Sie sind auf Seite 1von 352

Self-Reference in the Media

Approaches to Applied Semiotics


Mouton de Gruyter
Berlin New York
Self-Reference in the Media

edited by
Winfried Nth
Nina Bishara

Mouton de Gruyter
Berlin New York
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter GmbH & Co. KG, Berlin.

Printed on acid-free paper which falls within the guidelines

of the ANSI to ensure permanence and durability.

Library of Congress Cataloging-in-Publication Data

Self-reference in the media / edited by Winfried Nth, Nina Bishara.

p. cm. (Approaches to applied semiotics ; 6)
Includes bibliographical references and index.
ISBN 978-3-11-019464-7 (hardcover : alk. paper)
1. Mass media Semiotics. 2. Reference (Linguistics) 3. Meta-
language. I. Nth, Winfried. II. Bishara, Nina, 1977
P96.S43S45 2007

ISBN 978-3-11-019464-7
ISSN 1612-6769

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available in the Internet at

Copyright 2007 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin
All rights reserved, including those of translation into foreign languages. No part of this book
may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopy, recording, or any information storage and retrieval system, without per-
mission in writing from the publisher.
Cover design: Christopher Schneider, Berlin.
Printed in Germany.

Winfried Noth and Nina Bishara

Communication, the conveyance of messages, is the purpose of the media ac-

cording to the self-professed ethics of the mass communicators. Messages and
their communication imply otherness: they are about something other than mes-
sages and communication, something in some other place and time, addressed
to others by a self. Nevertheless, despite their dimensions of otherness, mes-
sages, communication, and the media have always been about themselves, too
self-referential messages about messages, communication about communica-
tion, media about the media. Street criers who once called out their public
announcements did not only attract the audiences attention to their messages
but also captured their imagination by means of their voices, rhetoric, gestures,
and appearance. The newspaper in its competition with other media does not
only inform its readers about the world of otherness, it also informs how and
why it informs so well. The movies do not only bring ever new stories about
heroes and heroines, they also raise an enormous interest and curiosity in the
private lives of those who convey the messages about these heroes and heroines,
i.e., the movie actors and actresses.
The topics of the present volume are the ways in which the media have
become self-referential or self-reexive (as some researchers prefer to call it)
and the degree to which they have ceased to mediate between the real or ctional
worlds about which their messages pretend to be and their audience which they
pretend to inform, to counsel, or to entertain. The self-referential networks in
which the media and their audiences are caught up indeed, by which we are
all so signicantly shaped will be investigated in the following chapters.
The papers are presented in seven sections. Part I on Theoretical Frameworks
introduces two theoretical approaches to reference and self-reference inspired
by the semiotics of Charles S. Peirce. In his keynote paper on Self-reference in
the media: The semiotic framework, Winfried Noth contextualizes the general
topic in its cultural background in postmodernity, gives a survey of its transdis-
ciplinary implications, and draws the outlines of a systematic framework for the
study of self-reference in the media as a matter of levels and degrees. Vincent Co-
lapietro, dealing with Distortion, fabrication, and disclosure in a self-referential
vi Winfried Noth and Nina Bishara

culture: The irresistible force of reality, investigates the concepts of reality, ref-
erence, and self-reference against the background of Peirces realism and shows
how media such as television, radio, and the world wide web constitute intricate
and arguably insular networks of self-citation and self-commentary.
Part II, Self-Referential Print Advertising, studies self-reference in the pic-
torial and verbal messages of advertisements of the print media. Siegfried J.
Schmidt introduces a systems theoretical perspective in his analysis of reex-
ive loops in advertisements in their relations to other social systems, and he
proposes a typology of Modes of self-reference in advertising. On the basis
of a distinction between Metapictures and self-referential pictures, Winfried
Noth shows how pictures in advertisements have become pictures about pic-
tures, and Nina Bishara, in Absolut Anonymous: Self-reference in opaque
advertising, argues how and why opaque elements in advertisements, which
make their comprehension more difcult, evince a mode of self-reference in the
Part III, on Self-Referential Photography, begins with Winfried Noths pa-
per with the metaphorical title The death of photography in self-reference, in
which the author examines the so-called loss of the referent in digital photogra-
phy, especially in art photography. Kay Kirchmann follows with the essay Mar-
ilyn: A paragone of the camera gaze, which studies Marilyn Monroes modes
of self-observation and self-presentation in photos for the media as presented
in the 1999 ARTE series Les cent photos du siecle / One Hundred Photographs
of the Century.
Part IV on Self-Referential Films is about the movies in the movies, lmic
allusions to other lms, quotations from lms in lms, and nostalgia created by
lmic self-reference. Gloria Withalm presents reections on The self-reexive
screen and draws the Outlines of a comprehensive model for the study of
many forms of self-reexivity and self-reference in the movies on the basis of
Rossi-Landis socio-semiotics. Andreas Bohns paper, Nostalgia of the media /
in the media, discusses nostalgia, memory, remembrance, and oblivion as forms
of lmic self-reference, and Jan Siebert, in his article on Self-reference in
animated lms, presents examples from the cartoons offering insights into
self-referential scenes and devices that testify to the close connection between
humor, paradox, and self-reference.
Self-Referential Television is the topic area of Part V. In On the use of self-
disclosure as a mode of audiovisual reexivity, Fernando Andacht presents
two studies, one of the television show Big Brother Brasil and the other of a
documentary lm by E. Coutinho, demonstrating the illusionary paradox that
self-reexivity is a means of the media to give additional evidence of the real
reality in the presentations of these programs. In The old in the new: Forms and
Introduction vii

functions of archive material in the presentation of television history on televi-

sion, Joan Bleicher shows how the visual language of television has become
self-referential in its more and more frequent presentations of archive material
recalling the history of television itself thus creating a collective memory of
the medium. From the point of view of media economics, Karin Puhringer and
Gabriele Siegert, in Theres no business without show-business: Self-reference
as self-promotion, give statistical evidence of how self-promotion has become
one of the most important forms of self-reference in the mass media.
Computer games are the topic of Part VI, entitled Self-Referential Games.
Computer games [are] the epitome of self-reference is Lucia Santaellas argu-
ment in her paper putting forward a typology of seven types of self-reference in
games. Bo Kampmann Walther proposes A formalistic approach to the study
of self-reference in computer games, dening rules, strategies, and interaction
patterns as their core elements and examining how and to what extent computer
games can be dened as complex dynamic systems. Britta Neitzel, in her paper
on Metacommunication in play and in (computer) games, shows that Gregory
Batesons theory of play is fundamental to the study of games, and Bernhard
Rapp, in Self-reexivity in computer games: Analyses of selected examples
concludes the section with exemplary analyses and proposals for future research
on the topic.
Part VII presents three papers on Other Self-Referential Arts in such diverse
elds as web art, body art, and music. Marie-Laure Ryan contextualizes self-
reexivity in the history of literature since Don Quixote and gives evidence
of the predominance of self-reexivity in digital art on the Internet in her pa-
per Looking through the computer screen: Self-reexivity in Christina
Ljungberg, in The artist and her bodily self: Self-reference in digital art/media,
constructs a typology of degrees and forms of self-reference in digital art exem-
plied by multi-media works of visual artist and performer Laurie Anderson,
video/digital artist Selina Trepp, and media artist Char Davies. Werner Wolf
concludes the volume with his paper entitled Metaction and metamusic: Ex-
ploring the limits of metareference. Based on a denition of meta-reference
in contrast to self-reference and self-reexivity in the narrower sense, Wolf
presents new typological tools for the comparative study of meta-music and
offers original proposals for a comprehensive program of future research on the
The volume is one of the main results of a research project on self-reference
in the media with special focus on advertising, the movies, and computer games,
carried out at the Interdisciplinary Center for Cultural Studies of the Univer-
sity of Kassel from 2003 to 2006. Supported by a grant from the Deutsche
Forschungsgemeinschaft (DFG), the project was directed by Winfried Noth,
viii Winfried Noth and Nina Bishara

whose collaborators were Nina Bishara (Kassel), Britta Neitzel (now Siegen),
and Karin Wenz (now Maastricht). With few exceptions, the papers presented
here were contributions to the international conference Self-Reference in the
Media organized in the framework of the aforementioned DFG project by Win-
fried Noth, Britta Neitzel, and Nina Bishara at the University of Kassel in July
Thanks are due to the DFG for their substantial support and encouragement
of this volume as well as to the University of Kassel for unbureaucratically
providing the necessary infrastructure. Especially worth mentioning is the DFG
supported collaboration of the research project Self-Reference in the Media
with the Postgraduate Program in Semiotics and Communication Studies of
the Catholic University of Sao Paulo, whose immediate results presented in
this volume are the contributions by Lucia Santaella, Vincent Colapietro, and
Fernando Andacht.
Thanks are also due to Dr. Renira Gambarato for improving several diagrams
and to Diena Janakat for editorial assistance.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Winfried Noth and Nina Bishara

Part I: Self-referential media: Theoretical frameworks

Self-reference in the media: The semiotic framework . . . . . . . . . . . . . . . . 3

Winfried Noth

Distortion, fabrication, and disclosure in a self-referential culture:

The irresistible force of reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Vincent Colapietro

Part II: Self-referential print advertising

Modes of self-reference in advertising . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Siegfried J. Schmidt

Metapictures and self-referential pictures . . . . . . . . . . . . . . . . . . . . . . . . . 61

Winfried Noth

Absolut Anonymous: Self-reference in opaque advertising . . . . . . . . . . 79

Nina Bishara

Part III: Self-referential photography

The death of photography in self-reference . . . . . . . . . . . . . . . . . . . . . . . . 95

Winfried Noth

Marilyn: A paragone of the camera gaze . . . . . . . . . . . . . . . . . . . . . . . . . . 107

Kay Kirchmann
x Contents

Part IV: Self-referential lm

The self-reexive screen: Outlines of a comprehensive model . . . . . . . . . 125
Gloria Withalm
Nostalgia of the media / in the media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Andreas Bohn
Self-reference in animated lms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Jan Siebert

Part V: Self-referential television

On the use of self-disclosure as a mode of audiovisual reexivity . . . . . . 165
Fernando Andacht
The old in the new: Forms and functions of archive material
in the presentation of television history on television . . . . . . . . . . . . . . . . 183
Joan K. Bleicher
Theres no business without show-business:
Self-reference as self-promotion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
Karin Puhringer and Gabriele Siegert

Part VI: Self-referential games

Computer games: The epitome of self-reference . . . . . . . . . . . . . . . . . . . . 207
Lucia Santaella
Self-reference in computer games: A formalistic approach . . . . . . . . . . . . 219
Bo Kampmann Walther
Metacommunication in play and in (computer) games . . . . . . . . . . . . . . . 237
Britta Neitzel
Self-reexivity in computer games: Analyses of selected examples . . . . . 253
Bernhard Rapp

Part VII: Other self-referential arts

Looking through the computer screen: Self-reexivity in . . . . . . . 269
Marie-Laure Ryan
Contents xi

The artist and her bodily self: Self-reference in digital art/media . . . . . . . 291
Christina Ljungberg
Metaction and metamusic: Exploring the limits of metareference . . . . . 303
Werner Wolf
Index of names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
Index of subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
Part I. Self-referential media:
Theoretical frameworks
Self-reference in the media: The semiotic framework

Winfried Noth

1. Self-reference in postmodernity and in the media

Self-reference is a much discussed characteristic of postmodernity (Lawson

1985; Noth 2001; Petersen 2003). In an era in which everything seems to have
been said, the grand narratives have lost their credibility, and representations
can no longer represent (Lyotard 1979: 27). To escape from this dilemma, liter-
ature, the visual and the audiovisual arts and media have become increasingly
self-referential, self-reexive, autotelic.
Instead of representing something heard about, seen, lived, or otherwise
experienced in social life, culture, and nature, journalists, commercial artists,
designers, and lm directors report increasingly what has been seen, heard, or
reported before in the media. The mediators have turned to representing repre-
sentations. Instead of narrating, they narrate how and why they narrate, instead
of lming, they lm that they lm the lming. The news are more and more
about what has been reported in the news, television shows are increasingly
concerned with television shows, and even advertising is no longer about prod-
ucts and services but about advertising. The messages of the media are about
messages of the media, whose origin has become difcult to trace. In literature,
ction has become metaction, novels have become metanovels, and texts are
being discovered as intertexts whose reference is not to life but to other texts.
Last but not least, art is now about art, and even architecture is about architecture.
The digitalization of pictures and lms, which has liberated the media from
the bonds of factual reference to a world which they used to depict, has con-
tributed to the increase of self-reference. No longer originating in a world which
leaves its documentary traces on the negatives of a lm, the pictures of the new
media have become the result of digital imaging and art work, whose origin is
in the software of the semiotic machines (cf. Noth 2002) by means of which
they are produced.
One of the most striking symptoms of the current concern with self-reference
in culture and in the media is probably the recent phenomenon of culture jam-
ming (Klein 2000, chapt. 12), the critical transformation of media messages by
4 Winfried Noth

activists who display their protest against the age of consumerism, globaliza-
tion, and social surveillance in public places and urban spaces in subversive
forms such as adbusting, grafti, ash mobs, hacktivism, cybersquatting, or
sousveillance (cf., 16.05.06), not without creating the
self-referential paradox that they depend on the media in their subversive at-
tacks against the media.

2. Self-reference as a multidisciplinary topic of research

The study of self-reference and related phenomena, such as self-similarity, self-
organization, autopoiesis, replication, or recursion is a topic of interest to various
elds of research. Bartlett (1987: 1024) gives a comprehensive survey of rel-
evant topics and studies in no less than twenty-one elds of research, from
mythology to neurophysiology, among them the following ones not dealt with
in detail below: linguistics (reexivity), space and time (loops, circles, Moe-
bius strip), law (self-referring and self-limiting laws, mutuality of contracts),
economics (business cycles), game theory (rules permitting self-modication),
anthropology (Sapir-Whorf hypothesis: culture determining language and hence
culture), mythology (cosmic cycles), psychiatry (narcissism), psychotherapy
(Batesons theories of play and double bind), neurophysiology (neuronal cir-
cuits), and general systems theory (see Noth 1977). The following survey of
more recent research in self-reference excludes systems theoretical approaches
to self-reference which have been reviewed elsewhere with particular reference
to semiotics (Noth 2000b; Jahraus and Ort 2003).
In the natural sciences, the theory of complex systems in physics and math-
ematics (chaos and fractals: Peitgen, Jurgens, and Saupe 1992), chemistry (dis-
sipitative structures: Prigogine and Stengers 1984), biology (self-reference, self-
description, autopoiesis: Hoffmeyer 1996: 3951), and even meteorology (but-
tery effect) are bringing more and more evidence of the omnipresence of self-
reference and related phenomena in nature: self-observation, self-description,
self-organization, self-replication, self-similarity, autopoiesis, feedback loops,
iteration, replication, recursion, or downward causation (Andersen et al. 2000)
are the key concepts in this context.
In computer science, the recursivity of Turing machines (Winkler 2004: 170
182) and the theory of autonomous agents (Pattee 1995; Noth 2002) are relevant
to the study of self-reference. The close afnity between recursion and self-
reference, for example, is evident when we consider the mathematical denition
of recursivity as a group using the own group or function that it calls to the
own function (
recursivity.htm, 16.05.06).
Self-reference in the media: The semiotic framework 5

Logic and the philosophy of language have given special attention to self-
reference with respect to tautology, the petitio principii (taking for granted
what should rst be proved), other semantic circularities (Myers 1966), or self-
referential propositions that lead to antinomies and paradoxes. Much attention
has been paid to forms of self-reference implied in metalanguage (Hofstadter
1979, 1985) and paradoxes (Whitehead and Russell 1910; Bartlett and Suber
1987; Fitch 1987; Bartlett 1992b; Scheutz 1995; Schoppe 1995). Other philo-
sophical aspects of self-reference are philosophical reexivity (Nietzsche, Hei-
degger, Derrida: Lawson 1985), the phenomenology of the self and its identity
(Buttner and Esser 2001), the problem of self-consciousness (Potthast 1971;
Colapietro 1989; Kienzle and Pape 1991), also a topic of cognitive science
(self-awareness: Brook and DeVidi 2001), and the topics of self-reection, self-
representation, autosymbolism, or the autotelic function in aesthetics (Shir 1978;
Luhmann 1984; Menninghaus 1987; Noth 2000a: 425, 432; Metscher 2003).
Literary studies are one of the elds of research (besides aesthetics) in which
the theory of self-reference has its longest tradition since the essence of liter-
ature has often been described in terms which imply self-reference. Key con-
cepts in this context are aesthetic autosymbolism (Shir 1978), self-representation
(Hempfer 1976: 70, 129; Jay 1984; Johansen 2002: 174288), literary autonomy,
autonymy, or the autotelic function of literature (cf. Noth 2000a: 458). While
most of these theories have been developed against the background of poetry,
often with reference to Jakobsons denition of the poetic language as a self-
referential language (Jay 1984; Whiteside 1987; Block 1999; Noth 2000a: 453;
Johansen 2002: 174182), self-reference in prose and drama is a more recent
topic. It has rst been approached in the 1970s under the heading of metalan-
guage (Smuda 1970), later as metatext, especially metaction (Waugh 1984;
Siedenbiedel 2005), or metanovel (Zavala 2000). In the study of narratives, the
topic has also been subsumed under the general heading of reexivity (Stam
1992), self-reexivity (Hempfer 1982; Scheffel 1997; Huber, Middecke, and
Zapf 2005), or self-reference proper (Wolf 2001; Krah 2005a, 2005b). Compre-
hensive surveys on the topic can be found in Scheffel (1997) and Wolf (2001).
Language about language, ction about ction, or the novel about the novel,
these are evidently topics which deal with self-reference at a very general semi-
otic level. The theory of intertextuality (Broich and Pster 1985) implies a sim-
ilarly general mode of self-reference since it deals with the way a text refers to
a text instead of to the adventures of its protagonists. Metaction containing re-
ections about the text in which these reections are narrated may be described
as evincing a higher degree of self-reference than intertextuality. Intertextual
references also evince references to texts, but these references are to other texts.
6 Winfried Noth

Like literature, music and the traditional visual arts have had self-reference
inscribed in their canonical denitions since the classics of philosophical aes-
thetics. Lart-pour-lart, autonomy and autoreexivity have been key concepts
in this tradition (cf. Noth 2000a: 434, 426427). The new trend since post-
modernity has been that artists have begun to reect programmatically about
art in their art works, so that art has become art about art (Lipman and Marshall
1978) and even architecture has become architecture about architecture (Wittig
1979). A conspicuous symptom of the increasing concern with self-reference in
the visual arts is the current interest in representing and exhibiting the artists
own bodily self in works of visual art (cf. Santaella 2004; Noth and Hertling
2005; Noth ed. 2006; Ljungberg, this vol.).
Media studies have discussed the argument that self-reference is at the root
of every medium. Each individual medium has a historical precursor to which
it refers back in media history. The more the media interact today and turn
intermedial, the more they refer to the media in self-referential loops. These
were some of the reasons why McLuhan (1964) declared that the medium is
the message. The famous tenet expresses among other things the view that each
message in the media refers both to its own medium and to other media, and
thus characterizes messages as partially self-referential. McLuhan (1964: 8)
develops this argument on the basis of his very broad concept of medium as an
extension of man, according to which even light is a medium:
The electric light is pure information. It is a medium without a message,
as it were, unless it is used to spell out some verbal ad or name. This fact,
characteristic of all media, means that the content of any medium is always
another medium. The content of writing is speech just as the written word
is the content of print, and print is the content of the telegraph. If it is asked,
What is the content of speech?, it is necessary to say, It is an actual
process of thought, which is in itself nonverbal.
Notice that in this description of how the messages in the media circulate
in a process of innite semiosis which even includes thought as a content of
a medium, the medium described as the most self-referential of all is light. A
medium without a message which nevertheless conveys pure information can
only be a medium that refers to nothing but to itself. All other media evince
self-reference to the degree that they refer to other media, which implies a
divided reference. To the degree that the media refers to the media, they are
self-referential, to the degree that they refer to other media, it is (allo)referential
(see below).
Intermediality (Muller 1996; Paech 1998; Spielmann 1998; Helbig 2001;
Rajewsky 2002), the media in the media (Liebrand and Schneider 2002), media
change (Ort 2003), as well as remediation (Bolter and Grusin 1999), i.e., the
Self-reference in the media: The semiotic framework 7

refashioning of the traditional media in the digital media, are the topics of
research in self-reference in the media related to McLuhans dictum in one
or the other way. From various other perspectives, self-reference in the media
has been approached in the contexts of lm (Karpf, Kiesel, and Visarius 1996;
Kirchmann 1996; Paech 1998; Buckland 2000: 5376), television (Withalm
1995; Frieske 1998; Bleicher 1999), journalism (Marcus 1997; Blobaum 1999;
Kohring 1999; Weber 1999), and advertising (Schmidt and Spie 1996). For
further references, see the papers of this volume.
Various aspects of self-reference concerning other domains of culture are
discussed by Hofstadter (1979, 1985), who has shown that self-reference is at
the root of cultural creativity (see also Schoppe 1995), in particular of humor
and paradox. Self-reference in popular culture from the comics to rock music
and video-clips is the topic of the book on metapop by Dunne (1992). Among
the topics of cultural semiotics with particular relevance to self-reference are
the semiotics of mirrors (Eco 1984; Ort 2003) and the semiotics of fashion.
It was Barthes (1967: 287) who described fashion as a tautological system
which denes itself reexively only through itself, a system of signs deprived
of content but not of sense, a kind of machine to operate sense without ever
xing it with the only goal of making the insignicant signicant, or, as
Goebel (1986: 476) put it, a system that keeps conveying the same message for
ever: fashion is hence a language that consists of nothing but synonyms.
In the interpretation of the phenomenon of ever increasing self-reference in
postmodern culture, we nd the apocalyptic critics opposing the integrated
ones. The former, among them Baudrillard (1976, 1981, 1991), deplore the
loss of referents in a more and more self-referential world in which reality has
degenerated to constructed, simulated or virtual reality. The latter interpret self-
reference as a symptom of increasing critical consciousness in a world that
has lost its condence in ultimate truths (Lawson 1985). However, while the
integrated ones may lack critical distance in face of the aporias of postmodern
self-reference, the apocalyptic ones run the risk of nding themselves involved
in paradoxes as long as they are unable to explain the nature of those referents
whose loss they deplore (Noth 2001; Noth and Ljungberg 2003).

3. Self-reference and reference: Semiotic premises

In the framework of the present research project on self-reference in the media
(cf. Noth 2005b), the concept of self-reference has been adopted in the very
broad sense similar to the outline proposed by Bartlett (1987: 6), whose point
of departure is the following reection on self-reference in human thought:
8 Winfried Noth

When we employ thought to understand the nature of thinking, when we

seek to know the presuppositions involved in knowing, we dene a task
that essentially involves the subjects we would study. Reexivities of this
kind are widespread: sociology, anthropology, biology, and many other dis-
ciplines, as we shall see, exhibit varieties of self-reference. Attempting to
understand reexivity gives one the sense of trying to lift oneself by the

Our own point of departure is a semiotic one: any sign that refers to itself or
to aspects of itself is a self-referential sign. Signs that do not refer exclusively
to themselves but only to parts, aspects, constituents, or elements of themselves
are self-referential to a degree that remains to be specied (see below on levels
and typology of s.-r.).
Self-reference in the broad sense adopted here includes a number of con-
cepts which are sometimes used as synonyms of this term as well as certain
concepts which some authors, in the context of media and cultural studies,
explicitly distinguish from self-reference. The most frequent synonym is re-
exivity (Whitehead and Russell 1910; Lawson 1985). Typically enough, both
concepts appear in title of the book by Bartlett and Suber (1987), which is
Self-reference: Reections on Reexivity. Some other terminological alterna-
tives are self-reexivity (Huber, Middecke, and Zapf 2005), self -representation
(Johansen 2002: 174288), or autoreferentiality (Pavlicic 1993).
Some authors distinguish these alternative concepts from the one of self-
reference. Connotations associated with such distinctions are the following:
reexivity and self-reexivity connote reections on the process of the au-
thors own writing or self-cognition and self-consciousness, for example, in
the philosophical tradition of the Romantics (Menninghaus 1987), the tradi-
tion of phenomenology (Lawson 1985), or anthropology (Babcock 1980); self-
representation is often preferred in the context of aesthetics (Metscher 2003). In
the context of literary semiotics, Johansen (2002: 174288) avoids the concept
of self-reference and uses the term self-representation instead (in opposition to
other-representation). This terminological decision is understandable from the
point of view of Peircean semiotics, since representation, and not reference is a
key concept of Peirces theory of signs. Wolf (2001: 56) distinguishes between
self-reference as a noncognitive and self-reection as a cognitive process,
using the former term to describe textual recurrences and repetitions and the
latter term to designate reection on the writers self (for details, see Wolf, this
vol.). In Luhmanns systems theory no such distinctions are drawn; the general
term is always self-reference. Luhmanns concept is very fundamental: it refers
to the capacity and tendency of a living system to establish reference to itself in
its interactions with the nonself, that is, its environment (cf. Noth 2000b).
Self-reference in the media: The semiotic framework 9

The study of self-reference requires an elucidation of its opposite, refer-

ence. By denition, reference, in contrast to self-reference, means referring
to something else. When the term is used in opposition to self-reference, it is
also called alloreference, heteroreference (Wolf 2001) or other-reference, which
comes closest to Luhmanns German term Fremdreferenz.
What is reference and what does it mean to refer? Despite much controversy
over the inscrutability of reference (cf. Geach 1970; Evans 1991; Katz 2004),
Bartlett (1987: 5) takes the concept for granted, when he considers reference
as a necessary constituent of human communication, stating that without a
wide range of abilities to refer, we would be bereft of thoughts, memories, and
sensations and that the world as we perceive it, remember it, and conceptu-
alize it would, in the absence of appropriate referring capacities, collapse into
impossibility. However, is reference really a necessary ingredient of human
communication and a necessary term in semiotics? Ferdinand de Saussure is
known to have banned reference from linguistic semantics for decades, Peirce
hardly uses the term, and the linguist Roman Jakobson, denes the referential
function of language as only one of six functions of verbal communication all
of which differ from the concept of reference although they do not exclude the
possibility of the message being referential to a certain degree either (see below
on the semiotic paradox and degrees of s.-r.).
In English, the concept of reference has been introduced in its current sense
with Max Blacks translation of Gottlob Freges dichotomy of Sinn vs. Be-
deutung as sense vs. reference (Munch 1992: 385; Noth 2000a: 15254). In
this tradition, reference is dened as the relation between a verbal expression
and the observable things or qualities to which it refers; that to which it refers
is called its referent, extension, or designatum. An expression that refers to a
referent identies it as an individual or a class of objects, actions, or events
(Kempson 1977: 13). For example, the word king, at the turn of the millennium,
refers to the present kings of Spain and Sweden and the past kings of these and
many other countries. Sense or meaning, by contrast, consists of the ideas or
concepts evoked in the mind of those who use or understand the word. The word
king has the conceptual meaning man who rules a country as a descendant
of a royal family.
In this tradition of logical semantics, it is possible for a word to have meaning
but no reference. In the year 2000, the expression the present king of France
is meaningful and makes sense because we understand the ideas associated with
its words, but the expression is without a referent because there is no individual
in France to whom it might presently refer. Although there are even abstract
objects, such as the object of imagination which refers to the class of all acts of
imagination, some words have no reference, since they refer to nothing that has
10 Winfried Noth

extension, for example, and, or, what, whether, of, unicorn, or the rst woman
to land on the moon.
The logical theory of reference as something in the external world to which
the sign refers, or points to, has not remained undisputed. In the framework
of Saussurean structuralism, linguists developed a semantic theory which ig-
nored the theory of reference for decades (cf. Noth 2000a: 7475). The semiotic
structure of a verbal sign was sought in its meaning only, which was studied
exclusively in its relation to other signs and not in relation to its referents. The
same aversion against approaching the dimension of reference is characteristic
of constructivism and systems theory. Niklas Luhmann, for example, justies
his exclusion of the referent from his theory of social and cultural systems as

There is indeed no reference for the sign as a form; which is to say: one can
either make use of the distinction between signier/signied or not. There
is no external point of reference that would force one to select either
option; neither is there any truth criterion for choosing a rst distinction as
a starting point. That is why a theory of language constructed as semiotics
must relinquish the idea of languages external referent. (Luhmann 1993: 24)

Nevertheless, despite this plea against the theory of reference, self-reference

vs. alloreference is a fundamental dichotomy of his systems theory of com-
munication, culture, the media, and the arts (cf. Noth 2000b). Both concepts
have to do with observation. While alloreferential observing is directed towards
phenomena in the environment of a system (or an observer), self-reference is
directed towards the observing system, the observer, the process of observation,
or the process of communication (Luhmann 1995: 15, 28). Furthermore, quite
against basic tenets of both the Peircean and the Saussurean semiotics, Luhmann
(1993: 24) even denes alloreference as external reference, when he states that
an operationally closed, language-using system [. . . ] must distinguish between
self-reference and external reference. However that may be, the concept of
self-reference can apparently be used without assuming the Fregean view of
Among the constructivists, S. J. Schmidt adopts a similar position. On the
on hand, he argues that signs are not anchored in a sphere beyond discourse,
hence, are not anchored in referents; on the other hand, he nevertheless uses the
term of reference, albeit in a different sense. Reference, according to Schmidt,
is not a matter of semantics, but one of pragmatics; it concerns the process of
communication and not the relation between the sign and its referent. Reference,
in this perspective, is a renvoi from communication to communication which
permits connections and relays (Schmidt 1994: 145), while self-reference is
Self-reference in the media: The semiotic framework 11

a matter of how communication refers to communication (and hence to itself).

Schmidt even goes so far as to postulate that signs and communication, being
essentially about signs and communication, are always self-referential in the
rst place.
Still other premises of a theory of self-reference derive from Charles S.
Peirces semiotics on which this outline of a program for research in self-
reference in the media and several papers in this volume are based. Peirce would
never have subscribed to Freges theory of reference, nor does the noun reference
or the verb to refer belong to Peirces basic vocabulary (Noth 2006). Instead of
the referent or extension, Peirce speaks of the object of the sign, and instead of
saying that the sign refers to its object, Peirce says that the sign represents its
The Peircean object, which a sign represents, does not necessarily have an
extension, and it does not need to be a piece of the so-called real world at all, since
signs or ideas can be the object of a sign. The object of the sign is something
which precedes and thus determines the sign in the process of semiosis as a
previous experience or cognition of the world (cf. Noth 2006). Such an object
of the sign can be a sign itself, and this is where self-reference begins with signs
representing signs.
Reference in the narrower sense of referring or even pointing to something
else is a semiotic characteristic of only one of Peirces major classes of sign,
the indices. For example, the deictic words you, there, or then refer to a person,
a place, or a moment which is distal in relation to the speaker and the place
and time of speaking. Indexical signs identify and in this sense refer to objects
and events in time and space in many other ways, for example by means of
adverbial descriptions or nonverbal gestures of pointing. However, indexical
signs can also evince self-reference, namely in the case of proximal deixis in
words such as I, here, and now, which refer to the speaker and the circumstances
of the utterance. Symbols, such as cat or speaker, by contrast, do not refer in
this sense; they represent general concepts with which our experience of these
objects is connected. Even less so can icons, such as the speakers picture, be
said to refer to their objects. Pictures represent or show; they do not refer to
their object.
In sum, instead of self-reference, Peirce would use the term self-representa-
tion, but out of consideration for the wider acceptance of the term self-reference
in media theory, this term will not be adopted here.
12 Winfried Noth

4. The semiotic paradox and degrees of self-reference in the media

A sign, according to its medieval denition, is something that stands for some-
thing else: aliquid stat pro aliquo (cf. Noth 2000a, 2000b, 2006). If we disregard
certain problems associated with the verb to stand for and admit a broader range
of relational verbs as its interpretation, such as referring to, representing, or
evoking the concept of, the formula is reduced to a dyad which is considered in
all denitions of the sign. Whether dyadic or triadic, the basic assumption of
the difference between the sign and something other than the sign to which it
refers or which it represents is a distinction drawn in all denitions: the signier
is not the signied, the sign is neither its referent nor its object, just as the map
is not its territory, as A. Korzybski (1933) put it. Self-reference thus creates a
semiotic paradox: the sign does no longer refer to or represent something else;
it is its own object, a map that is its own territory.
It is true that signs also have other functions in addition to the one of refer-
ence. Roman Jakobson, e.g., distinguished no less than ve other functions of
language besides the one of reference in the narrower sense: the expressive, the
conative (appellative), the metalingual, the phatic, and the poetic function (cf.
Noth 2000a: 105106). Some of them, for example the expressive, the poetic,
and the metalingual function, indeed evince characteristics of self-reference
since they are associated with messages about the sender of the message, or the
message itself and its signs, but language without a potential of representing
and referring to a world it represents and above all which is absent in time and
place would fail its evolutionary, cultural, and social purpose.
If it is the purpose of signs to represent or to refer to something else, this
purpose should be no less characteristic of the signs in the media. After all, the
concept of media implies mediation, and mediation is a process of semiosis,
the action of signs. Medium is even a synonym of sign in the framework of
Charles S. Peirces semiotics, and Peirce even considered substituting the con-
cept of sign for the term medium, when, in 1906, he exclaimed: All my notions
are too narrow. Instead of Sign, ought I not to say Medium? (MS 339: 526).
The media must be able to inform about, narrate or evoke events, persons,
places, and messages from elsewhere in time and space. Their potential to do
so has turned the world into a global village. Global communication without
reference is unthinkable. To fulll their function, the signs of the media must
evince the potential of reference or representation. Any message from the mass
media is referential by necessity as far as its enunciation is concerned, since it is
a message from elsewhere, the radio station for example, about an event which
happened or originated at still another place in the world. Even the music that
we hear is not without elements of reference to other times and other places;
Self-reference in the media: The semiotic framework 13

jazz refers to New Orleans, samba to Brazil, and Bach to 17th century Europe,
but music is essentially self-referential, in particular the art of the fugue which
is highly recursive and in this respect self-referential (Hofstadter 1979).
Reference and self-reference are thus evidently a matter of degree. Various
degrees of self-reference can be distinguished, from the sign that refers to nothing
but itself to the sign that refers only partially to itself and partially still to
something else. No message in the media is completely devoid of self-reference.
Even in everyday verbal communication, the speaker indicates himself or herself
as a speaker, whether intentionally or unintentionally. A message in the NewYork
Times refers self-referentially to the prole and status of this newspaper, and
each television picture that shows the station logo in its upper left or right hand
corner refers self-referentially to the station itself, but at the same time it is an
alloreferential message which serves to draw a distinction to all other stations. A
statement of the President of the USA about the Iraq is highly referential, since
it concerns events in very remote places, but it is also self-referential insofar as
it is a message referring to Bush, his own and the US politics as well as to his
own language, the English language in which he transmits the statement.
The media differ as to the degree to which their messages are typically self-
referential or (allo)referential. Consider advertising, lm, and computer games.
Advertising is referential at its roots, since it has the purpose of promoting
and selling products or services. For this reason, genuine self-reference would
be counterproductive; a genuinely self-referential message would be unable
to fulll its commercial purpose of propagating a message about goods and
services. Nevertheless, advertisements make use of the creative devices of self-
reference to draw the consumers attention towards the message. Feature lms,
by contrast, which have both ctional and aesthetic qualities, are referential
and self-referential at the same time. While their narrative plot is referential
or indexical (Bettetini 1971) insofar as it narrates events from the lives of its
protagonists, their aesthetic devices are based on self-reference, and if Lyotard
(1979: 27) was right when he proclaimed the end of the grand narratives, it is
only natural that self-reference in lms must have increased. In computer games
we are nally faced with a medium in which alloreference has been secondary
since its beginning, since playing and games create their own self-referential
worlds apart from the world of referential facts and realities.
14 Winfried Noth

5. Levels and typology of self-reference in the media

Self-reference occurs at different levels of the media and the message in which
it occurs. The degree of self-reference is related to the level in the hierarchy
from the level of elementary signs to the one of complex signs and from the
level of a text or message to the level of the media system as a whole. For
example, a newspaper article (level: text) that criticizes the media in general
(to which it belongs itself) is less self-referential than a newspaper article that
criticizes only the newspapers and not the other media; and an authors self-
referential comments on his or her own story are more (directly) self-referential
than reections of an author on the principles of narrating in general since these
refer only partially to the story in which it is included.
A distinction between different levels of self-reference is implicit in some
proposals that have been suggested for a typology of forms of self-reference.
Among other varieties of self-reference, Bartlett (1987, 1992a), for example,
distinguishes between self-reference at the level of indexical words, paradox-
ical and tautological sentences, and pragmatic or performative self-reference
in statements in which the speakers intentions are self-referentially involved.
Scheutz (1995: 24) proposes a typology beginning with self-referential symbols,
having self-referential sentences as its second, and self-referential theories as
its third level.
The levels of self-reference in the media distinguished in the following are
equally inspired by the ambition of establishing a hierarchy from the most ele-
mentary to the highest level of self-reference in the media. The rst three levels
are derived from Peirces trichotomy of the interpretant, which draws the distinc-
tion between the rheme, the dicent, and the argument (cf. Noth 2000a: 6567).
A rhematic sign or rheme is a verbal or pictorial sign at a level equivalent to the
one of the word or concept in language. The above-mentioned self-referential
symbols and indexical words are types of rhematic self-reference. A dicentic
sign corresponds to the level of sentences or statements in language. Paradoxical
and tautological sentences belong to this level of self-reference. An argument
presupposes a sequence of sentences in which a conclusion is derived from
premises. The logical fallacy of the petitio principii (the taking for granted what
should rst be proved) and similar argumentative circularities evince argumen-
tative self-reference.
In extension of these levels derived from the three Peircean categories of the
interpretant, forms of self-reference at the following higher levels will be dis-
tinguished: intratextual, intertextual, and intermedial self-reference and enun-
ciative self-reference. While (intra)textual self-reference concerns the level of
an individual text, a single advertisement, lm, or computer game, for example,
Self-reference in the media: The semiotic framework 15

intertextual self-reference concerns references from one text to other texts of

the same genre or medium, e.g., from one to another advertisement or lm or
game, respectively. The term intermedial will be used to refer to the relation
between different media or genre, for example painting and lm, lm and games,
or advertising and cinema.
In addition to these forms of self-reference distinguished according to the
hierarchical order of signs in the media, three other forms of self-reference will
be distinguished according to criteria which overlap or combine with the above
hierarchical typology: enunciative, iconic, and indexical self-reference (cf. Noth
2007). Enunciative self-reference involves the communicative situation and de-
scribes reference of the speaker, writer, composer, or producer of the sign but
also the role of the audience or spectators. Iconic and indexical self-reference
involves self-referential icons and indices. Self-reference that is not specic
of the media, such as the elementary self-referentiality of communication (we
communicate that we communicate), will not be considered in this context.

5.1. Rhematic self-reference: Examples from advertising

Maybe, just maybe was the advertising slogan of the British national lottery
of 1998 (Knowles 2004: 4) which illustrates well rhematic self-reference. The
slogan is a verbal rheme, a sign that afrms nothing but expresses a mere poten-
tiality, and the slogan is self-referential in its repetition to the same extent that
any repeated form refers back to itself in an iconically self-referential way.
Rhematic self-reference is a popular strategy in advertising. One of its most
frequent forms is the advertisement that attracts the consumers attention to
nothing but the brand name without saying anything about the product.A parallel
strategy is the mere showing of the product in the form of a picture. In both cases,
the message consists of a rhematic sign. Unlike a dicent, a rheme afrms nothing.
Without a predication, a praise of its qualities, for example, an advertisement of
his kind merely shows, and thus remains open to many interpretations. Like a
word without context, e.g., beer, the rheme refrains from designating anything
in specic. Its meaning is a mere possibility, and its context in time and space
is undetermined. In advertising, the meaning of the rhematic message about a
product is left to the consumers imagination, but their prior knowledge about
the product is important. A new product cannot be introduced with rhematic
The prototype of a rhematic advertising campaign is the classical Coca-Cola
sign at a countryside highway. It shows nothing but the Coca-Cola bottle with
the name of the soft-drink as its label. Coke, nothing but Coke or Coke
16 Winfried Noth

forever seems to be an implicit quasi-tautological message. Since it remains

unsaid whether the drink is good, desirable, or unique it seems to be assumed
that the consumer knows its qualities well enough. The tacit assumption is that
it would be a tautology to repeat what everybody knows anyhow. However, not
to repeat it makes the message equally redundant, for if it need not be said what
else is the purpose of the advertisement? This is the circularity which constitutes
the basic self-referentiality of these advertisements.
Rhematic signs of a product that conveys no other message than the one of its
name or picture have some afnity with signboards, for example, the pictorial
sign of a shoe indicating a shoemakers shop. However, the difference between
signboards and rhematic advertisements is semiotically important. The shoe-
makers shop sign is also a rheme since it corresponds to a mere word, but in
contrast to the Coca-Cola sign, the shoemakers shop sign refers indexically to
a specic place. It is a rhematic index, which conveys the alloreferential mes-
sage: Here is a shoemaker. Rhematic advertisements which merely show the
product, by contrast, indicate nothing. Without any reference to a specic object,
they are rhematic icons, signs which evince qualities of their objects without any
indication of it. Insofar as it shares the qualities of its object and insofar as it is
tautological in its reference to the well-known and hence presupposed qualities
of the product, the rhematically iconic advertisement is a self-referential sign.
Self-reference of this kind is frequent in current print advertising for fashion
labels which reduce their message to a mere showing of the clothes for sale with
the inscription of the designer logo (e.g., Joop, Boss, Gucci, etc.; cf. Bishara,
forthcoming). Without any further comment, the message suggests that neither
the name nor the product need any comment since they speak for themselves.
On its surface, it is an open and hence rhematic message, but in a certain way,
advertisements are never open; their message always presupposes or takes for
granted that quality and desirability are characteristics of the product. On this
assumption, the rhematic sign disguises a dicentic message afrming that there
is only one product that deserves consideration which is the best, the most
desirable, and the one that must be bought.

5.2. Dicentic self-reference in advertising

Dicentic self-reference can be illustrated with the famous German tautologi-

cal advertising slogan for Persil washing powder Persil bleibt Persil [Persil
remains Persil]. Explicit tautologies, such as the one of this slogan, or quasi-
tautologies are statements and hence dicentic signs. At rst sight, the claim has
the form of a predication. However, instead of a predicative and thus alloref-
Self-reference in the media: The semiotic framework 17

erential statement of the form A is B, we are confronted with a tautological

and consequently self-referential statement of the type A is A. The rhetoric of
tautology serves to remind of a quality that no longer needs to be stated. The
advertising slogan simply presupposes the knowledge of the quality inherent to
the product. In a more recent BMW campaign, the tautology is even a twofold
one: A BMW is a BMW is a BMW. . .
Other popular kinds of dicentic circularities are created by means of elliptical
constructions. A slogan for the detergent Domestos bleach of 1959 claims: Kills
all germs (Knowles 2004: 4). To understand this message, the reader is obliged
to ll the gap left by the ellipsis of the subject Domestos from the packaging
of the product and to substitute the missing proper noun from the product which
it designates. The product in its package utters, so to speak, the self-referential
message: I kill all germs.
Still another rather frequent kind of self-referential circularity caused by
elliptical constructions at the level of the dicent has the form of an open predicate
whose scope includes the argument from which it should differ. The slogan
Persil washes whiter is an example (Persil washing powder, GB, 2006). The
enigma is: whiter than what? The comparative clause left out at the end of this
slogan seems to be than everything, and this interpretation does not exclude
the paradoxically self-referential reading Persil washes whiter than Persil.

5.3. Argumentative self-reference in advertising

Self-referential arguments in advertising occur in many forms of circular rea-

soning. Most of them are elliptical and oblige the consumer to substitute the
missing links in the chain of arguments. In 2006, Unilever of Great Britain
launched a new advertising campaign under the name Dirt is good (http:// dig.asp, 16.05.06). The
paradoxical slogan Persil Dirt is good can only be understood as an elliptical
argument or more precisely pseudo-argument, whose pseudo-syllogistic line of
reasoning must be: Premise 1: Dirt is bad. Premise 2: Good Persil removes bad
dirt. Conclusion: Bad dirt removed by good Persil: how good!
Implicitly circular or quasi-circular arguments are quite frequent in adver-
tising. The quality of a product stated at the end of the elliptical argument is
already presupposed from its beginning. Such a rhetorical device suffers from
the fallacy of the petitio principii. Winston tastes good like a cigarette should
was a famous advertising slogan of the 1960s with a semi-circular way of argu-
ing (cf., 16.05.06). The
reason for the alleged quality of this cigarette is already only implied in its mere
18 Winfried Noth

being a cigarette. The conclusion only conrms what the general premise pre-
supposes: all cigarettes (should) taste good, and therefore this cigarette tastes
good, too. The world has changed. The dictionary also, was the slogan with
which Hachette launched a new dictionary. The two propositions of this slogan
sound like the major and the minor premise of a syllogism (All S is P and
Some S is P) calling for the conclusion that the new dictionary incorporates al
recent changes of the world of which it is a part. The second premise with its syn-
tactic and semantic parallelism to the rst creates an iconically self-referential
circular argument: the dictionary must be good since it reects the changes of
the changing world of which it is a part. Of course, the conclusion is not valid
since the dictionary could have changed from good to bad, or it could not have
changed at all. Furthermore, there is another circularity in the argument, since
a dictionary, being a part of the world that has changed, must change in a trivial
sense by necessity with every new edition.

5.4. (Intra)textual self-reference: Cinema and advertising

There are two major sources of (intra)textual self-reference, poetic features and
metatextual passages of a text about the text. As Jakobson has argued, poetic
features draw the readers attention towards the text as a text by means of recur-
rence, symmetry, rhyme, loops, or stylistic and rhetorical devices. The former
devices evince iconic self-reference since they are based on similarities and
forms of sameness, the latter testify to indexical self-reference, insofar as style
is indicative of an author, epoch, or otherness in general (cf. Noth 2005a). Self-
reference is particularly conspicuous in lmic loops (Manovich 1999: 187191)
in general and in the lm Run Lola Run (Tom Tykwer, D, 1998; http://www.lola-, 16.05.06), a lm that is self-referential already in its English title and
which returns to its beginning and anticipates its end several times.
Examples of metatextual self-reference are comments on the text, its narra-
tive form, its content and its structure, its plot, previous or subsequent chapters,
its beginning and its end. In the movies, textual self-reference occurs when the
lm begins and ends with a trailer marking its beginning by presenting its title
and its end by concluding with the message THE END in writing. In advertising,
the line Advertisement above the text refers to the text as a particular type of text
(and not one that belongs to the news reports, for example).Textual self-reference
of this kind in advertising runs the risk of being in conict with the goals of the
genre. The metamessage This it is an advertisement reminds the readers that
the message is one-sided and pursues the goal of inuencing the public for the
sole purpose of buying the product. Instead of saying (alloreferentially) Prod-
Self-reference in the media: The semiotic framework 19

uct X is good, the advertisement says: This message is an advertisement for

product X. Advertisements tend to avoid this kind of self-reference since the
credibility of commercials is generally low, and the admission that the message
is only publicity puts the efciency of the message at risk.

5.5. Intertextual and intermedial self-reference

Quotations, allusions, adaptations, inuences, borrowings from texts, lms, or

any other medium are the sources of intertextual self-reference. When several
media are involved, such as painting in the cinema, lms in games, or novels
in the lm, there is intermedial self-reference. In a way, borrowings from other
texts or media are certainly alloreferential, since there is reference from one
message to another so that the object of the quoting sign is a quoted sign from
which it differs. On the other hand, a lm A that quotes a lm B makes in-
tertextual reference to its own medium and not to the world which both lms
represent, and the TV spot that quotes another TV spot remains within the world
of advertising. These messages are intertextually self-referential to the degree
that its quotations remain within their own world beyond lms. Furthermore,
every quotation presupposes repetition and sameness which is the source of
iconic self-reference.
There are several kinds of loops and circularities in intertextual and inter-
medial self-reference which have been studied elsewhere. One of them is the
intermedial deja-vu effect which has often been exemplied with certain side
effects created by the news reports on September 11. In one of his Essays on
September 11 and related dates, Zizek (2002: 17) writes: The question we
should have asked ourselves as we started the TV screens on September 11 is
simply: Where have we already seen the same thing over and again? In our
context, only the aspect of self-reference addressed by Zizek is of relevance.
What Zizek reminds us of is that the TV pictures of the collapsing World Trade
Center on September 11 did not only arouse shock, horror, and despair but it
also created some feeling of deja-vu. In a way, the lm reports of the September
11 catastrophe in the news media only seemed to repeat the scenarios which the
genre of disaster movies had displayed for decades. The TV pictures seemed to
lack absolute novelty because the viewers had been all too familiar with similar
pictures of catastrophes, wars, destruction, and invasions by enemies and aliens,
some of them even in New York City. As Tim Dirkss (2006) list of the Greatest
disaster lm scenes demonstrates, the worlds highest glass tower building had
been aame before (although in a ctional version in which the towers were
located in San Francisco), namely in the movie The Towering Inferno of 1974,
20 Winfried Noth

and there are dozens of catastrophe lm scenarios resembling the September 11

events with plane crashes, terrorists of many kinds, out for control res, nuclear
annihilations, and even end of the world scenarios. On September 11, the media
had been ahead of the event; reality seemed to lag behind. In short, the deja-vu
effect on the screen accounts for a particular form of self-reference in the me-
dia, which consists in the repetition of the same scenario, whether ctional of

5.6. Enunciative self-reference: Examples from the movies

Enunciation has been a key concept in lm semiotics since Christian Metz

(Buckland 2000; Buckland 1995). It pertains to the communicative situation
of a message, the way the addresser interacts with the addressee of a message.
In verbal communication, the study of enunciation is concerned with the many
voices of the speaker, especially the narrator, their intentions and modes of
manipulating the addressee (Santaella and Noth 2004: 113126).
Enunciative (or communicative) self-reference occurs when the author, the
narrator, the actor, the reader, or the spectator become the topic of the mes-
sage. Instead of presenting or representing ideas or events in the world from
elsewhere, the text deals with its own communicative context, its function, the
presuppositions of its narration, and the text has thus its own communicative
situation as its topic. Alfred Hitchcock, for example, leaves his place behind the
camera to mingle with the actors on the screen, reminding the spectators of his
permanent presence in his lm, even when invisible. In a so-called screen pas-
sage, one of Woody Allens actors in The Purple Rose of Cairo (US, 1985) even
steps out from his role as an actor on the screen to mingle with the audience.
For decades, lms used to conceal the traces of their production, for example
the details of the lm studio and the staff behind the scenes, as much as possible
with the purpose of creating a perfect real-life illusion. Alloreference was on the
agenda. Modern digital lm technology has increased the potential for illusion
and enables the alloreferential representation of previously impossible reali-
ties. The audience is no longer restricted to viewing the sinking Titanic above
the sea level but can also participate in the drama below the water surface. As
a result of the new possibilities of digital picture manipulation, it is no longer
possible to distinguish real shots from digital additions (cf. Manovich 1999).
The alloreferential perfection of this pictorial manipulation makes us forget its
digital construction. More and more accurate representations and the increasing
possibilities of representing the world in all of its visual facets create the illusion
of a growth of alloreference of the medium.
Self-reference in the media: The semiotic framework 21

On the other hand, there are those new strategies and effects of illusion that
conduce from a world of the real to the awareness of a world of simulation.
The more the pictures distance themselves from reality, the more doubts in the
authenticity and plausibility of the feigned worlds arise. The ever repetitive ef-
fects of simulation shatter the audiences belief in the communicative contract
between lmmaker and audience. Films deal with the premises and conditions
of this communicative contract as a result of a critical reection of this situ-
ation. It eventually becomes the subject matter of lmed representation itself:
lmmakers appear on the screen in the role of actors, actors play the role of
the producer, and last but not least, they leave the screen entirely in an effec-
tive screen passage to join the audience in the cinema (cf. Stam 1992; Karpf,
Kiesel, and Visarius 1996, and Withalm, this vol.). New forms of pragmatic self-
reference are emerging with interactive lms in which the spectator becomes
the producer of his own viewing.
Enunciative self-reference is of a different kind in computer games. Not
unlike other games, reference to the world is secondary in computer games.
Games do not want to simulate real life. In contrast to other forms of play, the
computer game offers still more possibilities for the creation of new worlds.
Their virtual character is highly self-referential from the beginning on. Play-
ers can interact with the program code and thus control the referential action,
and they can become producers of the text. In which way communicative self-
referential autonomy of the players is actually attained remains open for further

5.7. Iconic self-reference: Loops, repetitions, and recursion

Among the most important iconic modes of textual, intertextual, and intermedial
self-reference are recursion and recurrence. Recursion, the circular or loop-like
return to an earlier point in the same text, in other texts, or media, is similar
to recurrence, the principle of repetition. There are diverse functions and ef-
fects. In music, art, and literature, the nontrivial recurrence of varied forms is
a source of aesthetic effects: repetitio delectat. As the trivial repetition of the
same, recurrence and recursion are signs of the trivial, for example in soap op-
eras. In games, recursion can even be a means of punishment, for example in
the classical ludo, where the return to the point of departure can be an element
of suspense, satisfaction, or disappointment.
In advertising, repetitive campaigns a la Marlboro exemplify best the prin-
ciple of intertextual recurrence and hence intertextual self-reference with their
permanent return to the same scenario. Evidently, the Marlboro man does not
22 Winfried Noth

only refer alloreferentially to scenes of the myth of the Wild West but also
self-referentially to the never changing world of the Marlboro posters.
In the movies, too, we have become accustomed to intertextual self-reference.
The most recent James Bond lms, for example, are hardly discussed in terms of
what they represent. Instead, intertextuality is the topic as public interest focuses
on the question of how these lms can be compared with those which preceded.
Nina Bishara, in a comment on this paper, describes this form of self-reference in
the most recent James Bond movie Die Another Day (UK/US, 2002) as follows:
Not only are well-established and recurrent James Bond themes taken up (e.g.
good against evil, the pre-titles sequences, My name is Bond James Bond
etc.), the 20th Bond movie also has strong allusions to the previous movies so
that the real connoisseur can indulge in a guessing game. One scene with Bond
girl Halle Barry resembles a scene with Ursula Andress from the rst Bond
movie Dr. No (1962) and props that played an important role in previous movies
reappear. Allegedly, each of the previous lms is included in the new Bond
movie in some form or other. Moreover, cases of intermedial self-reference can
be found in the product placements of cars (Ford, Jaguar, Aston Martin), Bonds
favorite champagne (Bollinger), spy tools such as the watch by Omega or the
Ericsson mobile phone. Even print advertisements self-referentially refer back
to these product placements, for example a BMW ad which advertises the fact
that the new BMW model appears in the James Bond movie The World is not
Enough (1999). Another intermedial form of self-reference can be observed in
the video clip for the title song by Madonna for the 20th Bond movie which is
also called Die Another Day and which re-enacts scenes from the movie.
One of the characteristic features of digital lm is the increasing possibility
of self-repetition in the form of loops, as in Run Lola Run, where several varia-
tions of the same event are connected by means of time loops. There is no true
beginning and no real end when this form of textual self-reference predomi-
nates. There is nothing but a sequence of recursive loops. Loops and recursivity,
however, are not only modes of repetition; they are as well loci of variation (cf.
Winkler 2004: 170182).
In computer games, recursion in the form of textual self-reference is still
more advanced. For example, the player can chose a certain point of departure
in the game and then try out a number of possible variations of the same strategy.
Furthermore, the well-known order Return to X (i.e., to a previous position)
clearly exemplies textual self-reference. Textually self-referential recursion is
probably the most characteristic feature of computer games, since the underlying
algorithms are not only the basis of the production but also of the execution of
the game.
Self-reference in the media: The semiotic framework 23

6. Self-reference between subversion, play, and art

Self-reference in the media is hardly subversive, as it might seem when it results

in paradox; its functions tend to be predominantly playful and aesthetic. Being
directed towards itself without any ambition to represent, the world beyond the
signs, a self-referential message cannot transmit a subversive message in the
sense of a message that wants to undermine social or cultural values.
However, a self-referential message can be subversive in the sense of breaking
with the codes and conventions of the genre. In this sense, self-reference is most
subversive in advertising, where the self-referential message is incompatible
with the goal of advertising services or products. In lms, too, self-reference has
been a subversive device of the genre when it was rst introduced as a stylistic
device. Both media, however, have always shown self-referential elements in
their poetic and aesthetic dimensions, since poetry and art are self-referential
by nature.
In games, by contrast, self-reference is not the exception, but the rule, since
play and playful conduct have always tended to be self-referential. The values of
chess gures, such as the king, the queen, the bishop, or the pawn, nd very little
correspondence in real life. Carnival seems to be revolutionary in allowing the
peasant to become the prince, but carnival has never been suppressed by those
in power, since they quickly recognized that playful conduct is self-oriented and
cannot develop a revolutionary impetus.
Computer games, however, have begun to create new realities and to simulate
virtual realities which raise the question of subversion in a new way. Do they
conduce to merely self-referentially playful activities, to play for plays sake,
as in chess, or do they create virtual realities with the potential to subvert the
conventional values of culture and society?


Andersen, Peter Bgh, Claus Emmeche, Niels Ole Finnemann and PederVoetmann Chris-
tiansen (eds.)
2000 Downward Causation. Aarhus: University Press.
Babcock, Barbara A.
1980 Reexivity: Denitions and discriminations. Semiotica 30: 114.
Barthes, Roland
1967 Systeme de la mode. Paris: Seuil.
Bartlett, Steven J.
1987 Varieties of self-reference. In: Steven J. Bartlett and Peter Suber (eds.),
Self-Reference: Reections on Reexivity, 530. Dordrecht: Nijhoff.
24 Winfried Noth

1992a The role of reexivity in understanding human understanding. In:

Steven J. Bartlett (ed.), Reexivity: A Source-Book in Self-Reference,
318. Amsterdam: North Holland.
Bartlett, Steven J. (ed.)
1992b Reexivity: A Source-Book in Self-Reference. Amsterdam: North Hol-
Bartlett, Steven J. and Peter Suber (eds.)
1987 Self-Reference: Reections on Reexivity. Dordrecht: Nijhoff.
Baudrillard, Jean
1976 Lechange symbolique et la mort. Paris: Gallimard.
1981 Simulacres et simulation. Paris: Galilee.
1991 La guerre du Golfe na pas eu lieu. Paris: Galilee.
Bettetini, Gianfranco
1971 Lindice del realismo. Milan: Bompiani.
Bishara, Nina
forthcoming Selbstreferenzielle Werbung.
Bleicher, Joan-Kristin
1999 Unterhaltung in der Endlosschleife. In: Michael Latzer, Ursula Maier-
Rabler and Gabriele Siegert (eds.), Die Zukunft der Kommunikation,
109114. Innsbruck: Studien-Verlag.
Block, Friedrich W.
1999 Beobachtung des ICH. Zum Zusammenhang von Subjektivitat und
Medien am Beispiel experimenteller Poesie. Bielefeld: Aisthesis.
Blobaum, Bernd
1999 Selbstreferentialitat und Journalismus. In: Michael Latzer, Ursula
Maier-Rabler and Gabriele Siegert (eds.), Die Zukunft der Kommu-
nikation, 181188. Innsbruck: Studien-Verlag.
Bolter, David and Richard Grusin
1999 Remediation: Understanding New Media. Cambridge, MA: MIT Press.
Broich, Ulrich and Manfred Pster (eds.)
1985 Intertextualitat. Tubingen: Niemeyer.
Brook, Andrew and Richard C. DeVidi (eds.)
2001 Self-Awareness and Self-Reference. Amsterdam: Benjamins.
Buckland, Warren
2000 The Cognitive Semiotics of Film. Cambridge: University Press.
Buckland, Warren (ed.)
1995 The Film Spectator: From Sign to Mind. Amsterdam: University Press.
Buttner, Stefan and Andrea Esser (eds.)
2001 Unendlichkeit und Selbstreferenz. Wurzburg: Konigshausen & Neu-
Colapietro, Vincent
1989 Peirces Approach to the Self. Albany: State University of New York
Self-reference in the media: The semiotic framework 25

Dirks, Tim
2006 Greatest Disaster Film Scenes. (16.05.06).
Dunne, Michael
1992 Metapop: Self-Referentiality in Contemporary American Popular Cul-
ture. Jackson: University Press of Mississippi.
Eco, Umberto
1984 Mirrors. In: Umberto Eco, Semiotics and the Philosophy of Language,
202226. Bloomington: Indiana University Press.
Evans, Gareth
1991 The Varieties of Reference. Oxford: Clarendon.
Fitch, Frederic B.
1987 Self-reference in philosophy. In: Steven J. Bartlett and Peter Suber
(eds.), Self-Reference: Reections on Reexivity, 221230. Dordrecht:
Frieske, Michael
1998 Selbstreferentielles Entertainment: Televisionare Selbstbezuglichkeit
in der Fernsehunterhaltung. Wiesbaden: Deutscher Universitatsver-
Geach, Peter T.
1970 Reference and Generality, 2nd ed. Ithaca, NY: Cornell University
Goebel, Gerhard
1986 Notizen zur Semiotik der Mode. In: Silvia Bovenschen (ed.), Die Lis-
ten der Moden, 458479. Frankfurt: Suhrkamp.
Helbig, Jorg
2001 Intermedialitat: Eine Einfuhrung. Frankfurt: Suhrkamp.
Hempfer, Klaus W.
1976 Poststrukturale Texttheorie und narrative Praxis. Munich: Fink. 1982
Die potentielle Autoreexivitat des narrativen Diskurses. In: Eberhard
Lammert (ed.), Erzahlforschung, 130156. Stuttgart: Metzler.
Hoffmeyer, Jesper
1996 Signs of Meaning in the Universe. Bloomington: Indiana University
Hofstadter, Douglas R.
1979 Godel, Escher, Bach. New York: Basic Books.
1985 Metamagical Themas. New York: Basic Books.
Huber, Werner, Martin Middeke and Hubert Zapf (eds.)
2005 Self-Reexivity in Literature. Wurzburg: Konigshausen & Neumann.
Jahraus, Oliver and Nina Ort (eds.)
2003 Theorie Prozess Selbstreferenz: Systemtheorie und transdisziplina-
re Theoriebildung. Konstanz: UVK.
26 Winfried Noth

Jay, Paul
1984 Being the Text: Self-Representation from Wordsworth to Roland
Barthes. Ithaca: Cornell University Press.
Johansen, Jrgen Dines
2002 Literary Discourse: A Semiotic-Pragmatic Approach to Literature.
Toronto: University of Toronto Press.
Karpf, Ernst, Doron Kiesel and Karsten Visarius (eds.)
1996 Im Spiegelkabinett der Illusionen. Filme uber sich selbst. Marburg:
Katz, Jerrold J.
2004 Sense, Reference, and Philosophy. Oxford: University Press.
Kempson, Ruth M.
1977 Semantic Theory. Cambridge: Cambridge University Press.
Kienzle, Bertrand and Helmut Pape (eds.)
1991 Dimensionen des Selbst: Selbstbewusstsein, Reexivitat und die Be-
dingungen von Kommunikation. Frankfurt: Suhrkamp.
Kirchmann, Kay
1996 Zwischen Selbstreexivitat und Selbstreferentialitat. Uberlegungen
zur Asthetik des Selbstbezuglichen als lmische Modernitat. In: Ernst
Karpf, Doron Kiesel and Karsten Visatius (eds.), Im Spiegelkabinett
der Illusionen. Filme uber sich selbst, 6786. Marburg: Schuren.
Klein, Naomi
2000 No Logo. Toronto: Knopf Canada.
Knowles, Elizabeth (ed.)
2004 The Oxford Dictionary of Modern Quotations, 2nd ed. Oxford: Oxford
University Press.
Kohring, Matthias
1999 Selbstgesprache: Der Begriff der Selbstreferenz und das Fallbeispiel
des Journalismus. In: Michael Latzer, Ursula Maier-Rabler and
Gabriele Siegert (eds.), Die Zukunft der Kommunikation, 189198.
Innsbruck: Studien-Verlag.
Korzybski, Alfred
1933 Science and Sanity. Lakeville, CN: Int. Non-Aristotelian Library.
Krah, Hans
2005a Selbstreferentialitat, Selbstbezuglichkeit, Selbstreferenz: Die Begriffe
und ihr Bedeutungsspektrum. Zeitschrift fur Semiotik 27(12): 322.
Krah, Hans (ed.)
2005b Selbstreferenz und literarische Gattung. Special issue of Zeitschrift
fur Semiotik 27(12). Tubingen: Stauffenburg.
Lawson, Hilary
1985 Reexivity: The Post-Modern Predicament. London: Hutchinson.
Liebrand, Claudia and Irmela Schneider (eds.)
2002 Medien in den Medien. Cologne: DuMont.
Self-reference in the media: The semiotic framework 27

Lipman, Jean and Richard Marshall (eds.)

1978 Art about Art. New York: Dutton.
Luhmann, Niklas
1984 Das Kunstwerk und die Selbstreproduktion der Kunst. Deln 3: 51
1993 Zeichen als Form. In: Dirk Baecker (ed.), Probleme der Form, 45
69. Frankfurt: Suhrkamp. Transl. M. Irmscher. 1999. Sign as form.
Cybernetics and Human Knowing 6.3: 2137.
1995 Die Realitat der Massenmedien. Opladen: Westdeutscher Verlag.
Lyotard, Jean-Francois
1979 La condition postmoderne. Paris: Minuit. Engl. (1984). The Postmod-
ern Condition, G. Bennington and B. Massumi (transl.). Minneapolis:
Minnesota University Press.
McLuhan, Marshall
1964 Understanding Media: The Extensions of Man. London: Routledge
and Kegan.
Manovich, Lev
1999 What is digital cinema? In: Peter Lunenfeld (ed.), The Digital Dialec-
tics: New Essays on the New Media, 172192. Cambridge, MA: MIT
Marcus, Solomon
1997 Media and self-reference: The forgotten initial state. In: Winfried Noth
(ed.), Semiotics of the Media, 1545. Berlin: Mouton de Gruyter.
Menninghaus, Winfried
1987 Unendliche Verdopplungen: Die fruhromantische Grundlegung der
Kunsttheorie im Begriff absoluter Selbstreexion. Frankfurt: Suhr-
Metscher, Thomas
2003 Ahnlichkeit und Selbstreprasentation. In: Silja Freudenberger and
Hans Jorg Sandkuhler (eds.), Reprasentation, Krise der Reprasenta-
tion, Paradigmenwechsel, 271299. Frankfurt: Lang.
Muller, Jurgen
1996 Intermedialitat: Formen moderner kultureller Kommunikation. Mun-
ster: Nodus.
Munch, Dieter
1992 Referenz, Referenztheorie. In: Joachim Ritter and Karlfried Grunder
(eds.), Historisches Worterbuch der Philosophie, Volume 8, 386387.
Basel: Schwabe.
Myers, C. Mason
1966 The circular use of metaphor. Philosophy and Phenomenological Re-
search 26: 391402.
28 Winfried Noth

Noth, Winfried
1977 Dynamik semiotischer Systeme. Stuttgart: Metzler.
2000a Handbuch der Semiotik, 2nd rev. ed. Stuttgart/Weimar: Metzler.
2000b Selbstreferenz in systemtheoretischer und semiotischer Sicht. In:
Achim Barsch, Gebhard Rusch, Reinhold Viehoff and Friedrich W.
Block (eds.), Festsite Siegfried J. Schmidt. (16.05.06) and 2002
in: etc: Empirische Text- und Kulturforschung 2.2002: 17.
2001 Autorreferencialidad en la crisis de la modernidad. Cuadernos: Re-
vista de la Facultad de Humanidades y Ciencias Sociales 17: 365
2002 Semiotic machines. Cybernetics and Human Knowing 9(1): 522.
2005a The art of self-reference in Edward Lears limericks. International
Journal of Germanic Linguistics and Semiotic Analysis 10(1): 4766.
2005b Formen der Selbstreferenz in den Medien. In: Sigrid Schade, Thomas
Sieber and Georg C. Tholen (eds.), SchnittStellen, 133146. Basel:
2006 Reprasentation und Referenz bei Peirce. In: Hans-Jorg Sandkuhler
(ed.). Theorien und Begriffe der Reprasentation (= Schriftenreihe der
von derVolkswagenstiftung geforderten Forschergruppe Reprasentati-
on 1), 4361. Bremen: Universitat.
2007 Narrative self-reference in a literary comic: M.-A. Mathieus LOrigi-
ne. Semiotica 165:173190.
Noth, Winfried (ed.)
2006 Semiotic Bodies, Aesthetic Embodiments, and Cyberbodies. Kassel:
University Press.
Noth, Winfried and Christina Ljungberg (eds.)
2003 The Crisis of Representation: Semiotic Foundations and Manifesta-
tions in Culture and the Media. (= Special Issue of Semiotica 143.14).
Noth, Winfried and Anke Hertling (eds.)
2005 Korper Verkorperung Entkorperung. Kassel: University Press.
Ort, Claus-Michael
2003 Medienwechsel und Selbstreferenz. Tubingen: Niemeyer.
Paech, Joachim
1998 Zur theoretischen Grundlegung von Intermedialitat. In: Jorg Helbig
(ed.), Intermedialitat.Theorie und Praxis eines interdisziplinaren For-
schungsgebiets, 1430. Berlin: Erich Schmidt.
Pattee, Howard H.
1995 Evolving self-reference: Matter symbols, and semantic closure. Com-
munication and Cognition 12(12): 928.
Pavlicic, Pavao
1993 What is the purpose of autoreferentiality? Neohelicon 20(1): 97106.
Self-reference in the media: The semiotic framework 29

Peirce, Charles Sanders

193158 Collected Papers, vols. 16, ed. Charles Hartshorne and Paul Weiss,
vols. 78, ed., A. W. Burks. Cambridge, MA: Harvard University
Press. Quoted as CP.
196366 The Charles S. Peirce Papers. 30 reels, microlm edition. Cambridge,
MA: The Houghton Library, Harvard University Microproductions.
Quoted as MS; see Robin, comp. 1967.
Peitgen, Heinz-Otto, Hartmut Jurgens and Dietmar Saupe
1992 Chaos and Fractals. Berlin: Springer.
Petersen, Christer
2003 Der postmoderne Text. Kiel: Ludwig.
Potthast, Ulrich
1971 Uber einige Fragen der Selbstbeziehung. Frankfurt: Klostermann.
Prigogine, Ilya and Isabelle Stengers
1984 Order out of Chaos. New York: Bantam.
Rajewsky, Irina O.
2002 Intermedialitat. Tubingen: Francke.
Robin, Richard S.
1967 Annotated Catalogue of the Papers of Charles S. Peirce. Amherst:
University of Massachusetts Press. (MS refers to the numbers of this
Santaella, Lucia
2004 Corpo e comunicacao. Sao Paulo: Paulus.
Santaella, Lucia and Winfried Noth
2004 Semiotica e comunicacao. Sao Paulo: Hacker.
Scheffel, Michael
1997 Formen des selbstreexiven Erzahlens. Tubingen: Niemeyer.
Scheutz, Matthias
1995 Ist das der Titel eines Buches? Selbstreferenz neu analysiert. Vienna:
Schmidt, Siegfried J.
1994 Kognitive Autonomie und soziale Orientierung. Frankfurt: Suhrkamp.
Schmidt, Siegfried J. and Brigitte Spie
1996 Die Kommerzialisierung der Kommunikation: Fernsehwerbung und
sozialer Wandel 19561989. Frankfurt: Suhrkamp.
Schoppe, Arno
1995 Theorie paradox: Kreativitat als systemische Herausforderung. Hei-
delberg: Carl Auer.
Shir, Jay
1978 Symbolism and autosymbolism. Journal of Aesthetics and Art Criti-
cism 37(1): 8189.
30 Winfried Noth

Siedenbiedel, Catrin
2005 Metaktionalitat in Finnegans Wake. Wurzburg: Konigshausen &
Smuda, Manfred
1970 Becketts Prosa als Metasprache. Munich: Fink.
Spielmann, Yvonne
1998 Intermedialitat. Das System Peter Greenaway. Munich: Fink.
Stam, Robert
1992 Reexivity in Film and Literature: From Don Quixote to Jean-Luc
Goddard. New York: Columbia University Press.
Waugh, Patricia
1984 Metaction. London: Methuen.
Weber, Stefan
1999 Das System Journalismus: Oszillieren zwischen Selbstreferenz und
Fremdsteuerung. In: Michael Latzer, Ursula Maier-Rabler and
Gabriele Siegert (eds.), Die Zukunft der Kommunikation, 161180.
Innsbruck: Studien-Verlag.
Whitehead, Alfred N. and Bertrand Russell (eds.)
1910 Principia Mathematica. Cambridge: Cambridge University Press.
Whiteside, Anna
1987 The double-bind: Self-referring poetry. In: Anna Whiteside and
Michael Issacharoff (eds.), On Referring in Literature, 1432. Bloom-
ington: Indiana University Press.
Winkler, Hartmut
2004 Diskursokonomie. Frankfurt: Suhrkamp.
Withalm, Gloria
1995 Fernsehen im Fernsehen im Fernsehen. . . : Selbstreferentielle Zeichen-
prozesse. Vienna: OGS/ISS.
Wittig, Susan
1979 Architecture about architecture: Self-reference as a type of architec-
tural signication. In: Seymour Chatman, Umberto Eco and Jean-
Marie Klinkenberg (eds.), A Semiotic Landscape, 970978. The
Hague: Mouton.
Wolf, Werner
2001 Formen der literarischen Selbstreferenz in der Erzahlkunst. In: Jorg
Helbig (ed.), Erzahlen und Erzahltheorie im 20. Jahrhundert, 4984.
Heidelberg: Winter.
Zavala, Lauro
2000 Una tipologa estructural de estrategias metaccionales en cine y
literatura. Xochimilco: Universidad Autonoma Metropolitana.
Zizek, Slavoj
2002 Welcome to the Desert. London: Verso.
Distortion, fabrication, and disclosure in a
self-referential culture:The irresistible force of reality

Vincent Colapietro

1. Introduction: The cultural frame

Reexivity has been taken to be a hallmark of postmodernity (Lawson 1985;

Bartlett 1992; Dunne 2001; Noth, this vol., Part I). There is unquestionably a
warrant for doing so. For ours is manifestly a time in which self-reference, espe-
cially in its most intensied forms and paradoxical implications, is a pervasive
and thus inescapable facet of our lives.1 The forms and effects of this reex-
ivity are far from supercial: they are, indeed, integral to the structures of our
lives, including our modes of embodiment, engagement, and affectivity (Lenoir
2000; Gitlin 2002: 6; Santaella 2003). Arguably, our propensity toward irony,
our incredulity toward metanarratives, our longing for immediacy, our ambiva-
lence regarding authenticity, and various other factors are intimately connected
to the reexivity so characteristic of our time. Whether or not these connections
can be established, there is little doubt about reexivity being a hallmark of
postmodernity. Though its character, forms, and signicance invite numerous
questions, this connection is itself beyond question.
This is nowhere more evident than in the forms and effects of reexivity
generated and intensied by the interlocking networks of contemporary media,
ranging from long-established ones such as newspapers to cutting-edge tech-
nologies such as computer simulation of war games (Lenoir 2000). The news
being broadcast via such media as television, radio, and blogs on the world
wide web (www) constitutes unquestionably intricate and arguably insular net-
works of self-citation and self-commentary. The news reports on the news as
much as anything else. Individuals are celebrities for no other reason than being
celebrated in the media. Popular entertainment constructs a world of complex
allusions to the fabrications of the entertainment industry itself. Nowhere is
intertextuality more manifest than, for example, in the structures of cinema. In
light of all of this, we seemingly inhabit a world of our own making, one in which
claims about reality are viewed with deep suspicion, if not outright dismissal.
32 Vincent Colapietro

Even the character of the real world of global politics has come increasingly
to take on that of the fabricated world of the entertainment industry (a point to
which I shall return).
The crisis of representation, insofar as it is generated by the inherent dynamic
of, and theoretical reections on, mass media, is inseparably connected to the
forms of reexivity so pervasive in a culture so radically structured by such
media. The possibilities of reference and representation in anything approx-
imating a straightforward sense seem to be increasingly and paradoxically
(Noth 2003: 911) limited to those of self-reference and self-representation.
Indexical signs in their most rudimentary form are assumed at every turn, but
theoretically erased.
In an increasingly self-referential culture, the relationship of the self to itself
paradoxically works to shatter the putative unity and stability of the self. The
self is, on Richard Rortys account at least, a centerless web of contingent beliefs
and desires (see, e.g., 1989: 10). Such an account seems to be tailored to the
sensibility of those who are mobile nodes in an expanding network of self-
referential media. For such selves, self-consciousness tends to assume a form
in which self-identication becomes increasingly problematic and awkward, but
at the same time a form in which self-estrangement does not ordinarily feel self-
lacerating. The self is affectively attached to an identity or endeavor from which
it can readily distance itself, very often through ironic self-commentary.
In an increasingly self-referential culture such as our own, our relationship
to a common reality is no less radically altered than our relationship to our in-
nermost selves. Indeed, our condence in our ability to refer to a shared reality
tends to be deeply shaken, if not entirely undermined. Insofar as we (at any rate,
you and I individually) lack such condence, the reality of community itself
becomes problematic.2 Our ever deepening skepticism regarding the ability to
appeal to a shared understanding or to refer to a common reality is, at once,
widely felt and communally debilitating.3 More and more of us feel this way
and the more we actually do the less we are possible (i.e., the less likely com-
munal norms and ideals are concretely embodied in shared habits, practices,
and institutions). The solidarity requisite for political opposition grounded in
shared concerns, grievances, and hopes can be, for example, more ephemeral (at
least less efcacious) than the fabricated sense of national unity. Actual forms
of solidarity are rendered more and more insubstantial and ineffective, while
fabricated or, more precisely, fantastic forms of identity become more and more
central and energizing.
In the writings of C. S. Peirce we encounter a theorist who combines a robust
sense of reality with an acute sense of the fragile and (in no small measure) illu-
sory character of the self. For this and other reasons, some of his most distinctive
Distortion, fabrication, and disclosure in a self-referential culture 33

contributions (above all, his doctrine of signs, his version of pragmatism, and
more broadly his account of inquiry) seem especially relevant to the task of
making sense out of our selves and our world, as both are encountered in an
increasingly self-referential culture. For Peirce, the self is in no small measure a
fabrication (CP 4.68; Short 1997: 307); and, in turn, our only access to reality is
through the mediation of signs. Even so, a fragile, ssured, and fallible self, as
much (if not more) illusory as real, might have adequate resources (if only barely
adequate resources) to know reality. But such a self can do so only in conjunction
with other such selves, forming thereby a community of inquirers dened by a
commitment to open-ended self-criticism. To paraphrase Peirce, the idea of real-
ity is inseparably connected to not only the idea but also the reality of community.

