Visit https://ebooknice.

com to download the full version and

explore more ebooks

(Ebook) Characters in Fictional Worlds:

Understanding Imaginary Beings in Literature, Film,
and Other Media (Revisionen Grundbegriffe der
Literaturtheorie 3) by Jens Eder, Fotis Jannidis,
_____ Click the link
Ralf Schneider below
(Hg.) to download
ISBN _____
9783110232417,
9783110232424, 3110232413, 3110232421
https://ebooknice.com/product/characters-in-fictional-
worlds-understanding-imaginary-beings-in-literature-
film-and-other-media-revisionen-grundbegriffe-der-
literaturtheorie-3-2472778

Explore and download more ebooks at ebooknice.com

Here are some recommended products that might interest you.
You can download now and explore!

(Ebook) Biota Grow 2C gather 2C cook by Loucas, Jason; Viles, James

ISBN 9781459699816, 9781743365571, 9781925268492, 1459699815,
1743365578, 1925268497

https://ebooknice.com/product/biota-grow-2c-gather-2c-cook-6661374

ebooknice.com

(Ebook) Matematik 5000+ Kurs 2c Lärobok by Lena Alfredsson, Hans

Heikne, Sanna Bodemyr ISBN 9789127456600, 9127456609

https://ebooknice.com/product/matematik-5000-kurs-2c-larobok-23848312

ebooknice.com

(Ebook) SAT II Success MATH 1C and 2C 2002 (Peterson's SAT II Success)

by Peterson's ISBN 9780768906677, 0768906679

https://ebooknice.com/product/sat-ii-success-
math-1c-and-2c-2002-peterson-s-sat-ii-success-1722018

ebooknice.com

(Ebook) Master SAT II Math 1c and 2c 4th ed (Arco Master the SAT
Subject Test: Math Levels 1 & 2) by Arco ISBN 9780768923049,
0768923042

https://ebooknice.com/product/master-sat-ii-math-1c-and-2c-4th-ed-
arco-master-the-sat-subject-test-math-levels-1-2-2326094

ebooknice.com
(Ebook) Cambridge IGCSE and O Level History Workbook 2C - Depth Study:
the United States, 1919-41 2nd Edition by Benjamin Harrison ISBN
9781398375147, 9781398375048, 1398375144, 1398375047

https://ebooknice.com/product/cambridge-igcse-and-o-level-history-
workbook-2c-depth-study-the-united-states-1919-41-2nd-edition-53538044

ebooknice.com

(Ebook) Fortschritte der Physik / Progress of Physics: Band 29, Heft 3

ISBN 9783112655887, 3112655885

https://ebooknice.com/product/fortschritte-der-physik-progress-of-
physics-band-29-heft-3-50956780

ebooknice.com

(Ebook) Vagabond, Vol. 29 (29) by Inoue, Takehiko ISBN 9781421531489,

1421531488

https://ebooknice.com/product/vagabond-vol-29-29-37511002

ebooknice.com

(Ebook) Fortschritte der Physik / Progress of Physics: Band 29, Heft 9

ISBN 9783112656006, 3112656008

https://ebooknice.com/product/fortschritte-der-physik-progress-of-
physics-band-29-heft-9-50956742

ebooknice.com

(Ebook) Fortschritte der Physik / Progress of Physics: Band 29, Heft 7

ISBN 9783112655962, 3112655966

https://ebooknice.com/product/fortschritte-der-physik-progress-of-
physics-band-29-heft-7-50956754

ebooknice.com
Characters in Fictional Worlds
 Revisionen
Grundbegriffe der Literaturtheorie

Herausgegeben
von
Fotis Jannidis
Gerhard Lauer
Matı́as Martı́nez
Simone Winko

De Gruyter
Characters in Fictional Worlds
Understanding Imaginary Beings in Literature, Film,
and Other Media

Edited
by
Jens Eder
Fotis Jannidis
Ralf Schneider

De Gruyter
 ISBN 978-3-11-023241-7
e-ISBN 978-3-11-023242-4

Library of Congress Cataloging-in-Publication Data

Characters in fictional worlds : understanding imaginary beings in literature,

film, and other media / edited by Jens Eder, Fotis Jannidis, Ralf Schneider.
p. cm. - (Revisionen. Grundbegriffe der Literaturtheorie ; 3)
Includes bibliographical references.
ISBN 978-3-11-023241-7 (acid-free paper)
1. Characters and characteristics in mass media. 2. Fictitious characters.
3. Influence (Literary, artistic, etc.) I. Eder, Jens. II. Jannidis, Fotis.
III. Schneider, Ralf, 1966-
P96.C43C47 2010
8091.927-dc22
2010037621

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 http://dnb.d-nb.de.

” 2010 Walter de Gruyter GmbH & Co. KG, Berlin/New York

Printing: Hubert & Co. GmbH & Co. KG, Göttingen
⬁ Printed on acid-free paper
Printed in Germany
www.degruyter.com
 Preface

Most of the contributions to this volume are based on papers presented

and discussed at the conference »Characters in Fictional Worlds: Inter-
disciplinary Perspectives«, which was held at the Centre for Interdiscipli-
nary Research (Zentrum für interdisziplinäre Forschung, ZIF) of the
University of Bielefeld, Germany, 28 February – 2 March 2007. We would
like to thank the Centre for the funding and the organization of the
conference. A number of contributions were added to the topics of the
conference to complement the present collection. The volume has been a
long time in the making – we are very grateful to all contributors for join-
ing this project and for their patience.

Our heartfelt thanks go to Marcus Willand for the editorial work on this
project. Without his indefatigable support, care, and patience this volume
would not have been printed. Sarah Böhmer, Mareike Brandt, Daniel
Bund, Anne Diekjobst, Sebastian Eberle, Christian Maintz and Maike
Reinerth joined forces with him, and we would like to express our grati-
tude to them, too.

Some of the chapters were translated into English, for which we thank the
translators, Wolfram Karl Köck, Alison Rosemary Köck and Michael
Pätzold. Thanks are also due to Wallace Bond Love for last-minute
language support.

The editors
 Content

Content

Introduction
JENS EDER / FOTIS JANNIDIS / RALF SCHNEIDER
Characters in Fictional Worlds. An Introduction .................................... 3

I General Topics
HENRIETTE HEIDBRINK
Fictional Characters in Literary and Media Studies. A Survey
of the Research ............................................................................................67
MARIA E. REICHER
The Ontology of Fictional Characters .................................................. 111
PATRICK COLM HOGAN
Characters and Their Plots ..................................................................... 134

II Characters and Characterisation in

Different Media
ALAN PALMER
Social Minds in Persuasion ........................................................................ 157
JONATHAN CULPEPER / DAN MCINTYRE
Activity Types and Characterisation in Dramatic Discourse ............ 176
SIMONE WINKO
On the Constitution of Characters in Poetry ...................................... 208
MURRAY SMITH
Engaging Characters: Further Reflections ................................................ 232
JOHANNES RIIS
Implications of Paradoxical Film Characters for Our Models
and Conceptualizations ........................................................................... 259
JÖRG SCHWEINITZ
Stereotypes and the Narratological Analysis of Film Characters ..... 276
viii Content

CHRISTIAN HUCK / JENS KIEFER / CARSTEN SCHINKO

A ›Bizarre Love Triangle‹. Pop Clips, Figures of Address and
the Listening Spectator ............................................................................ 290
FREDERIK LUIS ALDAMA
Characters in Comic Books .................................................................... 318
HENRIETTE C. VAN VUGT / JOHAN F. HOORN /
ELLY A. KONIJN
Modeling Human-Character Interactions in Virtual Space ............... 329

III Characters and Their Audiences

RICHARD J. GERRIG
A Moment-by-Moment Perspective on Readers’ Experiences
of Characters ............................................................................................. 357
CATHERINE EMMOTT / ANTHONY J. SANFORD /
MARC ALEXANDER
Scenarios, Characters’ Roles and Plot Status. Readers’
Assumptions and Writers’ Manipulations of Assumptions
in Narrative Texts ..................................................................................... 377
URI MARGOLIN
From Predicates to People like Us. Kinds of Readerly
Engagement with Literary Characters ................................................... 400
KATJA MELLMANN
Objects of ›Empathy‹. Characters (and Other Such Things)
as Psycho-Poetic Effects ......................................................................... 416
DAVID C. GILES
Parasocial Relationships .......................................................................... 442

IV Characters, Culture, Identity

MARGRIT TRÖHLER
Multiple Protagonist Films. A Transcultural Everyday Practice ....... 459
RUTH FLORACK
Ethnic Stereotypes as Elements of Character Formation ................. 478
MARION GYMNICH
The Gender(ing) of Fictional Characters ............................................. 506
 Content ix

V Transtextual and Transmedial Characters

BRIAN RICHARDSON
Transtextual Characters ........................................................................... 527
WERNER WUNDERLICH
Cenerentola Risen from the Ashes. From Fairy-Tale Heroine
to Opera Figure ........................................................................................ 542

Bibliography
JENS EDER / FOTIS JANNIDIS / RALF SCHNEIDER
Characters in Fictional Worlds. A Basic Bibliography ....................... 571
Introduction
 JENS EDER / FOTIS JANNIDIS / RALF SCHNEIDER

Characters in Fictional Worlds

An Introduction

1 Questions of Character Analysis and Theories of Character

Most kinds of fiction centrally feature characters – from ad hoc bedtime

stories to the most complex works of art. Some characters are known to
millions of people, such as Anna Karenina or Lara Croft, Ulysses or James
Bond, Mickey Mouse or R2-D2. The aim of this volume is to present a
survey of the varieties of international and interdisciplinary research on
characters in fictional worlds in different media. That such a survey does
not exist to date is perhaps due to the gaps between the disciplines, but
perhaps also to the apparent normality and ubiquity of characters: We
encounter them every day, and they are so familiar a phenomenon that
they do not seem to require closer inspection. Yet another reason could
be that once they are subject to closer scrutiny, characters prove to be
highly complex objects in a number of ways. They remind one of real
persons, but at the same time they seem to consist of mediated signs only.
They are ›there‹ but they do not appear to exist in reality – we do not meet
them on the streets, after all. They do exert an influence on us, but we
cannot interact with them directly. They are incredibly versatile, they
change over time and appear in different forms in different media. The
introduction to Ronald B. DeWaal’s Sherlock Holmes bibliography gives
an impression of this:
This bibliography is a comprehensive record of the appearances in books, periodicals
and newspapers of the Sacred Writings or Canonical tales (fifty-six short stories and
four novels), the Apocrypha and the manuscripts written by Sir Arthur Conan Doyle
between 1886 and 1927, together with the translations of these tales into sixty-three
languages, plus Braille and shorthand, the writings about the Writings or higher
criticism, writings about Sherlockians and their societies, memorials and memorabilia,
games, puzzles and quizzes, phonograph records, audio and video tapes, compact discs,
laser discs, ballets, films, musicals, operettas, oratorios, plays, radio and television
programs, parodies and pastiches, children’s books, cartoons, comics, and a multitude
4 Jens Eder / Fotis Jannidis / Ralf Schneider

of other items – from advertisements to wine – that have accumulated throughout the
world on the two most famous characters in literature.1
Terminology already posits a problem for a general or comparative
approach that wants to examine (maybe even equally ›famous‹) characters
across those media: We have to subsume readers, hearers, viewers, users,
and players under the heading of ›recipients‹, and books, paintings, radio
plays, films, video games, etc. under the heading of ›texts‹.2 (Coming from
literature and moving image studies, the authors of this introduction are
aware of their limited disciplinary perspectives in trying to give a general
survey of the field.)
Moreover, in any media, characters confront those who are concerned
with them – creators, audiences, critics and commentators – with
numerous questions. These questions can be clustered into three groups
concerning the analysis and interpretation of characters.
1. In the production phase of a media product, authors, filmmakers and
other media producers are mainly confronted with the question of how
characters can be crafted in a way that allows them to evoke certain
thoughts, feelings and lasting effects in the target audience. Sir Arthur
Conan Doyle had to invent Holmes in the first place, screenwriters had
to adapt him, casting agents had to cast an actor for the role, etc.
2. The interpretation of a work of fiction confronts critics and scholars
with the question of how characters can be understood, interpreted
and experienced, and by which stylistic devices they are shaped.
3. Studies in the fields of cultural theory and sociology consider
characters as signs of empirical production and reception processes
embedded in their socio-cultural contexts in different historical periods
and (sub-)cultures. The master sleuth Holmes, for instance, has been
read in connection with the socio-cultural developments of a modern,
industrialised society.
Each of these three fields of inquiry – production, interpretation and
cultural analysis – has prompted scholars to find answers and develop

_____________
1 De Waal: Holmes <http://special.lib.umn.edu/rare/ush/ush.html#Introduction>
(Jul. 21st, 2008).
2 When we use the term ›text‹ in this introduction, we include literature, everyday
language, film – and, indeed, all other utterances in which characters may occur.
Following Mosbach: Bildermenschen – Menschenbilder, p. 73, we might define text to
mean ›complex, coherent utterances based on signs, which are contained in a media
format, and, in their totality, communicative and culturally coded‹ (German original:
›komplexe, aber formal begrenzte, kohärente und [als Ganze] kommunikative, kulturell
kodierte Zeichenäußerungen‹; on film as text, see montage/av: Film als Text, and
Hickethier: Film- und Fernsehanalyse, pp. 23–25.
 Characters in Fictional Worlds: An Introduction 5

theories. For the first two thousand years of the debate, the first set of
questions was tackled mainly by practitioners – dramatists and directors,
artists and media producers – with a view to practical concerns. It was
only in the 19th century that a more theoretical, descriptive and systematic
analysis of characters was developed in various disciplines of scholarship,
such as literary studies, theatre studies, and later in film and media studies,
communication studies, the history of art, philosophy and psychology.
Each of these disciplines has produced diverse rival theories on which we
can only cast a passing glance in this introduction (for a more detailed
survey of the research, see the contribution by Henriette Heidbrink in this
volume).
Simplifying matters for the purposes of clarity, we can point to four
dominant paradigms that reach across disciplines but have different
tenets, emphases and methods.
1. Hermeneutic approaches view characters dominantly as representations
of human beings and emphasise the necessity of taking into considera-
tion the specific historical and cultural background of the characters
and their creators.
2. Psychoanalytic approaches concentrate on the psyche of both characters
and recipients. They aim at explaining the inner life of characters, as
well as the reactions of viewers, users, and readers with the help of
psycho-dynamic models of personality (e.g., those developed by Freud
and Lacan).
3. Structuralist and semiotic approaches in contrast highlight the very
difference between characters and human beings, focussing on the
construction of characters and the role of the (linguistic, visual,
auditive or audio-visual) text. They frequently regard characters
themselves as sets of signifiers and textual structures.
4. Cognitive theories, which have been established since the 1980s, centre
on modelling in detail the cognitive and affective operations of
information processing. In these approaches, characters are regarded
as text-based constructs of the human mind, whose analysis requires
both models of understanding text and models of the human psyche.
The rivalry between these approaches in various disciplines and regions
has contributed to the fragmentation of character theory and the co-
existence of viewpoints. The interdisciplinary and international survey we
envisage with this volume may help to remedy the situation. Most
contributors to this book have done extensive research in the field, and
are thus able to present their own established approaches and theoretical
results. We are hoping that this will facilitate a dialogue between different
positions. The essays are roughly clustered into five groups: (1) general
topics (the research on characters, their ontology, and their relation to
6 Jens Eder / Fotis Jannidis / Ralf Schneider

narrative plots); (2) characters and characterisation in different media

(prose fiction, drama, poetry, feature films, pop clips, comics, the
internet);3 (3) recipients’ cognitive and affective responses to characters
(from understanding to empathy and aesthetic evaluation); (4) relations of
characters to identity and culture (stereotyping, gender); and finally, (5)
characters that cross the borders of single texts or media. Clearly, this
clustering is far from being comprehensive. It is only giving a first
orientation and is not supposed to draw rigid lines. Many essays deal with
several topics and could have been located in a different group as well.
This introduction is intended to help to situate the contributions in a
more general context. We hope that our footnotes and references serve as
links for the readers, pointing to essays that deal in more detail with topics
we can only briefly mention here. Keeping our considerations on a rather
abstract level and leaving out extensive examples and historical case
studies, we start with some fundamentals: the definition and ontology of
characters, their relations to real people and to the media they are
represented in. We then turn to action and character constellations as two
important contexts of individual characters in fictional worlds. On that
basis, we examine somewhat more specifically how characters are re-
identified and characterised in different media. From a more global
perspective, characters can then be associated with recurring types and
media genres, as well as with certain functions they fulfill and meanings
they convey. Finally, we conclude with some thoughts on how recipients
respond to characters and what kinds of lasting effects characters may
have.

