz's semiology is that
__________________________________________________________ IS
he rejects the notion of a cinematic sign:
There is no cinematic sign. This notion, like that of "pictorial
signs," "musical signs," etc., stems from a naive classification
which proceeds according to material units (languages) and not by
units of a logical order (codes). (LC, p. 194)
A few pages later, however, he feels the need to
clarify the statement about the absence of cinematic signs
and to separate this issue from his examination of the
nature of the codes. He says:
The notion of the sign . . . has no right to play a more important
role in cinematic and filmic semiotics than in other areas of
contemporary semiotics and linguistics. Without rejecting the
notion of the sign as such, it must be realized that it only
represents, today, one tool of research, and that it no longer
enjoys the privileged and central status which it had with
Saussure or Peirce . . . A system of signification is not only a
system of signs; units larger or smaller than the sign play a
considerable role in it. (LC, p. 207)
This reluctance to deal with the cinematic sign and the
downgrading of its importance in the study of signification
in the cinema seems to bfe important in terms of my concern
for the reality of film. All of the approaches to the study
of cinematic signs which I am aware of cannot help but deal
with the image and through the image with the reality of
film. Metz has no desire to deal with something which will
involve the image and the reality it embodies. He prefers
to deal with the abstract analytic tools of his techne. Metz
admits this quite clearly when he summarizes what he has to
say about the "specific/non-specific" aspects of his codes:
In sum, the position which is adopted here includes two elements,
not one: (1) the specificity which interests semiotics is the
specificity of codes, hot the "crude" specificity of physical
signifiers; (2) the specificity of specific codes nevertheless
refers to certain features of the material of expression. (LC,
p. 219)
____________________________ 77
The crudity of the physical signifiers is rejected, but
Metz never explains clearly what is the nature of those
"features" of the material of expression. The material of
expression, the image, graphics, speech, music, and sound
effects, are merely identified and enumerated by Metz. He
never deals with them in his work. As indicated a number of
times, he is particularly reluctant to deal with the image.
So, the specificity of his codes remains a very vague
notion. The vagueness of the concept of the code and the
unreality of that to which Metz reduces movie films, the
cine-text, becomes apparent when one tries to distinguish
between his analytic tools and the object of analysis.
The object of analysis is an analytic tool, and the
analytic tool is the object of analysis. Metz occupies most
of his time analyzing his analytic tools and not film. The
natural object, the movie film, has been forced by Metz to
keep still.
78
FOOTNOTES
^his expression is borrowed from E. D. Hirsch who uses
it in defending the view that a text means what its author
meant it to say/ or what a reader finds it to mean; what the
linguistic signs themselves supposedly indicate that it
means: See E. D. Hirsch, validity in interpretation (New Haven:
Yale University Press, 1967).
2Hirsch, Validity in Interpretation, p. 2.
3My italics.
‘ ‘For an interesting analysis of the shortcomings of the
structuralist method and of that feature of structuralist
criticism which transforms form into content, see Fredric
Jameson, The Prison-House of Language.
5Gilbert Harman, "Semiotics and the Cinema: Metz and
Wollen," Quarterly Review of Film Studies, 2, No. 1 (February,
1977), 15-24.
79
CHAPTER III
METZ AND THE FILM AUDIENCE
The purpose of this chapter is to look into the method
Of analysis proposed by Metz in Film Language and Language and
cinema in terms of those aspects of the reality of film which
involve the film audience. The material in this chapter is
divided in two parts: (1) Metz and the Limitations of the
Phenomenological-Structural Linguistic Methods, and
(2) Metz’s Approach and the Film as Art.
Metz and the Limitations of the
Phenomenological-Structural Linguistic Methods
In his attempt to isolate the "filmic object" so that
he can study it scientifically, Metz banishes not only the
author but the film audience as well. This banishment is
unavoidable. It is imposed by the phenomenological and
structural linguistic approach which Metz takes.
The phenomenological approach is nearly always purely
theoretical and requires a detached, contemplative, and
intuitive observation of the object of study. This object
is subjected to a "reduction" which consists of trying to
put aside all subjectivity on the part of the observer, all
hypotheses and proofs from other sources, and all tradition.
............................ an
It seeks to isolate the "given object," the phenomenon.
Before the intellectual observation of the phenomenon
can begin, the object must be further "reduced" by disre
garding its existence and focusing on its essence.1
Phenomenological observation is analytical and descrip
tive. The analysis, an exegesis or hermeneutics, deals
nearly always with structures because it is claimed that an
important element of the essence of the phenomenon is its
structure.
Phenomenology excludes most utilitarian considerations.
The analyst contemplates the phenomenon for the purpose of
attaining knowledge. The analyst's thought is focused
exclusively on the object of his inquiry excluding every
thing that comes from himself in the form of attitudes,
feelings, emotions, desires, and biases of all sorts. Of
course, the objectivity of phenomenology is only an ideal.
The phenomenological aspects of Metz's approach create
some serious problems as far as the role of the audience is
concerned.
As we have seen, Metz disregards the natural object,
the film, and he contemplates a pure product of his mind—
"the intelligibility of the object that is itself made into
an object" (FL, p. 36).
Although in the natural sciences the term "phenomenon"
usually refers to something which can be observed directly
through the senses (or indirectly through instruments), in
81
phenomenology it is not necessary for the object to be
observed in this manner. It is enough for the phenomenon
to be imagined.
In Metz's writings, especially in Language and Cinema, the
structural elements making up the phenomenon and the rela
tionships between these elements are clearly imagined. Metz
repeatedly admits that his codes, sub-codes, and the various
systemic relationships are creations of the analyst and not
properties of the object of his study.
An audience, however, is not relating to an abstract
and imagined mental duplicate of the film. In a motion
picture theatre the audience is experiencing very vividly
through the senses of sight and hearing a "natural object"
which is vibrating with a very strong impression of reality.
Accepting Metz's contention that his phenomenological
relationship with the product of his techne explains how
films communicate to an audience becomes a matter of faith
and not scientific documentations. Metz and the film
audience are relating to two different things.
Metz supposedly studies only what is "out there" with
out allowing his emotions to interfere with his efforts,
and without incorporating anything in his analytic scheme
which can account for the role of emotion in film communi
cation.
On the other hand, the attitude of an audience cannot
possibly be phenomenologically purged of emotions and
£21
feelings. An audience does not respond to a narrative film
only intellectually. Emotions play a very important role in
an audience's relationship with a film, and it does not seem
possible to study how films communicate without taking into
account the emotional aspects of an audience's response.
Phenomenology serves as the overall philosophical
orientation of Metz's approach. The actual instrument of
analysis is structural linguistics. It is the nature of
this instrument that it excludes any consideration for the
film audience.
A major conceptual tool of Saussurean linguistics is
that of the sign and of the "internal" relationship between
the signifier and the signified.
Focusing on the sign, this type of linguistic approach
excludes any consideration for the object of reference in
real life. Then, the signified is isolated and usually put
aside. The analyst is then free to relate to the signifier
and to formulate the syntagmatic and paradigmatic relation
ships which he sees as operating in connection with the
signifiers.
Metz's work consists of theorizing about the nature of
the signifier and of the functioning of the various analytic
features which he invents for it. The relationship between
the film audience and the signifier is not examined. Metz's
analytic study takes into account only the relationship
between the analyst and the signifier. As for the signified
____________________________________________________________________________aa
in film, it is excluded from consideration as a matter of
principle. What is signified in film, to whom, in what
manner, and under what circumstances is not examined.
Related to the concepts of the signifier and the sig
nified are those of denotation and connotation. Structural
linguistics places emphasis on denotation and so does Metz.
However, because of the nature of the film image, Metz
is unable to distinguish between denotation and connotation.
As we have seen, he admits that there is practically no way
of keeping them apart in terms of the actual film-viewing
experience. This does not prevent him from claiming that
his interest lies in denotation and from proceeding to study
what he considers to be denotation.
The denotation Metz deals with is not related to
anything which is actually in the film. The denotation of
his analysis turns out to be the relationships he creates
for the analytic elements he invents for the mental dupli
cate of the natural object.
In this kind of scheme whatever Metz says about denota
tion in film is totally unrelated to what an audience comes
in contact with as it views a motion picture.
As for connotation, it is not examined by Metz in terms
of the analyst or the film audience.
Language changes over time and for this reason it is
not a stable object of "scientific" analysis. The ingenuity
of Saussurean linguistics lies in that it distinguishes
___________ BA
between the synchronic and the diachronic dimensions of
language and then focuses its attention on the synchronic.
This solution is considered a "breakthrough" in the study
of language because it allows language to be studied in a
"static" state which, supposedly, yields valuable observa
tions. At the same time the synchronic approach creates
problems.
A very serious shortcoming of the synchronic approach
is that the analytic methods it creates and utilizes may be
appropriate (and the results they produce may be meaningful)
for a given state in whatever is being analyzed, but not for
all stages in the development of the object of study.
Another serious shortcoming of the synchronic approach,
at least as practiced in most fields, is that it may yield
seemingly useful information about individual segments of
the object of study, but hot meaningful in terms of our
understanding of the subject as a whole.
Also, most synchronic studies tend to fragment the
object of study without attempting a synthesis. We are left
contemplating heaps of fragments which add little or nothing
to our understanding of the subject.
Metz's approach is synchronic. There is nothing in his
method of analysis which accounts for the fact that films
consist of expressive elements which are presented to an
audience and experienced by that audience diachronically.
There is nothing in any specific analytic procedure,
85
or in his overall method, which identifies the point in the
history of the narrative film to which his approach is
applicable. He allows us to think that his synchronic
approach is valid for all periods in the development of the
narrative film.
In interviews and while defending his work, Metz has
made a number of admissions which indicate that his Grande
Syntagmatique is an appropriate analytic tool only for a
specific period in the history of motion pictures. He has
indicated that this period may be that of from 1933/35 until
about 1955.2
As far as Language and Cinema is concerned, he has never
indicated that the procedures he describes in it are valid
only for a specific point in the development of the
narrative film. We must assume that these analytic proced
ures are valid for all periods in the history of film.
However, since the role of the audience is completely
eliminated in his approach, it is practically impossible to
verify that his procedures are indeed applicable to all
periods in the development of the narrative film.
The diachronic approach, which could have taken into
account the role of the spectator in the development of the
narrative patterns of film, is rejected by Metz because
through such an approach "filmic orderings are codified
primarily for purposes of connotation rather than denota
tion" (FL, p. 118). As we have seen, he is not interested
__________________________________________________________________E6
in studying connotation.
The limitations of the structural linguistic approach
which Metz assumes also become evident in his refusal to
deal with the audience in terms of the systems through
which an audience maintains a contact with the reality of
film.
In Language and Cinema, Metz identifies the main types of
systems through which the audience experiences the total
message of a film. These systems are an elaboration of
the systems he mentions in Film Language. They are:
1. Visual and auditory perception.
2. The capacity of the audience to recognize, identify, and
enumerate the visual and auditory objects which appear in film.
3. The ensemble of symbolisms and connotations which are
associated with objects or relations between objects within
the film.
4. The ensemble of principle narrative structures of a given
civilization which are present in a film.
5. The ensemble of properly cinematic systems which organize
the discourse of the elements presented to the audience by means
-of the preceding four elements. (LC, pp. 33-34)
Metz admits that elements (1) through (4) are cultur-'
ally acquired or influenced. Element (5) involves an
organization of elements (1) through (4). In this fashion
the dimension of cultural influence is very important in
this organization of elements (1) through (4).
Since Language and Cinema supposedly represents Metz's
major contribution to the development of the semiotics of
film, one expects him to deal with the above systems. How
ever, this does not occur. He simply fails to deal with
elements (1) through (4) . Language and cinema attempts to
S L 1
deal only with element (5), without any reference to the
natural object, cultural influence, or the audience. The
systems he describes in Language and Cinema are clearly
identified as creations of the analyst, and they function
according to rules borrowed from structural linguistic
analysis. These systems merely describe how the analyst
relates to the linguistic object/phenomenon.
In Metz’s work there are frequent references to
signification. In structural linguistics signification is
viewed as a process. It is the act of the signifier and the
signified combining to produce the sign. This process is
not studied in order to discover what sort of meaning is
conveyed through the act of semiosis. The object of the
study is to determine how the union of the signifier with
the signified produces meaning.
However, the problem is that structural linguistics has
not provided us with a clear explanation of how semiosis is
actually achieved. Exactly what occurs during the fusion
of the signifier and the signified has never been deter
mined. Fredric Jameson points out that
The emphasis on signification takes the form of a mystery, the
mystery of the incarnation of meaning in language, and as such
its study is a kind of meditation. This is what accounts for
the hermetic quality of the writers who deal with it. The sense
of the esoteric may be understood, in Barthes' sense, precisely
as a sign, as a way of signifying ritual and the presence of
mystery, of understanding through the very temporal unfolding
of the ritual language the sacred quality of the object itself.3
To the basic inability of structural linguistics to
explain exactly how signification is achieved, Metz adds the
__&8
problems created by the fact that he does not take into
account the signified. He is trying to demonstrate a
process of signification which involves only the signifier.
In general terms, according to this scheme, the follow
ing factors are involved in signification:
1. In the cinema the material of expression out of
which meaning arises is not reality itself but the channels
of information: images, graphic elements, recorded speech,
music, and sound effects.
2. Signification is the process of conveying messages
to the audience through codes.
3. Codes are the logical relationships, the rules,
which make possible the messages conveyed to the audience.
The codes organize the material of expression to generate
the messages or the meaning of the film.
4. Codes exist in systems in texts. The text can be a
single film or a number of films.
5. The orchestration of the codes in systems is what
causes the codes to "release" their messages (meaning) to an
audience.
In terms of the place of the audience in Metz's work,
the following comments can be made about his views on
signification: When Metz talks about signification, he does
not refer to the transmission of a so-called eidetic meaning
— what a sign means. To achieve this Metz would have to
take into consideration the signified and connotation. He
__________________________________________________ £3
is referring to a purely operational meaning of the sign
which consists of knowing how the sign can be used in a
film. This operational meaning depends on the codes which
are creations of the analyst. The organization of the codes
into systems is also the creation of the analyst.
In Language and Cinema, Metz admits that
Each film has its own structure, which is an organized whole, a
fabric in which everything fits together; in short, a system. But
this system is valid for one'film . . . To the extent that films
are considered as unique totalities, each contains within itself
a system which is as unique as the film itself. (LC, p. 63)
So, when Metz is talking about signification he is
actually referring to the unique manner in which the analyst
manipulates the analytic features which he has created for
the purpose of relating to a film.
The audience can come into all this only if it were to
assume the role of an analyst of Metz's persuasion and were
to perform on a film the abstract logical manipulations
suggested by Metz. It seems very doubtful that a human
being seated in a movie theatre and experiencing a narrative
film could be engaged in such an activity.
Instead of telling us how a film communicates meanings
to an audience, Metz's approach describes the "interior
monologue" of Metz communicating with himself as he relates
to films. It is this procedure which he presents to us as a
new, revolutionary, and "scientific" theory which will
finally enable us to understand how films communicate.
Time after time in Film Language and in Language and Cinema,
90
Metz points out that his is a purely descriptive enterprise.
There are no normative aspects to his work, he claims, and
no attempt to evaluate what is being analyzed. This de
scription, in typical structural-linguistic fashion,
involves an infinite regression into dimensions which are
progressively more and more abstract. Although Metz indi
cates that some day this descriptive enterprise will come to
an end and a synthesis of sorts will take place, he does not
want to predict when this synthesis will become possible or
what form it will take.
Metz's claim that there are no evaluative and normative
aspects to his work does not coincide with the facts. There
are both normative and evaluative aspects, and he owes some
of them to the nature of the structural-linguistic approach.
An important value judgment built into his work is the
emphasis on the synchronic. Logically there is no valid
reason why the synchronic approach yields more valuable
information than the diachronic. Saussure’s claim concern
ing the supremacy of the synchronic approach represents a
value judgment on his part. Actually, the synchronic is
merely another method of approaching and dissecting the
object of study.
The emphasis on denotation at the expense of connota
tion can be viewed both as an evaluative and a normative
feature of Metz's work which he owes■to the structural
linguistic method.
