Body, Language and Mind Volume 2: Sociocultural Situatedness

Body, Language and Mind
Volume 2: Sociocultural Situatedness
≥
Cognitive Linguistics Research
35.2
Editors
Dirk Geeraerts
René Dirven
John R. Taylor
Honorary editor
Ronald W. Langacker
Mouton de Gruyter
Berlin · New York
Body, Language and Mind
Volume 2:
Sociocultural Situatedness
Edited by
Roslyn M. Frank
René Dirven
Tom Ziemke
Enrique Bernárdez
Mouton de Gruyter
Berlin · New York
Mouton de Gruyter (formerly Mouton, The Hague)
is a Division of Walter de Gruyter GmbH & Co. KG, Berlin
앪
앝 Printed on acid-free paper
which falls within
the guidelines of the ANSI
to ensure permanence and durability.
Library of Congress Cataloging-in-Publication Data
Body, language, and mind. Volume 1, Embodiment / edited by Tom

Ziemke, Jordan Zlatev, Roslyn M. Frank.
p. cm. ⫺ (Cognitive linguistics research ; 35.1)
Includes bibliographical references and index.
ISBN 978-3-11-019327-5 (hardcover : alk. paper)
1. Language and languages ⫺ Philosophy. 2. Mind and body.
3. Semiotics. I. Ziemke, T. (Tom), 1969⫺ II. Zlatev, Jordan.
III. Frank, Roslyn M.
P107.B63 2007
401⫺dc22
2007028708
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.
ISBN 978-3-11-019618-4
ISSN 1861-4132
쑔 Copyright 2008 by Walter de Gruyter GmbH & Co. KG, D-10785 Berlin
All rights reserved, including those of translation into foreign languages. No part of this book
may be reproduced or transmitted in any form or by any means, electronic or mechanical,
including photocopy, recording, or any information storage and retrieval system, without
permission in writing from the publisher.
Printed in Germany
Table of contents
List of contributors vii
Introduction: Sociocultural situatedness

Roslyn M. Frank 1
Section A: The dynamics of cultural categorization
An interview with Mark Johnson and Tim Rohrer:

From neurons to sociocultural situatedness
Roberta Pires de Oliveira and Robson de Souza Bittencourt 21
Beyond the body: Towards a full embodied semiosis

Patrizia Violi 53
Properties of cultural embodiment:

Lessons from the anthropology of the body
Michael Kimmel 77
Distributed, emergent cultural cognition,

conceptualisation and language
Farzad Sharifian 109
Collective cognition and individual activity:

Variation, language and culture
Enrique Bernárdez 137
Section B: The sociocultural situatedness of scientific discourse
Entangled biological, cultural and linguistic

origins of the war on invasive species
Brendon M. H. Larson 169
vi Table of contents
In search of development
Joseph Hilferty and Óscar Vilarroya 197
The language-organism-species analogy:

A complex adaptive systems approach
to shifting perspectives on “language”
Roslyn M. Frank 215
Section C: Sociocultural situatedness in lexical and usage-based

approaches to metaphor
Toward a socially situated, functionally embodied

lexical semantics: The case of (all) over
Kurt Queller 265
The embodiment of Europe: How do metaphors evolve?

Andreas Musolff 301
Sociocultural situatedness of terminology in

the life sciences: The history of splicing
Rita Temmerman 327
Section D: Exploring the sociocultural situatedness of language

and cognition
Discourse metaphors
Jörg Zinken, Iina Hellsten and Brigitte Nerlich 363
The relationship between metaphor, body and culture

Ning Yu 387
Idealized cultural models: The group as a variable

in the development of cognitive schemata
Gitte Kristiansen 409
Index 433
List of contributors
Enrique Bernárdez is Professor of English linguistics at the Complutense

University, Madrid (Spain). He studied German Philology at the same Uni-
versity and Dutch Linguistics at Groningen University, The Netherlands.
He specialises in the history of English and other Germanic languages, as
well as Old and Modern Icelandic. He has been working on Textlinguistics
and in a Cognitive Linguistics framework for many years. Among his most
significant books on linguistics are: Qué son las lenguas? published in
1999, and reprinted six times, new edition 2004; Teoría y epistemología
del texto, 1995; Introducción a la lingüística del texto, 1982, as well as
many scholarly papers published in collective books and journals in many
countries. A new book, provisionally titled El lenguaje como cultura, is
currently under preparation. He has also been active for years in literary
translation, especially from Old and Modern Icelandic.
e-mail: ebernard@filol.ucm.es
Robson de Souza Bittencourt is a PhD student in Linguistics at the Eng-

lish Graduate Program, Universidade Federal de Santa Catarina (UFSC),
Brazil. Using a framework drawn from Cognitive Linguistics, he is cur-
rently writing his dissertation on the role of metaphors in economic dis-
course, an analysis which will bring into focus ideological aspects of the
data.
e-mail: pfrobson@gmail.com
Roslyn M. Frank is Professor Emeritus in the Department of Spanish and

Portuguese at the University of Iowa. She is co-editor of Cognitive Models
in Language and Thought: Ideology, Metaphors and Meaning (2003);
Language and Ideology, Vol. 2. Cognitive Description Approaches (2001)
and has published extensively in the field of cognitive linguistics as well as
in ethnoscience, most particularly in ethnomathematics and ethnoastro-
nomy. Her research on the Basque language has taken her to Euskal Herria,
the Basque Country, where she has done extensive fieldwork and given
numerous seminars. In addition she has given presentations on these re-
search topics throughout Europe.
e-mail: roz-frank@uiowa.edu
viii List of contributors
Iina Hellsten is a Research Fellow at the Royal Netherlands Academy of

Arts and Sciences (KNAW) in the research group Virtual Knowledge Stu-
dio (www.virtualknowledegstudio.nl). Her background is in science com-
munication and the social studies of science and technology. Her current
research deals with the anatomy of scientific and public controversies on
the Web. Her areas of expertise include metaphor theory, science commu-
nication, public understanding of science, media and communication sci-
ences. She has published articles on the role of metaphors in public contro-
versies on science in Metaphor and Symbol, Science Communication,
Science as Culture and New Genetics and Society, for example.
e-mail: iina.hellsten@vks.knaw.nl
Joseph Hilferty, a San Francisco Peninsula native, graduated from San

Francisco State University in 1987. He became involved with the cognitive
linguistics movement in the early 1990s. In 2004, he obtained his Ph.D.
from the University of Barcelona with the thesis “In Defense of Grammati-
cal Constructions”. He is the coauthor of Introducción a la lingüística cog-
nitiva (Ariel, 1999) with Maria Josep Cuenca. Currently, he teaches Eng-
lish linguistics at the University of Barcelona.
e-mail: hilferty@ub.edu
Michael Kimmel (PhD Vienna University 2002, MA 1995) is a researcher

based at the University of Vienna, Austria. His interests span cognitive
linguistic methods of metaphor and image schema analysis, socio-cultural
embodiment, cognitive narratology, as well as qualitative and mixed meth-
ods. He has conducted research on metaphor interaction in political dis-
course and literature with software-based analytical tools, and has been
working psycholinguistically on sensorimotor resonance in reading and
plot comprehension. From 2007–2010 Kimmel will be the principal re-
searcher of two research projects, one being a text-linguistic approach to
imagery in literary cognition (2007–2008), and the other an ethnographic
fieldwork approach to embodied imagery in dance apprenticeship (2008–
2010). The latter project will combine cognitive linguistic and phenomenol-
ogical methods with motion analysis to explore dance class interactions
and embodied learning. In the past, Kimmel has been a member of the
Austrian Academy of Sciences, the Institute of Advanced Studies, Vienna
and the University of Economics, Vienna, as well as a freelance researcher
in Vienna, Budapest, and Berlin.
e-mail: michael.kimmel@gmx.at
List of contributors ix
Gitte Kristiansen is Assistant Professor in linguistics at the Department of

English Language and Linguistics, Universidad Complutense, Madrid
(Spain). Her main research interests include diachronic linguistics, cogni-
tive sociolinguistics and cognitive phonology. She has taught courses on
Historical Sociolinguistics, Cognitive Semantics, Linguistic Change in
Contemporary English, Registers and Varieties of English, amongst others,
and is currently co-editing a volume entitled Cognitive Sociolinguistics:
Language Variation, Cultural Models, Social Systems (Mouton) with René
Dirven.
e-mail: gkristia@filol.ucm.es
Brendon Larson received an M.Sc. in evolutionary ecology from the Uni-

versity of Toronto (Canada) in 1997 and an Interdisciplinary PhD in Sci-
ence and Society from the University of California, Santa Barbara (USA)
in 2004. His dissertation, entitled “The Metaphoric Web of Science and
Society: Case Studies from Evolutionary Biology and Invasion Biology”,
emphasized the social resonance of competitive and progressive metaphors
in evolutionary biology and militaristic ones in invasion biology. In 2005-
2006, he continued his research and teaching on the linguistic and social
dimensions of invasion biology as an interdisciplinary post-doctoral fellow
in the Center for Population Biology, University of California, Davis
(USA). He is now an Assistant Professor in the Department of Environ-
mental Studies, University of Waterloo (Canada).
e-mail: blarson@fes.uwaterloo.ca
Andreas Musolff is Professor of German Language at Durham University

(UK). He has published widely in the fields of metaphor analysis, the study
of public discourse in Britain and Germany, and on the history of func-
tional linguistics, including the monograph Metaphor and Political Dis-
course. Analogical Reasoning in Debates about Europe (2004). He is cur-
rently researching the history of corporeal metaphors in political thought
and discourse in Germany as well as popular conceptualisations of evolu-
tion in the 19th and 20th centuries.
e-mail: andreas.musolff@durham.ac.uk
Brigitte Nerlich is a Principal Research Officer at the Institute for the

Study of Genetics, Biorisks and Society (IGBiS) at the University of Not-
tingham (UK). She has published numerous books and articles on the his-
x List of contributors
tory of semantics and pragmatics, cognitive semantics, figurative language,

polysemy and semantic change. She currently studies the uses of meta-
phorical models in the discourses about cloning, designer babies, GM food,
stem cells and genomics. She has recently concluded a project on the social
and cultural impact of foot and mouth disease in the UK and will shortly
start working on a new project “Talking cleanliness in health and agricul-
ture” which deals with MRSA and avian flu from a sociological and ap-
plied linguistics perspective. Like the foot and mouth one, this project is
funded by the Economic and Social Research Council.
e-mail: brigitte.nerlich@nottingham.ac.uk
Roberta Pires de Oliveira is a professor of semantics at Universidade

Federal de Santa Catarina (UFSC), Brazil, and a researcher at the Conselho
Nacional de Pesquisa (CNPq). Her MA dissertation and PhD thesis deal
with metaphor from the perspective of Max Black’s Interactionism, and the
more recent approach of cognitive linguistics. Since the beginnings of the
90’s, however, her research has moved away from metaphor and the cog-
nitive paradigm towards formal approaches to the semantics of natural
language. She has published an introduction to semantics, papers analyzing
several aspects of Brazilian Portuguese (in particular, the semantics and
pragmatics of quantification, free choiceness, tense and aspect), and also
articles on the epistemology of linguistics (comparisons between the for-
mal and the cognitive paradigms in linguistics).
e-mail: pires@cce.ufsc.br
Kurt Queller (PhD, Stanford, 1994) teaches linguistics and languages

(German, Italian, Spanish, Mandarin) at the University of Idaho (USA) and
elsewhere. His primary research field is cognitive semantics. Grounded in a
usage-based re-analysis of the English over network, his current work ar-
gues that most significant polysemy originates non-teleologically, resulting
not from intentional speaker innovation, but rather (as in Croft’s 2000
model of change) from hearers’ abductive inferences about the contextual
meanings of usage events. (See Queller’s contribution to the present vol-
ume, and references cited therein). The overall argument is presented in
Polysemy: A Usage-based Approach (in preparation). Other work includes
historical analysis of gendered language, e.g. “‘Whether man or woman’:
List of contributors xi
Gender-inclusivity in the town ordinances of medieval Douai”, with Ellen

E. Kittell, Journal of Medieval and Early Modern Studies 30 (2000): 63–
100.
e-mail: kqueller@uidaho.edu
Farzad Sharifian, PhD, is a Senior Lecturer in the School of Languages,

Cultures, and Linguistics at Monash University, Melbourne, Australia. He
has carried out research in the areas of psycholinguistics, language and
memory, cognitive anthropology, cognitive linguistics, cultural linguistics
and applied linguistics. He has published widely in international journals
such as Journal of Cognition and Culture, Journal of Psycholinguistic
Research, Anthropological Linguistics, Discourse Studies, Pragmatics and
Cognition, World Englishes, Language and Intercultural Communication,
Journal of Multicultural and Multilingual Development, and Language and
Education. He is the editor (with Gary B. Palmer) of a volume of essays on
Applied Cultural Linguistics (John Benjamins, 2007).
e-mail: farzad.sharifian@arts.monash.edu.au
Rita Temmerman is the coordinator of Centrum voor Vaktaal en Com-

municatie (CVC) (Centre for Special Language and Communication) at
Erasmushogeschool Brussels and teaches translation, terminology and
knowledge representation. She obtained her degree in Germanic Philology
from The University of Antwerp (Belgium), her Masters in Translation
from the State University of New York (USA) and her PhD in Linguistics
from the University of Leuven (Belgium). Based on case studies on catego-
risation and naming in the life sciences (DNA technology) she developed
the sociocognitive terminology theory. In 2000, she published Toward New
Ways of Terminology Description. The Sociocognitive Approach (John
Benjamins). Her latest research interest concerns the impact of changing
perspectives on knowledge representation. She has been involved in sev-
eral projects concerning the development of a methodology and software
for the creation of ontologically underpinned multilingual terminological
resources.
e-mail: rita.temmerman@ehb.be
Óscar Vilarroya earned his first degree in Medicine (1987) and his PhD in
Cognitive Science (1998). He currently heads a neuroimaging team in Bar-
celona and teaches Brain and Consciousness at the Universitat Pompeu
xii List of contributors
Fabra. He is the author of The Dissolution of Mind (Rodopi, 2002). Some

of his most recent articles are: Carmona et al. (2005) “Global and regional
gray matter reductions in ADHD: A voxel-based morphometric study”,
Neuroscience Letters; Vilarroya (2005) “A Categorial Mutation”, Behav-
ioral and Brain Sciences 28, 508–509; and Vilarroya (2005) “In Search of
Radical Similarity”, Behavioral and Brain Sciences 28: 35–35.
e-mail: 24678ovo@comb.es
Patrizia Violi is Professor of Semiotics at the University of Bologna and

Director of the International Center for Cognitive and Semiotic Studies at
the University of San Marino. She is currently Coordinator of the Ad-
vanced PhD program in Semiotics, run by the University of Bologna and
the Higher Institute of Human Sciences. She has published numerous
books and articles on theoretical and applied semiotics. In particular she
has worked on semantic theory (Meaning and Experience, Indiana Univer-
sity Press, 2001); text and discourse analysis; the relationship between
Semiotics and Cognitive Science; language and gender; semiotics, psycho-
analysis and dreaming. At the present she is engaged in an interdisciplinary
research project on early acquisition of semiotic competence in preverbal
children, and the role of embodiment in meaning development. Other cur-
rent research is on the semiotic construction of space in the Mediterranean
area; and on women’s identity in conflict and post-conflict situations (Bal-
kans and Palestine).
e-mail: violi@dsc.unibo.it
Ning Yu is an Associate Professor at the University of Oklahoma (USA).

His publications include the book The Contemporary Theory of Metaphor:
A Perspective from Chinese (John Benjamins, 1998) and numerous articles
in Cognitive Linguistics, Journal of Pragmatics, Pragmatics and Cogni-
tion, Journal of Literary Semantics, and Metaphor and Symbol. He is inter-
ested in embodied cognition and how it is manifested in language. His
research, focused on the relationship between language, culture, body and
cognition, attempts to reveal, via systematic study of language, how bodily
experiences contribute to human meaning, understanding and reasoning in
cultural contexts.
e-mail: ningyu@ou.edu
List of contributors xiii
Tom Ziemke is Professor of Cognitive Science in the School of Humani-

ties and Informatics at the University of Skövde, Sweden. His research is
mainly concerned with embodied and distributed cognition, i.e. theories
and models of how cognition is shaped by the living body and its interac-
tion with the material and social environment. He is coordinator of a large-
scale European project on robotic models of embodied cognition, called
Integrating Cognition, Emotion and Autonomy (www.his.se/icea), and a
member of the executive committee of euCognition – The European Net-
work for the Advancement of Artificial Cognitive Systems. He is also asso-
ciate editor of the journals New Ideas in Psychology and Connection Sci-
ence.
e-mail: tom.ziemke@his.se
Jörg Zinken is a Senior Lecturer in the Department of Psychology at the

University of Portsmouth, UK. His main research interests are in semantic
universality and diversity, discourse processes, relations between subjec-
tive state and verbal expression, and theoretical issues in cognitive linguis-
tics. His current projects include the spatial conceptualisation of time in the
Amazonian language Amondawa.
e-mail: joerg.zinken@port.ac.uk
Introduction: Sociocultural situatedness
Roslyn M. Frank
1. Background
This work constitutes the second volume of a two-volume set with the title
Body, Language and Mind. While the first volume focuses on the concept
of embodiment, i.e. the bodily and sensorimotor basis of phenomena such
as meaning, mind, cognition and language, the second volume addresses
sociocultural situatedness, i.e. the ways in which individual minds and
cognitive processes are shaped by their interaction with sociocultural
structures and practices. Naturally, the domain covered by the two volumes
overlaps significantly. In fact both of them have their genesis in a one-day
theme session entitled “Situated Embodiment: The Social and Biological
Grounding of Metaphorical and Symbolic Thought on ‘Embodiment’”,
organized for the 8th International Cognitive Linguistics Conference, held
July 20–25, 2003, at the University of La Rioja, Spain. Many of the con-
tributors to this volume participated in the original theme session, while,
subsequently, several additional authors were invited to take part in the
project in order to further expand the range of perspectives represented.
2. Sociocultural situatedness
Whereas Volume 1 concentrates on the concept of embodiment, understood

as “the material or bodily basis for mind, meaning and cognition”, Volume
2 explores the concept of sociocultural situatedness. Briefly stated, so-
ciocultural situatedness denotes the way(s) in which individual minds and
cognitive processes are shaped by their being together with other embodied
minds, i.e., their interaction with social and cultural structures, such as
other agents, artifacts, conventions, etc. and, more particularly, for the
purposes of this book, with language itself, as its central position in the
book title Body, Language and Mind suggests. In this sense the approaches
discussed in Volume 1 and Volume 2 might be compared to two sides of a
coin since their perspectives are complementary. The difference is that
2 Roslyn M. Frank
Volume 1 looks more at the “bodily” aspects of mind, perception and cog-
nition, while Volume 2 concentrates more on exploring the “social” side of
cognition and language.
This second volume offers a representative collection of new papers on
sociocultural situatedness and displays a variety of perspectives with re-
spect to the way that language can be understood to be socioculturally situ-
ated. While the concepts of embodiment and sociocultural situatedness are
closely linked, they are still evolving: they have overlapping areas of con-
sensus as well as aspects that are still being debated and elaborated by
researchers. In order to address this evolving set of perspectives, the pres-
ent volume brings together the work of well recognized authorities in the
field along with significant contributions by younger scholars, all of whom
are currently working in the field of Cognitive Linguistics and/or closely
related disciplines.
Seen from this wider interdisciplinary perspective, Volume 2 is a cog-
nitive linguistic contribution to the current theoretical and empirical re-
search being conducted in relationship to the concept of embodiment, so-
ciocultural situatedness and situated cognition. In four main sections, the
papers explore various dimensions of these notions as they apply to cogni-
tion and language such as: a) cultural categorization; b) scientific dis-
course; c) lexical usage-based approaches to metaphor; and d) the interac-
tion of culture and cognition.
In the past, different aspects of the notion of sociocultural situatedness
have been addressed (Dirven, Frank and Ilie 2001; Dirven, Frank and Pütz
2003; Ziemke 2001, 2002), however, without systematically exploring the
concept within a wider theoretical framework, e.g. accessing metaphor not
in the mind but in the cultural world (Gibbs 1999); the sociocultural role of
metaphors, frames and narratives (Nerlich, Hamilton and Rowe 2002); the
relation between culture and the embodiment of spatial cognition (Sinha
and Jensen de López 2000); and the cultural context needed to understand
the history of Dutch causal verbs (Verhagen 2000).
Certainly the debate over the role of culture in language is not a recent
phenomenon but rather one with a long and complex history (cf. Döring
and Nerlich (2005) for a recent review of the literature as well as the ear-
lier discussions by Geeraerts (1988, 2002) and Jäkel (1999)). This much
earlier debate about culture and language began heating up again in the
1980s when the theoretical framework was beginning to be reoriented to-
wards a more situated view of language. Over the past decade this reorien-
tation has been particularly evident in numerous investigations aimed at
Introduction: Sociocultural situatedness 3
exploring the complex relationship and influence of language on culture,

cognition and conceptualization, and vice-versa, e.g., the rekindling of the
Sapir-Whorf debate in cognitive linguistics and cognitive anthropology
(Lee 1996; Lucy 1992, 1996 a, 1996 b; Palmer 1996; 2006 a).
Indeed, as Palmer (2006 a: 265) has observed, we need to keep in mind
that “[t]he importance of the culture concept to the common enterprise of
cognitive linguistics and cognitive anthropology has been explicitly recog-
nized at least since 1987 (Lakoff and Kövecses 1987), and since empha-
sized by Geeraerts and Grondelaers (1995), Langacker (1994), D’Andrade
(1995) and Palmer (1996)”. And while an increasing number of studies
have started to focus on the role of cultural schemata (Palmer 2006b;
Sharifian and Palmer 2007), the goal of the present volume of essays has
been to explore the concepts of socioculturally situatedness and situated
cognition, specifically with respect to the broader implications of these
concepts and their application to language. In particular the volume dem-
onstrates the diverse ways in which these relatively new insights into the
unity of body, language and mind can be brought to play with respect to
investigations of language, culture and cognition.
3. Historical overview
The historical roots of the concept that we refer to as sociocultural situat-

edness can be traced back to at least the first half of the 20th century where
it played, although not under that name, a major role in the theoretical
work of various researchers in philosophy and psychology, who laid the
groundwork for the phenomenological turn in the history of scientific
thought. The philosophical foundation was laid in Heidegger’s Being and
Time (1927 [1962]) and Merleau–Ponty’s Phenomenology of Perception
(1945 [1962]), while the psychological foundation was set forth by Vygot-
sky (1930 [1978]). Within Cognitive Linguistics, some twenty years ago
Johnson (1987) and Lakoff (1987) introduced the notion of “image
schema” as the bodily basis of language and metaphorical reasoning. In
their book Philosophy in the Flesh (1999) Lakoff and Johnson (1999)
elaborated further upon the notion of embodiment as isolated from any
contextual situation.1 This perspective on embodiment was brought into
1. For a more detailed discussion of the evolution of the concept of “image sche-
mas” and “embodiment”, cf. Kimmel (2005, this volume); Johnson and Rohrer
4 Roslyn M. Frank
question by Zlatev (1997) and criticized for portraying the “embodied

mind” as separated from situational contexts, as if it were an entity floating
in the air. Countering what he perceived to be the drawbacks inherent in a
decontextualized approach to embodiment, in his 1997 work, Zlatev intro-
duced the notion of “situated embodiment” by bringing forward a different
theoretical framework
based on the idea that our mastery and use of language is crucially depend-
ent on the fact that we are beings which are embodied as well as situated
within a culture of shared practices. […] [T]he key to understanding the na-
ture of linguistic competence and its acquisition […] lies in the dialectical
relationship between bodily dispositions and activities on the one hand, and
sociocultural practices on the other. (Zlatev 1997: 1–2)
Following Lindblom and Ziemke (2002, 2007), the term “social situated-
ness” denotes the way(s) in which individual minds (cognitive processes)
are shaped by their interaction with social and cultural structures, including
language itself. Also, the notion of sociocultural situatedness dovetails
with Langacker’s description of language as “an essential instrument and
component of culture, whose reflection in linguistic structures is pervasive
and quite significant” (Langacker 1999: 16).
While Zlatev’s term “situated embodiment” served to focus attention on
the situated nature of language, the origins of this more culturally informed
debate along with the notions of “cultural models”, “schemas” and “cate-
gorizations” go back to the late 1980s. For example, discussions of the role
of culture in language have been a central component in several somewhat
more narrowly focused debates concerning metaphor, e.g., on the “hu-
moral” and hence “cultural” background of Lakoffian “anger” metaphors.
In this literature, too, the culturally situated view has consistently gained
ground. For example, while the discussion involving Geeraerts and Gron-
delaers versus Kövecses started in the first half of the 1990s (cf. Geeraerts
and Grondelaers 1995; Lakoff and Kövecses 1987) it has continued to
gather steam, so to speak, with the growing recognition on the part of re-
searchers, such as Kövecses and others, of the role of culture in language,
cognition, and conceptualization (cf. Kövecses 2005; Palmer 2006a; Shari-
fian et al. in press).
Until quite recently, the various fields of cognitive science, psychology,
phenomenology, semiotics and linguistics, where the concepts of situated
(2007); Rohrer (2005, 2007); Pires and Souza Bittencourt (this volume); Violi
(2004, this volume); Zlatev (2005, 2007).
embodiment, situated cognition and sociocultural situatedness have been

developed, were relatively isolated from each other. Moreover, until quite
recently, while discussions concerning the relationship between the indi-
vidual mind (agent) and its sociocultural environment were taking place in
one subfield, e.g., artificial intelligence (AI), the discussants themselves
were often unaware of the fact that similar discussions were being under-
taken in a different, but theoretically allied, discipline, namely, in Cogni-
tive Linguistics. Similarly, linguists have been relatively unaware of the
significance of their own research to work being done in the area of situ-
ated cognition and situated embodiment in the fields of AI, cognitive psy-
chology, developmental psychology and cognitive anthropology. This two-
volume set of essays attempts to break down these disciplinary walls and
initiate a cross-disciplinary dialogue concerning notions of embodiment
and sociocultural situateness.
At this juncture in the development of the field of Cognitive Linguis-
tics, we believe it is important to open things up by bringing into focus
these meta-theoretical concerns, that is, the way that Cognitive Linguistics
is redefining the field of linguistics and constituting the “limits” or
“boundaries” of linguistics as a discipline. Those working in Cognitive
Linguistics regularly traverse the borders of different disciplines and sci-
entific methods and therefore have supported the development of interdis-
ciplinary approaches and views. Yet, CL researchers may not fully recog-
nize the strategic importance of their field, the key location that their field
occupies within the expanding network of disciplines composing Cognitive
Science well as the field’s interdependence on ideological, scientific and
social trends which structure research in these adjoining fields (cf. Robbins
and Aydede in press). As an emerging field of research, Cognitive Lin-
guistics is not just growing out of itself, but is interwoven with several
other areas and disciplines.
In this sense, the research initiatives and theoretical frameworks dis-
cussed in Volume 2 form part of and contribute to wider discussions of
situated cognition that have been taking place in Cognitive Science for
some time. As Clark stated, nearly ten years ago, “[t]alk of embodiment
and situatedness has become increasingly frequent in philosophy, psychol-
ogy, neuroscience, robotics, education, cognitive anthropology, linguistics,
and in dynamical systems approaches to behavior and thought. There is
clearly a shift in thinking but the nature and importance of the shift is sur-
prisingly hard to pin down” (1999: 345). The incipient situated cognition
movement in Cognitive Linguistics, like the cognitive sciences themselves,
6 Roslyn M. Frank
represents a loose-knit family of approaches to understanding the mind,

perception and cognition. And in consonance with the larger debates cur-
rently taking place with respect to the concept of situated cognition and the
extended mind thesis in the cognitive sciences (cf. Clark 1997; Clark and
Chalmers 1998), the chapters in this volume bring together diverse ap-
proaches and explore a variety of theoretical perspectives. Nonetheless,
they share a common framework in that they are characterized by a con-
scious move away from individualistic and essentialist views of “lan-
guage”; the focus on the individual (ideal) speaker has given way to ap-
proaches that emphasize the role of collective, group and distributed
cognition and, hence, the sociocultural embeddedness of “language”, hu-
man subjectivity and intersubjectivity. To paraphrase Clark and Chalmers
(1998: 7) the contributions ask: “Where does language stop and the rest of
the world begin?2”
The question invites two standard replies. Some accept the boundaries of
skin and skull, and say that what is outside the body is outside the mind.
Others are impressed by arguments suggesting that the meaning of our
words ‘just ain't in our head’ […]. (Clark and Chalmers 1998: 7)
In this way the contributions to this volume reflect the increased attention
that is being paid to the sociocultural embeddedness of cognition and lan-
guage in general. More specifically, they elaborate upon the growing con-
cern with developing, articulating and applying research frameworks that
are in consonance with situated approaches. Additionally, the contributions
demonstrate how a situated perspective on language that moves away from
individualistic approaches to emphasize the role of collective, distributed
and group cognition can lead to substantive insights into broader questions
of concern to cognitive linguists. Most particularly, while operating within
the broad framework of situated cognition, the research opens up questions
of methodology and practice, the kind of questions that any scientific
community has to ask itself, e.g. concerning what defines and sustains the
field as such, as well as the need for shared conceptions and agreed-upon
terminology (Ziemke 2003).
Finally, we hope that this volume will contribute to a kind of increased
creativity within Cognitive Linguistics, a synergetic energy flow that is
important for the development of all new disciplines. In summary, the pre-
2. The question Clark and Chalmers (1998: 7) asked at the beginning of their
seminal article on the extended mind was: “Where does the mind stop and the
rest of the world begin?”
sent volume brings together valuable contributions from a cross-

disciplinary perspective on a new and vitally important concept in Cogni-
tive Science and related disciplines: sociocultural situatedness.
4. Overview of the various sociocultural perspectives in this

volume
In the present volume, the concept of sociocultural situatedness is ap-

proached from four perspectives. Contributions making up the first section
focus broadly on the dynamic nature of cultural categorization, providing a
variety of theoretical viewpoints. Of particular note is the call for greater
cross-disciplinary research approaches as well as greater attention to the
“collective” and “distributed” nature of cognition, language and knowl-
edge. In the second section the authors undertake an examination of the
socioculturally situated nature of scientific discourse, bringing into view
three case studies where the interface between cultural domains and lan-
guage is particularly salient. In the third section, the contributors allow us
to appreciate the application of the notion of sociocultural situatedness as a
tool of analysis in lexical and usage-based approaches to metaphor. The
papers in the fourth section investigate the role played by culture-specific
knowledge in discourse, exploring the situatedness of metaphor, the inter-
relationship between bodily grounded experience and cultural models as
well as the contributions of group-specific cultural and social mechanisms
in the development of cognitive schemata.
4.1. The dynamics of cultural categorization
The volume opens with an examination of the dynamics of cultural catego-

rizations, specifically, taking the position that the notions of categories and
categorization are in constant evolution since both bodily-based and so-
cioculturally-based experiences are operative as sources of categorization.
The chapters in this section also provide the reader with a broad historical
overview of the development and application of the concepts of embodi-
ment and situatedness to human cognition and language, along with an
examination of some of the complexities governing current debates on
situated cognition itself.
8 Roslyn M. Frank
The contribution of Roberta Pires de Oliveira and Robson de Souza

Bittencourt offers a critical survey of the various trends in “embodiment”
research by interviewing Tim Rohrer and Mark Johnson about Lakoff’s
and Lakoff and Johnson’s conception of embodiment, its interpretation in
various other sciences ranging from philosophy to AI and biology, and its
place and relevance in Lakoff’s new research orientation on the Neural
Theory of Language (NTL). In the interview human embodiment is por-
trayed not as “the fleshy boundary of the skin”, rather our body and brain
are viewed as extending out to the world beyond us, engaging in all sorts of
bodily and socio-cultural interactions, in experiences of meaning which are
not objectively out there.
Patrizia Violi begins her analysis by setting forth a wider semiotic
viewpoint. On the basis of the work of Eleanor Rosch (1999), she first
radically revises the traditional view of categorization as static, fixed enti-
ties and re-defines categories as flexible instruments that are actualized
according to the changing discourse context. Whereas many categories are
image-schematically based, and hence necessarily bodily-based, many
other categories are not bodily-based at all, but arise from and in social
interaction between care-takers and children, a process which implies the
transfer of cultural knowledge.
In his paper Michael Kimmel focuses on the very basis of embodiment,
image schemas and related problem areas, offering a sociocultural per-
spective on these concepts. He confronts the CL use and understanding of
the notion of image schemas, especially by Lakoff and Johnson, with
viewpoints drawn from other disciplines such as cognitive and phenome-
nological anthropology and developmental psychology. If CL is to handle a
more up-to-date understanding of the nature of image schemas, it must take
into account how this notion has been further developed in interdiscipli-
nary research, the latter being a topic reviewed at some length by Kimmel.
Farzad Sharifian concentrates on the other side of the coin, i.e. so-
ciocultural situatedness, more specifically on cultural categorization and
cultural models. Unlike bodily-based categories and their semantic exten-
sions through metaphor and metonymy into more abstract domains, cul-
ture-based categories and models grow through social interaction and are
not necessarily represented equally in the minds of each and every member
of the community in question. Rather they are socially distributed to differ-
ent degrees among the various members of a cultural community. While it
is the cultural community as a whole that possesses this cultural knowl-
edge, it is not necessary for every single member of the community to

know all of the cultural categories existing in it.
In his chapter Enrique Bernárdez focuses on the question of how to
bridge the gap between an individual’s cognitive system and, consequently,
language, and linguistic diversity, namely, the problem of the relationship
between the individual and the collectivity. He introduces the concept of
synergic cognition, related to the study of similar problems in biology and
complex systems theory, and emphasizes the importance of understanding
language as “collective” knowledge This type of knowledge is a collective
possession, even though unequally distributed over individual members.
The patterns of individual activity may, or even must, show variation be-
cause of this unevenly distributed knowledge. Yet, ultimately, collective
cognition and culture are the main determinants of each single individual’s
activities.
4.2. Sociocultural situatedness of scientific discourse
The themes of Section 2 are grounded in the very strong traditional ten-
dency of scientific thought and discourse to appropriate categories of a
totally different branch of science for its own heuristic purposes, self-
understanding and self-definition. The fact that perhaps two of the most
notorious cases of such scientific heuristic exchanges are biology and lin-
guistics cannot be a coincidence. Whereas (evolutionary) biology often
tries to understand its field of research in terms of a given, hidden code,
referred to metaphorically as “the book of nature” or more generally as
“the language of nature”, linguistics has often approached its own object of
research in terms of biological categories such as “language as an organ-
ism” without necessarily analyzing the socioculturally entrenched meta-
phorical processes in question. In short, the projection of linguistic catego-
ries onto biological categories, as well as the projection of biological
categories onto linguistic and non-linguistic categories as instantiations of
cultural situatedness is explored in this section.
In his paper Brendon Larson picks up the cultural categorization of
ecologically non-default species as “invasive species”, projecting the
metaphor of human army invasions onto the biological world of ecosys-
tems, thereby loading the newly incoming species with all the negative
associations of human invasions. This contribution reflects the increasing
sensitivity on the part of biologists and others to the role of metaphor in the
10 Roslyn M. Frank
discourse of science, most particularly, the uncritical use of the metapho-

rics of invasion, competition and warfare. Larson argues that these textual
megametaphors promote an antagonistic stance toward the natural world
and, in fact, may be counterproductive to the very goals espoused by con-
servation biologists.
The contribution by Joseph Hilferty and Óscar Villarroya concen-
trates not so much on biology itself, but on a linguistic theory (TGG) that
has proclaimed itself as strongly biologically grounded and oriented. They
analyze the various reductionist metaphors and metonymies adopted by the
Chomskyan paradigm. These metaphors of “development” instantiate the
same logic that is found in biological pre-transformationalist thought. By
adducing a series of concrete examples, e.g. the case of the KE family and
the FOXP2 gene, the authors show that such nativist arguments are either
misleading or are based on a misunderstanding of what genes do. The
authors conclude that nativist hypotheses of language acquisition and de-
velopment are unhelpful because they invite inferences that are not sup-
ported by the results of current molecular biology and genetics, e.g. that
genes do not code (in the sense of information theory) for phenotypical
traits.
The chapter by Roslyn Frank addresses another aspect of the interac-
tive and dynamic role of sociocultural situatedness by bringing forward a
new conceptual frame of analysis, one that emphasizes the importance of
reflexivity when examining the way that “language” itself has been
“imagined”, and its metaphoric instantiations. Using an approach informed
by complex adaptive systems thinking, she introduces the concept of dis-
course metaphor formations and then moves on to examine the historically
conditioned evolution of the “language-organism-species” metaphor. The
contribution highlights the continuously shifting definitions of “language”
over time, the sociocultural situatedness of discourse metaphor formations
and the fact that the way that we view this entity called “language” has
been influenced, repeatedly, by extensive discursive interactions between
the fields of biology and linguistics.
4.3. Sociocultural situatedness in lexical and usage-based approaches to

metaphor
The contributions composing Section 3 deal with sociocultural situatedness

in lexical and usage-based approaches to metaphor. Sociocultural situated-
ness and particularly actual usage as the source of people’s automatic

functioning in interaction provide the basis for social and cultural patterns
and their conceptualizations. Another aspect of this type of situated cogni-
tion is the ease with which we make cross-domain mappings in metaphori-
cal thought, whereby domains are interpreted as largely culturally defined
units of experience.
Concentrating fully on the link between embodiment and cultural situ-
atedness in language, Kurt Queller maintains that certain issues in lexical
semantic analysis can be resolved only by attending to the notion of “func-
tional embodiment”, i.e. the grounding of lexical meaning in usage and
discourse. Attention is paid especially to the lexical entrenchment of par-
ticular idiomatic, collocational and constructional routines in which a word
figures. These constitute an indispensable rather than an ancillary part of
lexical semantic description. Concretely, he proposes extending the Lakoff
(1987) notion of “functional embodiment” to comprise not only lexical
entrenchment of particular usages, but also the embedding of usage events
within recurring sorts of communicative context.
In his contribution Andreas Musolff looks at the ways the concept of
“cultural evolution”, developed in “naturalistic” approaches to cultural
studies, can be applied to metaphor with reference to mappings from the
source domain of the HUMAN BODY to the target domain of POLITICAL
ENTITIES. The study focuses on micro-historical changes in the use of the
HEART-OF-EUROPE concept in British and German Euro-debates during the
1990s, as documented in a special corpus drawn from large general cor-
pora. These historical data are applied to models of concept evolution,
concretely in order to propose a perspective on metaphor development as
an adaptation to argumentative trends in the respective discourse commu-
nities.
In her chapter Rita Temmerman begins by discussing the importance
of diachronic approaches to the study of scientific discourse and then goes
on to develop a detailed analysis of the term splicing. Specifically she con-
centrates the extension of reference of the English lexeme splicing from its
origins in Dutch to its current use in biotechnology. Focusing on the exten-
sions of its frames of reference as the term becomes part of the specialized
terminology of genetics, she argues that a scientific discipline can be un-
derstood as an interpersonal intelligent system; that it is also possible to
study the sociocultural embeddedness of language and thought through an
analysis of metaphor and analogy in scientific terminology.
12 Roslyn M. Frank
4.4. Exploring the sociocultural situatedness of culture and cognition
As two sides of the same coin, (situated) embodiment and (sociocultural)

situatedness are not in competition with one another. Rather they must be
seen as different but integrated sources of conceptualization. This Janus-
like aspect of language is explored in the final section of the volume, dedi-
cated to investigating the sociocultural situatedness of culture and cogni-
tion. More concretely, the contributions making up this section offer three
different perspectives on the role of culture-specific knowledge in dis-
course and in the process they bring into view new innovative conceptual
tools: (1) discourse metaphor, (2) the “Triangle Model” of metaphor, body
and culture, and (3) the group as a variable in the development of cognitive
schemata.
In their contribution Jörg Zinken, Iina Helsten and Brigitte Nerlich
investigate the relationship between conceptual bodily-based metaphor and
cultural metaphor. They examine work based on one of the cognitive meta-
phor theories, i.e. Conceptual Metaphor Theory (e.g. Lakoff and Johnson
1999) in a variety of fields, e.g. psychology, archaeology, anthropology,
robotics and communication studies, highlighting the successes and draw-
backs of this theory, particularly the neglect of social and cultural aspects
of cognitive activity in the theoretical modeling of metaphors and metaphor
use. They then put forward an integrated model of interaction between
universal, bodily-grounded knowledge and culture-specific knowledge in
discourse. In the process they introduce the concept of discourse metaphor.
Approaching the relationship of cognition and culture in the sense of
Lakoff and Johnson (1999), Ning Yu argues that many conceptual meta-
phors are universal, because they are based on universal bodily experi-
ences. At the same time he highlights many culturally-specific linguistic
realizations of conceptual metaphors. Moreover, Yu suggests that meta-
phor, body and culture may form a “circular triangle relationship” (the
“Triangle Model”). While conceptual metaphors are usually grounded in
bodily experiences, cultural models filter bodily experiences for specific
target domains of conceptual metaphors. Thus, cultural models themselves
are very often structured by conceptual metaphors. In this way Yu’s contri-
bution offers an exciting compromise between bodily and cultural situated-
ness.
Gitte Kristiansen’s paper offers an insightful set of linguistic reflec-
tions on cultural situatedness. She no longer associates cognitive (image)
schemata with the individual as in a narrow embodiment approach, but
rather introduces the social approach as one of the various category-

creating forces. The social group creates such cultural categories of self-
group identification and “otherness” identification, thereby specifying and
categorizing “otherness” as national, regional, social-class or occupational
otherness. Thus, social cognition, social categorization and successful so-
cial functioning constitute the main focus of this contribution. In sum,
Kristiansen asserts that the many aspects of sociocultural situatedness are
so massively represented in human experience that they necessitate and
command the attention of present and future research.
5. Conclusion
In summary, at this particular juncture in time it is far too early to bring to

closure the wide range of theoretical perspectives, approaches and applica-
tions of the concepts of sociocultural situatedness and situated embodi-
ment, some of which have been laid out in this volume. Rather, heuristi-
cally speaking, we are still at an exploratory stage in the search for a
theoretically unified framework that would adequately embrace the notion
of situated embodiment and the sociocultural situatedness of language. In
this respect, the papers included in this volume contribute to a fuller under-
standing of these basic concepts and to the importance of promoting
broader interdisciplinary approaches to achieving our overall goal: that of
defining the role of culture in language and cognition as well as in collec-
tive and distributed conceptualizations. Hopefully, the broad range of re-
search perspectives offered in this volume will move us closer to achieving
that goal.
References
Clark, Andy
1997 Being There: Putting Brain, Body and World Together. Cambridge,
Mass.: MIT Press.
1999 An embodied cognitive science. Trends in Cognitive Science 3 (9):
345–351.
Clark, Andy and David Chalmers
1998 The extended mind. Analysis 58(January) (1): 7–19.
14 Roslyn M. Frank
D'Andrade, Roy G.
1995 The Development of Cognitive Anthropology. Cambridge: Cam-
bridge University Press.
Dirven, René, Roslyn M. Frank and Cornelia Ilie (eds.)
2001 Language and Ideology. Vol. 2. Cognitive Descriptive Approaches.
Amsterdam/Philadelphia: John Benjamins.
Dirven, René, Roslyn M. Frank and Martin Pütz (eds.)
2003 Cognitive Models in Language and Thought: Ideology, Metaphors
and Meanings. Cognitive Linguistics Research 24. Berlin/New York:
Mouton de Gruyter.
Döring, Martin and Brigitte Nerlich
2005 Assessing the topology of semantic change: From Linguistic Fields to
Ecolinguistics. Logos and Language: Journal of General Linguistics
and Language Theory 6 (1): 55–68.
Geeraerts, Dirk
1988 Cognitive Grammar and the history of Lexical Semantics. In: Brygida
Rudzka-Ostyn (ed.), Topics in Cognitive Linguistics, 647–677. Am-
sterdam/New York: John Benjamins.
2002 The theoretical and descriptive development of lexical semantics. In:
Leila Behrens and Dietmar Zaefferer (eds.), The Lexicon in Focus.
Competition and Convergence in Current Lexicology, 23–42. Frank-
furt am Main: Peter Lang Verlag.
Geeraerts, Dirk and Stefan Grondelaers
1995 Looking back at anger: Cultural traditions and metaphorical patterns.
In: John R. Taylor and Robert E. MacLaury (eds.), Language and the
Cognitive Construal of the World, 153–179. Berlin/New York:
Mouton de Gruyter.
Gibbs, Raymond W.
1999 Taking metaphor out of our heads and putting it into the cultural
world. In: Raymond W. Gibbs and Gerard J. Steen (eds.), Metaphor
in Cognitive Linguistics, 145–166. Amsterdam/Philadelphia: John
Benjamins.
Heidegger, Martin
1962 Being and Time. New York: Harper and Row. (Trans. by John Mac-
quarrie and Edward Robins of Sein und Zeit, Tübingen, Germany).
Original version 1927.
Jäkel, Olaf
1999 Kant, Blumenberg, Weinrich. Some forgotten contributions to the
cognitive theory of metaphor. In: Raymond W. Gibbs and Gerard J.
Steen (eds.), Metaphor in Cognitive Linguistics, 9–27. Amster-
dam/Philadelphia: John Benjamins.
Johnson, Mark
1987 The Body in the Mind: The Bodily Basis of Meaning, Imagination
and Reason. Chicago: University of Chicago Press.
Johnson, Mark and Tim Rohrer
2007 We are live creatures: Embodiment, American Pragmatism and the
cognitive organism. In: Tom Ziemke, Jordan Zlatev and Roslyn M.
Frank (eds.), Body, Language and Mind. Vol. 1. Embodiment, 17–54.
Berlin/New York: Mouton de Gruyter.
Kimmel, Michael
2005 Culture regained: Situated and compound image schemas. In: Beate
Hampe (ed.), From Perception to Meaning: Image Schemas in Cog-
nitive Linguistics, 285–311. Berlin/New York: Mouton de Gruyter.
this vol. Properties of cultural embodiment: Lessons from the anthropology of
the body.
Kövecses, Zoltán
2005 Metaphor in Culture: Universality and Variation. Cambridge: Cam-
Lakoff, George
1987 Women, Fire and Dangerous Things: What Categories Reveal about
the Mind. Chicago: University of Chicago Press.
Lakoff, George and Mark Johnson
1999 Philosophy in the Flesh. New York: Basic Books.
Lakoff, George and Zoltán Kövecses
1987 The cognitive model of anger in American English. In: Dorothy
Holland and Naomi. Quinn (eds.), Cultural Models in Language and
Thought, 195–221. Cambridge: Cambridge University Press.
Langacker, Ronald W.
1999 Assessing the cognitive linguistic enterprise. In: Theo Janssen and
Gisela Redeker (eds.), Cognitive Linguistics: Foundations, Scope
and Methodology, 13–59. Berlin/New York: Mouton de Gruyter.
Langacker, Ronald
1994 Culture, cognition, and grammar. In: Martin Pütz (ed.), Language
Contact and Language Conflict, 25–53. Amsterdam/Philadelphia:
John Benjamins.
Lee, Penny
1996 The Whorf Theory Complex: A Critical Reconstruction. Amster-
Lindblom, Jessica and Tom Ziemke
2002 Social situatedness of natural and artificial intelligence: Vygotsky
and beyond. Adaptive Behavior 11 (2): 79–96.
2007 Embodiment and social interaction: A cognitive science perspective.
In: Tom Ziemke, Jordan Zlatev and Roslyn M. Frank (eds.), Body,
16 Roslyn M. Frank
Language and Mind. Vol. 1. Embodiment, 129–163. Berlin/New

York: Mouton de Gruyter.
Lucy, John A.
1992 a Grammatical Categories and Cognition: A Case Study of the Lin-
guistic Relativity Hypothesis. Cambridge: Cambridge University
Press.
1992 b Language Diversity and Thought: A Reformulation of the Linguistic
Relativity Hypothesis. Cambridge: Cambridge University Press.
1996 The scope of linguistic relativity: An analysis and review of empiri-
cal research. In: John Gumperz and Stephen C. Levinson (eds.), Re-
thinking Linguistic Relativity, 37–69. Cambridge: Cambridge Uni-
versity Press.
Merleau-Ponty, Maurice
1962 Phenomenology of Perception. London/Boston: Routledge and
Kegan Paul. (Trans. by Colin Smith of Phénoménologie de la
Perception). Original version 1945.
Nerlich, Brigitte, Craig A. Hamilton and Victoria Rowe
2002 Conceptualizing foot and mouth disease. The socio-cultural role of
metaphors, frames and narratives. metaphorik.de 02/2002: 90–108.
http://www.metaphorik.de/02/nerlich.htm
Palmer, Gary B.
1996 Toward a Theory of Cultural Linguistics. Austin: University of Texas
Press.
2006 a Energy through fusion at last: Synergies in Cognitive Anthropology
and Cognitive Linguistics. In: Gitte Kristiansen, Michel Achard,
René Dirven and Francisco Ruiz de Mendoza (eds.), Cognitive Lin-
guistics: Foundations and Fields of Application, 264–305. Berlin/
New York: Mouton de Gruyter.
2006 b When does cognitive linguistics become cultural? Case studies in
Tagalog voice and Shona noun classifiers. In: June Luchjenbroers
(ed.), Cognitive Linguistics Investigations across Languages, Fields
and Philosophical Boundaries, 13–45. Amsterdam/Philadelphia:
John Benjamins.
Pires de Oliveira, Roberta and Robson de Souza Bittencourt
this vol. An Interview with Mark Johnson and Tim Rohrer: From neurons to
sociocultural situatedness
Robbins, Philip and Murat Aydede
in press Cambridge Handbook on Situated Cognition. Cambridge, U.K.:
University of Cambridge Press.
Rohrer, Tim
2006 Three dogmas of embodiment: Cognitive linguistics as a cognitive
science. In: Gitte Kristiansen, Michel Achard, René Dirven and
Francisco J. Ruiz de Mendoza Ibánez. Cognitive Linguistics: Current

Applications and Future Perspectives, 119–146. Berlin/New York:
Mouton de Gruyter.
2007 The body in space: Dimensions of embodiment. In: Tom Ziemke,
Jordan Zlatev and Roslyn M. Frank (ed.), Body, Language and
Mind: Embodiment, 339–377. Berlin/New York: Mouton de Gruyter.
Rosch, Eleanor
1999 Reclaiming concepts. Journal of Consciousness Studies 6 (11/12):
61–77.
Sharifian, Farzad, René Dirven, Ning Yu and Susanne Niemier
in press Culture, Body and Language: Conceptualization of “Heart” and
other Internal Body Organs across Languages and Cultures. Berlin/
Sharifian, Farzad and Gary B. Palmer
2007 Applied cultural linguistics: An emerging paradigm. In: Farzad Shari-
fian and Gary B. Palmer (eds.), Applied Cultural Linguistics: Impli-
cations for Second Language Learning and Intercultural Communi-
cation, 1–14. Amsterdam/Philadelphia: John Benjamins.
Sinha, Chris and Kristine Jensen de López
2000 Language, culture and the embodiment of spatial cognition. Cogni-
tive Linguistics 11 (1/2): 17–41.
Verhagen, Arie
2000 Interpreting usage: Construing the history of Dutch causal verbs. In:
Michael Barlow and Suzanne Kemmer (eds.), Usage Based Models
of Language, 261–286. Stanford, CA: CSLI Publications.
Violi, Patrizia
2004 Embodiment at the crossroads between cognition and semiosis. Re-
cherches en Communication. 19: 199–234.
this vol. Beyond the body: Towards a full embodied semiosis
Vygotsky, Lev
1978 Mind in Society: The Development of Higher Psychological Proc-
esses. Cambridge, MA: Harvard University Press. (Trans. from Rus-
sian source text, 1930).
Ziemke, Tom
2001 The construction of ´reality´ in the robot. Foundations of Science 6
(1): 163–233.
2003 Embodied AI as science: Models of embodied cognition, embodied
models of cognition, or both? Embodied Artificial Intelligence 605–
615
Ziemke, Tom (ed.)
2002 Situated and Embodied Cognition [special issue]. Cognitive Systems
Research 3 (3).
18 Roslyn M. Frank
Zlatev, Jordan
1997 Situated Embodiment: Studies in the Emergence of Spatial Meaning.
Gotab: Stockholm.
2005 What’s a schema? Bodily mimesis and the grounding of language. In:
Beate Hampe (ed.), From Perception to Meaning: Image Schemas in
Cognitive Linguistics, 313–342. Berlin/New York: Mouton de
Gruyter.
2007 Embodiment, language and mimesis. In: Tom Ziemke, Jordan Zlatev
and Roslyn M. Frank (eds.), Body, Language and Mind. Vol. 1. Em-
bodiment, 297–337. Berlin/New York: Mouton de Gruyter.
Section A
The dynamics of cultural categorization

An interview with Mark Johnson and Tim Rohrer:
From neurons to sociocultural situatedness
Roberta Pires de Oliveira and Robson de Souza

Bittencourt
Abstract
As one of the three findings of Second Generation Cognitive Science, the notion of
embodiment changed radically the way the mind-body problem was viewed and
consequently the scope of many disciplines associated with it. Many cognitive
researching branches, from neural sciences to philosophy, concurred in saying that
“our human embodiment determines both what we think and how we think” (Mark
Johnson, in this interview). This is the core of this interview: a clarification of the
notion of embodiment and its relation to the many issues. Human embodiment
springs from the interview not as “the fleshy boundary of the skin”, on the contrary,
our body and brain extend out to the world beyond us, engaging in all sorts of bod-
ily and socio-cultural interactions, in experiences of meaning which are not objec-
tively out there. These are only affordances, that is, “they afford opportunities for
individuals to experience the meaning of things and situations and events”. Since
embodiment entails interaction in levels – bodily, socio-cultural, aesthetic, etc – it
rules out physicalist monism in the traditional sense. Thus, we end up with a new
way of seeing: an embodied mind in a “minded” body!
Keywords: Cognitive Science, Cognitive Linguistics, embodiment, embodied ver-

sus disembodied cognition, Experiential Realism, mind/body dualism, Social Em-
bodiment.
Thank you for the opportunity of interviewing you both, Tim and Mark. It’s
a great pleasure to have the opportunity to clarify the notion of embodi-
ment and at the same time to be able to gain access to a more neuro-
scientifically oriented perspective as well as a more philosophical one. The
notion of embodiment, generally associated with one of the three discov-
eries of the second generation of Cognitive Science, radically changed the
way not only Cognitive Linguistics understood the mind, but also many
22 Roberta Pires de Oliveira and Robson de Souza Bittencourt
other disciplines associated with contemporary Cognitive Science, evolu-

tionary biology, philosophy and the “life sciences”, as well as more spe-
cialized areas such as AI, artificial life and robotics.
Embodiment has no usual counterpart word in Brazilian Portuguese. In
fact, there is no good translation for it, since every alternative would imply
that something that had no body was embodied. In English, however, it
seems that embodiment is an every-day word. For instance: “She’s the
embodiment of honesty.” We could start by clarifying what (if any) rela-
tions exist between the ordinary use of the term and its use in the cognitive
paradigm.
MJ: The term “embodied” should be understood, first, as a contrast term to

“disembodied”. The idea of disembodied mind, concepts and thought en-
tails some form of ontological dualism, typically a mind-body dualism. On
this view, the body as a material thing is the locus of sensation, images and
feelings. It gives “input” to something called “the mind”, but it does not
determine what the mind does with that input. Our view, by contrast, is that
the fact of our human embodiment shapes both what and how we think. In
a Pragmatist vein, and also in line with recent developments in cognitive
neuroscience, we are denying disembodied thought. All dimensions of
human thought emerge from increasing levels of complexity in organism-
environment interactions, and all of these interactions require and are
grounded in our bodies.
Could you, perhaps, clarify what you mean by Pragmatist vein? Do you
mean American Pragmatism à la Dewey?
MJ: Yes, we are referring to classic American Pragmatism as articulated

by philosophers such as C.S. Peirce, William James and John Dewey.
Their work is characterized by a non-dualistic metaphysics that sees expe-
rience as an ongoing process of organism-environment interactions that are
at once physical, social, moral, cultural and spiritual. They argued that our
traditional Western dichotomies – mind vs. body, cognition vs. emotion,
fact vs. value, theory vs. practice, etc. – capture dimensions or aspects of
the more primordial flow of embodied experience, but that these divisions
do not represent ultimate metaphysical distinctions or types. Each of these
pragmatists was committed to the idea that “higher” cognitive functions
arose from our bodily engagement with our world. In other words, our ca-
pacities for conceptualization, reasoning and symbolic interaction were
An interview with Mark Johnson and Tim Rohrer 23
held to be continuous with our sensorimotor capacities. They saw human

thought and creativity within an evolutionary framework, and they drew on
the best science of their day in trying to understand how “mind” emerges
from embodied activity in the world.
How does the scientific meaning of embodiment emerge from the notion in
ordinary language?
MJ: In ordinary language, the term “embodied” can mean something as

general and abstract as “having a concrete instantiation in some physical
object”. Thus we say, “She is the embodiment of grace” and “He embodies
our highest ideals”. In common parlance the term “body” is also used to
identify certain material objects, with the human body as a prototype.
Moreover, the prototype is a living, breathing, acting human organism, and
not just the cold lump of flesh that rests in the morgue after someone dies,
although this is certainly one sense of “body”.
It is this living flesh that is at the heart of our conception of embodi-
ment. To say that cognition is embodied is to recognize multiple levels at
which the body shows itself. First, there is the physiological organism
made up of flesh, bones, blood, muscles, viscera and many organs of per-
ception and life-maintenance, all organized into complex interactive sys-
tems. Second, to be embodied is to have a brain and central nervous system
that establish further conditions for how we monitor our body-state and our
ongoing interactions with the environment. Third, the body does not termi-
nate with the fleshy boundary of the skin. It extends out into its environ-
ment, so that the organism and environment are not independent, but rather
interdependent aspects of the basic flow of bodily experience. Our em-
bodiment gives rise to felt qualities and emotional responses and phenome-
nology provides us ways of becoming aware of how we exist – bodily – in
our world. The neurosciences study how our organs, chiefly the brain and
nervous system, constitute our patterns of thinking. And the other cognitive
sciences can examine the role of our bodies in meaning, conceptualization,
reasoning and communication (or symbolic interactions).
Consequently, the scientific and philosophical senses of embodiment go
far beyond our commonsense conceptions of the body to encompass virtu-
ally every aspect of experience, meaning, thought and language. But what
runs through all of these various conceptions is the refusal to admit an
ontological difference in kind between mind and body. Hence, John Dewey
(1925 [1981]) coined the term “body-mind” to capture the idea that what
we call “body” and “mind” are, as Merleau-Ponty (1962 [1994]) also ob-
served, simply abstractions from the more primordial pre-subject and pre-
object flow of experience. To speak of embodied cognition is to say, "no
body, never mind”. It is to say that we think from and within our bodily
experience.
TR: I would just add to Mark’s description of the levels at which the body
shows itself by emphasizing that our interactions with the environment
include interactions with other organisms – particularly other human be-
ings and thereby our social, cultural and communicative systems.
As you both have made clear, the notion of embodiment is a refusal of the
traditional dichotomy of body/mind. At this stage there seems to be a con-
sensus that the Cartesian, dualist view of the body/mind antithesis can no
longer be accepted, though there are distinct non-dualist positions. Just to
give an example, Chomsky (2000, among others) argues that Descartes
was right about the mind, but wrong about the body. In his own terms,
since Newton’s physics, the ghost is in the machine. Moreover, Chomsky
postulates a clear distinction between natural phenomena, of which lan-
guage, i.e. syntax, is an example, and socio-cultural phenomena, which
cannot be explained naturalistically. It seems that the cognitive perspective
within which you work would not accept either of the two claims by Chom-
sky, but would rather rely on the notion of body. But if ultimately every-
thing must be explained in bodily terms, aren’t we back to some kind of
physicalism?
MJ: The crux of your question is whether everything must ultimately be

explained in bodily terms. In answering “yes” to this key question, what we
are rejecting is any proposed explanation that relies on the assumption of
disembodied meaning, conceptualization and reasoning. The apparent
counter example that is always cited, the one you mention from Chomsky,
is that there are aspects of language and also of social and cultural meaning
that cannot be explained via embodiment. However, it should be clear from
my response to question 1 above that “embodiment” is used in a very rich,
non-reductivist sense in the theory of embodied cognition. Meaning is
based on our human embodiment, but it is not locked up within individual
organisms. Meaning is public and shared. The principal reason it can be
shared is that we have similar sensory-motor systems, similarly structured
bodies, brains that function similarly and environments that afford us re-
curring, shared patterns of interaction. What leads some people to say that
meaning is somehow “outside” the body/mind is typically that our coordi-
nated human actions involve language, symbolic interactions, rituals,
shared practices, etc. There is thus meaning invested in what transcends the
confines of any particular body. Andy Clark (1999) calls this vast transper-
sonal dimension of meaning “scaffolding”.
These cultural forms and symbolic interactions are thus integral to
meaning, and we see them as our shared way of carrying meaning forward
from generation to generation. They make it possible for each person to
enter a world of funded meaning, in which the prior accumulated under-
standing of our ancestors is available to us. That is why each new infant
does not have to start from scratch to build the world anew or reconstitute
all our learning and inherited understanding. However, these forms are
always just affordances: they afford opportunities for individuals to experi-
ence the meaning of things and situations and events. Their meaning is not
written objectively on them. It is not something pre-determined and fixed.
Rather, they enter into our experience of meaning, which is organized by
the character of our bodies and brains, as we reach out actively to engage
what lies beyond us. These “objective” symbols and bodies of knowledge
must be enacted (as neuronal patterns) within and taken up by each person
for whom they become significant. And this requires embodied neural acti-
vations and the forming up of stable background knowledge, in the form of
what Paul Churchland (2002: 28) calls “the entire activation space for the
relevant population of neurons, a space that has been sculpted by months or
years of learning, a space that encompasses all of the possible instances of
which the creature currently has any conception”.
TR: If you have to carve the world up that way, I’m with the Monists be-
cause I accept evolutionary explanations. Given enough time, the socio-
cultural parts of embodiment have emerged for us as a result of evolution-
ary changes in our material embodiment. But is it still useful to carve
things up in this way? After all, the concepts of monism and dualism and
the like presume that notions like body and mind belong in some fixed and
final categories, not that one gradually emerges from the other in a long
Darwinian process. We humans now live in a milieu which is partly social
and cultural; to explain things in bodily terms is to explain them, at least in
part, in terms of the social and cultural. Experiential Realism is not just
physical Monism; it’s a subspecies of post-Darwinian philosophical Prag-
matism in which later phenomena, like our sense of “mind” and the other
sociocultural phenomena which Chomsky finds unamenable to naturalistic

explanations, are now bound up with the physical body.
So, it goes beyond traditional notions based on the mind-body dichotomy

…
TR: So neither monism, nor dualism, nor physicalism can really adequately
label this theory. As Mark argues, explaining everything in bodily terms is
no longer just explaining them only in the terms of the physical body. Only
when considering evolutionarily long time-scales might it be possible to
attempt that sort of reduction on a wholesale basis. The rest of the time, we
attempt to explain things using embodiment in its richly non-reductivist
and interactionist sense with the physical-body-to-mental/social-cultural
evolutionary sense in the background.
But on the other hand, there have been attempts – notably by the NTL re-
search group headed by Lakoff and Feldman – to explain language and
mind in terms of the brain. So, if we understood you correctly, when Lakoff
says that “all concepts are physical”, as he did in Logroño at the ICLC
2003, he does not mean that concepts are identical with neural struc-
tures/patterns of activation, right? Otherwise, how can he avoid reduc-
tionism?
MJ: To say that all concepts are “physical” is to deny that concepts are
attributes of some alleged immaterial substance or structure. It is to insist
that concepts are human creations and tools (and not just human, since
some animals have concepts, too). There can be no conceptualization with-
out a pattern of neural activation. That is the heart of the claim that con-
cepts are physical. It does not follow from this that every concept is corre-
lated with a single neural activation, even within the same person at
different times. There is too much neural plasticity (thankfully) for this to
be the case. Gerald Edelman makes this case very strongly in Bright Air,
Brilliant Fire (1992), and in his book with Giulio Tononi, A Universe of
Consciousness (2000). Still, human brains process certain concepts using
various parts of the sensory-motor system, although there may be great
variability in how these concepts are realized in different people.
As we mentioned earlier (in response to Question 2), however, it is im-
portant to always remember that grounding conceptualization in brain
events does not mean that a complete account of concepts can be given
without reference to social interaction and cultural symbols and institu-

tions.
TR: While physical, my neural experience of red is not identical to your

neural experience of red – simply sufficiently similar in their patterns of
activation due to evolutionary pressures. What it is not is some concept
which corresponds to some disembodied experience of “red” as it might
exist apart from the bodies interacting in and with the world.
Your answers, as well as the literature in Cognitive Linguistics, clearly

recognizes and advocates for a broader view of concepts and language,
one that ranges from physical to social and aesthetic experiences as well
as genetic and cultural heritages. One the other hand, as pointed out ear-
lier, one central research topic is the neuronal basis of concepts. Lakoff
seems to suggest this in his 2001 interview concerning blends, when he
says: “So for us who are working in NTL, blends are real but they are just
ordinary everyday phenomena. They are nothing special, no new theory is
needed for them, there is no need for any theory of blending” (Sánchez
2003: 259). Accordingly, the reason there is no need for a new theory is
because “blendings are bindings”, and this neuronal notion (binding) is
enough to account for the phenomenon. However, Fauconnier and Turner
(2002) use the term “binding” all the time, too. Do they mean the same
thing by it or something different? So, could you say how you see their
views: what is “binding” from a neuroscience point of view and how does
it affect blending theory? Is the main task of NTL that of ultimately unify-
ing mind and brain, and hence the complete elimination of this Cartesian
dichotomy?
TR: To neuroscientists, the “binding problem” is the problem of how the

fairly well-understood qualia of experience (i.e. color, shape, motion)
come together as a unitary conscious experience – i.e. my experience of the
red pen as it moves across the paper as I write these words. Speaking as a
cognitive neuroscientist, I can explain how and where perceptual “redness”
is instantiated in neural patterns, I can explain where in our cortical cir-
cuitry we map the shape of the pen and where we map its motion across the
paper, but no one has yet been able to explain how these different patterns
are synchronized or bound together to give a unitary conscious experience.
The debate comes up because Fauconnier and Turner (2002) have pro-
posed that conceptual blending theory can be extended to be a model of
how “red” is bound to “pen” in phrases such as “the red pen” that might
serve to call up some mental image of it. Without presuming to speak for
them, there seem to me to be at least two motives for extending blending
theory in this way: First, it highlights that this “binding” operation is lin-
guistic as well as perceptual. While cognitive neuroscientists have largely
focused on the “bottom-up” (perceptual to unitary conscious experience)
view of the binding problem, one can also ask how a more “top-down”
(language to unitary conscious experience to perceptual imagery) view of
neural binding could work. Second, this extension could serve to unify
blending theory with a problem in neuroscience, providing a possible neu-
ral basis for explaining how blending might take place in other, more com-
plex cases (e.g. their “the pope finding it hard to box with a mitre on his
head” example). By contrast, in a more “bottom-up” model of language
like NTL, linguistic expressions like “the red pen” are simply seen as the
natural outcome of how processes of color and objects interact in neural
terms.
For my own part, I am not happy with either camp’s claims. I am not
sure how pragmatically useful it is to think of these “perceptual” blends as
equivalent to the neural binding problem – partially because like most
problems in the study of consciousness, it remains unsolved and in my
opinion its mention is likely to contribute nothing more than a distraction
from a focus on unrelated problems that are solvable. Nor do I think that
NTL is a very tight neurocomputational model of the underlying neuroanat-
omy, so I don’t buy the argument that modeling perceptual blends, i.e.
solving the binding problem, will just be a natural outcome during the
course of modeling metaphor.
I can’t ever see any one single notation ever unifying “the mental” and
“the neuronal” in general. But we might eventually be able to agree that a
particular NTL-successor model is (i) a neurocomputational model of the
particular neuroanatomical processes (ii) that underlies a range of linguistic
expressions of a particular conceptual metaphor and (iii) how, on a specific
run, it can produce a particular conceptual blend. Doing that much would
be very impressive, and NTL isn’t really all that far away from doing that.
MJ: Saying that Conceptual Blending Theory requires a theory of neural

binding is true but not very startling, since, after all, all conceptualization,
cognition and reasoning would have to involve neural binding (or how else
would we have unified experiences and thoughts). I would simply observe
that, if there is no thought without an active brain in an active body en-
gaging its surroundings, then neural binding strikes me as the most plausi-
ble way to think about how various kinds of conceptual blending are possi-
ble. If concepts involve patterns of neural activation, then the ways those
patterns are connected (i.e. bound together) during a certain temporal win-
dow of co-activation will determine how the conceptual blends have the
meanings they do. Fauconnier and Turner (2002) do not attempt to give a
theory of neural binding as part of their theory of conceptual blends. One
of the things that most distinguishes the Neural Theory of Language (NTL)
project is that its supporters actually do take seriously the challenge of
modeling the neural architectures and dynamic processes that underlie
human cognition and language. This is, of course, a monumental under-
taking, far beyond the reach of current computational neuroscience. Lakoff,
Feldman, Narayanan, Regier, Gallese (2005) and other NTL proponents are
under no illusions about the tentative and partial nature of their current
proposals. They recognize that they are offering computational models that
are going to have serious shortcomings, they recognize that much of what
they say is and must be highly speculative (in light of the nascent character
of cognitive neuroscience), and they are fully aware that nobody can pre-
tend to have neural reductions of all cognitive phenomena. Yet, the merit
of their project is that they have accepted the assignment of actually de-
scribing neural processes underlying language. They don't just provide a
theory of syntax or semantics or pragmatics and then naively assume that
the elements of their theory must somehow have neural mechanisms.
They've tried to build their models on recognized neural architectures, and
then they refine these models on the basis of research coming out of Cog-
nitive Linguistics.
Could we concentrate a little bit more on the Neural Theory of Language

(NTL) model? There are some aspects of it that we would like to clarify.
Would you say that NTL view is holistic, and thus incompatible with a
modular view of the mind? Aren’t there different systems, which work in
different ways and interact with each other?
TR: This is basically a confusion resulting from misunderstanding what

concepts apply at which levels of investigation. There is world of a differ-
ence between saying the mind-brain is holistic and saying it is non-
modular. Holism is clearly false if by that word we mean that the entire
brain is significantly involved in every local computation. But modularism
goes awry because while there is reasonable neuroanatomical support for
modularity in the neurocomputational modeling of some small scale neural

structures, e.g. Marr’s original work on orientation-tuning columns, it is
not the case that the brain in general is merely the succession of such iso-
lated structures, each taking as their inputs the output of previous modules.
In fact, once you move to the scale of describing the brain at a systems
level – spatio-visual, auditory, sensorimotor, object-recognition and so on –
there is considerable interconnection between neural systems, even though
they are largely localizable to different brain regions. The notion of distinct
modules is just simply not useful at this level of scale, unless you purpose-
fully want to abstract away the complexities of those interconnections for
the purposes of producing a neurocomputational module of some large-
scale process. But unlike in the small-scale cases that modularity theory
was built on, there is generally no strong neuroanatomical motivation for
doing so.
This misunderstanding has real consequences even for good work on
language. For example, Levinson took modularity so seriously in analyzing
Tzeltal body-part prepositions that he assumed that “spatial-primitives” of
the visual system would be inaccessible to the language system (cf. Rohrer
2001).
However, our position is that language is not a neural system at the
same level of analysis as systems such as these. I believe that language and
mind use these same cortical subprocesses of these neural systems – just in
an off-line, emulative way. When neuroscientists write about the brain’s
“semantic system”, they typically mean something at the end of “object-
recognition system”, typically in the anterior temporal cortex, because our
prototype example of semantics is object-naming. Naturally, many of them
have been pleasantly surprised by my work and that of Hauk, Johnsrude
and Pulvermuller (2004) showing that primary and secondary sensorimotor
cortical areas respond to linguistic stimuli for body parts. Similar results
have been published for hand tools, and we are also getting similar reports
of neurological deficits involving body-part naming (Coslett, Shaffran and
Schwoebel 2002). From a modularist’s perspective this is terrible: seman-
tics belongs in its black box down in the anterior temporal cortex, not up in
motor cortex. Fortunately, however, most neuroscientists are more wedded
to the data than to a modularity theory of language. It is my belief that we
will ultimately show the same sort of results for syntax. My hunch is that
many syntactic relations rely on spatial neural subprocesses found largely
in the parietal lobes.
Let us move back to Lakoff’s views on embodiment. He extensively argued

that language is explained directly by our bodily movements/experience,
which are given a neuronal explanation. My language about love, for in-
stance, is explained by my physical reactions: I feel a warmth through my
body when I meet my beloved ones, my heart beats faster, … and this ex-
plains my making sense of sentences such as: You are the sunshine of my
life. But there certainly are different levels of “embodiment”: one is of my
conscious feelings about love, and another is my non-conscious being.
What do you believe to be the relationship between a “sub-conscious”
level of being, and a “conscious level”? Aren’t there intermediary levels?
TR: Well, I believe that when you consciously hear language such as
above, you subconsciously imagine sensations like warmth and so on – to a
degree, of course – you typically don’t consciously feel warm just as a
result of hearing such language. Similarly, we know that experimenters can
measure the activation of low-level visual cortical areas of subjects asked
to do visual imagery tasks, though the same doesn’t seem to hold true for
low-level auditory areas when subjects imagine auditory phenomena. From
my own research, I know that with language tasks we often have to build
up the theme of body-part language before we can measure any activation
(how much depends in large part on which neurophysiological measure-
ment method is being used). But I think it is important to note that we are
just at the beginning of figuring out how to design the right stimuli, how to
measure these activations and where exactly to look. I don’t see a need for
intermediate level to mediate between our sensations and the neural. We
can just have activation below the threshold of consciousness.
Of course, it is important to acknowledge that there is a common cul-
tural component underlying what I am saying are the expected neural acti-
vations. Most of this work is done on Indo-European speakers in North
America and Europe who have highly similar cultural models underlying
their conceptual metaphors. A really interesting experiment would compare
the neural activations of both westerners and non-westerners reading both
non-western cultural conceptual metaphor expressions and western cultural
conceptual metaphor expressions. However, we will have to wait for some
enterprising person to perform such an experiment to see what differences,
if any, could be found and attributed to cultural factors.
MJ: I don’t know of anyone who has a completely satisfactory explanation

of the relation of the unconscious to conscious processes, but I like Anto-
nio Damasio’s (1999) attempt to explain the processes by which the body
monitors changes in its own state, as the result of ongoing interactions with
its surroundings. He says that a moment of core consciousness consists in
the feeling awareness of what is happening in your body, as it is affected
by both its internal processes and events in the external world. But Dama-
sio doesn’t pretend to have an adequate neural theory of how this works in
all cases. Moreover, even if we had a good theory of consciousness, this
wouldn’t really address the claims Lakoff and I (and many others) have
been making about the unconscious activation of various sensorimotor
domains as the basis for different conceptual metaphors. We are seldom, if
ever, consciously aware of the neural activation of the sensorimotor source
domain of a conceptual metaphor. That is the primary reason why we need
the methods of the cognitive sciences to probe these unconscious proc-
esses, since we cannot rely merely on phenomenological reports of what
we are feeling or thinking.
Lakoff does not attempt to ground concepts and language in neurophysiol-

ogy, but instead he uses an intermediary “computational level” of struc-
tured connectionism, X-schemas etc. as implemented in the computer mod-
els of Regier (1995), Feldman and Narayanan (2004) and Feldman (2006),
among others. What status does, in your opinion, this “computational
level” have? If it is a higher-level description of, but otherwise identical
with the neurophysiology “beneath” it, isn’t Lakoff a philosophical func-
tionalist – “the software matters rather than the hardware” – despite fre-
quent claims to the contrary? If it is only an “approximation” or “model”,
then how can Lakoff claim that NTL really grounds language in neurobiol-
ogy?
TR: The NTL computational level has exactly the status it is defined to
have: it is comprised of programming constructs that are mathematically
reducible to known neural behaviors. But it is also no more than that: by
definition it is not a very tight model of the way any particular neuro-
anatomical structures perform whatever actions a particular NTL model
seeks to model. That’s why NTL typically claims to have a neurally plausi-
ble model, not a tight neurocomputational model of what this particular
brain region is doing. Furthermore, NTL isn’t philosophical functionalism
in its strict sense because in this case the software is being designed with
the constraints of the biology in mind, i.e. all constructs are mathematically
reducible to known facts about the neurobiology. The claim is never that
the software matters and the hardware doesn’t. In fact the hardware (or
more accurately the neurophysiology) determines what kinds of constructs
the software can have. In this sense NTL grounds its ability to comprehend
language in the neurobiology.
When we look at the details of exactly what constructs are used in one
particular NTL model (KARMA or SHRUTI for instance), the degree to
which the neurophysiological details constrain the software constructs is
often not enough to satisfy someone who wants a more neuroanatomically
tight model such as myself, but it is still a starting place. Because I would
define the levels of the investigative model in terms of the physical scale
that produces the phenomenon to be studied (or modeled), I would further
argue that NTL is mistaken to consider computation as a level rather than a
method. I would say it this way: NTL models use an adequately neurally-
plausible computational method to produce a model of linguistic activity in
the brain-where the brain is considered at a fairly high level of investiga-
tion where the interaction of neural systems is the focus of investigation. If
one wants more detailed models of linguistic activity in the brain at lower
levels of investigation, one needs to use constructs which have more direct
neurophysiological analogs. As it stands, I think NTL is a really interesting
case of bridging machine language efforts and neurocomputational model-
ing.
Closely linked is the issue of representations, which seems to play a role in

the cognitive view you advocate, right? Does (human) cognition depend on
the use of representations, and if so, what kind of representations? If, on
the other hand, one rejects the idea that representations play any (impor-
tant) role in cognition – as some proponents of embodied cognition such as
Maturana and Varela do – how can one account for capacities such as the
abilities to plan, imagine, believe, etc.?
MJ: We are arguing that the classical Representationalist view of cognition

is mistaken, even though it continues to exercise a virtual stranglehold on
commonsense and philosophical views of mind. The Representationalist
view is that thought involves the mental processing of internal mental enti-
ties (called variously, “ideas”, “concepts”, “representations”) that can stand
in a relation of intentionality (directedness at or toward) some mind-
independent object, person, event or state of affairs. However obvious it
might seem that we can entertain ideas (such as dog) that, through some
mental function of referring, allow us to pick out which objects in the
world are dogs, this account is seriously misleading. The problem is two-
fold: (1) It involves an incipient mental/physical dualism, and (2) it gener-
ates an inescapable and insoluble skepticism, since we can never be sure
that our inner ideas do actually represent correctly what is “out there” in
the world beyond our minds.
Consider the fact that there are topographical and topological “maps” in
our brains. We now know that there are neural maps in the visual cortices
that preserve structure and relations of objects “in the world”. While these
might appear to be prime examples of “inner representations of outer reali-
ties”, they are not. For an organism that has such visual maps, they just are
the structures of its visual experience. The maps don’t re-present anything;
rather, they are the neural activations that allow us to experience what we
experience and to think what we think. I suspect that we are too easily se-
duced into the Representationalist view, just because we have the gift of
language. Language makes it possible for us to name our concepts, and this
tempts us to treat them as if they were mental objects with various proper-
ties and relations (like the relation of “referring”). The fact that we can
abstract aspects of the ongoing flow of our experience and treat them as
general patterns capable of being instantiated in past and future experience
need not lead us into Representationalism. The neural account of how this
is possible will require accounts of reentrant mapping, feedback loops,
binding and other cognitive processes, and nobody has the full story on this
yet, but we are taking the first steps in this direction.
I do not object to the use of “representation” for any pattern of neural
activation, but this can be risky, insofar as it can lead us to mistakenly hy-
postatize concepts and to reinstate the inner vs. outer ontology of mind. We
can even say that our representations have the property of intentionality, as
long as we mean by this only that when we attend to some part of the flow
of experience and treat it as a generality that transcends its particular pres-
ent instantiation, then it can “point beyond itself” to aspects of past and
future experiences. As abstracted, it can become part of a reasoning proc-
ess that goes beyond the immediately given.
TR: Though its use is rampant in the neurosciences I try hard to avoid the
term – although when cognitive neuroscientists like Steve Kosslyn (1994)
take the pains to emphasize the term’s roots with hyphenation when ex-
plaining how visual images are re-presented to successive visual cortical
areas, I think he shows how to rectify this confusion by making the term a
little bit more active and dynamic. However, the presentational metaphor
still suffers from the problem that there is really nothing in neural terms for
the re-presentation to be presented to – no homunculus in the brain. That is
why I prefer to talk about neural cognition using terms like mapping, maps,
image-like wholes and topography/topology preservation (and omission!).
It is obviously not the case that the brain has static copies of every image
we have ever seen stored in our heads, or I could quote every page I’ve
ever read. As Nietzsche (1874) observed, there is real virtue in forgetting.
Some of the topographic and topological details drop out in every cortical
re-presentation. In fact, our brains learn to select for the parts of the image
that are useful for us to survive and flourish. (Similar observations hold for
the other perceptual modalities.)
I don’t think Maturana and Varela (1980, among others) are vulnerable
on this point either. We can dream, imagine, plan, believe and the like be-
cause we can use these same brain areas in an emulative and offline man-
ner to re-present future possibilities – to anticipate. In their terms, we are
enacting these possibilities as part of the normal process of living, just as
when we reach for a tool our neural structures are already forming the hand
into the appropriate shape to grasp it. In fact, recent fMRI studies show that
premotor and motor areas of the sensorimotor cortex are activated by sim-
ply viewing pictures of hand tools (Vingerhoets et al. 2002).
So we still need a better account of re-presentation, right?