2 Definition and Ontology of Character

How we define character is relevant not only with regard to theoretical

questions, but also in quite practical terms, for the definition influences
how we analyse characters: If we regard Sherlock Holmes as a person-like
being, we are likely to focus on his personality traits; if we see him as a
sign, we will concentrate on the textual structures of his presentation; if
we think of him as a mental construct, the psychological processes of his
recipients will move centre stage, and so on. Each of these approaches,
and some others, have been explicated in detail, and we can only gesture

_____________
3 Unfortunately, we did not succeed in including further important art forms and media
like painting or TV.
 Characters in Fictional Worlds: An Introduction 7

towards them here.4 Little explanation can be found in the etymology of

the term, and the languages differ to boot. The English term ›character‹
goes back to Greek charaktér, ›a stamping tool‹, meaning, in a figural sense,
the stamp of personality, that which is unique to a human being.5 The
French and Italian terms – personnage and personaggio, respectively – point to
Latin persona, i.e. the mask through which the sound of the voice of an
actor is heard. The German Figur in turn has its roots in the Latin figura,
and suggests a form that contrasts with a background.
In spite of the differences, in all of these languages characters are most
frequently defined as fictive persons6 or fictional analoga to human
beings.7 Such definitions are in accordance with the intuition that we
resort to knowledge about real people when we try to understand fictional
characters. Definitions of this type, however, are not entirely unproblem-
atic: they are too vague as far as the ontological status of fictive beings is
concerned, they are restricted to anthropomorphous characters and
exclude, e.g., animal characters, aliens, monsters and robots. This raises
two questions. First, there is the basic question of the ontology of
characters: What kind of object are they? Second, there is the question of
their specificity: What is the difference between them and other objects of
the same kind?
The ontology of characters has been discussed most widely in philoso-
phy and in literary scholarship.8 One position, according to which
characters are regarded as component parts of fictional worlds, has been
particularly prominent in this context. Fictional worlds are in turn
explained in the scholarly discourses of fictional worlds theories and the
philosophical possible worlds theories.9 Within this framework, a fictional
world is conceived of as a system of non-real but possible states, or as a
constellation, created by the text, of objects, individuals, space, time,
_____________
4 For more detailed discussions of the definition and ontology of characters, see Eder:
Fiktionstheorie; Jannidis: Figur, chap. 5 and 6; as well as the references in the
subsequent footnotes.
5 See the entry in the Oxford English Dictionary.
6 E.g. Wilpert: Figur, p. 298.
7 E.g. Smith: Characters, p. 17.
8 For introductions to the debate in the discourse of philosophy, see Proudfoot:
Fictional Entities; Howell: Fiction; Lamarque: Fictional Entities. The discussion in
literary theory can be found in Rimmon-Kenan: Narrative Fiction, pp. 31–34,
Margolin: Individuals and Margolin: Characters.
9 See Margolin: Individuals; Eco: Lector; Ryan: Worlds, Possible Worlds Theory;
Doležel: Heterocosmica; Pavel: Fictional Worlds; Ronen: Possible Worlds; Buckland:
Digital Dinosaurs. Cf. also the helpful surveys in Martinez / Scheffel: Erzähltheorie,
pp. 123–134, and Surkamp: Narratologie.
8 Jens Eder / Fotis Jannidis / Ralf Schneider

events, regularities, etc.10 Upon closer inspection, however, it becomes

apparent that the character problem is not fully solved by referring to
fictional or possible worlds, for their very status has itself been disputed.11
Models of fictional or possible worlds do allow for an integration of
characters into the larger structure of the world presented in, or created
by, the text, but they do not manage to clarify the ontology of characters
convincingly, because fictional or possible worlds are subject to
ontological problems themselves. What is more, the scholarly discourse
on characters is much older and more varied than that on fictional worlds.
Therefore, it makes sense to start from the perspective of character
proper.
There are four major positions on the ontological status of characters,
and they are highly controversial:
1. Semiotic theories consider characters to be signs or structures of
fictional texts.12
2. Cognitive approaches assume that characters are representations of
imaginary beings in the minds of the audience.13
3. Some philosophers believe that characters are abstract objects beyond
material reality.14
4. Other philosophers contend that characters do not exist at all.15
As we mentioned above, each of these positions has its own far-reaching
implications for the analysis of characters. Each definition thus entails a
particular perspective and a particular method.
This is not the place to deal with the pros and cons of the various
positions in detail, not least because the authors of this introduction are
not unanimous in their theoretical stance: Ralf Schneider conceives of

_____________
10 See, e.g., Doležel: Heterocosmica,pp. 16–23; Ryan: Narrative, p. 91.
11 For a survey of philosophical positions on the ontology of possible worlds, see Melia:
Possible Worlds.
12 Branigan: Point of View, p. 12 (›surface feature of discourse‹); Wulff: Charakter, p. 1
[French ed.: 32]; see also Jannidis’ criticism of (post-) structuralist varieties of this
position (Figur, chap. 5).
13 For psychological approaches in literary theory, see Grabes: Personen; Schneider:
Grundriß; Culpeper: Characterization; Gerrig / Allbritton: Construction, and the
cricitism in Jannidis: Figur, pp. 177–184. No comparably detailed version of this
theory has been put forward in the area of film studies, but it is implied in many
approaches, such as Bordwell: Cognition; Ohler: Filmpsychologie; Grodal: Film
Genres, or Persson: Understanding Cinema.
14 See Thomasson: Fictional Characters, and Reicher: Metaphysik; see also Howell:
Fiction, and Lamarque: Fictional Entities.
15 Künne: Abstrakte Gegenstände, pp. 291–322; Currie: Characters; see also Proudfoot:
Fictional Entities; Howell: Fiction; Lamarque: Fictional Entities.
 Characters in Fictional Worlds: An Introduction 9

characters as mental constructs, as in position 2 above, whereas Fotis

Jannidis and Jens Eder stand for different versions of position 3,
maintaining that characters are abstract objects; another variety of this
thesis can be found in Maria Reicher’s contribution to this volume.16
Despite such differences, we share a number of convictions. The
philosophic-semantic view that characters do not exist is jeopardised by
the fact that it requires extremely complicated logical re-formulations of
quite straightforward utterances about characters: every sentence about a
character would have to be translated into a sentence about the text – we
would not be talking about Sherlock Holmes at all, but about the books
and films in which he appears. Some hold the view that characters are
signs, mere words or a paradigm of traits described by words. A well-
known example of this approach is Roland Barthes’s S/Z (1970) in which
one of the codes, ›voices‹, substitutes for person, understood as the web
of semes attached to a proper name. In this view, a character is not to be
taken for anything like a person, yet on closer examination these semes
correspond to traditional character traits. Moreover, the reduction of
characters to words poses many practical problems in literary and media
criticism. In addition to that, every aspect of meaning of the term ›sign‹
leads to counterintuitive consequences when applied to characters:
characters simply cannot be reduced to signifiants or signifiés or relations
between them, because each of these aspects would imply that one
character is always restricted to the one text to which it belongs, as part of
the overall set of signs. It is, however, a well-known fact that characters
can appear in a number of texts, as the example of Holmes and Watson
clearly shows.
Given this situation, the series of essays by Uri Margolin, by combining
elements of structuralism, reception theory and the theory of fictional
worlds, proved to be a breakthrough. For Margolin, characters are first
and foremost elements of the constructed narrative world: ›character‹, he
claims, ›is a general semiotic element, independent of any particular verbal
expression and ontologically different from it‹.17 If, in a similar vein, we
consider characters to be elements of fictional worlds, which exist either
as subjective mental entities or as inter-subjective communicative
constructs, the question is what differentiates them from the other
elements of the text. To what extent is Sherlock Holmes different from
his pipe, the Thames or a lifelike Sherlock Holmes wax figure? This
_____________
16 See Schneider: Grundriß and Literary Character; Jannidis: Figur, chap. 5; Eder:
Fiktionstheorie; Reicher (in this volume).
17 Margolin: Characterisation, p. 7.
10 Jens Eder / Fotis Jannidis / Ralf Schneider

question has been addressed by referring to some closely connected

criteria that a character fulfills, including being animate, having an
intentional mind (in the phenomenological sense), being able to act, being
humanlike and having person status.18 Some of these criteria, however,
prove to be too broad or too narrow: the criterion of being animate would
on the one hand include the earthworm in the possession of an angler that
Holmes identifies as a clue as a character; on the other hand, it would
exclude inanimate characters, such as robots. Anthropomorphism and
person status would exclude many well-known characters such as Lassie
or the extraterrestrial plant Audrey II (The Little Shop of Horrors). In
contrast to these criteria, the ability to act and to have an inner life (of
whatever quality) appear to be more plausible. In addition to that, an
element of the text is more likely to be regarded as a character if it is a
particular, recognisable entity, not an indistinct part of a mass (of beings).
At the prototypical core of the concept of character, then, is a recog-
nisable fictional being, to which the ability to think and act is ascribed.
Individual characters can deviate from this prototype in a variety of ways
and to various degrees. Models in advertisements, for instance, can be
hard to identify (criterion of recognisability); a character can be a reference
to historical persons, such as Napoleon in historical novels and feature
films (criterion of fictionality); some cannot use their bodies to act, such
as the invalid Johnny in Johnny Got His Gun (criterion of being able to act);
others are even dead from the beginning of the story, such as Harry in
Hitchcock’s The Trouble with Harry (criterion of being animate). In addition
to that, Uri Margolin has pointed out that not all characters exist within
the main level of the fictional world at all. He reminds us that characters
can have various modes of existence: they can be factual, counterfactual,
hypothetical, conditional, or purely subjective.19 At the end of the mind
game movie Fight Club, for instance, Tyler Durden (Brad Pitt) turns out to
be the hallucinated alter ego of the nameless narrator-protagonist (Edward
Norton), who suffers from a split personality syndrome. Cases like this
highlight the relevance of some further questions: What is the relationship
between characters and real persons?

_____________
18 Eder: Fiktionstheorie, pp. 55–59.
19 Margolin: Characters, p. 375.
 Characters in Fictional Worlds: An Introduction 11

3 Characters and People

If we conceive of characters as beings in fictional worlds, to which the

audience ascribes intentionality or action, we must ask what precisely the
difference between characters and real persons is. The differences concern
especially the textual construction and fictional representation of
characters, their ontological incompleteness, and, in connection with that,
the difference between the audience’s knowledge about characters on the
one hand and about persons on the other.20 Obviously, the reception of
characters is quite different from the direct encounter with real persons:
Readers, listeners, or viewers focus on media texts, activate media
knowledge and communication rules, they cannot interact with the
represented persons but can think about their meaning, as well as about
causes and effects, and they can shift their attention from the level of
what is represented (Sherlock Holmes) to the level of presentation (the
words of the book, the actor’s performance). The symbolism and the
communicative mediation of characters mark fundamental differences to
the observation of persons in reality. In addition to that, the texts that
construct characters are fictional. Real persons can of course also be
represented in (non-fictional) texts, such as biographies or the news, but
they do not owe their existence to these texts.
This consideration is connected with the ontological incompleteness of
characters. Objects in the real world have certain properties. If such
objects are mentioned in a non-fictional text, all persons involved in the
communication process will assume that even those properties of the
object which the text does not name and specify explicitly are still
accessible in principle. This is even true in circumstances where there is a
lack of sources, so that the evidence cannot be provided. If, for instance,
the colour of Napoleon’s hair had not been mentioned in any of the
contemporary texts about him, we would still assume that his hair was of a
certain colour, and that this colour could still be found out, through the
discovery of hitherto unknown sources, an exhumation, etc.
The situation appears to be entirely different in the case of characters in
fictional worlds. If the medium that constitutes them provides no
information on a certain property, this property is simply lacking in the
fictional world – there is a gap, as it were, in that world. The recipient has
_____________
20 The term ›knowledge‹ is used in a wide sense here, including also erroneous beliefs,
pre-conscious dispositions, procedural or implicit knowledge, kinds of embodied
cognition, etc. The incompleteness of fictional characters has been discussed exten-
sively in analytical philosophy; see Eaton: Character; Crittenden: Fictional Characters;
Lamarque: How to Create.
12 Jens Eder / Fotis Jannidis / Ralf Schneider

no opportunity to fill this gap in a way that would allow him to consider it
an item of reliable knowledge. We simply cannot know how many
children Lady Macbeth had, or if Sherlock Holmes has a birthmark on his
back – to mention two cases in point which have been discussed
extensively. There is, of course, nothing that would stop the recipient
from contributing such pieces of knowledge, and each individual reading,
viewing, etc. is likely to differ from all other readings with regard to the
unmentioned details the recipient imagines in the process, but on the level
of the fictional universe the text creates, the information will remain
unavailable.
Things get more complicated because the above formulation that ›the
medium which constitutes a character provides information‹ is admittedly
vague. In the most straightforward case, the colour of a character’s hair is
simply mentioned explicitly (in the language-based media genres) or
shown (in the visual media genres). The case is less clear if a text presents
this piece of information implicitly rather than explicitly (see below for a
further discussion of this distinction). A character may, for instance, be
presented as a typical Frisian, or a typical Italian from the south of Italy –
in both cases, information on the colour of the hair is implied. The
question here is to what extent the perception of persons feeds into – or
ought to feed into – the perception of characters. As has become clear,
knowledge that comes from outside the text plays a crucial role in many
cases when a character’s behaviour is to be understood adequately.
Therefore, if we want to understand the text, film, etc. in its historical
context, we need to find out about the psychological and anthropological
knowledge that was available to the author and her or his contemporaries.
This process, however, is quite different from the way we approach
persons, for in a historically adequate interpretation it only makes sense to
fill in information that would have been available in the context of the
text’s original production and reception. If we read, for instance, a
historical report about the symptoms of an unknown disease, we may of
course say that according to today’s knowledge, it is likely that this or that
particular disease is meant; in the case of a fictional text, this procedure
would be anachronistic and meaningless: If the disease is unknown in the
fictional world and its context, the lack of information cannot be
remedied. Whether or not one wants to admit such potentially anachronis-
tic readings depends to some degree on the theoretical background one
chooses: On the one hand, it has been an established practice, e.g. in
psychoanalytical interpretations, to find prove of the symptoms described
by psychoanalysis in texts that precede the development of the discipline
itself by a few hundred years (consider, for instance, Freud’s famous
analysis of the Oedipus myth in Shakespeare’s Hamlet); on the other hand,
 Characters in Fictional Worlds: An Introduction 13