But the most prominent evaluative aspect of Metz's work
__________________________________________ ai
which he owes to structural linguistics and which is related
to the role of the audience, seems to be what he calls the
"principle of relevance" (LC, pp. 9-22, 71-73, 96-97, 122-
124, 143-144). This is the principle which the analyst
follows in making important decisions involving the materi
als of expression and the selection and manipulation of the
analytic elements which the analyst creates. Since Metz
claims that the work of the film viewer parallels that of
the analyst, this appears to be a very important principle
which lies at the heart of the question of how films com
municate. Despite the apparent importance of this princi
ple, Metz does not indicate what is its exact nature. He
merely lets us know that it exists, and he moves on to other
matters.
However, if the analyst can choose whatever he consid
ers "relevant" according to a principle of relevance which
has never been defined, so can the individual viewer of a
film. In this manner Metz introduces a very disturbing
impressionistic element which casts doubts about the valid
ity of his claim that his is a scientific approach to film.
Metz's Approach and the Film As Art
Art is a social phenomenon. It is a process which
involves the artist's experience of reality, the expression
Df this experience into an artistic medium, and the evoca
tion of an experience in whose who come into contact with the;
work of art.
92
The audience's reaction to a work of art becomes part
of the artist's experience of reality and it may influence
the artist as he creates other works of art.
The audience's reaction to works of film art very much
influences both the audience and the financial success of
these works. Through its reaction to the produced and
exhibited films the audience has exerted a tremendous
\
influence in the choice of thematic material, the shaping of
narrative patterns, the method of utilization of expressive
means, and even the development of the technological
aspects of film.
As it has been indicated already, Metz repeatedly
admits in Film Language and in Language and Cinema that cinema is
an art. But, when it comes to developing his analytic
method which supposedly deals with how films communicate,
he does not take into account that film is an art.
Also, Metz does not take into account that film is a
social phenomenon. This comes about by completely disre
garding the film artist (the "author"), by subjecting to
analysis an abstract mental duplicate of the natural object
(the film), and by substituting the analyst for the audience
in dealing with the object of analysis.
What remains to be seen is whether, or not the analyst
deals with the mental duplicate of film in a way which
accounts for the way an audience deals with the "real" film
which is projected in a motion picture theatre.
____________________________________________________________________________S3
Reflected in Metz's approach is a structuralist's
attachment to a cryptographic nature of reality which calls
for an emphasis on decoding and decipherement. Such an
approach requires that the analyst relates to the object of
analysis in terms of logic rather than emotions. The
phenomenological aspects of Metz's approach also call for a
relationship in terms of logic. As a result there is no
feature, aspect, or dimension of Metz's codes, sub-codes,
and systemic relationships which accounts for an emotional
involvement on the part of the analyst.
However, emotional involvement is a very important
feature of the relationship between an audience and a work
of art. In the narrative fiction film a great deal of the
information conveyed to an audience on the level of emo
tional involvement is through characters performing the
actions of the plot of the film. Consciously and uncon
sciously a film is made so that the audience identifies with
one or more of the characters of the story. Metz makes no
effort whatsoever in Film Language and Language and Cinema to
account for audience identification as a factor in film
communication.
In the arts, especially the non-representational ones,
the question of subject matter is a very troublesome one
resulting in nearly endless and largely inconclusive debates.
In the art of film perhaps the situation is more clear
because much of what a film is about is expressed by the
___________________________________94
plot and the events making up the plot. The representa
tional quality of most images of most films also aids the
audience in determining what a film is about.
Although Metz admits that "the event is still and
always the basic unit of the narrative" (FL, p. 24), he
completely ignores the event as an instrument of communica
tion in the narrative film.
One reason for not being able to deal with the event,
or with any other aspect of the plot, is his unwillingness
to deal with the image which is inextricably bound with the
subject matter of a film.
Metz's failure to deal with the image and with the
events depicted through the image eliminates any considera
tion for important means of communication such as composi
tion of the image, camera and/or subject movement, proxemic
patterns, color, costumes, make-up, texture, lighting, sets,
and the performance of actors.
Matters of tempo and rhythm within the shot and in the
arrangement of shots into scenes and sequences, which is
the source of great aesthetic pleasure and communication of
important information to an audience, are completely elimi
nated by Metz's approach.
The fact that images often have metaphorical signifi
cance cannot be taken into account by Metz's methods because
his semiological approach forces him to study only what he
considers to be the denotative aspects of film.
95
In the art of film two very important expressive means
at the disposal of the filmmaker are the manipulation of
space and time. Narrative films tell a story from selected
viewpoints which tend to fragment the dimension of space.
While this is happening, the dimension of time can be
expanded or contracted. This manipulation of space and time
is a key element in controlling the flow of information to
an audience and in determining the response of the audience
to the film.
Despite his emphasis on the syntagmatic dimension of
film and the structural relationships of the elements he
sees operating in it, Metz fails to take into account the
effect which the manipulation of space and time has on an
audience. Again, the main reason why this occurs is that he
refuses to deal with the image which contains a fragment of
the time-space continuum.
Although he identifies the soundtrack . as one of the
main channels of information, there is nothing in his
analytic tools which gives us even a clue as to how the
various soundtrack elements communicate meanings to an
audience.
The spoken word in the form of dialogue or narration
which is the source of a great deal of information in
narrative films is totally ignored by Metz. In view of his
linguistic approach this is a very curious omission.
In Film Language and in Language and Cinema, there is no
____________________________________________________________________________
analytic tool and no analytic procedure which accounts for
the manner in which music conveys to an audience extremely
valuable information on the intellectual and emotional
level. The same is true about the sound effects element of
the soundtrack. Metz merely identifies sound effects as one
of the elements of the channel of information which the
soundtrack represents. No attempt is made to examine how
information is communicated through sound effects.
Metaphorical uses of sound, which can be very rich
sources of information, are completely ignored by Metz.
The projection of a film in a motion picture theatre
constitutes a performance. It is, of course, a recorded
performance and as such quite different from one involving a
stage play. However, during the showing of a film there is
emotional interaction between the members of the audience
which is similar to the one that exists among members of an
audience experiencing a stage play. For an individual who
is viewing a film the experience is unique since in subse
quent screenings of a film he is bound to react to it in a
different manner. This could be due to the fact that the
audience composition may be different, or because the
individual has already been exposed to the filmed performance;
and he is communicating with it in a different manner.
In Metz's analytic procedures there is absolutely
'
nothing which accounts for the fact that in motion pictures
the spectator is a member of an audience witnessing a
B J i
performance.
Finally, Metz's analytic procedures fail to take into
account the physical conditions under which the spectator
comes in contact with a film and the psychological implica
tions of these conditions. For example, Metz completely
fails to take into account that the spectator is viewing a
film in a darkened theatre and that his attention is focused
on a two-dimensional picture within the boundary of a frame.
98
FOOTNOTES
XI. M. Bochenski, The Methods of Contemporary Thought
(Dordrecht, Holland: D. Reidel Publishing Co., 1965),
pp. 16-25.
2Daniel, "Metz's Grande Syntagmatique," p. 83.
3Jameson, Prison-House of Language, p. 169.
CHAPTER IV
THE SEVERED ROOTS
Narrative films are cultural objects within the social
process of communication.1
Making films, viewing films, and writing about films
are part of social interaction. As cultural objects films
reflect the ideas, beliefs, and values of the filmmaker. A
viewer relates to films in the context of his culture. The
film theoretician does not live in a social vacuum. He
approaches the study of films in terms of what is culturally
meaningful to him.
All of this represents an important aspect of the
reality of film and Metz is fully aware of it. However,
Metz's analytic procedures fail to take into account the
socio-cultural context of film.
The purpose of this chapter is to point out how Metz
indicates his awareness of the socio-cultural aspects of
film, and how his semiology operates in isolation from this
aspect of the reality of film.
Metz's awareness of the position which film occupies
within society and culture is demonstrated in Film Language
repeatedly. Some of the most obvious manifestations of this
100
awareness, other than the ones reflected in the extensive
quotes presented in the previous chapters, are the follow
ing:
What we call "the cinema" is not only cinematographic language
itself; it is also a thousand social and human significations
that have been wrought elsewhere in culture but that occur also
in films. (FL, p. 74)
The merging of the cinema and of narrativity was a great fact,
which was by no means predestined— nor was it strictly fortui
tous. It was a historical and social fact, a fact of civiliza
tion . . . a fact that in turn conditioned the later evolution
of the film as a semiological reality. (FL, pp. 94-95)
In discussing problems of denotation in the fiction
film he distinguishes between two general types of signify
ing organizations: "cultural codes," and "specialized
codes" (FL, p. 112):
The iconological, perceptual, and the other codes are cultural
codes, and they function in good part within photographic and
phonographic analogy. (FL, p. 113)
These codes are "natural" and they do not require that
the spectator be trained in any special way in order to
relate to them. It is sufficient for someone to be living
in, and to have been raised in, a given society.
On the other hand,
The purely cinematographic signifying figures studied here
(montage, camera movements, optical effects, "rhetoric of the
screen," interaction of visual and auditory elements, and so on)
constitute specialized codes . . . that function above and beyond
photographic and phonographic analogy. (FL, p. 113)
He acknowledges that films contain "fragments of
reality" which reflect cultural significations.
The mechanical character of the basic filmic operation (photo
graphic and phonographic duplication) has the consequence of
integrating into the final product chunks of signification whose
101
internal structure remains afilmic, and which are governed mainly
by cultural paradigms. When some of these "fragments of reality"
have been specially produced for the film (i.e., mise en scene),
this production itself . . . is never entirely obedient to systems
that are unique to the art of film, but rather in large part to
those same cultural significations that intrude into the filming
of an object. (FL, pp. 139-140)
As it has been indicated in previous chapters, Metz
acknowledges that the competence of the semiotics of the
cinema eventually encounters its limits because in trying to
discover the smallest unit of cinematographic signification
one eventually becomes confronted with the "myriad winds of
culture" (FL, p. 142) .
In putting forward his principle of the "see-saw,"
which operates between "the screen instance (which signi
fies) and the diegetic instance (which is signified)," Metz
points out that
This see-sawing . . . is nothing other than the consequence of an
underlying cultural and social fact: The cinema, which could
have served a variety of uses, in fact is more often used to tell
stories. (FL, pp. 143-144)
Metz's discussion of certain theoretical problems
involving "the modern cinema and narrativity" gives him the
opportunity to make a comment which is significant in terms
of his failure to consider the sociocultural dimension of
film:
Remove "drama," and there is no fiction, no diegesis, and
therefore no film. (FL, p. 194)
While dealing with the work of Pier Paolo Pasolini,
Metz admits that:
. . . when very broadly cultural codifications occur in films— as
they do frequently, especially when one thinks of the contents of
102
individual films— they are often present in the image itself (or
in the sound itself)— that is to say, within the "analogy," or at
a point that, in relationship to the total economy of the filmic
signification, is distinct from that occupied by the codifications
that constitute what one calls "cinematographic language." (FL,
p. 215)
Then, he attempts to make a distinction which points again
to the sociocultural dimension of film:
The image of the wheels of the train derives from society, not
from the cinema; when it appears on the screen it is identified
by visual analogy with the real wheels of a train, and it is
thanks to this resemblance that the film is able to carry all
the additional significations associated with the image in
culture. (FL, p. 215).
Afraid, perhaps, to open the door to the "myriad winds of
culture," he quickly points out:
But if the image is ordered along with other images in an alter
nate montage . . . another kind of codification emerges, one that
is specifically cinematographic and is no longer broadly cultural,
and is superimposed over the visual analogy and not merged with
it. Similarly, one must point out that cultural codifications,
when they occur in films, often appear on the level of the actual
filmed "objects," whereas cinematographic codifications mainly
affect the disposition of the objects once they are filmed. (FL,
p . ' - ' 215)
Forced once more by the inadequacy of his approach into
attempting another impossible distinction, Metz immediately
tries to find a way out of the difficult situation in which
he has placed himself. He says:
There remains, of course, a problem, and that is that cinemato
graphic language itself, in as much as it is a body of orderings,
must certainly be influenced by various sociocultural codifica
tions: Just as the "filmed object" retains the meaning it had
outside the film, the types of filmic orderings must in one way or
another refer to given patterns of intelligibility within society.
Thus, parallel montage is a peculiarly cinematographic figure, but
it is inconceivable that it could legitimately exist in the cinema
in a society that had no prior notation (in its language, its
writings, its "logic," etc.) of the symbolic and intelligible
value of certain very general types of relationships such as
alternation, parallelism, antithesis, etc. (FL, p. 215)
103
Metz admits that there are three types of influence,
"censorships" as he calls them, which determine the content
of narrative films: "censorship of moral standards"
(censorship proper), "economic censorship," and "ideological
censorship" (FL, pp. 236-237).
According to Metz, operating around, beside, beneath,
and larger than institutional censorship, there is the
"censorship of the Plausible," a "reduction of the possible,'
which determines the manner that film subjects are handled.
He claims that the Plausible is cultural and arbitrary and
varies according to the country, the period, and the genre
(FL, pp. 239-244).
Metz acknowledges that
Since the film-maker shoots films, to some degree he often shoots
the films of other people, believing that he is shooting his own.
To tear oneself away, even partially, from an attraction so
profoundly rooted in the fact of culture, in the shape of fields,
requires unusual strength of mind. Books reflect each other, so
do paintings, and so do films. (FL, p. 245)
In Film Language, Metz excludes the sociocultural dimen
sion of film through the following:
He fails to examine the role which visual perception
plays in experiencing films. He does not formally take into
account the numerous studies which have been conducted on
visual perception, and he does not incorporate in his work a
study of his own. The very few observations he makes on how
audiences perceive films are unsupported by evidence and
they appear to be nothing more than his impressions of the
subject.
_______________________________________________ 104
By not taking into consideration the mechanics of
perception, he is unable to deal with what an audience
brings to the viewing of films in the form of cultural
conditioning, and he cannot explain how an audience relates
to the aspects of culture contained in the images and sounds
of a film. Considering that he claims to be very much
concerned with the problem of the "impression of reality in
films," his failure to deal with something as fundamental as
perception constitutes a major weakness of Film Language.
It must be remembered, of course, that Metz does not
deal with the natural object but with a mental duplicate of
it which he has invented. He mentions visual perception a
number of times, but the product of his techne is unrelated
to visual perception.
After he states that there is a distinction between
"cultural" and "specialized" codes, he abandons the cultural
codes and attempts to examine signification in the cinema
only in terms of the specialized codes.
However, he completely fails to indicate how the
specialized codes function "above and beyond photographic
and phonographic analogy." His subsequent comments about
the "chunks of signification" which are governed by cultural
paradigms and which are contained in the "final product,"
indicate that he is unable to make a clear distinction
between the cultural and the specialized codes.
Although he is forced to admit that the "see-saw
_______________________________________________________________________________________________________________________________105
principle" operating between the signifier and the signified
is the consequence of the underlying cultural and social
fact that the cinema has been used to tell stories, he
fails to examine the cultural and social aspects of his
"see-saw principle." The same thing occurs in connection
with the dramatic-theatrical dimension of film. Metz does
remove "drama" and all the sociocultural aspects associated
with it.
Metz does not provide any element in his analytic
procedures which can deal with the four kinds of cultural
and institutional "censorships" which he sees operating in
film and which determine the handling of film subjects.
Since he indicates that these "censorships" are very
much related to the manner in which films signify, his
failure to account for these "censorships" constitutes a
major flaw of his analytic procedures.
He makes no effort whatsoever to determine how the
culturally conditioned Plausible influences signification
in the cinema.
In Film Language Metz does not suggest any method for
dealing with the "fact of culture" that films reflect one
another. However, this "reflection" is an important element
in terms of signification, and his failure to take it into
account is another major flaw of his analytic procedures.
Metz claims in Film Language that some of the central
questions in the semiology of the cinema are successivity,
1061
precession, temporal breaks, causality, adversative rela
tionships, consequence, spatial proximity and distance (FL,
p. 98) .
All of these matters are inextricably related to
perception and sociocultural conditioning. His Grande Syntag-
matique supposedly represents a master plan accounting for
intelligibility in filmic discourse. However, the proced
ures which gave birth to the Syntagmatique and the master plan
itself ignore the sociocultural dimension of all the central
questions in the semiology of the cinema.
In Language and Cinema some of the ways by which Metz
indicates his awareness of the social context within which
cinematic signification occurs are the following:
Language and cinema opens with the statement that the
cinema,is a vast and complex sociocultural phenomenon and
an established cultural fact (LC, p. 9).