2. A contemporary example in light of Peircean realism

A pragmatist is, according to the originator of this doctrine, a theorist for whom
the distinction of reality and ction is naturally highly prominent. But I cannot
see how such a mind can deny the reality of generals (MS 313, p. 19). Though
highly prominent the distinction between fact and fancy, reality and gment
is continuous, such that the one shades into the other (facts always begin in
some measure imaginative depictions and fancies real affairs). For a pragmatist
committed to maintaining this distinction, however, generals are in countless
instances no mere ctions: the reality of generals needs to be acknowledged
along with the ubiquity of imagination and the otherness of reality.
The mediated (or semiotic) realism of Peirce is able to do fuller justice to
the complex actualities of contemporary culture than more inuential theories
of radical constructivism (Colapietro 2000). For theoretical and practical pur-
poses, the language of disclosure must not be completely jettisoned in favor of
the language of distortion or that of fabrication (or construction). One of the
purposes of this paper is to make this argument as briey and yet as pointedly as
possible. Disclosure is inevitably partial, perspectival, and fallible, just as facts
are (as the root of the word suggests) something wrought (facts being, in a sense,
fabrications or constructions). Even so, our use of signs generates possibilities
wherein some facets of reality are disclosed, are revealed often to our surprise
and even frustration. I want to make this argument for Peircean realism in the
teeth of hyper reexivity in the media, as illustrated in an editorial in the New
York Times (one wherein the world of politics is mapped onto that of lm).
In an editorial entitled The two wars of the worlds (New York Times, July
3, 2005), Frank Rich compares George W. Bushs speech at Fort Bragg and
Steven Spielbergs just released War of the Worlds.4 The premise of Richs
34 Vincent Colapietro

piece is that both the president and the director are aiming at inducing terror
in their audience, Bush for the purpose of reversing the mounting skepticism
regarding his adventure in Iraq, Spielberg for the purpose of entertainment,
prot, and possibly edication.5 Both are in fact reruns or remakes: Spielberg
is offering (according to the website for this movie) a contemporary reading
of H. G. Wellss seminal classic, while Bush was giving at Fort Bragg almost
the identical televised address, albeit with four fewer 9/11 references, at the
Army War College in Pennsylvania in May 2004. The dazed response of the
military audience to Bushs summer rerun however stands in marked contrast
to the excited response of audiences to Spielbergs remake. Much of this might
be the result of Bushs ineptitude in crafting a narrative. At least, Rich argues
this point; and his argument is rich in allusions to the history of cinema and
Mr. Spielbergs movie illuminates [. . . ] how Mr. Bush has ubbed the basic
storytelling essential to sustain public support for his Iraq adventure. The
president has made a tic of hammering [home his points] in melodramatic
movie tropes: good vs. evil, youre with us or youre with the terrorists,
wanted dead or alive, bring em on, mission accomplished. When you
relay a narrative in that style [Rich continues], the audience expects you to
stick to the conventions of the genre; the story can only end with the cavalry
charging in to win the big nal battle. Thats how Mr. Spielberg deploys his
platoons, Saving Private Ryan-style, in War of the Worlds. By contrast,
Mr. Bush never marshaled the number of troops needed to guarantee Iraqs
security and protect its borders; he has now dened mission accomplished
down from concrete victory to the inchoate spreading of democracy. To start
off sounding like [General George] Patton and end up parroting [President]
Woodrow Wilson is tantamount to ambushing an audience at a John Wayne
movie with a nal reel by Frank Capra.

The conclusion of Richs piece is also especially worthy of our attention,

since he not only continues to frame his argument in reference to cinema but
also casts Bush himself in the role of an alien. The current wars unpopularity
now matches the Gallop ndings during the Vietnam tipping point, the summer
of 1968. For a number of reasons, its the Bush presidency, not the insurgency,
that will [soon] be in its last throes. The gure of a dying alien is, Rich implies,
an apt image for this failing president: Is the commander in chief so isolated
in his bubble that he does not realize this [that he, not the insurgency, is soon
to be in his last throes]? He accordingly concludes with a piece of his advice:
G.W.B., phone home.
The cartoon by Barry Blitt accompanying Frank Richs editorial underscores
the current war as a media event in which the signicance of the event itself
Distortion, fabrication, and disclosure in a self-referential culture 35

invites comparison to cinema. In this cartoon, War in the Gulf: The Sequel
is being Held Over at Quagmire Theatre. But, in the course of depicting
Bushs speech as a summer rerun and media event (albeit one of lackluster
quality), Rich asserts: Much of what Bush said in this speech was, as usual, at
odds with reality. The comparison between The Two Wars of the Worlds is
consequently intended not to erase the distinction between depiction and reality
(much less that between fabrication and disclosure) but to highlight, in reference
to Bushs words, the clash between what has been actually asserted and what
can be responsibly established. What Peircean realism tries to secure, above all
else, is the colloquial sense of what is meant when a journalist or anyone else
claims that an utterance, report, or other use of signs is at odds with reality.
The very meaning of reality is partly derived from that which has the ca-
pacity to thwart even our most strenuous endeavors or to disrupt our habitual
responses.6 Often, our present selves, as concrete representatives of the determi-
nate past carried into and beyond the actual present (above all) by the tenacity of
our habits, cannot effectively negotiate the scene in which we are entangled and,
of greater signicance, the one in which we will continue to be entangled. When
effort and habit prove to be ineffective (when we cannot negotiate the scene in
which we are entangled), we need to renegotiate the terms of our coexistence
with the things and persons around us. Indeed, the metaphor that seems best to
capture the complexity of this process is that of renegotiation. This metaphor
suggests a political as well as communal process, one in which power is brokered
and people repositioned vis-a-vis one another.
The real is that which brings to light unsuspected deciencies in the cumu-
lative results of our ongoing negotiations. For the Peircean realist, however, the
locus of the real is not so much at the origin of any process of renegotiation as
at the provisional conclusions of an ongoing process (CP 8.12, 8.208, 8.284).
The real is what the course of such renegotiations has forced us to acknowledge,
above all, to acknowledge as what we would need to consider if the pursuit of
our purposes is to be successful.
Human purposes abound and extra-human factors are interwoven with what
is, even in a seemingly simple, single deed, best seen as the dramatic, simul-
taneous enactment of multiple purposes. The multiplicity of our purposes and
the complexity of our circumstances, the propulsive force of our present habits
and the often upsetting presence of environing objects and their characteristic
dispositions, all point to the abiding need to acknowledge real externality and to
articulate in more concrete terms than we have thus a robust realism (Colapietro
2000). The irrepressible force of reality is such that, in however coded and dis-
torted a form, its traces7 are innumerable and (to some extent) both discernible
and decipherable (both identiable and interpretable). This force is also such
36 Vincent Colapietro

that we can position ourselves to feel more fully and to discern more nely its
efcacy, but only to the extent we adhere to rigorous ideals of self-critique.
We can observe, in the self-referential tendencies evident in various media,
a multiplicity of functions. Such tendencies seem to serve predominantly ludic
and aesthetic functions (Noth), though Hilary Lawson and others argue that
they contribute to the largely unacknowledged enhancement of critical con-
sciousness in a world where ultimate truths and absolute certainty are so widely
taken to be cultural illusions. Some theorists even take these tendencies to be
inherently subversive, to carry by their own momentum the power to undermine
the sanctity of traditional authorities and to disclose the historicity of allegedly
timeless truths. Others (and I agree with them) contend that subversion is not
so easily accomplished (Noth). But, in addition to the ludic, aesthetic, and sub-
versive functions just mentioned, I take it to be important, even urgent, to stress
the critical function made possible by the open-ended reexivity so character-
istic of distinctively human uses of signs and media (a function distinct from
that conceived by Lawson and others who seem to make critical consciousness
an inevitable outcome of the inherent dynamic of subversive tendencies char-
acteristic of reexive media). Self-reference is a condition for self-criticism
and, in turn, self-criticism is indispensable for responsibly establishing what
is so. In some contexts, self-reference does not serve to enclose us in an ever
more insular world, but rather works to expose us to an always somewhat un-
predictable domain in which ongoing revision, and at least occasionally radical
revision, of our habitual modes of description and explanation is the price to
be paid for framing a reliable account of the world (for disclosing as accu-
rately as our nitude and fallibility allow the contours, facets, and textures of
Such, at least, is the central claim of Peircean realism. As defended by Peirce
himself, it is inseparably connected to his fallibilism, synechism, and pragmati-
cism. But rather than argue for this claim in a theoretical manner, I want now to
illustrate it in a deceptively simple way. This illustration is, however, intended
to have the force of an argument; it is offered as a means of showing how self-
reference in the media can serve the ideal of self-criticism and, in turn, how
the ideal of such criticism is vital to the disclosure of reality. Such reference
serves a variety of other functions, ones often eclipsing and even counteracting
the critical function of reexive discourse. But it also generates the possibility,
however fragile and eeting, of self-criticism. Such criticism is, at bottom, the
result of a moral stance, that of holding oneself accountable to others, to both
what others have said and also what allegedly mute objects (Bakhtin 1981:
351) in effect say in response to our depictions of them (cf. Colapietro 2003:
1415). More or less reliable disclosures of some facet of the world depend
Distortion, fabrication, and disclosure in a self-referential culture 37

utterly upon such a stance. In the media as much as anywhere else, we discern
just this possibility, along with countless examples of effective distortions and
insulating fabrications. I now offer an example of this.

3. An example of journalistic reexivity

Thus, allow me to turn at this point to the task of reporting on a report about
the activity of reporters working in the mainstream media in the United
States (Gitlin 2005: 6). This might seem to be a case of hyperreexivity or
self-referentiality (terms I use synonymously here); moreover, hyperreexivity
appears to be characteristic of the very processes by which our relationship to
reality becomes ever more attenuated, if not entirely severed. My purpose in
doing so, however, is not to construct a hall of mirrors in which an endless
multiplication of reections frustrates the desire to locate the palpable source
of those reections. Rather my aim is to use this example of self-reference in
the media as a way of bringing into focus the irreducibly complex relationship
of reexivity, rationality, and reality, though in such a way that emphasis on
reexivity is not allowed to eclipse rationality and reality.8
So, let us attend to a report in the media on the activity of those in the
mainstream media in the U.S. More exactly, this report concerns less the activity
than the failure of reporters, less what they did than what they have failed to do.
It is a story about what until recently has been missing in the media and indeed
who have been missing (in the sense of not reporting for duty, not doing what
they have vowed to do). In a regular feature entitled Press Watch in an issue
of The Nation (July 4, 2005), Todd Gitlin constructs a compelling case for the
noteworthy failure of mainstream media simply to report a signicant fact.
The focus of his concern in the piece under consideration is, to repeat, what
the mainstream media have failed to report. The title of his article makes this
point immediately clear: MIA: News of Prison Toll. In this instance, the re-
porters are the ones missing in action (MIA). The news regarding the number
of those who have died while prisoners of the U.S. under suspicion of terrorism
is that this number has not been reported in the news. In the New York Times,
Thomas Friedman (an opinion columnist, not a news reporter) declared, the
abuse at Guantanamo and within the whole U.S. military prison system dealing
with terrorism is out of control. Tell me, how is it that over 100 detainees have
died in U.S. custody so far? Heart attacks? (May 27, 2005; quoted in Gitlin
2005: 6).
In Gitlins piece, one member of the media is chastising other members9 and
therein the media is exhibiting a seemingly irrepressible tendency to report on
38 Vincent Colapietro

itself. One of the effects of this is, however, not to sustain an ever more insular
network of self-reference (though one of the results is, arguably, to generate
an ever more intricate system of such references). At any rate, reexivity is not
operating here in the service of insularity. Rather it is pushing to render a system
vulnerable to the pressures ultimately, the reality of what this system has so
strenuously tried to exclude. In other words, reexivity here is operating in the
service of registering a reality resolutely ignored. Reexivity is fullling this
function through a process of self-questioning: a representative of the media is
questioning the media in the name of what they (the media themselves) promise
to convey. In turn, such questioning is integral to our understanding of rational-
ity. The acknowledgment of any complex reality (especially any controversial
reality) is impossible apart from responsiveness to ongoing interrogations, of-
ten ones of an increasingly reexive character. Whatever else we might mean
by rationality, responsiveness in this and indeed other respects is central to the
meaning of this word.

4. Conclusion: Vision of the cobblestones

Allow me to draw these reections to a close by recalling an arresting image

offered by a contemporary author. In the Preface to The Captive Mind, Czeslaw
Milosz confessed about himself as a younger man, Hitherto, I had had no strong
political afnities and was only too ready to shut myself off from the realities
of life. But reality would never let me remain aloof for long (1953: vi). Later
in this work he insists: The work of human thought should withstand the test
of brutal, naked reality. Probably only those things are worth while which can
preserve their validity in the eyes of a man threatened with instant death. He
valorizes this decisive perspective in an unforgettable way when he writes:
A man is lying under machine-gun re on a street in an embattled city. He
looks at the pavement and sees a very amusing sight: the cobblestones are
standing straight up like the quills of a porcupine. The bullets hitting against
their edges displace and tilt them. Such moments in the consciousness of a
man judge all poets and philosophers. Let us suppose, too, that a certain poet
was the hero of the literary cafes, and wherever he went he was regarded
with curiosity and awe. Yet his poems [or someone elses theories], recalled
in such a moment, suddenly seem diseased and highbrow. The vision of the
cobblestones is unquestionably real, and poetry based on an equally naked
experience could possibly survive triumphantly that judgment day of mans
illusions. (1953: 39)
Distortion, fabrication, and disclosure in a self-referential culture 39

At least one of the most critical tests of poetry and theories, news reports
and presidential addresses, is their ability to appear other than diseased when
exposed to such experience. Such experience might never be naked, in the sense
of unmediated. The mediated yet direct encounter with the displaced and tilted
cobblestones (or some analogue), however, secures in our experience a basis for
maintaining an irreducible (if not necessarily highly prominent) distinction
between reality and ction, disclosure and fabrication. That the ongoing medi-
ations of a self-consciously historical community of self-critical inquirers are
required to ascertain the signicance of these encounters is no argument against
a direct encounter with the actual world; for Peircean realism insists that all our
encounters with reality are direct yet mediated affairs (Smith 1992: 2025; cf.
Smith 1978: 8795; Colapietro 1995: 4244). The recognition of the ubiqui-
tous thirdness inherent in experience does not require us to ignore or deny the
salient secondness characteristic of experience (the direct confrontation with
irreducible otherness). A nuanced understanding of human experience suggests
rather the abiding need to accredit the disclosive potential of our direct encoun-
ters but also the equally persistent need to embrace the task of interminable
critique (Bernstein 1981: 116120).
Reality is, as Peirce suggests, what inquiry would disclose in the indenite
long run (see, however, Smith 1970: 104108). In the meantime, it is what the
most reliable of our investigations have intimated, however revisable these in-
timations turn out to be (Colapietro 1996: 13738). In this light, reality is at
once frustratingly elusive, brutally insistent, and inherently intelligible. Reex-
ive rationality, especially in the form of reexive interrogation (more simply,
in the form of self-questioning), is required for disclosing the contours, facets,
and dimensions of reality, insofar as this is possible. Self-referentiality in the
media can be an instance of nothing less than such rationality. But, then, it also
can be one of the principal forces by which an increasingly insular world of
autotelic systems maintains itself. The markedly aesthetic dimension of con-
temporary existence is nowhere more evident than in the degree and manner
in which the technological, communicational, educational, and other networks
assume the character of autotelic systems. Even so, the marked presence of the
aesthetic and ludic functions of increasingly reexive media does not preclude
the effective operation of a critical function, one operating often in the name of
the strenuously repressed, the systematically ignored, and the grossly distorted.
Such a function is necessary for disclosing the degree to which we humans and
our uses of signs are so often at odds with reality. In light of the distortions and
fabrications so vital to the maintenance of power, this everyday expression and
its numerous equivalents in their colloquial sense are, in turn, vital to the ever
unnished work of humane critique (Bernstein 1981: 118120). Is such critique
40 Vincent Colapietro

truly rendered impossible or unnecessary in such a radically self-referential

culture as our own?


1. This point is nely articulated by Winfried Noth in his introductory chapter to this
volume: The visual, and the audiovisual arts and media have become increasingly
self-referential, self-reexive, autotelic. Instead of representing something heard
about, seen, lived, or otherwise experienced in social life, culture, and nature, jour-
nalists, commercial artists, designers, and lm directors report increasingly what
has been seen, heard, or reported before in the media. The mediators have turned
to representing representations. Instead of narrating, they narrate how and why they
narrate, instead of lming, they lm that they lm the lming. The news are more and
more about what has been reported in the news, television shows are more and more
concerned with television shows, and even advertising is no longer about products
and services but about advertising. The messages of the media are about messages
of the media, whose origin has become increasingly difcult to trace. In literature,
ction has become metaction, novels have become metanovels, and texts are dis-
covered as being intertexts whose reference is not to life but to other texts, and in
the visual arts, art is now about art, and even architecture is about architecture. The
digitalization of pictures and lms, which has liberated the media from the bonds
of factual reference to a world which they used to depict, has contributed to the
increase of self-reference. In this essay, Noth also carefully distinguishes the forms
and means of self-reference. In addition, he helpfully stresses the point that there
are degrees of reexivity: Various degrees of self-reference can be distinguished,
from the sign that refers to nothing but itself to the sign that refers only partially to
itself and partially still to something else. Finally, he notes, self-reference concerns
different levels of the media and the message in which it occurs (emphasis added).
My own approach to this topic has been crucially informed by this paper and other
writings by Prof. Noth.
2. In one of the articles (Some Consequences of Four Incapacities) in the series in
which C. S. Peirce so brilliantly argued for a semiotic conception of cognition, con-
sciousness, and selfhood, he stresses: The real, then, is that which, sooner or later,
information [i.e., experience, V.C.] and reasoning would nally result in, and which
is therefore independent of the vagaries of me and you. Thus, the very origin of
the conception of reality shows that this conception essentially involves the notion
of a COMMUNITY, without denite limits, and capable of a denite increase of
knowledge. And so those two series of cognition the real and the unreal consist
of those which, at a time sufciently future, the community will always continue to
re-afrm; and those which, under the same conditions, will ever after be denied (CP
5.311; EP 1, 52).
Distortion, fabrication, and disclosure in a self-referential culture 41

3. An important practical question is whether such a widely felt matter can serve as
one of the bases for the recovery or reconstruction of local sites of effective
4. Rich points out: Ever since Jaws, a movie set on the July Fourth weekend, broke
box ofce records 30 summers ago, Independence Day has come to stand for terror
as much as for freedom (2005: 11).
5. Rich reports that, in Spielbergs movie, Tim Robbins (who else?) pops up to declare
that when aliens occupy a country, the occupations always fail. Even Tom Cruises
doltish teenage screen son is writing a school report on the French occupation of
Algeria. I do not have to remind Europeans, as I would have to remind Americans,
that Jacques Chirac told George Bush that Iraq will prove to be Americas Algeria,
to which Bush responded, I could not disagree with you more. Perhaps the doltish
adolescent in Spielbergs terrifying thriller is, as a result of his research, in a better
position than the adolescent President to appreciate the analogy.
6. This would be reality in its secondness, in the form of otherness, resistance, obsis-
tence, or oppugnancy (CP 1.322, 1.324, 2.79, 8.291). But, in Peirces lexicon, reality
is a subtle, nuanced term, and reality in its secondness is hardly exhaustive of its
7. The notion of trace is crucial to the argument of this paper. This implies that so too
is that of index, for the trace in the sense intended here is an instance of indexicality.
See Noth, The death of photography in self-reference (this vol., Part III).
8. One of the most crucial aspects of the relationship between rationality and reexiv-
ity is brought into immediate focus by Peirce when he suggests: rational means
essentially self-criticizing, self-controlling and self-controlled, and therefore open
to incessant question (CP 7.77).
9. The situation is even more complex than this since Gitlin is amplifying a criticism
rst voiced by Friedman in an editorial. He is also underscoring the fact that the
news regarding the number of deaths was in effect reported in an editorial. At a time
when the New York Times has made a renewed effort to draw a sharper demarcation
between news and opinion, the presumably hard side of the news has, concerning
a stark and consequential matter of fact, been missing in action (Gitlin 2005: 6). For
news regarding the number of deaths of those detained by US as terrorists individuals
in that country have had to rely on an opinion writer. Gitlins search on LexisNexis
database of television news yielded these results: No report of this number on CSP,
one brief mention on NBC, another on ABC, but nothing at all on CNN, Fox,
or MSNBC; also nothing in either Time or Newsweek. On Corporation for Public
Broadcasting (CPC), under intense re for allegedly liberal bias, Jim Lehrer in his
News Hour did give the number of such deaths as 100, counting 20 as homicides.
42 Vincent Colapietro


Bakhtin, Mikhail M.
1981 The Dialogic Imagination: Four Essays by M. M. Bakhtin. Michael
Holquist (ed.). Austin, TX: University of Texas Press.
Bartlett, Steven (ed.)
1992 Reexivity: A Source-Book in Self-Reference. Amsterdam: North-Hol-
Bernstein, Richard J.
1981 Toward a more rational community. In: Kenneth L. Ketner et al. (eds.),
Proceedings of the C. S. Peirce Bicentennial International Congress,
115120. Lubbock: Texas Tech University.
Colapietro, Vincent
1995 Immediacy, opposition, and mediation: Peirce on the irreducible as-
pects of the communicative process. In: Lenore Langsdorf and An-
drew Smith (eds.), Recovering Pragmatisms Voice, 2348. Albany,
NY: SUNY Press.
1996 The ground of semiosis:An implied theory of perspectival realism? In:
Vincent Colapietro and Thomas Olshewsky (eds.), Peirces Doctrine
of Signs: Theory, Applications, and Connections, 129140. Berlin and
New York: Mouton de Gruyter.
2000 Robust realism and real externality: The complex commitments of a
convinced pragmaticist. Semiotica 130: 301372.
2003 Signication and interpretation. Jornada CEPE [Sao Paulo] (October
30), 520.
Dunne, Michael
2001 Metapop: Self-Referentiality in Contemporary American Popular Cul-
ture. Jackson, MS: University Press of Mississippi.
Gitlin, Todd
2005 MIA: News of prison toll, The Nation (July 4), 68.
2002 Media Unlimited. New York: Metropolitan/Owl Book.
Lawson, Hilary
1985 Reexivity: The Post-Modern Predicament. London: Hutchinson.
Lenoir, Timothy
2000 All war is simulation: The military-entertainment complex. Congu-
rations 8: 289335.
Milosz, Czeslaw
1953 The Captive Mind. New York: Vintage Books.
Noth, Winfried (ed.)
1997 Semiotics of the Media. Berlin and New York: Mouton de Gruyter.
Noth, Winfried and Christina Ljungberg (eds.)
2003 The crisis of representation: Semiotic foundations and manifestations
in culture and the media. Semiotica 143(14). (Special Issue).
Distortion, fabrication, and disclosure in a self-referential culture 43

Peirce, Charles S.
19311958 Collected Papers of C. S. Peirce. Ed. Charles Hartshorne, Paul Weiss
and Arthur Burks. Cambridge, MA: Harvard University Press. Quot-
ed as CP.
19631966 1979 The Charles S. Peirce Papers, 30 reels, 3rd microlm edition.
Cambridge, MA:The Houghton Library, Harvard University, Microre-
production Service. Quoted as MS.
18671893 2003 The Essential Peirce. Ed. Nathan Houser. Bloomington, IN: In-
diana Unversity Press. Quoted as EP.
Rorty, Richard
1989 Contingency, Irony, and Solidarity. Cambridge: Cambridge University
Rich, Frank
2005 The two wars of the worlds. The New York Times (July 3), 11.
Santaella, Lucia
2003 Culturas e Artes do Pos-humano: Da Cultura das Mdias a Cibercul-
tura. Sao Paulo: Paulus.
Short, Thomas L.
1997 Hypostatic abstraction in self-consciousness. In: Jacqueline Brunning
and Paul Forster (eds.),The Rule of Reason:The Philosophy of Charles
Sanders Peirce, 289308. Toronto: University of Toronto Press.
Smith, John E.
1970 Themes in American Philosophy: Purpose, Experience, and Commu-
nity. New York: Harper & Row.
1978 Purpose and Thought: The Meaning of Pragmatism. New Haven: Yale
University Press.
1992 Americas PhilosophicalVision. Chicago: University of Chicago Press.
Part II. Self-referential print advertising
Modes of self-reference in advertising

Siegfried J. Schmidt

1. Media culture societies on their way to self-reference

In recent decades, there has been a clear tendency in the development of media
culture societies towards self-referential structures and processes. As far as I
can see, there are three different reasons for this tendency:
Studies of the origin as well as of the functioning of communities and soci-
eties have revealed the crucial role of self-reference in terms of reexivity.
Today, reexivity can be regarded as the fundamental process of generating
social structures. The best example is communication. Communication re-
lies on collective knowledge which bridges the gap between the cognitive
autonomy of individuals and their necessary social orientation. Collective
knowledge can be described in terms of reexive loops of expected expec-
tations in the domain of knowledge and imputed imputations in the domain
of motives and intentions. As part of each individuals stock of knowledge,
it is more or less shared by all members of a society and plays an essential
role in the coordination of their behavior and their communicative practice.
Self-referential reexivity has become part of everyday experience in the last
century through the development of a complex media system since media
conceived of as complex social systems primarily serve the purpose of
enabling specic self-observations of media culture societies. Furthermore,
media increasingly observe each other. Accordingly, an immensely complex
network or process system of observations has emerged, steered by the fun-
damental constructive processes of reexivity in terms of self-referentiality.
The more societies increase the degree of their observability through the
development of self-referential media systems, the more urgent becomes the
question for the functional and integrative performance of culture programs1
for actors and social subsystems alike, since reexive structures of observa-
tion inevitably lead to the experience of contingency. For this reason, soci-
eties whose models of reality and culture programs are a permanent subject
of discussion in complex media systems automatically develop media cul-
48 Siegfried J. Schmidt

tures marked by a high level of plurality and a low level of compulsion in

traditional problem solutions.
In light of these considerations, it becomes plausible that advertising in media
culture societies had to generate modes of self-reference, too. In this paper, ad-
vertising is conceived of in terms of a social system which can be characterized
as a subsystem of the economic system, which follows its own system specic
logic. The advertising system sells creativity for money. It aims at arousing atten-
tion for its media supply which is then transferred to the goods, services, persons,
and messages to which it seeks to draw attention. In order to arouse attention, ad-
vertising must permanently translate changing social events and developments
into communication contents and forms, into pictures and stories which promise
economic prot. Since advertising appears in public in the form of expensive me-
dia supply, it is necessarily related to three dimensions, viz., economy (money),
creative processes of advertisers and their target audience, and communication
processes initiated by advertising activities. These systematic interrelations need
to be taken into consideration when we talk about the media supply of adver-
tising, which is only the nal result of these interrelations and does not simply
speak for itself. Today, advertising is not only an important economic sector but
an inuential instrument of socialization since it forms an integral part of our
daily life. In 2002, for example, 2.6 billion commercials were shown on German
television; the print media are full of commercials; internet, radio, or city lights
confront us with hundreds of ads every day and everywhere. Furthermore, due
to the increasing commercialization of communication, advertising has become
an important power of motivation in the dynamics of the media system.

2. The semiotics of self-reference

Before turning to a description of self-referential modes in advertising, I would

like to clarify my concept of semiotic self-reference. I advocate a process-
oriented three-digit concept of sign which systematically combines (1) semi-
otic material, (2) the use of such material in accordance with the rules of a given
language, and (3) the result of these processes.
Semiotic material consists of structurally organized condensations of social
experiences which can trigger cognitive as well as communicative follow-up
processes. The arbitrariness of such triggering is kept in check by the fact that
language-users refer to collective knowledge about semiotic routines as an ef-
cient operative ction, that is to say collective knowledge which is regarded to
be mutually shared by all members of a society.
Modes of self-reference in advertising 49

The use of semiotic material can be modeled as a context-bound process of

referring to previous references or, in Charles S. Peirces phrase, as reference to
previous experiences or cognition of the world. In using semiotic material, we
refer to successful previous uses. In other words, we do not represent objects,
events etc. but we refer to previous descriptions (of objects and events) we
deem socially acceptable. Accordingly, these references mark a time difference
between previous and actual descriptions, they occur in social contexts, and
they necessarily rely upon collective cultural knowledge. In other words, signs
do not present stable reference-relations, but are used by communicators in real
situations to trigger follow-up operations in the cognitive or communicative
domain of communication partners which happen according to the conditions
of the respective systems.
In and through using semiotic material subject-dependent cognitive pro-
cesses are triggered. These processes result in self-referential constructions of
ordered cognitive states that rely on culturally programmed knowledge about
socially acceptable usages of the respective semiotic materials. That is to say,
expected expectations or reexive loops establish a socially successful relation
between semiotic materials on the one hand and subjective cognitive operations
on the other. Again the point bears repeating I do not model semiotic pro-
cesses in terms of representations of items beyond language but in terms of
references to previous references to cognitive processes and imputed collective
These considerations imply that self-referential operations presuppose sec-
ond-order observation, the observation of ones own or other peoples activities
performed on a rst-order level. This particular mode of observation is possible
because of the time-difference between rst-order and second-order operations,
since the more or less automatic rst-order processes have to be interrupted in
order to impose on them various kinds of structures and postprocesses. In my
view, this is the reason why semiotic self-reference and cognitive reexivity
both fall into the category of reexive looping.

3. Modes of reference in advertising

In accordance with its general aim to produce maximum attention for the
goods, services, messages and persons it advertises the advertising system is
distinguished by specic communication-processes or discourses; that is, by a
specic macro-form of communication, which competes with other macro-forms
of communication such as journalism, literature/art or public relations, each of
which possesses a different mode of reference. Journalism, e.g., claims to re-
50 Siegfried J. Schmidt

fer to the reality in an objective, reliable and authentic way; literature claims
to refer to a ctitious world according to aesthetic practices and expectations;
and public relations claim to refer to wishful images of persons, institutions or
rms which have been established by previous communications. Advertising,
by contrast, refers neither to truth nor to objectivity but to brand values on the
side of the consumers, hence to experiences, expectations, emotions, desires,
needs etc. and everybody in media culture societies can know that. There-
fore, everybody is assumed to know that advertising does not and cannot lie,
because everybody knows that ads are biased, one-sided and prejudiced in favor
of the items advertised for. Ads have to touch people emotionally; they have to
entertain. They tell what people want to believe and should believe in order to
become happy. Fascination comes before semantic information. Accordingly,
what matters in advertising communication is the meaning and importance of
the promises ads make, not the truth of their messages. Social actors in the
advertising system know that ads promise to solve insoluble problems here and
now you only have to believe their message and buy the product.
In addition, advertising refers to collective knowledge in those domains
which can be instrumentalized in ads: famous landscapes or buildings (the Riv-
iera or the Eiffel Tower), stereotypes of professions or national characteristics
(the doctor or the Frenchman) or well-known works of art (Michelangelos David
or Leonardos Gioconda).
Finally, advertising refers to collective knowledge about social practices re-
garding, for example, the interaction of genders and generations, the relevance
of religious or political activities and the evaluation of the female and male bod-
ies, of clothing, eating and drinking, of sports or trendy leisure-time activities.
In sum, advertising selectively refers to collective mentalities, needs and desires
of specic target groups.

4. Modes of self-reference in advertising

Generally, it can be said that only owing to self-referential maneuvers of or in

the advertising system we as the addressees of media supply are able to observe
the advertising system at all. This general hypothesis will be explained in the

(a) In his paper Self-reference in the media: The semiotic framework, Win-
fried Noth (this vol., I, 5.5) mentions self-referential practices in advertising
which may be subsumed under the heading of intertextuality, such as citations,
repetitions, recursions, or other kinds of reference to signs, texts, or other media.
Modes of self-reference in advertising 51

Such practices make use of what since the 1990s has been called recycling or
sampling (Figure 1a/1b).

Figure 1a. Stiebel (1995)2

Figure 1b. Nirvana (1991) 3

In other words, media supply refers to media supply, and the efciency of
this strategy depends upon whether or not the addressees recognize the recycled
components (Figure 2).
According to Noth (this vol., I, 5.1), another type of self-reference in adver-
tising is tautology or quasi-tautology. A good example is the famous German
slogan for Persil washing powder, Persil bleibt Persil (Persil remains Persil;
Figure 3).
52 Siegfried J. Schmidt

Figure 2. Die Marke [the brand]

This slogan claims that Persil will remain the same regardless of what will
happen in the future. This promise tames the paradox that on the one hand,
Henkel, the producer of the washing powder, will and has to develop the new
Persil (das neue Persil) as soon as possible, which of course has to be so much
better than the present one but which, on the other hand, will still remain Persil
and nothing else. In other words, as long as Persil is available, it will be of
outstanding quality a quality nevertheless to be always improved. In addition
to this tautology, another kind of self-reference can be observed in this example,
viz., reexive stabilization of ad communication. Guido Zurstiege (2003) has
Modes of self-reference in advertising 53

Figure 3. Persil remains Persil4

pointed out that the highly emotional relation between brands and consumers
is based on trust. Persil campaigns are famous for their repetition of the same
advertising pattern: my grandmother and mother, says the pretty young lady
in the TV commercial, have successfully used Persil, so I will continue this
successful tradition. The message is clear: Persils customer-brand relation is
dened by continuity in trust on both sides this is the aspect of redundancy.
At the same time, this trust is a challenge to the producer to improve the quality
of the product in order to ensure this condence by a continuity of successful
applications of this washing powder this is the variety aspect. Continuity and
variety are reconciled by the self-referentiality in Persils advertising throughout
the decades.
Another good example is the exact repetition of a TV spot presenting an Audi
Quattro on a ski-jumping platform on occasion of the 25th birthday of this car
(Figure 4).
54 Siegfried J. Schmidt

Figure 4. Audi Quattro

Modes of self-reference in advertising 55

A genuine example of self-reference is nally exposed by the following

Coca-Cola ad (Figure 5).

Figure 5. A Coca- Cola neon sign

(b) In another perspective of self-reference, advertising refers to the advertising

system. This strategy is realized in the following ways:
Advertising advertises for advertising. Every advertising media supply ad-
vertises for the advertising system as such and as a whole and in so doing
inevitably generates a public discourse about the pros and cons of adver-
tising: it generates the topic advertising, which then allows for or provokes
either approval or rejection (Figure 6).
Advertising has become its own best client. Today, the advertising system
invests more money in advertising than the automobile industry does.
Advertising becomes a topic in advertising media supply: the advertising
system demands attention for itself, in most cases in an ironical or humorous
way, playing with the collective knowledge of the public (Figure 7).
56 Siegfried J. Schmidt

Figure 6. Hingerichtet [executed]

Figure 7. Lucky Strike (left): Advertising gives a completely wrong picture, doesnt
it? Lucky Strike. Nothing else.

(c) Advertising observes its own practices by means of advertising research

carried out by the advertising agencies themselves; for example, through target-
group research and research into the efciency of campaigns or social trends.
For example, the advertising agency Grey has established a large archive in
which journals and ads of the last thirty years have been and are being collected.
It has also founded the so-called Grey Academy, a department for strategic
planning, and the Magic Lab, an institute for research in communication
Modes of self-reference in advertising 57

and its development. This kind of self-reference qua self-observation has to be

distinguished from the external observation of the advertising system by other
social systems including sociological research and communication studies.

(d) Another type of self-reference occurs in self-descriptive statements by ad-

vertising agencies concerning their own visions, missions, and philosophies. As
the following examples will show, such philosophies tend to be extremely poor.
McCann-Erickson believes in truth well told and in total communication.
Scholz & Friends answer the question how people can be fascinated by ideas.
But then, what is an idea? An idea is new. The DDB-Group promises: We be-
lieve that brilliant ideas can achieve extraordinary results. Springer & Jacoby
follow the strategy: simple, imaginative, exact, and so on.

(e) Finally, in recent years more and more stars of the advertising system have
published books with their ideas about best practices in advertising. The major-
ity of these books can be characterized as how-to-do-it literature but some of
the advertising stars have developed a specic kind of self-reexion one that
looks at the advertising system in terms of its present internal and external prob-
lems, the relation to scientic disciplines, efcient problem solving strategies
etc. Relevant examples include Jean Etienne Aebis, Einfall oder Abfall. Was
Werber warum erfolgreich macht (2003) [Idea or trash. What makes advertisers
successful and why], Wulf-Peter Kempers Brandholdervalue. Was Auftraggeber
zu mehr Werbe- und Markenerfolg beitragen konnen (2003) [Brand holder value.
What customers can contribute to more success of advertising and brands]; and
Holger Jung / Remy von Matts Momentum die Kraft, die Werbung heute
braucht (2002) [Momentum The power advertising needs today].

5. Advertising as a mode of self-reference

Until now the focus was on modes of self-reference in the advertising system.
The nal considerations will deal with the fact that the advertising system itself
is the result of self-referential relations.
The advertising system can be characterized as an instrument of observa-
tion and description; it is an instrument societies use for the purpose of self-
observation and self-description. Modern media culture societies can neither be
observed nor described as a whole and from the outside, for every discourse is
of necessity situated within this society; there is no social system which is not
part of it. Self-reference therefore takes the form of partial observations and
descriptions, which can be found in, among other things, academic disciplines,
58 Siegfried J. Schmidt

literature, and the advertising system. Advertising observes social phenomena

according to the logic of the advertising system. That is to say, it observes both
very precisely and very selectively. Observations have to be exact because only
collective knowledge, mentalities of target groups, their needs, aims, and dreams
can be detected and incorporated in ads, whose aim, after all, is to establish an
emotionally persuasive connection between these components and the promoted
item. Only in this way can advertising arouse attention for its message in the
highly competitive media system. The selectivity of such observations has a
double reason. On the one hand, only domains such as consumption, services,
lifestyle, and taste are relevant to the advertising system in the economic respect
which dominates its activities. On the other hand, as mentioned above, adver-
tising exclusively tells positive stories about its promoted goods and services.
Even Benettons advertising campaigns which focused on war, Aids, ecological
pollution, and erotic relations between Catholic priests and nuns, turned out to
be advertising for advertising. By breaking the iron rule of the advertising sys-
tem to exclude all negative components from the ads and banning its products
from these campaigns, it advertised for attention for advertising (Figure 8).

Figure 8. Benetton5
Modes of self-reference in advertising 59

In light of these considerations, it becomes clear that advertising cannot be a

mirror of society, as has been assumed by many authors. Instead, I think it can be
characterized as an indicator of those social phenomena which can be related to
goods, services, messages, and persons for which advertising makes publicity.
The advertising system may thus be regarded as an interface between the
capitalist economy and the media or, as Michael Schudson (1984: 232) once
said: Advertising is capitalisms way of saying I love you to itself. It provides
the various groups of society with a self-observation of those needs which are
deemed relevant by members of these target groups, and it publicly draws atten-
tion to the mechanisms of satisfying them. Advertising realizes a short circuit of
needs and satisfactions. It promises to solve all problems by the mere consump-
tion of goods and services. Thus, as Guido Zurstiege (2003: 77) has pointed out,
the advertising of a society is at the same time advertising for a specic society.

6. Conclusion

In this paper, I tried to describe different reexive loops in advertising and in the
relations between advertising and other social systems. It has become clear that
advertising works primarily as an instrument of observation. The video artist
Maciej Toprowicz has created a video in which he generates another interesting
reexive loop: he observes as an artist how advertising observes society, and
confronts this type of afrmative observation with irritating similarities or con-
traries, irritatingly similar and dissimilar ways of acting and observing during
the Nazi regime. Toprowicz deliberately counteracts the principles of the adver-
tising system. In this way, he confronts us with what we see when we observe
how advertising observes society.


1. By culture I understand the basic program for solving the essential problems of a
2. Figures 1a, 2, 4, 5, 6, and 7 are from the authors les of print, TV, and outdoor
3. Nirvana album cover Nevermind (1991).
4. See Wolfgang Feiter. 1997. 90 Jahre Persil. Die Geschichte einer Marke, 2nd ed.
Dusseldorf: Henkel, p. 49.
5. See Lorella Pagnucco Salvemini. 2002. Toscani. Die Werbekampagnen fur Benetton
1984 2000. Munchen: Knesebeck, p. 104.
60 Siegfried J. Schmidt


Schudson, Michael
1984 Advertising, the Uneasy Persuasion: Its Dubious Impact on American
Society. New York: Basic Books.
Zurstiege, Guido
2003 Zwischen Kritik und Faszination: was wir beobachten wenn wir die
Werbung beobachten, wie sie die Werbung beobachtet. Cologne: Ha-
Metapictures and self-referential pictures

Winfried Noth

1. Pictures, signs, absent, and present objects

Pictures are signs which represent the visual or visually imaginable world. Since
the cave paintings of the Stone Age, pictures have represented by similarity.
A drawing of an elephant, a painting of a landscape, and a photograph of
Sir Winston Churchill are pictorial signs which evince similarities to the ob-
jects they represent. Pictures which are similar to the object they represent are
iconic signs. Language, by contrast, consists of mostly symbolic signs since
words usually evince no similarity to their objects but must be associated with
what they refer to by learning, habit, and convention (cf. Santaella and Noth
Despite this essential difference between words and pictures, there are also
fundamental similarities in their semiotic potential. Both words and pictures
are signs which can evoke the image of, or refer to, absent objects. In his list
of design features of language, Charles Hockett (1977) introduces the term
displacement to describe the semiotic potential of language to refer to objects
remote in time and space (cf. Noth 1990: 236). Derrida (1972: 9) discusses this
feature as an aspect of the sign in general when he states: A sign represents
what is present in its absence. However, the generalization that all signs refer
to something absent cannot be sustained since signs can also refer to something
that is present. Mirror images and shadows, in spite of Ecos arguments to the
contrary (Eco 1986), are indexical signs, images which indicate the presence of
their object, and, as we will see below, self-referential signs are also signs which
refer to their objects in their presence.
Despite the potential of pictures to indicate objects in their absence or in
their presence, it would be wrong to say that reference to, or representation of,
objects is necessarily their primary function. For example, prehistoric rock art
most probably served magical and ritual purposes and not representational ones
(Anati 1994), and paintings of the more recent history of art are not art because of
what they show (referential function) but for how they show (aesthetic function).
Nevertheless, from the perspective of present day media culture, displacement
62 Winfried Noth

is certainly one of the most important features of pictures since it is evidently

the prerequisite of global communication.
Besides displacement, prevarication, the potential to deceive, is another
semiotic feature which pictures have in common with language. Whereas dis-
placement makes the representation of objects remote in time and place possible,
prevarication serves the opposite purpose; a message which lies represents an
object the temporal, local, or qualitative characteristics of which are not those
which the sign indicates. Evidently, the feature of prevarication does not only
enable language users and picture makers to lie but also to create illusions and
The present paper is about another way in which a picture may be nonrepre-
sentational; it deals with metapictures and self-referential pictures. Metapictures
are pictures about pictures (Mitchell 1994: 3582; Alessandria 1996). Instead
of referring to the world of nonpictorial objects, they refer to other pictures.
Self-referential pictures refer to themselves, that is, they have their object of
reference inside and not outside their own pictorial frame. Self-reference is the
opposite of alloreference or simply reference, in the sense which Derrida (1972:
9) has in mind when he writes about the absence of the object and how signs
are traditionally dened as referring to, or standing for, something else, which
is not the sign.

2. Pictures and metapictures

Self-referential pictures are often metapictures, that is, pictures about pictures,
but not all metapictures are self-referential. The two categories overlap, but
the difference between them has often been ignored, for example by Mitchell
(1994: 35) who denes metapictures as pictures about pictures that is,
pictures that refer to themselves, pictures that are used to show what a picture
is. Let us try to distinguish between metapictures and self-referential pictures
in analogy with the distinctions which linguistic terminology has established
between metalanguage and self-referential language.
The term metapicture is coined after the term metalanguage, which means
language about language. Terms such as vowel, consonant, word, sentence, con-
jugation, or declension are metalinguistic words, verbal signs which refer to
nothing but language. Metalanguage is the opposite of object language. Words
that belong to object language are words that have their referents in the nonver-
bal world, for example, duck, love, or freedom. In analogy, the term metapic-
ture should designate a picture about a picture or a picture of a picture. The
term object picture may be useful to designate ordinary pictures which are not
Metapictures and self-referential pictures 63

metapictures, such as the picture of a ying duck or a running cheetah. Instead

of object picture we may simply use the term picture whenever there it is no risk
of confusion. Examples of metapictures are the following:
(1) a picture of a room with a framed painting hanging on the wall
(2) a picture quoting a famous painting in a new style, as in Duchamps
transformed drawing of Leonardos Mona Lisa with the disrespectful leg-
end L.H.O.O.Q.
(3) a picture of a painter (not in a self-portrait) painting the portrait of a lady
(4) a picture of a photographer (not in a self-portrait) taking photos
(5) an ambiguous picture, such as the Necker cube
(6) a picture of a room with a teichoscopic mirror, showing a scene from the
otherwise invisible rear of this same room as in Brassas Au Bal-Musette
of 1932 (cf. Machado 1994: 87)
The criteria according to which these pictures are metapictures are the follow-
ing: pictures (1) to (3) are pictures depicting another picture; picture (2) refers
both to what and how the other picture represents; (3) is a pictures about painting
pictures; (4) is not a picture of a picture but a picture about taking a picture;
the difference between the metapictures of the kind (1) or (2) and (3) or (4)
can be described with Mitchell (1994: 37) as the difference between showing
showing and showing the shower; (5) is the case of pictures with two read-
ings in conict (rabbit or duck?; old woman or elegant lady? etc.; cf. Mitchell
1994: 4757). Such pictures do not depict other pictures; they include a sec-
ond picture as an alternative reading and thus create a visual dialogue between
the two readings; Example (6) requires two additional comments. (a) Teicho-
scopic mirror is probably a new word coinage; teichoscopia is the ancient dra-
matic device of extending the limitations of the dramatic scenery visible to
the audience by introducing a messenger who narrates what he, from his own
privileged perspective (for example while standing on a wall), is discerning
in the distance concealed to the audience. Analogously, a teichoscopic mir-
ror is a mirror which extends the pictorial scenario by adding a view from
an otherwise inaccessible perspective. (b) A mirror image is not usually con-
sidered to be a picture but it is not a real-world scene either; instead, it is
the reection of such a scene, and as such it is an indexical sign of such a
eld of vision which is not its own. Hence, a teichoscopic mirror is certainly
a sign, and the picture of such a mirror (indeed of any mirror) is certainly a
metasign. Let the depicted teichoscopic image thus be included in the category
of metapictures even though, as mentioned above, a mirror image is not really
a picture.
64 Winfried Noth

Most metasigns in the context of pictures are verbal messages, but these
are not the topic of this paper (but see Mitchell 1994 on talking metapictures;
Santaella and Noth 1999; Noth 2003b). Typical examples of how words function
as signs about pictures are the title of a picture, the painters signature, the caption
of a press photo, or the body copy of an advertisement.

3. Self-referential metapictures

Let us now introduce the term self-referential metapicture for those metapictures
which refer to themselves in a narrower sense. Such pictures are self-referential
because they are representations representing their own representation, that is,
they depict a picture of what they depict, how they depict, or under which
circumstances they came to depict. Usually, a self-referential depiction is only
a partial representation of the picture it represents. Examples are:
(7) a picture mise en abyme, i. e., one which represents a scene which contains
a picture of this scene (cf. Owens 1978; Conant 2005)
(8) a picture of a painter quoting an earlier work by himself
(9) a picture of a teichoscopic mirror reecting its own mirror image from
another mirror
(10) a picture of a photographer taking his own picture in front of a mirror
(11) a picture showing a lady looking into a mirror, which reects her own
mirror image
(12) a drawing of a hand (Escher) or a man (Steinberg) that draws itself resp.
In contrast to the pictures (1) to (6), pictures (7) to (12) do not depict other
pictures, but themselves; (7) contains itself in a smaller picture; (8) depicts itself
in a picture by its own painter; (9) is a self-reected mirror image; (10) is a self-
referential showing of the shower of his own picture in contrast to (4) which
was the showing of the shower of a different picture; (11) is self-referential
under the condition that the portrait of the person appears twice in a similar
view (and not, e. g., once in a frontal and once in a dorsal perspective; (12) is
self-referential in a metaleptic way (see below): the picture includes its own
representation which seems to be the product of its representation.
The preliminary denitions of metapictures and self-referential metapictures
may prot from the comparison with the distinctions between metalanguage,
object language, and self-referential metalanguage drawn in linguistics. For ex-
ample, of the metalinguistic terms vowel, syllable, word, sentence or text, only
the term word is a self-referential metalinguistic term, since only word is a word
Metapictures and self-referential pictures 65

itself, whereas syllable is not one syllable but two, and the mere word text is not
yet a text. At the syntactic level, an example of self-referential metalanguage in
written English is: This senntence contains a mistake. It is not only an example
of language about language (and hence metalanguage), but in addition, there
is self-reference because the referent of the word senntence, namely a spelling
mistake, is located within the very sentence that asserts the mistake it contains.
By contrast, the sentence The preceding sentence contained a mistake is met-
alinguistic but not self-referential since it refers to a sentence other than itself.
Metapictures thus give evidence of a semiotic feature which pictures have in
common with language, the feature of reexiveness, as Hockett (1977) calls it,
that is, the potential of language to create its own metalanguage. However, there
is an important difference between metapictures and metalanguage. While lan-
guage has a particular class of signs which serve exclusively as verbal metasigns,
pictures, with one exception, have no sign repertoire of pictorial metasigns. The
exception is the picture frame, a topic to which we will turn below.

4. Iconic and indexical self-reference

The vocabulary of metalanguage consists of symbolic signs; words such as

word, sentence, or preposition are signs which designate by convention.
Metapictures, by contrast, are essentially iconic; a picture can only depict another
picture if it is similar to it. Self-reference, in language as well as in pictures,
is iconic and indexical. The very term self-reference suggests both kinds of
sign. While reference is essentially indexical, involving a mode of pointing
(referring) from the sign to its object, self-reference, with its loop that suggests
a return from the sign back to itself, implies iconicity; the sign that reappears
in its own object is evidently an icon of itself. Despite this duality of all self-
referential signs, some are more iconic and others are more indexical. Let us
examine more in detail the role of icons and indices rst in self-referential
language and then in pictures. The following words are iconically self-referential
for different reasons:
(13) cock-a-doodle-doo
(14) quick
(15) English
(16) longest
(17) black
(18) bold
66 Winfried Noth

All of these words evince some similarity to what they refer. Cock-a-doodle-
doo is a sound symbolic word; its acoustic form is similar to the acoustic event it
refers to. (14) to (18) are words that evince the quality they refer to, either in its
phonetic or in its written form. The word quick is itself quick in its pronunciation,
consisting of a single syllable with a short vowel. The word English is self-
referential in a metalinguistic way, both in pronunciation and in spelling; it is
not only a word of the English language that refers to itself (English is evidently
an English word), but it also sounds English and is English in its spelling. The
superlative form longest is relationally (that is, diagrammatically) iconic. In
English, the word is not really very long (in contrast to its Portuguese translation
longssimo, a word with four syllables), but the superlative longest is longer than
the base form long. The written word black is self-referential in any type that is
not colored, and the word bold is self-referential whenever it is written in bold

Examples of indexically self-referential expressions are:

(19) I, me, we
(20) here, in this city
(21) now, today, this month
(22) I promise, I accept, I bet, I ask you to
(23) I am coming; we are arriving
(19), (20), and (21) belong to the so-called autodeictic expressions (cf. Har-
weg 1990). Such expressions refer to the circumstances of a speech act. The
rst person pronouns I, me, we are indices referring to the speaker, expressions
such as here, in this city refer to the place, and now, today, this month to the
time of the utterance at the moment of its utterance. Other modes of indexical
self-reference occur in illocutionary speech acts such as I promise. The very
utterance of a promise constitutes the obligation to which the promise refers.
Finally, the I who speaks in the continuous form, as in I am coming, does not
only refer to something that is being done but also to the simultaneity of the
discourse about what is being done. The mother who says I am coming is not
only coming, she also says (self-referentially) that she is coming. (Does the
son who says I am listening speak self-referentially? Can he listen while he is
Metapictures and self-referential pictures 67

5. Iconically and indexically self-referential object pictures

The criteria for distinguishing between iconic and indexical self-reference in

language can serve to elucidate the nature of self-reference in pictures, in par-
ticular those which are not metapictures. Consider these examples of iconically
self-referential object pictures:
(24) a triangle in a geometry book
(25) a monochrome painting in red
(26) the picture of a symmetrical form, e.g., a buttery
(27) a photograph of a photograph
A triangle in a geometry book does not only illustrate and hence represent
the features of this geometrical gure; it is itself a triangle. For the same reason,
a monochrome painting in red is self-referential; it represents a quality which it
evinces itself, namely the quality of redness. Iconic self-reference is a charac-
teristic of abstract art; deprived of the function of referring to something else,
abstract paintings show only their own chromatic and geometrical qualities (cf.
Noth 2003a, 2004, this vol., Part III). All symmetrical forms are self-referential;
the symmetrically reected form is an icon of the reecting form, which repeats
itself. A photograph of a photograph is self-referentially iconic to a lesser degree;
the representing medium is the same as the represented medium.
There is indexical self-reference in a picture whenever it indicates the cir-
cumstances under which it was produced (where, when, how). In one respect,
pictorial self-reference is weaker than self-reference in spoken language. The
I, the now, and the here of a pictorial enunciation are in the past and not in
the immediacy of the present, as in spoken language. However, this is not a
categorical difference between verbal and visual communication, since pictures
share their remoteness from their origo, as Karl Buhler (1934) called it, with
written texts. Analogies between pictorial and verbal indices of self-reference
must therefore be drawn from written language, for instance, documents which
indicate their authenticity and are authentic, the handwriting which identies
a writer and in a way constitutes partly the writers identity, narratives which
indicate their narrator who would not exist without this narrative, street signs
which indicate and are at the same time the name of a street. Let us consider
along these lines the following examples of self-reference in pictures (which are
not metapictures):
(28) a self-portrait
(29) van Goghs self-portrait
(30) an original Rubens
68 Winfried Noth

(31) a Rococo scene painted by Watteau

(32) my passport photo
Any self-portrait is self-referential in the same way as any product is an
index of its producer (28). A self-portrait shows its shower without being a
metapicture about making pictures. In addition, van Goghs self-portrait (29)
also shows its showing in a certain way; the portrait contains unmistakable
indices of van Goghs way of painting; the picture indicates the painter and
his style. The van Gogh self-portrait does not only depict van Gogh it is also
a van Gogh (painting). An original style is always indexically self-referential;
it identies its own authenticity; the Rubens painting communicates I am a
Rubens (30). In Watteaus Rococo scene (31) there is a further index of self-
reference. The Rococo painting depicting a scene with elegant ladies dressed in
the fashion of the time of its being painted is not only a picture which depicts
a Rococo scene; it is at the same time a Rococo picture. A passport photo (32)
is in its origin simply an indexical picture like all photos are. However, when I
use it to identify myself I use it in a self-referential way. I may say: this picture
is a document of the light rays that once emanated from my body; hence it is an
extension, maybe even part of, my body. In a self-referential loop, the picture
which is in this sense part of my body refers back to the body from where it
emanated. Notice that from the opposite perspective, the one of the border
ofcial, my passport photo, may even constitute my identity: the picture is the
legal document, and I should better look like my picture in the passport if I do
not want to get into trouble. The border ofcial who rst looks at the photo and
then at my face wants to check whether I, the person he looks at, am like (i.e.,
the icon of) the photo and whether I can prove the correspondence of my face
with my photo. From this perspective, the passport photo is not self-referential
at all. It is like an original (or even a piece of reality) of which I should be the
picture and true copy.
An indexical feature of pictorial self-reference which will be discussed in
the next section is the frame. Is it a metasign or a self-referential metasign? The
answer depends on whether the frame is considered as something that is part
of the picture or whether it is considered as something external to the picture.
In the rst case, the frame is a self-referential metasign, in the second case it
is a metasign, but not a self-referential one. The heading of the next section
presupposes the rst of the two ways of seeing a frame.
Metapictures and self-referential pictures 69

6. Frames: Self-referential pictorial metasigns

Every picture contains a self-referential pictorial metasign in the form of its

frame, which is an indexical sign conveying a metamessage such as: I am a
picture (and not, for example, a landscape seen through a window). There is
an analogy between this metasign at the root of pictorial representation and a
metasign that can be found at the root of verbal communication. Verbal mes-
sages, too, convey a fundamental message identifying the situation as a com-
municative one. Austin (1962) called the speech act in which such signs are
produced the locutionary act. Its message is more or less, I am speaking. Pri-
eto (1966), with reference to signs in general, called it the semic act. Roughly,
the semic act identies a message as a message and states I am a message.
There are two fundamental ways in which a picture can be said to be framed.
The rst is related to the space a picture takes up in a visual eld. Let us call it the
spatial frame. Its message is basically: The space I delimitate is a picture. The
second is a frame in a metaphorical sense. It refers to the circumstances which
make pictorial information a message emitted by an addresser and addressed to
an addressee. In analogy to the terminology introduced by Christian Metz (1991)
in the context of lm studies, this kind of frame will be called enunciative frame.
The frame which delimits the picture in a spatial sense is either a material or an
immaterial frame. The material frame does not only delimitate, it also occupies
a space itself, situated between the picture and its visual environment. It may be
a wooden or metal frame enclosing its margins or a pane of glass covering its
surface. In printed pictures, it is the white or colored margin between the picture
and the edge of the page or a white margin that separates one picture from the
next on the same page. The material frame marks the boundaries between the
picture and the visual eld in which it is inserted.
In one way, the material picture frame is part of the picture since it marks the
picture as an object of a certain value that deserves to be preserved, exhibited,
sold, or possessed. In another way, the material frame is not part of the picture.
For example, it can be exchanged for a new frame, and the picture will remain
essentially the same. This ambiguity of the material frame, which seems to be
part of the picture and not to be part of it at the same time, ceases to be one when
a framed picture is depicted and thus transformed into a metapicture. As part
of the metapicture, the painted frame is clearly an object picture of a material
The immaterial picture frame is the frame which every picture has, whether
it is materially framed or not. It determines the limits of the space available for
pictorial representation. Every picture has an immaterial frame even though this
frame may be vague, blurred, or otherwise difcult to determine. Sometimes,
70 Winfried Noth

this frame is coextensive with the contours of a gure which emerges from a
ground, for example, the gure of a political hero painted on a white wall. In
this case, the contours of the gure constitute the immaterial picture frame; the
ground of this picture is simply a painted wall and not a picture. Without an
immaterial frame in the sense of a spatial delimitation of the picture, the picture
would be coextensive with the visual universe and lose its character of being a
picture. Even a picture cut out of a larger picture is immaterially framed.
The enunciative picture frame identies the addresser as a painter, a pho-
tographer, an advertising agency, an editor, or a publisher, and the viewer as an
art connoisseur, a magazine reader, visitor of a museum, or as a consumer ex-
posed to publicity. Perspective is a signicant element of the enunciative picture
frame. It shows whether the addresser took or painted the picture from below
or from above, and it has to do with the addressees point of view in face of
the picture. The theories of verbal and pictorial enunciation are closely related
to the theory of narrativity with its search for narrative voices and their real or
implied readers.
Apart from its spatial frame, a pictorial message does not evince any other
specic segmental sign repertoire to identify self-referential or metapicto-
rial features of the picture. Unlike language, which has a rich vocabulary of
metalinguistic terms and of deictic words to identify the addresser and the ad-
dressee in time and space, pictures are without a repertoire of metasigns, and
indicators of self- or metapictorial reference must be inferred by the viewer from
indirect evidence. Such inferences may be ambiguous or misleading, and the
play with such ambiguities has been much used to create paradoxical pictures,
for example, pictures depicting windows that look like painted on the wall or
the opposite, pictures depicting pictures that look like windows with the land-
scape seen through them. We will return to such ambiguities between pictures
and metapictures or between pictures and self-referential pictures below, after a
brief exemplication of the analytical categories discussed so far.

7. Three cheetahs or one?

Let us examine more in detail the features of metapictures and self-referential

pictures in the pictorial message of the upper half of an advertisement for the
Brazilian internet provider terra (Figure 1). While the lower part of this adver-
tisement contains a verbal message with information about the services of this
company, the upper half shows the picture of a cheetah. The only connection
between both is the quality of rapidity which is ascribed explicitly to the inter-
net services offered by the provider and implicitly or by presupposition to the
Metapictures and self-referential pictures 71

Figure 1. A metapicture about the picture of a cheetah1

cheetah that is shown running at a high speed. The discrepancy between the
semantic elds of animal nature and high technology leads the reader to the
conclusion that the cheetah must be interpreted as a visual metaphor of rapidity.
No other connection between the technology provider and the worlds fastest
land animal seems plausible. As a visual metaphor, the picture of the cheetah
72 Winfried Noth

is a metapicture, a picture that serves to illustrate something else, namely the

idea of rapidity. In this respect, the cheetah picture is an alloreferential and not
a self-referential metapicture.
The pictorial message of this advertisement shows a stack of three overlap-
ping pictures, each of them representing a cheetah. The picture on top is an
incomplete instantaneous photo of a cheetah, taken in a shot whose focus cut
off the cheetahs head and tail on its right and left margins. Below, the other
two pictures supplement the missing parts with their overlap, in a way that the
missing head and tail of the rst photo are complemented to result in a complete
picture of the cheetah. The second picture in the stack is a charcoal drawing
of a cheetah of the same size of which we only see the tail while its trunk is
apparently hidden below the rst photo which covers it partially. The picture
below the rst two pictures is an oil painting of a cheetah with the painters
signature at the bottom right; it serves to supplement the head missing in the
instantaneous photograph, while its rest is covered by the two other overlapping
The visual message of this advertisement can be read in two ways. It can
either be viewed as a picture of three pictures, or it can be interpreted as one
picture made up of three partial representations of a cheetah. Read as a picture of
three pictures, the message is interpreted as a metapicture in which each single
picture contains a message concerning the two other pictures. For example:
(1) the three cheetah pictures complement each other from the left to the right;
(2) the rst cheetah picture is on top, the second between, and the third is below
two other pictures; (3) the third protrudes to the right, the second to the left, etc.
Read as one (patchwork) picture of a cheetah, the pictorial message is viewed
as a self-referential metapicture in two major respects. First, it is a picture about
taking pictures. It shows what happens when the photographer fails in selecting
the proper focus; it shows some of the specic differences between the three
pictorial genres of color photography, black-and-white charcoal drawing, and
oil painting; it shows three kinds of framing, framing by a white margin, by an
immaterial frame, and by a wooden frame; it creates the mental image of three
complete pictures of the same size, each of which are iconic signs one of the
other, etc.
Secondly, the photo on the top of the stack is self-referential in quite a
different way. It is an instant photo, which means, a rapidly produced one. Fur-
thermore, it was apparently taken too rapidly so that the result was unfortunately
an incomplete picture. The rapidity of picture taking inscribed in the photograph
with these faults is an image of the rapidity of the cheetah in fast motion. The
picture that depicts rapidity in a metaphorical way evinces itself the feature of
rapidity, and in this sense it is a self-referential picture.
Metapictures and self-referential pictures 73

Of the three pictures, the instant photo and the oil painting are clearly marked
by a material frame which is a white margin in the case of the photo and a wooden
frame in the case of the painting. The charcoal drawing, by contrast, has only
its immaterial frame, which is the edge of the white paper. In addition, the
rst and the second frames on the stack are marked by a shadow indicating the
direction of their illumination, which comes from the top left. The enunciative
frames indicate the rst pictorial addresser as a photographer, the second as a
draftsperson and the third as a painter. Considered as one picture, the addresser
is the designer of the advertisement who integrated the three messages into one;
it is the advertising agency addressing among others those by whom they were
contracted; it is the internet provider addressing its potential clients, etc.

8. Illusionist and deceptive enunciative frames

Let us now return to the enunciative picture frames and examine some exam-
ples that may better exemplify their relevance than the previous example in
which its description lead to a somewhat obvious result. The relevance of these
frames is most apparent when there is some kind of ambiguity with respect to
the enunciative situation of the picture, an uncertainty as to its addresser, its ad-
dressees, or even its being a picture at all. Four examples will serve to illustrate
such ambiguities, (1) photorealist paintings, (2) pictorial fakes, (3) Parrhasioss
super-illusionist curtain, and (4) Magrittes anti-illusionist pipe.

Photorealism. Can a picture also be a total self-representation, in which the

representing picture is coextensive with the represented one? It seems as if it
is impossible for a picture to be self-referential in its totality. Such a picture
would be self-referential of itself, and how can we distinguish between the
represented and the representing picture if both are coextensive? Nevertheless,
approximations to the borderline case of total pictorial self-representation seem
to exist. Consider a photo-realistic painting. Such a painting creates the illusion
of being a photograph, but it pretends at the same time to be nothing but a
painting. Being a complete picture of a picture, the photorealist painting is a
metapicture as any copy is, but it is also self-referential. It is a picture which
refers to two kinds of pictures at the same time which do not really exist as two
separate objects. Many photorealist paintings are not even copies of existing
photos, and even if they are, their reference to the photo of which it is a copy
is quite marginal. Since the reference to the photo is only a mental one, the
depicting and the depicted picture can be said to be co-extensive.
74 Winfried Noth

Fakes. Copies and fakes are metapictures when it is known they are not the
original. Both are pictures representing other pictures as faithfully as possible,
but as long as they do not conceal what they are, their metapictorial designation
as copy and fake emphasizes the difference between the representing and the
represented picture. A copy so to speak conveys the message I am not the
original. It thus refers to and at the same time emphasizes its difference from
the original. With this distancing, the copy declares itself to be an alloreferential
metapicture. The same holds true for the fake which is known to be a fake. A
fake that is taken to be an original, by contrast, is viewed as an original, it is not
a metapicture.

Parrhasioss superillusive curtain. In language, the enunciative frame is mostly

a matter of who is speaking; rarely does the question arise whether someone is
speaking at all. Situations are rare in which a speaker needs to signalize that he
or she is speaking and not, for example, breathing, listening, watching, eating,
or singing. While the question Do you hear me? is not unusual in everyday
conversation, a situation is hard to imagine where one would have to ask a
question such as: Are you speaking? Speech sounds do not compete with as
many acoustic signals emitted in the environment of a speaker. Pictures, by
contrast, are much more immersed in a visual eld of competing nonpictorial
signs, and it is no coincidence that the question picture or not? asked in the
context of the visual arts as art or not? has been one of the great topics of the
visual arts since Dada.
This is why pictures seem to be much more in need of an analogously fun-
damental self-referential message of the kind I am a picture. The question
was already raised in Greek mythology. Pliny tells the legend of two painters,
Zeuxis and Parrhasios, who competed for a prize for the best painting (Nat. hist.
xxxv: 65). Zeuxis exposed a picture of grapes painted so well that the birds ap-
proached to pick them. Proud of having deceived the birds to take his picture for
real, Zeuxis turned to his competitor and asked him to remove the curtain which
concealed his competitors picture. Parrhasios was triumphant. The curtain was
merely painted, and his work of art had not merely deceived birds but even his
fellow painter, who took a painted for a real curtain. The legend is about the
art deceiving by means of pictures (cf. Moeller 2003). Both painters succeed
in hiding the basic self-referential message by which their paintings convey the
message: I am a picture.
Metapictures and self-referential pictures 75

Magrittes anti-illusionist pipe. Magritte took the opposite road in questioning

the enunciative frame message. In his famous picture of a pipe with the seem-
ingly contradictory verbal legend Ceci nest pas une pipe, his strategy was to
destroy the pictorial super-illusionism a la Parrhasios. Read as a comment on the
enunciative picture frame, there is no contradiction at all in his verbal metames-
sage about the picture: the represented object must indeed not be confounded
with the representing picture.

9. Verbal paradoxes and metaleptic self-reference in pictures

Self-reference is one of the sources of paradox in language. Consider Epi-

menides, the Cretan, who says: All Cretans are liars. The utterance creates the
famous paradox according to which Epimenides either says the truth and thus
falsies the statement that all Cretans lie, since at least one Cretan is speaking
the truth, or he lies, like all other Cretans, but then his proposition would negate
itself and mean: It is a lie that all Cretans are liars, which by double negation
would mean the opposite, namely that no Cretan lies. The source of this paradox
is in its self-referential nature: the speakers utterance refers to the speakers
speech act, but the semantic level of the proposition enters into a conict with
the pragmatic level of what the speaker utters and the scope of what he refers
to, which is partly what others say, partly what he himself says.
The Epimenides paradox exemplies a mode of verbal self-reference which
is not metalinguistic in the narrower sense since the Cretan says nothing about
phonemes, words or sentences nor does he say anything about the Greek or the
Latin languages. It is true that the concept of metalanguage is occasionally also
extended to a broader sense which includes also self-referential speech acts such
as the one of Epimenides, and in this broader sense, one could argue that the
Epimenides paradox is also a metalinguistic one. However, there is an important
difference between metalanguage in the narrower sense and metalanguage in
such a broader sense. While metalinguistic terms such as word or sentence
always explicitly refer to language, the Epimenides paradox is only implicitly
self-referential. The source of the verbal self-reference is in the unexpressed
illocutionary act which can be made explicit in the form of the metalinguistic
paraphrase: I, the Cretan, say that all Cretans say untrue things.
There is a well-known parallel between Epimendess paradox and pictorial
representation: Eschers hand that draws itself. The kind of self-reference we
are confronted with in this picture is metaleptic self-reference (cf. Ryan 2004).
Metalepsis, a concept borrowed from rhetoric, is a narrative device by which a
narrator begins to participate in the life of the ctitious persons who are nothing
76 Winfried Noth

but his or her own creation, or vice versa, in which the characters of the authors
own narrative creation begin to enter into a dialogue with the author (as in
Pirandellos Six Characters in Search of an Author). The self-referential nature
of narrative metalepsis has its explanation in the enunciative frame: the addresser
turns into a self-addressed addressee. The metaleptic element in the picture of the
hand that is drawing itself consists in the transformation of the drawn hand into
a drawing hand. Eschers message seems to convey this paradoxical message:
I, the draftsmans hand am drawing a hand that draws itself. The device of
pictorial metalepsis enjoys some popularity in the genre of the cartoons where
gures of animals or humans occasionally begin to change their appearance
because the drawn gure begins to change its own drawing by adding, deleting
or changing the lines and shapes of its own drawing.