in the context of Foucault’s discourse theory it makes sense to regard the

moment in which a phenomenon – say, an illness – is first described as
the one in which the discourse brings forth the phenomenon in social
reality, so that an a posteriori interpretation of a phenomenon of a previous
epoch raises a number of epistemological and ideological questions. What
is more, we not only make use of our knowledge about persons in
understanding characters, but also our knowledge about character types,
genres and the protagonists they typically feature, and the rules of specific
fictional worlds: The utterance ›I want to see the sun‹ can be understood
adequately in rather different ways, depending on whether it comes from a
human being or a vampire.
Does this mean that characters are indefinitely changeable concepts
which can only be understood in the context of particular contemporary
knowledge about persons and characters? The answer is that in principle
they are, but in spite of all this flexibility there seems to be a core set of
properties, a common denominator that all presentations of characters
share. This prototypical core or ›base type‹ – or basic structure of mental
character models – is constituted by only very few and rather general
properties, which seem to be anthropological givens of the perception of
human (and humanlike) beings: In contrast to objects, characters have
mental states, such as perceptions, thoughts, feelings, and aims.
Accordingly, characters have both an outer appearance and an inner state
of the psyche that is not visible from the outside. This definition of the
base type is supported by recent research on person perception in early
childhood.21 Assumptions about stable features or traits appear to be
essential for most characters, too, so that it makes sense to include this
aspect in the definition of the basic type as well. Other approaches also
include the sociality of characters as a fundamental component beside
corporeality and inner states.22 The relationship between a character and
its environment may presuppose body and mind, but further particular
qualities emerge from social interaction, e.g., social roles. In all three areas
of the general structure of characters – corporeality, psyche, and sociality
– the features that characters are ascribed can be either stable (static) or
changeable (dynamic).
Even if this base type may be the same across cultures, it can only
provide a very general framework. How this frame is filled will depend to
a major extent, and perhaps entirely, on the respective cultural context,
_____________
21 On the base type, see Jannidis: Figur, pp. 185–195, with further references to the
relevant research; cf. also Tomasello et al.: Understanding.
22 Eder: Figur, pp. 173–185.
14 Jens Eder / Fotis Jannidis / Ralf Schneider

which is subject to historical change. The constitution of characters from

textual information and cultural knowledge is based on character
schemata. This concept refers to expectations of regular connections that
exist between two or more pieces of information. Such expectations direct
the inferences of the audience: If one piece of information is given in the
text, the schema allows the reader or viewer to fill in the second bit. Such
schemata include such everyday items of knowledge as the fact that the
consumption of alcohol will lead to intoxication.
The sources of such processes of inferencing consist of, on the one
hand, knowledge about the actual world, especially the social world.23 On
the other hand, there is media knowledge and narrative knowledge about
fictional worlds in general, and about the rules of the narrated world in
particular.24 Social knowledge includes person schemata; images of human
nature; social categories; prototypes and stereotypes; knowledge of
patterns of social interaction; groups and roles; folk psychology and
sociology; the dynamics of social cognition; attribution and the
interpretation of behaviour (e.g., the so-called fundamental attribution
error); the knowledge of prototypical persons and last, but not least, the
self-image of the reader/viewer/user. Media knowledge, on the other
hand, includes an awareness of a text’s communication processes and
fictionality; an awareness that is guided by the rules and aims of
communication as well as media-specific knowledge of genres, modes of
narrative, character types, dramaturgical functions, aesthetic conventions,
star images, contexts of production, intertextual references, and individual
popular characters (e.g., Sherlock Holmes as a pattern of later detectives).
The entirety of the character schemata formulated by, or implied in, a
text, constitutes its ›text-internal anthropology‹.25 Of central importance in
this context are traditional configurations such as the book-keeper, the
melancholic, the extrovert, the beau, the vamp, etc. Such character types
can emerge from a variety of sources: the knowledge of the specific kind
of narrated world, the knowledge about fictional worlds in general, and

_____________
23 On accounts of social perception or social cognition, see for instance Zebrowitz:
Social Perception; Lavine / Borgida / Rudman: Social Cognition.
24 On the interaction between different kinds of social and media knowledge, see Ohler:
Filmpsychologie; Eder: Figur, pp. 162–248.
25 Titzmann: Psychoanalytisches Wissen, p. 184. Titzmann correctly points out that terms
like ›psychology‹ and ›anthropology‹ ought not to be taken literally, because neither
should we project the concepts formulated by the specialist disciplines back onto the
text and its context, nor should we overestimate the coherence of such bits of
knowledge.
 Characters in Fictional Worlds: An Introduction 15

the knowledge about the actual world, including the habitus of social
groups. We will say more on such character types below.
In view of the abundance of knowledge about people and characters in
every society, it seems unlikely that there should still be gaps left in the
fictional world. Even information missing from the text could be filled in
from these knowledge stores. We should not forget, however, that
fictional worlds are not autonomous worlds; rather, they emerge from
processes of communication with their own particular rhetorical structure.
Some aspects of the presentation of characters may be part of aesthetic
structures that reach beyond the characters. Most importantly, characters
themselves can be signs in a number of ways: they can be instances of
exemplary behaviour, they can be symbols or in other ways representative
of feelings, attitudes, problems and the like. In addition to that, characters
are an important part of the emotional structure of literary texts, films, etc.
They influence the feelings, moods and emotions of the audience to a
considerable degree (see the remarks on ›Functions and Effects of
Characters‹ in this introduction). In accordance with the complexity of the
rhetorical structures, the reader or viewer may of course consider the
number of Lady Macbeth’s children. Many of the questions of this kind,
however, will look irrelevant, for the aesthetic structure sketched here will
determine the quality and quantity of the import of contemporary
knowledge.
The differences between characters and real persons come to the fore if
we systematically consider the ways we understand and talk about them.
Theories of reception stress the fact that we understand characters on
several levels:26 Viewers, readers, listeners or users do not only grasp a
character’s corporeality, mind, and sociality in the (fictional) world. They
are building on those processes to understand the character’s meanings as
sign or symbol, and to reflect on the character’s connections to its
creators, textual structures, ludic functions, etc. The latter processes
diverge from the social perception of real persons, and it would be
unusual (to say the least) to think about human beings in those ways.
Moreover, and in accordance with the different levels of reception, the
readers’ or viewers’ meta-fictional discourse about characters (e.g., talking
about them after leaving the cinema) contains sentences of different
logico-semantical structure:27 While the statement ›Holmes is a detective‹
stays safely in the boundaries of the fictional world and might also be
_____________
26 E.g., Persson: Understanding Cinema.
27 Künne: Abstrakte Gegenstände, pp. 295–296, and Currie: Characters, are proposing
different logical transcriptions of such sentences.
16 Jens Eder / Fotis Jannidis / Ralf Schneider

uttered by another character, sentences like ›Holmes stands for human

reason‹, ›Holmes was invented by Sir Arthur Conan Doyle‹, or ›Holmes is
more perceptive than any other fictional or real detectives‹ transcend the
fictional world in different ways, connecting it with reality. But again, we
usually do not talk about real persons in those ways.
Several theories have given accounts of that. Phelan has proposed the
description of characters as participating in a mimetic sphere (due to the
character’s traits), a thematic sphere (as a representative of an idea or of a
class of people), and a synthetic sphere (the material out of which the
character is made).28 In a similar way, but starting from a triangulation of
several theories, Eder distinguishes between analysing characters (a) as
artifacts (how are they represented, and what are their textual structures?);
(b) as fictional beings (what features do they possess in the fictional
world?); (c) as symbols (what do they stand for?); and (d) as symptoms
(why are they the way they are, and what are their effects?).29 According to
that distinction, we perceive real persons as inhabitants of a world (the
actual one), but we do not perceive them as artifacts, symbols or
symptoms. This also indicates a crucial difference between emotional
reactions to characters and to real humans: We not only emotionally react
to characters as fictional beings, but also react to their (brilliant or clumsy)
representation, to the (often controversial) meanings they impart, to the
intentions of their makers (e.g., propagandistic ones), or to the supposed
effects they may have (e.g., on minors). Those kinds of reactions in turn
may influence the feelings we have for the fictional being (see the section
on ›Recipients’ Reactions and Relations to Characters‹ in this volume).30
Distinguishing between different aspects of analysing characters might
also be helpful in understanding their media specificity, intertextuality, and
transmediality. It seems plausible that characters which are represented in
different media (like Holmes) may – or even have to – retain their core
properties as fictional beings and symbols (e.g., in many film adaptations,
Holmes is still a detective standing for human reason), while their meta-
fictional properties as artifacts and symptoms usually change (in film, for
instance, Holmes is represented by images of actors). The next section
elaborates on such questions.
_____________
28 Phelan: Reading.
29 Eder: Figur.
30 For a reformulation and elaboration of this model in terms of semantics and
pragmatics see Uri Margolin’s contribution to this volume. Margolin has been one of
the first to argue for considering characters’ different aspects as non-actual individuals
in fictional worlds, as thematic elements, as topical entities of discourse, and as
artificial constructs.
 Characters in Fictional Worlds: An Introduction 17

4 Trans-medial and Media Specific Aspects of Character

The section on defining character (above) has shown that characters are
entities in fictional worlds which are brought forth through signs, but are
not identical with those signs. Most characters in modern societies are not
created in face-to-face narrations, but with the help of media, such as
books, comics, theatre, film, TV, the radio, or computer games. The way
they are created is therefore subject to the conditions of the respective
media – an aspect systematically dealt with in a forthcoming anthology by
Rainer Leschke and Henriette Heidbrink.31 Some aspects of characters
and their presentation are the same across the media, while others are
media specific. In what follows, we will first deal with the media specific
aspects and then turn to the trans-medial ones. If we want to understand
the relationship between characters and the media, we need to define the
contested term ›media‹ first. Following Siegfried J. Schmidt, ›medium‹ can
be seen as a compound term which comprises four components: (1)
instruments of semiotic communication, such as natural languages and
images; (2) media technologies used by the producer and the recipient,
such as print or television technology; (3) the institutionalisation of the
media-technological dispositives, for instance by publishing houses, TV
stations, cinemas, etc; and (4) the individual media products, such as
books, newspapers or TV spots.32 Schmidt refers to the systemic
constellation of these factors as a ›medium‹, while other authors apply the
term to the individual components. In any case, the function of media is
to transmit, with the help of signs, certain experiences of a perceptual,
cognitive, emotional and bodily kind.33
Schmidt’s definition of the term ›media‹ points to fundamental factors
that contribute to the media specific forms of the production and
reception of characters: media sign systems, technologies and institutions.
The options for the production of character implied in this can be further
differentiated with the help of Marie-Laure Ryan’s narratological
categories, which can be applied to characters in different media.34 First,
_____________
31 Heidbrink / Leschke: Formen.
32 The original quotation reads: »(1) semiotische Kommunikationsinstrumente [z.B.
natürliche Sprachen oder Bilder], (2) die jeweilige Medientechnologie auf Produzenten-
wie auf Rezipientenseite [z.B. Druck- oder Fernsehtechnologien], (3) die sozialsystemi-
sche Institutionalisierung der medientechnischen Dispositive [etwa durch Verlage,
Fernsehsender, Kinos etc.] sowie (4) die jeweiligen Medienangebote wie z.B. Bücher,
Zeitschriften oder Fernsehspots« Schmidt: Medienkulturwissenschaft, pp. 351–369.
33 Cf. also Vogel: Medien, p. 292.
34 Ryan: Media, pp. 282–292.
18 Jens Eder / Fotis Jannidis / Ralf Schneider

the media can be differentiated according to their spatio-temporal

extension: the radio play creates characters in a specified time slot, but
without positing a particular space; photo-novellas in contrast are spatially
specific, but not temporally (you can look at the characters as long as you
wish); other media, including the cinema or the theatre, develop characters
over a specific time and space. Second, certain kinetic properties of media
and their characters are connected with the spatio-temporal ones:
characters in paintings are static; characters in computer games are
dynamic. The third aspect refers to the semiotic code of character
representation, and the sense of the recipients affected: On the one hand,
there is language as a »code which speaks to the mind through the
conventional meaning of its signs«, while on the other there are some
»sensory modes of expression, such as sounds and pictures, which convey
meaning without relying on a fixed semantic content«.35 Because of this,
characters develop varying degrees of sensory presence or distance,
concreteness or abstraction. Fourth, if several codes and sensory channels
are combined, they are related to each other in varying states of priority –
music, for instance is more significant in the creation of character in the
opera than in drama in general. The fifth differentiation deals with the
technological support and materiality of signs: television allows for live
broadcasts of theatre figures, while cinema does not. Finally, the cultural
role and methods of production/distribution is, according to Ryan, a
pragmatic differentiation: Individual media fulfil a range of functions
within the system of the media, and they follow rather practical
conventions: The production of a feature film is more expensive than, say,
the production of a novel, and the economic considerations this entails
frequently influence the creation of characters.
These differentiations can be illustrated briefly using the example of
Guy Ritchie’s action movie Sherlock Holmes (2009). The cinematic
representation of Holmes deviates from that of Doyles’ narratives in a
number of ways: In the film, Holmes is not represented by printed words
but by moving images and sound, which gives him a sensory specificity
that at the same time diminishes the range of individual imaginations by
the recipients. The film employs specific visual and auditory strategies of
characterisation, including the acting of Robert Downey Jr., the dynamic
montage and the urging music. Close-ups invite the study of Holmes’
facial expression, while long shots place him in his London surroundings.
Holmes’ actions are slotted into a precise temporal dramaturgy, and
despite the ironical mode of narration, the impression that the character is
_____________
35 Ibid., p. 291.
 Characters in Fictional Worlds: An Introduction 19

mediated through a narrator recedes into the background. This action-

oriented version of Holmes, which emphasises the bodily rather than the
mental capacities of the sleuth, owes much to economic constraints.
However, some similarities to the literary or dramatic representations of
Holmes do remain: as in novels and short stories, Holmes in the movie is
characterised through various, changing perspectives, times and places; as
in theatre performances, he is represented by an actor and addresses the
senses of the spectators in a finite temporal sequence. In none of these
media, however, the audience is in control of the character or interact with
him – ludic elements that are characteristic of computer games.
In view of such basic differences the question arises to what extent we
may still speak of the same character here. It is frequently the case that
characters are distributed via numerous texts, media and cultures over
time.36 But is the character the same in all of these cases? Is the hero of
Ritchie’s action movie identical with Doyle’s Holmes? Or did Doyle
himself perhaps create different Holmes characters in his various stories?37
Characters can be presented across the media, and in principle, any
character can appear in any medium. Their appearances in various media
products may differ according to their qualities as artifact, symbol or
symptom (i.e., their crafted-ness, their meaning, and their references to
reality), but it will still be the same character as long as the core features of
the fictional being remain the same. There are, however, kinds of
characters or character types that are specific to one medium and
therefore difficult to transfer. Computer game characters such as Pacman
who fulfil a predominantly ludic function and possess only few traits
would appear fairly uninteresting in other media – who would want to
read a novel about Pacman? In contrast to this, characters like Holmes are
present in various media; they are at the centre of transmedial storytel-
ling.38 This is not only the case with media adaptations of individual
_____________
36 See Werner Wunderlich’s contribution to this volume, in which he demonstrates this
with the example of Cenerentola, providing ample material.
37 The contributions by Brian Richardson and Maria Reicher to this volume deal with
questions of trans-textual and trans-medial identity of characters. According to
Richardson, truly trans-textual characters are legitimised through the author of the
original character, and they are congruent with it in terms of traits. Reicher offers an
explanation for this kind of continuity of traits, differentiating between two ways of
approaching the character: when experts interpret a character, they refer to a ›maximal
character‹, possessing a multitude of detailed traits; normal recipients, in contrast, tend
to speak about ›sub-maximal‹ characters with fewer differentiated traits. A reference to
the diegetic core elements of a character in the sense of a ›sub-maximal character‹ will
suffice to establish trans-textual identity.
38 Jenkins: Culture.
20 Jens Eder / Fotis Jannidis / Ralf Schneider

works, as e.g. literary adaptations, but even more so with the protagonists
of the major entertainment franchises (including James Bond, Lara Croft,
Harry Potter, or the characters from the Lord of the Rings), and even
characters from advertisements (Ronald McDonald) or the personae
created by successful comedians (Sacha Baron Cohen’s Borat). Such trans-
medial characters occur in a dense network of stories, games and other
entertainment products, giving them coherence, and in many cases also
their name. They can be found in the media cultures of many countries
and exist, beyond the original fictional world, in marketing, merchandis-
ing, fan fiction, news reports and recipients’ conversations.
The trans-medial presence of characters reaches its peak when charac-
ters are not even created through one original text in one medium, but
amalgamated from sources in various media. Henry Jenkins, in his
discussion of ›synergistic storytelling‹ emphasises that certain dispositions
in the characters ›The Kid‹ and Niobe in The Matrix Reloaded can only be
understood if the corresponding short films and computer games from
the context of the film are known.39 We are only able to hint at the
complex questions connected with the media specificity, intertextuality
and trans-mediality of characters – serious scholarly investigation of them
has only just begun.