Mind, computers and related issues bring us back to a more precise
definition of embodiment, which, according to Ziemke (2003), has been
used since the mid-1980s. The author identifies the following notions of
embodiment: i) structurally coupled embodiment (which does not require a
body), ii) historical embodiment, iii) physical embodiment, iv) organismoid
embodiment, and v) organismic embodiment. Ziemke understands that
Lakoff’s approach (Ziemke explicitly quotes a passage from Lakoff (1988))
belongs to the fourth type (organismoid embodiment), which supports the
idea that cognition “might be limited to organism-like bodies”, e.g. that
human-like cognition requires a human-like or humanoid body with, e.g.
eyes, hands, legs, etc., and the (neural) mechanisms for their control and
coordination. In contrast, the fifth type of embodiment, i.e. organismic
embodiment, differs from organismoid because it holds that cognition is
limited to organisms, i.e. living bodies. “According to this view, there are
crucial differences between living organisms, which are autonomous and
autopoietic, and man-made machines, which are heteronomous and allo-
poietic” (Ziemke 2003: 1308). This classification corresponds roughly to
that given in the Chrisley and Ziemke (2002) paper published in the Ency-
clopedia of Cognitive Science). First, do you think that Lakoff’s NTL fits
into the organismoid notion of embodiment?
TR: More or less. In my opinion, current versions of NTL are probably

most compatible with organismoid embodiment as Ziemke’s categorization
system cleaves the subject. However, allow me to point out that Ziemke’s
categorization system for embodiment is drawn with respect to issues in
cognitive robotics. That is not the only pragmatically useful way to cleave
embodiment theory – for example, I use the more biologically-based ap-
proach common within cognitive neuroscience to discuss different ap-
proaches to embodiment on the basis of the physical size of the phenomena
which produce them, as well as giving a purely descriptive linguistic ac-
counting of the kinds of uses of the terms that exist within the literature on
embodiment (Rohrer BLM volume 1). Even more importantly, NTL wasn’t
explicitly conceived in terms of robotics, but as a way to bridge linguistic
theory, neural behavior and neurocomputational modeling by way of solv-
ing difficult problems in machine language learning. That’s a significant
difference from Brooks and Stein’s (1994) Cog proposal for a humanoid
robot, for instance.
In fact, there’s nothing vaguely humanoid about the physical instantia-
tion of any of the current NTL models. The humanoid aspect is coded into
how the models are built and what they model. But these models are very
partial; they model only limited parts of cognition and language. In princi-
ple, could a vast collection of NTL models similar to current ones, operat-
ing together model all of human cognition and thought? I think that even
more important than attaching a body to them will be the practical prob-
lems engendered by having the wrong physical architecture – silicon – will
serve to limit that problem long before it could be accomplished. We have
engineered silicon in the image of vacuum tubes to embody an all-or-
nothing logic, and then we run analog simulations on top of it. I think
you’ll need a more flexible medium – one that self-organizes and is error
tolerant.
Are you envisaging the possibility that the living body is not a necessary
requirement for providing agents, with meaning, consciousness? So we
would get the same human mind if we put NTL models into humanoid ro-
bots and “raise” them like children?
TR: Isn’t this a false dichotomy? I see that you are trying to raise what you
see as the key difference between organismoid and organismic embodiment
– whether or not the living flesh is necessary for meaning and conscious-
ness like ours. But if an NTL robot were to succeed on the grandiose scale
you outline, it would for all practical purposes have a body – one whose
NTL organization was dictated by the human neurophysiology, one whose
humanoid robotic body was determined by developing sensors and motors
that are dictated by how human physiology interacts with the physical envi-
ronment, and one whose cognition was enculturated from birth in a human
socio-cultural world. Those are the three tenets of embodiment in cognitive
linguistics: the physiological and neurophysiological body, the interactive
body, and socially and culturally embedded body. So such a grandiose
success would then have a body.
But would the grandiose success be fleshy (as in organismic)? Well,
matter matters. In order to have cognition like ours, I think that it is a nec-
essary requirement to have a medium – a collection of matter – that is for-
mally organized quite similarly to ours. Our usual idea of what that entails
is at minimum a living body, where living is defined in terms of autopoi-
esis. Now, if there were just one, it would be hard to see how it could re-
produce. But given enough of them to make a cultural milieu, then perhaps
they could. However, I think it may be less important here that something
be capable of self-reproduction than self-organization, and I would cer-
tainly maintain that we humans have created some interesting self-
organizing machines. Unfortunately we are more apt at creating self-
organizing machines which emulate how crayfish swim (Rowat and Selver-
ston 1997) than machines which can emulate human cognition and lan-
guage. I think NTL is a one small step on the road to getting at that latter
set of problems. But to accomplish the grandiose dream one probably
would have to rethink the medium as well. Part of that would involve sim-
ply putting NTL in touch with a humanoid body, while another part would
involve re-engineering how the silicon has been optimized – or perhaps
rethinking whether silicon is the optimal substrate to solve these sorts of
problem. Couldn’t it be artificial, and yet still fleshy?
I guess so, but then the notion of body is too restricted, isn’t it? It is just
flesh, but that is not really the point, is it?
TR: Believe me, NTL is still a long way from such grandiose success. It
doesn’t even work that much like the neurophysiology yet. As a counter-
point, consider that the Darwin II robot at Edelman’s Neurosciences Insti-

tute is both much more neurophysiologically sophisticated and yet doesn’t
do very much at all yet that we would think of as human cognition – cer-
tainly not language.
Aren’t you reduplicating the dichotomy between mind and brain (mind
being software that can run on different hardwares)?
TR: No. I don’t buy the definition of artificial intelligence as mind equals
software, brain equals hardware. I would instead argue that an artificial
intelligence simply means that we have imbued something with some sort
of patterns that we can recognize as having cognitive processes close to our
own – as Marvin Minsky (1965) puts it, the problem is that no one is even
remotely close to coming up with an AI with whom we can talk. If anyone
were to build such an artificial intelligence, it would be intelligent in part
because it would have a certain kind of physicality – one that we could
interact with. (A nanorobot won’t cut it on this definition – just the wrong
physical size.) In my view the hardware (the robot body and brain) would
be as much a part of the intelligence as the programming (the model).
Another way to think of this view of the possibility of artificial intelli-
gence comes out of how I think of work in cognitive anthropology by re-
searchers like Ed Hutchins (1995). Consider a watch as an example of an
intelligent cognitive artifact. We certainly interact with them cognitively –
they embody very useful patterns to us. A good one at least is pretty much
self-organizing – we don’t have to constantly reset it. Moreover, it is artifi-
cial – not even William Paley (1802 [1986]) would dispute that. To me,
what most people mean by artificial intelligence is nothing more than an
extension of the class of intelligent cognitive artifacts which are at least
minimally self-organizing – the wristwatch, the compass, the gyroscope.
Such an artificial intelligence would just be much more self-organizing and
more sophisticated than the wristwatch. And perhaps even that much more
useful.
Putting aside the issue of the creation of artificial bodies, one still faces
the problem of how to bring together mind (and subjective experience) and
brain. Violi (2003: 217), for instance, claims that the reduction of em-
bodiment to the brain no longer allows us to cope with the phenomenologi-
cal realities of perspective and subjectivity. She concludes her paper with
these remarks: “We have a deeply paradoxical chiasmus: on the one hand,
there is a theory of embodiment without the subject, on the other a theory

of the subject without a body. To finally achieve a bringing together of
body and subject might well be the most challenging goal of all for a cog-
nitive semiotics to [have].” To what extent do you agree with this conclu-
sion?
MJ: Violi is absolutely correct that “reduction of embodiment to the brain”

is reductivist and leaves out key aspects of human experience. That is why
Lakoff and I (1999) have insisted on the necessity of multiple levels of
embodiment, and correlatively, multiple levels of explanation in cognitive
science and philosophy. As we have already seen above, “the body” is not
just the brain, it is not just the physical body, and it is not just the organ-
ism-environment coupling. It is all of these, and probably more. The phrase
that Lakoff and I repeatedly use to capture this richer sense of embodiment
is something like “a brain, in a living body, in a changing environment that
is at once physical, social, cultural, economic, religious, gendered, etc.”. I
believe that second-generation cognitive science (with its attendant non-
reductivist philosophical orientation) is trying to overcome the chiasmus of
which Violi speaks, where we have found ourselves with a “theory of em-
bodiment without the subject” and a “theory of the subject without a
body”. The most philosophically sophisticated cognitive neuroscientists –
people like Antonio Damasio and Gerald Edelman – would never, ever say
that you are your brain. They recognize that we need the combined dia-
logue of neuroscience, cognitive and developmental psychology, phenome-
nology, Pragmatist views of mind and language, cognitive anthropology
and other disciplines, if we are to have a realistic view of what it means to
be a human person. This cannot be an easy task, as anyone knows who has
ever tried to move among the very different methods, vocabularies and
assumptions of even the most open-minded versions of these different ap-
proaches. But we can’t, or shouldn’t, give up, just because it is so difficult
and frustrating. We have to remember that each of our little attempts to
make sense of human cognition, identity and values is necessarily perspec-
tival, highly limited, oversimplified and likely to be supplanted at some
future time. Once we realize this, one of our biggest errors would be not to
listen to voices that remind us of how much our precious stories leave out
about what it means to be human.
In the same vein, it is common to define being a human by “intentionality”

(see Davidson (1980, among others), Searle (1979, 1980), among others).
We are meaningful organisms because we have intentionality. Artificial

systems lack this property (they also lack qualia; see interview with
Sánchez García (2003)) and this is the reason why they are necessarily
meaningless. Is intentionality derived from our bodies? If so, in what
sense? Would it be possible to build a robot with intentions? Imagine, for
instance, that intentionality is fully derived from our bodily experiences
and that we are able to trace all the way from body to intention and to
reduplicate this in a machine. (Though this seems to be science fiction, it is
a hypothesis seriously entertained by Zlatev (2001).)1
MJ: I don’t want to deny that so-called “intentionality” is one of the great
glories of humanity. Our human ability to go beyond our present situation
by using symbols and signs that have meaning has made possible our most
impressive scientific, social, cultural and spiritual achievements. But I’m
convinced that this claim is overblown, since certain other animals surely
have intentionality. And I’m especially distressed by the impoverished and
limited ways in which intentionality has been traditionally understood in
philosophy. The main problem, as I see it, is that intentionality has been
defined in terms of the human capacity for processing concepts and propo-
sitional content. But, as Paul Churchland and others have been arguing for
many years now, on the basis of cognitive science and neuroscience, it is
just false that human meaning and thought are essentially propositional and
linguaform. We’ve mistakenly assumed that the fact that our capacity for
language distinguishes our species entails the claim that all thought must
have the form of linguistic statements. As Damasio (1999) shows, much of
our thinking doesn’t rely on propositions or anything like them. It works
via what he calls “images”, which are not just visual quasi-pictures, but
include all sensory modalities, motor programs and patterns of action that
are meaningful to an organism. Much of this cognition, of course, takes
place beneath the level of conscious awareness and doesn’t consist in
proposition crunching. It also involves qualities, feelings and emotions.
I would argue that we can still use the term “intentionality” for all of
these various dimensions of meaning and thought, just because they all can
involve structures that have some kind of directed character that points
beyond themselves. That is, an image schema (e.g., Source-Path-Goal or
Container) has meaning for us by virtue of the ways it leads to possible
inferences, plans for future action or anticipated future experiences. The
1. Zlatev does not hold this hypothesis as viable anymore. See Zlatev (2003).
image schema points beyond itself, as it is instantiated here and now in this
present experience, to other possible experiences. And even emotional
responses have intentionality, since emotions are part of our monitoring of
how it is going with our bodily states in relation to our flourishing. Emo-
tions involve evaluations of our situation and are part of our acting in re-
sponse to changes in our internal and external milieu.
On our account, then, intentionality is not a mysterious property of dis-
embodied mind, but is instead simply a consequence of the fact that we are
directed, interested, evaluative organisms in ongoing interaction with our
environment. We can select aspects of our experience that have a general
character and use them, together with other such abstractions, to think
about the nature and possibilities of our experience. The patterns of these
interactions can be described as “about” aspects of our experience, insofar
as their significance can transcend the confines of our present concrete
experience and make it possible for us to gather the meaning of our experi-
ence, to evaluate and to plan actions.
You mentioned that the role of intentionality, though important for human
developments in various aspects, cannot characterize human beings, since
other animals have it too. That leads us to the issue of the discontinuity
between human and non-human animals. According to Anderson (2003), a
central tenet in contemporary Cognitive Science is the critique of the Car-
tesian discontinuity between humans and animals. But without positing
some sort of discontinuity isn’t it difficult to understand why only human
beings have language, in the sense of a non-ostensively learned creative
system?
MJ: Once again, we have here a claim that has a kernel of truth, but that is
vastly overblown. Nobody could reasonably deny that language (which
John Dewey (1925 [1981]) called “the tool of tools”) is one of the keys to
what distinguishes humans from other animals. However, you can maintain
this truism without insisting that language marks the Great Ontological
Divide that places humans a little lower than the angels and quite a bit
above the so-called brutes. The use of words to coordinate our actions, to
plan, to bind us into communities, to express our feelings, to develop
knowledge, and to create new meaning is a grand and wondrous accom-
plishment.
Let us remember, though, that language is not the sole repository of
human meaning, conceptualization and thought. Work in the cognitive
sciences has revealed the vast territories of meaning-making that are em-
bodied and non-linguistic, and it has shown some of the ways that the lin-
guistic is grounded in the non-linguistic. Some of these ways we make
meaning are ways we share with other animals. But, once we developed
spoken and then written language, we so vastly increased our capacity for
abstract thought that we began to far outstrip our oxen, our dogs and our
cats (well, maybe) in our ability to understand and control aspects of our
environment. Much of Cognitive Linguistics is devoted to showing how
language is tied to embodied processes of perception and action that are
already meaningful without language. This is not to deny that the emer-
gence of language often comes later to shape our pre-linguistic experience
of meaning.
TR: You know, I’ve often thought that I would like to be able to run as fast
as a cheetah, but I can’t. Discontinuities in evolution are pretty standard
fare. Evolution isn’t always generous in her distribution of the best survival
strategies, and even her unique innovations don’t always work out for the
best either. There are many unique abilities which are part of our basis for
distinguishing organisms from one another; why should human language be
so different from them? Because humans are so special, so unique, so com-
plex, or at the endpoint of the evolutionary process? And as the novelist
Kurt Vonnegut (1985) asked in Galapagos, who is to tell whether evolu-
tion’s grand experiment in big-brained, talking chimpanzees will be suc-
cessful? Although we may seem to be at the moment, drowning in one’s
success remains always a possibility. In short, human language and reason
are no more radically discontinuous than the innovations of other animals;
consequentially it is not difficult to understand why we have them and they
don’t. Our ancestors simply happened upon a series of evolutionary inno-
vations which worked so well we managed to kill off all of our closest
competitor species (see Jared Diamond’s Guns, Germs and Steel), provid-
ing an illusion of evolutionary discontinuity.
So, you both agree that the discontinuity claim is not totally false. The
point is that language seems to introduce something (qualitatively) new in
primate cognition, for instance symbolicity (arbitrariness), systematicity,
hierarchical structure, narrative, extensive self-consciousness. Aren’t they
(largely) novel developments in Homo Sapiens, and doesn’t their emer-
gence depend on human language?
MJ: Everybody recognizes that, once you have language, you vastly in-
crease your cognitive resources, because you can transcend the present
moment in thought and action. Abstraction lets you go beyond the pres-
ently given, back to what has come before, and forward toward what may
come (and over which you might exert some influence). So, clearly, lan-
guage introduces something qualitatively and quantitatively different that
sets humans apart from higher primates and other animals. Maybe we can
teach bonobos to use signs to convey meaning, but the fact that it takes us
years to do this, and with only modest results, makes it plausible to think
that there is something special about humans.
However, we should not think that one of the things that distinguishes
human language is the so-called “arbitrariness of the sign”. Although the
use of a particular word (sign) for a concept might be mostly (but not en-
tirely) arbitrary, there is nothing arbitrary about meaning and conceptual
structure. To cite again the research from Cognitive Linguistics, our
meaning structures are grounded in embodied experience, which highly
constrains what can be meaningful and how it is meaningful. This extends
even to notions of form and syntax, which appear to be tied to the nature of
our embodied experience and are not the result of cognitive modules.
The task for Embodied Cognition theory is to explain the growth of all
forms of human symbolic interaction (including language, music, ritual
practice, architecture, visual art, dance and on and on) as emerging from
ever increasing complexity within the organism. As Dewey put it, in-
creased complexity of functions can result in qualitative changes for the
organism, from the emerging possibility of locomotion, to the capacity for
emotional response, all the way up to the ability to make abstract infer-
ences. An important part of this emergentist story will involve neurosci-
ence, since we will have to discover how ever more complex functions can
develop through neural binding and reentrant loops.
Another vexed question is the relation between concepts/categories and

experience. The best account for the relationship between concepts, cate-
gories and experience in CL is provided in Philosophy in the Flesh (1999).
There Lakoff/Johnson claim that categories are formed as part of our ex-
perience; they are not separate from it. They also say that concepts “are
neural structures that allow us to mentally characterize our categories and
reason about them” (1999: 19). Thought prototypes are not categories, but
the most salient member of them, they are also neural structures that per-
mit inferences to be carried out about categories. Similarly, categories are
not out there in the world, but rather come about because of the way one
ensemble of neurons groups together input patterns and passes them to the
next ensemble. Could you clarify how, according to Lakoff/Johnson, lin-
guistic meanings relate to neurally stored concepts (and prototypes)? And,
at the same time, what the relation is between neurally stored concepts and
gestalts?
MJ: In your formulation of the question, you have given a very nice sum-
mary of how categories are not something separate from experience. Con-
cepts are patterns of neural activation, and categories are concepts that
define the general kinds of things that populate our experience. As such,
categories are the stable neural activation patterns that provide the basic
structure of our shared experience. They constitute distinctions that we, as
a developing species, have found to be important to pursuing our needs,
interests, values and goals. They are not absolute structures written indeli-
bly into the nature of Being; rather, they are the cuts and demarcations in
our shared experience that we have found it most useful to make. Many
(most) of them are not going to change evolutionarily, because some parts
of our bodies and some dimensions of our environments aren’t likely to
change (such as our existing within a gravitational field, or our needing
nourishment, or our being erotically attracted to certain people). They
could change, in some imaginable, though unlikely, future scenario, but
their stability leads us to treat them as fixed givens (which has, unfortu-
nately, led to a mistaken ontology of the world as fixed).
You next raise the question of how “linguistic meanings relate to neu-
rally stored concepts”. Well, I am inclined to say that linguistic meanings
are those communally shared neuronally-realized concepts, which are ge-
stalts, or unified patterns. When I hear or read the word “dog”, this “turns
on” a fairly complex set of neural activation patterns that would include
whatever goes into processing the actual sound of the word, or its written
form, plus a large array of connections called up by that word – all the
related concepts that form a what Fillmore calls a “semantic field” or
“frame”. Some parts of this complicated network are going to be more
strongly activated, such as animal, four-legged, furry and domesticated, as
compared with more weakly activated patterns like “wild cat” and “man’s
best friend”. There will, of course, be associated images (with their dis-
tinctive neural patterns), feelings, emotions and possible metaphorical ex-
tensions connected with this concept as part of its semantics.
Both of you have emphasized that embodiment embodies our cultural life.
Thus concepts are not solely physical. In fact in your answer the concept of
dog is culturally oriented, being man’s best friend, for instance. But this
property is ascribed a less prominent role, as if we first learned what
really matters, and then culture. Could you elaborate on the issue of what
role culture plays in our cognition.
MJ: One of the more frequent criticisms leveled against Lakoff and me is
that we don’t have a place for culture in our account of cognition and
meaning, since we locate meaning in the body. As I have tried to indicate
above, meaning is located in the complex, dynamic arc of interactions that
includes brains, bodies, environments and cultural artifacts and institutions.
A culture involves various symbols, institutions, shared practices, rituals,
values and traditions. Cultures can appear to have an existence independent
of particular people, since so many aspects of culture transcend the living
and dying of individuals. But I want to suggest that culture exists only as
enacted by individuals and groups over time. And this enactment requires
that people take up the practices characteristic of a culture, that they utilize
and interpret its symbols, and that they carry the culture forward in their
lives. Buildings, written languages, paintings, sculptures, musical works,
scientific theories, technical discoveries, machines, tools, clothing and so
forth do not constitute a culture. People have to appropriate these objective
structures and live by means of them, in order to realize culture. So, culture
exists in the interaction, in the living out of meaning, and in the transfor-
mation of experience via what are known as “cultural resources”.
As soon as we start to investigate how these cultural resources (objec-
tive and observer-independent as they might be) shape our lives and our
understanding, then we are back in the domain of studying human under-
standing and cognition. We are back in the realm of embodied cognition,
and we can utilize the resources of Cognitive Linguistics and other parts of
cognitive science to study how we think, feel and act. These methods and
tools will shed light on how cultural artifacts, institutions and practices can
do what they do to shape our existence.
What you just said seems to be at the central core of the cognitive research
program since its very beginning, as can be seen from the following asser-
tion by Lakoff and Johnson which has been cited by many investigators in
the field:
In other words, what we call “direct physical experience” is never

merely a matter of having a body of a certain sort; rather, every experi-
ence takes place within a vast background of cultural presuppositions. It
can be misleading, therefore, to speak of direct physical experience as
though there were some core of immediate experience which we then “in-
terpret” in terms of our conceptual system. Cultural assumptions, values
and attitudes are not a conceptual overlay which we may or may not place
upon experience as we choose. It would be more correct to say that all
experiences is cultural through and through, that we experience our
“world” in such a way that our culture is already present in the very expe-
rience itself (Lakoff and Johnson 1980: 57)
But it is still not quite clear how culture contributes…
MJ: In light of what I’ve just said in answering the previous question, I
hope it is clear in what sense “experience is cultural through and through”.
The vast array of cultural symbols and ways of being and doing are not
added onto a culture-free experience. Young infants are probably good
examples of pre-cultural creatures, but, of course, from the moment of birth
they are being en-culturated or cultivated in the ways of being a member of
a particular culture. Perhaps they start with their animal ways of communi-
cating, but they gradually learn to grasp the meaning of a situation and to
interact with other people via cultural practices, using cultural resources,
such as language. Lorraine Brundige, in her doctoral dissertation on
Swampy Cree Philosophy (2004), explains how the Swampy Cree (of
southern central Canada) understand their world via the enactment of very
specific kinds of narratives that presuppose quite specific views of agency
and causation that are not universally shared by some other cultures. Brun-
dige shows, for example, how the Cree make sense of their identity and
define their values relative to the land they inhabit, so much so that they
will proclaim, “We are the land”. I don’t think that Swampy Cree babies
come into their world knowing this and having it as part of their self-
understanding, since they don’t even have a well-developed self-
understanding in the early months of life. But what I want to say is that, as
they progressively acquire this cultural self-understanding, it is not merely
an “add-on”, not merely an externality imposed on the child’s intrinsic
nature. The intrinsic vs. extrinsic dichotomy is not apt here. The Swampy
Cree child has myriad bodily interactions with their homeland – its vegeta-
tion, water, light, geography, weather, animals – that becomes part of the
meaning of “We are the land”, for that child. The child doesn’t have to
know this consciously in order to live it. So, to say that experience is cul-
tural through and through is to refuse to accept any rigid inner/outer, in-
trinsic/extrinsic, body/culture, individual/social dichotomy.
Let me push this question a little further. Take the notion of a container,
which Lakoff/Johnson claim is one of the most pervasive bodily linked
image schemata we have. However, this may not be universal, or homoge-
neous, since the experience of drinking or eating something may not be the
same as pouring or putting something into a box, or the same as entering a
house or a building. So it appears that there are different schemas for con-
tainer, and they differ in substantial ways: force-dynamics, causality. Frank
(2003) suggests that there is reason to believe that at least in Basque the
body is not conceptualized as a container and, consequently, that the CON-
TAINER/CONTAINED spatial image schemata is not as prominent as in
English. In short, in contrast to the Western mindset, in the Basque ontol-
ogy there is no fundamental ontological separation of “mind” and “body”
to begin with and, hence, to be overcome. Others have suggested that the
prevalence of CONTAINER/CONTAINED could be linked to the dominant
role played by “form” and “matter” in Western thought. Could you expand
the way you understand the relation between our (embodied) cognition, our
bodies and our socio-cultural embodiment?
TR: Although I don’t know Basque, I find it unlikely that they would lack
a container schema altogether. What about the physical interactional level
of embodiment? Surely the Basque people drink out of cups and glasses,
use bottles, carry water in buckets and the like. Just because the container
schema doesn’t map onto the skin boundary of body doesn’t mean they
don’t have the schema. And with respect to ingestion and excretion, I
would expect they must have a sense of it in terms of their physiology as
well – though it might not be one which is elaborated on linguistically.
This all within the realm of possible sociocultural variation, though it is
perhaps surprisingly different – or not so surprising considering how
unique a language Basque is. I see the socio-cultural, the physical interac-
tional and the physiological all as important elements in embodiment the-
ory.
What Frank (2003) meant is that the container/contained image schema
isn’t nearly as prominent in Basque as it is in English, for instance. The
body/mind dichotomy is far less obvious, if there is any at all. Thus, per-
haps the way the body is understood in Basque might also be reflected in
the fact that the word in Basque that would translate as “thought”, con-
cretely, gogo, is already embodied, i.e. it is conceptualized in a way that

would make Damasio proud, with the bodily sensation aspects internal to
the concept – whereas in English thoughts are generally considered to be
“mental” rather than bodily informed. At www.Metaphoric.de/04/frank.
pdf the reader may find a deeper description of this aspect of Basque lan-
guage.
Thank you very much for this opportunity to get both of you together in
elaborating the notion of embodiment. It made clear that this notion must
not be misunderstood as advocating physicalism or some sort of narrow
Embodied Realism. The very complexity of the matter – since embodiment
encompasses various levels of realization and scholars have to tie together
various levels of investigation – explains the difficulties the cognitive sci-
ences still face in addressing issues as consciousness, the general archi-
tecture of mind, the place of culture in embodiment, etc. Our hope is that
this interview made clear that there is still a lot to be done, a lot to be un-
derstood, and probably a lot to be re-done, and corrected, as it is always
the case in science.
References
Anderson, Michael
2003 Embodied cognition: A field guide. Artificial Intelligence 149 (1):
91–130.
Brooks, Rodney A. and Lynn Andrea Stein
1994 Building brains from bodies. Autonomous Robots 1 (1): 7–25.
Brundige, Lorraine
1995 tansi taisinisitohtamáhk kitaskino: Cre Philosophy awa
kagáshkyácimowin (translation: how we understand our world: Cree
philosophy and history). PhD Dissertation. University of Oregon.
Chomsky, Noam
2000 New Horizons in the Study of Language and Mind. Cambridge:
Cambridge University Press.
Chrisley, Ronald and Tom Ziemke
2002 Embodiment. Encyclopedia of Cognitive Science, 1102–1108. Mac-
millan Publishers.
Churchland, Paul
2002 Inner and outer space: The new epistemology. In: Proceedings and
Addresses of the American Philosophical Association 76 (2): 25–48.
American Philosophical Association.
Clark, Andy
1999 Being There: Putting Brain, Body and World Together Again. Cam-
bridge: MIT Press.
Coslett H. Branch, Eleanor M. Saffran and John Schwoebel
2002 Knowledge of the human body: A distinct semantic domain. Neurol-
ogy 59: 357–363.
Damasio, Antonio R.
1999 The Feeling of What Happens: Body and Emotion in the Making of
Consciousness. New York: Harcourt Brace.
Davidson, Donald
1980 Mental events. Reprinted in Davidson, Donald [2001], Essays on
Actions and Events, 206–224. Oxford: Claredon Press.
Dewey, John
1981 Experience and Nature. In: Jo Ann Boydston (ed.), John Dewey: The
Later Works of John Dewey, 1925–1953. Vol. 1. Carbondale: South-
ern Illinois University Press. Original version 1925.
Diamond, Jared
1997 Guns, Germs and Steel. New York: W.W. Norton and Company.
Edelman, Gerald M.
1992 Bright Air, Brilliant Fire: On the Matter of Mind. New York: Basic
Books.
Edelman, Gerald M. and Giulio Tononi
2000 A Universe of Consciousness: How Matter Becomes Imagination.
New York: Basic Books.
Fauconnier, Giles and Mark Turner
2002 The Way We Think: Conceptual Blending and the Mind’s Hidden
Complexities. New York: Basic Books.
Feldman, Jerome A.
2006 From Molecule to Metaphor: A Neural Theory of Language. Cam-
bridge, Mass.: MIT Press.
Feldman, Jerome A. and Narayanan, Srini
2004 Embodied meaning in a neural theory of language. Brain and Lan-
guage 89: 385–392.
Frank, Roslyn M.
2003 Shifting identities: The metaphorics of nature-culture in Western and
Basque models of self. metaphorik.de 4: 66–96. http://www.meta
phorik.de/04/frank.pdf
Gallese, Vittorio
2005 Embodied simulation: from neurons to phenomenal experience. Phe-
nomenology and Cognitive Sciences 4: 23–48.
Hauk, Olaf, Ingrid Johnsrude and Friedmann Pulvermüller

2004 Somatotopic representation of action words in human motor and
premotor cortex. Neuron 41 (2): 301–307.
Hutchins, Edwin
1995 Cognition in the Wild. Cambridge, MA: MIT Press.
Kosslyn, Steven
1994 Image and Brain: The Resolution of the Imagery Debate. Cam-
bridge, MA: MIT Press.
Lakoff, George
1988 Cognitive Semantics. In: Umberto Eco, Marco Santambrogio and
Patrizia Violi (eds.), Meaning and Mental Representations, 119–154.
Bloomington: Indiana University Press.
1980 Metaphors We Live By. Chicago and London: The University of
Chicago Press.
1999 Philosophy in the Flesh: The Embodied Mind and Its Challenge to
Western Thought. New York: Basic Books.
Maturana, Humberto R. and Francisco Varela
1980 Autopoiesis and Cognition: The Realization of the Living. Boston: D.
Reidel Publishing House.
1994 Phenomenology of Perception. Translated by Colin Smith. London:
Routledge and Kegan Paul. Original version 1962.
Minsky, Marvin Lee
1965 Matter, Mind and Models. Cambridge: MIT Press.
Nietzsche, Friedrich
1874 On the use and abuse of history for life. In: Daniel Breazeale, Karl
Ameriks and Desmond M. Clarke (eds.), Nietzsche: Untimely Medi-
tations, 57–124. Trans. R. J. Hollingdale. Cambridge: Cambridge
University Press. Original edition 1874.
Paley, William
1986 Natural Theology: Or, Evidences of the Existence and Attributes of
the Deity, Collected from the Appearances of Nature. Charlottesville,
VA: Lincoln-Rembrandt Publishing. 12th edition. Original edition
1802.
Regier, Terry
1995 The Human Semantic Potential. Chicago: The University of Chicago
Press at Berkeley.
Rohrer, Tim
2001 Pragmatism, ideology and embodiment: William James and the
philosophical foundations of cognitive linguistics. In: René Dirven,
Bruce Hawkins and Esra Sandikcioglu (eds.), Language and Ideol-
ogy: Cognitive Theoretic Approaches. Volume 1, 49–91. Philadel-

phia/Amsterdam: John Benjamins.
Rowat, Peter and Allen Selverston
1997 Oscillatory mechanisms in pairs of neurons connected with fast in-
hibitory synapses. J. Computational Neuroscience 4: 103–127.
Sánchez García, Jesús
2003 Language and cognition: George Lakoff on some internal and exter-
nal complexities. An interview. In: Francisco J. Ruiz de Mendoza
(ed.), Annual Review of Cognitive Linguistics 1: 233–267.
Searle, John
1979 Intentionality and the use of language. In: Avishai Margalit (ed.),
Meaning and Use, 181–87. Dordrecht: Reidel.
1980 Minds, brains and programs. Behavioral and Brain Sciences 3 (3):
417–457.
Vingerhoets, Guy, Floris P. de Lange, Peter Vandemaele, Karel Deblaere and Erik
Achten
2002 Motor imagery in mental rotation: An FMRI study. Neuroimage 17:
1623–1633.
Violi, Patrizia
2003 Embodiment at the crossroads between cognition and semiosis. Re-
cherches en Communication 19: 199–217.
Vonnegut, Kurt
1985 Galapagos: A Novel. New York: Delacourte Press/Seymour Law-
rence.
Wilson, Margaret
2002 Six views of embodied cognition. Psychonomic Bulletin and Review
9 (4): 625–636.
Ziemke, Tom
2003 What’s that thing called embodiment? In: Richard Alterman and
David Kirsh (eds.). Proceedings of the 25th Annual Meeting of the
Cognitive Science Society, 1305–1310. Mahwah, NJ: Lawrence Erl-
baum.
Zlatev, Jordan
2001 The epigenesis of meaning in human beings, and possibly in robots.
Minds and Machines 11 (2): 155–195. http://www.lucs.lu.se/People/
Jordan.Zlatev/Papers/Epigenesis.pdf
Beyond the body: Towards a full embodied semiosis
Patrizia Violi
Abstract
The notion of embodiment has become very prevalent in current research in a num-
ber of disciplines associated with cognitive science such as philosophy, computer
science, psychology, linguistics and semiotics. However, there is no unified theory
of embodiment, only many different uses of the term, each presupposing different
assumptions and conceptual frameworks. This paper reviews and discusses several
of these theories, and the different conceptions of body each implies. It is claimed
that for a fully embodied semiosis, able to account for the role body plays in our
processes of giving meaning to experience, we will need to overcome static, bio-
logical conceptions of the body, and open up to a phenomenological understanding
of it. This will imply taking into account crucial components of embodied experi-
ence not always accounted for within cognitive approaches so far, namely emotion,
affect, subjectivity and intersubjectivity. To fully understand the role of the body in
meaning-making processes, we will then have to, so to speak, go beyond the body
itself.
Keywords: affect, constructivist perspective, emotion, enunciation, experience,

intersubjectivity, Merleau-Ponty, Peirce, phenomenology, semiosis, Semiotics,
situated meaning, subject, subjectivity.
1. Body is not enough: the semiotic body
The notions of body and embodiment have become more and more preva-
lent over the last 20 years, in a number of disciplines associated with cog-
nitive science such as philosophy, computer science, psychology, linguis-
tics. Today, the centrality of the body in human cognition, meaning-making
and experience is broadly acknowledged and this has provoked a huge
quantity of research in this general area throughout a wide range of scien-
tific domains.
54 Patrizia Violi
This is certainly a more than welcome shift in our traditional Western

research paradigm, since it can help free us from the old, seemingly unre-
solvable dualisms between body and mind, between the internal world of
immaterial concepts and thoughts and the external world of objectivist
reality. However, the present widespread use of the notions of body and
embodiment across different fields and with different meanings makes it
particularly important to develop a better understanding and clarification of
these two notions, beginning with a rethinking of the first one, “body”
which sometimes appears to be, paradoxically, the most misleading.
Body is often taken as a “natural” concept, and one which does not need
any further elaboration. Apparently body is something easily accessible,
objective and physically defined. The body seems to be “there”, possessing
an immediate self-evidencing character which does not need to be ex-
plained.
But this is not the case. The body is not a self evident concept, but the
result of the various discourses that construct it. If the phenomenological
experience of the body can appear an immediate one, the concept of
“body” certainly does not. Rather, it appears to be seen in terms of the
construals made of it within any given disciplinary perspective. In other
words, the various meanings attributed to the notion of body are the sum of
the various effects on its sense of the different disciplines as they investi-
gate and define it. The body as described by neurosciences is not the same
body as the one described by psychonanalysis, or by experimental psychol-
ogy, and so on. All these different “bodies” are not reducible to one an-
other; on the contrary they produce a quite “heteroclitic” object, not very
different from how language appeared to be when Saussure first started
describing it. Many of the differences in the use of the very word “em-
bodiment” that I will discuss in this paper depend on the different dis-
courses that construct “body” in their respective ways as an object of re-
search.
So, the first point to be made here is that there is no such thing as a
body “in itself”, naively taken as a given, immediate object of inquiry.
Body cannot be described outside of the different discoursive practices that
define it: to forget this implies the risk of hypostatising the body, as if it
were endowed with an inherent essence, independent of the different prac-
tices, discourses and cultures that shape it. No “hard” science can escape
from this paradox: even the the body as it is described by the most sophis-
ticated technologies – radiography, magnetic resonance imaging and spec-
troscopy, etc. – is not a more basic level of description that reaches some
Beyond the body: Towards a full embodied semiosis 55
more essential hypothetical “structure” of the body, but just another way of
representing it.
Even the body as studied in medicine is a construal, so much so that dif-
ferent medical practices in different cultures construe as many different
bodies as there are cultures: the “Western” body studied in our medical
tradition is not the same as the body mapped by Chinese acupuncture.
This does not mean a denial of the very exsistence of bodies as material
entities, but rather, within a radical constructivist perspective, one which
would have appealed to Peirce, to recognize that we can only reach these
bodies through different practices and discourses, i.e. through semiosis.
“The” body in such a perspective becomes a kind of unreacheable Dynamic
Object, to use Peirce’s terminology, only approachable through a series of
partial descriptions, depending on the particular perspective or disciplinary
approach we decide to take. Such descriptions, which we can consider as
forming part of an open set of Immediate Objects in Peirce’s sense, will not
necessarily converge to form a completely homogeneous picture. Rather
they may continue to remain highly divergent as, for example, in the case
of the phenomenological body we perceive proprioceptively, and the body
as it appears to us on the basis of the results of a laboratory experiment.
Body is, then, a semiotic construal, and this remains the case even when
we attempt to describe its more basic, material levels of organization, such
as neurons or brain synapses, which are certainly “real”, but are not the
body. If we miss this point we risk a curious paradox, which could be de-
fined as “embodiment without the body”. To understand the role the body
plays in processes of producing and understanding meaning, i.e. in semio-
sis, we need much more than this.
In what follows I will discuss the issue of embodiment from a semiotic
perspective, starting with a (very brief) look at some of the main contribu-
tions to be found in this theoretical field, then going on to review some of
the different forms that embodiment has taken in cognitive science, and
concluding with a look at what I believe still remains to be investigated.
That the body plays a major role in semiosis is not a total novelty in
semiotic quarters. Semiotics, like all the other disciplines already men-
tioned, has in its recent developments begun to concern itself more and
more with issues related to the body, and semiotic investigations have also
been started into a related set of problems connected with the role that
feelings, emotions, and sensory and perceptual elements play in meaning
making processes – in a word: the embodied dimensions of meaning. If
such a “corporeal turn” is only quite recent in the post structuralist tradi-
56 Patrizia Violi
tion that gave birth to contemporary generative and narrative semiotics,

this is not the case for the other main tradition in semiotics, i.e. interpreta-
tive semiotics, as it is commonly referred to today, which may be traced
back to the work of the pragmatist philosopher Charles Sanders Peirce.
This is not the place to enter into an in depth discussion of the complex
philosophical approach advocated by Peirce; it will suffice here to mention
just a few points that are relevant for our present purposes. Peirce is often
remembered mainly for his cognitive semiotics, and for his important con-
tributions to the logic of abductive reasoning. However I believe that in his
phenomenology, which is perhaps less well known than his logic, an im-
portant theory of the role of the body in semiosis and a very innovative
intuition regarding the nature of the body-mind relation can be found.
Although Peirce does not thematize in an explicit way the role of the
body in semiosis, it is quite evident that for him, the body plays an impor-
tant role: it would be enough to consider that at the very basis of the semi-
otic processes that enable us to make sense of the world there is, for Peirce,
perception with its bodily based inferential processes. Perception, for
Peirce, far from being an automatic record of external reality, is a highly
constructive process, which requires exactly the same inferential and ab-
ductive devices as abstract forms of reasoning do, while being rooted
firmly in the basic physiological functioning of our bodies. Therefore,
semiosis begins in the body and in its perceptive and proprioceptive proc-
esses.
But this is not the only hint of embodiment we can find in Peirce’s
semiotics. Even more interesting is his theory of interpretants with its im-
plications of a potentially endless process of sign production and interpre-
tation that gives rise to meaning and sense. For Peirce all interpretation
implies an interpretant, which is always a sign, produced from a first, pre-
ceding sign, as its effect. According to Peirce, there are several kinds of
interpretants and more than one classification of these; interestingly
enough the first two levels of interpretation, before arriving at the level of
logical interpretant, which is the cognitive level of concepts, are the emo-
tional and the energetic interpretants. The first is concerned with the emo-
tions signs evoke in us, the second with the muscular bodily reactions they
evoke. Now, all these three levels of interpretants remain active during the
ongoing semiotic process, and this means that even in more cognitively
oriented tasks, such as abstract reasoning, emotions and bodily reactions
are always involved, although with different degrees of relevance with
regard to the specific task and situation in hand.
More generally speaking, Peirce does not conceive the mind as some-
thing qualitatively different from the body or other forms of matter: there
exists a fundamental continuity (referred to in his terminology as
“synechism”) between these, since both share some natural common char-
acteristics, as we can see from the following citation:
We ought to suppose a continuity between the characters of mind and mat-
ter, so that matter should be nothing but mind that had such indurated habits
as to cause it to act with a peculiarily high degree of mechanical regularity
or routine[...]. This hypothesis might be called materialistic, since it attrib-
uted to mind one of the recognized properties of matter, extension, and at-
tributes to all matter a certain excessively low degree of feeling, together
with a certain power of taking habits. (CP 6.277)
In this way body, mind and the world are not only connected, but funda-
mentally interdependent of one another in an endless process of sense
making which reminds us of the dynamics of self organizing systems in an
ongoing developmental relationship between organism and environment.1
The classical dualistic relationship between mind and matter is overcome,
as well as that between the internal and the external world, which are no
longer seen as being dramatically and irreducibly separate from one an-
other. There is mutual interpenetration in all directions.
If the role of the body forms the basis of Peirce’s notion of semiosis,
then the same cannot be said for classical structural semiotics, rooted in the
work of Saussure and Hjelmslev, where a formalistic approach to meaning
was dominant. However in Greimas’ latest works, as well as in the most
recent work by Fontanille2 the mind-body question is reopened, in particu-
lar through a rereading of Merleau-Ponty’s phenomenology.
According to Merleau-Ponty, meaning is in the first place articulated in
our body, through perception. Also for the French philosopher perception
is not merely the simple and passive record of an external world, already
structured and pre-given in its configuration; perception is rather the active
construction of a world already endowed with meaning and intentionality.
Through perception the subject meets the world in the first place and be-
1. For an elaboration of this point, see Coppock (2002), where there is a criticism
of simplistic naturalistic definitions of the notion of body. Also other forms of
embodied mind as in culturally produced material artefacts, bodily borne
protheses, communication devices or other types of new media technologies, all
take part in the continuity of the body-mind-world complex.
2. Cf. Greimas (1987); Fontanille (1999, 2004).
58 Patrizia Violi
gins to give meaning to it. Phenomenological and perceptive meaning is

transformed into linguistic meaning through the corp propre which founds,
at one and the same time, the subjectivity of consciousness and the exteri-
ority of the world. Here we can see another possible compatibility with
Peirce’s philosophy: in Merleau-Ponty’s phenomenology, too, external and
internal world are not separate and in opposition with one another, but
related to each other via the mediation of the corp propre that operates, in
a way, as the translator of perceptually constructed meaning into linguistc
and conceptual meaning.
But the body is also the place where affect and emotion are rooted, as
Freud and psycoanalysis have taught us, reminding us that the Ego is first
and foremost a corporeal Ego. Recent developments in semiotic theory3 are
insistent on the fundamental role emotions play on the very deep level of
sense structuring.
The basic approach to the body that emerges from such a background is
not always consistent with the way in which embodiment has been studied
in other cognitively oriented research domains. What I shall claim in the
present paper is that in order to fully understand the role that embodiment
plays in meaning construction and semiosis, we have, so to speak, to go
beyond the body itself. To develop a satisfactory theory of embodiment the
body is not enough, and we will need to incorporate not only issues related
to action and movement, but also those related to affect and emotion, a
move that will force us to open up to the crucial issues of subjectivity and
intersubjectivity.
At this point, however, it has become vitally important to look more
closely at some of the basic tenets of the notion of embodied cognition as
developed in various areas of the cognitive sciences, in order to see if we
can discover some possible links, overlappings, or differences relative to a
more semiotically oriented approach. In particular I would like to claim the
following: 1) there are today within the field of cognitive studies many
very different notions of embodiment, only some of which are of real theo-
retical interest from a semiotic perspective. It is therefore crucial to distin-
guish between these in order to specify which type of conception of em-
bodiment might be most productive for semiotics; 2) embodiment is related
in an important way to the problem of meaning processes, and it can help
in a decisive way to reframe some of the most controversial questions in
semantics. A context oriented, encyclopedic approach to meaning, which
3. Cf. Greimas and Fontanille (1991).

semiotics intrinsically offers, needs to take into account the role of the
body; 3) as I already suggested, the notion of “body” is not a self-evident
nor simple one, as is too often assumed in contemporary cognitive science;
on the contrary the body is a constructed concept, and as such, cannot be
reduced to purely neuro-physiological aspects nor to the brain. The kind of
body we need to incorporate into our theory of embodiment is more com-
plex than that; it has to be considered in its full phenomenological com-
plexity, as the place where affect and emotions are articulated, and, maybe
more importantly, it must to be tied in with the central issue of subjectivity
and intersubjectivity, a topic not often addressed in cognitive approches to
embodiment.
But it is now time to have a closer look at what is exactly meant by
“embodiment”, and how it might constructively be related to a more spe-
cifically oriented semiotic approach
2. Different embodiments
In very general terms we could say that the main idea behind embodiment
is that mind derives and takes shape from the fact that we have a body that
interacts with our environment. Such an assumption is generally seen as
drastically opposed to classic representational cognitivism, which is based
on functionalism and the computer-mind metaphor. According to function-
alism, mind is independent from its material implementation, as the com-
puter-mind metaphor suggests.
Implicitly connected to this position is a theory of concepts and seman-
tic categories which is generally referred to as the “classic” theory, where
it is claimed that it is possible to arrive at a precise definition of the se-
mantic categories over and above, and independently from, their uses and
contexts of application. In this perspective the body does not play an im-
portant role: it is essentially an output device, as often defined, merely
executing commands generated in the mind through symbol manipulation.
In the embodied perspective, on the other hand, cognition is seen as de-
pending in a fundamental way on the body and its perception and motor
systems, as well as on bodily-based experience and our interactions with
the world.
Before going on to discuss these matters, we must immediately point
out that there is no such thing as a unique theory of embodiment. On the
contrary, the concept of embodiment is a very polysemic one, and different
60 Patrizia Violi
authors use it in quite different ways. Rather than referring to a single the-
ory of embodiment, we ought to refer to different theories of embodiment,
often highly divergent from one another, and sometimes having very little
in common.
So let us now return to the issue of what might be considered the basic
idea underlying the various approaches to embodiment. What exactly does
it mean to say that the mind is embodied, and that it emerges and derives
from the body? If we look more closely, we can see that there are many
different readings of this same thesis, ranging from an extremely weak to
an extremely strong, which is theoretically more interesting, but also more
controversial. It will certainly prove useful to examine these various posi-
tions more closely, since, as has been stated, only some of them will turn
out to be of interest from a semiotic point of view.
A first and extremely weak interpretation would simply imply that all
cognitive processes have a material basis. This is such a generic option that
it would be difficult to disagree with it, but at same time it is so generic
that it is not very meaningful. A more interesting assumption would be to
say that cognitive processes cannot not have a material basis or, in other
words, that cognition is directly connected to the various structures and
biological processes that implement it. A somewhat similar version, still
rather weak, implies that in order to understand mental processes one can-
not ignore the way the nervous system and the brain work. In the last few
decades, both neuroscience and neuropsychology have made such a posi-
tion highly popular, and also widely accepted: today there are probably
very few researchers in cognitive science who would disagree with this
position, with perhaps the exception of few more orthodox functionalists.
From a semiotic point of view, however, this appears to be somehow a
more background type of issue, since a semiotic analysis is not directly
concerned with these more basic levels of description, but rather with the
higher levels of sense organization.
A third interpretation, defined as “material” embodiment (Núñez 1999:
55), also takes into account – in addition to the idea that the mind depends
on underlying neurobiological processes – the constraints imposed on cog-
nition by real-time bodily actions performed by an agent in a real environ-
ment. This is a quite popular position today in robotics, where research is
focused on low-level cognitive tasks such as visual scanning or motion.
Since it has to deal with the construction of robots able to perform real
actions in a real environment, robotics must necessarily develop models of
vision, perception and movement constrained by genuine perceptual-motor
interactions with the environment. Here embodiment means essentially

taking into account the spatial-temporal constraints implicit in real bodies,
but it does not imply any strong theoretical assumptions. Lakoff and John-
son (1999: 37) distinguish here between embodiment as realization and
embodiment as shaping.
Embodiment as shaping, often defined as full embodiment, or radical
embodied cognition, is certainly the more popular position in contemporary
cognitive semantics, and appears to be the one we should look at more
closely from a semiotic point of view. According to this view, all concepts,
even the most abstract ones such as those of mathematics4 are the result “of
the way the brain and body are structured and the way they function in
interpersonal relations and in the physical world” (Lakoff and Johnson
1999: 37).
Notice that in this quote from Lakoff and Johnson, brain and body are
used as substantially interchangeable; this kind of overlapping is found in
many fields of research on embodiment. According to Nunez, for example,
embodiment explains concepts “in terms of the non-arbitrary bodily expe-
riences sustained by the peculiarities of brains and bodies” (Núñez 1999:
56).
This is a crucial question, since there is a potential ambiguity in consid-
ering body and brain as equivalents – an ambiguity that could produce
potentially dangerous levels of confusion. Body and brain are not the same
thing, as the phenomenological tradition, both of Husserl and of Merleau-
Ponty, has taught us, a tradition to which most researchers today seem to
refer. So this would seem to be a vital issue if we want to incorporate an
embodied approach in a serious way into semiotics.
The body is something quite different from the brain, and if the latter
can be seen as an immediate object for scientific study, the body certainly
is not, at least not in any direct and transparent way. Indeed, I have already
made the opposite claim, i.e. that the body is not at all a self-evident con-
cept, as it might appear at a first sight.
For the moment I just want to make salient one specific ambiguity of
this kind which underlies most work on embodiment. While material em-
bodiment refers to the properties of the brain, and, therefore, in this model
the body may be described as a body-brain, when we are speaking of em-
bodied concepts or embodied cognition, a quite different meaning of
“body” is at stake, much closer to the notion of “corporeal schema” than to
4. Cf. Lakoff and Núñez (2000).

62 Patrizia Violi
that of the brain. Although embodied cognition might well have a neural
plane of implementation, we have here two different levels of description,
which do not coincide, and it would be helpful to keep them apart. Semi-
otics, with its phenomenological tradition, might very well play an impor-
tant role in clarifying these issues and distinguishing between these two
conceptual levels, of which only the second is, as I have already men-
tioned, of real semiotic concern.
Within the field of cognitive science, the picture is even more compli-
cated, however, since the new paradigm is pursued within different disci-
plines and by means of different methodological approaches, which do not
all necessarily share the assumptions of cognitive linguistics, not to men-
tion those of semiotics.
To simplify, three main research domains relevant for our present dis-
cussion might be designated: connectionism (and neo-connectionism), ro-
botics and cognitive semantics. These domains do not necessarily share the
same notion of embodiment.
For example, many of the neo-connectionist models which use a dy-
namic modelling approach are not at all necessarily embodied, in the sense
of having systematic, continuous relations with their actual perception and
motor referents. What we have here is rather a conceptual interpretation
that has little to do with empirical perceptive states, as Prinz and Barsalou
(2000) have shown. Connectionist nets do not guarantee embodiment, nei-
ther the radical embodiment of cognitive semantics, nor the weaker notion
of material embodiment.
Situated robotics, on the other hand, as I have already pointed out, has
necessarily to take into account actual bodily constraints, since, in order to
be fully operative the cognitive system underlying a robot must have an
efficient interface with perception and action data: a simple abstract com-
puting system would not be sufficient.
Maybe the main lesson we can derive from situated robotics is that to
perform perception and action we cannot use only the cognitive system
itself, we need also to exploit the resources inherent in the body and the
environment. As Clark (1997: 36) claims, intelligence is not based exclu-
sively on cognitive abilities rather it evolves from the dynamic interaction
between brain, body and world.
The concept of embodiment used in situated robotics is also different
from the one used in the more theoretical fields of cognitive semantics and
contemporary cognitive semiotics, which are crucially concerned with
embodied experience. Both cognitive semantics and semiotics see human
experience as fundamentally bodily based: concepts and cognition emerge

from our experience and are bodily grounded.
To conclude, there are probably more differences than similarities
among researchers who explicitly refer to the notion of embodiment. For
some, the “embodied” mind is still computational in a literal way, for oth-
ers it is not computational at all. Some refute completely the concept of
representation, generally preferring dynamic systems, others, like Barsalou,
refute dynamic systems and still use forms of representation. For some,
embodiment exists only in authentically living systems (and not in simula-
tions, not even connectionist ones), for others this is irrelevant; finally for
cognitive semantics and semiotics the crucial idea is that of phenomenol-
ogical bodily experience.
What then do all these different approaches have in common? Well,
probably the only real unifying aspect to be found is a critical one.
Embodiment theories are essentially a critical reaction to representa-
tional cognitivism, and in particular Fodor’s functionalism. Here, there are
two points of criticism: first, the non-consideration of body-based “mate-
rial” aspects of cognition; second, the reduction of cognitive processes to
purely syntactic symbolic manipulation.
From this point of view, theories of embodiment appear to be a natural
development of cognitive semantics and cognitive linguistics of the seven-
ties and eighties. Theoretical antecedents can be traced back to cognitive
grammars, especially Space Grammar and Mental Space theory5; research
on space and language6 and Force Dynamics, the system of forces that
Talmy (1988) posits as the ground of the linguistic system of modality,
which is essentially derived from embodied structuring.
A fundamental antecedent is also to be found in the critical review of
the classical category theory that goes under the generic name of prototype
theory7.
Since these seminal works first arrived, research in this field has con-
tinued to advance, reframing in a radical way some of its key concepts,
beginning with that of representation.
5. Cf. Langacker (1986); Fauconnier (1985).

6. See, among others, Talmy (1983).
7. It is impossibile to provide even a very concise bibliography on this topic. For a
critical reading of the theory, see Violi (2001).
64 Patrizia Violi
3. Body and situated meaning
The anti-representational controversy is more properly a controversy

against a particular type of representation: symbolic representation, in the
Fodorian sense. Such a criticism, as we will see, is not at all contradictory
to basic semiotic tenets, rather quite the opposite.
Rosch (1999: 62), for example, claims there is a need to distinguish
between two types of representation: the first is a device that mediates
between mind and world, close to Peirce’s idea of semiosis, connecting the
external and internal worlds; the second is based on a notion used in classi-
cal cognitivism, where symbols are seen as syntactic symbols – formal
operations within the closed system of a machine (or a mind, which is
nothing but a machine).
One of the most important differences between these two models is the
different ways they offer for looking at context. Traditional cognitive sci-
ence sees representations as stable, context-insensitive configurations that
cannot be affected by contextual change. The so-called classical theory of
categories was based on precisely such an assumption: a category might be
a node, a network, a set of features, or a mental world, but it was in any
case always a static and immutable entity. In other words the basic idea
was that one and the same invariant structure represented one particular
concept in all possible contexts.
Now such a conception of the matter seems highly problematic: there is
little doubt that natural cognitive systems exhibit a high degree of variety,
and that our functioning in the world is much more flexible than any fixed
structure could describe. Both our behaviours and our mental states adapt
continuously to changing contexts, responding in a highly flexible way to
environmental modifications. The traditional concept of representation thus
turns out to be radically inadequate.
This is not something new in semiotics: similar criticisms of the classi-
cal theory of representation have been developed within a semiotic per-
spective since the Seventies. Umberto Eco in his A Theory of Semiotics
(1976) had already pointed out the fundamental incapacity of any kind of
invariant, dictionary-like structure to represent meaning, and successively,
in 1984, he elaborated further the general notion of the encyclopedia as the
only viable alternative to dictionary based models. From this point of view,
semiotic perspectives, at least those developed within a Peircian interpre-
tative framework, and those of cognitive semantics based on prototype
theory, are certainly highly compatible, as I have discussed elsewhere

(Violi 2001).
At this point, however, my thesis is that developing the issue of em-
bodiment can help us to go even further and to develop a more sophisti-
cated approach to meaning and semiosis, and their relation to context, an
approach that is theoretically more radical than that presupposed in Eco’s
models.
Concepts are indeed sensitive to contexts because we are embodied or-
ganisms and we interact with the environment. Embodiment and interac-
tion are basic features of our semantic system, and more generally, of the
ways in which we make sense of all our ongoing experience.
Taking embodiment seriously in describing meaning can help a semiotic
approach to overcome some of the limitations that can still be found in the
encyclopedic model. Indeed the concept of encyclopedia, as elaborated by
Eco, is a cultural construct that can account, in terms of a regulative hy-
pothesis, for all possible cultural and social components of meaning. How-
ever, it has considerably less to say regarding the phenomenological side of
our experience, although it does not in principle exclude it.
I believe that if something such as a cognitive semiotics is to be estab-
lished as a field of study, it cannot avoid incorporating embodiment in its
basic definition of cognition, and indeed taking this very incorporation of
embodiment as its starting point.
Among the various embodied approaches we can already find some in-
teresting suggestions in this particular direction. Rosch, for example, em-
phasizes the role of situation and context in an embodied perspective. Ac-
cording to Rosch (1999: 72), even when concepts appear to be universal
and abstract, they always refer to specific and concrete situations. Real
situations are events rich in information and should be the real object of
study. Generally speaking, psychology tends to see contextual effects as
negative elements that invalidate experimental work, but this perspective
should be changed, and variations should become the main data for analy-
sis.
Interestingly enough, the adoption of a strong contextualism of this kind
parallels some recent positions in semiotics, where focus has been shifted
from the system, and therefore from structural regularities, to process and
text. The textual turn in semiotics implies making, and considering the text
as the real unit of analysis; this is compatible with Rosch’s positions,
where the single situation is considered to be the correct object of analysis.
66 Patrizia Violi
In both approaches we can find a common holistic component, which in

some semiotic approaches appears to be extremely radicalized.8
Today, Rosch’s broader assumptions regarding representations and the
nature of concepts are quite different from her previous work on proto-
types, and are embedded in a strongly holistic idea of the mind-world
whole. Concepts are now seen as intrinsically non-representational: they do
not have the function of representing the world in the mind, nor do they
mainly have an identifying function, as is generally taken for granted in
experimental research on naming tasks. Rather, concepts participate in
situations.
“Concepts and categories do not represent the world in the mind, they
are a participating part of the mind-world whole” (Rosch 1999: 72). Their
participative nature derives from their being a natural mediation between
mind and world, a mediation which is necessarily anchored into specific
and locally defined situations.
Concepts are the natural bridge between mind and world to such an extent
that they require us to change what we think of as mind and what we think of
as world; concepts occur only in actual situations in which they function as
participating parts of the situation rather than either as representations or as
mechanisms for identifying objects. (Rosch 1999: 61)
Even those who do not share such a radical position would agree to not
conceiving of representations primarily as structures that represent the
external world, but rather as control structures for the regulation of inter-
actions with the external world. This shift from mirror or encoding models
to action-device models is quite common in current research on embodi-
ment.
In robotics, for example, Clark describes representations as control
structures: “The idea here is that the brain should not be seen as primarily a
locus of inner descriptions of external states of affairs; rather, it should be
seen as a locus of inner structures that act as operators upon the world via
their role in determining actions” (Clark 1997: 47).
Representations become here oriented toward action, while at the same
time describing aspects of the world and prescribing possible actions, in a
fine balance between pure control structures and passive representations of
the external world.
With respect to the issue of representation it is worth noticing how
close an approach of this kind is to the basic tenets of Peirce’s pragmati-
8. Cf. Rastier, Cavazza and Abeille (1994).

cism. For the American philosopher, too, concepts (and representations)

are always correlated with actions: while concepts, seen as habits of mind,
have a regulative function in relation to the internal world, stabilizing the
process of unlimited semiosis; on the other hand when operative as beliefs,
they also constitute the basis for behavioral and communicative habits,
which are nothing but regularities in actions. In this way the very same
semiotic structures regulate both the internal world of concepts and beliefs
and the external world of actions, acting as a bridging system between the
two.
A similar idea can be found in the model for memory proposed by
Glenberg (1997: 1–55), where memory does not primarily have a repre-
sentative function “to store the past”, but is rather an embodied device for
facilitating interactions with the environment.
Such a perspective, largely shared among embodiment theorists, focuses
on the role of the larger environment and its interactions with the organism,
and on the relation between external and internal worlds. This explains a
growing interest in Gibson (1979) and his concept of affordances. For Gib-
son, too, representations and internal states that mediate the relationship
with the external world are centred on action, or, to use Gibson’s words,
connected to affordances. Affordances are nothing more than possibilities
for action and use offered by the local environment to a particular type of
embodied agent, equipped with specific bodily features. In this way per-
ception is always contextualized and constructed: the world is essentially
perceived by some given organism endowed with its own intentions in
some given context, and is seen as affording opportunities for goal directed
actions. Perception is therefore always connected to action, and both per-
ception and action are always connected to cognition.
This is a crucial point, because the action-perception-cognition link is
perhaps one of the most important acquisitions of embodiment theories.
Perception is never seen as a passive recording of information, but is im-
mediately connected to action potentials. Therefore any kind of rigid dis-
tinction between perception and cognition disappears, and they become
highly integrated and overlapping processes. Not surprisingly, such an
approach is very interested in results of neuro-physiological studies that
show a connection, even at neuronal level, between perception, action,
thought and imagination. Recent research on mirror neurons have shown
that in primates, and also in humans, the same neurons fire both when a
given action (like grasping a cup of coffee) is effectively executed by some
individual, and when it is observed while being executed by an other, and
68 Patrizia Violi
as well as when the subject merely thinks of executing it. Interestingly

enough, this does not happen just for any kind of movement, only for in-
tentional actions, finalized to a goal (such as grasping a cup), and thus only
for intentional interactions with the environment, or, to use Gibson’s
words: interactions connected to precise affordances.
The existence of underlying schemas common to perception, action,
language and cognition probably represents one of the most challenging
acquisitions of work on embodiment, and it is one that semiotics cannot
ignore, since it implies a highest possible level of integration between all
these systems. Perception, action, language cannot any more be considered
as totally autonomous and independent modules, they must become func-
tional specifications in a common unitary configuration.
This is also the ground of metaphorical concepts, so central in cognitive
semantics, in that they represent linguistic and conceptual projections of
bodily configurations of various kinds (perceptual, motor, spatial, and so
on). Metaphorical projections are always motivated; this is the second im-
portant lesson we can derive from embodiment studies. Together with the
motivational aspect, this offers a radical challenge to the dominant view of
language as a formal system, totally arbitrary and abstract. An important
consequence of this work is a shift from the study of linguistic forms to the
study of linguistic substances, a shift fully shared by contemporary cogni-
tive semiotics. As Petitot suggests:
Il s’agit d’abord de rompre avec l’idéalisme sémiotique à l’œuvre dans les
approches formalistes du sens qui auront dominé la grande période du
structuralisme logico-combinatoire. (Petitot 2000: 84) [What is at stake here
is a break with the semiotic idealism of the formalist approaches to meaning
that dominated the heyday of logic-combinatory structuralism.]
Idealistic formalism has several important consequences: first of all it im-
plies a totally disembodied approach to meaning :
Le sens perd tout rapport au monde naturel externe et au couplage percep-
tion-action qui fonde notre rapport écologique et ethologique à ce monde.
(Petitot 2000: 85) [Meaning loses all relationship with the external natural
world and the coupling of perception and action that grounds our ecological
and ethological relationship with this world.]
Secondly, meaning is deprived of all self-organizing systemic principles
and cannot but be purely logical and combinatory. A semiotic approach
based on embodiment should pursue a double program that we could define
at one and the same time as a de-formalisation and a de-mentalisation of
meaning and sense, reintroducing the study of substance as an essential

part of its project.
4. Intersubjectivity and the embodied subject
The new field of embodiment has brought to light many interesting con-
cepts and questions of central concern for semiotics. Firstly, there is a more
realistic idea of the way human beings perceive and interact with their
environment, and the way in which meaning emerges from these activities.
Next, there is the interconnection between cognition, perception and ac-
tion; the crucial relevance of situations and contexts, and a different and
more articulated idea of the relationship between external and internal
world. Finally, there is the central role of embodied structures in language
and cognition, and the embodied nature of metaphorical mappings. All this
points to a contextualist and pragmaticist conception of semiosis, in the
Peircian tradition, allowing an anti-idealisitic and anti-formalistic shift in
semiotics, such as the one advocated by Petitot.
Embodiment allows and indeed requires a superceding of the purely
logical and formal approach which had characterized semiotic structural-
ism in its initial period of development; meaning ceases to be a purely
negative value, as it has been conceived in the Saussurian tradition, for it
now acquires a living connection with our perceptional, phenomenological
and emotional experience of the world. In this way world, experience, body
and mind will all come to be seen as much more closely interconnected and
strictly related to one another than before, in a way highly consistent with
the Peircean tradition, as I have already indicated.
These are all very important acquisitions. However, there are still a few
points which will need to be more carefully considered, and where I be-
lieve that semiotics will be able to contribute an important series of clarifi-
cations to the wider study of embodiment. Indeed, in research on embodi-
ment, there are some possible “zones of confusion” that appear to be
particularly crucial in our current situation. The first zone of confusion has
already been mentioned and concerns the interchangeable use that is some-
times made of the terms “body” and “brain”. It is important to emphasize
once again the complete lack of coincidence between these two levels: the
body can certainly not be reduced to purely neural forms of activity. A
“body-brain” of this kind would exclude the whole phenomenological di-
70 Patrizia Violi
mension of experience, that live presence that Husserl called Leib, as op-
posed to the material Körper.
The second zone of confusion arises in relation to the distinction be-
tween body and corporeal schema. The confusion is more implicit than
explicit, since corporeal schemas are rarely mentioned, although the notion
might represent a crucial concept for the discussion of embodied experi-
ence. The concept of corporeal schema was first used by psychiatrists and
neurologists towards the end of the nineteenth century, and was then fur-
ther elaborated by Paul Schilder in the mid-1930s (Schilder 1935).
The corporeal schema is not only the general kinaesthetic experience
we have of our body, but it is also the spatial dimension that is occupied by
the body. According to Schilder, it is neither a sensation nor a mental rep-
resentation, but rather something intermediate between these two things.
Merleau-Ponty (1945) refers to the notion of corporeal schema in order to
define the corps propre and its relationship with subjectivity. According to
Merleau-Ponty the notion has a gestalt configuration and a dynamic char-
acter, implying an intentional dimension. The body is always endowed
with a project in the world; it has its own goals deriving from its interac-
tions with the environment.
The notion of corporeal schema seems crucial if we wish to investigate
the embodied grounding of concepts, since at that level what is at stake is
not the “body” as a material and natural object, but its schematic configu-
ration, as has been well demonstrated in studies on spatialisation in lan-
guage. On the basis of this type of embodied configuration, the body be-
comes the first place of meaning articulation, and its embodied schema are
the basic structures that organize meaning, even before language, as I will
discuss in a moment. However, to fully understand the role of embodied
configuration in semiosis, we have first to discuss a very important issue,
related to affect and emotion. Bodily states are always, and at the same
time, pathemic states, endowed and infused with feelings and emotions.
Body is where emotions have their primary space, and if we do not take
this aspect of embodiment into account in our analysis, we miss a crucial
dimension of meaning making, and risk ending up with a totally inadequate
and reduced conception of the body itself.
Affect and emotion are in the body from the very beginning, in all our
sensations and perceptions, which are always permeated by an affective-
emotional tone. We do not only feel sensations of warmth or coldness: we
feel pleasant, unpleasant, or unbearable temperature levels, and the same
also holds for perception: what we see, hear, taste or smell is never “neu-
tral”, but always endowed with some sort of emotional reaction along the
pleasure-displeasure scale. Body is, in other words, never pure “soma”, but
always soma animated by certain affective and emotional states, in other
words: soma and psyche are always simultaneously co-present. Here we
can see that it is precisely the notion of psyche that enables the overcoming
of body-mind dualism, unravelling the categorial distinction between the
two terms.
But this switch from a naturalistic body to a somatic-psychic one also
implies that we must enter into the domain of subjectivity and intersubjec-
tivity. The whole issue of subject and subjectivity is almost completely
absent in the North American tradition of work on embodiment. However
we can in several cases quite easily find implicit reference to something
that we more appropriately would have referred to as subjectivity, but
which is not always recognized as such.
Let us take as an example the otherwise excellent article by MacWhin-
ney (1999), where the author analyses some of the different forms in which
language emerges from embodiment. According to MacWhinney “language
comprehension and production are embodied processes whose goal is the
creation and extraction of embodied meanings […]. We can refer to these
processes of active embodiment as the perspective-taken system”
(MacWhinney 1999: 214).
The embodied perspectival systems operating in language are related to
four levels: 1) affordances, where language and cognition are related to
individual objects and actions through affordances; 2) spatio-temporal
reference frames, which refer to “the set of competing spatio-temporal
reference frames” (MacWhinney 1999: 215); 3) causal action chains, most
centrally involved in the emergence of grammar and the different perspec-
tives of nominative-accusative language or ergative-absolutive language; 4)
social roles, where the perspectival system allows us “to adopt the social
and cognitive perspectives of other human beings” (Mac Whinney 1999:
216).
What is of interest here is that all of these systems are not equivalent in
their relations to the issues of embodiment and subjectivity. If the first
level of affordances is certainly linked to the body and its grounding in the
linguistic perspectival system, since all the properties we can think of in
relation to an object are affordances grounded in the perspective of our
own body, the same does not hold for the other three levels, where it is not
so much the body that plays a role, but the point of view of the subject as
represented in language. Consider the spatio-temporal reference frames.
72 Patrizia Violi
MacWhinney explicitly mentions three alternative frames, an object-

centred, a speaker-centred, and an environment-centred frame. These
frames do not depend on the body, but on the way the position or perspec-
tive of the subject is framed within discourse. The same is true for the
other two systems: both the perspective a given grammatical construction
imposes on the action, and the perspective connected to interpersonal and
social frames, refer to subjectivity more than to embodiment. What we
have in these cases are traces left at the sentence level by the process of
enunciation. The notion of perspective can be framed in the wider issue of
linguistic subjectivity, which, in European post-Saussurian linguistics, has
most convincingly been elaborated in the Theory of Enunciation.9 Such a
theory unifies in one and the same framework a family of heavily intercon-
nected issues, ranging from pronominal, temporal and spatial reference
systems, to focalization, perspective, point of view, and so on.
So obviously the question is not whether or not we use enunciation the-
ory as formulated in post-Saussurian linguistics, but the possible overlap-
pings that may be found between two different issues, both of which are
extremely important. However, they are not necessarily interconnected.
Perspectival systems depend on the presence in every sentence of an un-
cancellable point of view which is the trace of the enunciation process.
This is something quite different to embodiment, which is the existence, in
semantic structures, of motivated configurations, all of which depend on
embodied experience.
Given the extent to which these two issues are not the same, the theory
of enunciation removes the issue of embodiment altogether, leaving only
reference to a transcendental subject, completely deprived of any form of
bodily qualification, gender difference or any other dimension which might
be linked to individual subjects.10 Here we have a deeply paradoxical chi-
asmus: on the one hand there is a theory of embodiment without the sub-
ject, on the other a theory of the subject without a body.
In order to develop a fully embodied theory of semiosis we certainly
need a bringing together of body and subject, and to do this we must de-
velop an approach to subjectivity which is quite different from the tran-
scendental Ego that is implicit in the classical structuralist framework. An
alternative approach of this kind will need to be more firmly connected to
the dynamic dimension of enunicative practices of subjects, and, above all,
9. Cf. Benveniste (1966, 1974 ).

10. Cf. Violi (1986).
to the interplay between the embodied subject and the relational dimension
of intersubjectivity.
Subjectivity is not the emergence of a transcendental subject revealing
himself (and here the masculine pronoun seems more than appropriate), but
rather the emergence of a subjective dimension within a complex, relation-
ally grounded interpersonal, social and cultural environment, in other
words: the realm of intersubjectivity, in which all embodied organisms
necessarily ground their meanings. This implies, in a way, going beyond
the individual subject itself, which cannot manage to exist in any kind of
isolated, solipsistic form, and even beyond the body itself, if considered
merely as an encorporalisation of mind. An embodied subject is more than
a body and more than an individual entity: it is a somatic-psychic organ-
ism, constituted by embodied affect and emotions and inextricably en-
meshed in a complex world of intersubjective relationships.
To exemplify this last point, I will conclude with some, necessarily very
brief, references to my current research on preverbal children. Working on
video of interactions of young children (aged less than 12 months) with
their mothers it becomes strikingly evident how meaning is inherently em-
bodied, in that it emerges from embodied interactions well before it begins
to manifest itself in language. Preverbal babies are already engaged in a
complex work of building meaning on the basis of their interactions with
their environment and the relationships they are involved in with the adults
around them, especially the mother. Their gestures, gazes and movements
can all be read as an already articulated kind of “language”, where the
emotional and mental world of the child manifests itself, not yet through
words but through embodied actions.
It is quite intriguing to notice in analyzing these materials the strong
interconnections that can be seen to exist between the ongoing intermin-
gling of intersubjective patterns – a kind of relational dance involving both
mother and child – and different bodily responses on the part of the child.
In order to understand the process of meaning construction at this very
early developmental stage it would be quite misleading to look only at the
body, without also taking into account the full range of intersubjective
practices within which it is created. Meaning seems to emerge as a series
of bodily and emotional responses to environmental interactions: a kind of
coupling of embodied actions on the part of the individual subject to a
wider pattern of intersubjective relations, a process which might be defined
as a coupling of subjective and objective components of meaning.
74 Patrizia Violi
From its very beginnings the embodied subject, far from being either a
transcendental ego or a purely neural brain, will emerge as the unique way
in which each individual body shapes emotions and feelings in the inter-
subjectivity of relations with the other.
References
Barsalou, Lawrence
1999 Perceptual symbol systems. Behavioral and Brain Sciences 22: 577–
609.
Benveniste, Emile
1966 Problèmes de linguistique générale I. Paris: Gallimard.
1974 Problèmes de linguistique générale II. Paris: Gallimard.
Clark, Andy
1997 Being There: Putting Brain, Body and World Together Again. Bos-
ton, MA: MIT Press.
Coppock, Patrick
2002 Semiotics and the body: C. S. Peirce on the mind-body-world rela-
tion. Versus. Quaderni di Studi Semiotici 93: 135–167.
Eco, Umberto
1976 A Theory of Semiotics. Bloomington: Indiana University Press.
1984 Semiotics and the Philosophy of Language. Bloomington: Indiana
University Press.
Fauconnier, Gilles
1985 Mental Spaces: Aspects of Meaning Construction in Natural Lan-
guages. Cambridge, MA: MIT Press.
Fontanille, Jacques
1999 Polisensorialità e autonomia della dimensione figurative. In: Pierluigi
Basso and Lucia Corrain (eds.), Eloquio del senso. Dialoghi
semiotici per Paolo Fabbri, 188–212. Milano: Costa e Nolan.
2004 Figure del corpo. Per una semiotica dell’impronta. Roma: Meltemi.
Gibson, James
1979 The Ecological Approach to Visual Perception. Boston: Houghton
Mifflin.
Glenberg, Arthur
1997 What memory is for. Behavioral and Brain Sciences 20: 1–55.
Greimas, Algirdas
1987 De l’imperfection. Paris: Pierre Fanlac.
Greimas, Algirdas and Jacques Fontanille
1991 Sémiotique des passions. Des états des choses aux états d’ames.
Paris: Seuil.

Lakoff, George and Rafael Núñez,
2000 Where Mathematics Comes from: How the Embodied Mind Creates
Mathematics. New York: Basic Books.
Langacker, Ronald
1986 Foundations of Cognitive Grammar. Stanford: Stanford University
Press.
MacWhinney, Brian
1999 The emergence of language from embodiment. In: Brian MacWhin-
ney (ed.), The Emergence of Language, 213–256. Mahwah, New Jer-
sey: Lawrence Erlbaum.
1945 Phénoménologie de la perception. Paris: Gallimard.
Núñez, Rafael
1999 Could the future taste purple? Reclaiming mind, body and cognition.
Journal of Consciousness Studies 6 (11/12): 41–60.
Peirce, Charles S.
1934–1948 Collected Papers. Cambridge: Harvard University Press.
Petitot, Jean
2000 Les nervures du marbre. Remarques sur le “ socle dur de l’être” chez
Umberto Eco. In: Jean Petitot and Paolo Fabbri (eds.), Au nom du
sens. Autour de l’oeuvre d’Umberto Eco, 63–102 Paris: Grasset.
Prinz, Jesse and Lawrence Barsalou
2000 Steering a course for embodied representation. In: Eric Dietrich and
Arthur Markman (eds.), Cognitive Dynamics: Conceptual and Rep-
resentational Change in Humans and Machines, 51–77. New Jersey:
Lawrence Erlbaum.
Rastier François, Cavazza Marc and Abeillé Anne
1994 Sémantique pour l’analyse. De la linguistique à l’informatique.
Paris: Masson.
Rosch, Eleanor
1999 Reclaiming concepts. Journal of Consciousness Studies 6 (11/12):
61–77.
Schilder, Paul
1935 The Image and Appearance of the Human Body. New York: Interna-
tional Universities Press.
Talmy, Leonard
1983 Language structures space. In: Herbert Pick and Linda Acredolo
(eds.), Spatial Orientation: Theory, Research and Application, 225–
282. New York: Plenum Press.
76 Patrizia Violi
1988 Force dynamics in language and cognition. Cognitive Science 12:

40–100.
Varela Francisco, Evan Thompson and Eleanor Rosch
1991 The Embodied Mind. Cambridge, Mass: MIT Press.
Violi, Patrizia
1986 L’infinito singolare. Considerazioni sulla differenza sessuale nel
linguaggio. Verona: Essedue.
2001 Meaning and Experience. Bloomington: Indiana University Press.
Properties of cultural embodiment: Lessons from
the anthropology of the body1
Michael Kimmel
Abstract
At present for a genuinely cultural theory of embodiment the first step should be to
bring together cognitive linguistic and anthropological discourses on embodiment.
The specific strength of cognitive linguistics is its analytic tool of embodied image
schemas. However, a cultural approach requires moving towards a more context-
adaptive analysis, as expressed in my notion of situated image schemas. The spe-
cific strength of the anthropology of the body, in particular cultural phenomenol-
ogy, is a contextually situated, qualitative and performative approach that views
embodiment as being-in-the-world. Based on both theoretical strands, I will argue
that cognitive theory should widen its purview (a) by looking at the integral relation
between embodied intentionality, agency and human selves, as well as the cultural
nature of the preconceptual; (b) by exploring “shared” or “distributed embodiment”
between agents; and (c) by modeling the body-discourse relation bi-directionally,
including how discursive imagery is implanted into body awareness.
Keywords: anthropology of the body, cultural phenomenology, distributed embod-

ied cognition, embodiment, image schemas, retrojection, socioculturally situated
cognition.
1. Introduction
Of late few buzzwords have kindled interest across so many diverse aca-
demic disciplines as embodiment, a vogue that has swept through the cog-
nitive sciences, philosophy, several social science disciplines and cultural
studies (Weiss and Haber 1999). Indeed, the term embodiment is on the
verge of generating what theorists of science have called a “theory net”
1. I would like to thank Roslyn Frank for reading several drafts versions of this
paper and for freely giving her advice and constant support.
78 Michael Kimmel
(Balzer, Moulines and Sneed 1987), that is, it acts to bring together several
relationally connected “theory elements” constitutive of a theory core.
However, in order for embodiment to become a viable and tightly knit the-
ory net, there is still one unfulfilled challenge, namely, that of bringing
together the divergent cognitive and cultural approaches to embodiment
under a unifying terminology. This rapprochement already is in the offing
in some quarters (Gibbs 1999; Geurts 2003). In particular, the framework
of “experiential realism” which originated in cognitive linguistics (Lakoff
and Johnson 1999) has many things to commend it as an integrative ap-
proach. Yet, – so I will argue – at least to date this framework: (1) holds a
too limited view of cultural variation in embodied learning and perform-
ance; (2) it demonstrates a narrow view of cultural experience and the pre-
conceptual; (3) it offers no comprehensive model of how cultural discourse
and the body relate to each other; and, finally, (4) it fails to take notice of
embodiment as something frequently involving interactions between cul-
tural agents. My task here is to indicate ways that experiential realism
could incorporate research from the anthropology of the body, particularly
the cultural phenomenology framework (e.g. Csordas 1990, 1993, 1994 a,
b, 1999; Kirmayer 1992, 1993), which allows for a better understanding of
what is “cultural” about embodiment.
This chapter is divided into four major sections. The first section
sketches, from a systemic standpoint, what speaking of cultural embodi-
ment implies. The following section addresses cognitive linguistic re-
search. It argues that only a more situated ontology of its key notion of
image schema will bring out cultural aspects of embodiment. The third
section introduces a phenomenological approach to embodied cultural ex-
perience, as originating in the anthropology of the body. Building on this,
the final section draws attention to several integral aspects needed for a
cultural theory of embodiment. These pertain to the body’s relation to the
self, discourse and collective cognition, respectively.
2. Where is culture in the embodiment literature?
Provisionally defined, the term embodied cognition designates the study of

how cognitive phenomena are informed by the body substrate or by bodily
experience in one way or the other, albeit at diverging levels of observation
and from differing disciplinary perspectives. Where culture sits in all this
Properties of cultural embodiment 79
has not been a major topic in the many recent attempts to bring order into
the vast field of approaches intent on reclaiming the notion of embodiment.
2.1. Surveying recent embodiment approaches
Ziemke (2003) asks what kind of body-substrate forms the precondition of

cognition and if it can also be realized in computational systems. He differ-
entiates physical, organismoid and organismic substrates. Further along,
Ziemke introduces the criterion of depth-in-time, if only implicitly, when
he speaks of “historical embodiment”. This refers to the view that the pres-
ent embodiment (at the level of an organism) is “a result of reflection of a
history of agent-environment interaction” (Ziemke 2003: 3). Notions of
embodiment can then also be distinguished according to how they are time-
related: Do they only refer to the present state of the unit of analysis or do
we need to backtrack into its history?
Rohrer’s (2001: 60–66, 2007) detailed typology surveys ten levels for
deploying notions of embodiment. First, by means of the criterion of disci-
plines and methods he distinguishes phenomenological and cultural views,
linguistic and psychological approaches to the cognitive unconscious, as
well as neurophysiological, neurocomputational and even evolutionary
senses of the word. Second, what is particularly useful in Rohrer’s ap-
proach is his heuristic which groups the analytic units under study by a
scale of magnitude. This ranges from the subcellular via neural regions
(brain) to the individual (mind) and finally the supra-individual levels of
communication and social behavior (collective representations). Cultural
approaches to embodiment are, according to this view, situated at the scale
of 1 meter or higher and include agent interactions and the socio-cultural
system as such.
Consider, as a third position, Wilson’s (2002) catalog of embodiment-
related claims in cognitive science. Her first five claims I would character-
ize as background assumptions rather than directly related to the body: (1)
“cognition is situated”; (2) “cognition is time-pressured”; (3) “we off-load
cognitive work to the environment”; (4) “the environment is part of the
cognitive system”; and (5) “cognition is for action”. Finally, there is, ac-
cording to Wilson, the best documented claim that (6) “off-line cognition is
body-based”. As Ziemke (2003) observes, it is the only claim of hers di-
rectly addressing the body as a physical entity.
80 Michael Kimmel
Do these wider philosophical frameworks model culture satisfactorily,

as ethnographers would understand it? Not really. True, part of the envi-
ronment is cultural, as are the settings that serve to situate cognition; and
much of action is directed to specifically cultural ends. Yet, spoken of ab-
stractly, all of these helpful insights target only generic properties of cul-
ture, culture in the singular, while the counterpart of specific cultures is not
addressed.
2.2. Culture, systemic complexity and structural coupling
Perhaps the philosophical term of structural coupling (Maturana and

Varela 1987) can open a window on the specific complexity any cultural
theory worth its name buys into. The concept of “structural coupling” is a
kind of shorthand that expresses the inescapable attunement of the individ-
ual’s cognitive system to her environment.
The complexities of cultural embodiment reflect different aspects of
structural coupling, overlapping aspects of what constitutes an “environ-
ment”: first, part of our individual environment is manifested in the other
individuals around us reciprocally making us their environment, and shar-
ing it intersubjectively. Second, we must look at the specific ways that
collectives “attune” their environment through technology, symbolic cul-
ture and symbolic action and thus how the latter shape minds and bodies in
turn. Third, environment includes an internal environment, the way we
reflexively perceive our body-selves. And, finally, the collective of internal
environments operates on partly, although never wholly shared dispositions
towards partly, but not wholly shared goals (Strauss and Quinn 1997). All
this adds distinct levels of complexity, subsumed under the label of culture.
While the anthropological notion of culture is a vague and perhaps debated
label, it does remind us of the need to treat these complexities of human
interaction integrally and with ethnographic sensitivity. In this spirit, the
following cultural account of embodiment will inquire into how embodi-
ment becomes phenomenal, intentional and action-directed, self-related, a
matter of shared or distributed interaction, and dialectically related to con-
ceptual knowledge.
3. A cultural perspective on image schemas and embodiment
Cognitive linguistics made its mark with the major insight that an imagery-
based descriptive framework is a crucial methodological resource for
studying embodiment (Johnson 1987). And indeed, its key notion of image
schema has proven to be an interface of great productivity. It reaches out
into neural, experimental and linguistic research dealing with general
structures of analog cognition, while aptly describing cultural gestalt repre-
sentations rooted in FORCE, PATH, CONTAINER, UP-DOWN, BALANCE,
CYCLE, etc. From its inception onward cognitive linguistics has seen itself
as contributing a theory of how conceptual cognition is grounded in em-
bodied image schemas. However, to what extent this theory is, at present,
fully able to address what is cultural about embodiment remains debatable
(cf. Sinha 1999; Kimmel 2002). Recognizing that image schemas are an apt
descriptive tool for any specific analysis, I will contend that this tool has to
be honed further in order to illuminate the cultural aspects of embodiment.
3.1. Image schema = embodied?
Let us put culture into parentheses for now and begin with some general
reflections about what makes image schemas embodied. To what extent
cognition at large may be deemed embodied can be divided up into several
methodologically and theoretically separate question (cf. Gibbs 2003: 13).
There is the claim that cognitive linguistics started out with and which
concerns the grounding of conceptual cognition in the bodily activity of
infants: language and thought are embodied because the primary units of
cognition called image schemas are acquired in kinesthetic experience
(Johnson 1987). Image schemas are then used in metaphorical mappings
and thereby extended to concepts. In this process non-sensory and abstract
meaning becomes grounded in sensory meaning. Developmental data con-
firms the role of image schemas in concept acquisition (Mandler 1992).
Complementarily, linguistic data demonstrates that a vast amount of ab-
stract notions that adults use can be legitimately interpreted as structured
by image schemas (Lakoff and Johnson 1999). However, are these two
sources of evidence sufficient to conclude that cognition is fully embod-
ied?
In order to address this question appropriately we need to go beyond the
previous claims and focus on a distinct, and stronger, embodiment hy-
82 Michael Kimmel
pothesis which targets embodied performance, i.e. the on-line cognition of

adults. Only with regard to the criterion of the immediate bodily activation
in conceptual processing does the Achilles’ heel of the image schema ar-
gument for embodiment become evident. The difficulty lies in a premature
equation of terms. First, evidence that a conceptual image is structured by
image schemas need not be tantamount to showing that the concept pro-
duces a strong sensori-motor resonance in on-line cognition. A phenome-
nological view in particular cautions that image schemas may be frequently
used without substantial co-activation of bodily awareness, if we take Le-
der’s (1990) work on the conspicuous absence of the body in experience
seriously.
Similar conclusions result from neuroscience methods which address
whether conceptual image schemas activate the same neural maps as per-
ception and motor action. PET scans indicate that imagining something in
action-oriented terms recruits elements of the sensorimotor cortex (Barsa-
lou 1999: 579, 585; Gibbs and Berg 2002: 8). However, neural activation
does not prove that what gets activated are the exact brain-state counter-
parts of image schemas. Importantly, “[m]ost scholars agree that motor
processes activated during perception and imagination are always a limited
subset of those activated during overt movement.” (Gibbs and Berg 2002:
8, my italics). Similarly, Barsalou et al. (2003: 4–5) concede that “this
process may range from simulation, to traces of execution, to full-blown
execution”. Presumably, at the far end of the continuum we will find ex-
tremely weak embodied activation.
What about experimental data dealing with analog cognition? Here,
there is good evidence for the on-line simulation of physical settings and
their sensori-motor affordances for an acting self, with ego imaginatively
placed into the simulation. There is also some evidence for the vicarious
experiencing of emotional, affective and proprioceptive states that are in-
herently linked to bodily states in real experience (Gibbs 2003). Yet, what
remains more open to debate is the wider (and more interesting) claim that
conceptualizing abstract entities that are prima facie situated outside the
body and its environment equally involves an on-line simulation of em-
bodied or perceptual states. Although Barsalou’s theory of “perceptual
symbol systems” (1999) spells out many particulars of this possibility,
pending more experimentation we don’t really know how widely analog
states in on-line conceptual processing are embodied. All this seems to
point to the necessity of defining the cognitive nature of image schemas in
real-life settings, of distinguishing situations of various types, and of in-

quiring into their experiential aspects in situ.
3.2. Returning culture to embodied image schemas
Next, we may ask how the aim of culture-sensitivity affects the way that
we think about image schemas. A culture-sensitive viewpoint is tied to a
widened view of meaning, which, as Johnson and Rohrer (Pires de Oliveira
and Souza Bittencourt, this volume: 21) say, is “located in the complex,
dynamic arc of interactions that includes brains, bodies, environments, and
cultural artifacts and institutions”. Yet, at present, several mutually rein-
forcing ontological and methodological assumptions still bias us against a
cultural differentiation of image-schematic embodiment. They do so either
by unduly de-emphasizing cultural variation in embodied learning or by
insufficient attunement to cultural aspects of embodied performance.
3.2.1. The universalist acquisition bias: “Image schemas are, by virtue of

pre-linguistic embodiment in infancy, developmental universals”
Experiential realism has a relatively a-cultural and universalist take on how

primary cognitive forms emerge. What looms large are presumably univer-
sal patterns of bodily experience that developmentally prefigure conceptual
discourse. The grounding of conceptual schemas is either envisaged to
issue from highly transcontextual kinesthetic experiences like FORCE
(Johnson 1987) or from primary scenes – experiential co-occurrences of
metaphorical source and target – like RELATIONSHIPS ARE ENCLOSURES
(Grady 1997). Both accounts involve image schemas and conceive of them
as grounded in early experiences of a universal nature. Yet, we need to
recognize that image schemas are also acquired and refined by culture-
specific practices throughout socialization. Bodily interaction with other
bodies, social space or artifacts as well as bodily participation in rituals and
everyday life substantially flesh out each individual’s image-schematic
inventory.
Ethnographically oriented studies reveal several mechanisms of culture-
specific concept formation that involve the body’s interactions with other
people or the environment. First, acquiring complex image schemas occurs
through the mediation of formative special situations or special practices.
84 Michael Kimmel
Shore (1991) describes ritualized postural techniques that are instrumental

in acquiring the twin concepts mana (“generative potency”, “luck”) and
tapu (“sacred”, “bound”, “set-aside”). When tapu is imposed on people or
objects “in the interest of rendering these people or objects intelligible and
redirecting personal potency for general or cosmic ends” (Shore 1991: 17),
body techniques are involved that can be connected with BINDING,
CONTAINMENT, CENTEREDNESS, RIGIDITY and STASIS schemas that must be
co-activated in body awareness. In another study of cultural learning, Shore
(1996: ch. 3 and 4) discusses how Aboriginal novices, over several years,
distill a complex “walkabout” schema (arguably having to do with CYCLE,
ITERATION and IN-OUT) of geographical and epistemic relevance, from a
multitude of overlapping episodic memories, procedural schemas from
ritual and connected semantic memories.
Beyond these specialized settings, image schema acquisition is also me-
diated by a mix of overt and covert body practices that are ubiquitous in
everyday activities. Bourdieu’s ethnography of Kabyle habitus (1977) does
not speak of image schemas proper, but it does describe a systematic array
of gendered homologies in which a whole system of postures, practices and
social space define OUTWARD and UP schemas as male and INWARD and
DOWN schemas as female. More recently, Geurts (2003) studied the image
schema of BALANCE across contexts in the Anlo–Ewe culture of Ghana.
Finally, acquisition is mediated through the body’s cultural environment
of artifacts or spatial arraying (cf. Toren 1993). According to Sinha and
Jensen de López (2000: 31), children employ social knowledge of the ca-
nonical use of objects in conjunction with their innate capacity for sche-
matizing spatial relations. The image-schematic nature of cultural objects
may be a prototypical ecological affordance that influences language
(Sinha and Jensen de López 2000: 22). Thus, Zapotec children are not as
quick as Danish or English children to notice linguistic differences be-
tween senses of “under” and “in” because they are not encouraged to play
with upright cups, and more generally because Zapotecs use a smaller vari-
ety of containers while tending to use them more multi-functionally.2
2. Although this does not form part of embodied learning proper, acquisition is
also mediated through language itself (Bowerman 1996; Zlatev 1997). For ex-
ample, when categorizing, Yucatec Maya speakers pay more attention to what
something is made of, while English speakers pay attention to its shape (Lucy
1996: 49 ff.). This is probably due to ontological commitments that are embed-
ded in linguistic marking.
The upshot of these studies is that image schemas need to be defined as

inherently culturally mediated and augmented. Hence, concept acquisition
involves a dialectical relationship between bodily dispositions and socio-
cultural practices.
3.2.2. The feed-forward bias: “Embodiment is rooted in general

kinesthetic experience in space, whereby the body constrains
culture, but not vice versa”
Experiential realism highlights particular relations between the body and

conceptual discourse, but downplays others. With a feed-forward logic it
emphasizes that the kinesthetic experiences of the body, notably in infancy,
constrain cultural concepts that develop later. This unidirectional view is
too limited because discourse and cultural practices also shape embodied
cognition in children and adults. One part of the reductionism results from
the view that, when an individual learns, it is necessarily the body that
comes first and that brings forth concepts. However, discourse just as often
plants metaphors into individual body awareness, a point that is examined
in detail in section 5.3.
A related reductionism holds that it is only the body that constrains cog-
nition. Thus, experiential realism most often traces the bottom-up nexus of
how conceptual metaphors are experientially motivated by universal body
physiology. This needs to be replaced by a framework that shows how
cognition is doubly constrained by embodied experiences and by cultural
ideology (cf. Bernárdez this volume). For example, Kövecses’ (2000)
cross-linguistic comparison of emotion concepts indicates that the con-
ceptual metaphor ANGER IS A HOT FLUID IN A CONTAINER is experientially
so well-motivated through blood pressure, body heat and muscle tension
that it occurs in almost all cultures he studied. Yet, the people of Ifaluk in
Micronesia seemingly lack this concept, simply because anger, to them, is
ontologically nothing intra-personal or psychological, but something situ-
ated in the social sphere (Lutz 1988). Apparently, cultural beliefs may con-
strain conceptual metaphors through feedback mechanisms, even when
they are so plausibly motivated by physiology as in the case of anger.
Kövecses is justified in arguing for culturally responsible multi-
dimensional models when he concludes that the “cultural models of anger
and its counterparts are the joint products of metaphor, metonymy, (possi-
86 Michael Kimmel
bly universal) actual physiology and cultural context” (Kövecses 2000:

162; cf. Gibbs 1999; Harkins and Wierzbicka 2001).
In addition to these blind spots which result from a selective view of the
empirical evidence, biases also result from the fact that image schemas
have been endowed with an overly non-situated ontology.
3.2.3. The maximal transcontextuality and schematicity bias: “Image

schemas are what maximally different settings share”
I agree with the view that defines image schemas as structures of cognitive
competence entrenched in long-term memory (Gibbs and Berg 2002; cf.
Johnson’s 1987: 183–190 discussion of the Searlean notion of “Back-
ground”). They acquire their profile not through the specifics of episodes,
but through what many contexts share, and, hence, comprise primary
building blocks of cognition, regardless of how these may combine in any
specific setting. Yet just how transcontextual does the origin of an image
have to be to make it an image schema? The simple image schemas for-
mulated by Johnson (1987) like FORCE or BALANCE only capture schematic
commonalities across the widest possible scope of differing situations.
They are as schematic as our imagination allows, without a trace of con-
textuality. But do image schemas qua schemas need to be maximally sche-
matic entities?3 In other words, is it possible for a limited set of contexts to
produce a more set-specific image schema that encodes how the image
schema is used in a specific type of action? Answering affirmatively, I
propose to go beyond the practice of describing image schemas through
maximally abstract formulas like FORCE, CONTAINER, or BALANCE which
have been distilled from the lowest common denominator of otherwise
highly different experiences. To achieve this aim, I argue in favor of two
important add-ons for the description of image-schematic variants which
are characteristic of a narrower class of experiential settings. Sensitivity for
such variants comes into focus in two ways: (1) from a detailed description
of the specific image-schematic intentionality that a given setting brings
3. The notion of schema per se does not enforce a commitment to maximal sche-
maticity. Schemas are spoken of at various levels of embeddedness. Hence, the
term is legitimate, even if it captures commonalities of a limited set of experi-
ential settings.
into play and, (2) from a characterization of compound image schemas and
their emergent effects in this setting.
Furthermore, there is reason to go beyond image schemas in entrenched
memory and to look at on-line cognition as well. Generally, we may always
take a double perspective on cognition, as non-situated competence and as
situated performance (a.k.a. on-line cognition). Exemplifying this,
Strathern (1996: 188–189) argues that in symbolic healing
[d]emonic possession […] begins with an inchoate (pre-objectified) feeling
of loss of control over the body […]. This is then objectified by a healer in
terms of what Johnson calls the “container schema” and is diagnosed as an
intrusion across a boundary, to be corrected by a suitable form of embodied
action in response. What emerges, then, is something quite particular and
also something comparable to other contexts in which the container schema
is similarly activated. [my italics]
A cultural perspective necessitates a “stereoscopic” view recognizing the
more context-bound as well as the fully transcontextual functions of image
schemas in cognition (Kimmel 2002: 162ff).
3.2.4. The Euclidean imagery bias: “Image schemas can be described

devoid of the intentionality, emotions or entire scenario they are
enacted with”
Image schemas reside in long-term memory. Yet, from another viewpoint

their ontological status is also that of contextual significance bestowing
devices, never actualized as pure idealizations or Euclidean abstractions
(Alverson 1991: 117). Recurrent cultural contexts add something to the
universal form of image schemas and unique experiences may further add
to the specifications that already come with cultural contexts. Gibbs (1999:
154) recognizes this:
containment is not just a sensori-motor act, but an event full of anticipation,
sometimes surprise, sometimes fear, sometimes joy, each of which is shaped
by the presence of other objects and people that we interact with. Image
schemas are therefore not simply given by the body, but constructed out of
culturally governed interactions.
How even simple image schemas are intentionally construed and may be
subject to cultural patterns, as Palmer (1996: 148) shows on the basis of the
Yaqui tendency to construe seemingly static scenes as dynamic. What is
88 Michael Kimmel
more, image schemas become imbued with emotion and motivation to the
degree that they are full carriers of intentionality for a certain kind of con-
text. Palmer (1996: 107, 109) argues that “emotions are complex configu-
rations of goal driven imagery that govern feeling states and scenarios,
including discourse scenarios”. This sits well with Paul’s (1990: 439) defi-
nition of drives as “cognitive mental images already endowed with an af-
fective tone that renders them motivational” (cited in Strauss 1992: 15). To
be emotion-imbued, image schemas must be goal-directed, and situated at
least at the level of some given scenario.
Depending on their situated intentionality and emotional valence basal
image schemas spawn sub-variants, e.g. the difference between conceiving
one’s glass half FULL or half EMPTY. Any simple image schema turns into a
more situated one through its intentional usage in context. Our own body
container, a thermos flask and an all-encompassing metaphysical entity are
not all simply CONTAINERS with the same intentional relation to the body.
Our descriptive ontology of a container will have to go beyond in-out and
boundary dimensions and become sensitive to such striking differences.
Above all, the locus of an image schema must be specified, i.e. whether it
is felt in one’s own body, attributed to the body of a conspecific whom we
can empathize with, projected into a perceptual scene or used in conceptu-
alizing something abstract. Since image schemas rarely occur in isolation,
we also need to recognize that the embodied intentionality is more strongly
connected with holistic experiential scenes (Alverson 1991: 112; Cienki
1997: 7ff), e.g. NEAR-FAR, MERGING and MASS in the experience of seeing
something recede, and not so much with any single image schema.
3.2.5. The micro-unit or primary gestalt bias: “Primary building-blocks of

cognition are ontologically or functionally prior to higher-level
gestalts”
Through its focus on simple, basal image-schematic building-blocks like

CONTAINER, CENTER-PERIPHERY, UP-DOWN, LINK, PATH or BALANCE expe-
riential realism makes these seem ontologically more “real” or at least
functionally more basic than complex gestalts. In fact, when image sche-
mas are combined in complex ways, many authors would no longer refer to
them as image schemas. Consider however that every complex body pos-
ture (not to speak of an extended dance choreography) involves the simul-
taneous activation of numerous image schemas, the spine being STRAIGHT,
the arms in BALANCE, the shoulders UP, the chest a rigid CONTAINER, etc.
Arguably, such a configuration is remembered as a complex image-
schematic gestalt. Compound image schemas may gradually become “psy-
chologically simple” (one of Lakoff’s [1987: 489, 525] criteria for a suc-
cessful image-schematic gestalt) although they are structurally complex.
Compound image schemas may include body postures, action sequences
and ritual, material culture and visual imagery, as well as complex thought
models (Kimmel 2002, 2005).
3.2.6. Single level bias: “There is a single preferred scope of contexts at

which image schemas are stored”
Although image schemas may be entrenched in long-term memory, this

need not imply that embodied experiences impact the conceptual architec-
ture at one given level. Image schemas may be encoded at multiple mental
hierarchies out of which situated or high-level gestalts can be just as easily
generated as simpler ones. Enacting a complex ritual dance in which my
body is perceived in balance, but does many other things at the same time,
will reinforce both the gestalt image of the entire ritual and activate the
basal BALANCE schema that forms part of many other contexts. Assuming
multiple hierarchies is congruent with neurocognitive evidence of sensori-
motor feature maps that are funneled into convergence zones. In these as-
sociative areas “mechanisms outside sensory-motor systems enter into
conceptual knowledge” (Barsalou 1999: 583). This points to processing at
multiple hierarchic levels as well as the meshing with situated knowledge.
3.2.7. The de-contextualized methods bias: “Image schemas emerge from

non-ethnographic or discourse analytic data”
The currently dominant methods for discovering image schemas mirror a

theoretical perspective that sees them as entirely transcontextual entities.
This view typically emerges from synchronic linguistic and laboratory
methodologies and, less so from more context-sensitive discourse-data, and
least from ethnography. More detailed ethnographies cast a somewhat dif-
ferent light on image schemas than the view that children all over the world
acquire them in a roughly comparable way from universal kinesthetic expe-
rience (see 3.1). Moreover, variation within a culture remains inaccessible,
90 Michael Kimmel
in cases where linguists study image schemas by collecting linguistic ex-

pressions across the widest possible number of contexts where, say, MORE
IS UP or PURPOSES ARE DESTINATIONS occurs. Decontextualized distilla-
tions from collections of metaphors or idioms can explore a more general
layer of the cognitive inventory of a cultural community, but not more situ-
ated usages and their intentional particulars. (Of course, metaphor formulas
of the above sort are not designed to capture discourse pragmatics.) What
precisely is embodied about image schemas in an on-line sense invariably
gets lost in this perspective.
3.3. Studying situated image schemas
What about embodiment as rich phenomenal reality and performance? Is

performance too qualitatively saturated to be within the scope of image
schema theory? Not necessarily. An example of symbolic healing, dis-
cussed below in section 4.2, will show that a phenomenological study of
cultural performance often brings to the fore image-schematic scenarios
and that these can in turn be enriched by phenomenological analysis.
For now, let me address some helpful theoretical steps. If we seriously
shift our focus onto performance as documented in discourse and ethno-
graphic data, this will immediately nudge us towards rethinking our non-
situated ontology of image schemas. Studying embodied performance is
tantamount to creating a notion of situated image schemas.
A situated view of image schemas makes sense in the face of a more
general connection drawn between embodiment and the fact that cognition
is inherently situated in environments (Zlatev 1997; Gibbs 1999). As the
active body extends out into and establishes an interdependency relation
with the environment (Pires de Olivera and Souza Bittencourt this volume),
not only universal affordances like standing, running and holding will
move into view, but also the ways in which the environment is culturally
adapted.
Several earlier points taken together contribute to making our analysis
situated. First, in targeting concept acquisition, we must attend to the spe-
cific embodied cultural practices whereby image schemas are acquired and
refined after infancy. Second, in targeting on-line embodiment in everyday
practices, we must differentiate sub-variants of generic-type image sche-
mas by descriptively specifying the embodied intentionality, the emotion
and motivation that typically emanate from the context in question. In dif-
ferentiating, we will also need to stop pretending that this intentionality is

independent of the body-related locus of image schema usage. Intentional-
ity depends on whether an image schema occurs in: (1) one’s own body;
(2) other bodies that we can partly “mirror”, but that remain external to us;
(3) external perceptual objects and events; or (4) wholly abstract notions.
Finally, the kind of analysis I envisage would examine how primitive im-
age schemas combine into compound experiential gestalts. Such a focus
means studying the embodied intentionality of holistically conceived im-
age-schematic scenarios or scenes, because only at the level of whole
scenes we can explain how action-related or conceptual affordances are
created.
A wider notional issue is that a view of experience is needed which
takes into account the inherent transformation of experience by “cultural
resources” (Pires de Oliveira and Souza Bittencourt this volume; cf. Alver-
son 1991, 1994). To date, it remains a paradox of experiential realism that,
despite its early recognition that “all experience is cultural through and
through” (Lakoff and Johnson 1980: 57), it fails to incorporate studies of
culture-specific embodied experience, notably from the anthropology of the
body.
4. Cultural experience and the anthropology of the body
What can cognitivists learn from the anthropology of the body? Scheper–
Hughes and Lock (1987) distinguish three perspectives on embodiment in
social and cultural anthropology. (1) The perspective of phenomenology
(“the individual body”) focuses on the lived body as experience. Marcel
Mauss was the first to embrace this perspective with the notion of “tech-
niques of the body” that constitute triggers for cultural experience. (2) The
perspective of structuralism and symbolism (“the social body”) in the work
of Mary Douglas and Victor Turner. Their research focuses on the human
body as a source of symbolism with which to think about nature, culture
and society. For example, a healthy body offers a metaphorical model of
organic wholeness that is applied to the “social body”. (3) The post-
structuralist perspective (“the body politic”) identifies the body as the
locus of regulatory social practice. Here, Michel Foucault’s history of dis-
cursive formations analyzes the body as an instrument of regulation of the
self in medical, penal, labor, reproductive and sexual systems.
92 Michael Kimmel
Of course, this tripartite distinction points “not so much to three bodies

as to three types of text produced by scholars. In any real event they form a
single system” (Kirmayer 1992: 324).4 However, while symbolism à la
Douglas speaks of the bodily source of representation and Foucault of the
bodily results of representation, phenomenology speaks of a process in-
stead of the body as an object. I will focus on phenomenology here because
of its singular perspective on embodied experience and the “living flesh”
(cf. Pires de Olivera and Souza Bittencourt this volume), which, in turn,
sits well with the present turn to a performance-focus and on-line cogni-
tion.
4.1. Culture within: Proprioception
Before introducing phenomenology, which deals with the cultural experi-

encing and performativity of the body, an approach coming from the eth-
nography of cultural sensoria should be mentioned. While this field tradi-
tionally had more to say about the five external senses than about the inner
bodily dimension, Geurts (2003) breaks new ground by framing embodi-
ment as an inner kind of sensory perception. For the purposes of studying
sensoria, the innovative move here is to highlight processes happening at
the level of proprioception, i.e. the senses of deep tissue, balance, kinesthe-
sia, body displacement and joint position, and thereby going beyond the
limitations dictated by our folk-model of the five senses. For embodiment
theory, classifying how people monitor internal phenomena with percepts
serves to highlight the fact that proprioception, the “inner sense”, is subject
to culture in the same way the externally perceived world is. Propriocep-
tion is not only a universal substrate, but also the locus of culture-specific
ways of monitoring one’s own body.
Geurts’ ethnography of the Anlo–Ewe of South Ghana culturalizes this
inner sense by describing a mode of embodied engagement, specifically the
kinesthetic and proprioceptive schemas relating to what the Anlo–Ewe
cultural theory calls seselelame (“attending to feeling within the body”).
Part-and-parcel of this is a generative principle of perception, thought and
enaction that clusters together variants of the BALANCE image schema.
Learning and maintaining proper bodily balance plays a key role in Anlo–
4. Cognitive linguists have focused more on conceptual relations (2) where the
body remains a metaphorical source for understanding body-external entities
than on lived body experience and performance (1) or on power relations (3).
Ewe life: infants get their joints flexed to develop an awareness for grace-
ful movement; toddlers are exhorted to balance. At all ages, posture and
walking express moral fortitude and psychic disposition. The conceptual
metaphor is both linguistically more varied and performatively more elabo-
rated than Euro-American counterparts such as “to show backbone”. Next,
in ritual the head-balancing of ritual objects is elaborated. What is more,
balance is also perceived as a dynamic relation: diachronic balance sche-
mas determine the embodied dramaturgy of ritual as it alternates between
heated and cool. Likewise, with respect to extra- and introversion an Anlo–
Ewe should achieve a balance between these two modes of being. Finally,
balance is not strictly intra-individual; it also refers to the necessary bal-
ance of the social and cosmic bodies. Living in balance therefore also re-
quires sensitivity to kinship relations, as it were, beyond one’s skin.
Thus, Geurts’ ethnography describes practices of body awareness mani-
fested in metaphors, everyday body habitus and ritual elaborations. Par-
taking of a culture-specific disposition for cultivating proprioceptive im-
agery, Anlo–Ewe BALANCE underlines the importance of differentiating
sub-variants of the more generic image schemas. Here, a generic BALANCE
schema is transformed and refined with regard to a cultural intentionality.
To be sure, this happens in various loci giving BALANCE various functions;
yet these appear as co-determined by an overarching cultural ethos of ap-
proaching one’s body.5
4.2. Being-in-the-world: culture and the preconceptual
Phenomenological anthropology is arguably the field that probes most

deeply into what is uniquely cultural about embodiment. Influenced by
Merleau-Ponty (1962), Csordas (1990, 1993, 1994 a, b, 1999) and Kir-
mayer (1992, 1993) emphasize the nature of embodiment as a cultural
mode of being in the world. This perspective transcends the mode of “rep-
resentation”. It does so by targeting an existential condition and it thereby
establishes a methodological perspective addressing culture and the self
(Csordas 1999: 147). Phenomenology’s aim has been to act as a counter-
5. One drawback of conceiving embodiment as an internal sensory process is that

more entrenched cognitive dispositions are neglected. A second drawback is
that speaking of a proprioceptive “sense” makes us think more of what is per-
ceived, and less of a bodily self that enacts reality. In this respect we have to go
further.
94 Michael Kimmel
balance to the exclusive focus on representational systems (or, in cultural

studies, on what is expressed by the metaphor of “texts”). Culture, con-
ceived as a semiotic fabric or lattice of public symbology, tells only half
the story. Our aching, craving or sick bodies remind us that there is a sec-
ond intrinsic order to experience; the conceptual dimension has a pre-
conceptual counterpart situated in the body. Phenomenology offers a dia-
lectical partner to textuality and representation. Merleau-Ponty was the
first to suggest that the “body’s influence in thought is more presentation
than representation, given in substance and action than imagination and
reflection” (Kirmayer 1992: 325). According to Merleau-Ponty, meaning
cannot be reduced to signs that represent, i.e. that stand for an entity exter-
nal to the signs themselves. Instead, some meaning is isomorphic with ex-
perience. Phenomenological anthropology is thus directed against the clas-
sical representationalist bias in epistemology harkening back to Descartes’
disembodied mind which resides in a transcendental ego. Phenomenology
specifies one point that cognitive linguistics has left very vague and to its
own disadvantage, namely, the nature of the preconceptual. For Csordas,
the preconceptual is much more than the raw material of later cognition.
Rather, the preconceptual body, in itself, brings forth cultural intentional-
ity. By consequence, calling an experience preconceptual does not neces-
sarily imply that it is also precultural.
The notion of somatic modes of attention as discussed by Csordas
(1993: 138) is a key for understanding how culture shapes pre-objectified
experience. Somatic modes of attention encompass culturally elaborated
ways of attending to one’s own body, to the bodies of others, and to other
people’s attention to our body. Pathological somatization disorders such as
hyper-vigilance in hypochondria or tolerance for self-mortification in ano-
rexia and bulimia involve somatic modes of attention. Culture also shapes
the way somatic modes of attention attribute special significance to bodily
processes such as pregnancy or menopause.
In summary, phenomenology is most interested in the interstitial zone
between body and concepts and the processes occurring there. Accord-
ingly, Csordas’ (1994 b, 1999) notion of embodied imagery emphasizes
that mind and body are intertwined and all imagery is to some extent em-
bodied. In many respects Csordas might have subsumed the notion of im-
age schema under this term, if only the preconceptual were better illumi-
nated in image schema theory and if we had a better model of how
preconceptual and conceptual modes interact in culture.
5. Three properties of embodied culture
Having set forth this review of the contributions of cultural phenomenol-

ogy, I would now like to elaborate three major theoretical aspects that I see
as integral to a theory of cultural embodiment: (1) intentionality, the self
and the performativity of the cultural body; (2) supra-individual embodi-
ment; and (3) the placement of discursive imagery into the body.
5.1. Intentionality, self and performativity
Speaking of modes of being-in-the-world has several advantages over other

views of embodiment. First of all, being in the world emphasizes the tie
with intentionality, that is, attending to and taking up the world. In this
perspective, embodied perception and cognition are inherently directed
towards action and prepare us for action. A good example from perception
is stepping on an escalator that unexpectedly remains immobile. We expe-
rience being slowed down since we intentionally had expected the escala-
tor to jolt us forward and had adapted our body to that expectation. Inten-
tionality directed at action is not confined to the individual. It may be
shared (see below) and emerge in cultural interaction.
Being-in-the-world inherently also engages the self, a recognition that
further increases this notion’s scope compared to non-phenomenological
theories. Instead of a disembodied mind that mediates between perception
and cognition, what takes place inside the living flesh is an interface be-
tween external stimuli, what we know, and, more fundamentally, what we
are. Another way to express this is that the proprioceptive sense is directly
tied up with the knowledge of being an integral body and thus an entity that
is distinct from others and endowed with a body image as well as a center
of existential awareness. Neurological impairments illustrate this deep
nexus to the self dramatically. Sacks (1986) describes the case of a woman
who, upon losing her sense of joint position, was on the verge of losing her
self-identity. Another patient repeatedly tried to toss his leg out of his bed
because he could not feel it and believed he had an alien appendage. This
sense was so strong that it overrode his visual knowledge that the leg was
attached to his trunk. Another kind of transformation of the self is frequent
in the case of those suffering from chronic pain, those who often end up
foregoing all attempts to communicate their experience to others (Scarry
1994). In a similar fashion, the rationale of torture is to destroy the self by
96 Michael Kimmel
inducing unbearable pain (Daniel 1994). Kafka’s (1948) short story The
Penal Colony epitomizes with remarkable premonition the Foucauldian
notion of “inscribing knowledge” via the flesh into the self. The verdict, at
first incomprehensible to the delinquent, is engraved in his back with nee-
dles over many hours until he dies in a state of embodied epiphany of his
guilt. There is ample evidence of the collective nature of such techniques
of inscribing, inaugurated by Foucault’s work on bodily power regimes in
the clinic, the prison and sexuality.
Another strength of speaking of modes of being-in-the-world is its em-
phasis on the inherent meaningfulness of embodied performance. “Modes
of action and ways of life” are the substrate of embodied metaphors (Kir-
mayer 1992: 380). One aspect of phenomenology that cognitive linguists
also emphasize is that the embodied imagination frequently projects itself
into the conceptual world. The flipside of the coin, and one that is often
neglected by cognitive linguists, is that the body may enact culture without
substantially engaging in conceptual representation. For instance, Scheper–
Hughes (1990) reports a mass-syndrome of involuntary seizures and trem-
bling legs among exploited female Brazilian sugar-cane workers. She in-
terprets this as a shared, collectively embodied manifestation that could be
expressed as: “I cannot carry my burden any longer so that my legs falter.”
Scores of other culture-bound collective syndromes – Indonesian amok,
Victorian hysteria, Western anorexia, female nervios or susto in Latin
America, and many more (Csordas 1994 a) – display what Dreyfus and
Dreyfus call “intentionality without representation” (1999: 110ff) and may
be interpreted as enacted metaphors of the body occurring at a yet precon-
ceptual level.
5.2. Supra-individual embodiment
Beyond embodied dispositions that are culturally shared between individu-

als, embodiment can be genuinely collective in the sense of being interac-
tive. Csordas emphasizes this supra-individual dimension by expressly
including in somatic modes of attention both our attention to the bodies of
others and the others’ attention to our own body. For example, the social
contexts of dance, team sports and sex involve distributed embodiment in
both of these senses. Considering the ways in which distributed embodied
cognition occurs, it seems to me that this notion encompasses both sym-
metrical (i.e. either fully mutually attuned or partially shared) as well as
more asymmetrical and therefore complementary modes of action or expe-

riencing.
Symmetrically attuned somatic modes of attention are present in cul-
tural techniques of “consubjectivity” (Csordas 1993: 144). For example, in
pulse diagnosis of the Siddha medical system of South Asia before the
healer gives a diagnosis, he must enter a state of synchronization of his
own pulse with the client’s and experience shared pulsations. Melanesian
couvade, i.e. men experiencing birth pain when their women are in labor,
can be understood as attuned bodily sensations and therefore as embodied,
rather than mere imitation or charade. Consubjectivity also occurs in trans-
fers from client to therapist in psychoanalytic therapy. In these and many
other contexts, we may say that modes of collective “selfing” via the body
occur.
Let us now look at embodied states that are asymmetrically distributed
among agents committed to a common social goal.6 Here the dissimilar
somatic modes of the participants contribute to a joint experience. This
means that differing somatic modes and actualizations of embodied im-
agery contribute to an integral event in such a way that each participant
role interlocks with the other and provides feedback for it. For example,
symbolic healing typically distributes somatic states both between the ex-
pert healer and the lay client, as well as over the stages of the ritual. Csor-
das’ (1990) study of a Charismatic Pentecostal healing session illustrates
this very well. The ritual aimed at casting out evil spirits starts with the
clients in the congregation sensing a particular and distressing thought,
emotion or behavior outside their control, but without them knowing ex-
actly what it is. This requires the intervention of a healer, an expert in ob-
jectification, who diagnoses their distress as a case of spirit possession.
The healer reifies the congregation’s pre-conceptually embodied experi-
ence by introducing the cultural concept “demon”. Only at the stage when
the clients accept this conceptually objectified image suggested by the
healer, do they confer a more explicit causal interpretation on the experi-
enced loss of control. The healer uses the semantic label demon and the
associated field of conceptual metaphors to set into motion a range of per-
formative acts for “expelling” the demon.
6. This is somewhat analogous to Hutchins’ (1995) finding that complex collective

tasks like navigating a ship are typically achieved because cognitive resources
are distributed between agents (and artifacts) in specific configurations.
98 Michael Kimmel
Specifically, the healer objectifies these anxieties as evil forces intrud-

ing from the outside, and by classifying them as “possession”, the healer
has his client dissociate the source of distress from the individual’s inside.
In cognitive linguistic parlance, this ascription of external agency draws on
a scenario of image-schematic metaphors in which the self is a bounded,
but semi-permeable CONTAINER, evils are intruding AGENTS that exert cap-
tivating FORCE, and healing takes effect as symbolic EXPULSION through
recourse to a more powerful AGENT (God). By applying this cultural model
of causal attribution, the healer is able to lift the burden of guilt from the
client.
Ritual healing typically displays both of the aspects we have just dis-
cussed: a sequential and an interpersonal distribution of somatic and con-
ceptual states. Initially, a client experiences subjective distress at the level
of body awareness. Upon this, an expert performs metaphoric predications
on the client’s inchoate experience to help her turn it into an objectified
body image (typically drawing on a conceptual model like demonic posses-
sion). Finally, the causal inferences emanating from the healer’s actions
feed back onto the client’s body awareness and even her physiology. In
short, Csordas’ example highlights the inherent two-way street that links
embodied states and conceptual ones (body awareness problem => body
imagery altered/created =>body awareness improved).
Thus, embodied states in wider social settings may be distributed, both
interpersonally and over time. For instance, concerning social roles like
healer and patient, we need to study the role-specificity of jointly orches-
trated embodied states, which may range from more bodily to more objec-
tivized states. We also need to cultivate a sensitivity for “feedback loops”
in which embodied states of an individual trigger conceptual imagery and
vice versa in, as it were, a cycle of objectivization.
5.3. Embodied learning and imagery: Projection, retrojection and

mimesis
The literature is replete with treatments of the relation between discourse

and the body (e.g. Featherstone, Hepworth and Turner 1991; Coupland and
Gwyn 2003). Among these, Bourdieu’s habitus theory (1977) has become
influential for its recognition of the dialectic between discourse and the
body, and for eschewing subjectivism and objectivism as false alternatives.
The notion of habitus illuminates the “circular process whereby practices
are incorporated within the body, only to be regenerated through the em-
bodied work and competence of the body” (Crossley 2001: 126). Social
constructivism and cognitive theory have also highlighted the dialectic
interplay between instituted models in the social sphere and mental models
(Berger and Luckmann 1967; Shore 1996; Strauss and Quinn 1997), a fact
that logically extends to embodied models. This interplay has been empha-
sized also by cognitive linguists such as Sinha (1999) and Harder (1999)
who point out that neither the embodied nor the discursive grounding of
cognition make sense in isolation.7
How can an approach rooted in imagery take this dialectic into account?
By assuming that bodily states create conceptual states, the view of em-
bodied realism operates rather unidirectionally. Johnson’s (1987) feed-
forward or bottom-up nexus – perhaps best glossed as “projection” – takes
a developmental perspective and emphasizes how embodied image sche-
mas provide basic units of discourse. This view remains silent on how dis-
course, ritual and material culture may conversely shape, refine and re-
combine basic image schemas and turn them into cultural experiences.
Quite plainly, it often happens that discourse “goes under our skin”. This
occurs whenever discursive imagery is taken in and appropriated by the
body. Although the notion of image-schematic mapping appears to be emi-
nently suitable for explaining this appropriation into the body, the preva-
lent feed-forward focus on image schemas has remained silent on it. Thus,
Johnson’s projection view needs a feedback counterpart explaining how
individual body awareness becomes a map onto which discursive imagery
is inscribed. For this reason I propose retrojection as an apt term to de-
scribe situations in which cultural metaphors are picked up in discourse
and then mapped back into the body. Retrojection is a process whereby
discursively objectified body images or other symbolic associations reso-
nate with proprioceptive body awareness and thus come to be felt inside
the body. Such embodied sensations may be triggered by speech, symbolic
action or visual symbols and may manifest themselves in muscle tonus,
kinesthetic readiness, metabolic flow, focus of somatic attention, relaxation
or arousal.
The retrojection of words or symbols into the body can account for
situations where instructors use metaphor to encourage embodied and
emotional experiencing. In “How words move people to dance” Felton
7. Cf. Frank (this volume), who incorporates the systemic back propagation from
the wider environment as an alternative to the linear, feed-forward framework.
100 Michael Kimmel
(2004) studies contact improvisation, in which teachers extensively use

varied metaphors to make dancers experiment with new kinds of bodily
experience. One teacher envisions the body as a container that holds the
dance to encourage an awareness of interiority and boundedness and then,
in counterpoint, as something permeable from which dance pours forth and
into which it trails back. Through metaphors of connectedness and reach-
ing out, the teacher seeks to create a continuous field of dance so that the
dancers may feel continuity with the room as well as with the others. An-
other teacher uses water imagery extensively and depicts dancing as a con-
tinuous flux and the bodies as “pouring through” their partners. Similar
patterns of infusing the body with experience through the use of metaphor
may be found in fields as varied as body therapy or other healing systems,
meditation in yoga, qi gong, or tai chi or even in military drill. While not
everything about body movement or body habitus is learned through lin-
guistic metaphors, we should become very sensitive to the many subtle
ways that cultural discourse shapes or refines embodiment.
In addition to such specialized instructional practices, the concept of
retrojection enlarges our view of conventional conceptual metaphors that
people live by. Of course the fact that conceptual forms may frequently
have systematic bodily counterparts is predicted by cognitive linguists. The
embodied and conceptual double-nature of metaphor is nicely exemplified
by morality metaphors in English. As discussed by Lakoff and Johnson
(1999: ch.14), MORALITY IS PHYSICAL STRENGTH and MORALITY IS RE-
SISTING A PHYSICAL FORCE, both of which relate to EVIL IS A FORCE. Phe-
nomenological introspection suggests that these conceptual images of force
mirror preconceptual qualities sensed in the body. After all, “resisting”
immoral temptations and keeping “baser” impulses “in check” can be felt
as sapping one’s strength. (In the special case of BEING GOOD IS BEING
UPRIGHT / DOING EVIL IS FALLING the embodied implication may be that
morality requires resisting the “pull” of gravity by applying force verti-
cally.) MORALITY IS PHYSICAL STRENGTH8 has a preconceptual dimension
insofar as the awareness of moral strength may be felt inside one’s body
container. Note also that conceptually realizing one’s immorality may go
with an embodied feeling of weakening.
8. Locating morality in the body is conceptually manifest in the MORAL ESSENCE

metaphor of the self, where character is evaluated in terms of bodily essence
(“He’s rotten to the core”).
But how should we interpret these correspondences between body

awareness and conceptual images? We cannot know a priori whether, in an
individual learner, a specific discursive metaphor is brought forth by an
embodied feeling or vice versa. Both, the retrojection of discourse and the
projection of pre-discursive embodied experiences, offer a plausible ac-
count. Grady’s (1997) theory of primary metaphor exemplifies the projec-
tion account. Here conceptual mappings in adulthood are prefigured by
several dozens of embodied co-occurrences in infancy. Hence, the proto-
typical infant experience of RELATIONSHIPS ARE ENCLOSURES linguistically
surfaces later as “I’m in this marriage”. As I see it, primary metaphors in
infancy alone cannot shed light on the acquisition of specific somatic
modes. They often underspecify how people embody morality in ways like
MORALITY IS STRENGTH.
The notion of retrojection may serve to fill in this gap in the current
embodied explanatory paradigm. It makes the body-discourse loop genu-
inely bi-directional. For example, growing up, a child will internalize cul-
turally appropriate body feelings by hearing linguistic metaphors over and
over again. A child with a strict upbringing will begin to enact what the
parental morality metaphors exhort it to do by “showing backbone”,
“keeping her chin up”, standing “upright”, keeping her poise, not flinching,
and/or trying to “pull herself together”. Often, young children may be
partly familiar with a conceptual metaphor but may not have experienced
its full implications yet in their own body awareness. Thus, retrojection
may infuse the imagery of cultural metaphors into the individual’s body
and thus let people feel the power of discourse within.
Clearly, the explanatory avenues of retrojection (discourse => body)
and projection (body => discourse) aren’t mutually exclusive. That a per-
son first hears discursive metaphors and only then achieves their embodied
resonance need not clash with a (partial) experiential motivation in more
fundamental primary metaphors. A fuller view would assume that moti-
vated, but underspecified body feelings enter into an elective affinity with
cultural discourse, through which their embodiment is further specified in
contextually appropriate ways. In this way, specific somatic modes of at-
tention resonate with systems of conceptual metaphor.
Finally, we should mention mimesis (Bourdieu 1977; Taussig 1993;
Maran 2003) as an equally indispensable explanatory avenue that usually
goes hand in hand with retrojection. Mimesis bypasses the discourse-body
nexus by mapping from the perceived body of another person to ego’s body
directly. To illustrate mimetic learning, a person who notices that individu-
102 Michael Kimmel
als with a radiant and strong character tend to stand upright may mimic this
postural habitus.
Together, retrojection and mimesis have empirical and methodological
implications we should seriously consider. Future research projects should
involve interdisciplinary teams of linguists and anthropologists who could
study how children – or, for that matter, soldiers, priests or company em-
ployees – acquire body habitus and/or somatic modes of attention. Since all
three mechanisms will usually interact, any study of embodied learning
will need to study retrojection (learning through exposure to discourse) and
mimesis simultaneously, while also holding in focus our linguistically de-
rived background knowledge about shared primary conceptual mappings
which retrojection and mimesis refine or transform. Perhaps this call for a
more complex and more integrated perspective amounts to a reinvigoration
of Bourdieu’s approach to embodied knowledge. At the same time, the
approach I advocate has a more cognitive bent. It benefits from imagery
theory and cognitive linguistic methods to the fullest, while also incorpo-
rating a strong phenomenological sensitivity (cf. Bernárdez this volume;
Crossley 2001).
6. Conclusion
This chapter has surveyed the mutually complementary analytical perspec-

tives needed for a genuinely cultural theory of embodiment. The indisput-
able merit of experiential realism in cognitive linguistics is its recognition
that an imagery-based framework is an indispensable resource for describ-
ing embodiment in detail. Image schemas also constitute an important no-
tional interface between disciplines by reaching out into neural, experi-
mental and linguistic research while also being apt for addressing cultural
facets. However, I have expanded this perspective. Acquisition studies
caution against a too universal view of embodiment through image sche-
mas, as these may be culturally refined. Moreover, from a viewpoint that is
interested not only in schematic structures but also in a degree of socio-
cultural situatedness, we will have to focus on image schemas combined in
holistic scenes rather than on maximally schematic micro-units of experi-
ence. To move part of the way towards performativity-oriented views like
cultural phenomenology, I have called for a descriptive apparatus that dif-
ferentiates experiential characteristics of particular activity types like dance
and symbolic healing. The notion of situated image schemas meets this
goal by describing the intentional, emotion-imbued and goal-directed sub-

variants of more generic schemas, as manifested in image-schematic sce-
narios or compounds.
A full theory of situated embodiment will also have to enlarge the scope
of image schemas with regard to their diverse functional roles. Here lies
the merit of cultural phenomenology. It highlights dimensions that a fully
situated perspective must bring into play, specifically, the fact that em-
bodiment oftentimes deals with what happens between people (its inter- or
consubjective nature), its deep involvement with cultural selves, the fact
that the pre-conceptual is cultural in itself, and the fact that we have to look
at embodiment not only from a developmental perspective but also by tak-
ing into account bodily performance in everyday environments.
In conclusion, it is encouraging that cognitive research is now turning to
an on-line and situated perspective on embodiment. This perspective
clearly dovetails with the existing anthropological sensitivities for perfor-
mativity, lived experience and context (Gibbs 1999; Frank 2004; Sharifian
this volume). The joint impact of these developments broadens our under-
standing of the subject matter of embodiment and makes it more and more
legitimate to speak of a socio-cultural embodiment view in cognitive re-
search. In my opinion, the current challenge of theorizing cultural embodi-
ment requires that we hold in focus complementary dimensions of cogni-
tion and that we do so at several levels: individual and collective cognition;
entrenched dispositions in memory and lived experience; cognitive com-
petence and cognitive performance.
References
Alverson, Hoyt
1991 Metaphor and experience. Looking over the notion of image schema.
In: James Fernandez. (ed.), Beyond Metaphor. The Theory of Tropes
in Anthropology, 94–119. Stanford: Stanford University Press.
1994 Semantics and Experience: Universal Metaphors of Time in English,
Mandarin, Hindi and Sesotho. Baltimore, MD: Johns Hopkins Uni-
versity Press.
Balzer, Wolfgang, Carles Ulises Moulines and Joseph D. Sneed
1987 An Architectonic for Science: The Structuralist Program. Dordrecht:
Reidel.
104 Michael Kimmel
Barsalou, Lawrence
1999 Perceptual symbol systems. Behavioral and Brain Sciences 22: 577–
609.
Barsalou, Lawrence W., Paula M. Niedenthal, Aron K. Barbey and Jennifer A.
Ruppert
2003 Social embodiment. In: Brian H. Ross (ed.), The Psychology of
Learning and Motivation: Advances in Research and Theory, 43–92.
San Diego, CA: Academic Press.
Berger, Peter L. and Thomas Luckmann
1967 The Social Construction of Reality. A Treatise in the Sociology of
Knowledge. London: Penguin.
Bernárdez, Enrique
this vol. Collective cognition and individual activity: Variation, language and
culture
Bourdieu, Pierre
1977 Outline of a Theory of Practice. Cambridge: Cambridge University
Press.
Bowerman, Melissa
1996 Cognitive versus linguistic determinants. In: John J. Gumperz and
Stephen C. Levinson Gumperz. (eds.), Rethinking Linguistic Relativ-
ity, 145–176. Cambridge: Cambridge University Press.
Cienki, Alan
1997 Some properties and groupings of image schemas. In: Marjolijn
Verspoor, Kee Dong Lee and Eve Sweetser (eds.), Lexical and Syn-
tactical Constructions and the Construction of Meaning, 3–15. Am-
sterdam: John Benjamins.
Coupland, Justine and Richard Gwyn (eds.)
2003 Discourse, the Body and Identity. Basingstoke: Palgrave Macmillan.
Crossley, Nick
2001 The Social Body: Habit, Identity and Desire. London: Sage.
Csordas, Thomas J. (ed.)
1994 a Embodiment and Experience: The Existential Ground of Culture and
Self. Cambridge: Cambridge University Press.
Csordas, Thomas J.
1990 Embodiment as a paradigm for anthropology. Ethos 18 (1): 5–47.
1993 Somatic modes of attention. Cultural Anthropology 8: 135–156.
1994 b Introduction: The body as representation and being-in-the-world. In:
Thomas J. Csordas (ed.), 1–24.
1999 Embodiment and cultural phenomenology. In: Gail Weiss and Honi
F. Haber (eds.), 143–162.
Daniel, Valentine E.
1994 The individual in terror. In: Thomas J. Csordas (ed.), 229–247.
Dreyfus, Hubert and Dreyfus Stuart

1999 The challenge of Merleau-Ponty's phenomenology of embodiment for
cognitive science. In: Gail Weiss and Honi F. Haber (eds.), 103–120.
Featherstone, Mike, Mike Hepworth and Bryan S. Turner (eds.)
1991 The Body: Social Process and Cultural Theory. London: Sage.
Felton, Lori
2004 How Words Move People to Dance. Ms.
Frank, Roslyn M.
this vol. The language-species-organism analogy: A complex adaptive sys-
tems approach to shifting perspectives on “language”.
2004 What do John Lucy's “unitizer” languages tell us about topological
image schemas? Presentation at the “Workshop on Image Schemas
and Linguistic Relativity: Empirical Methods in Cognitive Linguis-
tics” held in conjunction with the International Conference on Lan-
guage, Culture and Mind, Portsmouth, UK, July 17th, 2004.
Geurts, Kathryn Linn
2003 Culture and the Senses: Bodily Ways of Knowing in an African
Community. Berkeley: University of California Press.
Gibbs, Raymond W.
world. In: Raymond W. Gibbs and Gerard Steen (eds.), Metaphor in
Cognitive Linguistics, 145–166. Philadelphia ; Amsterdam: John
Benjamins.
2003 Embodied experience and linguistic meaning. Brain and Language
84: 1–15.
Gibbs, Raymond W. and Eric Berg
2002 Mental imagery and embodied activity. Journal of Mental Imagery
26 (1/2): 1–30.
Grady, Joseph E.
1997 Foundations of meaning: primary metaphors and primary scenes.
Unpublished Dissertation, University of California.
Harder, Peter
1999 Dual grounding of language in the physics and physiology of the
body (causal grounding) and in the social-interactional context
(functional grounding). In: Theo Janssen and Gisela Redeker (eds.)
(1999). Cognitive Linguistics: Foundations, Scope and Methodol-
ogy, 195–222. Berlin: Mouton de Gruyter.
Harkins, Jean and Anna Wierzbicka (eds.)
2001 Emotions in Crosslinguistic Perspective. Berlin: Mouton de Gruyter.
Hutchins, Edwin
1995 Cognition in the Wild. Cambridge, Mass.: MIT Press.
106 Michael Kimmel
Johnson, Mark
Kafka, Franz
1948 The Penal Colony: Stories and Short Pieces. Trans. Willa and Edwin
Muir. New York: Schocken Books.
Kimmel, Michael
2002 Metaphor, imagery and culture. Unpublished Ph.D. Dissertation,
University of Vienna.
2005 Culture regained: Compound and situated image schemas. In: Beate
Hampe (ed.), From Perception to Meaning: Image Schemas in Cog-
nitive Linguistics, 285–311. Berlin: Mouton de Gruyter.
Kirmayer, Laurence J.
1992 The body’s insistence on meaning: Metaphor as presentation and
representation in illness experience. Medical Anthropology Quarterly
6 (4): 323–346.
1993 Healing and the invention of metaphor: The effectiveness of symbols
revisited. Culture, Medicine and Psychiatry 17: 161–195.
Kövecses, Zoltán
2000 Metaphor and Emotion: Language, Culture and Body in Human
Feeling. Cambridge: Cambridge University Press.
Lakoff, George
1980 Metaphors We Live By. Chicago: University of Chicago Press.
1999 Philosophy in the Flesh: The Embodied Mind and its Challenge to
Lakoff, George and Mark Turner
1989 More than Cool Reason: A Field Guide to Poetic Metaphor. Chi-
cago: University of Chicago Press.
Leder, Drew
1990 The Absent Body. Chicago: University of Chicago Press.
Lucy, John A.
1996 Language Diversity and Thought: A Reformulation of the Linguistic
Relativity Hypothesis. Cambridge: Cambridge University Press.
Lutz, Catherine
1988 Unnatural Emotions: Everyday Sentiments on a Micronesian Atoll
and their Challenge to Western Theory. Chicago: University of Chi-
cago Press.
Mandler, Jean M.
1992 How to build a baby II: Conceptual primitives. Psychological Review
99: 587–604.
Maran, Timo
2003 Mimesis as a phenomenon of communication. Sign System Studies 31
(1):191–215.
Maturana, Humberto R., and Francisco J. Varela
1987 The Tree of Knowledge: The Biological Roots of Human Under-
standing. Boston ; London: Shambhala.
1962 Phenomenology of Perception. London: Routledge and Kegan Paul.
Palmer, Gary B.
Press.
Paul, Robert
1990 What does anybody want? Desire, purpose and the acting subject in
the study of culture. Cultural Anthropology 5 (4): 431–451.
Pires de Olivera, Roberta and Robson de Souza Bittencourt
this vol. An interview with Mark Johnson and Tim Rohrer: From neurons to
sociocultural situatedness
Rohrer, Tim
2001 Pragmatics, ideology and embodiment: William James and the philo-
sophical foundations of cognitive linguistics. In: René Dirven, Bruce
Hawkins and Esra Sandikcioglu (eds.), Language and Ideology. Vol.
1: Theoretical Cognitive Approaches, 49–81. Amsterdam ; Philadel-
phia: John Benjamins.
Sacks, Oliver
1986 The Man who Mistook his Wife for a Hat. London: Picador.
Scarry, Elaine
1985 The Body in Pain: The Making and Unmaking of the World. New
York: Oxford University Press.
Scheper-Hughes, Nancy
1990 The rebel body. The subversive meanings of illness. TAS Journal 10:
3–10.
Scheper-Hughes, Nancy and Margaret Lock
1987 The mindful body: A prolegomenon to future work in medical an-
thropology. Medical Anthropology Quarterly 1: 6–41.
Sharifian, Farzad
this vol. Distributed, emergent cultural cognition, conceptualisation and lan-
guage
108 Michael Kimmel
Shore, Bradd
1991 Twice born, once conceived: Meaning construction and cultural
cognition. American Anthropologist 93 (1): 9–27.
1996 Culture in Mind: Cognition, Culture and the Problem of Meaning.
Oxford and New York: Oxford University Press.
Sinha, Chris
1999 Grounding, mapping and acts of meaning. In: Theo Janssen and
and Methodology, 223–255. Berlin: Mouton de Gruyter.
Sinha, Chris and Kristine Jensen de Lopez
2000 Culture and the embodiment of spatial cognition. Cognitive Linguis-
tics 11 (1/2): 17–41.
Strathern, Andrew
1996 Body Thoughts. Ann Arbor: University of Michigan Press.
Strauss, Claudia
1992 Models and motives. In: Roy G. D'Andrade and Claudia Strauss
(eds.), Human Motives and Cultural Models, 1–20. Cambridge:
Strauss, Claudia and Naomi Quinn
1997 A Cognitive Theory of Cultural Meaning. Cambridge: Cambridge
University Press.
Taussig, Michael T.
1993 Mimesis and Alterity: A Particular History of the Senses. New York ;
London: Routledge.
Toren, Christina
1993 Making history: The significance of childhood cognition for a com-
parative anthropology of mind. Man 28 (3): 461–478.
Weiss, Gail and Honi Fern Haber (eds.)
1999 Perspectives on Embodiment: The Intersections of Nature and Cul-
ture. New York: Routledge.
Wilson, Margaret
9: 625–636.
Ziemke, Tom
2003 What’s that thing called embodiment? In: Proceedings of the 25th
Annual Meeting of the Cognitive Science Society. Mahwah NJ: Law-
rence Erlbaum.
Zlatev, Jordan
Stockholm: Gotab.
Distributed, emergent cultural cognition,
conceptualisation and language
Farzad Sharifian
Abstract
This chapter presents an integrative view of cognition as a system that emerges

from the interactions between the members of a cultural group. Members of a cul-
tural group negotiate and renegotiate their emergent cultural cognition across time
and space. Emergent cultural cognition is the locus of cultural conceptualisations
such as cultural models, cultural schemas and cultural categories. Another integral
aspect of emergent cultural cognition is language in that human languages largely
embody the cultural conceptualisations of their speakers.
In terms of their representation, cultural cognition, cultural conceptualisations
and language are heterogeneously distributed across the minds in a cultural group,
rather than being equally imprinted in the mind of each individual. Overall, cultural
cognition and language appear to reveal properties of complex adaptive systems.
This chapter elaborates on these notions and provides examples of cultural con-
ceptualisations and their instantiations in various aspects of human languages.
Keywords: complex adaptive systems, cognitive linguistics, cultural cognition,

cultural conceptualisations, cultural models, emergent cognition, heterogeneously
distributed cognition, schema.
1. Introduction: The locus of cultural cognition
In classical circles of cognitive psychology the word “cognition” has

largely been associated with mind and mental activity. Different paradigms
within cognitive psychology have, however, not agreed upon the nature of
the human cognitive system. Proponents of what came to be known as
classicism (Newell 1980) viewed cognition as a symbolic system whereas
advocates of connectionism (Davis 1992) viewed cognition as emerging
110 Farzad Sharifian
from interactions among networks of interconnected processing units

called neurons.
The notion of cognition did not remain a focus only within the field of
cognitive psychology but attracted the interest of scholars from other disci-
plines such as biology, linguistics and anthropology. This led to the devel-
opment of the mega-discipline called “cognitive science” and also to the
emergence of sub-fields such as cognitive anthropology and cognitive lin-
guistics.
One of the natural consequences of the development of interdisciplinary
approaches to the study of cognition was a revisiting and in fact expansion
of the notion of cognition. Not all scholars within the areas of cognitive
science have been interested in finding out about what goes on inside the
mind of an isolated individual. Some have been more inquisitive about
population-level and group-level correlates or consequences of cognitive
processes. The expansion of the notion of cognition took place along sev-
eral lines. One group of scholars took interest in the interaction between
the human mind and the environment. Hutchins, an anthropologist and a
cognitive psychologist, and his colleagues, for example, observed that hu-
man cognition constantly interacts with an environment that is rich in or-
ganizational resources (Hutchins 1994). For Hutchins, cognition is distrib-
uted across individuals, tools and artefacts.
Another departure from the limited scope of cognition in traditional
cognitive psychology has been equating cognition with action (see
Bernárdez this volume) as well as activity that is socially situated. In an
introduction to a field guide, Anderson (2003: 91) states that:
For over fifty years in philosophy, and for perhaps fifteen in Artificial Intel-
ligence and related disciplines, there has been a re-thinking of the nature of
cognition. Instead of emphasizing formal operations on abstract symbols,
this new approach focuses attention on the fact that most real-world thinking
occurs in very particular (and often very complex) environments, is em-
ployed for very practical ends, and exploits the possibility of interaction
with and manipulation of external props. It thereby foregrounds the fact that
cognition is a highly embodied or situated activity – emphasis intentionally
on all three – and suggests that thinking beings ought therefore be consid-
ered first and foremost as acting beings.
The above quote clearly highlights two directions in which the notion of
cognition has been expanded, that is, “situated” activity and “embodi-
ment”. The embodiment thesis, in general terms, views cognition to be
mediated by our bodily experience. The exact relation between the body
Distributed, emergent cultural cognition, conceptualisation and language 111
and human cognition and the interpretations given to the word “body”,
however, have varied from overlapping views to conflicting and contrast-
ing ones (e.g., Violi 2003, this volume; Wilson 2002). In this context,
again, different interpretations of the notion of “cognition” have had epis-
temological consequences for how the notion of “body” has been viewed
and for the role that has been attributed to it in relation to cognitive activi-
ties (see more in Violi 2003, this volume).
Another dimension along which the notion of cognition has been ex-
panded is the dimension of culture. Scholars with interest in both cognition
and culture have been exploring how culture and cognition interact with
each other and with other systems such as language (e.g., Cole 1996;
D’Andrade 1995; Hutchins 1994; Shore 1996; Strauss and Quinn 1997,
1995; Tomasello 1999). As in other approaches to the study of cognition,
various scholars in this area have not totally agreed on the nature of the
relationship between culture and cognition or even on what constitutes
culture and/or cognition. For some, cognition is an aspect of culture in that
culture influences various cognitive processes (e.g., Altarriba 1993;
Redding 1980). Sperber and Hirschfeld (1999: cxv) view the relationship
between culture and cognition along two dimensions, reflected in the fol-
lowing statement:
The study of culture is of relevance to cognitive sciences for two major rea-
sons. The first is that the very existence of culture, for an essential part, is
both an effect and a manifestation of human cognitive abilities. The second
reason is that the human societies of today culturally frame every aspect of
human life, and, in particular, of cognitive activity.
Within the paradigm of cognitive linguistics many subscribe to the view of
Langacker (1994), namely, that culture is primarily a cognitive phenome-
non, with individual minds as its locus. Langacker, however, acknowledges
that not all aspects of culture are represented in the human mind.
2. Emergent cultural cognition
I maintain that “cognition” may also be viewed as a property of cultural

groups, and not just individuals. I refer to this level of cognition as emer-
gent cultural cognition in the sense that what is being described as cogni-
tion here is an emergent system (e.g., Johnson 2001) resulting from the
interactions between the members of a cultural group across time and
space. This of course does not confine the scope of culture to the cognitive
domain. Emergent cultural cognition may be instantiated in various aspects

of people’s lives including aspects of their physical environments, arte-
facts, tools, rituals, painting, dance, etc.
Cultural cognition is heterogeneous in the sense that it is heterogene-
ously distributed across the minds in a cultural group. The distribution of
cultural cognition extends across the dimensions of time and space. Mem-
bers of a cultural group negotiate and renegotiate their cultural cognition
across generations, vertically, and, horizontally, through a multitude of
communicative events. The notion of cognition here encompasses complex
systems that are dynamic and ever evolving, rather than a fixed set of rep-
resentations that extend to a cultural group. Cross-sectionally, the notion of
distributed, emergent cultural cognition may be diagrammed as Figure 1.
Figure 1. Distributed, emergent cultural cognition.
This simple figure is perhaps the closest visual depiction that can be of-
fered of distributed, emergent cultural cognition. In this figure the top part
represents the “global” cultural cognition that emerges from the interac-
tions between the members of a cultural group while the lower part is
meant to represent the way in which cultural cognition is distributed “lo-
cally” across the individual minds of the group members. The overall fig-
ure here reflects how emergent properties of cognition at the group level
supersede what is represented in the mind of each individual. It should of
course be kept in mind that emergent properties arise from the interactions
between the group members, a process that does not lend itself readily to a
static visual sketch. A crucial point that needs to be kept in mind when
viewing Figure 1 is that the distribution that is being modelled extends to
the dimension of time, a diachronic aspect of cultural cognition that cannot
be visually shown in a simple two-dimensional picture. Another point that
is meant to be reflected in Figure 1 is that members of a cultural group
might share some but not every aspect of their cultural cognition with other
members and the pattern is not exactly the same for all individuals across
the cultural group (see Borofsky 1994), that is, two members may share
more from their cultural cognition than others. In other words, as men-
tioned earlier, cultural cognition is heterogeneously distributed across the
members in a cultural group.
The above-mentioned view of distributed cognition is an initial step in
the direction of constructing the type of ideational account of culture that
Keesing (1987: 371) had in mind when he said: “An ideational theory of
culture can look at cultural knowledge as distributed within a social sys-
tem, can take into account the variation between individuals’ knowledge of
and vantage points on the cultural heritage of their people.” It is this varia-
tion between individuals’ knowledge of cultural conceptualisations that my
use of the term “heterogeneously distributed cultural cognition” is intended
to highlight. It should be stressed here that I do not view the ultimate level
of cultural cognition in terms of fixed representations inside the mind of
individuals but as emergent properties resulting from the interactions be-
tween members of a cultural group. This conception of distributed cogni-
tion seems also to be implied in Kronenfeld’s (2002: 430) statement that
“culture has no existence outside of our individual representations of it,
and since these representations are variable, there exist no single place
where the whole of any culture is stored or represented. Thus, culture is
necessarily and intrinsically a distributed system.” Kronenfeld also ob-
serves that culture is not merely fixed knowledge, but productive repre-
sentations of a growing repertoire capable of generating new responses to
novel situations that still make sense to cultural groups. Such a view of
cultural cognition constitutes a challenge for “cultural determinism” in that
it allows for individual differences while acknowledging the existence of
collective cognition. Cultural orientation, from this perspective, is seen as a
continuum rather than either/or membership.
In terms of consciousness, members of a group may be conscious of the
influence that a particular “collective” cognition has on their thought pat-
terns and behaviour and in fact may try to opt out of it. What is at issue
here is that even in those cases, the individual is very likely to recognize
certain knowledge or conceptualisation to be characteristic of the culture
they belong(ed) to. Cultural cognition is usually the basis for many aspects
of our actions and behaviour in two senses: one is that our behaviour, in-
cluding our linguistic performance, largely derives from our cultural cog-
nition, and second is that we largely operate on the basis of the assumption
that other interactants’ behaviour draws on the same cultural cognition. In
general we may say that cultural cognition serves as the basis for the “hy-
potheses” that people make regarding what they encounter during their
cultural experience.
The above-mentioned view of cultural cognition is at least partly con-
sistent with certain versions of other expansions of the notion of cognition.
Hutchins (1994), for example, also views cognition as “distributed”,
though in a slightly different sense. Hutchins (e.g., 1994), mainly empha-
sizes the distribution of cognitive processes and includes the material envi-
ronment within the domain of cognitive processing. I emphasise the emer-
gent nature of cultural cognition, which is primarily cultural knowledge,
and I use the term “distributed” in conjunction with the term “heterogene-
ous” to highlight the view that cultural cognition is not equally imprinted
in the minds of the people in a cultural group. Despite these differences in
the focus of research, the two strands should be viewed as complimentary,
particularly given the fact that Hutchins acknowledges that cognition is a
cultural process (see also Lindbloom and Ziemke, BLM Volume 1).
The notion of cultural cognition presented here is also consistent with
the version of embodied cognition which regards “body” as a constructed
notion (see Violi this volume). Whatever the role of body in our cognitive
life, it should be kept in mind that conceptualisations of “body” may be
culture-specific and in general body takes part and acts as a conceptual
resource for our cultural experience. Even the number of senses that we
assign to our bodies may vary across different cultures. On the other hand,
the situations and contexts implied by the notion of “situated cognition”
are in fact largely social and cultural. Anderson (2003: 126) also stresses
the importance of the role of culture in situated and embodied cognition,
maintaining that:
Along with research in situated cognition, EC [embodied cognition] further
suggests that intelligence lies less in the individual brain, and more in the
dynamic interaction of brains with the wider world – including especially the
social and cultural worlds which are so central to human cognition – and
therefore suggests that fields like sociology and cultural studies can them-
selves be important resources for (and in some guises are part of) the cogni-
tive sciences.
2.1. Emergent cultural cognition as a complex adaptive system
It is to be noted at this point that emergent cultural cognition may be

viewed as a complex adaptive system (e.g., Frank this volume; Waldrop
1992) in that it has the properties that are generally associated with com-
plex systems. One of the main attributes of complex adaptive systems is
that they reveal emergent properties. As mentioned earlier, cultural cogni-
tion is also an emergent system in that it results from the interactions be-
tween the members of a cultural group across time and space. The emer-
gent properties of cultural cognition as a system at the global level (cf.
Frank this volume), are not mirror images of those that characterize the
cognition of each individual within the group.
A closely related property of complex systems is that the parts consti-
tuting the system cannot contain the whole. In this sense, also, cultural
cognition is a complex system in that an individual’s cognition does not
capture the totality of their cultural group’s cognition. Furthermore, when
analyzing the case of cultural cognition, we find that its control is distrib-
uted throughout the group; rather than it being subject to centralized
mechanisms of control.
Another characteristic of complex systems is that they are nested. That
is, the agents that are components of the system are themselves complex
adaptive systems. Similarly, members of a cultural group, as agents of cul-
tural cognition, are themselves complex systems, controlled by nervous
systems, endocrine systems, etc. Like other complex systems, cultural cog-
nitions have their own unique history of interactions that constantly con-
struct and reconstruct the system. Often small changes in the interactions
of cultural groups have had a remarkable influence on the future direction
of their cultural cognition. This view is largely reflected in the writings of
Vygotsky (e.g., Vygotsky 1978), who viewed cognitive phenomena as em-
bodying the characteristics of historically bound sociocultural relations.
One of the characteristics of complex systems is the difficulty involved
in determining their boundaries: they are “open systems”. The decision is
usually based on the observer’s needs and prejudices rather than any intrin-
sic property of the system itself. This aspect of complex systems also ex-
tends to cultural cognition in that the boundaries as to where one cultural

group ends and another begins are difficult, if not impossible, to determine.
In relation to cultural cognition, as is the case with other types of com-
plex systems, the role of an individual agent can be viewed as two-fold. On
the one hand, the individual is the locus of cultural cognition and can have
an initial causal role in its development, dissemination and reinforcement.
On the other hand, an individual’s performance can be influenced or de-
termined to a varying degree by the cultural cognition that characterizes the
cultural group. Thus, the role of individuals in a cultural group may be
described in terms of a circular pattern of cause and effect.
At this point, I would like to focus on conceptualisation and language as
two integral aspects of cultural cognition. The whole field of cognitive
linguistics is based on the assumption that various aspects of language
embody conceptualisation of experience. While it is acknowledged that the
locus of language and conceptualisation is the individual, the two ulti-
mately emerge at the cultural level of cognition. This thesis will be ex-
plored further in the following sections.
3. Cultural conceptualisations: Cultural models, categories and

schemas
Human conceptual faculties, which might be largely universal and innate,

derive from various sources of experience, including bodily and environ-
mental ones, that in turn enable new experiences to be made sense of and
organized. Such experiences lead to the development of our conceptual
knowledge, which is both complex and systematic. The units of organiza-
tion in our conceptual knowledge, such as categories (e.g., Rosch 1978)
and schemas (e.g., Arbib 1992; Bartlett 1932; Bobrow and Norman 1975;
Mandler 1984; Rumelhart 1980), appear to be based on certain associations
that may help us tell them apart from each other. Robinson (1997: 263)
maintains that such associations reflect “regularities in an organism’s per-
ception of and interaction with its environment”. He considers schemas and
categories to be higher-level representational networks that store concep-
tual relationships rather than simple stimulus-response patterns. He notes
that “all of these schemata, categories and other conceptual relationships
are probabilistic functions which are not specific to any instantiations of
the group they summarize” (Robinson 1997: 263).
Categories include concepts that enter into x is a kind of y association.

In the case of schemas, the basis for association is rather experiential in the
sense that elements of a schema may have co-occurred in the same context
or an event. In general, the relationships that hold between the elements of
a schema may be thematic, temporal and/or spatial. As an example, “bill”
and “food” are related schematically, as “food” may evoke the event
schema of paying a bill in a restaurant (Schank and Abelson 1977). Here
the relationship is more spatial and obviously experiential. On the other
hand, “food” and “pasta” are related to each other categorically, as “pasta”
is an “instance” of the category of “food”. Blewitt (1993: 104) makes a
distinction between schematic representations and categorical representa-
tions, which she calls “taxonomic”, in the following way:
Schematically organized representations preserve the temporal sequences
and the spatial and functional relations among units of experience. For ex-
ample, “spaghetti” and “bib” may be related in lexical memory, because
they label categories of objects that have been functionally connected and
thus experienced together in the same event. […] Taxonomically organized
representations are based on similarities among the units being represented,
that is, on shared meanings. For example, the nouns “apple” and “spaghetti”
may be related in memory because they refer to categories of objects that are
foods.
Conceptualisation of experience, of course, does not end in forming cate-
gories and schemas but also involves setting up mental models (Johnson–
Laird 1980) mapping across concepts, with the end result of metaphors,
and also perpectivizing what is being conceptualised (Verhagen forth.). A
major focus in cognitive linguistics is identifying such conceptualisations
and recruiting them when delving into people’s social experience (Dirven,
Frank and Pütz 2003; Frank 2003 a). The following important point needs
to be made regarding the nature of conceptualisations such as schemas and
categories: they have been conceived differently by the various and some-
times competing paradigms in cognitive and social psychology, and natu-
rally by scholars working in different (sub)disciplines. For example, gener-
ally speaking, earlier schools of psychology conceived of schemas as
“structures” in the mind, while connectionists view schemas as patterns of
activated knowledge (Rumelhart et al 1986).
Regardless of what the status of conceptualisations, such as schemas
and categories, is within the boundaries of an individual’s cognition, I
would like to argue that these conceptualisations also largely emerge at the
cultural level of cognition discussed above. People partly partake in similar
experience and as such constantly negotiate and conventionalise the way

they conceptualise their experience. Although, in all probability, no two
individuals conceptualise experience in exactly the same way, it is often
possible to perceive a collective cognition emerging from the interactions
between the members of a cultural group. I refer to such conceptualisations
as cultural conceptualisations (Sharifian 2003). The choice of “conceptu-
alisation” over “concept” is meant to reflect and highlight the dynamic
nature of such cognitive phenomena.
As aspects of cultural cognition, cultural conceptualisations appear to
be heterogeneously distributed across the minds of a cultural group. That
is, these conceptualisations are not equally imprinted in the minds of the
members in a cultural group at any given point in time. A cultural group is
not a collection of a number of individuals who live in a certain area, but
rather people who more or less conceptualise experience in a similar fash-
ion. As such, the notion of a cultural group is not intended to convey rigid
boundaries. Within the popular classifications of culture there are still
those who conceptualise experience more closely and as such create a sub-
culture within a culture. While one might object to the boundary fuzziness
of such notions as “cultural group”, our realities appear to be largely char-
acterized by “fuzziness” rather than by rigid boundaries and units.
Cultural conceptualisations usually develop into complex, dynamic
systems of knowledge, which are not totally and equally shared by the
members of the target cultural group. Over time, such dynamic systems
may act as major anchor points for people’s thought and behaviour and
may even constitute a worldview. In other words, cultural conceptualisa-
tions enable the individuals across a cultural group to think, so to speak,
with one mind. Often a simple clue or a gesture is enough to point to the
cultural conceptualisations that are acting as the basis for a social interac-
tion. The operation of such aspects of cultural cognition is often, but not
necessarily always, salient to those who come from outside the cultural
group. Stated differently, social interactions between the members of a
cultural group may suggest the operation of some sort of a collective cog-
nition to those who are not members of the cultural group, whereas the
members of the in-group can be quite unaware that such cultural conceptu-
alisation are being brought into play. It should be noted here that different
cultural groups differ with regard to the coherence of their cultural con-
ceptualisations. Some cultures, and some people within a given/single cul-
ture, develop more coherent conceptualisations.
To make a distinction between different forms of cultural conceptuali-

sation, imagine that in a given society people interact with each other in
conceptualising and establishing systems of kinship. One aspect of kinship
conceptualisation would be to use linguistic labels to categorize people
into “mum”, “dad”, “aunt”, etc. Another would be to develop norms of
conduct and responsibility towards each kin. These norms do not define the
category but are associated with the category thematically and as such
would need to be considered as schemas. A related notion that has been
used in cognitive anthropology and more increasingly in cognitive linguis-
tics is cultural model (e.g., D’Andrade 1995; D’Andrade and Strauss 1992;
Frank 2003 b; Holland and Quinn 1987; Wolf and Simo Bobda 2001). The
term, initially intended to be used instead of “folk models” (Keesing 1987),
has also been employed in the sense of “a cognitive schema that is inter-
subjectively shared by a social group” (D’Andrade 1987: 112). D’Andrade
constantly refers to the notion of “schema” in his explication of the term
“cultural model” (D’Andrade 1987: 112) and he regards models as com-
plex cognitive schemas. Strauss and Quinn (1997: 49) also maintain that
“another term for cultural schemas (especially of the more complex sort) is
cultural model”. Polzenhagen and Wolf (2007) have used the notion of
“cultural model” as more general, overarching conceptualisations that
would encompass metaphors and schemas that are minimally complex.
For the sake of this writing I view cultural models as conceptualisations
that hierarchically characterize higher nodes of our conceptual knowledge
and that encompass a network of schemas, categories and metaphors. An
example of such a model would be the cultural model of American Mar-
riage (Quinn 1987). This cultural model includes conceptualisations such
as GIVING AWAY schema, WEDDING GIFT category, and MARRIAGE AS
JOURNEY metaphor. Returning to the hypothetical case of kinship men-
tioned above, then, we may refer to the “cultural model of Kinship”. The
content and the relationship between these conceptualisations may be
summarized as follows in Table 1.
I would now like to make the observation that, although the locus of
such conceptualisations may be the individual, eventually they “spread”
among the group members and are then constantly negotiated and renegoti-
ated. The dynamics of such group interactions eventually lead to emergent
properties that may no longer be reduced to individual representations.
What this means is that schemas and categories become the objects of in-
teractions between the members of a given cultural group and as such
emerge as aspects of distributed cultural cognition. It is at this level that I
consider these conceptualisations to constitute cultural models, cultural

categories and cultural schemas. Such conceptualisations become “cul-
tural” not only because they may differ across different cultures but also
because they are the emergent properties of the interactions between the
members of a cultural group. Schematically, conceptualisations such as
cultural models may be visually represented using the same diagram that
was presented earlier in this chapter (see below). This is due to the fact that
these conceptualisations constitute an integral aspect of emergent, distrib-
uted cultural cognition.
Table 1.
CULTURAL MODEL OF KINSHIP

Kinship categories: categories such as “mum”, “dad”, “auntie”, “close relative”,
“in-laws” etc.
Kinship schemas: schemas that embody norms and values related to kinship, such
as behaviour rules for every member of the family in view of their status, etc. An
example of this would be RESPECT FOR PARENTS schema.
Kinship metaphors: Conceptual metaphors that are used in relation to kin, such as
Kwara’ae’s kin metaphor EXTENDED FAMILY MEMBERS ARE ALL ONE
HEARTH (Watson-Gegeo and Gegeo 1999: 230).
Figure 2 is an attempt to visually render the locus of a cultural conceptuali-

sation showing that such a cultural model has the two levels of abstraction.
Again, the top part of the diagram represents the “global” level of the
model, which emerges from the interactions between the members of a
cultural group, while the lower section depicts the way in which the “local”
level is instantiated in a distributed fashion across the individual minds
composing the group. This explanation provides an account of the way in
which some people know more than others about a given cultural model
and also that two people might share more elements from a cultural model
than some other members of the cultural group1. Factors such as age and
gender might contribute to what people have in common and share with
each other. One aspect of cultural development and, hence, the increased
stability of the model/overall system is movement from a state where
someone knows A to where the same person knows ABCD, for example,
1. See Borofsky (1994) for an account of intra-group diversity in cultural knowl-

edge.
from a cultural model. But of course, issues such as the extent to which one
enters into interactions with the members of their cultural group would also
determine how much a person knows from/about their cultural conceptuali-
sations.
Figure 2. A distributed, emergent cultural model
A point that needs to be made here is that in the above figure the person
who knows A and the one who knows CD do not appear to belong to the
same cultural group. This is because the figure only represents one cultural
model. In reality, those two people might share more from other cultural
models, and as such still belong to one cultural group. This pattern of
sharing from two cultural models, X and Y, is represented in Figure 3.
The figure depicts how two members may share more elements from
one cultural model than from another. This pattern of distributed cultural
cognition accounts for “fuzzy” understandings that characterise our daily
cultural interactions. As mentioned earlier, people coming from the same
cultural background generally work on the basis of the assumption that
they have shared cultural models, whereas in reality this might not be to-
tally the case, as has been discussed here. This situation often leads to mis-
understandings and can even create conflicts between people.
The situation can of course get much more complex in intercultural
communication contexts in which interlocutors may draw on different and
even contrasting cultural models. In such situations, every interlocutor is
likely to draw on the cultural models that characterise his/her “native”

culture. However, there is often the case that even one interlocutor may
draw on the elements of two or more cultural models to which they have
been exposed during their life (Frank and Susperregi 2001; Frank 2003 a,
2005). Conflict and miscommunication often takes place in such contexts
due to the assumption made by the interlocutors that they are all drawing
on the same cultural models. It should however be noted that often durable
contact between groups of individuals from different cultural backgrounds
results in the emergence of new, and in a sense “blended”, cultural models.
CULTURAL MODEL X
CULTURAL MODEL Y
Figure 3. Two distributed cultural models
4. Emergent cultural cognition and language
Language is intrinsically related to distributed, emergent cultural cognition

which has been discussed so far in this chapter. Cultural cognition is
largely, but not solely, transmitted through language. It is also instantiated
in the content and the use of language. Inherent within the system of every
language are categories, schemas, conceptual metaphors and propensities
for certain perspectives that reflect cultural cognitions of those who have
spoken the language over the history of its existence. As Tomasello (1999:
169) puts it,
[…] in collaboration over historic time human beings have created an in-
credible array of categorical perspectives and construals of all kinds of ob-
jects, events and relations, and they have embodied them in their systems of
symbolic communication called natural languages.
Indeed, the way and the degree to which these conceptualisations have
been encoded in human languages appear to differ from one language to
another (Palmer 1996). The following section gives examples of how vari-
ous features of human languages may instantiate conceptualisations that
have at one stage or another characterized the cultural cognition of their
speakers.
At the level of lexicon, lexical devices that are considered to be
equivalent in different languages, or even language varieties, may signify
different conceptualisation of experience for their speakers (e.g., Sharifian
2001). Sharifian (2005), for example, observed that many speakers of Abo-
riginal English and Australian English associate different conceptualisa-
tions with words such as “family” and “home”. For Aboriginal English
speakers, the word “home” gives rise to conceptualisations that would be
associated with the company of the extended family members whereas the
Anglo-Australian speakers largely associate the word with a building that
is being rented or owned by themselves or a member of their nuclear fam-
ily. For an Aboriginal person, for instance, the word “home” may refer to
the place of residence of one’s grandmother or aunt.
The word “family” for Aboriginal English speakers is associated with
the Aboriginal model of Family. This cultural model includes categories
that go beyond those associated with the same word in the case of Anglo
Australians. Family for an Aboriginal person includes members of the
“extended” family and largely whomever one comes into frequent contact
with. A word such as “mum” for an Aboriginal person may evoke a cate-
gory that includes people who are described as “aunt” by an Anglo Austra-
lian. Also responsibilities, obligations and behaviour rules that are often
observed between the members of an Aboriginal family would give rise to
schemas that appear to be largely culture-specific. In some Aboriginal cul-
tures, a person may not be allowed to converse with their mother-in-law or
whoever is regarded as a member of the same category.
Cultural conceptualisations may also be marked on morphosyntactic
features of some languages. Aboriginal Australians have systems of con-
ceptualisation of kinship that are often viewed as complex from the view-
point of the Anglo-Australian culture. Aboriginal cultural conceptualisa-
tions of kinship are encoded in certain morphosyntactic features of Abo-
riginal languages. For example, Murrinh–Patha has various second person
pronouns including those which categorise family members. These include
nhi “you singular”, nanku “you two brothers and sisters” and nanku ngin-
tha “you two who are not brothers or sisters and one or both are female”
(Walsh 1993). In Arabana, there are pronouns which signify categories that
highlight moiety as well as generation level, such as the following:
Arnanthara = we, who belong to the same matrilineal moiety, adjacent gen-
eration levels, and who are in the basic relationship of mother, or mothers’
brother and child. (Hercus 1994: 117)
Another reflection of kinship conceptualisations in the grammar of a num-
ber of Aboriginal languages is in the use of collective suffix forms (Dench
1987). The suffix is described as “a morpheme deriving a new verb lexeme
which requires a nonsingular subject and has the added meaning that the
activity is performed together by the participants denoted by the subject
NP” (Dench 1987: 325). However, there appear to be cases where the col-
lectiveness denoted by the suffix is more of a marker of kinship rather than
of any “collective activity”. Consider the following example:
1. a. Nyiya karlpa-nyayi-ku wiya-larta panti-jangu karnti-ka –ku

This clim-COLL-PRES see –FUT sit -REL tree -LOC-ACC
This one is climbing up to see that one sitting in the tree.
(Dench 1987: 326)
In the above example, the activity of “climbing up” does not appear to be
“collective”, at least in the usual sense of the verb, and thus the collective
suffix may perform a different function here. Dench maintains that in such
cases “the appearance of the suffix indicates that the participants are in
the same set of alternating generations [italics original]” (1987: 327). That
is, the speaker who has uttered sentence (a) above knows that the person
climbing up the tree and the one to be seen are relatives in the same set of
alternating generations, or people in a “harmonious kinship”, as Hale
(1966) would put it.
Another area of language that encodes cultural conceptualisations of
experience is the area of metaphor (e.g., Frank 2003 a; Kövecses 1999,
2000; Yu 2002, 2003 a, 2003 b, 2003 c, 2004, this volume). Yu, for exam-
ple, gives numerous examples from Chinese where the metaphors involv-
ing a body part somehow embody Chinese cultural conceptualisations of

experience and also of the human body. He maintains the relationship be-
tween body, culture and metaphor as “conceptual metaphors are usually
derived from bodily experiences; cultural models, however, filter bodily
experiences for specific target domains of conceptual metaphors; and cul-
tural models themselves are very often structured by conceptual meta-
phors” (2003 c: 29).
Cultural conceptualisations also provide analytic tools for explorations
of pragmatic aspects of language. First, the use of pragmatic devices, such
as pragmatic markers, may be associated with culture specific conceptuali-
sations (see Sharifian and Malcolm 2003: 335). Also, at the heart of the
usage of terms such as “inferencing”, “implied meaning”, etc., lies the
notion of “conceptualisation”. When we say the use of a certain linguistic
device has a given implied meaning, we are in fact referring to conceptuali-
sations that the speaker/hearer associates with the use of the device in a
particular context. It is of course well-known in the area of pragmatics that
different cultures may have different pragmatic norms and devices and thus
it may be stated that across different cultures, different devices might be
associated with similar or overlapping cultural schemas and in some cases
similar devices may give rise to contrasting cultural schemas.
For instance, in Persian, a speaker may use the phrase sharmandeh-am
“I am ashamed” in achieving speech acts such as Offering Goods and
Services, Making a Request, and Expressing Gratitude. In such cases, it
appears that the formulaic expression is associated with a Persian cultural
schema (Sharifian 2004). This schema encourages the speaker to consider
the possibility that the action referenced by the speech act may give or has
given some “burden” to the hearer, or the food that is being offered may
not be tasty or correspond to the status of the guest. This schema then en-
courages the speaker to express the negative feelings that could arise out of
such considerations in the form of an expression of “shame”.
At the discourse level, both the content of discourse and its rhetorical
organization may reflect cultural conceptualisations of experience
(Malcolm and Rochecouste 2000; Malcolm and Sharifian 2002, 2005,
2007). Malcolm and Rochecouste (2000), for example, analysed excerpts
of narrative produced by speakers of Aboriginal English and realised that
the texts were largely governed by event schemas that reflected Aboriginal
cultural experience. They named these schemas Hunting, Travelling, Ob-
serving and Encountering the Unknown, which encompasses the Spiritual
experiences of Aboriginal people.
As an example of how cultural conceptualisations may be instantiated

in discourse, the following excerpt reveals the way in which three Aborigi-
nal English speakers locate themselves and their interlocutors in terms of
their kin:
L: Armadale you know all the streets an you know where to go?
EH: I’s It’s not like down the xxxx xxx too many big mob go that way
M: I’ve got some um people live round Armadale
EH: In Armadale?
M: Ah no not Armadale at Perth
L: [Perth]
EH: In Perth, what’s the names down there?
M: um Davises2
EH: Oh yeah
M: an Coles
EH: That’s on my Mum’s side, my Mum related to Coles
M: Um do you know, do you know um, Shane Cole?
EH: Yeah that’s my cousin. Mum’s cousin I think
M: We’ ah yeah, thas my brother, cousin brother
EH: Well there’s um there’s an older one as well isn’t there?
M: Um Donny... and but they’re all sisters, um Marcia but we just call
her Marce, Marcia Collins an um um Kate and um... um got some
Davises um but only just um um from my niece, Jeanette Cole, she
goes um horse riding every day um cos she lives with her Nan an
Pop an her mother and father cos their mother an dad um lives with
them, so she stays with them an, ‘cross the road there are these peo-
ple who that um takes her horse riding
EH: Oh yeah
M: Um like on a station, an she just goes with em to um – cos um they
signed her in so she could go with em, bout every other– every day
EH: Yeah we – we were talking about Jim L__ (FAMOUS FOOT-
BALLER) and the boys said that’s your uncle, unna?
L: mmm
EH: xxx cos Jim’s my cousin xxx I got Elvis in there (laughs) they were
saying that, someone was saying that Jim’s real name was Elvis
(laughs)
L: Well but e’s my uncle but I don’t know him,
2. The names used in the texts are pseudonyms.

EH: Alright
L: He’s just know Dad an ’e might be a second cousin or something
EH: What’s your Dad’s last name?
L: Um Gordon
EH: Oh your Dad’s Gordon too what was your Dad’s first name
L: Gavin Gordon, he was- Dad is um Ronnie Gordon and is brother is
Ronnie and Nathan
EH: I know that, I know that name
L: Do you know Cherie and Lindy, they Gordon, that’s my Dad’s sis-
ters
EH: Alright. What cos my Dad’s related to old oh yeah, nah well my
Dad – Jim’s Mum and my Dad are like brother and sister, an my Dad
he got no sisters an they all first cousins
L: Well what’s ya last name?
EH: Um Haines (Y70, Yarning about Family)
The above conversation, which is between Aboriginal speakers coming

from some 400 kilometres apart, is a clear instantiation of the Aboriginal
cultural model of Family. First, the text represents an Aboriginal schema
that encourages the speakers to locate themselves and others with regard to
their possible kinship links. This often seems to be necessary among many
Aboriginal people in that it has implications regarding where they stand in
relation to their interlocutor and what they should do or say.
The text also reveals cases of instantiating Aboriginal cultural catego-
ries. For instance, speaker M refers to someone as “brother, cousin
brother”. The category “cousin-brother” includes people who are biologi-
cally cousin to the speaker but who have the same cultural status as a
brother and may simply be referred to as “brother”. Speaker L also refers to
someone as “uncle” and then proceeds to say that he “might be a second
cousin or something”. As mentioned earlier in Aboriginal cultures the
categories that are labelled as “uncle” or “aunt” may include people who
may be considered as “distant relatives” from the Anglo Australian per-
spective.
As mentioned above, cultural conceptualisations may also be instanti-
ated in the rhetorical organization of discourse. Carrell calls schemas that
include knowledge relative to the rhetorical organisation of a text formal
schemas (Carrell 1987: 461). She found that reading comprehension was
easiest when the texts were familiar to the readers in terms of their cultural
formal and content schemas. Some cultures draw on a formal schema that
is tied to the linear conceptualisation of “time”. That is, people in such

cultures largely narrativise their experience based on the chronological
order of the happening of events. Not all cultures, however, follow such
patterns of discourse organization (e.g., Kaplan 1966, 1987; Kintsch and
Greene 1978). It has been noted, for example, that Kuna Indians of Panama
do not construct their narrative structure based on temporal ordering (Sher-
zer 1987). In other words, the speakers do not seem to rely on temporal
schemas in their narrativisation of experience. Palmer (1996) attributes this
to the salience and valuation of the imagery in the narrator’s worldview.
Aboriginal English speakers also do not appear to rely very much on the
chronological sequencing of the events in their discourse production
(Sharifian 2002). Rather, in Aboriginal English discourse, events may be
ordered according to their salience and significance in the cultural con-
ceptualisations that speaker is drawing on.
It is to be noted finally that as an integral aspect of cultural cognition,
language itself is a complex adaptive system (see also Frank this volume;
Steels 1996, 2000) in the sense that it is a distributed, emergent, adaptive
system. The knowledge of a language is heterogeneously distributed across
the minds in a speech community. In a study of mass/count in Persian, for
example, Sharifian and Lotfi (2003) employed a Preference task that meas-
ured the acceptability of a number of sentences by a group of native speak-
ers of Persian. The data showed a high degree of variability in the degree to
which participants rated the sentences as “acceptable”. For example, one of
the sentences was rated as “fully acceptable” by 17.9%, “acceptable but not
preferred” by 32.1% and “unacceptable” by 50%. This pattern of data
shows how knowledge of language is heterogeneously distributed across
the members of a speech community.
Also, language is an emergent system in the sense that it evolves and
hence results from the communicative interactions between the individual
members of a speech community across time and space. If we map human
commutative interactions onto a network that extends across the dimen-
sions of time and space, then language is the emergent property of the net-
work as a whole. It is to be noted that the interactions that characterise the
network are not mirror images of one another, which makes language a
dynamic system with unpredictable properties. In the terminology of com-
plex adaptive systems, language is rarely in any long run equilibrium.
Language is a dynamic adaptive system in the sense that it can be
adapted to meet the communicative needs of its speakers. At one level,
speakers often adapt their language in specific situations to express certain
specific meanings. Also, studies in diachronic linguistics have shown that

certain features of human languages may be adapted to express a wide
range of new conceptualisations. It has been observed that a language im-
planted in new localities may be adapted and appropriated by its new
speakers to express their own native worldview and culture. This has, for
example, been observed in the case of Aboriginal people adapting English
to clothe their own worldview and cultural conceptualisations (e.g., Shari-
fian 2006, 2007).
5. Concluding remarks
In this chapter I have made an attempt to further expand the notion of cog-
nition along the dimension of culture. From the perspective that is intro-
duced in this chapter, cognition is viewed as a property of cultural groups,
and not just the individual. In this sense, cognition is a heterogeneously
distributed system with emergent properties that arise from the interactions
between the members of a cultural group. An integral aspect of this view of
cultural cognition is group-level conceptualisation. Conceptualisations
such as models, schemas and categories have an individual basis as well as
an emergent basis as the cultural level of cognition. These cultural con-
ceptualisations are often instantiated in various cultural artefacts and ac-
tivities. Language in this perspective is viewed as a distributed system as
well as a repository for cultural conceptualisations. Various aspects of
human languages may encode conceptualisations that reflect cultural expe-
riences of their speakers. It is hoped that this chapter will contribute to the
emerging integrative perspective that is reflected in the title of this volume,
as well as in the other contributions.
Acknowledgements
I wish to thank Roslyn Frank and René Dirven for their generous, helpful
comments on the earlier drafts of this chapter. Ian G. Malcolm also de-
serves my special thanks for his encouragement throughout the develop-
ment of the ideas presented in this chapter and also for his helpful sugges-
tions.
References
Altarriba, Jeanette (ed.)

1993 Cognition and Culture: A Cross-cultural Approach to Cognitive
Psychology. Amsterdam: Elsevier Science Publishers.
Anderson, Michael L.
2003 Embodied cognition: A field guide. Artificial Intelligence 149 (1):
91–130.
Arbib, Michael A.
1992 Schema Theory. In: Stuart S. Shapiro (ed), The Encyclopedia of
Artificial Intelligence, 2nd Edition, 1427–1443. New York: Wiley–
Interscience,
Bartlett, Fredrick C.
1932 Remembering. Cambridge: The Cambridge University Press.
Bernárdez, Enrique
culture.
Blewitt, Pamela
1993 Taxonomic structure in lexical memory: The nature of developmental
change. In: Ross Vasta (ed.), Annals of Child Development, Vol. 9,
103–132. London: Jessica Kingsley.
Bobrow, Daniel G. and Norman, Donald A.
1975 Some principles of memory schemata. In: Daniel G. Bobrow and
Allan M. Collins (eds.), Representation and Understanding: Studies
in Cognitive Science, 131–149. New York: Academic Press.
Borofsky, Robert
1994 On the knowledge of knowing of cultural activities. In: Robert
Borofsky (ed.), Assessing Cultural Anthropology, 331–348. New
York: McGraw-Hill.
Carrell, Patricia L.
1987 Content and formal schemata in ESL reading. TESOL Quarterly 21
(3): 461–481.
Cole, Michael
1996 Culture in Mind. Cambridge. MA: Harvard University Press.
D’Andrade, Roy G.
1987 A folk model of the mind. In: Dorothy Holland and Naomi. Quinn
(eds.), Cultural Models in Language and Thought, 112–148. New
York: Cambridge University Press.
1995 The Development of Cognitive Anthropology. New York: Cambridge
University Press.
D’Andrade, Roy G. and Claudia Strauss (eds.)

1992 Human Motives and Cultural Models. Cambridge: Cambridge Uni-
versity Press.
Davis, Steve
1992 Connectionism: Theory and Practice. Oxford: Oxford University
Press.
Dench, Allan
1987 Kinship and collective activity in the Ngayarda languages of north-
west Western Australia. Language in Society 16: 321–339.
Dirven, Rene, Roslyn M. Frank and Martin Pütz
2003 Cognitive Models in Language and Thought: Ideology, Metaphors
and Meaning. Berlin/New York: Mouton de Gruyter.
Frank, Roslyn M.
2003 a Shifting identities: The metaphorics of nature-culture dualism in
Western and Basque models of self. Metaphorik.de O4/2003 (elec-
tronic journal). http://www.metaphorik.de/04/frank.pdf
2003 b Shifting Identities in Basque and Western cultural models of self and
being. In: René Dirven, Roslyn M. Frank and Martin Pütz (eds.),
Cognitive Models in Language and Thought, 123–157. Berlin and
2005 Shifting Identities: A comparative study of Basque and Western
cultural conceptualizations. Cahiers of the Association for French
Language Studies 11 (2): 1–54. http://www.afls.net/cahiers.html
this vol. The language-organism-species analogy: A complex adaptive sys-
tems approach to shifting perspectives on ‘language’
Frank, Roslyn M. and Mikel Susperregi
2001 Conflicting identities: A comparative study of non-commensurate
root metaphors in Basque and European image schemata. In: René
Dirven, Roslyn M. Frank and Cornelia Ilie (eds.), Language and
Ideology. Vol. 2. Cognitive Descriptive Approaches, 135–160. Am-
sterdam/Philadelphia: John Benjamins.
Hale, Kenneth L.
1966 Kinship reflections in syntax: Some Australian languages. Word 22:
318–324.
Hercus, Luise A.
1994 A Grammar of the Arabana–Wangkangurru Language Lake Eyre
Basin, South Australia. Pacific Linguistics Series C–128. Australian
National University: Canberra.
Holland, Dorothy and Naomi Quinn (eds.)
1987 Cultural Models in Language and Thought. Cambridge: Cambridge
University Press.
Hutchins, Edwin
1994 Cognition in the Wild. Cambridge, MA: MIT Press.
Johnson, Steven
2001 Emergence: The Connected Lives of Ants, Brains, Cities and Soft-
ware. New York: Scribner.
Johnson-Laird, Philip N.
1980 Mental models in cognitive science. Cognitive Science 4: 71–115.
Kaplan, Robert B.
1966 Cultural thought patterns in intercultural education. Language
Learning 16: 1–20.
1987 Cultural thought patterns revisited. In: Ulla Connor and Robert B.
Kaplan (eds.), Writing across Languages: Analysis of L2 Text, 9–21.
Reading, MA: Addison-Wesley.
Keesing, Roger M.
1987 Models, “folk” and “cultural”: Paradigms regained. In: Dorothy
Holland and Naomi Quinn (eds.), 369–393.
Kintsch, Walter and Edith Greene
1978 The role of culture-specific schemata in comprehension and recall of
stories. Discourse Processes 1: 1–13.
Kövecses, Zoltan
1999 Metaphor: Does it constitute or reflect cultural models? In: Raymond
W. Gibbs and Gerard J. Steen (eds.), Metaphor in Cognitive Lin-
guistics, 167–188. Amsterdam: John Benjamins.
Kronenfeld, David
2002 Culture and society: The role of distributed cognition. In: Robert
Trappl (ed.), Cybernetics and Systems, Vol. 1, 430–431. Vienna:
Austrian Society for Cybernetic Studies.
Langacker, Ronald
1994 Culture, cognition and grammar. In: Martin. Pütz (ed.), Language
Contact and Language Conflict, 25–53. Amsterdam/Philadelphia:
John Benjamins.
2007 Embodiment and social interaction: Implications for cognitive sci-
ence
Malcolm, Ian G. and Judith Rochecouste
2000 Event and story schemas in Australian Aboriginal English discourse.
English World-Wide 22 (2): 95–123.
Malcolm, Ian G. and Farzad Sharifian
2002 Aspects of Aboriginal English oral discourse: An application of cul-
tural schema theory. Discourse Studies 4 (2): 169–181.
2005 Something old, something new, something borrowed, something

blue: Australian Aboriginal students’ schematic repertoire. Journal of
Multilingual and Multicultural Development 26 (6): 512–532.
2007 Multiwords in Aboriginal English. In: Paul Skandera (ed.), Phrase-
ology and Culture in English. Berlin/New York: Mouton De Gruyter.
Mandler, Jean
1984 Stories, Scripts and Scenes: Aspects of Schema Theory. Hillsdale,
NJ: Erlbaum.
Newell, Allen
1980 Physical symbol systems. Cognitive Science 4: 135–183.
Palmer, Gary B.
Press.
Polzenhagen, Frank and Hans-George Wolf
2007 A case study on culture-specific conceptualisations: Linguistic ex-
pressions of corruption in African English. In: Farzad Sharifian and
Gary B. Palmer (eds.), Applied Cultural Linguistics: Implications for
Second Language Learning and Intercultural Communication, 181–
197. Amsterdam/Philadelphia: John Benjamins.
Quinn, Naomi
1987 Convergent evidence for a cultural model of American Marriage. In:
Dorothy Holland and Naomi Quinn (eds.), 173–192.
Redding, S. Gordon
1980 Cognition as an aspect of culture and its relation to management
processes. An exploratory review of the Chinese case. Journal of
Management Studies 17 (2): 127–148.
Robinson, Edward A.
1997 The cognitive foundations of pragmatic principles: Implications or
theories of linguistic and cognitive representation. In: Jan Nuyts
(ed.), Language and Conceptualization, 253–271. Cambridge: Cam-
Rosch, Eleanor
1978 Principles of categorization. In: Eleanor Rosch and Barbara B. Lloyd
(eds.), Cognition and Categorization, 27–48. Hillsdale NJ: Lawrence
Erlbaum.
Rumelhart, David E.
1980 Schemata: The building blocks of cognition. In: Rand Spiro, Bertram
Bruce and William Brewer (eds.) Theoretical Issues in Reading
Comprehension, 33–58. Hillsdale, NJ: Lawrence Earlbaum
Rumelhart, David, Paul E., Smolensky, James L. McClelland and Geofrrey E.