He distinguishes between the cinematic and the filmic
"fact" and points out that at the center of the cinematic
fact lie technological, economic, and sociological dimen
sions (LC, p. 12).
He admits that because film is a "closed text" like a
myth, a play, or a novel, it is a "total cultural object"
(LC, p. 18).
As has been indicated, Metz claims that the total
message of the film becomes comprehensible through five
types of systems: visual and auditory perception, recogni
tion, identification, and enumeration of visual and auditory
__ 102
objects, the ensemble of "symbolisms" and connotations which
are associated with the objects, the ensemble of narrative
structures which are present in a given civilization, and
the ensemble of properly cinematic systems (LC, pp. 33-34).
Metz points out that the first four of these systems
are cultural and acquired by living in a given culture. As
a result of this later on he is forced to admit that a film
is not only an example of cinema but also of culture (LC,
p. 72).
He declares that cinema is a "rich language system" and
identifies such a system as "one which is open to all
social, cultural, aesthetic, ideological, etc., influences
and initiations" (LC, p. 35). Also, he declares that "the
cinema is one of the language systems endowed with some
sociocultural depth" (LC, p. 36), and that
The cinema . . . like all rich languages, is largely open to all
symbolisms, collective representations, and ideologies, to the
influence of diverse aesthetic theories, to the infinite play of
influences and filiations between different arts and different
schools, to all the individual initiatives of film-makers
("revivals"), etc. (LC, p. 37)
He admits the social status of film by saying that all
films are a work of art and that they function socially as
works of art (LC, p. 38). He emphasizes the social function
of the cinema by adding that
. . . the cinema, like literature or the theatre, is in principle
capable of saying anything, and conveys non-specialized signi-
fieds which are above all ideological and cultural, and could be
found just as well . . . in other language systems utilized by
the same civilization during the same period. (LC, p. 39)
Metz knows that there is an ideological-political
108
dimension in films, and, for this reason, consciously or
not, films reflect systems of political thought (LC, p. 99).
The fact that films reflect social behavior is admitted
by Metz a number of times. He locates the reflection of
social behavior in the "extra-cinematic material" which, he
claims, is "as extensive and varied as social life itself
(from which it directly stems) " (LC, p. 150) .
Some of the specific ways by which Metz fails to
consider the sociocultural dimension of film are the
following:
After he distinguishes between the "cinematic" and the
"filmic fact," he chooses to deal only with the "filmic"?
rejecting the "cinematic" at whose center lie the technolog
ical, economic, and sociological dimensions of film. Then,
he declares that the "filmic fact" isrstill too vast and
unmanageable an area because the "filmic fact" contains
psychological, sociological, and aesthetic phenomena (LC,
p. 16). He proceeds to eliminate from consideration these
dimensions of the reality of film by dealing with the
"filmic fact" only as a "language system."
After he identifies the five systems through which
the total message of the film becomes comprehensible, he
puts aside the first four, which are acquired through
cultural conditioning, and he develops his analytic proced
ures around the fifth which he labels "properly cinematic."
This is very similar to his ignoring of the "cultural codes"
--------------------------------------------- — -------------------------------------- 109
in Film Language and attempting to deal only with what he
calls the "specialized codes."
He fails entirely to deal with film as an art form and
with film aesthetics in terms of the traditional concerns
of film aesthetics. As a result, he provides no feature in
his analytic procedures for handling the social status of
film as an art, and he does not concern himself at all with
sociocultural aspects of film aesthetics.
He distinguishes between "cinematic" and "extra-
cinematic material" and points out that the "extra-cine
matic are the elements yrhich reflect social life. Then he
proceeds to examine only the "cinematic" and to leave out
everything which reflects social behavior, ideology, and
the important economic elements of the reality of film.
He does not examine the relationship between the codes
of sight and the codes of spoken language although he
acknowledges the existence of very close ties between the
"visible world" and language. In this manner he fails to
investigate the cultural dimensions of language and, through
them, the cultural dimensions of visual perception as they
relate to film.
Although he identifies cinema as a "rich language
system," he fails to provide any feature in his analytic
procedures which accounts for the social, cultural, aes
thetic, and ideological influences to which such "rich"
systems are susceptible.
He discusses extensively cinematic "systems," but he
— ------ --------------------- --------- -------_____— --- 1X0
views the concept of the system in the structuralist tradi
tion and in a manner which hermetically seals out the socio
cultural dimensions of film. He points out, characteristic
ally, that
What the word system implies, in all its uses, is that one is
thinking of a coherent and integrated ensemble— an "autonomous
entity of internal dependencies," as Hjelmslev said of the notion
of structure— of an ensemble within which all the elements hold
together and have a value only in relation to one another. (LC,
p. 84)
He admits that the notions of code and sub-code do not
directly express empirical reality but they are analytic
tools which are creations of the analyst (LC, p. 141).
However, he shifts his position and makes the codes
and sub-codes appear to be elements of the empirical reality
which films represent. He says:
The film must be lighted and must be edited. The cineast will
necessarily choose, in either a conscious or intuitive manner, a
sub-code of lighting and a sub-code of montage. (LC, p. 142)
Then, he continues:
The plurality of the sub-codes is due to the fact that the
solutions to these problems are in turn quite diverse: it is not
exactly the composite of the "cinematic" that it reflects, but its
historicity, its variations from one epoch to another, etc. The
ideal sum of sub-codes . . . constitute nothing more than the
history of the cinema, at least insofar as what is truly cine
matic is concerned. (LC, p. 143)
In the paragraphs which follow he excludes the "extra-
cinematic" from his approach, and in this manner he leaves
out the sociocultural dimensions of film which the "extra-
cinematic" represents. He says:
As for the extra-cinematic codes, we have mentioned them only in
order to recall that they play an important role in films. But a
study of these codes could not be a goal of the analyst of the
_______________: __________________________________in
cinema, nor of the analyst of films. Neither is it a question of
a unified study, i.e., a "discipline." The extra-cinematic
material found in films is as extensive and varied as social life
itself (from which it directly stems), and its analysis relies
upon quite diverse skills and a large number of pre-existing
disciplines. (LC, p. 150)
He makes no effort to examine the sociocultural aspects
of the "material of expression" of films although he points
out that
"The material of expression," as its name indicates, is the
(physical, sensorial) material nature of the signifier, or more
exactly of the "fabric" into which these signifiers are woven.
This fabric may be phonetic...aural but not phonetic...visual and
colored...visual but not colored...it may consist of movements of
the human body...etc. (LC, p. 208)
Although he observes correctly that meaning in film
communication is psycho-sociological in nature (LC, p. 211),
he fails entirely to deal with meaning in psycho-sociolog
ical terms. Instead, he approaches the matter of meaning in
structuralist terms and identifies it with the relationships
of the various signifying elements operating within the
systems.
He labels "extra-cinematic," and thus he excludes from
consideration, important aspects of the reality of the
content of films such as:
... .a given collective representation or some social imago (the
seductor, the model spouse, the wayward youth, the adventurer,
etc.) which appear in films as well as in books, newspapers, and
conversations, at least within a given cultural area. (LC,
p. 213)
In discussing the "specificity" of the cinematographic
codes, he points to the existence of "visual-iconic codes"
which are of great anthropological importance. He labels
1 1 2
these codes "codes of analogy" and he observes that they
represent
. . . a whole ensemble of psycho-physiological montages, inte
grated with the perceptual activity itself, and whose modalities
vary noticeably from one culture to another. (LC, p. 228)
Then, he proceeds to minimize the importance of these
"codes of analogy," and he eventually fails to take them
into account in his examination of cinematic signification
on the grounds that the "picture track" of the film is not
sufficiently defined by its iconicity.
In Film Language and in Language and Cinema, Metz's failure
to take into account the sociocultural dimension of film
appears to stem from his aversion to anything which might be
regarded as normative, and from his insistence on maintain
ing what he considers to be a purely descriptive and "scien
tific" approach to filmic signification. As he points out
time after time, the audience "reads" a film in a "naive"
(cultural) manner, while the film semiotician reads it
analytically and scientifically.
However, as the editors of cinethique point out in terms
of their own interest in marxist film theory, Metz's
"neutrality" and his concept of scientific experimentation
ss independent of social practice is an illusion.2 No
matter what Metz says, it is unavoidable that a person in
his position be subject to the same idealist determinations
that govern the film viewer.
Metz's scientificity bears the external signs of
1X3
seriousness, rigor, method, and precision, but he fails to
acknowledge that he is unavoidably dealing with film in
terms of what is culturally meaningful to him. The result
of pretending that he can remain "neutral" is that he fails
to provide for any feature in his analytic procedures which
can account for the semiotician's sociocultural viewpoint.
Under these circumstances, one wonders, in what set
ting, from what perspective, in terms of what time period
reference, influenced by what set of values does the semi
otician approach the study of film signification? Metz does
not provide any answers to such questions. The consequences
of failing to account for all of this is that Metz, who
eliminates the author of the film, the concerns associated
with film-making, and the manner in which an audience
perceives the filmophanic film, has also eliminated, in
essence, the film semiotician.
114
FOOTNOTES
1 Andrew Tudor, Image and Influence: Studies in the Sociology
of Film (London: Allen & Unwin, 1974), p. 137.
2Cinethique, "On 'Langage et Cinema' " Screen, 14, No. 1/2
(Spring/Summer, 1973), 189-213.
1151
CHAPTER V
THE IMAGINARY SIGNIFIER
In Chapter I there were a number of references to
Metz's psychoanalytic approach to film signification. The
purpose of Chapter V is to describe briefly the psycho
analytic approach which Metz establishes in "The Imaginary
Signifier," to relate this new direction to his earlier
structural-linguistic approach, and to point to the problems
that Metz encounters in terms of the reality of film.
This chapter is divided into three parts: "General
Characteristics and Features of Metz's Psychoanalytic
Approach," "From the Linguistic to the Linguistic-Psycho
analytic Approach," and "The Linguistic-Psychoanalytic
Approach and the Reality of Film."
General Characteristics and Features
of Metz's Psychoanalytic Approach
The purpose of "The Imaginary Signifier" is, according
to Metz, to try to answer the question, "What contribution
can Freudian psychoanalysis make to the study of the cine
matic signifier?" (IS, p. 28).
To achieve this he feels the need to first define what
he means by "contribution," "Freudian," and "cinematic
116
signifier."
Unlike the early days in his "intervention," when he
proclaimed that the time had come for a semiotics of the
cinema based solidly on semiotics, Metz is now much more
cautious. He speaks in vague terms of a "psychoanalytic
inspiration," and he admits that psychoanalysis cannot be
the only discipline concerned with the study of the cine
matic signifier. The "psychoanalytic inspiration" must be
combined, he says, with "linguistic inspiration" because
linguistics and psychology are two sciences both of which
deal with signification (IS, p. 28).
However, both these sciences have to be set "within the
horizon of a third perspective . . . the direct studies of
societies, historical criticism, the examination of infra
structures" (IS, p. 28). This is so, Metz finally admits,
because:
In cinematic studies as in others, semiology (or semiologies)
cannot replace the various disciplines that discuss the social
fact itself (the source of all symbolism), with its laws that
determine those of the symbolic without being, identical with
them: sociology, anthropology, history, political economy,
demography, etc. (IS, p. 29)
Viewing now film from this expanded perspective, Metz
comes to have a new hope:
A combination of linguistic and psychoanalytic inspiration may
lead gradually to a relatively autonomous science of cinema
(=semiology of the cinema). (IS, p. 31)
As far as psychoanalysis is concerned, Metz points out
that
What I shall call psychoanalysis will be the tradition of Freud
117
and its still continuing developments, with original extensions
such as those that revolve around the contributions of Melanie
Klein in England and Jacques Lacan in France. (IS, p. 32)
The very crucial matter of defining that on which the
linguistic-psychoanalytic inspiration will be applied, the
"cinematic signifier," becomes the source of many problems
for Metz.
Frankly, he admits, he is not sure yet what should be
the overall orientation which will determine what aspect of
film he will study. But he does know that he will not
assume a "nosographic" approach. He defines this as an
attempt to treat films as symptoms from which one finds
access to the neuroses of the filmmaker, the script writer,
etc. (IS, p. 32). Also, he knows that he does not want to
engage in "psychoanalytically inspired characterology" (IS,
p. 33). This is a version of the nosographic approach which
places less emphasis on the pathological aspects of the
filmmaker's personality.
What he finds wrong with these two approaches is that
they are interested in persons and not the "discursive
facts," which he defines as "filmic texts or cinematic
codes" (IS, p. 34).
As a first and "somewhat simplifying step," he decides
to call his approach "the psychoanalytic study of film
scripts" (IS, p. 34). "Film script" is defined "broadly" as
"the manifest thematic complex of the film," extending to a
large number of features which may not appear in the written
document commonly referred to as the screenplay, and which
118
involves elements related to plot, situations, characters,
"landscapes," "period details," etc. (IS, p. 34).
Defined in this manner, the "film script" is regarded
by Metz as one aspect among others in the "textual system"
of films. This "textual system," he claims, "shifts in the
direction of the signifier" (IS, p. 36). Metz goes on to
state that:
To study the film script from a psychoanalytic (or more broadly
semiotic) viewpoint is to constitute it into a signifier. In this
the script is like a dream, as are many human products. The
manifest dream, i.e., the dream as such— "dream content" for
Freud...— is a signifier for the interpretation. (IS, p. 36)
Since there can be psychoanalytically oriented studies
which treat the "film script" in terms of a signifier which
is not "cinematic," Metz points out that his approach will
consist of a "direct examination, outside any particular
film, of the psychoanalytic implications of the cinematic"
(IS, p. 41).
In linguistic terms, Metz claims
. . . what calls for psychoanalytic illumination is not just each
film (not just films) but also the pertinent features of the
matter of the signifier in the cinema, and the specific codes
that these features allow: the matter of the signifier and the
form of the signifier, in Hjelmslev's sense. (IS, p. 42)
The aspects and features of the cinematic signifier
which Metz claims need to be studied are the following:
1. Visual and auditory elements of the signifier since
the signifier is perceptual in nature (IS, p. 46).
2. Fantasy aspects of the signifier since it consists
of images appearing on a screen which is an "area" more
1131
closely associated with fantasy than is the theatrical
stage.
3. "Fictive" aspects of the "unfolding" of the images
on the screen, in the sense that the actor, the decor, and
the words one hears are all absent, recorded:
. . . a little rolled up perforated strip which "contains" vast
landscapes, fixed battles, the melting of the'ioe on the River
Neva, and whole life-times, and yet can be enclosed in the
familiar round metal tin, of modest dimensions, clear proof that
it does not "really" contain all that. (IS, p. 47)
Because of this "absence," this imaginary quality
of the signifier, Metz claims that "every film is a fiction
film" (IS, p. 47). However, unlike his work in Film Language
and Language and cinema where he studied signification in terms
of the narrative-fiction film, in his linguistic-psycho
analytic approach he wishes to deal mainly with significa
tion as it applies to cinema in general.
4. The "false" aspects of the perceptions of the
cinema in the sense that
The activity of perception in it is real (the cinema is not
fantasy), but the perceived is not really the object, it is its
shade, its phantom, its double, its replica in a new kind of
mirror. (IS, p. 48)
The next major section in "The Imaginary Signifier"
deals with the role of the spectator in the cinema and of
his relationship to the "mirror." Here Metz sees the factor
of identification operating on a number of levels:
1. The spectator identifies with the character of the
fiction in narrative-fictional films. However, according to
Metz, this type of identification is not valid for the
___________________________ 12D
"psychoanalytic constitution of the signifier of the cinema
as such" (IS, p. 49) .
2. The spectator identifies with the actors or with
the human figures appearing in films which do not employ
actors.
3. The spectator identifies with himself. This comes
about as a result of the place of the spectator's ego (in
the Freudian-Lacanian sense) during the projection of the
film.
The spectator is absent from the screen because the
screen is not a true mirror since it does not reflect the
spectator's image. Yet, the spectator is aware of two
things: That he is perceiving something imaginary, and that
he knows that it is he who is perceiving it. "In other
words," Metz concludes, "the spectator identifies with
himself" (IS, p. 50).