5 Character and Action

For a long time, theories of prose fiction, theatre and film have played
character and action off against each other, and they have tended to give
the preference to action. One of the oldest theoretical statements on
character reflects on the relation of character and action in this way: »for
tragedy is not a representation of men but of a piece of action […].
Moreover, you could not have a tragedy without action, but you can have
one without character-study«.40 What Aristotle said in relation to tragedy
became the origin of a school of thought which claims that in order to
understand a character in a fictional world, one need only analyse its role
in the action.41 The most common labels for character in use since Greek
antiquity refer in fact to the role, or function, of a character in the action
of a narrative: ›Protagonist‹ refers to the main character of a narrative or a
play, and ›antagonist‹ to its main opponent. In modern literary theory, the
_____________
39 Jenkins: Culture, pp. 103–110.
40 Aristotle: Poetics, 1450a.
41 Pfister: Drama, p. 220.
 Characters in Fictional Worlds: An Introduction 21

approach that reduced characters to mere functions in the action was put
on a new foundation, especially in the plot theories of structuralism and in
actant models.42 In a ground-breaking corpus study of the Russian
folktale, Vladimir Propp43 analysed a hundred Russian fairy tales,
constructing a sequence of 31 functions which he attributed to seven areas
of action or types of character: opponent; donor; helper; princess and her
father; dispatcher; hero; false hero. Greimas44 generalised this approach
with his actant model in which all narrative characters are regarded as
expressions of an underlying narrative grammar composed of six actants
ordered into pairs: the hero (also sujet) and his search for an object; the
sender and the receiver; the hero’s helper and the opponent. Each actant
is not necessarily realised in one single character, since one character may
perform more than one role, and one role may be distributed among
several characters. Schank’s concept of story skeletons also starts from the
idea that stories have an underlying structure, but in his model there are
many such structures and therefore many different roles for actors, e.g.
the story of a divorce using the story skeleton ›betrayal‹ with the two
actors: the betrayer and the betrayed.45
The models presented here stand in contrast to a number of ap-
proaches in drama theory and in practical film theory.46 The changing
number of potential positions that characters can have surrounding the
core constellation of protagonist and antagonist indicates that such
models are to some extent contingent, except for the core itself. If we
wish to explicate which of them lends itself best to the analysis of
functional roles, their theoretical foundations and their practical
applicability would have to come under close scrutiny and be compared
accordingly. Whether one takes as one’s starting point the syntax of the
language (see Greimas), analogies to mental problem solving processes (as
do the developers of Dramatica), or Jung’s theory of archetypes – none of
the theoretical bases is in itself unproblematic. In spite of the differences
between the models, they all contain similar core functions. This suggests
that these at least can be found in typical and wide-spread types of
narration.
The claim that the action of a narrative is more important than the
characters must, in any case, remain ambiguous as long as it remains

_____________
42 Chatman: Story, pp. 108ff.; Koch: Menschendarstellung.
43 Propp: Theory.
44 Greimas: Structural Semantics.
45 See Schank: Story, chap. 6.
46 See Asmuth: Dramenanalyse, pp. 99ff.; Pfister: Drama, pp. 234ff.; Dramatica 1999).
22 Jens Eder / Fotis Jannidis / Ralf Schneider

unspecified what ›action‹ is supposed to mean. The term action can refer,
in order of increasing specificity, to the following items:
1. The entire set of events in a story, including events which are not set in
motion by characters but by chance, nature, etc. Included in this
definition are also events in which characters are neither involved as
agents nor affected in any other way, e.g., a sunrise that only the
audience watches.
2. The overall activities of the characters, including the consequences of
these activities. This definition also includes the mental processes in
the minds of characters. Many modernist novels, with their focus on
the inner life of characters, present little action in terms of physical
activity and changes of situation, but much action in terms of the
characters’ psychological processes. Virginia Woolf summed up this
tendency of the modernist novel to ›turn within‹ in her programmatic
plea for a representation of the ›truth‹ of human existence beyond
surface action, and even beyond the seeming logic of a plot: »Let us
record the atoms as they fall upon the mind in the order in which they
fall, let us trace the pattern, however disconnected and incoherent in
appearance, which each sight or incident scores upon the conscious-
ness«.47
3. Only the intentional behaviour of the characters, their speech and
action. Some media genres favour narratives that focus on such
behaviour, either because they profit from the spectacular, such as the
adventure story, the war movie and a good number of computer
games; or because the very intentionality is the point of the narrative,
as is the case in the mystery, crime and detective story – the ›whodunit‹
usually includes the question why someone committed a crime.
4. Only the physical activities of characters, excluding their speech acts.
Although there will be few examples of fictional worlds in which
characters move and do things without ever speaking, some media
genres heavily rely on action as a means of characterisation and
narration, as can be seen in the suggestion of many screenwriting
manuals to tell the story by action rather than by dialogue.
The action of a story in the sense of item 1 may not be constituted
exclusively by characters’ actions in the sense of items 2 to 4, but in most
cases, such actions will account for the major part of the story: Stories are
always stories of and about someone, and they narrate the activities of
anthropomorphous characters as a rule.48 Character, in contrast, can in
_____________
47 Woolf: Common, p. 190.
48 See Eder: Dramaturgie, pp. 78–82.
 Characters in Fictional Worlds: An Introduction 23

principle be presented without any action, as is the case in portraits,

descriptions or sculptures. Even in the temporal media, such as film, some
minor characters are characterised not by action but, for example, by a
physiognomy that hints at the features of personality. At least in some
media and some phases in the narration of fictional worlds, then,
characters can be independent of action (in every sense), which is not true
the other way round. There are also features of characters that are
independent of action, and action-oriented approaches tend to overlook
these aspects. They fail to capture the audience’s understanding and
experiencing of the fictional world, which is frequently focused on the
characters. It is particularly in the cinema that a character has »a palpable
autonomy that seems to make action subordinate to his/her prior
existence«,49 and a similar statement can of course be made for theatre.
Features of characters that are independent of action may influence not
only the audience’s empathy, but also determine the thematic focus of a
text, as well as its ethics and rhetoric. Therefore, if the action-oriented
perspective is not complemented by one that focuses on character, central
aspects of fictional worlds are overlooked.
With reference to the dramaturgical functions of characters in the plot,
we can differentiate between two approaches. The traditional approaches
have tended to employ a genre- and period-specific vocabulary for action
roles. Examples that come to mind are the allegorical stock characters of
the morality play (everyman, vice, folly, etc.), confidant and intriguer in
traditional drama, the libertine and the fop in English Restoration
comedy, or villain, sidekick, and henchman in the popular media of the
20th century. Other approaches have looked at characters’ functions in
more generalised ways. In an influential work, Campbell, for instance,
described what he called the ›monomyth‹, using a term coined by James
Joyce. A monomyth is an abstraction of numerous mythological and
religious stories marking the stages of the hero’s way: separa-
tion/departure; the trials and victories of initiation; return and reintegra-
tion into society.50 According to Campbell, who bases his argument on
Freud’s and especially on Jung’s version of psychoanalysis, the monomyth
is universal and can be found in stories, myths, and legends all over the
world. The tradition of universalist approaches has continued until today.
For instance, Vogler has applied Campbell’s approach to the structure of
Hollywood films and has been influential in the screen trade.51 From a
_____________
49 Bordwell: Cognition, pp. 183–198.
50 Campbell: Hero, p. 36
51 See Vogler: Journey.
24 Jens Eder / Fotis Jannidis / Ralf Schneider

more theoretically founded perspective, Patrick Colm Hogan argues that

identifiable universal patterns of story-telling and emplotment are
generated by a restricted set of human emotions across time and space.52
Although many of the approaches to plot have focused on the action
rather than the characters, as we have seen, there are at least some
approaches in the bewildering diversity of plot theories,53 which have
emphasised the connections rather than the distinction between character
and action. E.M. Forster coined the wide-spread definition of plot as the
causal and logical connections which exist between the actions of the
characters in a story, and which underlie the temporal sequence in which
the story is presented. In this vein, we may look at the ›and‹ in ›character
and action‹ as signifying a connection rather than a separation. The logic
of the story then automatically implies the logic of the character’s
intentions and hopes as to future events. Consequently, in order to
understand the story, the reader or viewer needs to understand the wishes,
plans and motivations of the characters. Among the approaches to plot
that pay attention to the wish-world of the characters and the mental
projection of potential, non-actual actions by the audience are those put
forward by Bremond54 and the more elaborated formulation in terms of
possible worlds theory by Ryan.55 Furthermore, Palmer expands the
perspective to include the ›social mind in action‹.56
The ›motivation‹ of characters constitutes the interface between
characters and action. The term motivation usually refers to a part of the
psyche, the inner life and personality traits: the entirety of psychical
processes that initiate, maintain and regulate behaviour. This definition
includes aims, wishes, feelings and drives.57 We explain the actions of
characters by ascribing them such motivations, and we expect certain
actions once we know their motivations. This is why motivation tends to
be the motor and the centre of a story, transmits its theme and presents a
significant influence on emotional reactions. It is important both for
narrating characters and for interpreting them. Thus, even a rather
formulaic narrative that has traditionally been analysed in terms of
characters’ plot functions requires at least one character’s motivation in
order to set the action in motion. Consider, for instance, heroic epics
_____________
52 Hogan: Mind; see also Hogan’s contribution to this volume.
53 Dannenberg: Plot.
54 Bremond: Logic, pp. 387–411.
55 See Ryan: Worlds.
56 Palmer: Minds; see also his contribution in this volume.
57 On the psychological definition of 'motivation', see the relevant sections in Gerrig /
Zimbardo: Psychology, chap. 11.
 Characters in Fictional Worlds: An Introduction 25

centering around a quest plot, which can be found in Western narratives

from Beowulf to The Lord of the Rings. If the aim of saving the community
from an external or internal threat were not an important element in the
value system of the hero, he would never venture out to face the dangers;
there is either the fear of disturbances to the community (and, conversely,
the wish to keep disturbances away); also, he must hope for some reward
for his deeds, whether the acknowledgement of the community, the love
of a lady, or riches. Values, aims, wishes, hopes and fears therefore propel
the action of the hero, and they invite a psychological reading of the
character at the same time. Most modern narratives deal with problems that
motivate the characters’ actions. The plotting of mainstream films, for
instance, follows the aims of their protagonists, who are involved in
exciting problem-solving processes.58 Motivation is a precondition also in
other forms of narration: Episodic narratives deal with the momentary
problems of several characters; character studies focus on unconscious
needs of persons on a quest, and even characters who do not even try to
fulfill their wishes produce actions on a small scale to which we ascribe
motivation. Even when surreal narratives and films, such as Un chien
andalou, prevent us from reconstructing motivations, they generate their
potential to provoke from the fact that apparently we cannot help but
look for motivations.
Motivations in many cases are the core of the personality of fictional
beings, particularly in their interaction with other characters, so that the
basic motivations of characters are a major element of their evaluation and
interpretation. Once we turn to understanding narratives on the basis of
characters' motivations, it is helpful to differentiate between various kinds
and aspects of motivation. Concepts from psychology, philosophy, drama
theory and theories of fictional worlds can contribute to a more refined
definition of motivation.59 Motivations differ according to whether they
are localised in the person itself or in its environment, whether they are
egotistic or altruistic, whether the person is aware of and willing or able to
reflect on them, how stable and influential they are, and whether they are
consistent or contradictory. To take an example: Is a murder that Holmes
investigates explained by a momentary impulse, the personality of the
murderer, the external situation or determined by previous events? Is it a
planned homicide by a hired killer or an act of madness? Did the doer
pursue the plan for a long time? Did he kill for his own advantage or
because he wanted to protect others? Did he have any scruples, did he
_____________
58 Eder: Dramaturgie.
59 Stückrath: Literatur; Eder: Figur, pp. 428–463.
26 Jens Eder / Fotis Jannidis / Ralf Schneider

engage in an inner moral battle or is he evil to the core, as for instance

Holmes’ arch rival Professor Moriarty? Motivation is, finally, an important
factor in the constellation of characters, which places individual characters
in a network of relationships.

6 Character Constellations

Characters do not tend to appear on their own in narratives. In most cases

they are part of a constellation of at least two, and frequently many more,
characters. This constellation puts all the characters of a fictional world in
relation to each other. The traditional classifications of characters that go
back to antiquity are already based on the notion that characters are
related to each other: The term ›protagonist‹, derived from the term for
the first player in the classical Greek tragedy, classifies a character in terms
of his function in the action, as does the opposite term, ›antagonist‹. Other
terms point in the same direction: minor character, opponent, parallel or
foil character – all view characters as in some way related to other
characters.
An average 19th century English novel presents some two dozens of
interrelated characters, a novel by Charles Dickens easily goes up to fifty
or more, and in films like Ghandi or a TV series like Six Feet Under, these
systems of relationships comprise up to several hundred speaking parts. A
character constellation is, however, more than the mere sum of all the
characters. Its structure is determined by all relationships between the
characters: relations of importance, correspondences and contrasts of
properties and functions, interaction and communication, conflict and
agreement, mutual seeing and listening to, wishes and desires, power and
value systems, narration and being narrated, perspective and participa-
tion.60
This is why the character constellation of a text is an abstraction. One
way of reconstructing it is by looking at the changing configurations in the
scenes (scenes in drama, chapters or sequences in novels, sequences or
shots in film, etc.) in which certain characters are present at the same time.
The dynamics of the character relationships can be approached via the
succession of scenic configurations. The configuration profile of a theatre
performance, a film or an episode highlights, for instance, which

_____________
60 On character constellations see Platz-Waury: Figurenkonstellation, pp. 591f.; Pfister:
Drama; Tröhler: Hierarchien, pp. 20–27.
 Characters in Fictional Worlds: An Introduction 27

characters appear, and when and how many times they do, with which
other characters they appear together and with which they don’t.
Having said that, there are of course numerous examples in which the
quantitative analysis of only the configuration misses the point of the text.
Who would doubt, for instance, that the love relationship in Shakespeare’s
Romeo and Juliet is the central interest of the play? They appear together on
the stage, however, in only four of the scenes, and some of these
encounters are rather short. Another example is Graham Swift’s novel
Last Orders (1996), in which four elderly friends go on a journey together
in order to bring the ashes of their deceased friend, Jack Dodds, to the sea
and disperse them there, according to his will. During the journey, the
individual memories produce a kaleidoscopic view of Jack and his various
relationships to each of the four friends plus a large number of other
characters. The character constellation of this novel thus includes a range
of configurations that are ›only‹ remembered, in addition to the ones that
occur in the main story time. What is more important in these cases is the
qualitative analysis of the relationships and the properties of the
characters. To some extent, then, character constellations can be mental
constructs invited by the text but not necessarily projected there by the
configurations. Such qualitative aspects as slight contrasts between two (or
more characters), whether they have scenic appearances together or not,
are particularly usful for highlighting the individuality of each of them and
pointing to the thematic focus of the text.61
The concept of the character constellation is an important tool for the
analysis and interpretation of media products and works of art. It has
proved particularly useful in the analysis of novels, TV series or
transmedial narrative universes that contain complex social networks.
With the help of constellation analysis, various kinds of relationships can
be investigated, including their social relationships (conflicts and bonds),
their values and norms (moral and otherwise), their diegetic and aesthetic
similarities and differences (parallel and foil characters), the hierarchies of
relevance (main vs. minor characters), and their dramaturgical and
thematic functions.62 The analysis of those aspects is closely connected to
questions of ideology, politics, and understanding texts as indicators of
collective dispositions, problems, wishes and fears in a certain time and
culture.63
_____________
61 In their contribution in this volume Emmott / Sanford / Alexander distinguish
between principal characters and scenario-dependent characters.
62 Eder: Figur, pp. 464–520.
63 For instance, Margrit Tröhler connects the proliferation of multi-protagonist films in
recent years with increased cultural skepticism about heroes and a greater willingkess
28 Jens Eder / Fotis Jannidis / Ralf Schneider