Hinton
1986 Schemata and sequential thought processes in PDP models. In: James
L. McClelland, David E. Rumelhart and the PDP Research Group
(eds.), Parallel Distributed Processing: Explorations in the Micro-
structure of Cognition, Vol. 2. Psychological and Biological Models,
7–57. Cambridge, MA: MIT Press.
Schank, Roger C. and Robert P. Abelson
1977 Scripts, Plans, Goals and Understanding: An Inquiry into Human
Knowledge Structures. Hillsdale, NJ: Lawrence Erlbaum.
Sharifian, Farzad
2001 Schema-based processing in Australian speakers of Aboriginal Eng-
lish. Language and Intercultural Communication 1 (2): 120–134.
2002 Chaos in Aboriginal English discourse. In: Andy Kirkpatrick (ed.),
Englishes in Asia: Communication, Identity, Power and Education,
125–141. Melbourne: Language Australia.
2003 On cultural conceptualisations. Journal of Cognition and Culture (3)
3: 187–207.
2004 Cultural schemas and intercultural communication: A study of Per-
sian. In: James Leigh and Eric Loo (eds.), Outer Limits: A Reader in
Communication and Behaviour across Cultures, 119–130. Mel-
bourne: Language Australia.
2005 Cultural conceptualisations in English words: A study of Aboriginal
children in Perth. Language and Education 19 (1): 74–88.
2006 A cultural-conceptual approach and world Englishes: The case of
Aboriginal English. World Englishes 25 (1): 11–22.
2007 Aboriginal language habitat and cultural continuity. In: Gerhard
Leitner and Ian G. Malcolm (eds.), The Habitat of Australia’s Abo-
riginal Languages, 181–196. Berlin/New York: Mouton De Gruyter.
Sharifian, Farzad and Ahmad R. Lotfi
2003 “Rices” and “waters”: The mass/count distinction in Modern Persian.
Anthropological Linguistics 45 (2): 226–244.
Sharifian, Farzad and Ian G. Malcolm
2003 The use of pragmatic marker like in English teen talk: Australian
Aboriginal usage. Pragmatics and Cognition 11 (2): 327–344.
Sherzer, Joel
1987 A discourse-centered approach to language and culture. American
Anthropologist 89: 295–309.
Shore, Bradd
1996 Culture in Mind: Cognition, Culture and the Problem of Meaning.
New York, NY: Oxford University Press.
Sperber, Dan and Lawrence Hirschfeld

1999 Culture, cognition and evolution. In: Robert Wilson and Frank Keil
(eds.), MIT Encyclopaedia of the Cognitive Sciences, cxi–cxxxii.
Cambridge, Mass.: MIT Press
Steels, Luc
1996 Emergent adaptive lexicons. In: Pattie Maes, Maja J. Mataric, Jean
Arcady Meyer, Jordan Pollack and Stewart W. Wilson (eds.), From
Animals to Animats: Proceedings of the Fourth International Con-
ference on Simulation of Adaptive Behavior, 562–567. Cambridge:
The MIT Press.
2000 Language as a complex adaptive system. In: Marc Schoenauer, Kaly-
anmoy Deb, Günther Rudolph, Xin Yao, Evelyne Lutton, Juan Julian
Merelo and Hans-Paul Schwefel (eds.), Proceedings of the Sixth
Conference on Parallel Problem Solving from Nature, 17–26. Berlin,
Germany: Springer Verlag.
Strauss, Claudia and Quinn, Naomi
1997 A Cognitive Theory of Cultural Meaning. New York: Cambridge
University Press.
Tomasello, Michael
1999 The Cultural Origins of Human Cognition. Cambridge, Mass.: Har-
vard University Press.
Verhagen, Arie
forth. Construal and perspectivisation. In: Dirk Geeraerts and Hubert
Cuyckens (eds.), Oxford Handbook of Cognitive Linguistics. Oxford:
Oxford University Press.
Violi, Patrizia
2003 Embodiiment at the crossroads between cognition and semiosis.
Reserches en communication 19: 199–217.
this vol. Beyond the body: Towards a full embodied semiosis.
Vygotsky, Lev Semenovich
esses. Cambridge: Harvard University Press.
Waldrop, M. Mitchell
1992 Complexity: The Emerging Science at the Edge of Order and Chaos.
New York: Simon and Schuster.
Walsh, Michael
1993 Classifying the world in an Aboriginal language. In: Michael Walsh
and Colin Yallop (eds.), Language and Culture in Aboriginal Aus-
tralia, 107–122. Canberra: Aboriginal Studies Press.
Watson-Gegeo, Karen Ann and David W. Gegeo
1999 (Re)modeling culture in Kwara’ae: The role of discourse in children's
cognitive development. Discourse Studies 1 (2): 227–246.
Wilson, Margaret
9 (4): 625–636.
Wolf, Hans-Georg and Augustin Simo Bobda
2001 The African cultural model of community in English language in-
struction in Cameroon: The need for more systematicity. In: Martin
Pütz, Susanne Niemeier and René Dirven (eds.), Applied Cognitive
Linguistics II: Language Pedagogy, 225–259. Berlin: Mouton de
Gruyter.
Yu, Ning
2002 Body and emotion: Body parts in Chinese expression of emotion.
Pragmatics and Cognition 10 (1/2): 341–367.
2003 a Chinese metaphors of thinking. Cognitive Linguistics 14 (2/3): 141–
165.
2003 b Synesthetic metaphor: A cognitive perspective. Journal of Literary
Semantics 32 (1): 19–34.
2003 c Metaphor, body and culture: The Chinese understanding of gallblad-
der and courage. Metaphor and Symbol 18 (1): 13–31.
2004 The eyes for sight and mind. Journal of Pragmatics 36 (4): 663–686.
this vol. Metaphor, body and culture: A cross linguistic perspective.
Collective cognition and individual activity:
Variation, language and culture1
Enrique Bernárdez
Abstract
One of the most salient features of human language is its diversity; this begs for an
explanation, as language has to be viewed as a “product” of human cognition,
which is principally assumed to be inter-individually identical. As cognition is
taken to be restricted to the individual, thus ignoring the existence and possibility of
variation, the old problem of the Janus-like physiognomy of language and the rela-
tion between its social (external) and individual (internal) faces has to be posed
anew. This paper will focus on the question of how to bridge the gap between an
individual’s cognitive system (and, consequently, language), and linguistic diver-
sity, i.e., the problem of the relationship between the individual and the collectivity.
This problem is approached by introducing the concept of synergic cognition in
relation to the study of similar problems in biology and complex systems theory.
Language will be viewed as a “product” of a socially-conditioned, activity-driven
cognition. The justification of this proposal will be based on both sociology (esp.
Pierre Bourdieu) and psychology (esp. “activity theory”), and parallel results in the
organisation of biological systems and especially the interplay between individual
and social group among animals will also be considered. The similarities of my
approach with others will be pointed out.
Keywords: activity, embodiment, habitus, situated cognition, synergic cognition.
1. This paper is based on a plenary talk presented at the 8th ICLC, Logroño, 2003.
I thank all those who offered me their comments and criticism. Special thanks
go to R. Dirven, O. Lizardo, P. Quist, C. Sinha, and J. Zlatev, but especially to
Roz Frank.
138 Enrique Bernárdez
1. The inescapable reality of variation in language
Language exists only in variety: according to some estimates, between

150,000 and 500,000 languages have been spoken through the history of
modern man. But even if the precise number of existing languages is im-
possible to ascertain, it seems obvious that linguistic variety is a consub-
stantial feature of human life in both the social and the individual sphere.
Nonetheless, we usually prefer to talk about language in the singular so
that the construct termed human language ends up representing the central
concern of most linguists. It is, therefore, necessary to make an explicit
distinction between language in general (French langage, Spanish len-
guaje) and “a particular language” (French langue, Spanish lengua,
idioma); many others, for instance Fuchs (1997: 6), have made similar
observations. Chomsky (1986) proposed the terms I-language and E-
language respectively and somehow this particular distinction still seems
to be hanging around even among non-Chomskyans. I-language is the in-
ternalised, individual, mental side of human language, whereas E-
language, in contrast, is its externalised, social side. I-language is some-
how “inside the brain” whereas E-language is “out there in society”: ex-
amples are what we use to call English, Spanish, Ojibwa or Indonesian No
such thing as I–language or “human language” is accessible to observation,
of course, and human language as such is a mere construct. But it is just
that supposed, hidden phenomenon, the I-language, which is usually as-
signed the primary value, whereas the directly observable phenomenon –
the individual languages and linguistic varieties as they are really used by
real speakers in real situations – is seen as a mere epiphenomenon of I-
language. As a result, there has been a widespread tendency to consider
similar abstract constructs such as langue, competence or human language,
as the primary object of research. Maybe this is a consequence of the folk
theory of essences (Lakoff and Johnson 1999), or of the traditional ten-
dency to see what is most abstract, less in direct touch with “reality”, as
most important. Or perhaps it is a consequence of the theoretical reason, as
defined by Bourdieu, which, through its exclusive interest in artificial con-
structs, annihilates its object of study, a problem even more serious if we
are not conscious of it:
[D]ans la mesure où elle engage un mode de pensée qui suppose la mise en
suspens de la nécessité pratique et met en œuvre des instruments de pensée
construits contra la logique de la pratique, [...] la vision scolastique s’expose
à détruire purement et simplement son objet ou à engendrer de purs artefacts
Collective cognition and individual activity 139
lorsqu’elle s’applique sans réflexion critique à des pratiques qui sont le pro-
duit d’une tout autre vision. Le savant qui ne sait pas ce qui le définit en tant
que savant, c’est-à-dire le “point de vue scolastique”, s’expose à mettre dans
la tête des agents sa propre vision scolastique; à imputer à son objet ce qui
appartient à la manière de l’appréhender, au mode de connaissance.2 (Bour-
dieu 1994: 219)
Of course we always seek generalizations and, certainly, it is generalization
that science is about, not the mere observation and subsequent description
of directly perceptible phenomena, although there seems to be no reason
(apart from philosophical preferences) for the rejection of whatever is im-
mediately perceptible in exclusive favour of their assumed hidden reality.
In order to reach valuable generalizations about human language, that is, in
order to be able to understand what human language can be, we have to
study the variety of human languages – in the plural. This does not mean,
at any rate, that one should look back to induction as the only means for the
scientific study of language. But even if introspection has to be accepted as
one basic tool of linguistic and cognitive study (Gibbs and Matlock 1999),
within a general epistemological framework based on abduction
(Bernárdez 1995), introspection cannot be the exclusive tool, either: it has
to be supplemented by the careful scrutiny of (real) language data. Cogni-
tive and functional linguistics is a recognisably empirical discipline, and as
our object, language, is multiple, our empirical study must equally be mul-
tiple, i.e., multilinguistic.
2. Relating language(s) and cognition
Another undisputed fact about language is that it stands in very close rela-
tion to cognition. The problem, of course, is understanding and explaining
2. [In as far as it implies a way of thinking which suspends practical need and puts
to work tools of thought which were built against the logic of practice, […] the
scholastic point of view runs the risk of annihilating its object and engendering
mere artefacts whenever it is applied, without a previous critical reflection, to
practices which are the product of a completely different perspective. The
scholar who does not know what defines him as a scholar, that is, who ignores
the “scholastic point of view”, risks the danger of assigning his own scholastic
view to the heads of his agents; of assigning to his object what belongs to the
way of apprehending, to the way of knowing.] (All translations are by EB unless
otherwise noted)
the nature of such relation, as the answer depends, among other things, on
how we choose to define language and cognition. As things now stand, in
most varieties of Cognitive Linguistics (CL) it is an accepted fact that we
can get to at least some knowledge of cognition through the study of lan-
guage, and vice versa, i.e., that our knowledge of cognition, acquired by
other, non linguistic, means, will improve our understanding of linguistic
phenomena. Of course, the main question deals with the relation itself that
has to be taken to hold between language and cognition. The different an-
swers provisionally given to the question are responsible for the variety of
approaches to language, cognition and their interrelations.
It is assumed that through the study of language – standing as it does in
such intimate relation to cognition, and therefore wired into our brains in
some way or other – we can get to know more about cognition itself. We
tend to see cognition as a merely individual phenomenon and more or less
strictly determined by the human genome, which implies its universality
and invariability throughout the human species. The metaphor COGNITION
IS THE BRAIN could be formulated that would form the basis for this view
(Bernárdez 2005). Interestingly, this looks much like a revised version of
the Chomskyan view which keeps untouched the individual, innate, inter-
nal features that lurk in the I of I-language, although the scope of the in-
nate component is quite different in Generative Grammar and the CL ap-
proach, as a consequence of their following the premises of “first
generation” vs. “second generation cognitive science, as defined by Lakoff
and Johnson (1999)
In this context, variation seems principally impossible: one human cog-
nition – one human language. A refined version of Noam Chomsky’s Uni-
versal Grammar? How could interlinguistic variation be explained, if inter-
personal cognitive variation is precluded? In addition to its (probable)
epistemological implausibility, this approach is methodologically danger-
ous. If the study of human language – the construct, realized in individual
linguistic varieties – is a way toward the knowledge of cognition, we run
the risk of unduly generalizing from one single language to the whole of
human cognition.
The relation between language and cognition could be understood, then,
in an unjustifiable form: a particular language ≈ human cognition.3 The
3. Fuchs (1997: 6): It is quite risky “d’hypostasier cette langue, et de généraliser

indûment de la langue aux langues, puis des langues au langage” [to hypostasise
result of this kind of error could be summarized in the words of Bateson

(2002: 2212) in his devastating review of Pinker’s (2002) The Blank Slate:
“[W]hat Pinker happily calls human nature is in reality individual nature
and depends critically on the circumstances of that person’s life”. It might
be, too, that when we draw conclusions of (supposedly) universal cognitive
validity we are in fact repeating the same error, viz. seeing as universally
human what is restricted to the speakers of a particular language. A similar
caveat has been recently expressed by others, e.g. in relation to such fields
as evolutionary psychology (a field where Pinker is to be included), be-
havioural genetics (Ehrlich and Feldman 2003) and cross-cultural psychol-
ogy (Ratner and Hui 2003).
Studying language on the basis of one language or only a few languages
is indeed dangerous, as we could tend to assign a universal value to a cer-
tain language-specific feature. Unless we are fully conscious of the danger
implicit in undue generalizations on the basis of one or only a few – usu-
ally closely related – languages, our conclusions on human cognition will
be misled: we shall be calling “human cognition” what is in reality individ-
ual cognition which depends critically on the circumstances of a particular
social and cultural group and of a particular language.
2.1. Language structures vs. language use
There exist several reasons for the preference toward one single language –
English nowadays, while before it was Latin –, a preference both meth-
odological and philosophical but also, much too frequently, simply cultural
and ideological.4 Linguistics, including CL, needs to carry out an in-depth
this particular language and unduly generalize from that language to languages,
then from languages to human language].
4. Without entering into the necessary details, which will be the object of a more
detailed analysis to be published soon, the following can be briefly noted: (a)
we run the risk to assign what is idiosyncratic, both linguistically and culturally,
a general, universal status; a risk much too apparent and frequent to be ignored
and which affects other approaches to human cognition, as we saw above in our
quotation of Bateson 2002. (b) If such a wrong assignment takes place, we
could be assigning to human cognition certain features which might be exclu-
sive of the English language, as was the case in centuries past with Latin, whose
grammar was not only seen as the model for the grammar of any other language,
but was given the status of the “correct way of thinking”. (c) A trend can de-
reflection on the rationale of our scientific endeavours, in line with Bour-

dieu’s proposal cited above. Indeed, among of the reasons for this type of
preference is the fact that Language, human language, has been mainly
associated with the structures of language as an abstraction from their real
use. Although this perspective on language is often linked to the writings
of Ferdinand de Saussure, in fact it has dominated the field since the very
beginnings of the study of Greek, Latin and Sanskrit grammar. The struc-
tures of language – whatever they are taken to be – seem much simpler,
stable and easier to define than the chaotic appearance of linguistic usage.
In fact, since the beginnings of contemporary linguistics, we have tried to
keep questions of use neatly separated from Linguistics (or Grammar)
proper. There were a few attempts to overcome that situation, as in Tex-
tlinguistics (Bernárdez 1995, 1999). This neat separation of grammar and
use/usage is being abandoned nowadays in CL, with the development of
“usage based” models of grammar:
A usage-based theory, whether its object of study is internal or external lin-
guistic system, takes seriously the notion that the primary object of study is
the language people actually produce and understand. Language in use is the
best evidence we have for determining the nature and specific organization
of linguistic systems. (Kemmer and Barlow 2000: xv)
These authors (Kemmer and Barlow 2000: viii–xxii) characterize usage-
based models as sharing “a set of characteristic assumptions”:
(a) The intimate relation between linguistic structures and instances of
use of language; (b) the importance of frequency; (c) comprehension and
production as integral, rather than peripheral, to the linguistic system; (d)
focus on the role of learning and experience in language acquisition; (e)
linguistic representations as emergent, rather than stored as fixed entities;
(f) importance of usage data in theory construction and description; (g) the
intimate relation between usage, synchronic variation and diachronic
change; (h) the interconnectedness of the linguistic system with non-
velop to neglect the study of other, less well known languages because the re-
sults “would be there” at any rate in English (let's not forget that this was a
charge much too frequently done to Generative Grammar). And (d) as a lan-
guage – any language – is always accompanied by a certain culture and ideol-
ogy, the excessive generalization of English as the point of reference brings
about a similar expansion in the other fields: one more aspect of la pensée
unique and cultural globalization.
linguistic cognitive systems; (i) the crucial role of context in the operation
of the linguistic system.
Of course, such a view of language and grammar is not new, as most of
these assumptions have been a familiar element in many functional models,
as in most versions of textlinguistics, for quite a long time; but it does rep-
resent a significant shift in the “dominant trends” of linguistics. Be it as it
may, there is no longer any need to justify the possibility and convenience
of taking usage as a central element of language, including the much more
restricted area traditionally called “grammar”.
Now, instead of the traditional emphasis on form or the mere pairing of
form and meaning in the absence of any context or conditions of use, it is
this usage-based grammar that can serve as the focus of typological and
variationist research, instead of the traditional, noncontextual, abstract
models of grammar. The role of usage goes far beyond grammar, though,
for its relevance permeates the whole of language, for instance in the study
of metonymy, where the matter is not simply whether a certain type of
metonymy is possible for human cognition; let us mention the – much too
famous – metonymic expressions of the type The ham sandwich has left
without paying. Perhaps it would be much more interesting to analyse why
it is that some languages – some cultures – accept such metonymies
whereas in others, certain conditions must hold for them to be possible, and
why in still others, metonymic utterances like these would be rejected in all
circumstances. In general, why do languages differ so widely in the type
and extent of metonymic reference they are willing to accept: if it were just
a matter of being cognitively possible, once we discovered that a FOOD FOR
CUSTOMER metonymy is possible, or, instead, and more probably, a selec-
tion within the complete “restaurant script”, as proposed by Ruiz de Men-
doza and Otal Campo (2002: 30–31, 54ff), there would not be much else to
say. However, languages like Spanish impose considerable restrictions on
certain types of metonymic usage, as e.g. reference to a human being
through variable, non intrinsic or essential features; or the creation of verbs
on the basis of an instrumental argument (as in English to finger); all this
needs to be explained.
At the same time this usage-centred perspective could permit an expla-
nation of the existence of variation itself. Variation is the inescapable con-
sequence of use. We could perhaps be willing to accept that human cogni-
tion is invariable, although absence of variation may be too strong a
hypothesis, if due attention is paid to the reality of biological systems. This
would prevent variation, but the constantly varying conditions of interac-
tion, the basis of linguistic use, have variation as their immediate, inescap-
able consequence. Remember, by the way, that it was the acceptance of the
role of variation that led to the redefinition and modernization of historical
linguistics (the basic reference is Labov 1966), which had traditionally
suffered from the same exclusive preference toward the mere study of
structures.
Another interesting point is that, whereas we could, perhaps, see cogni-
tion as an internal, individual phenomenon, and study at least some parts of
language in that spirit, whenever we try to look at language use we enter
into the arena of activity that is necessarily associated with interaction:
between the individual and other individuals, between the individual(s) and
the environment.
2.2. Individual thinking, synergic action?
The distinction we have been dealing with is in the last term one between
the individual and its inner states on the one hand and, on the other – as
soon as language use, communication enters the picture – the collectivity,
i.e., the individuals in interaction, in an active, externalised state. Or be-
tween thought, which we assume to be a purely individual matter, and ac-
tion, which necessarily implies an outward movement of the individual:
toward its environment and toward other individuals. We assume that indi-
vidual cognition “produces” something, so to speak – of course, no conduit
metaphor is intended here – which is then “put to action”: for instance, a
certain grammatical construction which is then used in communicative
interaction with other individuals; or the plan for an action, which is then
carried out. Thinking would thus be an individual affair, whereas activity is
necessarily interactive. But note that whereas the interactive character of
activity is the direct result of observation, the purely individual character of
thinking is merely hypothetical.
We tend to assume that individual, “inward”, autonomous thinking en-
joys some kind of pre-eminence over the supraindividual, “outward” activ-
ity; and over any kind of cognition directed towards immediate action or
interaction, an idea, by the way, that has extremely old roots in Western
thinking. This view is not necessarily right, however; see for instance Peter
Harder’s (1999, 2003) comments on the limited autonomy of cognition and
language; others have emphasized the importance of the collective, social
component of human cognition, including its ontogenetical development
(Geeraerts 1999; Semin and Smith 2002; Tomasello and Rakoczy 2003).
Pierre Bourdieu’s (1980, 1994) emphasis on the “logic of practice”, as will
be shown later, is in a very similar vein. Many philosophers and psycholo-
gists have also emphasized the social, active nature of the human psyche
including its “higher cognitive functions”; and we should remember the
inseparability of cognition and emotion, as demonstrated by Antonio
Damasio (1994, 1999) and proposed much earlier by philosophers like
Maurice Merleau–Ponty (1945) and psychologists like Lev Vygotsky (1934
[1962], 1978), among many others. Sinha (1999) proposed the term neural
solipsism for the view of cognition as a purely neural issue, without any
consideration of things external; i.e., for a view of cognition as a purely
internal, individual phenomenon.
3. Embodiment
In the last twenty years we have witnessed a significant widening in the

scope of cognition: from a purely internal view as in first generation cog-
nitive science, to the nowadays firmly entrenched view of cognition as
embodied (for some recent discussions, see the contributions to Cognitive
Systems Research 3 (2002) and Cognitive Linguistics 13: 3 (2002); Sinha
and Jensen de López 2000; Kimmel this volume). Embodiment means that
cognition cannot function without the physical reality of the body, which is
open to the environment.
Unfortunately, there seems to exist no clear and universally accepted
definition of embodiment, as is shown in great detail by Chrisley and
Ziemke (2002), Ziemke (2003) and Kimmel (this volume, see also Rohrer
in BLM, volume 1). Certainly, it refers to the relation of cognition and the
body, an issue that has worried Western philosophers since the beginning,
as shown by Vesey (1965) in his historical review. But a basic concept like
embodiment clearly needs sufficient clarification because, as Vesey him-
self pointed out (1965: 11), “[p]hilosophical problems arise from the in-
adequacy of the concepts in terms of which we think of things”. The term
embodiment is so much in vogue nowadays in AI and the cognitive sci-
ences that we have to try to determine as precisely as possible what is
meant by the term. Unfortunately, a merely lexical analysis does not solve
the problem. The verb to embody has the following main meanings, ac-
cording to the Webster’s Dictionary (1971 s.v.):
1: To give a body to (a spirit): invest with a body: INCARNATE. 2a: to cause

to become material or sensual: deprive of spirituality b: to make concrete by
expression in perceptible form […] 3: to cause to become a body or part of a
body: INCORPORATE, ORGANIZE […] 4: to represent in human or animal
form: PERSONIFY […]
It is a loan translation of the Latin incorporare (through its French coun-
terpart), itself built on the basis of the word for body, thus meaning ‘to
bring into a body’.5 The Latin word also appears in English as incorporate,
albeit with a slightly different meaning, as the immediate reference to body
is lost. According to the Webster’s Dictionary, the word means:
1a: to unite with or introduce into something already existent usu. so as to
form an indistinguishable whole that cannot be restored to the previously
separated elements without damage […] b: to admit to membership in a cor-
poration […] 2a: to combine (ingredients) into one consistent whole […] b:
to bring together in an association […] 3: to give material form to: EMBODY
Only meaning #3 coincides with that of embody, so that in English both
terms are not synonymous.6 Embody occupies a special position in the
group it forms together with incorporate and incarnate, as it includes a
transparent reference to the body, so that the meaning can be reinterpreted
– and, as in this case, changed into a (semi-)technical term – in a rather free
way. That is, the technical meaning of embody is not necessarily closely
related to its non-technical meaning. The problem is that no single precise
definition seems to have gained widespread acceptance.
A negative, collateral consequence of this complex semantic picture is
the term’s correspondence in other languages and, hence, the problem of
adequately translating the term itself. In the Romance languages, forms like
incorporare have been in use for quite a long time,7 the translation of em-
body into these languages is not without difficulties, as incorporare and its
5. But originally in Latin body was understood, in the use of this verb, as a ‘mili-
tary body’, a corps, not a physical one!
6. The electronic edition of the Oxford Dictionary has as definition 3a in embody
the same as one of the definitions for incorporate, viz. “To cause to become
part of a body, to unite into one body; to incorporate (a thing) in a mass of ma-
terial, (particular elements) in a system or complex unity”.
7. In Spanish it is first attested in 1386, according to Corominas’ (1967) etymo-
logical dictionary; much earlier in French; the first English example is from
1398, while the first attested occurrence of to embody in the meaning that inter-
ests us is from 1601 (Shakespeare).
derivatives are not always usable, and the same happens with the descen-
dants of Latin incarnare, built on carnis ‘flesh’ (incarnate).8 In the oppo-
site direction, some problems in the reception of Bourdieu’s concepts, es-
pecially that of the habitus,9 in US sociology and anthropology may be due
to the lack of full correspondence between the French sociologist’s use of
incorporer and incorporation and their English rendering as embody. To
mention just one possible instance of such misunderstandings, Strauss and
Quinn (1997: 45) seem to understand the habitus as an internal state of the
individual, so to speak; they define the notion as “intrapersonal knowl-
edge”, and its extrapersonal component is apparently separated from it.
Their comments on the embodiment of the habitus refer thus to the notion
of embodiment as current in most of American discussions, as something
merely affecting the individual and his/her body. For Bourdieu, however,
as I understand his writings, the separation of the intra- and the extra-
personal is just of very secondary interest, as the habitus is an essentially
cultural and social object which is then incorporated in individuals; the
habitus is acquired by an individual through explicit and implicit learning,
but also through direct experience and imitation. Once acquired, the habi-
tus is internalised, i.e., incorporated, ‘embodied’. Embodiment is thus the
result, not the beginning, as Strauss and Quinn seem to imply.10 In fact, and
in consonance with our observations above on the dangers of taking Eng-
8. In the framework of European Existential philosophy, Nicola Abbagnano

(1942) used terms like corporeità, corporeizzazione, etc., which do correspond
quite closely to present-day embodiment terms. For Abbagnano, knowledge, as
the result of active “research” (ricerca) is only possible through the existence of
the body which, so to speak, opens up our mind to the existence of other entities
and, through it, to the knowledge of our own existence.
9. Bourdieu used the Latin term, habitus, which he always writes in italics, as a
means to avoid confusion with habitude, a distinction that has not always been
rightly understood. As Mounier writes (2001: 41), “Habitus but not habitude, in
order to signify clearly that it is no automatic mechanism for the reproduction of
preestablished schemas, but a generating principle for those products of action
which cannot be mechanically deduced from the objective conditions of their
production.” In this and other respects, Bourdieu’s habitus must not be con-
fused – equated – with William James’ notion of habit, which is a simple means
of (cognitive, but also sensorimotor) lowering of effort (cf. Moya Santoyo and
García Vega 2001)
10. See also Lizardo (2003) on the problems of correctly understanding the cogni-
tive character of the habitus.
lish as the linguistic point of reference, some of the problems involved in

the understanding and treatment of embodiment in other languages may be
due to the uncritical acceptance of one meaning of the word in English,
instead of the complete meaning of the word, as a whole. The same can
hold, perhaps, for the indeterminacy in the uses of the term, which Ziemke
(2003) analyses in detail. It would seem that what is sometimes sought is
not so much the adequacy of a term+concept as its fit to the lexical mean-
ing of an English word, which in turn is understood in a certain way. A
problem that has been plaguing certain areas of linguistics at least since the
onset of Generative Grammar.
3.1. What is the body
Central to the comprehension of embodiment is its body part. Nothing ap-

parently more obvious than the definition of body; but things are not so
simple, as the problems of interpretation have shown. Burkitt’s (2002: 227)
comments, on the background of his discussion of Bourdieu’s habitus,
come directly to our point:
[T]he body is not to be thought of as a discreet entity, for we can consider
the bodily habitus only insofar as we also consider the technological means
through which the body operates and turns itself into a self. This is also true
for the moral dispositions that moral habits inculcate, for these are depend-
ent on the social institutions in which people’s moral actions are located.
In fact, just as in the case of cognition, the body can be seen as an individ-
ual or a social reality. The body is socialized, it even reflects the social,
economic and cultural groups an individual belongs to, as Bourdieu em-
phasized repeatedly. We can talk about the body as a physical object or we
can choose to take into account also, in the first place, its functions: not
only the physical, but also the social functions of the human body. Not just
what the body is, but what the body does. And what the body does is also
outward bound, directed toward others, i.e., social. And that activity car-
ried out by the individual but in a social setting, modifies the body itself at
the same time. The body, just like cognition – and rather counterintuitively,
too – is also social, collective. And if cognition is unthinkable without the
body, the opposite also holds: the activity – and the mere external and in-
ternal reality – of the body is directly linked to the activity of cognition.
And if the architecture and organization of our brain is susceptible to
modification by our experience of the world, and this experience necessar-
ily has to include the individual’s activity, as in Edelman’s proposal

(Edelman 1992; Thelen and Smith 1994), there is no way out of this indis-
solubility of body and mind. Taken in isolation and/or as limited to the
individual, however, we know the many problems that this dichotomous
perspective has caused throughout the history of Western thinking but also
in the case of Cognitive Linguistics and Cognitive Science.
Certainly, this view reminds us of Maturana and Varela’s (1987) theory
of enaction (see also Varela, Thompson and Rosch 1991): even a biologi-
cal species is to be defined primarily by the perceptible enactions of its
members, not only by their genetic endowment. In terms of activity, far
from neural and even organismic solipsism, the body and the mind are
undistinguishable: we can talk about one or the other, study them sepa-
rately but just to make the analysis simpler.
4. Situatedness
But we have to widen our understanding of cognition even further. We

have to include situatedness in cognition. I am not going to enter into the
historical details of this view which, interestingly, goes back to some much
older proposals, especially those of the Soviet school of psychology of the
1930’s as represented mainly by Vygotsky, Luria, Voloshinov and Le-
ont’ev (for recent evaluations and a review of its historical background, see
Luria 1976–2003; Frawley 1997; Cole et al. 1997; Ratner 2000; Bedny,
Karwowski and Bedny 2001); but also to Dewey’s theory of action (Garri-
son 2001), or the French philosopher, Merleau-Ponty (Burkitt 2003;
Varela, Thompson and Rosch 1991, passim) and, more recently, Pierre
Bourdieu (see Burkitt 2002 for a brief comparative analysis of his and
Merleau-Ponty’s ideas in this respect; for a much more detailed introduc-
tion, Mounier 2001). Similarly, as Zlatev (2003: 306) points out, certain
ideas of Ludwig Wittgenstein’s are in a similar vein:
The conceptual framework of situated embodiment […] incorporates the
principle of embodiment […] emphasized within cognitive semantics, but
complementing it with Wittgenstein’s (1953) view of language as “forms of
life” embedded, or situated, within socio-cultural practices.
The notion of situatedness is a great step forward, for it now incorporates
the need to consider the specific, concrete sociocultural situation in which
the individual’s cognitive activity is to take place. Even in the traditional
views of embodiment, it is the individual alone who is referred to, in isola-
tion from (a) other individuals, (b) the activities to be carried out, and (c)
the sociocultural component of the environment. In contrast, situatedness
leads directly to interaction, to interindividual contact, because practically
any possible sociocultural environment includes interaction with other
individuals.
Thus, complex forms of non-individualistic cognition enjoy a long-
ranging tradition which can fruitfully be made use of in CL. In this view,
cognition is impossible to dissociate from interaction, understood as social
activity. That is, cognition is not just “something that takes place” inside
the individual’s brain, or only in relation to the individual body’s active
perception, or apperception, of the environment, i.e., embodiment, but
something that is done, enacted in relation with the individual’s whole
activity in a particular social and cultural setting or situation. As this view
of cognition implies collective activity and interaction, it can be seen as a
collective form of cognition (cf. also Sharifian, this volume)
Similarly, Michael Tomasello’s recent work (1999, 2000a, 2000b, To-
masello and Rakoczy 2003) points in this same direction when considering
the phylogenesis and ontogenesis of human cognition:
Following the lead of Vygotsky […], Bruner […], Cole […], and other cul-
tural psychologists, my view is that what makes human cognition unique,
more than anything else, is its collective nature (Tomasello 1999). That is,
all of the many artifacts that enable and empower human cognition […] are
the joint product of many people working over many years, combining and
accumulating skills and knowledge. (Tomasello 2000a: 357)
In fact, Tomasello’s view of imitation, attention to other people’s actions
and development of a “theory of mind” as the central element in the acqui-
sition of language by children, also as opposed to the shortcomings of
those same social activities in apes, witnesses the extraordinary importance
of social, i.e., collective activity, for the development of individual cogni-
tion.
Similarly, palaeontologists point to the richness of the social interaction
of Homo sapiens in contrast to that of Neanderthals (Homo
neanderthalensis) as the main reason for the prevalence of the former (Ar-
suaga 1999, 2001; Arsuaga and Martínez 1998), instead of some – impos-
sible to demonstrate – pre-eminence of any a priori cognitive abilities (after
all, Neanderthals had a bigger brain!).
In Cognitive Linguistics, mainly but not only when we are dealing with
the multiplicity of human languages and their nearly limitless variability,
we need to pay particular attention to the situatedness of cognition. Ac-

cording to the school of situated cognition,
behavior can only be understood in the context of complex real-world situa-
tions. An important focus of research should therefore be the relationship
between people and the external world (and how the behavior of people is
coordinated with the external world) without the mediation of mental plan-
ning (i.e., without explicit inferencing over descriptive models of the world
and human behavior). (Mandelblit and Zachar 1998: 253)
5. Beyond situatedness
In situated cognition special attention has to be paid to collective forms of

behaviour and activity. The individual’s cognition bears the imprint of the
types of social activity an individual can be expected to carry out in the
sociocultural group(s) s/he belongs to. And the individual’s cognition will
heavily depend on the conditions of the activity itself (see León 2002;
Semin and Smith 2002; Hirose 2002; Alterman and Garland 2000; Ratner
2000; Clark 1999). But what is more, the (individual) cognition of all the
individuals participating in similar collective activities will develop in
similar directions, in accordance with the needs imposed by those activi-
ties: a process of self-organization takes place. We can say that in relation
to a particular activity, all the participating individuals will collaborate in
such a way that one can speak of distributed or collective cognition. As the
action in question is carried out through the common activity of a number
of collaborating individuals, the term synergic cognition can be proposed.
It is to be understood as a special form of distributed cognition where the
interaction itself plays a prominent role, also over time, as a collectively
(and therefore synergically) established and socially accepted cognitive
activity; it is the result of a historical process (see also Musolff this vol-
ume). Mandelblit and Zachar (1998: 254) make these comments on distrib-
uted cognition:
Cognitive activity may involve processes internal to the single individual,
the individual in coordination with a set of tools, or a group of individuals in
interaction with each other and a set of tools […]. The different individuals
and tools constitute the unit of cognition rather than merely modifying or
amplifying the internal structures of a single mind.
Now, these observations fit human activity and cognition, but also lan-
guage. In fact, this view is the modernization and further development of
the paradigm of the study of language as (social) activity (in connection
with the Soviet psychological school, see Leont’ev 1969). We may summa-
rize things as follows:
1) Human languages exist only in the form of social activity

2) Linguistic activity is essentially collective, cross-individual, i.e., it is
not simply carried out inside a (social) group, rather the reasons for its
realization, the form of its realization and the results of the activity it-
self are collective, social in nature; in other words, the process of lin-
guistic activity cannot be understood solely in terms of the individual.
At the same time, language is a part of each individual's cognitive sys-
tem and the link between the individual and the collective aspects has to
be the centre of our research.
3) As a direct consequence of its social aspect, language is an inherently
historical phenomenon. Only if exclusive attention is paid to the indi-
vidual aspect can history be forgotten. But, at the same time, due atten-
tion has to be paid to the apparent atemporality of an individual's cog-
nition and language. The tension between both inseparable aspects of
the same phenomenon was in part the object of study of the Soviet
school of psycholinguistics in the 1920's and 1930’s, which was able to
show that even at the level of the individual, change is inescapable: in-
dividual cognition was in fact affected by changes that were primarily
social in nature. Phenomena that are nowadays usually examined solely
at the individual level, like metaphor and metonymy (but see Yu, this
volume, for an alternative non-individualistic view of metaphor), when
examined in historical depth, show the extent to which most of our in-
dividual, contemporary metaphors and metonymies are in fact the result
of social, historic crystallisation. This tension between the historical
and the ahistorical, apart from its philosophical and methodological in-
terest, is also the object of research by those investigating theories of
complex systems.
4) Linguistic activity (linguistic use) determines linguistic forms, i.e., lin-
guistic structures. In a double process:
a. as in any form of activity, a number of alternatives exist; one (or
more) of them are selected as the preferred form(s) of activity in
stipulated contextual conditions; and
b. these preferred forms of activity are then integrated in the whole life
of the individual: they can correspond to Bourdieu’s habitus. Being
“incorporated” – or embodied, in this sense –, they become a part of
that individual’s cognitive abilities for action.
5) Through cognitive integration in the individual mind, those preferred
forms of language activity, of language use, are then gradually en-
trenched in the individual’s mind, up to a point where their originally
immediate relation with activity is lost, and they become a part of what
we like to see as “cognition” tout court, i.e., individual cognition.
Many examples could be introduced in order to show that this is how

things probably work. Let me refer to a single case in very brief terms.
Metaphor and metonymy are usually analysed in terms of the individual, as
if they were the instantaneous or, better, atemporal application of certain
cognitive mechanisms. However, in most cases the metaphors and me-
tonymies accepted and used in/by a sociocultural community are the his-
torical product of the synergic cognitive activity of the community,11 which
determines – in the way that it is done with other habitus – the structure of
the metaphoric fields. In the case of metonymy, it is the synergically estab-
lished and allowed or disallowed forms of reference which determine the
forms, use and structure of metonymical reference. The individuals, in the
vast majority of cases, limit themselves to taking up and using socially
established “labels” and applying them without the need for any special
cognitive activity on their part. As is the case with all forms of distributed
and synergic cognition, the participation of the collectivity brings about a
significant decrease in the individuals’ cognitive effort (for more details,
Bernárdez 2005; and in a similar vein, in this volume, see also Kimmel,
Kristiansen, Yu and Zinken, Hellsten and Nerlich).
6. Is language variation possible at all?
In our view, linguistic variation is the direct result of the character of lan-
guage as a social activity, under the effect of the contextual conditions of
11. Sharifian (2003) clearly implies that his “cultural conceptualizations” are sus-
ceptible to historical change, as everything social. This can be interpreted in the
sense that metaphorical conceptualization, metaphors, change over time. The
same can be said of Zinken, Hellsten and Nerlich’s discourse metaphors (this
volume). See also Frank (this volume).
linguistic activity itself and of those imposed by the participation of a

number of individuals, over time, in a common activity. This is also the
origin of our current construct language as opposed to real languages: lan-
guage would be what is left after peeling away everything that is directly
related to social cognition and activity. Apparently, what is left is very
little indeed. According to Hauser, Chomsky and Fitch (2002), the only
element in human language that seems to satisfy this condition is recursiv-
ity.
7. The habitus: A tool for the study of collective, situated,

embodied cognition and language
In addition, we propose integrating Bourdieu’s concept of the habitus into

CL and situated Cognitive Science. Although the habitus, as all of Bour-
dieu’s basic concepts, has been the object of constant modification by its
author, the following can be used as its basic definition; the habitus are
systèmes de dispositions durables et transposables, structures structurées
prédisposées à fonctionner comme structures structurantes, c'est-à-dire en
tant que principes générateurs et organisateurs de pratiques et de représenta-
tions qui peuvent être objectivement adaptées à leur but sans supposer la vi-
sée consciente de fins et la maîtrise expresse des opérations nécessaires pour
les atteindre, objectivement “réglées” et “régulières” sans être en rien le
produit de l'obéissance à des règles et, étant tout cela, collectivement or-
12
chestrées sans être le produit de l'action organisatrice d'un chef d'orchestre.
(Bourdieu 1994: 88)
The habitus is simultaneously social and individual; it produces socially
established forms of practice, i.e., individual behaviours that agree with the
social preferences; the habitus is acquired by the individual through learn-
ing, experience and social practice, which leads necessarily to interper-
sonal variation: no two individuals can have exactly the same background,
12. [Systems of durable and transposable dispositions, structured structures ready to

function as structuring structures, i.e., as generative principles and organisers of
practices and representations that can be objectively adapted to their goal but
without the conscious knowledge of the means and the mastery of the operations
needed to reach that goal, objectively ‘ruled’ and ‘regular’ without being at all
the result of the following of rules and, being all that, collectively orchestrated
without being the product of the organising action of a conductor.]
so their set of habitus will always be partially different, as will any indi-
vidual form of carrying out any activity, i.e., an individual habitus. At the
same time the constant process of feedback and individual reelaboration of
the available habitus will insure that a high degree of similarity does arise,
especially among the members of a particular cultural, economic, profes-
sional or, in general, social groups. And this, in turn, brings about social
differentiation (Bourdieu 1979).
Carl Ratner’s comments (2000: 11) on this issue are especially enlight-
ening:
The habitus is a set of expectations, assumptions and dispositions to react
which result from particular forms of social experience with particular social
conditions. Therefore, people’s actions are not freely constructed, rather
they are guided by the socially built-up habitus. […] Social experience is
profoundly embedded in the habitus and in ensuing psychological functions
and behavior. Social experience is not only internalized intellectually; it be-
comes inscribed in our bodies.
The habitus, then, gets entrenched (embodied) in the individual’s mind,
and its functioning is mainly unconscious, although the individual can be-
come conscious of his/her realization of a particular habitus in certain con-
ditions, mainly when confronted with some anomalous circumstance or
when unexpected results are observed. But habitus is also firmly entren-
ched in the body:
[L’] importance du corps et de la posture, cette “géométrie dans le monde
sensible” […] l’analyse structurale la néglige totalement par préjugé intel-
lectualiste, […] parce qu’elle n’est pas pensée, mais simplement agie. Dans
la mesure en effet où le structuralisme s’intéresse avant tout aux représenta-
tions mentales et aux opérations logiques qui y sont inscrites, il ne peut pen-
ser le corps que comme représentation du corps, en ignorant la physique
corporelle qui découle de sa matérialité.13 (Mounier 2001: 25–26)
Interestingly, recent studies on the consciousness and awareness of actions
realized by oneself or by other individuals (Frith 2002) have shown that we
carry out our actions without much or any previous conscious planning and
13. [The importance of the body and its position, this “geometry in the sensitive
world” […] was neglected by structural annalists due to intellectualistic preju-
dice […] because it is not thought, but simply done. As structuralism is espe-
cially interested in the mental representations and the logic operations inscribed
in them, this school cannot think of the body but as a representation of the body,
while ignoring the bodily physics derived from its materiality.]
in a basically automatic fashion. Only when unexpected results or conse-

quences arise does awareness reappear. In fact, the speed of our actions is
too high for us to be able to really “feel” what we are doing, and we simply
assume what is happening: we do not “feel”, even less instruct the move-
ment of our arm when grasping something:
What we are aware of must be based on predicted rather than actual sensa-
tions. We are very surprised if the actual sensations do not match those we
predicted, as when we pick up an object that is much lighter than we antici-
pated. (Frith 2002: 483)
Similarly, the habitus must not be confused with the “habit”. Indeed, it was
to avoid this confusion that Bourdieu selected the Latin term from the be-
ginning. Nonetheless, precisely that cognitive aspect of the term, its confu-
sion in English with the concept of “habit”, was what led to the widespread
misinterpretation and consequent rejection of the concept in US sociology,
according to Lizardo (2003). On the other hand, a recent proposal (Roos
and Rotkirch 2003) opened up the possibility of having some habitus ge-
netically established in human beings, rather in the vein of evolutionist
psychology:
[W]e could treat habitus as something between nature and culture, as a
meeting point of the two in the sense that habitus contains both extremely
permanent elements of human nature and the variability brought about by
cultural and social adaptation. (Roos and Rotkirch 2003: 4)
According to their proposal, the following would have to be added to
Bourdieu’s characterization of the habitus:
1. The fact that lots of our bodily functions and emotions are based on
evolved characteristics
2. The fact that the ways in which habitus-based actions (instincts) function
and work back in the society are to some extent biologically bounded
and determined. (Roos and Rotkirch 2003: 5)
The notion of the habitus will prove fruitful as it can provide a unified
account of other recent cognitive concepts which, although in different
ways and with different goals, also try to link cognition and linguistic ac-
tivity, as Sharifian’s (2003) cultural conceptualizations and schemas,
Werth’s (1999) megametaphors,14 Zinken, Hellsten and Nerlich’s (this
14. “[M]etaphors can also be sustained, as a kind of undercurrent, over an extended

text, which allows extremely subtle conceptual effects to be achieved” (Werth
1999: 323).
volume) discourse metaphors and cultural metaphors, Kristensen’s (this

volume) idealized cultural models, or Kimmel’s (this volume) notion of
cultural embodiment.
Much remains to be studied about the habitus in relation to cognition,
activity, the human body and language. For example, Quist (2002) ap-
proaches the concept of the linguistic standard in terms of the habitus,
while Mounier summarizes the importance of the habitus as a mediator
between the individual and the social:
L’agent est donc comme la monade leibnizienne, à la fois individu singulier
et reflet d’une totalité à laquelle il appartient. Guidé dans sa vie quotidienne,
dans sa confrontation à l’événement même le plus inattendu par “un ensem-
ble de dispositions durables” inscrites en lui, ses actions ne se définissent ni
comme le pur produit de sa volonté consciente […] ni comme des réponses
automatiques à des stimuli, mais comme un processus continu d’invention
limité par les conditions objectives “appréhendées à travers les schèmes so-
cialement constitués qui organisent sa perception”.15 (Mounier 2001: 41)
8. The lessons from biology
Our proposals can find confirmation outside linguistics and cognitive sci-
ence proper in two areas: the study of collective animal behaviour and the
physiological means for collective interaction. Remember that one of the
main methodological tenets of the Cognitive Linguistics enterprise is that
confirmation has to be sought from different, independent sources and,
hence, converging lines of evidence. Some very brief notes will have to
suffice here.
First, animal behaviour. It is clear that animals – not only apes, or even
primates – are able to coordinate among themselves in order to carry out an
activity (Conradt and Roper 2003; Rands et al. 2003; Visscher 2003; Susi
15. [The agent is thus similar to Leibniz’s monad, at the same time a single individ-
ual and the reflection of the whole he belongs to. Guided in his everyday life, in
his confrontation with even the most unexpected event, by “a set of durable dis-
positions” inscribed [or: embodied!] in him, his actions are defined neither as
the mere result of his conscious will nor as the automatic response to stimuli,
but as a continuous process of invention, limited by the objective conditions
“apprehended through the socially constructed schemas which organize his per-
ception.]
and Ziemke 2001). Human beings dispose, thus, of an extremely old

mechanism for the coordination of behaviour among individuals. Further-
more, this type of coordination needs to be studied through recourse to the
tools of analysis afforded by theories of self-organization (Thelen and
Smith 1994; Bernárdez 1995; Kelso 1995).
Second, the physiological apparatus. Apart from other points of interest,
the so-called Mirror Neuron System identified first in apes, then in humans
(Stamenov and Gallese 2002; Kohler et al. 2002; Ferrari et al. 2000; Iaco-
boni et al. 1999; Rizzolatti and Arbib 1998) shows that we dispose of a
special system of neurons, situated in close vicinity to the brain areas in-
volved in language functions, specialized in the (visual or acoustic) identi-
fication of our own actions and those carried out by other conspecifics.
These neurons are activated not only when the action is fully carried out,
but also when it is simply intended.16
We are thus pre-wired for social interaction, for identifying ourselves in
other people’s actions. The existence and development of what has been
called “theories of mind”, which are now seen by many as one of the basic
features of human cognition (Tomasello 1999), and the proposal to see
much of our cognition in terms of “cognitive simulation” (Hesslow 2002),
are just two extremely significant aspects of the importance of social inter-
action for the functioning of human cognition. Both sets of facts point in
the same direction: that of a basic gapless continuity between the body and
the mind, the individual and the social group. In all these areas, including
imitation, understanding of others, etc., it is always necessarily the exter-
nal, perceivable element that has to be at the forefront: enaction, interac-
tion, linguistic activity and usage.
9. Conclusion
From time to time it is necessary to look back at what we are doing and try
to discern what might be wrong or, perhaps, simply less inadequate. From
this general perspective I have identified a couple of significant problems
that should be solved in the future developments of Cognitive Linguistics.
Both correspond to two types of reductionism.
16. Recall, by the way, what was said above about our awareness of our own ac-
tions: it is our intention, not the action per se, that is the object of awareness.
Firstly, there is the tendency to take one language as fully representative

of human language in general and draw general conclusions on language
and cognition from that single language – or from a very limited sample of
languages. The main reason for this tendency is the idea that any single
human being is fully representative for the whole of mankind.
Secondly, there is the tendency to limit the scope of our study to indi-
vidual cognition and language, abstracting away from both the element of
cognition as activity and the social settings in which all activity, and hence
most linguistic and cognitive processes, take place.
Both tendencies go back to Chomskyan principles, which tried to see all
of language and cognition in the mind of the individual. Through the years
a number of shifts in perspective have taken place leading to a much
broader and more open view of cognition. These shifts correspond to
trends set up in certain non-dominant areas of linguistic study (mainly
functional and textual linguistics); thus, nowadays cognition is seen as
embodied, i.e., as taking place not only in the mind or the brain but in the
whole body, including its functions, activities and contact with the individ-
ual’s environment. However, as we have seen, the notion of embodiment
still needs much refinement and clarification, as it is frequently used in a
rather intuitive, non-critical manner. In a further step forward, cognition is
also seen as situated, i.e., as implying activities carried out by the individ-
ual under certain socioculturally given conditions.
This enables us to go even further and break the limits of the individual,
even in contact with the environment. In this paper I have tried to show that
it is also necessary to include a form of cognitive activity carried out col-
lectively by a group of individuals, which might be termed synergic cogni-
tion, and which implies forms of interindividual collaboration for the
solving of problems with the least individual cognitive effort. I have briefly
shown that similar advances have been made both in the study of animal
behaviour and in the understanding of our brain functions, especially with
the discovery, in apes and humans, of what is called the mirror-neuron
system (Rizzolatti and Craighero 2004).
This proposal is also very close to a number of recent approaches such
as that of distributed cognition and most of the papers in this volume, es-
pecially those by Kimmel, Kristiansen, Yu, Zinken, Hellsten and Nerlich
who approach a number of significant issues in cognitive linguistics using
sociocognitive approaches. Finally, I have tried to show the potentiality of
Bourdieu’s theory of practice for the type of cognitive and linguistic study
approached here.
References
Abbagnano, Nicola
1942 Introduzione all’esistenzialismo. Torino: Taylor.
Alterman, Richard, and Andrew Garland
2000 Convention in joint activity. Cognitive Science 25: 611–657.
Arsuaga, Juan Luis
1999 El collar del neandertal. Madrid: Temas de Hoy.
2001 El enigma de la esfinge. Las causas, el curso y el propósito de la
evolución. Barcelona: Random House Mondadori.
Arsuaga, Juan Luis and Ignacio Martínez
1998 La especie elegida. La larga marcha de la evolución humana.
Madrid: Temas de Hoy.
Barlow, Michael and Suzanne Kemmer (eds.)
2000 Usage Based Models of Language. Stanford (CA): CSLI Publica-
tions.
Bateson, Patrick
2002 The corpse of a wearisome debate. (A review of Pinker 2002). Sci-
ence 297: 2212–2213.
Bedny, Gregory, Waldemar Karwowski and Marina Bedny
2001 Implications of Activity Theory for the study of human work. Inter-
national Journal of Cognitive Ergonomics 5(4): 401–420.
Bernárdez, Enrique
1995 Teoría y epistemología del texto. Madrid: Cátedra.
1999 Some reflections on the origins of cognitive linguistics. Journal of
English Studies 1: 9–27.
2005 La cognición y el lenguaje: Lo individual y lo social. In: Pedro A.
Fuertes Olivera (ed.), Lengua y sociedad. Investigaciones recientes
en lingüística aplicada, 39–60. Valladolid: Universidad de
Valladolid.
Bourdieu, Pierre
1979 La distinction. Critique sociale du jugement. Paris: Les Éditions du
Minuit.
1980 Le sens pratique. Paris: Les Éditions de Minuit.
1994 Raisons pratiques. Sur la théorie de l’action. Paris: Éditions du
Seuil.
Burkitt, Ian
2002 Technologies of the self: Habitus and capacities. Journal for the
Theory of Social Behaviour 32(2): 219–237.
2003 Psychology in the field of being. Merleau–Ponty, ontology and social
constructionism. Theory and Psychology 13(3): 319–338.
Chomsky, Noam
1986 Knowledge of Language. New York: Praeger.
2002 Embodiment. In: Encyclopedia of Cognitive Science: 1102–1108.
London: Macmillan.
Clark, Andy
1999 Where brain, body and world collide. Journal of Cognitive Systems
Research 1: 5–17.
Cole, Michael, Yrjö Engeström and Olga Vásquez (eds.)
1997 Mind, Culture and Activity. Cambridge: Cambridge University Press.
Conradt, L. and T.J. Roper
2003 Group-decision-making in animals. Nature 421(9 January): 155–158.
Corominas, Joan
1967 Breve diccionario etimológico de la lengua castellana. Madrid:
Gredos.
Damasio, Antonio
1994 Descartes’ Error. Emotion, Reason and the Human Brain. New
York: Grosset/Putnam.
Damasio, Antonio
1999 The Feeling of What Happens. Body and Emotion in the Making of
Consciousness. San Diego: Harvest.
Edelman, Gerald M.
1992 Bright Air, Brilliant Fire. On the Matter of the Mind. New York:
Basic Books.
Ehrlich, Paul and Marcus Feldman
2003 Genes and cultures. What creates our behavioral phenome? Current
Anthropology 44(1): 87–107.
Ferrari, Pier Francesco, Evelyne Kohler, Leonardo Fogassi and Vittorio Gallese
2000 The ability to follow eye gaze and its emergence during development
in macaque monkeys. Proceedings of the National Academy of Sci-
ences USA 97(25): 13997–14002.
Frank, Roslyn M.
tems approach to shifting perspectives on ‘language’
Frawley, William
1997 Vygotsky and Cognitive Science. Cambridge, Mass.: Harvard Univer-
sity Press.
Frith, Chris
2002 Attention to action and awareness of other minds. Consciousness and
Cognition 11: 481–487.
Fuchs, Catherine
1997 Diversité des représentations linguistiques: Quels enjeux pour la
cognition? In: Catherine Fuchs and Stephane. Robert (eds.) Diversité
des langages et représentations cognitives, 5–24. Paris: Ophrys.
Garrison, Jim
2001 An introduction to Dewey’s Theory of Functional “Trans-Action”:
An alternative paradigm for activity theory. Mind, Culture and Ac-
tivity 8(4): 275–296.
Geeraerts, Dirk
1999 Idealist and empiricist tendencies in cognitive semantics. In: Janssen
and Redeker (eds.): 163–194.
Gibbs, Raymond W. and Teenie Matlock
1999 Psycholinguistics and mental representations. Cognitive Linguistics
10(3), 263–269.
Harder, Peter
1999 Partial Autonomy. Ontology and methodology in cognitive linguis-
tics. In: Janssen and Redeker (eds.), 195–222.
Harder, Peter
2003 The status of linguistic facts: Rethinking the relation between cogni-
tion, social institution and utterance from a functional point of view.
Mind and Language 18(1): 52–76.
Hauser, Marc D., Noam Chomsky and W. Tecumseh Fitch
2002 The Faculty of Language: What is it, who has it and how did it
evolve? Science 298: 1569–1579.
Hesslow, Germund
2002 Conscious thought as simulation of behaviour and perception. Trends
in Cognitive Sciences 5(6): 242–247.
Hirose, Naoya
2002 An ecological approach to embodiment and cognition. Cognitive
Systems Research 3: 289–299.
Hull, David L.
2002 Nurturing a view of human nature (A review of S. Pinker: The Blank
Slate). Nature 419: 251–252.
Iacoboni, Marco, Roger P. Woods, Marcel Brass, Harold Bekkering, John C. Maz-
ziotta and Giacomo Rizzolatti
1999 Cortical mechanisms of human imitation. Science Vol. 286: 2526–
2528.
Janssen, Theo and Gisela Redeker (eds.)
1999 Cognitive Linguistics: Foundations, Scope and Methodology. Berlin:
Mouton de Gruyter.
Johnson, Mark
1987 The Body in the Mind. Chicago: University of Chicago Press.
Kelso, J.A. Scott

1995 Dynamic Patterns. The Self-organization of Brain and Behavior.
Cambridge, Mass.: The MIT Press.
Kemmer, Suzanne and Michael Barlow
2000 Introduction: A usage-based conception of language. In: Barlow and
Kemmer (eds.), vii–xxviii.
Kimmel, Michael
this vol. Properties of cultural embodiment: Lessons from the anthropology of
the body
Kristiansen, Gitte
this vol. Idealized cultural models: The group as a variable in the develop-
ment of cognitive schemata
Kohler, Evelyne, Christian Keysers, M. Alessandra Umiltà, Leonardo Fogassi,
Vittorio Gallese and Giacomo Rizzolati
2002 Hearing sounds, understanding actions: Action representation in
mirror neurons. Science 297: 846–848.
Labov, William
1966 The Social Stratification of English in New York City. Washington,
D.C.: Center for Applied Linguistics.
León, David de
2002 Cognitive task transformations. Cognitive Systems Research 3: 349–
359.
Leont’ev, Aleksej A.
1969 Sprache – Sprechen – Sprechtätigkeit. Übersetzt und herausgegeben
von C. Heeschen und W. Stölting. Stuttgart: Kohlhammer.
Lizardo, Omar A.
2003 The cognitive origins of Bourdieu’s habitus bringing together culture
and cognition. Journal for the Theory of Social Behaviour 34: 375–
401. http://www.nd.edu/~olizardo/papers/bourdieu.pdf
Luria, Aleksandr R.
1976–2003 Desarrollo histórico de los procesos cognitivos. Trans. Arturo
Villa; revision by Luis Cebrián Tornos. Madrid: Akal, 2003.
Mandelblit, Nili and Oron Zachar
1998 The notion of Dynamic Unit: Conceptual developments in cognitive
science. Cognitive Science 22(2): 229–268.
Maturana, Humberto, and Francisco J. Varela
1987 The Tree of Knowledge: The Biological Roots of Human Under-
standing. Boston: Shambhala.
1945 Phénoménologie de la perception. Paris: Gallimard.
Mounier, Pierre
2001 Pierre Bourdieu, une introduction. Paris: Pocket/La Découverte.
Moya Santoyo, José and Luis García Vega
2001 James. Madrid: Ediciones del Orto.
Musolff, Andreas
this vol. The embodiment of Europe: How do metaphors evolve?
Pinker, Steven
2002 The Blank Slate: The Modern Denial of Human Nature. London:
Allan Lane/Viking.
Quist, Pia
2002 Sproglig habitus, symbolisk magt og standardisering. Pierre
Bourdieus begreber anvendt på sociolingvistik. IDUN 15: 119–130.
Rands, Sean A., Guy Cowlishaw, Richard A. Pettifor, J. Marcus Rowcliffe and
Rufus A. Johnstone
2003 Spontaneous emergence of leaders and followers in foraging pairs.
Nature 432: 432–434.
Ratner, Carl
2000 Agency and Culture. Journal for the Theory of Social Behavior 30:
413–434. [communication.ucsd.edy/MCA/Paper/00_01/agency.htm]
(Accessed 3–09–02)
Ratner, Carl and Lumei Hui
2003 Theoretical and methodological problems in cross-cultural psychol-
ogy. Journal for the Theory of Social Behaviour 33(1): 67–94.
Riegler, Alexander
2002 When is a cognitive system embodied? Cognitive Systems Research
3: 339–348.
Rizzolatti, Giacomo and Michael Arbib
1998 Language within our grasp. Trends in Neural Sciences 21(5): 188–
194.
Rizzolatti, Giacomo, and Laila Craighero
2004 The Mirror-Neuron System. Annual Review of Neuroscience 27:
169–192.
Rohrer, Tim
2007 The body in space: Dimensions of embodiment. In: Tom Ziemke,
Jordan Zlatev and Roslyn M. Frank (eds.), Body, Language and
Mind. Vol. 1. Embodiment, 339–377. Berlin/New York: Mouton de
Gruyter.
Roos, J. P. and Anna Rotkirch
2003 Habitus, nature or nurture? Towards a paradigm of evolutionary
sociology. Paper presented at the European Sociological Association
Conference, Murcia, September 23–28, 2003. www.valt.helsinki.fi/
staff/jproos/habitusmurcia.htm
Ruiz de Mendoza, Francisco J. and José Luis Otal Campo

2002 Metonymy, Grammar and Communication. Granada: Comares.
Semin, Gün R. and Eliot R. Smith
2002 Interfaces of social psychology with situated and embodied cogni-
tion. Cognitive Systems Research 3: 385–396.
Sharifian, Farzad
2003 On cultural conceptualisations. Journal of Cognition and Culture
3(3): 187–207
guage
Sinha, Chris
1999 Grounding, mapping and acts of meaning. In: Janssen and Redeker
(eds.), 223–255.
tive Linguistics 11(1/2): 17–41.
Stamenov, Maxim I. and Vittorio Gallese (eds.)
2002 Mirror Neurons and the Evolution of Brain and Language. Amster-
dam: John Benjamins.
Susi, Tarja and Tom Ziemke
2001 Social cognition, artefacts and stigmergy: A comparative analysis of
theoretical frameworks for the understanding of artefact-mediated
collaborative activity. Journal of Cognitive Systems Research 2:
273–290.
Thelen, Esther and Linda B. Smith
1994 A Dynamic Systems Approach to the Development of Cognition and
Action. Cambridge, Mass.: The MIT Press.
Tomasello, Michael
2000 a Primate cognition: Introduction to the issue. Cognitive Science 24(3):
351–361.
2000 b First steps toward a usage-based theory of language acquisition.
Cognitive Linguistics 11(1/2): 61–82.
Tomasello, Michael and Hannes Rakoczy
2003 What makes human cognition unique? From individual to shared to
collective intentionality. Mind and Language 18(2): 121–147.
Varela, Francisco J., Evan T. Thompson and Eleanor Rosch
1991 The Embodied Mind. Cambridge, Mass.: The MIT Press.
Vesey, G. N. A.
1965 The Embodied Mind. London: George Allen and Unwin.
Vygotsky, Lev S.
1962 Thought and Language. Cambridge, Mass.: MIT Press. First pub-
lished 1934.
1978 Mind in Society. Ed. M. Cole. Cambridge, Mass.: Harvard University
Press.
Visscher, P. Kirk
2003 Animal behaviour: How self-organization evolves. Nature 421: 799–
800.
Webster’s Third New International Dictionary
1971 Chicago: Encyclopaedia Britannica.
Werth, Paul
1999 Text Worlds: Representing Conceptual Space in Discourse. London:
Longman.
Wittgenstein, Ludwig
1953 Philosophical Investigations. Oxford: Blackwell.
Yu, Ning
this vol. The relationship between metaphor, body and culture
Ziemke, Tom
Annual Meeting of the Cognitive Science Society, 1305–1310.
Mahwah, NJ: LawrenceErlbaum.
Zinken, Joerg, Lina Hellsten and Brigitte Nerlich
this vol. Discourse metaphors.
Zlatev, Jordan
1997 Situated embodiment: Studies in the Emergence of Spatial Meaning.
Stockholm: Gotab.
2003 Holistic spatial semantics of Thai. In: Eugene H. Casad and Gary B.
Palmer (eds.), Cognitive Linguistics and Non-Indo-European Lan-
guages, 305–336. Berlin and New York: Mouton de Gruyter.
Section B
The sociocultural situatedness

of scientific discourse
Entangled biological, cultural and linguistic origins
of the war on invasive species
Brendon M. H. Larson
Abstract
The language of invasion biology reflects its sociocultural situatedness with three
metaphorical elements: fears of invasion, an emphasis on competition, and preva-
lent militarism. These elements incorporate salient emotionally laden themes, which
help to convince biologists and their audience that invasive species (IS) are a
problem. I show that conceiving IS as invaders draws upon two congruent fears:
that our bodies will be invaded by disease and our nations by foreigners. Once IS
occur on a landscape, invasion biologists disproportionately perceive the interac-
tion between IS and native species as competitive – a bias that is common in biol-
ogy and alludes to the power of the competition metaphor. Finally, in concert with
prevailing militaristic approaches to problem-solving, invasion biologists use mili-
taristic language and actions to defend native landscapes and their species by ex-
terminating IS. While biologists may not consciously manipulate public opinion
about IS by using metaphors of invasion, competition and war, their uncritical use
naturalizes an antagonistic way of relating to the natural world that may be counter-
productive for conservation.
Keywords: competition, conceptual metaphors, CONTAINER image schema, evolu-

tionary biology, invasion biology, invasive species, militarism, rhetoric.
1. Introduction
Helicopters recently flew over Anacapa Island, one of the California Chan-
nel Islands, so that pellets of a deadly anti-coagulant could be dropped
along precise GPS gridlines to exterminate resident rats (Faulkner, Howald
and Ortega 2001). Because the rats were non-native,1 abundant, and had
1. Non-native species – also known as alien, exotic, introduced or non-indigenous

species – have been introduced by humans to “new, often distant, ranges”
170 Brendon M. H. Larson
been observed eating the eggs of rare (and native) seabirds, invasion biolo-
gists who oversaw the project could justify its $1 million cost. Invasion
biology was founded on concerns about species such as these rats, defined
as invasive species (IS) because they spread and become problematic after
humans introduced them. Only a small percentage of introduced species
become IS. However, these IS tend to have great effects on the pre-existent
community (see Mack et al. 2000; Baskin 2002 for reviews), so conserva-
tion biologists2 classify them as the second greatest threat to biodiversity
(Wilcove et al. 1998). They also have tremendous economic costs
(Pimentel 2002). In their influential review of biotic invasions, Mack et al.
(2000) advised that
Failure to address the issue of biotic invasions could effectively result in se-
vere global consequences, including wholesale loss of agricultural, forestry,
and fishery resources in some regions, disruption of the ecological processes
that supply natural services on which human enterprise depends, and the
creation of homogeneous, impoverished ecosystems composed of cosmo-
politan species.
Consequently, invasion biologists feel justified in eradicating IS; the rats,
for example, could gradually “homogenize” endemic communities of Ana-
capa Island.
Another classic case of an invasive species – the ruddy duck in Europe
– shows how invasion biologists justify the removal of a species. The
ruddy duck is native to North America, but escaped from wildfowl collec-
tions in the U.K. in the 1950s and began to spread through Europe (Milton
2000). They weren’t considered a threat until the early 1990s when they
entered Spain and began to hybridize with the rare, native white-headed
duck. Since hybridization with the ruddy duck could lead to extinction of
the white-headed duck, Spain began to kill its ruddy ducks. Trials to elimi-
nate ruddy ducks from the U.K. began in 1999, and were overseen by a
euphemistically named White-headed Duck Task Force. If ruddy ducks are
not removed from Britain, the argument goes, there will always be a source
for continued spread into neighboring European countries.
(Mack et al. 2000: 690). In contrast, native species occur in an area “naturally”,
having either evolved there or dispersed there from somewhere else.
2. Invasion biology is a major subdiscipline of conservation biology, which is
concerned with the more general issue of how to maintain biodiversity.
Entangled biological, cultural and linguistic origins 171
As a biologist I sympathize with these concerns, but I am also skeptical

because of how IS are framed.3 In the words of Takacs (1997: 8),
How can one feel about the natural world as strongly as I do, and as do the
biologists whose exploits I narrate, and not believe that those feelings ap-
proach the truth in some sense? How can I balance my healthy skepticism
about conservation biologists’ proselytizing on behalf of biodiversity against
my fervent hope that they succeed?
Invasion biologists derive substantial funding for and prestige from their
cause, but numerous critics have questioned whether this is warranted (e.g.,
Sagoff 1999; Subramaniam 2001; Chew and Laubichler 2003). Invasion
biology relies upon a narrative of native versus non-native that is seldom
questioned by invasion biologists. In a recent critique of my research on
potential implications of a metaphorical war against IS, for example, a
well-known conservation biologist wrote: “The bottom line for me is that,
given the abundant, massive, and seemingly insurmountable global conser-
vation problems that we face, the semantics of dealing with invasive spe-
cies is a low priority.” This comment belies a scientistic view that over-
looks the extent to which this issue is inextricable from pre-existent
cultural lenses. These lenses force us to think primordially in terms of “us”
and “them”, which is reflected in the use of linguistic categories such as
“native” and “invasive”, respectively.
There is extensive evidence that ecologists do not see the world “as it
is”, but through the eyes of their professional culture. These cultural influ-
ences have been documented by numerous historical studies (e.g., Fine and
Christoforides 1991; Journet 1991; Barbour 1995) and specific attention to
the over-representation of the notion of competition4 in ecological research
3. In a sense, conservation biologists have created IS, regardless of their effects.

Humans are inscribed within IS, not only because we introduced them, but also
because conservation biology itself is a human activity (Milton 2000). While
cultural and linguistic features partly constitute IS, they do have effects (just as
some native species do). Nonetheless, social problems only come about through
communication, and one of my primary concerns is with the transformations
that occur during this process.
4. Biologists classify competition, mutualism and predation as the three main types
of biotic interaction. Competition is defined as an interaction where both part-
ners are harmed by their interaction, whereas both benefit in mutualism (e.g.,
pollination systems, where insects derive nectar and/or pollen as “rewards” for
enabling sexual reproduction between plants), and one (the predator) gains and
the other (the prey) loses under predation.
(Boucher 1986; Keller 1991). The ecologist Keddy (1989: 163), for exam-
ple, proposed that: “Scientists can only draw models from the possibilities
of which they are aware, and perhaps ecology has been hampered by re-
stricted access to individuals (and ideas) offering co-operative models for
society and nature.” Related arguments have been made about the bias
towards militaristic metaphors in environmental science (Glotfelty 2000).
A recent paper in Science concluded that “we should be concerned
about what the frequent use of ‘natural enemies’5 (and the notable absence
of ‘natural allies,’ describing an equally familiar set of ecological interac-
tions) reveals about the ways in which we interpret nature through meta-
phorical lenses, especially in the current historical situation” (Chew and
Laubichler 2003: 53). Here, I argue that invasion biology unduly adopts
competitive and militaristic metaphors because of the cultural context in
which invasion biologists are situated. Specifically, invasion biology re-
flects three aspects of its sociocultural situatedness: contemporary fears of
invasion; a bias towards a competitive view of life; and the habit of apply-
ing militaristic metaphors to nearly every challenging situation. Invasion,
competition and war are large-scale metaphors that circulate nomadically
between segments of society, including science and society (Bono 1990;
Maasen, Mendelsohn and Weingart 1995). They also reinforce one an-
other, as small-scale individualistic competition is consistent with larger-
scale political militarism, which is often motivated by fears of invasion.
I employ the tools of Cognitive Linguistics to analyze these metaphors
(Lakoff and Johnson 1980), while also attending to their rhetorical (persua-
sive) effects (Eubanks 2000). The Lakoffian view of metaphor underscores
the extent to which our metaphors influence how we conceptualize and act
(Schön 1993). Bono (2003: 228) calls them “material metaphors: embodied
metaphors-in-action”. As an example, the invasion biologists Davis,
Thompson and Grime (2001: 3–4) observed that “ecologists during the past
few decades […] have focused on the headline invaders, a small group of
plants and animals that are not representative of the very large group of
species that are currently colonizing new areas of the globe [in part be-
cause] funding and publication pressures prompt ecologists to promote new
and exciting research themes”. However, they neglect the possibility that
the allure of “battling against invaders” itself creates the emotional excite-
ment of this field and its focus on dramatic cases and narratives.
5. “Natural enemies” are species that harm invasive species, but one of the points
made by Chew and Laubichler (2003) is that the phrase is often used vaguely.
I will not simply claim that biologists use these metaphors rhetorically
to convince the public of a problem; rather, in the spirit of Cognitive Lin-
guistics I will utilize examples from within the flagship journal of invasion
biology, Biological Invasions, to show how this “rhetoric” operates within
the field itself, revealing endemic patterns of thought. My approach fol-
lows Fine and Christoforides’ (1991: 377) study of the Great English Spar-
row War6: “Our claim is not that the proponents of attacks on sparrows
cynically manipulated nativist rhetoric in order to inflame passions, but
rather this set of nativist beliefs made sense in explaining the dangers of a
foreign interloper to the community of American birds.” While it may be
somewhat natural for invasion biologists to invoke prevailing metaphors
and narratives – discourse metaphors, as they are called by Zinken, Hell-
sten and Nerlich (this volume), this militaristic language not only restricts
the possibility of seeing their problem in other ways, but also links it to
large-scale political trends.
2. The conceptualization of “fears of invasion”
The term “invader” is culturally resonant because of its embodied basis;

that is, physiologically and mentally experienced fears that our bodies will
be invaded by disease and our nations by foreigners. These two issues af-
fect interpretation of IS because all three types of invasion are congruent,
particularly in their reliance on the CONTAINER image schema (see Rohrer
1995: 124–125; Chilton 1996: 197–198). Because of this schema, it is easy
to interpret the invasion of natural landscapes, simultaneously, as the inva-
sion of a metaphorically projected “person” and a “nation”. I will demon-
strate the first point by providing evidence for the conceptual metaphor
NATURAL LANDSCAPES ARE PERSONS, which allows IS to be understood as
a disease, and the second by considering linkages to NATION IS A PERSON,
where IS are interpreted as human invaders. In combination, these meta-
phors mutually reinforce one another and strengthen the case of invasion
biology within an unquestioned ontological framework.
6. English sparrows were introduced into North America from Europe in the early
1850s to control insects, but when they began to spread they were vilified and
attacked, just as IS today.
2.1. The image schemata structuring the conceptualization of invasion

biology
The notion of boundaries evoked by the CONTAINER image schema con-

tributes to fears of invasion, whether by disease, human invaders, or IS.
This schema derives from the experience of embodiment, which differenti-
ates our interior and exterior across a boundary (Johnson 1987). There is a
range of opinion within Cognitive Linguistics concerning the extent to
which image schemas are innate and individualistic versus developmen-
tally – and culturally – conditioned. In light of numerous critiques of the
former view (e.g., Gibbs 1999; Bono 2003; Zinken, Hellsten and Nerlich,
this volume), I will follow Santibáñez (2002), who defines image schemata
as “pervasive organizing structures in human cognition which emerge from
our bodily and social interaction with an environment at a preconceptual
level”. However, I will assume that cultural conditions during the ontogeny
of most biologists I am referring to (as well as Westerners in general) are
relatively consistent so that the schema is conventional even though it is
socioculturally situated. That is, this schema makes sense both to biologists
and those they try to reach out to rhetorically because it is so consistent
with everyday expressions and ways of relating.
The CONTAINER image schema distinguishes between inside and out-
side, a distinction that can be projected onto the world as a means to
structure and understand it (Lakoff and Johnson 1999: 32, 117, 380). In the
case of IS, this schema provides a powerful basis for reifying the boundary
between native and non-native species. Milton (2000: 242) employed a
case study of an IS, for example, to argue that “conservation [is] a bound-
ary maintaining exercise. In order to conserve the things that constitute
nature, the boundaries that separate them must be maintained, and in order
to conserve nature’s ‘naturalness,’ the boundary between the human and
the non-human must be preserved”. In some cases, these boundaries may
correspond with national boundaries, but they may also occur at smaller
scales, such as individual states, counties, biogeographic regions, national
parks, or local vegetation communities (see Figure 1). Even though ecolo-
gists currently doubt that communities are integrated wholes (see Soulé
1990: 234; Woods and Moriarty 2001: 172), invasion biologists continue to
metaphorically enforce their boundaries, which indicates how compelling
this schema has become for understanding biological systems.
Figure 1. Nesting of self within biogeographic region within nation. Each of

these levels depends upon an experiential CONTAINER image schema,
which is metaphorically projected in the case of biogeographic re-
gions and nations.
Invaders do not just equilibrate with their surroundings – they spread and
expand. This conceptualization derives from two additional kinesthetic
image schemas, PATH and FORCE, which depend on the CONTAINER schema
and contribute to the ease with which IS are associated with other kinds of
invaders. IS can expand into a predefined CONTAINER by expansion of their
own CONTAINER via the addition of a PATH schema. This schema “involves
structural elements such as starting point (origin), obstacle, destination
(endpoint), path and directedness toward the endpoint” (Chilton 1996:
199). The prevalence of this schema in invasion biology is indicated by
references to the “spread” and “expansion” of IS in 42 and 22 papers in
Biological Invasions, respectively.7 Typically, this is in terms of range
expansion, such as the “rapid expansion of this species’ range since its
arrival in North America” (Shurin and Havel 2002).
As their perimeter spreads, IS also exert a metaphorical force on natural
landscapes. The underlying schema of FORCE dynamics is constitutive of
7. I conducted an analysis of Biological Invasions because it is the only journal

solely dedicated to invasion biology. I searched for keywords within the first
five volumes (1999 through 2003) using the online Kluwer search engine. The
search captured occurrences of terms within abstract text, titles and keywords,
with each article counted only once in the totals given herein. A few matches
were deleted from the total if their usage was distinct from the examples cited.
There were a total of 166 substantive papers in the volumes covered here.
the field of invasion biology, as shown by use of the term “impact” in the
journal Biological Invasions.8 The first substantive article in the journal
was entitled “Impact: Toward a framework for understanding the ecologi-
cal effects of invaders” (Parker et al. 1999). In archetypal scientific prose,
the authors attempt to use unbiased language and to work objectively from
the evidence to conclusions. In this case, however, the authors reverse the
usual logic when they foreground the word “impact” (a negative effect) by
setting it off with a colon. Thereafter they refer to the potential “effects” of
invaders. The unstated enthymeme is that invaders exert a negative force,
and there is little need to discuss whether this is actually the case. Subse-
quently, another 37 papers refer to impacts of IS, and the term “impact”
constitutes fully 6% (13/219) of the words in one abstract (Forrest and
Taylor 2002).9 Invasion biologists created the journal Biological Invasions
in part to address their concerns about the expansive force of IS.
2.2. Invasive species conceptualized as disease
To understand IS as a disease, landscapes must first be personified. The

metaphorical projection NATURAL LANDSCAPES ARE PERSONS is supported
by two main lines of evidence. First, humans have utilized body-landscape
metaphors for millennia. As explained by Porteous (1986: 10):
The human body is the first landscape we encounter and explore. It is likely
that we carry the cognitive imagery in our heads as well as the actuality of
our own bodies as we approach the external environment. Landscape is our
second major encounter. For both practical and magical reasons, the appli-
cation of notions of the self to the environment of non-self makes sense. In
this way we humanize our environment, reduce its primeval unknownness
and terror, make it ours.
8. In the first five volumes of Biological Invasions there were two direct refer-
ences to species exerting pressure, and an additional three in the first issue of
volume 6. In some cases, however, native species exerted this pressure on inva-
sive ones.
9. In other places, force dynamics take a militaristic twist. For example, in a book
in the Worldwatch Environmental Alert Series on IS, Bright (1998: 24) reported
that “there is little consolation in the fact that 90 percent of these impacts are
‘duds,’ and only 1 percent of them really detonate. The bombardment is contin-
ual, and so are the detonations.”
But the strongest support for the NATURAL LANDSCAPES ARE PERSONS
mapping is invasion biologists’ use of the ecosystem health metaphor and
its entailments (for discussion, see Ross et al. 1997). Three papers in Bio-
logical Invasions referred to health, including the claim that “‘ecosystem
management’ strategies promoting healthy, undisturbed sites will not al-
ways be effective against invasive pest species” (Parker 2001) and two
papers by Bonneau, Shields and Civco (1999) that analyzed “the health of
hemlock forests infested by the hemlock woolly adelgid”. Also, Mack et al.
(2000: 693) discussed “community vulnerability to invasion”, which be-
speaks the idea of an integrated personified community. In each case,
healthy sites are relatively free of IS, and it follows that invasion biologists
can restore health and balance10 by removing them. As examples, nine pa-
pers referred to “restoration”, and Alpert and Maron (2000) entitled their
article, “Carbon addition as a countermeasure against biological invasion
by plants”. Even though invasion biologists may sometimes decry health
and balance metaphors they still help to define the field.
By extension from notions of human health, an ecosystem is considered
healthy if it contains few IS: IS ARE A DISEASE. Chilton (1996: 197; and
see Otis 1999), for example, observed that “[d]iseases are typically imag-
ined as invading the body from outside, a notion which rests both on the
CONTAINER schema and the warfare script”. The editor-in-chief of Biologi-
cal Invasions invoked this metaphor explicitly in his one page opening
editorial for the journal: “The resulting scale of hourly inoculations has led
to a proportional increase in successful introductions. The Earth is now
virtually itching with new invasions” (Carlton 1999, italics added). A total
of seven papers in the journal called IS an “infestation”, a term often used
to refer to parasitic disease, and Mack et al. (2000) included a section on
the “epidemiology of invasions”. By invoking the language of human
health and disease, invasion biologists lend support to the operation of
NATURAL LANDSCAPES ARE PERSONS, which provides one source domain
for preferring landscapes that are free of IS.
10. Implying the operation of the BALANCE image schema. Another abstract states
that introduced mammals have “pushed the competitive balance from native to
exotic species” (Holmgren 2002).
2.3. Invasive species conceptualized as human invaders
To understand IS as human invaders, biogeographic regions must first be

understood as nations. This is a natural association, since invasion biolo-
gists have been educated amidst realist political discourse, which presumes
the individual ‘in’ the state; […] the state itself with its containing and pro-
tective perimeters; and […] the outside world, the domain of the anarchic
international system. These elements, the inside, the boundary, and the out-
side, derive from a powerful and pervasive spatial image, that of an imper-
meable container. (Chilton 1996: 195)
Rohrer (1995: 125) has elaborated the resultant NATION IS A PERSON meta-
phor and its entailments in the context of political “rape”, which can be
applied to IS. Just as he observes that “The ‘rape of Kuwait’ is the rape of
the body of a metaphorically projected person via the ‘NATION IS A
PERSON’ metaphor,” I claim that for invasion biologists the invasion of
natural landscapes is the invasion of a metaphorically projected nation.
Although biologists may not be patriotic in the usual sense, their active
defense of biogeographic boundaries suggests that these are partly con-
ceptualized in terms of the culturally-prevalent NATION IS A PERSON meta-
phor. As Smart (1996: 276) has observed, “[t]he body of the nation is its
land, and this is often the object of national piety”. In the German context,
for example, Eser (1998: 102) explains that
Historically […] the idea of nature conservation has been part of the broader
concept of the conservation of ‘Heimat’. ‘Heimat’ means the place, where
people feel at home. It is not pure nature but a place where humans and na-
ture live together in harmony, dependent on each other. […] Spreading non-
indigenous plants are not a part of ‘Heimat’ in every sense of the word. They
are ‘aliens’, they ‘don’t belong’, they are unfamiliar to the people. They
seem to change the landscape more rapidly than humans are able to adapt to
[these changes]. Thus, they afflict the major function of Heimat: to guaran-
tee stability, safety and identity.
Once natural landscapes have been personified as nations, they can come
under threat from others: IS ARE HUMAN INVADERS. Numerous similarities
between IS and human invaders support this biological-political mapping
(Table 1), and thus encourage repeated use of the term “invader” within
invasion biology.11 This term ascribes purposiveness to the movement of