10. Conclusion and prospects

The differences and overlaps between metapictures and self-referential pictures

were the topic of this paper. Despite the many distinctions set up in this study,
it remains to be pointed out there are still other forms of pictorial self-reference
which deserve to be discriminated (cf. Noth, this vol., Part II). Furthermore, the
study of the categories established in this paper may reveal the necessity of ad-
mitting gradual transitions between metapictures, self-referential pictures, and
self-referential metapictures. From a more general semiotic perspective, it may
be even true that all metapictures evince some kind of self-reference and that
all self-referential pictures may, in a way, be metapictures. After all, metapic-
tures are pictures of pictures, which suggests a certain self-referential loop.
Both metapictures and self-referential pictures invite reections on the nature
of pictorial representation. However, the very general common denominators
of metapictures and self-referential pictures do not invalidate the more subtle
distinctions that can and should be drawn between them.


1. See Veja 38.33 [Aug 17], 2005, p. 140.

Metapictures and self-referential pictures 77


Alessandria, Jorge
1996 Imagen y metaimagen. Buenos Aires: Universidad, Faculdad de Filoso-
a y Letras
Anati, Emmanuel
1994 Constants in 40,000 years of art. In: Winfried Noth (ed.), Semiotics of
the Media, 385403. Berlin: Mouton de Gruyter.
Austin, John L.
1962 How to Do Things with Words. Oxford: University Press.
Buhler, Karl
[1934] 1978 Sprachtheorie. Frankfurt: Ullstein.
Conant, Chloe
2005 Mise en abyme / mirror text. Dictionnaire international des termes
litteraires. (18.05.06).
Derrida, Jacques
1972 Marges de la philosophie. Paris: Minuit.
Eco, Umberto
1986 Mirrors. In: Paul Bouissac, Michael Herzfeld and Roland Posner (eds.),
Iconicity: Essays on the Nature of Culture, 215237. Tubingen: Stauf-
Harweg, Roland
1990 Studien zur Deixis. Bochum: Brockmeyer.
Hockett, Charles
1977 The View from Language. Athens: University of Georgia Press.
Machado, Arlindo
1984 A ilusao especular. Sao Paulo: Brasiliense.
Metz, Christian
1991 Lenonciation impersonnelle ou le site du lm. Paris: Klincksieck.
Mitchell, W. J. Thomas
1994 Picture Theory. Chicago: University Press.
Moeller, Hans-Georg
2003 Before and after representation. Semiotica 143: 6978.
Noth, Winfried
1990 Handbook of Semiotics. Bloomington: Indiana University Press.
2003a Photography between reference and self-reference Fotograe zwi-
schen Fremdreferenz und Selbstreferenz. In: Ruth Horak (ed.), Re-
thinking Photography I+II: Narration and New Reduction in Pho-
tography Narration und neue Reduktion in der Fotograe, 2239.
Salzburg: Fotohof Edition.
78 Winfried Noth

2003b Press photos and their captions. In: Harry Lonnroth (ed.), Fran Narpes-
dialekt till EU-Svenska: Festskrift till Kristina Nikula, 169188. Tam-
pere: University Press.
2004 Semiotic foundations of the study of pictures.  : Sign
Systems Studies 31(2): 377392.
Noth, Winfried and Christina Ljungberg (eds.)
2003 The Crisis of Representation: Semiotic Foundations and Manifesta-
tions in Culture and the Media. (= Special Issue of Semiotica 143.14).
Owens, Craig
1978 Photography en abyme. October 5: 7388.
Prieto, Luis J.
1966 Messages et signaux. Paris: Presses Universitaires de France.
Ryan, Marie-Laure
2004 Metaleptic machines. Semiotica 150: 439469.
Santaella, Lucia and Winfried Noth
1998 Imagem: Cognicao, semiotica, mdia. Sao Paulo: Iluminuras. Trad.
Roque Graciano. 2003. Imagen: Comunicacion, semiotica y medios.
Kassel: Reichenberger.
Absolut Anonymous:
Self-reference in opaque advertising

Nina Bishara

1. Referential and self-referential advertising

According to the old AIDA-formula of marketing strategists, advertising aims at

attracting attention, raising interest, creating desire, and initiating action. With
these goals, advertising is a means to an end but never an end in itself. The means
begin with attracting attention for a message which praises available goods or
services for their alleged positive qualities. The end of getting action is to
create consumption and prot. Both means and ends testify to the referential
nature of advertising.
Advertising wants to convey a message about something. As soon as we
discover that we are confronted with an advertisement, we know that we are
faced with a referential, or, more precisely, alloreferential message a message
referring to a segment of the market. Self-referential advertising, that is, adver-
tising that refers to itself instead of referring to products and services, creates a
paradox: since a self-referential message is one that refers to nothing but itself,
how can such a message be about goods and services, and how can it result in
economic action? Can such a message be an advertisement at all?

2. Opaque advertising as a form of self-reference

The topic of this paper is opaque advertising, i.e., advertising in which the prod-
uct or service, the trademark, the advertiser and the advertising message itself
are disguised and can only be discovered through a special effort of advertising-
literate interpreters. Depending on the degree of the interpreters advertising
literacy, such ads may trigger incomplete processes of semiosis as dened by
Peirce: they may fail to result in the triadic process in which a sign, determined
80 Nina Bishara

by its object, generates an interpretant. The object by which the advertisement

is determined may remain opaque, without any indices, except for those which
point back to the message itself. The object of the sign which pretends to adver-
tise remains enigmatic, and the interpretant incomplete. The enigma, as will be
argued, does not so much concern the identication of the text genre itself since
formal indicators such as placement, layout, typography, and other graphic de-
vices clearly distinguish the advertisement from its surrounding messages and
so announce, in a rst step towards semiosis, that this is advertising. However,
in interpreting what has been identied as an advertisement, the readers may not
be able to identify the real object of the advertising message without investing
more time and effort in decoding it.
The semiotic framework of this paper is Charles S. Peirces triadic sign model.
What other semioticians have called reference is the relationship between the
sign and its object, according to Peirce, while self-reference means that the sign
refers to itself, instead of to an object which is elsewhere. Peirce distinguishes
two kinds of objects, the immediate and the dynamical object, and he further
differentiates three kinds of interpretants: the immediate, dynamical, and the
nal interpretant.

3. Semiotic premises

Nothing is a sign unless it is interpreted as a sign (CP 2.308). With these

words, Charles S. Peirce describes the pragmatic and functional character of the
sign process. A sign can be described as a mediator between an object and a
mind. In advertising, this process, called semiosis, occurs as follows: potential
customers must become aware of the advertisement (the sign) whose object is
a product, a service, a market, a company, etc. The consumers minds must be
affected by the message about the object and perhaps become motivated to buy
or consume the product promoted by the advertisement. These are results of
the sign which constitute its interpretant. The semiotic and economic process
cannot be triggered if the potential consumers are unable to interpret the message
According to Peirce, any process of semiosis involves the three universal
categories of rstness, secondness, and thirdness. Firstness refers to the mere
suchness of still undifferentiated and independent possibilities and qualities.
It is a mode of being without reference to anything else. An advertising message
perceived in its mere rstness would not yet be a message at all since its verbal
and pictorial elements would be perceived without being related to anything else.
At the level of secondness, a rst is related to a second. The rst is the adver-
Absolut Anonymous: Self-reference in opaque advertising 81

tisement, the second the product, the service, etc. Secondness predominates in
indexical advertising which points to the existence of the product or commodity
and appeals to its purchase. Thirdness presupposes rstness and secondness; it
is the category of mediation (between a rst and a second), of meaning, expec-
tation, and habit. In advertising, it is the phase of familiarity with the message
and with the product. Thirdness is fully developed in an advertisement when it
is anchored in memory and habit.
For Peirce, signs constitute a triadic relation of a signier, a thing signied,
and an interpretant created in the mind of an interpreter. According to one of
his most frequently cited denitions:

A sign, or representamen, is something which stands to somebody for some-

thing in some respect or capacity. It addresses somebody, that is, creates in
the mind of that person an equivalent sign, or perhaps a more developed
sign. That sign which it creates I call the interpretant of the rst sign. The
sign stands for something, its object. It stands for that object, not in all
respects, but in reference to a sort of idea. (CP 2.228)

Applied to advertising, these three correlates of the sign are the sign, the
object, and the interpretant. The sign (or representamen) is the advertising text
as such, for example, a magazine advertisement or a TV commercial; elsewhere,
Peirce calls it the perceptible object (CP 2.230) which functions as a sign.
The object of the sign is essentially the product, the service, the corporate
image of a company or an ideology which the advertisement conveys in its
message. However, the sign cannot establish the recognition of the object by
itself; it rather presupposes the acquaintance with its object. In semiosis, the
object thus actually precedes the sign. Furthermore, Peirce distinguishes two
kinds of objects, the immediate and the dynamical object:

We have to distinguish the Immediate Object, which is the Object as the

Sign itself represents it, and whose Being is thus dependent upon the Repre-
sentation of it in the Sign, from the Dynamical Object, which is the Reality
which by some means contrives to determine the Sign to its Representation.
(CP 4.536)

The immediate object is the object as it is represented by the sign whereas

the dynamical object is in a way the reality from which the sign originates.
However, the latter is inaccessible; it can only be indicated by the sign and
requires for its recognition the interpreters collateral experience of this object
(CP 8.314). According to Peirce, no sign can be fully understood unless the
interpreter has collateral acquaintance with every Object of it (CP 8.183).
This knowledge of the dynamical object, or, in Peirces terms: its reality is one
82 Nina Bishara

which can only be apprehended by approximation. In their unlimited scientic

research, scholars may approach the dynamical objects of the signs more and
more.1 If, in advertising, the immediate object is that commodity or service
as represented by the advertisement, then the dynamical object includes all the
economic forces of the market, the most detailed product information, price
regulations and mechanisms etc. intricacies of which even economists or
market researchers are never fully aware.
The ideas, thoughts, conclusions, impressions, and actions created or trig-
gered by the advertising text in the mind of an interpreter constitute its inter-
pretant, which can in its turn become a new sign in the unlimited process of
semiosis (CP 2.303). The interpretant of a sign is a sign itself. When the sign
evokes an idea, for example, the interpretant is a mental sign, when it evokes a
question or a reply, it is a verbal sign, and when it evokes an action, the inter-
pretant is a nonverbal sign. Each sign thus becomes a new sign in an unlimited
process of semiosis (CP 2.303). Peirce differentiates three types of interpretants:
the immediate, the dynamical, and the nal interpretant (CP 8.315).2 The im-
mediate interpretant is any impression or feeling that may be produced by the
quality of a sign (e.g., the color design of or the music in an advertisement). It is
the rst effect a sign produces but not yet any concrete reaction or direct effect
produced by a sign, the latter being the dynamical interpretant, for example,
to comply to the appeal Come in and nd out, as the slogan of a perfumery
demands. Among the effects of the dynamical interpretant are mental agitation
and physical action, the consumers curiosity and his movement towards a shelf

Figure 1. The sign in advertising in a Peircean perspective

Absolut Anonymous: Self-reference in opaque advertising 83

in supermarket in order to buy a product. The logical interpretant is determined

by law or habit. In science, it is the nal result of research for which truth and
consensus can (ideally) be claimed (e.g., a dictionary denition of a word).
In advertising, it is the consumers habit to consume what the advertisement
Figure 1 represents the three correlates of Peirces sign model as applied to
the genre of advertising.

4. Exemplications

The following examples of opaque advertising show how the object and the
interpretant of the advertising message are not immediately apparent. Readers
may even fail to be affected by the object and so create an interpretant unrelated to
it when attempting to decode the advertisement. The semiotic process in opaque
advertising may result in confusion, puzzlement, and astonishment since the sign
seems to represent nothing at all. The sign does not fulll its straightforward
function of representing its object; instead, it refers back to the advertisement
itself, thus coming close to being self-referential.
Figure 2 shows an example of an (initially) self-referential advertisement in
the form of a pictorial riddle. Whereas the text can clearly be identied as an

Figure 2. McDonalds company logo in disguise3

84 Nina Bishara

advertisement due to the contextual frame (here: a page in a magazine that is

markedly different from the surrounding editorial texts), the product advertised
for and the sender of the message remain opaque to the average recipient due
to the apparent lack of verbal or pictorial cues. The immediate object of this
message is an indenable red background with a green sign depicting three
white pictograms: the one of a running man, of an arrow pointing to the right
and of an image resembling two arches or the capital letter M. From their
world knowledge, collateral experience, as Peirce would call it, the readers
may identify the green sign with the white pictograms as an emergency exit
sign this format being the international convention for emergency exit signs
and so may translate the pictograms to indicate one of the following messages as
its interpretant: In case of an emergency run in the direction of (a) . . . two arches
(unspecied) / (b) . . . a meeting point (unspecied) / (c) . . . M (unspecied).
It remains open in where one should run to in case of emergency. The reader
can therefore conclude that the solution to the riddle lies in the identication of
the unspecied destination. Recipients familiar with the advertising world and
the world of company logos, that is, consumers for whom a prior collateral
acquaintance of the object can be presupposed, will decipher the two arches as
a stylized letter M, the symbol for the fast-food chain McDonalds which is
usually yellow against a red background. By change of color, the advertisers have
simulated an object of the sign which is difcult to access and they have created
an interpretant which will mislead its readers for a while. Those recipients who
succeed in recognizing the company logo will create an interpretant which relates
it to their experience of those places in which they became acquainted with the
fast-food company and its services, thus reversing the initial self-reference into
reference to the world of McDonalds. Their prior acquaintance with the company
logo has made its re-cognition possible. To those who recognize McDonaldss
message, the immediate object, the object as the sign represents it, is closer to
the dynamical object of this message. Those who are misled either because
they fail to recognize the disguised company logo or simply because they are
not acquainted with it as an established symbol remain unaffected by this
dynamical object. They must fail to recognize the logo for lack of sufcient
collateral experience with it and can take the advertising message only as an
incomplete rhematic sign whose message remains open. Their interpretation is
restricted to the level of an immediate interpretant creating feelings of confusion,
surprise, or lack of relatedness to anything.
Figure 3 shows an advertisement that deliberately masks the identity of the
product which it promotes by concealing the product name which is completely
blurred on the label of the depicted bottle. Hence, the slogan beneath the picture
of the bottle is nothing but redundant since it only repeats what we have already
Absolut Anonymous: Self-reference in opaque advertising 85

Figure 3. An opaque product name (Absolut Vodka)4

concluded from the blurred label: this message is absolut(ely) anonymous.The

object represented in the sign remains enigmatic. Although the advertisement
succeeds in attracting the viewers attention and maybe even in raising their
interest, desire and action cannot be elicited as long as the identity of the product
remains opaque and the advertisement only refers to itself.5
To avoid the complete failure of the message, the advertisement contains
some clues which can help to decode its message. A rst clue is provided by the
slogan absolut anonymous which strikes its readers as grammatically incor-
rect and which should read absolutely anonymous since the word preceding
anonymous should be an adverb (even in the form of an adjective it should be
spelled absolute). The word absolut evinces a grammatical mistake which
serves to attract the attention of its readers to this slogan. These might recall the
name of the Swedish vodka called Absolut, and this piece of previous knowl-
edge (collateral experience) allows a better interpretation of the advertisement.
Those readers, however, who are unfamiliar with the brand name Absolut still
face the riddle of a self-referential message. However, if they pay attention to the
86 Nina Bishara

small print text on the label of the bottle, they will also succeed in identifying
the product. This text informs them that the bottle depicted in the advertisement
contains liquor (40% vol.), specically superb vodka that has been sold
under the name Absolut since 1879. For readers who do not participate in the
search for the object, the message remains self-referential and lacks any indices
that point to something other than itself.

Figure 4. An opaque brand (Volkswagen New Beetle)

(bottom left reads: Seen at the mall, bottom right reads: Drivers wanted VW)6

Verbal anchorage plays an important role in the advertisement depicted in

Figure 4. At rst sight, the advertisement seems to follow traditional advertising
strategies. It shows a picture, albeit not exactly an attractive one, a company logo
(VW), and an appeal in the form of the slogan Drivers wanted (bottom right).
Consumers can quickly deduce that this is an advertisement for Volkswagen
automobiles. The message fulls the complete triad of a sign (the advertisement
itself), its object (VW automobiles, available for purchase) and the interpretant
(consumers eager to purchase VW cars and convinced of their unquestionable
A legend in the lower left-hand corner of the advertisement, however, gives
rise to confusion. Its elliptical message is: seen at the mall. The statement
arouses the readers curiosity: what was seen at the mall? The text can only refer
to the picture depicting a tiled oor with some cracks in it and a shopping bag in
Absolut Anonymous: Self-reference in opaque advertising 87

its right-hand corner. There is no apparent relationship between this picture and
VW automobiles. Nevertheless, the legend asserts that there is something to be
discovered in the picture. At a closer look, the object of this pictorial message can
be identied: one dent in the tiled oor is especially prominent; it has the shape of
a Volkswagen Beetle. As in the other cases of opaque referents discussed so far,
the dynamical object of the sign which seemed to be invisible was indeed present
all the time. However, the identication of this cryptic object of the message
does not only require the readers efforts to decode the advertisement. It is
also necessary to bring in knowledge about the world of products, markets, and
advertisements. Such knowledge necessarily precedes the sign and is determined
by its dynamical object. In the VW example, consumers must be familiar with
the form of the legendary Beetle car in order to solve the riddle. Readers who
do not recognize the shape of this car will still be able to extract a very general
message, namely that this is an advertisement for VW automobiles, but they
would not grasp its additional ludic value. In other words, the self-referential
feature of this advertisement is rooted in the legend and its comment on the
picture seen at the mall. This comment remains enigmatic and self-referential
as long as readers cannot identify the object that leaves its trace in the picture
seen at the mall.
Let us consider one last example. In the summer of 2000, a new energy
provider on the German market began to make future customers aware of its
existence by launching a campaign that attracted everybodys attention. The
campaign consisted of print advertisements, billboards, and TV commercials,
all of which showed nothing but the color red, i.e., there was a completely red
page in a magazine, a red poster, or, for a few seconds, there was nothing but
red as a commercial on TV. As the name of the energy provider was not included
in the campaign and since the unidentied company was absolutely new to
the market, nobody could draw upon any prior knowledge about the rm and
its advertising strategies. The consumers were faced with an utterly enigmatic
and nothing but self-referential advertising message. Not a single clue as to
the meaning, sender, or intention of the advertisement was provided on the red
surface. Nevertheless, extratextual markers clearly identied the messages as
advertisements: on TV, the red surface appeared during the commercial break;
in news magazines, the headline Advertising distinguished the red page from
its surrounding editorial texts, and posters were placarded on the appropriate
hoardings for outdoor advertising.
The mere presentation of the color red without reference to anything par-
ticular in space or time leaves the advertising message open to and indenite
number of interpretations. The message is a rhematic qualisign, characterized
twice as a sign of rstness and mere suchness. A sign of this kind that remains
88 Nina Bishara

without reference to anything else is self-referential in every respect. No object

relation (secondness) can be established and no triadic process of semiosis can
take place. Although consumers may identify the color red as the symbol of
love, danger, or of the political parties of the left, there are no signs in this ad-
vertisement that indicate and support any interpretation of this kind. Reference
to the tradition of monochrome painting, e.g., Monochrome Crimson by Claude
Tousignant (1981), does not lead to any solution of this opaque message either.
The consumers are confronted with a genuine riddle which, in contrast to the
above examples, does not entail any stylistic or formal indices conducing to a
meaningful advertising message. The object of the message must remain enig-
matic, although there might be some kind of secondness in it based on the color
red, e.g. attention, compulsion to act (red trafc light), etc. Figure 5 illustrates
this with reference to Peirces categories of the object and interpretant.

Figure 5. Opaque object and enigmatic interpretant in a self-referential advertisement

What is the aim of such an advertisement? It certainly attracts attention

by violating the conventional means of advertising. It also requires a follow-up
campaign that offers a solution to the riddle. A few weeks upon appearance of the
red-color-campaign, the red surface was complemented by the company name
E.ON and the motto Neue Energie (new energy). As the energy provider
E.ON was new to the market, the advertisers wanted to achieve public awareness
of the new company and to ensure future associations of the color red with E.ON.
The companys homepage commented on its advertising strategy as follows:
Absolut Anonymous: Self-reference in opaque advertising 89

Germany literally saw red. After the company color was acquired, the secret
was disclosed.7 Red became a part of the companys corporate identity and is
still dominant in its advertising campaigns.

5. Riddles and opaque advertising

The examples of opaque advertising discussed so far were all classied as enig-
matic. Research in riddles has distinguished between true and false riddles such
as joking or wisdom questions (Abrahams and Dundes 1973). The latter, al-
though enigmatic, cannot be answered from the content presented in the ques-
tion, the solution to joking questions being entirely arbitrary8 and the answer to
wisdom questions depending on the riddlees general knowledge.9 True riddles,
by contrast, although also designed for the purpose of confusion, are questions
with descriptions or cues that allow the referent to be guessed. Riddles are re-
quired to use their wits and to interpret or read the signs correctly in order to
arrive at the proper solution.10
Enigmatic advertisements which leave the object of their message opaque are
like true riddles. A completely enigmatic advertisement incapable of conveying
a message about its object must necessarily be unsuccessful. Advertisers cannot
afford the risk of empty messages. They must incorporate reliable clues to help
the consumers in the decoding of their messages. The consumers should get the
message and are indeed prepared for it. After all, they have a clear notion of the
aims of advertising and know its core message, which is the promotion of goods
and services. They know that an advertisement is never without this purpose
and will therefore be alert to discover the slightest textual clue of this intent.
Riddles without a solution cannot be admitted. It is true that E.ON used this
device for a time, but it was only a means of temporary suspense. The solution
was ahead in the continuation of the advertising campaign.

6. Style and function of opaque advertising

Enigmatic advertising is a stylistic device aiming at creating curiosity by means

of deviations from the standard patterns of the genre which suggest to the con-
sumers that they are faced with something different, worth their attention. Rid-
dles have something in common with metaphors, as Aristotle observed. Both
can be used with prot in the art of persuasion. The consumers are invited to
invest an extra effort in the decoding of the message. However, the consumers
willingness to dedicate time to solving riddles posed by advertisers is limited.
90 Nina Bishara

If the message remains too opaque and turns out to be merely self-referential
the purpose of the message will fail. A self-referential message that only points
back to itself cannot convey any advertising message. Without a specic referent
its message is open to many possible readings.
Deviation from the standard patterns of advertising messages is a means of
attracting attention. It is one of the strategies of poetic semiosis: not only does
the advertising text which deviates from the conventional expectations about
the genre attract the consumers attention; it also demands more effort in the
process of decoding. Assmann (1988) has introduced the term wild semiosis
to describe such processes of creativity: by violating the standard patterns of
expectation something unexpected and unknown is being created which turns the
viewers attention to the materiality and form of the text as such. Self-reference
is one of the means of wild semiosis.
In addition to its poetic quality and its potential to capture the consumers
attention, opaque advertising is likely to create intellectual pleasure. Riddling is
fun and makes the advertisement attractive as well as entertaining. Furthermore,
it demonstrates the consumers competence and advertising literacy. Above all,
opaque advertising, despite its risk of failing, is likely to contribute towards a
higher memory value of the message and thus to the success of the advertising

7. Conclusion

Advertising is essentially a means to an end but never an end in itself. The mes-
sages of the marketplace are signs whose objects are the products and services
which the advertisers want to promote. In order to promote a product, a mes-
sage must refer to it. The referential function, a highly indexical sign process,
tends to be the primary function of advertising. However, creative advertis-
ers have discovered the advantages of hiding their messages in the disguise of
self-reference. Their messages are opaque and seem to point to nothing but to
themselves. Such messages seem to constitute a paradox. Can they still achieve
their purpose of promoting a product? Furthermore, they seem to be counterpro-
ductive to the primary aims of transmitting a message to the consumer. Opaque
advertising contains indices of an object which remains concealed and creates
an interpretant which is likely to remain incomplete. Its enigmatic message is
self-referential as long as the consumers do not succeed in discovering the ob-
ject of reference. However, opaque advertisements never remain truly opaque
and self-referential. Consumers who pay attention to their message are bound
to discover the referent in disguise. Enigmatic advertisements are thus likely
Absolut Anonymous: Self-reference in opaque advertising 91

to capture better the consumers attention, arouse their interest and desire, and
implant their message better in the memory of those who have to pay for what
they saw when they decide to purchase the product.


1. An example of an immediate and dynamical object given by Peirce is the statement

The sun is blue of which the perception or sensation of the suns blueness constitutes
the immediate object, while the physical phenomenon of wave lengths causing the
blue appearance is the dynamical object (CP 8.183).
2. These terms are sometimes used synonymously with the emotional, energetic, and
logical interpretant. Some scholars, however, understand the emotional, energetic,
and logical interpretant as a subdivision of the dynamical interpretant, and for some
it is a threefold subcategorization of the triad of the immediate, dynamical, and nal
interpretant (Noth 2000: 65).
3. See Willi Schalk, Helmut Thoma and Peter Strahlendorf (eds.). 2000. Jahrbuch der
Werbung 2000. Berlin: Econ, p. 144.
4. See Der Spiegel 46/2002, p. 62.
5. Advertising-literate recipients will of course immediately grasp the pun of this ad-
vertisement and will identify its referent since Absolut vodka has maintained its
advertising style for several decades: depiction of the vodka bottle and a two-word
slogan always beginning with Absolut. This repetitive style of advertising is self-
referential in a different way.
6. See Volkswagen Advertising Database, Nr. BE020USPR.
7. (03.02.03).
8. Example: What is red and green and goes round and round? A frog in a mixer.
9. Example: Which is the rst book of the Bible? Genesis.
10. It is interesting to note that the words riddle and to read are related etymologically.


Abrahams, Roger D. and Alan Dundes

1973 Riddles. In: Richard M. Dorson (ed.), Folklore and Folklife, 129143.
Chicago: University of Chicago Press.
Assmann, Aleida
1988 Die Sprache der Dinge. Der lange Blick und die wilde Semiose. In:
Hans Ulrich Gumbrecht and K. Ludwig Pfeiffer (eds.), Materialitat
der Kommunikation, 237251. Frankfurt: Suhrkamp.
Noth, Winfried
2000 Handbuch der Semiotik, 2nd rev. ed. Stuttgart: Metzler.
92 Nina Bishara

Peirce, Charles S.
19311958 Collected Papers of C. S. Peirce. Ed. Charles Hartshorne, Paul Weiss
and Arthur Burks. Cambridge, MA: Harvard University Press. Quo-
ted as CP.
Part III. Self-referential photography
The death of photography in self-reference

Winfried Noth

1. The death of photography and the birth of the

postphotographic era

The Death of Photography is the provocative title of an anthology with papers

on the role of photography in the visual arts of the 19th and the 20th centuries
(Beckley and Aguilar 2007). At the occasion of its publication, art critics, art
historians, semioticians, and ironically also photographers (Bill Beckley,
Samuel Jamier, Siri Hustvedt, Winfried Noth, Max Kozloff, and Renee Cox)
discussed the topic at a symposium in the Museum of Modern Art in New
York on May 4, 2005. It was not the rst time that the death of photography
had been declared in face of the technological transformations of the medium
since the advent of digital imaging. Ritchin (1990) foresaw a revolution in
photography more than a decade ago, Mitchell (1992) announced the beginning
of a postphotographic era in the 1990s (cf. Santaella 1997; Santaella and Noth
1998: 166; Carani 1999), and David Acton (2004) even declared the end of
photography in the last chapter of his book on Photography at the Worcester
Art Museum.
Ironically, photography itself had once been considered the cause of the death
of another pictorial medium, namely painting. In 1839, after examining pho-
tography commissioned by the French Academy for Daguerrian photography,
Paul Delaroche announced: From this day on, painting is dead (apud Weibel
2002: 611).
It is, of course, a paradox when those who continue producing photographs
and exhibiting them publicly declare the death of photography, but it is well-
known that we are merely confronted with a metaphorical death. The death
which this metaphor evokes means the birth of the new digital technologies of
image making and processing by means of cameras. Photography is dead, long
live photography! But if photography continues to exist as a technologically and
semiotically transformed medium, what is it that has died in postphotography?
96 Winfried Noth

The now defunct photography is the one about which Roland Barthes (1980:
87) once said: Photography never lies: or rather, it can lie as to the nature of the
thing, being by nature tendentious, never as to its existence. The traditional pho-
tograph does not lie in the sense that the light rays emitted by the object are pro-
jected via the lens and registered by the lm in a process of optical and chemical
causality. In Mitchells words (1992: 24): A photograph is fossilized light, and
its aura of superior evidential efcacy has frequently been ascribed to the special
bond between gurative reality and permanent image that is formed at the instant
of exposure. It is a direct physical imprint like a ngerprint left at the scene.
The faith in the truthfulness of photography which derives from this natural
causality is as old as the history of the medium. Early in the 19th century, the
painter Eugene Delacroix (17991863) praised the daguerreotype, the precursor
of the photograph, for the reliability of the picture in its faithful reection of the
object: Daguerrotypy is more than a blueprint, it is the reex of the object,
and Hypolyte Taine, in his Philosophie de lart of 1865 praised the truthfulness
of the new medium: Photography is the art which imitates on a plane surface,
with lines and tones, with perfection and without possibility of error the form
of the object which it must reproduce (apud Arrouye 1978: 74).
The truthful photograph, causally connected to its referent and dying in
the age of digital simulation, is that medium whose pictures, despite all their
similarity to their referential object, Charles Sanders Peirce dened as primarily
indexical signs:
Photographs, especially instantaneous photographs are very instructive, be-
cause we know that they are in certain respects exactly like the objects they
represent. But this resemblance is due to the photographs having been pro-
duced under such circumstances that they were physically forced to corre-
spond point by point to nature. In that aspect, then, they belong to the second
class of signs [i.e., to the indexical signs], those by physical connection. (CP
2.281, circa 1895)

2. The loss of the referent and the crisis of representation

Some have hailed the revolutionary changes which digital photography has
brought about, while others have deplored the losses that the postphotographic
media have been suffering with the death of photography. The praise of the
new medium is for the expanded creative potential of digital image making. As
summarized by Kevin Robins (2007: 22): There is the sense that photography
was constrained by its inherent automatism and realism, that is to say, by its
essentially passive nature; that the imagination of photographers was restricted
The death of photography in self-reference 97

because they could aspire to be no more than mere recorders of reality. From a
historical perspective, the postphotographic era also leads back to the precursor
of photography, painting, which it once seemed to have dethroned:
Since captured, painted, and synthesized pixel values can be combined
seamlessly, the digital image blurs the customary distinctions between paint-
ing and photography and between mechanical and handmade pictures. A
digital image may be part scanned photograph, part computer-synthesized
shaded perspective, and part electronic painting all smoothly melded
into an apparently coherent whole. It may be fabricated from found les, disk
litter, the detritus of cyberspace. Digital imagers give meaning and value to
computational readymades by appropriation, transformation, reprocessing,
and recombination; we have entered the age of electrobricolage. (Mitchell
1992: 7)
Parallel with the praise of the enhanced potential of visual expression, the
postphotographic era has brought about theoretical encomia of the new semiotic,
aesthetic and even psychoanalytic horizons of digital image making, such as:
The certainties of the photographic era have been deconstructed, and we are
now ready, it seems, to come to terms with the fragility of ontological distinctions
between imaginary and real (Robins 2007: 24).
From the semiotic perspective, the most important revolutionary feature of
the postphotographic era seems to be that the new images have become in-
dependent of referents in the real word (Robins 2007: 22). Although this
independence may enhance the artists creative potential, the ensuing loss of
the referent has also given rise to discourse about a crisis between reality and
its image in the age of electronic simulation (Grundberg 2007: 63). In a crit-
ical perspective, this loss of the referent can disturb and disorient by blurring
comfortable boundaries and by encouraging transgression of rules on which we
have come to rely (Mitchell 1992: 223). The new potential of absolute ma-
nipulation and simulation of merely virtual worlds seems to have subverted
traditional notions of authenticity and originality (Robins 2007: 27).
However, if the death of photography is the advent of pictures which have
lost their referent this death must have occurred many times in the history
of photography. The death of photography is a manifestation of the crisis of
representation (cf. Noth and Ljungberg 2003) whose roots in the visual arts are
the roots of modernity with which we have been familiar since impressionism,
pointillism, cubism, abstract painting, Dada, etc.
It is indeed nave to assume that there has been an uninterrupted tradition of
indexical photographs which point to a referent. The loss of such referents has
begun early in the history of photography, and there have been many forms and
modalities of this death, e.g., deletion or insertion by retouch with the purpose of
98 Winfried Noth

deceit, defamiliarizing the referent by representing it incompletely, in a distorted

way, by double exposure, or by abstraction from the referential object with an
aesthetic purpose.
The loss of the referent in traditional photography can have its cause in
alloreference or in self-reference. Alloreference, that is, reference to something
else, is the expected mode of reference of an ordinary photograph. The photo
is an indexical sign of its referent, a person, a building, a landscape, etc. When
this object disappears because it is distorted, fuzzy, hidden, unrecognizable, or
falsied by retouch, the loss of the referent is an alloreferential one.

3. Seven self-referential deaths of the photographic referent

A different kind of disappearance of the referent in photography is due to self-

reference. Instead of being the indexical sign of a referential object, the self-
referential photo shows only itself, it draws attention to nothing but the photo,
and invites the eye to get immersed in the forms and colors of the photographic
composition. By negating alloreference, the self-referential picture negates the
referent as such, either partially or entirely. Various forms and strategies of
pictorial self-reference have been invented in the history of photography. In the
following, we will discuss seven of them.

3.1. The negated self in a paradoxical self-portrait

In 1839, the photographer Hippolyte Bayard (18011897) portrayed himself in

a photo entitled Self-portrait of a drowned man. It shows him in a reclined
position half naked and with closed eyes (see Mitchell 1992: 194). Like any
self-portrait, the picture is to a certain degree self-referential: the sender of the
message is part of the message himself so that the picture refers to the source
of its own production. The self-referential message of this portrait is true and
false at the same time. It is true that we see a picture of the photographer, but it
is false that he is dead, as the title claims. The referent is thus partially hidden
by the false claim of the title. Parallel with this false indexical message attached
to the referent, we are faced with a logical paradox, as it often occurs in self-
referential statements. It requires no indexical reference at all to realize that
the self-referential title which is supposed to identify its referent is false. It is
a matter of logic that a dead man cannot portray himself. The solution to this
paradox leads back to the photo, and we readily conclude that this photo is not
a self-portrait of a drowned man.
The death of photography in self-reference 99

3.2. The camera in the picture

About 1880, August Giraudon took the photo Mirrors in Versailles Caste (see
Govignon 2004: 37). It is a picture of a large mirror in the Versailles mirror hall,
which does not only show a highly ornamented wall reected in the mirror on
the opposite wall, but also, right in the center of the picture, the large camera
on a tripod which is taking this very picture. Strangely, the photographer is
absent in the picture. The self-referential aspect of the photo is evident. The
camera, instrument of taking a picture other than itself, represents itself, but
it shows only partially the referent it is supposed to depict. There is a loss of
the referent, the mirror in the palace, insofar as the view of the reected palace
wall is obstructed by the camera. Again, the result is a paradoxical relationship
between the verbal and the pictorial message. While the verbal message states
Mirrors in Versailles Castle, the pictorial message is: This picture does not
(only) show mirrors in Versailles Castle, but it shows the taking of a picture of
mirrors in Versailles Castle. Furthermore, the photographers absence leaves
the eye with the suspicion of a second paradox: can a photo be taken in absence
of a photographer?
The self-referential presence of the photographers instrument of image mak-
ing, the camera in the picture, violates the ancient principle of ars est celare
artem [art requires the hiding of art], which goes back to Quintilian and Ovid (cf.
Wetzel 2003). The artist should show in an alloreferential way something other
than himself. He should avoid self-reference by concealing any indices of his
artistic doing. The violation of this ancient principle makes this self-referential
picture a picture and metapicture at the same time.

3.3. Abstraction from the referent

The third death of the photographic referent can be illustrated with Paul
Strands Abstraction. Porch Shadows of 1917 (see Govignon 2004: 77). The
picture shows an aesthetically well-formed pattern of parallel stripes of black
and white, but it seems impossible to recognize the form of an object. Despite
the indication given by the title, which tells us that we see porch shadows, the
referential object remains indiscernible. It even remains unclear whether the
enigmatic pattern represents shadows of a porch or in a porch.
However, we are not disappointed about the absence of the referent since the
aesthetic quality of the photo resides entirely in its material and formal qualities.
Although we are informed that there is reference, the impression of a missing
referent predominates. Once more, we are confronted with a paradox. The title
100 Winfried Noth

afrms the existence of a porch, whereas the pictorial message is that there is
none. The loss of the referent is total, but the self-referential gain in aesthetic
quality compensates this loss entirely.

3.4. Double exposure

Under the inuence of cubism and Dada, experimental photography of the 1920s
and 1930s discovered the technique of double exposure with the result of pictures
which were indexical signs of two referents. Such double exposure seems to
result in the opposite of the loss of the referent, since a photo exposed twice
is a sign of at least two referents. However, the pictorial space of a photo is
restricted, and the doubling of the referent can only be achieved at the cost
of the pictorial space of one of the two overlapping pictures. The overlapping
causes the disappearance of the part of the underlying referent that is hidden
by the overlap. In this sense, there is a loss of the referent. Consider Maurice
Tabards Composition (Double Exposure) of 1931 (see Govignon 2004: 277).
It shows the lateral view of a face overlapped by the front view of the same
face. The referent that disappears partially is the one of the face in its lateral
The relation between the photo and its referential object is one of double
indexicality insofar as the two views of the same face refer to its two perspectives
seen at two different moments from two different points of view. Internally,
however, the photo evinces the repetition of the same of the face from its two
perspectives, which makes one perspective the iconic representation of the other
and vice versa. This internal iconicity is a mode of self-reference. Once again,
pictorial self-reference results in a pictorial paradox: the photo cannot be an
indexical photo since it reects an impossible perspective of the real face, but
at the same time it is nothing but a photo.

3.5. Aesthetic deconstruction

The partial destruction and similar forms of manipulation of the paper surface
of a photograph for the purpose of its exhibition in an art gallery leads to another
mode of self-reference, which exemplies our fth death of the referent in
photography.Arnulf Rainers self-portrait of 1951, entitled The empty painting
(see Weibel 2002: 598), shows the artist in a gallery between one of his paintings
and two empty frames. In addition to being self-referential in the sense that every
self-portrait is a self-referential message, the photo evinces a different kind of
self-reference in so far as it was torn into two pieces by the artist himself, who
The death of photography in self-reference 101

put the fragments once more together to exhibit the fragmentary result as a
new deconstructed whole. The photo is a photo and a metaphoto at the same
time. On the one hand, it is a portrait of the artist, on the other hand, it conveys
the message: This is a photo torn to pieces and reassembled once more. The
paradox consists in two contradictory messages at the metalevel, which can be
paraphrased as: This is a photo and not a photo at the same time. It is the
destruction of the photo by the artist which conveys the message This is no
more a photo (but scraps of paper), whereas his gesture of putting it together
once more conveys the opposite message: This is a photo. The loss of the
referent occurs at the metalevel. It is the loss of the referent which results from
the destruction of the photo, since a fragmented photo is impeded from fullling
its function of indicating its referent.

3.6. Mise en abyme

The sixth type of the death of the referent appears in the strategy of self-
referential photos in the photo (mise en abyme). The semiotic implications are
similar to the ones discussed in the case of double exposure. Consider the ex-
ample of the photo entitled Kassel is everywhere or: Where am I? of the
Arbeitsgruppe Fotoforum of 1979 (see Heyne 2003: 53). A giant photo of a
street is inserted in the picture of the same street taken from the same point of
view. The method of insertion is not double exposure, but photographic self-
depiction. The group of artists appears in the photo carrying their giant photo
of the same street. Everything except for the artists is represented twice, which
constitutes genuine iconic self-reference. At the same time, the photo inserted
by the artists in the foreground is hiding part of the background scenery. It ob-
structs the view of the street right in the middle of the picture. However, the
resulting loss of the background referent can be recuperated from the scenery
represented by the photo in the foreground which is the picture of what is in-
visible. Several paradoxes are involved. The photo is a photo, but at the same
time it is two photos. The photo makes visible and invisible at the same time.
Strangely, the two photos in one have one and the same referent. In their double
reference, the two photos are indexical signs of their referential objects, in their
mutual sameness, they are iconic and self-referential signs.
102 Winfried Noth

3.7. Self-obliteration

Photo-collage, photomontage, and intermedially hybrid photographs are para-

photographic genres which extend the method of photographic deconstruction
considered above in the example of Arnulf Rainers self-portrait. While Rainers
destruction of his own photo was only a partial one after all, the pictorial
loss by the tear in his picture is only a small impediment to our eye Timm
Ullrichs went all the way in the destruction of his own self-portrait with his
Self-effacement by painting [Selbstausloschung durch Malerei] of 1973
1976 (see Weibel 2002: 618). The technique by which this work was produced
is described as heavy paint-overpainting of a sheet of glass as a sequence in
ten photos laminated onto cardboard (Weibel 2002: 618).
In the top left corner we see the artists complete photographic self-portrait
behind a window-like sheet of glass. In the sequence of the ten pictures from
left to right and top to bottom, Ullrichss portrait gradually disappears while the
artist is shown in a sequence of photos in which he is overpainting the sheet
of glass between himself and the camera with white paint, which makes him
gradually disappear. The background behind the glass shows the artist in the
different stages of self-effacement. Only the rst picture is a full portrait. From
the second to the last but one frame only parts of his body can be discerned
while his body gradually disappears behind the painted portions of the glass.
In the tenth picture, the portrait becomes altogether invisible. It is no longer a
photo of the artist, but one of a sheet of glass painted white.
Ullrichss metapicture abounds with self-references and paradoxes. Each of
the ten pictures is a portrait of the artist, even the last one where the artist is
invisible because he disappears behind the coat of white paint, which creates the
paradox of a portrait of an invisible man. In the end, the self-referential gesture
of effacing himself and hence making the referent of the photographic portrait
disappear leaves us to solve the paradoxical doubt whether the last picture is a
photo of a monochrome painting or a deleted photo, hence the mere negation of
a photo? Furthermore another major paradox all the while is: How can an artist
go on to produce a self-portrait while he himself disappears in the process of
doing so?

4. Postphotographic pictures without a referent

Despite the various forms of the disappearance of the referent in its seven self-
referential forms of death distinguished so far, traditional photography always
left some indexical traces of the referents which it depicted. The shadows which
The death of photography in self-reference 103

we perceived as an abstract array of black and white stripes were really the
emanation of sunrays in or of a porch captured by the lens of a camera, and the
artists disappearance behind a white sheet of glass, left us not only with the
indexical picture of this glass, but also with the possibility of reconstructing the
artists image behind the glass by a process of visual inference, which leads us
back to the rst photo in the sequence, where we saw him fully.
Only in digital photography have photos entirely devoid of indexical an-
chors in the visual world become possible. Being programmatically without any
referent, they have become self-referential right from the beginning. Photos of
this trend in postphotography did not lose their referent, they never had one.
In contrast to Abstract Photography, which used to abstract from the referent,
they have become Concrete Photography, which creates its own images without
abstracting (cf. Jager 2003: 178). In a manifesto about his own postphotography,
which can be exemplied with his black-and-white photographs on canvas of
1993 and 1995 entitled Sources (see Selichar 2003: 270), the Austrian digi-
tal photographer Gunther Selichar declares: These images look like paintings,
speak about new media and are photographs: reproductions of a grammar of
media, located on the paper-thin boundary between the gurative and the ab-
stract, which ultimately shows that categories like this have perhaps become
obsolete (2003 268).

5. Nonrepresentational photography as self-referential genuine


From a semiotic perspective, the question remains to be answered whether such

pictures without a referent can still be considered as signs. Does not the notion
of sign require the correlate of a referent, and if so, can semiotics still contribute
to the study of such signs without a referent? Is the concept of the nonreferential
sign a contradiction in terms?
In more detail, I have given answers to these questions from the perspective
of Peirces semiotics elsewhere (Noth 2005). In a very brief summary, signs do
not require a referent in the sense of an object in the real world. Peirce does not
even use the concept of the referent. The term that he uses instead of referent
is the object of the sign, and the object of the sign can be something real,
but also a mere thought, a concept, or an idea. Even in concrete photographs
without any indexical anchors in the real world, the pictures are signs. Their
objects are our previous experience of the visual world, of the forms, colors, and
textures with which we have become acquainted in our pictorial culture.
104 Winfried Noth

The principal semiotic innovation of the postphotographic era is not the dis-
appearance of the object of the photographic sign, but the shift from indexical
to genuinely iconic pictures. The self-referential photos of Concrete Photogra-
phy, whose visual message is in their formal design only, are genuine icons,
according to Peirce (Noth 2003). A genuine icon does not depict in the tradi-
tional sense of mimesis. It refers to nothing but its own simple visual qualities
of form, luminosity, contrast, or texture. The genuinely iconic sign constitutes
a kind of degree zero of semioticity since it is reduced to the category of rst-
ness, the mode of being of that which is such as it is, positively and without
reference to anything else (CP 8.328). Such an icon is a sign merely by virtue
of qualities of its own, and since it is not yet distinguished from its object, it
does not refer to or stand for it at all. Peirce says that the genuine icon does
not draw any distinction between itself and its object since it is a sign by virtue
of its own particular qualities: [Genuine] icons are so completely substituted
for their objects as hardly to be distinguished from them. [. . . ] The distinction
of the real and the copy disappears, and it is for the moment a pure dream not
any particular existence, and yet not general (CP 5.74, 4.447).

6. Conclusion

The death of photography began, and ran parallel, with the crisis of represen-
tation. It begins with the emergence of pictures which are losing their indexical
reference to an object in the visual world. There are several forms and modalities
of this death of traditional photography. Some are due to alloreference, some to
self-reference. Self-reference in photos rst meant a partial loss of the referent,
in digital photography we are faced with photos that have never had a referent.
We identied seven kinds of such losses of the referent in the history of pre-
digital photography. The postphotographic era is not the end of photography,
but the beginning of a new postphotographic photography. Pictures in Con-
crete Photography are signs without a referent in the real world, but not signs
without an object in the sense of Peirces denition of the sign. Although these
photographs represent nothing, they are nevertheless signs. Whereas the now-
defunct classical photography produced indexical signs, the postphotographic
art of Concrete Photography is a self-referential art of genuine icons.
The death of photography in self-reference 105


Acton, David
2004 Photography at theWorcesterArt Museum: Keeping Shadows. Worces-
ter, MA: Worcester Art Museum and Ghent: Snoeck.
Arrouye, Jean
1978 Semio-photo ou la mort de lanalogie. Critique 368: 7287.
Barthes, Roland
1980 La chambre claire: note sur la photographie. Paris: Cahier du cinema.
Transl. 1982. Camera lucida: Reections on Photography. New York:
Beckley, Bill and Katherine Aguilar (eds.)
2007 The Death of Photography And Other Modern Fables on the Visual
Arts. New York: Delano Greenidge.
Carani, Marie
1999 Au dela de la photo positiviste: de la photo post-moderne a la post-
photographie. Visio 4(1): 6791.
Govignon, Brigitte (ed.)
2004 La petite encyclopedie de la photographie. Paris: Matiniere. Transl.
2005. Kleine Enzyklopadie der Photographie. Munich: Knesebeck.
Grundberg, Andy
2007 Photography in the age of electronic simulation. In: Bill Beckley and
Katherine Aguilar (eds.),The Death of Photography, 6168. NewYork:
Delano Greenidge.
Heyne, Renate (ed.)
2003 Kunst und Fotograe: Floris Neususs und die Kasseler Schule. Mar-
burg: Jonas.
Jager, Gottfried
2003 Abstract photography. In: Ruth Horak (ed.), Rethinking Photogra-
phy I+II: Narration and New Reduction in Photography, 162195.
Salzburg: Fotohof Edition.
Mitchell, William J.
1992 The Recongured Eye: Visual Truth in the Post-Photographic Era.
Cambridge, MA: MIT Press.
Noth, Winfried
2000 Handbuch der Semiotik. 2nd rev. ed. Stuttgart: Metzler.
2003 Photography between reference and self-reference. In: Ruth Horak
(ed.), Rethinking Photography I+II: Narration and New Reduction in
Photography, 2239. Salzburg: Fotohof Edition.
2005 Warum Bilder Zeichen sind. In: Stefan Majetschak (ed.), Bild-Zeichen:
Perspektiven einer Wissenschaft vom Bild, 4961. Munich: Fink.
106 Winfried Noth

Noth, Winfried and Christina Ljungberg (eds.)

2003 The Crisis of Representation: Semiotic Foundations and Manifesta-
tions in Culture and the Media. (= Special Issue of Semiotica 143(14).
Peirce, Charles Sanders
19311958 Collected Papers, vols. 16, ed. Charles Hartshorne and Paul Weiss,
vols. 78, ed., A. W. Burks. Cambridge, MA: Harvard University
Press. Quoted as CP.
Ritchin, Fred
1990 In Our Own Image: The Coming Revolution in Photography. New
York: Aperture.
Robins, Kevin
2007 Why images move us still. In: Bill Beckley and Katherine Aguilar
(eds.), The Death of Photography, 2144. New York: Delano Greenid-
Santaella, Lucia
1997 The prephotographic, the photographic, and the postphotographic im-
age. In: Winfried Noth (ed.), Semiotics of the Media, 121132. Berlin:
Mouton de Gruyter.
Santaella, Lucia and Winfried Noth
1998 Imagem: Cognicao, semiotica, mdia. Sao Paulo: Iluminuras. Trad.
Roque Graciano. 2003. Imagen: Comunicacion, semiotica y medios.
Kassel: Reichenberger.
Selichar, Gunther
2003 Photography remixed. In: Ruth Horak (ed.), Rethinking Photogra-
phy I+II: Narration and New Reduction in Photography, 248287.
Salzburg: Fotohof Edition.
Weibel, Peter
2002 An end to the end of art? On the iconoclasm of modern art. In: Bruno
Latour and Peter Weibel (eds.), Iconoclash, 570684. Cambridge, MA:
MIT Press.
Wetzel, Michael
2003 Artem celare The cryptic mediality of the photographic. In: Ruth Ho-
rak (ed.), Rethinking Photography I+II: Narration and New Reduction
in Photography, 5881. Salzburg: Fotohof Edition.
Marilyn: A paragone of the camera gaze

Kay Kirchmann

In his study The Reality of the Mass Media, sociologist Niklas Luhmann asserts
a dual status of reality in the contemporary mass media. The rst reality of the
mass media, in Luhmanns words its real reality (Luhmann 1996: 12), consists
of its own operations, that is in the fact that there is printing, broadcasting, and
reading, that programs are received, movies watched. All this is permeated and
framed by pretexts and recursive elements, by countless communications of
preparation and talking about it afterwards (Luhmann 1996: 13). This is why,
according to Luhmann, the real reality of the mass media has to be understood as
the communications that exist within it and run through it (Luhmann 1996: 13).
From the point of view of systems theory, something else is far more important
for the operative functionality of the media:
One can also talk of a second meaning of the reality of the mass media,
namely in the sense of what passes as reality for it and through it for others.
[. . . ] This meaning implies that the actions of the mass media are not simply
considered to be a sequence of operations, but a sequence of observations.
[. . . ] In order to reach this understanding of the mass media, we have to
observe it observing. For the meaning rst introduced, an observation of the
rst order is sufcient, as if it were all about facts. For the second option of
understanding, one has to take up the position of an observer of the second
order an observer of observers. (Luhmann 1996: 1415)
However, it is well known that in his own theoretical work, Luhmann has
kept a critical distance from empiricism, and in a very general sense his concept
of observation only implies that distinctions are made according to the binary
codiers (+/) generated in a system. His concept therefore cannot be equated
with the act of (visual) perception as such (cf. Baraldi, Corsi, and Esposito 1997:
124). But in the context of (audio-)visual media observation, this term will in-
evitably take on this second meaning as well, since the differentiations at hand
are also (though not exclusively) articulated in visual form. And it is only in this
form that they, in turn, are accessible to us as observers of the second order
since, according to another famous distinction of Luhmanns (form/medium),
the medium itself is not accessible, it is only perceptible through the forms it
108 Kay Kirchmann

generates (cf. Luhmann 1997: 195215). Accordingly, to observe the obser-

vations of (audio-)visual media means quite literally to observe (audio-)visual
forms; forms in which the distinctive systemic operations of the respective mass
media are condensed into objects of perception (Baraldi, Corsi, and Esposito
1997: 58).
What generally interests me in the following is not so much the theoreti-
cal edice of systems theory as such: I am specically interested in the gure
of a medial function of observation implicit in it and this only in the sense
of a heuristic base category; a methodological starting point for concrete in-
quiries in the eld of media analysis. In the context of a number of current
phenomena of media reality (in the sense introduced above), the mode of me-
dial self -observation appears to me particularly signicant. Luhmann himself
has conceded with reference to the base dichotomies of system and environ-
ment, of external- and self-reference (cf. Baraldi, Corsi, and Esposito 1997:
128) that as a system, the mass media can observe itself, albeit only through
quite complex operations:

A special case of self-observation presents itself when the observation is

itself an operation of the observing system and takes part in its autopoiesis.
By autopoiesis we do not mean an operation which observes itself while
observing (something that is impossible), but rather an operation that ob-
serves something of which it is also a part (another operation of the system
of which it partakes). (Baraldi, Corsi, and Esposito 1997: 127)

Even if the epistemological problems that come with such an operation are
ignored for a moment, it remains worth considering whether the mass media
can be taken so collectively as the subject, instrument, and object of such a self-
observation. After all, such an observation can only take place in one medium
and through its instruments. In view of the multiplicity of contemporary media,
it is highly unlikely that the self-observation of the medial realm could extend to
all media including itself. Hence one should rather speak of a gure of medial
media-observation that is, rst, necessarily selective and, second, oscillating be-
tween external- and self-observation. And, nally, one would have to investigate
whether, as Luhmann thinks, such medial self-observation has already reached
its purpose in causing a system to become auto-dynamic (cf. Baraldi, Corsi, and
Esposito 1997: 128), or whether we can nd entirely different intentions and
I would like to consider these questions by looking at a paradigmatic exam-
ple: an approximately six-minute-long feature broadcast by the French-German
television station ARTE in 1999 as part of the series Les cent photos du siecle /
One Hundred Photographs of the Century (cf. Robin 1999). The series belongs
Marilyn: A paragone of the camera gaze 109

to a new sub-genre, the retrospective view of the century (cf. Filk and Kirch-
mann 2000), which surfaced on diverse TV channels in the wake of the recent
turn of the millennium. Other than most series in this format, the ARTE series
approached history from the outset in a way that emphasized and reected on the
comprehensive saturation of our cultural memory by the media. Accordingly,
the series did not even attempt to reconstruct a past that would be accessible
without any recourse to media. Instead, it focused exclusively on historys media
documents; in this case, famous photographs. While individual episodes center
on the history of production of individual photographs, the photograph itself is
elevated to an object in which individual biographies randomly intersect at a
particular moment in time, namely those of the photographers and of the per-
son photographed. Thereby history is interpreted as a contingent eld of eeting
events and of meetings that receive historic prominence only in retrospect. From
a conceptual point of view, One Hundred Photographs of the Century draws on
a wide denition of historical relevancy according to which the history of pop-
ular culture and therefore also the mass media is a constitutive element of
history as well. For this very reason the series also includes an episode entitled
Marilyn 1960 that deals with the photograph shown in Figure 1.

Figure 1. The central photograph in the episode Marilyn 1960: Eve Arnolds photo-
graph of Marilyn Monroe on the set of The Mists
110 Kay Kirchmann

Depending on a viewers pre-existing knowledge, this photograph, with

which the episode opens, will call up different pretexts: all viewers will probably
recognize Marilyn Monroe as the person depicted, which in turn will conjure up
the much-publicized biography of the actress her problematic personal life,
her mysterious suicide, etc. A large section of the audience will also recognize
in the mystifying apparatus on the left margin of the picture a boom, or they will
at the very least associate it with cinematic recording technology that is to
say, they will properly place the photograph into the context of lm production.
Veritable lm buffs, on the other hand, might immediately recognize the land-
scape and identify it as the setting of The Mists, a John Huston picture Monroe
shot in the Nevada desert in 1960. And they might furthermore remember that
the lms legendary status in the history of cinema partly derives from the fact
that it was Marilyn Monroes and Clark Gables last lm; that Marilyn Monroes
then-husband, dramatist Arthur Miller, had written the lm script for her as a
birthday present in a nal attempt to secure for her the serious role she des-
perately desired; that the marriage of the two had nevertheless suffered its nal
break-down during lming and that they shortly after divorced; that despite all
the problems and personal crises on the set, the lm would go on to win multiple
Academy Awards, etc. (cf. Miller and Toubiana 2000: 4996). Those who are
also well-versed in the history of aesthetically ambitious press photography will
furthermore know that the rights for all set photography on this lm were given
exclusively to the famous Magnum agency, which had dispatched some of its
most highly-regarded members to the lm set (cf. Miller and Toubiana 2000:
6774), including Henri Cartier-Bresson, Ernst Haas, Cornell Capa, and, last
but not least, Eve Arnold, who is credited with having taken the picture under
consideration here.
The pre-knowledge called up by this opening is thus layered, and the layers
correlate with the complexity of the attendant gure of observation: in this case,
the medium television observes the medium photography as it observes a lm
production (in retrospect, a legendary one) and a celebrated lm star who in
turn is one of the most photographed people of the 20th century and we nd
ourselves in the position of trying to observe this almost innite regression.
After the opener, the television episode deals with the multiple preconditions
of the history of the photograph and the biographies of the protagonists tied
to it, as it were, in fast forward mode. It dedicates a mere 45 seconds to the
genesis of The Mists, almost a full minute to the biography of Monroe, and
20 seconds to the importance of the Magnum agency in the history of photo-
graphy. Clearly, the episode is not so much concerned with lling in substantial
gaps in the pre-knowledge of its audience; it rather attempts to activate this
pre-knowledge once more. In a marked difference to the other episodes of the
Marilyn: A paragone of the camera gaze 111

series, this section does not set out to provide a contemporary reassessment of
signicant historical fragments included in this photography. This episode of
One Hundred Photographs of the Century makes something entirely different
the object of its observation of medial observations, and it is left to Eve Arnold,
by now almost 90 years old, to provide the sound bites for this deeper theme:
Marilyn loved the photo camera just as she loathed lm cameras. When
she worked for the cinema, she had to put on make-up, prepare, learn her
lines, be on time, and, above all, she had to listen to people telling her
what to do. During a photo shoot, it was she who controlled everything.
Her true profession was to have her picture taken. Photography was her
toy. It allowed her to present herself as she really was. She knew without
prompting to affect the appropriate look and how to achieve the result she
wanted. She knew how to move her hands, how to throw back her hair, and
how to make use of her famous pout, which made her appear even more
sensuous. She was extremely gifted. She not only knew how to take in the
photo camera but also, and in particular, the photographer. (Transcript of
the TV program by the author)

Quite obviously, this medial media observation serves the purpose of de-
veloping a discourse about different kinds of camera gazes and therefore of
different visual media, a discourse that clearly favors photography. Television
here contends that photography by contrast with the lm camera has passed
down the correct gaze at Monroe. I will specify in more detail below how this
theory is developed. For now, I would like to stress that the critical comparison of
different camera gazes stays within a classical discursive tradition of European
cultural history: the paragone. From the Renaissance to Lessings Laokoon, the
paragone (literally a critical comparison, often the result of an argumentative
dispute, Reck 1992: 120) served as a means to divide up spheres of inuence
between the arts themselves and between the arts and other social sub-systems,
such as the sciences. Studying the aesthetic order of the semiotic material unique
to them was an attempt to establish the distinctiveness of individual art forms.
While, according to art historian Hans-Ulrich Reck, the historical paragone in
the long run created an awareness that the arts do not all deal with the same ma-
terial or have the same meaning (Reck 1992: 120), the paragone that resurfaces
here is quite at odds with this option of shared labor within different, co-existing
media. In the so-called media age, every object has to be potentially accessible
and adequate to every medium. In the face of a highly competitive market and
the bully mentality that comes with it, any notion of a co-existence or a careful
discrimination of objects have long been superseded by claims for omnipotence
and plain competitiveness of individual media forms. Marilyn Monroe, and in
this respect the thematic choice of the ARTE short feature is anything but ran-
112 Kay Kirchmann

dom, is herself a highly signicant example because already in her own lifetime,
she was a medial object embroiled in competitive struggles between photogra-
phy, the print media, cinema, and albeit in a weakened form television of
the time.
Against this background, it is only logical that the paragone discourse of our
example shifts from a semiotic onto an ontological level. We are no longer deal-
ing with the traditional question of which medial sign system is the proper one
for Monroe, we rather have to ask which amongst the camera gazes (all of them
potentially relevant to the object in question) will be able to show us the true
Monroe. The respective verdict of the ARTE episode in favor of the photo cam-
era and against the lm camera draws implicitly on the popular understanding
that behind Monroes image on the screen hides a genuine, fragile, and lonely
woman; a woman even who was victimized by being continuously typecast by
a ruthless studio system as a blonde sex goddess, etc. However, this argument
could just as easily be reversed particularly in light of the pin-up photographs
which had brought Marilyn Monroe fame even before her lm career started (cf.
Arnold 2005: 2436). As observers of this medial media observation, however,
we cannot simply focus on a question that must necessarily remain speculative,
the question about the ultimate truth behind Marilyn Monroe. Rather, we must
focus exclusively on the discursive gure as developed by television in her spe-
cic case. Again, having been trained in Luhmanns constructivism, the question
whether mass media products refer to reality can of course never refer to an on-
tological, existing, objectively accessible, unconstructively recognizable reality
(Luhmann 1996: 20). It can only be: How do the mass media create reality?
The construction affected by this TV program works in part through a de-
liberate reference to the pertinent position that Eve Arnold takes. By virtue of
its rhetorical structure, Arnolds above quoted statement constantly brings into
play the opposition between photo and lm camera and joins it with an onto-
logical question. Let us consider the quote once more with a special focus on
its rhetorical devices (marked by italics):

Marilyn loved the photo camera just as she loathed lm cameras. When
she worked for the cinema, she had to put on make-up, prepare, learn her
lines, be on time, and, above all, she had to listen to people telling her
what to do. During a photo shoot, it was she who controlled everything.
Her true profession was to have her picture taken. Photography was her
toy. It allowed her to present herself as she really was. She knew without
prompting to affect the appropriate look and how to achieve the result she
wanted. She knew how to move her hands, how to throw back her hair, and
how to make use of her famous pout, which made her appear even more
sensuous. She was extremely gifted. She not only knew how to take in the
Marilyn: A paragone of the camera gaze 113

photo camera but also, and in particular, the photographer. (Transcript of

the TV program by the author; emphasis added.)

These sound bites by Eve Arnold only paraphrase again what she reiterated
before and after in the records of her work with Marilyn Monroe: When you
photographed her, she controlled and manipulated the whole set: me, the cam-
era. . . She knew her way around cameras and enticed reactions from it like I have
seen no other person do. In this set-up, she got whatever she wanted simply be-
cause the pressure of shooting a lm, which would threaten to bury her, was
missing (Miller and Toubiana 2000: 71). Similarly, in another passage: Having
her picture taken was a safe way for her of being loved and admired (Arnold
2005: 137). Arnold even calls upon her colleague at the Magnum agency, Inge
Morath, who was also on the set of The Mists, to bear witness once more to
Marilyns entanglement with the two kinds of cameras: She was in charge of
the still camera she was the animal tamer, the photographer was the beast.
She fought constantly with lm cameras, but in front of the photo camera, she
was free (Arnold 2005: 72). Because of the amount of printed and reprinted
statements, the producers at ARTE presumably had to have been very aware of
what they could expect Eve Arnold to say about Marilyn and photo cameras,
and this awareness must have inuenced their choice in making these specic
photos and this specic photographer the focal point of this episode.
Both in written statements and in the sound bites taken from the TV pro-
gram under consideration here, the main argument always stays the same: Eve
Arnold shifts the question onto an ontological level, and her comments locate
the functional relation between visual image and visual object in a decidedly
anthropomorphic dimension. In this dimension, cameras and actress interact
against a backdrop of personalized, inter-human and thus de-medialized and
de-technicalized patterns of action. A visual object is infused with the power
to love or to hate cameras, to control them or turn them into toys, to manipulate
them or force them into an unexpected reaction and vice versa. In the process,
the bond between Marilyn and the camera carries all the traits of an eroticized
relationship; one in which the camera surrenders voluntarily and passionately
to the actresss art of seduction and manipulation. The stylized erotic pas de
deux introduces a moment of secret complicity between Marilyn, the camera,
and the photographer which neatly balances what has been conceded before
that Marilyn has control over the camera and the photographer. This complicity
in turn allows Monroe to grant the photo camera and only the photo camera
what seems to be a privileged glance into her innermost self. She does so by
voluntarily choosing to concede control in order to display her true nature
to the camera. It is this claim of a reciprocal willingness to be seduced that
114 Kay Kirchmann

legitimizes a specic photographic gaze on a companion and friend a gaze

that seems not to be entangled in hierarchies and is therefore as appropriate as
it is intimate.
While making comments to this effect on Marilyns predilection for the photo,
which the Magnum-photographer took in the Nevada desert, Eve Arnold also
supplies pertinent instructions on how to make sense of it:

This photo she [Marilyn] loved especially since it transmitted a sense of

loneliness. [. . . ] When she is putting her hands to her face like this, you
have a feeling that she cries for help. For me, the main point this photo is
making is a cry for help. You can think she cries Help me, Im in danger!
(Transcript of the TV program by the author)

The TV program leaves little doubt that this danger does indeed originate
from the medium lm. While the bond between Marilyn and the photo camera is
characterized, as we have seen, by a sense of passion and complicity, the actresss
interactions with a lm camera are dominated by such aspects as force, control,
and a lack of authenticity. These interactions cannot be characterized as a recip-
rocal bond between lovers: instead, they feature as the frightening suppression
of a young woman who in her daily life is subjected to an uncomfortable but sen-
sible arrangement in short: we are witnessing the passion play of married life.
What we can observe is a calculated transformation of a photograph into a nar-
rative, complete with the established narrative topoi: the classic melodramatic
narrative involving the plight of a blonde, beautiful, and suffering heroine who
is caught between an understanding lover and a husband who tries to dominate
her, between an erotic liaison that offers her fulllment and the silent suffering
of married life constantly regulated by the intricate mechanisms of patriarchal
domination. Since there is no way out of this conict, the only logical result
seems to be the untimely death of the woman.
This narrative frame changes our perception of Arnolds photo and structures
our reading of it. The program cuts to the respective photo again and again and
favors, by means of selection and camera movement, the established pattern
of meaning. This dramaturgy of repetition seems to demonstrate in actu, as it
were, the change in perception mentioned above: the boom suddenly seems to
turn into a frightening, almost animalistic instrument; a pensive and attentive
Marilyn changes into a desperate one, the bizarre beauty of the desert landscape
metamorphoses into a gloomy site of existential struggle.
In retrospect, it becomes apparent just how much the selection of this photo
(and this photographer) was owed to the cold rationale of dramaturgy. If this had
been just about Marilyns life, other and possibly even more famous photographs
would have been more than tting in the context of a short feature for example,
Marilyn: A paragone of the camera gaze 115
116 Kay Kirchmann
Marilyn: A paragone of the camera gaze 117

Figure 2af. Six photographs by Eve Arnold taken on the set of The Mists
118 Kay Kirchmann

one from the legendary last session with George Barris shortly before Monroes
death, or even the photo that Elliot Erwitt took on the set of Billy Wilders
The Seven Year Itch which has since been reprinted countless times: Marilyn
standing on a ventilation shaft, her white dress blown up into the air. But only
in the photographs taken by Eve Arnold on the set of The Mists was and is the
paragone already implied which so obviously interested theARTE programmers.
Therefore, the editorial decision in favor of the pictures of this photographer was
anything but arbitrary. If you look at further pictures by Eve Arnold taken during
the 1960 production (Figures 2af), the repetitive pattern of staging becomes
fully obvious. These, by the way, are images that were not shown in the episode,
probably so as not to subtly qualify the uniqueness claimed for the relevant
Whether Eve Arnold takes photos of Marilyn Monroe preparing for her role
and studying her text (Figure 2b and 2f) or whether she takes her picture as she
listens to the instructions of the director, John Huston (Figure 2a) Marilyn
is always portrayed as suffering because she is caught up in the wheels of lm
production (Figures 2a and 2e). Each element of the production, the battery
of lights (Figure 2c) just as much as the lm camera (Figures 2a), metonymi-
cally functions as pars pro toto for a larger apparatus whose main message
spells force. Arnold stages this apparatus with her photo camera, for example
through her preference for mild low-angle shots, as a monstrous giant, complete
with all threatening connotations. The same applies in no small measure to the
boom (Figures 1 and 2d), which we have seen in the photo and which in itself
does not appear to be particularly threatening. However, Arnold develops an
obsessive-compulsive predilection for this specic combination of motifs. The
faint associations it carries of the Sword of Damocles or an executioners axe
may have motivated her choice. The fact that the pattern of motifs also very
subtly plays with phallic connotations, however, adds yet another nuance to the
sexual subtext of this discourse.
In this respect, the TV program merely seems to voice once more the inherent
structural feature and main topic of Eve Arnolds photography. The motive for
embracing Arnolds paragone in such an uncritical fashion is, of course, far less
altruistic. The gure of discourse is tied up with a well-established technique of
reciprocal medial observations which for Siegfried J. Schmidt are a constitutive
feature of the increasing complexity of highly distinctive media systems: So-
cieties that contain complex and open media systems expand dramatically their
(partial) observability. Media observe everything and everywhere, they observe
the fact that they observe, and they observe themselves observing (Schmidt
1998: 68). These observations remain partial in that the objects of their gaze
are usually only other media which is not just owed to the basic epistemo-
Marilyn: A paragone of the camera gaze 119

logical aporias of self-observation mentioned above. Another reason is the fact

that such observations are never without purpose. In fact, such operations lend
an air of objectivity to a medium that observes another medium. Therefore,
the medial self-observation of (or within) a media system is always a calcu-
lated strategy to achieve believability (cf. Todorow, Grampp, and Ruhe 2004:
202226). In a society suspicious of medial operations, it is designed to rein-
tegrate these suspicions into the media system and thus reestablish credibility
and legitimacy.
In the case of our object of observation, the strategy works as follows: Just as
TV readily observes the medium lm and the medium photography and offers
them to us for inspection, it takes care to avoid any form of self-observation; as
a medium, it thrives to remain invisible. Just as photo cameras and lm cameras
become a topic for discussion, the TV camera remains an anathema. Anthropo-
morphizing the cameras of other media means turning them into subjects and
thus subtly but permanently discrediting them as objective instances of world
observation. Conversely, TV presents itself as a medium that seems strangely
devoid of both camera and gaze as if there was no mediating, interpreting, and
selecting instance between visual objects and their representations; as if the TV
screen was a space of purely transparent manifestations. This way, television
establishes itself as a kind of super-medium, and, at the same time, as a non-
medium; it pretends to be a neutral space of the paragone when it is de facto its
silent partner and even its secret winner. The apparent alliance between tele-
vision and photography should not deceive us about the result of this paragone
because for television, photography is no serious competition: photography al-
ways needs, as our example has amply shown, another medium to put back into
motion and into a sequence the time it arrested in order to make readable and
discoursiable the fertile moment of a photo through a contextualizing narra-
tive. The medium lm could achieve all of this just as well as television, and
televisions real combatant in this paragone, the one that needs to be discredited,
is therefore lm. In an interview in another section of the episode, Eve Arnold
mentions, more or less in passing, that after the end of the production of The
Mists, it took her another ten years before she felt comfortable enough to watch
the lm in a cinema the reason, she claimed, was that she had the best images
of the lm memorized on her inner screen. In the context of the episode as a
whole, this statement may appear to be out of place, but it actually points towards
the unique background of the paragone. Apparently her statement is not about
the question of which medium can best transform photography into narrative.
Instead, she raises a much more general question of what is the appropriate
20th -century medium for remembering a pertinent comparative perspective
given the sub-format of this episode, a retrospective view at the century.
120 Kay Kirchmann

Again, the line of argument in our object under inspection is clear: while lm
seems to pose a threat to Eve Arnolds memory images, television seems to be
just the place to expand these individual memory images, make them speak and
thus transfer them into the collective memory. The conclusion implied here is
that television is the only medium capable of adequately storing contemporary
history and reecting on it (cf. Filk and Kirchmann 2000).
An explicit paragone like the one portrayed here may be a special case in the
spectrum of medial media observations, but it is by no means unique. Media-
referential genres and formats, currently in high demand in very many media,
do not occur in a vacuum and are never free of both purpose and aim. They
are symptoms of an increasingly competitive stance in a segment of the market
that is still very lucrative. It may not be a coincidence that all of this happens
at a time when the discourse on media attests to quite contrary tendencies.
While catchphrases like convergence, compatibility, and multimedia promise
a peaceful union of formerly discrete media (or at least their uncontested and
uncomplicated dissolution in the all-embracing binary code of the computer), the
individual media themselves seem less than willing to stand by and observe how
their dramaturgies and programs, genres, and modes of reception and perception,
all of them distinct and with their own histories, are simply smoothed over. The
opposite seems to be the case: they seem to insist on a re-evaluation of their
merits and competence, and it is no mere accident that they revive the tradition of
the paragone, and with it a discourse whose historical achievement of providing
levels of differentiation is perhaps too readily negated in the current state of
multimedia euphoria.