7 Identifying Characters

Some media texts and works of art such as paintings, sculptures or one-act
plays display characters in a temporally continuous, unified or even static
representation. In most narratives however, characters are presented in
separate sequences interrupted by parts in which they do not appear. The
question is how a character is recognised in each of the sequences, and
how the individual representations are related to each other. This point is
even more acute if these representations occur in different media formats
or texts: if, for instance, a drawing of Holmes illustrates a novel, or if we
identify Holmes across several novels or serial instalments.64
The fact that in real life humans ascribe a spatio-temporal continuity to
bodies and assume that persons continue to exist even if they cannot be
seen or heard anymore, is a precondition for this ability to identify
characters. Fictional worlds constructed in the media can rely on our
willingness to keep the characters existent if they are not shown or
mentioned. For this to happen, each individual medium needs to establish
the link between the initial presentation of a character and the subsequent
occurrences. Obviously, the way characters are presented differs across
various forms of media. In purely language-based texts nothing acquires
an existence of its own if it is not explicitly mentioned. The link between
one presentation of a character and the next will differ according to the
technique of mentioning. Usually, the link is established by a stable
referent, whether in the form of a name (Holmes) or a repeated
description (›my friend‹ – as Sherlock Holmes is sometimes called by the
narrator Watson). The mention or presentation of objects, especially
articles of clothing, can have the same effect of supporting the identity of
the character referred to.
Scenes play a major role in the process of identity construction within
individual media texts. Literary and audiovisual narratives can, following
Catherine Emmott, be understood as sequences of scenes or situational
frames, of which only one is active at a particular point in the narration.65
Scenes integrate spatio-temporal aspects with events and information
_____________
to consider multiple perspectives on complex modern worlds. See Tröhler: Welten;
also her contribution to this book.
64 See Maria Elisabeth Reicher’s contribution on the ontology of characters and Brian
Richardson's chapter on transtextual characters in this volume.
65 See Emmott: Comprehension, and the contribution by Catherine Emmott, Anthony J.
Sanford and Marc Alexander in this volume; cf. also the reformulation of this
approach in Jannidis: Figur, pp. 109–149, which contains a more detailed discussion of
the various aspects of identifying characters.
 Characters in Fictional Worlds: An Introduction 29

about the characters. Links to previous character representations in scenes

follow special rules; in a scene in which two female and one male
character are present, the personal pronoun ›he‹ will suffice to refer to one
of the characters unequivocally. Identifying characters with the help of
(anaphoric) references in language-based texts can be described
systematically as follows. References to characters can be direct or
indirect. Direct references include the use of names, definitive descrip-
tions (e.g., ›the boy‹, ›the visitor‹) and personal pronouns.66 Indirect
references can be made by the presentation of direct character speech
without any introductory fomulae, the description of actions that do not
refer to any one character directly (›a hand grabbed‹) or the use of the
passive voice (›the window was opened‹). The first reference to a character
can be termed the ›introduction‹; if reference to the character is made in a
subsequent frame, we speak of ›identification‹.67 We can differentiate
between normal, false, impeded, and deferred identifications. A ›false
identification‹ occurs when a previously mentioned character is identified
but it turns out later that some other character was in fact being referred
to. An ›impeded identification‹ does not refer unequivocally to any specific
character, and a clear reference to the character or characters is never
given in the text, while in the case of ›deferred identification‹ the recipient
is ultimately able to establish the identity of an equivocally presented
character. Deferred identification can further be broken down into an
overt form in which the recipient knows that he is kept in the dark and a
covert form in which he does not.68
In purely auditory media like the radio play, voice, typical sounds or
musical leitmotifs are added to the above list of linguistic linking devices.
In purely visual media, such as comics or the silent movie, identification is
usually (i.e., except for intertitles or captions) established by the similarity
or identity of the visual appearance of a character, which is created
through the bodies, and especially the faces, presented by the actor, or the
drawing or animation.69 In silent movies, characters are three-dimensional
entities, many of whose outer attributes (looks, beauty, dress, age, gender)
are immediately apparent, while other features such as names, traits, social
position and relationships only emerge in the course of the film, if at all.
Sound film and other audiovisual media (video games, internet clips)
combine stills, moving images, spoken and written language, sound and
_____________
66 Margolin: Literary Narrative, p. 374.
67 In film studies, Murray Smith uses the term ›re-identification‹ to distinguish it from
›identification with‹ characters. See Smith: Characters, pp. 110–116.
68 Jannidis: Figur, chap. 4 and 6, based on Emmott: Comprehension.
69 Smith: Characters, pp. 110–141 on the re-identification of characters in film.
30 Jens Eder / Fotis Jannidis / Ralf Schneider

music, and can therefore also combine different techniques of identifica-

tion. Rick Blaine in Casablanca, for instance, is first introduced by linguistic
means, for his name appears on the lettering on his café, and other
characters talk about him. This will arouse the audience’s curiosity and
make them expect his first appearance, which happens, however, only ten
minutes after the film has started. The combination of language and image
is established in this scene by having Rick write his name on a cheque and
by having the camera pan on the upper part of his body after a cut. As in
the case of Rick Blaine, the identification of characters in film, and also in
music videos, is frequently predetermined by additional contextual cues.
This not only concerns the text itself, but also certain para-texts. All cues
which are apt to activate social and narrative knowledge in certain
communicative contexts (pre-release information, advertising, paratexts,
etc.) play a role in identifying characters, and they can also contribute to
characterisation (see the following section).70 When characters are
represented by actors or pop stars who have had appearances in other
media products, a complex network of identifications emerges.71
Characters in film can even be re-identified when they are presented by
different actors with entirely different looks, as is the case in Todd
Solondz’ Palindrome.72 The narrative context is of great importance for
establishing continuity here. A character that has been introduced by the
text and identified by a recipient is attributed properties in the course of
the reception. This is the area of characterisation, which we will discuss in
the next section.

8 Characterisation

Until recently, characterisation was understood to mean the ascription of

mainly psychological or social traits to a character by the text.73 In fact,
however, texts ascribe all manner of properties to characters, including
physiological and locative ones. Thus, the term ›characterisation‹ can refer
to different phenomena, not all of which are distinguished from each
other clearly enough in common usage. In many cases, the term is used
when a person is ascribed a certain number of stable properties. These
_____________
70 Eder: Figur, pp. 326–254.
71 On characters and stars in the music video, see the contribution by Christian Huck,
Jens Kiefer and Carsten Schinko in this volume.
72 See the contribution of Johannes Riis in this volume for a discussion of this
phenomenon.
73 E.g. Chatman: Story.
 Characters in Fictional Worlds: An Introduction 31

properties tend to be character traits, as for example in ›he was shy and
anxious‹. Many languages have a rich stock of words for such character
traits at their disposal.74 This dimension of the term ›characterisation‹
includes statements about the outer appearance of a character. These can
be rather unspecific – ›his beauty attracted men and women alike‹ – and
perhaps in some way related to character traits, but they can also refer to
certain specific physical markers, such as a high forehead. Whether such
remarks are decoded to refer to character traits depends on the cultural
contexts: In the 19th century, the popular pseudo-sciences of physiognomy
and phrenology provided a code system for creating and understanding
literary character via the descriptions of the shape of the head, nose, chin,
mouth, etc.75
In a broader sense, the term characterisation is frequently used to refer
to information about habitual actions, the circumstances of a person and
his or her social relationships: e.g., ›he lived a very secluded life‹, ›his flat
was furnished in a plain style‹, ›the family doctor was his only friend‹. Such
pieces of information tend to be indirect characterisations: they invite the
recipient to look for a character trait that motivates the action, the
circumstances or the relationships. Usually, however, more information
comes in to play, for one piece of information may also hint at other
motivations: A shy person may be homeless and ashamed, and somebody
may be living a secluded life because they suffer from paranoia.
The broadest definition of characterisation includes all information
associated with a character in a text. In addition to the usages already
listed, this includes information about time, place, actions, and events
connected to the character. In many cases, this is contingent information
that does not hint at stable character traits. Part of this information can of
course be understood to contribute to the narrative identity of the
character – comparable to other operations of identity construction. It
even makes sense to use this broadest sense of characterisation, simply
because it is not easy to decide where to draw the line between informa-
tion that is characterising and information that is not, and because the
question arises whether we would not have to regard all information as
contributing to narrative identity. The word ›character‹ in ›characterisation‹
puts the emphasis on a specific mental entity that is imagined as stable
(e.g., ›he is a difficult person‹). In this sense, only statements about
_____________
74 Allport and Odbert found nearly 18.000 English words for character traits, and since
then there has been considerable further research on that topic. Cf. John: Taxonomy,
pp. 66–100.
75 For example, Lavater’s physiognomy was very successful and influential; Graham:
Lavaters’ Essays.
32 Jens Eder / Fotis Jannidis / Ralf Schneider

character traits ought to count as characterisations. Having said that, there

is the danger that one particular conception of character/personality,
which may be culturally specific, forms the basis of all characterisations.
Therefore, even if we would argue for a very broad definition of
characterisation, we have to acknowledge that the usage of the term in the
second sense is more common, both in everyday language and in literary
scholarship. In any case, whoever uses the term in whatever sense ought
to be aware of the implications.
Characterisation (in the wide sense) can then be defined as the process
of connecting information with a figure in a text so as to provide a
character in the fictional world with a certain property, or properties,
concerning body, mind, behaviour, or relations to the (social) environ-
ment. From the perspective of reception, this distribution of information
about a character corresponds to processes of understanding the
character: Textual cues or signs activate inferences based on different
kinds of knowledge about reality as well as about media and communica-
tion.76 In the language-based media, characterisation works by directly and
verbally ascribing to characters certain traits which are not visible from the
outside, particularly psychological traits. Indirect characterisation also
ascribes traits directly in words, or it reports actions which may hint at
invisible states and properties. This includes all manners of activity, from
habitual movements to the description of actions a character performs in
his or her working life, down to the report of habitual movements or gait.
In addition to that, when characters are focalizers, their mind style, i.e.
habits of reasoning and of formulating thoughts in the mind, may hint at
traits.
Literary scholarship has introduced the differentiation of modes of
characterization, some of which are media-specific, while others are not.
As early as the 19th century, literary critics spoke of the difference between
direct and indirect characterisation. Scherer,77 for instance, differentiates
between direct presentation, which involves the explicit naming of the
traits (»Ich zähle die Eigenschaften auf, die jemand besitzt«) on the one
hand, and indirect presentation, in which the traits, and the entire
personality of a character or a person need to be guessed from words,
opinions and actions, on the other (»bei welcher man aus Worten,
Gesinnungen, Thaten gewisse Eigenschaften und so den ganzen
Charakter errathen lässt«). He argues that writers in his day prefer the
_____________
76 Cf. also Jannidis chap. 6; Schneider: Grundriß, pp. 80–90; Eder: Figur: pp. 69–107;
168–232; 326–372.
77 Scherer: Poetik, pp. 156–157.
 Characters in Fictional Worlds: An Introduction 33

indirect mode to the direct one, which is used in historiography, and

sometimes in the novel. The indirect mode, according to Scherer, is both
more skilled (»kunstvoll«) and closer to life, because this is the way we
infer personality traits of other human beings. This technique of having
the audience guess (»die Technik des Errathenlassens«) relies on the
activity of the audience (»Selbstthätigkeit des Publicums«), whose interest
can be increased by precisely this strategy.78
To conceptualize direct and indirect characterization as a binary
opposition, however, is less convincing than meets the eye. All of the
above characterisation strategies in both the language-based and the visual
and audiovisual media may appear more or less direct to the recipients, and
in many cases it is all but impossible to define which is the case: If a
character in a novel is described as ›plain‹, it will to some extent depend
on the specific cultural and social milieu of the character – and the
reader’s knowledge of its conventions – whether this verbal ascription is
directly understood to refer to a certain disposition or not. In a similar
vein, can the close-up of a care-worn face in a film still be called ›indirect‹?
So it is probably more accurate to view direct vs. indirect as a continuum;
characterisation can then be described as more or less direct.
Another important differentiation is that between altero-
characterisation and self-characterisation, for information about characters
can be provided by agencies other than the character (the narrator, for
instance, or other characters in the same fictional world), or the character
may ascribe properties to him- or herself. Other characters may pass on
perfectly reliable information about the character in question, and there is
no reason to assume that each and every altero-characterisation is entirely
wrong simply because characters’ perspectives happen to be naturally
restricted by subjectivity. The other characters, however, may also pursue
their own aims and motivations in the social interaction when they make a
characterising statement. Such statements can therefore reveal their own
value systems, so that every explicit altero-characterisation can also be
read as an indirect self-characterisation of the utterer. Moreover,
characterisation is not always reliable. Some textually explicit ascriptions
of properties to a character may turn out to be invalid, as when this
information comes from an unreliable narrator or fellow-character.
Explicit ascriptions may also turn out to be hypothetical or purely
subjective.
The crucial issue in the process of characterisation is thus what
information, especially of a psychological nature, a recipient is able to
_____________
78 Ibid., pp. 156f.
34 Jens Eder / Fotis Jannidis / Ralf Schneider

associate with any character as a member of the fictional world and where
this information comes from. There are at least three sources of such
information: (a) textually explicit ascription of properties to a character,
whether in altero-characterisation or in self-characterisation; (b) inferences
that can be drawn from textual cues (e.g. ›she smiled nervously‹); (c)
inferences based on information which is not associated with the
character by the text itself but through reference to historically and
culturally variable real-world conventions (e.g. when the appearance of a
room reveals something about the person living there or the weather
expresses the feelings of the protagonist).79
Most inferences can be understood in terms of abductions. Abduction
is a more heuristic form of reasoning than deduction and induction.80
Abductive reasoning infers the case from the result and the rule: (1)
Socrates is mortal. -- (2) All humans are mortal. -- (3) Socrates is human.
Such reasoning is not obligatory; if ›Socrates‹ is the name of a dog, (1) and
(2) are still valid, but (3) is false. The inference is therefore based mostly
on likelihood and it is part of a heuristic which produces well-founded
assumptions rather than certified knowledge in the best case. The question
in each case is whether the right rule was used for the phenomenon, i.e.,
the observed result. If we understand inferences in terms of abductions,
the fundamental role of character models and of the character encyclope-
dia as rules becomes obvious: (1) is given in the text, but (2) is usually only
presupposed by it and has to be supplied. Characterisation can therefore
be re-conceptualised as a process to which both the text and the recipient
contribute (see below).
Another key issue concerns the limits and underlying rules of such
inferences when they are applied to fictional beings. Ryan, noting that
recipients tend to assume that a fictional world resembles the real world
unless explicitly stated otherwise, adopts the philosopher David Lewis’s
›principle of minimal departure‹.81 In a thorough criticism of this and
similar hypotheses, Walton points out that this would make an infinite
number of inferences possible, and he comes to the conclusion:
There is no particular reason why anyone’s beliefs about the real world should come
into play. As far as implications are concerned, simple conventions to the effect that
whenever such and such is fictional, so and so is as well, serve nicely […]. 82