An important element in Lacan's psychoanalytic
theory, which Metz incorporates in his work, is Lacan's
concept of ego formation. According to Lacan the ego is
founded on the opposition and identity between Self and
Other: a child discovers that there is a difference between
himself and the world and through the "imaginary" relation
ship to others this difference becomes an opposition. The
child, which is born as an undifferentiated "a-subjective"
being, becomes an individual by passing through a phase
during which it feels as "he" or "she" and eventually comes
121
to feel as "I." The imaginary relationship to others
centers around the Oedipus Complex. As the child matures
and comes to feel as an individual this relationship becomes
a symbolic one.1
The relationship between the child and the external
world described above is somewhat similar to that described
by Melanie Klein when she speaks of a "fantasy relationship"
between the maturing individual and the external world.
4. The spectator identifies with the camera. During
the showing of a film the projector replaces the camera.
But, according to Metz, the projector is at the back of the
spectator's head, "precisely where fantasy locates the
'focus' of all vision" (IS, p. 52).
5. The spectator identifies with the screen. To
explain this aspect of identification Metz has to expand
his theory of spectator identification with himself and the
camera-projector. He claims that:
During the performance the spectator is the searchlight . . .
duplicating the projector, which itself duplicates the camera,
and he is also the sensitive surface duplicating the screen,
which itself duplicates the film-strip. (IS, p. 53)
In this manner, Metz claims, the cinematic signifier
"depends on a series of mirror-effects organized in a
chain" (IS, p. 53).
Metz is very insistent and quite defensive on this
matter of a relationship to a mirror-like surface because,
it seems, without it he cannot bring into his psychoanalytic
approach the ego-formation mechanism which Freud and Lacan
122
see as operating in life; which in turn may not allow him to
deal with spectator identification in psychoanalytic terms
at all.
What causes Metz to operate under stress is that he
wants, once again, to be precise and "scientific," by util
izing this time the analogy that "film is like a mirror"
(IS, p. 4 8), although he knows very well and admits that
film "differs from the primordial mirror in one essential
point . . . there is one thing, and one thing only that is
never reflected in it: the spectator's own body" (IS,
p. 4 8). Metz also knows that "the cinema spectator is not a
child" (IS, p. 49). But since he is determined to build on
this analogy although film is not a true mirror and the
spectator is not a child, he claims that what makes possible
"the intelligible unfolding of the film" is the fact that
the spectator "has already known the experience of the
mirror (of the true mirror), and is thus able to constitute
a world of objects without first having to recognize him
self within it" (IS, p. 49).
Since Metz wants to introduce to his scheme the
dimension of the symbolic, which may allow him to incorpo
rate in his work Lacan's ideas on the symbolic, he claims
that the spectator "knows himself and he knows his like: it
is no longer necessary that this similarity be literally
depicted for him on the screen, as it was in the mirror of
his childhood" (IS, p. 49). This, claims Metz, causes the
______________________________________ 123
cinema to be "on the side of the symbolic" (IS, p. 49).
A key weakness of this enterprise is that Metz provides
no evidence whatsoever that what he is saying is based on
any kind of research in the field of motion pictures or any
of the other fields which are involved. Also, these
impressions of his are not supported by any examples from
films, or by the reactions which someone has had while
viewing any specific film.
The importance of the analogy of the "mirror" becomes a
very critical matter for Metz when he discusses certain
"sub-codes of identification" which, according to him,
operate in the cinema. Examples of such sub-codes are the
various kinds of subjective shots, and various kinds of
"looks." The unusual framings and uncommon angles, which
are often associated with subjective shots, merely intensify
the feeling of the spectator's identification with himself.
As for "looks" (the way a character looks at someone else
on the screen, or the point of view of a character indicated
by a certain camera angle), Metz claims that they are
related to the psychoanalytic concepts of "primary" and
"secondary" identification which also play a role in ego
formation.
In the cinema, Metz claims, identifying with one's own
look must be considered "primary cinematic identification"
(IS, p. 58), although in terms of Freudian and Lacanian
psychoanalysis such identification constitutes secondary
___________________________________________________________________________ 124
identification. On the other hand, Metz continues:
. . . identification with the screen characters with their own
levels (such as looks at out-of-frame characters or scenes which
represent points of view of out-of-frame characters) represent
secondary, tertiary cinematic identifications, and as such they
are sub-codes. (IS, p. 58)
Only after he has prepared the ground in this manner
can Metz attempt to explain what is involved in the process
of signification as it operates in his linguistic-psycho
analytic approach to film communication. He describes the
process in this manner:
. . . in order to understand the film (at all) I must perceive
the photographed object as absent, its photograph as present, and
the presence of this absence as signifying. The imaginary of the
cinema presupposes the symbolic, for the spectator must first of
all have known the primordial mirror. But as the latter insti
tuted the ego very largely in the imaginary, the second mirror
of the screen, a symbolic apparatus, itself in turn depends on
reflection and lack. However, it is not fantasy, a "purely"
symbolic-imaginary site, for the absence of the object and the
codes of that absence are really produced in it by the physis of
an equipment: the cinema is a body (a corpus for the semiolo-
gist), a fetish that can be loved. (IS, pp. 58-59)
With this reference to a fetish, Metz moves on to the
last two major sections of his essay where he discusses the
relationship between spectator and film in terms of
voyeurism and fetishism.
He claims that the cinema becomes possible through two
"perceptual passions" of a sexual nature: the desire to see
(scopic drive, scopophilia, voyeurism), and the desire to
hear (the "invocative drive") (IS, p. 59). Both of these
irives are dependent on a lack which marks them on the side
cf the imaginary. This is so, we are told, because these
irives always remain more or less unsatisfied even when the
__________________________________________________ 125
object of desire has been attained:
The lack of what it wishes to fill, and at the same time what it
is always careful to leave gaping, in order to survive as desire.
In the end it has no object, at any rate no real object (a "lost
object") which is its truest object, an object that has always
been lost and is always desired as such. (IS, p. 60)
Metz associates all of this to Freud's observation that
voyeurism always keeps apart the object looked at, and then
he proceeds to build a case for the spectator relating to
film in the manner of a voyeur. Then he goes on to associ
ate cinematic voyeurism ("unauthorized scopophilia") with
the concept of the "primal scene." To prove that such a
relationship exists, he points to the following aspects of
experiencing motion picture films:
1. The obscurity surrounding the onlooker-spectator
during the projection of a film.
2. The keyhole effect of the screen aperture.
3. The less accepted and more shamefaced nature of
cinematic voyeurism. This is due to the fact that, unlike
the theatre which is an ancient art which grew out of
ceremonial and religious practices giving a respectability
to its voyeurism, the voyeurism of the cinema has not been
aiccepted as "legitimate."
4.' The manner in which the anonymous movie audiences
enter and exit movie theatres has a "furtive" quality to it
much like the visits to maisons de tolerance. The cinema, Metz
claims,
. . . is based on the legalization and generalization of the
prohibited practice . . . going to the cinema is one licit
—^^
activity among others with its place in the admissible pastimes
of the day . . . and yet that place is a "hole" in the social
cloth, a loophole opening on to something slightly more crazy,
slightly less approved than what one does the rest of the time.
(IS, pp. 65-66)
The imaginary dimensions of the cinematic signifier,
and the voyeuristic aspects of the relationship between the
spectator and the projected film, are examined by Metz in
terms of exhibitionism. He observes that voyeurism rests on
the fiction that the object of the voyeur's attention
"agrees" to be observed and it is, therefore, practicing a
form of exhibitionism. But, unlike the theatre where the
spectator and the actor are at the same time present one to
another, in the cinema:
. . . the actor was present when the spectator was not (=shooting),
and the spectator is present when the actor is no longer ^projec
tion) : a failure to meet of the voyeur and the exhibitionist
whose approaches no longer coincide (they have "missed" one
another). The cinema's voyeurism must (of necessity) do without
any very clear mark of consent on the part of the object. (IS,
p. 63)
One of the most important "roots" of the cinema in the
unconscious mind, which justifies a psychoanalytic approach
to cinematic signification, is, according to Metz, the
fetishistic aspects of the relationship between the specta
tor and the film. Metz informs us that for Freud and Lacan
fetishism is linked very closely with castration and the
fear which castration inspires. Embracing the classic
Freudian views on the subject and the shift toward the
symbolic aspects of castration as expounded by Lacan, Metz
points out the following:
122
1. Castration is the mother's castration who, accord
ing to the child, has lost her penis.
2. The terrified child attempts to "arrest its look"
at what becomes its fetish. The fetish is a negative signi
fier and represents the penis. It is always a substitute
for it.
3. The cinematic aspects of fetishism include film
equipment and film technique, the enchantment for the cinema
shown by the film connoisseur, the pleasure derived by the
moviegoer, and the desire of the film theoretician to write
about film (IS, pp. 71-73).
Metz closes "The Imaginary Signifier" by stating in
very lyrical terms that
Psychoanalysis does not illuminate only the film, but also the
conditions of desire of whoever makes himself its theoretician.
Interwoven into every analytical undertaking is the thread of
self-analysis. (IS, p. 75)
From the Linguistic to the
Linguistic-Psychoanalytic Approach
The following brief comments, relating Metz's earlier
linguistic approach to the linguistic-psychoanalytic, seem
to be necessary for a proper transition to the last part of
this chapter which deals with the psychoanalytic approach
and the reality of film.
In "The Imaginary Signifier," Metz admits that,
compared to a more classical semiological approach such as
the one he had attempted earlier, in a psychoanalytic
approach attention shifts from the enonce to the enonciation;
12£
from what is uttered to the act of uttering.
This shift drives Metz away from the isolation of a
"filmic object" which can be studied phenomenologically.
However, he does not appear willing to abandon his search
for such an "object." "The Imaginary Signifier" is full of
references to the "manifest content" of films which serve as
the "object" of study. Exactly what is the nature of the
"manifest content" of films is never made clear.
An equally vague term which he uses often is "manifest
filmic material." This material is said to have signifiers
and signifieds. As indicated earlier, Metz's psychoanalytic
approach deals with "the matter of the signifier" of this
manifest filmic material and with the codes which are ‘
allowed by the "pertinent features" of the signifier (IS,
p. 42). Exactly what these "pertinent features" happen to
be is never explained.
As for the codes which these mysterious "pertinent
features" allow, the impression is given that the codes
could be properties of these features and not creations of
the analyst. Here Metz seems to be engaging again in the
same deceptive practice which characterized Language and
Cinema by making the codes appear to be either properties of
the manifest cinematic material, or analytic tools; depend
ing on what best suits his purpose at a given point in the
analysis.
The distinction between the "filmic" and the "cine
______________________________ _ _ _ __________________________________________________________________________________ .12. 9
matic" which was attempted unsuccessfully in Film Language and
in Language and cinema, is attempted again in "The Imaginary
Signifier." Metz indicates that his psychoanalytic approach
will be applied to the cinematic.
Although he has never been able to make a convincing
case for a distinction and a difference between the signi
fier and the signified in the cinema, he proceeds here to
utilize this distinction, and he points out that his purpose
is to study psychoanalytically the cinematic signifier and
not the signified.
In Film Language the purpose of his study was to examine
the narrative fiction film. In Language and Cinema his
analytic procedures were meant for the narrative fiction
film and for other types of films. In "The Imaginary Signi
fier" Metz tries to establish an approach which is suitable
for the analysis of the cinematic signifier in general
because, as he hints, the process of signification is
identical in the various types of films and film genres.2
However, he provides no evidence whatsoever that this obser
vation about the process of signification is based on a
formal study of any kind.
Metz claims that psychoanalysis explores the "primary
process," and linguistics (and modern symbolic logic) the
"secondary process." Together they cover the entire field of
signification (IS, p. 28). "Primary processes" are the
experiences, resulting from frustrations, which stimulate
___________________________________________________________________________13-0
the development of the Id. "Secondary processes" are the
activities of the Ego through which reality is discovered
i
and produced by means of action developed by thought and
reason.3
There are important links between linguistics and
psychoanalysis. But, it seems to me, the claim that
psychoanalysis deals with the "primary process" and linguis
tics with the "secondary process," a claim made so that
these two fields can be combined in the study of cinematic
signification, represents an over-simplification and a
distortion of what these two fields explore. An elaborate
case could be made questioning the wisdom and validity of
linking the two fields in the manner in which Metz links
them in order to study the cinematic signifier. Such a
"case" falls outside the limits of this study. However, it
should be pointed out here that Metz's method of linking
these two fields consists of applying to the examination of
the cinematic signifier a mixture of linguistic and psycho
analytic methods, principles, and concepts, without provid
ing any evidence whatsoever that the methods, principles,
and concepts of these two fields (of any two fields) can be
combined in this manner. One is tempted to remind Metz of
what he said in Language and Cinema about such an uncritical
blending of analytic techniques:
Methods are things which cannot be interchanged and which cannot
be "combined" without great danger of giving rise to monstrocit-
ies. (LC, p. 20)
131
What makes Metz's mixture of linguistics and psycho
analysis particularly disturbing is that he attempts to
incorporate in it the ideas and writing style of Jacques
Lacan.
Lacan is perhaps the most controversial writer in the
field of psychoanalysis who is suspected of making his
writings impossible to understand as a matter of conscious
intellectual strategy.1 *
Characteristic of the reaction which Lacan evokes are
the views of Anthony Wilden. In his impressive study
System and Structure, Wilden says that Lacan's works
. . . read more like a "schizophrenic discourse"— or like poetry,
or nonsense, depending on your prejudice and your tendencies
towards positive or negative transference— than anything else.
Lacan's hermeticism cannot be excused on any grounds.5
Lacan's overall "scientific" method is described by
Wilden in this manner:
Lacan takes bits and pieces from everywhere and anywhere and
jumbles them up— like the text of a dream— playing on ambiguities,
etymologies, puns, analogies, poetic metaphors— again like the
text of a dream— as well as on the reader's benevolent desire to
understand. Lacan's argument is presumably that he is represent
ing his theory in the very "language" the theory is designed to
analyze: if you understand the theory, you will then understand
his exposition, and vice versa. Within all this, there appear
strikingly important concepts, such as the theory of the Imagin
ary. Such an attitude nevertheless defies the simplest criterion
of all "science": that any theory must be demonstrably less
complex than what it is intended to explain.6
Even Claude Levi-Strauss, from whom Lacan has borrowed
key elements of the structural linguistic approach, attacks
some of Lacan's major positions as being "imposture" and
"sleight of hand."7
132
The result of this reliance on the ideas and writing
style of a writer such as Lacan is that Metz describes the
function of the psychoanalytic approach, which "The Imagin
ary Signifier" represents, in terms which often defy
comprehension.
As for the influence of Melanie Klein on Metz's
approach, in "The Imaginary Signifier" it is limited to
certain relatively minor concepts such as the "good" and
"bad object" relationship between a subject and an object.
However, Metz saying that the spectator has a "good" or a
"bad object" relationship with a film, instead of saying
that the spectator liked or disliked a film, hardly amounts
to a contribution of psychoanalysis to our understanding of
how films communicate.
The Linguistic-Psychoanalytic Approach
and the Reality of Film
It has been indicated that "The Imaginary Signifier"
represents Metz's attempt to set the foundations of his new
linguistic-psychoanalytic approach. This approach is still
in the process of being developed. For this reason, "The
Imaginary Signifier" must be viewed in terms of establishing
the direction towards which Metz is heading and not in term
of specific achievements made possible by this approach.
Examined from the point of view of the reality of film,
this essay reveals that Metz basically does not plan to
levelop anything other than a psychoanalytically flavored
1331
reinterpretation of the signification process which he
described in Film Language and Language and Cinema. This appears
to be the case because his "psychoanalytic inspiration"
turns out to be not a real integration of psychoanalysis and
film, but a labeling with psychoanalytic jargon of certain
aspects of his earlier structural linguistic approach.
The similarity between his earlier approach and the new
psychoanalytic one is also demonstrated by the fact that in
"The Imaginary Signifier" the reality of film is acknowl
edged and then disregarded. Since the pattern of this
practice has been demonstrated in the previous chapters
there is no need here to give examples of how Metz acknowl
edges and then disregards the reality of film. Instead,
what follows is a brief presentation of what occurs in this
essay in terms of the various aspects of the reality of
film.
The Director as the
Author of the Film~
In "The Imaginary Signifier" the banishment of the
director as the author of the film is as complete as in Film
Language and in Language and Cinema. This is so despite the fact
that Metz claims that in a psychoanalytic approach attention
shifts from enonce to enonciation. Concern for the spectator
as a participant in the act of enonciation appears to be an
important part of the new approach. However, concern for
the director as a participant in the "act of saying" is
____________________________ 134
replaced by concern for the role of the analyst.