(Translated from German by Gerd Bayer and Christian Krug)


Arnold, Eve
2005 Marilyn Monroe: Eine Hommage von Eve Arnold. Munich: Schirmer/
Baraldi, Claudio, Giancarlo Corsi and Elena Esposito
1997 GLU: Glossar zu Niklas Luhmanns Theorie sozialer Systeme. Frank-
furt: Suhrkamp.
Filk, Christian and Kay Kirchmann
2000 Wie erinnerungsfahig ist das Fernsehen? Thesen zum Verhaltnis von
Geschichte, Medien und kulturellem Gedachtnis. Funkkorrespondenz
42: 39.
Marilyn: A paragone of the camera gaze 121

Luhmann, Niklas
1996 Die Realitat der Massenmedien. Opladen: Westdeutscher Verlag. [The
Reality of the Mass Media, translated by Kathleen Cross. Cambridge:
Polity, 2000.]
1997 Die Gesellschaft der Gesellschaft. Frankfurt: Suhrkamp.
Miller, Arthur and Serge Toubiana
2000 The Mists: Die Entstehungsgeschichte eines Films von Magnum-
Fotografen dokumentiert. Munich: Kehayoff.
Reck, Hans-Ulrich
1992 Der Streit der Kunstgattungen im Kontext der Entwicklung neuer Me-
dientechnologien. In: Klaus Peter Dencker (ed.), Interface 1: Elektro-
nische Medien und kunstlerische Kreativitat, 120133. Baden-Baden:
Robin, Marie-Monique
1999 Die Fotos des Jahrhunderts: Das Buch zur arte-Serie. Cologne: Ta-
Schmidt, Siegfried J.
1998 Medien. Die Kopplung von Kommunikation und Kognition. In: Sibylle
Kramer (ed.), Medien, Computer, Realitat: Wirklichkeitsvorstellungen
und neue Medien, 5572. Frankfurt: Suhrkamp.
Todorow, Almut, Sven Grampp and Bernd Schmid-Ruhe
2004 Medien unter Verdacht. Selbstreexivitat als Glaubwurdigkeitsstrate-
gie. In: Aleida Assmann, Ulrich Gaier and Gisela Trommsdorff (eds.),
Zwischen Literatur und Anthropologie: Diskurse,Medien, Performan-
zen, 202226. Tubingen: Gunter Narr.
Part IV. Self-referential lm
The self-reexive screen:
Outlines of a comprehensive model

Gloria Withalm

In the movies, reference from the lm to the lm itself is as old as the history of
lm. The device can be found in all times, in all lm genres, and at several levels
of cinematic communication. There are many forms, functions, devices, and
textual strategies of self-reference or self-reexivity, for example, the strategy
of creating an ironic or critical distance or even a sense of alienation (in Brechts
sense), the mere fascination with cinematographic possibilities, the device of
creating emotional bonds between the audience and the movies or movie stars,
the device of humor, or the attempt at attracting the attention of an audience
sated with watching the media.
In order to deal with the topic in a comprehensive way, a model will be pro-
posed to cover the entire range of self-reexive textual strategies and practices
and to relate them to the lm as a sociocultural and a sign system. A theory of
self-reference in the movies must provide a broader framework without restrict-
ing itself to the lm as a text. It requires an approach that is able to cover lm in
its entirety. A promising framework for such an endeavor can be derived from
the work of the Italian philosopher and sociosemiotician Ferruccio Rossi-Landi.

1. The sociosemiotics of Ferruccio Rossi-Landi

Ferruccio Rossi-Landis position in semiotics can best be characterized by the

term sociosemiotics. Rossi-Landis semiotics is rooted in a dialectic and materi-
alist philosophy, or, as he himself characterized it once, a materialist, philosoph-
ical, and anthropological approach. Although he never explicitly dealt with lm
or cinema, the concepts Rossi-Landi elaborated can be used as the foundation
for a semiotic reection on the phenomenon lm as a whole able to cope with
the dual character of lm as a semiotic and a sociocultural system.
Despite the wealth and complexity of Rossi-Landis semiotics, only two
aspects of his oeuvre can be briey dealt with, his concept of sign system and
his model of social reproduction.
126 Gloria Withalm

The key concept of Rossi-Landis semiotics is certainly work. Rossi-Landi

conceives of work in an anthropological sense and with regard to both material
and sign production. Thus, sign work is a special case of work as such. According
to Rossi-Landis denition, work involves six characteristics combinable to a
set of three which constitutes the fundamental triad of work: (1) the materials
on which work is performed, (2) the operations, comprising the worker, the
actual working operations, the instruments used, and the aims of work and
(3) the product. This basic triad can be interpreted in a dialectical sense with
the material functioning as thesis, the operation(s) as their antithesis, and the
product as the synthesis (cf. Rossi-Landi 1985: 13). When sign work is just
a special case of work in general, the product, which is the sign itself, is not
only the result of working operations but also a dialectic totality of a particular
kind, since the sign is a synthesis of signans and signatum1 and a mediation
between the material (in the usual sense of this term) and the social (Rossi-
Landi 1979: 3031).
In dialectic terminology, what happens when a sign is used is that a social
thesis is mediated by means of a material antithesis. The signans as
an antithesis has immobilized that social piece [pezzo soziale], and it has
brought it at a new level as a signatum [. . . ]. The antithesis is a piece of
matter which is used socially [. . . ] in order to identify the thesis as something
which had reached the social level already. [. . . ] The signans imposes its
own restricting power on the signatum as its opposite. The contradiction
between two different levels of organization is overcome, and the provisional
peace [pace provvisorio] of a synthesis is reached. This synthesis is the
social result that we call a sign. (Rossi-Landi 1979: 3031; cf. 1985: 165)
In his denition of sign systems, which is extremely fruitful with regard to
complex media such as the movies, Rossi-Landi goes beyond the usual descrip-
tion in terms of set and elements. In a single paragraph, he demonstrates that
the various concepts in his semiotics are not separate parts but interconnected
elements of one integrated theory of signs and society. The premise of his def-
inition is once again his concept of sign work with instruments and materials
performed by a worker according to rules. A sign system must not be reduced
to a mere code since it comprises also the semiosic context, the communicative
situation including all those who actually exchange the messages:
A sign system comprises at least one code, that is the materials which one
works and the instruments with which one works; but it comprises also
the rules to apply the latter to the former (the loci of the rules are two:
in a sense they are in the code, but even more so, they are inside the one
who uses them); it comprises the channels and the circumstances that make
communication possible, and furthermore, the senders and receivers who
The self-reexive screen: Outlines of a comprehensive model 127

make use of the code. A sign system includes furthermore all the messages
which are exchanged or could be exchanged in the universe instituted by
the system itself. (Rossi-Landi 1985: 242)

Rossi-Landi then introduces his key concept of social reproduction, ex-

pounding the two theses that there is no social reproduction without sign systems
and that no human sign system can exist outside social reproduction. In con-
clusion, Rossi-Landi argues that a sign system is a slice of social reality and
certainly not just a symbolic machine which stays there waiting (Rossi-Landi
1985: 242). In another context, Rossi-Landi expresses this sociosemiotic princi-
ple in an even more radical way: We must never raise the sign systems above the
reality of social reproduction. Sign systems are not a sort of skeleton of social
reproduction. Instead, they are social reproduction themselves (Rossi-Landi
1985: 144).
Social reproduction is a pivotal notion of sociosemiotics which refers to the
sum total of all processes by means of which a community or society survives,
grows, or, at least, continues to exist (Rossi-Landi 1985: 175). Rossi-Landi
describes several factors and processes that constitute social reproduction, but
only the most fundamental process, which will also serve as the basis of the
model developed in this paper, will be presented as it is summarized in the
Schema of Social Reproduction (Figure 1; Rossi-Landi 1975: 65, 1985: 38). It
is a model of the cycle of production, exchange, and consumption, which are
the three indissolubly correlated moments which social reproduction always
comprehends in a constitutive way (Rossi-Landi 1975: 65).

Figure 1. Rossi-Landis schema of social reproduction

In the middle of Rossi Landis model of social reproduction (Figure 1), ex-
change is shown as consisting of both external material exchange and sign
128 Gloria Withalm

exchange or communication, and the latter as comprehending a threefold sub-

division into the indissolubly correlated factors of sign production, sign ex-
change, and sign consumption.
In its printed form, Rossi-Landis schema presents a two-dimensional dia-
gram which presents triadic relations at two points. A representation of Rossi-
Landis model in graphic form requires by necessity a triadic graph, and it
is informative that the very term of triad occurs frequently throughout Rossi-
Landis writings on social reproduction. The three basic factors of this dialectical
triad are not only correlated but interrelated, they belong to the same totality,
one does not exist without the other [. . . ]. Their unity is dialectic (Rossi-Landi
1985: 180).2

2. Self-reference and self-reexivity: From shooting to showing

In the Preface to the second edition of his Reexivity in Film and Literature,
Robert Stam discusses the various terms related to the key concept of his book
as follows:
The broad notion of reexivity has generated a swirling galaxy of satellite
terms pointing to specic dimensions of reexivity. The terms associated with
reexivity belong to morphological families with prexes or roots deriving from
the auto family, the meta family, the reect family, the self family, and
the textuality family. (Stam 1992: xiv)
What strikes the reader most in the literature on the topic of self-reference
and self-reexivity is the plurality of notions used to cope with the various ways
a text can evince a relation to other texts, to the modes of text production, to the
genre, to the medium, to itself, to its own discourse, etc. No less astonishing are
the various relations constructed among the concepts adopted, either by strictly
excluding certain textual modes or by adopting an umbrella term under which
a network of different textual modes is subsumed.
The most frequently used terms in this context are: intertextuality, intratextu-
ality, and intermediality; self-reference or self-referentiality; self-reexivity or
auto-reexivity; self-conscious, self-begetting, or self-aware ction; metatextu-
ality, metaction; metalm, metacinema, metacinematographic; metareference;
metacodal or metacommunicative; foregrounding, the device of revealing; es-
trangement, deautomatization, or defamiliarization; mise-en-abyme, etc. In ad-
dition to this rst unordered and by far not exhaustive list, another list of related
terms describing particular modes of textual relations can be set up, such as
allusion, parody, pastiche, quote, etc.
The self-reexive screen: Outlines of a comprehensive model 129

Based on Rossi-Landis sociosemiotics, the model to be presented and dis-

cussed in the following tries to encompass all of these phenomena. As mentioned
above, lm, in the course of its history, has developed many ways of making
reference to lm. Film is characterized by its double nature of lm as a text
which is always and necessarily embedded in lm as a sociocultural (and eco-
nomic) system, and both aspects are the basis of self-referential and self-reexive
discourses and stories.
First of all, it has to be clear that any lm is subject to the fundamental cycle
of production, exchange (or distribution), and consumption (or reception) as
described by Rossi-Landi, both in the material (including the economic) and the
semiotic sense. Hence, the triadic graphic representation of this cycle suggested
above will be used as a model of this cycle (Figure 2a). The three indissolubly
correlated factors can actually describe the various phases in the life of a
lm from shooting to showing.

Figure 2a and b. Cycle of production, distribution, reception, and the product

As mentioned above, the movies fulll all criteria of Rossi-Landis deni-

tion of a complex sign system. This means that all materials and instruments,
all messages as the products of sign work which are produced, exchanged, and
consumed, received, or reproduced, and the entire communicative universe es-
tablished by the system itself belong to the sign system. Accordingly, the dis-
cussion of self-referential texts created by and within the sign system lm has
to combine the cycle of production, exchange, and consumption with the factor
of lm as a product of complex sign work (Figure 2b).
The combination of both models, the one of the fundamental cycle of pro-
duction, distribution, and consumption (Figure 2a) and the one representing the
product of sign work (and the sign system) (Figure 2b) lead to the model that
will be used in the following for the discussion of self-referential processes in
130 Gloria Withalm

Figure 3. Cycle of production, distribution, reception, and the product

lmic texts. This schema of lmic self-reference (Figure 3) represents all do-
mains, phases, or states in the overall cycle constituting the lm, and it covers
all forms of self-reference and self-reexivity in individual lms as they have
occurred since the beginning of lm history.
However, the actual modes of self-reference that can be found in the movies
are not conned to forms of lmic reference to the lm in general. In addition,
the model also takes into consideration a special case of lmic self-reference
which I would like to dene as self-reexivity. A self-reexive lm is a lm
which focuses or reects on itself, that is, on the specic lm that is being
watched. Various cinematic devices are used to draw the spectators attention to
the lm itself in this sense: lines of the dialog, the materialization of lmic
means, and in some less frequent cases, to the showing of the dispositif, the
technical device of lm production and lm showing.
Although self-reexivity concerns primarily the lm as a product or, more
precisely, one actual product it focuses on, it cannot be reduced to this factor
since self-reexivity can also concern the three stages of the cycle production,
distribution, reception. Hence, the entire cycle is reduplicated, as Figure 3 shows.
However, self-reective reference to aspects of these phases is always more spe-
cic than in lmic self-reference; it is always restricted to the lm under consid-
eration. The following sections will present examples of lmic self-reference
and self-reexivity concerning all stages of the sociosemiotic cycle of lmic
The self-reexive screen: Outlines of a comprehensive model 131

3. Production work and life in Tinseltown

The production of a lm in the world of the movies opens the entire cycle
(Figure 4a). This stage covers the institutions of production and the people
working in the movie business as well as the actual production which comprises
preproduction, shooting, and postproduction.

Figure 4a and b. Self-referential and self-reexive production

Making movies about movie making often labeled lm-in-lm or Hol-

lywood on Hollywood is probably the most popular and certainly the most
common, reference to lm and cinema. Already in 1899, the audience could
see what it looked like when a cameraman was shooting a parade of decorated
cars (Concours dautomobiles euris, Lumiere Brothers, FR 1899, n 1009) or
a group of people leaving a navy arsenal somewhere in Indochina (La sortie de
larsenal, Lumiere Brothers/Gabriel Veyre, FR 1899, n 1279). Not only do the
lms show the scene lmed by an invisible cameraman, they show a second cam-
eraman lming these scenes as well. Less than ten years later, the entire cycle of
lmmaking from script to screen was depicted in Making Motion Pictures:A Day
in the Vitagraph Studio (US 1908, Vitagraph). Soon, self-reference in the form
of a look behind the scenes became a matter of course, as in the 1910s slapstick
comedies mostly showing the havoc caused by a character on the set in the lm,
as Charlie Chaplin did in His New Job (Charles Chaplin, US 1915, Essanay).
Following the ups and downs in the lives and careers of actresses and actors
has always been a favorite pastime of movie fans. Accordingly, when it comes to
the lmic presentation of movie people, the rst and foremost genre is the one
that deals with biography, nicknamed biopic. From the 1910s on, biopics have
been part and parcel of Tinseltowns self-portrait. The genre is most suitable to
satisfy the fans desire for a closer look at the life of real or ctitious stars as
well as to promote a particular inside look manufactured by the industry itself.
132 Gloria Withalm

Among the most popular motifs is the beginning of the career of a lm star, the
rst steps into motion pictures in the Hollywood Cinderella story, but also
the downfall in a career due to personal problems or to the drastic changes in
the business, such as the radical changes from silent movies to the talkies. The
success story of a girl making it is in the center of the early one-reelerAVitagraph
Romance (US 1912; Vitagraph). Against the will of her father, a Senator, a young
woman elopes with the man she is in love with. He is an aspiring author, and
both nd work with Vitagraph Film Company. He starts to write screenplays,
and she soon becomes a leading actress. When her father nds out about her
movie career, he visits the Brooklyn studios, meets the companys heads (played
by the actual executives), and is reunited with his daughter.
Apart from the plot, this early example has several ingredients that reappear
in many other lm-in-lm movies: showing the studio premises (the actual
Vitagraph studios), cameo appearances of studio bosses (in this lm Albert E.
Smith, J. Stuart Blackton, and William T. Rock), and the tension between the star
and the role of the star, intensied by the spectators interest in the stars private
life. Clara Kimball Young, the female protagonist, was actually the daughter of
Edward Kimball who played her senator father in the lm. The latter form of
lmic self-reference characterizes one of the most famous examples of ctional
biopics, Sunset Blvd. the lm on the aging silent star Norma Desmond (Billy
Wilder, US 1950). The star is played by another silent star, Gloria Swanson, and
when playing bridge with old friends, her partners are the real life aging movie
stars Buster Keaton, Anna Q. Nilsson, and H. B. Warner.
Self-reference in biopics is not conned to ctional or real-life actresses and
actors. In addition to the many examples of lms on the life of stars, for example,
the lms on Jean Harlow, Marilyn Monroe, or Rudolph Valentino, there are also
movies focusing on members of the crew, for example, writers (Barton Fink, Joel
and Ethan Coen, US 1991; Adaptation, Spike Jonze, US 2003), set designers
(Good Morning, Babilonia, Paolo and Vittorio Taviani, FR/IT/US 1987), or
producers (The Bad and the Beautiful, Vincente Minnelli, US 1952; The Player,
Robert Altman, US 1992).
The second category of lms whose self-referential features pertain to the
production phase are those offering a glance behind the scenes of moviemak-
ing, allowing us to watch ctitious or real lm directors during their work.
Well-known examples of lms showing the actual shooting phase are Singin
in the Rain (Gene Kelly and Stanley Donen, US 1952), La nuit americaine
(Francois Truffaut, FR 1973), or The French Lieutenants Woman (Karel Reisz,
UK 1981). Other lms that are self-referential with respect to its production
focus on the phase of preproduction including the planning and writing of a
lm, the preparations before shooting (Otto e mezzo, Federico Fellini, IT 1963),
The self-reexive screen: Outlines of a comprehensive model 133

or the rst rehearsals with the actresses (Reperages, Michel Soutter, CH 1977).
Postproduction is in the focus of self-referential lms which deal with synchro-
nizing, as in Blow out (Brian de Palma, US 1981) or the actual editing and even
what can go wrong in this phase, as in Wenn die Filmkleberin gebummelt hat (G
1925).3 Finally, there are lms that ctionalize the shooting of really existing
lms, for example, Shadow of a Vampire (E. Elias Merhige, US 2000) on F. W.
Murnaus Nosferatu (G 1922) or also some scenes of Insignicance (Nicolas
Roeg, US 1985) which relate to Billy Wilders The Seven Year Itch (US 1955).
As discussed above, self-reexivity unfolds along the same phases inside
the overall cycle (Figure 3). Hence, there are also self-reexive lms focusing
on production aspects (Figure 4b). In contrast to self-referential lms which deal
with the production phase, self-reexive lms related to the same phase do not
merely present the production of just any movie, but the production of the very
lm which represents the production of this lm. One of several possibilities
of lming the production phase self-reexively is to show parts of the studio
and/or members of the crew, as is the case at the end of E la nave va (Federico
Fellini, IT 1983). The device is not restricted to feature lms. There are also
examples from the genre of video clips. Genesiss I Cant Dance (Jim Yukich,
US 1991) shows not only Phil Collins in the lmed story but also how he is
prepared for the shooting getting his face powdered and his hair combed. It
goes without saying that even if the crew members are real professionals, this
make-believe glance behind the scenes is staged just like the rest of the lm.
Another group of self-reexive lms related to the production phase has its
focus on the shooting camera. Against the tradition of the invisible camera in
the classical Hollywood style, these scenes draw attention to the circumstance
that it is due to a camera that we are able to see this scene and the entire movie,
as is the case in La tarea (Jaime Humberto Hermosillo, MX 1990). Examples of
lms showing the shooting camera are The Man with the Movie Camera (Dziga
Vertov, SU 1929), or Jane B. par Agnes V. (Agnes Varda, FR 1988). In Boogie
Nights (US 1997) Paul T. Anderson lets us literally look into the camera. There
are even scenes in which the shooting camera gets in direct physical contact,
or rather confrontation or collision, with the characters or other crew members.
There is a very early example of this plot element in How It Feels to Be Run Over
(Cecil M. Hepworth, GB 1900) in which the camera lming a car is apparently
run over by this automobile.
The climax of self-reexivity with regard to the production phase is a lm
shown in the process of its own production. The subject matter of these lms is to
show a lm in the process of its own production. Among the narrative strategies
is one that could be labeled When words turn into moving images. The movie
unfolds simultaneously with the telling of the story by a character, like in Never
134 Gloria Withalm

Give a Sucker an Even Break (Edward F. Cline, US 1941). A further device is

to show interventions in and changes of the lm in the phase of its production
made by the participating characters, as in Spaceballs (Mel Brooks, US 1987)
or Comicalamities (Pat Sullivan, US 1928, animator: Otto Messmer).

4. Distribution: Movies on the marketplace

The second phase of the sociosemiotic cycle of lmic reproduction concerns

the lm in its phase of distribution (Figure 5a and b). In relation to other phases
of the cycle, such as production or reception, the examples for the distributive
eld are relatively rare. Whereas the production aspect is easy to mystify by
means of a glance behind the scene strengthening the bonds between lms,
lm stars, and spectators, and whereas the reception phase, that is, going to
the movies, depicts an experience with which the audience is very familiar,
the distributive eld is the least known phase. Their marketing and promotion
strategies constitute a topic which the lm business is not too eager to disclose in
detail. The topics to be addressed include the institutions dealing with the actual
distribution, the diverse marketing and advertising campaigns related to movies
and lm stars, the documents necessary to promote a lm, such as trailers, lm
magazines, reviews, fanzines, posters, etc., and the various forms of critical
evaluation of the lm, from praise at a lm festival to censorship.

Figure 5a and b. Self-referential and self-reexive distribution

The earliest example of a lm focusing on its own documents is the lm

Les colleurs dafches (FR 1897, n 677) by the Lumiere Brothers. In this lm,
we see posters announcing the proper Cinematographe Lumiere being pasted on
top of other posters advertising the lms of a competitor called Cinematographe
Grand Four. As far as Hollywood marketing strategies are concerned, there are
The self-reexive screen: Outlines of a comprehensive model 135

lms in which Hollywood itself and its landmarks can be found at the center
of the activities of promotion agents. One of these landmarks which can be
seen rather frequently in Hollywood lms is Graumans Chinese Theatre with its
famous forecourt in which all the famous Hollywood stars have left their imprints
in cement. The rst and the last scenes of A Star Is Born (William A. Wellman,
US 1937), show this location. The domain of assessment and evaluation ranges
from festivals as in Cannes or Venice and great awards like the Oscar to
self-regulation and censorship.
There is also a self-reexive (Figure 5b) variant of the censorship motif in
combination with the motif of on-screen messages from the media institutions,
e.g., the movie theater management. In Hellzapoppin(H. C. Potter, US 1941),
for example, the possible use of four-letter words by the characters is stopped
with the projectionists verbal reference to the Hays Ofce and an inserted title
card that reads censored.
A further case of self-reexivity in the distributive stage is the on-screen
presentation of (real and ctitious) companies in different varieties. The Pathe
trademark, the rooster, applied on every set to prevent illegal copying, is an
example of an early and rather useful variant that began to appear in the 1900s.
Another strategy is to let the opening company logo, for instance the mountain
in the Paramount logo, segue right into the movie, as in the beginning of all three
Indiana Jones lms. The well-known brand images are also subject to parodies
as is the case with the MGM logo. In two very different, though equally self-
referential, television series, the roaring lion is replaced by a meowing kitten,
in Mary Tyler Moore Show (US-CBS 197776) and several episodes of the
Austrian cop comedy series Kottan ermittelt (Peter Patzak, ATORF 1982).
Finally, all the TV characters who talk about their own show and particular
features of their network belong to another subcategory of self-reexive lms.
A striking example is the 1980s ABC series Moonlighting with Cybil Shepherd
and Bruce Willis. The last episode even ends with the characters talking about
their show being cancelled.

5. Reception, or: A lm is shown

The third domain in which self-reference and self-reexivity can be found in

the sociosemiotic cycle of lm concerns the consumption or reception in the
world of the audience (Figure 6a and b). The stories focus on the one hand on
people watching a lm either in a cinema or at home and talking about movies,
and on the other hand, on the movie theaters as such, the people working there,
the showing of a lm.
136 Gloria Withalm

Figure 6a and b. Self-referential and self-reexive reception

Since the very beginning of lm history, lms have shown the audience. As
early as 1896, LEntree du cinematographe (Lumiere Brothers, n 250) shows
the crowd leaving Empire Theatre (on Londons Leicester Square) after a movie
show. Only ve years later, lms such as The Countryman and the Cinemato-
graph (Robert William Paul, UK 1901) or Uncle Josh at the Moving Picture
Show (Edward S. Porter, US 1902) let us step inside the movie theater and
watch the show in the show. The plots starting point is the depiction of the
strange behavior of spectators who are unable to distinguish real events from
those presented on the screen. Jean-Luc Godard quotes the scene in his Les
Carabiniers (FR/IT 1962).
In the second group of self-referential lms dealing with reception, there are
many lms which do not only depict movie theaters and people working there
but pay homage to the bygone days of cinema, as in The Last Picture Show
(Peter Bogdanovich, US 1971), Nuovo Cinema Paradiso (Giuseppe Tornatore,
IT 1989), Splendor (Ettore Scola, IT 1989), or The Majestic (Frank Darabont,
US 2001).
As to self-reexivity (Figure 6b), the audience related motif of addressing
oneself to the screen, like spectators applauding in a theater, is extended to spec-
tators actually communicating with the screen and with on-screen characters.
Such interactions culminate in the temporary dissolution of the barrier separat-
ing the spectators from the cast, the fourth wall, in the so-called screen passages.
The rst lm character who entered the screen in this way, although only in a
dream, was the projectionist in Sherlock, Jr. (Buster Keaton, US 1924). More
recent examples of such transgressions are Purple Rose of Cairo (Woody Allen,
US 1984), Last Action Hero (John McTiernan, US 1993), or Pleasantville (Gary
Ross, US 1998).
The self-reexive screen: Outlines of a comprehensive model 137

6. The lm as the product

The result of the sociosemiotic process of lmic work and sign work is the
lm as a product. It is situated at the core of the entire cycle of production.
Self-reference with regard to the product (Figure 7a) occurs in a relatively small
group of movies presenting lmic lm history as well as in its most frequent and
most conspicuous form constituted by the various procedures of intertextuality.
Examples of intertextuality are the play with lm genres, such as in parodies,
allusions to famous scenes, citations of music and dialog lines, or the actual and
material (or today digital) quotation of other lms.

Figure 7a and b. Self-referential and self-reexive aspects of the lm as a product

A subcategory of lms relating self-referentially to the lm as a product

makes a very particular use of the device of lmic citation. The scenes inserted
as a quotation are combined with the new scenes in a shot and reverse shot mon-
tage creating the illusion that the two characters separated in time are directly
interacting. Film examples are The Last Remake of Beau Geste (Marty Feldman,
US 1977) with Marty Feldman talking to Gary Cooper (William Wellman, US
1939) or Carl Reiners Dead Men Dont Wear Plaid (US 1982) using clips of al-
most twenty movies to show Steve Martin in interaction with almost everybody
once famous in the genre of lm noir of the 1940s, from Humphrey Bogart to
Lana Turner. In these lms, the quoted characters and the original characters
can only interact because of the spectators ability to read the editing in such a
way that their imagination places the characters in shots and reverse shots in the
same space and time.
The progress of digital editing has brought fundamental changes in this con-
text. Nowadays, lmmakers can combine old and new footage within one and
the same shot. Even characters who could never have met in their lifetime can
138 Gloria Withalm

now be represented side by side. Examples of this relatively new practice of

digitextuality range from the ad for Diet Coke teaming Elton John with Louis
Armstrong (Night Club, Steve Horn, US 1991) to feature lms such as Forrest
Gump (Robert Zemeckis, US 1994).
Self-reexive discourse concerning the lm as a product (Figure 7b) can
be exemplied with lms focusing on the cinematographic codes. In order to
qualify as self-reexive, the use of the various stylistic devices has to be related
to the lm in which it is self-reexively applied. Extra-diegetic elements, such
as the credit lines, the end title, or subtitles, become part of the lmic narra-
tive, visible to the characters who then comment on something they are usually
not supposed to be aware of. Sometimes, the devices even materialize and are
suddenly physically present within the diegetic universe of the characters. In
Volunteers (Nicholas Meyer, US 1985), two characters bend over to read the
subtitles at the bottom of the frame, and at the end of History of the World, Part I
(Mel Brooks, US 1981) The History / of the World Part I / The / End appears
in chiseled letters on a mountain.
Such discursive strategies of materializing also appear in transition devices,
such as the device of iris-out in scenes in which the characters try to keep open
the iris diaphragm of the camera, as done by Felix in Comicalamities (Otto
Messmer, US 1928), or Digby Geste (Marty Feldman) who is almost choked
by the diaphragm in Last Remake of Beau Geste (Marty Feldman, US 1977).
The textualization of certain physical properties of the lm strip, the integration
of the lm in its material sense, for instance, when parts of the lm strip such
as frame borders or sprocket holes enter the world of the characters, is another
mode of self-reexivity concerning the lm as a product. Examples are the scene
in Hellzapoppinwhen the lm seems to run through the projector and suddenly
a group of characters is separated by the horizontal border lines between the
frames, or the Wolf in Dumb Hounded (Tex Avery, US 1943) who runs beyond
the sprocket holes of his own lm into the white nowhere (of the projector
light?). The climax of self-reexivity focusing on the lms own materiality
is certainly to show how the lm strip as such breaks, as in a Popeye cartoon
entitled Goonland (Max Fleischer, US 1938), or even starts to burn as in Persona
(Ingmar Bergman, SE 1966).
One of the most striking self-reexive forms of a lm representing itself
as a lm is the lm that makes use of the device of the recursive loop, for
example, by returning to its beginning at its end. The lm narrative that is about
to end transforms itself into another lm narrative, as in the lms Pee-wees
Big Adventure (Tim Burton, US 1986) or in Get Shorty (Barry Sonnenfeld, US
1995), and Wes Cravens New Nightmare (Wes Craven, US 1994). Pee-wees Big
Adventure ends in a drive-in theater during the screening of a movie narrating
The self-reexive screen: Outlines of a comprehensive model 139

the protagonists own adventures during the search for his bicycle shown in
the preceding lm. In Get Shorty, the story of the gangster Chili Palmer (John
Travolta) and the lm director-producer Harry Zimm (Gene Hackman) is about
to reach its climax with its last scene, when, without any prior notice and without
any transition, this nale turns into a movie scene in the course of being shot,
and the lm ends with the wrap of the production of Chilis movie. Finally, at
the end of the movie Wes Cravens New Nightmare, it turns out that it was the
movie itself, via its protagonist Wes Craven, that wrote the end of its own script
of the last Nightmare installment. After having defeated Freddy Kruger, Heather
(Heather Langencamp) nds a copy of the script with the following dedication
by Wes Craven, Heather | Thanks for having the guts | to play Nancy one
last time. | At last Freddys back | where he belongs | Regards | Wes. When
she starts reading the rst page of this script to her little boy, the scene repeats
exactly the opening scene of the lm which is about to end.

7. The full cycle, or: Who says theres nothing good on TV?

Since the phases of the sociosemiotic process of lm are interlinked and some-
times even overlap, there are examples of self-reference and self-reexivity that
can be attributed to more than just one of the phases. Moreover, some lms delib-
erately go full circle covering many, if not all, aspects of the process (Figure 8).

Figure 8. Self-reference and self-reexivity in all phases of lmic reproduction

140 Gloria Withalm

One masterpiece of this kind was already mentioned, The Man with the Movie
Camera by Dziga Vertov (SU 1929), but there are also less artistically renowned
lms that fulll the criteria in question. Among them is a 60 second commercial
for Pepsi Cola directed by Joe Pytka, under the title Set Piece (US 1995; BBDO,
New York). The spot opens with a view over a TV control room in which the
crew is busy with the broadcasting of a basketball game. During a time-out,
the producer starts a Pepsi commercial. The basketball star Shaquille ONeal,4
presented in a close-up on one of the monitors, looks up and turns his head
towards another monitor at the other side of the room where the ad appears
as if he had heard the theme music and the zzing of the soda. Shortly after,
Shaq ONeal leaves his (prolmic) world, the basketball playground, and also
disappears from the screen on which he and the game are being presented. In a
continuous screen passage, he is shown striding through a dozen lms and TV
programs. Right in the middle of this circuit, he enters the scene of the Pepsi
ad, grabbing a Pepsi Cola. Then, he is back on the court to sink a perfect shot,
which allows him nally to have a break, drink his Pepsi, and exclaim Who
says theres nothing good on TV?
This last example combines almost every form of self-reexivity discussed
so far, and it even goes beyond, creating further innovations. The phase of pro-
duction is self-reexive since it shows a team doing their job in a control room
of the television network. It is self-reexive as to its phase of distribution in
two respects. Like in all live television shows, production coincides with distri-
bution, so that the show shows its own distribution. In addition, however, this
commercial evinces self-reexivity as to its distribution since it shows an ad
aired during a commercial break which is itself broadcasted as an ad during a
commercial break. Furthermore, the spot is also self-reexive as to its phase
of consumption. The television crew is watching their own program, not only
because it is their job to do so but also because the broadcast they are producing
is a basketball game. Finally, there is self-reexivity in the form of the digital
quotations of the movies and TV shows which we see during the protagonists
circuit and in the form of a hitherto unseen screen passage of the basketball star.


1. Rossi-Landi uses the recently revived Augustian terms signatum (plural: signata)
and signans (plural: signantia) reintroduced by Jakobson for the two constituents
of the sign in order to avoid the mentalistic ambiguity of Saussures signie and
signiant respectively (Rossi-Landi 1979: 21).
The self-reexive screen: Outlines of a comprehensive model 141

2. This relation was already discussed by Marx (1961: 623) in his characterization of
consumption as giving the product the nishing touch: The conclusion we reach
is not that production, distribution, exchange and consumption are identical, but that
they all form the members of a totality, distinctions within a unity (Marx 1961:
3. In this lm, a splicing girl has to nish the editing of a lm by the evening. She
completes it in the last minute, takes a cab to the movie theater, delivers the lm,
and sits down in the audience. But after the title card Lil Dagover at Breakfast, she
realizes that something went terribly wrong: instead of the well-known movie star,
a black woman with bare breasts appears on the screen playing with the child in her
arm and drinking from a calabash.
4. In 1995, Shaquille ONeal, the famous basketball player and long-term spokesman
for Pepsi, still played with Orlando Magic. When it was rst aired, the spot got an
extra reexive twist because of the context in which it was presented, the NBA 1995
playoffs. The effect of this context was uncertain because it depended on the success
of Shaquille ONeals team in the playoffs but, as Gary Hemphill, then manager of
public relations at the Pepsi-Cola company, stated: We couldnt have been more
fortunate that Magic made it so far, it literally looks like the commercial is part of
the game. Theyre playing, and then it segues right into the spot (quoted in Winters


Marx, Karl
1961 Einleitung [zur Kritik der Politischen Okonomie]. In: Karl Marx and
Friedrich Engels. Werke, vol. 13, 615642. Berlin: Dietz. Engl. Grund-
risse: Foundations of the Critique of Political Economy. London: Pen-
guin 1973, transl. by Martin Liclaus. available online at: Marx & En-
gels Internet Archive.
download/Marx Grundrisse.pdf, 2036. (31.01.03).
Rossi-Landi, Ferruccio
[1975] 1977 Linguistics and Economics, 2nd ed. The Hague: Mouton.
1979 Towards a theory of sign residues. Versus 23: 1532.
1985 Metodica losoca e scienza dei segni. Nuovi saggi sul linguaggio e
lideologia. Milan: Bompiani.
1995 Work, time, and some uses of language. In: Jeff Bernard (ed.), Zei-
chen/Manipulation, 141159. Vienna: OGS.
Stam, Robert
1992 Reexivity in Film and Literature: From Don Quixote to Jean-Luc
Godard, 2nd ed. New York: Columbia University Press.
142 Gloria Withalm

Winters, Patricia
1995 Vodka marketers from overseas are partying it up here on American
soil. Newstimes [source: New York Daily News, AP-NY-06-19-95]. (06.12.01).
Nostalgia of the media / in the media

Andreas Bohn

The media have always been a means of bringing back memories, but they have
also become the object of memory. The cultural and technological development
of the media has brought about great changes; some of the media have even
disappeared. As a result, the media of the past become objects of cultural mem-
ory. The way the media were in the past and how they have changed has become
represented and reected in the media, and this is where self-reference comes in.
Media can refer to themselves as they exist in time and how they have changed
in time. More and more, they do so in rather nostalgic ways, or they reect a
nostalgic way of looking at the media as it can be found in our society.

1. Nostalgia and the paradox of memory

Nostalgia seems to be a feature of our time, but neither the concept nor the
phenomenon is new. The word nostalgia rst appeared in 1678 as the title of
a medical dissertation by Johannes Hofer (Fischer 1980: 268, ref. 8). It derives
from the Greek words nostos, coming home, and algos, pain. In the word
nostalgia, the general sense of longing for something lost, or at least not
at hand, is expressed by means of a spatial image. Nostalgia is considered as a
disease caused by being away from home. Quite early, for example in Rousseaus
correspondence, the role of symbols is mentioned as a cause of nostalgia. The
Swiss soldiers abroad hearing a Swiss melody were reminded of their native
country and became nostalgic; therefore it was forbidden, on penalty of death,
to play this melody during their service (Fischer 1980: 12). However, what we
currently understand by the concept of nostalgia is a predominantly temporal
notion which cannot yet be found before the 1970s (Fischer 1980: 1516). Since
nostalgia has come to mean the longing for something far away, not necessarily
in space, but in time, symbolic representation and objects as mediators between
the past and the present have gained more and more importance.
Nostalgia has to do with what can be called the general paradox of memory.
On the individual level, our consciousness consists of immediate states, each of
144 Andreas Bohn

them dissolving into the next in the course of time, but each state of conscious-
ness is also related to the preceding and the following state. Husserl (18931917:
28, 43) has described the two directions in which consciousness extends as re-
tention and protention. The associated temporal processes determining our
consciousness allow us to construct our personal continuity and guarantee our
identity as conscious beings in time; each immediate state becomes transformed
into a representation of itself, and by the retention of this representation in the
following state of consciousness we are enabled to operate with it.
On the collective level, there is no such automatically functioning device
of memorization. Humans, social groups, and societies had to develop other
strategies of creating memory to make social continuity and identity possible.
But such strategies have always been endangered by the possibility of failure.
Time is a constant threat to social stability, the more so when a society changes
rapidly and when its members are conscious of the changes. One of the strategies
of creating collective memory has been the attempt to eliminate time, to relate
the present very closely to a past which is held in great social esteem.An example
is the glorication of a heroic age as a phase of social and cultural foundation
of the present. The paradox of such strategies of glorication of the past lies
in the circumstance that memory would not be necessary if the past were not
absolutely gone and that memory tries to represent the past as something which
is still present. Cultural strategies of dealing with the past can emphasize either
side of this paradox. Rituals which re-enact scenes from the past and have us
participate in by-gone events show the past as something which is still there;
mourning over the dead, by contrast, does not prevent that the mourners remain
conscious of the death of the deceased.1
At a higher level of reexivity, of course, cultural ways of representing mem-
ory can also deal with many other kinds of events. Nostalgia as a relation of
pain and longing for the past at a mainly personal and emotional level is di-
rected towards objects which cannot only represent but also evoke the past. It
is directed towards objects which allow, at least for a certain time, for a lustful
revival of this past in a process which can be seen as a rst of three steps. The
second step is taken at the collective level where cultural objects are produced
that serve in an analogous way for larger sections of the population. The third
step is the reection on these tendencies in advanced cultural discourse, be it in
a theoretical or an artistic way.
Nostalgia as a cultural phenomenon has become a topic of research, which
has gained more and more attention since the 1970s. It has often been linked
to postmodernism in general, and more specically to ways of getting in touch
with the past. For example, period pictures in the eld of the cinema, remakes,
quotations, re-adoptions of seemingly outdated lm genres, etc. Whereas a pe-
Nostalgia of the media / in the media 145

riod picture refers to a bygone world at the level of representation, a remake

refers to a previous movie and implicitly or explicitly to a past of the mediat-
ing lm. Thus, some of the strategies of nostalgia obviously evince aspects of
self-reference, too, but they are self-referential to a much higher degree when
the object of nostalgia is a specic medium or when the medium itself is used
for representing nostalgia.

2. Media, memory, and musealization

The media participate in an expanding culture of memory evincing traces of

musealization, like the objects in a museum which have lost their use value
having become mere signs of their former use values (cf. Bohn 2005), and nos-
talgia, which can be regarded as a counterpart to the process of modernization.
In the process of modernization, the media have been inuential instruments of
social change not least because they have altered the techniques of cultural mem-
ory. Drawing and, even more so, writing were the rst methods of exteriorizing
memory in an enduring and sustainable manner. The modern audiovisual media,
such as photography, the phonograph, and the lm, have extended cultural mem-
ory to bring back also sensual impressions, which individuals in earlier times
had difculties to remember and to pass on to the next generation. The timbre of
the voice, the typical gestures, facial expressions, and body movements are hard
to remember if you met a person only once or twice; in words and sentences,
such impressions cannot be evoked in a way that others can fully imagine what
the description is about. However, when we see a historical lm, for example,
Oliver Hirschbiegels movie Der Untergang (English: Downfall, G 2004), we
cannot only compare Bruno Ganzs impersonation of Hitler with our personal
memory of the dictator (which not many people still have of Hitler), but we
can also compare it with original audiovisual documents and with our personal
memory of the externalized cultural memory of Hitlers last days.
Important elements of our personal memory are such memories of cultural
memory which have become part of our own biography. Even the time, place, and
circumstances of the situation in which we had the knowledge or experience that
became remembered can itself remain associated with the memory of it. When
did we rst see our favorite movie, in the cinema, on TV, on video, etc.? When did
we rst hear a recording of a certain piece of music which impressed us deeply?
Was it a live or a studio version, which orchestra was it, which conductor, from
which year was the recording? Did we hear it from an old and scratchy LP which
made an ugly noise or from our MP3-player while jogging through a park? The
circumstances of the moment of our rst hearing will often be remembered when
146 Andreas Bohn

we hear the same music again. The examples show that media experiences tend
to go together with the circumstances in which we acquire the memory of them.
Since the media belong to the world of everyday life, they are also the object of
our personal memory. Evolving and transforming themselves in the ux of time,
they are no longer what they used to be in a former stage of our life. Some of us
still remember the time when cinemascope was new and astonishing. A friend
of mine once told me what an erotic disturbance it caused in his adolescence
when he rst saw I Dream of Jeannie on color television (after his parents
had substituted the old black-and-white television set) because now he found
Jeannie so much sexier.
Media products have become the object of nostalgia because they are linked
to so many personal memories and biographies or, more precisely, to individ-
uals constructions of their personal biographies. Whereas younger people are
eager to see new movies on TV, older people are happy when movies from the
past are shown which remind them of their youth. The tendency to their own
musealization which the media have developed is as much a reaction against
the rage for the new as it gives an additional impetus to it. In an exhibition on
the topic of the history of computer games in the year 2002 in Kassels Museum
for Sepulchral Culture, visitors became really sentimental when they saw the
computer model of the old days when they played their rst computer games.
Since computer games have not yet been on the market for a long time, such
nostalgic reactions may seem to come very early but they are understandable
since things have changed so rapidly. Nostalgia seems to depend not only on
the period of time between the event and the moment of its nostalgic recollec-
tion but also on the amount of change. The change can be so great that people
are simply unable to cope with it and search, instead, for a withdrawal into an
articial world of nostalgic remembrance. As Gottfried Fliedl (1990: 171) has
pointed out, situations of abrupt political change, combined with the destruction
of former social structures and hierarchies, have always favored nostalgia and
musealization (cf. Fliedl 1996).
Wolfgang Beckers movie Good-Bye, Lenin! (G 2002) draws on this tendency
with respect to ostalgia or eastalgia, the nostalgia for the good old days of the
GDR (Bohn 2005). The tendency towards the musealization of the GDR in this
lm does not only extend to material culture, but also to the media. With the
help of his friend Denis, a would-be movie director, the protagonist Alex gathers
recordings of GDR television, such as recordings of the daily news program
Aktuelle Kamera or the political magazine Schwarzer Kanal which they
use to produce their own news programs. Alexs mother, a staunch follower of
the communist regime, recovering from a heart attack, has to be prevented from
receiving the news that the GDR has collapsed during the time when she was
Nostalgia of the media / in the media 147

in a coma. The old lady has to stay in bed unable to move. In this position, she
can see the outside world only through her bedroom window. This situation, not
unlike the one of Platos Allegory of the Cave, makes it easy for the two friends
to withhold the ongoing political transformations from the bedridden mother.
But when she watches television, another window to the world is open which
has to be manipulated. First, Alex and Denis simply show her old programs, but
then they begin to experiment themselves with montages of old with new scenes
of their own production. In the end, they even create an alternative history of the
German unication in which they ctionally make come true a third way of
a German republic between former socialism in the East and capitalism in the
West. The reasons why all this became possible are partly in the media politics
of the former GDR, as Paum has pointed out in his following assessment:

Good Bye, Lenin! demonstrates, in an excellent and clever way, the com-
pliance of pictures and tones. The lm goes beyond its own story. The fake
succeeds all the better considering that the GDR, in the course of its forty
years of existence had been in a habit of self-glorication which made the
falsication of alleged documents easy enough. (Paum 2003: 12)

3. Nostalgia of/in the movies

After this little example of TV in a lm in a nostalgic reproduction of GDR

culture, let us now consider more specic examples of nostalgia concerning the
medium of lm itself. A movie can create nostalgia by means of lm history
in general, certain periods of lm history, or by making use of old-fashioned
genres. An example of the latter would be the lm musical, which has been
readopted recently in several productions differently but always with a look back
in nostalgia. Francois Ozon combined his retrospective whodunit in Huit femmes
(F 2002) with elements from the musical tradition. Woody Allens Everyone
Says I Love You (US 1997) with a dozen songs and several dancing scenes
gives the impression of a musical from the beginning to its end. In fact, it
is a compilation and adaptation of material from the classical era of the lm
musical. The score has been arranged by Dick Hyman, the songs are performed
by the actors themselves. Especially the dancing scenes are full of parody. In
the opening sequence, the song Just You, Just Me performed by two young
lovers is accompanied by a choir of three women with baby carriages, an elderly
lady with a nurse, a beggar, and a ballet of three dummies in the shop-window
of an Yves Saint-Laurent shop. There is a highly exalted dancing scene in a
hospital with the song Makin Whoopee by Walter Donaldson and Gus Kahn
148 Andreas Bohn

taken from the Busby Berkeley movie Whoopee of 1930, and there is a ballet of
ghosts in a funeral chapel singing Enjoy Yourself. Near to its end, the movie
makes direct reference to its precursors, the Marx Brothers and their absurd
choreographies, with the song Hooray for Captain Spaulding by Bert Kalmar
and Harry Ruby from Animal Crackers (US 1930). Later, Im Through with
Love is taken up for the third time as the tune of a dancing scene with slow-
motion effects. The ease of Fred Astaires style in his famous scenes with Ginger
Rodgers and others is imitated and exaggerated as Woody Allen is doing nearly
nothing and his partner Goldie Hawn is literally oating in the air. The title song
Everyone Says I Love You by Kalmar and Ruby from the Marx Brothers lm
Horse Feathers (US 1932) accompanies the nale.
Woody Allens Everyone Says I Love You holds the balance between parody
and homage evincing elements of irony and parody in the tradition of the classical
American lm musical itself. These elements can be found in the early musical
comedies of the Marx Brothers or also in Hellzapoppin (US 1941), which is
quoted in the hell scene of Allens Deconstructing Harry (US 1997). These early
examples were more direct parodies than the more recent ones. Woody Allens
lm uses elements from the musical to characterize persons and situations and
to borrow tunes and moods. Emotional qualities are presented by means of well-
known expressions from the history of the lm musical. On the one hand, they
seem to be perfectly natural as expressions of emotions, on the other hand, they
are obviously not spontaneous expressions of genuine feelings, but stereotypes
which, used as quotations, create an ironic distance. There is a shift from the
expression of an emotion to the mere mentioning of its precursor, which attaches
a historical marker to this expression. The nostalgic undertone or mood results
from the feeling that those expressions are no longer useable except in an ironic
and distanced way or as a quotation.
In On connat la chanson (F 1998), Alain Resnais, who had already used
musical elements in La vie est un roman (F 1983), borrows songs from the
tradition of the French chanson instead of taking them from musicals (Ochsner
2004). At the formal level of the combination of story and music, the deviations
from the originals are even greater than in Allens movie. The songs are borrowed
in the acoustic form of their original performance with the voices of well-known
singers like Edith Piaf, Charles Aznavour, Gilbert Becaud, or France Gall. The
actors in the movie apparently do not sing with their own, but with someone
elses voice. With a different voice, they sing songs in a voice recorded from
entertainers of the past according to different technical recording standards.
Sometimes a man even sings in the voice of a woman and vice versa. For example,
in the opening sequence, the commanding ofcer of the German occupation
forces, having just received the order to destroy Paris, sings the song Jai deux
Nostalgia of the media / in the media 149

amours, Paris et ma patrie in Josephine Bakers voice. Though the songs go

very well with the situation in which they are sung, they create an ambiguous
effect. By themselves, they aim at enacting certain affections in order to express
and to intensify the emotions. Using these songs as elements of a narration could
serve the same purpose as in an opera or in a classical lm musical. However,
the aforementioned ruptures and divergences between story and song, actor and
voice are incompatible with the traditional purposes of combining a narrative
with music. Emotional intensication and historical distance are in an unstable
balance, which makes it very difcult to concentrate on one of the two and
to forget about the other. However, as much as these discrepancies may be an
impediment to our identifying with the protagonists and getting emotionally
involved in the story, it gives us the possibility of concentrating on the songs
and of remembering the possible emotional importance they may have to us.
The movie turns into a sort of living museum of French chanson culture,
with Everyone Says I Love You as its counterpart in the classical American lm
Maurizio Nichettis Ladri di saponette (IT 1988) does not only quote a spe-
cic form typical of a genre of lm history but a whole set of formal character-
istics of a historical period, the style of Italian postwar neorealism. Neorealism
is not simply a period style among others. It was intended and perceived as a
counterpoise against fascist monumentalism and distraction by means of glam-
orous settings and sceneries. The neorealist movies tell stories about people in
a humble social environment, stories about the living conditions of ordinary
people, about moral values, and matters of conscience. The neorealist movies
construct the image of new nonfascist Italy, and this new image was not only
highly esteemed in postwar Italy, but also internationally well liked and honored.
The title of Nichettis lm is a reference to Vittorio de Sicas Ladri di biciclette
(IT 1948), the most famous example of this style, which won an Oscar in 1949,
already the second for de Sica.
The change of its title is indicative of what is happening in the lm. Saponette
are the commercials for soap, washing-powder, etc. which rudely interrupt the
movies shown on TV. Nichettis lm begins with a TV presentation of a lm
in the neorealist style, paying homage to neorealism in black and white. The
commercials in the breaks are, of course, in color. There are elements of par-
ody in both the lm and the commercial. At a certain point the lm gets mixed
up with the commercials when a gure in color from a commercial enters the
movie and the black-and-white scenes now have one of its actors in color. At this
point, the forms of neorealism and current television advertisements are in di-
rect confrontation. Film and commercials are no longer separate. They are both
integrated within the one diegetic framework of Nichettis lm in which they
150 Andreas Bohn

nevertheless remain clearly marked as two different modes of representation.

The combined scenario now evinces divergent goals for the narrative personae
in their respective stories, such as preservation of basic moral principles under
difcult circumstances, family values, or modest prosperity in the frame of neo-
realism versus consumerism in the frame of the TV commercial. The mixing-up
of the two frames has effects on the narrative plot, too. The postwar neorealist
family wants to escape their misery and live in the consumers paradise of post-
modern advertising. In this way, the anachronism of neorealism in the eighties
becomes evident but also the lack of realism and morals of the commercials.
Both worlds operating separately in the peaceful coexistence of the usual TV
program with the movies, but they shatter into pieces when they clash, as in
Ladri di saponette.
The quasi-neorealist lm-in-the-lm is in itself, like the movies it imitates,
highly sentimental. By means of its nostalgic touch it is able to involve the view-
ers emotionally with the values it promotes. At rst glance, the arrangement of
feature lm, TV studio scenes, commercials, and TV viewers at home is clearly
committed as to its values: the good old times of the feature lm versus the
contemporary perversion, but this initial emotional orientation of the spectator
gets more and more into trouble. It ends up in an utter boundary crossing at
different levels, a crossing that nally erodes the moral hierarchy and the affec-
tive structures established in the beginning. The result is mixed emotions and a
sense of affective ambivalence on the side of the viewers who have followed the
development of the lmic narrative.

4. Nostalgia and self-reference

The analysis of the above examples has shown that besides the ongoing produc-
tion of genre lms and the continuous modication of genre rules, and besides
the nostalgic imitation of historical styles, for instance in period pictures, there
are other ways of incorporating genre traditions and elements of style into a lm.
Quotations of formal elements can be used to play with the emotions associated
with these forms or to create a historical distance from them which may cause
a nostalgic longing for their restoration. These are some of the means of the
cinema to reect on its own history. The tendency of the lm to deal with itself
and reect on its impact on its viewers has even been passed on to television.
Pleasantville (US 1998), directed by Gary Ross, is a movie about a boy who
loves certain old TV family series from the 1950s (Dika 2003: 201). His nostal-
gia is represented like the nostalgia for neorealism in Ladri di saponette. Both
movies leave us with the impression that the past is much better than the present
Nostalgia of the media / in the media 151

because it is past and we do not really have to live it any more. The protagonist
of Pleasantville is nostalgic for the world of the 1950s family series, which is
not only ctitious but was already outdated when he rst saw it, apart from the
fact that he never knew the society and the state of mind that produced it. The
only thing he knows are the episodes from the series which are repeated on TV
and that they are old because they are coded with past.
Not only does the cinema serve as a cultural archive for television and has
become a source of nostalgia, but also has the reverse become true. The above
discussed On connat la chanson does not only refer to the tradition of the
musical but also to the French chanson and its phonographic recordings. Other
media can be added, for example, the radio, as in Woody Allens Radio Days (US
1987), a lm that associates nostalgic reminiscences of the radio as it was before
the advent of TV, with personal memories from the narrators childhood. In a
similar way, movies such as Giuseppe Tornatores Nuovo cinema Paradiso (F/US
1989) or Ettore Scolas Splendor (F/IT 1989) intermingle the recollections of an
individual life (in the former case beginning with the protagonists childhood)
with the personal memories of a particular movie house, the lms shown there,
the technical equipment which was used etc. Rather differently, in Agnes Vardas
Les cent et une nuits (F 1995), the history of the lm is narrated to a young woman
from the memory of an allegorical Monsieur Cinema.
In Tornatores and Scolas movies, general history, personal history, and me-
dia history are interrelated, as Splendor is with the end of World War II, the
protagonists return, and the showing of Frank Capras Its a Wonderful Life (US
1946). At the moment when the owner of the cinema enters it for the rst time
after the war, the nal scene of the movie is presented. It is Christmas, the com-
munity and the protagonist are reconciled, and everybody sings For Auld Lang
Syne. The tune is taken up at the end of the movie when the village community
gathers to defend the old cinema which is in danger because it is no longer prof-
itable. The movie changes from color to black and white, and although we are
in the middle of summer, snow begins to fall. Few people nowadays remember
this, but Christmas in its ritual essence is a means of recalling something from
the deep past to the present in order to share its effects. The Christmas scene
taken from Its a Wonderful Life associates the demise of the cinema in a small
town with the powerful history of the medium in order to convey the message
of its endurance and resistance against current transformations. However, as
the lm Splendor demonstrates, that which is greater than life works only in
the movies. In this respect, Splendor is the most nostalgic of all lms dealing
with the lm culture of the past; it is self-referential with regards to the medium
which it represents and of which it evokes the feelings of nostalgia.
152 Andreas Bohn

5. Conclusion

In this paper, the focus was on examples from the lm which do not only follow
the trend towards nostalgia but also reect it in a rather complex manner. Below
the surface of the current trend towards media nostalgia, there is a broad cur-
rent of musealization to counterbalance the hype about progress which the new
media cause. Even among those who are euphoric about the internet or even
addicted to computer games, nostalgia can be found, for example, nostalgia for
oppy disks, which seems to be the most recent manifestation of media nos-
talgia. The nostalgia of the media does not only extend to the material remains
which have been collected in media archives, personal collections, or which
have been exhibited in museums and cultural centers. Media nostalgia is also
apparent in the way the media represent the media and in the way they let us
see the world narrated by them. More and more, the media devote themselves to
media nostalgia, relying on different historical ways of positioning themselves
in relation to other media. Media nostalgia in the media is a manifestation of
self-reference in the media because the media refer to themselves, show how
they have been the source of entertainment, how they have been subject to his-
torical changes or even destruction, and how they have been remembered or
consigned to oblivion.


1. In this context, Boym (2001: 41) distinguishes between restorative and reective
nostalgia: Restorative nostalgia puts emphasis on nostos and proposes to rebuild
the lost home and patch up the memory gaps. Reective nostalgia dwells in algia,
in longing and loss, the imperfect process of remembrance.


Bohn, Andreas
2005 Memory, musealization and alternative history in Michael Kleebergs
novel Ein Garten im Norden and Wolfgang Beckers Film Good Bye,
Lenin! In: Silke Arnold-de Simine (ed.), Memory Traces: 1989 and
the Question of German Cultural Identity, 245260. Frankfurt: Peter
Boym, Svetlana
2001 The Future of Nostalgia. New York: Basic Books.
Nostalgia of the media / in the media 153

Dika, Vera
2003 Recycled Culture in Contemporary Art and Film. The Uses of Nostal-
gia. Cambridge: University Press.
Fischer, Volker
1980 Nostalgie. Geschichte und Kultur als Trodelmarkt. Luzern: Bucher.
Fliedl, Gottfried
1990 Testamentskultur: Musealisierung und Kompensation. In: Wolfgang
Zacharias (ed.), Zeitphanomen Musealisierung. DasVerschwinden der
Gegenwart und die Konstruktion der Erinnerung, 166179. Essen:
Fliedl, Gottfried (ed.)
1996 Die Erndung des Museums. Burgerliche Museumsidee und Franzosi-
sche Revolution. Vienna: Turia & Kant.
Husserl, Edmund
[18931917] 1969 Zur Phanomenologie des inneren Zeitbewusstesens, ed. Rudolf
Boehm (=Husserliana X). The Hague: Nijhoff.
Ochsner, Beate
2004 Jai deux amours: la musique et le lm . . . Intermediale Verschran-
kungen von Musik und Film in On connat la chanson (1997) vonAlain
Resnais. In: Susanne Schlunder and Scarlett Winter (eds.), Korper-
Asthetik-Spiel. Zur lmischen ecriture der Nouvelle Vague, 157
182. Munich: Fink.
Paum, Hans Gunter
2003 Der diskrete Charme der Ostalgie:Wolfgang Beckers Good Bye, Lenin!
konkurriert noch um den Berlinale-Baren und kommt schon ins Kino,
Suddeutsche Zeitung 36 (February 13, 2003), 12.
Self-reference in animated lms

Jan Siebert

1. Introduction

If we look at the nature of animated or drawn lms1 , we can often nd traces

that point to the articiality of the images in a number of ways, thus enabling
the lmmakers to inscribe themselves in the story of the lm. It seems that these
lmmakers have an attitude towards lmmaking that is different from the view
taken by live-action directors. To a much higher degree, animated lms give
their creators the freedom to experiment with physical and biological laws and
to adopt new perspectives on reality in ways impossible in live-action lms. This
becomes most obvious when animated lms use animals as protagonists to tell a
story from an alternative point of view. On a more general level, animated lms
often include a comment on the borders between the individual media which
are being crossed again and again, since drawn lms show inuences by comic
books, literature, paintings, computer games, and (classical) live-action lms.
Methods of disillusioning the viewer are pointing to other media, letting the
actors address the viewers, or pretending that the lm is in the making at the
moment the viewer is watching it, to mention only a few.2
When the rst animated lms were produced, the illusion of moving pictures
was far from perfect. For example, the animators did not always cut out every
frame in which their hands accidentally got photographed while adding another
stroke. They were probably even proud to present themselves as the creators.
It is part of the charm of the very early animated lms that from time to time
a nger here and there shows up for a millisecond. It is interesting to see that
although new technology can provide perfectly rendered images, animators to-
day in continuation of the tradition of their predecessors still think of ways of
including self-referential effects.
156 Jan Siebert

2. References to the paper medium in old and new lms

One of the most successful animated lms from the era of silent lms is a series
of lms with the remarkable title Out of the Inkwell (USA 1918ff.). This lm is
about its own making, full of indices of how it was produced. In every episode,
the protagonists climb out of an inkwell, and they go back inside after the story
has ended, just like the opening and closing of a book.
Another highly appraised series from that time is Felix the Cat (USA 1922).
Whenever Felix is in a difcult situation, a thinking bubble with a question mark
appears above his head. Felix then grabs this question mark and turns it into an
object useful to help him. This is a reference to the paper medium comic book
that preceded the moving images. In the era of silent movies, the presentation of
devices such as speech bubbles etc. can be described as a creative replacement
for the words coming out of the characters mouth. Until today animated lms
use stars, bubbles, and bouncing letters surrounding a character. In the German
lm Werner Beinhart (1990), for example, the protagonist gets so angry that
his words come out of his mouth in huge red letters traveling through corridors
(Figure 1). Since the animated lm very often uses the aesthetics of the comic
book (and this lm is an adaptation of a highly successful German comic book),
this usage does not seem out of place.

Figure 1. Comic book aesthetics in Werner Beinhart (G 1990)

Many years later, when the animated lm had sound and color, a similar
reference to the paper medium was used to create comical effects in the ani-
mated lm Flatworld (GB 1997). This lm features the perfect impression of
two-dimensional cut-out objects; even the markings on the paper can be seen.
Flatworld creates its own rules, and its cut-out forms dominate the narration
and inuence the action. It allows the protagonist sitting in a car to slide slant-
wise beneath another car during a wild chase (Figure 2). Holes in the street are
repaired with a stapler which is also used as a weapon!
Self-reference in animated lms 157

Figure 2. Two-dimensional paper objects in Flatworld (GB 1997)

The lm Felix the Cat and the Out of the Inkwell lm series work in much
the same way. The protagonists react to their special environment, and the story
explicitly crosses borders to other media and becomes bizarre, even abstract.
Many animated lms use this exible interpretation of media borders and dis-
tort the realism that most live-action lms are based on. Generally speaking,
animated lms have never been able to challenge the authenticity of the images
of live-action lms. Instead, they seek to create a different atmosphere and a
world with its own rules in which not only the characters are very exible by
using the squash-and-stretch-animation but also the borders of the media can
be stretched to a very high degree. The protagonists are not tied to the physical
laws of gravity; they do not die, not even if their body is lled with bullets.
As we learn in the live-action-animated lm-mixture Who framed Roger Rabbit
(USA 1988), animated personnel can in fact die; they just have to be erased.

3. Interactions with the viewer

The talkative Bugs Bunny character is highly self-referential mainly because of

his many contacts and pacts with the viewers. Even in the middle of a typical
chase scene, he might pause and talk to them. For example, Bugs Bunny might be
unhappy with the story because it uses a gag for the second time. This inclusion
158 Jan Siebert

Figure 3. The spectator included in the lm Daffy Duck and Egghead (USA 1938)

of the viewer is another way of creating a self-referential effect. In another very

early Warner Brothers lm, the viewer is not only suspected somewhere among
the audience, but actually plays a part in the lm: Daffy Duck and Egghead (USA
1938) includes a scene that shows a duck hunter (later known as Egghead) who
is irritated by a spectator from the audience moving through the rows (Figure 3).
The spectators whole body almost blackens the scene and the hunter is afraid
the spectator might chase his prey away. Since the spectator does not go away,
the hunter nally shoots him dead, he stumbles and falls. The bullets have struck
through the fourth wall of the lm and have deployed their effect on the other
side of the screen.

4. A product in the making

Very often, the lm presents itself as a product that is still in the making when it
is presented to the audience, so we can see parts of the lm crew or the creators
hands etc. Sometimes we can even see parts of the material of the lms itself. In
Tex Averys Dumb Hounded (USA 1943), one character is moving so fast that
he ends up running out of the frame and even the holes of the lm reel become
visible. After having realized this, the character quickly turns around because
there is nothing but a white background obviously the white projection screen
(Figure 4). Of course, this consequence lacks any logic because the characters
picture could not be projected had it run out of frame. Nevertheless, the gag
works despite its lack of logic since, as Lindvall and Melton (1997: 210) observe,
animated lms do not need the consistency or internal logic of a realist lm;
[. . . ] the super-textual can break into the text at any moment.
Self-reference in animated lms 159

Figure 4. Cartoon character leaves the lm frame in Dumb Hounded (USA 1943)

The same is true of a seemingly never-ending chase scene in Lucky Ducky

(USA 1948). The characters are running past a road sign, and, all of a sudden,
they nd themselves in a black-and-white scenario. The music stops as they look
at each other, turn around, and then notice the road sign, which says: Technicolor
ends here. As soon as they re-enter the world behind the road sign, they are
colored again, and music continues. This is a brilliant comment on the endless
chases so typical of the animated lms of that time; this one has gone on for so
long that the story loses its color and its sound. The protagonists have reached
the end of their (animated) world.3
Each time lmmakers have their protagonists turn to the camera or have them
think about their status as an integral part of a lm, they bring them a bit closer
to the audience. The drawn characters move out of the diegetic frame, and when
they address their creator, the viewers get the impression of a live broadcast,
with the screen turning into a theater stage. The lm seems to be in the making.
Space Jam (USA 1994) shows children watching a classical TV series with
Warner Brothers characters. Suddenly the two protagonists are stopped in the
middle of a ght by a third who asks them to join him to a conference of actors
of animated lms. They all leave, and the scenery is completely empty. This
suggests that all Warner Brothers characters have a job in endless re-runs. In
Space Jam the viewer can actually see what the characters are doing in their spare
time. They even complain about not getting any money from the merchandising
campaigns that use their names and faces.
160 Jan Siebert

5. New developments

During the last decade, many computer-generated lms have been released that
have succeeded in pushing the genre to new heights. Toy Story (USA 1995)
was the rst feature lm to be entirely produced with a computer, and other
very successful lms like A Bugs Life (USA 1998), Shrek (USA 2001, 2004)
and Finding Nemo (USA 2003) soon followed. A new and most interesting
development is that we are now being better informed about what the lmmakers
think in relation to the device of self-reference and its comical effects in their
cartoons, since the DVD versions of these highly professionally produced lms
bring us relevant interviews in making-ofs. Since DVDs are on the market,
out-takes, making-ofs, and trailers have become part of the medium. Out-takes
typically show scenes originally taken out for various reasons, for example
because actors forgot their texts, slipped, or failed in other ways. The DVDs
of Toy Story, A Bugs Life and Shrek are among those which include out-takes,
which are mere fakes or just copies of the way live-action lms are presented.
Out-takes from animated lms are by no means bad material, nothing taken
out. They are special productions which provide the spectators with bonus gags.
In A Bugs Life we can see a baddy who is worried not to create the impression
of being a tough person. In another scene the lmmakers borrow the protagonist
Woody from Toy Story, the rst successful lm produced by the same company.
Woody wants to help directing the lm with the result that other parts of the
lming material become visible (microphones, a camera, etc.). The characters
in these mock out-takes act as if they were humans of esh and blood. At one
point, a gigantic, intimidating bird can be seen, spreading its wings that must
look enormous to the small bugs. All of a sudden, the birds movements come
to a grinding halt accompanied by the noise of an old, malfunctioning machine.
The contrast between the perfection of the computer-animated surroundings and
the archaic aura of the machine age is an explicit reference (or even homage) to
a time when live-action images had to be created without blue-screen techniques
and without the computer-aided postproduction process.
In recent DVDs of animated lms one can occasionally nd not only fake out-
takes but also mock interviews with actors. This is not an entirely new device;
since the early times of the medium, the makers of animated lms have been
using it much more frequently than the directors of live-action movies. However,
the more recent implementation of self-referential effects seems to express in
a kind of tongue-in-cheek attitude: Yes, we all know that we are dealing with
highly articial images. Our protagonists can do impossible things, so let us
comment on these exible interpretations of the world. Let us experiment with
ways of crossing the borders of our medium with the medium of the comics and
Self-reference in animated lms 161

cartoons, or with live-action lms to make the viewers aware of the processes of
the many possibilities involved. Even the animated characters are aware of these
processes and make use of their knowledge. Films like Who framed Roger
Rabbit? even feature animated characters who meet real persons and both
learn from each other. In The Mask (USA 1994), a live-action lm based on a
comic book hero, a bank clerk puts on a mask and turns into a creature half-way
between human being and animated character. He can benet from the most
important features of a comic book hero; he is invincible, his body can deform
and adapt to each new situation. We can see that it is not only the animated lm
that learns from the predominance of the live-action lms but also vice versa.


1. An animated lm is one that is created frame-by-frame (Stephenson 1973: 14f) in

contrast to live-action lms that are recorded live action in front of the camera. In this
essay, I will concentrate on those animated lms which dominate the genre: those
that are based on drawn images (hand-made, also computer-assisted) in contrast to
sand or clay animation etc.
2. These are only a few examples of the ways animated lms use self-referential effects
in order to create comedy. See Siebert (2005) for a detailed description.
3. It might also be the painter who is fed up with providing more and more scenery,
which would be an ironic remark about the practice of using the same scenery over
and over again.


Lindvall, Terrance and J. Matthew Melton

1997 Toward a post-modern animated discourse: Bakhtin, intertextuality
and the cartoon carnival. In: Jayne Pilling (ed.), A Reader in Animation
Studies, 203220. London: John Libbey.
Siebert, Jan
2005 Flexible Figuren: Medienreexive Komik im Zeichentricklm. (Kul-
turen des Komischen 2.) Bielefeld: Aisthesis.
Stephenson, Ralph
1973 The Animated Film. London: Tantivy Press.
Part V. Self-referential television
On the use of self-disclosure as a mode of audiovisual

Fernando Andacht

1. On the use of reexivity in the documentary genre

The present study is part of an ongoing research on some contemporary me-

dia representations of reality1 , such as documentary lm and TV reality show,
which are genres based on the predominance of indexical signs.2 These two
kinds of nonction are characterized by their opposite though complementary
use of reexivity: in the case of the documentary lm, reexivity is a self-
conscious, earnest, and explicit strategy with the aim of making the viewers
aware of the illusory nature of the reality effect created by the very lm they are
watching, in a way akin to the epistemological core of many studies in social
constructionism (Hacking 1999). In the case of the TV reality show, which will
be studied on the basis of the successful format of Big Brother, reexivity is
also crucial for its normal functioning but it works in a self-mocking, mundane,
and ostensive manner to enhance the entertainment value of this television pro-
gram. Based on Lynchs (2000) carefully argued critique of the use and abuse
of reexivity as a methodological tool in the social sciences, I will argue that
the emancipatory use made of reexivity in the quality documentary lm genre
with its ethical concern is ultimately related to dualism as its metaphysical
This paper posits an analysis of a kind of media reexivity that corresponds to
what Noth describes as communicative self-reference (see I.5.6.). It involves
the ethically and aesthetically justied inclusion of scenes of the making of the
documentary in the lm itself. Self-disclosure of this kind is a key feature of
the poetics of reexivity in artistic documentaries such as those by Coutinho in
Brazil and Comolli or Rouch in France. Allen (1977: 37) denes this practice
as follows:

Self-reexivity is dened here as any aspect of a lm which points to its

own processes of production: the conceptualization of a lm, the procedures
necessary to make the technology available, the process of lming itself.
166 Fernando Andacht

[. . . ] By presenting them self-reexively, a documentary lm can make

an audience aware of the processes of production as a limitation on the
lms neutral stance, its ability to document objectively. In doing so the
lm draws attention to the process of selecting and reconstructing events to
convey meaning. Self-reexivity becomes then a reaction against or a way
of countering the traditional mode of the documentary which emphasizes

The documentary Edifcio Master (Brazil, 2002, henceforth EM ) sets off

with scenes of the laborious arrival of the lm crew led by director Eduardo
Coutinho (Sao Paulo, 1933- ) in the Master building in Copacabana, Rio de
Janeiro. Three sets of cameras are simultaneously at work in the lm, only two
of which are visible: those carried by the main crew, the electronic cameras that
survey the building and, last but not least, the reexivity creating cameras which
are lming the lm making activity itself but which, of necessity, must remain
unlmed. This multiplication of perspective is distinctive of the documentarists
cinematic style, as lm critic Mattos (2003) describes in his study of Coutinhos
work: Part of that person-to-person cinema is the exhibition of the process of
documenting inside the very lm. The arrivals of the camera crew which are
always documented by a supporting camera which duplicates the axis of the
main camera have become a trademark since Cabra marcada para morrer.
The audience gets to see some crucial moments of the actual backstage of the
lm, of its material production, e.g., the negotiations made to be admitted inside
the apartments whose inhabitants are invited to participate in the lm. It soon
becomes clear that several attempts end up in failure. Thus, lm viewers join
the research crew in their strenuous efforts to visit many of the 276 ats which
nally result in 35 interviews of which only 27 were kept in the edited lm,
as the director comments in an initial voice-over sequence. In the documentary,
reexivity is also implemented during the encounters when the auxiliary camera
switches from the dwellers, who do most of the talking, to their engrossed
interlocutor thus enabling the viewers to catch a glimpse of the director and of
the camera crew behind him.
In the intellectually informed practice of the quality documentary genre (e.g.,
cinema verite), reexivity of this kind results in acts of self-disclosure which
will be dened as self-critical reexivity. My analytical aim is to relate this
mode of media reexivity with two different postulates of the representation
of the real. Although documentary makers are not semioticians, their implicit
metaphysical views do exert a key role in how they proceed in their creative
tasks. The documentarists use of reexivity owes much to an unwitting belief
in the doctrine of dualism, the kind of thought which, in the words of Peirce
(CP 7.570), performs its analyses with an axe, leaving as the ultimate elements
On the use of self-disclosure as a mode of audiovisual reexivity 167

unrelated chunks of being. My critique is an attempt to disclose a fallacy

at the heart of the poetics of reexive documentary which is characteristic of
postmodern anti-objectivistic thought.4 The nondualistic premise defended here
is that there is no unbridgeable gap between reality and its representation, be they
qualitative, factual, or general. Such is the upshot of Peirces formidable critique
of Cartesian reductionistic dualism which is inseparable from his semiotic and
which implies an alternative method based on an absolutely original concept,
the concept of thought as a sign, as Santaella (2004: 24) explains. What makes
something a sign is not material;5 it is a capacity, since its essential function is
to render inefcient relations efcient (CP 8.332). A sign can do that on account
of its power of serving as intermediary between its Object and a Mind (CP
6.455). This is the synechistic way of all signs, which results in the continuous
growth of reasonableness (CP 5.4). Thus, it may not be inappropriate to add to
these examples proposed by Peirce, which include a daily newspaper (6.455),
a TV reality show.
To close this introduction, I bring in the voice of an important, unwitting
contributor to self-critical reexivity in the media, namely William James, a
life-long friend of Peirce and a not too felicitous interpreter of his doctrine of
pragmatism. Concerning the question of the possibility of a reality independent
of human thinking James wrote:

We may glimpse it, but we never grasp it; what we grasp is always some
substitute for it, which some previous human thinking has peptonized and
cooked for our consumption.6 If so vulgar an expression were allowed us
we might say that wherever we nd it, it has been already faked. (James
1963/1906: 109, emphasis in the original)

The metaphysical basis of self-critical reexivity lies in a nutshell in Jamess

peculiar account of representation: James states that what signs do with the real
as it is apart from its being represented is to distort it to the point of disso-
lution, like enzymes do with the food we eat, to extend his digestive metaphor.
Although no actual, historical inuence may be claimed, there is enough in-
tellectual kinship in Jamess assertion to construe it as a virtual philosophical
basis of self-critical reexivity such as it is embodied in the melancholy poetics
of contemporary documentary. Jamess image of irrecoverable loss, of a radical
change for the worse the use of the word faked at the climax of his account
entails the impossibility of ever beholding reality face to face. This decit-like
conception of representing the real entails our falling prey to an ersatz, a faked
version of reality, thus becoming helpless dupes while trying to represent it,
since we have nothing else to go by. From the self-critical reexive viewpoint,
all a documentarist can do is struggle against the illusion of the reality they
168 Fernando Andacht

inevitably help to create with their lms about an ever elusive real. This hope-
lessly Sisyphean task will be illustrated below with the work and thoughts of
the Brazilian director Coutinho. Jamess construal of representation is the very
opposite of Peirces (CP 5.607) proposal, which involves a direct and mediated
way of perceiving the world, one which Ransdell (1986: 68) calls the doctrine
of representative perception.
In the same year in which James (1906) gave his above quoted lectures on
Pragmatism at Bostons Lowell Institute, Peirce addressed himself to a gifted
British correspondent to explain his mature conception of signs as elements
which are in contact with the outward clash, with the real as it is when it is not
being represented and which do bring such an acquaintance effectively to us
interpreters. In a letter dated March 9th, 1906 we nd the sketch of a synechistic
theory of signication according to which inner and outer world are not separate
but in a living communi(cat)ion:
Sign [is] any medium for the communication or extension of a Form
[. . . ] In order that [a form] be extended or communicated, it is necessary
that it should have been really embodied in a Subject independently of the
communication [. . . ] The Form, (and the Form is the Object of the Sign),
as it really determines the former Subject, is quite independent of the sign;
yet [. . . ] the object of a sign can be nothing but what that sign represents it
to be. [T]o reconcile these apparently conicting Truths, it is indispensable
to distinguish the immediate object from the dynamical object. (Hardwick
1977: 196 my emphasis)
What James described as the unavoidable degradation of reality on account of
its representation is the paramount instance of communication, of contact with
ourselves and with others, i.e., with reality independent of the sign according
to Peirces realist, synechistic semiotic. The distinction between two kinds of
semiotic objects is the technical solution to the supposed dilemma faced by
documentarists who pursue reexivity: they can never grasp the entirety of
the real which they fallibly and partially register in their lms, regardless of
their editing or reconstructing. The actual limitation has more to do with the
Heraclitean nature of reality and with the fallible way of signs. Through their
evolving nature, signs manage to depict and portray, albeit imperfectly, enough
of the real for viewers and documentarists alike, to do more than just glimpse
the real. We are able to grasp enough of it to criticize whether a documentary is
doing its job well or not and how it fails, if that is the case, to fulll its indexing
(Carroll 1996: 238). A solution to the riddle posited by our limited, human way
of cognizing the world out there is offered by Peirce by means of the rainbow
image (CP 5.283): the metaphor shows how the world outside, i.e., everything
which is present to us is both a phenomenal manifestation of ourselves,
On the use of self-disclosure as a mode of audiovisual reexivity 169

and the manifestation of something external, that lies outside of signs, i.e., its
object. The world works its dynamic, forceful limits through signs. Peirce has
bequeathed us a realist analysis of representation, in stark contrast to Jamess
dualism, which involves our giving up any hope to grasp the real which is forever
faked by our signs.
In conformity with Peircean synechistic semiotic, I argue that you can have
the real thing and represent it too. Curiously enough, the excruciating self-doubt
of the ethically and politically conscious poetics of reexive documentary nds
a positive resolution in the TV reality show. This is so because life involves
the steady growth of meaning, which occurs every time reality is represented,
no matter in how articial and biased a manner (Andacht 2005). The indexical
ground provides the main input for the symbolic production of the two audio-
visual genres and the public meaning generated and consumed as their public
meaning. The ambivalence at the heart of the self-critical documentary its aim
to represent faithfully what these directors believe its representation cannot but
miss is contrasted with the jolly assurance of the reality show everyday life
is both real and represented in a ow of signs of ever increasing complexity.
This does not imply an ethical or aesthetical judgment of either genre; instead,
I attempt to give a semiotic account of media reexivity based on its general
consequences according to Peirces pragmatic maxim (CP 5.89). What makes
the quality documentary fascinating to its audience is precisely what makes its
poetics of self-critical reexivity awed. In the case of the Big Brother reality
show, which stands in the opposite aesthetic pole of the indexical range, what
accounts for its popularity is the continuity between the signs which are closest
to our complex human and animal nature, namely, the indices emitted by our
bodies and the elaborate symbols into which these indices are incorporated and
permanently transformed for the sake of entertainment. This is so in life as it is
on the television or movie screen.