IS, which is enhanced by explicit personification. Because IS are “invad-
ers”, they are given malicious intent, even if unconsciously, which makes
them to some extent guilty simply by their name.
Table 1. Mapping between the concept of a human invader and that of

a biological invader.
Human invader Biological invasion

soldier or invader species
originate from another country originate from afar
cross national boundary cross biological boundary
expand within new country expand within new biological range
overcome citizens overcome native species
threaten native culture threaten native ecosystems
The term “invader” is culturally resonant because of fears that nations will
be literally invaded. Davis, Thompson and Grime (2001: 3) posit that the
founder of invasion biology, Charles Elton, was influenced by Britain’s
vulnerability to invasion:
There is another reason why the war may have transformed Elton’s perspec-
tive on invasions. Throughout the war years, British people were much more
concerned about a very different kind of invasion, one far worse than a ro-
dent infestation. They feared invasion by Germany. For Elton, invasion was
at the center not only of his work but also of his country’s psyche.
These authors demonstrate that Elton increasingly distinguished invading
species from normal ecological processes over the middle decades of the
20th century, which reflected his nationalistic concerns. Given concerns
about a “world without borders”, Mack et al. (2000: 689) raise this fear in
the present day when they claim that the spread of IS could create “homo-
geneous, impoverished ecosystems composed of cosmopolitan species”.
Fears of invaders have only intensified since September 11, 2001, which
may increase the appeal of the anti-IS campaign for many people.
11. Note that the cultural model of human invasion adopted here is that of Asiatic
hordes overflowing Europe in the sixth century or Spanish or Anglo immigrants
massively settling in the Americas and taking Indians’ territories, but it is not
compatible with the model underlying World War II type of invasions (R. Dir-
ven, personal communication).
Fears of invasion are reinforced in the journal Biological Invasions by

the intentional choice of metaphors whose entailments are potentially
frightening to the reader. The second article in the journal, for example,
was entitled “Positive interactions of nonindigenous species: Invasional
meltdown?” (Simberloff and Von Holle 1999). It reviewed beneficial inter-
actions among invaders that could “well lead to accelerated impacts on
native ecosystems – an invasional ‘meltdown’ process”, which invokes
images of nuclear disaster. While this rhetorical ploy is clearly aimed at
instilling fear, it also functions as a “clever diversionary tactic […] for
social control” (Rediehs 2002: 76). Whether patriotism attaches itself to a
nation or to biogeographic regions, when people identify with these
bounded spaces their own human vulnerability is exaggerated. Militaristic
language may help draw attention to an issue that is initially invisible to
non-biologists. However, it may also exaggerate the emotional intensity of
the situation.
Finally, since “immigration” is often portrayed as another form of “in-
vasion”, numerous writers have critiqued IS policy as having xenophobic
tendencies (Pollan 1994; Sagoff 1999; Subramaniam 2001; but see Sim-
berloff 2003). Fine and Christoforides (1991: 388) demonstrate that “the
sparrow issue ‘piggy-backed’ on the larger issue of how to protect the
American community from the presence of outsiders”. Given that the U.S.
media commonly invokes the metaphor IMMIGRANTS ARE ANIMALS (Santa
Ana 1999), it should not be surprising that the reverse mapping (ANIMALS
ARE IMMIGRANTS) can be interpreted as xenophobic. Xenophobic people
have a dislike for “other” people that is somehow rationalized. Similarly,
the ultimate cause of concern about IS is a dislike of what they do to native
species (including humans). Although the strength of the charge that biolo-
gists are xenophobic is limited by the analogy between IS and people, it is
supported by the ease with which this association can be made.
3. The conceptualization of a competitive bias within biology
Competition is a prevalent organizing metaphor within both contemporary

culture and invasion biology. Keddy (1989: 161–165) hypothesized that its
frequency in biology may result from cultural factors, i.e. its ability to pro-
voke drama, conflict and excitement, the dominance of male researchers, a
taxonomic bias, and ultimately, the level of competitiveness found among
scientists themselves. While each of these is probably a contributing factor,
competitive views of life partly derive from evolutionary thought, which

naturalizes them. Numerous scholars have discussed how Darwinism –
particularly through the metaphor of a “struggle for survival” – became
associated with competitiveness and militarism in both popular culture and
science (e.g., McIntosh 1992; Maasen, Mendelsohn and Weingart 1995;
Lakoff and Johnson 1999: 557–561). According to Keller (1991: 87), these
implications derive from how “much of contemporary evolutionary theory
relies on a representation of the ‘individual’ […] [whose] first and fore-
most need [is] the defense of its boundaries”. Because of the affinity be-
tween invasion biology and evolutionary theory (see Ludsin and Wolfe
2001 for review), invasion biologists are prone to emphasize competitive
interactions resulting from the occurrence of IS in a given region.
To demonstrate the embedding of competitiveness within the culture of
modern biology, I surveyed12 three contrasting groups of biologists: the
Society for the Study of Evolution (SSE, evolutionary biologists), the Hu-
man Behavior and Evolution Society (HBES, evolutionary psychologists)13
and the National Association of Biology Teachers (NABT, American biol-
ogy teachers). I asked members to respond to two questions about an array
of metaphorical statements:14
12. I administered a web survey in November-December 2003 using the email dis-
tribution lists of four organizations (one of which is excluded here; additional
details about my protocol are available upon request). I was unable to survey
ecologists or invasion biologists directly. There were 1892 respondents in the
final data set, with minimum response rates of 16% (NABT), 33% (SSE) and
44% (HBES). For further details on methodology and results, see Larson (2004,
2006).
13. In contrast to evolutionary biologists, who predominantly restrict their studies to
evolution among non-human species, evolutionary psychologists search for evi-
dence of why humans are the way they are now because of their evolutionary
history.
14. In this chapter I present their response to two statements – concerning struggle
for survival and cooperation – that I claim are metaphorical based on extensive
historical evidence (e.g., Maasen, Mendelsohn and Weingart 1995; Ruse 1996).
The actual survey contained numerous metaphorical statements about competi-
tion (and progress), and the results of a preliminary factor analysis suggests that
these statements reflect conceptual metaphors EVOLUTION IS A COMPETITIVE PROCESS
and EVOLUTION IS A PROGRESSIVE PROCESS. My brief discussion here is consistent with
overall results presented elsewhere (Larson 2004).
1. Do you believe this statement to be factually true? In your opinion, has

biological research provided sufficient evidence to support it?
2. Do you believe it would be beneficial if applied within society? Would
it be a good thing if people were to use this statement as a guide for so-
cial practices?
Figure 2. This figure shows responses to the statement “A struggle for survival
characterizes evolution”. The mean values along the left axis correspond
to response options in the survey: 1 = strongly disagree, 2= disagree, 3=
neutral, 4 = agree, 5 = strongly agree (that is, higher means greater
agreement). The mean response (with standard error) is given for both
question 1 (pale bars) and question 2 (darker bars). The organizations
are all statistically different from one another (p<0.001, Kruskal–Wallis
test), as are the responses to questions 1 and 2 for each organization
(p<0.001, Wilcoxon signed-ranks test).
The results show that these groups disproportionately project competitive

metaphors onto the natural world. On average, all three groups agreed that
biological research supports the statement “A struggle for survival charac-
terizes evolution”15 (Figure 2, pale bars), whereas they disagreed that this
15. I used this statement about “a struggle for survival” as a metric of a competitive
view of life because of the long association between these terms (McIntosh
is the case for the statement “Cooperation typifies the interaction between
animals” (Figure 3, pale bars). Despite the historical linkage between
“struggle for survival” and social Darwinism, the former is still accepted as
a relatively accurate reflection of reality even by scientists.
Figure 3. This figure shows responses to the statement “Cooperation typifies

the interaction between animals.” Other details as in legend for
Figure 2.
The results of question 2 concerning application to society show a similar

pattern among groups (Figures 2 and 3, dark bars). However, the respon-
dents differentially evaluated the two statements. On average, all three
groups disagreed with the assertion that it would be appropriate to apply
the statement about a struggle for survival within society, whereas they
agreed that it was factually true. In contrast, they agreed that it would be
beneficial to apply the statement about cooperation even though they felt it
was factually incorrect. Taken together, these results demonstrate that bi-
ologists recognize the potentially negative implications of applying com-
1992) and also because the expression “struggle for survival” is such a popular
metaphor for evolution.
petitiveness within the social realm,16 and thereby reveal a major dilemma
of modern biology. Biologists often communicate scientific results meta-
phorically, and there is a bias towards presenting them in a competitive
light, even though it is recognized that this bias could have undesirable
implications. This conclusion, based on empirical data, is of fundamental
importance in understanding the way in which such “scientific” metaphors
actively recruit from and resonate within larger frames of reference (Bono
1990).
These results demonstrate that biologists generally personify the inter-
actions between organisms as competitive. It is important to note, however,
that each of the statements I asked centered around value-laden meta-
phors.17 For example, what does competition describe? Consider “scram-
ble” (or exploitation) competition,18 which results from the passive use by
more than one species of a common resource that is in short supply. The
classic experimental test for scramble competition is to exclude a “com-
petitor” and to observe whether the remaining species does better. How-
ever, no competition has actually been observed, and this is also the case
for the more general term “struggle for existence”. Hence, the imposition
of these terms on the experimental or observational setting reflects attune-
ment of an observer to competition as a prevalent cultural metaphor that
can be applied to the biological world (Keller 1991). If biologists and oth-
ers uncritically adopt the idea that nature is competitive, competition be-
comes “naturalized”. It is just this type of bias towards competition that
partially creates the problem of IS.
The bias towards competition in invasion biology is revealed in two
main ways. First, there are many more studies of competition than of mutu-
alism, indicating that invasion biologists preferentially project competi-
16. Note that the question of how it might be applied was left undefined to allow the
respondents to provide their gestalt impression. It also forced them to recognize
that biological statements can be and are applied in the social realm.
17. Respondents were given the option to choose a “not applicable” box rather than
to respond on the disagree-agree scale, but fewer than 5% of them chose this
option for the statements discussed here. Although many qualitative comments
about the survey complained that statements were difficult to evaluate as scien-
tific facts, only a handful of people stated that they were metaphorical, and in
any case the results indicate that most individuals were content to agree with a
competitive view of life.
18. A similar case could be made for contest (or interference) competition, which is
a more direct behavior that limits access to a resource.
tiveness onto the interaction between native and non-native species. The
aforementioned review by Mack et al. (2000), for instance, neglected the
possibility of any population-level benefits arising from introduced species,
while it provided a long list of assumed competitive interactions. They
describe numerous cases of competition for resources, including introduced
ant species that “devastate large fractions of native ant communities by
aggression” (2000: 697). Plants are also portrayed as competitive: “inva-
sive plants have diverse means of competing with natives. Usurping light
and water are probably the most common tactics” (2000: 696). More gen-
erally, the first five volumes of the journal Biological Invasions mentioned
mutualism and cooperation only twice, whereas competitive interactions
were addressed in 25 papers. The emphasis on competitive interactions was
implicit in both papers on mutualism, which examined whether mutualism
between IS may intensify their effect on native species and communities
(Simberloff and Von Holle 1999; Morales and Aizen 2002).
Finally, invasion biology often assumes that invading species compete
with native ones, despite the limited evidence for this assertion. Hager and
McCoy (1998), for example, demonstrate that frequent assertions about the
competitiveness of the European species purple loosestrife in north Ameri-
can wetlands are over-stated. Similarly, a recent analysis of the effects of
IS concludes: “Taken together, theory and data suggest that, compared to
the effects of intertrophic interactions [predation] and habitat loss, compe-
tition from introduced species is not likely to be a common cause of ex-
tinctions of long-term resident species at global, metacommunity, and even
most community levels” (Davis 2003: 488). In conclusion, the often un-
tested hypothesis that IS compete with native species is in part ideologi-
cally-driven by the dominant competitive outlook in biology.
4. The conceptualization of militarism against invasive species
Once invasion biologists conceptualize IS as invaders that compete with

native species, it becomes natural to think of them as having a negative
valence and to defend “our” landscapes against them; that is, lands con-
ceived as belonging either to us as individuals or as members of a nation.
Drawing on the CONTAINER image schema, inside and outside are reified:
what is inside is inherently good whereas what is outside is unknown,
scary, an enemy. In Biological Invasions, IS are characterized with an array
of negative descriptors such as exotic, alien, weed and pest,19 which con-
trasts with the purity of natural landscapes (Milton 2000; Lien 2005). In
the words of Douglas (1966: 4): “Ideas about separating, purifying, demar-
cating and punishing transgressions have as their main function to impose
system on an inherently untidy experience. It is only by exaggerating the
difference between within and without, above and below, male and female,
with and against, that a semblance of order is created.” Invasion biologists
engender order on the biological world with a good-bad opposition that is
revealed by the prevalent frame of comparison between native species and
IS,20 one which relies on the establishment of a problematic distinction
between them (see Woods and Moriarty 2001).
Once a duality is created between personified friends (natural commu-
nities and their species) and foes (IS), their imputed competitiveness can
quickly escalate into militarism. This is demonstrated by ten papers in
Biological Invasions that refer to “aggressive” interactions between native
species and IS (e.g., Usio, Konishi and Nakano 2001). Another 22 papers
refer to the “threat” that they pose. IS also govern a “sea under siege”
(Galil 2000) and adopt a “‘sit and wait’ strategy” (Greenberg, Smith and
Levey 2001). Finally, in their article entitled “Biotic resistance experienced
by an invasive crustacean in a temperate estuary”, Hunt and Yamada
(2003) extend the war metaphor by attributing acts of resistance to the
native species themselves. In summary, these language choices attribute
agency to IS, which personifies them as competitive and thereby intensifies
our perception of their effect and our antagonism toward them.
Consequently, biologists feel justified in waging a war against IS (see
Larson 2005). For example, Webb et al. (2000: 350) stated that “[t]he third
front in the war on invasives is restoration”, and similar, yet more subtle
references typify well-cited review papers (such as Mack et al. 2000). Al-
though less common, militaristic metaphors were still detectable in Bio-
logical Invasions. Six papers referred to IS as “targets”, including Camp-
bell and Echternacht (2003), who envisioned introduced species as
“moving targets”. Ten papers invoked “strategies” for removing IS, and
eleven referred to their “eradication”. Two papers posited “non-target ef-
19. Each of these was used between 20–50 times in Biological Invasions. These
terms make it easy to confound native and invasive with notions of good and
evil. M. Chew (personal communication) has collected many examples of this
phenomenon, including an article in a children’s science magazine entitled:
“Those wicked weeds.”
20. 17 titles in Biological Invasions directly compared native and IS.
fects” in reference to the unintended effects of biological control agents21

on native species. In four papers, these agents allowed an indirect “attack”
on IS. It is apparent that while invasion biologists defend biodiversity and
non-human species as having an intrinsic right to exist, they proclaim, Ja-
nus-like, that IS don’t have these same rights. Indeed, they recommend that
a “‘guilty until proven innocent’ approach” be used against them (Mack et
al. 2000: 689).
5. Concluding thoughts
Bono (2003: 225) suggests that we “regard metaphor as a contingent, his-

torical ‘tool’ which we use (and which ‘uses’ us) to approach, ultimately to
inhabit, the unstable flux of things from which our world must emerge”.
This chapter, for example, shows that invasion biology is an expression of
three metaphors used to conceptualize and respond to novel species. Inva-
sion, competitiveness and militarism are interwoven into a narrative meta-
metaphor (or perhaps “root metaphor”) of contemporary American culture.
One consequence of their resonance with that larger narrative is that those
most committed to conservation may begin to have doubts about the inten-
tions of invasion biologists who use these metaphors. In the words of Un-
derhill (2003: 154): “A recurrent insistence on warfare metaphors does,
therefore, tend to imply a fundamental (though perhaps largely uncon-
scious) sympathy with, and desire for, the conflict and power struggle that
warfare allows.”
At a larger scale, militaristic metaphors may lessen the reality of war so
it can be further used as political trope. Underhill (2003) has demonstrated
a “switch” in the media whereby real war is presented with non-militaristic
metaphors (WAR IS X), while everyday occurrences are transformed into
wars via militaristic metaphors (X IS WAR). If these latter wars are por-
trayed uncritically, then they indirectly support the occurrence of actual
wars. By using militaristic metaphors, invasion biologists create an artifi-
cial similarity that contributes to a semantic field of war. For example,
President Bush recently merged part of the Animal and Plant Health In-
21. Biological control agents are species known to prey upon an IS in its native
range and which have been introduced purposefully to control it in its new
range. The term “control” is common in IS literature, occurring 38 times in
Biological Invasions.
spection Service – which is responsible for IS among other tasks – into his
new Department of Homeland Security. The Union of Concerned Scientists
criticized this move, observing that “It’s hard to imagine that a department
rightfully focused on preventing terrorist activity will pay much attention
to the movement of pests and weeds” (UCS 2002). Unfortunately, the way
invasion biologists present IS may have contributed to the ease with which
this link to international terrorism was made. If invasion biologists are
deeply committed to conservation they may need to oppose all wars, espe-
cially given their tremendous ecological costs (Austin and Bruch 2000).
Invasion biologists need to carefully reconsider their language if they
truly want people to connect with nature and to care for it. These objectives
may not be met by employing metaphors of invasion, competition and
militarism, which are founded on implicit dualities between self-other and
good-bad. In this respect, Waldron (2003: 166) has observed that
The nation has been sacralized by the same processes through which indi-
viduals, societies, and cultures are reified into selves or entities: by creating
boundaries dichotomizing the world into us and them, coercing homogeneity
within and excluding foreignness without, and imbuing all this with an emo-
tionally charged aura of eternal truth and goodness that simultaneously
sanctifies and obscures its contingent, constructed nature.
Although IS create problems for humans in certain circumstances, so do
some native species. Rather than siring a scapegoat, founded in prevailing
modes of relating, perhaps invasion biologists could better attend to just
how much IS are like us, as a means to break-down the distinction between
native-invasive, self-other. By any definition, humans are IS (Woods and
Moriarty 2001: 177–178). We are sometimes competitive, but we also
sometimes cooperate, even if the former is accentuated in our current cul-
tural context. In the words of Rodman (1993: 152), “[w]hen we look at […]
invasion, we look as if in a mirror and realize that restoring the balance
must, in large part, come from within”. Invasion biologists need to be
aware of the entailments of the unconscious metaphors that they adopt, live
by, and defend, and be open to alternative possibilities.
Acknowledgements
This chapter is an outgrowth of a talk given at ICLC 2003 in Logroño. I am

grateful to Matt Chew, Paul Chilton, René Dirven, Brigitte Nerlich and
especially Roz Frank for comments and encouragement that helped to sig-
nificantly improve it.
References
Alpert, Peter and John L. Maron

2000 Carbon addition as a countermeasure against biological invasion by
plants. Biological Invasions 2: 33–40.
Austin, Jay E. and Carl E. Bruch (eds.)
2000 The Environmental Consequences of War: Legal, Economic and
Scientific Perspectives. New York: Cambridge University Press.
Barbour, Michael G.
1995 Ecological fragmentation in the fifties. In: William Cronon (ed.),
Uncommon Ground: Rethinking the Human Place in Nature, 233–
255. New York: W. W. Norton and Company.
Baskin, Yvonne
2002 A Plague of Rats and Rubbervines: The Growing Threat of Species
Invasions. Washington: Island Press.
Bonneau, Laurent R., Kathleen S. Shields and Daniel L. Civco
1999 Using satellite images to classify and analyze the health of hemlock
forests infested by the hemlock woolly adelgid. Biological Invasions
1: 255–267.
Bono, James J.
1990 Science, discourse and literature: The role/rule of metaphor in sci-
ence. In: Stuart Peterfreund (ed.), Literature and Society: Theory and
Practice, 59–89. Boston: Northeastern University Press.
2003 Why metaphor? Toward a metaphorics of scientific practice. In:
Sabine Maasen and Matthias Winterhager (eds.), Science Studies:
Probing the Dynamics of Scientific Knowledge, 215–233. Transcript
Verlag.
Boucher, Douglas H.
1986 The idea of mutualism, past and future. In: Douglas H. Boucher (ed.),
The Biology of Mutualism: Ecology and Evolution, 1–28. London:
Croom Helm.
Bright, Christopher
1998 Life out of Bounds: Bioinvasion in a Borderless World. New York:
W. W. Norton.
Campbell, Todd S. and Arthur C. Echternacht
2003 Introduced species as moving targets: Changes in body sizes of in-
troduced lizards following experimental introductions and historical
invasions. Biological Invasions 5: 193–212.
Carlton, James T.
1999 A journal of biological invasions. Biological Invasions 1: 1.
Chew, Matthew K. and Manfred D. Laubichler
2003 Natural enemies – Metaphor or misconception? Science 301: 52–53.
Chilton, Paul A.
1996 The meaning of security. In: Francis A. Beer and Robert Hariman
(eds.), Post-Realism: The Rhetorical Turn in International Relations,
193–216. East Lansing: Michigan State University Press.
Davis, Mark A.
2003 Biotic globalization: Does competition from introduced species
threaten biodiversity? Bioscience 53: 481–489.
Davis, Mark A., Ken Thompson and J. Philip Grime
2001 Charles S. Elton and the dissociation of invasion ecology from the
rest of ecology. Diversity and Distributions 7: 97–102.
Douglas, Mary
1966 Purity and Danger: An Analysis of Concepts of Pollution and Taboo.
New York: Frederick A. Praeger.
Eser, Uta
1998 Assessment of plant invasions: Theoretical and philosophical funda-
mentals. In: Uwe Starfinger, K. Edwards, Ingo Kowarik and Mark
Williamson (eds.), Plant Invasions: Ecological Mechanisms and
Human Responses, 95–107. Leiden: Backhuys Publishers.
Eubanks, Philip
2000 A War of Words in the Discourse of Trade: The Rhetorical Constitu-
tion of Metaphor. Carbondale: Southern Illinois University Press.
Faulkner, Kate, Gregg Howald and Steve Ortega
2001 Eradicating rats from Anacapa Island. Natural Resource Year in
Review – 2001. http://www.aqd.nps.gov/pubs/yir/yir2001/05_risks/
05_5_faulkner.html
Fine, Gary A. and Lazaros Christoforides
1991 Dirty birds, filthy immigrants and the English sparrow war: Meta-
phorical linkage in constructing social problems. Symbolic Interac-
tion 14: 375–393.
Forrest, Barrie M. and Michael D. Taylor
2002 Assessing invasion impact: Survey design considerations and impli-
cations for management of an invasive marine plant. Biological Inva-
sions 4: 375–386.
Galil, Bella S.
2000 A sea under siege – Alien species in the Mediterranean. Biological
Invasions 2: 177–186.
Gibbs, Raymond
world. In: Raymond W. Gibbs, Jr. and Gerard J. Steen (eds.), Meta-
phor in Cognitive Linguistics: Selected Papers from the Fifth Inter-
national Cognitive Linguistics Conference, 145–166. Philadelphia:
John Benjamins.
Glotfelty, Cheryll
2000 Cold war, silent spring: The trope of war in modern environmental-
ism. In: Craig Waddell (ed.), And No Birds Sing: Rhetorical Analy-
ses of Rachel Carson’s Silent Spring, 157–173. Carbondale: South-
ern Illinois University Press.
Greenberg, Cathryn H., Lindsay M. Smith and Douglas J. Levey
2001 Fruit fate, seed germination and growth of an invasive vine – an
experimental test of ‘sit and wait’ strategy. Biological Invasions 3:
363–372.
Hager, Heather A. and Karen D. McCoy
1998 The implications of accepting untested hypotheses: A review of the
effects of purple loosestrife (Lythrum salicaria) in North America.
Biodiversity and Conservation 7: 1069–1079.
Holmgren, Milena
2002 Exotic herbivores as drivers of plant invasion and switch to ecosys-
tem alternative states. Biological Invasions 4: 25–33.
Hunt, Christopher E. and Sylvia Behrens Yamada
2003 Biotic resistance experienced by an invasive crustacean in a temper-
ate estuary. Biological Invasions 5: 33–43.
Johnson, Mark
and Reason. Chicago: Chicago University Press.
Journet, Debra
1991 Ecological theories as cultural narratives: F. E. Clement’s and H. A.
Gleason’s “stories” of community succession. Written Communica-
tion 8: 446–472.
Keddy, Paul A.
1989 Competition. New York: Chapman and Hall.
Keller, Evelyn Fox
1991 Language and ideology in evolutionary theory: Reading cultural
norms into natural law. In: James J. Sheehan and Morton Sosna
(eds.), The Boundaries of Humanity: Humans, Animals, Machines,
85–102. Berkeley: University of California Press.
Larson, Brendon M. H.
2004 The Metaphoric Web of Science and Society: Case Studies from
Evolutionary Biology and Invasion Biology. Ph.D. Dissertation, Uni-
versity of California, Santa Barbara.
2005 The war of the roses: Demilitarizing invasion biology. Frontiers in
Ecology and the Environment 3: 495–500.
2006 The social resonance of competitive and progressive evolutionary
metaphors. BioScience 56: 997–1004.
Lien, Marianne E.
2005 “King of fish” or “feral peril”: Tasmanian Atlantic salmon and the
politics of belonging. Environment and Planning D: Society and
Space 23: 659–671.
Ludsin, Stuart A. and Andrea D. Wolfe
2001 Biological invasion theory: Darwin’s contributions from The Origin
of Species. Bioscience 51: 780–789.
Maasen, Sabine, Everett Mendelsohn and Peter Weingart (eds.)
1995 Biology as Society, Society as Biology: Metaphors. Dordrecht: Klu-
wer Academic Publishers.
Mack, Richard N., Daniel Simberloff, W. Mark Lonsdale, Harry Evans, Michael
Clout and Fakhri A. Bazzaz
2000 Biotic invasions: Causes, epidemiology, global consequences and
control. Ecological Applications 10: 689–710.
McIntosh, Robert
1992 Competition: Historical perspectives. In: Evelyn F. Keller and Elisa-
beth A. Lloyd (eds.), Keywords in Evolutionary Biology, 61–67.
Cambridge: Harvard University Press.
Milton, Kay
2000 Ducks out of water: Nature conservation as boundary maintenance.
In: John Knight (ed.), Natural Enemies: People-Wildlife Conflicts in
Anthropological Perspective, 229–246. New York: Routledge.
Morales, Carolina L. and Marcelo A. Aizen
2002 Does invasion of exotic plants promote invasion of exotic flower
visitors? A case study from the temperate forests of the southern An-
des. Biological Invasions 4: 87–100.
Otis, Laura
1999 Membranes: Metaphors of Invasion in Nineteenth-Century Litera-
ture, Science and Politics. Baltimore: Johns Hopkins University
Press.
Parker, Ingrid M.
2001 Safe site and seed limitation in Cytisus scoparius (Scotch broom):
Invasibility, disturbance and the role of cryptogams in a glacial out-
wash prairie. Biological Invasions 3: 323–332.
Parker, Ingrid M., Daniel Simberloff, W. Mark Lonsdale, Karen Goodell, Marjorie
Wonham, Peter M. Kareiva, Mark H. Williamson, Betsy Von Holle,
Peter B. Moyle, James E. Byers and Lindley Goldwasser
1999 Impact: Toward a framework for understanding the ecological effects
of invaders. Biological Invasions 1: 3–19.
Pimentel, David (ed.)
2002 Biological Invasions: Economic and Environmental Costs of Alien
Plant, Animal and Microbe Species. New York: CRC Press.
Pollan, Michael
1994 Against nativism. New York Times Magazine (May 15): 52–55.
Porteous, J. Douglas
1986 Bodyscape: The body-landscape metaphor. Canadian Geographer
30: 2–12.
Rediehs, Laura J.
2002 Evil. In: John Collins and Ross Glover (eds.), Collateral Language:
A User’s Guide to America’s New War, 65–78. New York: New
York University Press.
Rodman, John
1993 Restoring nature: Natives and exotics. In: Jane Bennett and William
Chaloupka (eds.), In the Nature of Things: Language, Politics and
the Environment, 139–153. Minneapolis: University of Minnesota
Press.
Rohrer, Tim
1995 The metaphorical logic of (political) rape: The New Wor(l)d Order.
Metaphor and Symbolic Activity 10: 115–137.
Ross, Nancy, John Eyles, Donald Cole and Adele Iannantuono
1997 The ecosystem health metaphor in science and policy. Canadian
Geographer 41: 114–127.
Ruse, Michael
1996 Monad to Man: The Concept of Progress in Evolutionary Biology.
Cambridge: Harvard University Press.
Sagoff, Mark
1999 What’s wrong with invasive species? Report from the Institute for
Philosophy and Public Policy 19: 16–23.
Santa Ana, Otto
1999 ‘Like an animal I was treated’: Anti-immigrant metaphor in US pub-
lic discourse. Discourse and Society 10: 191–224.
Santibáñez, Francisco
2002 The object image-schema and other dependent schemas. http://www.
atlantisjournal.org/24_2/santiba.pdf
Schön, Donald A.
1993 Generative metaphor: A perspective on problem-setting in social
policy. In: Andrew Ortony (ed.), Metaphor and Thought, 137–163.
Cambridge: Cambridge University Press.
Shurin, Jonathan B. and John E. Havel
2002 Hydrologic connections and overland dispersal in an exotic freshwa-
ter crustacean. Biological Invasions 4: 431–439.
Simberloff, Daniel
2003 Confronting invasive species: A form of xenophobia? Biological
Invasions 5: 179–192.
Simberloff, Daniel and Betsy Von Holle
1999 Positive interactions of nonindigenous species: Invasional meltdown?
Biological Invasions 1: 21–32.
Smart, Ninian
1996 Dimensions of the Sacred: An Anatomy of the World’s Beliefs. Ber-
keley: University of California Press.
Soulé, Michael E.
1990 The onslaught of alien species, and other challenges in the coming
decades. Conservation Biology 4: 233–239.
Subramaniam, Banu
2001 The Aliens have landed! Reflections on the rhetoric of biological
invasions. Meridians 2: 26–40.
Takacs, David
1997 The Idea of Biodiversity: Philosophies of Paradise. Baltimore: Johns
Hopkins University Press.
Union of Concerned Scientists (UCS)
2002 Homeland Security Department could open floodgates to biological
invaders. http://www.ucsusa.org/news/press_release.cfm?newsID=
269
Underhill, James William
2003 The switch: Metaphorical representation of the war in Iraq from
September 2002 – May 2003. metaphorik.de 05/2003: 135–165.
http://www.metaphorik.de/05/underhill.htm
Usio, Nisikawa, Motoharu Konishi and Shigeru Nakano
2001 Species displacement between an introduced and a ‘vulnerable’ cray-
fish: The role of aggressive interactions and shelter competition.
Biological Invasions 3: 179–185.
Waldron, William S.
2003 Common ground, common cause: Buddhism and science on the af-
flictions of identity. In: B. Alan Wallace (ed.), Buddhism and Sci-
ence: Breaking New Ground, 145–191. New York: Columbia Uni-
versity Press.
Webb, Sara L., Marc Dwyer, Christina K. Kaunzinger and Peter H. Wyckoff
2000 The myth of the resilient forest: Case study of the invasive Norway
Maple (Acer platanoides). Rhodora 102: 332–354.
Wilcove, David S., David Rothstein, Jason Dubow, Ali Phillips and Elizabeth
Losos
1998 Quantifying threats to imperiled species in the United States. Bio-
science 48: 607–615.
Woods, Mark and Paul V. Moriarty
2001 Strangers in a strange land: The problem of exotic species. Environ-
mental Values 10: 163–191.
Zinken, Jörg, Iina Hellsten and Brigitte Nerlich
this vol. Discourse metaphors.
In search of development
Joseph Hilferty and Óscar Vilarroya
Abstract
Some authors have asserted that the high incidence of familial aggregation in a
certain case of specific language impairment (SLI) provides “strong evidence for
the genetic transmission of specific, strictly grammatical traits” (Newmeyer 1997:
59). In this paper we show that such radical nativist claims are either extremely
misleading or reveal a basic conceptual confusion stemming from often-used de-
velopmental metaphors. Despite their currency, the inferences invited by such por-
trayals of development are not supported by current molecular biology and genet-
ics: genes are not codes or programs for phenotypes (Oyama 1985; Nijhout 1990;
inter alia). In the case of language, we argue that, just as genes do not code for
behaviors such as reading or walking, genes do not represent the details of grammar
or fragments thereof.
Keywords: FOXP2, genetic dysphasia, genetic misattributions, innateness (nativ-

ism).
1. Introduction
On the face of things, behavioral-genetic studies represent a very attractive

research strategy, in that they promise to chart the “developmental pathway
that runs from gene to brain to behavior” (Pennington 1995: S69). This is
an alluring promise, to be sure, and one to which many researchers in the
cognitive sciences cling. Yet, much of the work relating behavioral pa-
thologies and genes is very controversial (Flint 1999), and nowhere is this
more clear than in so-called “experiments of nature”.
A case in point is a form of specific language impairment (SLI) known
as genetic dysphasia (GD). GD is seen by defenders of grammatical nativ-
ism (i.e., Chomskyan universal grammar) as a straightforward corrobora-
tion of their position. As one commentator has put it, “GD does provide
evidence that seems to point incontrovertibly to a genetic basis for
198 Joseph Hilferty and Óscar Vilarroya
autonomous grammar” (Newmeyer 1997: 47). In this paper, we conduct a

critical examination of such assertions. In particular, we show that nativist
metaphors focus on gene-trait correlations at the expense of development.
This, we argue, is especially true in the case of heritable pathologies such
as GD, which nativists routinely interpret as clear evidence in favor of their
position. We adduce a number of examples that show that such arguments
are either misleading or are based on a misunderstanding about what genes
do.
2. The case of the KE family
Arguments for and against the innateness of behavioral competences have

always been the source of acrimonious polemics. Certainly, within the field
of linguistics, where nativism is a key part of the dominant paradigm, em-
pirical evidence supporting nativist claims of a genetically endowed uni-
versal grammar has been rather scant. Thus, the debate had largely cen-
tered on rationalist arguments to the best explanation (e.g., the argument
from the poverty of the stimulus and the like). This essentially led to a sort
of “theoretical deadlock” since what constitutes a “best argument” is
largely ideological.
Some fifteen years ago, however, a surprising development occurred,
one that many felt had tipped the scales in favor of nativist thought. In
1990, the prestigious journal Nature published a piece of science corre-
spondence describing a large pedigree (known as the KE family) with a
heritable grammatical deficit (Gopnik 1990). The kindred, according to
nativists, suffered from “feature blindness”, i.e., an impairment in regular
inflectional morphology. Theoretically, this is a very striking claim: regular
morphology in English is rule based, whereas irregulars are learned by
rote. So, it was reasoned that the affected KE family members were im-
paired in morphosyntactic rule formation (Gopnik and Crago 1991). While
the appropriateness of this characterization was promptly contested
(Fletcher 1990; Vargha-Khadem and Passingham 1990), nativists wasted
little time in declaring that this was key empirical evidence in their favor
(see, e.g., Jackendoff 1993; Pinker 1994).
The impairment is genetic, and it specifically affects the ability to construct
a mental grammar, leaving other cognitive abilities intact. In order for this to
be possible, there must be at least one gene that is responsible for a special-
purpose endowment for language acquisition. The part of Universal Gram-
In search of development 199
mar having to do with acquiring inflectional endings must not be a general-

purpose learning strategy. (Jackendoff 1993: 116)
If these remarks proved to be true, then GD might support the notion of an
innate universal grammar. This did not pan out, however.
In spite of the emphasis placed on the grammatical impairment of the
KE family’s phenotype, it seems quite certain that the deficit is not solely
an impairment of morphosyntactic inflection (Vargha-Khadem et al. 1995,
1998; Watkins, Gadian and Vargha-Khadem 1999). First, the affected
family members suffer from severe orofacial dyspraxia, which is accompa-
nied by an articulatory disorder. Because of this, the affected members’
speech is largely unintelligible to naive listeners. Second, both verbal and
nonverbal IQs are, in general, considerably lower in affected family mem-
bers than in those that do not carry the mutation (Watkins, Dronkers and
Vargha-Khadem 2002). While the significance of this finding is not en-
tirely clear (Gopnik and Goad 1997; Ullman and Gopnik 1999), it would
seem to suggest that the impairment is not specific to grammar or language.
Third, although the most recent results (Watkins, Dronkers and Vargha-
Khadem 2002) confirm that the family members have a clear morphosyn-
tactic deficit, the dissociation between regular and irregular forms is less
than clear. Contrary to the previous reports, the affected family members
are also impaired in their production of irregular past-tense forms. Cer-
tainly, given such findings, it is difficult to see how GD supports nativist
contentions.
This said, it is true that the aggregation pattern in the KE family is un-
ambiguous. The syndrome has occurred in at least three different genera-
tions of the family and affects both sexes equally (Hurst et al. 1990). Such
an inheritance pattern suggests that the syndrome is autosomal dominant.
After a decade of intense study, it was finally discovered that a point mu-
tation in the FOXP2 gene was implicated in the KE-family phenotype (Lai
et al. 2000). Compared to unaffected members, as well as a control group,
it appears that several regions of gray matter are abnormal bilaterally
(Watkins, Dronkers and Vargha-Khadem 2002). In particular, the left cau-
date nucleus is smaller in the affected members. Intriguingly, the decreased
volume of the left caudate nucleus correlates to a lower score in oral-praxis
tests. This is the first piece of direct evidence indicating a relationship
between a brain abnormality and the behavioral phenotype in the KE fam-
ily, though the causal nature of this correlation is still to be determined. Be
this as it may, nobody disputes the genetic basis for the KE family’s disor-
der (Fisher 2005; Fisher et al. 1998; Lai et al. 2000; Vargha-Khadem et al.
1998; Vargha-Khadem et al. 2005; Watkins, Gadian and Vargha-Khadem

1999).
Evidence indicating that GD is an inherited syndrome does not grant, in
our view, the more radical nativist claims that some authors freely make in
their of discussions of genetic-behavioral associations.
[W]e have strong evidence for the genetic transmission of specific, strictly
grammatical traits [...]. (Newmeyer 1997: 59)
In our opinion, such interpretations of GD do not cohere with the findings
of molecular biology.
3. Genetic impostures and neo-preformationalist metaphors
Assertions such as those above stem directly from what is often referred to
as the innateness hypothesis. Frederick Newmeyer states the hypothesis in
the following terms:
Central aspects of this autonomous system [i.e., grammatical competence]
are provided by the human genome. (Newmeyer 1997: 49)
Other formulations of the thesis can be found scattered throughout the
literature. For example, Smith and Tsimpli (1995) make it very clear that,
when some aspects of language (i.e., universal grammar) are described as
innate, this plainly means “genetically determined” (1995: 22; see also
Smith 2003). Noam Chomsky has been even more explicit about the point:
This seems to me what we should hope to discover: that there is in the gen-
eral initial cognitive state a subsystem (that we are calling S0 for language)
which has a specific integrated character and which in effect is the genetic
program for a specific organ (here it is the program for the specific organ
which is human language). It is evidently not possible now to spell it out in
terms of nucleotides, although I don’t see why someone couldn’t do it, in
principle. (cited in Piattelli-Palmarini 1980: 124)
Such views entail the possibility that a gene (or a set of genes) could some-
how encode a behavioral faculty (or some fragment thereof). In the case of
language, a gene such as this would, without a doubt, be a sort of program
for the cortical representations that implement our grammars. Just as we
might say that an architect’s blueprint codes for a building, genes could be
thought of as a program for constructing a phenotype. Given this assump-
tion, we could, in a very real sense, attribute the faculty in question to the
gene. Unfortunately, genes do not work this way.
Conceptualizing genes as codes for phenotypes is quite commonplace,
to be sure. Nonetheless, its pervasiveness does not grant its truthfulness. In
fact, we believe that the code metaphor constitutes a mistaken outlook on
development, as it invites inferences that are entirely preformationist in
nature:
Today we think of preformationism as an archaic relic of outmoded thought,
and we snicker at the absurd idea that there are little people curled up in
sperm or egg cells. But replacing curled-up people with curled-up blueprints
or programs for people is not so different. [...] What is central to preforma-
tionist thought is not the literal presence of fully formed creatures in germ
cells, but rather a way of thinking about development – development as
revelation of preformed essence rather than as contingent series of construc-
tive interactions, transformations, and emergences. It is a way of thinking
that makes real development irrelevant because the basic “information” or
form, is there from the beginning, a legacy from our ancestors. (Oyama
2000: 136, italics in original)
The assumption that phenotypical traits are represented in the genome
turns genes into what Schaffner (1998) labels traitunculi, i.e., copies of a
trait codified in certain stretches of DNA. We believe that “provided by the
genome” hypotheses, such as the innateness hypothesis (see above), are
susceptible to such an interpretation, viz., as entailing that behaviors or
competences are somehow “encoded” in genes. This, however, does not fit
the facts. Genes have no representational resources to specify phenotypical
traits, since the only thing that genes code for is the primary structure of
proteins (Nijhout 1990; Oyama 1985; inter alia).
Providing proteins is only a small portion of development. Nonetheless,
the local effects of genes can set off a cascade of further gene actions,
which in turn may trigger a multi-level interaction of biochemical, cellular,
physiological and behavioral events (see Figure 1). Therefore, the way a
gene is expressed as, say, a behavioral output is the result of a complex
intermeshing of processes requiring many levels and components. In short,
the developmental processes leading to a given trait constitute a highly
dynamic configuration, in which feedback mechanisms (both positive and
negative) interact. The emergence of traits – even so-called monogenic
traits – is no simple matter (Scriver and Waters 1999): the collaboration of
many genes is necessary for a given phenotype to appear. Traits are thus
rooted in combinations of genes (e.g., polygeny and epistasis) or in single
genes that may also affect other traits (i.e., pleiotropy). Clearly, the inter-
actions involved in such a system are a many-to-many process, not a simple
hierarchical control structure (Schlichting and Pigliucci 1998).
Figure 1. Depicted here is a very simplistic and idealized characterization of

the complex regulatory processes involved in gene expression. For
clarity, cross-influences have been restricted to two or three possible
actions; likewise, the number of levels depicted might actually em-
bed an unknown number of intermediate levels. Environmental influ-
ences, including learning stimuli, have also been simplified. None-
theless, even in this idealized form, it can be seen that a mutation of a
single gene can be sufficient to change or even eliminate one of the
behavioral outputs.
Construing genes (or genomes) as preencoded instructions for traits does

not provide much insight for understanding development (Nijhout 1990).
In reality, such a view is not much more than a modern-day folk theory,
whose inaccuracy grants it distressingly little heuristic value. Take, for
instance, the following two examples:
1. Homologous genes. Homologous genes preserved on a phylogenetic

scale show that the morphology of organisms has changed at a much
faster rate than the evolution of DNA sequences (Schlichting and Pigli-
ucci 1998). This is clearly the case of the group of homologous Pax-6
genes. Pax-6 homologues are recognized as a family of regulatory genes
related to the development of eyes in fruit flies, frogs and many mam-
mals. In mice, however, Pax-6 also plays a role in the formation of the
nose (Wakayama et al. 1998) and, in squids, Pax-6 is active in the ani-
mal’s tentacle development (Tomarev et al. 1997). The phenotypical re-
sults deriving from Pax-6 homologues are thus highly context-
dependent and inextricably tied to the genetic background.
2. Environmental effects. The homeotic gene Antennapedia is known to

induce legs in place of antennas in the fruit fly Drosophila (Schneuwly,
Klemenz and Gehring 1987; Halder, Callerts and Gehring 1995). In or-
der for this to happen, Antennapedia cDNA must be heat pulsed during
the embryonic stage or during the late third larval instar; no ectopic leg
structures are obtained if the heat shock is given during the first or sec-
ond larval instars (see Nijhout 1990: 442). Because of such contingen-
cies, it is impossible to attribute developmental primacy to either genes
or environment (Gottlieb 1995; Griffiths and Knight 1998; Oyama
1985; inter alia). Instead, these two facets of development are better
seen as coacting to produce phenotypical expressions (Gottlieb 1995).
Of course none of this implies that development is an inherently unstruc-

tured process; rather, it merely points to the fundamental fact of develop-
ment that nothing occurs in a vacuum. The nativist metaphor of develop-
ment does not really capture the nature of such contextual effects.
Genes do not shape the course of developmental or behavioral events by
themselves. The reason for this is simple. While genes can be viewed as
instructions for the production of proteins, they do not encode how proteins
interact, much less the more distal effects such as how cells and tissues
communicate, how organs come into being, or how the central nervous
system forms. The following passage from Gilbert Gottlieb is instructive in

this regard:
In [the 1950s] we could think of genes making neural structures that under-
lie behavior. Today we realize that genes do not make finished neural
structures but rather produce protein (DNA Æ RNA Æ protein) which then
differentiates further to eventually become first a neuroblast, then a neuron,
and finally a specific kind of neuron located in a particular place in the
nervous system. The developmental “fate” of the protein is not determined
upon its production. Influences from higher supragenetic levels in the devel-
opmental system play a role in specifying the development of the protein
into a fully mature, functional unit, integrated with other parts of the system
[…]. (Gottlieb 1995: 132)
Genes are not a blueprint or program for the development of an organism.
A fortiori, genes definitely are not a blueprint or program for the specific
functions or behaviors that the complete developmental process satisfies.
4. Gene-pathology associations
Gene-trait correlations of various pathologies are routinely cited by nativ-

ists as incontrovertible evidence that genes “provide” developmental out-
comes. As we have seen in section 2, GD is a prime example of this. While
it is clear that certain genetic anomalies may change the path of develop-
ment in quite remarkable ways, it is also clear that this should not be inter-
preted as proof that genes actually control phenotypical development. To
paraphrase Nijhout (1990: 442), such an interpretation of gene actions
amounts to equating the steering wheel with the driver.
In order to avoid such faulty attributions, it is essential to understand
what specific gene-trait correlations can, and cannot, tell us: all a connec-
tion between a gene and a trait shows is that the gene produces a protein
that is necessary for the development of the phenotype. Finding the corre-
lation, however specific it may be, does not tell us where and when the
particular developmental process begins or ends. Rather, it is, at best, only
a first methodological step that can be used to manipulate and explore the
developmental process at hand. Consequently, gene-pathology associations
are correlations in search of a cause.
To put the matter another way: a specific correlation between a gene
and a given phenotypical effect can be used only as a measure of probabil-
istic differences that are related to the presence, absence, or mutation of the
gene. It is an analysis of variance, not an analysis of causes (Lewontin

1974). Nativist inferences seem to overlook this fact time and time again.
This is unfortunate, because such misattributions have significant causal
and methodological repercussions.
EXAMPLE 1: PHENYLKETONURIA
A specific correlation between a malfunctioning gene and the disruption of
a given behavior is anything but a confirmation of the genetic “encoding”
of the behavior. Rather, the relationship between a genetic mutation and a
behavioral deficit merely indicates where we might start to look for the
cause of the pathology. Nonetheless, the connection does not tell us where
the precise cause is, nor does it reveal the whole story about what causes
normal functioning. In this sense, a genetic anomaly that obstructs a certain
developmental process might be compared to a house of cards that col-
lapses because a single card has been taken out. All the cards are necessary
to build the structure, but perhaps only one is needed to make it fall. Simi-
larly, a genetic disturbance that is sufficient for a given malfunction to
develop does not imply that the corresponding normal gene determines the
normal phenotype.
Consider phenylketonuria (PKU). This syndrome is associated with a
serious genetic disorder that produces light-skinned infants with severe
mental retardation (e.g., Scriver et al. 1995). Over the years, researchers
have found a number of mutations in what is known as the phenylalanine
hydroxylase (PAH) gene, and currently there is little doubt that a distur-
bance of this gene can lead to the development of PKU (Eisensmith and
Woo 1992).
In the case of PKU, a false inference concerning the specificity of a be-
havioral-genetic connection could – potentially at least – be quite grave if
we were to assume that a gene might actually give rise to a faculty or abil-
ity. Were we to hold such a belief, then we might well contend that the
PAH gene is the “gene for intelligence”. It is highly implausible that a gene
could provide such a faculty. The reason for this is not that we have many
other forms of intelligence deficits (though this certainly is a legitimate
objection). Instead, the true problem lies in the fact that a deficit in a given
cognitive faculty or ability does not presuppose that the implicated gene
really accounts for the faculty itself.
EXAMPLE 2: DEVELOPMENTAL DYSLEXIA

Genetic correlations also present an important methodological concern. As
is the case with most descriptions, phenotypical profiles are the product of
a number of analytical decisions. These decisions come, however, with no

guarantees that the descriptive profile corresponds to the functional level at
which the gene actions should be explained. In other words, although there
may be a specific connection between a gene and a given trait, it may be
the description of the trait that makes the connection specific, rather than
the actual gene actions. In such cases, the issue of domain-specificity be-
comes rather clouded.
Take, for instance, developmental dyslexia (DD). Like specific lan-
guage impairment, DD seems to have a genetic basis, and indeed it has
been known for almost a century that the deficit is heritable (see Penning-
ton 1995). Some linkage studies have implicated certain small regions on
various chromosomes, especially chromosomes 6 and 15 (see Fisher and
DeFries 2002). The effort currently underway to trace DD to chromosome
6 is very interesting, because it suggests that the relationship between a
genetic anomaly and the phenotype is both specific and complex. For ex-
ample, Gayán et al. (1999) argue that the linkage to a small region on
chromosome 6 involves an impairment of several important component
skills of reading, namely, (a) word recognition, (b) orthographic coding, (c)
phonological decoding, and (d) phoneme awareness.
If the results of Gayán et al. are correct, it seems likely that there is a
specific relationship between a particular genetic disorder and a disorder in
reading acquisition. However, because the decoding of written language is
greatly dependent on abilities taken from the domain of oral language (e.g.,
phoneme awareness), the specificity of the connection is rather deceptive.
Thus, even if there is a specific connection between a given gene and dys-
lexia, it is really the decision to describe the functional anomaly at the level
of “reading” that makes the connection specific.
5. Development
The pathway of development leading from genes to phenotypes is ex-

tremely complex. It is important not to lose sight of this fact. The correct
development and functioning of any specific domain of cognition depend
on the appropriate interaction of all the relevant elements that comprise its
expression. As Bishop (1997) has remarked, this is true whether the faculty
in question is walking or talking:
Like language, walking is a “species universal” – that is, common to all
normal humans but not seen in other primates – which develops without
overt instruction in a wide range of environmental circumstances. [...] No

other primate shares this skill, yet all normal humans master it without spe-
cific instruction within the first few years of life [...]. [...] Just as with lan-
guage, though, some unfortunate individuals have a specific single-gene dis-
order – muscular dystrophy – which selectively interferes with this ability.
(Bishop 1997: 911)
The operations involved in moving one’s body bipedally over a wide vari-
ety of terrains as diverse as slippery ice and jagged mountainsides are quite
complex. Nonetheless, if a single crucial underlying link is disrupted in the
intricate web of motor and proprioceptive processes, it can be enough to
impair quite profoundly our faculty for bipedal locomotion:
It is apparent that walking depends on the integrity of a wide range of un-
derlying systems, involving muscles, nerves, and central control processes
that regulate balance, proprioception, and motor planning. Muscular dystro-
phy has a specific effect on just one of these systems – the muscles – but this
is sufficient to make walking difficult or impossible. (Bishop 1997: 911)
Surely nobody would want to suggest that the genetic mutation connected
with muscular dystrophy points incontrovertibly to the “genetic transmis-
sion of specific locomotive traits.” The very thought would be completely
reckless.
Behavioral faculties, whether they be reading, talking, or whatnot, are
the result of a complex sequencing of many developmental levels, ranging
from DNA, genes, and proteins to the formation and maturation of the nec-
essary body parts, biomechanics, and the appropriate interaction with the
environment. Consequently, behavioral phenotypes are not properly
thought of as the output of a genetic program that somehow finds its home
in the human brain and is blindly triggered by certain environmental stim-
uli. Such a view creates an explanatory void that can only be termed im-
mense (Tomasello 1999). Moreover, it is precisely this type of elliptical
reasoning that generates the illusion that there are actually genes for com-
plex behaviors (see Figure 2).
The import of such considerations for grammatical development should
be evident. Just as it is a fallacy to surmise that there is a gene for intelli-
gence or for walking on the basis of pathologies such as phenylketonuria
and muscular dystrophy, so too is it unsound to believe that the nucleotides
of the FOXP2 gene might spell out grammatical properties on the grounds
that GD is a heritable deficit. After all, this gene, it turns out, is found in
mice (Fisher 2005; Vargha–Khadem et al. 2005) and is expressed in organs
as diverse as the lungs, guts and heart (Marcus and Fisher 2003; Vargha–
Khadem et al. 2005). Of course, in saying this we are in no way denying
hadem et al. 2005). Of course, in saying this we are in no way denying the
involvement of genes in the acquisition of language or other behavioral
competences: all behaviors have a genetic component. In fact, the work on
GD shows this quite handily. However, it is no more accurate to think that
a gene (or a set of genes) might encode a behavior than it is to think that
the quartz battery of a watch tells time.
Figure 2. Depicted in this diagram is a caricature of gene-behavior associations as

a sort of black box. Only two levels of development are taken into account (the
genetic level and the behavioral level); all intervening levels of biological processes
are disregarded. The omission of these levels and their interactions with the envi-
ronment leads to the illusion that genes encode behavioral competences.
5. Conclusion
Does the genetic basis of GD constitute incontrovertible evidence that

genes actually transmit specific grammatical traits? Nothing could be fur-
ther from the truth. Despite the robust connection between a particular case
of GD and FOXP2, there is absolutely no reason to believe the gene actu-
ally encodes any sort of grammatical competence. Such information is

clearly far beyond the pale of a gene’s sphere of operation (which is to
produce protein), and to believe otherwise is to fall into the sheerest of folk
theories. In sum, statements to the effect that genes provide behavioral
competences are unfounded and grossly misleading.
Acknowledgment
Funding for the first author was provided by the BioLex Project (Spanish
Ministry of Science and Education grant: DGICYT PB 97-0887).
References
Bishop, Dorothy V.M.

1997 Cognitive neuropsychology and developmental disorders: uncom-
fortable bedfellows. The Quarterly Journal of Experimental Psy-
chology 50A(4): 899–923.
Eisensmith, Randy C. and Savio L.C. Woo
1992 Molecular basis of phenylketonuria and related hyperphenylalanine-
mias: Mutations and polymorphisms in the human phenylalanine hy-
droxylase gene. Human Mutation 1: 13–16.
Fisher, Simon E.
2005 Dissection of molecular mechanisms underlying speech and language
disorders. Applied Psycholinguistics 26: 111–128.
Fisher, Simon E. and John C. DeFries
2002 Developmental dyslexia: Genetic dissection of a complex cognitive
trait. Nature Reviews Neuroscience 3: 767–780
Fisher, Simon E., Faraneh Vargha–Khadem, Kate E. Watkins, Anthony P. Monaco
and Marcus E. Pembrey
1998 Localisation of a gene implicated in a severe speech and language
disorder. Nature Genetics 18: 168–170.
Fletcher, Paul
1990 Speech and language defects. Nature 346: 226.
Flint, Jonathan
1999 The genetic basis of cognition. Brain 122: 2015–1231.
Gayán, Javier, Shelly D. Smith, Stacey S. Cherny, Lon R. Cardon, David William
Fulker, Amy M. Brower, Richard K. Olson, Bruce F. Pennington and
John C. DeFries
1999 Quantitative-trait locus for specific language and reading deficits on
chromosome 6p. American Journal of Human Genetics 64: 157–164.
Gopnik, Myrna
1990 Feature-blind grammar and dysphasia. Nature 344: 715.
Gopnik, Myrna and Martha Crago
1991 Familial aggregation of a developmental language disorder. Cogni-
tion 39: 1–50.
Gopnik, Myrna and Heather Goad
1997 What underlies inflectional error patterns in genetic dysphasia. Jour-
nal of Neurolinguistic 10: 109–137.
Gottlieb, Gilbert
1995 Some conceptual deficiencies in “developmental” behavior genetics.
Human Development 38(3): 131–141.
Griffiths, Paul E. and Robin D. Knight
1998 What is the developmentalist challenge? Philosophy of Science
65(2): 253–258.
Halder, Georg P., Patrick Callerts and Walter J. Gehring
1995 Induction of ectopic eyes by targeted expression of the eyeless gene
in Drosophila. Science 267: 1788–1792.
Hurst, Jane A., Michael Baraitser, Elizabeth Auger, F. Graham and S. Norell
1990 An extended family with a dominantly inherited speech disorder.
Developmental Medicine and Child Neurology 32: 347–355.
Jackendoff, Ray
1993 Patterns in the Mind: Language and Human Nature. Harvester–
Wheatsheaf.
Johnston, Judith R.
1997 Specific language impairment, cognition, and the biological basis of
language. In: Myrna Gopnik (ed.), The Inheritance and Innateness of
Grammars, 161–180. Oxford: Oxford University Press.
Lai, Cecilia S.L., Simon E. Fisher, Jane A. Hurst, Elain R. Levy, Shirley Hodgson,
Margret Fox, Stephan Jeremiah, Susan Povey, D. Curtis Jamison,
Eric D. Green, Faraneh Vargha-Khadem and Anthony P. Monaco
2000 The SPCH1 region on human 7q31: Genomic characterization of the
critical interval and localization of translocation associated with
speech and language disorder. American Journal of Human Genetics
67: 357–368.
Lewontin, Richard C.
1974 The analysis of variance and the analysis of causes. American Jour-
nal of Human Genetics 26: 400–411.
Marcus, Gary and Simon E. Fisher

2003 FOXP2 in focus: What genes tell us about speech and language.
Trends in Cognitive Sciences 7(6): 257–262.
Newmeyer, Frederick J.
1997 Genetic dysphasia and linguistic theory. Journal of Neurolinguistics
10(2–3): 47–73.
Nijhout, H. Frederik
1990 Metaphors and the role of genes in development. BioEssays 12(9):
441–446.
Oyama, Susan
1985 The Ontogeny of Information: Developmental Systems and Evolu-
tion. Cambridge: Cambridge University Press. (2nd edition: Durham,
NC: Duke University Press, 2000).
2000 Evolution’s Eye: A Systems View of the Biology-Culture Divide.
Durham, NC: Duke University Press.
Pennington, Bruce F.
1995 Genetics of Learning Disabilities. Journal of Child Neurology 10,
supplement n. 1: S69–S77.
Piattelli-Palmarini, Massimo (ed.).
1980 Language and Learning: The Debate between Jean Piaget and
Noam Chomsky. Cambridge, MA: Harvard University Press.
Pinker, Steven.
1994 The Language Instinct: How the Mind Creates Language. New
York: William Morrow.
Schaffner, Keneth
1998 Genes, behavior and developmental emergentism: One process, indi-
visible? Philosophy of Science 65: 209–252.
Schlichting, Carl D. and Massimo Pigliucci
1998 Phenotypic Evolution. A Reaction Norm Perspective. Sunderland,
MA: Sinauer.
Schneuwly, Stephan, Roman Klemenz and Walter J. Gehring
1987 Redesigning the body plan of Drosophila by ectopic expression of
the homeotic gene Antennapedia. Nature 325: 816–818.
Scriver, Charles R., Stuart Kaufman, Randy C. Eisensmith and Savio L.C. Woo
1995 The phenylalanine hydroxylating system. In: Charles R. Scriver,,
Arthur L. Beaudet, William S. Sly and David Valle (eds.), The Meta-
bolic and Molecular Basis of Inherited Disease, 1015–1075. 7th edi-
tion. New York: MacGraw-Hill.
Scriver, Charles R. and Paula J. Waters
1999 Monogenic traits are not simple: Lessons from phenylketonuria.
Trends in Genetics 15: 267–272.
Smith, Neil
2003 Dissociation and modularity: Reflections on language and mind. In:
Marie T. Banich and Molly Mack (eds.), Mind, Brain, and Lan-
guage, 87–111. Mahwah, NJ: Lawrence Erlbaum.
Smith, Neil and Ianthi-Maria Tsimpli
1995 The Mind of a Savant: Language Learning and Modularity. Oxford:
Basil Blackwell.
Tomarev, Stanislav I., Patrick Callaerts, Lidia Kos, Rina Zinovieva, Georg Halder,
Walter Gehring and Joram Piatigorsky
1997 Squid Pax-6 and eye development. Proceedings of the National
Academy of Science USA 94: 2421–2426.
Tomasello, Michael
1999 The Cultural Origins of Human Cognition.Cambridge, MA: Harvard
University Press.
Ullman, Michael T. and Myrna Gopnik
1999 Inflectional morphology in a family with inherited specific language
impairment. Applied Psycholinguistics 20: 51–117.
Vargha-Khadem, Faraneh and Richard E. Passingham
1990 Speech and language defects. Nature 346: 226.
Vargha-Khadem, Faraneh, Kate Watkins, Katherine Alcock, Paul Fletcher and
Richard E. Passingham
1995 Praxic and nonverbal cognitive deficits in a large family with a ge-
netically transmitted speech and language disorder. Proceedings of
the National Academy of Science USA 92: 930–933.
Vargha-Khadem, Faraneh, Kate E. Watkins, C.J. Price, J. Ashburner, Katherine J.
Alcock, A. Connelly, Richard S.J. Frackowiak, Karl J. Friston, M.E.
Pembrey, Mortimer Mishkin, David G. Gadian and Richard E. Pass-
ingham
1998 Neural basis of an inherited speech and language disorder. Proceed-
ings of the National Academy of Science USA 95: 12695–12700.
Vargha-Khadem, Faraneh, David G. Gadian, Andrew Copp and Mortimer Mishkin
2005 FOXP2 and the neuroanatomy of speech and language. Nature Re-
views Neuroscience 6: 131–138.
Wakayama, Teruhiku, Anthony C.F. Perry, Maurizio Zuccotti, Kenneth R. Johnson
and Ryuzu Yanagimachi
1998 Full-term development of mice from enucleated oocytes injected with
cumulus cell nuclei. Nature 394: 369–373.
Watkins, Kate E., David G. Gadian and Faraneh Vargha-Khadem.
1999 Functional and structural brain abnormalities associated with a ge-
netic disorder of speech and language. American Journal of Human
Genetics 65: 1215–1221.
Watkins, Kate E., Nina F. Dronkers and Faraneh Vargha-Khadem.

2002 Behavioural analysis of an inherited speech and language disorder:
comparison with acquired aphasia. Brain 125 (3): 452–464.
The language-organism-species analogy: A complex
adaptive systems approach to shifting perspectives
on “language”
Roslyn M. Frank
[A]ll language is metaphoric [and] if we look at the
implications of recent discussions of the theory laden-
ness of observation, of realism and the use of scientific
models, we find that the use of language in scientific
theory conforms closely to the metaphoric model. Sci-
entific revolutions are, in fact, metaphoric revolutions,
and theoretical models should be seen as metaphoric
redescription of the domain of phenomena. (Arbib and
Hesse 1986: 153, 156)
Abstract
The human mind is not only embodied, that is, individually situated in its own
body, it is also situated socioculturally together with other embodied minds. This
chapter addresses the interactive and dynamic role of sociocultural situatedness by
examining the way that “language” itself has been “imagined” in its various meta-
phoric instantiations in discourse. The chapter brings forward a new conceptual
frame of analysis that concentrates on the way metaphors, especially in scientific
discourses, have come about, expanded, disappeared or been replaced by new ones.
Divided into four parts, the paper begins with an introductory section in which the
concept discourse metaphor formation is introduced and discussed. It then moves
on to a detailed examination of a discourse metaphor, namely, the “language-as
organism-species” metaphor, which has dominated the metaphoric repertoire of
linguistics for several centuries (cf. Zinken, Hellsten and Nerlich this volume). The
analysis is further informed by thinking of language and metaphor formations as
complex adaptive systems. The characteristics of the latter are taken up, explicitly,
in the third section of the paper. The final section looks at the way the metaphor of
“language-organism-species” is undergoing shifts in its meaning and application to
language and language change, shifts that coincide in certain ways with those taking
place in the discourse of the biological sciences in the post-genomic era.
216 Roslyn M. Frank
Keywords: complex adaptive systems, discourse metaphor formation, emergence,

evolutionary linguistics, linguistic organicism, race, species.
1. Introduction: The notion of “discourse metaphor formation”
The point of departure for this article is the notion of discourse metaphor,
which is defined by Zinken, Hellsten and Nerlich (this volume) as “rela-
tively stable metaphorical mappings that function as a key framing device
within a particular discourse over a certain period of time”. In an elabora-
tion of, and also in contrast to, their contribution, the present chapter fo-
cuses on a somewhat broader concept, that of discourse metaphor forma-
tions, and, more concretely, on a single example: the evolution and
discourse career of the 19th century “language-organism-species” metaphor,
which has not lost its power of attraction and, like metaphor formations in
other sciences, continues to function as a source for heuristic inferences in
contemporary investigations of language and language change, particularly
in the case of those attempting to incorporate an evolutionary or Neo-
Darwinian perspective in their overall approach.
Discourse metaphor formations provide evidence for the sociocultural
situatedness of metaphorical reasoning along with the characteristic fea-
tures of context-boundedness, strategic fuzziness, and polyvocality. As I
shall attempt to show, such discourse metaphor formations not only have a
rich social and cultural history, they can also demonstrate an uncanny con-
ceptual staying power, which reflects their status as highly entrenched,
albeit constantly changing, entities, given that the sociocultural ground
under them is always shifting (Nerlich and Hellsten 2004: 262; Zinken,
Hellsten and Nerlich this volume). This set of conditions allows the dis-
course metaphor formation to interact and hence co-evolve with its so-
ciocultural environment. On the one hand, this environment acts to provide
stability for the formation, i.e., there are cognitive constants that seem to be
discursively embedded in a relatively stable reservoir of cultural beliefs
and social representations. On the other hand, these environmental factors
can act to destabilize the dynamics of the construct, given that the dis-
course metaphor formation simultaneously provides sites for conflict,
resolution and cooperation. In sum, the meanings associated with a given
discourse metaphor formation are socioculturally situated and co-evolve in
conjunction with the cultural constructs in which it is embedded (Zinken,
Hellsten and Nerlich this volume).
The language-organism-species analogy 217
The expression discourse metaphor formation was inspired, in part, by

Condit’s discussions of rhetorical formations, a concept she has used to
illuminate the shifting meanings of the notion “gene” in the 20th century
discourses of eugenics and genetics (cf. Condit 1999, 2001; Henze 2004).
Condit explains that it is useful to think of groups of circulating utterances
in terms of rhetorical formations, defined as the “relatively co-occurrent
sets of discourse – metaphors, narratives, [and] values – surrounding a
given theme in a particular period” (Condit 1999: 14). By substituting the
expression discourse metaphor formation for Condit’s more general one of
rhetorical formation we can focus our attention on the advantages that a
socioculturally situated perspective on metaphor and meaning-creation
provides us, especially with respect to its value as an instrument for ana-
lyzing these complex dynamic conceptual systems. In addition, it is a con-
struct that serves to highlight processes of change, stabilization and desta-
bilization, as well as conflict, in which “alternative definitions and
perspectives struggle against each other, producing brief and tenuous mo-
ments of stasis rather than monolithic, permanent formations. This view of
discourse is useful because it recognizes the ongoing tensions and opposing
forces, rather than the moments of apparent stability, as the most salient
features” shaping the interpretative strategies brought into play for under-
standing a given metaphor at a specific juncture in time as well as over
longer periods of time (cf. Henze 2004: 315). In this way a discourse meta-
phor formation provides a localizable framework of interpretation which,
through explicit specification, can be assigned an upper temporal bound-
ary, a point of departure for the analysis of its evolution (Musolff 2006,
this volume).
Furthermore, a closer look at the complex interactions taking place in
the subdynamics of the system will reveal how modifications are brought
about, and the way that various types of attractors can operate upon each
other. In some cases, given the sociocultural situatedness of the language
agents, the resulting environmental resonances may contribute to stabiliza-
tions within networks making up the overall meaning-making system.
Moreover, a confluence of opinions, beliefs and motives can allow certain
nodes to be selected for globalization, that is, discursive prominence can be
given to a specific node or cluster of meaning(s) within the formation, e.g.,
the new node produced by the conflation of the notions of “language”,
“species” and “race”. These shifts in salience are the end result of aggre-
gate actions of individual language agents operating at the local level over
time – and these can be integrated into the discourse metaphor formation at
218 Roslyn M. Frank
a global level. Thus, a discourse metaphor formation is not a monolithic

entity. Instead, it should be viewed more as a phenomenon whose dynamic
structure is, always (somewhat) unstable – always in flux – as a result of
the continual process of re-negotiation taking place between competing
voices, microstructures which in turn must operate in local environments
conditioned by the top-down influence of (pre-)existing emergent global-
ized macrostructures.
In other words, as we shall see, some meanings, more than others, cre-
ate resonances between the different domains in society and across differ-
ent disciplinary and epistemic communities. At the same time nonlinear
interactions are set in motion giving rise to feedback loops, and these, in
turn, produce emergent meanings that can become causes of further stabili-
zation or destabilization of the (pre-)existing structures of the discourse
metaphor formation. While activated through microstructural acts of indi-
vidual language agents, the cumulative effect can lead not only to local
stabilization, i.e., within the idiolect of the language agent or a (special-
ized) epistemic community within the larger community, but also to the
modification of the macrostructure, the global level. Nonetheless, the point
at which such globalized stabilization takes over is fuzzy and at this stage it
is probably still far too early to attempt to determine what factors should be
taken into consideration in order to establish when a particular transition
has taken place.
Stated differently, in the case of discourse metaphor formations, par-
ticularly those used in scientific fields, because of the concrete sociocul-
tural situatedness of individual language agents which inevitably places
them within a given location and time frame, as well as the heterogeneous
and distributed nature of cultural conceptualizations in general (Sharifian
this volume, forth.), any given set of language agents operating, locally, can
employ anachronistic interpretive frameworks and/or conceptualizations
that are partial or limited in some respect, e.g., ones that are not shared by
the members of the “expert” scientific community in question. In sum,
there are different levels of awareness of the historically conditioned use of
the terms in question and hence different understandings of their accepted
meanings.
2. The language as organism-species metaphor: A case study
2.1. History of the “language as organism” metaphor
As we begin to examine the organicist (organic or organismic) view of

language, we must first of all recognize that as a foundational root meta-
phor of Western thought, “organicism” has a long history:
For twenty-five hundred years a single metaphoric conception of change has
dominated Western thought. Drawn from the analogy between society and
organism, more specifically between social change and the life-cycle of the
organism, this metaphor very early introduced into Western European phi-
losophy assumptions and preconceptions regarding change in society that
have at no time been without profound influence on Western man’s contem-
plation of past, present and future. (Nisbet 1969: 211)
As is well recognized, there has been a constant tendency to see organicism
as the prototype of all dynamic wholes and, consequently, to attribute a
cyclical development to individuals and institutions alike (Nisbet 1970:
70), and to language, also. This fact is clearly demonstrated in the literature
(cf. Janda and Joseph 2003: 6–13). A typical instance is offered in an early
work by Herder (1767), where language “geminates, bears buds, flowers
and eventually withers away” (Herder 1877–1913, i, 151–152, cited in
Morpurgo-Davies 1992: 84). For this reason 19th century interpretations of
the organicist metaphor and their applications to the concept of “language”
must be viewed in light of earlier conceptual networks, their interactions
with each other and the resulting habits of thought, as well as the so-
cioculturally situated and contextualized variations on these much earlier
incarnations of the metaphor. Consequently, at no time did a single writer
(speaker) ever have access to the full range of prior social usages of the
term, that is, the global level of usage consisting of centuries – actually
several millennia – of reformulations brought about by the aggregation of
individual utterances over time. In other words, the individual language
agent always has had a limited vantage point from which to construct her
cultural conceptualization(s) of the notion of organicism.
These understandings are, at a local level, microstructures in the sense
that the individual’s usage of the term feeds into the larger discourse meta-
phor formation. Yet, the language agent’s knowledge remains local and
socioculturally situated in a concrete time and place. Then there is an addi-
tional problem: in the 21st century, these 19th century texts tend to be read
through a different sociocultural lens and, in this fashion, actualized by it.
220 Roslyn M. Frank
In order to grasp what was going on, we need to be aware of the way that
the term organicism was connected to other constantly evolving cultural
networks and domains. Also we should recognize the fact that these latter
elements were interacting not only with each other but also acting to pro-
duce emergent structure which was incorporated into the discourse meta-
phor formation as a whole: that the nodes and clusters composing it were
interconnected and dynamic in the way that they functioned. In summary,
because of the complex interactions taking place among the various parts
that make up the overall system of any discourse metaphor formation, its
causality needs to be viewed as nonlinear, as opposed to being a simple
feed-forward or linear type of causality. Otherwise, we will be trying to fit
linear theory into a narrative that requires recognition of dynamic nonlinear
changes where causation patterns involve both feedforward and back-
propagation relations, as well as effects brought about by dynamic interac-
tions between the parts of the overall system (cf. Clark 1997; Strohman
1997).
Finally, because of the great time-depth involved with respect to the
root metaphor of organicism, an upper temporal limit must be set which
will allow us to focus on its subsequent development, as certain networks
composing it are restructured to form the “language-organism” aspect of
the “language-organism-species” metaphor. This must be done in order to
reduce the complexity that otherwise would be involved in the task of ex-
plicating the wide range of meanings applied to it and evoked by it. For our
purposes we shall concentrate our attention primarily on its reformulations
during the 19th century with occasional backward glances. When speaking
of the “language-organism” aspect of the discourse metaphor formation, we
must also consider the epistemological power of the metaphor itself, in-
cluding the remarkable ambiguities concerning its literal and figurative
existence, that is, whether a given author is using it solely rhetorically to
argue a position or, strategically, in order to gain ground and disciplinary
prestige; or whether, in fact, the individual is truly committed to its literal
interpretation.1
1. In the 19th century, among those promoting the organic view of language, espe-
cially among those who were writing in English, there was yet another factor
that was beginning to exert an influence on the discourse metaphor formation,
namely, the narrowing of the scope of meaning of the term “science” itself to
include only the “natural sciences”, whereas earlier the term included, unprob-
lematically, the “moral (historical or social) sciences”, those “sciences” that
dealt with phenomena caused by freely choosing agents. This process intensi-
Moreover, as is well documented (Alter 2005; Morpurgo-Davies 1992:

83–97), depending on the circumstances, the same writer might move back
and forth in her positions vis-à-vis the metaphoricity or literality of the
expression “organism”. Therefore, literal and figurative uses of the meta-
phor can exist simultaneously and are often more a matter of perspective
than of absolute division. Consequently, it can be extremely difficult to
determine which pole of meaning is the dominant one in the work of a
given writer. For instance, at one point Müller appears to insist that lan-
guage is to be understood literally as an “organism”. Yet later he explicitly
rejects anything but a metaphoric usage of the term. As Alter (2005: 122–
206) observes, Müller was willing to change positions, strategically, de-
pending on the advantages he perceived would be gained from each type of
usage.
Somewhat earlier in the 19th century, just after the publication of Dar-
win’s Origin of the Species (1859), we find the writings of Schleicher,
perhaps best remembered for his work on linguistic geneaology, but who
was nonetheless one of the foremost advocates of linguistic organicism.
His writings show that by promoting the organicist metaphor he was also
attempting to establish the bases for a strict dichotomy between older phi-
lological pursuits and the new science of Linguistik:
The discipline which has language as its object but uses it as a way of access
to the spiritual nature and life of one or more Volksstämme is Philologie,
which belongs in essence to History. In contrast with it there is Linguistik
which has language as such as its object and has nothing to do with the his-
torical life of the people who speak the languages; it is part of the natural
history of men. (1850: I, cited in Morpurgo–Davies 1992: 196)
Schleicher’s insistence on defining language as a natural object is also
linked to his desire to put forward an organicist explanation of agency and
hence causality with respect to language change. Thus, the object of lin-
guistics is
not a free mental activity (i.e. history), but language established by nature,
subject to inalterable laws of formation, which is impossible to determine
through individual will, just as it is impossible for the nightingale to change
fied after mid-century. Alter (2005: 123–145) discusses at some length the rhe-
torical strategies that Müller and Whitney brought into play, each in his own
way, in order to counter this progressive semantic narrowing of the concept
“science” and, in turn, to defend the validity of the discipline of linguistics.
222 Roslyn M. Frank
its song; in other words, the object of Glottik is a natural organism. (1860:
120, cited in Morpurgo–Davies 1992: 196)
Moreover, the question of agency has been a central concern in discussions
about the way in which “language” is to be portrayed metaphorically since
the early 19th century and the debate continues, unabated, still today (cf.
sections 3 and 4 below). Indeed, it constitutes a core element, albeit often
operating in the background, in the elaboration of competing explanatory
models of language change (cf. Janda and Joseph 2003; Steels 2004;
Döring and Nerlich 2005; Croft 2006, forth.).
2.2. Linguistic organicism
When examining the development of linguistic organicism in the 19th cen-

tury we must keep in mind that it requires a kind of reflexivity on our part:
a recognition of the processes that allow for the discourse of linguistics to
be viewed as part of a larger body of shifting discourses, all of which are
socioculturally embedded. That is, rather than being bracketed off in a
hermetically sealed discursive space, 19th century linguistic models of lan-
guage were in constant contact with other extra-linguistic domains and,
therefore, engaged in complex intercourse with co-existing conceptual
frameworks, including those that would give rise to the “race” metaphor
and to racialist positions later. Similarly, individuals engaged in linguistic
pursuits often took up positions within these debates, frequently without
full awareness of the (future) implications of their writings, e.g., the way
their words would be appropriated, later, by others. Clearly, the language
as organism metaphor was a powerful and highly flexible discursive re-
source that would be recruited in different fashions and for different pur-
poses, providing rhetorical support first for one ideological stance and then
another, often opposed one (Gal and Irvine 1995). Chameleon-like, it could
change colors, depending on the nature of its contextualization. In short,
the term’s conceptual flexibility allowed it to be recruited in support of
differing even contrasting scenarios. Thus, it was capable of lending per-
suasive force to an argument at the same time that it blended, quite unob-
trusively, into the rhetorical backdrop provided for it (cf. Musolff this vol-
ume).
However, the interpretation of the metaphor of linguistic organicism
varies not only in the works of different authors but even in different pas-
sages of the same author. Moreover, as Morpurgo–Davies (1992: 86) ex-
plains, in Germany the metaphor was used in “at least one of three senses
and sometimes in all”.2 In other words, the term was polysemous from the
beginning. These three basic meanings include the aforementioned equa-
tion of language with an “organism” where the emphasis is on develop-
ment, possibly autonomous (orthogenetic) development, and the associated
entailments relating to its life history, e.g., birth, growth, decay and death.
Secondly, the notion of organicism can be recruited to speak of the basic
unity and mutual dependence or common purpose of all of the parts of
language and at times that “the whole is sometimes said to be greater than
the sum of its parts” (Morpurgo–Davies 1992: 87).3 Then, there is the fact
that
language like law, art, religion, etc. can be seen as an ‘organic’ expression
of the people or the nation; here what matters is the natural, non-mechanical,
non-superficial aspect of the connection. No contradiction is seen between
this ‘organic’ connection and the fact that language may be treated as an or-
ganism in its own right. (Morpurgo–Davies 1992: 86)
2. Cf. also Moss (2004: 9–15) and Keller (2005) for a discussion of Kant’s con-
ceptualization of “organism” and also the entry under Organismus in Eisler
(1930 [2004]).
3. Another node or cluster that would need to be taken into account if one were to
undertake a thorough analysis of the development of the language-organism-
species metaphor and its associated discourse formation, is the role played by
the term Organismus in German and the way that it came to be translated into
English. As is well known, outside of biology and medicine, the term Organis-
mus is generally used to refer to a system, in a metaphorically extended mean-
ing of “(cohesively organized) system”, rather than as the direct equivalent of
the English words “organism”, “organization” or “organ” (cf. Janda and Joseph
2003: 10, ff. 9). That extended meaning is apparent in Bertalanffy’s (1968,
2001–2007) work on general systems theory, written in German. So the ques-
tion must arise as to which of the polysemous meanings was intended in the
original German texts; when was even the contextualized meaning of the origi-
nal text still ambiguous and to what extent did translators of these German texts
consciously decide that the term should be translated, consistently, into English
as “organism”? Furthermore, was it the choice of a single translator that caught
on – setting up resonances with the “organisms” found, for example, in English
texts by Darwin, Wallace and Lyell – or were there other motivations operating
in the background? Or does this choice on the part of translators date back to
earlier German-English translation practices?
224 Roslyn M. Frank
In the 17th and 18th centuries the organic analogy had already been linked to
the expression “evolution”, the latter being understood in the vernacular of
the time as “a kind of unfolding”, the supposed series of changes that a
species was predetermined to undergo, like an embryo is preprogrammed
to develop. Stated differently, on this view organisms develop and change
through pre-programmed inner forces, a theory known in biology as ortho-
genesis: that evolutionary change is predetermined by the constitution of
the germ-plasm and independent of external factors. When applied to cul-
tures, peoples, nations and languages, it becomes a theory that alleges all
cultures (and languages) pass through the same sequential periods or stages
of growth (and decay) in the same order. It is the inner spiritus of the or-
ganism that manifests itself, unfolding over time. Change is unidirectional
and uniform: determined by the inner nature of the organism, just as an
embryo passes through predetermined and irreversible stages, so do all
other entities defined as organisms, e.g., nations and languages. Such pre-
formationalist views persisted into the first half of the 20th century, despite
the fact that Darwin’s own theory asserted no such predetermined series or
stages (Mayr 1982).
2.3. Semantic shifts in “species” and “race”
The conflation of language with the concepts of “people”, “nation” and

even “race” leads us to another robust metaphorical network operating
within this discourse metaphor formation, one that was active in the 19th
century but which today has more or less disappeared from view, even
though its influence continues to shape the way that we understand and use
the term “race” (Ascroft 2001). In order to appreciate what has happened,
let us begin by looking at the way Darwin’s first major work is commonly
cited today, namely, as if its title were simply The Origin of Species
(1859). Some might argue that the shortened version of the title is used
merely for convenience. However, I suspect that other factors may be in
play, ones that in the 20th century contributed to a studied avoidance of
Darwin’s original choice of nomenclature: On the Origin of Species by
Means of Natural Selection, or the Preservation of Favoured Races in the
Struggle for Life. It is the second half of the title that might cause a visceral
reaction on the part of a contemporary reader, who might not immediately
recognize that the races in question were primarily pigeons.
The following excerpt, taken from the letter that Alfred Russel Wallace
sent to Charles Darwin, in 1858, explaining his own theory of the evolution
of species, could evoke a similar response in a naïve reader:
But this new, improved and populous race might itself, in course of time,
give rise to new varieties, exhibiting several diverging modifications of
form, any of which, tending to increase the facilities for preserving exis-
tence, must, by the same general law, in their turn, become predominant.
[…] It is not, however, contended that this result would be invariable; a
change of physical conditions in the district might at times materially modify
it, rendering the race which had been the most capable of supporting exis-
tence under the former conditions now the least so, and even causing the ex-
tinction of the newer and, for a time, superior race. (from Wallace 1858:58,
cf. Reveal, Bottino and Delwiche 1999)
Our initial negative reaction to Wallace’s fundamentally innocent phrasing
is brought about, in part, because of the major shift that has taken place
with respect to the meaning of the words “species” and “race” since the
time of Darwin. And, that shift, in turn, is related to what was taking place
within a highly salient network of the language-organism-species meta-
phor. Indeed, the shift demonstrates how reorganization can take place
inside a discourse metaphor formation, and more concretely, how the
meaning-making potential of a given aspect of the language-organism-
species metaphor could be exploited.4 So at this juncture, we shall turn our
4. Wallace, a committed socialist who rejected the notion of Lamarckian acquired

features from the onset, uses the term “superior” in concrete reference to any
race that at a given point in time and space has numerical superiority over an-
other, e.g., over another species of birds. Into a remarkably modern sounding
discussion of the complex interrelationship of niche construction, access and
availability of food sources, habitat and other environmental influences on
population size, he inserts this comment: “Now it is clear that what takes place
among the individuals of a species must also occur among the several allied
species of a group, – viz. that those which are best adapted to obtain a regular
supply of food, and to defend themselves against the attacks of their enemies
and the vicissitudes of the seasons, must necessarily obtain and preserve a supe-
riority in population; while those species which from some defect of power or
organization are the least capable of counteracting the vicissitudes of food, sup-
ply, andc., must diminish in numbers, and, in extreme cases, become altogether
extinct. Between these extremes the species will present various degrees of ca-
pacity for ensuring the means of preserving life; and it is thus we account for the
226 Roslyn M. Frank
attention away from the analogy formed by “language” understood as an

“organism”, and move on to examine the conflation of “language” with
“species”, and with the concept of “race”.
Briefly explained, the story goes as follows. As has been noted, the
identification of a language with a people or nation has deep roots in West-
ern thought (Ascroft 2001; Gal and Irvine 1995; Poliakov 1974). Further-
more, in the early 19th century, although the fixity of species was still not
being challenged, at least not frontally, the genealogical study of languages
was well advanced in terms of recognizing their changing nature over time.
Hence, Schleicher and others were keenly aware of the fact that the ge-
nealogical ranking of languages and their interrelationships provided a
ready-made analogue for the emerging Darwinian consensus about the
changing nature of other aspects of the natural world, namely, the origin of
species. And at the same time, by bringing the linguistic analogy to bear in
support of the Darwinian thesis, it could also be used to promote the cause
of the emerging field of comparative linguistics. This line of reasoning is
patently evident in Schleicher’s widely publicized essay, published in
1863, shortly after the German translation of Darwin’s work in that same
year. Schleicher’s German language contribution was entitled, relatively
modestly, Die Darwinsche Theorie und die Sprachwissenschaft (Darwin-
ian Theory and the Science of Language). Schleicher sent a copy of his
brief tract, written in response to the publication of The Origin of Species,
to Darwin himself. Shortly later an English edition appeared under the
much more striking title of Darwinism Tested by the Science of Language.5
Then a few years later, in the Descent of Man (1871), Darwin returned the
favor, picking up on the heuristic value of incorporating the linguistic ar-
guments laid out for him earlier and ever so neatly by Schleicher, Lyell and
abundance or rarity of species” (from Wallace 1858:53, cf. Reveal, Bottino and
Delwiche 1999).
5. Schleicher, who was also an ardent gardener, wrote a separate review of Dar-
win’s work called “Die Darwinsche Theorie und die Thier- und Pflanzenzucht”
[“Darwinian Theory and Animal and Plant Breeding”] (1864) which was pub-
lished in an agricultural journal. In it Schleicher “summarized Darwin’s argu-
ment and added elements that he undoubtedly thought rounded out the theory,
including the suggestion that human beings descended from the ‘higher apes’
and differed from them only by reason of language and ‘high brain develop-
ment’. Schleicher neglected to mention that Darwin himself did not discuss hu-
man evolution in the Origin” (Richards 2002a: 169–170).
others (Alter 1999: 73–79; Koerner 1995; Nerlich 1990; Richards 2002a:
26–40; Taub 1993).
The contributions of Schleicher and others to Darwin’s thought proc-
esses and Darwin’s own often subtle process of argumentation by analogy
should not be underestimated (Alter 1999). Even in his first major work of
1859 Darwin was already conscious of the rhetorical power of the linguis-
tic analogy. For instance, in a chapter dedicated to classification and sys-
tematics, Darwin cites the classification of languages and then moves on
directly to the classification of species, varieties and subvarieties:
It may be worth while to illustrate this view of classification, by taking the
case of languages. If we possessed a perfect pedigree of mankind, a genea-
logical arrangement of the races of man would afford the best classification
of the various languages now spoken throughout the world; and if all extinct
languages, and all intermediate and slowly changing dialects, were to be in-
cluded, such an arrangement would be the only possible one. Yet it might be
that some very ancient language had altered little, and had given rise to few
new languages, whilst others (owing to the spreading and subsequent isola-
tion and states of civilisation of the several races, descended from a common
race) had altered much, and had given rise to many new languages and dia-
lects. (Darwin 1859: 422)
Darwin goes on to say that
[t]he various degrees of difference in the languages from the same stock,
would have to be expressed by groups subordinate to groups; but the proper
or even only possible arrangement would still be genealogical; and this
would be strictly natural, as it would connect together all languages, extinct
and modern, by the closest affinities, and would give the filiation and origin
of each tongue.
In confirmation of this view, let us glance at the classification of varie-
ties, which are believed or known to have descended from one species.
These are grouped under species, with sub-varieties under varieties; and
with our domestic productions, several other grades of difference are requi-
site, as we have seen with pigeons. (Darwin 1859: 422–423)
In this passage, as can be seen, Darwin clearly recognized the isomorphism
between language descent and human biological descent. “Not only could
the human pedigree serve as a model for tracing linguistic development, as
he here emphasized, but also the reverse, as he implied, could be the case:
the descent of language might serve as a model for the descent of man”
(Richards 2002a: 24).
228 Roslyn M. Frank
Darwin’s suggestion about a similar genealogy holding for human be-

ings and language appeared casually in only one paragraph of the Origin
and as Richards (2002a) and Alter (1999) have shown, at this stage Darwin
did not really bring the linguistic analogy into play in any systematic fash-
ion. Nonetheless, the “bare suggestion of this apparent isomorphism be-
tween the development of language and the development of human varie-
ties […] caught fire almost immediately” (Richards 2002a: 25). Adding to
the furor that this analogy provoked were other passages in Darwin’s work
where he labored over how to define the concepts of species and variety. In
the end he concluded that since there was no clear distinction between
“species” and “varieties”, we could regard human groups as forming either
several species of one genus or several varieties of one species, although he
preferred the latter way of putting it (Richards 2002b).
While Darwin recognized physical and intellectual differences among
the different human groups, his assumption of their common descent made
these unimportant. Yet the fact that he left the distinction between species
and varieties ambiguous allowed for the interpretation that humankind was
formed of several species. And that in turn opened the door to more devi-
ous polygenetic interpretations. This is because a “species” according to
prior definition, that is, the commonly accepted biological definition of-
fered by Buffon over a hundred years earlier, was: a group that could inter-
breed without difficulty or otherwise negative consequences (Farber 1972).
So by allowing for the possibility that human groups formed several spe-
cies of a single genus, the following logic could be applied: by assigning a
polyphyletic origin to human “species” (equated semantically with
“races”), it follows that the latter are not meant to interbreed.
In order to appreciate the way Darwin’s ambiguous stand on defining
species and variety fits into the larger picture, we need to briefly examine
several other aspects of 19th century classification systems in biology and
linguistics, most particularly how their taxonomic categories interacted and
became structurally aligned. The results of this alignment, in turn, created a
conceptual feedback loop that operated between the two disciplinary do-
mains (as well as between other related disciplines). By the end of the 18th
century attention was focused on an aspect of the Great Chain of Being that
had been used by naturalists to explain the natural world by means of a
series of interrelated forms arranged on a graduated scale of complexity,
varying from the simplest inanimate organisms to the most complex mam-
mals, a taxonomy that placed human beings as the link between animal
creation and the divine. By the turn of the century there was increased fo-
cus on how the varieties of humans should be classified, a concern that

reflected an historical moment when colonial and economic expansion and
missionary work was rapidly contributing new evidence of the vast differ-
ences among human types and human languages. In short, by the end of the
18th century, human difference had become a well-developed topic of in-
quiry, featured in the work of natural historians and taxonomists, and com-
plemented by discussions of the diversity of the world’s language(s) (Gal
and Irvine 1995; Henze 2004: 312; Stepan 1983: 6–7).
The Linnaean (1737) taxonomic system, which gained massive accep-
tance starting in the first part of the 18th century, was based on a series of
hierarchical categories: each kingdom was a descending hierarchy com-
posed of four levels: classes, order, genus and species while later a family
level was introduced between order and genus, so that the resulting se-
quence was broken down into: class, order, genus, family and species. At
the same time, there were other terms – more informal taxa – that contin-
ued to be used to characterize resemblance and difference between organ-
isms, terminology rooted firmly in discourses as diverse as natural history,
philosophy and agriculture. These included: stock, kind, type and especially
race. Moreover, by the 19th century the two sets of terms were regularly
intermingled and their meanings hopelessly confused, at times deliberately
and at others inadvertently, a situation that arose, in part, from their wide-
spread use in pre-Linnaean discursive practices.6 Of these terms the most
polysemous was undoubtedly “race” (Henze 2004), followed closely by the
combination of “species” and “variety”.
In addition, almost from the moment that Linnaeus brought forward his
classification system there were overt efforts to align it structurally with a
model that would classify languages into a similar hierarchy. And at the
same time, there was the recognition that the discursive interactions be-
tween the system naturalists used to organize the natural world, the Lin-
naean system, and the more informal taxa had produced a dangerous situa-
tion: the ambiguous referentiality of the terms, most specifically, “species”,
6. The ethnologist James Cowles Prichard (1786–1848) was unquestionably the

most persistent and determined writer of his time concerned with correcting this
problem. In 1813, when he composed his major opus Researches into the
Physical History of Man, we see him struggling valiantly to take rhetorical con-
trol of the terms race, species and variety. In fact, according to Henze (2004),
Prichard’s Researches was arguably the most influential and certainly the most
detailed pre-Darwinian account of human variation. For a detailed discussion of
the rhetorical strategies used by Prichard, cf. Henze (2004).
230 Roslyn M. Frank
“variety” and “race”. Many 19th century discussions revolved, albeit tacitly,
around efforts to stabilize (and destabilize) the meanings of these three
terms through recourse to different types of rhetorical strategies of defini-
tion and redefinition, and motivated by often opposing points of view, most
particularly with respect to the ongoing debate over monogenism and poly-
genism.
2.4. Monogenism vs. polygenism
In terms of our historical narrative there is still a missing link, namely, the
node that acted to connect the metaphor of linguistic organicism to the
heated discussion taking place in other disciplinary domains in natural
science, ethnology and the incipient field of anthropology, namely, the
aforementioned debate over monogenism and polygenism. As stated, the
subdynamics of the language-organism-species metaphor was impacted by
the conflation of the concepts of people, nation, tribe and race, all of which
were identified, in turn, with the concept of language.
By the middle of the 19th century, the debate concerning the unity of
mankind and, consequently, the origins of language as deriving from a
single source was heating up, as opponents and defenders took up positions
and readied their rhetorical ammunition. Although at first glance, one
might assume that the two questions were not that intimately related, this
was not the case. Rather the language-organism-species analogy provided a
ready-made weapon which could be brandished, curiously enough, by de-
fenders of either side of the debate. The Biblical view espoused a mono-
phyletic or monogenetic origin for humankind and human languages: all
languages derived from a single source, the original pair. This theory was
based on the idea that the morphological diversity of humans could be
traced back to one primordial family, and all languages to a single lan-
guage. In contrast, those defending the multiple, polyphyletic or polyge-
netic origin theory dismissed the orthodox Biblical account and held that
from the beginning humankind had been divided into separate and unre-
lated “species” or “races”, even though the exact number and nomenclature
of these divisions varied.
The debate was framed in such a way by the proponents of polygenism
that those supporting monogenism frequently were portrayed as unscien-
tific, as backward religious thinkers who would not give way to new ideas.
Thus, the “scientific” solution was polygenism, the idea that different ra-
cial groups were the result of different initial pairings and implicit in that
viewpoint was the belief in the innate inequality of human racial groups.
Yet many in this camp also relied on deviant (pre-Adamite) literalist inter-
pretations of the Bible to lend support to their polygenist views (Living-
stone 1992; Poliakov 1974). In addition, the polygenist perspective tended
to integrate an evolutionary model in which change was viewed as progres-
sive, as opposed to the older model based on a degenerative view of differ-
ence, so that the species or races placed on the bottom rungs of the hierar-
chy were portrayed as more primitive and, hence, less evolved, rather than
having arrived there through regression or degeneration.
Moreover, given the organicist point of view, languages and nations,
tribes and peoples tended to be defined by and inseparable from the lan-
guages they spoke, the latter being natural objects. Because of the struc-
tural alignments inherent to the discourse metaphor formation, when “spe-
cies” was used interchangeably with “race” and the terms “species”, “race”
and “language” were all conflated with each other, what resulted was a
language-species-race isomorphism. Yet in discursive practice these analo-
gies were often vague, diffuse, and at times their impact was almost imper-
ceptible. In short, they were recruited in contradictory ways and with dif-
ferent rhetorical goals. Nevertheless, they were available.
Although one might assume that the defenders of polygenism lined up
on Darwin’s side, this was not the case. Instead, it was the group of mono-
genists who, defending a modified Biblical account, rejected the multiple
origin theory for humankind and for languages (Henze 2004). There were
many monogenists who were already inclined to believe that humans were
shaped by their environment and when Darwin published his theory of
evolution by natural selection, they supported it. Not only did they believe
in monogenism, they tended to be politically liberal, especially on matters
related to race. In contrast, the defenders of polygenism tended to lend
support to pre-Darwinian views, espousing the doctrine of Lamarckian and
Malthusian individualistic “struggle for existence”. These views would
later come to be known collectively as Social Darwinism and eventually
linked to what would become overtly racialist theories. Hence, mono-
genism and polygenism did not begin with Darwin, but rather much earlier
(Augstein 1996). However, by mid-century the forces of polygenism were
rapidly gaining ground against those whose orthodox Christian views
and/or acceptance of Darwinian theory bound them to a monogenist posi-
tion.
232 Roslyn M. Frank
2.5. Conflating “species” and “race”
Against this backdrop of Linnaean classification increased attention was

being focused on defining the groups situated between humans and pri-
mates and in that sense categorizing mankind. Following the still prevalent
degenerationist viewpoint which dated back to the Middle Ages, in the 18th
century, adherents of the Biblically-based view of the unity of mankind and
the origin of human language(s) positioned Homo sapiens europeaeus at
the top of the Linnaean hierarchical order. Although not yet identified with
the term “race” or “species”, the human groups in question were classified,
implicitly, in a descending order. And over time what was originally a geo-
graphically-based taxonomy would come to be replaced by another, based
primarily on skin-color.
The early emphasis on geography as a determining factor in this ranking
came about because human differences were viewed as the result of the
(adverse) effects of climate, diet and other environmental factors on the
original populations, changes that brought about what some viewed as
diversity, while most others saw them as reflecting the negative effects of
degeneration (Poliakov 1974). Indeed, the term “race” was applied to “va-
rieties” of Homo sapiens in the middle of the 18th century, concretely, in
1745, by Buffon (1708–1788), who subscribed to a monogenist view in
which the geographically distributed diversity of humankind was explained
by a process of dégradation or dénaturation, attributed to the combined
effects of climate, eating habits, and environment. Still the fixity of the
species themselves was not openly questioned and in Buffon’s writings he,
too, uses the terms “race” and “species” almost interchangeably.
Among the very first to break with the belief in the fixity of species, al-
beit tentatively, was a protégé of Buffon, Jean Baptiste Lamarck (1744–
1829) who at the beginning of the 19th century brought forth his arguments
in a volume called Zoological Philosophy (1809 [1999]), his most famous
work. In it he describes his “theory of transmutation” which was under-
pinned by the following principle: that Creation is in a constant state of
advancement and that, therefore, there is an overall “tendency to progres-
sion”. To explain the instability of species he laid out what has become
known as the “theory of acquired characteristics”, that through the giraffe’s
determined efforts he was able to stretch his neck, longer and longer, and
7
that trait was passed on to his offspring making them more fit.
For our purposes, however, there is another aspect of Lamarck’s work
that concerns us, namely, the way in which he incorporates the Linnaean
principles of classification, most specifically the explicit semantic equiva-
lency that he establishes between the concepts of “species” and “race”.
This he does in the first chapter of his Zoological Philosophy: “We give
the name genus to the groups of races, called species, [that are] brought
together following a consideration of their interconnections […]. When a
genus is created well, all the races or species which it includes are similar
in their most essential and most numerous characteristics […]” (Lamarck
1809 [1999]).
By equating “species” and “race” Lamarck was following in the foot-
steps of his mentor Buffon who earlier had set forth what would become
the most widely-embraced definition of “species”, although one that was
contested by many even at the time (Henze 2004).8 Buffon was among the
very first to speak of human beings as a “natural species”, while assigning
to the term “species” a biological (rather than morphological) meaning that
has not been abandoned even today: that of an interbreeding population,
one capable of producing fertile offspring. And, in spite of efforts to the
contrary, the lack of specificity attached to the terms “species” and “race”
(as well as “variety”) persisted. In fact, this lack of specificity gave rise to
7. Lamarck’s doctrine of the “inheritance of acquired characteristics” represents

another component of the concept of orthogenesis, i.e., the vitalist notion of the
workings of the inner spirit. Also the notion of “the survival of the fittest” re-
lates to Lamarck’s position concerning the transmutation of species, specifi-
cally, his doctrine of struggle which placed the onus of change on the individ-
ual. In 1809, when his major work Philosophie zoologique was first published,
his evolutionary theory was not particularly well-received by his scientific col-
leagues, e.g., Lyell and Cuvier, although it resonated strongly with the vitalist
tendencies of his time, e.g., those of the German romantic movement, and was
embraced enthusiastically by the so-called Social Darwinists. Indeed, through-
out the 19th century and into the 20th, among non-biologists, the Lamarckian
(pre-Mendelian) explanation of the mutation of species was far more popular
than Darwin’s more population-based views.
8. As Henze (2004: 329–330, ff. 6) notes, the present-day concept of species is no
less polysemous. There are at least four major “species concepts”, each with
one or more species definitions, currently vying for recognition (cf. also Mayr
1996).
234 Roslyn M. Frank
such a degree of conceptual flexibility that the terms would be recruited

over and over again in differing and even contrasting scenarios and where
their very ambiguity was a rhetorical asset (Musolff 2006: 70).
We should keep in mind that previously “race” was used primarily to
describe breeds of domestic animals and plants, their group membership or
descent from a common ancestor, that is, the vertical aspect was empha-
9
sized. Furthermore, Richards (2002b: 697) observes that
[t]he term ‘race’ and its equivalent in several languages gained currency in
the seventeenth century to describe descendents of the same family or house.
The word was also used to refer to a tribe or nation, as in the Germanic
races. Only in the nineteenth century did the term take on the taxonomic
meaning of a distinctive group or variety within a human or animal species.
In other words, until the 19th century meanings associated with the word
“race” tended to emphasize descent and offspring, whether human, plant or
animal.10 In summary, it was only in the 19th century that the term “race”
came to acquire an essentialist meaning, that is, as a distinct category of
human beings with physical characteristics transmitted by descent. As
Augstein (1996, 1998) demonstrates, what a late 19th century speaker of
English understood as natural division by race had not been at all natural
9. Under “race” The American Heritage Dictionary of the English Language