_____________
79 For a systematic description of such inferences employed in characterisation, cf.
Margolin: Characterization.
80 Keller: Theory, chap. 9, based on Peirce.
81 Ryan: Fiction.
82 Walton: Mimesis, p. 166.
 Characters in Fictional Worlds: An Introduction 35

This approach, in turn, increases the number of conventions without

necessity and without providing any convincing argument as to how
recipients go about accessing these conventions, apart from drawing on
their real-world knowledge, despite the fact that many conventions apply
only to fictional worlds. Even so, this does not invalidate Walton’s
criticism, which can probably be refuted only by including another
element: the fact that characters are constitutive of fictional worlds which
are not self-contained, but communicated. Recipients’ assumptions about
what is relevant in the process of communication determine the scope and
validity of inferences.83 Conan Doyle’s readers may of course infer that
Holmes’ chin is stubbly when he does not shave; that his bodily functions
are comparable to those of Watson, and, indeed, the reader; that he wears
underwear under his characteristic garments – but both agents in the
communication process, Doyle and his readers, appear to follow a tacit
understanding that these points are unlikely to be what the story is about.
The presentation of characters is a dynamic process, as is the construc-
tion of characters in the recipient’s mind. Schneider,84 building on
concepts developed by Gerrig & Allbritton,85 has proposed a model for
describing the psychological or cognitive dynamics that come into play
here, based on the ›top-down‹ and ›bottom-up‹ processes observed during
empirical studies on reading comprehension. The processing of
information in the top-down mode involves the activation of a knowledge
structure, such as a schema or a category, stored in long-term memory;
this structure is initially triggered by a piece of textual information and will
then guide the further processing as long as possible. Bottom-up
information processing, in contrast, involves the successive accumulation
of textual information in working memory, where it is kept accessible until
it can be connected with prior knowledge or turned into a category or
schema itself. In understanding a character in a fictional world, the
recipient builds a mental model of the character into which all characteris-
ing information is integrated. Depending on the quality and quantity of
the information on the character presented in the text on the one hand,
and the availability of knowledge structures of the recipient on the other,
the mental model construction can proceed in either of two ways: A top-
down process occurs in the application of a category to a character, by
slotting the information given in the text into this category, while a
bottom-up process results from the successive integration of information
_____________
83 See Sperber / Wilson: Relevance.
84 Schneider: Grundriß; Schneider: Theory.
85 Gerrig / Allbritton: Construction.
36 Jens Eder / Fotis Jannidis / Ralf Schneider

on a character, which will lead to an individualised representation. At the

beginning of a character presentation, textual cues may trigger various
types of categorisation: social types (›the teacher‹, ›the widow‹); literary or
media types (the hero in a Bildungsroman); text-specific types (characters
that do not change throughout the story and are introduced in a way that
evokes precisely that expectation). The mind, following a principle of
parsimony, will tend to slot as many characters into such categories as
possible. Successful top-down processing, however, depends on the
availability of knowledge structures, and the varieties of types just
mentioned involve different types of knowledge that feed into the
processing: social knowledge and categories from folk psychology; literary
or media knowledge, especially narrative knowledge (of genres with their
typical character constellations, character types, actants, and characters’
narrative functions, etc.).86 While producers of characters can use
categorisation cues for efficient characterisation, on the side of the
recipient the categorised mental model is felt to be complete at a fairly
early stage.
In contrast to the top-down processing that takes place in these forms
of categorisation, bottom-up processing occurs when the recipient is
unable to integrate the given information into an existing category,
resulting in a personalised mental model of the character. This is
frequently reached by producers of characters through the distribution of
characterising information over the text rather than providing it initially,
by presenting many pieces of information that do not easily fall into any
one category – or both. Personalised characters can of course also be
members of a category, but this is not the focus of their description.
Reading a text, watching a film, or using other narrative media involves
building up either categorised or personalised character models, and the
type of processing elicited accounts for different experiences: Characters
that can be personalised tend to trigger more emotional involvement and
more tolerance for change and development (they are therefore
comparable to what is called a ›round‹ and ›dynamic‹ character in
structuralist approaches). Categorised characters, in contrast, tend to be
the reliable representatives of a property, and are likely to attract increased
attention only when information is presented that does not fit the
category, for this will make the recipient aware of his or her previously
unconscious inferencing. Generally, information encountered in the text
_____________
86 See Ohler: Filmpsychologie; Eder: Figur, pp. 168–248; see also the sections on
›Characters and People‹, ›Character and Action‹ and ›Character and Genre‹ in this
introduction.
 Characters in Fictional Worlds: An Introduction 37

subsequent to the categorisation or personalisation may change their

status and possibly decategorise or depersonalise those characters.
The role of names in interpreting characters may be taken as an
example for the functioning of characterisation devices, because it has
been treated repeatedly, resulting in different ways of classifying name
usage.87 Telling names occur in many guises: there are names that quite
overtly hint at individual characteristic qualities, features or habits (e.g.
allegorical figures, or Mrs. Malaprop in Richard Brinsley Sheridan’s
comedy The Rivals, 1775). In the tradition of Western literature, characters’
names have often referred to figures from the Bible, so that the characters
in question have inherited more complex sets of properties, both in the
sense of psychological dispositions and their function in the story – Mary
and Ruth, Thomas and Stephen are names of characters that can be found
in a large number of novels of the 18th and 19th centuries, when authors
could still count on their recipients’ knowledge of the sacred texts. Telling
names are characterisation devices that appear on a graded scale between
the two poles of direct characterisation on the one hand, and indirect
characterisation on the other: If a name closely follows the label for a
property – as it has for instance been formulated in folk psychology – the
name functions as a direct characterisation. In contrast to that, many
characters have names that acquire a characterising function only in the
course of the text, when the actions and properties of the character turn
out to be reflected by its name. In accordance with the reception-oriented
approach sketched above, the most likely effect of a directly characterising
telling name is that the recipient constructs a mental model of the
character based on this quality from the moment when the character is
first mentioned. The model is then likely to be activated top-down in the
further information processing. Whether or not the model is kept stable as
a categorised one, and whether the one quality remains the only one in
focus, depends on the further interaction between textual information and
the recipient’s activation of knowledge. Some texts use a telling name to
elicit an expectation only to frustrate that expectation later on.
In theatre and audiovisual media, names are also a common device, but
more importantly, characterisation is intimately connected to casting and
acting. The casting of an actor or actress for a role will determine not only
what the character looks like, how he or she speaks and moves (stature,
facial expression, gestures, body language, proxemics, etc.) These outer
attributes also provide hints to further properties, such as age, gender,
ethnicity, character and social milieu. A close analysis of the performance
_____________
87 See e.g. Lamping: Name; Birus: Vorschlag.
38 Jens Eder / Fotis Jannidis / Ralf Schneider

of actors and actresses, i.e. the individual way in which they execute body
movements, speech acts and actions is therefore of central importance for
an understanding of theatrical and audiovisual characterisation.88 On a
general level, various forms of over- or underacting in comparison with
real-life repertoires of expression are as important as the differentiation
between a self-referential autonomous performance (as for instance in
musicals) and a more unobtrusive integrated performance,89 which shapes
different acting styles, as for instance the repertoire style or method-
acting.90 If the actors are well-known, audiences can activate knowledge
about star images and role biographies.91
The different media all make use of many further specific techniques of
characterisation, for instance camerawork, production design, sound
design, music, and editing in audiovisual media; styles of painting; or styles
of drawing in cartoons. Our cursory remarks in this section can only give
a vague impression of the multifarious forms of characterisation, many of
which also rely on the recipients’ knowledge of certain character types.

9 Character Types

The term ›character types‹ is used here to refer to a fixed set of character
traits, in the sense of a gestalt, which feature repeatedly in certain media
products which belong to at least one specific cultural milieu and its
recipients’ collective knowledge. Common examples of such types include
the mad scientist, the altruistic nurse who works herself to exhaustion, the
femme fatale, the schemer at court, the trickster, etc. Such character types
frequently occur as parts of larger social constellations, so that they are
related to each other. In many comedies, for instance, we find the jealous
but slightly stupid old cuckold in combination with a significantly younger
wife and her young and daring lover. Such constellations also account for
typical action sequences (see the section on ›Character Constellations‹).
In the literature on the subject, a number of specific terms are used for
characters types, including ›archetype‹ for types of trans-historical and
trans-cultural distribution, and ›stereotype‹ for the ideological and
derogative typification of social groups.92 Two aspects of such types must
be differentiated: the prototypical constellation of character traits within
_____________
88 See Fischer-Lichte: Semiotik; Dyer: Stars; Naremore: Acting.
89 Maltby / Craven: Hollywood, pp. 249–257.
90 Dyer: Stars, pp. 136ff.
91 Ibid., pp. 127ff.
92 Eder: Figur, pp. 375–381.
 Characters in Fictional Worlds: An Introduction 39

the fictional world on the one hand, and the typical representation of the
character on the aesthetic level on the other. The two aspects can
coincide, but this is not necessarily always the case: Social stereotypes may
lurk behind an extraordinary and formally complex representation. Two
common assumptions about character types are false anyway: the
assumption that they always possess only few traits and the assumption
that they never change in the course of the story.93 A typified character
can possess a quite complex system of traits and be represented in much
detail, as in the case of Molière’s imaginary invalid (Le Malade imaginaire). It
is not the simplicity of a character, but the degree to which it agrees with
established schemata which turns it into a type. Typified characters can
also change, though they tend to change in a typical way: The cuckold
tends to realize he has been duped in the last scene in a comedy of
manners, and – with enervating regularity – the typified career woman of
many romantic comedies is overwhelmed towards the end by the insight
that marriage and household are her best choice, after all.
In the history of media fictions, it was a matter of course for a long
time that the construction of fictional characters would revert to types.
This even increased the plausibility of character presentation and
supported claims of mimetic authenticity.94 This attitude changed radically
in the course of the 18th century, when the concept of individuality gained
central importance.95 Since then, character types have been rejected for
being non-realistic, at least in the areas of so-called high culture. The ideal
of a representation of individuality has been approached by using
constellations of traits that stand in apparent contrast to each other, or of
which some are surprising to find in combination with the others.96 A
character construction of such complexity, however, is in most cases
possible only for very few characters in one work, and even these
characters tend to be innovative variations of well-known types.97
There are various sources for character types. Some of them are
without a doubt based on everyday social types that a society develops and
that can be studied as stereotypes by social psychology. Age, gender,
_____________
93 This is the opinion of Dyer: Images, pp. 12ff.
94 See Ruth Florack’s contribution in this volume where she discusses how the use of
common knowledge condensed in national and ethnic stereotypes contributed to the
plausibility of the poetic fiction.
95 Luhmann: Individuum, pp. 149–258.
96 See the famous definition of round characters by Forster: Aspects.
97 Shorter texts, for example poems, rely on using types and typified character
constellations; cf. Simone Winko’s contribution in this volume which discusses
characters in poetry.
40 Jens Eder / Fotis Jannidis / Ralf Schneider

origin and social status have been central aspect of such stereotyping since
Aristotle – there is a line of tradition that formulates and discusses such
types which goes back to antiquity. The ›Characters‹ of Theophrastus, and
its extended 17th-century revision by La Bruyère, are the most well-known
examples. Another source is the store of fictional media representations of
characters that exist in a society.98 Such stock media characters – which
turn into stock characters partly through frequent use in widespread
fictions – in many cases owe their existence to some very influential works
or genre contexts. The character of the hard-boiled detective, for instance,
acquired its popularity in US-Amerian film of the 1930s and 40s through
the iconic representation in The Maltese Falcon. Upon closer inspection,
some character types prove to be typical for specific genres; Sherlock
Holmes, e.g., is not only a figure in a range of media representations
located in Victorian England, the type also turned into a blueprint of a
large number of detectives in all periods and locations (see the section on
›Character and Genre‹).99 Since the representation of characters relies
heavily on such types, a lot of research has been done in the field, whether
on types in one particular medium or in transmedial fictional worlds like
Star Trek or Sherlock Holmes. The questions to what extent the creation of
types is influenced by media specific features, or whether there are certain
media specific types, have received significantly less attention.100
The fictional media representations are part of the overall media
representations of a culture, and as such they absorb social stereotypes of
gender, national character or the habitus of certain professions. Due to
the pressure to be innovative, the fictional representations modify the
stereotypes and feed them back into the circulation of social patterns. It is
therefore hard to overemphasise the contribution of fictional media to the
distribution and modification of such auto- and hetero-stereotypes.
To be sure, there is a whole range of stereotypes which a particular
society regards as purely fictional, but there is also a large area in between
the fictional and the non-fictional representations involved in the creation
of social types. This is the reason why they play a major role in the
creation and distribution of ideologies. As a consequence, clichés,
including anti-Semitic ones, gender-images or national stereotypes, have