2. Self-critical reexivity: Melancholy over peptonized,

documentary reality

Let us consider two approaches to self-critical reexivity in documentaries. The

rst extends a critique of methodological reexivity in the social sciences to
this genre, the second is a semiotic discussion of the oral poetics of reexivity
as practiced by the contemporary documentarists Coutinho and Comolli. We
will rst approach reexivity from the viewpoint of Jamess dualistic account of
representation as a peptonizing agent which paradoxically undoes what he aims
170 Fernando Andacht

to convey. We will then contrast this approach to Peirces synechistic construal

of signs as fallible but powerful and, in the long run, reliable media for reaching
the real, both in life and in lm. Adopting the latter approach, we hope to be able
to undo the ambivalence at the heart of self-critical lm reexivity and of its
doleful, self-appointed Sisyphean task of grasping what these documentarists
say cannot be grasped.
In a paper on reexivity as a method in the social sciences, Lynch criticizes
the excesses to which an infatuation with this procedure may lead, particularly
in constructionist analyses:
Like any other effort to expose, uncover, reveal or disclose surprising, coun-
terintuitive, and potentially unsettling matters, a reexive analysis must be
entrusted to an uncertain fate. There are no guarantees of success and no in-
herent advantages to doing reexivity or being reexive. Consequently,
a project that deconstructs objective claims should be no more or less prob-
lematic, in principle, than the claims it seeks to deconstruct. [. . . ] There
is no particular advantage to being reexive, or doing reexive analy-
sis, unless something provocative, interesting, or revealing comes from it.
(Lynch 2000: 42 emphasis in the original)
Lynch (2000: 34) acknowledges that the remarkable academic success en-
joyed at present by reexivity is indebted to the Enlightenment conception of
self-reection as a uniquely human cognitive capacity that enables progressive
understanding of the human predicament. This is a timely caveat against the
risks involved in any contribution to scientic knowledge: an excessive reliance
on reexivity may involve a displacement from the fourth method, the scien-
tic one, which follows the other three ways of xing belief in Peirces classic
epistemological statement (CP 5.358387). In such cases, inquiry is no longer
determined by the object itself, by some external permanency, by something
upon which our thinking has no effect (CP 5.384), but by a priori convictions,
by fashionable theoretical trends, or by sheer tenacity. When this happens, the
outcome is no longer a scientically valid procedure for advancing from doubt
to belief. Naturally, such scientic rigor is not a requirement for documentary
making, not even of the self-critical kind. Still, there is more than a passing re-
semblance in the procedure of these lm directors to justify an analogy with sci-
entic inquiry. Documentarists are concerned with the methodological matters
involved in their work, which they make explicit in their self-critical statements
on their lms. Thus, I propose a partial parallelism between the reexive meth-
ods used by social scientists and the poetics enacted by reexive documentarists
both in their lms and in their reection thereupon.
From the extensive list of variants of methodological reexivity drawn by
Lynch (2000: 29), I will consider three reexive methods and examine their
On the use of self-disclosure as a mode of audiovisual reexivity 171

approximate cinematic counterparts in the self-critical documentary EM. The

parallelism brings to light a procedure which enjoys great favor in postmodern
culture and in the media (Noth 2001). First, there is the method of philosophical
self-reection, it is similar to Cartesian introspection, and it makes the director
Coutinho ponder on his own identity as he is meeting and lming the participants
of his documentary. Who am I, who are they, my lm subjects, as they are talking
to me in the way they do? These are some of the questions involved in this
reexive methodology, and Coutinho deals with them by attending to what he
calls his ethical problem (cf. Caldeira 2001). He tries to resolve these doubts
by becoming what I would like to describe as a caretaker or curator of the self
of the Other: As I deal with helpless people, my ethical problem in editing [this
lm] is the following: Will I harm the life of this person? This is my concern.
The lm does not make things any better, but it can make them worse (Caldeira
Next in Lynchs (2000: 29) taxonomy there is methodological self-con-
sciousness, a common strategy in qualitative studies, which has become a
canonical feature of participant observation. It tries to undo the distortions
introduced by subjective bias in observations of reality. This is a central concern
for Coutinho as he reects on his own middle class, intellectual origin and
ideology, when he sets out to lm the Other, typically the underdog of shanty
towns (e.g., in Babilonia 2000 or Santo Forte) or the even more destitute folks
of Boca de Lixo, a lm which takes place in a garbage lot where people make
a living by scavenging. Coutinhos reexive self-consciousness manifests itself
in strong self-doubts, as he reects on the risks which haunt his lming of
the real as he nds it, but not truly as it is (outside his gaze). In an interview,
he explains how he takes pains not to include picturesque images (Caldeira
2001) in his work or how he avoids presenting the lmed materials in such a
way that a common theme, a funny or a sad effect, may emerge (e.g., two or
more successive episodes in which people sing, or talk about having tried to
commit suicide, in EM ). As an antidote against the temptation of cinematic
commonplaces, Coutinho submits himself to a strict, frugal economy of signs,
and he willingly sacrices any possible attention grabber in the nal version of
the lm.
With this aim, the director suppresses pictures which, in his view, would
make his lm similar to a normal TV interview as it occurs in the news or in
a talk show. Neither does Coutinho allow the insertion of any extra evidence
which was not originally present during the actual lming of the encounter,
as he chooses to describe the central goal that he pursues in his lms. Given that
it is so simple nowadays to insert an attractive picture or sound at the editing
stage, he makes a point of telling how he has to ght against this impulse.
172 Fernando Andacht

For example, Coutinho relates that he decided to omit pictures that two of the
women in EM found signicant because he had not reacted quickly enough,
and had not included them at that particular point in the lming process. In
allusion to the utter simplicity and care deployed by the German photographer
of the 1920s famous for his natural portraits which show the depicted people
in an environment which corresponds to their individuality7 , I have called this
procedure the August Sander effect (Andacht 2005: 113). According to this
frugal, reexive lm poetics, any later accretion does not pertain to the Others
individuality but is a sign of the directors will, of an arbitrary, subjective choice,
and not the result of the lmed interaction between the director and the lm
The third reexive mode is methodological self-criticism. According to
Lynch (2000: 29), this mode of self-criticism often seems to follow naturally
from self-consciousness and is not limited to confessional ethnography. In
Coutinhos poetics, this reexive method manifests itself as a relentless self-
critical gaze which can be traced back to an inuential gure in European doc-
umentary and whose importance for the ideology of the genre is acknowledged
by the Brazilian director, namely, the French lm critic and documentarist Jean-
Louis Comolli.8 In a long interview Comolli, the former editor in chief of the
Cahiers de Cinema, eschews the presumption of voyeurism for documentary
art by giving a detailed account of a method which I nd remarkably akin to
methodological reexivity and which also ts nicely Coutinhos own cinematic
There is always voyeurism in the operation of lming. But I wouldnt speak
of voyeurism in this case. The word does not seem adequate. Id rather
say there was a great force of listening [une force d ecoute]. What you
can feel in that lm is not just that people talk (as in television) but
it is rather that we feel the listening of someone. [. . . ] To listen is labor,
it is something which involves us and transforms us as well. There is an
analytical dimension in the lm because the analysis is likewise the place
of listening. When I say analysis, I am also speaking of transference [. . . ];9
things happen between those who are lmed and the one who is lming,
and this pertain to transference, which is possible through an act of listening
[ecoute]. (Comolli 1995: 68)

Similar to Coutinho who rejects as a subject matter for his lms the cunning
talk of the bourgeois class, of those who know only too well what they are about
but also what they do not want the Other to know about them,10 his ideological
French mentor Comolli expresses an eloquent critique of the frightening and
deadening effect of generalities, of the commonplaces of a social class, of its
certainties, of the dead semiotic weight which prevents the documentary from
On the use of self-disclosure as a mode of audiovisual reexivity 173

exploring untrodden ground. Evidently, Comolli does not only mean that those
with material and symbolic capital should be shunned as lm participants but
anyone who lacks spontaneity in front of the camera. The French director ex-
plains why he omitted from his 1993 documentary on ofce life and work11 the
scenes which featured a union worker, one of the very few men who appear in
the lm. The reason was the high risk of inserting a ready-made speech into
his lm which would have altered the overall quality of reexive inquiry for the
viewers and for himself: We realized that there was a ready-made discourse, a
sort of wooden language (because wooden language, unfortunately, is not only
with the powerful), a discourse in which there was no tremor, in which one could
not feel the crisis of a subject through which one could have tried to put in crisis
the one who is watching (Comolli 1995: 78).
There is still a further instance of reexivity in Comollis interview. After his
remarks that listening constitutes the center of the lm (1995: 78), Comolli
proposes a different, more radical kind of self-knowledge which for him is the
natural upshot of the reexive documentary genre:

What I nd interesting in people is that they bring me not just what had
attracted me in them but also what I did not know before, what I am dis-
covering while lming them and which is their way of thinking the lm.
[. . . ] Film functions always as a kind of revealing agency [une sorte de
revelateur]. Cinema is not an image of things; it creates an image different
from the image which you had before. That is what happens to the people
who have seen and heard themselves as they had never observed themselves
before. (Comolli 1995: 66, 70)

This view of the reexive documentary as a privileged semiotic agency brings

to bear a different kind of reexivity, namely, metatheoretical reexivity, which
Lynch (2000: 29), following Berger (1963), described as a matter of stepping
back from full engagement in cultural activity, and which is said to be em-
blematic of the sociological attitude. The French and the Brazilian lm makers
agree in that they both seek to make a discovery, to pursue an earnest audiovisual
inquiry which they pit against the already known and the already said and which
they believe to be characteristic of TV news interviews. This commercial format
seeks only to conrm what the journalists know well or think they know. The
Interviewers only have to repeat, after some rehearsal, what is expected of them,
in each situation. No epiphany may be envisaged the, and this lack, according to
the documentarists, is caused by the absence of reexivity in commercial TV, of
a method which runs upon the very work that is being carried out by the camera,
whose aim ought to be to problematize what otherwise threatens to become a
routine, production format.
174 Fernando Andacht

3. On the anguish of peptonizing reality in the self-critical


Let us consider once more the dualistic metaphysics underlying the poetics of
documentary representations of the real. William Jamess approach to reality will
help us to understand the remarkable ambivalence, even contradiction, between
the lm-makerspractice in their lms with their reexive strategies and their
reections on their own production. The forlorn sense of lack at the kernel lm
poetics derives from a basic misconstrual of the way in which representation
When asked whether the essence of a documentary is in very act of docu-
menting (Figueiroa et al. 2003: 21617), the reply that Coutinho gives deserves
to be part of any future manifesto of reexivity since it contains the warrant of
the lm makers inclusion of the process of lming in the nal documentary:
We are always lming encounters. [. . . ] It is the verbal act which is ex-
traordinary, an act provoked, catalyzed, by the moment of lming without a
conscious deliberation neither from me nor from the person. To lm is [. . . ]
to provoke, to catalyze that moment. It is in the interaction that takes place
in the lming process that a great character is born.

Just before that statement, the lm director had spoken about the reexivity
of his lms, which always tell that they are lms, always reveal, somehow, to
the spectator, their own conditions of production by revealing the presence of
the camera lming that encounter (Figueiroa et al. 2003: 215). Coutinho also
posits a critical contrast between his reexive method and the one of a related
but nonreexive audiovisual genre, the TV interview:
A [TV] interview tries hard to seem objective and supposedly to show the
real. The documentary, on the contrary, is shaped by the questioning
of that objectivity, of that possibility of dealing with the real. The great
documentary is not only based on this presupposition but it has also the very
impossibility of dealing with whatever might be called real as its issue. Faced
with this real, every documentary, deep down, is precarious, is incomplete,
is imperfect, and it is precisely from that imperfection that its perfection is
born. The documentary is always a subjective view. The documentary is the
very act of documenting. (Figueiroa et al. 2003: 216)

How can one of the most private human relations, namely face-to-face inter-
action, the encounter which Coutinho turns into the centerpiece of his art, be
not an experience of the real? Why does Coutinho speak of a character to refer
to the person who accepts to part of the lmed representation of life as it takes
place, whether it becomes the subject matter of a documentary or not? Is the
On the use of self-disclosure as a mode of audiovisual reexivity 175

real indeed something alien, something that is doomed to remain incompatible

with a faithful representation of it? Is the destiny of outward reality to be sadly
peptonized by its representations, to be altered so that all we can hope for is only
a meager, faked version of the real? The tentative answer I offer to this apparent
dilemma echoes in many respects Lynchs (2000: 42) critique of the excesses of
reexivity in social constructionist analysis.
In view of its relevance to the poetics of self-critical documentary, I now
propose to analyze the notion of encounter through its semiotic effects, or in-
terpretants, which are both verbal (the reections of the lm-makers) and au-
diovisual (the lm scenes). An example of the former is Coutinhos description
of his role as a catalyst of the interaction with the lm participants. The latter
may be exemplied by the opening pictures showing the arrival of the camera
crew as an intrusion led by the director himself, at the moment when they all
enter the building where they are going to live and lm for a month. All inter-
pretants of encounter construe this event as a shaping inuence. However, two
kinds of encounter must be distinguished on the basis of the ethical outcome of
documenting the real, encounter as interference, and encounter as shock.
Encounter as interference is the one which undergoes a negative evaluation; it
is the uncanny echo of the Jamesian concept of representation as the peptonizing
of the real. In the construal of the documentary as a mildly negative interference,
encounter is the upshot of the intervention of a foreign presence which sensibly
and irreversibly alters an environment, namely, the pro-lmic (Souriau 1953)
material of the documentary. In this perspective, there is the suspicion that the
lming of the real and thus the representation of the real, might peptonize it,
dissolve it, so that the real can no longer be grasped as it truly is. Such is
the forlorn outcome of making a documentary, according to the self-critical
tradition, regardless of how honest and respectful of the Other the director may
be. This lmic representation is conceived of as a construction, a frustrating
substitution of the real out there, which the lm only seems to show objectively,
but which it cannot but fake, as James stated in 1906.
This is the kind of encounter that Comolli has in mind, when he comments on
the work of the founding father of the documentary, Robert Flaherty, the director
of Nanook of the North (USA, 1922). Comolli (1995: 54) voices a skeptical,
anti-objectivist view typical of the self-critical reexive approach: The lm
really produces the reality, which it seems to show by legerdemain. But it is a
red herring, the bait of ction [. . . ] in the documentary; the viewers belief is
somehow warranted by the idea that reality exists.
In contrast to the dysphoric interpretation of the lmed encounter, the second
interpretation of encounter as shock, as described by Comolli, evaluates the
encounter as the most sought after achievement of this reexive genre. In his
176 Fernando Andacht

semiotic, Peirce (CP 8.266) has employed the notion of shock to account for
the experience of reaction, of the semiotic object, as that which offers resistance
or sets limits to the tasks of the interpreter. In conclusion of his construal of the
genre of the documentary as a powerful act of listening to the Other, Comolli
(1995: 71) says: As long as you listen to someone with intensity, there is always
a shock. To be shocked in a lm is therefore something positive. A lm is made
to destabilize the viewer. Comolli also speaks of the importance of the energy
which comes from the real people with whom the documentarist comes into
contact. This second interpretant of encounter fullls a goal of the documentary
which does not belong to its symbolic but to its indexical dimension: It is not
what is said that touches; it is always the presence of a subject within his own
utterance which will reach the other subject who is watching the lm (Comolli
1995: 76). In this interpretation, the documentary encounter functions as a
record of a unique and authentic contact with the participants. What is pursued
in the lming act is the shared here and now which engrosses the maker and
the participants of the lm. The purpose of the genre is to furnish a public trace
of the resistance, a shock which the encounter with another person inevitably
brings about.
The two distinct interpretants of the lmed encounter are the source of an
essential contradiction in the poetics of self-critical reexivity, which docu-
mentarists are trying to resolve by means of their self-referential practice of
self-disclosure through lming the act of lming and giving it a place of honor
in the nal work. The following reection may illustrate their self-conscious or-
deal. Coutinho recalls an occasion when he sat among the public in a rural town
in northern Brazil as he watched one of his documentaries. The lm was Santo
Forte which includes a scene of the payment of its participants. The director tells
us a revealing anecdote: There was a young woman, who came to talk to me
afterwards [and said:] I enjoyed it very much. When I saw the payment scene,
I thought it was all staged. I told her: If I was able to nd people who said
those things for that money, I would be a genius! (Caldeira 2001). Coutinho
justies the inclusion of the payment scene as follows: For some people, it
takes away a bit of the poetry, of the poetical atmosphere of that encounter. It is
rather tough but real to say of that man, who was wonderful, that he was paid
but not to say that he was paid to talk in front of the camera. By the inclusion
of what is normally unseen, the director has self-referentially exposed part of
the backstage as a way of disclosure of the lms inuence on its subject.
It is ironical, though symptomatic, of our indexical audiovisual age that the
voiced suspicion of that viewer of Coutinhos lm brings out an unexpected but
undeniable kinship between the earnest, self-critical reexivity of documentaries
and the light, facetious reexivity of TV reality shows. A typical suspicion about
On the use of self-disclosure as a mode of audiovisual reexivity 177

what is going on, e.g., in Big Brother, is the following: all that transpires there
in both senses of the word, namely, the overall activity, and the index appeal that
generates a number of existential signs in connection with the conned bodies
of the participants all is based on a pre-arranged, secret script so that nothing
that is said or done in the house of the TV format ought to be taken as the
real thing. Documentarists found their way out of the dilemma of their reexive
genre: the crux of the matter is not a choice between a faked represented real
and the pure facts out there. It is the wholehearted dedication to the Other, the
heartfelt attempt to become a curator of the Others self which distinguishes the
indices recorded in a lm in a self-critical way from those gathered by the 24/7
observation TV program: When I am there lming, I am totally available, the
people feel it. At last they will be listened, [. . . ] but they dont even know me;
when they look at me, they just see someone there who has only eyes and ears
for them (Coutinho in Figueiroa et al. 2003: 226).
The fear of altering the real which the documentary seeks to represent by
peptonizing or faking it is unfounded. There is no more reliable albeit fallible
contact with the world than that which we attain in truth representations of it.
The people whom we watch and listen to as they are passionately being listened
to in Coutinhos or in Comollis lms do not become characters, ctional beings
due to their appearing in the documentary. This representation is but one more
relationship, albeit a very special one, in which the self evolves and reveals
certain of its aspects whose authenticity it is our endless task to nd out, as we
must in our everyday relations, wherever meaning grows.

4. Conclusion: Documentary reexivity in the age of mass

indexical representation

Reexivity has been hailed as a noble component of high culture and equally as a
valuable method of high theory. In the realm of culture, reexivity serves to make
explicit and to explore artistically the self-consciousness that a creators put into
their works of representation. EM, for example, begins with the pictures and the
noise of the lm crew entering the building. In the directors view, this moment
of interference entails an irreversible alteration of normal life, an alteration hard
to gauge but denitely transforming what is to be represented audiovisually.
Self-critical poetics assumes that what is jeopardized by the act of lming is
the objectivity of what is represented, i.e., the dwellers of the building Master
and the life stories they willingly offer. Reexivity is also part and parcel of an
indexical genre placed at the aesthetic antipodes. In Big Brother, the purpose
178 Fernando Andacht

of self-referential practices is not to dispel the illusion of reality fostered by

this program of 24/7 electronic surveillance but to enhance the entertainment
value of the format. Reexivity is central in a TV program that makes much
of the opposition between what is faked or a mere self-enactment and what is
the authentic semiotic transpiration given off by the bodies secluded in a social
space without almost any backstage (Andacht 2004: 126). Distinct as the two
are, indexical signs are at the basis of the aesthetic effect of both documentary
and reality show, and this semiotic ingredient is what motivates the practices of
reexivity in them.
The strategy of reexivity used by self-critical documentarists as a way of
dealing with the paradox of authentically depicting the Other by means of what
they deem to be a device of falsication, namely, representation, reveals indeed
the fallacy at the heart of their dualistic metaphysics. In the reections on their
own work, there is a serious, albeit theoretically awed, attempt to destroy the
illusion of the reality in their lmic representations and the narratives of every-
day existence. Far from demonstrating that documentaries depict nothing but
staged stories enacted by characters drawn from life, the careful work of these
artists reveals, like the Peircean rainbow image, that the signs they use are a
manifestation of both their producers and the shock of otherness.
Once again, Lynchs critique of unfounded uses of reexivity falling prey to
the reductionistic trap of dualism is relevant: If reexivity shines for nobody
in particular and its illumination is controlled by no special theory, method, or
subject position, it loses its metaphysical aura and becomes ordinary (Lynch
2000: 48).
Hopefully, this journey through the mirror-like land of media reexivity
that now ends has shown that relevant and valuable as reexivity certainly is
qua analytical tool in the sciences and in the arts, it is only reliable to the
degree that it helps us to understand the growth, complexity, and the dynamics
of sign action. If reexivity is used simply as one of the other three ways of
xing belief described by Peirce (CP 5.377), i.e., as a means of establishing its
own, self-willed superiority to other methods of understanding reality and its
representation, then it will only perpetuate some current fallacies such as the
impossibility of reliably representing the real in the genre of the documentary.
The attempts at representing the Other do not necessarily peptonize the real,
let alone fake it, as William James would have it. Instead, the indexical genre
discussed here endlessly translates the aspects of the reality which it represents
in qualitative, factual, and general ways. Thus, documentaries do reveal aspects
of the real in a way still little understood but which should not be obscured by a
reexive analyses which accounts for everything which is shown by these kinds
of representation as a ction or a mere fabrication.
On the use of self-disclosure as a mode of audiovisual reexivity 179


1. The present work was done with the support of a Research Grant of CNPq in Brazil
and is part of the ongoing research A representacao do real na epoca de sua espetac-
ularizacao midiatica.
2. For the basic analytical notions concerning these indexical genres see Andacht 2002,
2003, 2004, 2005.
3. For simplicitys sake and to avoid a certain redundancy in Allens notion of self-
reexivity, I will refer to this phenomenon throughout this work as reexivity.
4. See Colapietro (1998) for a ne account of this position in postmodern thinking as
opposed to Peircean semiotic realism.
5. Every mention of sign in this work should be taken as synonymous with represen-
6. According to a 1912 edition of the Webster Dictionary to peptonize is a transitive
verb which means To convert into peptone; to digest or dissolve by means of a
proteolytic ferment; as, peptonized food.
7. August Sander, in The Encyclopedia of Photography (1984). http://www.masters-of- (01.12.04).
8. There is a reference to Comolli as an authority gure for Coutinho in an interview
(2003: 217).
9. Reference is made here to Comollis 1993 lm La vraie vie (dans les bureaux).
10. I dont make lms on the rich, middle class, because one cant do that. They defend
themselves, they are the specialists. Im not interested in talking to a specialist,
because they have an image to defend. Then the specialist will certainly not say
things which may cause trouble for him, every specialist is boring in that sense
(Coutinho in Caldeira 2001).
11. See note 10 above for full reference.


Allen, Jeanne
1977 Self-reexivity in documentary. Cine-Tracts 1(2): 3743.
Andacht, Fernando
2002 Big brother te esta mirando. La irresistible atraccion de un reality show
global. In: Paiva Raquel (ed.), Etica, Cidadania e Imprensa, 63100.
Rio de Janeiro: Mauad.
2003 Uma aproximacao analtica do formato televisual do reality show Big
Brother. Galaxia. Revista Transdisciplinar de Comunicacao, Semioti-
ca, Cultura 6: 245264.
180 Fernando Andacht

2004 Fight, love and tears: An analysis of the reception of Big Brother in
Latin America. In: Ernest Mathijs and Janet Jones (eds.), Big Brother
International, 123139. London: Wallower Press.
2005 Duas variantes da representacao do real na cultura midiatica: O exor-
bitante Big Brother Brasil e o circunspeto Edifcio Master. Contem-
poranea: Revista de Comunicacao e Cultura 3(1): 95122.
Caldeira, Joao Bernardo
2001 Entrevista a Eduardo Coutinho. Curta Ocurta, 0201.asp (03.06.01).
Carroll, Noel
1996 From real to reel: Entangled in nonction lm. In: Noel Carroll (ed.),
Theorizing the Moving Image, 224252. Cambridge: Cambridge Uni-
versity Press.
Colapietro, Vincent
1998 Natural processes and historical practices. Semiotica 119: 10555.
Comolli, Jean-Louis
1995 Les questions de cinema de Jean-Louis Comolli. (Interview by Rene
Predal, Youri Deschamps, and Delphine Goupil). CinemAction 76:
Figueiroa, Alexandre, Claudio Bezerra and Yvana Fechine
2003 O documentario como encontro. Entrevista com o cineasta Eduardo
Coutinho. Galaxia. Revista Transdisciplinar de Comunicacao, Se-
mioti-ca, Cultura 6: 213232.
Hacking, Ian
1999 The Social Construction of What? Cambridge, MA: Harvard Univer-
sity Press.
Hardwick, Charles S. (ed.)
1977 Semiotic and Signics: Correspondence Between Charles S. Peirce
and Victoria Lady Welby. Bloomington: Indiana University Press.
James, William
[1906] 1963 Pragmatism and humanism. In: Pragmatism and Other Essays,
101117. New York: Washington Square Press.
Lynch, Michael
2000 Against reexivity as an academic virtue and source of privileged
knowledge. Theory, Culture, and Society 17(3): 2654.
Mattos, Carlos A.
2003 Eduardo Coutinho. O homem que caiu na real. Santa Maria da Feira:
Edicoes do Festival de Cinema de Santa Maria da Feira.
Noth, Winfried
2001 A auto-referencia na perspectiva da teoria dos sistemas e na semiotica.
Revista de Comunicacao e Linguagens 29: 1328.
On the use of self-disclosure as a mode of audiovisual reexivity 181

Peirce, Charles S.
19311958 Collected Papers of C. S. Peirce. Charles Hartshorne, Paul Weiss and
Arthur Burks (eds.). Cambridge, MA: Harvard University Press.
Quoted as CP.
Ransdell, Joseph
1986 On Peirces concept of iconic sign. In: Paul Bouissac, Michael Herzfeld
and Roland Posner (eds.), Iconicity: Essays on the Nature of Culture,
5174. Tubingen: Stauffenburg.
Sander, August
1984 The Encyclopedia of Photography,
Santaella, Lucia
2004 O metodo anticartesiano de C. S. Peirce. Sao Paulo: UNESP.
Souriau, Etienne
1953 Lunivers lmique. Paris: Flammarion.
The old in the new: Forms and functions of archive
material in the presentation of television history on

Joan K. Bleicher

In my research on the history of self-representation in the media, I have studied

self-reference with respect to genres, modes of representation, and sign systems,
and I have shown that self-reference is relevant to media reception. Television
tends to present its own history in its own programs. By means of its specic
visual language, the medium passes on its own history to the memory of its

1. The construction of collective memory by the media

Long before the present strong impact of the media, history and narration had
been interconnected as collective memories. Historians such as Reinhart Kossel-
leck (1990) and Hayden White (1990) have argued that history must be trans-
formed into stories to be transferred to individual and collective memories.
Television has been able to make use of a long tradition of visual and narrative
presentations of history. By means of stories, individual memories are con-
nected to the collective memory of social groups. The strong visual impact of
the blockbuster movies testies to the re-shaping of individual and collective
memories of historical events in the media. In contrast to lm, television is
much more concerned with constructing social relations between the individual
and society. Gerbner et al. have described this social function of television as

Television is a centralized system of storytelling. Its drama, commercials,

news, and other programs bring a relatively coherent system of images
and messages into every home. [. . . ] Television has become the primary
common source of socialization and everyday information (mostly in the
form of entertainment) of otherwise heterogeneous population. (Gerbner et
al. 1986: 18)
184 Joan K. Bleicher

Self-reference is closely related to, and intertwined with, intermediality. One

of the strong points of television in its competition with other media is in its
ability to visualize history. In contrast to books, which lack pictures and the
auditory channel, and unlike the way archive material is presented on the inter-
net, television can transform history into stories which convey a live experience
of past events. The role of the media in society is constructed by the media
system itself. By continuously referring and adding to the world view of others,
the media centralize such perspectives and are at the same time the source of
emotional experience.

2. The inuence of television modes of presentation on individual

and collective memory

One of the functions of language is to create memory. Television makes special

use of its visual language to create its own collective memory of media imagery
in combination with traditional methods of story telling.The story telling devices
of television serve to organize historical events along a chain of actions guided
by the intentions of special persons.
Media theorists such as Jan and Aleida Assmann (2003) have shown that the
concepts of monument and repetition are fundamental to the construction of
historical consciousness. Monuments serve to visualize history in architectural
space; repetition is a method of memorizing events which are narrated again and
again. Both methods of historical memorization are used to construct history in
the current formats of historytainment. Monuments are visual landmarks, signs
of historical signicance, which make use of the principle of repetition in space.
Their everlasting visibility marks their cultural relevance. The historical event
or personality is represented by the monument to be anchored in the collective
memory. The fall of the Berlin Wall on November 9, 1989, is an example of
how a monument can represent history. The scene has become an integral part
of many programs dealing with the history of the German Democratic Republic
Nostalgic television shows make use of fragments of narratives to create
pieces of memory. Short lm strips open time windows on television history
without any relation to other genres, programs, contents, or other developments
in media history. Such fragments, sometimes combined with slow motion pre-
sentation, serve to highlight the signicance of a specic historical moment,
which may thus be presented as a scenic monument in the long sequence of
past events. The principle of repetition guides the viewers orientation, and the
The old in the new 185

unwritten rule of the program, seen forever, known forever, seems to serve as
the key to effective communication.
Historytainment shows make use of fragments of historical memory by fo-
cusing on everyday objects popular in the past, such as cars, toys, furniture,
or clothes. Various discourses of national history are combined. In its shaping
of the history of the GDR, for example, sports events of the past are used as
elements in the creation of a GDR identity, and winners of past Olympic Games,
such as Katharina Witt, are stylized as national heroes. The focus is on popular
culture, not on politics, and the past is used as a source of present entertainment.

3. General modes of representing history in television genres

Its all about entertainment,

and you know it.
(Tom Kummer)

The representation of history on television involves self-reference insofar as

the means of presenting history are historical television documents and history
includes the history of television. In the evolution of narrative media, television,
especially the documentary genre, has emerged as the predominant format of
visualizing history since it is able to use its own archive material in repetitive
presentations. Documentaries dealing with historical topics combine elements
of visual presentation with forms of oral narration which include commentaries
as well as personal observation. The messages from archive material of diverse
origins are presented as viewing strips, a term coined by Horace Newcomb
(1994) to describe the visual fragments which television makes use of. The
construction of history on television follows a stereotypical pattern which can
be found in documentaries as well as in show formats; it is characterized by the
stereotypical sequence of the following four elements:
1. viewing strip from archive material with voice-over narration,
2. oral narration by eye witnesses,
3. comments by eye witnesses or experts (historians),
4. viewing strip from archive material.
In general, documentaries use this stereotypical sequence in their presentation
of history as television history, the history of daily life, and of pop culture in new
formats. Recently, diverse program formats of historytainment have grown in
popularity in German television. A typical example is Die Burg [The Castle],
186 Joan K. Bleicher

ProSieben, in which well-known or less well-known celebrities are obliged to

live a medieval lifestyle. Many formats have made use of fragments of archive
material to convey nostalgic memories, especially in shows involving the history
of the GDR.

4. Narrative presentation of history in information formats

My studies of TV programs have revealed that narration has become increasingly

relevant as a means of presenting history on television (cf. Bleicher 2003).
Since the 1990s, documentaries on episodes from the history of mankind or
documentaries dealing with epochs such as the Third Reich in Germany have
been included in the news and magazine formats in charge of presenting current
events. In addition to documentaries, there are the so-called living history
formats, in which television uses its own forms of narration in the presentation
of history as the present-day experience of a serial world.
Docutainment formats present episodes from natural history such as the
genealogy of the dinosaur. In such documentaries, information is conveyed by
means of narration, and history is presented as a construction which can be
segmented into chains of events connected by the principle of cause and effect.
Historical facts and elements of ction are continually mixed up. In this way,
new documentary formats do not focus on what actually happened but on what
could have happened in the past. The result is counterfactual history as Guido
Knopp, editor of the history department in the CNN network has called it,
the presentation of what we have been spared. Archive material is used to
represent what did not happen in order to convey a sense of collective relief
to those who imagine themselves victims of history. On December 1, 1998,
for example, the German public television network ZDF aired the documentary
The Third World War. Real pictures of real events related events which had
never taken place. In this ctional presentation pseudo-historical monuments
were set up as props to convey the impression of historical authenticity, and
facts were used to stage ction. Such are the mixtures of information, parody,
and entertainment in diverse formats of historytainment which testify to the
general shift from fact to ction on television.
Not only has the borderline between fact and ction become blurred but
the differences between television genres are also disappearing. Various live
history formats have been using features of lm and television genres as props
or structural elements in their presentations of real-life stories. In diverse en-
tertainment programs, staging is assuming the role which archive material used
to play in documentaries. Similar to the aforementioned documentaries, the
The old in the new 187

show formats of historytainment make use of elements of personalization. A

frequent device is the palimpsestic overlay of pictures of media stars featuring
as commentators on strip scenes running in the background of the television
screen. By such means, the past is represented in a subjective and occasionally
even nostalgic light; sometimes, there is even some ironic distancing when new
perspectives on well-known material are presented.
In Germany, public television has offered a mixture of documentaries with
traditional lm genres, like heimatlms, the sentimental lm genre in idealized
regional settings of the 1950s, or series such as Das Schwarzwaldhaus 1902
[The Black Forest House 1902], Abenteuer 1900 Leben im Gutshaus [Ad-
venture 1900 Life in the Manor House], both ARD, or Die harte Schule der
50er Jahre [Tough School of the 1950s], ZDF. Reality show formats such as
Die Alm [The Alpine Pasture] or Die Burg (see above), have presented a
mixture of movies, shows, and computer games in which the viewers could par-
ticipate via telephone to substitute members of the cast like puppets in a Punch
and Judy show. By such devices, television transmutes into computer game and
can also meet the demands of the new competitor in the media system, the

5. The inuence of television on social discourse

Television discourse is in many ways related to other modes of discourse in mod-

ern society. In television archives, many kinds of discourse of all periods of the
history of the medium are assembled. From the perspective of discourse theory,
such archives can be described as systems of forming and transforming existing
assertions (Foucault 1981: 188). How does the highly inuential medium of
television inuence its audience when presenting its archive material from the
past? Television programs aim at mirroring the present-day world in a complex
system of narratives, and its never-ending ow of narration keeps transforming
the present into past and the past into the present.

6. Time spheres of history on television

The presentation of history on television concerns two time spheres, the short-
term history of very recent events and the long-term history of events of past
decades. Examples of short-term historical shows on German television are
The Best of. . . , Zapping, Switch, Kalkofes Mattscheibe [K.s Tube], or
TV Total. Among the long-term historical shows are Die 100 nervigsten TV
188 Joan K. Bleicher

Shows [The 100 Most Annoying TV Shows], ProSieben, even though this show
mainly focuses on more recent shows, Kenn ich die witzigste Serienshow
[I Know The Funniest Series Show], Kabel 1, Schwarzwaldhaus 1902, or
Abenteuer 1900 Leben im Gutshaus (see above). Short-term history is evi-
dently presented in show and reality formats, whereas long-term history appears
in hybrid formats of reality and ction genres such as the live history series.

7. Self-presentation of television and the collective memory

of society

Based on the above observations, let us now reect on the difference between
self-presentation and reality presentation on television. Luhmann (1996) has
agued that entertainment programs construct a limited world of its own existing
apart from the so-called objective reality while the program is on the air. Bau-
drillard (1978) went further with his claim that media reality creates a separation
between the individual and the so-called objective reality. The visible reality of
the shows implies the invisibility of the objective reality. Thus, it seems to be
an important task for media research to analyze the principles on which media
reality is founded.
According to Foucault (1981), archives create world views from existing
resources of collective memory. Archive material presented on television staging
history certainly shapes the collective memory of our times. Television has
assumed the role of a collective archive of society. Against the threat of its most
recent competitor, the internet, television is trying everything to maintain its
central position in the media system.

8. Contexts and economic strategies

The presentation of history on television can be studied in three respects, (1)

economic contexts and strategies, (2) general modes of presentation, and (3)
programming trends, especially in program formats such as nostalgic shows or
ranking shows.
As far as the economic contexts and strategies are concerned, the numbers of
channels in the German media scene have been increasing since the 1990s, cli-
maxing in an excess of verbal and pictorial content. In order to attract and keep
the viewers attention, network marketing has to develop unique selling proposi-
tions, for example by promising experience values in slogans such as Powered
The old in the new 189

by Emotion, Sat.1, or Total Entertainment, ProSieben. For the realization of

the goals of such slogans, programs are needed to fulll the expectations they
From the perspective of media companies and networks, archive material is
a source for the reproduction of existing programs with material almost free
of charge. One market-orientated form of reproduction is the presentation of
network history in documentaries and shows. Television past is used as a brand
image to sell the present programming to advertising companies as well as to
the viewers. Program history and program marketing are closely interlinked in
a process of telling and selling.

9. The archive as the source of self-reference on television

In my recent research, I have been investigating the relevance of self-reference

in the domain of television entertainment. One of my special interests has been
to study the strategies and forms of self-reference as methods of reproducing
and hence renewing existing entertainment values. The present article focuses
on the archive as the source and material basis of these various formats. The
archive as the collective memory of television is used both for the reproduction
and continuous ow of attractive images. But there are many other functions.

10. Looking back in anger: Short-time TV history as a form

of media criticism

The comedy shows Kalkofes Mattscheibe [K.s Tube], premiered in 1994, and
since 2002 Kalkofe Sport (ARD, since 2003 ProSieben) present highlights of
the worst moments on television programs during the past few weeks. The tele-
vision studio is used as a stage for television criticism. Kalkofes Mattscheibe
mainly presents and restages original viewing strips. Oliver Kalkofe, dressed in
the same costume as the original television announcer, reporter, or show master,
steps forward and parodies, by exaggeration as well as commentaries, what has
been shown before on television. In this way, he generates a palimpsest rewriting
the old visual material with his verbal attacks and his impressive body.
Combining archive material with his own comments, Kalkofe uses elements
of humor known from television comedy as well as the aggressiveness of his
critical comments to reveal the absurdity of the situation he satirizes (cf. Lam-
bernd 2000: 37). His acting and talking inserted into the original scenario is in
190 Joan K. Bleicher

sharp contrast with the message of the program he satirizes and its original pro-
duction values. The recycling of the original material deconstructs its intended
meaning by means of the critical comments which served to replace it.
Similar forms of combining archive material with critical comments char-
acterize the comedy shows TV Total, ProSieben, and Kruger sieht alles [K.
Watches/Sees Everything], RTL. Viewing strips have their origins in diverse for-
mats of different networks. In their commentaries, the comedians Stefan Raab
and Mike Kruger criticize and sometimes even insult the original programs from
which their viewing strips are taken. Such viewing strips function as quotations
in a more general television scheme of self-criticism (cf. Bleicher 2005).
Considering the current phase of transformation of television, such forms
of critical presentation may mark the end of an old, and the hailing of a new
kind of program. Kalkofes criticism of traditional show formats has prepared
the audience for new kinds of reality shows, which have meanwhile become a
new focus of Kalkofes critical eye.

11. Program history as network history?

From the perspective of media economy, nostalgic shows serve the purpose of
constructing a memory network related to the collective memory. The principle
of chronology of program history is abandoned in favor of a system representing
isolated moments of specic emotionalized event as well as experiences related
to a network. Such emotions are essential to a network marketing based on
slogans like Powered by Emotion (Sat. 1). Networks create their own canons
of what they consider to be the important events in television history. Even the
time of the event presented in a program is closely related to the history of the
network presenting the program.
Self-reference in television is not restricted to the presentation of televi-
sion history on television. In all historytainment shows, television uses its own
archive, its own modes of presentation as a rich source of collective and indi-
vidual memories. In this way, television tries to establish itself as the collective
memory of society.

12. Forms and functions of nostalgic show formats

The presentation of history in new entertaining formats overwrites traditional

construction forms of objective history with its entertaining reproduction of
historical events of daily life. History is evaluated in terms of nostalgia or also
in negative terms, as in Die 100 nervigsten TV Shows (see above).
The old in the new 191

Nostalgic show formats, which started to spread in diverse programs during

the last decade, create a connection between the collective memory of television
with the viewers subjective memory. The additional element of interactivity
created by cross-media relations is of special importance. On the webpage of
the network ProSieben, for example, viewers can vote for the most annoying TV
show. Such forms of connecting the two media to create interactivity allow the
extension of the specic media aesthetics of television with the one of internet.
Francesco Casetti and Roger Odin (2002) have described interactive show
concepts as the essential difference between the old and the new television.
Traditional forms of intentional communication have been replaced by the in-
troduction of interactive processes:

In paleotelevision, the program ow [. . . ] presents itself as a continuous

sequence of programs each of them functioning according to a specic
communicative contract. [. . . ] Neo-television breaks with the educational
communication model of paleo-television. One of the most obvious aspects
of this change is the explicit refusal of all forms of intentional communica-
tion as well as the introduction of interactive processes. (Casetti and Odin
2002: 31214)

The shows that present history in negative terms evoke elements of both visual
and verbal memory in a nostalgic vein accompanied by ironic commentaries.
The commentators are voice-over narrators, VIPs shown in a blue screen studio
while watching the archive material that is being presented, or host in dialogue
with prominent guests in front of a studio audience.

13. Strategies of creating humorous effects

There are various devices of creating humor on self-referential television. Parody

creates humor by exaggerating the messages of everyday television programs.
In Die 100 nervigsten TV Shows, for example, commentaries are substituted
by verbal description of the visual material being presented. Redundancy caused
by a host is another source of humor. The contrast between the content of the
original message and its actual presentation with ironic comments is the source
of humor in historytainment. VIPs chatting about old TV shows fraternize with
the audience in their nostalgic reminiscences of past times on television, self-
ironically in mutual agreement about the missing quality of the television shows
of former times. Self-evaluation is a popular variant of humorous self-reference.
In Die 100 nervigsten TV Shows, television history is evaluated by jurors in
the guise of the choir of ancient Greek drama. Its members are VIPs known from
192 Joan K. Bleicher

other television shows of the same network. The presentation and criticism of
bad shows in the past suggests news about what will be a good show in the
future. Using ranking numbers, subjective comments of the past are changed
into objective evaluations.
Archive material acquires new meaning when old pictures are accompanied
by funny or ironic commentaries. The resulting contrast between the original
meaning and the message of its ironic commentary is another source of humor.
Quite often, such ironical commentaries serve to distort the original message,
for example, by offering a sexual reading of the viewing strip completely absent
from the original. By such means, new meanings are generated for no other
purpose than attracting the viewers attention.

14. Conclusion

The trends in the development of contemporary television formats described in

this paper can be related to general changes in programming which date back to
the 1990s. Television, which used to present itself as a window to the world in
the 1950s, has gradually transmuted to a mirror of its own world construction
during the past decade. The events presented in the programs are no longer
sought in the world outside the medium but taken from the world of television
itself, and instead of supplying forms of knowledge, television is supplying
forms of experience in programs offering emotional involvement.
Self-reference on television is of special relevance to the theory of media
effects. Faced with the uncertainties of current social transformations, the au-
dience demands paths of nostalgic escape into a safe past. This public demand
is met by the offer of media formats relating the past to the present. History as
a way of conveying meaning and form to bygone events is used as a source of
orientation in the past and in the present; it is a way of structuring the present
which offers orientation in many spheres of life.


Assmann, Aleida
2003 Erinnerungsraume. Formen undWandlungen des kulturellen Gedacht-
nis. Munich: Beck.
Baudrillard, Jean
1978 Agonie des Realen. Berlin: Merve.
Bleicher, Joan Kristin
2003 Darstellungsformen von Mediengeschichte im Fernsehen. Medien &
Kommunikationswissenschaft 51(34): 366281.
The old in the new 193

2005 Grenzgange zwischen Kritik und Humor. Fernsehkritik im Fernsehen.

In: Michael Beuthner and Stephan Weichert (eds.), Die Selbstbeobach-
tungsfalle. Grenzen und Grenzgange des Medienjournalismus, 127
146. Wiesbaden: Verlag fur Sozialwissenschaften.
Casetti, Francesco and Roger Odin
2002 Vom Palao- zum Neo-Fernsehen. In: RalfAdelmann, Jan-Otmar-Hesse
and Judith Keilbach (eds.), Grundlagentexte zur Fernsehwissenschaft.
Theorie Geschichte Analyse, 282311. Constance: UVK.
Foucault, Michel
1981 Archaologie des Wissens. Frankfurt: Suhrkamp. Engl. 1990. Arche-
ology of Knowledge. London: Routledge.
Gerbner, George, Larry Gross, Michael Morgan and Nancy Signorielli
1986 Living with television: The dynamics of the cultivation process. In:
Jennings Bryant and Dolf Zillmann (eds.), Perspectives on Media
Effects, 1740. Hillsdale, NJ: Lawrence Erlbaum.
Kosselleck, Reinhart and Wolf-Dieter Stempel (eds.)
1990 Geschichte, Ereignis und Erzahlung. (Politik und Hermeneutik 5.)
Heidelberg: Fink.
Lambernd, Jochen
2000 Die sieben As humoristischer Attraktivitat. Grimme 1: 3538.
Luhmann, Niklas
1996 Die Realitat der Massenmedien. Opladen: Westdeutscher Verlag.
Newcomb, Horace and Paul Hirsch
1994 Television as a cultural forum. In: Horace Newcomb (ed.), Television:
The Critical View. New York: Oxford University Press.
White, Hayden
1990 Die Bedeutung der Form. Erzahlstrukturen in der Geschichtsschrei-
bung. Frankfurt: Fischer.
Theres no business without show-business:
Self-reference as self-promotion

Karin Puhringer and Gabriele Siegert

In this paper on self-reference as self-promotion in the media, we are not arguing

from a linguistic point of view but rather from a communication studies, or more
precisely, from a media economics perspective. Our interest is not only to deter-
mine the many forms and varieties of self-thematization in the media but also
to examine the responsible economic agents and their motives in this context.
Our characterization of the media is a classical one. The media are the
journalistic output providing topics for public communication. We agree with
Siegfried J. Schmidt (1996) that the concept of media comprises a variety of
phenomena, such as materials of communication and information (e.g., news-
papers), organizations producing, supplying, and distributing the media (e.g.,
publishing houses or broadcast stations), and often also this media system as
a whole. According to this concept, thematization occurs in quite different do-
mains of the media. Thus, self-reference or self-thematization in the media con-
cerns media organization, that is, the newspapers or journals, the radio or tele-
vision stations and programs, as well as the media system as a whole, and these
various manifestations of the media mean rather different forms and formats.

1. Background

The media have become increasingly self-referential. In advertising, for ex-

ample, self-thematization and self-promotion are constitutive elements of the
medium. Media pages in newspapers, commercial spots or ads for media prod-
ucts, TV parodies, and more and more program trailers testify to the omnipres-
ence of self-reference in the media.
It is a commonplace that we live in a media and information society. Hence,
the media (and its supply, i.e., programs and actors) are signicant in all spheres
of social life. If it is the function of the media to monitor society and to reect
this task, self-thematization is a direct consequence of this function because
by observing themselves, the media observe a signicant sector of social life.
196 Karin Puhringer and Gabriele Siegert

With their critical eye, the media strengthen the decision-making skills and the
citizensresponsibilities. In many important ways, the media act as mediators for
authorities by informing the population about what is happening in the world. In
his study of recipient-ratings on media coverage, Thomas Quast (1999) found
out that 96.5% of the persons asked quote audience-oriented media as their
source of knowledge and 63.7% report them to be their primary source. In other
words, almost everything we know about the media, we know from the media.
Different forms of self-reference evince different possibilities of develop-
ment on the market. Promotional types of self-thematization, especially in the
form of house ads (see below), benet from the current general conditions of
media markets. As far as media rms are concerned, the much discussed econ-
omization and commercialization of the media entail a general engagement in
activities aiming at prot and prot-orientation. The media want to make money
in the rst place. Return on investment (ROI) measures how much prot an owner
makes relative to the amount of investment required to make that return. Firms
seek the highest possible return on investment. Thus, when a media company
compares two competing investment options, it usually invests in the one with
the higher ROI potential. Furthermore, media rms are highly competitive about
audience favor and advertising investments.
Although strong competition cannot be found at all levels, economic pressure
has consequences. Media rms do not only want to produce economically; they
also tend to be active as to marketing measures. The cost orientation calls for self-
reference. Using archive materials, for example, is a means of cost saving, since
the pictures need not be bought elsewhere (cf. Bleicher 1994, 1999 and this vol.).
Even though media companies make great efforts to distinguish themselves
from their competitors and communicate in which respect they differ from them,
it is not easy to express these differences, especially when products and per-
formances are very similar. Furthermore, media companies are often accused
of differentiating their own products insufciently from other products. When
media companies evince too little product differentiation, only the strategy of
communication remains as a means of accentuating the existing distinctions in
programming and the resultant added value for the consumers. Brand identica-
tion is another important potential for differentiation. Consumers tend to develop
brand loyalty over time, and therefore, having a clearly dened brand is a long
term advantage. According to Jacobs and Klein (2002), media managers began
only in the mid-1990s to develop distinct brand identities in order to differentiate
their stations, networks, and publications from those of their competitors.
Due to the information overload in present-day information society, attention
has become a rare commodity, so that it is decisive for the survival of competi-
tors to attract as much consumers attention as possible. The situation is further
Theres no business without show-business: Self-reference as self-promotion 197

aggravated by a basic media economic principle: media products are experience

and credence goods. Therefore, the recipients possibilities of evaluating media
supply are restricted.The quality of an entertainment show, for example, can only
be assessed after watching the program; the evaluation of the quality of a news
program is practically impossible as long we know nothing else about the events
reported in it.Thus, the consumersexperience of the product gives only a limited
possibility of deciding which channel or program to select. For this reason, media
organizations offer, as an alternative, information about the quality of their prod-
ucts, for example by signaling and creating expectations, trust, and credibility.
To sum up, due to the economic circumstances and the professionalism in
the methods of media marketing, it has become crucial for media companies
to introduce a mix of communication instruments including direct advertising,
personal selling, sales promotion, trade fairs, or exhibitions as well as sponsoring
and event marketing. Our interest in this study lies in widely used formats, such
as media house advertising (see below), cross-promotion, media PR (public
relations), or editorial references. These formats are not always clearly dened
and they differ considerably as to their levels and strategies of self-thematization.
Figure 1 represents these levels, in the form of a scale extending from low to
high degrees of self-reference (Siegert and Puhringer 2001: 255).

Figure 1. Self-thematization in selected instruments of communication

Depending on the stage of the development of the station and its programs,
the instruments of communication vary. In the companys phase of foundation
when its brand is still relatively unknown, more external media will be used. In
subsequent phases of development, the company tends to use more frequently
house ads or editorial references. In its period of consolidation, advertising
spending can be reduced because the company can use its own media as vehicle
for advertising. Branding, a key concept in media economics, is used by the
media companies as a way of creating identity awareness in connection with the
content of the products. Most audiences and most advertisers recognize brands,
and for this reason, larger media companies have invested billions of dollars to
develop and acquire the different brands on the market (Albarran 2004: 300).
198 Karin Puhringer and Gabriele Siegert

House ads are advertisements which the company positions in its own media
supply, in spots for its own brand, programs, shows, titles, stars, show masters, or
news casters. House ads are self-referential insofar as advertisers, the advertising
vehicle or medium, and the object of advertising are, or belong to, one and
the same media. As Karstens and Schutte (1999: 109) demonstrate, television
is both a media product and a media for advertising this product. The same
can be said of most other media. Among the various forms of self-reference
by means of house ads, two types can be distinguished, the more informative
ones and the more manipulative ones. Station promos evidently have a strong
promotional character, but it is more difcult to describe, for example, the forms
of self-reference of trailers and teasers. Their function, by the way, is similar
to newspaper editorials. Trailers, teasers as well as newspaper editorials inform
and give orientation, and they are also rather manipulative. According to Siegert
and Puhringer (2001: 261262), two subtypes of house ads can be distinguished,
house ads with program-reference and house ads without program reference.
Forms of house ads with direct program reference are:
teaser before and after commercial breaks; references to the following pro-
gram, to commercials, or to other forms of intermission
teaser in split screen, e.g., during end titles; visual/verbal (voice over) ref-
erence to next, daily, monthly, or other programs
episode or serial trailer: reference to next serial or next newscast
traditional program announcement (separate from the program)
trailer: has replaced traditional program announcement announcing daily or
weekly program
horizontal trailer: weekly or monthly thematic orientation (no particular
Forms of house ads without direct program-reference are:
passage: separates program from commercials before and after breaks
station promos: image advertising to create awareness, identity, and relation-
merchandising spots: advertising for articles or services of the broadcast
event advertising for organized or co-organized events of various kinds (e.g.,
cultural or sporting events)
consumer invitation for consumer participation: give us a call, visit our
website, etc.
Theres no business without show-business: Self-reference as self-promotion 199

2. Research questions

According to the results of previous studies, it remains unclear to which degree

house ads lead to a fragmentation of the TV program in which they are inserted.
Our studies are hence guided by the following research questions:
RQ1: Which house ads are presented by the different broadcast stations?
RQ2: To what extent are house ads introduced on daily TV program?
RQ3: Did the frequency and features of house ads change between 2000 and

3. Methods of Study 1 and Study 2

Study 1: In 1999/2000, in a rst content analytic study, we examined approxi-

mately 240 hours of TV program of twice 24 hours from a midweek day and a
weekend day of the following television stations: ORF 1, ORF 2 (two Austrian
public-service broadcasters), ARD (Germany, public-service broadcaster), RTL
and ProSieben (two German commercial broadcasters).

Study 2: The second study of 2005 was a 250 hour content analysis study of the
program of eight Swiss TV stations, the public-service stations SF 1 and SF 2
(German), TSR 1 (French), and TSI 1 (Italian), and the commercial stations Tele
Zuri (German), Leman Bleu (French), Tele Ticino (Italian), and Sat.1 Switzer-
land (from Germany with Swiss license). This research was commissioned and
nanced by the Swiss Federal Ofce for Communication (BAKOM).

4. Findings of Study 1

In Study 1 we found 1,365 units of content with self-referential features outside

the actual program, which can be divided into units with and units without
program reference (Figure 2).
The self-referential breaks and intermissions in commercial broadcasters
are between two and three times as many as with public-service broadcast-
ers (ORF 1, ORF 2, ARD). However, there is no major difference between
public-service and commercial television stations as far as the use of form of
self-reference without program-reference is concerned. In all of the TV station
analyzed in this study there was a high frequency of program related forms of
self-reference. The reason behind these ndings is the stations increasing effort
to attract regular and loyal costumers for their programs in the future.
200 Karin Puhringer and Gabriele Siegert

Figure 2. Self-referential units with and without program reference (n = 1,365)

Based on the distinction between self-reference in house ads with and with-
out direct program reference introduced above, the study adopted the following
coding categories in its analysis of the units of self-reference: trailer and teaser,
trailer and teaser in split-screen, opener (the former program announcement),
passage, image and media spot (station promo), merchandising spot, and con-
sumer invitation (cf. Figure 3).

Figure 3. Dimensions and frequencies of self-referential house ads (n = 1,365)

As shown in Figure 3, the most frequently used types of self-referential house

ads are the trailer and teaser forms, followed by the passages. The frequency
of the indirectly program referential teasers and trailers reect the stations
efforts to address the recipients memories and develop their program loyalty.
The frequency of the directly program referential passages reects the (still)
Theres no business without show-business: Self-reference as self-promotion 201

high number of commercial breaks of the traditional kind whose beginning and
end the passages indicate. The number of almost every type of self-referential
house ads is signicantly higher for commercial television stations. For example,
RTL broadcasts nearly four times as many teasers and trailers than ORF 1 (240
to 67) or six times as many passages than ORF 2 (132 to 21).

5. Findings of Study 2

In Study 2 we found 2,713 units of content with characteristics of self-reference

outside the actual program. The results are similar to the ndings of Study 1, but
Study 2 also investigated television broadcasts themselves. We found 1,039 self-
referential units within the television programs but for reasons of comparability
with Study 1, the diagram of the frequency of the self-referential units in Study 2
(Figure 4) excludes the latter instances of self-referential advertising within the
With the exception of SF 2, the second German speaking public-service
broadcasting channel, the results of Study 2 show that there has been a reversal
of the trend documented in Study 1: In 2005, seven of eight stations show a
higher frequency of self-referential units of content without program reference
and a lower frequency of units with program reference, with the peak of Tele
Zuri, the commercial local TV station in the Zurich region.

Figure 4. Frequency of self-referential units with and without program-reference (n =


The programs of the commercial television stations evince 583 units with
program reference and 876 units without program reference. The public-service
202 Karin Puhringer and Gabriele Siegert

stations evince 519 units with program reference and 735 units without program
reference (Figure 4). Distinguishing the same coding categories as Figure 3 (of
Study 1), Figure 5 shows that the most frequently used types of self-referential
house ads found in the sample of 2,713 units are once again the trailer/teaser
and the passage.

Figure 5. Dimensions and frequencies of self-referential house ads (n = 2,713)

Figure 6 shows the development in the frequency of the use of the various
forms of self-referential and self-promotional house advertising and allows the
comparison of the frequencies of Study 1 (19992000) with those of Study 2
(2005). The data concerning the relative frequencies testify to a decrease of the
trailer and teaser type of self-reference and an increase of the passage type of
self-reference or self-promotion without program reference from 2000 to 2005.

Figure 6. Types of house ads in Study 1 and in Study 2

What are the answers to our three research questions in the face of these nd-
ings? RQ1 raised the question which types of self-referential house ads are used
by the different broadcast stations. Answers are given in Figure 3 and Figure 5
Theres no business without show-business: Self-reference as self-promotion 203

where eight different types are distinguished, trailer and teaser, trailer and teaser
in split-screen, opener (the former program announcement), passage, image and
media spot (station promo), merchandising spot, and consumer invitation. An-
swers to RQ2 concerning the frequency according to which the house ads are
introduced in daily television programs can equally be found in Figures 3 to 5
and in the ensuing comments.
Let us now propose answers to RQ3 concerning the development of the
features of self-reference in the house ads between 2000 and 2005. Despite the
differences in the development of the self-referential house ads between 2000
and 2005 in the stations investigated in both studies, the similarities remained
relatively stable as far as the comparison between public-service and commercial
broadcasters are concerned. Our ndings indicated a number of interesting new
trends in the use and the structural features of house ads of 2005 as compared
to those of 2000. For instance, the types of house ads changed from 2000 to
2005. In 2000, trailer and teaser was the most frequent format. In 2005 it was
displaced by the passage type, and this development conrms, as discussed
above, the trend to a more segmented and structured television program.
Overall, there has been a signicant increase in the form and frequency of
self-referential house advertising with the purpose of self-promotion. In 2000,
an average of 5.7 units were identied in any of the 240 hours of broadcasting
investigated in Study 1. In 2005, the weighted average rose to 10.3.

6. Conclusions

The media play an important part in our society, addressing citizens as consumers
and customers and providing a major source of knowledge.
The media compete for the audience as well as for advertising investments,
and the resulting pressures, among others, have consequences for the structures
of the daily program. The two studies presented in this paper have given evidence
that the pressure for commercial television stations is stronger than the one
for public-service stations. For this reason, commercial stations have a more
fragmented program structure and there has been a signicant increase in their
self-referential promotional content with the goal of creating brand awareness to
make brands distinctive. Study 2 has also shown an increase in the frequency of
program-integrated forms of self-reference for the purpose of self-promotion.
Like in similar general trends in advertising, this increase reects the great
efforts made to attract the consumers attention and to avoid their channel-
hopping during program breaks.
The two studies presented in this paper do not pretend to give a complete
picture of the many ways in which broadcasters have become self-referential,
204 Karin Puhringer and Gabriele Siegert

nor do they claim to offer conclusive evidence about current and future trends in
television, not even as far self-promotional house ads are concerned. Neverthe-
less, the paper may have given an instructive example and informative empirical
evidence of the role and the increasing importance of self-reference in the media.


Albarran, Alan B.
2004 Media economics. In: John D. Downing, Denis McQuail, Philip Schle-
singer and Ellen Wartella (eds.), The SAGE Handbook of Media Stud-
ies, 291307. Thousand Oaks, CA: Sage Publications.
Bleicher, Joan Kristin
1994 Das Fernsehen im Fernsehen. Zur Rolle von selbstreferentiellen Sen-
dungen im Programm. Medien + Erziehung 36(5): 295299.
1999 Unterhaltung in der Endlosschleife oder wie das Fernsehen mit sich
selbst spielt. In: Michael Latzer, Ursula Maier-Rabler, Gabriele Siegert
and Thomas Steinmaurer (eds.), Die Zukunft der Kommunikation.
Phanomene und Trends in der Informationsgesellschaft, 115128.
Innsbruck: StudienVerlag.
Jacobs, Ron D. and Robert A. Klein
2002 Cable marketing and promotion. In: Susan T. Eastman, Douglas A.
Ferguson and Robert A. Klein (eds.), Promotion and Marketing for
Broadcasting, Cable, and the Web, 127151. Boston, MA: Focal Press.
Karstens, Eric and Jorg Schutte
1999 Firma Fernsehen. Wie TV-Sender arbeiten. Reinbek: Rowohlt.
Quast, Thomas
1999 Reexive Medienberichterstattung in der Legitimation und Selbstre-
gulation des (Systems) Journalismus Selbstreferenz oder Selbst-
reverenz. In: Kurt Imhof, Otfried Jarren and Roger Blum (eds.), Zerfall
der Offentlichkeit?, 208223. Opladen: Westdeutscher Verlag.
Schmidt, Siegfried J.
1996 Die Welten der Medien. Grundlagen und Perspektiven der Medien-
beobachtung. Brunswick: Vieweg.
Siegert, Gabriele and Karin Puhringer
2001 Programm- und Eigenwerbung Narzissmus im Fernsehen. In: Ju-
lia Neissl, Gabriele Siegert and Rudi Renger (eds.), Cash und Con-
tent. Popularer Journalismus und mediale Selbstthematisierung als
Phanomene eines okonomisierten Mediensystems, 255301. Munich:
Reinhard Fischer.
Part VI. Self-referential games
Computer games: The epitome of self-reference

Lucia Santaella

Until recently, computer games used to be considered a vulgar kind of popular

entertainment. The only topic worth being discussed was the aggression or
even violence games were supposed to instigate in children. Today, computer
games have become the success of the market and a topic of media studies.
The growth rate of the games industry is astounding; it represents the third-
largest branch of industry in the world right after the war and the automobile
industries. The growth of this market is due to the wide appeal that games exert
on their players, mostly boys and young men, although in recent MMORPGs
(Massive Multiplayer Online-Role-Playing Games), a remarkable increase in the
percentage of female players has been observed. Games are greatly responsible
for the technological development of the whole entertainment business. The
games industry pioneers the use of advanced technological research, and it is
the rst to make these advances available to the public. In fact, the last decade
has been one of enormous creative experimentation and innovation in the games
industry (Jenkins 2004: 120).
As a reection of these indicators of their increasing cultural relevance,
games have become the topic of multidisciplinary research, in particular in
media and cultural studies. One of the rst places to embrace game design
and game culture as a subject of scholarly study was the MIT, where the
rst computer game, Space War, was created as an independent hack by com-
puter science PhD students. In a recent interview with Cicero Ignacio da Silva,
which circulated in the Rhizome site, the game researcher Wardrip-Fruin de-

Although I think this is changing, there is a sense that games were a kind
of Other, the separate thing in digital media. Games were very successful
commercially, but very uninteresting from an artistic or scholarly point of
view. So I also wanted to challenge that a little bit and say: Yes, games are
one of the most popular forms of digital media, but they are also interesting
art work, interesting writing, and this is happening and is related to games,
and I think that scholars and artists have to contribute to our discussion
about making and criticizing games. (Wardrip-Fruin 2005)
208 Lucia Santaella

According to Henry Jenkinss January 2001 presentation at the occasion of

a conference at the University of Southern California on Entertainment in the
Interactive Age, the most signicant evolutionary leap in the lm craft occurred
when people started writing about it (Pearce 2004: 143). This is exactly what
we can expect to happen with games.
One of the major topics of current theoretical discourse on games has been
the dispute between the ludologists, who focus on the mechanisms of game play,
and the narratologists, who declare that games should be studied in the context
of other storytelling media (Jenkins 2004: 118). Even authors who do not side
with any of the two theories cannot escape from participating in the discussion.
The present paper, however, is deliberately decentralized and peripheral as far as
the dispute between ludologists and narratologists is concerned. The argument
to be discussed is that self-reference is important to the understanding of the
semiotic processes involved in playing computer games. The argument to be
developed is that in computer games, self-reference is taken to its extremes:
computer games are the epitome of self-reference.

1. The increase of self-reference in digital culture

Self-reference occurs when a text or a sign process, in some respects, refers

to itself instead of referring to something beyond the message it conveys. In
Roman Jakobsons well known theory of language functions, the metalinguistic,
the poetic, and, to a certain extent, also the phatic functions of language evince
elements of self-referentiality, each of these functions in its own way. Self-
reference has not been invented in contemporary literary discourse, although
it has been considered a characteristic feature of postmodern culture. A well-
known classical example of literary self-reference is the play within the play in
Shakespeares Hamlet.
However, since the advent of mass media culture with the phenomenal growth
of the print media, the rise of mass advertising, followed by the invention of pho-
tography and the movies, radio and television, the occurrence of diverse types
of self-reference in the media grew considerably. The more the media multiply,
the more the interaction among them increases. The proliferation of the media
tends to accelerate the dynamical interchange among them, resulting in a myr-
iad of quotations, repetitions, intertextuality, and mutual references generating
the phenomena of intermediality and hybridization, that is, mixtures of texts,
discourses, and sign processes, which constitute one of the characteristics of
postmodern culture.
Computer games: The epitome of self-reference 209

With the rise of the information society, hybridization has reached its climax
with cyberspace. The most prominent semiotic concomitant of cyberspace is
media convergence. Distinct from the mere coexistence of media, media con-
vergence implies the integration of all kinds of individual media sound, image,
text, informatics programs within one and the same universal language, the
digital language. The result is in the convergence of the four main modes of
human communication media, the print media (such as newspapers, magazines,
books), the audiovisual media (such as television, video, cinema), telecommu-
nication (such as telephones, satellites, cables), and informatics (computers and
software) in a new complex media, the hypermedia. In this new space of inter-
change between and overlap of media and sign processes, self-referential sign
processes proliferate, leaving the impression that the media do nothing but talk
about media. As Noth states in the beginning of the introductory chapter to this
volume (paragr. 1),
the mediators have turned to representing representations. Instead of nar-
rating, they narrate how and why they narrate, instead of lming, they lm
that they lm the lming. The news are more and more about what has been
reported in the news, television shows are increasingly concerned with tele-
vision shows, and even advertising is no longer about products and services
but about advertising. The messages of the media are about messages of the
media, whose origin has become difcult to trace.
The three media in which self-reference is most prominent are undoubtedly
lm, advertising, and computer games.According to Noths introductory chapter
(paragr. 4),
the media differ as to the degree to which their messages are typically self-
referential or (allo)referential. [. . . ] Advertising is referential at its roots,
since it has the purpose of promoting and selling products or services.
[. . . ] A genuinely self-referential message would be unable to fulll its
commercial purpose of propagating a message about goods and services.
[. . . ] Feature lms, by contrast, which have both ctional and aesthetic
qualities, are referential and self-referential at the same time. While their
narrative plot is referential [. . . ] insofar as it narrates events from the lives
of its protagonists, their aesthetic devices are based on self-reference, and if
Lyotard [. . . ] was right when he proclaimed the end of the grand narratives, it
is only natural that self-reference in lms must have increased. In computer
games we are nally faced with a medium in which alloreference has been
secondary since its beginning, since playing and games create their own
self-referential worlds apart from the world of referential facts and realities.
A further aspect of the self-referentiality of computer games is addressed
by Azevedo, who points out that each new computer game tends to refer to the
210 Lucia Santaella

previous ones: Nowadays when a new game appears, it is almost impossible not
to compare it to other games to explain some of its characteristics, since games
are referring to each other all the time (in: Santaella 2005). A good example
is Matrix, whose games are a complement to the lm, expanding certain scenes
which were not explored in the lm (cf. Azevedo in: Santaella 2005).