What banishes the author/director is Metz's determina
tion to avoid the "nosographic" and the "psychologically
inspired characterology" type of approach. As we have seen,
these approaches are rejected because they are indifferent
to the "filmic text."
Apparently, Metz fails to see that his claim that the
psychoanalytic approach shifts attention from enonce to
enonciation, is in conflict with his claim that what needs to
be studied psychoanalytically is the "filmic text" which is
clearly an enonce.
An additional reason given for the rejection of the
"nosographic" and "psychological characterology" types of
approaches is that, according to Metz, they remain indiffer
ent to the social aspects of signification (IS, p. 33).
Metz is clearly in error here because most of the studies of
this nature view the neuroses and other aspects of the
personality of the author/director in a sociocultural
perspective.
In his discussion of sub-codes of identification, such
as "subjective images" which express the viewpoint of the
filmmaker, it seems that Metz places himself in the position
of having to deal with the author/director as an important
alement in the process of signification.
However, Metz extricates himself from this situation by
resorting to a manipulation of Lacanian psychoanalysis which
123
permits him to deal only with the spectator and maintain the
banishment of the author. The following description of how
the spectator supposedly feels when confronted with "subjec
tive images" is characteristic of the dexterity by which
Metz manages to stay on course with his approach:
Precisely because it is uncommon, the uncommon angle makes us
more aware of what we had merely forgotten to some extent in its
absence: an identification with the camera (with "the author's
viewpoint"). The ordinary framings are really felt to be non
framings: I espouse the film-maker's look...but my consciousness
is not too aware of it. The uncommon angle reawakens me and (like
the cure) teaches me what I already knew. And then, it obliges
my look to stop wandering freely over the screen for the moment
and to scan it along more precise lines of force which are
imposed on me. Thus for a moment I become directly aware of the
placement of my own presence-absence in the film simply because it
has changed. (IS, pp. 56-57)
Beyond this point in "The Imaginary Signifier," Metz
avoids any discussion which deals with aspects of significa
tion which may involve the author of the film.
The Relationship Between
Spectator and Film
Unlike Film Language and Language and Cinema, where Metz had
practically nothing to say about this relationship because
the object of his investigation was not really the filmo-
phanic film, "The Imaginary Signifier" contains extensive
material on the relationship between spectator and film.
These are the sections where Metz talks about the nature of
the "perceptual passions" of sight and hearing, the aspects
of voyeurism in film viewing, and the disavowal-fetishism
involved in seeing and writing about films.
The length of these sections, amounting to approximately
: ________________________________________________
half the material in the essay, and the detail in which he
describes the relationship, give the impression that Metz
finally intends to deal with this aspect of the reality of
film in a meaningful way. However, both the length and the
intricacy of the material are deceptive as far as a signif
icant contribution to our understanding of how films commun
icate is concerned.
The element of deception in Metz's work is contained in
two aspects of his presentation: (a) What he has to say
about the relationship between the spectator and film is not
related to the announced aim of his new approach which is
the examination of the cinematic signifier in terms of
Hjelmslev's ideas, and (b) his approach indicates a great
lack of critical distance from Freudian psychoanalysis.
The relationship described in the sections which deal
with perception, voyeurism, and fetishism, is that between
a human being, the spectator, and the filmophanic film.
The relationship suggested in the statement of purpose of
the psychoanalytic approach to film, concerns a relationship
between an analyst such as Metz and analytic concepts which
are the creations of the analyst; it is a study of the
matter of the signifier in Hjelmslev's sense.
In this mixture of approaches no pattern exists, no
criteria, and, in the final analysis, no apparent logic for
all the mental acrobatics involved in his talking about a
spectator's relationship to the filmophanic film in one
______________________________________ 137
paragraph, and an analyst's relationship to the matter of
the signifier in another paragraph.
The lack of critical distance from Freudian psycho
analysis which characterizes the psychoanalytic dimension of
Metz's new approach has been the subject of harsh criticism
from a number of quarters.8 This lack of critical distance
appears to stem from Metz's assumption that psychoanalysis
is a science, and from the assumption that the methods of
interpretation of a patient's responses to the probing of a
psychoanalyst can be applied to the analysis of a spec
tator's reactions to films. However, in the case of cinema,
Metz does not have the benefit of the responses of the film,
the author of the film, or the spectator.
In view of all this, what Metz has to say in "The
Imaginary Signifier" about the guilt-ridden spectator/voyeur
who is secretly entering movie theatres as if visiting
brothels in a symbolic search for his lost penis, may be
illuminating in terms of coming to know Christian Metz but
not of how films communicate. Metz has again ended up in a
dialogue with himself.
Film Content and the
Linguistic-Psychoanalytic
Approach
When Metz talks about the spectator relating to the
screen as if to a mirror-like surface through which rapport
is established with a Lacanian symbolic "other," he is
suggesting that the spectator relates to a content of films
__________________________________________________________1 3fil
which consists of images and sounds that make up a motion
picture film. He is suggesting that a relationship exists
with what I have identified in this study as an important
aspect of the reality of film.
However, this is another of the deceptive qualities of
Metz's work. The "content" which his linguistic-psycho-
analytic approach actually sets out to investigate is the
matter of the signifier in Hjelmslev's sense.
Here it must be noted that Hjelmslev claims that
language manifests itself on three basic levels: form,
matter (or purport), and substance.9 Form is the pure
network of relations which define a set of elements. Matter
is the initially amorphous instance in which form is
manifested. Substance is the configuration which is
obtained by inscribing form in matter.
These levels of language are examined by Hjelmslev in
terms of another major distinction: expression-content.
Expression and content are designations of the functives of
the linguistic sign. The entity called the linguistic sign
is generated by the union between expression and content.
Expression and content necessarily presuppose each other.
An expression is expression only by being a content of
expression.
Then, form, matter, and substance are related to
expression and content, allowing Hjelmslev to distinguish in
language six fundamental levels: form of expression, form
1IQI
of content, matter of expression, matter of content, sub
stance of expression, and substance of content.10
Armed with such distinctions and levels, Hjelmslev
moves into even more abstract territory employing a maze of
terminological innovations in which he fails to show the
relevance of all these distinctions in describing a real
language.11 Hjelmslev's concept of language as a purely
deductive system and his inconsistent use of an extravagant
and forbidding terminology has been criticized rather
severely, especially by American linguists.12
For his new approach, Metz takes Hjelmslev's concept of
language, his terminology, and his distinctions, and applies
to them Jacques Lacan's version of Sigmund Freud's
psychoanalytic concepts and procedures.
In "The Imaginary Signifier" Metz provides no clues
whatsoever on how this concoction of Hjelmslev-Lacan-Freud
will deal with the images and sounds making up the content
of motion picture films shown in movie theatres. If Metz .
does not intend to deal with this aspect of the reality of
film, where do his comments about the voyeur/spectator fit
in his new approach? One has great difficulty imagining
how a search for the lost penis can be accomplished in
Hjelmslev's sense.
The Film As Art
It is obvious that in "The Imaginary Signifier" Metz
is not interested in examining film as an art form. But,
_______________________________________________ 140
in applying psychoanalytic concepts to film he is forced to
deal with a subject which has been one of the major concerns
of traditional film aesthetics: the framed' movie image and
the spectator's relationship to it.
Realist and formalist approaches to traditional film
aesthetics have used metaphors such as "window," "mask,"
and "frame" in describing the functions of the photographed/
projected image. Thus, the movie mask/frame has been
regarded, among other things, as a device for editing,
delimiting, organizing, composing, and ordering reality for
the benefit of the spectator.
Metz adds to the extensive literature on this subject
new metaphors and analogies and he examines the spectator's
response in terms of such devices. His contribution is the
concept of the movie frame (the screen image) as a "mirror"
in the Lacanian sense, and the "imaginary relationship"
between the spectator and what is reflected in the "mirror"
as a kind of dream.
This seems to be an important contribution and a step
in a useful direction. Finally, Metz appears to come to
grips with the reality of film.
Unfortunately, upon closer examination this does not
prove to be the case.
Through the concept of the mirror and the "imaginary
relationship" of the spectator, Metz is merely re-stating
in terms of Lacan's concepts and terminology what has been
XAll
examined in traditional film aesthetics as far back as 1916.
Generations of writers on the cinema, including such names
as Hugo Munsterberg, Ilya Ehrenburg, Hugo Mauerhofer, and
Suzanne Langer, have spoken of the dream-like quality of the
relationship between the spectator and projected film: the
relative passivity of the spectator in the darkened theatre
which brings about a receptiveness which induces a
suspension and a withdrawal from reality during which the
unreal is somehow accepted in a way as real.
It is true, of course, that the conventional approaches
to film aesthetics have left many questions unanswered. For
example, exactly what is involved in the heightened
receptiveness of the spectator and how that which is
obviously (logically) unreal comes to be accepted as real
has never been explained in a very satisfactory manner. It
is also true that the works of the conventional approach
to film aesthetics very often contain passages which are
too lyrical and "impressionistic."
But, stripped of its Lacanian psychoanalytic jargon
and rephrased in more conventional terms, Metz’s approach
adds nothing to what has been said already about film as
an art form. Metz merely better disguises his "impression
ism. "
Since reference has been made to Metz's metaphors of
the screen image being a "mirror," and the spectator's
experience being a "dream," it must be pointed out again
142
that the entire linguistic dimension of Metz's work is
based on the analogy that film is like a language. Metz's
problems are those of a man who attempts to develop a
science which is based on analogies and metaphors.
The Individual Film As
a Unique Determination
In Film Language Metz proclaimed that the image is a
hapax, a unique determination, and that it is not possible
to create a "dictionary" of images since there is an
infinite number of them. In Language and Cinema he admitted
that the codes tell us nothing about the film as a whole
and he proceeded to demonstrate that what accounts for a
film's organizing principles is the textual system. This
system is a unique mixture of codes and as such it is
unrepeatable. The manner in which these codes combine to
form the unique system was something which Metz failed to
determine. His failure prevented him from dealing in any
meaningful manner with "filmic writing," the last major
undertaking in Language and Cinema, and seriously weakened
the value of the codes as keys to cinematic signification.
Certainly, counting and recounting codes within a film
is not a very productive undertaking because codes are
keyboards and of themselves they indicate nothing about how
they can be used to produce meaning. In the absence of a
system for knotting these codes into the unique systems
which make up the individual film, the codes cannot help
___________________________________________________________ 143
us deal with signification as it applies to the complete
film.13
In "The Imaginary Signifier" Metz tries to establish
a case for codes of the dream but again he fails to indicate
how these codes combine to account for the film as a whole.
Freud, to whom Metz has turned for assistance, is not
of much help when it comes to accounting for the uniqueness
of a given work of film art. This is so because Freud was
unable to explain how the process of displacements (replace
ment of blocked object-cathexes with substitutes) often
leads to what he called a "sublimation"; the deflection of
psychic energy into the creation of works or the pursuit of
other intellectual goals.14 For Freud "sublimation"
remained a mystery, the mystery of creation. For Metz the
sublimation of the codes of the dream into meaningful and
unique works of cinema also remains a mystery. But, unlike
Freud, who admitted his inability to explain sublimation,
Metz remains absolutely silent about this very serious flaw
of his analytic procedure.
What makes Metz's silence on this matter even more
disturbing is his eloquence on the subject of changing his
mind about the existence of textual systems in films. He
announces that one of the contributions of psychoanalysis
to linguistics is that he no longer believes that each film
has a textual system, or even a fixed number of such
systems as he claimed in Language and cinema. In fact, the
_________________________________________ 144
concept of the textual system seems to be abandoned. Now
he sees the existence of a kind of "working conveniences,"
or "sorts of blocks of interpretation" which constitutes a
new kind of infinitely "thick" textual systems which are
created by the analysis as a result of a mysterious force
labelled by Metz "significatory pressure" (IS, p. 50).
The result of this change in Metz's procedures is that
now there is nothing left to account for the uniqueness of
the individual film. Not even the mystery of creationI
Everything has come now under the control of the analyst.
The codes are his creation and so is the system. The
analyst does not care to explain how the codes combine to
create the system. There is only the matter of the
"significatory pressure" as a faint ray of hope that some
thing exists which may be independent of the analyst. But
Metz is very cryptic about this "pressure" and he drops the
subject after mentioning it only twice.
The Sociocultural Roots of Film
The introduction of psychoanalysis to the structural
linguistic approach for the study of cinematic signification
could loosen upon Metz1s method the "myriad winds of cul
ture" which he avoided so carefully in Film Language and
Language and cinema. This is so because Freud related the
development of the human psyche to a large number of
sociocultural influences, and because he incorporated in
his psychoanalytic procedures numerous insights on the
.............................. 145
nature of artistic activity which point to the social
function of art.15
The social function of art, as perceived by Freud,
involves the
. . . transmission of the artist's ideas and psychic states by the
use of symbols capable of carrying both conscious and unconscious
stimuli which together evoke in the appreciator a combined
intellectual and emotional response. Their power is enhanced by
their patterns (artistic form) especially when these approximate
the patterns of the basic human experiences which both artist and
audience have as their common heritage.16
Some key concepts of Freudian psychoanalysis which deal
with the relationship between the individual and the exter
nal world are those of Ego, Superego, Dream, the Reality
Principle, and Identification. Ego is the psychological
system which is governed by the "reality principle" and
which transacts between the person and the outside world.
Superego is the moral and judicial dimension of personality.
It develops out of the child's assimilation of his parents’
standards, and enables the person to control his behavior
and adjust it to what is socially acceptable. The impulses
most often controlled by the Superego are those of sex and
aggression because they are considered to be the most
dangerous to the stability of society. A dream is a
succession of mental images having as its function the
reduction of tension by reviving memories of past events and
objects associated with the gratifications of needs. The
"reality principle" is a mechanism of the Ego through which,
in the interest of reality, there is a postponement in the
1 _ 4 _ 6 I
discharge of energy towards the satisfaction of needs until
the actual object, which will be able to satisfy the needs,
has been discovered or produced. Serving the "reality
principle" is the "secondary process" which consists of
discovering reality through a plan or action developed by
thought and reason. Identification is the incorporation of
the qualities of another person into one's personality. A
very common type is "goal-oriented identification" where a
frustrated individual achieves vicarious satisfaction by
identifying with someone who is successful in reaching his
goals.
Freud's analysis of the sociocultural aspects of
artistic activity examined the frame of mind of the artist,
the manner in which the artist's vision is reflected in the
work of art, and the nature of the relationship between the
artist and the appreciator of works of art.
In terms of the reality of film, it seems that a
complete and proper utilization of Freudian psychoanalysis
in the study of cinematic signification should take into
account all of the important elements involved in the com
munication process: artist, work of art, appreciator of
the work of art, and social setting.
However, in his linguistic-psychoanalytic approach Metz
banishes the artist/author, does not treat film as art, and
he ignores the social setting. Also, he substitutes the
analyst for the appreciator/spectator, focuses only on
_ __________________________________________ 1. 4. 7
selected psychoanalytic concepts, and then he manipulates
these concepts out of context of the Freudian-Lacanian
approach in order to retreat into the dialectical abstrac
tions of structural linguistics.
Of interest in this section is that Metz manipulates
certain psychoanalytic concepts and that this act eliminates
the sociocultural roots of the cinema. This manipulation is
most evident in connection with the analogies of Screen/
Mirror and Film/Dream.
A major problem seems to be that Metz cannot claim that
the projected movie image is a true mirror in the Freudian-
Lacanian sense. He admits this and explains in considerable
detail the many ways in which the screen image is not a
mirror. But, since he insists on using this psychoanalytic
concept without being able to refer to a relationship with a
mirror, he builds an elaborate case for the analogy of the
screen image being like a mirror.
The relationship of the individual with the mirror, as
described by Lacan, occurs within a sociocultural setting
and accounts for the development of the ego and the sense
of identity which a person has as a social being.
Metz claims that the relationship between the spec
tator and the mirror-like screen is a very special one.
The spectator, he says, has already known the true mirror
and his ego is formed. The screen/mirror has nothing to do
with the ability of the spectator, in terms of real life,
_____________________________________________148
to make the transition from the imaginary to what Lacan
calls the "symbolic" level.