_____________
98 See Schweinitz: Stereotypes (in this volume).
99 As is well-known, Sherlock Holmes is based to a large extent on Edgar Allen Poe's
Dupin. See also the contribution by Wunderlich in this volume, who sketches the
process of Cinderella's typification.
100 See the section on ›Transmedial and Media Specific Aspects of Character‹.
Exploring the Variety of Random
Documents with Different Content
in the British is nothing compared to the fact that if we leave our
Legation the Boxers and Chinese soldiers will immediately burn
them and loot them, and this may give them such a lust for loot and
pillage that it may become an incentive strong enough to overcome
their national fear of attacking, and make it most terribly difficult for
us to hold out until the troops come. Until the troops come! What a
wail that is! and it is heard at all times, and all people take their turn
in asking somebody else, “When will they come?”
This afternoon we were in Mrs. Coltman’s room, and her sweet
baby was asleep in a funny, old-fashioned, high-backed crib.
Although the sound of exploding bullets was to be heard outside the
house, we were much startled to feel one—you can’t see them, they
come so fast—enter the room, hit the headpiece of the baby’s crib,
detaching it from the main part, and bury itself in the opposite wall.
An inch lower and it would have cut through the baby’s brain. His
mother picked him up, and all of us flew into a room on the other
side of the house, where we felt we would be free from shot, at any
rate coming from that direction.
We were accompanied by the wife of the Chief, Mrs. Conger,
conspicuous for her concise manner, and an open follower of Mrs.
Eddy. She earnestly assured us that it was ourselves, and not the
times, which were troublous and out of tune, and insisted that while
there was an appearance of warlike hostilities, it was really in our
own brains. Going further, she assured us that there was no bullet
entering the room; it was again but our receptive minds which falsely
lead us to believe such to be the case. With these calming (!)
admonitions she retired, and I can honestly say that we were more
surprised by her extraordinary statement than we were by the very
material bullet which had driven us from the room.
All women are busy sewing up sandbags to strengthen our
defence, while bullets are raining into the compound like hailstones.
A man comes rushing to where we are working, and tells whoever is
in charge of filling the sandbags that a hundred, or as many as
possible, must be taken to such and such a barricade, or it cannot
hold out. We get snatches of the real state of affairs very often in this
way.
 June 23.
Yesterday, the 22nd, the Austrian marines vacated their Legation,
although Von Rostand, the Austrian Chargé d’Affaires, and other
people greatly criticized them for having left too soon. These marines
then went to the French Legation, and M. and Mme. von Rostand
became Lady Macdonald’s guests at the British Legation. The
Belgians stayed with the Austrians until they left, when they came to
this compound, and the Belgian Minister also became a guest at the
Legation. The Dutch compound and the Austrian compound are still
burning.
Yesterday at ten o’clock in the morning a sort of terror, almost
unaccountable, seemed to sweep over the entire length and breadth
of our lines; the French soldiers got in a terrible funk, left their
Legation to Boxers, fire, or anything else that might appear, and ran
all the way without stopping to the British Legation, where they said
everything was “lost.” The Germans also got the fright, but after
coming up Legation Street half-way, they turned back, and not only
took a stand in their own Legation again, but they sent men into the
deserted French Legation and kept it manned, so that if the Boxers
came they would be resisted, and not be allowed quietly to take
possession.
The Russian compound is the only passage-way by which the
American marines can escape and retire to the British Legation, and
it was understood that in case of an attack from the Chinese serious
enough to necessitate everyone leaving their Legations, the
Russians would not close their big gate opening on Legation Street
until our American soldiers had entered, when they would hold out
there (in the Russian compound) as long as possible, and then
retreat all together to the British Legation. Our Russian friends,
however, forgot this little arrangement, and when our men were also
seized with this panic and left the Wall, and retreated through our
Legation across Legation Street to the Russian gate, they found it
not only locked and barred against them, but no one near enough
even to hear them knocking. They excused themselves afterward by
saying they had left a tiny gate open farther down the street, but as
none of our people knew there was such an entrance, we thought
this a rather poor excuse.
However, in an hour’s time, after this terror had passed over the
entire line, our marines had returned to the United States Legation,
and had manned the Wall again. The French returned later to their
Legation which the Germans had kindly guarded for them in the
interim, rather disheartened to think that the scare they had started
should prove to have been only in their own overwrought minds. As
the French and German Legations occupy two important positions,
and are constantly being attacked by Boxers and soldiers, the
French Legation could have been taken very easily by the Chinese
had the Germans not occupied both Legations. They are directly
opposite each other in Legation Street.
Our men already have the reputation of being the crack shots of
any of the guards in Peking. It has been noticed that when our men
aim they bring down their game—whether the game is a Chinese
soldier’s head or a Boxer.
Yesterday it seemed too hard that, after the nervous excitement
and fright to everyone in the morning, Providence did not withhold
the terrible fire that broke out almost in our very midst in the park
directly next the Wall. Each hour seemed more terrible than the one
before. A huge column of smoke went up into the air, and in its
centre forked tongues of flame burst out. It seemed impossible that
this enormous fire—one so large or so near I have never before
seen in my life—would not in an hour or so completely burn us up.
The Boxers or soldiers who had so successfully started it must have
been overjoyed to see their work, knowing it would take almost
superhuman power to put it out, although I am sure they could not
have thought it possible that we could extinguish it.
There was little enough hope written on people’s faces in our
compound to make us feel, for a time at least, that perhaps the
Chinese might be successful, and by burning one wall that played so
important a part in our defence, they could enter and massacre us
without having to attempt an attack by scaling. Had there been a
wind blowing this enormous column of fire in our direction, we could
not have fought it at all, and the entire long wall which divides the
British Legation from the Empress-Dowager’s carriage park would
have fallen.
Our men scaled ladders and worked like New York firemen in the
way they strove together and in the good sense they exhibited. I
suppose man is able to keep his head clear when he knows that this
may be his last chance in this world to save his skin from Chinese
savages, and that his arm develops in consequence a good deal of
strength. Men who were on top of the wall, throwing down buckets of
water on the fire, and handling with as much care as possible the
small rubber pipe that we are using as a hose, came down every
fifteen minutes, to be relieved by others, for they were half scorched,
some badly burned from cinders and falling débris, and all of them
had lost their breath in that terrible heat.
It must be remembered that while these men were on this wall
they were beautiful targets for Chinese sharp-shooters, and we
found afterwards there were many in the Chinese troops. There were
three wells in the compound, and from the two biggest there was a
line of men and women passing buckets, ewers, and any other kind
of vessel that was available, filled with water, to the men who were
actually fighting the fire on the wall. One realizes the heroism it takes
to continue working at a fire though half scorched, but what shall one
say of these men who worked under the ordinary danger of a
scorching fire, and who knew they were the target for the continuous
rifle-fire and sniping that was kept up throughout? The sky was grey,
and the men on the Wall made agonizingly big and black silhouettes
for the Chinese to aim at.
If I live to be a thousand, I could never see a queerer collection of
people working together to extinguish a fire, and with the object to
save themselves from a massacre—coolies, missionaries, soldiers,
and Ministers Plenipotentiary working and straining every muscle for
the same object. Surely Peking never before saw such unanimity of
her foreign residents. I was in that line of men and women passing
buckets, and so was the wife of the French Minister, and many other
well-known women.
 MRS. R. S. HOOKER

Fargo Squiers, Dr. Martin, and Dr. Poole, surgeon for the British
Legation, were three soot-covered people who came to our rooms
after the fire was entirely out,—which meant they had worked
desperately for many hours without stopping. To say they were
thirsty would not be truthful—they were parched. Dr. Poole
whispered that the only cup he knew big enough to quench his thirst
was a big loving-cup that was in a small closet in a corner of the
room (this house having been his before the siege), and that if I
would fill it with Apollinaris he would put in the whisky. I filled my
order, and he poured out about four fingers of Scotch into the bottom
of that big loving-cup, and as he drank it slowly, holding it by both
hands, I thought I had never seen such thankful eyes as were his
during that long and pleasant well-earned drink.
Again to-day thousands of sandbags have been made by the
women. Shooting continues all the time, and to-day a cannon was
fired from the Ch’ien Men Gate, which we hope may mean that our
troops are coming and the Chinese resisting them. Prince Ching is
supposed to have under his command in China fifty thousand troops,
and he must be friendly to us, or we feel he would have ordered half
of his troops to Peking before this to finish us. It is stated that some
of them have shot at the Boxers, but this is hardly credible. Prince
Ching is a Prince of the first Order, and head of the Tsung-li Yamen.
Dr. Morrison is the most attractive at our impromptu mess; he works
wherever a strong man is needed, and he is as dirty, happy, and
healthy a hero as one could find anywhere.

June 24.
Two weeks ago to-day the troops started from Tien-tsin. Yesterday
by 11.30 a.m. the Hanlin Library, directly behind Sir Claude’s house
in the compound, was fired by the Chinese, and the way we fought
the flames I described yesterday, only perhaps the men felt a little
stronger. They have succeeded once in putting out an enormous fire,
so why should they not be able to do so to-day? This time, however,
the wind was against us, so that from the morning until seven o’clock
at night we were fighting it desperately.
How absurd it is to have any “consideration” for people like the
Chinese! After the big and dangerous fire of the day before
yesterday, the committee on fortifications and defences suggested
that the world-famous Chinese College (the Hanlin Library) should
be burnt by us in such a way that the Chinese could not use it as a
position to fire on us from. There was danger, too, that they would
fire it themselves, taking it for granted that the fire would surely
spread to such an extent—aided by themselves with kerosene—as
to burn this entire end of the Legation. The Defence Committee was
afraid of this, and at a conference of the Ministers it was discussed,
and more or less unanimously disapproved of. “Such vandalism!”
they said. “This trouble will soon be over, and then what a disgrace
to have to acknowledge to the world that we deliberately burnt one of
the finest, if not the finest, libraries in the East!” We only had to wait
twenty-four hours to see that our consideration for the famous library
was thoroughly thrown away, for, notwithstanding the troubles “will
be over in a few days,” the Chinese seem so anxious to destroy us
before these troubles have passed that they themselves burned this
gorgeous old library, containing as it did all their oldest and most
revered literature, in the hope that they could burn out a large
enough part of our Wall to facilitate their getting in.
The great danger was over by seven o’clock, but careful sentries
watched all night in case a strong wind should start, and small
isolated buildings were burning all night, so that, looking down from
our house to that end of the compound, it made one think of the
blazing flames one sees at night in the oil districts of Pennsylvania.
With these terrible fires the Chinese are clever enough to keep up a
volley of rifle-fire, so our labour is a frightful danger to every man
working. The suspense was hard to bear, because it was over five
hours before the most optimistic dared say, “We are comparatively
out of danger;” and nobody knew just what would happen if this end
of the compound was to go, for this British compound is looked upon
by all as the strongest and last resort in Peking, and that is why, of
course, all of the women and children and stores of every description
have already been sent here.
Twenty-five Chinese Sisters, who were rescued from the Nan-
t’ang, come to our tiny little courtyard at the back of our house—on
which charming view, by the way, our windows look—and cook in a
big caldron their portion of rice that is allowed them by the General
Committee. These people and all of the families of Mrs. Coltman’s
“boys,” and Mrs. Squiers’s “boys,” fill up our tiny backyard with their
cooking, etc., until, from the propinquity of these people, one is
almost convinced that one is living and sleeping in the heart of the
Chinese settlement of San Francisco.
The marines at our Legation, who naturally will not come here until
they are forced to, are in a very bad way about food. From May 29,
when they arrived in Peking, they were fed by a Chinaman who
contracted to feed them all at so much per man, and he fed them
splendidly, but since we have been besieged he naturally has no
market to call upon. Mr. Squiers has fed them for some days out of
his own storeroom, but each meal makes a terrific hole in his
supplies. There are fifty men and two officers, and naturally they do
not get satisfied on one tin of sardines and a loaf of bread. We have
cooked rice in great quantities, putting many tins of corned beef into
it, cooking it in the same big caldron that the Sisters use. Preparing
the food over here makes it very difficult getting it to them, as there is
constant sniping going on, and it is extremely dangerous to walk
from one Legation to another.

June 25.
So far the moral of the Legation, or, I should say, of this
compound, is decidedly good. The weather is very warm, but the
heavy rains that generally come at this time of summer are not here
yet. Only a few babies are sick with dysentery, and there are some
cases of scarlet fever and malignant malaria. The hospital, a house
of four rooms, only holds a comparatively small number of patients.
Let us pray it will not have time to fill up. Dr. Velde, a surgeon of the
German army, who has been detailed for three years to the Legation
in Peking, is a man who for very clever and consecutive work has
already been decorated by his Emperor. His forte is surgery, and it
looks as if he would save the medical day here in Peking. Dr. Poole,
I think, will consult and work with him. One of our marines has
already been killed, and two are at the hospital wounded. These
people, who are the first to lose their lives and get hurt, make one
feel that truly this is war.
I was at the hospital with Mrs. Squiers this morning. Several men
were brought in, and they all had to wait their turn to be operated on,
and the two nurses were so busy assisting with the work in
connection with the operation of the moment that nothing was done
for a wounded Cossack who was laid on the floor. He was covered
with blood, and it trickled down his chest and formed into a pool all
around him, his face an olive-green—the colour one sees in
unskilfully painted pictures of death—so livid, I never believed even
dying people could look that way. He lay there for some time,
everyone in authority too busy except to tell me to do what I could for
him, and keep the flies from bothering him until he should die,
probably in twenty minutes. He was shot through the lungs.
People continue to be cheerful, but it is strange considering that
we have death around us morning, noon, and night. The gossip, if
one can so call the reports and rumours that are circulating
throughout the compound nearly every few hours, is that a Russian
declares he knows their troops are coming, because during the night
a sentry saw a green rocket go up into the sky. It is supposed that
the Chinese have no green rockets; therefore, as the Russians
constantly use green rockets, it must be a signal from the Russian
troops to let us know they are practically at the door. And so on and
so forth.
To-day Dr. Morrison went over to the Fu, where the Chinese
Christians are, to assist Colonel Shiba in some difficult and
dangerous barricading work, and incidentally to take a part in a
sortie. He was in command of a squad of Japanese and Italian
soldiers, the latter most ineffective, and the former magnificent. They
cleared the Chinese out of some alleys which Colonel Shiba decided
must be added to their lines for the protection of the Chinese
converts. The brunt of the fighting fell on the Japanese, and one was
killed and three wounded. Such a clever idea it was of Dr. Morrison’s
and Dr. H. James’s to put these poor wretches in Prince Su’s park,
which, owing to its close proximity to the Japanese Legation, seems
now to fall upon the Japanese to defend.
Dr. H. James met with such a terrible end yesterday! From the
gate of the British Legation facing the canal, he looked down towards
the Imperial Wall, and seeing there several Chinese officers carrying
a regimental flag with which he was familiar, he started out, as if on
the impulse of the moment, to parley with them. He was watched
with breathless interest. Although from the time he left our wall until
he reached them he held his hands up to show he was unarmed,
they grasped him in the fiercest way, dragging him over the bridge
beyond our range of vision. The horror of his too probable fate is
hanging like a pall over the compound. We cannot understand how a
man, knowing the Chinese as well as he does, could have been so
mistaken in their character as to trust himself to them with such
confidence.
During the two fires in the Mongolian Market Place and in the
Hanlin University a great many Chinese were shot by us, and when
possible we straightway threw their bodies into the flames.
Unfortunately, some Boxers were captured during the almost hand-
to-hand fighting that has taken place, and confined in this compound.
They were all shot at dawn this morning.
Captain Myers has been in command for two days and two nights
on the Tartar Wall, with no sleep. This afternoon the marine quarters
in the United States Legation caught fire and for a time it looked as if
the whole American compound would go, but with hard fighting it
was put out.
Mr. Cheshire, of the United States Legation, is willing to take the
most difficult and dangerous work wherever an interpreter is needed,
and for some nights now he has been on the Tartar Wall directing
and encouraging the picked Chinamen forming the gang of labourers
who nightly help our marines to strengthen the barricades. Many
Chinamen who advance towards our lines too rashly, are killed every
night, and after hours of this work the number of corpses that
accumulate is astounding. For the sake of the health of the
community, Mr. Cheshire has to spend much of his time
superintending his gangs in throwing dead bodies over the Wall, and
to-day he facetiously remarked he thought he should be dubbed
Major-General of the Corpses, as he comes in touch with so many.
Such gruesome tales as these do we hear and talk of daily!