2. The rules of the game

Without rules, no game is possible. Every game begins with a set of rules to
guide the players throughout the game in all its different states towards a goal.
Rules are the foundation of the meaning and structure of a game. A game has to
be self-explanatory; its rules are the elements that guarantee this requirement.
Rules are at the core of the design of a computer game. This implies that all
subject matter of a game has to be formalized and created as rules before the
game can start (Juul 2004: 141). The main difference between the computer
game and its nonelectronic precursors is that computer games add automation
and complexity they can uphold and calculate game rules on their own, thereby
allowing for richer game worlds (Juul 2004: 140).
All games obviously develop in time, but the electronic ones do so in a way
that their rules do not have to be necessarily explicit from the start; they can be
homeopathically inserted in the course of playing and with the players increas-
ing experience. In modern games with cut scenes1 such as American McGees
Alice, every mission that is accomplished by the player is rewarded with a cut
scene which gives the player information about the next task.
In storytelling games, Celia Pearce (2004: 145) has identied six different
narrative operators that can exist in a game. One of these operators is a rule
based story system, a kit of generic narrative elements which allows the players
to create their own narrative. Story systems can exist independently of, or in
conjunction with, a metastory. Metastories are obviously one of the types of
self-reference since they are stories about stories. They are rather frequent in
games, as we shall see later on. In the MMORPG genre,
players take actions that construct their character on the y. [. . . ] These
games, because they are highly improvisational in nature, require constant
attention from their operators. EverQuest, for example, has a Command
Central at its San Diego headquarters where its customer service staff wan-
ders about the virtual game world assisting the players, and creating narra-
tive events, conicts and missions for players to engage in. They carefully
watch what players are doing and constantly evolve the game, the game
rules, and the game narrative accordingly. [. . . ] The result is an emergent
Computer games: The epitome of self-reference 211

narrative, a story that evolves over time as a result of an interplay between

rules and players. (Pearce 2004: 149)
A game genre in which rules are presented in another interesting way is The
Sims, designed by Will Wright of Maxis. The Sims is a simulator game of a
psychological kind:
Rather than employing purely player-inhabited characters or purely auto-
nomous characters, the game puts the players in the role of inuencing semi-
autonomous characters.They are semi-autonomous because while they have
their own innate behaviors, they depend on player inuence to dictate their
actions. The view-point is isometric rather than rst person, allowing players
to have a God-like view over the game terrain. (Pearce 2004: 150)
The story system of this game is a kind of narrative Lego. The Sims has now
spun off into a variety of add-ons and enhancements, but the original game was
basically a domestic drama, a sitcom. The players create a family, The Sims, and
place them in a house that must then be enhanced and furnished with a variety
of domestic items to improve the familys standard of living and their level of
comfort. In addition to the story system, there is a built-in descriptive component
which allows players to take snapshots of their game whenever they wish. They
can then make descriptive storyboards and post them together with the snapshots
on the Sims website for others to view and participate. Players can also upload
their games onto the site so that other players can continue playing their game.
In this way, multiple versions of a players family story may come into existence
with developments leading into different directions according to the various
players. The story system lets the player drive the story experience within a
set of carefully crafted rules, processes, and constraints (Pearce 2004: 151).
Such rules guarantee that a socialized ctional construction with an open-ended
framework does not degenerate into chaos.

3. Internal indexicality of games

Associated with the rules is another important aspect of self-reference in games

which may be dened as its internal indexicality. The consecutive moves of a
game in their rule governed sequence are mutually connected in a way that each
present move or state of the game indicates the actually preceding and the possi-
bly or probably subsequent states of the game. Juul (2004: 132) describes these
determinants of the development of a game as follows: The more fundamen-
tal part of games is a change of state, the movement from the initial state (the
outcome has not been decided) to another state (the outcome has been decided).
212 Lucia Santaella

If the initial or any other state of the game anticipates the following state and
the following state retains the memory of the previous state, each of the states in
this sequence contains an index of the preceding and the following states which
it indicates, and since all of these indexical signs imply reference from one state
to another state of the game, these indices are self-referential. In the course of the
game, a continuous, not necessarily linear, set of indexical interconnections is
created. The process is similar to the effect of the constraints which the grammar
of a language creates in the process of speaking. For instance, in the utterance
The boy whom you see over there is my brother, the relative pronoun whom,
before indicating the boy over there, indicates the antecedent word boy in
the main clause of this utterance, and the former indication is linguistically self-
referential insofar as it is a reference from words to words, and not from words
to objects or persons.
A move is a transition from one to the next state in a game. A move is
hence a change of state. To extend the conception of the game as a sequence
of states, Juul, adopting the terminology of computer science, interprets the
computer game as a state machine: It is a system that can be in different states,
it contains input and output functions and denitions of what state and what
input will lead to what following state (2004: 133). Juul concludes that playing
a game means interacting with the state machine of the game. In a board game,
each state between two moves consists of the present position of the pieces on
the board; in sports, the game state is the players and the balls position in the
playing eld; in computer games, the game state consists of the of data presently
stored in the machine and the form of their representation on the screen.
To play a game, a player must be able to inuence the game state to per-
form a move, and in some games, the player is even obliged to do so without
risking a penalty. According to this difference between games that do not re-
strict the players freedom to move or not to move and those that require action
in a predetermined time frame, Juul (2004: 133) distinguishes between turn-
based and real-time games. In turn-based games, the game state only changes
when the player takes a turn. In a real-time game, not doing anything also has
consequences. In sum, playing a game means interacting with the game state.

4. The immersive and interactive conditions of games

The concepts of immersion and interactivity, relevant to the study of any game,
have become the focus of attention in the context of cybercultural studies. Im-
mersion seems to be a magic word able to explain everything that goes on in
cyberculture. Elsewhere, I have distinguished four forms and degrees of im-
Computer games: The epitome of self-reference 213

mersion (Santaella 2004: 4647). The rst degree, experienced in virtual reality
(VR) environments, is the one of perceptive immersion. VR gives the participant
the sensation of being inside the environment and of acting within the virtual
scenery. The second degree is telepresence, which is mediated by robotic sys-
tems conveying a feeling of presence in a distant location. The third degree is
representative immersion; it is obtained in environments constructed by means
of the VRML language. In representative immersion, the participant, mostly by
means of an avatar, is represented in the virtual environment of the screen. The
fourth and most frequent degree of immersion occurs when the user is connected
to the web. To get into the web means to navigate in an immaterial parallel world
which consists of bytes of data and particles of light.
In computer games, the most frequent mode of immersion is representative
immersion, but other modes of immersion are also involved in games. In a
general psychological and perceptual sense, not exclusively in the cybercultural
sense, immersion is a requirement of any kind of game. Players must concentrate
on the game, be absorbed in action and the planning of moves, be immersed
in the states of the state machine of their game. This means that the playing
of computer games involves two kinds of immersion operating simultaneously,
a deeper psychological and perceptive absorption, like in any other game, and
the immersion in the environment of a cyberworld. This double involvement
increases the players subjective sense of immersion and is probably one of the
reasons why computer games are so overwhelmingly attractive or even hypnotic.
The high concentration required by the players of computer games results
from the circumstance that players, as soon as they begin a game and become its
agents, enter a parallel, self-sufcient world whose sufciency increases with the
self-referentiality of its rules. The notion of parallel world does not necessarily
mean a world that is of an entirely articial design, as in computer games, whose
virtual environment is of a completely new design. What it means is that the
player has to enter another plan of reality, a ctive plan involving the pretense
of being a character in a story. Even the players of checkers, tennis, or of the
computer game Tetris are immersed in an autonomous world, a self-referential
parallel world, and even in a realistic computer game with a design imitating the
real-life environment of our everyday world the players are faced with a parallel
and self-referential world.
What matters in a game is neither the realism nor the ctionality of its
scenario. It does not matter whether it is a science ction story or as trivial as a
cartoon. Games do not have to make sense at all; they have to be enjoyable and
make fun. The more the players are immersed in their game, the more enjoyable
they will nd it. In fact, the immersion of players in their game is much deeper
214 Lucia Santaella

than the one of the movie spectators or novel readers because only games project
their players interactively into their world.
Interactivity is another term that has been much used in cyberculture studies.
The word is quite appropriate in this context since every computer interface
is an interactive program and to work with a computer requires by necessity
interactivity. Like in immersion, from which it is inseparable, interaction is
present in any computer game as well as in man-machine interaction in general.
Hence, there is a double interaction. It is not by chance that a great part of
the present discussion about computer interactivity involves the comparison of
game interactivity with the philosophical and anthropological notions of playing
(see Neitzel, this vol.). As in all games, a requirement of interaction is that the
players perform an act, such as moving a piece on a board or pressing a key on
the keyboard with which a specic meaning in the game world is associated. The
performance involves the players interaction with the game state in a process
in which one state refers to the next and so on.

5. Seven types of self-referentiality in games

After the discussion of the general framework of self-reference in computer

games, we will now propose a typology to distinguish seven types of self-
reference in games.

5.1. Commands, missions, and discontinuities

The rst and rather common type of self-reference in games occurs in the form
of commands. Commands may be considered a rudimentary version of the rules
of a game. They are indexically self-referential insofar as they are in a way
self-addressed. In contrast to a command such as Ground arms!, in which a
soldier is addressed by a superior, his corporal, who has a real authority over
him, the players, who are themselves the readers of the command directed at
them, are not really addressed by anybody form outside the game. In a game, the
order in a way, is given by the players to themselves, since they have submitted
themselves voluntarily to the rules of the game.
In modern pinball games, for instance, the basic rule is expressed in the
imperative form Hit all the ashing things! Such orders may also be given in
indirect or nonverbal ways. In Star Trek: Next Generation (1993), for example,
the order appears in the form of a small display telling the player to destroy
the asteroid, which threatens the ship, by hitting a ashing object with a ball
Computer games: The epitome of self-reference 215

(cf. Juul 2004: 140). As Juul points out, modern adventure games basically
follow the model of the detective game. There are not only cut scenes, but also
artifacts in the game world (event time) that tell the player what happened at
a previous point in event time (2004: 136). Hence, the discontinuous times
and worlds of these games point strongly to themselves as being games rather
than believable ctional environments (2004: 140).

5.2. Metagames: Games within games

Another elementary type of self-reference can be found in games within games

(cf. Rapp, this vol.). The embedding of a game within a game is iconically
self-referential like a picture in a picture, the device of mise-en-abyme. The
embedding game framework refers to the embedded framework and vice versa,
but since both frameworks are game frameworks, there is self-reference insofar
as a game refers to a game. A simple example can be found in Eraser Turnabout,
which is based on a lm of the same title. In some of the phases of the game, the
player has to solve puzzles in order to be able to continue playing. In The Sims,
one of the possible entertainments of the avatar is to play with a computer.

5.3. Metastories: Stories within stories

One of the most famous metastories of world literature is certainly A Thousand

and One Nights, the story of Queen Scheherazade who prevents her cruel hus-
band, King Shahryar, from killing her by amusing him with a different story
every night for a thousand and one nights. As in the case of games in games,
the self-referentiality of this framework story is an iconic one: A Thousand and
One Nights is a story about story-telling. The framework story with an outer
narrative frame of the Queen telling stories to the King in order to save her life
and an inner frame with 1,001 stories have the structure of a mise-en-abyme,
in which the outer narrative frame is an icon reecting its own inner narrative
In classical games, a good example of a story-in-a-story scenario can be
found in chess. As Pearce (2004: 147) observes, chess has a metastory of two
battling kings with their armies and minions. The outer framework of this game
is the story of the players from the rst move, whose meaning is attack, to the
last move, whose meaning is defeat or victory; the inner narrative framework
of chess is the one of the battle between the black and the white king and their
216 Lucia Santaella

In story telling computer games, the device of stories within stories is rather
frequent. Pearce (2004: 14849) gives the example of the MMORPGs. There is
metastory, primarily in the form of a predesigned story world and various plots
within it, with a story system that allows players to evolve their own narratives
within the games story framework. [. . . ] Most of the original MMORPGs
metastories focus on medieval fantasy/Dungeons and Dragons style themes,
although more mainstream themes are forthcoming.

5.4. The personalization of games

Mods are modiers which allow expanding the possibilities of a game. By

means of this device, the source code of a game is altered, and a new version
of the game is implemented. Mod packages of game expansion allow players
to personalize their game just as they like. The self-referential nature of a mod
consists in this device of personalizing the game. A game personalized by the
player is indexically self-referential to the degree that the personalized game
evinces traces of the players participation. Some games have been developed
from mod versions of previous games. For example, the action game Quake
Rally is a run game derived from a mod of Quake, or Half Life, a game about
modern life, allows a journey to the Second World War thanks the mod Day of

5.5. The games own materiality

A fth type of self-reference can be found in games which give evidence of their
own materiality. This type evinces hence again indexical self-reference. In more
recent games, such as Quake III or Counter-Strike, there are jumps between
different levels of the game which remain unexplained but are indicated by a
display referring to the digital processes going on in the game at the present
moment. For example, the display indicated loading goes on, or the display
indicates awaiting game state. Such messages are self-referential insofar as
they indicate what the machine is doing and not what is going on in the game

5.6. Intermediality

The sixth type of self-reference can be found in games which are related to texts
(lms, novels, advertising, music, etc.) in other media which, for their part, also
Computer games: The epitome of self-reference 217

refer to the game.The dialogue of games with other media, especially the movies,
is rather frequent. Many game designers have drawn plots or story elements from
existing lms or works of literature. Games are apt to tap those media since they
share with them the genres fantasy, adventure, science ction, horror, war, etc.
All games on the theme of medieval fantasy represent the evolution of about
forty years of popular culture converging on the computer. Semiotic translations
of works of literature are also frequent (Jenkins 2004: 122). These are cases of
semiotic translation because the games do not only retell a lm story but expand
our previous experience of a story and the way to interpret it.
Especially the interaction between lms and games is thriving. Not only
the marketing campaigns of lms and games but also their production and
development are becoming more and more interconnected and interdependent,
as Jenkins (2004: 124) points out:

Increasingly, we inhabit a world of transmedia storytelling, one that de-

pends less on each individual work being self-sufcient than on each work
contributing to a larger narrative economy. [. . . ] One can imagine games
taking their place within a larger narrative system with story information
communicated through books, lm, television, comics, and other media,
each doing what it does best, each a relatively autonomous experience, but
the richest understanding of the story world coming to those who follow
the narrative across the various channels. In such a system, what games do
best will almost certainly center around their ability to give concrete shape
to our memories and imaginings of the story world, creating an immersive
environment we can wander through and interact with.

A good example is Matrix, whose games are a complement to the lm and

expand certain scenes which were not explored in the lm (cf. Azevedo 2005).

5.7. A game theory of games

The seventh and last case of self-reference in games has been described by
Celia Pearce (2004) in her paper Towards a game theory of game in which
she develops a play-centered theory of games. Nothing could be more telling
of the argument of our own paper that computer games are the epitome of self-
reference than a game theory that is a theory of games.
218 Lucia Santaella


1. Editors note: According to Wikipedia, a cut scene is a sequence in a video game

over which the player has no control. Cut scenes are used to advance the plot, present
character development, and provide background information, atmosphere, dialogue
and clues. Cut scenes can either be animated or use live action footage
( scene, 26.04.06).


Jenkins, Henry
2004 Game design as narrative architecture. In: Noah Wardrip-Fruin and
Pat Harrigan (eds.), First Person. New Media as Story, Performance,
and Game, 118130. Cambridge, MA: MIT Press.
Juul, Jesper
2004 Introduction to game time. In: Noah Wardrip-Fruin and Pat Harrigan
(eds.), First Person. New Media as Story, Performance, and Game,
131142. Cambridge, MA: MIT Press.
Pearce, Celia
2004 Towards a game theory of game. In: Noah Wardrip-Fruin and Pat
Harrigan (eds.), First Person. New Media as Story, Performance, and
Game, 143153. Cambridge, MA: MIT Press.
Santaella, Lucia
2004 Navegar no ciberespaco: O perl cognitivo do leitor imersivo. Sao
Paulo: Paulus.
2005 E-mail interview with Theo Azevedo (July 7, 2005).
Wardrip-Fruin, Noah
2005 Games-language. An Interview with Noah Wardrip-Fruin by Cicero
Ignacio da Silva, (26.04.06).
Self-reference in computer games:
A formalistic approach

Bo Kampmann Walther

1. Introduction

As its title indicates, this chapter is about self-reference in computer games, and
a primarily formalistic approach will be adopted. I shall argue that computer
games can be self-referential in at least three ways. On a ctional or content
level, games often refer to other games or other types of media. The monsters
in Doom 3 pay tribute to the original scary polygon creatures in Doom. Villains
and good guys in Grand Theft Auto: San Andreas refer more or less to actual,
present-day characters in mass mediated pop culture. Plots and key actors in
Myst IV: Revelation t nicely into the overall cosmology of the much celebrated
adventure game Myst.
However, when we move beyond the sphere of narrative, plot, and vast game
worlds, we nd that computer games by themselves, on a structural or formal
level, are inherently self-referential as to their ontology. To put it bluntly: games
are games because they are fundamentally self-referential. To eliminate or fail
to recognize this highly specic and, to a large extent, technological and scien-
tic feature of computer games is to ignore the invariant base of the computer
medium. This is intimately correlated with the important concept of recursivity
(Noth, this vol., Part I) which explains computer games as mutually dependent
linear and circular systems. It also alludes to the fact that all games are dynamic
and temporal systems evolving, among other things, about the tension between
rules and strategies. In the following, the purpose is rst and foremost to inves-
tigate this innate feature of self-referentiality in computer games.1 The method
derives from economic game theory, computer science, and systems theory.
The chapter is divided into three parts. The rst denes and discusses the
core elements of any game, namely rules, strategies, and interaction patterns.
The second will examine how and to what extent computer games can be de-
ned as complex dynamic systems. The argument is that gaming is a higher level
220 Bo Kampmann Walther

activity incorporating the act of playing (hence the term game-play) into its
very structure. This second part draws on some previous work on the philosophy
of games (Walther 2003). According to Espen Aarseth (2003), there are three
components of games in virtual environments: game play is about the players
actions, strategies, and motives; game structure contains the rules, including
simulation rules and physics; and game world includes ctional content, level
design, textures, etc. All three aspects will be covered as they are intimately
interwoven, but special emphasis will given to game structure. First, the ontol-
ogy or gameness of games will be dened. Secondly, an explanation of the
epistemology of games (and play) will be given which pretends to zero the dis-
tinctiveness of game play. Thirdly, the chapter will seek to set out to illuminate
the level of self-referentiality in the form of recursivity in computer games by
paying attention to the relationship between rules and game world.
Before embarking on the formal journey into the heart of games, it should
not remain unnoticed that games are clearly self-referential also on a broader
cultural level2 which constitutes the third sense of self-reference in computer
games. Not only do games point to other more specic games while borrow-
ing themes, characters, plot, and back stories. As modern leisure artifacts and
carriers of intellectual value, they further subscribe to a wide ranging bricolage
culture in which texts, images, motion pictures, games, commercials, and brands
cite each other at a rapid pace. This citation praxis arises horizontally through
the instantaneous replication across the borders of various current media and
vertically through the reshaping of the old media. The former mode of citation
may be called transmediality, while the latter may be referred to as remediation
(Bolter and Grusin 1999; Walther 2005a).
A nice example of this dual mode of cultural self-reference in the present-day
media landscape is the television series 24 by the American Fox channel, star-
ring Kiefer Sutherland. Clearly, the series points towards a number of classic
issues and plot congurations imported from the history of television drama and
cinematic entertainment. This is remediation, old media rethought, recongured,
and, in a sense, made new again. In addition, however, 24 tells the story of the
way in which new media (TV, web, games, chat, etc.) speak to each other
along a synchronous axis, constantly pushing and blurring the demarcations
which used to dene the specicity of media and their contents hence offering a
rich platform for the users cut-paste-and-consume approach to media (Walther
2005b). This is hence transmediality. To come to a partial conclusion, computer
games are self-referential in three ways:
(1) Semantic self-reference: games refer to other content matter or ctional
elements in games and (new) media.
Self-reference in computer games: A formalistic approach 221

(2) Formal self-reference: the dynamics of recursivity is part of the genuine,

formal nature of games.
(3) Cultural self-reference: games can be self-referential as they point to a
surrounding culture of transmediality.

2. Rules, strategies, and interaction

Economic game theory is a set of mathematical methods of decision making

in which a competitive, risky situation is analyzed to determine the optimal
course of action for a player. John von Neumann and Oscar Morgenstern were
the founders of game theory. According to von Neumann and Morgenstern, a
game consists of a set of rules governing a competitive situation in which two to
n individuals or groups of individuals choose strategies designed to maximize
their own winnings or to minimize their opponents winnings. The rules specify
the possible actions for each player, the amount of information received by each
as play progresses, and the amounts won or lost in various situations. Neumann
and Morgenstern restricted their attention to games in which no player can gain
except at anothers expense (so-called zero-sum games).
Later, John F. Nash revolutionized game theory by demonstrating that in
noncooperative games there are sets of optimal strategies (so-called Nash equi-
libria) used by the players in a game such that no player can benet by unilat-
erally changing his or her strategy if the strategies of the other players remain
unchanged (Nash 1997).
Drawing on economic game theory we can now dene games as complex, rule
based interaction systems consisting of these three key mechanisms: absolute
rules, contingent strategies, and possible interaction patterns.
Game rules are absolute in the sense that while the players may question the
rationality of the rules at hand, they are nevertheless obliged to obey, to play by
the rules. Rules are therefore absolute commands (Neumann and Morgenstern
1953) and unquestionable imperatives. They transcend semantic issues, cultural
signication, moral agendas, etc. This does not, incidentally, preclude the fact
that game rules are discussed in a cultural or ethical milieu.
In contrast to rules, strategies are contingent, nonabsolute entities since they
count as the more or less detailed plans for the execution of turns, choices, and
actions in the game. Other strategies than the ones actually carried out could
have been outlined and performed. Both in the shape of short term tactics and
as long term schemes, strategies are contingent. In economic game theory, a
strategy is an overall plan for how to act in the assembly of different states that
the game may be in (Juul 2004: 56). Game theory studies the afliations of the
rules and the strategic behavior in competitive situations (Smith, 2006).
222 Bo Kampmann Walther

Finally, interaction patterns are the moves and choices which become part of
the game being played, thus interfering with the restrictions and options of the
game. As the implementation of game strategies tend to cluster in selected re-
gions of the possibility space of the game (in approximation of what is known as
the dominant strategy in game theory) forming a path through the game space,
we may even insinuate that the interaction patterns, taken as a whole, are the
game itself especially if we view it from the perspective of the player (Holland
1998). Interaction patterns are the possible as opposed to necessary combina-
tions or the emergent outcome of rules and strategies. This differentiation can
be listed even more briey:
Rules are commands.
Strategies are plans for game executions.
Interactions patterns dene the actual path through the game and specify the
topography of human-computer (or player vs. rule) dynamics.
Clearly, the interaction patterns work as middle ground as they occupy a
domain located between the machine that upholds the rules (the computer) and
the human player who has to nd and optimize the best way to accomplish the
goal of the strategy (Figure 1).

Figure 1. The relation between rules, interaction patterns, and strategies

The notion of game play, which we shall pursue in depth in the subsequent
section, involves all three levels of a game, which also explains the difculty
in dening the concept properly. Game play is the actualization of a specic
stratication of rules, strategies, and interactions as well as the realization of a
certain amalgamation of commands, plans, and paths. For a player, a successful
game play means a delicate balance between knowing the rules and mapping
ones strategy in accordance with both rules and the possible actions of oppo-
nents. Games should be equally challenging and rewarding, hovering between
boredom and anxiety hereby assuring a space of ow through the network of
choices. For a computerized game system, a successful game play implies a
balance between xed rules and the control of player input in variable settings.
Self-reference in computer games: A formalistic approach 223

3. Rule system and interaction system

What denes a rule? A rule, being algorithmic in its core design, consists of a
simple, unequivocal sentence, e.g., you are not allowed to use hands while the
ball is on the pitch. Hereby, a rule constitutes the possibility space of a game
by clearly stating limitations (not use hands) as well as opportunities (the ball is
on the pitch). It is always possible to dene a game both in negative and positive
terms: rules limit actions; they determine the range of choices in the possibility
space; they encircle the arenas to be played in; yet they also frame what can be
At this point, I am speaking of all games, i.e., both traditional games, in-
cluding sports, and computer games. Heroes of Might and Magic rests on rules
stored in and processed by a computer. Chess or Monopoly, by contrast, rely on
rules not accumulated in the database and algorithms of a computer but writ-
ten down on paper and stored in the players mind during the play. In a game
of soccer, for example, such rules are administered by a referee. Implicit rules
that are normally considered exterior to the real rules (e.g., clock in chess
matches) must be engaged explicitly in digital games. These rules have to be
programmed as well. Weather conditions or the general physics of a soccer game
are usually taken as out-of-game features in the real world. When we simu-
late a soccer game in a computer, however, the rules of soccer and the general
physics (including random variables such as surface granularity, crowds, time
of day, etc.) must be built into the rule algorithms and the input-output control
of the computer.
Rules specify the constitution of the playing deck or, more broadly, the
playing eld.3 In games, behavioral patterns inside this eld are limited, con-
strained, and highly codied (Huizinga [1938] 1994; Caillois [1958] 2001;
Walther 2003). Rules are guidelines that direct, restrict, and channel behav-
ior in a formalized, closed environment so that articial and clear conditions
inside the magic circle of play are created (Salen and Zimmerman 2004).
The outside of this circle, reality or nonplay, is essentially irrelevant to game
play. Confronted with unambiguous rules, strategies (or tactics) might entail
best practice solutions variable to the given rule constraints. Hereafter, interac-
tion patterns map the various player interventions and can hence be viewed as
a texture of moves and choices overlain on top of the possibility space of the
game. Furthermore, interaction patterns can refer to the social and competitive
intermingling of players during the fulllment of the game. In that respect, the
patterns correspond to the outcome of absolute rules and social dynamics.
224 Bo Kampmann Walther

Rules have the following qualities:

They limit and restrict player action. Thus, they tell what can be done and
what cannot be done with the objects associated with the game.
They are unambiguous, explicit, and nite (which is why they are easily
incorporated in computer algorithms).
All players of a game must share them.
Rules are xed, i.e., unchangeable (if they do change, we rather refer to local
or house rules).
They are binding, i.e., nonnegotiable.
They can be repeated, which means that they are portable and work indepen-
dent of technology platform or ctional representation.
The formal organization of games can be regarded as a parameter space. In this
space, the current state of the game counts as a point and ultimately a dimension
in the parameter space. A played game has therefore n possible state dimensions.
In Tic-Tac-Toe, for instance, the nine squares constitute the parameter space of
the game and thus the possibility domain for the arrangement of the board pieces.
The rules of the game dene the possible edges in the space connecting states.
Rules dene the possible game, whereas a particular game is a path through the
state space. The crucial factor is that there can be no variability or multiple paths
through the possibility space of a game without the compulsory parameters of
the game. Hence, the parameter space constitutes the transcendental level of the
game, whereas the particular game path expresses the contingent realization of
the space.4
This dialectic between parameter space and actual game path also sheds some
light on why games are complex; basically it is because there is an uneven relation
between the unchanging set of rules and the actual and changing realization of
a particular game. This asymmetrical tie between rules and realization (or rules
and strategies) can be termed game emergence. Most often it is impossible to
predetermine the actual moves and outcome of a game only by knowing the set
of rules.5 Also, most games are games of imperfect information (Nash 1997).
At the outset, the rules of chess are simple, and yet the wealth of distinct chess
playing tactics is enormous. A child can memorize chess rules, but to master all
grand openings in the actual game is probably a lifetime achievement.
When it comes to computer games we must be careful not to confuse two
distinct yet closely associated levels of rules. One level, which is the algorithmic
source code of the game, consists of an unambiguous list of specications for
what can be done and what cannot be done, i.e., what counts as edge in the
parameter space. On another level, rules designate the ability of the computer
to keep track of the players interaction with the different states that the game
Self-reference in computer games: A formalistic approach 225

system can be in. We can specify the former level the rule system of the computer,
and we will name the latter level its interaction system. While the rule system
contains the data structures that enable the initial set-up of the game as well as
determine the constraints and possibilities of the game, the interaction system
evidently operates in a dynamic framework whose prime function is to control
the executing of new outputs relative to the players real-time inputs.
Another way of explaining the difference between the two levels is that the
rule system is responsible for the initial framing of the game by setting up the
possibility space for the game and for the players actions and choices, whereas,
in a slightly different way, the interaction system links to the actual game play.
The latter is the realization of, or a given path through, the possibility space
(Figure 2).

Figure 2. Rule system and interaction system imply a combination of linear and circular
movement, i.e., recursivity

Furthermore, we can model the relation between rule system and interac-
tion system by considering also the machine domain and the player domain
(Figure 3).

Figure 3. Computer and player overlap in the interaction domain as a kind of middle
226 Bo Kampmann Walther

4. Rules and recursivity

The movement from rules to interaction occurs in the medium of time. However,
for this forward processing to be effective, the system needs to perform backward
or looped executions as well. The events occurring in the possibility space of
the game have to be measured constantly against the initial rule system (see
Figures 2 and 3). In order for the computer to respond adequately to player inputs
(which derive from the players strategy) it has to check the viability of input in
accordance with the specied rule set.This rapid intersection of forward linearity
and backward loop circularity denes the elementary recursivity function of a
computer game. A recursive system, such as the computer, is thus a dynamic
system consisting of both linear and circular operations. The computer handles
progress because it also has a memory.
We may further rene the concept of recursivity by comparing it with what
is known in computer science as the state machine (Selic, Gullekson and Ward
1994: 223ff.; Juul 2004: 57ff.). A state machine is a computing device designed
with the operational states required to solve a specic problem. Automatic ticket
dispensing machines are state machines, and so are computer games. There are
several aspects of a state machine but we need only consider two for our present
purpose, state transition, and output function:
The state transition function maps states and inputs to states. This function
denes, limits, and makes possible what happens in response to a given input.
The output function maps states and inputs to outputs. This function denes
the machine outputs at a given time. Y is thus a function that maps states and
inputs to outputs (S I O).
When we look at the game as a state machine we nd that the machine (i.e., the
game) consists of an array of cells, each of which can be in one of a nite
number of possible states. The cells are updated synchronously in discrete time
steps according to a local and identical interaction rule (which we identied
above as the interaction system of a game). The state of a cell at the next step in
time is determined by the current states of a surrounding neighborhood of cells.
The transitions are usually specied in the form of a rule table that denes the
next state of the cell for each possible neighborhood conguration.6
According to Juul (2004), the concept of rules corresponds to the notion of
the state transition function that determines what will happen in response to a
given action at a given time. The transition function is thus a specication of a
set of deep rules, i.e., algorithms that determine the possible output relative to
the current game state and the current player input at time t. Next, the output
function sends a specic view of the game state to the player; a view or a piece
Self-reference in computer games: A formalistic approach 227

of information that is mediated through the interface of the computer (e.g., a

specic screen image, a textual message, etc.).
However, all this clearly involves rules that attest the capacity of the computer
for response or adaptiveness in a variety of settings and a number of game
states. Thus, the state transition function and the output function relate strongly
to the computer interaction system whose primary function, as stated above, is
to control the execution of new outputs relative to the players real-time inputs.
When viewing the computer as a state machine, we can further identify the rule
system (see above) as the possible input events, which the machine accepts and
at this level, the machine determines the constraining elements of the game (the
edges in the possibility space) specied by the rules; in short:
Input events rule system
State transition function and output function interaction system
The recursivity of digital games therefore implies a linear as well as circular rela-
tion between input events, state transition function, and output function. Phrased
differently, recursivity results from the complex intersection of the parameter
space (input events and possibility space) of the game and its interaction system
(output function).
Traditionally, the notion of recursivity is used in group theory, where it des-
ignates a group that is dened as using the group itself or a function that is
called to its own function. Furthermore, the term is used in certain program-
ming languages (such as C++) where recursivity is the property that functions
have to be called by themselves. Here, however, we will stick to the broader and
more general denition of recursivity indicative of systems that entail a dynamic
oscillation between linearity and circularity.
My point is that this general classication of recursivity is indeed a formal
or structural denition of self-reference. We could say that self-reference, for
example, in literature operates in the form of codied or semantic relations
between possible input events and output function. However, for the input events
to be functionally effective, they have to be stored and made actively operational
in the readers (players) mind so that they may act as base for the current output
function. This means, in essence, that the mind corresponds with the memory of
the computer as well as it resembles the state transition function of a computer.
The analogy is this: I stumble across a character in a novel and think that this
character is somehow connected with a character or a situation in another novel;
or I may begin to wonder whether the entire design of a given novel might not be
implicitly allegorical of some textual element from elsewhere. More advanced,
I might postulate that the novel I am reading (as a series of mental outputs that
228 Bo Kampmann Walther

determine my path through the narrative) on a deeper, hidden level invokes

the transcendental conditions of its own being-narrative.
Whether I am concerned with identifying passages linking ctional charac-
ters, or whether I am trying to demonstrate the particular poetological modality
of self-reference of a novel, I always link possible input events (what is referred
to) to the current output function (i.e., what is considered to be signs of self-
reference). However, in the case of reading a novel and scrutinizing its complex
web of reference, this procedure needs to be implemented in a purely hermeneu-
tic framework; self-reference in nondigital media does not possess an automatic
state transition function that maps states and inputs to states. In nondigital me-
dia, a response to a given input (the state transition function) belongs neither to
the rule nor to the interaction system.
Rules formulated in and controlled by computers always hinge on algorithms
that only react to very selected aspects of the world, e.g., the state of the system
or the well-dened inputs (Juul 2004: 61). A game thus has a predened and
nite number of input events whereas the input events that act as referential
markers in the self-referential circularity of a novel are clearly innite in
number. The computer controls the niteness of algorithms and output functions
using a necessary decontextualization, which means that no element or only a
restricted number of elements of a given context is relevant to the game system.
On the contrary, the human interpreter controls the relation between input event
and output function by deploying a potential contextualization that allows for
a principally innite number of parts of a given context to be relevant for an
understanding of the system.7 This allows for the following denitions:
Computerized game recursivity implies an automatic, cybernetic process
in which only a nite number of input events are accepted as base for possible
output functions. The dynamic system therefore presupposes a trivial relation
between the initial possibility space and the information (or output) shown
to the player in a given game state.
Hermeneutic ction recursivity implies a nontrivial process in which an
innite number of input events (which together form the possibility space
of referentiality) can be potentially linked to a likewise innite number of
output functions (which are the self-referential signs).
It is vital to be aware that this double denition could have been made without
inducing any content or ction oriented phenomena. Any mention of novel,
text, etc. was only provided for exemplication. The difference lies solely in
the formal nature of computer games vis-a-vis nondigital media such as codex
literature and cinema. We can illustrate the concept of recursivity as in Figure 4.
Self-reference in computer games: A formalistic approach 229

Figure 4. The recursive system depends on an oscillation between linearity and circularity

5. Game epistemology: Game play and recursivity

In the previous section, we saw how rules and interaction system together dene
the gameness of games. Now we will enquire more deeply into the logic of
game play, the playability or ludic structure of gaming.
Using Niklas Luhmanns systems theory as well as George Spencer-Browns
(1969) form theory, I have tried to categorize and reect on the difference be-
tween playing and gaming (Walther 2003). The trick is to view gaming as
something that takes place on a higher level, structurally as well as temporally.
When it comes to play, the installation of the form of the play world vs. nonplay
world distinction must performatively feed back on itself during play, continually
rearticulating that formal distinction in the play world so as to sustain the inter-
nal ordering of the play world. However, in the game mode, this rearticulation is
already presupposed as a temporal and spatial incarceration that protects the rule
binding structure of a particular game from running off target. In other words,
games should not be play; but that does not imply that they do not require play.
This means, in effect, that in the play mode the deep fascination lies in
the oscillation between play and nonplay, whereas game mode presses forward
ones tactical capabilities to sustain the balance between a structured and an
unstructured space. In the play mode, one does not want to fall back into reality
(although there is always the risk of doing so). In the game mode, it is usually
a matter of climbing upwards to the next level and not losing sight of structure.
Play is about presence, while game is about progression.
In play, the deep fascination therefore lies in the oscillation between play
and nonplay, which is the other of play usually considered to be reality. In
the playing of games, we are more xated on progressing in the prior structure
which is the game (Kirkpatrick 2004: 74). Gaming presupposes the tension, or
230 Bo Kampmann Walther

the initial transgression, in which we constantly resist falling out of the fan-
tasy context of play, and gaming presupposes further focus on a second, higher
transgression in which success and failure is measured against our achievement
of dened objectives. Thus, in playing a computer game, we work in a second
simulacrum, an as if structure overlain on top of the rst initial transgression
that makes play possible in the rst place.
Two things are particularly important with respect to our investigation of
self-reference or recursivity in computer games. First, we can note that the
act of gaming or game playing involves the fabrication of willed illusions that
support the progress from initially stepping into the magic connes of play
and, subsequently, trusting and acting in accordance with the xed rule set and
structured topology of games. Second, as Kirkpatrick writes in his interpretation
of my research in these matters, it also
involves a certain self-understanding; players know that they are responsible
for maintaining the illusion that is the game world and the sense of play that
supports it. This knowledge ultimately threatens the game and play itself,
giving it a kind of ontological insecurity. This is why play is often repetitive,
since repetition reinforces the reality of the game world. However, this same
repetitiveness results in a kind of disenchantment for the player [. . . ] and an
inability on her part to continue foregrounding the game play experience.
(Kirkpatrick 2004: 74f.)
In systems theoretical terms, this self-awareness of ontological insecurity
translates into the players ultimate understanding and therefore constant han-
dling of the other-reference in the game itself which is simultaneously part of
the self-reference of the game. It is a fundamental sign of the game itself that the
threat of a nongame domain or a nongaming situation is forever intrinsi-
cally tied to the construction of the game itself, and the players have to be aware
of and even stay alert to this fact. Thus, a certain level of self-referentiality or,
at the very least, a minimal awareness of the logical organization of play and
nonplay is required. Game play requires reference to the way in which a game
feeds from its own negative preconditions; this reference is obligatory for any
actualization of a game. In psychological terms, when a game becomes uninter-
esting it is probably because the player fails to sense a presence from the inside
of deterritorialization of the presence (Walther 2003). The player falls out
of the constraints and the negatively dened territory that is the game. In the
terminology of systems theory, to play means to engage in a dynamic oscillation
between levels of transgression without getting caught in the ontological uncer-
tainty that is part of the set-up of the game. To play also means to master the
critical coincidence of reference and self-reference, that is, the ability to toggle
between what the game is about and what it takes for the game to come about.
Self-reference in computer games: A formalistic approach 231

A certain, although not always explicitly articulated, level of self-referentiality

is hence an essential element of game play.
The success of transforming games (e.g., board games) into computer games
might stem from the fact that a digital computer is a discrete state machine. It thus
bears, in its very design, a strong resemblance to formalized game systems, most
notably rules for discrete sequential operations. In contrast, play seems to focus
on investigations of semantics, since the task is not only to measure its space but
also to elaborate upon its modes of interpretation and means for reinterpretation.
Not only do we explore a world while playing. Its potential meaning and the
stories we can invent in that respect also drive us. Play spaces tend to expand,
either in structural complexity or in physical extent. This expansion is further
reected in the praxis of play, for instance when players argue over the exact
thresholds of a play domain (Tosca 2000). Again, this must be understood in a
double sense of both the physical closure and the mental activities attached to it.
Why is this simultaneous division between an intermingling of play and
games important for the study of computer games? Because it touches upon
the concept of game play. One can get immersed in the playing mood that is
needed to get into the game in the rst place (the rst distinction that enables
one to identify with an effective killer) but one can also be caught up in a certain
area of the game where one begins to question its criteria for structure (the
second distinction that focuses on transitions). Too much self-reference spoils
the game play! The plot is exactly to balance playing and gaming while gaming.
One must hold on to the initial distinction (otherwise one is swallowed up by
the other of play), and one needs constantly to accept the organization, the rule
pattern, of the game. When one disregards this complementary balance, a ow
is interrupted. Then one begins to speculate: why am I playing; and what exactly
is the objective of the game?
A game play works precisely to assure this ow by serving as a potential
matrix for the temporal realization of particular game sequences. A game se-
quence may lead to wondering how one got into the game in the rst place (this
is to observe the rst transgression, and to be in the play mode), or the actual
sequence might force one to reect upon the criteria for the design of the space-
time settings (this is to observe the second transgression, and to be in the game
232 Bo Kampmann Walther

6. The recursivity of rules and game worlds:

A kind of conclusion

I began this chapter by pointing out that self-reference in computer games splits
into three distinct forms: (1) games can be self-referential as part of the way in
which they handle import and export of content or ction related elements; (2)
games refer to a larger and immensely complex horizon of cultural bricolage,
a kind of cut-copy-and-consume culture; and, nally, and most importantly,
(3) the intrinsic fabric of computer games points to an all-necessary level of
self-reference or recursivity without which games, both ontologically and epis-
temologically, would simply cease to be games.
Next, we found that formal recursivity in computer games can be linked to
two different modes:
On the level of game ontology, there is a recursive dynamic between rule
system and interaction system, i.e., between the possibility space or input
events and the actual path through the game states.
On the level of game epistemology, we ascertained a recursive dynamic
between the transitional differentiations of play mode and game mode that
both together make up game play. The fact that there is a dynamic (temporal
as well as logical) relation between playing and gaming also indicates a
certain level of recursivity mandatory of game play.
Furthermore, we must be aware that there is a vital discrepancy in the con-
cept of self-referentiality when we regard it as an intrinsic constituent in the
workings of the computer as opposed to the idea of a hermeneutical relocation
of input references and output signs of self-reference. In the former case,
self-reference is something that is performed and trivially executed, while in the
latter case, self-reference is something that needs to be perceived and actively
Actually, things are very simple. All digital games are naturally cybernetic,
self-referential systems (Kucklich 2003) whereas all nondigital media, including
ction and cinema, are basically semantic, referential systems that can be per-
ceived as self-referential entities. This does not exclude the subtle fact that digital
games in addition to being formally self-referential may be self-referential on a
content related or cultural level, too.
However, what can be said about the game world? What is the relation be-
tween game rules and ctional representation? As a kind of conclusion, let me
briey show how the formal requirements of a game may be inscribed in the
ction of the game and perhaps vice versa.
Self-reference in computer games: A formalistic approach 233

Many games seem to disrupt the unfolding of narratives in game worlds

in order to assist the player in how to control the keyboard, how to set up the
buttons on the joystick, etc. One example (Juul 2004: 158f.) among many is
the GameCube game Pikmin in which our avatar is a scientist stranded on an
unknown planet. In the course of game play, the scientist takes notes in a diary
that is displayed on the screen and includes notes about the handling of the
controller. According to Juul, there is nothing artistic about this deliberate mix
of ctional representation and game control commands. In fact, this confusion
even strengthens the ction: since the player is the avatar, notes about the
controller are exactly the kind of thing we would write down if we were to take
notes about our playing of the game (Juul 2004: 159).

Figure 5. Max Payne realizing he is, horrifyingly, a character in a computer game

Another example, however, tells us that the reference system of games is

not always that straightforward. In the adventure based rst-person shooter Max
Payne (Figure 5) we are, as noted by Sren Pold (2005), caught in a stratied
maze controlled by drug lords and corrupt police on the level of the plot and
the cybernetic game engine on the structural level. More than allocating the
in-game story as a motivation for game play, which is typical of the genre, Max
Payne designates the narrative as a cliche; the Hollywood signs point towards
narrative structure in general rather than support the particular narrative (Pold
2005: 16). Pold continues:
234 Bo Kampmann Walther

The game could be interpreted as a self-conscious intervention in the on-

going debate about the roles of narrative in computer games. Narrative
becomes an effect that the game self-consciously alludes to and puts on but
does not fulll in the deep Aristotelian way imagined by the proponents of
interactive narrative. This is narrative surface or skin that does not attempt
to become hegemonic, covering all aspects of the game, but like postmodern
novels and cinema alludes to narrative, quotes it, without fully enacting it.
(Pold 2005: 16)

In a graphic novel sequence in the drugged opening of the third act, Max Payne
nally realizes and reveals to the player that he is nothing but a pixilated avatar
in a computer game (Figure 5). Suddenly, Max Payne, as a precondition for the
game plot, questions the initial and vital transgression of play. Consequently,
through the metactional confession we are thrown into play mode. Why play
if the character that is supposed to glue together playful praxis and structured
game space is genuinely untrustworthy?
Paynes existence serves only the endless repetition of the game which is
the at once dull and sophisticated blend of realism to the max and max
pain, advertised through the graphical user interface with its weaponry, red
bar, and bullet time on-off button. Pold concludes by categorizing Max Payne
as illusionistic media realism (Pold 2005: 20), a realism that simultaneously
engages in illusion and can be viewed as a self-reexive exploration of its own
representational techniques and media. In light of the ndings of this article, we
may further hypothesize that Max Payne knows and plays with its own recursive
dynamic and places it amidst the ctional elements as a self-conscious cue to
its own rule structure and level of progression. Games like Max Payne therefore
ironically mock at, yet at the same time celebrate, a self-awareness of how the
necessary recursivity of all games (not just the intentionally artistic ones) gets
immersed into the ction while clearly belonging to the trivial, nonsemantic
level of rules.


1. This emphasis on the formal side of games and game theory is denitely not con-
structed so as to denounce any kind of study that enquires into the ctional or
intermedial self-referentiality in computer games. Readings of self-referentiality of
the latter kind, on the contrary, will provide us with much knowledge of the culture
of narrative transfer in the contemporary media system. However, it is my genuine
belief that the initial distinction between structural and ctional self-reference in
games as well as in other types of media claries discussions that otherwise tend
Self-reference in computer games: A formalistic approach 235

to obscure the levels of analysis: computer games are by necessity self-referential

(or recursive), dynamic systems; yet not all of them need to be self-referential in a
ctional sense. What does the game Tetris refer to? Nothing really, unless, that is, one
would claim that the image of falling polygons in a vectorized eld is an indicator
of a cultural dynamic (text) of some sort.
2. I say advisedly cultural and not sub cultural since games nowadays are the norm
of mediated communication and not just a more or less esoteric sub-branch that
connects to the entire media ecology.
3. The notion of deck and eld also alludes to the common sense comprehension of
games board games and sports count as archetypes.
4. Here, we may note that the game in itself is set in a spatial realm. The nature of the
spatial structure pre-determines the organization of edges in the topology. Contrary
to this, the particular playing of a game is functionally operative solely in the domain
of time. A game exists; but a game also evolves.
5. This does not mean that the player is incapable of optimizing his or her strategy by
knowing and, essentially, anticipating the rule-based responsiveness of the computer.
In the game Need for Speed awareness of features such as catch up effect and
spawn mechanisms effectively aid the player in obtaining the primary objective of
the game to win.
6. Thus, we could formally dene a game as the sum of all states of cells at time t, which,
in turn, is a function of the state of a nite number of cells called the neighborhood
at time t-1.
7. I guess this is another way of claiming that the mind works in mysterious ways and
that a computer operates in entirely pre-determined ways.


Aarseth, Espen
2003 Playing research: Methodological approaches to game analysis. Pro-
ceedings from Digital Arts and Culture (DAC), Melbourne 2003, (18.01.06)
Bolter, Jay David and Richard Grusin
1999 Remediation. Understanding New Media. Cambridge, MA: MIT Press.
Caillois, Roger
[1958] 2001 Man, Play, and Games. Urbana: University of Illinois Press.
Holland, John H.
1998 Emergence: From Chaos to Order. Oxford: Oxford University Press.
Huizinga, Johan
[1938] 1994 Homo ludens. Vom Ursprung der Kultur im Spiel. Reinbek:
236 Bo Kampmann Walther

Juul, Jesper
2004 Half-real. Video games between real rules and ctional worlds. Un-
published doctoral dissertation, The IT-University of Copenhagen.
Kirkpatrick, Graeme
2004 CriticalTechnology:A SocialTheory of Personal Computing. London:
Kucklich, Julian
2003 Perspectives of computer game philology. Game Studies 3.1, (18.01.06).
Nash, John F.
1997 Essays on Game Theory. London: Edward Elgar.
Neumann, John von and Oskar Morgenstern
1953 Theory of Games and Economic Behavior. Princeton: Princeton Uni-
versity Press.
Pold, Sren
2005 Interface realisms: The interface as aesthetic form. Postmodern Cul-
ture 15.2, culture/v015/15.2pold.html
Salen, Katie and Eric Zimmermann
2004 Rules of Play: Game Design Fundamentals. Cambridge, MA: MIT
Selic, Bran, Garth Gullekson and Paul T. Ward
1994 Real-Time Object Oriented Modeling. New York: Wiley.
Smith, Jonas Heide
2006 The games economists play: Implications of economic game theory
for the study of computer games. Game Studies 6.1., smith (01.04.07).
Spencer-Brown, George
1969 Laws of Form. London: Allen & Unwin.
Tosca, Susana Pajares
2000 Selbsreferentialitat in Computer-Spielen. Dichtung Digital, October,
Walther, Bo Kampmann
2003 Playing and gaming. Reections and classications. Game Studies
3.1, (18.1.06).
2005a Konvergens og nye medier [Convergence and New Media]. Copen-
hagen: Academica.
2005b A hard days work: Reections on the interfacing of transmedialization
and speed in 24. In: Klaus Bruhn Jensen (ed.), Interface://Culture
The World Wide Web as Political Resource and Aesthetic Form. Copen-
hagen: Samfundslitteratur Press/NORDICOM.
Metacommunication in play and in (computer) games

Britta Neitzel

It has often been argued that play and games are in a way self-referential.1
According to Friedrich Schiller ([1801] 1967), play drive creates an autonomous
aesthetic domain with its own living forms (lebende Gestalten) which are in
themselves both eternal and transitory.2 Following this line of thought, Johan
Huizinga ([1938] 1994) argued that play takes place in a realm of its own,
separate from the rest of the world because of its own rules and boundaries. Hans
Scheuerl ([1954] 1990) introduced the concept of circular movement to describe
the nature of play, while Roger Caillois ([1958] 2001) established the criterion
of separation in space and time as a distinctive feature of games. Boundaries
and frames which separate games from their social environment and establish a
world in which play activities have only a meaning in themselves seem to be an
important attribute of games.
Gregory Bateson ([1955] 1972) has given convincing evidence that commu-
nication between players is self-referential in another respect. In his Theory of
Play and Fantasy, Bateson describes play as an autonomous sphere of human and
animal behavior which differs from nonplay by the feature of metacommunica-
tion. Communication in play is a form of communication about communication,
and the circularity which is apparent in such communicative processes evinces
a mode of communicative self-reference.
Batesons theory of play as metacommunication is of great interest to the
emerging research eld of computer game studies (cf. Salen and Zimmerman
2004). Although his theory only focuses on play and Bateson restricts himself
to stating that games are more complex than mere play, we will consider games,
too, in the present article. The hypothesis will be developed that games differ,
among other things, from play with respect to metacommunication. The paper
studies metacommunication and various forms of framing in play, games, and
digital games.
238 Britta Neitzel

1. Metacommunication in play according to Bateson

Inspired by his observations of monkeys in the San Francisco zoo, Bateson put
forward the hypothesis that play behavior is accompanied by metacommunica-
tive signals which are noticed and interpreted both by players and observers:
I saw two young monkeys playing, i.e., engaged in an interactive sequence
of which the unit actions or signals were similar to but not the same as those
of combat. It was evident, even to the human observer, that the sequence
as a whole was not combat, and evident to the human observer that to the
participant monkeys this was not combat. Now, this phenomenon, play,
could only occur if the participant organisms were capable of some degree
of metacommunication, i.e., of exchanging signals which would carry the
message this is play. (Bateson [1955] 1972: 179)
Bateson distinguishes between metalinguistic and metacommunicative mes-
sages. Metalinguistic messages refer to language. An example of a metalinguis-
tic message is the sentence: The word cat has no fur. A metacommunicative
message, on the other hand, refers to the communicative situation in which a
speaker and hearer (or players) are involved. According to Bateson, the meta-
communicative message This is play establishes a paradox of the kind which
is also known as Russells paradox or Russels antinomy.3 The formula which
denes this paradox is: M = { A | A / A }. This formula, in which A designates a
set and M the set of all sets that do not contain themselves as members states
that A can only be an element of M if A is not an element of A.
Now, if M , the set of all sets that do not contain themselves, contained itself,
M , by denition, could not be the set of all sets that do not contain themselves
since it would contain itself despite its claim to the contrary. However, if M did
not contain itself as one of its elements, M could not be the set of all sets that
do not contain themselves. This is the paradox: on the one hand, the set M must
include itself as one of its elements; while on the other hand, it must not contain
itself. The set M would paradoxically contain itself and would not contain itself
at the same time, which would assert that the self-contradictory statement A and
not-A is true despite its being incompatible with the principle of the excluded
middle, which postulates that only A or not-A can be true. Hence, A is an element
of M if and only if A is not an element of A.4 The logical problem underlying this
paradox has been known since antiquity, which discussed it as Epimenidess
paradox, derived from Epimendes, the Cretan, who came to Rome and declared
self-referentially that all Cretans are liars. The rule which Russell established
against paradoxes of this kind postulates that sets (or classes) and their elements
must be strictly kept apart since they cannot be dealt with at the same level of
Metacommunication in play and in (computer) games 239

According to Bateson, the message This is play implies a negative metas-

tatement such as These actions in which we now engage do not denote what
those actions for which they stand would denote (Bateson [1955] 1972: 180).
Since standing for, according to Bateson, is a synonym of denoting, the
sentence may hence be paraphrased as: These actions, in which we now en-
gage, do not denote what would be denoted by those actions which these actions
denote, and. hence, The playful nip denotes the bite, but it does not denote
what would be denoted by the bite (1972: 180).
Since the underlying paradox evinces the logical contradiction that this is
a bite and not a bite at the same time, we are faced with an infringement of the
law of the excluded middle. But on which level does the paradox emerge in play?
Consider the example of two girls boxing in play. The bodily actions may
be quite the same as in a real ght; nevertheless, the girls are not ghting at
all, even though their sts may be clenched and they may even hit each other.
Playful boxing is an iconic sign of real boxing with the difference that players
in contrast to ghters will not end up with a bloody face or a broken nose.
Evidently, there is a difference concerning the consequences of the two modes
of behavior. The agents motives and intentions differ, too. While a real ght is
carried out because of anger, fear, or hatred, a playful ght has no such causes.
Playful ght can hence be interpreted as a sign of real ght. Signs of action
differ from the action they refer to especially in their pragmatic dimension,
which concerns the effects on our lives. For example the statement Smoking
30 cigarettes daily will cause lung cancer can be used to frighten people, but
it cannot cause lung cancer; only the actual act of smoking may do so. Only
performative speech acts of the subtype of the declaratives (as Austin called
them) do more than refer to an effect; if uttered appropriately, they are able to
cause the effect which they refer to. The utterance I herewith declare you man
and wife, spoken by a registrar, really makes the couple husband and wife. As
long as we do not confound signs with their objects there is no paradox. Words
and utterances can mean objects and actions, but they do not exert the same
inuence on our lives as the objects and actions they refer to, and they do not
have the same consequences. Signs and their objects are of a different kind or,
as Bateson put it with reference to Alfred Korzybski: the map is not the territory.
In play, the distinction between the map and the territory is not as radical as
in geography. The two levels of play and serious life are marked as different
by metacommunication. The iconic representation of the bite in play does not
mean the same as a real bite, but the playful bite does not simply negate the real
bite. Signs in play communication negate their objects through afrmation. A
merely playful act denotes and at the same time it does not denote the real
action to which it refers since it differs in meaning and in its consequences. The
240 Britta Neitzel

action to which the players iconic nonverbal sign refers is really performed,
but the meaning which this action has in a nonplay context is negated with the
performance of this action. In this sense, there is a paradox.
For this reason, Bateson argues that play marks a step forward in the evo-
lution of communication the crucial step in the discovery of mapterritory
relations. In primary processes, map and territory are equated; in secondary
processes, they can be discriminated. In play, they are both equated and dis-
criminated (Bateson [1955] 1972: 185). Both in therapy and in play, metacom-
munication is part of communication:
As we see it, the process of psychotherapy is a framed interaction between
two persons, in which the rules are implicit but subject to change. Such
change can only be proposed by experimental action, but every such exper-
imental action, in which a proposal to change the rules is implicit, is itself a
part of the ongoing game. It is this combination of logical types within the
single meaningful act that gives to therapy the character not of a rigid game
like canasta but, instead, that of an evolving system of interaction. The play
of kittens or otters has this character. (Bateson [1955] 1972: 192)
In play, participants have to be aware of this paradox, which is especially
evident when we consider role play or acting. Actors have to play their roles as
convincingly as possible, but at the same time, they have to be aware that they
are just playing their roles.5 (The audience, too, must be aware of this fact; there
is the well-known example of illiterate audiences yelling at the hero to warn him
of the hidden aggressor.) An actor or actress who fails to realize the difference
between theater and life is no longer an actor or actress. They behave like a
schizophrenic who actually believes to be another person. The connection be-
tween play behavior and psychiatric anomaly is apparent, as Bateson has shown.
In sum, Bateson considers play to be metacommunicative because of the
self-referential way in which the players signalize that they are playing. Their
metacommunicative message This is play is inherently paradoxical since it
afrms and negates at the same time what the players are doing.
Metacommunication in play is self-referential communication. Players per-
form actions and simultaneously they refer to the way they perform these actions,
afrming that what they are doing is just play. In the context of the semiotics
of pictures, Noth (this vol., Part II) distinguishes between metapictures and
self-referential pictures: a metapicture refers to another picture, for example,
by alluding to it or by quoting it, whereas a self-referential picture is a picture
that refers to a specic picture, namely to itself. Every self-referential picture
is hence also a metapicture because it is a picture (or a pictorial element) about
a picture, but not every metapicture is self-referential since a metapicture can
also be a picture about another picture.
Metacommunication in play and in (computer) games 241

In analogy to the distinction between metapictures and self-referential pic-

tures, the message This is play can be dened as both metacommunicative
and self-referential: it is metacommunicative since it is about play, which is a
form of communication; it is self-referential since it refers to the very situation
in which this message is conveyed. We can even go one step further: play ac-
tivities (not framed by games) must be self-referential; otherwise play cannot
take place at all. There is no play without self-reference. Metacommunicative
self-reference sets the frame of reference for play.
Bateson argues that play involves a further paradox which is due to the
ctionality of play behavior:
Not only does the playful nip not denote what would be denoted by the
bite for which it stands, but, in addition, the bite itself is ctional. Not only
do the playing animals not quite mean what they are saying but, also, they
are usually communicating about something which does not exist. (Bateson
[1955] 1972: 182)

However, from the perspective of Peirces semiotics, it is evident that this

Batesonian paradox is not a paradox at all since the object of a sign, according to
Peirce, must not necessarily be a real object but can also be idea, an imagina-
tion, or even a mere ction to which a sign refers (cf. CP 4.531, 8.314). Neither
play nor sign behavior presuppose any real object of reference; both can also
refer to previous knowledge or experience in the realm of imagination or ction.
The potential for creating and referring to ctional events is common to play
and other modes of ctional communication. According to Bateson ([1955]
1972: 181), histrionic play, bluff, playful threat, teasing play in response to
threat, histrionic threat [. . . ] form together a single total complex of phenom-
ena. Games of adults such as gambling or playing with high risk have their roots
in the combination of threat and play. Bateson also considers the afnity of play
with art and rituals. All these modes of communication do not refer to reality
in the sense of something really existent but to possible worlds, be they alterna-
tive, imaginary, impossible or perhaps in some indenite future. Games certainly
belong to this complex of play and fantasy. In the following we will examine
how both metacommunication and self-reference is characteristic of games.

2. Metacommunication in a card game

The distinguishing feature between games and play is that games are played
according to rules, whereas play is spontaneous and has no previously estab-
lished rules. The rules of a game determine the range of the players possible
242 Britta Neitzel

moves and in some games, their temporal and spatial order. According to Salen
and Zimmerman (2004: 125), game rules limit the players actions; they must
be explicit and unambiguous, shared by all players, xed, binding, and repeat-
able. While in play, every single action must give evidence that it is play, games
have rules that set a frame for all activities. It is not the players who establish
the sphere of the game but the rules, which create a magic circle6 in which
all and only game actions take place. Game actions are thus dispensed from
metacommunicative and self-referential discourse, whereas play is not. Is there
metacommunication or self-reference in games at all?7 I will try to answer this
question by means of the example of a popular German card game called Skat.
Skat is usually played by three players with a pack of 32 German or French
cards. Let us imagine that three girls have decided to meet for a game. Each of
them receives a hand of ten cards from the dealer; two cards remain in a stack
on the table. To begin with, one of the players turns to her neighbor to the right
and begins to negotiate, in a dialogue, the value of their hands. Since each player
may prot from not announcing the real value of her cards immediately because
the risk of losing (and winning) increases with each value, the rst negotiator
will begin with the lowest possible value, 18, and continue in the sequence of
the next possible values, that is, 20, 21, 22, 23, 24, 27. . . , up to the highest
value at the risk of which she is still willing to play. After each number proposed
by the negotiating party, the co-player answers yes or no according to her
calculation of the value of her hand. When no player is willing to take a higher
risk, the negotiation ends and the two parties are determined: the player who
declared the highest value has to play against the two co-players who form a
team against her. The single player has the right to substitute the two remaining
cards for two cards of her hand and to discard the cards with the lowest value.
She announces which suits will be trumps, and the player who sits on the left
hand side of the dealer leads with the rst card, the others will follow in playing
their cards in clockwise direction. Each party tries to win the cards on the table
by playing a card of higher value. The single player has to obtain at least 61 of
the 120 possible points to win the game, otherwise she loses.
According to the above description, there is self-referential communication
about the game before the game begins, i.e., when the players arrange to meet
to play the game. The rules of the game exist before the actual game is played.
They are constitutive rules, which are valid independently of whether the game
is played or not. The game situation is completely framed before the players
begin to play.
Players who know the rules well need not discuss these rules in any meta-
communicative discourse during the game, but the rules as such are metacom-
municative in a way, since they determine what the players may or may not do.
Metacommunication in play and in (computer) games 243

In Skat, they even contain a constitutive procedure to determine the formation

of the two teams of the game.
Neither the agreement to play nor the rules as such are self-referential. The
rules of the game are unquestionable instructions determining the players per-
mitted and the prohibited moves. They have no reference to themselves but they
refer to all games that will and have been played.8 The individual moves and
the initial negotiation of the teams are neither metacommunicative nor self-
referential. Insofar as they are rule-governed activities, they are alloreferential,
having reference to conventions established long before the game begins.
Such alloreference at the level of the individual moves does not preclude
that the game as a whole might not be nevertheless a self-referential activity.
Games, unless played for money, have no purpose beyond the game itself. This
is what they have in common with poetry whose self-referential nature has
been emphasized by Roman Jakobson, among many others. Games are as self-
referential (and alloreferential) as poetry is.
Furthermore, other forms of self-referential metacommunication take place
when a game is played, for example, when players change their communicative
role from ally to opponent and begin to speak like friend or foe, attering each
other or using playful verbal injuries against the opponent. Metastrategic dis-
course of this kind is not prescribed by the rules of the game; hence it is not
part of the game although it is still a mode of play. Play has a very fragile com-
municative situation because there is always the danger that playful rudeness or
simulated verbal injuries might be taken seriously as a personal offense.
Thus, games evince a kind of double framing. First, the game is framed by
its own constitutive rules. Secondly, but only optionally, a game may also be
framed by play accompanying the game. In addition to the constitutive rules,
which dene the game, play may thus introduce additional regulative rules,
which determine the players activities in various, mostly ad hoc, ways. For
example, the atmosphere of the play regulated by its social setting as informal,
relaxed, funny or competitive and even professional. The social setting is an
important incentive to the players, but it is not constitutive of the game.9
In sum, playing and gaming must be distinguished. A game is not play, but
play tends to occur concomitantly with games. Games differ from play with
respect to the features of metacommunication and self-reference. Game activity
or gaming is a rule-governed activity guided by the intention to win.10 Play
activity or playing, by contrast, refers to an activity not framed by constitutive
and xed rules but by metacommunication. It is unavoidable that the expression
playing a game also contains the verb play which should theoretically be
distinguished from the concept of game.
244 Britta Neitzel

3. Rules, communication, and metacommunication in computer


Let us now turn to computer games, which are essentially games according
to rules since, with each move, the players have to follow a rule established
by the game, and these rules are unambiguous, repeatable, xed, and binding.
Furthermore, the player must give unambiguous commands, and, as in other
games which allow only a limited number of moves, computer games permit
only a restricted range of commands; a players shouting at the monitor, for
example, will not make a computer react since it is programmed to input from
its keyboard.11 This situation differs from the one of play, which is based on
ambiguity, the frame of play being uid because it is only established in the
course of the play.
There are two basic types of computer games, multiplayer and single player
games. In how far and to which degree is communication self-referential in these
two types of game?
The social frame which characterizes the situation of players playing a game
comes to existence whenever players meet for a game, not only in a game of Skat,
but also in multiplayer computer games. Digital games evince a kind of framing
similar to the one of nondigital games. They can be played by two players on
one console, in LANs12 , or via internet. Played on one console, the framing
of the game is not unlike the one of Skat. The players do not only play against
each other on the console, for example in a racing game, but also together.
The setting may also be one of small LANs in which the game is played in
the presence of the players who can communicate with each other. The players
verbal comments may also have support via the chat function which allows the
use of written messages. Online games in which direct oral contact is not possible
have chat functions for written messages. Some games also include the so-called
teamspeak function for oral communication via an internet telephone device.
With these characteristics, multiplayer computer games, even when played in the
bodily absence of other players, fulll the prerequisites of metacommunicative
and self-referential play.
Single player computer games, by contrast, have only one player. There is
nobody with whom the individual player can discuss his or her moves so that
no metacommunication can be expected unless the player assumes two selves
in a soliloquy. Can there be metacommunication and self-reference when singly
players interact with their computer games in the absence of other players at
all? Indeed, textual strategies have been devised in such games which simulate
metacommunication even in single players games. They can be found at various
Metacommunication in play and in (computer) games 245

levels of the game and involve the single participants in their different roles as
gamers or players. The game situations in question are either self-referential or
metacommunicative or both. Let us examine two such strategies in the computer
games Zork and Metal Gear Solid.
Can there be metacommunication or self-reference when players interact
with computer games in the absence of other players at all? Indeed, textual
strategies have been devised in such games which simulate metacommunication
even in single players games. They can be found at various levels of the game
and involve the single participants in their different roles as gamers or players.
The game situations in question are either self-referential or metacommunicative
or both. Let us examine two such strategies in the computer games Zork and
Metal Gear Solid.
The strategy of ctional metacommunication can be traced back to the early
times of computer games; it can be found in text based games such as Zork
(Infocom 1977), which is a so-called text adventure. As the genre name suggests,
Zork operates with written text only (cf. Rapp, this vol.). The game starts as
follows (player input being marked as > and lines added for further reference):
1 West of House
You are standing in an open eld west of a white house, with a boarded front door.
There is a small mailbox there.
>open mailbox
5 opening the small mailbox reveals a leaet.
>read leaet
leaet, taken
Welcome to Zork
Zorkis a game of adventure, danger and low cunning. In it you will explore some of the
10 most amazing territory ever seen by mortals. No computer should be without one.
>go north
North of House
You are facing the north side of a white house. There is no door here, and all the 14
windows are boarded up. To the north a narrow path winds through the trees.
15 >go east
Behind House
You are behind the white house. A path leads into the forest to the east. In one corner
of the house there is a small window which is slightly ajar.
>open window
20 With great effort, you open the window far enough to allow entry.
Kitchen Score: 10
You are in the kitchen of the white house. A table seems to have been used recently for
the preparation of food. A passage leads to the west and a dark staircase can be seen
25 leading upward. A dark chimney leads down and to the east is a small window which is
On the table is an elongated brown sack, smelling of hot peppers.
246 Britta Neitzel

A bottle is sitting on the table.

The glass bottle contains:
30 A quantity of water
>open sack
Opening the brown sack reveals a lunch, and a clove of garlic.
>eat lunch
35 Thank you very much. It really hit the spot.
>open bottle
>drink water
You have to be holding the glass bottle rst.
40 >take bottle
>drink water
Thank you very much. I was rather thirsty (from all this talking, probably).
The player can navigate through the ctional Zork territory by typing two-
word commands, which make the program reply with a verbal description of
a state, an event, action involving the player, a command, request, or comment
addressed to the player. This resulting sequence of orders and responses evinces
self-reference at different levels. In the beginning of the dialogue (line 9), the
player reads the line Zork is a game. This self-descriptive statement is already
part of the game; it does not belong to a pre-text that precedes the beginning
of the game like an imprint precedes the beginning of a book. Hence, the text
begins with a self-referential metatextual statement. However, this rst sentence
is not metacommunicative since it does not concern the relationship which the
text establishes with its player in this game.
The usual way of navigating through the Zork world is by typing orders, such
as go north (l. 11), which are given to an unspecied addressee in the world
of Zork. In such moves, the player is an addresser who utters the order, and there
must also be an addressee to comply with the order, but who is this addressee?
Since the player is faced with nobody else, the commands seem to be addressed to
a ctional character in the game world by the mediation of the computer, but the
answer which appears on the screen conveys a different impression. An unknown
voice writes back: You are facing the north side of a white house (l. 13).
This time, the addressee can be nobody else than the single player, that is, the
same person who gave the previous order. Apparently, the computer refuses to
become the addressee of the players order and returns this role to the addresser,
the player. As a result, the single player turns out to be both addresser and
addressee and is entangled in a self-referential communicative loop. As a result,
the player is both inside and outside the game. As the participant who gives
the order, the player is outside, as the one who is addressed by the text of the
Metacommunication in play and in (computer) games 247

program, the player is inside the ctional game world. In terms of systems theory,
the player is an observer who is observing him or herself. This textual strategy
introduces an element of play into the game since the self-referential system
of address exemplies well the dilemma which characterizes play according to
Bateson, the dilemma of being and not being in a given role at the same time.
On the operational level, the commands of the player and the answers of the
program can be compared with performative speech acts, even if they do not
have the form of a statement but of an imperative. The imperatives typed by a
player do not operate like commands but immediately have factual results in
the ctional game world. Typing open window means that the you in the
ctional world is opening the window.
However, at the level of the players interaction with the machine, there is no
self-reference. As far as the computer is concerned, players who type orders such
as go north actually produce a sequence of electronic signals whose effect it
is to trigger a sequence of digital operations which function like commands to
the computer and hence have an utterly alloreferential semiotic effect.
A new communicative scenario begins with the kitchen scene (l. 22). The pro-
grammed addresser now speaks in the voice of a counselor thanking the player
(l. 35, 44) and giving advice (You have to. . . l. 39). In line 43, with the remark
I was rather thirsty (from all this talking, probably,) the addressers voice as-
sumes the new role of a personal speaker who does not only refer to his own
bodily needs (thirst), but also turns self-referential and metacommunicative
with a comment on his own talking. There is hence a situational catachresis,
a break in the continuity of the participants roles. Now, the addressee is no
longer the same as the addresser, and the player, no longer isolated in soliloquy,
is faced with an addresser who seems to be a true interlocutor.

4. Metalepsis and ctional metacommunication

The strategies of metacommunication in the game of Zork, in which different

diegetic frames are manipulated against all conventions of narration, are well-
known from literature. In literary theory they have been described as metalepsis.
Metalepsis is a narrative device that manipulates the level of narrating with the
level of the narrated events. As Marie-Laure Ryan (2004: 441) puts it: Metalep-
sis is a grabbing gesture that reaches across the levels and ignores boundaries,
bringing to the bottom what belongs to the top or vice versa. Examples are c-
tional characters who address their author or their readers or also narrators who
enter the world of ction created by themselves. Ryan distinguishes between
rhetorical and ontological metalepsis:
248 Britta Neitzel

Whereas rhetorical metalepsis maintains the levels of the stack13 distinct

form each other, ontological metalepsis opens a passage between levels
that results in their interpenetration, or mutual contamination. These levels,
needless to say, must be separated by the type of boundary that I call onto-
logical: a switch between two radically distinct worlds, such as the real
and the imaginary, or the world of normal (or lucid) mental activity
from the worlds of dream or hallucination. (Ryan 2004: 442)

In rhetorical metalepsis, the levels of narrating and the narrated world remain
distinct, although there is some rhetorical reference from one to the other. In
line 8 of the Zork excerpt, this was the case. An unidentied addresser uses the
medium of a leaet in a mailbox to greet the player with the words Welcome to
Zork. Who is this mysterious addresser? Was the message sent by mail from an
agent outside of the game, or is there some addresser within in the Zork world
who sent this message? In the latter case there would be no metalepsis in this
message, but apparently there is no mysterious addresser in this world of ction,
and the former case is more plausible. The addressers of the message are really
the authors and the publishers of this game who interfere in the world of ction
with self-referential product placement.
Ontological metalepsis, which results in real life interferences from the world
of the narrator to the narrated world or vice versa, is even at the root of Zork as
well as of many other computer games. The player who, at the desk in front of a
home computer, types orders such as open window (l. 19), open sack (l. 31),
or open bottle (l. 36) is rewarded with immediate obedience not only of undis-
closed agents but also of inanimate objects, such as windows, sacks, or bottles.
Players of computer games thus seem to have the power of metaleptic interfer-
ence into the world of ction that, in principle, should exist independently of the
world of their own social environment. Metacommunication which is the basis
of play and which can also be found whenever people play together is integrated
in single player computer games by the textual gure of metalepsis, which can
be called a simulation of metacommunication or ctional metacommunication.
Fictional metacommunication cannot only be found in verbal but also in
visual messages of computer games. Instead of typing open window, for ex-
ample, the player may simply have to press a button on the keyboard or use a
game controller to open a window. Recent examples of ctional metacommu-
nication in single player computer games can be found in games which contain
so-called training missions. Training missions are game scenes designed for the
purpose of getting players acquainted with the rules of the game and the way
it operates. They serve to make them familiar with the use of the avatar of the
respective game world and want to teach the basic commands of the game. In
Tomb Raider I (Eidos 1996), for example, the player has the opportunity of
Metacommunication in play and in (computer) games 249

practicing with Lara Croft in a special training mission to Lara Crofts house
before the adventure starts. In Metal Gear Solid I (Konami 1998), the protago-
nist receives a virtual training on the way to his mission. In Half Life (Sierra
Online 2001),14 the players experience how the newly employed protagonist
becomes familiar with the security requirements of his enterprise.
In training missions, the players are not addressed directly but through the
protagonist of the game. The games conceal their training purpose by creating
a ctional learning environment. Instead of letting the players know that they
are being taught to press a button for the purpose of becoming acquainted with
a new type of game, the illusion of the training for a more important mission
is created. The ctionalization of the rules to be learnt also makes use of the
device of ontological metalepsis for which the Metal Gear Solid series (Konami
19982004) is a good example.
The Metal Gear Solid games belong to the genre of Stealth Shooters or
Sneakers. The protagonist, whose role the player assumes, has the task to conduct
important secret missions in foreign territories while avoiding contact with the
enemy. To obtain the goal of the mission rescuing an ally or destroying the
enemys weapons the real gamer has to save the game occasionally.15 In the
series, this game activity is integrated in the games diegesis. The ctional and
the operational levels are thus interconnected.
At the ctional level of all the games of this series, Snake, the protagonist,
has to sneak into buildings of the enemy all alone, but he remains connected
with his headquarters and also with a paramedic by radio. The headquarters
advise him how to nd his way through the enemys territory; the paramedic
keeps Snakes state of health under surveillance.
Shortly after the beginning of the mission, Snake receives a call from the
headquarters. In addition to information about the mission, Snake and the gamer
who has assumed his identity are advised how to ask the headquarters for help or
information: Press the select button of your controller. Of course, Snake, the
protagonist, has no select button to press, so that the advice is evidently directed
at the gamers with the controllers in their hands. Snake is addressed, but the
gamers are the real addressees. After having learned how to make a phone call,
Snake can ask the paramedic for a report on his health. When the paramedic
complies, the game is saved. The action of saving the game by recording the
state of health has two addressees, the ctional character Snake and the gamer
who wants to have control of the game.
The paradox created by the metacommunicative message This is play in
play is particularly evident at the operational level of the game: addressing the
gamer means addressing the protagonist, and addressing the protagonist means
addressing the gamer, while recording the state of health at the ctional level
actually means saving the game at the real world level.
250 Britta Neitzel

5. Resume

Batesons paradox, according to which play simultaneously afrms and negates,

can account for the manifold shifts between communication, metacommunica-
tion, and self-referential communication in computer games. In play, the bor-
derline between real life and its negation in the sphere of mere play must be ex-
plored since there is no distinct marker to distinguish between play and nonplay.
According to Bateson, playing involves permanent self-referential metacommu-
nication, which sets up a frame for play and occurs simultaneously within that
very frame. More clearly than in play, this interpenetration of frame and content
of the frame can be observed in role play or acting.
Games mark their boundaries very distinctively by their own rules which
determine what is allowed as a game activity and what not. Rules are meta-
communicative but not self-referential. The rules of a game set game activities
free from set up a play sphere by metacommunication. They make game activ-
ities possible at a functional level whose aim is to win the game. Games can
go parallel with play. Nevertheless, there are also game activities that have a
symbolic meaning in addition. For example, placing a piece of a board game
on a certain eld of the board can mean buying a street or occupying a city.
Gaming a term that can be used to describe playing a game with respect to its
pure functionality is almost always accompanied by play that uses the game
only as starting point for play amongst the players.
Single player computer games do not require, but the may simulate metacom-
munication by the device of metalepsis. Some single player computer games set
up a ctional play situation in which metacommunication from the ctional level
to the player world can take place.16 The difference between this kind metacom-
munication and metacommunication in play as described by Bateson is due to
the uid frame of play. While play is only established in the process of playing,
being constantly subject to possible changes, metaleptic metacommunication is
part of the game program and a central issue of computer games.