The analogy of the mirror-like screen is manipulated
by Metz so that he can prove that the spectator identifies
with himself. This, in turn, is used by him to point out
that while the spectator is in this frame of mind signifi
cation in the cinema springs out of the "presence" of the
spectator perceiving the photographed (and projected)
object as absent.
From this point on the "psychoanalytic inspiration" is
discarded and Metz lapses into familiar structural linguis
tic methodology. Close examination of the path which he
takes in order to arrive at the conclusion that significa
tion is generated out of the presence of perceiving some
thing as absent indicates that he could have achieved the
same result without the aid of psychoanalytic concepts. In
this respect the use of psychoanalysis is entirely super
ficial and deceptive. Metz is merely decorating with
psychoanalytic jargon what he has said numerous times in
Film Language and Language and Cinema.
The Screen-Mirror analogy as used by Metz does not
permit him to indicate how social influences and aspects of
culture can be reflected in the projected film, and how the
spectator (or even the analyst) can relate to them. Metz's
failure to deal with the sociocultural aspects of the
reality of film is very disappointing in view of the fact
14 9 1
that Lacan, with his emphasis on symbolic aspects of signif
ication, offers him an opportunity to involve this reality.
It is true that Metz mentions a number of times that in
viewing a film the spectator makes the transition from an
imaginary to a symbolic level. But Metz does not indicate
how this occurs or what features of the perceived film
allow the spectator to make this transition.
As indicated earlier, Metz claims that films are like
a dream and that the spectator relates to them as if
relating to a dream. Metz even claims that as a film
theorist he relates to the linguistic-psychoanalytic
approach as if relating to a dream. He says, for example,
that the central question which "The Imaginary Signifier"
tries to answer is a dream/question. He also claims that
the entire essay is the interpretation of that dream/
question and that he examines the matters of "contribution,"
"Freudian," and "cinematic signifier" by associating freely
between them by following the method suggested by Freud in
his works on dream interpretation (IS, p. 28).
In the passages where he talks about aspects of
voyeurism and fetishism in the relationship between the
spectator and film, Metz engages in a rather extensive
analysis of his own "dream" of wanting to examine cinematic
signification. The "textual system," which becomes the
target of his investigation, is claimed to possess a
"latent" quality which, according to Freud, operates on
___________________________________________ 15H
both the unconscious and the preconscious level. Therefore,
this latent quality of the "textual system" must be sub
jected to analysis and interpretation similar to the one
applied to dreams (IS, p. 3 8).
This dream approach to film theory is disappointing in
terms of the sociocultural aspects of the reality of film
because Metz rejects the "dream analysis" of all the
aspects of film which are in any way related to what he
calls "the story." This eliminates the analysis of the
images and sounds that make up a film along with characters,
events, settings, time periods, and the orchestration of
these elements. All of this is rejected because, according
to Metz, "the story" is "indifferent to the cinematic
signifier" (IS, p. 40).
Since Metz has never been able to present a convincing
case that there can be a distinction in the cinema between
signifier and signified, one could argue that "the story" is
not merely a signified and it cannot be rejected for pur
poses of dream analysis. Of course, Metz’s dialectical
polarities and transformations do not permit arguments of
this kind. The circularity of his approach permits him to
accept everything and nothing and to proceed in any direc
tion he chooses to proceed.
However, with "the story" eliminated from dream analy
sis, he is unable to deal with anything which is associated
with matters of ideology. By "ideology" what is meant is
151
not just political orientation but the system of ideas and
connotations which human beings build up and which provide
them with a sense of direction in life and a guide for their
actions.17 To put it in psychoanalytic terms, ideology is
the orientation provided to the individual by the Superego.
Specifically, Metz's dream analysis is unable to deal
with the following aspects of what is called here ideology:
1. Transmitted and inherited through the family group
aspects of ideology.
2. Social group and social standing aspects of
ideology.
3. Ethnic and racial identity aspects of ideology.
4. Political and religious aspects of ideology.
5. Features of ideology which are related to period
in history.
6. Features of ideology which are related to geo
graphic location.
7. Sexual identity aspects of ideology.
8. "Technical" aspects of ideologies such as their
high degree of dependence on inner evidence and rejection of
empirical proof, their durability, and their inflexibility.
9. Aesthetic aspects of ideologies such as people's
concepts on the nature of beauty, pleasure, etc.
Despite the rejection of all of the above dimensions of
ideology, in his dream analysis Metz does focus upon
something which is related to the reality of life and the
reflection of that reality in ideology: the sexual act and
______________________________________________ __________ 3_5_2
the concerns which human beings have in connection with it.
The problem is that Metz gives the impression that the
sexual act is not just the most important element in human
existence, but the only one which is relevant in connection
with cinematic signification.
It is true that Freud believed that all dreams contain
a very large number of disguised representations of
repressed sexual needs and images of the tension-reducing
objects which represent wish-fulfillments. But, to make the
sexual act, and the "complexes" which Freud associated with
it, the most important element in dream analysis and the
key to cinematic signification, represents a distorted view
of the Freudian approach and an embarrassingly naive
proposition as to where the heart of cinematic signification
lies.
The great reliance on the analysis of sexual matters
seems inappropriate for another reason which has to do with
the structural linguistic aspect of Metz’s approach. The
search for the lost penis, literal or symbolic, is related
to the thematic material of the rejected dimension of "the
story," and not to the matter of the signifier in
Hjelmslev's sense.
153
FOOTNOTES
Anthony Wilden, System and Structure (London: Tavistock,
1972), p. 261.
2Metz, "The Imaginary Signifier," pp. 41-42, 45.
3Calvin Hall, A Primer of Freudian Psychology (New York:
World Publishing Co., 1963), pp. 24-29.
“Buscombe, et al., "Psychoanalysis and Film," p. 120.
5Wilden, System and Structure, pp. 5-6.
6Wilden, pp. 14-15.
7Buscombe, et al., "Psychoanalysis and Film," p. 120.
8For example, see Julia LeSage, "The Human Subject:
You, He, or Me?" Screen, 16, No. 2 (Summer, 1975), 77-82.
9Louis Hjelmslev, Prolegomena to a Theory of Language,
Revised English Edition (Madison: University of Wisconsin
Press, 1969).
10Stephen Heath, "Metz's Semiology: A Short Glossary,"
Screen, 14, No. 1/2 (Spring/Summer, 1973), 222.
11R. R. K. Hartmann, The Language of Linguistics (Tttbingen:
Tilbinger BeitrSge zur Linguistik, 1973), pp. 57-59.
12David Crystal, Linguistics (Harmondsworth: Penguin,
1971), p. 246.
13J. Dudley Andrew, "Film Analysis or Film Therapy: To
Step Beyond Semiotics," Quarterly Review of Film Studies, 2,
No. 1 (February, 1977), 33-41.
14Hall, Primer of Freudian Psychology, pp. 82 — 83.
15References to the nature of artistic activity appear
throughout his works. In addition to these references,
Freud conducted studies on Dostoevsky, Goether, Leonardo da
Vinci, and Shakespear.
16Louis Fraiberg, Psychoanalysis and American Literary Criti
cism (Detroit: Wayne State University Press, 1960), 44-45.
1 7Paul Schilder, Psychoanalysis, Man, and Society (New
York: W. W. Norton & Co., 1951), 61.
154
SUMMARY AND CONCLUSIONS
Christian Metz is considered to be one of the most
important film theoreticians of our times, who has made
great contributions towards the establishment of a scien
tific approach to the study of how films communicate.
Speaking of his own work Metz has declared that he inter
vened in the course of film studies in order to help
eliminate the impressionistic excesses of conventional film
criticism and theory.
To achieve the task which he set for himself, Metz
initially attempted to apply to the study of film communi
cation the principles of structural linguistics on the
assumption that film is a kind of language. After more than
a decade of efforts along this line, he turned to an
approach which combines structural linguistic theory and
psychoanalysis. This time the added assumptions were that
the unconscious mind is structured like a language and that
film is like a dream.
The purpose of this study was to determine to what
extent Metz has contributed something of value towards our
understanding of how films communicate; something which
indeed eliminates the alleged impressionistic excesses of
155
conventional film theory. To deal with this problem I felt
that Metz's work should be confronted with "the reality of
film" as it pertains to the fiction narrative film which has
been the focus of Metz's investigations.
The basic elements of this reality are that film is an
art, that individual films have an author, and that they are
cultural objects to which an audience relates on a personal
level and as part of social interaction.
The uniqueness and significance of this study hopefully
lies in that an attempt was made to examine fairly closely
three important works by Metz in order to determine-whether
or not Metz really studies film and not something which
he claims to be film when in reality it is not. The reason
for taking this approach was the suspicion that the basic
assumptions on which Metz bases his investigations inevit
ably lead him to an analysis of something which is not
film, with the result that his work does not contribute
anything of value to our understanding of how films
communicate.
In this study an attempt was made to confront Metz
with his own words through extensive quotes which indicate
that he is fully aware of the reality of film. This was
done to focus attention on a certain pattern which was
discerned in his works involving an acknowledgement of the
reality of film followed by the manipulation of the elements
he borrows from structural linguistics and psychoanalysis
for the development of analytic procedures which do not
______________________________________________________________________________________' _____________________________________________________ 13£
permit him to deal with this reality.
The fairly extensive quotes from his works were also
thought to be necessary because I felt that a considerable
number of the existing studies and references to his work
exhibited a lack of concern for the inconsistencies, and
flaws of his analytic procedures. As a result, most\of
these existing studies and references have probably exag-
\
gerated the importance of Metz's contribution to film
theory, and they may have been detrimental to the work of
those who have depended on these studies in conducting their
own investigations and analyses of films.
In the opening remarks of the first chapter, which
was entitled "The Tortured Mind," I pointed out the reasons
for Metz's "intervention," and I tried to describe what Metz
sees as his goal in creating analytic procedures which
were based initially on structural linguistics and later
on, linguistics and psychoanalysis. A second (section in
that chapter consisted of extensive quotes from the primary
sources for this study which indicate Metz's awareness of
the reality of film. The major portion of that chapter
dealt with some very disturbing features of Metz's overall
approach and style which explain in part why he is unable
to deal with the reality of film.
All of these difficulties seem to stem from the fact
that in Film Language and in Language and Cinema Metz bases his
approach on the analogy that film is a kind of language,
157
and then he proceeds to use the concepts and methods which
apply to the study of language in order to develop his
analytic procedures. The fact that film is not a language
and that there are very fundamental differences between
film and language, which Metz acknowledges fully, places
him in a very difficult position and imposes upon his
supposedly precise and scientific procedures an impression
istic quality which is very similar to the one he attempted
to eliminate with his "intervention." Considering that Metz
has a very good background in film theory and criticism and
that he seems to be an intelligent and sensitive person, it
is quite astonishing that he fails to realize how imprecise
and impressionistic his approach turns out to be.
Especially in Film Language, Metz seems to be literally
in an agonized state of mind battling against the windmills
of his own creation. Chapter I attempts to point to the
great uncertainty, timidity, hesitancy, vagueness, and
circularity of his thought. The chapter also gives examples
of some of the impossible distinctions which he attempts to
make, and some of the problems he encounters by an outpour
ing of neologisms and by creating numerous and intersecting
elements, levels, and ranks for models of a whole which he
never constructs.
Since his Grande Syntagmatique is supposed to be one of
his major achievements, this chapter includes a number of
comments on this model of the major analytic units for the
158
examination of cinematic intelligibility. As I pointed
out, this model is incomplete in that it cannot account
for all possible types of syntagmatic relationships, it
is applicable to certain epochs only in the development of
film, and in general, it seems to be nothing more than a
presentation in linguistic jargon of types of shots and of
relationships between shots which have been discussed in
the works of the conventional film theorists decades before
Metz appeared on the scene.
As far as his impossible distinctions is concerned, it
seems that the most troublesome are those of denotation/
connotation, cinematic/filmic, and code/message. Metz is
simply unable to establish that there can be valid distinc
tions between these elements. And yet, he bases much of
what he is attempting to do in Film Language and Language and
cinema on> these impossible distinctions.
The major reason for his difficulties in this area
seems to be that, as far as the motion picture image is
concerned, he cannot establish that it is possible to
distinguish between denotation and connotation. The same
seems to apply to the sound track elements although he does
not examine the problems of the sound track in any detail.
Unable to focus properly on the object of his study,
and influenced by what he considers to be methodological
features of structural linguistics and of the dialectical
approach, Metz examines matters through what I call a
159
"drifting" point of view. In this approach the study of
cinematic signification acquires the flavor of an intellec
tual game with a seemingly infinite number of continuously
changing rules. In this game, the reader is almost always
asked to go along with whatever argument Metz is presenting
and not to pay attention to disturbing contradictions and
inconsistencies. Metz simply asks us to have faith in him
and to trust that improvements will come about in some
unspecified time in the future.
This chapter also deals briefly with two more very
disturbing qualities of Metz's overall style: His proof
via manifesto and the lyrical quality of his writing. The
almost total lack of documentation which plagues his work
when it comes to supplying evidence that signification in
motion pictures occurs as described by him, is concealed by
the fact that he presents the premises of structual linguis
tics as self-evident truths. In this manner trying to deal
with Metz is equivalent to engaging in an argument with a
religious zealot. Coupled with what is a very obvious
quality of self-affectation in all of his writings, this
proof via manifesto technique and the accompanying lyrical
style give to his works the quality and flavor not of
scholarship but of some sort of fiction or poetry.
The second chapter of this study, "From Movies to
Cine-Text," deals with Metz's "reduction" of film into
something which is unrelated to the "filmophanic film"; the
160
film as experienced by the spectator during projection.
Metz’s "reduction" takes the form of banishing the film
author, of failing to take into account the content of the
filmophanic film, and of constructing and dealing with a
product of his mind which has nothing to do with what an
audience perceives on the screen during the projection of
a motion picture film.
The interesting and at the same time very deceptive
thing is that Metz calls the product of his techne a
"duplicate" of the natural object (the filmophanic film),
and that he claims that the study of this "duplicate" will
enable us to understand how films communicate.
However, a "duplicate" is a thing that exactly
resembles another in appearance, pattern, or content. A
"duplicate" is a copy. Metz provides no evidence whatsoever
that his systems, structures, and activities between these
systemic elements, which are creations of the analyst, are
in any way related to the expressive elements of motion
picture films. Also, he provides no evidence whatsoever
from communication research or from any other source that
the film audience engages in the kind of mental activities
which he describes and which could explain how signification
is achieved in films.
As for the manner in which the codes and sub-codes
function to create the "cinematic language system," Metz
provides no definite plan which sets the rules of behavior
161
of the various systemic elements which are involved. The
activities of the various systemic elements appear to be
made up by Metz as he goes along. The matter of the cine
matic system becomes even more troublesome later on because
Metz indicates that the system of each film is unique and
unrepeatable. If that is the case, much of what he has
said about systems cannot possibly be a blueprint for the
cinematic systems in general which applies to all films.
Chapter III, "Metz and the Film Audience," examined
Metz's approach in terms of that aspect of the reality of
film which involves the audience's relationship with the
filmophanic film.
In terms of the limitations of the phenomenological-
structural linguistic aspects of Metz's work, in this chap
ter it was pointed out that Metz and the film audience are
relating to two different things: Metz is relating to the
"duplicate" product of his techne, and the audience is relat
ing to the reality of film. Nothing that Metz proposes is
in any way related to the emotional response involved in
experiencing a film as projected in a motion picture
theatre. He fails to examine the relationship between the
audience and the signifier. He also fails to examine the
relationship between the audience and anything even remotely
associated with the connotation and the signified. The
synchronic aspects of the structural linguistic approach
trap Metz in such a way that he cannot account for any
162
dimension of film which involves the diachronic aspects of
the reality of film and of the experience of viewing films
as they unfold diachronically.
As for the fact that film is an art form, Metz fails
entirely to account for any aspect of the relationship
between the audience and the filmophanic film which explains
the manner in which human beings relate to a work of art.
In describing the manner in which an analyst such as Metz
relates to the mental "duplicate" of film, Metz also fails
entirely to take into account that film is an art form. In
Film Language and Language and Cinema, there is simply no
analytic feature and no systemic relationship which in any
way accounts for the analyst's relationship with any aspect
of artistic inspiration, creation, or enjoyment.
In Chapter IV, "The Severed Roots," it was pointed out
that Metz fails to deal with film as a cultural object
functioning within the social process of communication.