(Morris 1971) lists the following entries, which reflect the semantic shift that
has taken place in the term’s core meanings since the mid-19th century, and a
point in time when the fourth and sixth entries would have been among the first
to come to mind: “1. A local geographic or global human population
distinguished as a more or less distinct group by genetically transmitted physical
characteristics; 2. Mankind as a whole; 3. Any group of people united or
classified together on the basis of common history, nationality or geographical
distribution; 4. A genealogical line, lineage, family; 5. Any group of people
more or less distinct from all others, the race of statesmen; 6. Biology a. a plant
or animal population that differs from others of the same species in the
frequency of hereditary traits, subspecies; b. a breed or strain of domestic
animals (Morris 1971: 1074–1075). An even greater appreciation of depth of
these shifting currents can be gained by consulting the relevant entries in the
Oxford English Dictionary.
10. With respect to the contemporary range of uses of the term Rasse in German:
“In German you can’t really use the word Rasse in the [more recent] 19th and
20th century meaning, I think, without shuddering; however we still talk about
Rassehunde ‘pedigree dogs’, for example, where the word is still used in its 18th
century sense and has no negative connotations” (Nerlich 2007).
only a century earlier. Indeed, especially toward the end of the 19th century,
publications specifically ranking different groups of people became
extremely popular. For example, Gobineau’s Essai sur l'inégalité des races
humaines (1853–1855) was touted as one of the milestones in the new
racialist discourse, a discourse that took full rhetorical advantage of the
conflation of species and race, as well as the language-organism-species
metaphor (Poliakov 1974).
2.6. Taking a second look at the language-organism-species metaphor
Keeping this general historical overview in mind, we can examine in more

detail other implications of the monogenism-polygenism debate and how
the structural alignments and even conflation of language, organism, spe-
cies and race played a major role. As we have noted, in 1863, after reading
a German translation of Darwin’s work, Schleicher responded by writing a
short work that, in 1869, was translated into English and published with the
openly confrontational title of Darwinism Tested by the Science of Lan-
guage. The English translation of Scheicher’s commentary appeared three
years before Darwin would publish his Descent of Man (1871), giving the
latter ample time to take full rhetorical advantage of the provocative lin-
guistic analogies that Schleicher (and others) had brought into sharp focus.
Languages are organisms of nature; they have never been directed by the
will of man; they rose, and developed themselves according to definite laws;
they grew old, and died out. They, too, are subject to that series of phenom-
ena which we embrace under the name of ‘life’. The science of language is
consequently a natural science; its method is generally altogether the same
as that of any natural science. […] Now we observe during historical periods
how species and genera of speech disappear, and how others extend them-
selves at the expense of the dead. I only remind you, by way of illustration,
of the spread of the Indo-European family […]. (Schleicher 1869; in Ko-
erner 1983: 20, 60)
Schleicher continues to build on his argument by quoting two sentences
taken directly from Darwin’s 1859 work, citations set off here by double
quotation marks:
“If any group has once been extinguished it can never appear again, because
a chain in the link of generation has been broken.”
236 Roslyn M. Frank
“This explains how the extension of dominant species which admit of the
greatest variation, peopled the earth in the course of time with other forms of
life, closely related though modified; and how these generally succeed in
supplanting those groups of species which succumb to them in the struggle
for existence.”
Not a word of Darwin’s need be changed here if we wish to apply this rea-
soning to the languages. Darwin describes here with striking accuracy the
process of the struggle for existence in the field of human speech. In the pre-
sent period of the life of man the descendents of the Indo-Germanic family
are the conquerors in the struggle for existence; they are engaged in contin-
ual extension, and have already supplanted or dethroned numerous other
idioms. The multitude of the Indo-Germanic species and sub-species is il-
lustrated by our genealogical tree. (Schleicher 1869; in Koerner 1983: 63–
64)
In 1865, once again we find Schleicher attempting to promote language as
the definitive biological marker for classifying the “races” or “species” of
mankind, surpassing in importance the variations in morphological features
commonly brought to bear in such discussions. In order to do this, lan-
guages had to be viewed as “organisms of nature”, as an innate endowment
of the members of each human species, while the latter are divided up ac-
cording to their language family. As he argues in his work Über die Be-
deutung der Sprache für Naturgeschichte des Menschen (On the Signifi-
cance of Language in the Natural History of Mankind):
How inconstant are the formation of the skull and other so-called racial dif-
ferences. Language, by contrast, is always a constant trait. A German can
indeed display hair and prognathous jaw11 to match those of the most dis-
tinctive Negro head, but he will never speak a Negro language with native
fluency […]. Animals can be ordered according to their morphological char-
acter. For man, however, the external form has, to a certain extent, been su-
11. The adjective “prognathous” refers to a jaw that projects forward to a marked
degree. Maxillary prognathism is a protrusion of the maxilla, and is also a
common feature of many populations. It affects the middle third of the face,
causing it to jut out. In the context of Schleicher’s comment, the term is linked
to studies of “craniofacial anthrometry” which came into vogue in the 19th
century when anthropologists began to measure human skulls in their attempts
to categorize race. More technically, today the term “prognathism” is used to
describe the positional relationship of the mandible and/or maxilla to the
skeletal base in cases where one of the jaws protrudes beyond a predetermined
imaginary line in the sagittal plane of the skull (cf. Wikipedia 2007 b).
perseded; as an indicator of his true being, external form is more or less in-
significant. To classify human beings we require, I believe, a higher crite-
rion, one which is an exclusive property of man. This, we find, as I have
mentioned, in language. (Schleicher 1865: 16, 18–19, cited in Richards
2002a: 30)
In 1868, we discover Haeckel taking rhetorical advantage of Schleicher’s
(1863) observations and drawing out the conclusion that derives from this
overt conflation of nation, species and race with languages, namely, in his
Die Natürliche Schöpfungsgeschichte (The Natural History of Creation).
The conclusion brings the position Haeckel is espousing into line with that
of the supporters of a polygenetic origin for humankind and language:
We must mention here one of the most important results of the comparative
study of languages, which for the Stammbaum of the species of men is of the
highest significance, namely, that human languages probably had a multiple
or polyphyletic origin. Human language as such probably developed only
after the species of speechless Urmenschen or Affenmenschen had split into
several species or kinds. With each of these human species, language devel-
oped on its own and independently of the others. At least this is the view of
Schleicher, one of the foremost authorities on the subject. […] If one views
the origins of the branches of language as the special and principal act of
becoming human, and the species of humankind as distinguished according
to their language stem, then one can say that the different species of men
arose independently of one another. (Haeckel 1868: 511, cited in Richards
2002a: 45)
In the later editions of his Natürliche Schöpfungsgeschichte (1868–1911),
Haeckel identified twelve species that derived from the original “ape-man”,
maintaining that the “Mediterraneans” (consisting of the Indo-Germans,
Caucasians and the Hamo-Semites) were the most evolved. As can be ob-
served in the previous quotation, he also “believed the hereditary effect of
language to be the engine producing the various species. Languages that
had the most potential for human thought produced races with brains hav-
ing the greatest capacity for thought” (Richards 2002b: 697). With this
final example of rhetorical slight-of-hand – Haeckel’s conflation of the
Stammbaum of the “species of men” with that of the implied “species of
human languages” – we can appreciate the way that the dynamics intrinsic
to the metaphor of language as an “organism” and “species” not only inter-
acted but actually co-evolved with its sociocultural environment; that on
the one hand there were conditions acting to provide stability for the for-
mation, cognitive constants discursively embedded in a relatively stable
238 Roslyn M. Frank
reservoir of cultural beliefs and social representations, while on the other

there were environmental factors which destabilized the dynamics of the
construct and/or exploited (sub)networks of meaning-making emergent
within it.
2.7. Agency in language: “Phenomena of the third kind”
Leaving aside the more ideologically charged aspects of the evolution of

the language-organism-species metaphor, aspects which nonetheless un-
doubtedly contributed to its success and staying power, we need to look at
the advantages the organismic metaphor12 provided with regard to the no-
tion of agency in language. If language was a natural object, then change,
too, was natural and internal to it. Since natural phenomena exist inde-
pendently of human will, no further explanation was needed. On this view,
agency was internal to the organism itself, that is, to the system itself. The
heated exchanges that took place between Müller and Whitney in the latter
half of the 19th century reflect the two opposing views of agency in lan-
guage: the organismic, essentialist view and one that assigns agency only
or primarily to speakers. For the most part Müller defended the first posi-
tion while Whitney was an ardent defender of the second.
According to Whitney, the minute changes introduced (by conscious or
unconscious choices on the part of individual speakers) at one juncture in
time can be responsible for cumulative effects produced over longer
stretches of time (Alter 2005: 130–135): “It is indeed true that the individ-
ual has no power to change language. But it is not true in any sense which
excludes his agency, but only so far as that agency is confessed to be inop-
erative except as it is ratified by those about him” (Whitney 1867 [1901]:
12. Another factor shaping the development of the language-organism-species anal-

ogy is found in the reasons behind the apparent preference for using the term
“organism” (or German organismus) to describe the phenomenon of “language”
in the first place. The logical candidate – at least from today’s vantage point –
would have been the term “system”. However, it would seem that in the 19th
century the term “system” was somewhat out of favor, reflecting perhaps “the
latter’s residual but strong connotations of ‘grandiose overarching speculative
scheme’ […], with which it had been tinctured during the preceding 100–150
years, as the pendulum swung away from such schemes” and the construction of
grand and comprehensive theories, at least that appears to be the case with the
expression in French, l’esprit de système” (Janda and Joseph (2003: 11, ft. 9).
45). Here Whitney is subscribing to an incipient population-based view of

language agency, similar to that of Croft (2000) and earlier to the approach
found in the pioneering work of Labov (1996, 2001). This suggests what
are localized actions of individual language agents, in the aggregate, give
rise to larger global patterns of language.
The difficulty in locating the site of agency in language has often been
compared to the problem of agency associated with “the theory of the in-
visible hand”, while language itself has been categorized as “a phenome-
non of the third kind”, based on the fact that it looks like something that
was brought about by prior design, but was not (Keller 1990 [1994]: 61–
107). According to Keller, “phenomena of the third kind” can be perceived
and described on the micro-level as well as on a macro-level, while he
compares language itself to something much more highly complex than a
system of footpaths, yet similar in its constitution, an analogy that, as we
shall discover in the next section, resonates strongly with complex adaptive
systems thinking and the notion of circular causality (cf. also Mufwene
2003). Moreover, today the systems that Keller lists as belonging to this
class of “phenenoma of the third kind” are regularly modeled using a com-
plex adaptive systems framework.
3. A different metaphor: Languages as complex adaptive systems
3.1. Rise of the notion of “complex adaptive systems”
Examples of complex adaptive systems include social insect and ant colo-
nies, the biosphere and the ecosystem, the brain and the cell, the immune
system and financial markets, social networks, the Internet, and also, in
general, any human social group-based endeavor forming part of a cultural
and social system. Over the past decade the study of complex adaptive
systems, a subset of dynamic nonlinear systems, has become a major focus
of interdisciplinary research in the social and natural sciences (Lansing
2003: 183) and more recently in the field of “evolutionary linguistics” (cf.
Steels 1999, 2004, 2005).13 Broadly defined, a complex adaptive system
13. Perhaps the most well known initiative in this direction is that of Luc Steels and
his team of researchers working at the Free University of Brussels and the
research units of Sony CSL in Paris. Over the past decade, they have
investigated ways in which artificial agents can be used to self-organize
languages with natural-like properties and how meanings can co-evolve with
240 Roslyn M. Frank
(CAS) is one that is self-organizing in which there are multiple interactions

between many different components while the components themselves can
consist of networks that in turn operate as complex (sub)systems. CAS
thinking is concerned with understanding the global behavior arising from
local interactions among a large number of agents. Very often, this global
behavior or emergent dynamics is complex; it is neither specified by prior
design nor subject to centralized mechanisms of control, and, consequently,
it is often difficult or impossible to predict solely from knowledge of the
system’s constituent parts what the emergent global level properties of the
system will be. Complex systems are, therefore, systems in process that
constantly evolve and unfold over time. Change is an integral element of
their functioning. In the case of complex adaptive systems, they are adap-
tive in that they have an innate capacity to change and learn from experi-
ence, so to speak. In short, they are endowed with the ability to evolve and
adapt to a changing environment.
Since complex adaptive systems arise in a wide range of contexts, this
theoretical framework is rapidly gaining ground in a variety of disciplinary
areas, not only in the biological and physical sciences, but also the social
sciences and to some extent as a tool for the study of artificial and natural
language evolution. Of particular note is the close working relationship that
already exists between the field of complex adaptive systems thinking and
artificial life (A-Life), while applications of CAS and related developmen-
tal systems approaches to 21st century post-genomic and other types of
research problems in the biological sciences are becoming increasingly
common (Griffiths 2002; Griffiths and Gray 2000, 2004; Kay 2000: 326;
Lansing 2003; Strohman 1997). In all of these areas the principles of emer-
gence and self-organization are fundamental: complex global patterns with
new properties can emerge from local interactions. CAS thinking and the
related term complexity science are used to refer to the loosely organized
and highly interdisciplinary academic field that has grown up around the
study of such systems, even though the specific theoretical frameworks of
the disciplines, fields or subfields in question may differ significantly.
Although CAS oriented investigations often tend to be of a highly
quantitative nature, a less quantitatively oriented CAS modeling approach
language. Central to their research project is the hypothesis that language is a

complex adaptive system, one that emerges through adaptive interactions
between the artificial agents and that over time continues to evolve, as a self-
organizing system, adapting itself to the needs and capabilities of the agents.
can be adopted for investigating natural language and discourse metaphor

formation, as has occurred in the case of other disciplines. In fact, the lan-
guage-organism-species discourse metaphor formation we just outlined
may be viewed as a prototypical example of a socioculturally situated mul-
tiagent system, that is, as a complex adaptive system. Thus, rather than
functioning solely as a tool for understanding the dynamics of artificial
factual worlds and computer simulations of language evolution, as has
been the case in “evolutionary linguistics” (Steels 2004), a CAS approach
can also be appropriated to explore the evolution and sociocultural entail-
ments found in natural languages (Sharifian this volume, forth.) and associ-
ated discourse metaphor formations, especially highly entrenched scientifi-
cally oriented ones, those that tend to leave behind abundant traces in the
written record.14
3.2. Natural language applications of complex adaptive systems thinking
As Luc Steels has pointed out, for some time now “linguists have been
trying to pin down what kind of object a language is, but this has turned out
to be far from obvious” (1999: 143). Moreover, Steels goes on to ask
whether that question itself is properly formulated: whether we would not
be better off asking a different question. Following Steels, instead of ask-
ing what type of object language is, we should consider asking what type
of activity language is.15 Such a reformulation of the question leads us to a
CAS response to it and to an examination of the significant advantages that
accrue when a CAS approach is adopted and applied to our understanding
of “language” in general and to the study and description of discourse
metaphor formation(s) in particular.
14. In the case of discourse metaphor formations whose historical tracks are less
obvious and, hence, harder to follow in the written record and/or involve greater
time-depths, more circuitous approaches must be attempted in which the so-
ciocultural traces (and off-loaded material metaphors) left behind by the pas-
sage of the linguistic artifact often play a larger role in the reconstruction proc-
ess (cf. Frank 2005, in press, in prep.-b).
15. Nearly twenty years ago, Nerlich (1989) put forward a very similar suggestion
calling for the definition of “language” as “activity”. Cf. also Nerlich and
Clarke (1988).
242 Roslyn M. Frank
3.2.1. Linear vs. nonlinear
Let us begin our formal examination of the characteristics of a complex

adaptive system by defining it as a dynamic nonlinear system: 1) it is
constantly constructed and reconstructed by its users; 2) it is self-
organizing; and 3) it is characterized by multiple mechanisms of control,
that is, control is distributed throughout the system rather than residing in a
single centralized command and control center. Furthermore, in such a
system “global order” derives from local interactions. Hence, when we
apply the theoretical model of a complex adaptive system to natural
language and/or, less expansively, to discourse metaphor formations, the
overall system can be viewed from two perspectives at the same time:
– from the local level which allows for description and analysis of the
activity of the (individual) language agent and her cognitive architecture
(idiolect + the sociocultural situatedness of the agent herself, viewed as
embedded in and, hence, inseparable from the influence of an
environment that itself is subject to constant alteration);
– from the global level which allows for the description and analysis of
the global order while the latter, in turn, is the result of the combined
activities (utterances) of heterogeneously distributed agents over time.
Then, with respect to the meaning of the terms “linear” and “nonlinear”
they are not entirely clear cut since they mean different things in different
scientific settings, while at the same time most of us are far more familiar
with linear (Newtonian) systems than nonlinear ones. In mathematics,
linearity simply means that we can know the value of the whole by adding
up the sum of the parts. So, for example, if we know the value of the initial
condition of a system and these conditions don’t interact with one another,
we can predict the system’s future behavior. In contrast, a nonlinear system
is one in which “initial conditions interact so that outcome prediction is
difficult at best, even when a complete knowledge of the initial conditions
is possible” (Strohman 1997: 200, ff. 13).16 Depending on the number and
nature of the initial conditions and the intensity of interactions between
them over time, including the effects of positive and negative feedback
loops, the inherent complexity of the system, coupled with our frequent
16. For a more detailed discussion of complex adaptive systems thinking, cf. Lans-
ing 2003; Holland 1995; Kauffman 1993.
inability to determine accurately the initial conditions of the system itself,

can make predictions concerning the future behavior of the system
extremely difficult if not impossible.
Even in the case of our attempts to describe the evolution, expansion
and modification of a relatively simple discourse metaphor formation, e.g.,
such as the language-organism-species one, we are often obliged to make
assertions concerning the initial conditions of the system, albeit without
full assurance that such an assessment is either accurate or complete. Yet
when viewed retrospectively, the pathway(s) laid down by a given
discourse metaphor formation can often be charted with significant
accuracy and precision (e.g., Hellsten 2005; Maasen and Weingart 2000;
Moss 2004: 1–50; Nerlich and Hellsten 2004). Nonetheless, it is often
difficult to determine where to set the upper temporal boundary of a
discourse metaphor, a difficulty that is related to the general problem of
units and levels of analysis (or selection), as well as to the degree of
granularity that should be employed when describing a given discourse
metaphor formation (Musolff 2006, this volume).
At the macrostructural (global) level, “language” may be viewed as an
emergent phenomenon, the cumulative effect of the heterogeneous and
distributed behavior of socioculturally situated language agents.17 Or, more
concretely, the global or macrostructural level can be viewed as an emer-
gent phenomenon resulting from the utterances produced by these agents;
the utterances in turn being based on the individual agent’s “idiolect”.
Hence, by shifting our perspective, we can focus either on the microstruc-
tural, individual, or local level, the behavior of the language agents, or on
the macrostructural, collective, or global level, keeping in mind that global
properties flow from the aggregate behavior of individuals, although the
actions of the latter are not the sole source in bringing about changes in the
system.
In the first instance, we are looking at usage and variation, whereas at
the macrostructural level, global structure(s) can be viewed as having their
own internal support networks, rhizome-like interdependencies, some of
which are more tightly connected to other nodes in the overall network or
in a given subsystem within the overall network than others. Generally
speaking, we can argue that those networks which are the most interlinked,
17. For a visual rendition of this type of phenomena, cf. Sharifian’s diagrams con-
cerning his “distributed, emergent cultural model” (this volume, cf. figures 1, 2,
3).
244 Roslyn M. Frank
the most entrenched within a given system or set of subsystems, are the
ones most resistant and least likely to be subject to collapse. For example,
in the case of the discourse metaphor formation that we have analyzed the
higher connectivity of certain usages, e.g. nodes that linked to definitions
and redefinitions of “species”, “variety” and “race”, eventually brought
about a shift in the prototypical meaning of “race” itself, a shift that was
already taking place in the second half of the 19th century and one that be-
came more firmly entrenched, at least among certain groups, in the 20th
century. And, this occurred in spite of the valiant efforts on the part of
some writers, e.g. Prichard (Henze 2004), to fix and/or otherwise control
the direction of these semantic movements. In this process the absence of
any form of central control is self-evident. Briefly stated, language as well
as discourse metaphor formations are best understood as examples of com-
plex adaptive systems, constantly in the process of being constructed, de-
constructed and reconstructed by users, in which there are multiple
interactions between many different components.
3.2.2. The role of circular causality
In order to understand the functioning of the system overall, we need to

remember that the emergent phenomenon described above has a strong
causal impact on the behavior and learning of each individual language
agent. Consequently, there is a kind of “circular causality” operating: re-
ciprocal causality is an intrinsic aspect of the system. At the “local level”
the individual language agent’s behaviors (utterances) determine “lan-
guage”, that is, they act, cumulatively, to determine language understood at
the “global level”. At the same time, when viewed from the “local level”
the resulting emergent global level structure co-determines the range of
behaviors of the agents, that is, it acts to constrain and shape their interac-
tions at the “local level”.18
To paraphrase Steels (1999: 144–145), this top-down/down-up mutual-
ity of influence, i.e., circular causality, is established in several ways:
– The global level systemic structures of language are already in exis-

tence before the individual language agent comes into contact with
18. “In a way this is what Saussure tried to get at with the langue-parole distinction
– at least in his unpublished notes” (cf. Nerlich 2007 and also 1986).
them. These global level structures act as a strong constraint on the lin-
guistic behavior of individual language agents, while the language
agents acquire their “local level” understandings of this already existing
system as their idiolect, understandings that can be renewed, restruc-
tured over and over again during the course of the individual’s lifetime.
– At the same time the local level systemic structure of language con-
stantly acts to bring about emergent structure and change. While the
speaker – the language agent – has to abide by the structures provided
by the system at the risk of not being understood, there is always a de-
gree of flexibility to expand the existing system.
– Although the structures are to some extent in constant flux, in commu-
nicative practice, the speaker is capable of choosing to draw, con-
sciously or unconsciously, from among them, a selection informed by
the nature of her idiolect, her “microstructural (local) knowledge” of the
global level macrostructures.
Finally, we might note that bilingual and multilingual language agents can
draw on additional microstructural knowledge that, in turn, can act to set in
motion perturbations in the emergent global level structures. At the same
time, we can appreciate that the sociocultural situatedness of the language
agents themselves is in a constant state of flux: that changes taking place in
their environment feed into the global system, that is, through the local
level behaviors and interactions of the language agents with this ever
changing localized environment. In short, it is an open rather than closed
system, constantly subject to perturbation. Additionally, since global level
structure is informed, epigenetically, by the aggregation of socioculturally
situated utterances, the resulting global level structure should also be un-
derstood as socioculturally situated.
In summary, circular causality is a fundamental aspect in the function-
ing of language and the constitution of discourse metaphor formations and
is not unusual in other types of living systems which are themselves self-
organizing and complex in nature. Therefore, understanding the bottom-up
and top-down exchanges between local and global levels of a complex
system, as each provokes emergences and constraints upon the other, is not
only the “holy grail” of artificial life research, as Gessler (2003: 76) puts it,
but also a fundamental goal of research models designed for the study of
natural languages, evolutionary change and metaphor formation.
Until recently the CAS approach to language has been more widespread
as a model in the field of “evolutionary linguistics”, most particularly in
246 Roslyn M. Frank
computer simulations of the evolution of language, rather than as part of

the theoretical toolkit of cognitive linguists examining natural language(s).
So, the next logical step would seem to be its design application to the
problems facing “evolutionary cognitive linguistics” where explorations of
natural language processes would be the goal. Here the CAS model could
help reduce or even alleviate certain conceptual complexities that have
been brought about recently by more “gene- or meme-centric” and “spe-
cies-” based models of language (Chilton 2005; Croft 2002, 2006, forth.;
Deacon 2004; Mufwene 2001, 2005; Steels 2004). When applied to natural
languages the CAS approach assumes that (emergent) linguistic structures
– language change or equilibrium – depend on and are brought about by the
agent’s cognitive architecture, including the individual’s embodied and
situated “idiolect” on the one hand, and on the other, on the resilience and
complexity of the matrixes of linguistic structures (features) that are impli-
cated primarily at the global level, some being more robust and hence more
resistant to change than others; some being more isolated in terms of their
connections to nodes within other networks, and to trajectories of attrac-
tors, composing the overall system.
4. Shifting perspectives
When viewed through the lens of a CAS model, “language” is no longer

understood in terms of its ability to reproduce, but rather in terms of its
ability to self-produce: it is a self-organizing system. Language is under-
stood as a constantly evolving system that defies simplistic taxonomic,
essentialist categorization. Thus, language is best characterized as a multi-
agent complex adaptive system in which emergent phenomena results from
behaviors of embodied, socioculturally situated agents. Stated differently,
the highly entrenched metaphor equating “language” with “organism” is
expanded. Moreover, allowing the older discourse metaphor formation to
adopt a CAS perspective does not negate the possibility of continuing with
the heuristically productive identification of other updated and expanded
versions of the language-organism-species metaphor, e.g. the use of “spe-
cies” analogies (Mufwene 2001, 2005; Kristiansen this volume) as well as
Croft’s “Generalized Analysis of Selection” model with its “linguemes”
and “lineages” (Croft 2000; 2002, forth.). What the shift in theoretical
perspective does do is create a larger more expansive conceptual platform
for research without rejecting the advantages that accrue from the use of a
highly entrenched and familiar organic metaphor of language.
From my vantage point there are several other distinct advantages that
would accrue from the adoption of a CAS approach. First, for those of us
concerned with developing models for the investigation and analysis of
discourse metaphor formations, the CAS model provides conceptual tools
of significant agility and ready application to the task at hand, Most par-
ticularly, the theoretical model could contribute positively to (although it
certainly doesn’t solve) the debate over units and levels of selection by
separating global and local levels of a given discourse metaphor formation
(Griffiths and Gray 2004; Hull, Langman and Glenn 2001).19 For those
interested in pursuing ways to model discourse metaphor formations, com-
plex systems approaches and a good dose of lateral thinking could contrib-
ute significant theoretical and methodological resources along with a prac-
tical interdisciplinary base for heuristic inferences. In short, although CAS
approaches have had remarkable successes in the area of computer simula-
tion of language evolution, the usefulness of a complexity-based heuristic
has not been fully recognized, yet alone explored, by those concerned with
change in natural languages, and, most particularly, with charting the
pathways through which discourse metaphor formations come into being
and evolve over time.
Secondly, while the CAS model brings new conceptual tools into play it
also allows for continuity with another emerging avenue of research,
namely, the reformulation of memetics, a process that has been taking
place in the work of linguists such as Musolff (2006, this volume), Chilton
(2005), Croft (2002) as well as Mufwene (2001, 2005). These researchers,
as well as others in adjacent disciplines of cognitive science, have noted
that the concept of a “meme”20 needs to be fleshed out; that as a concept it
19. Although the focus of this chapter has been on discourse metaphor formation,
CAS approaches have applications in other areas where situated cultural con-
ceptualization and sociocultural situatedness are emphasized in the analysis of
linguistic data. As Kristiansen (this volume), Queller (this volume) and Yu (this
volume) have demonstrated, once sociocultural factors are taken fully into con-
sideration, the units and levels of selection can become greatly varied.
20. Following Dawkins (1989: 192), a “meme” refers to a unit of cultural
information that is transferable from one mind to another: a unit that leaps from
one mind to another. For Dawkins, examples of memes are “tunes, catch-
phrases, beliefs, clothes fashions, ways of making pots or of building arches”.
Furthermore, a memetics framework asserts that a meme “propagates itself as a
248 Roslyn M. Frank
is vague, lacking in specificity and operates with a misplaced sense of

agency (Deacon 2004; Gatherer 1998; Ingold 2004; Marsden 1999; Wil-
kins 1998, 1999).21 Moreover, the history of the concept of a meme (cf.
Frank forth.; Musolf this volume) dates back to metaphoric extensions of
what today we must view as overly simplistic definitions of the concepts of
gene and gene agency, extensions which are simply no longer viable in the
post-genomic era, but rather tied to earlier gene-centered discourses which
in turn were based on even earlier variations of preformationist gene con-
cepts, grouped under the rubric of “gene-determinism” (Moss 2002, 2004;
Hellsten 2005; Hilferty and Vilarroya this volume; Strohman 1997, 2001).
In other words, the notion of agency that was assigned to the gene in the
1960s and later transferred to Dawkins’ (selfish-gene) meme has been
called into question by advances in today’s systems biology and related
fields of complexity science. At the same time philosophers of science as
well as researchers in the field of biology and environmental science (cf.
Larson this volume) are increasingly attuned to the important role played
by extended metaphor formations in guiding research directions and ex-
perimental practices. (cf. Keller 2000 a, b; Moss 2002, 2004; Sidler
2006).22 In short, when Dawkins first came up with his notion of a meme,
based analogically and phonologically on a gene, biologists still thought
unit of cultural evolution and diffusion – analogous in many ways to the

behavior of the gene (the unit of genetic information). Often memes propagate
as more-or-less integrated cooperative sets or groups, referred to as memeplexes
or meme-complexes. […] Proponents of memes suggest that memes evolve via
natural selection – in a way very similar to Charles Darwin’s ideas concerning
biological evolution – on the premise that variation, mutation, competition, and
‘inheritance’ influence their replicative success” (cf. Wikipedia 2007 a).
21. “The core problem of this theory [of memetics], I think, is a kind of misplaced
agency, that gives the impression that both genes and memes – replicators – can
be understood without considering their embeddedness in a dynamic system
which imbues them with their function and informational content. This, then, is
not just a problem with memes, but a problem with the replicator concept in
general, inherited from Dawkins’ short-circuited description of information pro-
cesses in biology” (Deacon 2004: 20).
22. While discussion of the successive meanings assigned to the term “gene” from
its inception at the beginning of the 20th century to the present is outside the
scope of this paper, for examples of just how truly metaphorical, even hypo-
thetical, its meaning was from the beginning, cf. Keller (2000 a, b, 2002); Moss
(2004: 1–51); and for a brief and relatively non-technical overview of the
problem cf. Turney (2005).
that the “program” was in the genes, and then in the proteins “encoded” by
genes. This type of gene-determinism is now being challenged by a broader
context-bound model, the “new epigenetics” (cf. Moss 2004: 52) and
shaped by a complex adaptive system approach (Strohman 2001), as well
as by a far greater awareness of the complexity of gene-protein-
environment interaction (Hilferty and Vilarroya this volume; Nerlich and
Hellsten 2004; Temmerman this volume).
Consequently, even though memetics was based originally, as I have in-
dicated, on earlier and now, in many senses, outdated formulations of gene
agency, recent attempts by (cognitive) linguists to reformulate memetics
are both interesting and promising. Their research agendas should be seen
as a means of testing, albeit very tentatively, how the highly entrenched
and wildly popular term “meme” might be appropriated, expanded and
recast, in short, how it might be rhetorically redefined and appropriated as
part of a terminological toolkit that could be employed when discussing the
“discourse career of a metaphor” (Musolff 2006: 70),23 “discourse meta-
phor networks” (Zinken, Hellsten and Nerlich this volume), the relationship
between collective cognition and individual activity, and the distributed
nature of language (Bernárdez this volume), as well as “cultural conceptu-
alization” (Sharifian 2003, this volume, forth.) and, finally, from a slightly
different point of view, when discussing the somewhat more individualized
notion of “situated conceptualization” (Barsalou 2005; Barsalou et al
2003).
Over the past decade, population approaches to language have become
more common, inspired, in part, by the writings of Dawkins (1976, 1982)
with his memes, replicators and vehicles and, later, by Hull (1988), a well-
known philosopher of science, with his discussion and revision of Dawk-
ins’ theoretical framework, specifically, its application to the study of the
evolution of primarily scientific concepts, and more concretely, to the way
in which the scientific model of a given group of scientists is elaborated
and evolves over time. However, along with the diffusion of the concept of
memes has come significant criticism of the basic tenets of memetics, or
23. In this respect Musolff’s recent study is particularly valuable terms of its at-
tempt to convert memetics into a tool for the investigation of metaphor and con-
ceptual evolution, bringing into view the functioning of “conceptual clusters”,
roughly equivalent to the “nodes” and “networks” operating in the
(sub)dynamics of a discourse metaphor formation. Musolff (2006: 69–71) also
explores the methodological advantages of employing “a metaphor-meme’s
point of view”.
250 Roslyn M. Frank
lack of said, as well as a lack of consensus concerning what a meme actu-

ally is and what memetics really stands for as a field of research. Also, the
proliferation of studies in which “mind viruses” (cf. Dawkins 1991) or
other forms of “epidemiological” transmission are present, as conceptual
frames of analysis, has brought into question the effectiveness of the origi-
nal gene-meme-virus analogical construct (Deacon 2004; Gatherer 1998;
Wilkins 1998, 1999, 2002). For example, Chilton has explored the possible
applications of the meme and virus analogies to the analysis of metaphor
and in the process accepts, at least momentarily, the (misplaced) agency
attributed to the gene-meme-virus concept:
But here I would hypothesize that conceptual constructs become meme-like
and ‘infect’ the mind (under the right social conditions) when they have
complex blending potential that recruits fundamental knowledge domains
along with the core mechanisms of the metaphor. There is a further ingredi-
ent that seems to go along with textualized memes of this kind – the delivery
of some kind of credibility assurance and epistemic warrant. (Chilton 2005:
40)
Picking up on the implications of the virus analogy for metaphor studies
Chilton brings forth a new name for it, “ideational epidemiology”, and then
returns to the key question, that of agency:
[I]deational epidemiology will study the spread of ideas in the population.
Over time the distribution may change – may shrink or spread, so ideational
epidemiology will be interested in patterns of spread and retreat. […] Why
do some ideas or idea-clusters propagate more than others? (Chilton 2005:
17)
Chilton (2005: 41) concludes his discussion on a more circumspect note,
stating that: “if there is such a thing as meme propagation one of its main
modes of operation lies in the properties of metaphorical expressions […]”.
Or, to bring this statement more in line with the arguments laid out in this
chapter, we might take it a step further, and assume that the recognizably
vague referentiality of the term “meme” is better understood to mean
“metaphor” (in its broadest sense) or a situated “cultural conceptualiza-
tion” (cf. Musolff 2006: 69; Sharifian this volume, forth.). In other words,
conceptual constructs, such as metaphors and analogies, are constantly
shifting entities, that are constructed, deconstructed and reconstructed as
they expand and contract in terms of their spreading activation into new
domains or as they retreat from old ones. While these old subsystems can
act to build up network connectivity, emergent nodes with slightly new
(expanded or contracted) meanings can attach themselves and move into

place. And, as Chilton (2005) has suggested, under the right social condi-
tions the end result can be a highly entrenched and enduring discourse
metaphor formation.
5. Conclusion
In summary, we can state that once particular metaphors become part of the
very fabric of scientific discourses, i.e., once they become deeply embed-
ded metaphors that have taken up permanent residence in the back-
grounded knowledge-base of a community of speakers, a knowledge com-
munity or epistemic culture, then, as Bono (1990: 81) has alleged, the ca-
pacity of individuals, or even scientific communities to control them is, at
best, limited (cf. Frank 2003, 2005; Maasen and Weingart 1995). Rather
than subjecting themselves to unerring conscious design and authorial
control, such scientific metaphors adapt themselves to a larger ecology of
affirming or contesting social and cultural values, interests and ideologies:
the discourse metaphor formation emerges without a centralized command
and control center. Or as Kay has characterized this situation: “Some
[metaphors], like the information and code metaphors, are exceptionally
potent due to the richness of their symbolism, their synchronic and dia-
chronic linkages, and their scientific and cultural valences” (Kay 2000: 3).
The same can be said of the staying power of the 19th century language-
organism-species metaphor: its networks resonated with the cultural, social
and scientific concerns of the period, setting up subsystems, nodes and
clusters of concepts that interacted with each other in complex ways; while
some never achieved more than a fuzzy boundedness in terms of their fixed
or consensual meanings, others continue to live among us today.
As has been noted, the knowledge that the global system has of these
networks rarely coincides with the local level knowledge-base of individual
language agents, the conceptual system that speakers bring into play (Bar-
salou 2005; Frank 2003, 2005, in prep-a, -b.). Even the understandings that
different speakers have of a discourse metaphor formation can vary signifi-
cantly, depending on the disciplinary community to which they belong.
And because of the fact that cultural conceptualizations are heterogene-
ously distributed in any given population of speakers (Sharifian this vol-
ume), normally no one individual can be fully aware of past lives of the
construct, or able to accurately predict where it might go next. As a result,
252 Roslyn M. Frank
interactions taking place over time between the local and global levels of
the system, e.g., in a given discourse metaphor formation, can display a
complexity that may only be described as transcalculational, a mathemati-
cal term for mind-boggling (Strohman 1997: 197).
Finally, I would emphasize that the fine-grained research currently be-
ing carried out on discourse metaphors and other aspects of natural lan-
guage, often using corpora studies and revealing a keen sense of the im-
portant role played by sociocultural situatedness in cognition, is slowly
bringing into focus the myriad of pathways open to this type of meaning-
making and, in the process, preparing the ground for more CAS oriented
approaches to the study of language and metaphor formation (cf. Morgan
forth.; Sharifian this volume). Moreover, just as Strohman (1997), Kay
(2000) and Moss (2004) have repeatedly argued that a new more contextu-
alized philosophy of metaphor is needed to capture more effectively the
complexity of the gene-protein-environment interaction in the post-
genomic era, the same can be said for the need to revise and update the
language-organism-species metaphor of language. In short both fields are
undergoing a major shift in their metaphoric repertoire in which greater
emphasis is being placed on “nonlinear, adaptive properties of complex
dynamic systems, where visions of linear causality [are being] replaced by
analyses of networks interacting with the environment and operating across
[different] levels” (Kay 2000: 326).
Acknowlegements
I wish to express my thanks to René Dirven, Farzad Sharifian, Andreas

Musolff and especially Brigitte Nerlich for their helpful comments on the
earlier drafts of this chapter.
References
Alter, Stephen G.
1999 Darwinism and the Linguistic Image: Language, Race, and Natural
Theology in the Nineteenth Century. Baltimore and London: John
Hopkins University Press.
2005 William Dwight Whitney and the Science of Language. Baltimore
and London: The John Hopkins University Press.
Arbib, Michael A. and Mary B. Hesse

1986 The Construction of Reality. Cambridge: Cambridge University
Press.
Ascroft, Bill
2001 Language and race. Social Identities 7 (3): 311–328.
Augstein, Hannah Franziska (ed.)
1996 Race: The Origins of an Idea, 1760–1850. Bristol, England:
Thoemmes Press.
1998 Aspects of philology and racial theory in nineteenth-century Celti-
cism: The case of James Cowles Prichard. Journal of European
Studies 28 (4): 355–371.
Barsalou, Lawrence W.
2005 Situated conceptualization. In: Henri Cohen and Claire Lefebvre
(eds.), Handbook of Categorization in Cognitive Science, 619–650.
Amsterdam: Elsevier Ltd.
Barsalou, Lawrence W., Paula M. Niedenthal, Aron K. Barbey and Jennifer A.
Ruppert
Learning and Motivation, 43–92. San Diego: Academic Press.
Bernárdez, Enrique
culture
Bertalanffy, Ludwig von
1968 General Systems Theory: Foundations, Development, Applications.
New York: Georges Braziller. http://www.panarchy.org/vonbertalan
ffy/systems.1968.html
Bono, James J.
ence. In: Stuart Peterfreund (ed.), Literature and Science: Theory
and Practice, 59–89. Boston: Northwestern University Press.
Chilton, Paul
2005 Manipulations, memes and metaphors: The case of Mein Kampf. In:
Louis de Saussure and Peter Schultz (eds.), New Perspectives on
Manipulative and Ideological Discourse in Pragmatics and Dis-
course Analysis, 15–43. Amsterdam: John Benjamins.
Clark, Andy
1997 Being There: Putting Brain, Body and World Together Again. Cam-
bridge, Mass./London: The MIT Press.
Condit, Celeste Michelle
1999 The Meaning of the Gene. Madison, WI.: University of Wisconsin
Press.
254 Roslyn M. Frank
2001 Rhetorical formations of genetics in science and society. Rhetoric

Review 20: 12–17.
Croft, William
2000 Explaining Language Change: An Evolutionary Approach. Essex:
Pearson Education.
2002 The Darwinization of linguistics. Selection: Molecules, Genes, Me-
mes 3 (1): 75–91.
2006 The relevance of an evolutionary model to historical linguistics. In:
Ole Nedergård Thomsen (ed.), Competing Models of Linguistic
Change: Evolution and Beyond, 91–132. Amsterdam/Philadelphia:
John Benjamins.
forth. Evolutionary models and functional-typological theories of language
change. In: Ans van Kemenade and Bettelou Los (eds.), Handbook of
the History of English. Oxford: Blackwell.
Darwin, Charles
1859 On the Origin of Species by Means of Natural Selection, or the Pres-
ervation of Favoured Races in the Struggle for Life. London:
Murray. http://darwin-online.org.uk/content/frameset?itemID=F373&
viewtype=side&pageseq=1
1871 The Descent of Man, and Selection in Relation to Sex. London: John
Murray. http://darwin-online.org.uk/content/frameset?itemID=F937.1
&viewtype=text&pageseq=1
Dawkins, Richard
1976 The Selfish Gene. Oxford: Oxford University Press.
1982 The Extended Phenotype: The Gene as a Unit of Selection. Oxford
University Press. (Reprint 1989 edition’s title: The Extended Pheno-
type: The Long Reach of the Gene).
1991 Viruses of the mind. http://cscs.umich.edu/~crshalizi/Dawkins
/viruses-of-the-mind.html
Deacon, Terrence W.
2004 Memes as signs in the dynamic logic of semiosis: Beyond molecular
science and computation theory. In: Karl Erich Wolff, Heather D.
Pfeiffer and Harry S. Delugach (eds.), Conceptual Structures at
Work. 12th International Conference on Conceptual Structures, 17–
30. Berlin: Springer–Verlag.
Döring, Martin and Brigitte Nerlich
2005 Assessing the topology of semantic change: From Linguistic Fields to
Ecolinguistics. Logos and Language: Journal of General Linguistics
and Language Theory 6 (1): 55–68.
Eisler, Rudolf
1930 Organismus. In: Rudolf Eisler, Kant-Lexikon: Nachschlagewerk zu
Kants sämtlichen Schriften, Briefen und handschriftlichem Nachlass.
Berlin. Reprint Hildesheim: Georg Olms, 1969. Available online:

http://www.textlog.de/33087.html
Farber, Paul L.
1972 Buffon and the concept of species. Journal of the History of Biology
5 (2): 259–284.
Frank, Roslyn M.
2003 Shifting identities: The metaphorics of nature-culture dualism in
Western and Basque models of self. metaphorik.de 4/2003: 66–96.
http://www.metaphorik.de/04/frank.pdf
2005 Shifting identities: A comparative study of Basque and Western
cultural conceptualizations. Cahiers of the Association for French
Language Studies 11 (2): 1–54. http://www.afls.net/Cahiers/11.2/
Frank.pdf
in press Hunting the European sky bears: A diachronic analysis of Santa
Claus and his helpers. In: Paul G. Bahn (ed.), An Enquiring Mind:
Papers in Honour of Alexander Marshack. Boston/Leiden: Brill
Academic Publications.
in prep. a Linguistic relativity and the question of language-specific ontology:
An exploration of the ontological commitments of the Basque lan-
guage. http://www.uiowa.edu/~spanport/personal/Frank/FrankHm.
htm
in prep. b The complexity of sociocultural situatedness and material metaphors:
Badger paws and bear claws. http://www.uiowa.edu/~spanport/
personal/Frank/FrankHm.htm
forth. Shifting identities: Metaphors of discourse evolution. In: Andreas
Musolff and Jörg Zinken (eds.), Metaphor and Discourse. London:
Palgrave-MacMillan.
Gal, Susan and Judith T. Irvine
1995 The boundaries of languages and disciplines: How ideologies con-
struct difference: Defining the boundaries of social inquiry. Social
Research 62 (4): 967–1001.
Gatherer, Derek
1998 Why the thought contagion metaphor is retarding the progress of
memetics. Journal of Memetics – Evolutionary Models of Informa-
tion Transmission 2. http://cfpm.org/jomeit/1998/vol2/gatherer_d.
html
Gessler, Nicholas
2003 Evolving cultural things-that-think. Computational Synthesis: From
Basic Building Blocks to High Level Functionality. Papers from the
2003 AAAI Spring Symposium, Technical Report SS-03-02, 75–81.
Menlo Park, AAAI Press.
256 Roslyn M. Frank
Gobineau, Arthur
1967 Essai sur l’inegalité des races humaines. (An Essay on the Inequality
of the Human Races). Paris: Nizet. Original version 1853–1855.
Griffiths, Paul E.
2002 The philosophy of molecular and developmental biology. In: Peter K.
Machamer and Michael Silberstein (eds.), Blackwell Guide to Phi-
losophy of Science, 252–271. Malden: Blackwell Publishing.
Griffiths, Paul. E. and Russel D. Gray
2000 Darwinism and development systems. In: Susan Oyama, Paul E.
Griffiths and Russell D. Gray (eds.), Cycles of Contingency: Devel-
opment Systems and Evolution, 195–218. Cambridge, Mass.: MIT
Press.
2004 The developmental systems perspective: Organism-environment
systems as units of development. In: Katherine Preston and Massimo
Pigliucci (eds.), Phenotype Integration: Studying the Ecology and
Evolution of Complex Phenotypes, 409–431. Oxford and New York:
Oxford University Press.
Haeckel, Ernst
1868 Die Natürliche Schöpfungsgeschichte (The Natural History of Crea-
tion). Berlin: Reimer.
Hellsten, Iina
2005 From sequencing to annotating: Extending the metaphor of the book
of life from genetics to genomics. New Genetics and Society 24 (3):
283–297.
Henze, Brent
2004 Scientific definition in rhetorical formations: Race as “permanent
variety” in James Cowles Prichard’s Ethnology. Rhetoric Review 23
(4): 311–331.
Hilferty, Joseph and Óscar Vilarroya
this vol. In search of development
Holland, John H.
1995 Hidden Order: How Adaptation Builds Complexity. Addison–Wesley
Hull, David L.
1988 Science as a Process: An Evolutionary Account of the Social and
Conceptual Development of Science. Chicago and London: Univer-
sity of Chicago Press.
Hull, David L., Rodney E. Langman and Sigrid S. Glenn
2001 A general account of selection: Biology, immunology and behavior.
Behavioral and Brain Sciences 24 (3): 511–573.
Ingold, Tim
2004 Beyond genes and memes: A relational approach to the evolution of
language and culture. Plenary talk presented at the conference on
Evolutionary Epistemology, Language and Culture, Vrije Universiteit

Brussel, Brussels, Belgium, 04/05/26–28. Full talk audio [mp3: 6.3
Mb] http://www.comdig2.de/Conf/EELC04/
Janda, Richard D. and Brian D. Joseph
2003 On language, change and language change – or, of history, linguistics
and historical linguistics. In: Brian D. Joseph and Richard D. Janda
(eds.), Handbook of Historical Linguistics, 3–180. Oxford: Black-
well.
Kauffman, Stuart A.
1993 The Origins of Order. Oxford: Oxford University Press.
Kay, Lily E.
2000 Who Wrote the Book of Life: A History of the Genetic Code. Stan-
ford, CA.: Stanford University Press.
Keller, Evelyn Fox
2000 a The Century of the Gene. Cambridge, Mass.: Harvard University
Press.
2000 b Is there an organism in this text? In: Phillip R. Sloan (ed.), Control-
ling our Destinies: Historical, Philosophical, Ethnical and Theologi-
cal Perspectives on the Human Genome Project, 273–290. Notre
Dame, Ind.: University of Notre Dame.
2005 Ecosystems, organisms and machines. BioScience 55 (12):1069–
1074.
Keller, Rudi
1994 On Language Change: The Invisible Hand in Language. Translated
by Brigitte Nerlich. London and New York: Rutledge. Original Ger-
man version: Sprachwandel. Von der unsichtbaren Hand in der
Sprache. Tübingen: Francke, 1990.
Koerner, E. F. Konrad
1983 Linguistics and Evolutionary Theory: Three Essays by August
Schleicher, Ernst Haeckel and Wilhelm Bleek. Amsterdam/Philadel-
1995 The natural science impact on linguistic theory. In: E. F. Konrad
Koerner (ed.), Professing Linguistic Historiography, 47–76. Amster-
Kristiansen, Gitte
Labov, William
1994 Principles of Linguistic Change. Volume 1: Internal factors. Oxford
UK ; Cambridge USA: Blackwell Publishing.
2001 Principles of Linguistic Change. Volume 2: Social Factors. Oxford,
UK ; Cambridge USA: Blackwell Publishing.
258 Roslyn M. Frank
Lamarck, Jean Baptiste

1999 Zoological Philosopy. Translation by Ian Johnston of Philosophe
zoolologique. Paris: Dentu. Original version 1809. http://www.mala.
bc.ca/~johnstoi/ LAMARCK/lamarck1.htm
Lansing, J. Stephen
2003 Complex adaptive systems. Annual Review of Anthropology 32: 183–
204.
Larson, Brendon
this vol. Entangled biological, cultural and linguistic origins of the war on
invasive species
Livingstone, David N.
1992 The Preadamite Theory and the Marriage of Science and Religion.
Transactions of the American Philosophical Society 82 (3).
Philadelphia: American Philosophical Society.
Maasen, Sabine and Peter Weingart
1995 Metaphors – Messengers of meaning: A contribution to an evolution-
ary sociology of science. Science Communication 17: 9–31.
2000 Metaphors and the Dynamics of Knowledge. London/New York:
Routledge.
Marsden, Paul
1999 A strategy for memetics: Memes as strategies. Journal of Memetics –
Evolutionary Models of Information Transmission 3: 42–44.
http://cfpm.org/jom-emit/1999/vol3/marsden_p.html
Mayr, Ernst
1982 The Growth of Biological Thought: Diversity, Evolution and Inheri-
tance. Cambridge, MA: Harvard University Press.
1996 What is a species, and what is not? Philosophy of Science 63: 262–
277. http://members.aol.com/darwinpage/mayrspecies.htm
Morgan, Pamela S.
forth. Competition, cooperation and interconnection: ‘Metaphor families’
and social systems. In: Gitte Kristiansen and René Dirven (eds.),
Cognitive Sociolinguistics. Berlin: Mouton de Gruyter.
Morpurgo–Davies, Anna
1992 History of Linguistics. Vol. IV: Nineteenth-Century Linguistics. Lon-
don and New York: Longman.
Morris, William
1971 The American Heritage Dictionary of the English Language. Boston:
Houghton Mifflin Company.
Moss, Lenny
2002 From representational preformationism to the epigenesis of openness
to the world? Reflections on a new vision of the organism. Annals of
the New York Academy of Sciences 981: 219–229.
2004 What Genes Can't Do. Cambridge, Mass.; London: The MIT Press.
Mufwene, Salikoko S.
2001 The Ecology of Language Evolution. Cambridge: Cambridge Univer-
sity Press.
2003 Analogs anywhere: The flow of highway traffic and language evolu-
tion. In: Saikoko S. Mufwene and Sylvain Neuvel (eds.), Contempo-
rary Linguistics, 39–57. Chicago: Department of Linguistics, Univer-
sity of Chicago.
2005 Language evolution: The population genetics way. In: Günter Hauska
(ed.), Gene, Sprachen und ihre Evolution, 30–52. Regensburg: Uni-
versitätsverlag. http://humanities.uchicago.edu/faculty/mufwene/pub
lications/languageEvolution-populationGeneticsWay.pdf
Musolff, Andreas
2006 Metaphor scenarios in public discourse. Metaphor and Symbol 21
(1): 23–38.
Nerlich, Brigitte
1986 Saussurean linguistics and the problem of meaning: From dynamic
statics to static dynamics. Language and Communication 6 (4): 257–
276.
1989 Elements for an integral theory of language change. Journal of Liter-
ary Semantics 18 (3): 163–186.
1990 Change in Language: Whitney, Bréal and Wegener. London/New
York: Routledge.
2007 Personal Communication.
Nerlich, Brigitte and David D. Clarke
1988 A dynamic model of semantic change. Journal of Literary Semantics
17 (2): 73–90.
Nerlich, Brigitte and Iina Hellsten
2004 Genomics: Shifts in metaphorical landscape between 2000 and 2003.
New Genetics and Society 23 (3): 255 –268.
Nisbet, Robert A.
1969 Social Change and History: Aspects of the Western Theory of Devel-
opment. London: Oxford University Press.
Oxford English Dictionary
2000 Oxford, UK: Oxford University Press.
Queller, Kurt
this vol. Toward a socially situated, functionally embodied lexical semantics:
The case of (all) over
Poliakov, Léon
1974 The Aryan Myth: The History of Racist Ideas in Europe. New York:
St. Martin’s Press.
260 Roslyn M. Frank
Reveal, James L., Paul J. Bottino and Charles F. Delwiche

1999 The 1858 Darwin-Wallace Evolution Paper. http://www.life.umd.
edu/emeritus/reveal/pbio/darwin/darwindex.html
Richards, Robert J.
2002 a The linguistic creation of man: Charles Darwin, August Schleicher
and the missing link in 19th century evolutionary theory. In: Matthias
Dörries (ed.), Experimenting in Tongues: Studies in Science and
Language, 21–48. Stanford, CA: Stanford University Press.
2002 b Race. In: John Heilbron (ed.), Oxford Companion to the History of
Modern Science, 697–698.. Oxford: University of Oxford Press.
home.uchicago.edu/~rjr6/articles/Race--Oxford%20Companion.doc
Schleicher, August
1850 Die Sprachen Europas in systematischer Übersicht: Linguistische
Untersuchungen (A Systematic Overview of the Languages of
Europe: Linguistic Investigations). Bonn: H. B. König.
1860 Die Deutsche Sprache (The German Language). Stuttgart: Cotta.
Reprinted Niederwalluf: Sandig, 1974.
1863 Die Darwinsche Theorie und die Sprachwissenschaft (Darwinian
Theory and the Science of Language). Weimar: Böhlau.
1864 Die Darwinsche Theorie und die Thier- und Pflanzenzucht (Darwin-
ian Theory and Animal and Plant Breeding). Zeitschrift für deutsche
Landwirthe 15: 1–11.
1865 Über die Bedeutung der Sprache für die Naturgeschichte des Men-
schen (On the Significance of Language for the Natural History of
Mankind). Weimar: Böhlau.
1869 Darwinism Tested by the Science of Language. Trans. Alex Bikkers.
London: John Camden Hotten. (Reprint in: E. F. Konrad Koerner
(ed.), Linguistics and Evolutionary Theory: Three Essays by August
Schleicher, Ernst Haeckel and Wilhelm Bleek. Amsterdam/Philadel-
phia: John Benjamins, 1983.)
Sharifian, Farzad
2003 On cultural conceptualizations. Journal of Cognition and Culture 3:
187–207.
guage
forth. On collective cognition and language. In: Hanna Pishwa (ed.), Social
Cognition and Language. Berlin: Mouton de Gruyter.
Sidler, Michelle
2006 The rhetoric of cells: Understanding molecular biology in the twenty-
first century. Rhetoric Review 25 (1): 58–75.
Steels, Luc
1999 The puzzle of language evolution. Kognitionswissenschaft 8 (4):
143–150. http://www.csl.sony.fr/downloads/papers/1999/steels-kog
wis1999.pdf
2004 Analogies between genome and language evolution. In: Jordan Pol-
lack, Mark Bedau, Phil Husbands, Takashi Ikegami and Richard A.
Watson (eds.), Proceedings of Artificial Life IX: Proceedings of the
Ninth International Conference on the Simulation and Synthesis of
Living Systems, 200–206. Cambridge, Mass.: MIT Press.
2005 The emergence and evolution of linguistic structure: From lexical to
grammatical communication systems. Connection Science 17(3/4):
213–230. http://www.csl.sony.fr/downloads/papers/2005/steels-05h
.pdf
Stepan, Nancy
1982 The Idea of Race in Science: Great Britain 1800–1960. Hamden,
CT: Archon.
Strohman, Richard
1997 Epigenesis and complexity: The coming Kuhnian revolution in biol-
ogy. Nature Biotechnology 15: 194–200. http://bialystocker.net/files
/kuhn.pdf
2001 Human genome project in crisis: Where is the program of life?
http://www.biotech-info.net/StrohmanMarch09.pdf
Taub, Liba
1993 Evolutionary ideas and ‘empirical’ methods: The analogy between
language and species in the works by Lyell and Schleicher. British
Journal for the History of Science 26: 171–193.
Temmerman, Rita
this vol. Sociocultural situatedness of terminology in the life sciences: The
history of splicing
Turney, Jon
2005 The sociable gene. EMBO reports 6 (9): 808–810. http://www.cogs.
indiana.edu/pdf/embo.pdf
Whitney, William Dwight
1901 Language and the Study of Language: Twelve Lectures on the Prin-
ciples of Language Science. 6th ed. New York: Scribner and Co. First
edition 1867.
Wilkins, John S.
1998 What’s in a Meme? Reflections from the perspective of the history
and philosophy of evolutionary biology. Journal of Memetics –
Evolutionary Models of Information Transmission 2: 2–33.
1999 Memes ain’t (just) in the head. Journal of Memetics – Evolutionary
Models of Information Transmission 3: 48–55.
262 Roslyn M. Frank
2002 Darwinism as metaphor and analogy: Language as a selection proc-

ess. Selection: Molecules, Genes, Memes 3 (1): 57–74. http://www.
uq.edu.au/~uqjwilk2/papers/metaphor.htm
Wikipedia
2007 a Meme. http://en.wikipedia.org/wiki/Meme
2007 b Prognathism. http://en.wikipedia.org/wiki/Prognathism
Yu, Ning
this vol. The relationship between metaphor, body and culture
this vol. Discourse metaphors
Section C
Sociocultural situatedness in lexical and

usage-based approaches to metaphor
Toward a socially situated, functionally embodied
lexical semantics: The case of (all) over
Kurt Queller*
Abstract
Functional embodiment is “[t]he idea that certain concepts are not merely under-
stood intellectually: rather, they are used automatically, unconsciously and without
noticeable effort as part of normal functioning” (Lakoff 1987: 15). Linguistic re-
percussions include lexical entrenchment in functionally salient usage contexts of
numerous phrasal routines in which a word figures – a phenomenon here argued to
be crucial for lexical semantics. Analyzing a fragment of the English over network
within a usage-based framework, I show that similar usage constraints on a variety
of phrasal routines involving (all) over attest to entrenchment of a distinct “chaotic
dispersal” sense, not subsumable under “multiplex covering”. To account for such
innovation, I propose a non-teleological, socially and situationally embedded model
of semantic radial extension. First, situated speech comprehension yields gestalt
meanings for assemblies containing the relevant item, e.g., [{spill} {milk} all over
{the floor}]. Connotations of “chaotic dispersal”, compositionally licensed by verbs
like spill, become “distributed” (Sinha and Kuteva 1995) over the verb-preposition
collocation. Subsequently, considerations of functional embodiment trigger inde-
pendent association of the “dispersal” sense with the preposition. The model’s
implications are considered in the context of an evolving interdisciplinary under-
standing of the lexicon as a usage corpus, with lexical senses as emergent schemati-
zations over clusterings of usages.
Keywords: abduction, backformation, embodiment, lexical semantics, over, poly-

semy, radial extension, reanalysis, semantic change, usage-based model.
* Besides the editors (especially Roz Frank), I’d like to thank Elizabeth Traugott
and Bill Croft for reading the manuscript and offering useful insights and criti-
cism. Thanks also to René Dirven for encouraging me to include corpus analy-
sis, and to Beate Hampe and John Taylor for providing useful corpus data. Any
remaining infelicities are entirely my own.
266 Kurt Queller
1. Functional, situated and social embodiment
Lakoff (1987: 12–15) distinguished two sorts of embodiment: conceptual

and functional. The former asserts the impossibility of understanding
thought and language apart from their embedding in bodily experience.
Appreciation of conceptual embodiment has fostered understanding of how
linguistic conceptualization is grounded in non-propositional image sche-
mas that generalize over more concrete perceptual images, and how even
the most abstract language and thought are metaphorically rooted in bodily
experience. Functional embodiment is instead the idea that “certain con-
cepts are not merely understood intellectually: rather, they are used auto-
matically, unconsciously, and without noticeable effort as part of normal
functioning” (Lakoff 1987: 15, emphasis in original).
The linguistic implications of functional embodiment find articulation
in Langacker’s Usage Based Model (1987, 1988, 1999). One corollary is
that functionally salient conceptualizations tend to be associated with
highly routinized forms of linguistic expression. Syntagmatically complex
items, originally assembled from smaller constituents according to the gen-
eral patterns of the language, over time become entrenched in lexical con-
vention as unitary expressions (sometimes with variable slots). For routine
sorts of complex conceptualization, such entrenchment has obvious utility.
It reduces speakers’ on-line processing load, providing ready-made assem-
blies, immediately available “off the shelf”. It also streamlines comprehen-
sion, allowing direct matches between whole stretches of input and lexical
knowledge, obviating microprocessing of each minimal lexical item. At a
semantic level, moreover, it shortcuts the otherwise necessary contextual-
ized pragmatic inferencing from the minimal “compositional” meaning(s)
of an assembly to a richer conceptualization of meaning appropriate to the
particular speech event. This becomes possible because of direct incorpo-
ration into the unitized assembly’s semantic specification of elements that
are “extracompositional” in origin, originally contributed by gestalt-level
contextual inferencing (cf. Langacker 1999).
These notions are carried further in Zlatev’s (1997, 2003) model of
“situated embodiment”. Like Langacker’s model, Zlatev’s is grounded in a
Saussurean concept of linguistic symbols as pairings of phonological form
and semantic content. It treats the semantic pole, however, as involving not
conceptualizations, but actual “situation types” (and their component se-
mantic categories). Accordingly, linguistic communication is not the
transmission of conceptualizations from one head to another via the con-
Toward a socially situated, functionally embodied lexical semantics 267
duit of phonological form (cf. Reddy 1993), but rather the collaboration of
speakers and hearers in coordinating their attributions of contextual mean-
ing to situations, against a backdrop of shared sociocultural practices. As
basic unit within this negotiative process Zlatev proposes the utterance or
speech event, conceived as a “minimally differentiated language game”
(MDLG). The “game” designation incorporates the late-Wittgensteinian
understanding of language as involving interactive “forms of life”, situated
within and deriving meaning from complexes of conventionalized so-
ciocultural practices. Zlatev’s characterization of MDLGs as “minimal”
embodies the striking claim that the contextually situated utterance is not
merely “the smallest move in discourse”, but indeed “the smallest inde-
pendently meaningful unit of language” (2003: 454). The qualification
“differentiated” acknowledges that both utterance form and utterance
meaning, while preserving their holistic character, are analyzable into
smaller component elements. That the situated utterance nonetheless re-
mains the basic unit of semantic analysis is partly a function of the as-
sumption that within the utterance, mappings between semantic categories
and lexemes are typically many-to-many. A given lexeme frequently con-
flates more than one semantic category (cf. Talmy 1985); conversely, the
meaning of a single category may be distributed over more than one lex-
eme (cf. Sinha and Kuteva 1995). Both notions – situated utterance mean-
ing as a holistic yet componentially analyzable gestalt, and multiplicity or
“non-biuniqueness” of mapping relations between elements of form and
meaning – will figure crucially in the analysis that follows.
The present paper explores the significance of functional and situated
embodiment for lexical semantics, Arguing for a highly “granular” ap-
proach to word meaning (Sandra and Rice 1995), it proposes a model
whereby innovative lexical senses, often incompatible with the semantics
of their diachronic prototypes, emerge as schematizations over local clus-
ters of usages, absorbing from context semantic features unconnected with
compositional utterance meaning. Concretely, I examine that part of the
radially extended semantic network for English over once designated as
“multiplex coverage” (Brugman 1981; Lakoff 1987: 428–430). Building on
Queller (2001; cf. also Taylor 2002: 478–479, 2003 a: 40–41), section 2
argues that a wide range of routine phrasal usages containing all over in-
stantiate a “chaotic dispersal” meaning at odds with the semantics of cov-
ering (“multiplex” or otherwise). In support of this claim, I propose and
exemplify a functionally embodied methodology for lexical semantic
analysis that is grounded in routine phrasal lexical usage (collocations,
268 Kurt Queller
constructions, idioms). Section 3 offers a two-phase model for the emer-

gence of such innovative lexical meanings. Rather than positing direct
schema-to-schema mappings (image schema transformations, metaphorical
extensions), the model treats lexical semantic innovation as the product of
“semantic backformation” from extracompositional gestalt utterance
meanings. The approach is socially and situationally embodied, grounded
in a model of communication that treats hearers as creative co-participants
in the solution of concrete problems of communication.1 Guided by princi-
ples of functional and social embodiment, hearers arrive at linguistic con-
struals of usage events that are significantly distinct from those explicitly
coded by speakers’ lexical choices. While such construal differences in no
way impede speaker-hearer coordination around contextual meanings of
usage events, they lay the groundwork for subsequent lexical semantic
reanalysis. Section 4 considers the model’s place in an emerging interdis-
ciplinary understanding of the lexicon as a vast usage corpus and of lexical
senses as secondary schematizations over local patterns of usage.
2. MULTIPLEX COVERING or CHAOTIC DISPERSAL? A functionally

embodied approach to lexical meanings.
2.1. Previous approaches.
On the standard account, the usages in question elaborate the basic

COVERING schema for over. The classic Brugman / Lakoff presentation,
inclining toward a “maximalist” approach to schema specification, posited
two distinct subschemas:
1. a. There are flies all over the wall. MULTIPLEX COVERING

b. The spider crawled all over the wall. MULTIPLEX COVERING PATH
The first was assumed to derive from the basic image schema for
COVERING via a mass-multiplex transformation. In the new image schema,
the trajector consists not of a single, continuous entity (like a blanket) that
1. “Social embodiment” does not refer here to the body’s role in social cognition
(cf. Barsalou et al. 2003), but to the indispensability of taking communicative
interaction into account when doing (e.g.) lexical semantics. Some may prefer
to call this notion “social embeddedness” (cf. Chrisley and Ziemke 2002: 1103).
covers the landmark, occluding it from view, but of many individual enti-
ties (like flies).2 The landmark is conceived as containing “numerous small
regions which jointly cover its surface (or most of it)”, with the multiplex
TR distributed over the LM in such a way that “there is at least one trajector
in each region” (1987: 428; I shall refer to this notion as “sectoral cover-
age”). The MULTIPLEX PATH schema was in turn conceived as a minimal
variant on MULTIPLEX COVERING “in which the points representing the
multiplex entity of [the latter] are joined to form a path which ‘covers’ the
landmark”.
Seeking principled limits on the proliferation of polysemous schemas,
Kreitzer (1997) proposed constraining the notion of image schema trans-
formation so as to preclude (for example) distinct MULTIPLEX COVERING
schemas for over. The model prohibits derivation of new “relational”
schemas (image schemas in which distinct entities are related to one an-
other as trajectors and landmarks). Image schema transformations are in-
stead understood as construal operations involving only one or another
component of a full relational schema (e.g., the trajector). In the present
case, the transformation in effect applies in reverse (multiplex-mass). A
multiplex entity figuring as trajector (e.g., a collection of flies or a single
spider’s “multiplex” path) is “conceived (though not necessarily per-
ceived)” as a continuous surface, thus allowing the scene to be construed as
instantiating the existing COVERING schema for over.
Recently, the anti-maximalist reaction has taken a pragmatic turn. Fol-
lowing Fauconnier and Turner, Tyler and Evans (2003) emphasize that the
meanings of utterances are radically underspecified by the lexical expres-
sions that constitute them. The latter, they note, serve largely to prompt for
meaning construction in on-line speech processing. Much of the informa-
tion present in an utterance meaning, rather than being directly coded, is
thus contributed by contextual inferencing. Tyler and Evans (2003: 104–
106) accordingly seek to establish methodological principles for distinguish-
ing a minimal set of word senses that must necessarily be entrenched as
separate meanings of an expression from the much larger set that are ex-
plainable in terms of contextual inferencing. For prepositions with a basic
spatial meaning, they propose the following two criteria:
2. As usual in CL analyses, “landmark” (LM) refers to the entity with respect to

which one locates or specifies another entity, the “trajector” (TR).
270 Kurt Queller
2. a. The sense “must involve a meaning that that is not purely spatial
in nature and/or in which the spatial configuration between the
TR and LM is changed” vis-à-vis other senses, AND
b. “there must be instances of the sense that are context-
independent”, i.e., “instances in which the distinct sense could
not be inferred from another sense and the context” in which the
preposition occurs.
By these criteria, Tyler and Evans distinguish 14 distinct senses for over,
including one undifferentiated COVERING node. We can infer from their
discussion of the COVERING complex (2003: 132–133) how the criteria in
(2) may be taken to exclude independent entrenchment of “multiplex cov-
ering”. With respect to criterion (2a), one might assume changed spatial
configuration insofar as the TR, while still in a sense covering the LM, no
longer occludes it from view. Tyler and Evans however suggest that “the
occlusion interpretation is a contextual implicature of the covering sense
and real world knowledge of the properties of objects such as tablecloths
and blankets” that typically figure as “covering” trajectors (2003: 153,
footnote 30). Conversely, the non-occlusive covering of (1a and b) may be
seen as representing not a distinct lexical sense, but rather a contextual
implicature based on real-world knowledge of what “covering” of a surface
by less typical sorts of TRs like a swarm of flies or a spider’s path would
look like. Exclusion of distinct senses implies that all such “non-occlusive
coverage” cases are likewise inferentially derivable from context and real-
world knowledge (2b).
Numerous arguments thus suggest that the “multiplex covering” usages
straightforwardly elaborate the basic COVERING schema for over. The fol-
lowing section will nonetheless argue that usage evidence indeed requires
us to posit distinct schemas for the domain in question. The relevant sche-
mas, while diachronically derived from COVERING, no longer even elabo-
rate it synchronically, reflecting instead a distinctive CHAOTIC DISPERSAL
sense.
2.2. A bottom-up, usage-based approach.
The principle of functional embodiment predicts that frequent, functionally

salient complex conceptualizations will tend to find expression in unitized,
lexically entrenched linguistic assemblies. This suggests a methodology for
functionally embodied lexical semantic analysis. For a given lexical item,

one begins by identifying a reasonable sample of entrenched phrasal-
lexical patterns, establishing these heuristically as the formal or
“phonological” pole of a corresponding set of linguistic symbols. One then
examines characteristic contexts of use and constraints on usage to deter-
mine approximate values for the “semantic” poles of these syntagmatically
complex units. Working upwards from individual phrasal lexical units, one
then sees what more general patterns emerge at phonological and semantic
poles as one gradually schematizes away from the formal and semantic
details of the individual expressions. At any given level of schematization,
those aspects of form and meaning not shared by the instantiating units are
factored out, while aspects that are shared (however idiosyncratic) are
schematically retained.
It is well known that prepositions tend to contract special collocational
relationships with verbs. An obvious place to start, then, is to consider
what verbs typically collocate with (all) over. Examples are listed in (3):
(3) daub / dribble / drip / dump / pour / scatter / smear / spatter /

spill / splash / splatter / spread / sprinkle TR (all) over LM
Of course, none of these collocations demonstrates the existence of a dis-

crete “chaotic dispersal” sense for all over. That sense is clearly part of the
semantics of the verbs in question, and use of over in the “covering” sense
in the context of such collocations is presumably sufficient to prompt a
hearer to infer the appropriate sort of trajectory. More interesting are sta-
tive / resultative formulations using presentative or possessive construc-
tions, in which processes normally lexicalized by such “chaotic dispersal”
verbs, though evidently in the background, are not explicitly mentioned:
4. a. There are crumbs ( / *? tiles) all over the floor.

b. You’ve got chocolate ( / *? skin) all over your face.
c. This tablecloth has bloodstains ( / *? red and white squares) all
over it.
That something like “chaotic dispersal” is here signaled by all over be-
comes evident when one compares typically suitable trajectors (crumbs /
chocolate / bloodstains…) with less felicitous ones (tiles / skin / red
272 Kurt Queller
squares…) that are constitutive of the landmark or its design.3 On a “cov-

ering” interpretation, the latter ought to be ideal instantiations, but in fact
they are quite odd. The former (good) sorts of instantiation represent far
less ideal exemplars of covering; one indeed wonders whether covering is
even at issue. The problem is not merely that it takes only a relatively few
crumbs (for example) to warrant the predication that they are “all over the
floor”. More problematic is that there need not be even minimal “sectoral
coverage” of the landmark; crumbs that are construed as being all over the
floor may in fact be strewn over a very small portion of its total surface
(and quite likely elsewhere). The notion of “chaotic dispersal” captures the
fact that, in typical uses of all over, the distribution of the TR has little
regard for the boundaries of the LM surface, with respect to either reaching
them or remaining confined within them. It also captures the non-
topological, subjective sense of a loss of control that results in things end-
ing up where they don’t belong, creating “a mess”. (See Appendix 1 for
some relevant corpus analysis.)
One might want to preserve the COVERING sense here, arguing that
given our encyclopedic real-world knowledge, particular trajectors (like
crumbs) and particular landmarks (like floors) interactively “coerce” par-
ticular interpretations of the sort of covering involved (cf. Pustejovsky
1995). From a purely decoding perspective, and with respect to the fully
acceptable usages alone, the argument is appealing. The problem is with
the infelicitous usages. On such an account, usages that constitute good
exemplars of prototypical covering should, a fortiori, remain fully accept-
able. An adequate theory must account not only for acceptability of appar-
ently marginal cases, but for the problematic character of others that (on
the given account) should be unproblematic. This would seem possible
only on the assumption of a distinct CHAOTIC DISPERSAL schema for all
over.
The very same issue recurs at a more abstract level with respect to the
encoding idiom have {guilt} written all over {one’s} face.4 Unproblemati-
3. Exceptions prove the rule. For example, the “tiles” variant of (4a) becomes
felicitous if the tiles in question are scattered in a fairly random fashion across
the floor’s surface, rather than systematically (and constitutively) “covering” the
floor.
4. “Encoding idioms” (Makkai 1972) are entrenched assemblies that are semanti-
cally transparent from a decoding perspective, but whose status as conventional
routines cannot be predicted by speakers (or learners) apart from specific
phrasal-lexical knowledge. The conceptualization conventionally expressed in
cally acceptable instantiations of this item can certainly be construed as

involving “coverage” of the face by an emotional display; the problem is to
account for the dubious status of certain other usages that on a “covering”
interpretation should be equally acceptable. For example, usages like ??He
had rage / amazement / indifference written all over his face, though spo-
radically attested in large corpora, are perceived by many native speakers
as odd, compared with more prototypical instantiations involving words
like guilt. Otherwise comparable constructions in other languages seem
unencumbered by such nuances. With respect to amazement, for example,
it is perfectly normal in German to say Die Verwunderung stand ihm ins
Gesicht geschrieben, or in Italian Gli si leggeva in faccia lo stupore. Cor-
pus analysis confirms that, in contrast to the German and Italian construc-
tions, the English one is strongly preferred in contexts where an experi-
encer would like to conceal inner thoughts or feelings behind a façade of
indifference or composure, but cannot. (See Appendix 2.)
Such preferences are explainable in precisely the same way as are those
in (4) above. The best instances are those in which traces of the emotion
are construable as being “chaotically dispersed” across the face. Guilt is
characteristic for this expression precisely because it is the emotion that we
most typically attempt to contain behind a façade of nonchalance or impas-
sivity. When it nevertheless gets “chaotically dispersed” in such a way as
to make our true feelings visible to others, we perceive this subjectively as
caused by a loss of composure or control, and as resulting in a regrettably
messy situation (Queller 2001). Yet where does this “chaotic dispersal”
frame come from? The participle written in itself no more reflects the se-
mantics of chaotic dispersal than do the very general predicates in (4)
above. The only constituent that might plausibly contribute this nuance is
again the prepositional expression all over.
Going a step further, consider idioms for which “chaotic dispersal” is
the only reasonable interpretation of all over’s semantic contribution, with
a “covering” interpretation potentially yielding misconstrual:
English as George has guilt written all over his face can be expressed gram-
matically in other ways, e.g., it appears as if there is something for which
George feels guilty and which he’d like to conceal from people, if only he could
keep that chaotic dispersal of affect over his face from revealing his true feel-
ings. Only phrasal lexical knowledge allows a speaker to express the notion in a
way that native English speakers recognize as conventional. See Taylor (2002:
546–548) for useful discussion.
274 Kurt Queller
5. a. {This paper} is (just) all over the place.

b. {The data} are all over the map.
Though learners might take (5a) to imply systematic “coverage” of a wide

range of topics, it in fact refers to a piece of writing whose structure is
chaotic to the point of incoherence. Likewise, (5b) implies a chaotic distri-
bution that makes it hard to arrive at any clear conclusion, and attempts to
force a “covering” reading lead to confusion (see Queller 2001, appendix
2). Both expressions represent lexically entrenched instantiations of a
schema involving the evaluation of a summarily scanned path as chaoti-
cally dispersed.
All such expressions evincing a CHAOTIC DISPERSAL meaning in the ab-
sence of any chaotic dispersal verb suggest that, by the criteria in (2)
above, all over has developed a distinct sense not subsumable under
COVERING. Capturing the relevant generalizations in a descriptively ade-
quate fashion is perhaps the chief advantage of the present sort of function-
ally embodied analysis. However, it also has other merits. First, it reveals
much about the nature of idiomaticity. Expressions like those in (5) are
shown, despite their idiomatic status, to participate in networks of semantic
motivation linking them not only with one another, but also with less obvi-
ously idiomatic usages like those in (4). Second, it yields payoffs in ac-
counting for network relational structure. The standard Brugman / Lakoff
image schema for MULTIPLEX (COVERING) PATH, corresponding to utter-
ances like (1b) above, reflects what I call a CHAOTICALLY DISPERSED PATH,
even though the “connect-the-dots” transformation putatively deriving this
image schema from so-called MULTIPLEX COVERING proper might equally
well yield a path that “covers” the landmark systematically, from side to
side. Brugman’s original intuition of a prototypically chaotic path was in
fact correct, but nothing in the traditional account motivates this. On the
present account the chaotic trajectory, rather than being mysteriously
added when the path usage emerges, characterizes the entire CHAOTIC
DISPERSAL usage complex from the outset. (See Queller 2001: Figure 2 for
a visual representation.)
3. COVERING ⎯//→ CHAOTIC DISPERSAL: A socially situated,

functionally embodied approach to semantic radial extension.
To account for such semantic innovations, I propose a socially and situa-

tionally embodied discourse-based model in which speaker-hearer coop-
eration in resolving concrete communication problems entails inferencing
grounded in assumptions that hearers bring to the interpretation of usage
events. An indispensable first step is thus to consider the specific contribu-
tions made by hearers to the communicative enterprise.
3.1. Decoding vs. coordination: the hearer’s role in communication
Much psycholinguistic research suggests that language comprehension is a

constructive process in its own right, involving types of creative construal
and inferencing quite distinct from the processes involved in speaker cod-
ing (Bransford and McCarrell 1974; Straight 1986; Cutting 1998). Scholars
from Sperber and Wilson (1986) to Fauconnier and Turner (2003) moreo-
ver note that the radical underspecification of meaning in linguistic utter-
ances and the corresponding insufficiency of a purely decoding-oriented
approach to comprehension require increased attention to the contextual-
ized inferential reasoning that specifically characterizes the interpretation
process. Hearer construal is informed by assumptions about contextual
relevance and probable speaker intent that channel the interpreter’s con-
structive process of imputing meaning to an utterance. Interpreters rou-
tinely attribute to utterances contextually influenced extra-compositional
meanings that are substantially enriched vis à vis whatever bare-bones
compositional meaning might result from a strict reconstruction of the
speaker’s solution to the coding problem. Traugott and Dasher (2002) have
shown that hearer inferencing based on such contextually enriched inter-
pretation crucially shapes the nature and direction of lexical semantic
change.
A socially and situationally embodied alternative that avoids many of
the pitfalls of a pure decoding approach may be found in a model of lin-
guistic communication as joint action in the service of solving “coordina-
tion problems” (Lewis 1969: 5–8; Clark 1996: 62–65; Croft 2000: 95–
115). Given a range of options present in a given situation, a coordination
problem confronts participants with the task of deploying shared knowl-
edge and practical intelligence in such a way as to settle jointly on one
276 Kurt Queller
particular option, to the exclusion of others. In such a framework, linguistic

conventions serve as an inventory of devices available for selection and
arrangement by speakers, in specific situations and against the backdrop of
assumptions about world knowledge and common ground shared with
hearers, in order to prompt the latter to construct an interpretation consis-
tent with the speaker’s communicative intent. (One may note how well this
“coordination problem” approach to meaning and communication, recently
integrated by Croft into a compelling theory of language change, dovetails
with Zlatev’s situated embodiment model.)
Crucially, exploitation of linguistic convention in the service of jointly
solving a communicative coordination problem does not require hearers to
reconstruct a model of the usage event fully equivalent to that assumed by
the speaker. At a relatively trivial level, speaker and hearer may have dif-
ferent implicit understandings of the phonological structure of constituent
words. For example, a speaker may implicitly construe a nasalized vowel
as reflecting conditioning by an adjacent nasal consonant that was not
separately articulated, while a hearer reanalyzes the nasalization as an in-
trinsic feature of the vowel itself (Ohala 1989: 186, Croft 2000: 77). Such
differences need not impede coordination with regard to which lexical
items are being invoked. More dramatically, the entire morphosyntactic
structure of an utterance may be construed in radically different ways
without vitiating coordination around an essentially shared utterance
meaning. An utterance like I’m going to post this letter, for example, may
be construed by a speaker as containing a locomotion verb in present con-
tinuous form (am going) with a following subordinate infinitival purpose
complement (to post this letter), and by a hearer as involving instead the
main verb post… preceded by a grammaticalized pseudoauxiliary future
intent marker (am going to…). Even on the former construal, as Hopper
and Traugott (1993) point out, the very use of such an utterance in the
usual sorts of context warrants an inference of future intent. When hearers
innovatively impute to such an utterance a linguistic structure in which this
inference finds explicit morphological realization, they are thus by no
means misconstruing the speaker’s communicative intent. What results is
radical reanalysis (by one party) of the linguistic conventions assumed to
sanction the usage event, all in the service of successful speaker-hearer
coordination regarding its situated meaning (cf. Croft 2000).
Speakers and hearers may also achieve coordination regarding situated
meaning of a usage event while differing markedly in their implicit under-
standings of the meaning of a single lexical constituent. The associated
reanalysis involves a degree of radicality in speaker-hearer divergence

intermediate between the sorts of phonological and morphosyntactic
reanalysis just discussed. A familiar example is the Middle English
reanalysis of the word beads, originally meaning “prayers”, to mean “small
round objects perforated for threading on a string” (Jespersen 1922: 175;
Stern 1931: 326, 351; Langacker 1987: 383; McMahon 1994: 177). In the
case of utterances like Don’t bother her right now, she’s telling [counting]
her beads or This one here is your first Hail Mary bead, where the word
could at one time, in the context of praying the rosary, be taken indiffer-
ently as referring either to the prayer itself or to the associated object,
communicative success (i.e., coordination around a broadly shared under-
standing of the usage event’s contextual meaning) in no way depends on
speaker and hearer agreeing on one or the other lexical sense for bead.
Queller (2003) argues that such speaker-hearer discrepancies with regard to
situated understanding of lexical meaning may arise routinely in on-line
communication, and that the corresponding abductive reanalysis better
explains the sense shift than do alternative accounts based on direct sense-
to-sense mapping.
3.2. Emergence of new lexical meanings in discourse: A two-phase

model
Discussion of changes like the English beads shift in Queller (2003) was
restricted to the domain of metonymic extension, the argument being that
metonymic sense shift can occur without application of any metonymic
operation to the word in question. Abductive reanalysis of lexical meaning
is a two-phase process whose point of departure is not an atomistically
disembodied lexical meaning, but a richly contextualized, socially situated
utterance meaning. First, during on-line speech processing, hearers impute
to the relevant sort of usage event a gestalt utterance meaning in which the
conventional lexical sense is implicitly replaced by an extracomposition-
ally inferred sense that accords well with context and presumable speaker
intent. The clash between the hearers’ implicit lexical sense and the con-
ventional one is unproblematic precisely because it is merely implicitly
present within an inferred gestalt utterance meaning. It is only during a
second phase (perhaps offline) that the newly implicit lexical sense be-
comes the focus of linguistic attention. As utterance-level form-meaning
pairings like ((She’s telling her beads) / (SHE’S COUNTING HER BEADS –
278 Kurt Queller
MODERN SENSE)) get analyzed in terms of their implicit compositional lexi-

cal semantics, there emerges an innovative word-level form meaning pair-
ing ([beads] / (BEADS – MODERN SENSE)). I refer to this second phase as
“semantic backformation”, since it reflects innovation via compositional
reanalysis “backwards” from a syntagmatically complex expression, much
as in standard morphological backformation (cf. babysitterN [< babyN +
sitterN] Æ babysitV), but on the semantic rather than the formal side of the
form-meaning dyad. Its output is not a new word, but a new sense for what
happens to be an existing word.
This process involves no direct mapping from the conventional word-
level form-meaning pair; semantic innovation is crucially mediated by
contextualized interpretation of usage events. This distinguishes the pres-
ent model from most previous cognitivist accounts of semantic radial ex-
tension.5 To clarify the difference, I propose a new graphemic convention,
shown schematically in (6a), and specifically for the beads shift in (6b):
6. a. A —//→ B, or A —/ {Utterance type} /→ B

b. PRAYERS —//→ BEADS, or
PRAYERS —/ {She’s telling her beads} /→ BEADS
Speaker-hearer discontinuity is iconically represented by the double back-

slash that interrupts the arrow leading from the conventional to the innova-
tive lexical sense. Optional insertion of an utterance type (with curly
brackets indicating that it is one of several relevant types) suggests the sort
of usage on the basis of which the reanalysis process may be understood as
having operated. Such “link usage” types are characteristically construable
as instantiating with full sanction either the conventional or the innovative
word sense, without materially affecting the communicative import of the
usage event. The process accounts for emergence of new lexical meanings
not only in the domain of metonymic extension, but quite generally. Spe-
cifically, it permits a socially and functionally embodied account of the
emergence of the CHAOTIC DISPERSAL usage complex for over.6
5. One exception is Warren (1998). Cf. also Tyler and Evans (2003), whose em-
phasis on the role of what Traugott has called “pragmatic strengthening” of
lexical meanings in discourse reflects a socially and situationally embodied ap-
proach in some ways consistent with that proposed here.
6. The notion of “link usage” types proposed here is akin to that of “bridging con-
texts” as articulated by Enfield (2003) and Evans and Wilkins (2000). Thanks
3.2.1. Phase one: Link usage types and extracompositional gestalt

utterance meanings
The first step is to identify plausible link usage types. Taking a sentence
like (7a) as typical for standard COVERING and ones like (7c) as represent-
ing MULTIPLEX COVERING, Dewell (1994: 373) notes that (7b) represents a
“transitional” type:
7. a. She poured syrup over the pancakes. [cf. Dewell’s (28)]

b. She sprinkled water over the plants. [ = Dewell’s (57)]
c. She scattered seeds over the field. [ = Dewell’s (58)]
All such usages involve dispersal of a liquid or a particulate substance (the