1. For a survey of the history of the theories of play and games from various perspectives
and disciplines see Sutton-Smith ([1997] 2001).
2. Schillers approach to play is strongly connected with his aesthetic ideal and can be
associated with Kants notion of beauty as evoking disinterested benevolence. But
Schillers inuence is not restricted to aesthetics. His idealistic notion of play had an
inuence on the conception of kindergarten by Friedrich Frobel (cf. Scheuerl [1955]
1964: 57).
Metacommunication in play and in (computer) games 251

3. See paradox

4. We nd the same problem in Epimenidess paradox. And an example of the paradox
closer to real life is the one of the barber who shaves all men of a village that do
not shave themselves and none else. If he did not shave himself he would be a man
who did not shave himself and therefore would shave himself. If he shaved himself
he would not only shave the men who do not shave themselves but also a man who
shaves himself (namely himself).
5. The concept of mimicry, as described by Roger Caillois ([1958] 2001), is very similar
to Batesons concept of metacommunicative play.
6. Cf. Huizinga ([1938] 1994); Salen and Zimmerman (2004).
7. Bateson (1955: 182) argues that games are constructed around the question Is this
8. On the other hand, there is self-reference at a higher level of any game design.
Games are based on constitutive rules. This means that they owe their being played
exclusively to the rules which the players follow and which have no existence inde-
pendently of the game. The kind of self-reference which becomes apparent at this
higher semiotic level is the kind of self-reference which any sign evinces as a token
that refers to its own type of which it is but an instantiation.
9. In surveys dealing with the reasons for playing digital games, playing with others
has often been given as one of the main reasons, see, e.g., the English summary of
Ermi, Satu, and Frans (2004).
10. The term gaming is usually a synonym of gambling or playing for a stake. It
is also used by players of digital games to describe their hobby; they even refer to
themselves as gamers. In gaming, usually a certain amount of money is at stake.
In the sense it is used here, only winning the game is at stake. Gaming can describe
the seriousness of a player who wants to win a game.
11. Of course, there are attempts at broadening and multiplying the input mechanisms
for computers, but they are still very limited.
12. LAN = Local Area Network.
13. In her discussion of metalepsis, Ryan (2004: 439) uses the metaphor of a stack, a
multileveled data structure whose components are processed in an order known as
LIFO: last in, rst out.
14. I am referring to the Play Station 2 version.
15. The goal of a computer game is usually not attained with the rst try. On the contrary,
gamers have to try to master certain tasks along the way to the goal again and again.
To avoid the repetition of the many moves each time they replay the game, gamers
can save the state of the game they have reached in the previous game, which spares
them having to start again from the beginning and allows them to proceed from the
save point.
16. This is not the case in all computer games but only in games that create a game world
and do not merely show objects on the screen which can be moved around by the
player. The latter, which do not create a ctional world, have been called games with
an opaque interface by Bolter and Grusin (2000).
252 Britta Neitzel


Austin, John L.
1970 How to do Things with Words. New York: Oxford University Press.
Bateson, Gregory
[1955] 1972 A theory of play and fantasy. In: Gregory Bateson, Steps to an
Ecology of Mind, 143153. New York: Ballantine Books.
Bolter, Jay David and Richard Grusin
2000 Remediation. Understanding New Media. Cambridge, MA: MIT Press.
Caillois, Roger
[1958] 2001 Man, Play, and Games. Urbana: University of Illinois Press.
Ermi, Laura, Helio Satu and Mayra Frans
2004 Pelien voima ja pelaamisen hallinta Lapset ja nuoret pelikulttuurien
toimijoina, Tampere: Hypermedia Laboratory,
http://tampub.uta./tup/951-44-5939-3.pdf (02.01.06).
Huizinga, Johan
[1938] 1994 Homo Ludens. Vom Ursprung der Kultur im Spiel. Reinbek:
Peirce, Charles S.
19311958 Collected Papers of C. S. Peirce. Charles Hartshorne, Paul Weiss and
Arthur Burks (eds.). Cambridge, MA: Harvard University Press.
Quoted as CP.
Ryan, Marie-Laure
2004 Metaleptic machines. Semiotica 150(1/4): 439469.
Salen, Katie and Eric Zimmerman
2004 Rules of Play. Cambridge, MA: MIT Press.
Scheuerl, Hans
[1955] 1964 Beitrage zur Theorie des Spiels, 4th ed. Weinheim: Beltz.
[1954] 1990 Das Spiel, 11th ed. Weinheim: Beltz.
Schiller, Friedrich
[1801] 1967 On the Aesthetic Education of Man in a Series of Letters, ed.
and transl. by Elizabeth M. Wilkinson and L. A. Willoughby. Oxford:
Sutton-Smith, Brian
[1997] 2001 The Ambiguity of Play. Cambridge, MA: Harvard University
Self-reexivity in computer games:
Analyses of selected examples1

Bernhard Rapp

This paper will present examples and specic forms of self-reference in com-
puter games. The rst example is from Monkey Island 4 Escape from Monkey
Island (LucasArts 2000), a typical Graphic Adventure. In the scene shown in
Figure 1, the player has assumed the role of a pirate who has to persuade other
pirates to join him as the crew of his (still unmanned) pirate ship. Challenging
two uninterested adventurers playing darts in a bar room, the player in the role

Figure 1. Screenshot (PC) from Monkey Island 4 Escape from Monkey Island (Lu-
casArts 2000; German version): The protagonist pirate challenging two other pirates
254 Bernhard Rapp

of the pirate addresses one of the rogues as follows: I bet you wont hit this guy
up front, with a look in the direction of the barkeeper in the background of the
room with whom the pirate player himself is faced.
The rogue, however, is faced in the opposite direction and apparently under-
stands the opposite of what the pirate player means. Accepting the challenge, he
throws his dart not towards the bar keeper in the background but in the direction
of his own line of vision, which is the direction where the real player is sitting
in front of the monitor. He throws, and all of a sudden it seems as if the dart
were hitting the real players screen from within the monitor, dashing through
its glass (Figure 2). The dart seems to leave the ctional Caribbean game world
and to enter the real world of players family home. Of course, the breaking of
the glass of the computer screen remains ction, and the player is not really
impeded from continuing the game.

Figure 2. Screenshot (PC) from Monkey Island 4 Escape from Monkey Island (Lu-
casArts 2000; German Version): Smashing the screen from within the game scenario

In this surprising way, the ctional and the real frames of the game and of
gaming become the topic of gaming: The player in front of the playing device
manipulating the virtual world2 is all of a sudden affected by, and seems to
Self-reexivity in computer games: Analyses of selected examples 255

become the victim of, actions that originate in the ctional world behind the
screen. The pirates that the player should be able to command by his moves
attack their commander in the other real world of the players private home
in front of the computer screen. The director of the play is directed by the
characters of his creation. This ctional device is known as metalepsis (cf.
Genette 1994: 16869; Ryan 2004 and this vol.).

1. Self-reexivity

Players of games have always been confronted with self-reference. By dening

space, time, and types of action of the magic circle, as he calls the game,
Huizinga ([1938] 1980: 11) reminds us that the rules of a game generally re-
fer back to the game as a game. Many other game theories show how playing
games involves reection and permanent self-reection. Bateson (1983), e.g.,
described how games constitute a framing of the players minds (cf. Neitzel,
this vol.), and Niklas Luhmann (1996: 9697) has focused on self-reference in
games from the perspective of systems theory. Players are permanently called
upon to reect their status as players in the course of the game, which testies
to a feedback loop at the roots of gaming. In the context of video and computer
games, the (somewhat problematic) notion of interaction with the game sug-
gests a self-referential dialogue of the player with the game played by the player.
These are some of the basic forms and categories of self-reference which have
been issues of theoretical research in games.
The present paper is less focused on the fundamental categories devel-
oped in game theory; its topic are rather certain selective and explicit forms
of self-reference popular in the computer gaming community which may be
more specically described by the concept of self-reexivity as developed by
Kirchmann. Self-reexivity (or simply reexivity) in the movies, according to
Kirchmann (1996: 76), occurs when the medium lm becomes the topic of a
lm, and the lmic narrative no longer represents the world existing beyond and
apart from the lm. The purpose of this paper is to illustrate various forms of
self-reexivity in games in which playing games is the topic.3 The following
examples were chosen as typical of self-reexive strategies in computer games.
They should reveal a good variety in terms of period, genre,4 content, aesthetics,
and function.
256 Bernhard Rapp

2. Advertising the game to the game world: Zork I

Our second example is from Zork I (Infocom 1982). This game is an early, almost
archaic example of a computer game. It belongs to a genre which is probably
best known under the label of Interactive Fiction or Text Adventure. Such games
consist only of writing and for this reason literary scholarship has shown interest
in them (cf. Aarseth 1997: 97128). The player is faced with the typographical
text of ctional content on the screen (Figure 3) and interacts with the game world
by typing instructions such as go west or open door. These commands are
then translated and processed by the parser (a special part of the program). The
parser changes the text and develops it in response to what the player typed
into the machine. The player, or rather his (unspecied) game representation in
writing, moves through a literary world that emerges with the ow of the written
text. According to the players instructions, a story about the search for treasures
and their defense against thieves, robbers, and other villains unfolds.

Figure 3. Screenshot (PC) from Zork I (Infocom 1982): Plugging the product

In the actual in-game situation shown on the screenshot of Figure 3, we nd

ourselves at the very beginning of the game. Lines 1 to 4 show its title, the
Self-reexivity in computer games: Analyses of selected examples 257

copyright and trademark notes and its serial number. These are the paratextual
elements of the game, as Genette (1993: 1113) has dened such elements in
literature. Beginning the game in this way apparently imitates the beginning of a
book with its impressum page. Lines 5 to 8 describe the scenery, which is West
of House and includes a mailbox. The diegetic5 space is being unfolded. As
the player interacts with the text, the story develops. In line 9, the curious player
is opening the mailbox in front of the house and nds a leaet. After following
the instruction to read it (l. 11), however, the reader nds nothing of narrative
interest, no information about the ctional world of the game, nor any advice that
might help to survive the player in the enigmatic literary environment. Instead,
the message from the mail box is Welcome to Zork (l. 14), followed by further
lines advertising this game of adventure (l. 15).
All this might look quite simple at rst: the game contains an in-game ad-
vertisement for the game itself. However, from a metaperspective, we are faced
with a paradox: an element of the diegetic space (the message in the letter box)
refers to something not situated within but rather outside of this space, but in a
space nevertheless essentially connected with the game space, viz., the space in
which the device of the computer creates the text of the game. The inner space
of the game refers so to speak to its outer space, or its outside appears in the
inside. Indeed, the paradox cannot become reality. The player experiences the
world of Zork as a closed space whose boundaries are the one of ction. There
is no way out of ction into reality for ctional characters. But then, how did
this advertising message get into the letter box? How can the inventors of Zork
become ction?

3. The end of the game in the game and paradoxical

self-advertising: XIII

XIII (Ubisoft 2003) is a so-called Ego- or First-Person-Shooter (FPS). This sub-

genre of the action game is currently highly popular, but it is being discussed very
controversially in the media.6 The players experience the game world through
the eyes of a character whose role they assume. The complete body of this self-
guided character remains invisible like the body of a real person remains partly
invisible to the seeing self. On the screen, only a hand is visible which carries
a weapon. The aim of this kind of game is mainly to kill opponents or to avoid
them in order to reach a goal or achieve some other mission. The aesthetics of
XIII is the one of a comic book design, and the narrative subject is the one of a
secret agent.
258 Bernhard Rapp

Figure 4. Screenshot (PC) of in-game advertising in a scene from XIII (Ubisoft 2003)

In the scene shown in the screenshot in Figure 4, the player has just suc-
cessfully killed or beaten up a guard who lies dead or perhaps only knocked
out on the oor in an ofce room with two computers on a table. At this very
moment, the screen saver function appears on the two computer screens in the
ctional ofce. The screen to the right shows the publishers logo Ubisoft while
the one on the left confronts the player with a message well-known to everyone
who has played computer games: game over. With this kind of product place-
ment in the computer game, the player is faced with a threefold paradox of kind
with which the players found themselves confronted at the beginning of Zork I.
Game over is a message about the end of a game, but the game is not over;
it continues. Read as a message about this game itself, the announcement of its
end in the in-game action is only the rst paradox. Read as an advertisement,
the message (the publishers logo) implies a second paradox, for consumers of
a game need no advertising of this game. (Books occasionally have pages with
advertisements, but never advertise the book which the readers are reading.)
Thirdly, game over is also and ironical or even cynical metaphorical comment
on the guards death whose life ended through the players action, but how can a
machine comment on such an event even though it occurs in the very room in
which the machine is set up?
Self-reexivity in computer games: Analyses of selected examples 259

4. Played player scolding: Barbarian

Barbarian (Palace Software 1987) is a ghting game of the Beat-em-up type.

Designed for the Commodore 64 home computer in the 1980s, it was banned in
Germany by the Bundesprufstelle fur jugendgefahrdende Schriften because of
its realistic and brutal display of violence (Wirsig 2003: 49). In the scenario
of an ancient arena, two warriors are dueling with swords. The game requires
dexterity and quick reaction if the players want to be successful in their strikes
or counterstrikes against their enemies. Figure 5 shows the screenshot of a scene
in which the players did not interact with the game for some time. It was taken in
the double player mode at a moment when neither player players was interacting
with the game. At this moment, the two warriors on the screen address the real
players in front of the computer to blame them for their lack of participation. The
warriors directed in their actions by the players playing seem to be conscious
of what the players do (or fail to do while leaning back). The creatures of ction
are beginning a double play. They do not only play against each other in their
ghts they also seem to play against the player who plays them when directing
their actions on the screen.

Figure 5. Screenshot (C 64) from Barbarian (Palace Software 1987):A protest movement
against lazy players
260 Bernhard Rapp

5. The game in the game: Day of the Tentacle

The last example is from Day of the Tentacle (LucasArts 1993), an Adventure
Game in which the player is faced with pictures and written messages. Usually,
in this genre, the storyline develops as the players, in the guise of the in-play
characters solve riddles, seek precious objects, combine pieces to a whole, etc.
Day of the Tentacle, or DotT, as it is often called, is announced as the ofcial
successor of an extremely popular Graphic Adventure of the 1980s, Maniac
Mansion (LucasArts 1987). In Maniac Mansion, the players task was to put
together a group of kids to rescue a young girl held prisoner by a mad scientist.
DotT follows this plot and develops it further. Again, the player has to enter
the scientists house, and the same funny characters appear, heroes and villains
alike. In the situation shown in Figure 6, the player has just moved his character
Bernard into the room of the scientists son, the character to the right who looks
like Frankenstein. Among various items we notice an old home computer in the
background. To the players surprise, this computer in the game scene can be
used by the player from without. Once turned on, the old Maniac Mansion
game appears on the screen within the screen; the pictures shift completely
from the DotT scenario to the Maniac Mansion scenario. Players unfamiliar
with Maniac Mansion can play it now for the rst time, while those who know
it from their old home computer might be curious to play it once more.7

Figure 6. Screenshot (PC) from Day of the Tentacle (LucasArts 1993; German Version):
A game in the game
Self-reexivity in computer games: Analyses of selected examples 261

6. Findings

In our examples, self-reexivity appeared in a variety of forms. In Monkey

Island 4, there was a metaleptic transgression from the level of the creatures
of ction to the level of their (co-)creators in which the former seemed to in-
terfere in the life of the latter. In Zork I and XIII, we encountered metalep-
tic advertising: the creatures created by the designers of the software indus-
try became the addressees of advertisements for the world of the games in
which they acted and from which they could never escape to buy the product
they were invited to buy. In Barbarian, the creatures directed in their actions
by the players began to direct the players themselves. In Day of the Tenta-
cle the players began to act like characters in the play starting a computer
game within their game. Where does this predilection for self-reexivity in
computer games come from; when did it rst appear; is it a specic of the
Self-reexivity does not depend on a specic medium, period, or genre. There
has been self-reexivity early in the history of literature and the media in drama,
narrative literature, and lm, and thus it is not altogether surprising to nd it in
computer games as well. It occurs rather early in the history of the computer
game in the late 1970s and early 1980s.8 Hence, one can hardly consider it an
effect of maturity of the medium as it has been claimed for other media.
Many self-reexive strategies shown in computer games are well-known
from literature and the movies so that inuence from the literature and lm is
likely. In literature, lm, and computer games, the medium in its material or
technical possibilities and limitations as well as in its historical, cultural, and
economic context have long since been the subject of self-referential reections
and allusions. A signicant parallelism in the self-reexive strategies of games
and literature besides the one of metalepsis is mise en abyme (mirror text), which
occurs in the form of the play in the play (Hamlet) or the movie in the movie
(see Withalm, this vol.).
An important source of self-reexivity in games is the difference between
the world of the players and the creatures of their games. To understand a self-
reexive strategy, the players must be aware of this difference between the
two levels and interact accordingly. Essential to understanding self-reexivity
in computer games are the actions of the players, i.e. the congurative prac-
tices which determine the gaming situation as a combination of ends, means,
rules, equipment, and manipulative action (Eskelinen 2001). Furthermore,
self-reexivity is much indebted to the devices of the ctional world of a
262 Bernhard Rapp

Self-reexive strategies in computer games offer a special bonus to the play-

ers. The close association between self-reference and paradox as well as between
paradox and humor testies to the comic potential of self-reexive devices in

7. Perspectives for further research

Let us conclude with the proposal of some guiding questions for the future study
of self-reexivity in computer games.
(1) What is the subject of a self-reexive strategy? The parallelisms with the
medium lm in this respect suggest the relevance of a typology of interest
to both media (for the lm, see Withalm 1999: 149 and this vol.).
(2) How is a self-reexive strategy enacted? Again, the parallelisms with the
movies promise to be of relevance. A pertinent example from the lm is
the screen passage in which an actor or an object in a movie shown within
the movie transgresses, or at least touches the screen. Another cinematic
technique of self-reexivity relevant to game studies is the device of facing
the camera, which means, facing the audience.
(3) When or in which situation of the gaming process does a self-reexive strat-
egy occur? What has, or must have, happened just before a self-reexive
scene began? Possible candidates for answers to these questions are missing
player input, intense exploration of the game space, cul-de-sac situations,
or critical situations of search for solutions.
(4) To what extent does self-reexivity appear in a computer game? How much
space does a self-reexive strategy occupy? Does it occur only once as
a marginal joke; or does it extend over the whole game, as is the case of
games in games? How can the relation between a computer game and its
self-reexive strategies be described?
(5) What are the functions of self-reexive strategies in the game; are they
intended or can they occur otherwise? What do self-reexive strategies
mean for the computer game as a special form of communication? Can
they help to keep up the magic circle of the game, as in the example of
Barbarian in which the game tries to pull the player back into its spell? Self-
reexive strategies certainly provide extra motivation and entertainment
for the players, but can they also help in building up coherence within an
incoherent structure?9 In the study of the history of computer games, self-
reexive strategies are likely to function as some kind of remembrance of
a hybrid medium that had, until the breakthrough of the internet in the late
1990s, almost no public, institutionalized memory.
Self-reexivity in computer games: Analyses of selected examples 263

These guiding questions concerning the future study of self-reexivity in

computer games are far from delimitating the full scope of the research eld,
but they should be suitable as a rst sketch of the outlines of the eld of further


1. In this essay I use computer game as a collective term for all kinds of game
software that has been developed for contemporary personal computers (PC) as
well as older home computer systems, such as the Commodore 64 or Commodore
Amiga. To delimit the boundaries of research, I decided to exclude video games. This
term refers rather to the category of digital screen-games which run on consoles (for
instance Sonys Playstation 2, Microsofts Xbox or, in the 1980s, Nintendos Nintendo
Entertainment System (NES)); they are usually played in front of a TV or in a living
room environment. (For the differences between computer and video games, see
Rouse III (2001).) Nevertheless there is certainly no clear-cut distinction between
both types of digital games. Most observations and results presented in this article
should be transferable to the sector of video games as well.
2. Neitzel (2000: 54) has commented on the characteristic feature of computer games
(in contrast to all other sorts of games) to double the object of play.
3. For the purpose of this paper, simplications had to be made and other forms of
self-reexivity had to be excluded.
4. The term genre is used here not as a xed category but rather as a unit for basic
orientation without any theoretical claim.
5. For space and time in the narrative universe, see Genette (1994: 313).
6. Since the mid-1990s, games such as Doom, Quake, Unreal-Tournament or the very
successful Counterstrike made also FPS famous and notorious. The still ongoing
debate concerning the inuence of violence and brutality in computer games on
children has focused on these kinds of games.
7. Actually, this is not exactly a hidden feature (a so-called Easter Egg) for the infor-
mation where to nd Maniac Mansion is provided in the DotT user manual. Even
on the package, there is a sticker announcing that Maniac Mansion appears in the
game. In this way, it becomes an economic incentive whose message is: Buy this
one and get another free!
8. Warren Robinetts game Adventure of 1978 (Atari) is seen as the rst game that con-
tains a kind of a self-reexive strategy. The programmer hid his name in a secret room
of the labyrinth game world since Atari was keeping us game designers anonymous,
which I found irritating. Also, I was kind of proud of the game (Robinett 2003: xvii).
9. The argument of the incoherent structure of computer games has repeatedly been
brought forward, for instance by Poole (2000: 5054) and, in a more detailed way,
by Newman (2004: 7190).
264 Bernhard Rapp


Aarseth, Espen J.
1997 Cybertext: Perspectives on Ergodic Literature. Baltimore: The John
Hopkins University Press.
Bateson, Gregory
1983 Eine Theorie des Spiels und der Phantasie. In: Gregory Bateson,
Okologie des Geistes. Anthropologische, psychologische, biologische
und epistemologische Perspektiven, 241261. Frankfurt: Suhrkamp.
Eskelinen, Markku
2001 The gaming situation. Game Studies 1.1, (02.01.04).
Genette, Gerard
1993 Palimpseste: Die Literatur auf zweiter Stufe. Frankfurt: Suhrkamp.
1994 Die Erzahlung. Munich: Wilhelm Fink.
Huizinga, Johan
[1938] 1980 Homo Ludens: A Study of the Play-Element in Culture. London:
Routledge & Kegan Paul.
Kirchmann, Kay
1996 Zwischen Selbstreexivitat und Selbstreferentialitat. In: Ernst Karpf
(ed.), Im Spiegelkabinett der Illusionen: Filme uber sich selbst, 6786.
(Arnoldsheimer Filmgesprache 13). Marburg: Schuren.
Luhmann, Niklas
1996 Die Realitat der Massenmedien. Opladen: Westdeutscher Verlag.
Neitzel, Britta
2000 Gespielte Geschichten: Struktur- und prozessanalytische Untersuch-
ungen der Narrativitat von Videospielen. Weimar: Diss. phil., (16.01.05).
Newman, James
2004 Videogames. London: Routledge.
Poole, Steven
2000 Trigger Happy: Videogames and the Entertainment Revolution. New
York: Arcade Publishing.
Robinett, Warren
2003 Foreword. In: Mark J. P. Wolf and Bernard Perron (eds.), The Video
Game Theory Reader, viixix. New York: Routledge.
Rouse III, Richard
2001 The console and PC: Separated at birth? Computer Graphics 2: 59.
Ryan, Marie-Laure
2004 Metaleptic machines. Semiotica 150: 439469.
Self-reexivity in computer games: Analyses of selected examples 265

Wirsig, Christian
2003 Das Groe Lexikon der Computerspiele: Spiele, Firmen, Technik,
Macher von Archon bis Zork und von Activision bis Zipper In-
teractive. Berlin: Schwarzkopf & Schwarzkopf.
Withalm, Gloria
1999 Der Blick des Films auf Film und Kino. In: Michael Latzer, Ursula
Maier-Rabler, Gabriele Siegert and Thomas Steinmaurer (eds.), Die
Zukunft der Kommunikation: Phanomene und Trends in der Informa-
tionsgesellschaft, 147160. (Beitrage zur Medien- und Kommunika-
tionsgesellschaft 4.) Innsbruck: Studien Verlag.
Part VII. Other self-referential arts
Looking through the computer screen:
Self-reexivity in

Marie-Laure Ryan

As a feature of period style, self-reexivity is generally considered a sign of

old age. When it runs out of stories to tell, of new territories to explore, of
discoveries to make, thought can always turn back upon itself. The postmod-
ern fascination with self-reexivity can be attributed to the sense of pastness
that permeates turn-of-the-century, or rather, turn-of-the millennium culture
a sense that will endure until the new millennium nds its own cultural identity.
But self-reexivity could also be a response to the curiosity aroused by the de-
velopment of a new medium. A case in point is Don Quixote, the foundational
novel of modern Western literature, whose subject matter is the danger of read-
ing too many novels. Janet Murray (1997: 97) ascribes Don Quixotes madness
to the newly introduced practice of silent reading, which is itself a consequence
of the invention of print and of the ensuing spread of the book.
The insecurity of both old cultures and young media regarding their purpose
or direction explains why self-reexivity is such a prominent feature of digi-
tal texts. As part of an old culture, digital texts are caught in the postmodern
episteme, and they participate in the ideological, political and aesthetic preoc-
cupations of their time; while as part of a new medium, they are still unsure of
their contribution to art, to thought, and to culture. Both factors lead to a quest
for identity that takes the form of a playful interrogation of the technology that
supports them. In the present essay I will focus on the patterns of self-reexivity
found in Web-based art (or, arguably the form of new media that has
pursued the scrutiny of its technological foundation the most persistently. To
prepare, theoretically, the ground for this investigation, I will start by offering
an overview of the various forms of self-reexivity.
270 Marie-Laure Ryan

1. Types of self-reexivity

The term self-reexivity covers a wide range of phenomena diversied along

three continuums: the continuum of explicitness, the continuum of scope, and the
continuum of individuation. The continuum of explicitness runs from a strong
pole of literal self-reference through an intermediary zone of self-reexivity to a
weak pole of artistic self-awareness. Literal self-reference is illustrated by the fa-
mous paradox-creating sentence this sentence is false. Genuine self-reference,
as opposed to mere self-reexivity, is a feature limited to semiotic systems ca-
pable of making propositions or issuing commands. Outside natural language,
we nd it in mathematics, for instance in Godels proof of the incompleteness
of axiomatic systems, and in computer code, such as the recursive function that
computes the Fibonacci number series by launching multiple copies of itself.
Images cannot literally refer, since they lack the indexical power of language,
but they can represent themselves through recursive self-embedding. The clos-
est we nd to self-reference in the visual domain are consequently pictures
that contain copies of themselves, as in the heraldry gure known as mise en
abyme, or on the box of the Laughing Cow brand of cheese, where we see a
cow with earrings representing the Laughing Cow box of cheese. The middle
of the spectrum of explicitness is occupied by works that present what I will
call symbolic or emblematic forms of self-representation. Whereas the type of
straight self-reference that we nd in this sentence is false represents nothing
outside itself, symbolic or emblematic self-reexivity represents both the text
of which it is a part, and something situated in the world created or described
by the text. In a narrative text, for instance, the description of an object or a
conversation between characters may both play a role within the plot, and tell
us how the text should be read, and in a poem, a metaphor may both participate
in the concrete thematics of the text, and offer an image of poetry. At the weak
pole of the continuum of explicitness we nd the self-awareness that Roman
Jakobson (1960) calls the poetic function of language. Jakobson divides acts
of communication into six parameters (the sender, the receiver, the message, the
context, the code and the physical channel that puts the sender in contact with
the receiver), and he associates each of these parameters with a specic func-
tion. Among these, the poetic function is the one that focuses on the message for
its own sake. (Here message must presumably be understood as an inseparable
union of form and content.) Verbal art, in other words, is language that attracts
attention to itself, but it can do so in a subtle way, through either pleasant sound
patterns or creative imagery, without explicitly taking itself as referent.
The continuum of scope diversies self-reexivity according to how much
of the text the self-reexive elements capture in their mirror, and how dominant
Looking through the computer screen: Self-reexivity in 271

they are in the global economy of the text. Linguistic self-reference, as we nd in

this sentence is false, illustrates perfect scope, since the range of the indexical
element this encompasses the entire sentence, and since the sentence does
nothing else than reect upon itself. Visual self-reference, by contrast, is always
incomplete. An image can only show the image in which it is embedded if
it also shows itself as part of the larger image; but for this copy of itself to be
faithful, it must contain a third copy, and so on in an innite regression. Whereas
the Laughing Cow box illustrates explicit but incomplete self-reexivity, the
opposite situation is represented by an emblematic text that models itself entirely
through symbolism, as is arguably the case with Mallarmes hermetic poems
about poetry. A text may be self-reexive throughout in which case it becomes
an allegory of itself or blend reexive and non-reexive elements.The scattered
reexive elements may furthermore represent particular aspects of the text,
rather than trying to mirror it in its totality.
The third continuum concerns the focus of the reexive activity. It runs
from texts that reect specically on themselves, highlighting their distinctive
features, to texts that include a broader class in their self-mirroring, such as their
medium or their genre. I will call these two poles individuated and categorial
self-reexivity. As an example of categorial self-reexivity, consider the lexia
This writing from the hypertext Patchwork Girl, by Shelley Jackson (1995),
which reects on the difference between reading from a book and reading on a
screen in a hypertext environment:
When I open a book I know where I am, which is restful. My reading is
spatial and even volumetric. I tell myself, I am a third of the way down a
rectangular solid, I am a quarter of the way down the page, I am here on
the page, here on this line, here, here, here. But where am I now [reading
hypertext]? I am in a here and a present moment that has no history and no
expectations for the future.
Or rather, history is only a haphazard hopscotch through other present
moments. How I got from one to the other is unclear. Though I could list my
past moments, they would remain discrete (and recombinant in potential if
not in fact), hence without shape, without end, without story. Or with as
many stories as I care to put together.
While these remarks outline a theory of hypertext that purports to describe the
medium itself, rather than one of its particular instantiations, Patchwork Girl also
includes self-reexive elements that distinguish it from other works of hypertext
ction: for instance, the text map for the section Crazy Quilt is deliberately
shaped like a patchwork quilt. This image alludes to the narrative thematics of
the text, which describes how a ctional counterpart of Mary Shelley creates a
female monster by sewing together the body parts of various women. The sewing
272 Marie-Laure Ryan

activity of Mary Shelley functions in turn as an allegory of the writing activity

of the author, Shelley Jackson, who stitches together the body of a text out of
heterogeneous (and often recycled) textual fragments. In a movement leading
from individuated to categorial self-reexivity, the shape of the map allegorizes
the particular story, and the story allegorizes the type of writing promoted by
the Storyspace authoring system, with which Patchwork Girl was composed.


By, I mean any artwork available for free on the World Wide Web that takes
advantage of the computer, not only as a mean of production and dissemination,
but also as a support necessary to the performance of the text. In other words, I
restrict to works that need to be executed by code. This denition excludes
any artwork meant to be printed (such as Photoshop art or standard literary texts
posted on the Web), as well as any work sold in CD form (hypertext ction,
computer games), but it accepts both works that can be run directly from the
Web, and works meant to be downloaded and executed on the users computer. was born in the nineties, when the Internet developed from a resource
mainly used by a technologically savvy elite into a widely accessible forum of
mass communication, information, entertainment, and commercial activity. It
represents the revenge of the hackers, who previously owned cyberspace, over
the general public who now crowds (and spoils) the formerly guarded terri-
tory. Most is indeed created by artists with an extensive knowledge of
programming, or alternatively, by teams that include both artists and program-
mers. Fiercely anti-commercial it cannot be sold to collectors and museums,
hung on a wall, or placed on a bookshelf and generally anti-utilitarian,
restores the old cliche art for arts sake to its full meaning. Its spirit is gen-
erally subversive, if not destructive, and its aesthetics tends to sacrice pure
beauty to conceptual interest. The vast majority of the works reproduced in
Rachel Greenes book Internet Art (2004) give little pleasure to the eye, but
the best of them stimulate the mind through the cleverness of their generative
idea. While few of these works directly reect about themselves, a very large
proportion of them alludes to the features, protocols and utilities of the Internet:
browsers, e-mail, and search engines. Others take shots at commercial applica-
tions, such as computer games or, as we will see, graphics programs. It could
perhaps be argued that by commenting on other Internet applications, works of direct reference away from themselves, and do not consequently qualify
as self-reexive. To this objection I reply that a net-supported work that takes
as its subject matter a use of the Internet engages in a categorial form of self-
Looking through the computer screen: Self-reexivity in 273

reexivity, since it is itself a product of the technological environment toward

which it directs attention.
In what follows, I will examine several of the original ways in which
comes to grips with the net: parody, codework, creative destruction, and map-
ping. Though I will classify my examples under one or the other of these head-
lines, they may participate in more than one category.

3. Parody

My example of parody does not come from an artist who specializes in,
but from a distinguished novelist with a predilection for technological subjects:
Richard Powerss novel Galatea 2.2 (1995) deals with articial intelligence
and the Turing test, and Plowing the Dark (2000) with virtual reality. Powerss
web-based story They come in a steady stream now represents for him an
incursion into a new territory. A spoof of e-mail, the story combines reection
on the technological medium with a more individuated form of self-reexivity:
the text not only takes the proliferation of spam as its subject matter, it also
mimics the interface of a standard e-mail program (Figure 1).
When we rst open (or rather, execute) the text we are faced with a dis-
play that looks like a mailbox with various folders: inbox, drafts, sent,
and trash. As the reader clicks on a mail to read it, another message (or
rather, its headline) appears on the screen. At the end of the reading process,
there will be 17 mails in the inbox, but, ironically, none in the trash can, even
though ten of them are spam: the users agency is limited to reading the in-
box, and in keeping with the theme of the story, the ctional system is unable
to lter out the junk. The spam letters run the familiar gamut of pornography,
drug offers, and investment opportunities. Iris Suarez peddles a catalog of sin-
gles available for dating, Cora Triplett advertises http.//,
Christian Mortgages USA tells the user Jesus loves you renance now!,
Candrgs sells 6000 medicines at substantial price savings, Evidence Elim-
inator warns the reader that he is in serious trouble its a proven fact, but
offers an absolutely safe protection against this danger, and I leave the mes-
sage of Manure E. Griddlecake to the readers imagination. In addition to
the junk mail, the mail program is plagued by pop-up messages, which readers
must close one by one before opening a new mail, and each screen contains a
clickable animated ad. Both of these features promote obsessively (but rather
refreshingly, given its non-commercial character) the literary Web site Ninth
Letter of the University of Illinois at Urbana-Champaign, where the story is
274 Marie-Laure Ryan

Figure 1. Screenshot from Richard Powerss They come in a steady stream now

Counterbalancing the humor of the junk mail, the seven legitimate letters,
addressed to the reader by Richard Powers himself, contain a melancholic med-
itation on aging triggered by the spam letters incessant hawking of drugs that
promise to reverse the damage of time. The narrator sees himself on the brink of
a brave new world inhabited by a posthuman species that enjoys eternal youth,
constant state of sexual desire, and perfect memory, but he realizes that, like
Moses, he will never enter this Promised Land:
Lifestyle drugs, theyre called: and who is going to argue? Not you, at 65, the
last member of the last generation of humans still barred from returning to
the garden, the last who will have to grow old, with nothing to look forward
in retirement but Internet come-ons from the eternal future. . . What will it
feel like, to be another species? Nothing that your species might compare it
to. Soon well be whatever comes after people. And puzzled by the hunger
that weve nally outgrown. (Letter 4)
Looking through the computer screen: Self-reexivity in 275

While the spam outlines a dystopic future (at least for those who value our
present condition of pre-posthumans), it also opens windows onto the past by
jogging the memory of the narrator, not through drugs, but, quite inadvertently,
though the randomly generated names of the fake senders. A mail reminds
the narrator of the rst girl he loved at age fteen, and he wonders whether
she ended up as graceful as she began, but unfortunately, her name is too
common to Google. Another message the sixth of the seven letters of the
series mentions a mail that bears the name of a boy from your conrmation
class, struck by lightening when scrambling out of a lake one summer, but all
the narrator remembers about the boy is his auburn hair, his kindness, and his
goofy smile that declared a standing state of total bafement at the passage of
This sentence foreshadows the reversal of times arrow that will happen at the
end of the story, but not before the reader submits to a common Internet ritual.
In the last of the seven letters we read: PLEASE REGISTER. The content you
requested is available only to registered members. Registration is FREE and
offers great benets. The user is asked to enter his e-mail address in a box, and
to submit it by clicking a button. At this point I hesitated, wondering what kind
of plague I would bring upon my system by following these instructions, but in
the end, curiosity prevailed over caution. I was rewarded with a response in the
best tradition of Thank you! You will receive your conrmation
e-mail shortly.
The real e-mail sent to the user consists of a link to an Adobe le that can
be downloaded and then printed. This le contains the text of the previous six
fake mails, together with a very Proustian conclusion. In the new segment, the
narrator recaptures the lost time with a glance outside the window that liberates
him from the dystopic future of the screen, sends him back to the present,
refreshes his memory (without drugs!), and eventually leads to an absorption
of the past by the present, allowing the narrator to relive in its full intensity the
glorious day at the lake before the boy was struck by lightening. By including all
the previously read installments, the nal delivery invites the reader to reect
on the difference between the print and the electronic medium. The text that
came to us as a collection of fragments in the e-mail simulation achieves a
closure and unity in the printable le that gives rise to an entirely new reading
experience. Straddling two media, the text contrasts the continually interrupted
reading that takes place on the screen with the appreciation of the poetic quality
of its language that becomes possible when we hold the whole story in our
hands. The originality of Powerss achievement lies in the complementarity of
the comic experience of the screen version and of the lyrical experience of the
276 Marie-Laure Ryan

print version. In its play with two media, the text is truly more than the sum of
its parts.

4. Codework

Codework is a reaction to the so-called WYSIWYG (what you see is what

you get) aesthetics that has dominated the design of software and operating
systems since the Macintosh personal computer inaugurated the graphic user
interface in the early eighties. Before that time, users communicated with the
machine by typing instructions on the (infamous) command line of the DOS
operating system. These instructions, which had to be memorized and typed
exactly, could be regarded as a high-level programming language. It took some
knowledge of its functioning to operate a computer, and the difference between
user and programmer was much smaller than it is today. With the introduction
of the graphic interface, all the user has to do is to click on an icon to launch an
application, and anything resembling coded instructions becomes invisible. For
the common user, this was a blessing; for the hackers, who saw themselves as
the guardians of an esoteric knowledge, this was a profanation of the machine.
Icons are perfectly opaque buttons, and clicking on them requires no more
knowledge of the inner working of the machine than choosing an item on the
touch-operated menu of your microwave oven. Codework is an attempt to restore
the users awareness of the hidden layers of machine instructions that make it
possible for data to travel from the depth of computer memory to the surface of
the screen.
The play with code in takes various forms. The most supercial with
regard to the deeper layers of computer architecture is a blend of typograph-
ical signs borrowed from human and computer languages. Here are samples
of the pidgin languages invented by two practitioners of this technique, Mez
(pseudonym for Mary Ann Breeze), and Talan Memmott:
prealphanumeric//pre network n-cluded use ov com.put
[ty/llah]ers ofine
n-turr-rest in nework system[ic]z stemmed fromme a more organic base,
collaborationz via real-time eshmeat N n-stallation based
From out of NO.where, Echo appears in the private space of Narcissus.tmp to form a
solipstatic community (of 1, ON) with N.tmp, at the surface. The two machines the
Looking through the computer screen: Self-reexivity in 277

originating and the simulative collapse and collate to form the terminal-I. a Cell.f,
or cell. . . (f) that processes the self as outside of itselfin realtime.
If the hybridization of human subjectivity and computer intelligence proph-
esized by the theorists of the cyborg and the posthuman (Haraway 1989; Hayles
1999) ever becomes reality, this kind of language could develop into the literary
idiom of the new species. But the use of typographic elements borrowed from
computer languages will remain a purely cosmetic phenomenon as long as the
text cannot be run by the computer. For a school of net.artists that includes
Florian Cramer, Eric Andreychek, and John Cayley, codework should not only
address human concerns when read as a text, it should also change the state of
the system when executed as code; otherwise we could just as well read regular
code as a literary text; or feed the binary version of a literary text to the com-
puter as executable program and watch it cause the run-time error of unknown
Yet another form of play with code consists of revealing the actual com-
mands that underlie a text. This was the purpose of CODeDOC, an exhibi-
tion organized in 2002 by Christiane Paul at the Whitney Museum of Ameri-
can Art in New York City. Paul, the adjunct curator of New Media Arts, gave
a dozen artists the assignment to write a computer program whose purpose
was to connect and move three points in space, a theme that could be in-
terpreted either literally or guratively. The exhibit inverted the usual hier-
archy between code and output, by making visitors (as well as users of the
Web site where the project now resides) scroll through the code le, until
they reached a button at the bottom that triggered the execution of the pro-
The projects vary widely in their faithfulness to the given theme, and most of
them limit self-reexivity, beyond the fact that the code is made visible before
its output can be experienced, to the embedding of a description of the purpose
of the program as non-executable comments within the code le. But two of the
projects carried self-reexivity beyond telling us look, Im made of code by
creating an individuated connection between the code and its output.
In the rst of these two projects, Jack & Jill by John Klima, the code
produces an imitation of the low-resolution computer games of the eighties,
such as Lode Runner or Donkey Kong. The task of connecting and moving
three points in space is ingenuously and humorously fullled by turning the
three points into the protagonists of the well-known nursery rhyme Jack and
Jill. The purpose of the game is to enact the plot of the nursery rhyme, by taking
Jack and Jill up a slope to fetch a pail and by making them tumble down the hill,
once the pail has been reached (Figure 2).
278 Marie-Laure Ryan

Figure 2. Screenshot from John Klimas Jack & Jill

In contrast to standard computer games, the user cannot use the keyboard
to control the characters, but he can inuence their movements indirectly by
assigning values to a number of variable parameters: the choice of a Chau-
vinist or Feminist attitude decides which character is ahead of the other; the
assignment of an intensity value to Jacks and Jills desire controls the speed at
which the characters climb the hill (with a low desire, they never get to the pail);
and the specication of pail allure (which gives a choice of repulsive, moder-
ate or undeniable) dictates the magnetic force exercised by the pail. To win the
game, the user must nd the proper combination of values for the parameters.
The game is too easy to really challenge the player, but the real programming
coup lies in the duplication of the game story by the text of the code. In other
words, the story is both dramatically enacted on the screen, and verbally nar-
rated in the code. In contrast to most of the other projects of the exhibit, Jack
& Jill makes it rewarding, not only to look at the code, but to actually read
Looking through the computer screen: Self-reexivity in 279

Sub Main()
The Story.Show
While True
If YourAttitude = CHAUVINIST Then
If Fetch(pail, jack, jill) then GoUpHill jack, jill
If FellDown(jack) and BrokeCrown(jack) then TumblingAfter
jill, jack
Else YourAttitude=FEMINIST Then
If Fetch(pail, jill, jack) then GoUpHill jill, jack
If FellDown(jill) and BrokeCrown(jill) then TumblingAfter
jack, jill
End if
The Story.Draw
End Sub

What enables digital code to tell stories (or to produce poetry) is the fact
that computer languages consist of two types of elements: names and opera-
tors. While the operators are expressed through a xed vocabulary of reserved
words specic to the language, the names (which stand for variables, constants,
programs and subprograms) can be freely chosen by the programmer. In the
Jack & Jill example, the story is suggested by the variables Jack, Jill and Pail,
as well as by the subprogram names Fetch, FellDown, BrokeCrown and Tum-
blingAfter, but the operators If. . . Then are detrimental to narrative meaning,
because a story is a report of facts, and as such, it cannot be told, at least not
literally, in the conditional mode (even less through embedded conditionals).
The only operator that contributes to the narrative reading is =, which can be
read as the verb to be. It would be an extraordinary achievement to enroll both
names and operators in the production of a story, and Klima can be forgiven for
not achieving what is probably an impossible feat.
While in Jack & Jill the code mirrors the story told in the output, Brad
Paleys Codeproles performs the reverse operation: here the output of the
program is an image of its own code. Not only does the program display a
listing of itself, it also fullls the requirement set by the organizers of the exhibit
by moving three points across the display according to a logic described in a
comment section of the code le:
// This code reads in its own source and displays it in a tiny font, then //
// It moves three points in code space. It essentially comments on itself .//
// The white Insertion Point traces the code in the order it was written. //
// The amber Fixation Point traces word by word as someone might read it. //
// The green Execution Point shows a sample of how the computer reads it. //
// The code lines themselves gradually get brighter as they execute more. //
280 Marie-Laure Ryan

Figure 3. Screenshot from W. Bradford Paleys Codeproles (detail)

Figure 3 shows a portion of the screen (three columns out of four). The amber
point corresponds to the bright area in the left column. It runs linearly from the
rst to the last line, and then returns to the top, simulating the reading of a stan-
dard print text. If the user moves the cursor on one of the lines, it is magnied and
made legible; if the user clicks, the execution restarts from there. The trajectory
of the white point corresponds to the curved line that runs all over the image; at
the moment shown in Figure 3, the point is highlighting text in the third column.
Writing code is always a relatively linear process, because programmers must
simulate in their mind the operation of the computer, which takes and executes
the instructions sequentially, but a well-structured computer program consists
of various self-contained modules, known as procedures or subroutines, which
can be written in any order. This freedom explains the capricious arabesques of
the white line. The movements of the green point trace the order of execution,
whose sequentiality is frequently broken by commands implementing transfers
of control, such as go-to statements and calls to subroutines that make the pro-
gram jump across computer memory, where the instructions are stored before
being brought to the processor to be executed. When I captured the program, ex-
Looking through the computer screen: Self-reexivity in 281

ecution followed a loop represented by the triangle between the rst and second
Though Codeproles takes self-reexivity further than any of the other
projects of the CODeDOC exhibit, the author claims in a discussion of the pro-
gram available on the exhibits Web site that it was not written to be computer-
clever, nor postmodern reexive, but to compare and contrast three modes of
parsing: the laymans, who is tempted to read the le linearly, like an ordinary
text; the programmers, who composes the code module by module in a rela-
tively free order; and the computers, whose order of execution bounces back
and forth between modules and travels code space in all directions.

5. Creative destruction

There is perhaps no better way to make people appreciate what they have
or rather, what they had than to take it away. Alan Liu (2004) suggests the
term of creative destruction for the application of this principle in art. A
practice that originated in Dadaism and Surrealism but exploded in new media,
especially in, creative destruction draws attention to cultural, commercial
and technological phenomena by taking them apart.
In Auto-illustrator, Adrian Ward combines the idea of creative destruction
with parody and reection on code into a humorous piece of dysfunctional soft-
ware. Auto-illustrator (Figure 4) mimics graphic programs, such as Photoshop
or Corel PhotoPaint in the same way Richard Powerss text mimics e-mail, but
with the signicant difference that the interface is actually operative: you can
produce your own artwork by using the program, and you can even buy a li-
censed copy, which contains more features than the free demo version available
on the Internet. The main reason for buying a license is to support the cause
of, for I cannot imagine that anybody would have sufcient need for
Auto-illustrator to pay to $100 for it. But dont expect to enjoy the program for
a long time if you dont buy the license: every time you run your free copy, its
performance deteriorates, until you become unable to do anything with it.
Auto-illustrator subverts the utilitarian spirit of commercial software by turn-
ing the graphic tools into autonomous agents with a will of their own. If you
select the freehand pencil tool, the system does not use the position of the mouse
cursor to draw a line, but rather follows its own rules, merely taking clues from
your mouse coordinates. The exact nature of these clues remains a mystery:
the line you draw stubbornly refuses to follow the line you wanted to draw. If
you select the text tool, the system picks the letters, inventing nonsense words,
and your control is limited to making a selection among the options terse,
282 Marie-Laure Ryan

Figure 4. Screenshot from Adrian Wards Auto-illustrator 1.2

verbose, creative, and slightly foreign. The square and the oval tools let
you draw regular geometric shapes, but it gives you a choice between shabby
and precise shapes, as well as between childish, artistic, and regular.
Artistic does not draw anything there is no such thing as an artistic square
or circle, according to the program but childish brings delightful surprises:
the circles will be funny faces, and the squares will be turned into the kind of
houses that a four-year-old may draw (especially if you combine the childish and
shabby options). As for the bug tool, it will place moving creatures randomly
on your screen, and they will create art for you by crawling around and drawing
lines. If you dont like the result, a tool will let you exterminate the creatures.
The parody of serious art programs extends to the systems comments on the
choices of the user (this tool is boring), and to the zany options offered on the
preferences menu: here the user can click boxes labeled Death penalty for
poor designs, Exta-verbose KJX routines, Do cool things. Her curiosity
will be tested with a Pandora box labeled dont push this button. If she suc-
cumbs to the temptation, the programs behavior will become totally erratic, but
fortunately, licensed users can undo the damage by hitting a certain key on the
next run of the program.
Looking through the computer screen: Self-reexivity in 283

The rebellious behavior of the tools of Auto-illustrator reminds the user of

the additional level of mediation that distinguishes drawing on paper with hand
and pencil from drawing on a screen with computer software. In a graphics
program, the hand does not draw, but rather activates a hidden code. The user of
commercial, utilitarian software takes it for granted that the code listens to her
input: if she selects the straight line tool, she does not expect the program to draw
an arabesque. Auto-illustrator breaks this basic contract between the software
designer and the user, and draws attention to the hidden code by complicating
(rather than severing) the relationship between the movements of the hand and
the behavior of the tools. The program does listen to the user, but it does so in an
indirect, unpredictable way. This disturbingly lively machine to paraphrase
a much quoted formula by Donna Haraway (1989: 176) does not produce a
frightfully inert user, to conclude the formula, but on the contrary, distributes
authorship among three agents: the programmer, who designs the code and
invents imaginative new tools, the computer, whose unpredictable operation is
regulated by random numbers invoked by the code, and the user, who retains
modest control over the picture by choosing tools and colors, by letting the
program duplicate or animate objects, and by deciding when the output is worth
saving as an artwork.
Auto-illustrators reection on code does not take the form of making it
directly visible, but rather, of asserting the artistic dimension of the program-
mers activity. In other words, it is not codework, but rather, what Christiane
Paul calls software art (2003: 124). In an article included in the users guide
to Auto-illustrator, Florian Cramer observes that in commercial applications,
programmers are frequently considered to be mere factota, coding slaves who
execute other artists concepts (2002: 102). Software art liberates programmers
from the tyranny of corporate work by letting them express their own vision,
using code as a meta-medium to control other media: language, sound, color,
shapes, and animation.

6. Mapping

The development of maps of cyberspace by this I mean visual representations

of the information contained in the Internet is an area of teeming activity, both
in and in practical programming. The mapping projects inspired by the
Internet (many of which are shown in Dodge and Kitchins fascinating Atlas
of Cyberspace [2001]) range from purely functional navigational tools through
projects that combine usefulness and artistic self-awareness to artworks totally
devoid of practical purpose. Here I will discuss an attempt to map the Internet
284 Marie-Laure Ryan

aimed at resolving a paradox that has fascinated authors as illustrious as Lewis

Carroll and Jorge Luis Borges: the paradox of the map that achieves perfect
self-referentiality by becoming indistinguishable from the represented territory.
A fusion of map and territory would necessitate a complete image of a ter-
ritory at a 1 to 1 scale that includes the map itself. Why 1 to 1? Because any
reduction would require the omission of some features. And why should the
map be part of the territory? Because if it werent, it would point to something
external to itself: one can for instance imagine a complete map of the earth at a
1 to 1 scale spread out on another, larger planet. Both of these conditions lead
to paradoxes. As Borges has shown ([1951] 1983: 195196), if a map is part
of the world, it can only represent the world completely by representing itself,
which means that it must represent its own self-representation, in an innite
regression similar to the case of the Laughing Cow box of cheese. Moreover, if
the map were at a 1 to 1 scale, it would cover the whole world, and according to
Lewis Carroll this would lead to inevitable contradiction. A perfect map should
contain an image of every blade of grass, but if it were spread out over the world,
the sun would be blocked, the grass would die, the farmers would be mad, and
the map would be unfaithful. And if the map were not spread out. . . it could not
be consulted, and it would become useless. Carroll suggests, tongue in cheek, a
luminously simple solution to this problem: So we now use the country itself,
as its own map, and I assure you it does nearly as well ([1893] 1982: 726.). But
if we think of maps as navigational aides, this is a ludicrous proposal, because
we would have to traverse the territory to see what its map looks like, when in
fact the purpose of maps it to help us nd our way in the territory.
The digital artist Lisa Jevbratt proposes to reconcile functionality and ex-
haustive coverage of the territory with a mapping of the Internet appropriately
titled 1:1. According to comments by Jan Ekenberg posted on the projects Web
site, 1:1 becomes not only the map, but the environment itself. Referring to
Lewis Carroll, whose text is quoted in his commentary, Ekenberg concludes:
Lets hope the farmers dont object. Jevbratts own on-line description of the
project concurs with Ekenbergs assessment: The interfaces/visualizations are
not maps of the Web, but are, in some sense, the Web. They are super-realistic
and yet function in ways images could not function in any other environment or
The project that inspires such hyperbolic statements is an attempt to visu-
alize the Web as a system of IP addresses. The IP address of a Web site is the
numeric translation of its domain name; in other words, what is for human users could be for the computer Since IP ad-
dresses are made of four eight-bit words, for a total of 32 bits (or at least were
made in 1999 and 2001, when 1:1 was created), there could be as many as 232
Looking through the computer screen: Self-reexivity in 285

distinct pages on the Web; but many IP addresses are not claimed, and attempts
to reach them leads to the message: cannot nd server, or DNS error. Other
addresses are claimed, but the user is not authorized to access them.

Figure 5. Screenshot from Lisa Jevbratts 1:1 (detail)

The 1:1 project consists of ve different visualizations, but I will limit my

discussion to Every, the design that makes the strongest claim of being the Web
itself (Figure 5). To produce her images, Jevbratt used Web crawlers programs
that search the Web address by address to determine which IP numbers have
active servers. The crawlers returned 186 100 active addresses for the sampled
areas, and each of these addresses is represented on the screen by a distinct pixel.
The pixels are color-coded on the basis of the numerical value of the address
they represent, so that, by looking at the image, one can tell the distance (in
numerical value) between occupied addresses: sharp contrasts in color mean
that there are large intervals between active IPs, gradual contrast means that a
region is densely populated. Each pixel is a hot link, and by clicking on it the user
can reach the corresponding IP. This provides an interface to the Web radically
different from the modes of navigation offered by standard browsers. As Jevbratt
explains on the projects Web site, Instead of advertisement, pornography, and
286 Marie-Laure Ryan

pictures of peoples pets, this Web is an abundance of inaccessible information,

undeveloped sites and cryptic messages intended for someone else. The user
gets an idea of how small the proportion of the information stored on the Web
is publicly accessible, and of how much the Web has changed since the creation
of the project. I clicked about 20 times on the visualization, and my random
selection yielded only one accessible web site: the home page of Marjorie Orr,
top international astrologer.
How should we understand the title 1:1? One obvious interpretation is that
each unit on the screen corresponds to a distinct IP address, in a one to one
relation. But this relation is very different from the scale of a map, where 1:1
means that a certain area of the map corresponds to the same area in the world.
The units on the screen are made of one pixel, but they stand for addresses
made of 32 bits. Nor can we interpret 1:1 as meaning that the design represents
the information available on the Net in its totality. The image on the screen
admittedly provides access to every active IP address, but we have to traverse
the image to see the content of these addresses, which means that we cannot see
all of this information in one glance, as a map would let us do. Nor does the
visualization show what makes the Web a web: the complex system of links that
interconnects its various elements.
All this should make it clear that, while 1:1 could in principle be extended to
cover the entire address eld of the Web, it remains a long way from achieving
the self-referentiality inherent to the claim that it is not a map of the Web, but
rather the Web itself. By subjecting Jevbratts comments to a critical assessment,
rather than accepting them at face value (as most commentators seem to do, with
the exception of George Dillon), we learn that the tendency of conceptual art to
produce auto-descriptions does not guarantee the validity of these descriptions.
A representation of an artwork is liable to be considered inaccurate, whether it is
contained in the artwork itself, or describes an external referent. But if 1:1 does
not really fuse the map and the territory, it remains an impressive achievement
in data visualization, not only because it reveals the hidden geography of IP
addresses in this sense it is truly a map but also because its combination
of representation and active interface to Web sites creates a type of image that
could only exist in a digital environment.

7. Conclusion

Let me return, in conclusion, to the question of what makes self-reexivity so

dominant in I believe that we cannot achieve a proper understanding
of self-reexivity in art in general, and in new media in particular, without
Looking through the computer screen: Self-reexivity in 287

taking into account the force that it is trying to resist, namely the immersive
power of representations and their ability to create an illusion of reality (Ryan
2001; Wolf 2004). The self-reexivity of Don Quixote was a warning against
the tendency of readers to immerse themselves in the world of chivalric nov-
els, and to mistake these ctional worlds for reality. In the nineteenth century,
the development of the powerful illusionist techniques of realism led the novel
away from self-reexivity, and steered it back toward immersion, until postmod-
ernism denounced any attempt to make the medium invisible (a prerequisite to
immersion) as robbing the reader of his critical faculties. For those who regard
immersion as a low-brow pleasure (unjustly in my view, for the experience re-
quires a highly active involvement of the imagination), replacing transparent
windows into imaginary worlds with the mirrors of self-reexivity is a proven
key to artistic respectability. It is indeed by developing self-reexive features
that computer games, a fundamentally immersive use of digital technology, have
recently tried to promote themselves as an art form to be taken seriously.
In contrast to the novel and to computer games, never developed
immersive features; what it is trying to undermine is not its own power to create
illusion, but rather the kind of immersion in digital technology that limits our
attention to the surface of the computer screen, and fools us into believing that
we fully control this technology, when in fact our agency is restricted to what
the system was programmed to let us do. As part of this attempt to provoke
reection on the role of digital technology in our lives, lls the World
Wide Web with images and inverted images of its own utilities. By inspiring,
enabling, and hosting these multiple and varied images, the Web as a whole
becomes a system that thinks about itself. Do not expect to grow into
an immersive art form any time soon: there are already enough of these in the
media landscape. For, reecting on its supporting medium is not a search
for identity, it is identity.


Borges, Jorge Luis

[1952] 1983 Partial magic in the Quixote. In: Donald A. Yates and James E.
Irby (eds.), Labyrinths, 19396. New York: Modern Library.
Carroll, Lewis
[1893] 1982 Sylvie and Bruno. The Complete Illustrated Works of Lewis
Carroll. Ed. Edward Guillaro. New York: Avenel Books.
CODeDOC exhibit.
2002. (08.02.06).
288 Marie-Laure Ryan

Cramer, Florian
2002 Concepts, notations, software art. Auto-illustrator Users Guide: 101
112. Downloadable from (08.02.06).
Dillon, George
2002, 2003 .Writing with Images: Toward a Visual Semiotics of the Web,
Dodge, Martin and Rob Kitchin
2001 Atlas of Cyberspace. New York: Addison-Wesley.
Ekenberg, Jan
2001 Prologue to 1:1, to 1/jan.html (08.02.06).
Greene, Rachel
2004 Internet Art. London: Thames & Hudson.
Haraway, Donna
1991 Simians, Cyborgs, and Woman: The Reinvention of Nature. London:
Hayles, N. Katherine
1999 HowWe Became Posthuman:Virtual Bodies in Cybernetics, Literature,
and Informatics. Chicago: University of Chicago Press.
Jackson, Shelley
1995 Patchwork Girl. Hypertext software. Cambridge, MA: Eastgate Sys-
Jakobson, Roman
1960 Closing statements: Linguistics and poetics. In: Thomas A. Sebeok
(ed.), Style in Language, 35077. Cambridge: MIT Press.
Jevbratt, Lisa
2001 1:1, to 1/index ng.html (08.02.06).
Klima, John
2002 Jack & Jill,
Liu, Alan
2004 The Laws of Cool: Knowledge Work and the Culture of Information.
Chicago: University of Chicago Press.
Memmott, Talan
2000 Lexia to Perplexia, memmott/
Mez. [Mary Anne Breeze]
2000 The Art of M[ez] constructing polysemic & neology c/facti-
ons online, (08.02.06).
Looking through the computer screen: Self-reexivity in 289

Murray, Janet
1997 Hamlet on the Holodeck: the Future of Narrative in Cyberspace. New
York: Free Press.
Paley, W. Bradford
2002 Codeproles,
Paul, Christiane
2003 Digital Art. London: Thames and Hudson.
Powers, Richard, Jenifer Gunji, Joseph Squier, Jessica Mullen, Lauren Hoopes, Chad
Kellenberger and Val Lohmann
2004 They Come in a Steady Stream Now, (08.02.06).
Ryan, Marie-Laure
2001 Narrative as Virtual Reality: Immersion and Interactivity in Litera-
ture and Electronic Media. Baltimore: The Johns Hopkins University
Ward, Adrian
2003 Auto-illustrator 1. 2. Downloadable from Signwave: (08.02.06).
Wolf, Werner
2004 Aesthetic illusion as an effect of ction. Style 38(3): 32551.
The artist and her bodily self:
Self-reference in digital art/media

Christina Ljungberg

1. Introduction

Questions of self-reference have always been fundamental to art. Iconic self-

reference could even be said to be typical of the aesthetic sign, since one of
its characteristics is that it calls attention to various aspects of itself, above all,
its sensuous qualities and formal structures, actual materiality and rhetorical
strategies. At least, this becomes evident and explicable if we approach self-
reference in general and the self-referentiality in aesthetic signs in particular with
the help of C.S. Peirces doctrine of signs. Peirces second trichotomy of signs
(icon, index, and symbol), based upon the character of the relationship between
a sign and its dynamical object, is especially illuminating here since it provides
(1) a way of understanding reference in terms of indexicality (the indexical sign
being dened as that in which there is a spatio-temporal or causal relationship
between sign and object) and (2) a way of understanding self-reference (at least
in part) in terms of iconicity. All iconic signs are self-referential, which could
appear paradoxical since signs should really stand for something else. The reason
why a sign can represent a sign is explained by Winfried Noth in the introductory
chapter to this volume:

the Peircean object which a sign represents does not necessarily have an
extension, and it does not need to be a piece of the so-called real world at
all, since signs or ideas can be the object of a sign. The object of the sign
is something which precedes and thus determines the sign in the process
of semiosis as a previous experience or cognition of the world. (Noth, this
vol., I.3).

In other words, in this way, the signs referent can be another sign, and self-
reference can be a chain of signs referring to other signs.
292 Christina Ljungberg

Digital art is self-referential primarily because its images are mathemati-

cally generated. It is, put differently, a synthetic image presenting a numerically
based reality that we can see on a computer screen only because the screen
is composed of pixels, which are small discreet fragments that correspond to
numerical values. These values enable the computer to appoint a precise posi-
tion in a two-dimensional screen space within a Cartesian coordinate system to
which chromatic coordinates are added. As anyone who has been working with
digital photography knows, each pixel is a separate entity that can be fully con-
trolled and changed. In digital art, the computer can even synthetically generate
a picture; as Lucia Santaella points out, the numerical image is under perpetual
metamorphoses, oscillating between the image that is actualized on the screen
and the virtual image or innite set of potential images (1997: 126). This image
is highly iconic: there is no analogy between the algorithms that generate it and
the image on the computer screen. In addition, as she notes with reference to
Arlindo Machado (1993: 117), there are two keywords for synthetic images:
model and simulation. A model is a diagram, which, in Peircean semiotics,
is not only a subcategory of the icon, but which is also ideal for testing: a di-
agram or model can generate experiences that are not real but formalized
and repeatable [. . . ] calculations. Therefore, in Santaellas words, the essential
features of the synthetic image lie in its virtuality and simulation (1997: 127).
Another factor contributing to the self-referentiality of digital art is that
artists working in digital media generally put an emphasis on the very process
involved in the production of such art. This is something that digital art shares
with other postmodern art forms. It becomes particularly evident when hybrid
forms of art and media are used, which heighten the degree of self-reference: the
switching between or among various media not only forces its viewing or, rather,
participating audience1 to make comparisons among them but it also exposes
the particularities of the various semiotic systems that each media embodies. In
the case of virtual reality, the aids you have to use and your obviously techno-
logical environment make you aware of the necessary procedures that virtual
reality still entails. In addition, there is increased self-reference since digital art
often focuses attention on the artist and her bodily self, as both generating and
participating in the work of art. At the same time, works of digital art have a
marked indexical ingredient, too, in the sense of referring to other real works,
contexts or bodies. Even in virtual reality, an awareness of the physical body
is necessary for orienting ourselves in and understanding the particular digi-
tal work of art. Hence, there are varying degrees and forms of self-reference
characteristic of various types of digital art and media.
How can these various degrees of self-reference be determined? What are the
kinds of self-reference typical of digital art and how can they be differentiated?
The artist and her bodily self: Self-reference in digital art/media 293

That is what my contribution will attempt to chart, with the examples of artists
working in various kinds of digital art and media: the multi-media works of visual
artist and performer Laurie Anderson, video/digital artist Selina Trepp, and
media artist Char Davies whose interactive installations immerse participants
in an all-enveloping virtual reality in the ow of life through space and time
(Davies 2004: 70).

2. Self-reference in multimedia

Laurie Anderson is one of todays premier performance artists. Known primarily

for her multimedia presentations, she has cast herself in roles as varied as visual
artist, composer, poet, photographer, lmmaker, electronics whiz, vocalist, and

Figure 1. Scene from Laurie Andersons Songs and Stories from Moby Dick

Andersons performances show several forms of self-reference. There are

those of intertextuality, in the sense of referring not only to works by other
authors/artists and contexts but also in the form of repetitions and recursions in
the text that refer back to the artists earlier work; and there are those of inter-
294 Christina Ljungberg

mediality, in which the confrontation with hybrid forms of media increases the
degree of self-reference. Intertextuality would seem to be a dening feature of
her performances, which contain not only direct references such as Songs and
Stories from Moby Dick or Home of the Brave (a late 1940s movie about a war
veteran), but also explicit musical quotes (for example, from Hector Berlioz,
Jules Massenet, Frank Sinatra, Elvis Presley or Dolly Parton); quotes from lms
and TV series, such as her famous performance White Lily based on Rainer
Werner Fassbinders Berlin Alexanderplatz; literary quotes in abundance, for
example, William Burroughs (who also has done the vocal part in several of her
works); Italo Calvino, Don de Lillo, Buckminster Fuller, Franz Kafka, Thomas
Pynchon, William Shakespeare, and Ludwig Wittgenstein, just to mention a
few. In Moby Dick (Figures 1 and 2), Anderson takes Melvilles narrative tour
de force and translates it into a multimedia performance that attempts to mir-
ror the complex encyclopedic structure of the work by creating several visual
levels working in counterpoint to auditory ones. According to Anderson, ap-
proximately ten percent of the show is actually Melvilles text: she nevertheless
faithfully retells the tale of Captain Ahabs monomaniacal pursuit of the white
whale. Some passages are quoted in their entirety, and in yet other ones the sense
of Melvilles words are rephrased. Since the major intertext of Moby Dick is the
Bible, Andersons intertextuality inescapably refers to the tropes, themes, and
gures derived from Melvilles own intertextual universe.
While Anderson closely follows Melville in terms of narrative and imagery,
her performance also suggests certain ongoing thematic concerns in her work re-
fracted here through her reading of the literary classic. Anderson (1993) has long
explored the question that Melville raises concerning the human predicament
and human values, which in her work manifests itself though her ambivalent re-
lationship to the technologies she uses in her performances. In previous works,
for example, she has created duplicates of herself including a male video
clone and a digital puppet as if to suggest that the technology on which
her work depends may ultimately usurp her own presence, transforming a real
person into an unnecessary duplicate of different gender or into an inorganic
replica. This ambivalence also includes her use of sound: the talking stick, an
instrument she especially designed for the performance, is not only a wireless
instrument that can access and replicate any sound but can also disassemble
sound into tiny segments, called grains, and then play them back in different
ways, on the principle of granular synthesis. The computer then arranges the
sound fragments into continuous strings or random clusters which are played
back in overlapping sequences to create new textures, much like pixels.
Quotations and allusions are generally considered referential, since they refer
to something else, an object, in a different context. That would make Andersons
The artist and her bodily self: Self-reference in digital art/media 295

Figure 2. Laurie Anderson performing Songs and Stories from Moby Dick against a
projection of quotes from Melvilles novel

literary quotations referential and indexical since they have a special context in
mind. On the other hand, when her music refers to other pieces of music, and
the visuals to other visuals, and not to any world beyond the world of music
or visual representation and I would argue that Andersons music and visuals
do precisely that her quotations of other musical and visual works are self-
referential. In addition, her work is pervaded with repetitions and recursions of
words, phrases or ideas which are typical and striking forms of self-reference
whether in music, texts or images/lms since they always refer back to the
preceding instances. So are her reuse and the quotes of her own work, which
make up much of her oeuvre and ought to be pure self-reference.
But her work is also characterized by the use of different media, which would
mark it as eminently intermedial. Examples of intermediality are a lm referring
to a book, or an opera to a theatre play, which has the quoting and the quoted sign
differ in medium. As with intertextuality, intermediality can and often does refer
to a different sign in a different medium; but when it functions so as to create
a different dimension for those who recognize the quote what Winfried Noth
(this vol., I.5.5) calls an intermedial deja-vu effect it is self-referential. This
pertains to a high degree to Andersons work, too: not only does she constantly
296 Christina Ljungberg

quote from different media but her quotes also challenge her spectators to search
for their original sources, which will give them access to the different levels of
Therefore, in Andersons performances, it is the various expressions of self-
reference that mark them as typically postmodern and contribute to their high
degree of self-reference. At the same time, the artists pervasive presence in her
work gives her performance a pronounced referential and indexical character.

3. Self-reference in digital manipulation

What is the relationship between reference and self-reference in a performance

that involves digital manipulation? In the case of Selina Trepp, for instance,
a visual artist who uses a wide range of technology lm, video, computer
animation for her performances, the presence of the artist functions as a link
to the indexical and referential while its digital manipulation would seem to
determine its character as predominantly self-referential.

Figure 3. Split screen video still from Trepps and Bitneys performance Spectralina

Trepps Spectralina (Figure 3) is a computer animation; its music and im-

ages were created in collaboration with the musician Dan Bitney of Tortoise and
Trepp. The general tenor is sometimes abstract and poetic, sometimes political
with inspirations from the artists life together. This declared autobiographi-
cal source of origin gives the performance a strong authentic and referential
character from the start. The performance consists of Bitney playing several
instruments, while Trepp is composing computer-animations utilizing Arkaos,
The artist and her bodily self: Self-reference in digital art/media 297

a new video jockeying software which enables the creation of real-time and
recorded visuals and offers an interesting way of editing and manipulating a
visual performance while it is running.
Using a video still, Trepp and Bitney are sitting in front of a video camera
going through a video mixer (analog), creating a split screen (Figure 3). Then
they move their heads in unison trying to create a seamless transition: a new face
made up of half of each face, created by their looking into the monitor adjusting
the position of their faces in real time as they see it in the screen merging into
one. Trepp herself calls it actually a really lo- real-time method of image
manipulation (2005).
At the same time as the performance involves manipulation and distortion
which always involves self-reference (and iconicity) as it refers back to itself
there is still a relation of contiguity, an indexical correspondence between the
signs on the screen and the objects they refer to, albeit a distorted one. One could
even say that the image oscillates between the iconic and the indexical, or the
self-referential and the referential, as the faces contort and change expression
and shape. It is precisely this oscillation that gives the performance its particular
tension, making it doubly self-referential: not only does it bring questions of
authenticity and manipulation into play, as it refers back to the original, but
also the relationship of art and representation to reality that has characterized
Trepps work from the very start.