Metz's exclusion of the sociocultural dimension of film
was accomplished by not examining the role of visual percep
tion in film communication, by excluding whatever may
function as a cultural code, by failing to deal with socio
cultural influences in producing, exhibiting, and perceiving
films, and in general by choosing to deal with film as a
linguistic object rather than as an art form.
An important matter which this chapter also tried to
point out is that as a film theorist Metz cannot function in
1£l 3 j
a social vacuum. He must be approaching the study of film
in terms of what is culturally meaningful to him. And yet,
in Film Language and Language and Cinema, Metz makes no refer
ence whatsoever to the sociocultural perspective from which
an analyst like himself views and analyzes film, and he
makes no provisions in his analytic procedures for such a
viewpoint. The "duplicate" of the natural object is
analyzed in an ideological and sociocultural vacuum.
In Chapter V, "The Imaginary Signifier," there is a
brief description of Metz's new psychoanalytically inspired
approach, certain comments on the relationship between his
earlier phenomenological-linguistic approach and the new
psychoanalytic approach, and an examination of the problems
he encounters in terms of the new approach and the reality
of film.
His new psychoanalytically inspired approach turns out
to be not so new after all. The old phenomenological-
structural linguistic approach, with all its weaknesses such
as the impossible distinctions, the circularity of thought,
the great vagueness and the impressionistic qualities, has
been incorporated into the new approach. In this manner the
ill-fated analogy that film is like a language is used now
to support the new analogies that the unconscious mind is
structured like a language and that film is like a dream.
These analogies, however, combined with Metz's lack of
interest in the filmophanic film prevent him from focusing
164
on the object of his study and make the new approach as
impressionistic as the earlier one. The "film script"
which Metz intends to study psychoanalytically is a
confused and vague notion which becomes progressively even
more vague and confused as he attempts simultaneously to
enlarge and to limit it in a futile effort to bring it
under control. In this effort the "matter of the signifier"
remains hopelessly entangled in the very abstract distinc
tions "in Hjelmslev's sense." No attempt is made to
describe what is manifested in the "manifest filmic mate
rial," and how this manifestation comes about. The
"pertinent features" of the signifier are shrouded in
mystery.
His dependence on the ideas and writing style of Lacan
and on a view of Freudian psychoanalysis which, in its lack
of critical perspective, resembles the vulgarized interpre
tations of Freud's work in popular magazines, introduce to
the linguistic-psychoanalytic approach a great lack of
precision and make it even more impressionistic than the
approach taken in Language and Cinema.
•In terms of specific aspects of the reality of film,
in "The Imaginary Signifier" Metz employs some familiar
methods of acknowledging these aspects and then rejecting
them in the creation of the analytic procedures. The author
of the film is banished again with the excuse that dealing
with the author may lead to a "nosographic" or other type
of characterology. The content of film in terms of images
__________ 165
and sounds is rejected again. Despite the fact that he
uses the film/dream analogy which lends itself to a
psychoanalytic approach of film as an art form, Metz again
does not wish to deal with film as an art. He repeats in
"The Imaginary Signifier" the admission he made in Film
Language and Language and Cinema that film is a hapax. But,
just as he failed in his earlier approach to demonstrate how
the codes and sub-codes combine to produce the unique and
unrepeatable system of a film, in his new approach he fails
again to indicate how the codes (now re-named "codes of the
dream") combine to produce meaning and the unique quality of
a given film. Although the individual is viewed in psycho
analysis as a social being, and Freud analyzed artistic
activity in terms of its sociocultural dimension, Metz
attempts to use psychoanalytic concepts for the examination
of cinematic signification which ignore the sociocultural
roots of film.
In terms of the reality of film a very disturbing
aspect of his new approach is the handling of the matter of
spectator identification. According to Metz the spectator
identifies with the screen characters, the camera, the
screen, and with his own self. In examining this relation
ship Metz has an opportunity to make use of psychoanalysis
in order to deal with the reality of film. Instead, he
manipulates the psychoanalytic principles involved and the
concepts of linguistics which he has retained from the
166
earlier approach in order to retreat once more to the
abstract domain of structural linguistics. In this domain
the reality of film has no place. The only relationship
which is really examined is that between an analyst like
Metz and the analytic procedures which he invents for the
mental "duplicate" of the natural object.
Since the "duplicate" which he creates is not a
duplicate of film at all but an analytic entity for some
thing which is not related to the reality of film, Metz's
entire effort represents a waste of time as far as helping
us understand how films communicate.
The study of Metz's "duplicate" seems to have some
value only in terms of the pleasure derived from manipulat
ing a product of one's imagination which pulsates with that
special excitement which characterizes mental structures and
from sharing with others the experience involved in
constructing and manipulating such structures.
It seems to me quite possible that one could use the
Metz "method" to fill thousands of pages with equally
fascinating mental acrobatics describing the transformations
and contaminations of codes, sub-codes, and systems without
reference to film or anything else of any specific nature.
The question is, Why .does Metz involve the cinema in such a
tortuous and unproductive enterprise? It is conceivable the
answer lies in that, despite his repeated professions of
great love for the cinema, Metz may not have real respect
162
and affection for the seventh and liveliest art. Before
attempting to analyze his written treatises on film, perhaps
one should pressure Metz to elaborate a little more on a
statement which he made once in response to questions put
to him by Cinethique. Said Metz:
I do not consider the film-maker as a creative individual
at all.1
168
FOOTNOTES
1Cinethique, "On Langage et Cinema-,” p. 209.
16 9
170
B IB L IO G R A P H Y
Altaman, Charles F. "Psychoanalysis and Cinema: The Imag
inary Discourse." Quarterly Review of Film Studies, 2
No. 1, August, 1977, 257-272.
Andrew, Dudley J. The Major Film Theories: An Introduction.
Oxford: Oxford University Press, 1976.
_________________ . "Film Analysis or Film Therapy: To Step
Beyond Semiotics." Quarterly Review of Film Studies, 2,
No. 1, February, 1977, 33-41.
Barthes, Roland. Elements of Semiology. London: Jonathan
Cape, 1969.
Baudry, Jean-Louis. "Ideological Effects of the Basic
Cinematographic Apparatus." Film Quarterly, 28, No. 2,
Winter, 1974/75, 39-47.
Bauxandall, Lee. Marxism and Aesthetics: A Selective Annotated
Bibliography; Books and Articles in the English Language. New
York: Humanities Press, 1968.
Bellour, Raymond. "The Obvious and the Code." Screen, 15,
No. 4, Winter, 1974/75, 7-17.
Benoist, Jean-Marie. "Structuralism: A New Frontier."
Cambridge Review, 93, No. 2204, October 22, 1971, 10-17.
Bettetini, Gianfranco. The Language and Technique of the Film.
The Hague: Mouton, 1973.
Bizet, Jacques-Andre. "Les Structuralistes, la Notion de
Structure et l'Esthetique du Film." La Pensee, 137,
February, 1968, 38-50.
Bochenski, I. M. The Methods of Contemporary Thought. Dordrecht,
Holland: D. Reidel Publishing Co., 1965.
Boon, James A. From Symbolism to Structuralism: Levi-Strauss in a
Literary Tradition. Oxford: Blackwell, 1972.
Bowlt, John. "Russian Formalism and the Visual Arts." 20th
171
Century Studies, 7/8, December, 1972, 131-146.
Brewster, Ben. "Structuralism in Film Criticism." Screen,
12, No. 1, Spring, 1971, 49-58.
Broekman, Jan M. Structuralism: Moscow, Prague, Paris. Boston:
D. Reidel Publishing Co., 1974.
Burch, Noel. Theory of Film Practice. New York: Praeger,
1973.
Buscombe, Edward, and others. "Ideas of Authorship."
Screen, 14, No. 3, Autumn, 1973, 75-85.
, and others. "Psychoanalysis and Film."
Screen, 16, No. 4, Winter, 1975/76, 119-130.
_________________ , and others. "Why We Have Resigned from
the Board of Screen." Screen, 17, no. 2, Summer, 197 6,
106-109.
Casey, John. The Language of Criticism. ’ London: Methuen,
1966.
Cegarra, Michel. "Cinema and Semiology." Screen, 14,
No. 1/2, Spring/Summer, 1973, 129-188.
Chatman, Seymour. "Discussion of Gilbert Harman's Paper,
'Semiotics and the cinema. '" Quarterly Review of Film
Studies, 2, No. 1, February, 1977, 25-32.
Cinethique. "On Langage.et Cinema." Screen, 14, no. 1/2, Spring/
Summer, 1973, 189-213.
Comolli, J. L., and Narboni, J. "Cinema/Ideology/Criticism:
Examining a Critique at Its Critical Point." Screen,
13, No. 1 Spring, 1972, 120-131.
Cornforth, Maurice C. Marxism and the Linguistic Philosophy. New
York: International Publishers, 1966.
Crystal, David. Linguistics. Harmondsworth: Penguin, 1971.
Culler, Jonathan D. Structuralist Poetics: Structuralism, Lin
guistics, and the Study of Literature. London: Routledge &
Kegan Paul, 1975.
Daniel, Jack. "Metz's 'Grands^Syntagmatiqner^. '-Summary.and'-. -
Critique." Film Form, 1, No. 1, Spring, 1976, 78-90.
DeGeorge, Richard T. , and Fernande, M. , Eds. The Structural
ists from Marx to Levi-Strauss. Garden City, N. Y.: Anchor
Books, 1972.
1 7 4
Dembo, L. S., Ed. Criticism; Speculative and Analytical Essays.
Maddison, Milwaukee: The University of Wisconsin
Press, 1968.
Eckert, Charles W. "The English Cine-Structuralists." Film
Comment, 9, No. 3, May-June, 1973, 46-51.
Eco, Umberto. "On the Contribution of Film to Semiotics."
Quarterly Review of Film Studies, 2, No. 1, February, 1977,
1-14.
______________ . A Theory of Semiotics. Bloomington: Indiana
University Press, 1976.
Ehrenzweig, Anton. The Psychoanalysis of Artistic Vision and
Hearing. New York: George Braziller, 1965.
Ehrmann, Jacques. structuralism. .Garden City, N. Y.:
Doubleday & Co., 1970.
Eikhenbaum, Boris. "Problems of Film Stylistics." Screen,
15, No. 3, Autumn, 1974, 7-34.
Erickson, Stephen A. Language and Being: An Analytic Phenomenol
ogy. New Haven: Yale University Press, 1970.
Feibleman, James Kern. Scientific Method. The Hague: Nij-
hoff, 1972.
Fieandt, Kai von. The World of Perception. Homewood, Illinois:
The Dorsey Press, 1966.
Fischer, Ernst. Art Against ideology. London: Penguin Books,
1969.
Fraiberg, Louis. Psychoanalysis and American Literary Criticism.
Detroit: Wayne State University Press, 1960.
Freud, Sigmund. The Ego and the id. • ' London: Hogwarth Press,
1947.
Beyond the Pleasure Principle. London: Hog-
warth Press, 1948.
' _________ ___. • An Outline of Psychoanalysis. New York: W. W.
Norton & Co., 1949.
_____________________ Collected Papers. 4 vols. London: Hog
warth Press, 1950.
The Interpretation of Dreams. London: Hog-
warth Press, 1953.
113
Frondizi, Risieri. What Is Value? An Introduction to Axiology.
LaSalle, Illinois: Open Court Publishing Co., 1971.
Fyock, James A. "Content Analysis of Films: New Slant of
an Old Technique." Journalism Quarterly, 35, No. 4, 1968,
687-691.
Gardies, Rene. "Structural Analysis of a Textual System:
Presentation of a Method." Screen, 15, No. 1, Spring,
1974, 11-32.
Glucksman, Miriam. Structural Analysis in Contemporary Social
Thought; A Comparison of the Theories of Claude Levi-Strauss and
Louis Althusser. London: Routledge & K. Paul, 1974.
Graff, Gerald. "What Was New Criticism? Literary Interpre
tation and Scientific Objectivity." Salmagundi, 27,
Summer-Fall, 1974, 72-93.
Guzzetti, Alfred. "Christian Metz and the Semiology of the
Cinema." Journal of Modern Literature, 3, No. 2, April,
1973, 272-308.
Hall, Calvin S. A Primer of Freudian Psychology. New York: The
World Publishing Co., 1963.
Handy, Rollo. Methodology of the Behavioral Sciences. Spring
field, Illinois: Charles C. Thomas Publisher, 1964.
Handy, William J. , ed. A Symposium on Formalist Criticism.
Austin, Texas: The University of Texas Press, 1965.
Harman, Gilbert. "Semiotics and the Cinema: Metz and
Wollen. " Quarterly Review of Film Studies, 2, No. 1, Febru
ary, 1977, 15-24.
Harpole, Charles, and Hanhardt, John. "Linguistics, Struc
turalism, Semiology: Approaches to Cinema with a
Bibliography." Film Comment, 9, No. 3, May-June, 1973,
52-59.
Harre, R. An Introduction to the Logic of the Sciences. London:
MacMillan & Co., 1960.
Harris, Marvin. The Rise of Anthropological Theory; A History of
Theories of Culture. New York: Thomas Crowell Co., 1968.
Hartmann, Reinhard. The Language of Linguistics: Reflections on
Linguistic Terminology. Ttibingen: Ttibingen Beitr&ge zur
Linguistik, 1973.
174
Hayes, E. Nelson, and Hayes, Tanya, eds. Claude Levi-Strauss:
The Anthropologist as Hero. Cambridge: Massachusetts
Institute of Technology Press, 1970.
Heath, Stephen. "Film/Cinetext/Text." Screen, 14, No. 1/2,
Spring/Summer, 1973, 102-128.
_______________. "Metz's Semiology: A Short Glossary."
Screen, 14, No. 1/2, Spring/Summer, 1973, 214-226.
_______________. "Comments on 'The Idea of Authorship.'"
Screen, 14, No. 3, Autumn, 1973, 86-91.
_______________. "The Work of Christian Metz." Screen, 14,
No. 3, Autumn, 1973, 5-28.
_______________. "Film and System, Terms of Analysis, Part
I." Screen, 16, No. 1, Spring, 1975, 7-77.
_______________. "Film and System, Terms of Analysis, Part
II." Screen, 16, No. 2, Summer, 1975, 91-113.
Heaton, J. M. "Insight in Phenomenology and Psychoanalysis.'
Journal of the British Society for Phenomenology, 3, No. 2,
May, 1972, 135-145.
Heller, Louis G. , and Maoris, James. Toward a structural
Theory of Literary Analysis; Prolegomena to Evaluative Descrip-
tivism. Worcester: Massachusetts Institute for Systems
Analysis, 1970.
Henderson, Brian. "Toward a Non-Bourgeois Camera Style."
Film Quarterly, 24, No. 2, Winter, 1970—71, 2—14.
_________________. "Two Types of Film Theory." Film Quarterly
24, No. 3, Spring, 1971, 33-41 .
_________________. "Metz: 'Essais I' and Film Theory."
Film Quarterly, 28, No. 3, Spring, 1975, 12-33.
Hendricks, william O. Essays on Semiolinguistics and Verbal Art.
The Hague: Mouton, 1973.
Hill, Anthony. "A Structuralist Art?" 20th Century Studies,
3, May, 1970, 102-109.
Hirsch, E. D. Validity in Interpretation. New Haven: Yale
University Press, 1967.
Hjelmslev, Louis. Prolegomena to a Theory of Language. Madison,
Milwaukee: The University of Wisconsin Press, 1963.
175
Hjelmslev, Louis. Language: An Introduction. Madison, Mil
waukee: The University of Wisconsin Press, 1970.
Hobsbaum, Philip. Theory of Criticism. Bloomington: Indiana
University Press, 1970.
Hogg, James, ed. Psychology of the Visual Arts. Baltimore:
Penguin Books, 1969.
Ihde, Don. Hermeneutic Phenomenology. Evanston: Northwestern
University Press, 1971.
Jacobson, Roman, and Hale, Morris. Fundamentals of Language.
The Hague: Mouton, 1956.
Jameson, Predric. Marxism and Form: Twentieth-Century Dialectical
Theories of Criticism. Princeton: Princeton University
Press, 1972.
. The Prison-House of Language: A Critical Account
of Structuralism and Russian Formalism. Princeton, New
Jersey: Princeton University Press, 1972.
Jarvie, I. C. Movies and Society. New York: Basic Books,
Inc., 1970.