TR) across a surface (the LM). Given our usual understanding of the type of
real-world situation involved (the viscosity of syrup, the properties of pan-
cakes, and the function of one with respect to the other), the action de-
scribed in (7a) will normally result in continuous coverage of the LM.
Given the discrete, particulate nature of seeds and the size of a typical
field, default readings of (7c) will instead involve multiplex coverage.
Lexical knowledge of the nature of “sprinkling” and real-world knowledge
of how water beads up on leaves yield for (7b) an image intermediate be-
tween (7a) and (7c) with respect to LM coverage.
Again, none of this alone shows emergence of a new sense for over.
Dewell’s treatment suggests that a unitary COVERING sense may be re-
tained for over in all the above cases, with inferences about the particular
nature of the “covering” prompted both by linguistic context (the nuances
of the particular “dispersal” verb) and by encyclopedic knowledge (cf.
Tyler and Evans 2003).7 But the point here is not to justify a new sense;
that has already been done in section 2.2, above. The point is to suggest,
given the demonstrable emergence of a distinct CHAOTIC DISPERSAL sense,
what sorts of conventionally sanctioned usage events may have yielded
material for the corresponding abductive semantic reanalysis.
to Elizabeth Traugott (p.c.) for bringing this to my attention. Similarities and

differences between the two concepts are discussed in Queller (in prep).
7. Dewell’s reference to a construal operation whereby summary scanning of the
TR serves to link its dispersed parts into a “virtual mass” also anticipates Kreit-
zer’s (1997) argumentation (2.1 above.).
280 Kurt Queller
To that end, consider potential link usages like that in (7b).8 Such utter-
ances are construable in two ways, depending on thematic focus. Focusing
on the landmark as theme, the utterance can be taken as answering the im-
plicit question “what happened to the plants?” The evident answer is that
they ended up covered with water. If one focuses on the trajector, the
question instead becomes “What happened to the water?” The answer (in-
ferred from the lexical meaning of sprinkle and from our real-world knowl-
edge of how water behaves) is that it got dispersed across the surface of the
plants. The first construal, consistent with a conventional COVERING sense
of over, would account for the original appearance of that word in such
contexts. The second, consistent with an innovative (CHAOTIC) DISPERSAL
sense, would account for the emergence of such a sense for over.
Initially, to be sure, the DISPERSAL sense is not explicitly attached to the
lexical item over. It is an aspect of the gestalt utterance meaning attributed
to the usage event as a whole, and is compositionally motivated by the
presence of a “dispersal” verb like sprinkle. But as Zlatev (2003: 454–459)
observes, mappings between lexical items in an utterance and conceptually
prominent aspects of a corresponding situation are often not strictly one-to-
one. Among the alternative possibilities is that of construing two or more
(possibly discontinuous) lexical items within the utterance as exponents of
a single conceptual/semantic component of the situation, a relation that
Sinha and Kuteva (1995) call “distributed” meaning. In the present case,
attribution of thematic prominence to the trajector favors construal of the
preposition over as signifying dispersal (rather than coverage) with respect
to the landmark. The result is an interpretation in which the single semantic
notion of dispersal gets distributed over two lexical elements within the
utterance: verb and associated preposition. However, a completed process
of lexical semantic reanalysis, through which the preposition becomes
capable of expressing this sense independently of the verb, depends on a
second phase: that of semantic backformation.
8. All three usage types in (7) can in principle serve as input to the abductive
reanalysis here schematized as COVERING —//→ (CHAOTIC) DISPERSAL.
3.2.2. Phase two: From gestalt utterance meaning to innovative lexical

sense
For an utterance like (7b), implicit construal of the semantic notion of

“dispersal” as distributed (in the sense of Sinha and Kuteva 1995) over
both the verb and its associated preposition may be represented approxi-
mately as in (8b); (8a) represents a more conventional construal, based on
compositional analysis of the original semantic contribution of over.
8. a. ((She sprinkled {the water} over {the plants}) / (SHE SPRINKLED

{THE WATER} IN SUCH A WAY THAT IT ENDED UP COVERING THE
SURFACE OF {THE PLANTS}))
b. ((She sprinkled {the water} over {the plants}) / (SHE SPRINKLED
{THE WATER} IN SUCH A WAY THAT IT ENDED UP DISPERSED
ACROSS THE SURFACE OF {THE PLANTS}))
In (b), the notion reflected by the gloss “…IN SUCH A WAY THAT IT ENDS
UP DISPERSED ACROSS THE SURFACE OF…” fits the same slot in a repre-
sentation of the larger assembly’s meaning as is filled in (8a) by the gloss
“…IN SUCH A WAY THAT IT ENDS UP COVERING…”. Just as the latter, on a
compositional reading, articulates the semantics of over, so likewise (8b) is
susceptible to an analysis in which the dispersal trajectory is taken to rep-
resent the preposition’s particular semantic contribution.
Such an analysis yields backformation of a new lexical form-meaning
pair from a larger, syntagmatically complex form-meaning pair. Expression
of a dispersal trajectory at this point becomes part of the preposition’s in-
herent usage potential, apart from the associated verb. Though that poten-
tial may reside primarily in memory traces of actual usages (what Croft
2000: 99 calls “a lineage of rich, context-specific meanings for which the
expression has been used”), schematization over such wholly or partly
remembered usages may yield a more abstract representation, roughly for-
mulated in (9):
9. [[…(all) over LM] / […(CHAOTICALLY) DISPERSED ACROSS THE

9
SURFACE OF THE LM]]
9. One might treat the CHAOTIC portion of (CHAOTIC) DISPERSAL as composition-

ally contributed by the all of (all) over. Support might be adduced from other
collocations where a prefixed all… at least reinforces a “chaotic” sense – e.g.,
282 Kurt Queller
It is through such a process, I suggest, that usages like (4) and (5) above,
which lack any chaotic dispersal verb and are in some cases actually incon-
sistent with a “covering” interpretation, become fully sanctioned within the
linguistic system.10
3.3. The role of functional embodiment in lexical semantic reanalysis
No obvious social or communicative pressure motivates the second, crucial

step in abductive lexical semantic reanalysis – that of semantic backforma-
tion from gestalt utterance meanings to innovative lexical meanings.
Nonetheless, a powerful dynamic may impel language users to match up
bits of semantically extracompositional material contained in their gestalt
utterance meanings with particular formal constituents of the correspond-
ing utterances. In this section, I endeavor to explicate the nature of this
dynamic, suggesting that what is involved is again an aspect of functional
embodiment.
Consider Tyler and Evans’ argument against positing the sort of distinct
“above-across” sense for over that many analysts (myself included) would
claim is instantiated in sentences like The cat jumped over the wall (2003:
118).11 They suggest that assumption of a dynamic “above-across” sense,
distinct from the static “above” sense, is based on a “logical fallacy” that
has essentially the following structure. (I have however minimally altered
all screwed up / all bent out of shape. (Compare, however, instances in which
all… reinforces notions of orderliness: all sorted out / all squared away). Provi-
sionally, I suggest that all reinforces the prototypically chaotic nature of disper-
sal events, but with local variations partly depending on idiomatic entrench-
ment. See Taylor (2003 a: 40–41) for further discussion.
10. The proposed process, whereby semantic elements originally intrinsic to a par-
ticular lexical item get “swapped out” in favor of elements originally inhering in
the usage context, reflects what Croft (2000: 130–134) calls metanalysis. I
would in fact suggest that the present model lays the basis for a theory of how
and why metanalysis occurs.
11. Four of the over senses that Tyler and Evans accept (by the criteria in (2)
above) as separately entrenched in “semantic memory” evidently reflect exten-
sions from such a dynamic spatial “above-across” schema. It is remarkable that
they nonetheless rule out entrenchment of the latter. For discussion, see Iwata
(2004: 289–292), and the chapter in Queller (in prep) on the OBSTACLE
SURMOUNTING usage complex for over.
the terms of the syllogism so that they refer instead to our link usage type
(7b), in a stative / resultative / existential formulation such as There is
{water} all over the {plants}.)
(10) a. a spatial scene is conceptualized in which {water} is dispersed

across the surface of {some plants};
b. there is nothing in the sentence, other than (all) over, which indi-
cates the trajectory followed by the {water};
c. therefore, (all) over must prompt for a trajectory involving dis-
persal across the surface of a landmark.
Tyler and Evans suggest that such reasoning wrongly assumes “that the
lack of formal expression coding trajectory information implicates a lack of
trajectory information per se. On this view, all elements that are salient in
the interpretation of a scene are encoded linguistically” (2003: 118; em-
phasis added). The trajectory information, as they point out, may be de-
rived contextually and constructed on-line; it need not be encoded in any
particular formal element of the utterance itself. This is all quite true. Nev-
ertheless, the “logical fallacy” in question has a natural appeal. It is indeed
the fallacy that lies at the root of abductive lexical semantic reanalysis.12
The heuristic can be formulated roughly as in (11):
(11) THE COMPOSITIONALITY ASSUMPTION: Given an utterance

meaning for a routine sort of usage event that corresponds well
with context and ostensible speaker intent, assume that each ele-
ment salient within that meaning is linguistically encoded by
some formal (lexical. phrasal, constructional) constituent of the
utterance.
12. Abductive reasoning, in a strictly logical sense, is by nature fallacious. Consider

the abductive version of the standard syllogism regarding the mortality of Soc-
rates, in which one starts from an observed result (the fact that Socrates has
died) and invokes an apparently relevant general principle (that all men are
mortal) to infer what may well therefore be the case (that Socrates was a man).
Nothing excludes the possibility that the principle invoked is inapplicable to the
case at hand (Socrates may have been a horse). But as Peirce pointed out, the
strictly logical weakness of abductive reasoning in no way diminishes its sig-
nificance for the emergence of cognitive innovation. (See Hopper and Traugott
1993.)
284 Kurt Queller
Such an assumption will tend to favor semantic backformation to new lexi-

cal senses especially in cases like (8b) above, where parallelism suggests
pairings between particular formal constituents of the utterance and par-
ticular elements of its contextually derived, inferentially constructed, ex-
tracompositional meaning.
The compositionality assumption is best understood as a natural mani-
festation of functional embodiment – the expectation of a correspondence
between routinely encountered experiences and ways of conceptualizing
them, on the one hand, and routine forms of expression, on the other. Se-
mantic backformation is a consequence of this expectation. Consider the
present case. Situations in which a liquid or particulate substance escapes
from a container and/or from one’s subjective control, dispersing chaoti-
cally across the surface of a landmark, are a recurring and salient aspect of
our experience. It is thus no accident that a whole series of English verbs,
including those in (3) above, should lexicalize the notion of chaotic disper-
sal. Absent any preposition specifically encoding a dispersal trajectory,
speakers can of course prompt for the desired sort of construal by selecting
a preposition that prompts for a “covering” construal. But once hearers,
routinely interpreting the relevant sort of speech event, have gotten used to
constructing gestalt utterance meanings in which the trajectory is appropri-
ately understood as involving not coverage but dispersal, they will natu-
rally tend to reanalyze the preposition as directly encoding a dispersal tra-
jectory.13
If the relevant situations are consistently associated with non-
topological nuances, moreover, these become part and parcel of the new
semantic usage potential. The “chaotic” aspect of the “chaotic dispersal”
designation for all over, for example, reflects not just the physical nature of
the trajectory, but also the subjective sense (prior to the dispersal event) of
a loss of control over the trajector and (following and resulting from it) of a
“mess” having been created. Importation of such non-topological, subjec-
tive nuances from the functionally salient internalized cognitive model for
dispersal events into the lexical semantics of all over (or, equivalently, of
13. Typological considerations (à la Talmy 1985) may be relevant. The typical

Germanic lexicalization pattern, with manner nuances incorporated into the se-
mantics of the verb and much of the work of specifying trajectory assigned to
prepositions, may tend to favor construals of collocations like scatter / spatter /
sprinkle / dribble… TR all over LM that treat the chaotically dispersed trajec-
tory nuance as uniquely encoded by the preposition.
constructions like have TR … all over {one’s} LM ) helps account for ac-
ceptability differences like those in (4) above, as well as for the preference
of an expression like have {guilt} written all over {one’s} face for situa-
tions involving loss of composure and resulting in an undesired display of
affect that was meant to be kept under wraps.
4. Conclusion
The proposed two-stage model of lexical semantic innovation may seem

cumbersome compared with one involving direct mapping from one se-
mantic schema to another via image schema transformations and/or figura-
tive extensions. Answering this objection requires deeper exploration into
the relations of cognitivist lexical semantics with other linguistic disci-
plines, including historical linguistics, lexicography, natural language
processing (NLP) and language acquisition.
Consider first the issue of goal-directedness in language change. The
“mapping” model implies that new senses arise as speakers creatively
stretch existing lexical resources to meet new expressive needs. Much dia-
chronic work suggests however that innovation largely results from re-
analyses of utterances originally formulated in purely conventional terms.
In the grammaticalization case cited in 3.1, it is unlikely that speakers in-
tentionally extended the meaning of {am} going to… in order to convey
future intent. Future intent just happens to be a functionally salient aspect
of the extracompositional gestalt utterance meanings that hearers naturally
attribute to speech events involving continuous-aspect locomotion verbs
followed by infinitival purpose clauses; the new pseudo-auxiliary usage
emerges as this extracompositional sense becomes explicitly aligned via
reanalysis with {am} going to…. Although the change results from the
goal-directed behavior of language users, it emerges not as intended out-
come, but rather as unintended by-product (cf. Keller 1994). The “hidden
hand” behind the change involves no teleology; it involves speakers and
hearers simply doing what they ordinarily do while routinely pursuing
other goals.
Just as in grammaticalization, each phase in the present model is inde-
pendently motivated in terms of ordinary language processing and use. The
first phase – attribution to speech events of gestalt utterance meanings
richer than the compositional sums of their parts – is motivated by the se-
mantic underspecification inherent in utterances, and by the resulting need
286 Kurt Queller
for inferential meaning construction in the process of speech comprehen-

sion. The second phase – backformation of new lexical senses from such
gestalt utterance meanings – is motivated by the compositionality assump-
tion (11), itself underwritten, as we have seen, by the principle of func-
tional embodiment. The present two-stage model is thus in a sense simpler
than the traditional direct mapping approach. While the latter seeks some
specific teleology behind each innovation, the former posits nothing be-
yond conventional usage on speakers’ part and routine inferencing on hear-
ers’ part.
Consider next the domains of lexicography, NLP and language acquisi-
tion. Until recently, one tended to assume that words had denumerable sets
of senses, and that the lexicon’s primary task was to specify these accu-
rately. More comprehensive dictionaries, especially if intended for non-
native speakers, might add illustrative phrasal usage examples, but these
were secondary. As computerized corpora become more central to the lexi-
cographic enterprise, lemmas typically devote proportionally less space to
defining senses and more to instantiating usage patterns. Senses are in-
creasingly construable as emergent generalizations over usage data. This is
consistent with a tendency among NLP specialists to eschew traditional
top-down processing models involving sense listing and disambiguation in
favor of bottom-up models involving “clustering” of corpus data into dis-
tinct “contextualization patterns” (Schütze 2000; Kilgarriff 2003; Taylor
2002: 472–474, 2003 b). Rather than try to pair input with one of several
listed abstract senses, such approaches directly match it against such em-
pirically established usage clusters.
The model is likewise consistent with a usage-based approach to acqui-
sition that no longer takes the primary targets to be word senses (let alone
mappings among senses), but rather concrete, lexically entrenched colloca-
tional and constructional usage patterns (Tomasello 2000; Nerlich, Todd
and Clarke, 2003). For innumerable encoding idioms like {This paper} is
all over the place and {The data} are all over the map, it is insufficient to
assign the item one or another lexical sense (Taylor 2003 b: 63). These are
idiosyncratic routines for talking about poorly organized papers and hard-
to-interpret data, respectively, and even correct matching of all over with
the CHAOTIC DISPERSAL usage cluster cannot account for the fact that
these expressions respectively require the complements place and map (as
opposed e.g. to field or chart). Schematization over such usage clusters
doubtless yields something resembling traditional lexical senses; what
seems increasingly doubtful is the notion that it is such senses, rather than
the highly specified situational usage patterns for which they are sche-
matic, that essentially constitute the mental lexicon. To put the matter con-
cretely: one can become a competent speaker of English without ever real-
izing that the above-mentioned encoding idioms and others jointly
instantiate a broader CHAOTIC DISPERSAL schema for (all) over (let alone
that they have a less direct relationship with COVERING usages). But the
converse is not true; familiarity with higher-level schemas cannot alone
assure active control of the particular usages, which must in any case be
individually learned.
The upshot may be that the lexicon is essentially a corpus, with lexical
senses reflecting secondary, higher-order schematizations over usage clus-
ters. This would explain why native speakers generally find it easier to
provide examples of how a word is used than to specify its meaning(s). It
would also explain why polysemy – the existence of multiple senses for a
given expression – is more problematic for certain traditional NLP ap-
proaches than for ordinary language users (Taylor 2003 a, 2003 b). Ordi-
nary speech comprehension does not normally involve computing and se-
lecting among all the possible compositional meanings that would result
from the various senses of an utterance’s minimal lexical constituents.
Entrenchment of collocational and constructional patterns entails direct
lexicalization of those aspects of their routine situated use that are most
functionally salient. Issues of lexical sense disambiguation rarely arise,
since competent users directly access such entrenched usage patterns and
their associated meanings.
None of this is shockingly new. The present article’s contribution is
simply to propose a route for the emergence of new lexical senses that is
not only consistent with the emphasis in diachronic linguistics on contex-
tual inferencing and reanalysis, but is also motivated in terms of processes
likely to occur independently in the course of communicative problem
solving and language processing. The emerging picture may appear to
threaten cognitivist advances in the modeling of polysemy – advances
grounded in the notion of networks of cognitively linked senses radiating
outward from conceptually embodied prototypes. I nonetheless hope to
have shown that network models that invoke direct schema-to-schema
mapping seriously overestimate the influence of prototype schema seman-
tics on the semantics of extension schemas, while underestimating the role
of situated, usage-based inferencing in lexical semantic innovation. The
road ahead, I would suggest, will involve carefully rearticulating the impli-
288 Kurt Queller
cations of conceptual embodiment for cognitive lexical semantics, while

elaborating a functionally, socially and situationally embodied model that
is even more fully consistent with insights into the nature of lexical mean-
ing that are emerging from cognate linguistic disciplines.
Appendix 1
Corpus analysis for “chaotic dispersal” all over
I have claimed that a large number of spatial all over usages involve situa-
tions that are better characterized in terms of “chaotic dispersal” than in
terms of the standard “sectoral coverage” account. The present appendix
provides empirical corpus evidence for this claim.
A sample of 100 instances of all over (spoken and written, predomi-
nantly the former) was taken from the British National Corpus sampler
edition, accessed via ICAME. Using Filemaker Pro, items were sorted into
the following categories: Chaotic Dispersal (CD), Sectoral Coverage (SC),
Ambiguous, Other Constructions, and Unclear. Criteria for distinguishing
CD from SC instances included the particular predicate, trajector and
landmark, as well as contextual cues. Ambiguous instances were those that
by these criteria might be construed with roughly equal plausibility as be-
longing to either category. Unclear instances (9 in all) were those that were
simply not interpretable with sufficient clarity to allow categorization.
Other Constructions (6 instances in all) included 4 Iteratives ({start} all
over (again)), and 1 each of Completive (“{the game’s} (all) over”) and of
the “that’s {him} all over!” construction. For present purposes, the Unclear
and OC categories were eliminated, leaving a total of 85 clearly spatial
uses. For those, the breakdown was as follows:
Chaotic Dispersal: 37 instances (43%)

Sectoral Coverage: 27 instances (32%)
Ambiguous: 21 instances (25%)
(Total): 85 instances (100%)
Following is a breakdown of the 37 CD instances by predicate category,

with trajector (TR) and landmark (LM) arguments specified for each in-
stance, separated by a backslash (absence of an explicitly articulated LM is
indicated by an X):
PREDICATES: Items: TRAJECTOR /LANDMARK Arguments

(there) be 7 food /clothes (x2); chickens /X; bitsoffibreglass
/street; hydraulicfluid /carpark; oil /cupboard; fuck-
ingpaper /floor
see [that there are] 1 lovebites /myneck
have/get/got/with 4 bodilyfluid/sheet; bloodymud /floor; mud /expensive
coat; tins /sittingroom
[absolute, no pred.] 2 [lemonade] /table; [mucus] /face
piss/pee/wee 3 dog /there; baby /me; baby /yourT-shirt
vomit (“go”) 1 I /kitchenfloor
trample (tr & intr) 2 brakefluid/carpet; [people] /cableundercarpet
spill/tip/upset 3 chips /floor; slides /floor; coffee /guests
dribble (intr) 1 dog /me
pour (intr) 1 petrol /forecourt
squirt (intr) 1 hooverdust /mypiggin’leg
spread (tr) 1 printingink /bench
wipe 1 spunk /somebody
drop (tr) 1 earth /floor
fly/go (intr) 2 paper? /theplace; cards /theplace
blow (tr) 1 burningdebris /X
water (intr) 1 [negative imperative] /foliage
run (intr) 1 [negative imperative?] /board
walk 1 thoseinpower /thoseweakerthanthemselves
sit there 1 he /her
climb 1 someone /us
Notable is the wide range of chaotic dispersal verbs (spill, dribble, squirt,
etc.), as well as of cases in which the TR is a substance construable as out
of place in all but rather restricted contexts (bodily fluids, motor-related
fluids, burning debris, vacuum cleaner dust, printing ink, etc.). Many of the
LMs involve surfaces (tables, floors, human bodies, body parts or cloth-
ing…) that may well not be sectorally covered by the TR, but for which
even relatively sparse dispersal of the TR is liable to be construed as cre-
ating a considerable mess. Consistent with this are the occasional exple-
tives (fucking, piggin’, bloody…) attached to both trajectors and landmarks.
Among the 27 Sectoral Coverage usages, the most striking regularity
involves the LMs. 18 (67%) of these explicitly involve a geographical do-
main (the world, the country, the UK, the island…), while 9 (18%) involve
the human body or a specified subdomain of it. If one includes cases of
290 Kurt Queller
non-specified LM where one of these two categories is nonetheless con-

textually understood, they comprise almost the whole SC category.
The Ambiguous category is most interesting for the examples it pro-
vides of possible link usages. Consider for example the following two:
…Look at those the marks all over the the window! After that
man cleaned them didn’t he Vicki? Oh! Pity they can’t do inside
as well!…
7.377 d:\icame\bnctex~1\spoken\kcu 59
…next door, but the herb must have come from that No, er Jane
spotted cos she said that’s where the seed must have come from
Yeah she said they seed freely , so we’ll have to watch we shall
have them all over the garden next year…
3.187 d:\icame\bnctex~1\spoken\kc2 19
Such usages may plausibly be construed in terms of either “sectoral cover-

age” or “chaotic dispersal”. A speaker may say such things with “cover-
age” of the LM (the window, the garden) in mind, while a hearer focuses
on the “dispersal” of the TR (the marks, the weeds) and the chaotic mess it
creates with respect to an LM normally expected to be free of such entities.
Functional embodiment in fact suggests that these considerations may fig-
ure more prominently in interpretation, since a hearer in the given context
will likely be more concerned with them than with the primarily topologi-
cal relationships implied in the “covering” schema. Note here that the
“chaotic dispersal” sense does not bleed over from the predicate, which is
neutral (zero in the first case, have in the second), but is simply a function
of situation and of encyclopedic knowledge about things like windows and
gardens and how people use them.
Of course, both senses are currently available in English. But they
weren’t always, and the independent “chaotic dispersal” sense of …all
over… argued for in section 2 above emerges historically from the “cover-
ing” sense. (For more on this, see Queller, in prep.) Ambiguous contempo-
rary usages, by definition construable in either sense, offer a window on
the sorts of link usages that ex hypothesi must have constituted the bridge
over which the transition to the new usage cluster occurred. (Many of the
corpus usages in the CD category also represent potential link usages, in-
asmuch as one may still imagine a speaker having formulated them on the
basis of the older “covering” schema; cf. discussion of the items in (7),
section 3.2.1 above.) Only so, it seems, can we account for the ultimate
emergence of usage patterns not adequately explained in terms of the “cov-

ering” schema, as discussed in section 2. (For further relevant corpus evi-
dence, see the following appendix.)
APPENDIX 2
Corpus analysis for “…written all over {one’s} face…” and (nearly)
equivalent expressions in Italian and German
I have claimed above (cf. also Queller 2001) that the English encoding
idiom to have {a feeling/thought} written all over one’s face shares with
numerous other all over expressions a prototypical “chaotic dispersal”
sense not directly derivable from the more basic “covering” sense of the
preposition, and glossable roughly as “to have signs of {a feeling/thought}
chaotically dispersed across one’s face (despite one’s attempts to maintain
a façade of composure)”. I have further suggested that ostensibly equiva-
lent expressions in other languages lack this specific connotation, which
must be attributed to a prototype “chaotic dispersal” sense inherent in
many uses of the prepositional expression (all) over. The present appendix
offers empirical corpus evidence in support of these claims.
A Google web search was done on Dec. 27, 2004 on the phrases
“…written all over {my/your/his} face…” (roughly 4,560, 8,440 and
10,900 hits, respectively). For each pronoun category (my/your/his), the
first 20 distinct, interpretable items were selected for analysis. (For exam-
ple, redundant citations of the same song lyric and unclear uses of the
phrase as a rubric were automatically eliminated. Also eliminated were rare
instances not instantiating the construction in question, e.g. a reference to a
man having words physically written all over his face.). Items were catego-
rized as to whether or not they reflected a “chaotic dispersal” (CD) context;
items that were ambiguous between “CD” and “non-CD” were classified as
“ambiguous”. In each case, the entity functioning as “figure” within the
…all over… expression was identified. For purposes of clarity, brevity and
cross-item comparability, these were sometimes given glosses deviating
from the precise wording, which was often lengthy and/or heavily depend-
ent on context for interpretation.
In assigning category values, not only was “figure” considered, but also
the context. Criterial was evidence for an inner state construed as some-
thing that the experiencer would prefer not be revealed to others. This
sometimes meant assigning different values in different contexts to essen-
292 Kurt Queller
tially the same figure. With reference to one’s sexual orientation being
“written all over one’s face”, for example, the one case in the English sam-
ple was classified “CD” because both context and word choice made clear
that someone’s deepest secrets were being revealed against his will (the
text involved a gay man’s memory of being publicly “outed” and hectored
in a religion class at the conservative bible college he had attended). In
contrast, the two references in the German corpus to the possibility of dis-
cerning sexual orientation from a person’s face involved neutral contexts
and wordings.14 Though it was tempting to assign these latter to the neutral
“non-CD” category, I classified them as “ambiguous”, that being somewhat
less favorable to my hypothesis.
Following are the results of analysis of the English corpus:
…written …written all …written all (All three)

all over my over your over his face…
face… face…
CD 15 (75%) 13 (65%) 10 (50%) 38 (63%)
Ambiguous 3 (15%) 5 (25%) 5 (25%) 13 (22%)
Non-CD 2 (10%) 2 (10%) 5 (25%) 9 (15%)
(Total) 20 (100%) 20 (100%) 20 (100%) 60 (100%)
“Figure” entities in English corpus (“…written all over your face…”

section) are:
CD sexual infidelity (x2) / wrongdoing (x2) / a history of being abused

(x2) / sexual desire / inability to forget an old lover / resentment
while checking out the sexual competition / gang membership /
anxiety, dejection, confusion / humiliation / a secret
Ambiguous love and pain (x2) / disappointment / bigotry / state of health and
of inner self
Non-CD love for one’s partner / distinction and sex appeal
14. For example: "Weder kann ich mir vorstellen, dass es dort viele (einheimische
und aufdringliche) Homosexuelle gibt, noch dass man einem seine sexuelle
Ausrichtung im Gesicht ansehen kann. Geht ja auch keinen was an!" [“I can't
imagine either that there are that many (local and importunate) homosexuals
there, or that you can tell a person’s sexual orientation just by looking at their
face. It’s none of anybody’s business, anyway!”] http://13313.rapidforum.com/
topic=100184128260
Occasionally, a gloss offers direct evidence for the specific “chaotic dis-
persal” connotations here proposed for the construction:
Idiom: Written all over your face. If someone has done some-
thing wrong or secret, but cannot hide it in their expression, it is
written all over their face. ... [italic emphasis added]
www.usingenglish.com/reference/ idioms/show.php?idiom=678
Note, for example, the following two song lyric passages, both referring to
being hopelessly in love with someone but desperately wanting not to show
it:
There’s a mask on the wall / That I should be wearin’

To keep you from seein’ / How I’m really feelin’.
I’d like to be cool / And tell you goodbye
(I think you’d better run now) / I’ve been such a fool
And now it’s written all over my face / That I’m about to cry.
(Belinda Carlisle, “I Need A Disguise”
home.att.net/~Bangles Com/Xinad.html)
…I wish I could be the girl at his side, / The one who has taken
my place,
Can everyone see what I’m trying to hide, / Isn’t it written all
over my face?
(Nina Simone, “That’s Him Over There”
lyricsplayground.com/ alpha/ songs/t/thatshimoverthere.shtml)
Another lyric contains one of the relatively few instances classified as

“non-CD”, there being no clear sense of any desire to conceal the inner
state from view:
I love the way you carry you

You have a lot of class and good taste
And you don’t have to say how much you care for me
Because it’s written all over your face
(Rude Boys, “Written All Over Your Face”
http://www.leoslyrics.com/listlyrics.php;jsessionid=6215EACC
68F33D95153DF2805AEEA46B?hid=z4cy37ZjW20%3D)
Even more clearly devoid of CD connotations is the following comment by

a photography instructor:
294 Kurt Queller
People don’t ask me if I like what I do. It’s written all over my
face.
www.wpja.org/quotes/index.shtml
(Even here, though, there may be a connotation of “I couldn’t hide it, even
if I tried”.)
Some non-CD cases represent attempts by advertising copy writers to ex-

ploit the idiom’s expressivity, though with dubious effects, since its “cha-
otic dispersal” connotations clash with the positive image they are trying to
project (cf. discussion in Queller 2001 Appendix 2 of similarly dubious
advertising copy uses of the idiom {The data} are all over the map):
“A distinguished and intriguing appeal will be written all over

your face in these Donna Karen [sunglasses]”
www.bizrate.com/buy/products_ _att259--256255-,cat_id10070
000.html
One might argue that the general preference of this English idiom for “cha-
otic dispersal” contexts results not from its intrinsic semantics, but rather
from a general human tendency, when talking about inner states being visi-
ble on people’s faces, to focus on situations experienced as uncomfortable
or chaotic. If this were so, then the “chaotic dispersal” connotations here
argued to be part of the expression’s semantics would be better understood
as mere artifacts of the contexts in which people happen to use the expres-
sion; it would accordingly be sufficient to posit only the basic “covering”
sense for (all) over. Such an argument further entails that expressions in
other languages that refer to inner feelings or thoughts being visible on
someone’s face should show a similar (nonlinguistic) propensity for “cha-
otic dispersal” contexts.
I would suggest that the latter entailment is false, and that this invali-
dates the counterargument in question. To test this, comparable corpora
were collected for the Italian expression {gli} si leggeva in faccia {lo
stupore} [literally: “{to-him} one read in face {the amazement}”] and the
German expression {man konnte} (ihm} {die Verwunderung} im Gesicht
ansehen [literally: “{one could} {to-him} {the amazement} in-the face
see”]. On Jan. 3, 2005, using the same selection and classification criteria
as for the English corpus, the first 20 distinct, interpretable instances were
collected and analyzed for “…si leggeva in faccia…” and “…im Gesicht
ansehen…” [about 399 and 189 total hits, respectively]. The results are as
follows:15
ITALIAN: GERMAN:
…si leggeva in faccia… …im Gesicht ansehen…
CD 3 (15%) 2 (10%)
Ambiguous 9 (45%) 6 (30%)
Non-CD 8 (40%) 12 (60%)
(Total) 20 (100%) 20 (100%)
“Figure” entities in Italian corpus (“…{gli} si leggeva in faccia…”) are:
CD loss of interest in sexual partner / worry (that narrator might beg

money or mug him) / unease over boyfriend’s internet porn habit
(discovered by snooping on his computer)
Ambiguous tension / unease / preoccupation / suffering / disbelief / hope / an
ostensibly mutual erotic interest / an identity crisis / a need for
help
Non-CD happiness (x2) / goodness of character / athletic character / opti-
mism / reminiscence / rage / evaluation of someone as mentally
unhinged
“Figure” entities in German corpus (“…{ihm} im Gesicht ansehen…”)

are:
CD sexual infidelity / dislike (of one’s dancing partner)

Ambiguous sexual orientation (x2) / physical strain / fear (of losing a game) /
cocaine addiction / a high opinion of oneself
Non-CD joy (x2) / physical strain (x2) / stupidity or intellectual slowness
(x2) / relief / eager anticipation / mood and state of health /
evaluation of so. as inept / dislike of a task / pain (in a horse’s
face)
Here is a typical “non-CD” example for each language:
15. Far more common (about 35,100 hits) is the expression {die Verwunderung}
stand {ihm} ins Gesicht geschrieben [“{the amazement} stood {to-him} into-the
face written”]. Analysis of results from a Google search on May 19, 2005, on
the phrase “ins Gesicht geschrieben” yielded the following breakdown: CD: 3 /
ambig: 9 / non-CD: 8.
296 Kurt Queller
Il generale guardava il sergente con gli occhi fissi e la bocca

aperta. Gli si leggeva in faccia un pensiero preciso: “Hanno
proprio ragione, è completamente matto!” [The general stared,
open-mouthed, at the sergeant. One could read in his face exactly
what he was thinking: “They’re right, he’s completely out of his
mind!”]
http://icsrobilante.scuole.piemonte.it/SCUOLE/elementari/ele_R
OBIL/PROGETTI/pace/pace_rc.htm
Die Skipper im Hafen schüttelten nur die Köpfe über uns, als wir
einliefen! In ihren Augen einfach unverantwortlich! Sie mussten
uns wohl für ganz unerfahrene Skipper halten. Bei so einem
Wetter fährt man doch auch nicht! Man konnte es ihnen im Ge-
sicht ansehen, was sie dachten. [The skippers in the harbor just
shook their heads over us as we were docking. Utterly irresponsi-
ble, in their view! They must have taken us for total rookies. Go-
ing out in weather like this – it’s just not done! You could see
from their faces what they were thinking.]
http://www.board-server.de/cgi-bin/foren/F_1361/forum.pl?
forum=29&thread=69
Both cases involve one party’s negative evaluation of another party’s

judgment or mental capacity. Context suggests that the evaluators have no
interest whatsoever in hiding their reactions. In the first case, distinctions
of military rank make it unnecessary for the general to do so; nor does the
description of his facial expression suggest any struggle to maintain a fa-
çade. Likewise, in the German example, there is no hint of concealment.
The ostentatious head-shaking, far from reflecting a failed attempt to hide
feelings, likely functions as a communicative gesture, signaling the evalu-
ators’ presumption of superior knowledge (Besserwisserei) and their open
disdain for the “rookie” fishermen they are observing.
Though it is not strictly wrong to translate such usages by saying that
the evaluators’ thoughts are “written all over their faces”, some native
English speakers (myself included) feel that such a rendition imports con-
notations not present in the original. More important than such subjective
reactions, however, is the English construction’s strong (though not abso-
lute) statistical preference for “chaotic dispersal” contexts, in contrast to
(nearly) equivalent expressions in other languages which show little or no
such skewing. I argue that this skewing is ultimately explainable only with
reference to the specific semantics of “chaotic dispersal” inherent not only

in this English expression, but also in a wide range of other expressions
using all over.
References
Barsalou, Lawrence W., Paula Niedenthal, Aron Barbey and Jennifer Ruppert
Learning and Motivation. Vol. 43: 43–92. San Diego, CA: Academic
Press.
Bransford, John D. and Nancy S. McCarrell
1974 A sketch of a cognitive approach to comprehension: Some thoughts
about understanding what it means to comprehend. In: Walter W.
Weimer and David S. Palermo (eds.), Cognition and the Symbolic
Processes, 189–229. Hillsdale: Lawrence Erlbaum Associates.
Brugman, Claudia
1981 Story of “over”. MA thesis, University of California, Berkeley. Re-
printed (1988) as The Story of Over: Polysemy, Semantics and the
Structure of the Lexicon. New York: Garland.
2002 Embodiment. Encyclopedia of Cognitive Science, 1102–1108. Mac-
millan Publishers.
Clark, Herbert H.
1996 Using Language. Cambridge: Cambridge University Press.
Croft, William
2000 Explaining Language Change: An Evolutionary Approach. Harlow,
England: Longman.
Cutting, J. Cooper
1998 Comprehension vs. production. In: Jef Verschueren, Jan-Ola Östman,
Jan Blommaert and Chris Bulcaen (eds.), Handbook of Pragmatics
(1998 Instalment). Amsterdam: John Benjamins.
Dewell, Robert
1994 Over again: Image-schema transformations in semantic analysis.
Cognitive Linguistics 5: 351–380.
Enfield, Nick J.
1993 Linguistic Epidemiology: Semantics and Grammar of Language
Contact in Mainland Southeast Asia. London: Routledge Curzon.
Evans, Nicholas and David Wilkins
2000 In the mind’s ear: The semantic extensions of perception verbs in
Australian languages. Language 76 (3): 546–592.
298 Kurt Queller
Fauconnier, Giles and Mark Turner

2003 Polysemy and conceptual blending. In: Brigitte Nerlich et al., 79–94.
Hopper, Paul J. and Elizabeth C. Traugott
1993 Grammaticalization. Cambridge: Cambridge University Press.
Iwata, Seizi
2004 Over-prefixation: a lexical constructional approach. English Lan-
guage and Linguistics 8 (2): 239–292.
Jespersen, Otto
1922 Language: Its Nature, Development and Origin. London: Allen and
Unwin.
Keller, Rudi
1994 On Language Change: The Invisible Hand in Language. Lon-
don/New York Rutledge.
Kilgarriff, Adam
2003 “I don’t believe in word senses.” In: Brigitte Nerlich et al. (eds.),
361–392. Berlin: Mouton de Gruyter.
Kreitzer, Anatol
1997 Multiple levels of schematization: A study in the conceptualization of
space. Cognitive Linguistics 8: 291–325.
Lakoff, George
the Mind. Chicago: The University of Chicago Press.
1987 Foundations of Cognitive Linguistics, Volume 1: Theoretical Pre-
requisites. Stanford: Stanford University Press.
1988 A usage-based model. In: Brygida Rudzka-Ostyn (ed.), Topics in
Cognitive Linguistics, 127–161. Amsterdam: John Benjamins.
1999 A dynamic usage-based model. In: Michael Barlow and Suzanne
Kemmer (eds.), Usage Based Models of Language, 1–63. Stanford:
CSLI Publications. (Reprinted in Langacker 1999b, Grammar and
Conceptualization, 91–145. Berlin: Mouton de Gruyter.)
Lewis, David
1969 Convention. Cambridge, MA: MIT Press.
Makkai, Adam
1972 Idiom Structure in English. The Hague: Mouton.
McMahon, April M.S.
1994 Understanding Language Change. Cambridge, MA: Harvard Uni-
versity Press.
Nerlich, Brigitte, Zazie Todd and David D. Clarke
2003 Emerging patterns and evolving polysemies: The acquisition of get
between four and ten years. In: Brigitte Nerlich et al. (eds.), 333–
360.
Nerlich, Brigitte, Zazie Todd, Vimala Herman and David D. Clarke

2003 Polysemy: Flexible Patterns of Meaning in Mind and Language.
Berlin: Mouton de Gruyter.
Ohala, John
1989 Sound change is drawn from a pool of synchronic variation. In: Leiv
Egil Breivik and Ernst Håkon Jahr (eds.) Language Change: Contri-
butions to the Study of its Causes, 173–198. Berlin: Mouton de
Gruyter.
Pustejovsky, James
1995 The Generative Lexicon. Cambridge, Mass: MIT Press.
Queller, Kurt
2001 A usage-based approach to modeling and teaching the phrasal lexi-
con. In: Martin Pütz, Susanne Niemeier and René Dirven (eds.), Ap-
plied Cognitive Linguistics, Volume 2: Language Pedagogy, 55–83.
2003 Metonymic sense shift: Its origins in hearers’ abductive construal of
usage in context. In: Hubert Cuyckens, René Dirven and John R.
Taylor (eds.), Cognitive Approaches to Lexical Semantics, 211–241.
Berlin/New York: Mouton de Gruyter.
in prep. Polysemy: A Usage-Based Approach. Berlin/New York: Mouton de
Gruyter.
Reddy, William
1993 The conduit metaphor: A case of frame conflict in our language
about language. In: Andrew Ortony (ed.), Metaphor and Thought,
164–201. Cambridge: Cambridge University Press. (First edition,
1979.)
Sandra, Dominiek and Sally Rice
1995 Network analyses of prepositional meanings: Mirroring whose mind
– the linguist’s or the language user’s? Cognitive Linguistics 6: 89–
130.
Schütze, Hinrich
2000 Disambiguation and connectionism. In: Yael Ravin and Claudia
Leacock (eds.), Polysemy: Theoretical and Computational Ap-
proaches, 205–219. Oxford: Oxford University Press.
Sinha, Chris and Tania Kuteva
1995 Distributed lexical semantics. Nordic Journal of Linguistics 18: 167–
199.
Sperber, Dan and Deirdre Wilson
1986 Relevance: Communication and Cognition. Oxford: Blackwell.
Stern, Gustaf
1931 Meaning and Change of Meaning. Bloomington: Indiana University
Press.
300 Kurt Queller
Straight, Stephen
1986 The importance and irreducibility of the comprehension/production
dialectic. In: Graham McGregor (ed.), Language for Hearer, 69–80.
Oxford: Pergamon Press.
Talmy, Leonard
1985 Lexicalization patterns. In: Timothy Shopen (ed.), Language Typol-
ogy and Syntactic Description, Vol. 3, 36–149. Cambridge: Cam-
Taylor, John
2002 Cognitive Grammar. Oxford: Oxford University Press.
2003 a Cognitive models of polysemy. In: Nerlich et al., 37–47.
2003 b Polysemy’s paradoxes. Language Sciences 25: 637–655.
Tomasello, Michael
2000 First steps toward a usage-based theory of language acquisition.
Cognitive Linguistics 11: 61–82.
Traugott, Elizabeth C. and Richard B. Dasher
2002 Regularity in Semantic Change. Cambridge: Cambridge University
Press.
Tyler, Andrea and Vyvyan Evans
2003 Reconsidering prepositional polysemy networks: The case of over.
In: Brigitte Nerlich et al, (eds.), 95–159. Berlin/New York: Mouton
de Gruyter. (First published in slightly different form in Language 77
(4): 724–765.)
Warren, Beatrice
1998 What is metonymy? In: Michael Hogg and Linda van Bergen (eds.),
Historical Linguistics 1995, 301–310. Amsterdam: John Benjamins.
Zlatev, Jordan
Stockholm: Gotab.
2003 Polysemy or generality? Mu. In: Hubert Cuyckens, René Dirven and
John R. Taylor (eds.), Cognitive Approaches to Lexical Semantics,
447–494. Berlin/New York: Mouton de Gruyter.
The embodiment of Europe: How do metaphors
evolve?1
Andreas Musolff
Abstract
The paper looks at ways in which the notion of “cultural evolution” can be applied
to metaphor, with particular reference to mappings from the source domain of the
HUMAN BODY to the target domain of POLITICAL ENTITIES. From antiquity to the
Renaissance, the concept of the BODY POLITIC served as the basis for prominent
theories of political systems as corporeal entities. Since the Enlightenment, how-
ever, BODY POLITIC theories seem to have disappeared from mainstream political
science, and only a few expressions, such as head of state, have survived in current
usage. On the other hand, corpus data for the use of metaphors in British and Ger-
man public debates on European politics show that BODY-POLITICS mappings are
still productive, especially in scenarios involving the source concept of the HEART.
This finding confirms insights of cognitive theory into the central function of
BODY-based source concepts for modern folk theories of state and society and re-
veals patterns of BODY-based metaphor use in public discourse that can be inter-
preted as evolutionary developments of conceptual and argumentative traditions in
the respective discourse communities.
Keywords: body politic, chain of being, corpus-based analysis, embodiment, evolu-

tion, heart of Europe, meme, metaphor, replication.
1. The BODY as a source concept in political discourse
The mapping A POLITICAL ENTITY IS A (HUMAN) BODY belongs to the com-

plex of the (GREAT) CHAIN OF BEING metaphors, whose central role in the
Western philosophical traditions was brought to prominence in classic
studies by A.O. Lovejoy (1936) and E.M.W. Tillyard (1943 [1982]). In the
1. I would like to thank Roslyn Frank, René Dirven and Michael White for their
helpful comments on draft versions of this paper.
302 Andreas Musolff
world-view of the Middle Ages and the Renaissance, the CHAIN OF BEING
linked “lower-order” entities (e.g. animal organisms) and “higher-order”
ones (e.g. human institutions) through a system of ontological correspon-
dences (Lovejoy 1936: 55–66; Tillyard [1982]: 95–108). Within this sys-
tem, the concept of the state as a “body politic” constituted a central part as
the interface between “macrocosm” and “microcosm” (Tillyard [1982]:
96–106; Hale 1971: 47). One strand of the BODY POLITIC tradition focused
on the person of the ruler, as epitomised in the theory of the King’s two
bodies, analysed in the seminal study by E. Kantorowicz ([1997]). In this
tradition, the ruler was seen, as having “in him” both a “Body natural [...],
subject to all Infirmities that come by Nature or Accident” and a “Body
politic” that “cannot be seen or handled, consisting of Policy and Govern-
ment, and constituted for the Direction of the People, and the Management
of the public weal” (Kantorowicz [1997]: 7). This analogy between the
concrete, natural body of the monarch and his abstract political and legal
powers served to separate the individual person of the ruler from the im-
mortal, supposedly divinely legitimised system of authority, justice and
dynasty (Kantorowicz [1997]: 7–23).
A second strand of body politic theory focused on explicating the func-
tions of parts of the political entity by reference to the parts and organs of
the body and their state of health. The medieval political philosopher and
bishop, John of Salisbury (ca. 1120–1180) in his treatise Policraticus, for
instance, assigned the most powerful position to the head, i.e. the prince,
who “is subject only to God and to those who exercise His office and rep-
resent Him on Earth, even as in the human body the head is given life and
is governed by the soul”; the senate, i.e. the council of state, occupied the
place of the heart, “from which proceeds the origin of good and bad
works” (Bass 1997: 206–207; for the original Latin passages cf. John of
Salisbury 1965, 1: 282–283). Thus, contrary to modern associations of the
HEAD with rationality and of the HEART with emotionality (as in the state-
ment he did not allow his heart [= feelings] to rule his head [= thoughts]),
the medieval Bishop used the BODY-STATE analogy to argue in favour of
the head, i.e. the prince, taking advice from the heart, i.e. the senate, as
well as being spiritually guided by the soul, i.e. the church (cf. Bass 1997:
208–212; Struve 1984: 309–315).
We can regard the mapping A POLITICAL ENTITY IS A (HUMAN) BODY as
a special case of embodiment: abstract concepts from the sphere of politics
are explicated by way of “translating” them into concepts of BODY
The embodiment of Europe 303
PARTS/ORGANS and their PHYSIOLOGICAL FUNCTIONS.2 The use of experi-

entially based BODY-related concepts for mappings that target abstract
entities has been shown by Mark Johnson to be part of a general cognitive
mode of organising knowledge in schemata based on “bodily movements
through space, our manipulation of objects and our perceptual interactions”
(Johnson 1987: 29). Cognitive analyses (Kövecses 1986, 1990, 2000;
Sweetser 1990; Lakoff and Johnson 1999) have highlighted the fundamen-
tal importance of bodily experience in conceptualising emotions, in the
spatial organisation of argumentative inference and in the semantics of
modality. Whilst this aspect of embodiment mainly concerns basic con-
ceptual metaphors that are acquired at an early stage in cognitive develop-
ment (Grady, Taub and Morgan 1996; Grady and Johnson 2002), BODY-
related concepts have also been shown to provide sources for highly com-
plex mappings in idioms and public discourse (Pauwels and Simon-
Vandenbergen 1995; Boers 1999; Niemeier 2000; Charteris-Black 2000;
White 2003). Lexical traces of the BODY POLITIC concept can still be found
in fixed lexical expressions, such as body politic itself, as well as head of
state or government, organ of the party or the long arm of the law (Deig-
nan 1995: 2, Brewer’s Dictionary of Phrase and Fable 1999: 149, 713).
These BODY-based metaphors fit into a modern, extended version of the
CHAIN OF BEING complex of mappings, which acts – at least in the Western
world – as an “unconscious cultural model indispensable to our under-
standing of ourselves, our world and our language” (Lakoff and Turner
1989: 167).
However, the existence of BODY-related source concepts in modern idi-
oms and public discourse is not necessarily proof of a continuous tradition
reaching back to the famous philosophical and poetic formulations of the
BODY POLITIC and GREAT CHAIN OF BEING theories that were the objects of
study in the “History of Ideas” research tradition, which predominantly
focused on the pre-Enlightenment era. David Hale, for instance, claimed
that the BODY POLITIC concept saw “its final flourishing” during the Ren-
aissance before it was made obsolete by “challenges to the anthropomor-
phic view of the universe” in the Enlightenment (Hale 1971: 47). Tillyard
and Lovejoy also concentrated on the classical traditions of the GREAT
CHAIN concepts leading up to the Renaissance, although they were aware
of the sinister modern revival of aspects of these concepts in Nazi ideology
2. For overviews of the different strands of embodiment theory cf. Wilson 2002
and Ziemke 2003.
304 Andreas Musolff
(Lovejoy 1936: 313; Tillyard [1982]: 117).3 Lovejoy even regarded the
nationalistic and racist (per-)versions of the CHAIN OF BEING metaphor in
Nazi ideology as a “wheel [come] full circle” (Lovejoy 1936: 313), but this
remark is more of a speculative hint than a conclusion.
How can we know whether modern mappings between the domain of
BODY-related concepts and the sphere of politics are indeed a “full circle”
return to pre-modern concepts? Unless we have a continuous “chain” of
statements linked by inter-textual allusions and cross-references, the asser-
tion of a coherent conceptual tradition is merely a supposition and can in
principle be challenged by the assumption of a basic BODY schema that is
activated from scratch in each instance of use. In fact, the two explanatory
perspectives need not even be considered to be in competition or mutually
exclusive. In the cognitive embodiment-oriented perspective, the emphasis
would be on investigating the conceptual origin of BODY-source imagery
(including its “extensions” in political and other discourse domains) in
universal, experientially based BODY schemata. In a complementary, his-
torical perspective, both classical and modern mappings of the metaphor
THE STATE IS A (HUMAN) BODY can be studied, like the examples analysed
by Zinken, Hellsten and Nerlich (this volume) as “discourse metaphors”
that evolve “over time, across topics and across different domains”. Thus,
without denying the importance of experientially based, “primary” con-
ceptualisation strategies for the study of the universal origins of BODY-
based metaphors, we can treat the socio-historically situated versions of
BODY-STATE mappings as phenomena of discourse and concept evolution.
2. Evolutionist approaches to conceptual history
The interpretation of metaphor traditions as stages in concept evolution can

build on theories that view conceptual history as comparable or even akin
to “evolution” in the sense developed in the biological sciences since
Charles Darwin. Irrespective of the many disagreements about the specific
biological mechanisms involved, “evolution” can be broadly characterised
as a chain of minimal “adaptive” changes in the genetic make-up of organ-
3. For detailed analyses of the conceptualisation of society or the nation as a body

that must be shielded from disease and parasites in Nazi-ideology and racist
theories in general cf. Sontag 1991: 82–84 and passim; Schmitz-Berning 1998:
460–464; Hawkins 2001: 44–47; Chilton 2005; Rash 2005, 2006: 125–156.
isms, which can be linked to ecological pressures (Dennett 1995: 39–60).

To match such an evolutionist theory model, cultural history needs to relate
the diachronic variation of conceptual entities to pressures in their envi-
ronment.
The application of a biological model of evolution in the humanities is
not a recent phenomenon: some 19th century linguists, for instance, seized
upon Darwin’s theory to construe historical narratives of national lan-
guages as organisms that had familial lines of descent, life-cycles, etc.
(Hoenigswald and Wiener 1987; Nerlich 1989; Frank, this volume). Whilst
such attempts were motivated by the classical version of Darwinist theory,
the focus of more recent evolutionist approaches has shifted to the applica-
tion of insights gained from modern genetics. Richard Dawkins, in The
Selfish Gene, has proposed the concept of “memes” to characterise “tunes,
ideas, catch-phrases, clothes fashions, ways of building pots or of building
arches” as the cultural counterparts of genes: like the latter, memes are to
be thought of as “replicators” of strings of information (Dawkins 1989:
192). If genes “propagate themselves in the gene pool by leaping from
body to body via sperms or eggs”, then memes can be regarded as “[propa-
gating] themselves in the meme pool by leaping from brain to brain via a
process which, in the broad sense, can be called imitation” (Dawkins 1989:
192).
If this analogy of biological and conceptual (“memetic”) evolution as
different types of the same underlying mechanism of “replication” were to
be validated, its potential for application to symbolic structures could in-
deed revolutionise the historical study of language and thought. Indeed,
Hull (1988, 2000), Blackmore (1999), Croft (2000) and Worden (2000)
have, with varying degrees of variation from Dawkins’ model, used the
notions of “meme” and “replication” to propose new approaches to lin-
guistic and conceptual evolution. Croft and Cruse (2004) even propose
such a perspective for metaphor:
When [a metaphor] is first coined, the only way to interpret it is to employ
one’s innate metaphorical strategy, which is subject to a wide range of con-
textual and communicative constraints. Once a metaphor takes hold in a
speech community and gets repeated sufficiently often, its character
changes. First, its meaning becomes circumscribed relative to the freshly
coined metaphor, becoming more determinate; second, it begins to be laid
down as an item in the mental lexicon; third, it begins a process of semantic
drift, which can weaken or obscure its metaphorical origins. […] At some
point along this path of change, the expression acquires a capability to act as
306 Andreas Musolff
a literal basis for further metaphorical extensions, which is not possible for a
fresh metaphor. (Croft and Cruse 2004: 204–205)
This description likens metaphor to a meme – Croft himself has introduced,
following a suggestion by Martin Haspelmath, the coinage lingueme as the
linguistic counterpart of Dawkins’ term (Croft 2000: 28) – in that it applies
the two-step model of evolution to it, i.e. innovation (“altered replication”)
and propagation “differential replication”). As regards innovation, Croft
and Cruse assume the working of a fundamental “innate metaphorical
strategy” that generates a new metaphor; its further “life cycle” is then
determined by the conditions of its propagation in the respective speech
community. The distinction of the two evolutionary phases of innovation
and propagation may be seen as a way of dealing with the problem of rec-
onciling the research perspectives outlined earlier. As cognitive phenom-
ena, “conceptual metaphors” must be motivated by recourse to fundamen-
tal cognitive strategies such as the mapping and blending of various
conceptual inputs to achieve a semantic innovation. As a complement to
this structural analysis, the study of the same conceptual entities as “dis-
course metaphors” explains their patterns of propagation within the com-
municative “environment” by revealing the specific communicative pur-
poses of their users in specific situational and social-historical contexts.
We can thus differentiate between two meanings of an “evolutionary ap-
proach” to cognitive metaphor research, one concerned with basic condi-
tions of metaphor creation/coinage, and a second one concerned with the
socio-historical conditions of its diffusion in speech communities and with
the concomitant changes to its structure and communicative function. In
the remainder of this article we shall focus on the second approach. It re-
mains to be seen, however, whether the empirical data to be analysed bear
out the neat distinction between “innovation” and “propagation” aspects.
Croft and Cruse’s account of metaphor evolution reifies to some extent
the notion of its object of analysis, turning successfully propagated meta-
phors into agents of further change that are able to “act as a literal basis for
further metaphorical extensions”. The agentive metaphor is inherited from
Dawkins’ original gene-meme analogy, which in turn was based on a con-
ception of the gene itself as a selfish “replicator” that behaves (statistically)
as if it was intent on propagation. This view of the gene is, as Dawkins
himself (1989: 50, 1999: 9–11) acknowledges, in itself metaphorical and
scientifically useful only as a shorthand way of expressing statistical prob-
abilities, not a “realistic” description. This analogical character of Dawk-
ins’ concept of genetic replication is multiplied by the further analogies of
the meme and its linguistic counterparts.4 In view of meme-oriented theo-

ries’ dependency on multiple analogies, Dennett (1995: 369) has concluded
that “the prospects of elaborating a rigorous science of memetics are doubt-
ful” but he insists that meme theory “provides a valuable perspective from
which to investigate the complex relationship between cultural and genetic
heritage”. This perspective is characterised by the switch from focusing on
the vehicle’s supposed interest (i.e. the survival and/or well-being of an
organism) to concentrating on the replicators’ (i.e. genes’ or memes’) evo-
lutionary success.
The focus on the replicators’ “interest” in optimal propagation helps to
explain seemingly odd cases where evolutionary adaptation appears to
work against or even destroy the “vehicle” organisms, but such cases may
still be evolutionarily explicable as changes that ensure the propagation of
the gene. This type of argument should, Dennett (1995: 365) argues, also
be used in memetic approaches: “Only if meme theory permits us better to
understand the deviations from the normal scheme will it have any warrant
for being accepted”. As regards conceptual evolution in general and the
evolution of conceptual metaphor in particular, this argument would sug-
gest that it might be profitable to pay special attention to variations of con-
cepts that may appear as exceptional or counter-productive from the user’s
viewpoint but that can be interpreted as ways of improving the meme’s
replication.
Whilst Dennett still defends the gene-meme analogy as a heuristic per-
spective, Dan Sperber (1996, 2000) insists that the differences in the repli-
cation of genes and memes demand a radical reconceptualisation of the
evolutionist (in his terminology: “naturalistic”) approach. Whereas genes
are normally replicated with extremely high fidelity, exact copying occurs
rarely in the cultural sphere (Sperber 1996: 102–104). Concepts have a
vastly higher rate of change than genetic mutation, due to their dependency
on continuous transformation from “mental representations” to “public
representations” and vice versa as the only mode of concept reproduction
available to humans. In order to survive, representations have to be com-
municated, i.e. they “get […] first transformed by the communicator into
public representations, and then re-transformed by the audience into mental
4. This would make Croft’s above quoted summary of a metaphor’s “life history”
a three-fold analogy: it is based on the concept of the lingueme, which is de-
picted in analogy to the meme, which is seen as analogous to the gene, which is
discussed in analogy to a “selfish” agent.
308 Andreas Musolff
representations”, and a “very small proportion of these communicated rep-

resentations” undergo this process repeatedly (1996: 25). Whilst Sperber’s
approach shares the basic biological analogy with meme theory (although
he prefers to draw parallels with viruses rather than with genes), his model
of conceptual change as a series of “transformations” from mental to public
representations (and back) reintroduces to some extent the users’ perspec-
tive: in order to be able to compete with other representations, a represen-
tation depends on the success of its vehicles’ (i.e. speakers’) competition
with other speakers (by using this particular representation rather than an-
other one). The “evolution” of representations is thus determined not only
by their “own” need to survive and propagate, but also by their usefulness
for speakers. This usefulness is, in turn, defined by Sperber as a corollary
of the “principle of relevance” (Sperber and Wilson 1995), i.e. as a “ten-
dency to optimize the effect-effort ratio” of the processing and transmis-
sion of contents (Sperber 1996: 53). Such a tendency in the evolution of
conceptual representations requires an explanation both in terms “of some
global macro-mechanism” and in terms of “the combined effect of count-
less micro-mechanisms” (1996: 54).
Building on Dennett’s and Sperber’s comments, we can formulate two
questions that will help to guide its application to the study of conceptual
metaphors:
a) In which way does a memetic/naturalistic approach provide an explana-

tion for seemingly odd or exceptional cases of conceptual evolution?
b) In which way can the conceptual evolution of metaphors be interpreted
in terms of micro-mechanisms combining to form conceptual traditions?
These questions will serve to orient the discussion of corpus-based meta-

phor data in the following sections.
3. HEART-based metaphors in EU debates
The data in question are derived from a bilingual corpus of British and
German press texts that contain metaphorical passages relating to European
politics, specifically those that conceptualise the project of socio-economic
and political integration. The renewal of this vision, which was launched in
the 1950s with the foundation of the “European Economic Communities”
(EEC), since the end of the Cold War has led to a proliferation of meta-
phors in public discourse.5 The concept for this emerging multi-national

entity is often expressed by a kind of synecdoche. As we shall see shortly
(cf. examples 6–22 below), the entity is often called Europe, and its offi-
cial names in the period following the EEC-stage – until 1993, the Euro-
pean Community (EC), since then, the European Union (EU) – suggest a
continent-wide referent. However, the economic-political institution
EEC/EC/EU has so far never included all the nations of the European con-
tinent: its pan-European appeal formulates a programme of turning the
geographic-cum-cultural entity “Europe” into an economic-political fed-
eration. Metaphorical mappings that are built onto this synecdoche provide
interpretations and evaluations of this political vision of a united Europe.
The corpus consists of a pilot version, called EUROMETA I, which in-
cludes some 2100 passages from 28 British and German newspapers from
the years 1989 to 2001,6 and a larger version, called EUROMETA II, which
covers the same period of time but was compiled from two general corpora,
i.e. the “Bank of English” (BoE) at the University of Birmingham and
“COSMAS” at the Institute for German Language in Mannheim (Ger-
many);7 it consists of more than 19,000 text passages. The leading research
interest in collecting the EUROMETA corpora was to find out if and how the
conceptual metaphors employed in Euro-related debates reflect and per-
haps also influence the differing public attitudes towards European inte-
gration in Britain and Germany. Viewed from an evolutionist perspective,
the corpus entries can be understood as public representations that compete
– metaphorically speaking – for evolutionary success in terms of widest
possible distribution among members of the two national discourse com-
munities. When the texts are read and interpreted by the members of the
public, they have to be turned into mental representations, and some of
these are in turn re-introduced into the pool of public representations by
discourse community members who participate actively in the debate (e.g.
politicians, media commentators and writers of letters to editors). We can
thus treat the corpus as a manifestation of conceptual structures that are
present in and perhaps characteristic of a specific discourse community.
Provided there are sufficient discourse data in our corpus that can be
5. Cf. Chilton and Ilyin 1993; Schäffner 1996; Musolff 2000, 2004.
6. The pilot corpus is accessible at http://www.dur.ac.uk/andreas.musolff/
Arcindex.htm.
7. For introductions to the “Bank of English” and “COSMAS” cf. the Internet
web-sites: www.cobuild.collins.co.uk/boe_info.html, www.ids-mannheim.de/kt/
corpora.shtml/.
310 Andreas Musolff
grouped together as belonging to the same conceptual source domain or

sub-domain, we can reconstruct emerging patterns of change as evidence of
conceptual evolution.
For the present study, I shall concentrate on one element of the source
domain of LIFE-BODY-HEALTH concepts, i.e. the concept of the HEART OF
EUROPE, which is by far the most frequently used concept of this domain in
EUROMETA II, accounting for 45% (i.e. 545 out of 1189) tokens from that
domain (cf. Tables 1 and 2 in the appendix). This strong representation of
the HEART-concept is hardly a surprise, considering the fact that it is one of
the most salient BODY-related concepts in the Western cultural tradition
and has given rise to a vast number of metonymic and metaphorical con-
cepts represented in popular idioms and proverbs.8
Niemeier (2000) has proposed that the metonymic link of HEART and
EMOTION, which is grounded in salient bodily experiences (i.e. of the
heartbeat quickening or slowing), underlies the mappings between the con-
cepts HEART and PERSON in English idioms: “the heart was taken as a me-
tonymy for the whole body and thus it stands for the whole person experi-
encing a specific emotion. It is on the basis of this archetypal metonymy
that the other understandings could arise and flourish” (Niemeier 2000:
210). Echoes of this metonymic mapping can be found in the “Bank of
English”, e.g. in a passage such as “Euro-sceptics will take heart from an
ICM poll” (The Guardian, 16 February 1995). However, such examples
based on the HEART-PERSON metonymy have got little to do with the notion
of a HEART as PART of the (political) BODY of Europe/the EU and have
therefore not been included in EUROMETA II.
The remaining metaphorical mappings of the HEART concept as a source
onto the target notion of Euro-political entities can be broadly grouped into
two types: 1) a general understanding of the HEART as the CENTRAL AND
MOST IMPORTANT PART of the BODY of the geographic and/or political
entity Europe; 2) a more specific notion of the HEART as a LIVING ORGAN
that can suffer damage from INJURY or DISEASE, which, given the HEART’S
central physiological function, is often conceived of as LIFE-THREATENING
to the whole ORGANISM. In the following sections, we shall discuss repre-
sentative examples of these two types of metaphor with special regard to
the question of whether their conceptual development in public discourse –
8. Cf. Niemeier (2000) and the entries for heart in Brewer’s Dictionary of Phrase
and Fable 1999: 557–558 and for Herz in German in Röhrich 2001, 2: 704–
708.
as reflected in the corpus data – can be interpreted in terms of an evolu-

tionist perspective.
3.1. The HEART as the CENTRE
The CENTRALITY aspect of the HEART concept is evident in references to

cities, regions or countries as being situated geographically in the heart of
Europe. These are by far the most frequent uses of the HEART OF EUROPE
concept in the German sample (with 252 out of altogether 336 tokens), and
they make up a sizeable portion in the smaller English sample (34 out of
209). Nearly half, i.e. 116, out of the 252 German tokens picture Germany
as a whole or German regions and cities as constituting the heart of Europe
or as being situated in the heart of Europe:
(1) Auch der Präsident der Industrie- und Handelskammer [...] richtete [...]
einen “dringenden Appell” an die Adresse der Politik [...]: “Berlin ist keine
Insel mehr, sondern liegt im Herzen Europas.” (die tageszeitung, 21 No-
vember 1992) [Even the president of the chamber of commerce made an ur-
gent appeal to the politicians: “Berlin is no longer an island but is in the
heart of Europe.”]
(2) Milosevics Entscheidung, sich an Deutschland zu wenden, ist eine weite-
re Bestätigung für die wachsende Macht dieses Landes im Herzen Europas.
(Die Zeit, 24 June 1999) [Milosevic’s decision to appeal to Germany again
underlines the growing power of this country in the heart of Europe.]
There are no references to Britain as being in the heart of Europe in either
the British or the German sample of EUROMETA II. This finding may seem
to be motivated solely on the grounds of geography; however, the corpus
data suggest that the HEART = CENTRE equation extends not just to the
countries of central Europe (i.e. Poland, the Czech Republic, Austria,
Croatia, Slovenia and Switzerland, apart from Germany), but also includes
Belgium, Franco-German border regions, and is extended to peripheral
regions such as Denmark or the Ukraine and Belarus. It even features in
references to the wars in the former Yugoslavia as taking place in the heart
of Europe, with the implication that what happens in the heart is – or
should be – close to, and of special concern for one’s emotional centre:
(3) Von 1991 bis 1995 wurde im Herzen Europas ein Krieg geführt, dessen
Brutalität und Menschenverachtung wir der Vergangenheit angehörig
glaubten. (Die Zeit, 18 February 1999) [Between 1991 and 1995, a war was
312 Andreas Musolff
fought in the heart of Europe, with a brutality and inhumanity which we had
thought belonged in the past.]
(4) Headlines about this war [in Kosovo] being in the ‘heart of Europe’ [...]
and other similar comments [...] have the implication that if this was hap-
pening thousands of miles away it would be more explicable and almost
normal. (The Guardian, 5 April 1999)
This emotive dimension of positioning a nation in the heart of Europe is
also discernible in references to candidate states for the EU enlargement
process, such as the Czech Republic, Poland and Hungary (which in the
following example are represented metonymically by their capital cities):
(5) [E. Diepgen, the Lord Mayor of Berlin]: “Prag, Warschau und Budapest
gehören zum Herzen Europas”, sagte er. (die tageszeitung, 2 January 1995)
[“Prague, Warsaw and Budapest belong in the heart of Europe”, he said.]
The appeal of relating a nation to the heart of Europe is even more evident
when we study uses where the notion of CENTRALITY that is embodied in
the HEART concept is extended beyond POSITIONAL to FUNCTIONAL as-
pects. The heart is, together with the brain, the most important organ as
regards the survival of an organism. It is this notion of FUNCTIONAL
CENTRALITY that also underpins the metonymic mapping HEART AS AN
OBJECT OF VALUE (Niemeier 2000: 204–206). In this sense, we speak for
instance of the “essence” of something as the heart of the matter (Concise
Oxford Dictionary 1979: 496). In this context, Britain finally comes into
the picture (i.e. into the corpus data). Indeed, the British public debate
about EC-/EU-politics over the course of the 1990s can be summarised
largely as a dispute about Britain’s relationship to the heart of Europe.
There is no question of Britain being in that heart, rather the issue is
whether Britain should or should not be at the heart of Europe, i.e. at the
functional centre of influence and power within the EU. The starting point
for the British heart of Europe debates in the 1990s was a speech given in
Germany in spring 1991 by the then Prime Minister, John Major. He com-
mitted Britain to supporting further integration of the “European Commu-
nity” (soon to become the “European Union”):
(6) John Major last night signalled a decisive break with the Thatcherite era,
pledging to a delighted German audience that Britain would work ‘at the
very heart of Europe’ with its partners in forging an integrated European
community. (The Guardian, 12 March 1991)
Major’s speech optimistically suggested an active role (working, forging)

at the most central part (the very heart) of the decision-making EC institu-
tions. Initially, most reports and comments interpreted his statement along
the lines of the FUNCTIONAL CENTRALITY perspective, even though, of
course, the political evaluations of Major’s vision differed according to the
Euro-political preferences of the commentators:
(7) Most galling of all, the British prime minister has decided that Britain is
at the `very heart of Europe’. Here is a dangerous new twist to British prag-
matism. (The Economist, 23 March 1991).
(8) Britain is still pulled both ways. It is not ‘at the heart of Europe’ – geo-
graphically and temperamentally, it is on the periphery. (The Economist, 26
September 1992)
(9) [Iain Vallance, chairman of British Telecom] urged the Government to
put Britain at the heart of Europe and play a full part in debates over mone-
tary union, employment, social costs, innovation and regional aid. (The
Guardian, 16 November 1993)
(10) Statt außen vor zu bleiben, versucht Großbritannien unter ihm, “im
Herzen Europas” Politik zu machen. (Die Zeit, 22 May 1992). [Rather than
staying on the sidelines, Britain under [Major] attempts to influence [liter-
ally: “make”] the politics “at the heart of Europe.”]9
In these examples, the HEART OF EUROPE is the politically most important
place in the European Community/Union, i.e. the place where decisions are
taken and where the relevant debates take place, as well as its tempera-
mental centre. Britain’s position vis-à-vis the heart is being defined in
terms of CLOSENESS or DISTANCE, and it is evident that this definition is
contentious, for otherwise it would make little sense for commentators to
insist or “urge” that Britain should be, work or play a full part in whatever
is decided at heart of Europe. This contentious status of Major’s statement
became especially prominent during the mid-1990s when his government’s
rejection of integration projects such as the currency union and EU-wide
social policies suggested that it wanted to move away from rather than
coming closer to the political CENTRE of the Union. Commentators now
used references to Major’s 1991 promise as proof of a change of govern-
ment policy or even of deception:
9. NB: the German texts consistently translate at the heart of Europe as im Herzen
Europas.
314 Andreas Musolff
(11) Britain […] as an old, offshore Euro-doubter that has improbably pro-
claimed itself to be at the ‘heart of Europe’ […]. (The Economist, 28 No-
vember 1992)
(12) Mr Major seems not to recall that his original project was to place Brit-
ain ‘at the heart of Europe’. His eyes are increasingly fixed on […] the next
British general election. (The Economist, 4 February 1995)
(13) Der Regierungschef, der einst Großbritannien ‘im Herzen Europas’
verankern wollte, hat keine feste Überzeugung. (Die Zeit, 13 December
1996). [The head of government, who once claimed he wished to lodge
Britain ‘at the heart of Europe’, has no firm convictions.]
Unlike the Tory government, the Britain at the heart of Europe formula
survived the landslide Labour victory in the general election of 1997. New
Labour’s new Prime Minister, Tony Blair, inherited the formula and was
soon credited and criticised for it in ways similar to his predecessor, some-
times to the self-conscious ennui or disbelief of the journalists commenting
on its use:
(14) The litany passes from government to government. A Britain at the
heart of Europe. We’ll hear the chant 1,000 times again this month […].
(The Guardian, 1 December 1997)
(15) Tony Blair’s attempts to place Britain at the heart of Europe faced a di-
rect challenge [...] (The Times, 23 March 1998)
(16) Blair will, im Kontrast zu den britischen Konservativen, sein Land
wirklich ‘im Herzen Europas’ ansiedeln. (Frankfurter Rundschau, 24 March
1999) [In contrast to the British Conservatives, Blair really does want to put
Britain ‘at the heart of Europe’.]
In these examples, the notion of Britain as a nation that is at the heart of
Europe has become independent from the initial use by Major. It now rep-
resents a specific (pro-European) political stance irrespective of whether it
is a Tory or a Labour government that is said to be promoting it. If we fol-
low Dennett’s suggestion and take the metaphor’s viewpoint (figuratively
speaking), we may conclude that the first phase of its evolution was suc-
cessful: it had been replicated often enough and become sufficiently
prominent to serve as a blueprint for new uses. Pursuing the evolutionist
perspective further, we shall now explore its conceptual variation across
the British and the German samples.
3.2. The heart of europe as an organ
When we compare the British and German EUROMETA II data for instantia-
tions of the concept HEART OF EUROPE the most prominent difference con-
cerns the political bias regarding the target topic. Whilst German politi-
cians and media generally assume the desirability and feasibility of one’s
own nation being at/in the heart of Europe, British commentators are at
best divided and often strongly negative (cf. examples 7, 8, 11, 15). How-
ever, there is another aspect to the British heart of Europe debate that sets
it apart from the German discussion, i.e. the frequent occurrence of HEART
tokens based on a physiological/medical source scenario. Physiologically, a
human body cannot function without a heart and it is in grave danger if the
heart is diseased or in some other way organically dysfunctional. This as-
pect of the source domain is highlighted in a second type of scenario where
the status of the HEART as a body organ that can be healthy or diseased is
foregrounded, so as to allow specific inferences about topics in the target
domain of EU politics:
(17) [....] if Mr. Major wanted to be at the heart of Europe, it was, presuma-
bly, as a blood clot. (The Independent, 11 September 1994)
(18) Britain may be advised that it can’t be at the heart of Europe if it is de-
tached from its arteries. (The Guardian, 10 June 1997)
(19) The Rotten Heart of Europe [title of book by B. Connolly, published in
1995]
(20) The European Commission is undemocratic. The truth is the rotten
heart of Europe will never be cleaned out. (The SUN, 17 March 1999)
In these examples,10 the reassuring promise that Britain would be/work at
the heart of Europe (as expressed by Major in 1991 and by his successor,
Tony Blair after 1997) is an implicit precedent for a critical comment by
way of a recontextualisation of the HEART concept within crudely put sce-
narios of HEART ILLNESS or DISEASE. In examples (17) and (18) the sce-
narios of the BLOOD CLOT and DETACHMENT FROM ARTERIES serve to re-
fute the promise of closer British involvement in EU decisions by
highlighting the discrepancy between the presupposition of a HEALTHY
HEART contained in the promise on the one hand and the LIFE-
THREATENING CONSEQUENCES of new government policies on the other. In
10. For analyses of further examples in the context of a general discussion of

ILLNESS scenarios in EU-related public debates cf. Musolff 2003.
316 Andreas Musolff
(19) and (20) the crass notion of the ROTTEN (i.e. DEAD or DYING) HEART is
applied to the allegations of mismanagement, nepotism and corruption
against the EU commission that led to the commission’s resignation in
March 1999. In these quotations, the commission itself is identified as the
HEART OF EUROPE that is not functioning properly on account of its
ORGANIC DEFECTIVENESS. The inferences at the level of the target topic are
strongly Euro-sceptical: if the Commission, as the EU’s heart, is rotting or
irreparably rotten, then the whole body (= the EU) is in danger or perhaps
even past hope of recovery. Hence, any further involvement in it or close-
ness to its heart is presented as foolish and dangerous.
This sarcastic and dismissive assessment also comes through in face-
tious uses of the phrase heart of Europe in the context of further BODY-
related terminology:
(21) These are just a handful of the issues which echo around Brussels’ con-
ference and dinner tables. There are many more in a similar vein – and one
thing binds them together. They bear no relationship to the British “debate”,
hearts, livers, gall bladders and all. (The Guardian, 1 December 1997)
(22) The contempt with which the French government treats Britain [in the
dispute over an immigrant camp near Calais] is beyond belief. Tony Blair
says he wants Britain to be at the heart of Europe. Well it looks this morning
as if Europe is showing us its backside. (The SUN, 3 September 2001)
The body parts of LIVER, GALL BLADDER or BOTTOM in these examples are
not seriously considered as parts of a European BODY POLITIC, but their
conceptual proximity to the HEART as parts of the general BODY concept is
used to achieve the main function of these passages, i.e. to ridicule the
promise of Britain working at the heart of Europe. They also remind read-
ers of a value-system attached in folk-theories to various body parts: the
humoristic effect of these quotations depends to a large extent on the con-
trast between the high-value HEART concept and the lower-value notions of
“inner” and “lower” organs/parts of the body.
Overall, tokens of the organic HEART scenario account for the largest
part of the 209 HEART tokens in the British sample, i.e. 175 tokens (=
84%), 31 of which belong to the HEART DISEASE/FAILURE scenario version.
By contrast, only 84 (= 25%) of the 336 German HEART tokens can be in-
terpreted in terms of organic scenarios. Included in these are the only two
cases fitting a DISEASE/FAILURE scenario,11 as well as 19 tokens of (non-
11. One example refers to a row between the French and German governments over
the euro currency introduction as revealing “the faulty cardiac valve behind the
committal) reports of the British debate about being/working at the heart of

Europe and six references to the title of Connolly’s 1995 book The Rotten
Heart of Europe, none of which endorses its damning diagnosis. Even on
the occasion of the 1999 nepotism scandal, which elicited as much critical
coverage in Germany as in Britain, we find no equivalent of the British rot
at the heart of Europe tokens. Although these statistical differences cannot
be regarded as fully validated on account of the different size and structure
of the source corpora of EUROMETA II,12 it seems plausible to conclude that
whereas both national samples rely on the source concept of the HEART OF
EUROPE as embodying a geographic-cum-political notion of CENTRALITY,
the British sample is characterised specifically by an emphasis on organic
scenarios, in particular the notion of the HEART OF EUROPE as SUFFERING
FROM A DISEASE or some other ORGANIC DETERIORATION.
4. Conclusions
The HEART DISEASE/FAILURE scenario as it appears in the British sample of

EUROMETA II carries with it, so to speak, an implicit reference to the phra-
seologism of Britain working/being at the heart of Europe, as used initially
by John Major. The main notion underlying that phraseologism was the
sense of FUNCTIONAL CENTRALITY of the HEART for the SURVIVAL of any
ORGANISM. If the EU is seen as a political BODY, then its HEART is its
POLITICAL CENTRE, and it is of crucial importance for Britain, if not to be
in it, then at least to be at or as close as possible to it – assuming, of
course, that the centre functions properly. This functional (rather than
merely positional) understanding of the phrase heart of Europe paved the
way for reinterpretations in terms of organic scenarios, especially ILL-
NESS/DISEASE scenarios. These scenarios are not entailed by the HEART
concept itself and they were certainly not intended by Major, or later,
Blair. However, the HEART concept used in the Prime Ministers’ optimistic
promises proved an easy object of caricature for Euro-sceptical critics, who
fainting fit” (Süddeutsche Zeitung, 16 June 1997: “Herzklappenfehler hinter

dem Schwächeanfall”). The second case is a denunciation of the heart of
Europe as ill, which is a quotation from an allegation by an extremist right wing
party that “Germany, as the heart of Europe, is ill due to its humiliation after
World War II” (die tageszeitung, 12 January 1990: “Wenn das Herz Europas
krank ist”).
12. Cf. the explanatory note regarding Table 2 in the appendix.
318 Andreas Musolff
only needed to pick up on its latent organic connotations to introduce the

ILLNESS/DISEASE aspect. From Major’s and Blair’s points of view, this
distortion of “their” metaphor source (HEART as an OBJECT OF VALUE ->
HEART as a DISEASED, DYSFUNCTIONAL ORGAN) must have been unwel-
come, and in view of the overall preponderance of positive sounding
HEART-related scenarios in EUROMETA II, the ILLNESS/DISEASE scenario
constitutes an exception, if not an aberration. But when we regard the con-
ceptual changes in the use of the phrase of Britain being/working at the
heart of Europe in British public discourse over the 1990s from what Den-
nett might call the “metaphor-meme’s point of view”, the ILLNESS/DISEASE
scenario looks less like an exception. In fact, its occurrence and statistical
rise over time helps to explain how the concept of the HEART OF EUROPE
survived, once the initial optimistic appeal of BEING AT THE CENTRE of the
political institution EU had worn out. In its changed appearance as part of
an ILLNESS scenario, the HEART OF EUROPE metaphor became the focus of
renewed debate and dispute.
This ability to turn into a contested notion heightened the metaphor’s
chances of “replication” among the competing political concepts. We are
thus dealing with a metaphor, which in Croft and Cruse’s model has been
propagated sufficiently to acquire “a capability to act as a literal basis for
further metaphorical extensions, which is not possible for a fresh meta-
phor” (Croft and Cruse 2004: 205). However, it is debatable whether this
change should be viewed as a mere “extension”, for the conceptual change
concerns not only the political bias (pro-EU → anti-EU) but also involves
the cognitive operation of (re-)introducing the “organic” aspect into the
conventional FUNCTIONAL CENTRALITY meaning of the political HEART
metaphor. This result seems to confirm the general hypothesis that “dis-
course metaphors” (in the sense introduced by Zinken, Hellsten and Ner-
lich) can undergo evolutionary conceptual change. It raises, however,
questions regarding the more specific distinction between innovation and
propagation: its discourse history belongs to the propagation aspect, as a
matter of “differential replication” (i.e. the “non-organic” and “organic”
scenarios exist side by side but their relative distribution differs across a
period of time and in particular across the two discourse communities). On
the other hand, the organic scenario, if judged to be more than a mere “ex-
tension” of the HEART OF EUROPE concept, also incorporates an element of
genuine conceptual innovation, i.e. “altered” metaphor-replication. Would
this innovative impulse have to be described as the start of another “life
cycle” of the metaphor? Or should we distinguish degrees of “innovative-
ness”, some of which might be considered to be a product of propagation?