June 26.
Yesterday afternoon, at four, five gorgeously costumed Imperial
Standard bearers appeared on the bridge in Legation Street with a
flag of truce, saying the Emperor would send later a despatch to the
bridge for us to read, and that there was in consequence an
armistice. It was brought later, and it read: “The Emperor desires the
Ministers to be protected. Therefore, firing must cease, and a
despatch will be handed to them later on the bridge.” It was
apparently not brought; but on seeing some mounted Chinese
officers belonging to Jung Lu’s regiment passing over the Imperial
bridge, we hailed them with a white flag, and with some soldiers to
back up the meaning of the flag we spoke to them long enough to
find that they were going the rounds of this part of the town, telling
their people not to shoot this night on us, as there was an armistice.
We told them to send the Emperor’s letter or despatch (which has
not yet arrived on the bridge) to the British Legation. They promised
that it should be brought to us, but it has not yet arrived at noon to-
day.
Last night I was talking to M. Pichon, the French Minister, when
the French Interpreter of Legation came up to us in great excitement,
saying the Russian officers had heard, without any possible doubt,
les sonneries du canon of the Russian troops. It is in this way we
hear so many tales that one is lost when one tries to think. The
captains of all nationalities have had a council of war, and they agree
that with great care and hard work we can hold our own for eight or
ten days longer, but after that we are lost.
 Mrs. Coltman, the mother of six lovely children, was speaking of
the impossibilities of clean linen or having any washing done. “But
after all,” she said, “what does it matter? If the troops come within
ten days, my children can wear what they are wearing; if Peking is
not relieved within that time, we will all be dead.” She was not
melodramatic, but spoke very quietly. A hundred other remarks of
this sort that one hears daily go to show how the people really feel
about our condition. Women with husbands and children suffer
horribly. They dread lest their children may die of disease or by
torture, as certainly would be the case if the Chinese get in—as they
are notoriously cruel and without mercy even to babies—and fear for
their husbands, who may be killed during any attack.
At one o’clock this morning a terrific firing began, apparently
coming from all sides at once, which proved to be the case later,
when the officers in charge of the defence compared notes. At this
Legation the air hummed with bullets, but the noise was so frightful
one could not tell if all the Legations were being attacked or just the
British. They tried to frighten us, and they certainly succeeded with
women, children, and some men, but, thank heavens, the officers in
charge of defending us and the sentries—most of them, at least—
know that our high walls and strong barricades are our safety, and
that, unless good and well-aimed artillery is brought to shell them
down, with our soldiers and soldier-sailors to man them, it will be
hard for the Chinese to get over the Wall and end our lives.
It all seems like a story from the Middle Ages to be able to place
such confidence in the strength and manning of our walls. Certainly
the foreign-drilled Chinese soldiers must be down at Tien-tsin, and
we are owing our present immunity from properly aimed artillery-fire
to the fact that the Chinese gunners here are utterly incompetent.
After this fiendish attack had been in progress long enough for
everyone to get up and dress, Mrs. Conger came back to our room,
and her manner was more than tragic when she saw me lying on my
mattress on the floor, not even beginning to dress for what I suppose
half of the women in the compound believed to be the beginning of
the final fight. She said: “Do you wish to be found undressed when
the end comes?” It flashed through my mind that it made very little
difference whether I was massacred in a pink silk dressing-gown,
that I had hanging over the back of a chair, or whether I was in a golf
skirt and shirt waist that I was in the habit of wearing during the day
hours of this charming picnic. So I told her that for some nights I had
dressed myself and sat on the edge of the mattress wishing I was
lying down again, only to be told, when daylight came, that the attack
was over, when it was invariably too late for anything like sleep
(which way of living is distinctly trying), and after a week of it, when
one has so much to do in the day hours, I had come to the
conclusion that, as it was absolutely of no benefit to anyone my
being dressed during these attacks, I was going to stay in bed unless
something terrible happened, when I should don my dressing-gown
and, with a pink bow of ribbon at my throat, await my massacre. This
way of looking, or I should rather say of speaking, did not appeal to
the Minister’s wife, but I must say that at such terrible moments
during the siege it is a great comfort to be frivolous. By making
believe that one is not afraid one really lessen one’s own fear.
“Assume a virtue if you have it not,” says our beloved Shakespeare.
After Mrs. Conger’s visit on this same terrible, ear-deafening night
came Clara, Mrs. Squiers’s German nursery governess, and she
needed all sorts of assurances to convince her that a massacre was
not in progress at that very moment.
These attacks are very terrifying, and to talk to a person two feet
away one has to shriek. People one sees are either apparently most
optimistic or desperately pessimistic, nothing between. It is a horrid
thing to see big, strong men unable to hide their innate cowardliness,
and shirking all duty of the slightest personal danger.

Friday, June 29.

One or two days have passed without my opening my diary, but
they are very much like the days that I have already written about.
The weather is very warm, but all able-bodied men are working
desperately hard over trenches, bomb-proofs and barricades, or
putting out more fires that have started at different places. Several
attacks are constantly being made in opposite parts of our lines, with
perhaps one big general attack in the afternoon and one during the
night, which causes great excitement and sometimes great fear in
everybody’s heart. Then comes more excitement when the rumour
arrives that a big fire is breaking out in the French and American
Legations, or we hear that someone, who has just come from one of
these Legations, says that he has heard the officer say they probably
cannot hold out much longer, with fire to fight as well as the Chinese,
and in ten minutes the report is all over the compound that these
Legations have been abandoned, and half of the soldiers defending
them killed. Such a quantity of rumours that are circulating every
day, only to be denied and proved untrue an hour later! It is
incredible!
In the case of the Legations who are still holding their own, it is
very hard on the women whose husbands are still staying with the
soldiers until they finally evacuate. These poor women naturally
wonder, “What is my husband doing? Is he dead, and when they
evacuate will he be amongst the lucky number to retire to the British
compound alive?”
Hard work is kept up on the important barricades, and men do
hours and hours of manual labour. The women make thousands of
sandbags daily, and help at the hospital, and make short rations go
as far and look as attractive as possible under the circumstances.
The only strong men in the compound who have no special work to
do are Ministers Plenipotentiary. There is no head-work to be done
now, and some of them don’t take kindly to physical work.
The British Legation Library is a complete one, and occasionally
some inquisitive soul will go to it and try to find, compared with other
sieges and massacres, what place this one will have in history. The
nearest similar harrowing siege seems to be that of Lucknow, where
a heterogeneous multitude, closed up in the Residency, were holding
out against fearful odds in expectation of relief by Havelock’s
Highlanders, resolved to die of starvation rather than surrender, for in
surrendering the fate of Cawnpore awaited them; and in thinking of
these things we recollect that the Tartar rulers of China are of the
same tribal family as the Great Mogul, who was the head of the
Indian Sepoy Mutiny. But the King of Delhi was trying to regain his
throne, whereas the Empress Dowager has no such excuse in
making war on practically all the nations of the civilized world.
The Tartars and Manchus are an alien race, although the rulers of
China for many centuries, and have always been inimical to
everything which tends to increase the power of foreigners; whereas
the Chinese are cleverer, from being so constantly in contact with
Europeans on the sea-coast, or anywhere where they can find gain
or advantage in trading with them, and have become, compared, at
least, to the ruling Manchus in Peking, progressive and modern. The
Emperor and his party for progress were completely snowed under
in 1898 by the Empress-Dowager and her old Manchu
Conservatives, who, lacking the desire to accept anything modern—
even diplomatic relations of the most simple kind—decided, in a
childlike and unreasoning rage, that everything foreign must be
swept down into the sea, and it really looks now as if the first steps
of her policy may be realized.
Lady Macdonald has forty Europeans in all to feed three times a
day including servants, and at table they sit down thirty-three. She is
very sensible, and has only one dish. Nobody thinks of dressing for
dinner, except the Marquis Salvago, and I think it shows things are
truly far gone when English people dine, but do not dress.
Our little mess is very attractive, and as our stores are much more
numerous and of a greater variety than those of almost any mess
here, we manage to have, up to the present at least, a most
satisfactory one. We have tinned beef as our pièce de résistance,
and rice is our mainstay—of a necessity, as it is that of which we
have most. Tomato catsup tastes very good in this hot weather.
Oatmeal is another staple that we have, and as luxuries we have a
good stock of jams, tinned fruits, tinned vegetables, sardines, tinned
mackerel, Liebig’s extract, a big box of Stilton cheese, coffee, tinned
butter, and white flour. Mr. Squiers has a large supply of champagne,
and every night we have one or two quarts with our siege dinner.
The men work so hard, and the women’s nerves are so much on
edge, that a small amount of stimulants is surely a blessed help.
 Our mess being comparatively small, these delicacies are lasting
nicely, as we use them with discretion, for we remember in the old
days before the siege a dollar was a dollar, and would buy a tin, but
in these days a tin has no market value—they cannot be bought.
When one’s tins are gone one can eat horse-meat and rice. We
brought a small lead-lined ice-box with us from our Legation, which
seemed foolish at the time, but which is a great comfort to us now.
We keep our wine and drinking-water in it, and also well-water, which
is very cool, so that our drink is somewhat cooled, and is not the
same temperature as the air. No other mess can attempt to have
things cool, and this is one of the features of our room—that we are
as comfortable as we can be under these extraordinary
circumstances. During this sizzling weather cool water is a great
comfort. It is so hot that a tin of meat, if left open all night, spoils by
morning.
There is an English newspaper-man, who, when he can spare a
few moments from the siege-work, gets his camera and takes a few
photographs of things as they are. He is fond of chaffing, and to-day
the Committee on Fortifications are of opinion that the house used
by the French Minister, M. Pichon, is being undermined by the
Chinese from outside, though indistinct noises, etc., are as yet the
only proof of it. The Minister was more than usually perturbed about
this new personal danger, and was not pleased, or at all amused, at
the remarks addressed him. “I am making photographs for the Paris
Figaro of this siege. Very soon your quarters will be blown up by
dynamite. My camera is ready to take the photographs, and as you
will be the principal person in it, how would you prefer me to take you
—as your Excellency is going up wholesale, or as you are coming
down retail?”
At this time people are not well-balanced, it seems to me. Some
take the daily horrors as a matter of course, are more callous than
they should be, and the others are so miserably pessimistic and
mournful that one shuns them, fearing to catch this infection. There
is a young man here who has been known to indulge in temporary
aberrations, usually at night, following long, hard days of work in the
broiling sun. On one occasion he was on his sentry beat, and on
being relieved by his chief, the sight of whom was too much for him
after having walked some hours on his dangerous sentry route
(which seemed doubly dangerous in the pitch-black night) he,
doubtless brooding over his probable approaching death, pointed the
muzzle of his gun straight at his relief. “C’est à cause de vous,
misérable, que je suis venu à Pékin et encore c’est à cause de vous
que ces belles années de ma jeunesse seront salement terminées
ici!” By not moving an inch the man thus threatened undoubtedly
saved his life, and most intelligently agreed with his attacker,
“Probably so; let’s talk it over.” In a few minutes the crisis had
passed, but the following day the man who had been in such danger
requested the General Committee to change his night sentry duty to
a different part of the compound, so that his young secretary should
not again be tempted to hold him responsible.

Sunday, July 1.
I have been quite under the weather, to use a civilized expression,
and I assure you that things have got (not are getting) to such a state
that to live and act and talk as one would do at home is quite out of
place. How soon people get accustomed to an idea! Now that we
have prepared our minds for a possible massacre we seem to be
getting back, to some degree at least, our old spirits. Now that I am
well, how much nearer seem the soldiers who are coming to relieve
us!
What a place this compound would be for an epidemic! There are
barely enough mattresses for the wounded and dying at the hospital,
so that, should we have one, and take a house for those taken sick, I
am sure that there would be no ordinary comforts of any kind for
them; they could only be isolated. Let us pray that we will have no
such horror to add to the already long list.
The hospital is already full, men lying on straw bags in halls—
crowded in every conceivable corner. They are brought in dying and
wounded every day. Dysentery has its grip on almost everybody
here. The treatment is almost to stop eating and to drink rice-water in
large quantities. Our four-times-divided cook—the other three
messes in the United States bungalow have a lien on him too—is off
for some hours daily on work which all personal servants have to
give to the General Committee. When the kitchen is comparatively
free, Mrs. Squiers, my maid, and I make gallons of rice-water, thick,
nutritious but tasteless, which we bottle in quart-bottles and place to
cool in our zinc-lined, cold-water-filled box. It is placed in a corner of
our two-roomed quarters, and the constant stream of men coming
and going to that box would lead an uninitiated observer to believe
that at least a Hoffman House bar was hidden there and doing a
steady business.
The rainy season and the bad time of the year par excellence has
begun, and the temperature is like a Turkish bath without the clean
smell. Apropos of smell, a whole story-book could be written about
the Peking smell. The dry heat was nothing compared with this damp
temperature, that seems to soak out of Mother Earth the most
incredibly disgusting odours. There are so many dead dogs, horses,
and Chinese lying in heaps all around the defended lines, but too far
for us to bury or burn them. The contamination of the air is
something almost overpowering. All men who smoke have a cigar in
their mouths from morning until night as a protection from this
unseen horror, and even the women, principally Italians and
Russians, find relief in the constant smoking of cigarettes.
On the 29th Dr. Lippitt, who came up from Taku with our marines,
was sitting in front of the Minister’s house smoking a cigarette, when
a bullet struck a limb of a tree nearby, and, glancing down, struck
him in the thigh, fracturing the bone. He is most dangerously ill, and
we shall not know for several days whether he will have to have his
leg amputated or not. He is an attractive man and a thorough
Virginian. We used to play tennis with him and Captain Myers before
the times got so terribly out of joint.
To-day the Germans were driven off the Tartar Wall close to their
Legation, which caused a great deal of excitement. They were driven
off by Chinese soldiers, some of whom were Tung Fu-hsiang’s men,
and others were Prince Ching’s especial troops, which seems queer,
as we have supposed all along that Prince Ching was friendly.
 The Germans could see from the Wall that the Ha Ta Men Gate is
being strengthened, and people who know say that the troops who
are closing the gates in such a warlike way are doing it as much
against the violent and uncontrolled soldiers of Tu Fu-hsiang, who
are notorious for the manner in which they loot and murder, as
against the allied Powers. They say that all Chinese families in
Peking who have anything to lose have left the capital, as they
realize that if the foreign troops come there will be great looting, and
if the Chinese troops are successful there will be looting and worse.
Mr. Pethick tells me that during the Japan-China War, when it was
considered highly probable that the Japanese would march on to the
capital, thousands of Mandarins and people of wealth left Peking
with their families and with as much treasure as they could carry. It is
natural to suppose that the same fright exists to-day.
This morning our men, the Germans following, retired in a panic
from their barricades on the Wall to the United States Legation,
momentarily expecting to see Chinese hordes occupy the German
position and theirs. After an hour’s wait they retook the Wall. This
example, however, was not followed by the Germans. During this
hour the excitement was intense in the British compound. The report
that the Wall had been evacuated caused a panic, for this
abandonment of the Wall would enable the Chinese to mount their
guns on this portion of it, directly commanding the British Legation,
and to fire down on us, and no one can say how long we could hold
out against such an attack. In such an event we will put women and
children into deep bomb-proofs that have been made for that
purpose, which are covered with logs, sandbags, and dirt, and are
shell-proof. These trenches we have made as near as possible like
those used in the siege of Ladysmith.
As the Germans have been unable to regain their positions on the
Wall, the difficulty for Uncle Sam’s men has been increased fifty per
cent., as they must now be prepared at all times, either during the
day or night, for an attack by Chinese from both directions. This
sentence, “to give up the Wall,” could be, translated into siege
language, “the beginning of the end,” and this news was most
terrifying to us. I think that there are few who in their heart of hearts
have given up hope of the troops coming soon. Nevertheless, the
facts remain that if we cannot hold the place it would not take very
long for us to be annihilated, and if the troops come a day after we
are finished, a miss is as good as a mile, and we don’t care then
when they come. If we had not had the greatest luck in the world we
could never have held out like this to the present date, and what the
Powers can be thinking about not to send a column to our immediate
relief, knowing, as they must, that we could never hold out against
artillery, is beyond the reasoning power of the people in this
Legation. Are the allied Powers fighting each other, or are they
fighting their way up here?
Yesterday an unsuccessful sortie was made by Colonel Shiba from
the Fu to capture a gun, and six men were killed. These offensive
measures seem to gain us nothing, and we always lose men.
Apropos of Colonel Shiba, he is a splendid, small person. He has
taken his position here by the strength of his intelligence and good
right arm, solely because the Ministers and the guard captains were
not especially inclined at the first morning conference to listen to him
—in fact, I don’t know that he tried to talk, but it is all changed now.
He has done so splendidly in his active and continuous fighting in the
Fu, and has proved himself such a general, that his opinion and help
are asked by all the commanders. His men are all so patient and
untiring in their long, long hours behind the barricades, and are so
game, in great contrast to the Italians who are with him defending
the Fu. One can only hope for Italy’s sake that her soldiers in Peking
are the worst she has.
Now that we have got down to the primitive motif of all nationalities
fighting for their lives, the racial friendships and animosities are very
obvious. The British and American are almost one people here;
although the expressions, “D—— Yankees!” and “D—— lime-
juicers!” are interchanged, they are used in a spirit of affection. The
dislike of the Russians for the British is so cordial that it is only
equalled by the feeling the British entertain toward them. The
frankness of this avowed enmity is delightful. Our compound joins
the Russians, and they love us and we love them in as strong a
fashion as they hate their English neighbours on their other side.
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebooknice.com