4. Self-reference in virtual reality

Digital or postphotographic images belong, in the words of Lucia Santaella

(1997: 130), to a both individualized and, at the same time, global transmission
of information. This becomes nowhere more apparent than in virtual reality
(VR), in which images can be both indenitely stored and increasingly accessed
from everywhere and at all times. In addition, these images only make sense
in interaction in which those taking part can often not determine whether they
are looking at an image, or whether the image is looking at them. A particularly
interesting specimen of VR is the one created by the team headed by media
artist Char Davies, who has also written an extensive amount of criticism on
both digital art and media.
Char Davies became well-known for her VR installations Osmose (Figure 4)
and Ephemere (Figure 5) which challenge conventions about representations of
articial worlds and interface. In contrast toAnderson and Trepp, who are mainly
performance artists, Davies works exclusively in immersive virtual space, a
computer-generated articial environment that one seemingly enters with the aid
298 Christina Ljungberg

of various technical devices. But Daviess synthetic worlds do not look like the
customary polygon-dominated space of traditional 3-D computer graphics. They
are populated with organic shapes that suggest plants, landscapes, body, and
water. Similarly, the interface for navigating the world indicates the breathing
and balance of the visitor.
Why are so many artists fascinated with VR? Davies (2004: 69) calls her
own work a subversion of conventional approaches to VR on the basis that
they reinforce an outdated dualist worldview. She has written numerous pa-
pers and made several presentations on her particular philosophical focus on
the experience of the physical body in cyberspace. In her work, she explores
paradoxes of embodiment, being, and nature in immersive virtual space. She
uses this new medium as a philosophical arena for constructing architectures of
enveloping material form, working with transparency, luminosity, spatial am-
biguity and temporality as well as a body-centered user interface of breath and
balance, with the intent of emphasizing the role of the subjectively felt by a
physical body in virtual space.
In immersive virtual environments like those created by Davies, the agent of
production is no longer an artist who leaves the mark of his or her subjectivity
and ability on the surface of a support, nor a subject acting on reality, though
he or she may transmute it by means of a machine. Digital images are above
all interactive: they enable the creation of an almost organic relationship to
those with whom they interact, in an interface that is both immediate bodily and
mental. It is a communal activity; yet still deeply private, which is also what
attracts artists. For example, Davies (2004: 73) says that she wants to re-present
the world behind the veil of appearances as immaterial, interrelated and
dynamic ux [making] habitually perceived distinctions between things dissolve
and boundaries between interior self and exterior world become permeable and
intermingled. Instead of a visually determined world, she creates a radically
different spatiality in which normally perceived boundaries between objects and
surrounding spaces are dissolved in light, disposing of the usual perceptual cues
by which we objectify the world.
In digital art, the abolishment of reference to the real world as we know
it therefore also changes the role of the artist. The digital artist is, above all,
a programmer whose visual intelligence interacts with the potentials of arti-
cial intelligence using technology to prosthetically blur the boundaries between
different realities,2 for example, the spherically-enveloping environment cre-
ated through the use of HMD (head mounted display) that Davies designed for
Osmose (Figure 4). Using transparency and luminous particles, Davies bases
the interface on breath and balance to allow participants to simply oat by
breathing-in to rise, to fall and to lean to change direction. In addition, the
The artist and her bodily self: Self-reference in digital art/media 299

hands-off interface frees participants from the urge to handle things and from
habitual gravity-bound modes of interaction and navigation.

Figure 4. Char Davies, Forest grid, Osmose (1995), digital image captured in real-time
through head-mounted display during live performance of immersive virtual environment

Daviess longstanding career as a painter becomes particularly evident in

her second famous VR installation, Ephemere (Figure 5), which she describes
as an exploration of the ephemerality of being and the symbolic equivalence
of body and earth (2004: 83). The works iconography is grounded in nature
as metaphor: archetypal elements of root, rock, and stream recur throughout.
Spatially, the work is structured into three parts: landscape, subterranean earth,
and interior body.The body of esh and bone functions as the substratum beneath
the fertile earth and the natural seasonal processes of the land.
The images in digital works of art in VR such as those by Davies are highly
self-referential: they are all self-generated and, although they look like fantastic
nature photography, they are infographic images, a numerically based reality.
What is interesting is that, in her creating synthetic images, Daviess goal is not to
project articial worlds but to remind people of their connection to the natural
(rather than man-made) environment not only biologically but spiritually and
300 Christina Ljungberg

Figure 5. Char Davies, Body (Egg), Ephemere (1998), digital image captured in real-time
through head-mounted display during live immersive journey/performance

psychologically, as regenerative source and mythological ground (Davies 2004:

75). As she points out, her method involves circumventing the conventions
of linear perspective, Cartesian space and objective realism [. . . ] in order to
collapse a culturally-created distance between subject-viewer and the world
(2004: 75).3 Accessing her VR installations, the participant or interactor rst
enters a three-dimensional grid (Figure 4), which then soon fades, leaving her
or him in a non-Cartesian place where everything is dematerialized and semi-
transparent there are no solid surfaces, hard edges, or separate objects in empty
space. Daviess images thus operate under the sign of metamorphosis as they
open up a portal to new and virtual worlds to those participating. Put in a safe
but unfamiliar environment, these virtual worlds allow participants (so far, more
than forty thousand) to experience their bodies and their perceptual faculties in
new ways, quite akin to meditation and mystic experiences. Common comments
by participants are that it feels like being in another place or like losing track
of time.4
The lack of reference that characterizes the virtual world would therefore
seem to enable participants to cut loose their bearings and take part in an al-
most other-worldly experience. They nevertheless need a physical body for the
interactive experience, which means that they need to be indexically, that is, ref-
erentially anchored. In putting on the HMD prosthesis which is what allows
her or him to enter the virtual world the participant becomes, in so doing, a
biocybernetic body, divided into two complementary media: one body which
remains carnal and real in the environment in which it exists, and its avatar,
which is the virtual, disembodied projection of the real body (Santaella 2003).
Obviously, in Daviess installations, we seem to momentarily lose ourselves in
The artist and her bodily self: Self-reference in digital art/media 301

cyberspace but once we take off the HMD, our physical body remains carnal
and real. That is what makes it possible for us to maintain proprioception, the
sensation of self from within the body. Although the medium of digital art is
fundamentally self-referential and may seem virtually non-indexical, there must
still be reference in order for us not to lose ourselves in cyberspace.
To conclude, the degree of self-reference in the digital arts varies according to
its mode of production and performance. Three ways of gauging self-reference
can be discerned. In the case of Laurie Anderson, degrees of self-reference can
be measured in terms of intertextuality and intermediality, including striking
forms of self-reference such as recursion and repetition. In Selina Trepps digital
manipulation, the degree of self-reference can be determined by the oscillation
between the referential and the self-referential, the indexical and the iconic as
Trepps subjects are both the role of representation and the role of the artist.
Finally, the highly self-referential virtual reality created by Char Davies enables
participants to immerse themselves in a virtual world functioning as a portal to
other spaces and realities. In so doing, it not only challenges our habitual modes
of perception, interpretation, and evaluation but also functions as a means to
create an alternative awareness. The self-referentiality at the heart of such work
overturns preconceived patterns of reference.


1. One major difference between digital art and other visual art forms (e.g., its pre-
decessors painting and photography) lies in its means of transmission and how this
affects the role of the addressee. Whereas objects of prephotographic art forms are
mainly contemplated in special places designed for this purpose, e.g., museums,
churches, and galleries, as unique pieces of art, photographs, being innitely repro-
ducible, belong to the space of mass media. In contrast, not only can digital art forms
be accessed anywhere and at all times but they also demand interactivity on the part
of the addressee (Santaella and Noth 1998: 175).
2. One of the reasons why contemporary artists are fascinated with VR may be as a way
to confront the challenge identied by Verena Conley (1993: xii) in the preface to her
Rethinking Technologies, in which she asks how critics of culture, philosophers, and
artists will deal with technologies in view of the menaces of the twenty-rst century:
How do they contend with expansionist ideology, and the accelerated elimination of
diversity and of singularities? How do they resist and act? [. . . ] Now, in a world where
the notion of space has been completely changed through electronic simultaneity,
where the computer appears to go faster than the human brain, or where virtual
reality replaces reality, how do philosophy, critical theory, or artistic practices
deal with those shifts?
302 Christina Ljungberg

3. Davies (2004: 73) has repeatedly spoken out against conventional VR and its ten-
dency towards disembodiment: As a realm ruled by mind, virtual reality as con-
ventionally constructed is the epitome of Cartesian desire, in that it enables the
construction of articial worlds where there is the illusion of total control.
4. Davies draws on psychological theories of deautomatization (Deikman 2005) that
suggest that destabilizing psychic structures can both achieve increased attention
and perceptual expansion.


Anderson, Laurie
1993 Stories from the Nerve Bible. New York: HarperCollins.
Conley, Verena
1993 Preface. In: Verena Conley (ed.), Rethinking Technologies, ixxiv.
Minneapolis: University of Minnesota.
Davies, Char
2004 Virtual space. In: Francois Penz, Gregory Radick and Robert Howell
(eds.), Space in Science, Art and Society, 69104. Cambridge: Cam-
bridge University Press.
Deikman, Arthur
2005 Deautomatization and the Mystic Experience, (06.09.2005).
Machado, Arlindo
1993 Maquinas e imaginario. Sao Paulo: Edusp.
Santaella, Lucia
1997 The prephotographic, the photographic, and the postphotographic im-
age. In: Winfried Noth (ed.), Semiotics of the Media. State of the Art,
Projects and Perspectives, 121132. Berlin and New York: Mouton
de Gruyter.
Santaella, Lucia and Winfried Noth
1998 Imagem: Cognicao, semiotica, mdia. Sao Paulo: Iluminuras.
Trepp, Selina
2005 Email correspondence 3 and 4 July.
Metaction and metamusic:
Exploring the limits of metareference1

Werner Wolf

1. Introduction: Metareference as a transmedial eld and the

apparent absence of music in it

In reections centered on self-reference in the media music is perhaps the

medium which should come to ones mind rst. For unlike literature and other
representational media music has only very restricted possibilities of pointing
beyond itself. In other words, it has difculties in engaging in alloreference2 or,
as I prefer to term it, heteroreference. Instead, musical compositions abound
with self-reference, at least in the Western tradition of art music, which will be
in the center of this essay. Indeed, each repetition or variation of a theme within
a fugue or a sonata can be regarded as an instance of self-reference, since a later
occurrence of a theme points back to its earlier appearance or original form. As
Jakobsons poetic function reminds us (cf. 1960: 356), such self-reference can
also variously be encountered in literature (e.g., in the recurrence of semantic
isotopies and themes, but also in lyric rhymes and generally in parallelisms).
However, in verbal texts self-reference hardly ever occurs with the same density
or with such dominance as it does in art music.
Yet in explorations of self-reference the focus is frequently not so much on
such general self -reference as on a special variant, namely metareference.
This is indicated in the current proliferation of terms such as metatextuality,
metanovels, metalms, metapainting, or even metaarchitecture, all of
which were referred to in Winfried Noths call for papers (this vol., Part I)
of the symposium which forms the basis of the present collection of essays.
Interestingly, in this list of potential metamedia music was conspicuously absent
and only occurred in the context of recursion and repetition. All of this the
list and the absence of music in it is no coincidence but indicative of two things:
rst, metareference is a transmedial phenomenon and by no means restricted
304 Werner Wolf

solely to literature and language3 (although these were the media in which it
rst received theoretical attention); and second, music appears to be somehow
located at the margins of the eld of metareference if not beyond its limits.
In the following, I would like to explore these limits and see whether music
occupies a place inside or outside the transmedial eld of metareference, in
other words: whether there is such a thing as metamusic that could be regarded
as an analogy to literary metaction.4 For metaction can surely be said to be
located somewhere near the center of the metaeld, and I will therefore use it
as a point of reference.The remarkably scant research concerning metamusic,5
indeed the all but absence of this term in musicology,6 leads to the expectation
that music should be relegated to the area beyond the connes of metaland.
Compared with this, my thesis is that under certain circumstances, music can be
positioned within this eld. I will argue that music is in fact able to approach the
condition of metareference, albeit only to a limited and often debatable extent
and with more difculties than other media. The following contribution is thus
dedicated not only to an intermedial comparison between ction and music with
regard to self- and metareference, but also to the exploration of the limits of the
metareferential eld as a whole.

2. Self-reference, self-reexivity and

metareference/metareexivity: Terminological distinctions
illustrated with literary ction

Discussing the metareferential potentials of music presupposes the clarication

of the concept of metareference as opposed to self-reference.7 This will be done
in this section, in which relevant conceptual and terminological explanations
and distinctions will be illustrated predominantly with literary ction as this
is the medium for which theoretical investigations of self-reference are most
The rst distinction, self-reference vs. its opposite, is essential for any
reection on self-reference. However, in practice, it is less a strict binary op-
position than a continuum with many gradations in between two poles: One of
the poles, self-referentiality, can be dened as the quality of signs and sign sys-
tems that point to themselves or to identical or similar elements within one and
the same semiotic system in contradistinction to the opposing pole, hetero-
referentiality, which denotes the normal quality of signs, namely to point to
what conventionally is conceived of as reality outside semiotic systems. Of
course, the term system also requires clarication in this context: in the case
Metaction and metamusic: Exploring the limits of metareference 305

of literature, music and other media I propose to differentiate between a nar-

row and a broad denition of a semiotic system: in the narrow sense it would
coincide with a given medial work (a literary text, a musical composition, a
painting, a lm). I call self-reference operating within these narrow connes
intracompositional self-reference. In contrast to this, self-reference in the broad
sense operates within the entire area of the media and arts. This variant, which I
term extracompositional self-reference, includes intermusical references be-
tween different compositions as well as intermedial references, for instance the
relation between literary texts and music embodied in verbal descriptions of
musical compositions.
Independently of the extension of the self-referential system which has just
been outlined, self-reference can occur in basically two variants. This second
basic distinction has rarely been made in research on self-referentiality but is
crucial for interpretation from a functional point of view:8 self-reference can
mean, rst, the fact that a sign (system) merely points at itself or to similar (or
identical) elements within the same system and, second, a signifying practice
that creates a self-referential meaning, in other words, elicits a cognitive process
or reection on itself, on other elements of the system or on the system as a
The rst case covers a vast eld, namely all variants of self-reference that
do not consist of, or imply, a self-referential statement. It comprises simple
symbolic references, e.g., intracompositional grammatical self-reference (e.g.,
between noun phrase and related relative pronoun) as well as extracompositional
intertextual and intermedial references (where a pre-text or another medium is
merely identied or mentioned in a text, e.g., as a part of its ctional world). In
addition, it is also applicable to all sorts of self-referential, extra- and intracom-
positional iconic references or similarities, for instance recurrences, intertextual
quotations or imitations (as opposed to symbolic identications of pre-texts) and
the mise en abyme of storytelling in stories within stories.9 Part of these variants
correspond to devices which Jakobson mentioned in explaining what he termed
the poetic function of texts (cf. 1960). Moreover, mere self-referential point-
ing at includes devices that foreground the sign (sequence) in question through
deviations. All of this, however, remains mere self-reference as long as it only
serves to classify, e.g., a text as belonging to literature or advertising etc.
and remains below the threshold of (intentionally) triggering certain reections
in the recipient: reections that are centered on the medium and related issues.
Triggering self-referential reections on (elements of) the semiotic system
under consideration, on other semiotic systems or on semiosis and the media
in general is the characteristic functional feature of the second main form of
self-reference, which I want to concentrate on here. For this variant, which goes
306 Werner Wolf

beyond mere self-reference in the above sense, the term self-reection should
be reserved, as it bears the connotation of a cognitive activity. This activity
is caused by implying if not by explicitly containing self-referential state-
ments. In order to elicit such semantic self-reection the aforementioned devices
of self-referential pointing at may be used and to that extent the border be-
tween self-referential pointing at and self-reection may appear fuzzy10
but other devices may also be employed. Following Scheffel (1997) one could
further differentiate according to whether such self-reection focuses on hetero-
referential elements that happen to occur within the same system11 or on the
medium as such and related issues. However, an exhaustive discussion of all
the systematic intricacies is impossible within the framework of the present
essay (for an overview see Figure 1); instead, the following remarks will con-
centrate on the latter kind of self-reection for which the terms metareection
or metareference are appropriate.

Figure 1. Reference in literature and other media

Such metareection always implies an awareness of the medial status of the

work or system under consideration and thus also an awareness of a logical dif-
ference between a metalevel and an object level. This consciousness concerns
the recipient as well as the author and the work. In fact, metareference like
so many other critical concepts in the eld of the media is actually a bipolar
Metaction and metamusic: Exploring the limits of metareference 307

phenomenon: it is not restricted to simple givens within a work, text or arte-

fact these are mere potentials that may have metaeffects but metareference
also requires the actualization of such potentials by recipients who are willing
and able to cooperate, for it is in the recipient that the essence of metareference,
the eliciting of a medium awareness, takes place. Metareference can thus be
described by the following three distinctive traits:
(1) the existence of an intrasystemic reference (self-reference);
(2) the semantic quality of this intrasystemic reference; in other words: the
fact that it consists of, or implies, a self-referential statement and is thus
(3) a kind of medium awareness on the part of both producer and recipient
which is implied or explicitly thematized in self-reection and, thus, gives
it a metadimension (this also implies the existence of a logical difference
between the object level and the level of the metastatement).
It should be noted that such metareference is rst and foremost applicable to
individual phenomena within certain works (metaelements). Yet, if metaphe-
nomena become salient features of a work as a whole, one may speak as has
been done in the aforementioned call for papers of a metatext, a metadrama
etc., and if several metaworks exist within one and the same medium, they may
even be said to form a metagenre. Thus, metaction can refer to individual
passages of a novel, to a novel as a whole or to a novelistic genre.
Metaelements occur in a remarkable variety of forms, for which some ty-
pologies have been devised with reference to ction. However, I would like to
concentrate on only four pairs of oppositions, which I have derived from research
on metaction by Linda Hutcheon ([1980] 1984) and myself (Wolf 1993: chap.
3.2). In order to ensure the transmedial applicability of these forms, the original
terminology shall, however, be adapted so that it avoids an exclusive reference
to ction. The four pairs of forms are:
(1) intracompositional or direct vs. extracompositional or indirect metarefer-
(2) explicit vs. implicit metareference,
(3) ctio vs. ctum (or medium- vs. reference-centered) metareference,
(4) critical vs. noncritical metareference.
Individual metaelements can, of course, be classied according to more than
one of these pairs of categories.
The rst of these pairs has already been introduced in the discussion of the
extension of self-reference. In ction, intracompositional metareference can,
for instance, be observed in metalinguistic comments by a narrator on his or her
308 Werner Wolf

style, while extracompositional metareferences include parodies of preexisting

texts, but also metaremarks that are not or do not seem to be immediately
applicable to the work in which they occur. An example of the latter can be
found in Sternes Tristram Shandy, namely in the narrators complaint about
readers who are eager to be let into the secrets of the story: I know there are
readers in the world [. . . ] who nd themselves ill at ease, unless they are let
into the whole secret from rst to last (Sterne [1767] 1967: 3738). Of course,
such indirect metareference is frequently only a disguised form of the direct,
intracompositional variant. Yet this indirection nevertheless merits a separate
typological entry because it can trigger different responses in the recipients.
The second opposition, explicit vs. implicit metareference, refers to the se-
mantic distinctness of the metareference as a quotable or nonquotable element:
Thus, the numerous discussions of storytelling in, for instance, Sternes Tristram
Shandy are all examples of explicit metareference, for they contain quotable
metareferential phrases such as my reader or my work (Sterne [1767] 1967:
9495). Such metareferential expressions are apt to remind the reader of the
(print) medium as such. In contrast to this, there are more covert devices which
may also elicit reections on the ontological status of the text as a medium with-
out, however, using explicitly metareferential expressions. In Tristram Shandy
such implicit metareferences can, for instance, be observed in the manifold ty-
pographical devices which not only foreground the conventional symbolic use
of novelistic language by deviating from it through the employment of iconic or
indexical signs but also imply an awareness of the medial conventions as such.12
Implicit metareference shows the necessity of the recipients cooperation in a
particularly clear way, for it is in principle possible to overlook the metaim-
plication of, for instance, typographical devices and consider them as a mere
ornament or oddity.13 Consequently, markers are requisite in order to ensure a
metareferential reception. Such markers can vary in their obviousness and can
range from the foregrounding of medial devices to the supplementary employ-
ment of explicit metareference in the vicinity of implicit elements14 (the many
metalinguistic and metactional comments on the uses and abuses of language
in Tristram Shandy fulll this condition).
The third pair of opposing terms, ctio vs. ctum metareference, uses the
content of the metareection as its criterion of differentiation. In all cases
metareference per denition elicits the idea of mediality and of the ontological
status of the work in question as an artefact. I have termed this medium centered
facet of the ctionality which is thereby implicated ctio metareference in
order to distinguish it from the special case of ctum metareference. In the lat-
ter, optional variant an additional facet of the term ctionality comes to bear,
namely a certain relation to reality. As a rule, ctionality in this sense denotes
Metaction and metamusic: Exploring the limits of metareference 309

a negative reference to reality, i.e., a merely imaginary quality. It should,

however, be noted that the reference centered variant of metareection under
discussion can also extend to positive relations to reality (e.g., in suggestions of
authenticity), although indications of a merely imaginary reference are perhaps
more frequent, particularly in recent literature. A by now classic example of this
can be found at the opening of the much quoted chapter no. 13 of John Fowless
novel The French Lieutenants Woman, where the narrator admits: This story I
am telling is all imagination. These characters I create never existed outside my
own mind (Fowles [1969] 1977: 85).
The quotation from Fowles leads to the fourth, also content related pair of
terms, since it not only exemplies ctum metareference but also self-critical
metareection. In the face of a tendency to overstress such critical laying bare
of the works ctionality so often encountered in criticism (in particular on
postmodernism) one should, however, emphasize that metareference can also be
noncritical. Noncritical metareference can be used, for instance, for explaining
aesthetic innovations but also in order to suggest that the story one is reading
is authentic: such an assertion of the truth of a story would be a noncritical
ctum metareference.

3. Applicability of metareference and its subforms to music:

A case study

In literary ction, the application of the concept of metareference in all of its

typological subforms does not present difculties. In the realm of music, a com-
parable ease can only be found where music appears together with words. In fact,
in all kinds of vocal music, metareference is not much of a problem either
thanks to the support of verbal language. Thus, songs can use explicit meta-
musicality by thematizing singing and music making,15 and metaoperas (such
as Wagners Die Meistersinger von Nurnberg or Richard Strauss Ariadne auf
Naxos16 ) and metamusicals (such as The Phantom of the Opera) can comprise
extensive comments on, and presentations of, musical and operatic activities.
Yet as soon as we come to the eld which will be in focus here, namely
pure, instrumental music (mostly referred to in the following as music),
metareference becomes quite problematic. Examples are rare, and the appli-
cation of the metaforms outlined in the preceding chapter with reference to
ction becomes debatable. Indeed, before one can engage in a discussion about
whether instrumental music can realize such metaforms, the question must be
answered whether it is susceptible to metareference in the rst place. As stated
above, in the rst chapter, music as such is certainly the most self-referential
310 Werner Wolf

medium thanks to its ability to create intrasystemic similarities on the level of

its signiers. However, music is also the medium which has most difculties in
relating signiers to signieds. This holds true of hetero-referential signieds
but also of metareferential ones. Indeed, the virtual impossibility of explicit mu-
sical statements rules out at least one of the aforementioned metaforms, namely
explicit metareference: music cannot explicitly and quotably comment on its
own medium in the way ction can.17 This leaves us with the variant of implicit
metareference as the only one we can explore for metamusic.
As stated above, this variant is by itself not unproblematic. Yet I would like to
claim that under certain circumstances, music can in fact approach the condition
of implicit metareference. An example may substantiate this: it is a short com-
position in three movements probably for organ and attributed to Johann Se-
bastian Bach18 entitled Kleines Harmonisches Labyrinth [Little Harmonic
Labyrinth] (BWV 591). Douglas R. Hofstadter, in his once famous book Godel,
Escher, Bach: An Eternal Golden Braid, used it as an, albeit abortive, illustration
of musical recursivity.19 My thesis is that this composition is not so much an
example of recursivity but of metamusic and that it implies a metareferential
statement, namely: listen to what extent and in what short time one can lose tonal
orientation! This idea is chiey triggered by the device of salient deviations
from the traditional style of baroque composition. These deviations playfully
foreground an aspect of the musical medium as employed at the period, namely
tonality. This is achieved by chromaticism and breathtaking modulations, in par-
ticular enharmonic ambiguities (chords that have an instable meaning because
they can be conceived of as belonging to different tonal scales at the same time).
According to the musical conventions of the period, compositions started
in the key in which they were written, as a rule by opening with the tonic and
often also by initially establishing the key through a full cadence, and they
also nished on the tonic of this key. In between, modulations could occur,
but were usually employed sparingly (in particular as long as well-tempered
keyboard instruments were not yet in common use that solved the problem of
older instruments, which in remote keys produced notoriously jarring chords).
Thus, within the 52 bars that comprise the Kleines Harmonisches Labyrinth,
contemporaries would perhaps have wandered off to no more than two or three
related keys.
Bach almost ironically overfullls the convention of establishing his key. In
our case, this is the key that can most easily be played on keyboard instruments,
namely c major. Within the rst six bars all Bach does is establish c major with
one intermittent excursion into the neighboring key of f major. But from bar 7
onwards, he leads the listener astray in an incredible way into a labyrinth of
foreign keys and ambiguous dissonances: until bar 10, where we arrive at a d
Metaction and metamusic: Exploring the limits of metareference 311

minor chord, that is, at a chord that can again be formed in a c major scale, we
are exclusively confronted with chords that can not be formed within c major.
What is more, c major is left in the short time of only three bars and replaced
by such remote keys as c sharp minor (notated with four sharps), g sharp minor
(ve sharps), d sharp minor (six sharps) and g minor (two ats).

Figure 2. J. S. Bach, Kleines Harmonisches Labyrinth (BWV 591), bars 111

Figure 2 illustrates only a small part of the harmonic intricacies which oc-
cur throughout this musical labyrinth from its entry (Introitus) to its center
(Centrum) right through to the exit (Exitus) until the last few bars. In his
conclusion, Bach again obeys the rule that one ought to end the entire com-
position in the home key and returns to c major while the rst movement
(Introitus) ends on c minor and the second one (Centrum) on g major. In-
terestingly, the salient deviation principle that operates in this composition as a
marker of metareference can also be seen on the level of motivic coherence or
rather in the relative lack of such a coherence. With the exception of the middle
part, a fugato on a chromatic theme, this composition appears to be surpris-
ingly heterogeneous with only very loose and unobtrusive motivic unity (in the
Introitus the motiv f-d-e, which occurs in the soprano in bar 5 for the rst
time, is repeatedly varied, but there is no counterpart to this in the Exitus).
So the focus is clearly not so much on the recurrence and variation of themes
and motives as on the foregrounding of the musical system of tonality, which is
metareferentially laid bare as such.
Of course, all of this only works for listeners who know about the conven-
tions and are able to perceive modulations as something extraordinary (unfor-
tunately, owing to later, in particular Romantic, music both chromaticism and
modulations have become such staples of musical composition and have been
so exhausted by overuse that much of the startling effect Bachs composition
must once have had has worn off for us). Thus, as with literary parodies, which
also heavily rely on contexts, the perceptibility of the implied metareference
312 Werner Wolf

of Kleines Harmonisches Labyrinth depends on a specic context and the

recipients ability and willingness to actualize it.
However, as far as the recipients willingness is concerned, Bachs compo-
sition contains powerful incentives: The extreme deviation from the traditional
employment of tonality as a means of foregrounding this crucial element of
the contemporary musical compositional system was arguably in itself a strong
stimulus to trigger metareferential reections. Yet, it is neither the only trigger
nor the only metamarker: the unusual title must also be taken into account. As
opposed to conventional titles such as prelude and fugue or fantasy it ex-
plicitly points to the alienating employment of tonal harmony. However, as
a verbal text, this paratext already leaves the realm of pure music. The use of
language arguably indicates a feeling, on the part of the composer, that a clear
thematization of the metareferential content of his composition was perhaps
necessary in order to ensure the establishment of a metalevel in the listeners

4. Further potential forms of instrumental metamusic

The discussion of Bachs Kleines Harmonisches Labyrinth has shown that

metamusic is an option which is not restricted to vocal music but can in fact
extend to instrumental music. As will appear, it even occurs in many, albeit not
all of the subforms of metareference outlined above and in a surprisingly wide
variety of compositions. Indeed, the Kleines Harmonisches Labyrinth with
its intracompositional metareference is not a unique example of metamusic.
Arguably, all kinds of compositions that are based on salient deviations20 and
all compositions that sport some sort of virtuoso use of musical material and
point to music as such are also, at least in principle, open to metareference.
Thus, theme with variations is a well known form in which composers not
only employ their professional skill but openly demonstrate it thereby implying
something like the following metacomment: Listen, what I have made out of a
simple tune (the theme)!
A similar metareference could be surmised not only with respect to the
composing of music as a cognitive activity but also and perhaps foremost to some
kinds of music as performance. Indeed, wherever composers or improvisers such
as Paganini and Liszt but also Jazz musicians leave the constraints of, for
instance, thematic work and formal conventions and engage in performances that
explore the limits of what is still playable on a given instrument, we arguably
enter the metadomain, for the focus of such activities is more on the medium
as such (the instrument and its player as well as the musical system) than on
Metaction and metamusic: Exploring the limits of metareference 313

the musical message. Here it is indeed true that the medium becomes the
message at least in the listeners mind.
In all of these cases, the problem is, however, where to draw the line be-
tween real metamusic and the usual kind of art music that frequently also
employs unusual techniques and brilliant devices. Certainly, not all deviations
are metareferences that somehow foreground the medium as such. Neither is
all exquisite thematic work meant to metareexively foreground the through-
composed quality of a piece of music. As in verbal metaction, the implied
variant always poses difculties, and as with implicit metaction, metamusic
which is always implicit also requires clear markers. As a rule, these markers
are intermedial ones, that is, they are taken from the verbal medium, as is the
case in the title of Bachs Kleines Harmonisches Labyrinth: when such mark-
ers point to certain musical qualities we may feel justied in suspecting some
kind of metaquality.
This is also true of an area in which a search for metamusic may prove most
promising, namely in the realm of what, following a phrase coined by Adorno
(see [1949] 1975: 165189), has been called music on music. The term sounds
remarkably analogous to one of the shorter denitions of metaction as c-
tion on ction which have been employed in literary studies (Hutcheon [1980]
1984: 1). Yet, curiously, in musicology it has rarely, if ever, acquired such meta-
connotations and has mostly been restricted to intermusical references or stylis-
tic imitations. It thereby apparently excludes such compositions as Kleines
Harmonisches Labyrinth, in spite of the fact that, strictly speaking, this com-
position is also music on music. Be that as it may, there nevertheless seems to
be an emergent, as yet covert awareness of the richness of this eld, at least in
musicology in German, as the recent publication of Schneiders Lexikon Musik
uber Musik: Variationen Transkriptionen Hommagen Stilimitationen
BACH shows. In many compositions mentioned in this encyclopedia the
quality of music on music is already indicated in the title, as for instance in
Brahmss Variationen uber ein Thema von Haydn (op. 56a).21 In spite of the
restriction in scope and a lack of explicit discussions of metamusic Schneiders
Lexikon22 is a unique tool for identifying potential examples of metamusic.
As the title of Schneiders book indicates, homages to other composers, no-
tably to Johann Sebastian Bach, form an important part of music on music.
We indeed approach the realm of metamusic to the extent that such homages
not only refer to composers as human beings but to their styles of composition.
This is particularly the case where recognizable references not only to individ-
ual works but also to stylistic features of a composer or even an entire musical
epoch are discernible (in literary intertextuality theory this would correspond to
intertextual system reference [cf. Broich and Pster 1985: ch. III.2]). For the
314 Werner Wolf

understanding of such compositions requires the awareness of a historical di-

mension of the musical system as such. A good example of such an intermusical
system reference, which also involves a meta-awareness of music as a histori-
cally developing system and consequently also of anachronisms, is Prokoevs
famous Symphonie classique. This work, composed in 1917, is, according to
the composer, an audible reference to Haydn with some modern elements (cf.
Schneider 2004: 154). The metafacet of the work is clearly indicated in its title,
which points to the traditional nature of the work.
If Prokoevs Symphonie classique is also a perhaps somewhat nostalgic23
homage to Haydn as well as to the classical symphony at large, this composition
combines intracompositional metareference (its audible anachronism) with ex-
tracompositional metareference (the homage allusion to a system of the past).
Owing to the positive nature of this allusion the Symphonie classique can at the
same time be employed as an example of noncritical musical metareference.24
One may now ask whether metamusic similar to metaction can also
contain critical metareference. I think it can.25 One may think of carnivalesque
musical pastiches in which the style or whole passages of usually classical music
are playfully imitated and distorted in order to achieve a comic effect. One may
also think of humorous self-critical devices in which musical conventions are
too openly fullled or used as a basis for playful deviation. Haydns often self-
ironic music as epitomized in the second movement of his Symphonie mit dem
Paukenschlag provides good classical examples in both respects, but one may
also think of Mozarts Dissonanzenquartett or Stravinskys backward-oriented
music on music, which was, at least for the highly critical Adorno, an example
of how music could be employed in a destructively parodistic way.26
As has already become clear, there is a whole range of possibilities between
noncritical homage and the kind of destructive parody Adorno had in mind.
In contemporary postmodernist compositional practice a brilliant example of a
playful kind of metamusic which implies an ironic self-criticism that is more
tongue-in-cheek than destructive is Friedrich Guldas Concerto for myself
composed in 1988. The ironic distance which this composition creates towards
itself is twofold: on the one hand and this is the surprise at the beginning
Gulda employs the anachronistic language of classical art music and, in the
course of the concerto, fascinatingly (and again nostalgically) imitates styles
from Bachs age up to the nineteenth century. In 1988, this anachronism was
even more salient than in Prokoevs time. On the other hand, Gulda comple-
ments the external distance between his work and contemporary musical style by
internal distance markers: these consist in the tensions created by the heteroge-
neous combination of classical elements and borrowings from twentieth-century
popular music, in particular jazz, borrowings that audibly undercut the serious-
Metaction and metamusic: Exploring the limits of metareference 315

ness of the classical imitations. The ironic distance thus created correlates with
the cognitive distance that is always implicated in metareference. For the re-
sulting, not quite neoclassical concerto for piano and orchestra may once again
be said to presuppose a musical medium awareness, in particular a historical
one, namely the competence of identifying the different compositional styles,
forms and devices as well as their historical incongruity. This awareness is most
powerfully activated in the remarkable transitions between different styles. It
is here that the concerto most clearly reveals its truly metamusical nature: it is
a composition that, in its ssures, openly lays bare an aspect of its ctio na-
ture, namely its pastiche character. Yet, at the same time it is also a homage to
concerto music of the past and an equally metamusical self-celebration of the
compositional and performative skills of the man who played the piano part
and had proudly referred to himself in the title, namely Friedrich Gulda. Thus,
the ironic, metareferential distance which informs the entire concerto does not
actually aim at self-destruction but is ultimately employed in a self-protective
way which enables Gulda to once again revive old compositional forms without
incurring the reproach of regressive restoration.27
To sum up, it has hopefully become clear that musical metareference can
occur in some forms which are analogous to metactional forms: extra- and in-
tracompositional, critical and noncritical. As for the pair of oppositions implicit
vs. explicit metareference it must be remembered that only the implicit variant
can be actualized in music. This leaves us with the remaining pair of forms that
are potentially analogous to metaction: ctio vs. ctum metareference.All of the
musical examples mentioned imply a comment on some ctio aspect, yet none on
the ctum dimension (the ctionality or truth of an artefact). This is no coin-
cidence since music cannot create ctional, invented worlds like ction, painting
or lm. Therefore, all metareference in music must be restricted to questions re-
lated to the medium, its production and reception but cannot deal with questions
of real or ctitious reference, in other words, it can only be ctio metareference.

5. Conclusion: Other neglected areas within the eld of

metareference and perspectives for further research

The preceding exploration of the limits of the eld of metareference has shown
that this eld also contains a medium which has so far been neglected in this
context, namely instrumental music. Yet, music is not the only medium that has
suffered from such neglect. While the metapotentials of literature and to some
extent also of lm (cf. Stam 2000) and painting (cf. Stoichita [1993] 1997)
have met with various degrees of attention, this is much less so with respect
316 Werner Wolf

to other media such as comics, sculpture, and architecture. Indeed, the eld of
metareference is by far not yet as cultivated as it deserves to be. There is even
a whole range of issues that are still waiting to be discussed. Some of these
are being treated in this volume, some have already found a certain amount of
interest elsewhere (cf. Hauthal et al. 2007) but would merit more attention, while
others must be left to future research. The following survey of perspectives for
research on metareference may serve as guidelines:
Formal problems, metareference, and related phenomena:
verifying and, if necessary, revising or elaborating on, the proposed typolo-
gies of self-referential and metareferential forms
exploring the connection between metareference and common forms of self-
reference such as intertextuality, intermediality and mise en abyme
investigating metalepsis as a transgeneric and transmedial special case of
examining strategies for the naturalization of metareference (from a trans-
medial as well as a monomedial perspective)
exploring forms of marking, in particular, implied metareference (cognitive
framing of metareference)
Intensication of transmedial research on metareference:
expanding the perspective on areas (media and genres) that have hitherto
been more or less neglected, such as: lm (in particular animated cartoons
and animation lms), the arts (including sculpture), architecture, comics,
computer games and virtual realities
intensifying research on (instrumental) metamusic (my preceding remarks
could be no more than an introduction to the eld)
Functional history of metareference:
exploring the functions of metareference as a phenomenon that, in the West-
ern world, has informed more and more works and media since the eighteenth
century as well as investigating the conditions that further (or hinder) an in-
crease in metareference
concerning metareference and postmodernism: research in this eld should
begin by relativizing this connection, which is frequently seen as too ex-
clusive; yet one should also attempt to account for the intimate correlation
which nevertheless can be observed between postmodernism and metaref-
erence; the fact that metareference has recently inltrated even popular
Metaction and metamusic: Exploring the limits of metareference 317

media and genres such as comics and mainstream lm to an amazing extent

is of particular interest and in need of explanation (it may well be that the
contemporary increase in metareferentiality is a symptom of a late stage in
cultural development if not of decadence)
examining metareference in the literatures, arts and media across cultures
(metareference as an object of comparative cultural studies)
exploring metareference in the context of other discourses such as the sci-
ences, philosophy and epistemology
investigating the contents of metareference from a historical, monomedial
as well as transmedial perspective
carrying out research on the relationship between metareference, rational and
comic distance (the earliest as well as the most intensive forms of metaref-
erence occurring predominantly in comic genres and media)
continuing the research on the relationship between metareference and aes-
thetic illusion (cf. Wolf 1993)
providing a critique of metareference: the gains and losses of metareference
(as opposed to hetero-reference)
In literary studies, the history of metaresearch is at least half a century old
(if we take Wayne C. Booths 1952 essay on the self-conscious narrator as a
starting point) or even older (if we consider Russian formalism as the point of
departure). But even after such a relatively long period, there is still much new
ground for future research. This is particularly the case since Western culture
which, e.g., in philosophy and notably in epistemology, has shown marked meta-
tendencies at least since the seventeenth century seems currently to be engaged
in a veritable metaturn, a genuine explosion of metareference in all kinds of
media and discourses and on all levels, highbrow and popular. Investigating
such cultural turns and trying to account for them is indeed one of the most
fundamental tasks of the humanities.


1. My thanks are once again due to Ingrid Pfandl-Buchegger and Sarah Mercer for
diligent proof-reading as well as to Ingrid Hable for expert assistance in formatting
the essay.
2. This is the term used in the call for papers of the conference which led to the present
volume (Noth, this vol., Part I).
318 Werner Wolf

3. For the concept of transmediality see Rajewsky (2002: 206) and (2003: 362363)
and Wolf (2002: 18).
4. As opposed to Nunning, who distinguishes between metaction and metanarra-
tion (see 2004), I conceive of metaction in a broad sense, as detailed in Wolf
(1993: chap. 3.2.).
5. Scholars who have approached the eld mostly concentrating on music on music
include: Adorno ([1949] 1975: 165189; who introduced this phrase in his critique
of Stravinsky); Karbusicky (1986; who acknowledges the possibility of music pos-
sessing analytical metalanguages [analysierende Metasprachen, 21]); Danuser
(1996; who treats a part of the eld, namely homage compositions, albeit without
theoretical ambitions; 2001; one of the best discussions of self-reexivity in mainly
vocal music); Mittmann (1999; his point of departure is linguistics; his attitude
concerning what he has in focus, namely what I will term critical intracomposi-
tional metareference, is predominantly sceptical, but his whole enterprise also its
failures shows to what extent musicological reection could benet from narrato-
logical investigations on metareference); Schneider (2004; a dictionary of examples
of music on music, which does, however, not include the type of metareference
to music as a system or medium which I think is particularly intriguing); Neubauer
(2005: 203205, who restricts Meta-Reection in Music [203] exclusively to the
verbal component of vocal music). All in all, musicological research in the eld (to
which Xenakis [(1967) 1971] does not belong in spite of the misleading title of his
essay Towards a metamusic, where metamusic simply stands for musicology)
appears to be largely untheoretical; music on music is mostly regarded as a mere
intramusical reference (e.g., also by Dibelius [(1966) 1998]), without proceeding to
the question of whether such intramusical self-reference in the sense of pointing
at (see below) could also become metareference.
6. Entering the term metamusic into the internet search machine Google leads to
several references to music therapy (e.g., as conducted by the jazz band The Di-
amond Jubilators, who advertise their services with the slogan Metamusic: Life
enhancement through music, as well as to a composi-
tion by Valentin Silvestrov entitled Metamusik, in which this title is explained as
being used in the sense of beyond music
(; both accessed
07.02.05). All of this has nothing to do with metamusic as a form of musical self-
reexivity. One should, however, mention that such self-reexivity is attributed by
Mittmann (cf. 1999: 236237) to another composition by Mauricio Kagel, enti-
tled Metapiece (incidentally, this title is identical to a piece by the Australian
composer Rainer Linz): in Kagels Metapiece Mittmann sees one of those com-
positions that constitute a metamusical reection on the performative act of music
playing (Kompositionen [. . . ], die den performativen Akt des Musizierens meta-
musikalisch reektieren).
7. Cf. also Wolf (2001a, 2007a).
Metaction and metamusic: Exploring the limits of metareference 319

8. I am here drawing on, but also modifying, Michael Scheffels research and a typol-
ogy which I have published elsewhere. Scheffel was among the rst to attempt some
systematic ordering in the vast eld of terms such as self-reference, autoreexiv-
ity, metaction etc. which had mostly been used as mere synonyms (see Scheffel
1997, notably 4649); in Wolf (2001a) I elaborated on Scheffels distinctions.
9. All of these cases are instances of iconicity in a broad sense, that is, iconicity that not
only occurs as a similarity between signier and signied (as expressed in the title
Form Miming Meaning of Nanny and Fischer 1999), but also as a similarity between
signiers (form miming form) and signieds (meaning miming meaning or
metaphorical iconicity). In my previous research (Wolf 2001a: 57) I focused on
the creation of self-referentiality through such iconicity as the only form of self-
referential pointing at and am here adding further possible variants.
10. This fuzziness can, for instance, be seen in the self-referential employment of the
device of foregrounding, which can be used both for the intensication of hetero-
referential meaning and for metareferential purposes. A certain fuzziness could also
appear in the fact that Jakobsons denition of the poetic function, namely to focus
on the message for its own sake (1960: 356), may be said to imply a statement, too
(for instance this text is literature). However, this would be a very weak statement,
which is so frequently (if not always) encountered that it does make sense to set it apart
from strong statements, in particular those that elicit metareections. It should also
be noted that Jakobsons two other self-referential functions of language, namely the
PHATIC function and in particular the METALINGUAL [. . . ] function (355
356), would more often trigger such strong metastatements.
11. Thus, a narratorial evaluation of the actions of a character which have just been told
would be an example of such hetero-referential self-reection (one should note
that the hetero-referentiality here only refers to the object or the signied of the self-
reference; in our example this is a character who is regarded as a being imaginatively
located beyond the text).
12. A famous example of such implicitly metactional typographical devices is the black
page inserted into chapter I/12, which comments on the death of parson Yorick and,
by a highly unusual deviation from the traditional dominance of symbolic signs in
the medium of print ction, may trigger metareections on the potential and limits
of the medium as such.
13. In other cases (e.g., mise en abyme) implicit metareference may just be regarded as
a form of self-referential similarity.
14. As a consequence, the implicit metareference will be noted with more or less intensity.
15. Examples of this kind would be the German hymn Singt dem Herrn ein neues Lied
or the last song in Schuberts Die Winterreise (cf. Wolf 2001b).
16. For this opera and musical self-reexivity in Strauss in general see the excellent
article Danuser (2001).
17. As music is also a nonrepresentational medium, it is unable to represent its medium
or the creation of a composition in the way a painting may do which self-reexively
shows, for instance, a painter at work.
320 Werner Wolf

18. Cf. Breig (1999: 631) and Keller ([1950] n. d.: 5657), who convincingly argues in
favor of Bachs authorship (which also seems to be conrmed by the soprano melody
in bar 30, which, in German notation, contains the signature B-A-C-H).
19. Hofstadter has his character Tortoise erroneously explain the g major chord on
which the second movement, the Centrum, ends as a disorientation (cf. [1979]
1980: 122123). G major, the dominant of c major, is, however, anything but unusual
for the concluding (transitional) chord of the middle section of a composition written
in c major, and therefore we are here, as justly remarked by Keller, actually near the
exit of the labyrinth (cf. [1950] n. d.: 57). Apart from that, it is not clear what
in Kleines Harmonisches Labyrinth justies Hofstadters idea of nested [. . . ]
structures (ix), for this would imply the existence of markedly different levels that
are used for recursive devices; such levels, which instrumental music can hardly
produce, are, however, not discernible in this composition.
20. See also Mittmann (1999: 236), who regards defamiliarization (Verfremdung)
as one possible device of metamusic, yet remains highly wary of any undifferentiated
afrmation of the existence of musical metareference.
21. This composition shows a clear intermusical self-reference, since it quotes a passage
from a divertimento for wind instruments by Joseph Haydn (more specically, the
reference is to the Chorale St. Antoni from Joseph Haydns rst of the Sei di-
vertimenti a due oboi, due corni, due fagotti obl. fagotto e serpent). The step from
intramusical self-reference to musical metareference is arguably made by the state-
ment implied in this composition, listen, how this old theme by Haydn can still be
brilliantly employed at the end of the 19th century!, or alternatively this statement
may be paraphrased as a homage to Haydn, not the man, but the composer. In any
case, Brahms may be said to presuppose a medium awareness on the part of his lis-
teners, even a historically differentiated one concerning different ways of actualizing
the system music in different epochs, and this indirectly implies a reference to the
medium music as employed by Haydn and Brahms, thus fullling the principal con-
dition of metareference. In this case, extracompositional metareference is combined
with intracompositional metareference. Once again it must, however, be noted, that
for most nonspecialists this metareference is, if at all, only discernible owing to the
verbal title of the composition.
22. Thus, there is only one single instance in which the author, in his introduction, uses
a collocation with meta- (cf. Schneider 2004: 7).
23. For further explorations of the relationship between nostalgia and metareferences
see Andreas Bohns contribution to this volume.
24. The aforementioned virtuoso music provides further examples, as does Bachs
Kleines Harmonisches Labyrinth, in which, as in his Die Kunst der Fuge (The
Art of the Fugue), the composer explores the potentials of the musical medium or
compositional forms and at the same time sports his own expertise without any
critical or self-critical intentions.
Metaction and metamusic: Exploring the limits of metareference 321

25. For Mittmann (1999) a critical function restricted to intracompositional metaref-

erence seems to be the only form of metamusic, at least it is the only one which he
discusses (thereby unduly restricting the realm of potential musical metareference).
26. Adorno speaks of a Beschadigungsaktion ([1949] 1975: 168). Incidentally, in
Adornos view, which is as one-sided as Mittmanns (see the preceding note), parody
is the basic form of music on music (Parodie ist die Grundform der Musik uber
Musik, 170).
27. For more details on this kind of protective irony see Wolf (2007b).
28. First explorations of this subeld have been carried on by Pier and Schaeffer (2005),
and Wolf (2005).


Adorno, Theodor
[1949] 1975 Philosophie der neuen Musik. (Gesammelte Schriften 12.) Frank-
furt: Suhrkamp.
Booth, Wayne C.
1952 The self-conscious narrator in comic ction before Tristram Shandy.
PMLA 67: 163185.
Breig, Werner
1999 Freie Orgelwerke. In: Konrad Kuster (ed.), Bach-Handbuch, 613712.
Kassel: Barenreiter.
Broich, Ulrich and Manfred Pster (eds.)
1985 Intertextualitat:Formen, Funktionen, anglistische Fallstudien. Tubin-
gen: Niemeyer.
Danuser, Hermann
1996 Homage-Kompositionen als Musik uber Musik. In: Gunther Wagner
(ed.), Jahrbuch des Staatlichen Instituts fur Musikforschung Preui-
scher Kulturbesitz, 5264. Stuttgart: Metzler.
2001 Musikalische Selbstreexion bei Richard Strauss. In: Bernd Edel-
mann, Birgit Lodes and Reinhold Schlotterer (eds.), Richard Strauss
und die Moderne: Bericht uber das Internationale Symposium Mun-
chen, 21. bis 21. Juni 1999, 5177. Berlin: Henschel.
Dibelius, Ulrich
[1966] 1998 Musik uber Musik. In: Ulrich Dibelius, Moderne Musik nach
1945, 607628. Munich: Piper.
Fowles, John
[1969] 1977 The French Lieutenants Woman. London: Triad Panther.
Hauthal, Janine, Julijana Nadj, Ansgar Nunning and Henning Peters (eds.)
2007 Metaisierung in der Literatur und anderen Medien. Berlin and New
York: Walter de Gruyter.
322 Werner Wolf

Hofstadter, Douglas R.
[1979] 1980 Godel, Escher, Bach: An Eternal Golden Braid: A Metaphor-
ical Fugue on Minds and Machines in the Spirit of Lewis Carroll.
Harmondsworth: Penguin.
Hutcheon, Linda
[1980] 1984 Narcissistic Narrative: The Metactional Paradox. London: Me-
Jakobson, Roman
1960 Closing statement: Linguistics and poetics. In: Thomas A. Sebeok
(ed.), Style in Language, 350377. Cambridge, MA: MIT Press.
Karbusicky, Vladimir
1986 Grundri der musikalischen Semantik. (Grundrisse 7.) Darmstadt:
Wissenschaftliche Buchgesellschaft.
Keller, Hermann
[1950] n. d. Die Orgelwerke Bachs. Leipzig: Peters.
Mittmann, Jorg-Peter
1999 Meta-Musik: Zum Problem musikalischer Selbstreferenz. In:
Christoph Asmuth, Gunter Scholtz and Franz-Bernhard Stammkotter
(eds.), Philosophischer Gedanke und musikalischer Klang: ZumWech-
selverhaltnis von Musik und Philosophie, 229238. Frankfurt: Cam-
Nanny, Max and Olga Fischer (eds.)
1999 Form Miming Meaning: Iconicity in Language and Literature. Ams-
terdam: John Benjamins.
Neubauer, John
2005 The return of the repressed: Language and music in the nineteenth cen-
tury. In: Lilo Moessner and Christa M. Schmidt (eds.), Anglistentag
2004, Aachen: Proceedings of the Conference of the German Associ-
ation of University Teachers in English 26, 199209. Trier: WVT.
Nunning, Ansgar
2004 On metanarrative: Towards a denition, a typology and an outline of
the functions of metanarrative commentary. In: John Pier (ed.), The
Dynamics of Narrative Form: Studies in Anglo-American Narratology,
1157. (Narratologia 4). Berlin and New York: Walter de Gruyter.
Pier, John and Jean-Marie Schaeffer (eds.)
2005 Metalepses: Entorses au pacte de la representation. Paris: Editions
de lEcole des Hautes Etudes en Sciences Sociales.
Rajewsky, Irina
2002 Intermedialitat. Tubingen: Francke.
2003 Intermediales Erzahlen in der italienischen Literatur der Postmod-
erne: Von den giovani scrittori der 80er zum pulp der 90er Jahre.
Tubingen: Narr.
Metaction and metamusic: Exploring the limits of metareference 323

Scheffel, Michael
1997 Formen selbstreexiven Erzahlens: Eine Typologie und sechs exem-
plarische Analysen. Tubingen: Niemeyer.
Schneider, Klaus
2004 Lexikon Musik uber Musik: Variationen Transkriptionen Hom-
magen Stilimitationen BACH. Kassel: Barenreiter.
Stam, Robert
2000 The politics of reexivity. In: Robert Stam, Film Theory: An Introduc-
tion, 151153. Oxford: Blackwell.
Sterne, Laurence
[1767] 1967 The Life and Opinions of Tristram Shandy, Gentleman, ed. by
Graham Petrie. Harmondsworth: Penguin.
Stoichita, Victor I.
[1993] 1997 LInstauration du tableau. Librairie des Meridiens. Paris: Klinck-
sieck. Engl. The Self-Aware Image: An Insight into Early Modern
Meta-Painting, Anne-Marie Glasheen (transl.). (Cambridge Studies
in New Art History & Criticism.) Cambridge: Cambridge University
Wolf, Werner
1993 Asthetische Illusion und Illusionsdurchbrechung in der Erzahlkunst:
Theorie und Geschichte mit Schwerpunkt auf englischem illusion-
sstorenden Erzahlen. (Buchreihe derAnglia 32.)Tubingen: Niemeyer.
2001a Formen literarischer Selbstreferenz in der Erzahlkunst: Versuch einer
Typologie und ein Exkurs zur mise en cadre und mise en re-
et/serie. In: Jorg Helbig (ed.), Erzahlen und Erzahltheorie im zwan-
zigsten Jahrhundert. Festschrift fur Wilhelm Fuger, 4984. Heidel-
berg: Winter.
2001b Willst zu meinen Liedern deine Leier drehn? Intermedial metatex-
tuality in Schuberts Der Leiermann as a motivation for song and
accompaniment and a contribution to the unity of Die Winterreise.
In: Walter Bernhart and Werner Wolf (eds.), Word and Music Studies:
Essays on the Song Cycle and on Dening the Field Proceedings of
the Second International Conference on Word and Music Studies at
Ann Arbor, 1999, 121140. (Word & Music Studies 3.) Amsterdam:
2002 Intermediality Revisited: Reections on Word & Music Relations in
the Context of a General Typology of Intermediality. In: Suzanne M.
Lodato, Suzanne Aspden and Walter Bernhart (eds.), Word and Music
Studies: Essays in Honor of Steven Paul Scher and on Cultural Identity
and the Musical Stage, 1334. (Word & Music Studies 4.)Amsterdam:
324 Werner Wolf

2005 Metalepsis as a transgeneric and transmedial phenomenon: A case

study of the possibilities of exporting narratological concepts. In:
Jan Christoph Meister (ed.), Narratology Beyond Literary Criticism:
Mediality, Disciplinarity, 83107. (Narratologia 6.) Berlin and New
York: Walter de Gruyter.
2007a Metaisierung als transgenerisches und transmediales Phanomen: Ein
Systematisierungsversuch metareferentieller Formen und Begriffe in
Literatur und anderen Medien. In: Janine Hauthal, Julijana Nadj, Ans-
gar Nunning and Henning Peters (eds.), Metaisierung in der Literatur
und anderen Medien, 2564. Berlin and New York: Walter de Gruyter.
2007b Schutzironie als Akzeptanzstrategie fur problematische Diskurse:
Zu einer vernachlassigten Nahe erzeugenden Funktion von Ironie. In:
Thomas Honegger, Eva-Maria Orth and Andra Schwabe. (eds.), Irony
Revisited: Spurensuche in der englischsprachigen Literatur. Festschrift
fur Wolfgang G. Muller, 2750. Wurzburg: Konigshausen & Neu-
Xenakis, Iannis.
[1967] 1971. Towards a metamusic. In: Jasia Reichardt. (ed.), Cybernetics,
Art and Ideas, 111123. London: Studio Vista.
Index of names

Aarseth, Espen J., 220, 256 Baudrillard, Jean, 7, 188

Abrahams, Roger, 89 Bayard, Hippolyte, 98
Acton, David, 95 Becaud, Gilbert, 148
Adorno,Theodor, 313314, 318, 321 Beckley, Bill, 95
Aebi, Jean Etienne, 57 Bernstein, Richard J., 39
Aguilar, Katherine, 95 Bettetini, Gianfranco, 13
Albarran, Alan B., 197 Bishara, Nina, 16, 22, 79
Alessandria, Jorge, 62 Bitney, Dan, 296297
Allen, Jeanne, 165, 179 Black, Max, 9
Allen, Woody, 136, 148, 151 Bleicher, Joan Kristin, 7, 183, 186,
Anati, Emmanuel, 61 190, 196
Andacht, Fernando, 165, 169, 172, Blitt, Barry, 34
178179 Blobaum, Bernd, 7
Andersen, Peter Bgh, 4 Block, Friedrich W., 5
Anderson, Laurie, 293296, 301 Bohn, Andreas, 143, 145, 146, 320
Andreychek, Eric, 276 Bolter, Jay David, 6, 220, 251
Arnold, Eve, 112114, 118120 Booth, Wayne C., 317
Arrouye, Jean, 96 Borges, Jorge Luis, 284
Assmann, Aleida, 90, 183 Boym, Svetlana, 152
Austin, John L., 69, 239 Brahms, Johannes, 313, 320
Azevedo, Theo, 209210, 217 Breig, Werner, 320
Aznavour, Charles, 148 Broich, Ulrich, 5, 313
Brook, Andrew, 5
Buckland, Warren, 7, 20
Babcock, Barbara A., 8 Buhler, Karl, 67
Bach, Johann Sebastian, 13, 310314, Bush, George W., 13, 3335, 40
320 Buttner, Stefan, 5
Bakhtin, Mikhail M., 36
Baraldi, Claudio, 107108
Barris, George, 118 Caillois, Roger, 223, 237, 251
Barthes, Roland, 7, 96 Caldeira, Joao Bernardo, 171, 176,
Bartlett, Steven J., 4, 5, 79, 14, 31 179
Bateson, Gregory, 4, 237241, 247, Carani, Marie, 95
250251, 255 Carroll, Lewis, 284
326 Index of names

Carroll, Noel, 168 Feiter, Wolfgang, 59

Casetti, Francesco, 191 Figueiroa, Alexandre, 174, 177
Cayley, John, 276 Filk, Christian, 120
Chaplin, Charlie, 131 Fischer, Olga, 319
Churchill, Winston, 61 Fischer, Volker, 143
Colapietro, Vincent, 5, 31, 33, 35 Fitch, Frederic B., 5
36, 39, 179 Fliedl, Gotted, 146
Comolli, Jean-Louis, 165, 169, 172 Foucault, Michel, 187188
173, 175177, 179 Fowles, John, 309
Conant, Chloe, 64 Frans, Marya, 251
Conley, Verena, 301 Frege, Gottlob, 9, 11
Corsi, Giancarlo, 107108 Friedman, Thomas, 37, 40
Coutinho, Eduardo, 165179 Frieske, Michael, 7
Cramer, Florian, 277, 283 Frobel, Friedrich, 250

Danuser, Hermann, 318319 Gall, France, 148

Davis, Char, 293, 297302 Geach, Peter T., 9
Deikman, Arthur, 302 Genette, Gerard, 255, 257, 263
Delacroix, Eugene, 96 Gerbner, George, 183
Delaroche, Paul, 95 Giraudon, August, 99
Derrida, Jacques, 5, 6162 Gitlin, Todd, 31, 37, 40
DeVidi, Richard C., 5 Godard, Jean-Luc, 316
Dibelius, Ulrich, 318 Godel, Kurt, 310
Dika, Vera, 150 Goebel, Gerhard, 7
Dillon, George, 286 Govignon, Brigitte, 99100
Dirks, Tim, 19 Grampp, Sven, 119
Dodge, Martin, 283 Greene, Rachel, 272
Dundes, Alan, 89 Grundberg, Andy, 97
Dunne, Michael, 7, 31 Grusin, Richard, 6, 220, 251
Gulda, Friedrich, 314315
Eco, Umberto, 7, 61 Gullekson, Garth, 226
Ekenberg, Jan, 284
Epimenides, 75, 251 Hacking, Ian, 165
Ermi, Laura, 251 Haraway, Donna, 277, 283
Erwitt, Elliot, 118 Hardwick, Charles S., 168
Escher, Mauritz Cornelis, 310 Harweg, Roland, 66
Eskelinen, Markku, 261 Hauthal, Janine, 316
Esposito, Elena, 107108 Haydn, Joseph, 314, 320
Esser, Andrea, 5 Hayles, N. Katherine, 277
Evans, Gareth, 9 Heidegger, Martin, 5
Index of names 327

Helbig, Jorg, 6 Kienzle, Bertrand, 5

Hempfer, Klaus W., 5 Kiesel, Doron, 7, 21
Hertling, Anke, 6 Kirchmann, Kay, 7, 107, 109, 120,
Heyne, Renate, 101 255
Hockett, Charles, 61, 65 Kirkpatrick, Graeme, 229230
Hofer, Johannes, 143 Kitchin, Rob, 283
Hoffmeyer, Jesper, 4 Klein, Naomi, 3
Hofstadter, Douglas R., 5, 13, 310, Klein, Robert A., 196
320 Klima, John, 277, 279
Holland, John H., 222 Knopp, Guido, 186
Huber, Werner, 5, 8 Knowles, Elizabeth, 15, 17
Huizinga, Johan, 223, 237, 251, 255 Kohring, Matthias, 7
Husserl, Edmund, 144 Korzybski, Alfred, 12, 239
Hutcheon, Linda, 307, 313 Kosselleck, Reinhart, 183
Kruger, Mike, 190
Jackson, Shelly, 271272 Kucklich, Julian, 232
Jacobs, Ron, 196
Jager, Gottfried, 103 Lambernd, Jochen, 189
Jahraus, Oliver, 4 Lawson, Hilary, 3, 5, 78, 31, 36
Jakobson, Roman, 5, 9, 12, 140, 208, Lenoir, Timothy, 31
243, 270, 303, 305, 319 Liebrand, Claudia, 6
James, William, 167168, 174, 178 Lindvall, Terrance, 158
Jay, Paul, 5 Linz, Rainer, 318
Jenkins, Henry, 207208, 217 Lipman, Jean, 6
Jevbratt, Lisa, 284286 Liszt, Franz, 312
Johansen, Jrgen Dines, 5, 8 Liu, Alan, 281
Jung, Holger, 57 Ljungberg, Christina, 67, 97, 291
Jurgens, Hartmut, 4 Luhmann, Niklas, 5, 8, 10, 107108,
Juul, Jasper, 210212, 215, 221, 112, 188, 229, 255
226, 228, 233 Lynch, Michael, 165, 170173, 175,
Kagel, Mauricio, 318 Lyotard, Jean-Francois, 3, 13
Kant, Immanuel, 250
Karbusicky, Vladimir, 318 Machado, Arlindo, 63, 292
Karpf, Ernst, 7, 21 Mallarme, Stephane, 271
Karstens, Eric, 198 Manovich, Lev, 18, 20
Katz, Jerrold J., 9 Marcus, Solomon, 7
Keller, Hermann, 320 Marshall, Richard, 6
Kemper, Wulf-Peter, 57 Marx, Karl, 141
Kempson, Ruth M., 9 Matt, Remy von, 57
328 Index of names

Mattos, Carlos A., 166 Ort, Claus-Michael, 4, 67

McLuhan, Marshall, 67 Owens, Craig, 64
Melton, J. Matthew, 158
Melville, Herman, 294 Paech, Joachim, 67
Memmott, Talan, 276 Paganini, Nicolo, 312
Menninghaus, Winfried, 5, 8 Paley, W. Bradford, 279
Metscher, Thomas, 5, 8 Pape, Helmut, 5
Metz, Christian, 20, 69 Pattee, Howard H., 4
Mez (Mary Anne Breeze), 276 Patton, George, 34
Middecke, Martin, 5, 8 Paul, Christiane, 276, 283
Miller, Arthur, 110, 113 Pavlicic, Pavao, 8
Milosz, Czeslaw, 38 Pearce, Celia, 208211, 215217
Mitchell, W. J. Thomas, 6264, 96 Peirce, Charles S., 812, 3236, 39
98 40, 49, 8084, 88, 91, 96, 103
Mittmann, Jorg-Peter, 318, 320321 104, 166170, 178, 241, 291
Moeller, Hans-Georg, 74 Peitgen, Heinz-Otto, 4
Monroe, Marilyn, 110114, 118 Petersen, Christer, 3
Morgenstern, Oskar, 221 Pster, Manfred, 5, 313
Mozart, Wolfgang Amadeus, 314 Paum, Hans Gunter, 147
Muller, Jurgen, 6 Piaf, Edith, 148
Munch, Dieter, 9 Pier, John, 321
Murray, Janet, 269 Pold, Sren, 233234
Myers, C. Mason, 5 Pool, Steven, 263
Potthast, Ulrich, 5
Nanny, Max, 319 Powers, Richard, 273275, 281
Nash, John F., 221, 224 Prieto, Luis J., 69
Neitzel, Britta, 237, 241, 255, 263 Prigogine, Ilya, 4
Neubauer, John, 318 Prokoev, Sergei Sergeyevich, 314
Neumann, John von, 221 Puhringer, Karin, 195, 197198
Newcomb, Horace, 185
Newman, James, 263 Quast, Thomas, 196
Nietzsche, Friedrich, 5
Noth, Winfried, 315, 18, 20, 3132, Raab, Stefan, 190
36, 40, 5051, 61, 64, 67, 76, 91, Rainer, Arnulf, 100, 102
95, 97, 103104, 165, 171, 209, Rajewsky, Irina, 6, 318
219, 240, 291, 295, 301, 303 Ransdell, Joseph, 168
Nunning, Ansgar, 318 Rapp, Bernhard, 215, 245, 253
Reck, Hands-Ulrich, 111
Ochsner, Beate, 148 Resnais, Alain, 148
Odin, Roger, 191 Rich, Frank, 3335, 40
Index of names 329

Ritchin, Fred, 95 Shir, Jay, 5

Robin, Amrie Monique, 108 Short, Thomas L., 33
Robinett, Waren, 263 Siebert, Jan, 155, 161
Robins, Kevin, 96, 97 Siedenbiedel, Catrin, 5
Rorty, Richard, 32 Siegert, Gabriele, 195, 197198
Rossi-Landi, Ferrucio, 125129, 140 Silva, Cicero Ignacio da, 207
Rouch, Jean, 165 Silvestrov, Valentin, 318
Rouse III, Richard, 263 Smith, John E., 39
Rousseau, Jean-Jaques, 143 Smith, Jonas Heide, 221
Russell, Bertrand, 5, 8, 238 Smuda, Manfred, 5
Ryan, Marie-Laure, 75, 247, 251, Souriau, Etienne, 175
255, 269, 287 Spencer-Brown, George, 229
Spielberg, Steven, 3334, 40
Salen, Katie, 223, 237, 242, 251 Spielmann, Yvonne, 6
Salvemini, Lorella Pagnucco, 59 Spie, Brigitte, 7
Sander, August, 179 Stam, Robert, 5, 21, 128, 315
Santaella, Lucia, 6, 20, 31, 61, 64, Stengers, Isabelle, 4
95, 167, 207, 210, 213, 297, Stephenson, Ralph, 161
300301 Sterne, Lawrence, 308
Satu, Helio, 251 Stoichita, Victor I., 315
Saupe, Dietmar, 4 Strahlendorf, Peter, 91
Saussure, Ferdinand de, 9 Strand, Paul, 99
Schaeffer Jean-Marie, 321 Strauss, Richard, 309
Schalk, Willi, 91 Stravinsky, Igor Fyodorovich, 314,
Scheffel, Michael, 5, 306, 319 318
Scheuerl, Hans, 237, 250 Suber, Peter, 5, 8
Scheutz, Matthias, 5, 14 Sutton-Smith, Brian, 250
Schiller, Friedrich, 237, 250
Schmid-Ruhe, Bernd, 119
Tabard, Maurice, 100
Schmidt, Siegfried J., 7, 1011, 47,
Taine, Hypolyte, 96
118, 195
Thoma, Helmut, 91
Schneider, Irmela, 6
Todorow, Almut, 119
Schneider, Klaus, 313314, 318
Toprowicz, Maciej, 59
Schoppe, Arno, 5, 7
Tosca, Susana Pajares, 231
Schubert, Franz, 319
Toubiana, Serge, 110, 113
Schudson, Michael, 59
Tousignant, Claude, 88
Schutte, Jorg, 198
Trepp, Selina, 293, 296297, 301
Selic, Bran, 226
Selichar, Gunther, 103
Shakespeare, William, 208 Ullrich, Timm, 102
330 Index of names

Visarius, Karsten, 7, 21 Wilson, Woodrow, 34

Winkler, Hartmut, 4, 22
Wagner, Richard, 309 Winters, Patricia, 140
Walther, Bo Kampmann, 219220, Wirsig, Christian, 259
223, 229230 Withalm, Gloria, 7, 21, 125, 261
Ward, Adrian, 281 262
Ward, Paul T., 226 Wittig, Susan, 6
Wardrip-Fruin, Noah, 207 Wolf, Werner, 5, 89, 287, 303, 307,
Waugh, Patricia, 5 318319, 321
Weber, Stefan, 7
Weibel, Peter, 95, 100, 102 Xenakis, Iannis, 318
Wells, Herbert George, 34
Wetzel, Michael, 99 Zapf, Hubert, 5, 8
White, Hayden, 183 Zavala, Lauro, 5
Whitehead, Alfred N., 8 Zimmerman, Eric, 223, 237, 242, 251
Whiteside, Anna, 5 Zizek, Slavoj, 19
Zurstiege, Guido, 52, 59
Index of subjects

100 nervigsten TV Shows (Die), 187, August Sander effect, 172

190191 authentic(ity), 296297
9/11, see September 11 autobiographical, 296
autodeictic, 66
Auto-illustrator, 281283
abstract, abstraction, 9899, 103
autonomy, 5, 6, 21
Adaptation, 132
autopoiesis, 4, 108
adbusting, 4
autoreferentiality, 8
addressee, 301
autoreexivity, 6, 319
advertising, 3, 7, 13, 1518, 21, 23,
autosymbolism, 5
4759, 7991, 209, 257258, 261
autotelic, 3, 5
advertising system, 4850, 5559
aesthetic(s), 5, 8, 36, 39, 61
aesthetic illusion, 317 Bad and the Beautiful (The), 132
AIDA, 79 Barbarian, 259, 261262
algorithm(s), 223 226, 228 Barton Fink, 32
aliquid pro aliquo, 12 Best of... (The), 187
alloreference, alloreferential, 910, Big Brother, 169, 177
13, 20, 62, 72, 74, 9899, 104, biocybernetic, 300
303 biology, 4, 8
Alm (Die), 187 biopic, 132
alternative awareness, 301 Blow out, 133
American McGees Alice, 210 Boogie Nights, 133
Animal Crackers, 148 bootstrap (phenomenon), 8
animated lm, 155161 brand identication, 196
anthropology, 4, 8 Bugs Life (A), 160
architecture, 3, 6 Burg (Die), 185, 187
archive material, 184192, 196
argument, 14, 17 camera, 99, 102103, 133
argumentative, see self-reference Captive Mind, 38
art(s), 3, 6, 23, 99, 104 Carabiniers (Les), 136
art-pour-lart, 6 carnival, 23
attention, 4849, 55, 58, 79, 85, 90 Ceci nest pas une pipe, 75
audiovisual media, 145, 209 censorship, 134135
332 Index of subjects

Cent photos du siecle (Les), see One convention, 61, 65

Hundred. . . copy, 68, 74
cinema, 1819 counterfactual history, 186
circuits, 4 Counter-Strike, 216
circularity, circularities, 5, 14, 1619 Countryman and the Cinematograph
code, 126127 (The), 136
cinematographic, 138 creative destruction, 281
digital, 279 crisis of representation, 32, 9697,
codework, 276277 104
cognition, 49 critical function, 36, 39
cognitive science, 5 cross-promotion, 197
collateral experience, 81, 84 cultural knowledge, 49
collective knowledge, 4750, 55, 58 culture, 34, 7, 10, 33, 177
Colleurs dafches (Les), 134 culture jamming, 3
comic book, 155156 cyberspace, 209
Comicalamities, 134,138 cybersquatting, 4
comics, 316317 cyborg, 277
command(s), 214 cycles, 4
commercial, 149150
communication, 911, 15, 20, 47 Dada, 97, 100
49, 197 Daffy Duck and Egghead, 158
communication studies, 195 daguerreotype, 96
communicative, see self-reference Day of the Tentacle, 260261
community, 3233, 3930 Dead Men Dont Wear Plaid, 137
computer animation, 296 Death of Photography (The), 95
computer game(s), 1314, 2123, 208, Deconstructing Harry, 148
210, 212213, 219220, 224, 226, deconstruction, 100, 102
231232, 244245, 248, 250, 255, defamiliarization, 320
261263, 277278, 287, see also deja-vu, 1920, 295
self-reference depict, 6365, 68
computer game studies, 237 designatum, 9
computer science, 4, 219, 226 diagram, 292
conative, 12 dicent, 1415, 17
Concours dautomobiles euris, 131 dicentic, see self-reference
Concrete Photography, 103104 Die Another Day, 22
consciousness, 5, 78, 36, 143144, diegetic frame, 159
306 digital, 9597, 269, 287
constructivism, 10, 112 art, 292293, 298
consumer invitation, 198, 200, 203 culture, 208
consumption, 127129, 135, 140, 142 game(s), 223, 227, 232, 244, 263
Index of subjects 333

manipulation, 296 fact, 186

media, 3, 7, 20, 22, 207 faked, 167, 169
photography, 96, 103104 fakes, 7374
digitextuality, 138 fallibilism, 36
disaster lm, 19 fashion, 7
discourse theory, 187 feedback, 4
displacement, 6162 Felix the Cat, 156157
documentary, 165169, 174178, 185 ction, 39, 62, 186, 188, 248, 304,
186 307, 309
docutainment, 186 ctionality, 241
Don Quixote, 269, 287 lm, 7, 13, 1823, 119, 125, 129,
Doom 3, 219 147, 183, 209, 315317, see also
double bind, 4 self-reference
double exposure, 98, 100101 lm camera, 111114, 119
downward causation, 4 lmic citation, 137
dualism, 165169, 178 lm-in-the-lm, 150, see also metalm
dualistic metaphysics, 174, 178 Finding Nemo, 160
Dumb Hounded, 138, 158159 rstness, 80
dyadic, 12 xing belief, 170, 178
Flatworld, 156157
E la nave va, 133 foreground(ing), 305, 308, 31013
economic strategies, 188 Forrest Gump, 138
economic system, 48 fractals, 4
economics, 4 frame, 62, 6870, 7475
Edifcio Master, 166, 177 Fremdreferenz, 9
editorial reference, 197 French Lieutenants Woman (The),
encounter, 174176 132, 309
entertainment, 152, 185186,188189, function (of language), see conative,
207 expressive, metalinguistic, phatic,
Entree du cinematographe (L), 136 poetic, referential
enunciation, enunciative, 67, 6970,
7375, see also self-reference game(s), 208, 210212, 219224, 228,
Epimenides paradox, 75 237, 241, 243, 250
episode, 198 game episte