Jean, B., and Lewis, T. J. "Structural Linguistics and
Literature in France." Journal of the British Society for
Phenomenology, 2, No. 3, October, 1971, 27-36.
Kitses, Jim. Horizons West. Bloomington: Indiana University
Press, 1969.
Klein, Melanie. Psychoanalysis of Children. London: Hogwarth
Press, 1933.
_________________ . Contribution to Psychoanalysis, 1921-1945. Lon
don: Hogwarth Press, 1948.
Envy and Gratitude. New York: Basic Books,
Inc., 1957.
Kristeva, Julia. "The Semiotic Activity." Screen, 14,
No. 1/2, Summer, 1973, 25-39.
Kroeber, A. L. Culture: A Critical Review of Concepts and Defini
tions. New York: Peabody Museum of American Archaeol
ogy, 1952.
Kuntzel, Thiery. "The Treatment of Ideology in the Textual
Analysis of Film." Screen, 14, No. 3, Autumn, 1973, 44-
54.
176
Lacan, Jacques. "Some Reflections on the Ego." international
Journal of Psycho-Analysis, 34, 1953, 11-17.
_________________. Ecrits. Paris: Seuil, 1966.
_________________ . The Language of the Self: The Function of Language
in Psychoanalysis. Baltimore: The Johns Hopkins Press,
1968.
"The Mirror-Phase . " New Left Review, 51,
1968, 71-77.
_______________. "On Structure As an Inmixing of an Other
ness Prerequisite to Any Subject Whatever." The Lan
guages of Criticism and the Sciences of Man: The Structuralist
Controversy, eds. R. Macksey and E. Donato. Baltimore:
Johns Hopkins Press, 1970, 186-200.
_________________ . Les Quatre Concepts Fondamentaux de la Psych-
analyse. Paris: Seuil, 1973.
Lane, Michael, ed. Structuralism: A Reader. London: Jona
than Cape, 197 0.
Leach, Edmund. Levi-Strauss. London: William Collins & Co.,
1970.
Lepley, Ray, ed. The Language of Value. New York: Columbia
University Press, 1957.
Lepschy, Giulio C. A Survey of Structural Linguistics. London:
Faber, 1970.
LeSage, Laurent. The French New Criticism. University Park:
The Pennsylvania State University Press, 1967.
LeSage, Julia. "The Human Subject— You, He, or Me?" Screen,
16, No. 2, Summer, 1975, 77-82.
Levi-Strauss, Claude. structural Anthropology. New York:
Basic Books, Inc., 1963.
The Elementary Structures of Kinship.
London: Eyre & Spottiswood, 1969.
Lovell, Alan. "Notes on British Film Culture." Screen, 13,
No. 2, Summer, 1972, 5-16.
_____________. "Robin Wood— A Dissenting View." Screen, 10,
No. 2, March/April, 1969, 42-55.
Lovell, Terry. "Cultural Studies." Screen, 14, No. 3,
____________________________________________ 17 7
Autumn, 1973, 115-122.
Lyas, Colin, ed. Philosophy and Linguistics. London: Mac
millan & Co., 1971.
MacBean, James Roy. "Godard's 'Weekend,' or the Self-Crit
ical cinema of Cruelty." Film Quarterly, 21, No. 2,
Winter, 1968-1969, 35-43.
. "Politics, Painting, and the Language
of Signs in Godard's 'Made in U.S.A.'" Film Quarterly,
22, No. 3, Spring, 1969, 18-25.
"See You at Mao." Film Quarterly, 24,
No. 2, Winter, 1970-1971, 15-23.
. Film and Revolution. Bloomington: Indi-
ana University Press, 1975.
Macksey, Richard, and Donato, Eugenio, eds. The Languages of
Criticism and the Sciences of Man: The Structuralist Controversy.
Baltimore: The Johns Hopkins Press, 1970.
Matejka, Ladislav, and Titunik, Irwin, eds. Semiotics of Art:
Prague School Contributions. Cambridge, Mass.: Massachu
setts Institute of Technology Press, 1976.
Metz, Christian. Essais sur la signification au Cinema. Paris:
Editions Klincksieck, 1968.
__________________ . Langage et Cinema. Paris: Larousse, 1971.
_______"Jean Mitry's 'L'Esthetique et Psychologie
du Cinema.'" Screen, 14, No. 1/2, Spring/Summer, 1973,
40-88.
______ "Methodological Propositions for the
Analysis of Film." Screen, 14, No. 1/2, Spring,Summer,
1973, 89-101.
_____________. Film Language: A Semiotics of the Cinema. New
York: Oxford University Press, 1974.
Language and Cinema. The Hague: Mouton,
1974.
___________. "The Imaginary Signifier." Screen, 16,
No. 2, Summer, 1975, 14-76.
_______ . "L'incandescence et le code." Cahiers du
Cinema, 274, March, 1977, 5-22.
178
Miel, Jan. Jacques Lacan and the Structure of the Unconscious.
Yale French Studies, Nos. 36-37. New Haven: Yale
University Press, 1966, 104-111.
Moles, Abraham. Information Theory and Esthetic Perception.
Urbana: The University of Illinois Press, 1966.
Monaco, James. How to Read a Film. New York: Oxford Univer
sity Press, 1977.
Morick, Harold, ed. Wittgenstein and the Problem of Other Minds.
New York: McGraw-Hill, 1967.
Morris, Charles. Signification and Significance: A Study of the
Relations of Signs and Values. Cambridge: Massachusetts
Institute of Technology Press, 1970.
__________________. Writings on the General Theory of Signs. The
Hague: Mouton, 1971.
Morse, David. "Criticism As a Science." 20th Century Studies,
9, September, 1973, 32-50.
Mullahy, Patrick. Oedipus Myth and Complex. New York: Hermit
age House, 194 9.
Murray, John C. "Robin Wood and the Structural Critics."
Screen, 12, No. 3, Summer, 1971, 101-110.
STajder, Zdzislaw. Values and Evaluations. Oxford: Clarendon
Press, 1975.
SJowell-Smith, Geoffrey. "Cinema and Structuralism," 20th
Century Studies, 3, May, 197 0, 131—139.
_______________________ . "I Was a Star-Struck Structural
ist." screen, 14, No. 3, Autumn, 1973, 92-100.
Pasquier, Sylvain du. "Buster Keaton's Gags." Journal of
Modern Literature, 3, No. 2, April, 1973, 269-291.
Pettit, Philip. "Wittgenstein and the Case for Structural
ism." Journal of the British Society for Phenomenology, 3,
No. 1, January, 1972, 46-57.
_________________. The Concept of Structuralism: A Critical Analysis.
Berkeley: University of California Press, 1975.
Petrie, Graham. "Auteurism: More Aftermath." Film Quarterly,
27, No. 3, Spring, 1974, 61-62.
_iiaJ
Piaget, Jean. structuralism. New York: Basic Books, 1970.
Poirier, Richard. The Performing Self: Compositions and Decomposi
tions in the Languages of Contemporary Life. New York:
Oxford University Press, 1971.
Pool, Roger C. "Structuralism and Phenomenology: A Liter
ary Approach." Journal of the British Society for Phenomenol
ogy, 2, No. 2, May, 1971, 3-16.
Pryluck, Calvin. "Toward a Psycholinguistics of Cinema."
AV Communication Review, 15, No. 1, Spring, 1967, 54—75.
________________. "Structural Analysis of Motion Pictures As
a Symbol System." AV Communication Review, 16, No. 4,
Winter, 1968, 372-402.
________________ . "Motion Pictures and Language: A Compara
tive Analysis." Journal of the University Film Association,
21, No. 2, 1969, 46-51.
________________ . "The Film Metaphor-Metaphor: The Use of
Language-Based Models in Film Study." Film Literature
Quarterly, 3, No. 2, Spring, 197 5, 117-123.
Richards, I. A. Principles of Literary Criticism. New York:
Harcourt, Brace & Co., 1950.
Rifflet-Lemaire, A. Jacques Lacan. Brussels: Dessart,
1970.
Riley, Gresham, ed. Values, Objectivity and the Social Sciences.
Reading, Mass.: Addison-Wesley Pub. Co., 1974.
Robey, David. Structuralism: An Introduction. Oxford:
Clarendon Press, 1973.
Rohdie, Sam. "Movie Reader, Film As Film." Screen, 13,
No. 4, Winter, 1972/73, 135-145.
Ross, S. D. The Scientific Process. The Hague: Martinus
Nijhoff, 1971.
Rossi, Ino, ed. The Unconscious in Culture: The Structuralism of
Claude Levi-Strauss in Perspective. New York: E. P. Dutton
& Co., 1974.
Salt, Barry. "Statistical Style Analysis of Motion
Pictures." Film Quarterly, 28, No. 1, Fall, 1974, 13-21.
Sarris, Andrew. "Where I Stand .on the New Film-Crit." The
Village Voice, 20, No. 32, August 11, 1975.
1801
Saussure, Ferdinand de. Course in General Linguistics. London:
Peter Owen, 1960.
Scheffler, Israel. Science and Subjectivity. New York: Bobbs-
Merrill & Co., 1967.
Schilder, Paul. Psychoanalysis, Man, and Society. New York:
W. W. Norton & Co., 1951.
Scholes, Robert. Structuralism in Literature. New Haven: Yale
University Press, 1974.
Sebag, Lucien. Marxisme et structuralism. Paris: Payot, 1964.
Segal, Hanna. Introduction to the Work of Melanie Klein. New
York: Basic Books, 1964.
Segre, Cesare. Semiotics and Literary Criticism. The Hague:
Mouton, 1973.
______________. "Space and Time of the Text." 20th Century
Studies, 12, December, 1974, 37-41.
Shands, Harley C. "Semiotics Approaches to Psychiatry."
Approaches to Semiotics, ed. T. A. Sebeok. The Hague:
Mouton, 1970.
Sieminska, Ewa. "Connotation and Denotation in a Work of
Film Art." Sign, Language, Culture, ed. A. J. Greimas.
The Hague: Mouton, 1970.
Silverstein, Norman. "American Film Reviewers and French
Critics." Salmagundi, 14, Fall, 1970, 105—112.
Simon, Julian L. Basic Research Methods in Social Science: The Art
of Empirical investigation, New York: Random House, 1969.
Smith, Alfred, ed. Communication and Culture: Readings in the
Codes of Human Interaction. New York: Holt, Rinehart &
Winston, 1966.
Todorov, Tzvetan. Litterature et Signification. Paris:
Larousse, 1967.
_________________ . "Freud sur 1'enonciation." Languages, 5,
March, 1970, 34-41.
___* "Some Approaches to Russian Formalism."
20th Century Studies, 7/8, December, 1972, 6-19.
_________________ . "Semiotics." Screen, 14, No. 1/2,
Spring/Summer, 1973, 15-24.
181
Tudor, Andrew. Image and Influence: Studies in the Sociology of
Film. London: Allen & Unwin, 1974.
Theories of Film. London: Seeker & Warburq,
1974.
Uitti, Karl D. Linguistics and Literary Theory. New York:
Norton & Co., 1974.
Viet, Jean. Les Methodes Structuralistes dans les Sciences Sociales.
La Haye: Mouton, 1969.
Vogel, Amos. "The Structuralist Incursion." Film Comment,
2, No. 4, July-August, 1975, 37.
Waugh, Butler. "Structural Analysis in Literature and Folk
lore." Western Folklore, 25, 1966, 153-164.
Weber, Max. The Methodology of the Social Sciences. G lencoe,
Illinois: The Free Press, 1949.
Wellek, R. Concepts of Criticism. New Haven: Yale University
Press, 1963.
___________. The Literary Theory and Aesthetics of the Prague School.
Ann Arbor: The University of Michigan Press, 1969.
___________. Discriminations. New Haven: Yale University
Press, 1970.
Wetherill, P., M. The Literary Text: An Examination of Critical
Methods. Oxford: Basil Blackwell, 1974.
"Why We Have Resigned from the Board of Screen." Screen, 17,
No. 2, Summer, 1976, 106-109.
Willemen, Paul. "Reflections on Eikhenbaum's Concept of
Internal Speech in the Cinema." Screen, 15, No. 4,
Winter, 1974/75, 59-70.
Williams, Alan. "Structures of Narrativity in Fritz Lang's
'Metropolis.'" Film Quarterly, 27, No. 4, Summer, 1974,
17-23.
Whitney, Edward, ed. Symbology: The Use of Symbols in Visual
Communication. New York: Hastings House, 1960.
Wilden, Anthony. "Freud, Signorelli, and Lacan: The
Repression of the Signifier." American imago, 23, No. 4,
Winter, 1966, 332-366.
________________. "Jacques Lacan: A Partial Bibliography."
182
Yale French Studies, 3 6-37, October, 1966, 263-268.
Wilden, Anthony. System and Structure: Essays in Communication and
Exchange. London: Tavistock Publications, 1972.
Wolff, Janet. Hermeneutic Philosophy and the Sociology of Art.
London: Routledge & Paul, 1975.
Wollen, Peter, ed. Working Papers on the Cinema: Sociology and
Semiology. London: British Film Institute, 1968.
_. Signs and Meaning in the Cinema. Blooming
ton:__Indiana University Press, 1972.
. "Structuralism Implies a Certain Kind of
Methodology." Film Heritage, Summer, 1974, 21-29.
Wood, Robin. "Ghostly Paradigm and H. C. F.: An Answer to
Alan Lovell." Screen, 10, No. 3, May/June, 1969, 35-48.
____________. "Art and Ideology: Notes on 'Silk Stockings.''
Film Comment, 11, No. 3, May-June, 1975, 28—31.
____________. "In Defense of Art." Film Comment, 11, No. 4,
July-August, 1975, 44-51.
____________. "Against Conclusions." Film Comment, 11,
No. 5, September-October, 1975, 30-32.
____________. "Old Wine, New Battles: Structuralism or
Humanism?" Film Comment, 12, No. 6, November-December,
1976, 22-25.
Worth, Sol. "The Development of a Semiotic of Film."
Semiotica, 13, 1969, 282-321.
Wright, Will. Six Guns and Society: A Structural Study of the
western. Berkeley: University of California Press,
1975.
Youngren, William H. Semantics, Linguistics, and Criticism. New
York: Random House, 1972.
183
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
A working theory of film genre
PDF
A historical-critical study of the films of Richard Brooks: With special attention to his problems of achieving and maintaining final decision-control
PDF
Aristotle And The Movies: A Critical Study Of Unity In The Film
PDF
Clint Eastwood: An ideological study of his films, star image, and popularity
PDF
John Steinbeck In Films: An Analysis Of Realism In The Novel And In The Film--A Nonteleological Approach
PDF
A historical study of cooperative Protestant religious film in America from 1914 to 1972
PDF
A historical and analytical study of the advent and development of cinema and motion picture production in Iran (1900 -1965)
PDF
A history of the screen writers' guild (1920-1942): the writers' quest for a freely negotiated basic agreement
PDF
A critical study of the image of marriage in the contemporary American cinema
PDF
A study of the "eastern" writer in Hollywood in the 1930's
PDF
Beyond clowns and kings: Aesthetic and ideological subversion in baroque tragicomedy
PDF
Cinema-verite in America
PDF
A descriptive study of the dramatic function and significance of the clown during Hopi Indian public ceremony
PDF
A critical study of the screwball comedy film, 1934-1941
PDF
Cinema bulimia: Peter Greenaway's corpus of excess
PDF
Beyond skill development: The effects of training and development on the attitudes and retention of employees
PDF
The portrayal of the Jew in American cinema, 1900-1953
PDF
Anglo agonistes: English masculinities in British and American film
PDF
A behavioral comparison of the nonverbal and verbal communication of distressed and nondistressed marital couples
PDF
An enticement to knowledge: Documentary spectatorship and a theory of performatives
Asset Metadata
Creator
Cozyris, George Agis
(author)
Core Title
Christian Metz and the reality of film
Degree
Doctor of Philosophy
Degree Program
Communication/Cinema
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cinema,OAI-PMH Harvest
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Blacker, Irwin (
committee chair
), Kaufman, Edward K. (
committee member
), Lazar, Moshe (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c17-671650
Unique identifier
UC11343887
Identifier
DP22243.pdf (filename),usctheses-c17-671650 (legacy record id)
Legacy Identifier
DP22243.pdf
Dmrecord
671650
Document Type
Dissertation
Rights
Cozyris, George Agis
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
cinema