At this point we can only formulate a basic hypothesis about the evolution-
ary survival of a metaphor, adopting (in Dennett’s sense) the “metaphor’s
viewpoint”. Its successful competition in discourse history seems to depend
on the interplay of two factors: 1) an experiential grounding, which ensures
that an essential meaning consistency is preserved in the conceptual varia-
tion and 2) on the conceptual flexibility that allows for its use in differing
or contrasting – even negatively or ironically biased – scenarios.
With regard to the second question put at the end of section 2 concern-
ing how discursive micro-mechanisms combine to form conceptual trends
or traditions, it is clear that the data presented here need a great deal of
further empirical corroboration from more extensive long range diachronic
studies. These will have to include a reassessment of some findings about
the BODY POLITIC concept that were formulated by the “History of Ideas”
and “conceptual history” approaches. For now, we can state that the an-
cient conceptual metaphor of the BODY POLITIC seems to have survived not
only in a few lexicalised expressions such as head of government etc., but
also in the form of metaphor scenarios based on the HEART concept, which
are applied to the target domain of political entities. The examples from the
EUROMETA corpora show that these scenarios allow for considerable varia-
tion in the conceptualisation of the European Union. In addition, charac-
teristic patterns of distribution of the two main scenarios – i.e. a prepon-
derance of geo-political positioning in the German media and a preference
for organic scenarios in the British debate – appear to indicate attitudinal
differences between the respective discourse communities. Further re-
search is needed to establish whether these hypotheses can be confirmed
and whether a chain of conceptual “adaptations” can be reconstructed that
links political HEART scenarios in present-day debates to the history of
BODY POLITIC-metaphors in Western thought and discourse.
320 Andreas Musolff
Appendix
Table 1. Conceptual elements of the LIFE-BODY-HEALTH domain in EUROMETA II
Source concepts English lexemes German lexemes

LIFE, SURVIVAL to live, life, alive, Leben, leben, lebendig,
survival überleben, Weiterleben, ins
Leben rufen
BIRTH/BABY birth, rebirth, born, Geburt, geboren, Wiederge-
still-born, premature burt, Frühgeburt, Missgeburt,
birth, abortion, bap- Kind, Baby
tism, baby, (bouncing)
child
DEATH death sentence/ war- Tod, tot
rant/ knell
ILLNESS/DISEASE (general) ill, illness, sick (sick krank, kranker Mann Euro-
man of Europe) pas, kränkelnd
I/D: EUROSCLEROSIS Euro(-)sclerosis Eurosklerose
I/D: MADNESS (Euro-)madness
I/D: INFLUENZA Asian (economic) flu Grippe
I/D: VIRUS virus
I/D: COLIC colic
I/D: WOUND Wunde, Narbe
I/D: WASTING/TBC Schwindsucht
I/D: HURT Wehtun
CURE/THERAPY/CARE therapy, diagnose Pflege, pflegen, Nachsorge
HEALTH/FITNESS/ to recover, recovery, Gesundheit, gesund, gesün-
RECOVERY revive, health, healthy der, gesunden, sich erholen,
Fit, Fitness
BODY PART: HEART heart Herz
BODY PART: gall bladder
GALL BLADDER
BODY PART: LIVER liver
BODY PARTS: EYES Augen
BODY PART: HEAD Kopf
BODY PARTS: LEGS Beine
BODY PARTS: FEET Füße
BODY PARTS: MUSCLES Muskeln
BODY PART: BOTTOM backside
Table 2. Tokens for conceptual elements of LIFE-BODY-HEALTH source concepts

in EUROMETA II in order of frequency
Source concepts Tokens in Sub-totals Tokens in Sub-totals Tokens

English German overall
sample sample
BODY PARTS (BP) 212 377 589
BP: HEART 209 336
BP: EYES 19
BP: HEAD 9
BP: LEGS 6
BP: FEET 5
BP: MUSCLES 2
BP: LIVER 1
BP: GALL BLADDER 1
BP: BOTTOM 1
ILLNESS/DISEASE (I/D) 60 137 197
I/D: SICK/ILL 40 92
I/D: EUROSCLEROSIS 12 32
I/D: MADNESS 4
I/D: INFLUENZA 2 3
I/D: VIRUS 1
I/D: COLIC 1
I/D: WOUND 5
I/D: WASTING/TBC 3
H/I: HURT 2
BIRTH/BABY 56 100 156
HEALTH/FITNESS/ 37 111 148
RECOVERY
LIFE, SURVIVAL 23 55 78
DEATH 4 8 12
CURE/THERAPY/ 2 7 9
CARE
Totals 394 795 1189*
* The 2:1 difference in absolute numbers of German and British tokens should not
be taken as prima facie evidence of greater general popularity of LIFE-BODY-
HEALTH metaphors in German press language. Rather it is most probably due to the
fact that the German corpus contains many more texts for the same period (1989–
2001) than the BoE. Overall, COSMAS (1500+ million word forms) is more than
three times larger than the BoE (450+ million).
322 Andreas Musolff
References
Bass, Allen M.
1997 The metaphor of the human body in the political theory of John of
Salisbury: Context and innovation. In: Bernhard Debatin, Timothy R.
Jackson and Daniel Steuer (eds.), Metaphor and Rational Discourse,
201–213. Tübingen: Niemeyer.
Blackmore, Susan
1999 The Meme Machine. Foreword by Richard Dawkins. Oxford: Oxford
University Press.
Boers, Frank
1999 When a bodily source domain becomes prominent. In: Raymond W.
Gibbs and Gerard Steen (eds.), Metaphor in Cognitive Linguistics,
47–56. Amsterdam and Philadelphia: John Benjamins.
Brewer’s Dictionary of Phrase and Fable
1999 Adrian Room (ed.). London: Cassell.
Charteris-Black, Jonathan
2000 Metaphor and vocabulary teaching in ESP economics. English for
Special Purposes 19: 149–165.
Chilton, Paul
2005 Manipulation, memes and metaphors: The case of Mein Kampf. In:
Saussure, Louis de and Peter Schulz (eds.), Manipulation and Ide-
ologies in the Twentieth Century, 5–45. Amsterdam and Philadel-
Chilton, Paul and Mikhail Ilyin
1993 Metaphor in political discourse: The case of the “Common European
House”. Discourse and Society 4 (1): 7–31.
Concise Oxford Dictionary
1979 J. B. Sykes (ed.). Oxford: Oxford University Press.
Connolly, Bernard
1995 The Rotten Heart of Europe. London: Faber.
Croft, William
2000 Explaining Language Change: An Evolutionary Approach. London:
Longman.
Croft, William and D. Alan Cruse
2004 Cognitive Linguistics. Cambridge: Cambridge University Press.
Dawkins, Richard
1989 The Selfish Gene. Oxford and New York: Oxford University Press.
(First ed. 1976).
1999 The Extended Phenotype. The Long Reach of the Gene. Oxford:
Oxford University Press. (First ed. 1982).
Deignan, Alice
1995 COBUILD English Guides. Vol. 7: Metaphor Dictionary. London:
HarperCollins.
Dennett, Daniel C.
1995 Darwin’s Dangerous Idea. Evolution and the Meanings of Life.
London: Penguin.
Frank, Roslyn M.
this vol. The language-species-organism analogy: A complex adaptive sys-
Grady, Joseph, Sarah Taub and Pamela Morgan.
1996 Primitive and Compound Metaphors. In: Adele Goldberg (ed.), Con-
ceptual Structure, Discourse and Language, 177–187. Stanford:
CSLI.
Grady, Joseph and Christopher Johnson
2002 Converging evidence for the notions of subscene and primary scene.
In: René Dirven and Ralf Pörings (eds.), Metaphor and Metonymy in
Comparison and Contrast, 533–554. Berlin and New York: De
Gruyter.
Hale, David George
1971 The Body Politic. A Political Metaphor in Renaissance English Lit-
erature. The Hague and Paris: Mouton.
Hawkins, Bruce
2001 Ideology, metaphor and iconographic reference. In: René Dirven,
Roslyn M. Frank and Cornelia Ilie (eds.), Language and Ideology.
Volume II: Descriptive Cognitive Approaches, 27–50. Amsterdam
and Philadelphia: John Benjamins.
Hoenigswald, Henry M. and Linda F. Wiener (eds.)
1987 Biological Metaphor and Cladistic Classification. An Interdiscipli-
nary Perspective. London: Francis Pinter.
Hull, David L.
1988 Science as a Process: An Evolutionary Account of the Social and
Conceptual Development of Science. Chicago: University of Chicago
Press.
2000 Taking memetics seriously: Memetics will be what we make it. In:
Robert Aunger (ed.), Darwinizing Culture. The Status of Memetics as
a Science, 42–67. Oxford: Oxford University Press.
John of Salisbury
1965 Policraticus sive De nugis Curialium et vestigiis philosophorum.
Clemens C. I. Webb (ed.). Reprint: Frankfurt am Main: Minerva
(Original ed.: Oxford: Clarendon Press 1909).
324 Andreas Musolff
Johnson, Mark
Kantorowicz, Ernst H.
1997 The King’s Two Bodies: A Study in Mediaeval Political Theology.
With a new preface by William Chester Jordan. Princeton, N.J.:
Princeton University Press. (Orig. ed. 1957)
Kövecses, Zoltán
1986 Metaphors of Anger, Pride and Love: A Lexical Approach to the
Study of Concepts. Amsterdam and Philadelphia: Benjamins.
1990 Emotion Concepts. New York: Springer.
Lakoff, George and Mark Turner
1989 More than Cool Reason: A Field Guide to Poetic Metaphor. Chicago
and London: University of Chicago Press.
Lovejoy, Arthur O.
1936 The Great Chain of Being. Cambridge, Mass.: Harvard University
Press.
Musolff, Andreas
2000 Mirror Images of Europe. Metaphors in the Public Debate about
Europe in Britain and Germany. Munich: iudicium.
2003 Ideological functions of metaphor: The conceptual metaphors of
health and illness in public discourse. In: René Dirven, Roslyn M.
Frank and Martin Pütz (eds.), Cognitive Models in Language and
Thought. Ideology, Metaphors and Meanings, 327–352. Berlin and
2004 Metaphor and Political Discourse. Analogical Reasoning in Debates
about Europe. Basingstoke: Palgrave.
Nerlich, Brigitte
1989 The evolution of the concept of ‘linguistic evolution’ in the 19th and
20th century. Lingua 77: 101–112.
Niemeier, Susanne
2000 Straight from the heart – metonymic and metaphorical explorations.
In: Antonio Barcelona (ed.), Metaphor and Metonymy at the Cross-
roads. A Cognitive Perspective, 195–213. Berlin and New York: De
Gruyter.
Pauwels, Paul and Anne-Marie Simon-Vandenbergen

1995 Body parts in linguistic action: Underlying schemata and value
judgements. In: Louis Goossens, Paul Pauwels, Brygida Rudzka–
Ostyn, Anne-Marie Simon-Vandenbergen and Johan Vanparys (eds.),
By Word of Mouth: Metaphor, Metonymy and Linguistic Action in
Cognitive Perspective, 35–69. Amsterdam and Philadelphia: John
Benjamins.
Rash, Felicity
2005 Metaphor in Hitler’s Mein Kampf. metaphorik.de 9: 74–111.
http://www.metaphorik.de/09/
2006 The Language of Violence. Adolf Hitler’s Mein Kampf. New York:
Peter Lang.
Röhrich, Lutz
2001 Das große Lexikon der sprichwörtlichen Redensarten. (New edition).
3 vols. Darmstadt: Wissenschaftliche Buchgesellschaft.
Schäffner, Christina
1996 Building a European house? Or at two speeds into a dead end? Meta-
phors in the debate on the united Europe. In: Andreas Musolff,
Christina Schäffner and Michael Townson (eds.), Conceiving of
Europe – Unity in Diversity, 31–59. Aldershot: Dartmouth Publish-
ers.
Schmitz-Berning, Cornelia
1998 Vokabular des Nationalsozialismus. Berlin and New York: W. de
Gruyter.
Sontag, Susan
1991 Illness as Metaphor. Aids and its Metaphors. London: Penguin.
Sperber, Dan
1996 Explaining Culture. A Naturalistic Approach. Oxford: Basil Black-
well.
2000 An objection to the memetic approach to culture. In: Robert Aunger
(ed.), Darwinizing Culture. The Status of Memetics as a Science,
163–173. Oxford: Oxford University Press.
Sperber, Dan and Deirdre Wilson
1995 Relevance: Communication and Cognition. 2nd ed. Oxford: Black-
well.
Struve, Tilman
1984 The importance of the organism in the political theory of John of
Salisbury. In: Michael Wilks (ed.), The World of John of Salisbury,
303–317. Oxford: Basil Blackwell.
Sweetser, Eve E.
1990 From Etymology to Pragmatics. Metaphorical and Cultural Aspects
of Semantic Structure. Cambridge: Cambridge University Press.
326 Andreas Musolff
Tillyard, E.M.W.
1982 The Elizabethan World Picture. Harmondsworth: Penguin. (First ed.
1943).
White, Michael
2003 Metaphor and economics: the case of growth. Journal of English for
Special Purposes 22: 131–151.
Wilson, Margaret
9: 635–636. http://philosophy.wisc.edu/shapiro/PHIL951/951articles/
wilson.htm
Worden, Robert P.
2000 Words, memes and language evolution. In: Chris Knight, Michael
Studdert-Kennedy and James R. Hurford (eds.), The Evolutionary
Emergence of Language. Social Function and the Origins of Lin-
guistic Form, 353–371. Cambridge: Cambridge University Press.
Ziemke, Tom
2003 What’s that thing called embodiment? In: Richard Alterman and
David Kirsh (eds.), Proceedings of the 25th Annual Meeting of the
baum.
this vol. Discourse metaphors
Sociocultural situatedness of terminology in the life
sciences: The history of splicing
Rita Temmerman
Abstract
We trace the origin of the metaphorical term splicing in its new usage in biotech-
nology. This case illustrates how a thought can remain relatively invisible, almost
hiding in the shadows, until it encounters the appropriate verbal counterpart and
how it is possible to study the sociocultural situatedness of terminology which can
be traced in the textual archives of human experience, a repository of collective
human memory.
In trying to gain more insight into the mechanisms behind lexicalisation we in-
terpret the use of the term splicing in the life sciences taking into account the meta-
phorical models discussed in Temmerman (2000) (DNA is information, coding, a
language, the book of life, a map, a film, software). We first examine the nature of
the diachronic study of scientific discourse and then concentrate on the polysemy of
splicing through a historical, diachronic, semantic and discourse analytic linguistic
analysis.
Keywords: cognitive modelling, discourse studies, distributed emergent cultural

cognition, embodiment, lexicalisation, metaphor, neologisms, polysemy, sociocog-
nitive terminology, sociocultural situatedness, splicing.
1. Introduction
In this paper we attempt to add the perspective of sociocognitive terminol-

ogy theory (Temmerman 2000) to the discussion concerning cognition, and
the role of sociocultural situatedness, taking the history of the English lexi-
calisation splicing as a case study. This case study proves that contempo-
rary novelist Jeanette Winterson’s motto not words for things but words
which are living things with the power to move applies to the language of
the life sciences. In traditional terminology theory (Felber 1984; Felber and
328 Rita Temmerman
Budin 1989; Arntz and Picht 1989; Wüster 1991) the emphasis was on
concepts and their representations. The classical theory of terminology
(especially the Vienna school) studied terms in special language only in so
far as they were indications for things, linguistic signs designating objects
in the real world. The conviction was that the function of terminology the-
ory should be to facilitate objective communication about the real world by
offering principles for the standardisation and description of concepts and
terms. The first principle was the onomasiological perspective which im-
plied that the linguistic unit (the term) should not be the starting point for
terminological analysis, but rather the concept that was believed to exists
in the mind. The second and third principles were that concepts have clear
boundaries and that in order to arrive at an objective understanding of the
world, exact definitions of the concepts are feasible and should be aimed
at. The fourth principle was the univocity ideal: terms should ideally refer
to one concept and one concept should be referred to by a unique term. In
order to achieve this, the creative potential of language − its power to move
− was to be either ignored (in descriptive terminology) or curtailed (in
prescriptive terminology). Scientific terminology was to be monosemous
and devoid of figurative meaning (tropes). The fifth principle was that
terminology had to be studied synchronically. Meaning evolution was not
part of terminological analysis. In Temmerman (2000) all of these princi-
ples were questioned and shown to be unrealistic if, for instance, applied to
the study of life sciences terminology.
Much of this ongoing discussion in terminology theory has of course
been studied in the cognitive sciences in general and in cognitive linguis-
tics more specifically. Temmerman (2000) extensively studied a number of
neologisms in life science terminology related to laboratory techniques
(sequencing, blotting, cloning, mapping). This consisted of a historical,
diachronic, sociological, discourse-oriented and multilingual analysis of
special language vocabulary as witnessed in neologisms.
1.1. Theoretical considerations
In this paper I would also like to relate the history of the term splicing to
the role of sociocultural situatedness (e.g. Lindblom and Ziemke 2003),
“collective cognition” (Bernárdez this volume) and “distributed, emergent
cultural cognition” (Sharifian 2003 this volume). The concept of sociocul-
tural situatedness, i.e. the idea that cognition requires a social and cultural
Sociocultural situatedness of terminology in the life sciences 329
embedding, has recently received much attention in cognitive science and

artificial intelligence research. Of particular significance to this paper is
the work of Sharifian who views cognition as a culturally distributed sys-
tem that emerges from the interaction between the members of a cultural
group. In our case, the cultural groups in question would be researchers in
general, researchers in the field of recombinant DNA technology along
with those working in the field of philosophy of science.1 These cultural
groups could be split up into subgroups, e.g. based on the natural language
they are most familiar with for communicating their results or based on the
school of thought of their background.
Sharifian emphasises that members of a cultural group negotiate and re-
negotiate their cultural cognition across time and space. This view of cog-
nition provides a basis for understanding group-level conceptualisation of
experience. Cultural conceptualisations are emergent in the sense that they
emerge between the members of a cultural group and they are distributed
as they are “not equally imprinted in the mind of each member”. Within
this perspective, language is considered both “a system of distributed cog-
nition” and also a system “that largely embodies cultural conceptualisa-
tions of experience by their speakers”. Sharifian prefers the term “concep-
tualisation” over “concept” as it reflects the dynamic nature of cognition.
He distinguishes different types of conceptualisations like schematisations,
categorisation, metaphors and conceptual blends. Language is viewed as a
distributed system as well as a repository for cultural conceptualisation.
In studying the history of the lexeme “splicing” in English we attempt
to chart the history of the meaning of a word. We study this word through
time as it is used by different cultural groups (e.g. sailors, film produces,
joiners, molecular biologists, Recombinant DNA technologists) sharing
one or more natural languages. In order to identify the different meaning
evolutions we consulted different types of discourse. (see also Calsamiglia
and Van Dijk 2004). Our vantage point has been to try to gain insight into
cognition as it emerges from terms and descriptions in scientific publica-
tions (both in original scientific articles and in popularising literature).
Initially, we studied the term splicing as part of a larger project that ex-
amined the process of naming, i.e. the lexicalisation in the English lan-
1. Researchers working in the field of genetics are becoming more aware of the
importance and the role of metaphors and analogies in shaping the direction of
their investigations as well as the models they use to acquire these understand-
ings. In short far more attention is being paid to these issues than in the past (see
e.g. De Chaddarevian 2002; Kay 2000; Van Dijck 1998).
330 Rita Temmerman
guage of newly developing categories in life science disciplines like mo-

lecular genetics, genetic engineering and biotechnology. Splicing was en-
countered in the texts as a term referring to a laboratory technique in ge-
netic engineering (gene splicing) and to a biological process (mRNA-
splicing) (see section 2.1).
For this project a corpus of English language texts written after 1953
(discovery of the double helix structure of DNA by Franklin, Watson and
Crick) was compiled. The senses of splicing related to the biological do-
main were recovered from this textual corpus, but the earlier senses of
polysemous splicing had to be looked for in texts related to different do-
mains (sailing, carpentry, cinema techniques, etc.) and in etymological
resources.
The larger project had as its underlying hypothesis the following: that
the creative forces of linguistic reasoning are part and parcel of the creative
mechanisms applied (consciously or not) when attempting the advancement
of science. Metaphorical modelling can be seen as one of these creative
mechanisms (see e.g. Bono 1990, 1995, 2001). To understand the essence
of life and genetics the analogies between “life” and “information proc-
essing” have been elaborated. In Temmerman (2000, chapter five), we
showed that sensitivity to analogical thinking and metaphor has an impor-
tant role to play in the development of new terminology in the life sciences.
We also pointed out how social, cognitive and technological changes are
accompanied by changes in the language and that the understanding of
these four simultaneous and mutually influential processes in the life sci-
ences can be penetrated more deeply by studying the textual archives of the
life sciences.
In this article we describe the complicated nature of the category2 of
splicing. We shall present the prototype structure of splicing in the fol-
2. Sharifian (2003, this volume) defines “conceptualisation” as a cover term that

refers to fundamental cognitive processes such as schematisation and categori-
sation. ”Schematisation” refers to a process that involves the systematic selec-
tion of certain aspects of a referent scene to present the whole, disregarding the
remaining aspects while “categorisation” is a process by which distinct entities
are treated as somehow equivalent. These cognitive processes naturally lead to
the development of schemas and categories. Sharifian refers to such products of
human cognition collectively as conceptualisations.
There are also other kinds of conceptualisations such as metaphors and
conceptual blends. In Temmerman (2000) we distinguished between “concepts”
lowing way, namely, by examining how the flexibility of this

prototypically structured lexical item is, to a large extent, the result of a
mechanism of polysemisation and the role that metaphorical models have
played in the development of the category.
1.2. Overview of study
Historically, the English word splicing traces its etymology back to a term
that was borrowed from Dutch before establishing its own identity in Eng-
lish. The Dutch term left traces of itself in English in idiomatic usages. In
English it acquired a specific meaning in the special language of wood
repair, metal repair, film repair, and finally gene repair. Our hypothesis is
that the motivation for assigning the name splicing to the insertion of for-
eign genetic material into a plasmid is the result of a number of analogies
reinforcing one another. Furthermore, the core meaning of splicing (to cut
and paste) can now be pictured in a metaphorical frame of text editing (text
repair), a new understanding within the metaphorical model of under-
standing genes as a text written in a language with a four letter alphabet.3
The latter observation shows how the “code metaphor” (Nerlich and Ding-
wall 2004) can interact with other metaphoric resources.
We apply the methods of componential analysis and diachronic sche-
matic representation (Geeraerts 1983, 1985, 1992, 1996) to the semasi-
ological analysis of the lexeme splicing. The history of the polysemisation
of splicing is illustrated by the phases of meaning extension of the term.
Onomasiologically considered, in an attempt to gain more insight into the
mechanisms behind lexicalisation, we interpret the use of the term splicing
in the life sciences, taking into account the metaphorical models we de-
scribed elsewhere (Temmerman 2000, 2002). We attempt to uncover the
existing cognitive frames which may have played a role in understanding
and creating new cognitive frames.
In section 2 we reconstruct the history of splicing in biotechnology and
molecular biology and indicate when and how two distinct units of under-
standing occurred (gene splicing and mRNA splicing) (section 2.1). In sec-
which allow for traditional definitions as they can be clearly delineated and
“categories” which show prototype structure.
3. The metaphorical “code” or “text” model, e.g. THE GENOME IS THE BOOK OF LIFE
has received significant attention of late (see e.g. Nerlich and Dingwall 2004;
Kay 2000; Hayles 1999).
332 Rita Temmerman
tion 2.2, we describe the diachronic meaning extension process and indi-
cate the models of experience or cognitive cultures of which rope splicing,
wood splicing, metal splicing and film and tape splicing are a part. These
may have served as domains of inspiration for the metaphorical naming of
gene splicing. In section 3 we present our combined method for analysis:
Componential analysis (section 3.1), diachronic schematic representation
(section 3.2) and cognitive model representation (3.3). Section 4 deals with
the influence of the term mRNA splicing on the continued existence of the
term gene splicing and section 5 goes into the possible impact of the meta-
phorical model on the naming of the splicing categories. Finally, in section
6 we state our conviction that terminological theory could gain from de-
tailed case studies like this one in its development of more diverse and
situated methodologies for terminological databases of different types. The
methodology developed in sociocognitive terminology for studying con-
ceptualisation in special languages for several languages, could at the same
time offer interesting case studies for cognitive studies.
2. The history of splicing
Not every native speaker of English is likely to know the words splicing
and to splice. Yet the word’s appearance in even less familiar idiomatic
expressions like to splice in marriage4 and to splice the main brace5, are an
indication for the complex history of the word. A lexicographer would not
consider the meaning components contained in these phrases to be the most
salient of the lexeme splice. Special language users will also have different
perceptions of the meaning constituents of this lexeme, according to
whether they are involved in sailing (rope splicing), film editing (film
splicing) or molecular biology and genetic engineering (mRNA splicing and
gene splicing). We can ask what it is in the nature of splicing that permits
its use for such a diversity of activities.
Lacking the experience of trained molecular biologists, but taking an
interest in the special language of this new discipline, we have been ex-
4. “splice in marriage, (colloq.) to join in marriage” (The New Oxford Illustrated

Dictionary 1981: 322).
5. “splice the main brace, (hist.) serve out extra rum ration on special occasion,
celebrate such occasion by drinking (said to be because splicing so thick a rope
would justify a special reward)” (The New Oxford Illustrated Dictionary 1981:
322).
cerpting relevant texts on the life sciences. In order to understand the terms
in these contexts, we have had to decode the terminology, i.e. to perform
the opposite process to what specialists in the life sciences do when they
are creating neologisms.6
We intend to show that precision in naming is often a problem because,
initially at least, the unit of understanding itself is not clearly delineated
and consequently, during the process of creating a new unit of under-
standing, its name is transitory as well. In reality, few units of scientific
understanding get delineated once and for all because phenomena based on
human activities are in constant evolution and so their understanding and
the linguistic means to refer to them are also further negotiated and un-
dergo change (see Frank this volume). Term formation occurs in a particu-
lar environment and situation that influences the naming activity itself, e.g.
in a laboratory. At some point, when a new insight occurs, when a tech-
nique is developed, or when a phenomenon is discovered, there is under-
standing of some kind. The subject specialists can describe and name what
they understand and create a term based on e.g. the intuitive meaning ex-
tension of an existing lexeme. This process will be illustrated on the exam-
ple of the naming history of two concepts both called ‘splicing’.
2.1. Two new units of understanding in the life sciences
The New Scientist of 24 April 1993 features a cover story celebrating the
40th anniversary of Watson and Crick’s discovery of the structure of DNA,
an event which marked the beginning of modern biotechnology. The cover
story highlights some landmarks in the history of DNA and genetic engi-
neering. Included in their survey are two events of special interest to our
present analysis.
6. In order to gain as precise an understanding as possible of “splicing” in the life

sciences, we have sifted a collection of English texts and stored contexts of the
concepts named “splicing” (including “to splice”, “a splice” and synonyms,
paraphrases, derivations and compounds). These examples were compiled from
different text types published since 1972, when foreign DNA was first spliced
into a plasmid. The text types include scientific papers and books, popularising
scientific literature in magazines and books, articles in general interest maga-
zines and newspapers (see bibliography). The most relevant examples are
quoted further on.
334 Rita Temmerman
First, in 1972,
By judicious use of restriction enzymes Herbert Boyer and Stanley Cohen
splice foreign DNA into a plasmid (a small DNA molecule often found in
bacteria) and slip it into the bacterium E. coli. They were opening the way
for cloning of any DNA in bacteria. (Kahn 1993: 24)
Second, in 1977,
Researchers realise that the genes of higher organisms are interrupted by re-
gions called introns, which do not carry instructions for assembling proteins.
Once a gene has been transcribed into messenger RNA, those unwanted
stretches of transcript have to be deleted in a process called mRNA splicing.
(Kahn 1993: 25)
These two statements indicate that in 1972 and 1977 two different units of
understanding were distinguished and both were named splicing. Let us
take a closer look at both discoveries.
2.1.1. The recombinant DNA technique gene splicing
Gene cloning is a process of genetic engineering with the following steps

(see fig 1).
1. The living cells are opened.

2. Genetic information is removed from the cell.
3. Specific DNA sections (genes) of interest are cut away from the
rest of the DNA.
4. These specific genes are inserted into, i.e. spliced into, plasmids
from bacteria.
5. The splicing process produces what is called a recombinant DNA
molecule.
6. The plasmid containing the specific genes is transferred into a
cell that is normally the host for the plasmid.
7. The host cell is allowed to multiply.
2.1.2. mRNA splicing (a step in protein synthesis)
Studies have confirmed that the general biochemical principles of life are
the same in humans and bacteria. However some remarkable differences
Figure 1. An overview of the gene-cloning process (Drlica 1984: 6).

336 Rita Temmerman
exist. Unlike bacterial DNA, human DNA (like all DNA in eucaryotes)
contains long stretches that do not code for anything (the introns). The
International Dictionary of Medicine and Biology (1986) defines mRNA
splicing as:
the natural process by which transcribed mRNA matures to become mRNA
that will be translated. The process involves excising transcribed intron re-
gions and rejoining the ends of each transcribed exon region.
The path from genes to protein has several steps (see fig. 2).
Figure 2. mRNA splicing (The British Medical Association 1992: 39).
1. A string of alternating exons and introns is transcribed.

2. The introns in the full length mRNA (unspliced mRNA) are ex-
cised, i.e. spliced out. Cells have mechanisms for removing the
introns.
3. Information in the exons is recombined, i.e. spliced together, to
form a new string: mature messenger RNA.
4. The processed mRNA is translated into protein.
2.2. Phases in the meaning extension process
Now that we roughly know what splicing refers to in biotechnology, we

can attempt to draw up a model which aims to illustrate the semasiological
meaning extension process of the lexeme splicing in English. This gives us
an indication of the basic domains of experience which may have served as
sources of inspiration in the naming act (see fig. 3) of both types of splic-
ing in the life sciences.
Use A to join by untwisting and interweaving the ends.

Rope splicing (OED 1524)
Use B to join by overlapping and securing the ends, e.g., pieces of timber,
metal girders or rails, concrete beams, etc. (OED 1626)
Use C to join film or tape.
Film (tape) splicing (OED 1912)
Use D D1: gene splicing (OED 1975)
D2: mRNA splicing (not in OED 1989 ed.)
Figure 3. Areas of use of “splicing”.
2.2.1 Splicing A: Rope splicing
The Oxford English Dictionary (1989) defines splicing as “the action or

operation of making a splice or splices”: a splice is “a joining or union of
two portions of rope, cable, cord, etc., effected by untwisting and inter-
weaving the strands at the point of junction (Chiefly Nautical)”.
To splice is
a) to join (ropes, cables, lines, etc.) by untwisting and interweaving the

strands of the ends so as to form one continuous length; to unite (two
parts of the same rope) by interweaving the strands of one end into
those of another part so as to form an eye or loop; to repair (rigging) in
this way. (Chiefly Nautical).
b) to form (an eye or knot) in a rope by splicing.
338 Rita Temmerman
The Encyclopedia Britannica (1979) has an illustration showing the se-

quence of steps in making a short splice (see fig. 4).
Figure 4. Rope splicing (The New Encyclopaedia Britannica in 30 Volumes 1979:

432).
2.2.2. Splicing B: Timber and metal splicing
In the OED (1989) splicing is also defined as: “the joining of two pieces of
wood, metal girders or rails, concrete beams, etc., formed by overlapping
and securing the ends; a scarf-joint” or “to graft by a similar process” and
“in various transferred and figurative uses: To unite, combine, join, mend.”
2.2.3. Splicing C: Film and tape splicing
To splice is also “to make a splice or joint in (a length of film or magnetic

tape); to join film or tape) in, on or up” (OED 1989).
The OED quotes F. A. Talbot, Moving Pictures (1912: xii, 137):
Occasionally when a film is being run through the projector it becomes sev-
ered by some means or other. Before it can be used again the break must be
repaired by splicing the two pairs together.
And L. Davidson (1978), Chelsea Murders, (1978: xxiii:141):

He put in six solid hours at the editing... He compared and cut and spliced
till two in the morning. (OED 1989)
A splice is “a joint made in editing or repairing film or magnetic or paper
tape.”
The Focal Encyclopedia of Photography (Stroebel 1969: 152) shows an
illustration (fig. 5).
Figure 5. Splicing cine film (Stroebel 1969: 152).
2.2.4. Splicing D1 (gene splicing) and D2 (mRNA splicing)
The OED (1989) does not (yet) mention mRNA splicing. To splice (for
gene splicing) is given under
c. In various transferred and figurative uses: to unite, combine, join, mend.

Also spec. in Biol., to join or insert (a gene or gene fragment). [...]
1975 Nature 18 Dec., 563/1 The genes to be cloned would first be spliced
on to either bacterial plasmid. or on to the DNA of bacteriophage lambda
which would then infect the bacterium. 1977 Sci. News 29 Jan. 70 The con-
troversial research in question is a class of experiments that include splicing
the genes of a virus or bacteria to partially purified DNA from mammals or
340 Rita Temmerman
birds… known to produce potent toxins or pathogens. 1979 Newsweek 4

June 64. One valuable product has already resulted from the work: human
insulin, manufactured by splicing fragments of DNA that manufacture the
hormone in humans into an intestinal bacterium.
The chronology indicates splicing A and B to be the domains of experience
for the analogical naming of the operation referred to in splicing C.
For splicing D1, it is most likely that rope splicing served as the source
domain for naming this technique gene ‘splicing’. Splicing C served on the
one hand as an educational metaphor (see Temmerman 2000), and on the
other hand as a more appropriate image for understanding splicing D2, in
which only one strand or ribbon or tape is involved. In some of the publi-
cations on molecular biology the film metaphor is explained.
In de Duve (1984: 307) splicing D2 is described as follows:
In a process reminiscent of cutting a movie film or editing a tape, a number
of segments are excised from the RNA ribbon and the remaining ones are
stitched back together.
Drlica (1984: 74) writes:
Since the gross aspects of information organization in DNA are easily de-
scribed by means of analogues between DNA and motion picture film, film
metaphors are used to begin describing gene cutting and splicing.
He shows the pictures we reproduce in fig. 6 to illustrate his point
Figure 6. Comparison of DNA and motion picture film (Drlica 1984: 74).
3. A combined method for analysis
In order to get more insight into the polysemisation process of splicing we

confront the results of a componential analysis (3.1) with the results of a
diachronic schematic representation (3.2) and of cognitive model repre-
sentations (3.3). The componential analysis gives a synchronic picture of
the senses of splicing whereas the schematic representation shows the dia-
chronic development. The cognitive model representations show the
meaning relationships.
3.1. A componential analysis reveals prototype structure
In trying to understand the polysemy of splicing one might wonder which

meaning components the five uses listed in figure 7 a have in common. By
componential analysis we split each sense into basic features, on the basis
of which the similarities and dissimilarities can be more easily understood.
A. B. C. D1 D2
Rope Wood and Film and Genes mRNA
metal tape
a Strands yes no yes (1) yes (2) yes (1)
(2 or
more)
b Overlap y/n y/n y/n yes no
c Repair y/n y/n yes yes yes
d Editing No no yes no yes
e Insert y/n no y/n yes no
f Loss of material y/n no yes no yes
g Human act Yes yes yes yes no
h Step 1: Yes y/n y/n yes yes
to separate
i Step 2: Yes yes yes yes yes
to rejoin
Figure 7 a. Contrastive componential analysis of five senses of “splicing”.

342 Rita Temmerman
On the level of senses (intensionally) we find salience effects, i.e. an inter-

nal structure of core and periphery. The core meaning component of all
senses of splicing (A, B, C, D1 and D2) is to rejoin. For some of the senses
the separation of the material prior to the rejoining (rope splicing) is im-
plied (A, D1 and D2), whereas for B and C the separations happened be-
fore the splicing and are the result of a weakness in the material.
The meaning of the lexeme splicing is a clustering of senses into family
resemblances and radial sets (see fig. 7 b). Splicing cannot be defined in
terms of necessary and sufficient characteristics since the only necessary
attribute which all the senses we considered have in common is ‘rejoining’.
To define splicing as ‘rejoining’ is incorrect as there are other ways of
rejoining (e.g. knotting) which are different from splicing.
rope splicing a (b) (c) - (e) (f )g h i

wood and metal splicing - (b) (c) - - - g (h) i
film and tape splicing a (b) c d (e) f g (h) i
gene splicing a b c - e - g h i
mRNA splicing a - c d - f - h i
Figure 7 b. The family resemblance of the lexeme “splicing”.
The foregoing analysis brings us to the conclusion that a traditional defini-

tion of gene splicing and mRNA splicing is impossible. The reasons are:
the prototype structure of both categories, the role encyclopaedic informa-
tion plays in the understanding of splicing, and the historical semasiologi-
cal evolution of the lexeme.
3.2. A diachronic schematic representation
A diachronic schematic representation in combination with the componen-

tial analysis can help us in finding which meaning components were diver-
sified and/or altered in the polysemisation history of a lexeme. The repre-
sentation can help us see which meaning changes occurred and which
subsets of meaning components were involved. This whole process appears
to be a good representation of negotiation and renegotiation of conceptuali-
sations of splicing across time and space (Sharifian this volume).
splicing: the joining

of two pieces of a
long-shaped object A. rope, cable, cord, etc.: the
ANALOGY:
(human act) joining of two pieces of a string-
FUNCTIONAL
like object by untwisting and in-
h ANALOGY: terweaving the strands (OED
i VISUAL
s B. timber, metal beam, etc.: the
t joining of two pieces by overlap-
o ping or scarfing the two ends to-
r gether (OED 1626)
i
c
C. film, tape, etc.: the joining of
a
two pieces of film or audio tape
l
ANALOGY: (OED 1912)
l INFORMATION
i D1. genetic material (DNA,
n RNA) (Boyer and Cohen, 1973):
e the joining of two pieces of a
strand of genetic material after the
ANALOGY: insertion of new genetic material
PROCESS
splicing: joining the D2. mRNA splicing (Berget,
pieces that are left 1977): the joining of the pieces of
after removing other a string-like object after removing
pieces (spontaneous introns
chemical process)
Figure 8. The meaning extension of splicing.
3.3. Cognitive model representations
Figure 8 illustrates two types of analogical understanding. The first type

occurs when the analogy between two different domains of experience
leads to new understanding resulting in metaphorical naming. In the second
type the analogy occurs within the same domain of experience. The new
naming can be an example of generalisation or specialisation as the case
344 Rita Temmerman
may be. We shall discuss two examples of analogical understanding and

naming: The gene splicing case (3.3.1), an example of metaphorical nam-
ing and the mRNA splicing case (3.3.2), an example of generalisation.
3.3.1. Metaphorical naming
Gene splicing is a case of metaphorical naming based on an analogy estab-

lished between two domains of experience. The source domain is the
structure of rope and the target domain is the structure of DNA.
The following quotes are just a few examples of the analogy between
the structure of DNA and the structure of rope:
− “the two strands of DNA form a double helix” (Lewin 1983: 32)
− “the whole DNA molecule consists of two such chains wrapped round
each other like strands in a rope, a ‘double helix’” (Bains 1987: 17)
− “double-helical DNA molecules are long flexible, threadlike structures”
(Berg and Singer 1992: 42)
− “intertwined strands” (Jones 1993: 40)
Figure 9 a shows the parts of the cognitive models which are highly analo-
gous.
source domain: structure of rope target domain: structure of genetic material

has has
rope 1 or 2 or x genetic 1 or 2
strands material strands
can be
can be can be
have can be have
spliced (double) helix spliced (double) helix

structure structure
Figure 9 a. A productive analogy resulting in the metaphorical naming of “gene

splicing”.
Other analogies which are used simultaneously in order to make the struc-
ture of DNA understood are: DNA IS A CHAIN and DNA IS A SPIRAL
STAIRCASE.
− “the density of DNA suggests that the helix must contain two polynu-
cleotide chains” (Lewin 1983: 28)
− “Each molecule of DNA consists of two distinct DNA strands joined by
weak hydrogen bonds to form a graceful tandem geometric structure.
This is the famous double helix discovered by Watson and Crick. Coil-
ing in parallel ascent, like a spiral staircase, each strand of the double
helix is composed of four kinds of molecular subunits called nucleo-
tides – each with a distinctively different shape. Each nucleotide con-
tains a sugar, a phosphate and one or four kinds of nitrogen-containing
bases: adenine (A), guanine (G), cytosine (C) and thymine (T). The
sugars and phosphates, linked end to end by strong chemical bonds,
form the spiraling double spine of the staircase. The bases, projecting
inwards from each spiral, are joined near the central axis by weaker
chemical bonds to create a soaring flight of stairs bridging the gap be-
tween the two strands.” (Suzuki and Knudtson 1988: 50)
− “the three-dimensional structure of DNA […] is a double helix. It has
been likened to a spiral staircase: the ‘banisters’ are composed of alter-
nate phosphate and sugar molecules, but the key to DNA is the ‘steps’
made of pairs of nitrogenous bases locked together crosswise.” (Hodson
1992: 97)
− “2 chains in a coiled embrace” (Levine and Suzuki 1993: 122)
The spiral staircase analogy, for instance, is not productive in the under-
standing and naming of gene splicing (see fig. 9 b).
source domain: spiral staircase target domain: DNA
spiral stair- 2 banisters DNA 2 banisters

case has has
Figure 9 b. An analogy that is not lexically productive for the understanding of

“gene splicing”.
346 Rita Temmerman
The metaphorical naming of gene splicing is reinforced by the film analogy

(see fig. 10 a).
source domain: film target domain: DNA
single se- is a sequence of

is a
film quence of DNA nucleotide
frames pairs
can be can be
constitute constitute
spliced a scene spliced a gene
Figure 10 a. The analogy between part of the film cognitive model and the DNA
cognitive model.
Splicing is as much part of the cognitive model of film-making as it is of

the cognitive model of rope-making. The analogy between, on the one
hand, rope and DNA, and, on the other, hand film and DNA is based on
different characteristics. It is the structure of DNA that is similar to the
structure of rope. Moreover, both a film tape and the “tape” of DNA are
perceived as carrying information, a fact that is essential for the analogy to
work (see fig. 10 b).
source domain: film target domain: DNA
nucleotide
film images DNA pairs
are part represent are part of
of represent
information information
Figure 10 b. How the reinterpretation of gene splicing as a domain metaphor with

“film splicing” as its source domain is reinforced by the information
analogy.
3.3.2. Generalisation or specification?
The naming of mRNA splicing is a case of naming based on analogy in the

same domain of experience, the domain of molecular biology. The compo-
nential analysis shows an important difference between gene splicing and
mRNA splicing. For gene splicing the agent is human, whereas for mRNA
splicing it is not. The analogy is in the process as represented in fig. 11.
This process consists of consecutive steps which are numbered chronologi-
cally in the schematic representation. What both processes have in com-
mon is that the genetic material is first cut by endonuclease and later an-
nealed by ligase.
source domain: molecular biology target domain: molecular biology
cuts eukaryotic
genetic 1 endonuclease RNA endonuclease
material 1
(DNA or has cuts
has
RNA) gets
2 exons introns
insert
3 anneals
3
ligase 2
anneals anneals
ligase
mRNA
gene splicing splicing
Figure 11. The analogy between “gene splicing” and “mRNA splicing” within
the same domain of experience.
Source domain and target domain coincide. Therefore this is not a case of
metaphorical transfer but a type of meaning extension. This could be
named specialisation or generalisation depending on the point of view one
takes. It is a case of generalisation in the sense that the lexeme splicing is
more widely used because it obtains more and more senses. The new sense
given to splicing (of mRNA), however, is only valid in a very specialised
348 Rita Temmerman
domain of experience and for that reason this procedure of meaning exten-
sion, based on an analogy within the same domain of experience, could be
named specialisation.
4. Implications for molecular biology
The detailed historical and diachronic analysis of the lexeme splicing pro-
vide us with a wealth of factual information which illustrates Bono’s claim
(1995, 2001) that metaphor can be studied in many more ways than as a
result of stable and physical embodiment. The example we described in
section 2 shows how metaphorical lexicalisations are “a contingent histori-
cal ‘tool’ which we use (and which ‘uses’ us) to approach, ultimately to
inhabit the unstable flux of things from which our world must emerge”
(Bono 2001: 225). It makes sense to redefine metaphor in language and
thought and action as a cognitive process having an embodied dimension
but our experience as bodies is situated in a (at various degrees) shared
physical, cultural and discursive world. As Sharifian (this volume: 109)
puts it, “Whatever the role of body in our cognitive life, it should be kept in
mind that conceptualisations of ‘body’ may be culture-specific and in gen-
eral body takes part and acts as a conceptual resource for our cultural expe-
rience”.
We might wonder whether insights like the above could or should have
an impact on the terminology of molecular biology. What could be the
effect of “metaphorical awareness raising” in the cultural group? In other
words, is the contingent historical tool which metaphor is believed to be
only working at a culturally distributed unconscious level? Does the crea-
tive power of metaphor cease to exist when attempts are made to con-
sciously use the “tool”? The example of the shift from the metaphorical
naming gene splicing to gene insertion can be observed in the archives of
texts on the subject. The same term splicing designates two different ac-
tivities in the domain of genetic engineering and molecular biology: Gene
insertion and mRNA cutting and pasting. The question is whether the ex-
istence of the same term for two concepts in molecular biology is not going
to cause confusion.
Gene splicing (D1) implies three steps (see fig. 1). Firstly, the plasmid
and the DNA-fragments are cut; secondly, the DNA-fragment is spliced
into (or inserted into) the plasmid; and thirdly, the ends are spliced to-
gether (or joined together). mRNA splicing (D2) (see fig. 2) is defined as a
process involving both excising introns (to splice out) and rejoining the
exons (to splice together). The overall process of excision and rejoining is
called splicing and the excision of the introns only is called to splice out,
while the rejoining of the exons is referred to as to splice together. The use
of the verb splice with different prepositions is visualised in fig. 12.
step 1 step 2 step 3

gene splicing to cut the plasmid and to splice the to splice the ends
the DNA fragment DNA fragment together
into the plasmid
mRNA splicing to cut the RNA to splice out the to splice the ends
introns of RNA together
Figure 12. Splice is used with several prepositions.
We witness a case of further extension of meaning of the verb splice which

can mean: a) to perform the overall process of gene splicing; b) to perform
the overall process of mRNA splicing; c) to splice a fragment of DNA into
a plasmid; d) to splice out introns (RNA); e) to splice together ends of
DNA or of RNA.
Careful study of a body of texts leads us to suggest that spontaneous
simplification processes are at work in the community of language users.
The first filtering occurs in the sub-domain of gene splicing (D1), where
the expression “to be spliced into” has its congruent or literal meaning in
“to be inserted into”, and is gradually replaced by it. We checked in some
of the first publications of the researchers on gene splicing (i.e. Hershfield
et al. 1974; Morrow and Berg 1972; Morrow et al. 1974; Cohen et al.
1973; Jackson, Symons and Berg 1972) and found the following instead of
“to splice”, “to insert” and “to cleave and ligate”:
[T]he restriction EcoRI was used to insert DNA fragments into the bacterial
plasmid pSC101, which served as the molecular vehicle. […] The […] DNA
is cleaved 11 times by the EcoRI endonuclease to produce 12 fragments
[…]. The EcoRI-generated fragments were ligated as described. (Hershfield
et al. 1974: 3455)
The fact that the specialists do not use “to splice”, but the (literal) syno-
nyms “to insert” or “to cleave and ligate” does not necessarily disclaim the
functioning of a metaphorical model of understanding. The reason for the
absence of “to splice” may lie in the fact that the writers of the articles (or
350 Rita Temmerman
even more likely the editorial board of the periodicals Proceedings of the
National Academy of Science in the USA and Journal of Virology) function
in an objectivist tradition which makes a point of replacing figurative lan-
guage by a literal equivalent. While some manuals use splicing for both
gene splicing and mRNA splicing, the more recent manuals use splicing
only for mRNA splicing (D2) while using insertion or a synonym, e.g.
building in for gene splicing (D1). In order to disambiguate gene splicing
and mRNA splicing, the first designation based on metaphor tends to be
replaced by some kind of hyperonym, e.g. gene insertion.
The confusion arising from the same term being used to refer to a) the
overall process of mRNA splicing, b) the first step of the process i.e. the
splicing out of introns and c) the second step of the process, i.e. the splic-
ing together of exons, seems to get filtered out in several ways in the most
recent publications (see fig. 13). The total process is always referred to as
mRNA splicing. But either the two steps are called e.g. “excision and
stitching back together” (de Duve1984: 308), “excision and suturing” (de
Duve 1984: 308) (this is a reference to surgery: “splice: a surgical proce-
dure or the site where a severed tissue, such as a tendon, is rejoined by
overlapping and suturing” (International Dictionary of Medicine and Biol-
ogy 1986)), “cleavage and ligation” (Lewin 1983: 407; Krainer and Ma-
niatis 1988: 131) (ligase being the enzyme responsible for the reaction),
“breakage and reunion” (Lewin 1983: 413) (the most general or unmarked
paraphrase); or splicing is used as a synonym for joining (so it names step
2), as in e.g. “splitting and splicing” (de Duve 1984: 310), “shedding of
introns and splicing” (de Duve 1984: 310), “the introns are chopped out of
[...] and the exons are spliced together” (Rennie 1993: 96), which suggests
the joining or repairing to be the strongest component in the meaning
analysis (see fig. 13).
So, either splicing is only used for the overall process of mRNA splicing
and the two steps are named differently, or the term splicing refers to the
overall process of mRNA splicing and for the second step in the process:
the joining. The shift from the whole process being named splicing to the
joining step being called splicing could be interpreted as a case of me-
tonymisation.
By studying metaphor as an analytical tool for a robust sociological ac-
count of science as a situated social activity (Bono 2001) we acquired in-
sight into how scientific conceptualisation (Sharifian 2003) functions. The
further history of naming is determined by regulatory processes inherent to
natural language evolution. These regulatory processes can be observed but
they are difficult to predict. One of the implications is that standardisation

of terminology in science has the advantage of streamlining the communi-
cative flow but the disadvantage of shortcutting the creative potential of
language as an emergent distributed cultural cognitive system (Sharifian
this volume).
process or mechanism: step 1: splicing out of introns

MrNA splicing step 2: splicing together of exons
two tendencies:
either: mRNA splicing = excision and stitching back together
or: excision and suturing
or: cleavage and ligation
or: breakage and reunion
or: mRNA splicing = splitting and splicing
or shedding of introns and splicing
or the introns are chopped out and the exons are
spliced together.
Figure 13. Paraphrases of “mRNA splicing”.
5. The metaphorical cognitive model
In section 2 we mentioned that gene splicing could be explained as a meta-

phorical naming with two possible source domains. In this section we de-
fend the hypothesis that the motivation for designating the insertion of
foreign genetic material into a plasmid by the term gene splicing is not
only the result of the rope analogy and the film analogy reinforcing one
another but also of the fact that the film analogy is part of the basic meta-
phorical model for a large part of present-day molecular genetics: the DNA
7
IS INFORMATION model, which we discussed in Temmerman (2000).
7. Evelyn Keller (2000) provides a fascinating historical account of various meta-

phors invested in past and present concepts of the gene. Unlike in the past our
modern discourse on heredity is pervaded by analogies with language and com-
puters. Yet, several biologists no longer believe in the existence of a non-
ambiguous entity that can be called a gene (Gayon 2000). In this alternative
352 Rita Temmerman
This metaphorical model was shown to have at least four submodels:

DNA IS A LANGUAGE WRITTEN IN A FOUR LETTER ALPHABET, DNA IS
AN ATLAS OF MAPS, DNA IS A TAPE OF FILM and DNA IS THE SOFT-
WARE OF THE CELL. In this metaphorical model a similarity is expressed
between inheritance information and well-known domains of information
representation and management, such as, written language, geography
maps, film tape and computer software. In the following quote we find
several traces of the submodel having become a cultural conceptualisation
(as shows from creative lexical variations on the basic metaphorical
schema): DNA IS A LANGUAGE WRITTEN IN A FOUR LETTER ALPHA-
BET.
DNA could be treated in many ways just as if it were a text written in Eng-
lish. […], DNA, RNA and protein, share one important feature with English.
These substances are made of subunits connected in a line, just as English is
written by placing a selection of letters, numbers and punctuation marks one
after the other in a row. DNA and RNA use only four letters, however, while
proteins use twenty. The letters in all of these biochemical substances run
continuously without spacing or punctuation [.…]. The single chromosome
present in the best-known bacteria has a letter content several times the
length of this book, while the smallest human chromosome, as we noted al-
ready, approximates a mammoth library dictionary. The problems in dealing
with such substances were simplified by the discovery of versatile new tools:
text cutters, text splicers and text-matching methods.
The names I have just used differ from the technical ones selected by
scientists. (Shapiro 1991: 79)
Shapiro’s didactic metaphor is an illustration of the metaphorical model
DNA IS INFORMATION within the framework of which the metaphorising
of splicing in this scientific domain should be understood. Shapiro’s text
cutters are known to specialists as restriction enzymes, his text splicers are
DNA ligases, his text matching is hybridization. It is not surprising that
Shapiro should come up with the metaphorical lexicalisations text cutter,
text splicer and text matching for categories which in specialists publica-
perspective, genes are important, and indeed ubiquitous in all forms of life as
we know it today; but they are not in themselves the “secret of life” (Stewart
2001). Ontogeny is accomplished not by the genes but by the organism as a
whole. The issue concerns “the epistemological foundations of biology − the
structural formalism of Mendelian genetics and neo-Darwinism versus an alter-
native organismic paradigm in which the notion of process is central” (Stewart
2001: 106)
tions are referred to by restriction enzymes, ligases and hybridisation. If

the submodel DNA IS LANGUAGE underlies the understanding of heredity,
then the link with how written language is processed nowadays in text
processors is easily made. In word processing programmes it is possible to
cut a particular fragment and to insert it elsewhere in a text.
The difference between this type of cutting and pasting and splicing B
and C is that the possible metaphorical lexicalisation for this aspect of
experience and understanding only occurs marginally. The British National
Corpus (http://www.natcorp.ox.ac.uk/) shows the following two examples
of splicing which may bear on language splicing:
“[…] for their effect on a kind of intimate twinning, a splicing between the
contexts on the basis of an actual or assumed […]”
“It consists of the splicing together of unrelated conversations […]”
Yet, even though the lexicalisation splicing for text in computer word
processing has only marginally become part of current English,8 it is con-
ceivable that the underlying analogy of cutting and pasting may have rein-
forced the conceptualisation which lead to the lexicalisation of mRNA
splicing.
The overall metaphorical environment concerning information, which is
at the basis of the new understanding of genetics, reinforces the explana-
tion of the choice for splicing in the naming of the unit of understanding.
This can be paraphrased as “the insertion of a piece of DNA into a plas-
mid”. As we mentioned in section 3, the discovery of mRNA splicing
threatens the survival of the term splicing for splicing D1 (gene splicing).
The following context shows that gene splicing is replaced by insertion
while splicing refers to mRNA splicing.
While no one has systematically constructed introns with altered sequences
and tested them as splicing substrates, it has been shown that the 21 nucleo-
tide insertion in a yeast tRNA Leu intron has no effect on splicing. (Cech
1983: 713)
The question remains as to whether splicing D2 (the excision of the introns
and the joining of the exons in mRNA) would ever have been referred to as
mRNA splicing if it had not been for the analogy with gene splicing (see
8. We checked eight recent manuals on text processing and found cutting and
pasting has become the technical term.
354 Rita Temmerman
section 2). It seems unlikely that this would have been the case. The suc-
cess of the metaphor lies in the reminiscence of the term to both the fact
that DNA looks like double stranded rope and the fact that splicing is part
of the word processing model. As RNA is one-stranded (also called a rib-
bon (de Duve 1984; Drlica 1984, 1992), the rope analogy seems less obvi-
ous than the film analogy. The film analogy in turn gets reinforced by the
text editing analogy. The history of splicing in its totality explains why the
D2 unit of understanding is named mRNA splicing. mRNA splicing is
called such because gene splicing has been named in this way. There is
analogy within the same domain of experience. We can not really see this
as metaphor because for metaphor there needs to be two distinct domains
of experience. What was known to be possible in a laboratory (gene splic-
ing) appeared to happen in nature as well. It is a case of meaning extension
of the term splicing, a consequence of further scientific discoveries.
6. Conclusion
In summary, the splicing case we have concentrated on shows the interplay

between historical, interlingual and cognitive aspects of metaphorical lan-
guage. Awareness of the system behind all of this has its consequences in
applied linguistics, more specifically in multilingual terminology descrip-
tions and knowledge representation. To recognise the creative force of the
language system, taking the sociocultural situatedness of language into
consideration, may have an impact on the management of dynamic termi-
nological resources which can account for meaning change (Temmerman
2002; Temmerman and Kerremans 2003). In short, this detailed study of the
lexeme splicing has enabled us to acquire some insight into meaning change
related to the study of domain specific languages: how the process of bor-
rowing terminology can take place, moving from one domain of experience
to the next (from rope splicing to film splicing to gene splicing and mRNA
splicing).
We have pointed out naming mechanisms based on analogy, i.e. how an
analogy established in the understanding of one domain of experience can
lead to an analogy with another domain of experience and result in the use
of the same terminology in both domains. When the analogy is seen be-
tween two different domains of experience we have cases of metaphorisa-
tion. This naming mechanism has to be distinguished from the conscious
use of metaphor. We have described metaphorisation and generalisation as
possible mechanisms of conceptualisation which go hand in hand with the

process of understanding. Something which is new and has not been named
before, except perhaps in paraphrase, is now named in a motivated way,
e.g. within the analogical model of understanding that has been established.
Thus, over time the terminology will emerge and stabilize, i.e. as a result of
the collective acts of the research community in question which in turn can
be viewed as a cultural group characterized by its own heterogeneously
distributed cognition (Sharifian 2003, this volume).
The understanding of metaphor happens through decoding at the word
level. Metaphorisation and generalisation are the result of encoding starting
from the analogical understanding of new categories. Initially, the resulting
name or term for the concept cannot be fully understood in its new mean-
ing without understanding the basis for the naming, i.e. without under-
standing the cognitive models and their sociocultural embeddedness. What
metaphorisation and generalisation have in common is that they are best
understood when the historically entrenched cultural conceptualisations
informing them are taken into consideration.
Acknowledgements
The author wishes to express her gratitude to Professor Roslyn Frank for
her advice and feedback on earlier versions of the article.
References
Arntz, Rainer and Heribert Picht

1989 Einführung in die Terminologiearbeit. Hildesheim: Georg Olms
Verlag.
Bains, William
1987 Genetic Engineering for Almost Everybody. London: Penguin Books.
Berg, Paul and Maxine Singer
1992 Dealing with Genes: The Language of Heredity. Mill Valley, CA:
University Science Books, Blackwell Scientific Publications.
Bernárdez, Enrique
culture
356 Rita Temmerman
Bono, James J.
ence. In: Stuart Peterfreund (ed.), Literature and Science: Theory
and Practice, 59–90. Boston: Northeastern University Press.
1995 Locating narratives: Science, metaphor, communities and epistemic
styles. In: Peter Weingart (ed.), Crossing Boundaries in Science,
119–151. Baden-Baden: Nomos Verlagsgesellschaft.
2001 Why metaphor? Toward a metaphorics of scientific practice. In:
Probing the Dynamics of Scientific Knowledge, 215–234. Bielefelt:
Transcript Verlag.
Calsamiglia, Helena and Teun Van Dijk
2004 Popularization discourse and knowledge about the genome. (Special
issue Genome Discourse. Guest ed. Brigitte Nerlich, Robert Ding-
wall and Paul Martin). Discourse Studies 15 (4): 369–389.
Cech, Thomas R.
1983 RNA splicing: Three themes with variations. Cell 34: 713–716.
Chadarevian, Soraya de
2002 Designs for Life: Molecular Biology after World War II. Cambridge:
Cohen, Stanley, Annie Chang, Herbert Boyer and Robert Helling
1973 Construction of biologically functional bacterial plasmids in vitro.
Proceedings National Academy of Science USA 70 (11): 3240–3244.
2002 Embodiment. In: Encyclopaedia of Cognitive Science 1102–1108.
London: Macmillan Publishers.
de Duve, Christian
1984 A Guided Tour of the Living Cell. New York: Scientific American
Books.
Drlica, Karl
1984 Understanding DNA and Gene Cloning. New York: John Willey and
Sons.
1992 Understanding DNA and Gene Cloning. New York: John Willey and
Sons. (second edition).
Felber, Helmut
1984 Terminology Manual. Vienna: Infoterm.
Felber, Helmut and Gerhard. Budin
1989 Terminologie in Theorie und Praxis. Tübingen: Gunter Narr Verlag.
Frank, Roslyn M.
this vol. The language-organism-species analogy: A complex adaptive systems
approach to shifting perspectives on “language”
Gayon, Jean
2000 The human genome project: Archaeology and prospected, cognitive
and practical. A commentary on “Is there an organism in this text?” In:
Phillip Sloan (ed.), Controlling Our Destinies: Historical, Philo-
sophical, Ethical and Theological Perspectives on the Human Ge-
nome Project, 291–299. University of Notre Dame Press.
Geeraerts, Dirk
1983 Reclassifying semantic change. Quaderni di Semantica 4 (2): 217–
240.
1985 Diachronic extensions of prototype theory. In: Geert Hoppen-
brouwers, Pieter Seuren and Anton Weijters (eds.), Meaning and the
Lexicon, 354–362. Dordrecht: Floris.
1992 Prototypicality effect in diachronic semantics: A round-up. In:
Günter Kellerman and Michael Morrissey (eds.), Diachrony within
Synchrony: Language, History and Cognition, 183–204. Frankfurt
am Main: Peter Lang.
1996 Diachronic Prototype Semantics. Oxford: OUP.
Hayles, N. Katherine
1999 How We Became Posthuman: Virtual Bodies in Cybernetics, Litera-
ture and Informatics. Chicago: University of Chicago Press.
Hershfield, Victor, Herbert Boyer, Charles Yanofsky, Michael Lovett and Donald
Helinski
1974 Plasmid Co1E1 as a molecular vehicle for cloning and amplification of
DNA. Proceedings National Academy of Science USA 71 (9): 3455–
3459.
Hodson, Anna
1992 Essential Genetics. London: Bloomsbury.
International Dictionary of Medicine and Biology
1986 New York: Wiley and Sons.
Jackson, David, Robert Symons and Paul Berg
1972 Biochemical method for inserting new genetic information into DNA of
Simian Virus 40: Circular SV40 DNA molecules containing Lambda
Phage Genes and the Galactose Operon of Escherichia coli’. Proceed-
ings National Academy of Science USA 69 (10): 2904–2909.
Jones, Glyn
1994 The quiet genius who decoded life. New Scientist 143 (8 October): 32–
35.
Kahn, Patricia
1993 Genome on the production line: 40 years of the double helix. New
Scientist 138 (24 April): 23–32.
Kay, Lily E.
2000 Who Wrote the Book of Life. Stanford: Stanford University Press
358 Rita Temmerman
Keller, Evelyn
2000 The Century of the Gene. Cambridge, MA: Harvard University Press.
Krainer, Adrian and Tom Maniatis
1988 RNA Splicing. In: B. David Hames and David Glover (eds.), Tran-
scription and Splicing, 131–206. Oxford and Washington DC: IRL
Press.
Lakoff, George
1987 Women, Fire and Dangerous Things. Chicago: University of Chi-
cago Press.
Levine, Joseph and David Suzuki
1993 The Secret of Life. Boston: WGBH Educational Foundation.
Lewin, Benjamin
1983 Genes. New York: John Wiley and Sons.
Morrow, John and Paul Berg
1972 Cleavage of Simian Virus 40 DNA at the unique site by a bacterial
restriction enzyme. Proceedings National Academy of Science USA 69
(11): 3365–3369.
Morrow, John, Steven Cohen, Annie Chang, Herbert Boyer, Howard Goodman and
Robert Helling
1974 Replication and transcription of eukaryotic DNA in Escherichia coli’.
Proceedings National Academy of Science USA 71 (5): 1743–1747.
Nerlich, Brigitte and Robert Dingwall
2004 Deciphering the human genome: The semantic and ideological founda-
tions of genetic and genomic discourse. In: Réné Dirven, Roslyn M.
Thought. Ideology, Metaphors and Meanings, 395–428. Berlin, New
Oxford English Dictionary
1989 2nd edition. Oxford: Oxford University Press.
Rennie, John
1993 Trends in genetics: DNA’s new twist. Scientific American (March):
88–96.
Sager, Juan Carlos
1990 A Practical Course in Terminology Processing. Amsterdam: John
Benjamins.
Shapiro, Robert
1991 The Human Blueprint: The Race to Unlock the Secrets of our Genetic
Script. New York: St. Martin’s Press.
Sharifian, Farzad
2003 On cultural conceptualisations. Journal of Cognition and Culture 3
(3): 187–207.
guage
Stewart, John
2001 Radical constructivism in biology and cognitive Science. (Special
issue) The Impact of Radical Constructivism on Science. Guest ed.
Alex Riegler) Foundations of Science 6 (1/3): 99–124.
Stroebel, Leslie and Richard Zakia
1969 The Focal Encyclopedia of Photography. Boston: Focal Press.
Suzuki, David and Paul. Knudtson
1988 Genethics. The Ethics of Engineering Life. London: Unwin Hyman.
Temmerman, Rita
1995 The process of revitalisation of old words: “Splicing”, a case study in
the extension of reference. Terminology 2 (1): 107–128.
2000 Towards New Ways of Terminology Description. The Sociocognitive
Approach. Amsterdam/Philadelphia: John Benjamins.
2002 Metaphorical models and the translation of scientific texts. Linguistica
Antverpiensia 1: 211–226.
Temmerman, Rita and Koen Kerremans.
2003 Termontography: Ontology building and the sociocognitive approach
to terminology description. In: Proceedings of the XVII International
Congress of Linguistics (24–29 July 2003) Prague: Universita Kar-
lova (CD–ROM).
The British Medical Association
1992 Our Genetic Future: The Science and Ethics of Genetic Technology.
Oxford: Oxford University Press.
The New Encyclopaedia Britannica in 30 Volumes
1979 Micropaedia IX. Chicago ;London ;Toronto ; Geneva ; Sydney ;
Tokyo ; Manila ; Seoul: Encyclopaedia Britannica, Inc.
Van Dijck, José
1998 Imagenation. Popular Images of Genetics. New York, New York
University Press.
Winterson, Jeanette
1994 Art and Lies. London: Jonathan Cape.
Wüster, Eugen
1991 Einführung in die allgemeine Terminologielehre und terminologis-
che Lexikographie. 3 Aufl. Bonn: Romanistischer Verlag.
360 Rita Temmerman
Ziemke, Tom
2003 What’s that thing called embodiment? In: In: Richard Alterman and
David Kirsh (eds.). Proceedings of the 25th Annual Meeting of the
baum.
Section D
Exploring the sociocultural situatedness

of language and cognition
Discourse metaphors
Jörg Zinken, Iina Hellsten and Brigitte Nerlich
Abstract
The article introduces the notion of discourse metaphor, relatively stable meta-
phorical mappings that function as a key framing device within a particular dis-
course over a certain period of time. Discourse metaphors are illustrated by case
studies from three lines of research: on the cultural imprint of metaphors, on the
negotiation of metaphors and on cross-linguistic occurrence. The source concepts
of discourse metaphors refer to phenomenologically salient real or fictitious objects
that are part of interactional space (i.e., can be pointed at, like MACHINES or
HOUSES) and/or occupy an important place in cultural imagination. Discourse
metaphors change both over time and across the discourses where they are used.
The implications of focussing on different types of source domains for our thinking
about the embodiment and sociocultural situatedness of metaphor is discussed, with
particular reference to recent developments in Conceptual Metaphor Theory. Re-
search on discourse suggests that situatedness is a crucial factor in the functioning
and dynamics of metaphor.
Keywords: conceptual metaphor theory, discourse analysis, discourse metaphors,

schematicity, sociocultural situatedness.
1. Introduction
This article introduces the notion of discourse metaphor to the cognitive

and social study of metaphor. By discourse metaphor we mean
a relatively stable metaphorical projection that functions as a key framing
device within a particular discourse over a certain period of time.
Examples of discourse metaphors are FRANKENFOOD, EUROPE IS A HOUSE,
NATURE IS A BOOK, or THE STATE IS A MACHINE.
In Cognitive Linguistics, metaphor has attracted immense interest as a
pervasive process of meaning creation. Highly schematic metaphorical
364 Jörg Zinken, Iina Hellsten and Brigitte Nerlich
mappings, motivated by the experience of correlations between sensori-

motor functioning and subjective judgement, have been hypothesised to be
at the core of much of human cognition. Examples of these so-called pri-
mary conceptual metaphors are KNOWING IS SEEING, STATES ARE LO-
CATIONS, AFFECTION IS WARMTH, or IMPORTANT IS BIG (Grady 1997; La-
koff and Johnson 1999).
In the following we want to explore the differences between discourse
metaphors and (primary) conceptual metaphors, and discuss their implica-
tions for our understanding of embodiment and sociocultural situatedness
within a cognitive linguistic framework. We believe that such a discussion
is useful, because ever more research on metaphor is carried out on the
basis of naturally occurring text and talk. Much of what seems central to
the study of metaphor in discourse (context-boundedness, strategical fuzzi-
ness, ideological bias) has, until recently, received little attention in the
cognitive linguistic literature, but things are beginning to change.1
The structure of this article is as follows. First, we provide a brief over-
view of various on-going case studies of metaphor in discourse (2). This
will illustrate some of the characteristics of discourse metaphors, and allow
us to explicate the differences in comparison to proposals about more
schematic mappings made in the literature. We discuss the cross-linguistic
occurrence of particular mappings, the ontological status of mappings on
different levels (phenomenological salience of discourse metaphors vs.
hypothetical status of primary metaphors), the evolution/life-span of par-
ticular metaphorical mappings, and the cultural imprint of metaphors (3).
We then discuss the implications of different approaches to metaphor for
our understanding of the embodiment of figurative language, and argue that
discourse metaphors provide evidence for the sociocultural situatedness of
metaphorical reasoning (4). Finally, we position our argument within the
wider discussion on the “dual grounding” (Sinha 1999) of human cognition
in the cognitive sciences (5).
1. Cf. Frank’s contribution (this volume) which explores various aspects of the
applicability of the discourse metaphor framework outlined here. See also
Zinken (in press) and Musolff and Zinken (in press).
Discourse metaphors 365
2. Case studies
This section is intended to highlight particular insights from discourse

studies on metaphor that are important to the objective of our article. Spe-
cifically, we wish to make three points that characterize discourse meta-
phors. These regard the cultural component in discourse metaphors, their
basis in interactional as well as individual cognitive processes (cf.
Bernárdez this volume), and their occurrence across languages. The sum-
mary of some case studies in this section will allow us to point out the dif-
ferences between discourse metaphors and primary metaphors. The results
described here stem from Nerlich’s research on cultural scripts in discourse
on animal and human diseases, such as FMD and SARS (see Larson, Ner-
lich and Wallis 2005; Wallis and Nerlich 2005; Larson this volume), Hell-
sten’s research on systemic and contested properties of metaphors in dis-
course (Hellsten 2002, 2003; Nerlich and Hellsten 2004) and Baranov and
Zinken’s research on the cross-linguistic comparison of metaphors as dis-
course practices (Baranov and Zinken 2003; Baranov and Zinken 2004;
Zinken 2004; cf. also Zybatow 1998).
2.1. Discourse metaphors employ cultural knowledge
Larson, Nerlich and Wallis (2005) have analysed media discourses sur-
rounding policies of biosecurity, implemented when nations or the world as
a whole are faced with biorisks, such as invasive species or invasive dis-
eases. The examples studied were foot and mouth disease (FMD) (an old
animal disease that broke out in the UK in 2001) and SARS (a new form of
pneumonia or flu which broke out in China in 2003 and spread to the
West). Metaphor schemas preserving a relatively high level of specificity
and relatively rich cultural knowledge in the source domain such as
HANDLING A DISEASES IS A WAR or A VIRUS IS A KILLER can be used in
these circumstances as a way of expressing a (preliminary) understanding
as well as evoking an emotive response. They can also be used to frame
policies intended to halt the spread of the disease in question. Using the
wrong policy framed by the wrong metaphor can have devastating social,
economic, psychological and animal welfare consequences. In such con-
texts the import of metaphor extends beyond individual cognition, into the
realm of society and culture.
Scientists and policy makers might use certain culturally available and
historically entrenched metaphors to frame scientific discoveries or poli-
cies, without initially reflecting on the wider implications their choices
might have – for instance the killing of around eight million animals in the
“war” against FMD. In this case a relatively harmless animal disease virus
that poses no risks to human health was framed, for mainly political and
economic reasons, as a deadly killer and invisible enemy that had to be
“stamped out” at all costs. This shows that “[m]etaphors, which entice us
‘to understand and experience one kind of thing in terms of another’ [...]
play a central role in the construction of social and political reality” (Annas
1995: 744, quoting Lakoff and Johnson 1980: 156). Using well-entrenched
metaphors and policies of war has, however, various drawbacks in the
framing of disease control programs, the most serious of which is perhaps
that “[m]ilitary thinking concentrates on the physical, sees control as cen-
tral, and encourages the expenditure of massive resources to achieve domi-
nance” (Annas 1995: 746).
The use of metaphors is not innocuous – it can have social costs and so-
cial benefits. It is therefore not only necessary to investigate the content of
a metaphor and ask
What does a particular metaphor express, and how? There is […] another
question that needs to be asked: How felicitous is a particular metaphor in a
particular context (e.g., solving a problem, obtaining consensus, elucidating
difficult subject matter, and so on)? (Mey 2001: 62)
This is a question asked by a discursive or pragmatic approach to meta-
phor.
Some of the most important schemas that can be used to “obtain con-
sensus” in certain socio-political and discursive situations are those of
CONTAINMENT, FORCE and BALANCE. In the discourses on FMD and SARS
we have found metaphors based on such (image) schemas. However, unlike
Lakoff and Johnson, who seem to embrace what some call an “unsituated
view of embodiment” (Bono 2001: 219; cf. also Zlatev 1997), we think
metaphors based on such schemas need to be explored in the cultural con-
text in which they are used, specifically, in terms of their sociocultural
situatedness.
As Paul Chilton has pointed out in an article on “The meaning of secu-
rity”: “Diseases are typically imagined as invading the body from outside, a
notion which rests both on the CONTAINER schema and the warfare script.”
(Chilton 1996: 197). Scripts and schemas interact to give metaphors dis-
cursive potency and to make certain metaphors plausible in certain situa-
tions. However, conceiving diseases as invaders and the control of disease

as war might not have always been the case in the past, and might not have
to be how we conceive of disease in the future (see Chilton 1996: 201). We
found for example that the “war” metaphor was much less used in the UK
media reporting on SARS than in the UK media reporting on FMD because
the metaphor was a more plausible framing device in the latter situation
than in the former (Wallis and Nerlich 2005). Discourse metaphors have a
social and cultural history and they influence social and cultural futures.
Take, for example, the metaphors of ‘balance’ and ‘warfare’ that have char-
acterized different epochs of medical thought in the West. The Hippocratic
and Galenic ideals of health as a balance of humours, or active bodily fluids
authorized a particular set of relationships between individual bodies, and
their external environment, and led to the cultivation of certain regimes of
bodily care and control. By contrast, the ‘embattled’ body of modern germ
theory adopts a quite different set of relations to its hostile external envi-
ronmental and enforces on itself – and on society more generally – a strin-
gent medicalized, socio-political regime. (Bono 2001: 225)
The study of conceptual metaphors has proliferated since the 1980s. How-
ever, cognitive linguists have rarely examined the repeated or continued
use of such metaphors in times of emotional turmoil or in times of scien-
tific or political uncertainty. This is a gap that needs to be filled if we want
to understand how general and local aspects of culture and cognition inter-
act in the ways people think and act in “the real world” (see Zinken, Hell-
sten and Nerlich 2003). Here, metaphor is frequently used not only to un-
derstand inherently unstructured abstract concepts, but also as a heuristic
device for exploring something global which is beyond normal comprehen-
sion and/or might directly threaten our health, well-being or survival.
More research is needed to find out whether in times of scientific or
political uncertainty, or during times of social upheaval, discourse meta-
phors, such as the WAR AGAINST DISEASES metaphor, become attractors for
cultural commonplaces, cultural myths and salient events of the past. This
“cultural” motivation of metaphor could be described as a kind of inter-
textuality (Zinken, Hellsten and Nerlich 2003). On the one hand, these
metaphorical and cognitive constants seem to be discursively embedded in
a relatively stable reservoir of cultural myths and social representations
available in social memory – e.g., memories of past wars and past epidem-
ics, or of weeds, plagues and displacements (see Cresswell 1997). On the
other hand, they can draw on knowledge of current social and political
events, such as, in the case of the fight against SARS, the concurrent “war
against terror”, the war in Iraq and threats of bioterrorism. SARS can there-
fore be conceptualised either as an ancient plague with all the old imagery
that surrounds this concept or as a “bioterrorism of nature” (Riddell 2003),
evoking much newer concepts and fears. Discourse metaphors seem to be
stable over long periods of time but they evolve and adapt to changing
socio-political circumstances (see Nerlich and Hellsten 2004; Musolff
2004, this volume; Nerlich 2005; Frank this volume). They might also tie
up with and reinforce long traditions of political thought, ideologies or
entrenched cultural values (see White and Herrera 2003: 277). Sustained
use of certain discourse metaphors contributes to giving a discourse or
discursive practice “its overall coherence and communicative edge” (ibid.).
2.2. Discourse metaphors evolve in historical time
Hellsten (2003, 2005) has been tracing the metaphors of FRANKENFOODS

and THE GENOME IS THE BOOK OF LIFE over time and across different dis-
courses. The FRANKENFOOD metaphor was used in agricultural biotechnol-
ogy discourse, while the BOOK OF LIFE metaphor was and is pervasive in
the medical biotechnology discourse (Kay 2000). The main point we wish
to make in this section is that the meanings of discourse metaphors co-
evolve with the cultures in which they are used. Conceptual metaphors are
considered universal, independent of time while discourse metaphors
change with the ongoing discourses (see also Musolff this volume), and are
used for specific purposes (Hellsten 2000).
Discourse metaphors reflect the cultural and social preoccupations of
the time. New topics and events are often discussed in terms of cultural and
mythical commonplaces; the target domain of the metaphor may be new
while the source domain is much older. The metaphor of GM-FOODS ARE
FRANKENFOODS, for instance, was coined only in 1992, while the source
domain, the myth of FRANKENSTEIN'S MONSTER has triggered people’s
imagination ever since Mary Shelley’s novel was published in 1818, and
has been used in various text traditions. The cultural image of scientists
creating potentially dangerous, new Frankenstein monsters in their labora-
tories is readily applicable to certain aspects of science and technology.
This image has been used in public debates on genetically modified foods,
for example. The metaphor of FRANKENFOODS gained its “momentum” in
Europe after 1996 as a reaction against the US import of genetically ma-
nipulated crops. In other words, it became a one-issue metaphor within the
debate on GM-foods. It was first used by environmental and consumer

related NGOs, in particular, Friends of the Earth, and spread to the mass
media in between 1998 and 1999. In the UK, the metaphor was also used in
the political decision making on GM-foods. The metaphor faded away
from the public agenda when the debate on GM-foods calmed down after
2000, but is still with us today. The UK tabloid newspaper The Daily Mail
still runs, for example, a so-called Frankenstein Food Watch campaign.
Successful discourse metaphors can resonate across a wide variety of
discourses, topics, and over time. The source domain of Frankenstein’s
monster can be mapped onto a wide variety of target domains, but it seems
to carry a relatively fixed set of associations and connotations with it, often
referring to the unpredictable negative outcomes of scientific activity.
Hence, it is readily made use of in new cultural situations, such as the in-
troduction of genetically manipulated crops into the European markets.
Because of this relatively stable set of associations, the metaphor of Frank-
enfoods and the related metaphors of Frankenfish, Frankencorn and
Frankenmilk can be effectively used to call for action against GM foods in
general or against particular types of GM food.
In a similar way, the metaphor of THE GENOME IS THE BOOK OF LIFE,
widely used in the debate on genomics, is both novel and old. The source
domain of THE BOOK OF LIFE has been in use ever since Antiquity, and has
a long history within the Judeo-Christian tradition where it refers to natu-
ral, eternal and universal texts (Kay 2000: 31).2 In the Book of Revelation,
the names of those to be saved from the Apocalypse are written in the
“book of life”. Parallel to the BOOK OF LIFE runs the BOOK OF NATURE,
common in the history of the natural sciences, where science was perceived
as an effort to read and write the book of nature. For Galileo, the book of
nature was written in the language of mathematics (Cohen 1994).
According to Kay (2000), the metaphor of the BOOK OF LIFE gained its
current scientific legitimacy in the debates on genetics when it was con-
nected to the discourse of information: genes carry the information, the
instructions for the formation of organisms. The connection between cel-
lular systems and the alphabet first became popular in the 1960s when
molecular biologists started using the metaphor for understanding the
2. The metaphor would deserve a more detailed study, which might investigate the
emergence of a discourse metaphor from “errors” or “variations” in translation.
The phrase “book of life” in Revelation has a contested history as it migrated
from various Latin versions of the Bible to its English instantiation.
working of the DNA (van Dijck 1998: 123), as composed of the four nu-
cleotides represented by their initial letters, adenine (A), thymine (T), cyto-
sine (C) and guanine (G). A, T, C and G became the alphabet of life. Dur-
ing the Human Genome Project, between 1990 and 2003, the metaphor was
effectively used to promote the research project to increase public aware-
ness of the research project, and ever new formulations of it are now being
used to promote post-genomic research (Hellsten 2005). The metaphor has
co-evolved with the genome project, from discussing DNA in terms of the
alphabet to comparing the different genomic books of humans and mice,
for example.
Resonance over time, across topics and across different domains of use
in society makes discourse metaphors apt tools for communication. The
metaphor of “the book of life”, for instance, has moved diachronically from
the Bible to modern sciences and to the genome, in particular from gene
sequencing to genome annotating, and from lexical to semantic structures.
Across topics, the metaphor has been used in the debates on genetics and
genomics as well as in the debates on biodiversity where nature is some-
times considered as “the library of life” (Väliverronen and Hellsten 2002).
The metaphor has also provided resonance across the different societal
domains that participate in the debates, such as the sciences, the social
sciences and the mass media (see also Hellsten 2000). The metaphors of
FRANKENFOOD and THE BOOK OF LIFE carry familiar cultural images
(Frankenstein myth and apocalypse myth) that gain negative or positive
resonance when reformulated to fit into new contexts (GM foods and ge-
netics/genomics).
Discourse metaphors evolve as part of communication and text tradi-
tions, in the social use of the metaphors. Some of these metaphors become
narrative metaphors3 (NATURE IS AN OBJECT; NATURE IS A BOOK) and gain
a very prominent position within a given culture while other, one-issue
metaphors have a shorter life-span (FRANKENFOOD).
Thus far we have focussed on diachronic aspects of discourse meta-
phors, but discourse metaphors can be traced synchronically as well by
comparing the width of discourses that use a certain metaphor as a key-
3. The concept “narrative metaphor” is introduced in Hellsten (2002) and refers to

very strongly entrenched metaphorical cultural models. Where discourse meta-
phors may vary from one-issue metaphors, i.e., metaphors such as Frankenfood,
that are purposefully coined to advance certain interests at the expense of oth-
ers, to more general metaphors such as “nature is a book”, narrative metaphors
often provide wider cultural views on the issue.
grasp. The novel metaphor STEM CELLS ARE LIFE’S MAGIC CAULDRON is
probably part of a specialist discourse, and therefore is much more re-
stricted than CLONES ARE COPIES, which, in turn, is more restricted than
NATURE IS A BOOK. But this novel metaphor links the new phenomenon of
stem cells to old cultural knowledge about magic and miracle.
Discourse metaphors are communicative and cultural tools, and as such
potentially more variable than the highly schematic mappings proposed
within Conceptual Metaphor Theory (CMT).
2.3. Discourse metaphors occur across languages
Baranov and Zinken (2003, 2004) have conducted cross-linguistic research

on the metaphors used in Russian and German newspapers to talk about
political transformations in the two countries in the late 1980s and early
1990s. In this project the focus was on source domains, the goal was to
compare the scope of target domains understood via a particular source
domain in these languages, as well as investigating the similarities and
differences within source domains between the languages.
Russian and German are genealogically and typologically related Indo-
European languages, and the cultural heritage shared within the area is
considerable. It is therefore not surprising that there is a vast number of
metaphors that are common in both Russian and German public discourse.
However, there is also considerable diversity.
Firstly, there is diversity in the use of metaphors for a specific target
domain. E.g., in the Russian media, the discourse metaphor for the target
domain transformation was TRANSFORMATION IS REBUILDING (perestro-
jka), the source domain being BUILDING. In the German discourse, the dis-
course metaphor for the same target domain was TRANSFORMATION IS A
TURN (Wende), using the source domain MOVEMENT. Obviously,
BUILDING-metaphors are also documented in the German discourse, and
MOVEMENT-metaphors are documented in the Russian discourse. However,
both quantitative and qualitative data suggest that understanding socio-
political change as a MOVEMENT was a discourse practice in the German
discourse of the time whereas it wasn’t one in the Russian discourse, and
conversely understanding socio-economic change in terms of (RE-)
BUILDING was a discourse practice in the Russian, but not in the German
media (Baranov and Zinken 2003).
Secondly, there are differences in the use of particular source domains.

As an example, let us look at the source domain FLORA. Table 1 shows part
of the semantic frame of the metaphor model FLORA:4
Table 1. Compiled sections of the semantic trees for the metaphor model FLORA.
Descriptor/Russian database Descriptor/German database ‘English gloss’
FLORA (LEVEL 0) ‘FLORA’
1.1. Flora (level 0)
ЧАСТЬ РАСТЕНИЯ-ДЕРЕВА TEIL EINER PFLANZE/EINES ‘PART OF A

(LEVEL 1) BAUMS (LEVEL 1) PLANT/TREE’
ветка-ветвь (level 2) Zweig (level 2) ‘branch’
разветвление (level Verästelung (level 3) ‘ramification’
3)
корень (level 2) Wurzel (level 2) ‘root’
As can be seen in table 1, both the Russian and the German discourse use
the concepts of a branch and of roots on level 2 in the subdomain hierarchy
metaphorically in public discourse. Furthermore, both languages have the
same metaphors, mapping branches onto economic domains, so that differ-
ent branches of an economy denote different economic domains, and map-
ping roots – as in her political roots – onto traditions or the beginning of a
political process. These are typical discourse metaphors in the sense of our
definition in (1). However, ramification, a subdomain of branch, is used in
the German corpus only, where it is a common metaphor in talking about
the target domain of questionably close institutional links. (e.g., between
companies). Not only is there no metaphor CLOSE INSTITUTIONAL LINKS
ARE RAMIFICATIONS in the Russian corpus, there is no mapping whatsoever
of X ARE RAMIFICATIONS with X being any target domain. In other words,
the gaps in discourse mappings vary cross-linguistically. Whereas ramifi-
cations as part of the domain branch is mapped onto the domain of econ-
omy in German discourse, such a mapping does not occur in Russian dis-
4. Inverted letters mean that no metaphor in the corpus was coded in the database
using the respective descriptor. Small Caps indicate level 1 in the subdomain
hierarchy, indentations indicate level 3.
course. This is a problematic case for attempts to account for the details of
complex metaphors by reducing them onto hypothesized universal primary
metaphors.
Generally speaking, cross-linguistic occurrence on levels 1 and 3 in the
subdomain hierarchy turned out to be more restricted than on level 2. This
allows for the hypothesis that basic level concepts (Rosch et al. 1976) are
cross-linguistically more salient as metaphorical source concepts than con-
cepts on the superordinate and subordinate domains. One implication of
this is that lexicalisation patterns have to be given more prominence in
accounting for the motivation of metaphor (cf. Evans 2004 for a related
plea). E.g., the fact that there is a ramification-metaphor in the German
corpus (the German word is Verästelung) but not in the Russian one is
probably best explained by the productivity of the German prefix Ver- in
metaphorical meaning extension.5
It is important to point out that we did not a priori claim a link between
the hierarchy of (sub-) domain levels and the notion of discourse meta-
phors. The nesting of levels within a domain is a cognitive phenomenon
that is part of conceptualisation (Croft 2003; Langacker 1987). Discourse
metaphors were initially defined in social terms: they are mappings that
regularly appear in discourse on the actual linguistic “surface” (social sta-
bility), which indicates a certain phenomenological salience of discourse
metaphors to speakers. Although these factors (social stability and middle
level in subdomain hierarchy) are logically independent, there does in fact
seem to be a relation between them, so that discourse metaphors usually
make use of source concepts from the middle level of categorisation.
3. Embodiment and metaphor theory: discourse metaphors and

primary metaphors
The latest elaborations of Conceptual Metaphor Theory (Grady 1997; La-

koff and Johnson 1999; Grady and Johnson 2003) emphasize that metaphor
is a strong source of evidence for the embodiment of cognition. In this
context, both the terms metaphor and embodiment are understood in a par-
ticular way. In this and the next sections, we will briefly spell out our
reading of this understanding of metaphor and embodiment. As we do so, it
will become clear that the type of metaphor we have found to be most sali-
5. We would like to thank René Dirven for drawing our attention to this.
ent in discourse in section 1 (discourse metaphors) differs in various ways

from primary metaphors. We argue that the particular characteristics of
discourse metaphors add to our understanding of the phenomenological
aspect of embodiment and our understanding of the relation between em-
bodiment and sociocultural situatedness.
We mentioned three case studies in order to make three points about
discourse metaphors: they use knowledge associated with basic level con-
cepts; they evolve in social interaction; and they are firmly linked to cul-
tural scripts and stereotypes. These three findings taken together seem suf-
ficient to make the claim that discourse metaphors are a distinct
phenomenon that needs to be accounted for in a cognitive theory of meta-
phor.6
3.1. Phenomenological salience
As mentioned at the beginning of the article, examples of primary meta-

phors are KNOWING IS SEEING, STATES ARE LOCATIONS, AFFECTION IS
WARMTH, or IMPORTANT IS BIG. These are regarded as primary in two
senses. They are primary in the sense that they are the first conceptual
metaphorical mappings acquired in childhood as a result of recurrent cor-
relations between sensori-motor experience and subjective judgement of
this experience. But they are also regarded as primary in the sense that all
or nearly all the metaphors that we use or hear in communication are
thought to be derived from a relatively small set of these primary meta-
phors. As the term primary metaphor indicates, in this line of research it is
this type of metaphor that is regarded as most important, or basic in under-
standing the cognitive functions of metaphor:7 We need to know how pri-
mary metaphors work, if we want to understand why we have the secon-
dary or tertiary metaphors that we do.
An example of this approach is Grady’s analysis of the proposed meta-
phorical mapping THEORIES ARE BUILDINGS (Grady and Johnson 2003).
6. By a cognitive theory of metaphor we mean any theory that tries to account for
the role of metaphor in conceptualisation, as opposed to the term Conceptual
Metaphor Theory relating to the school of Lakoff and colleagues.
7. Consequently, Özçalişkan (2003) uses the term “basic level” to denote the level
of primary metaphors and primary scenes in the sense of Grady and Johnson
(2003). Here, the term is used referring to a middle level of conceptualisation
(e.g., Rosch et al. (1976)).
Grady noticed that there are gaps in this mapping, i.e., not all of our expe-
rientially acquired knowledge about buildings is mapped onto theories –
e.g., we don’t conventionally talk about a theory having no windows.
Talking about theories in terms of buildings (e.g., laying the foundations
for a new theory) is a complex mapping that is constrained by two primary
metaphors: PERSISTING IS REMAINING ERECT and ORGANISATION IS
PHYSICAL STRUCTURE. Only the inferential patterns of primary metaphors
are mapped onto the complex metaphor.
A similar reduction of discourse metaphors onto primary metaphors
would surely be possible. However, while discourse studies do not provide
any evidence against the possibility of the existence of primary metaphors,
they certainly do not suggest that discourse metaphors are motivated by
such simpler mappings. In fact, there are reasons for claiming that in dis-
cursive reasoning, arguing and framing, metaphors like BELONGING IS
HAVING ROOTS, CLONES ARE COPIES, NATION-STATES ARE HOUSES etc. are
the basic imaginative acts. One reason for this is that discourse metaphors
are, as illustrated above, very frequent and cross-culturally wide-spread,
while the link between hypothesised abstract metaphor schemas like
PERSISTING IS REMAINING ERECT and observable linguistic behaviour is
much weaker. Notions of belonging, cloning and nation-states are in the
very focus of discourse, while a general notion like persisting never is.
Framing belonging as rootedness, cloning as copying, or nation-states as
houses is contested in discourse, framing persisting as remaining erect
never is.
But the point is not just to say that more specific phrasings of a map-
ping are more likely to appear on the linguistic surface of text and talk than
very abstract generalisations. The important point is that it is possible, by
means of looking at the linguistic surface, to identify a level of conceptual
projection from a source domain that seems to be most likely to become
entrenched in a discourse and that is most stable cross-linguistically,
namely the level of discourse metaphors, based on source concepts from
the basic level of categorisation. Moreover, the source domains of dis-
course metaphors have a high degree of phenomenological salience, while
the source domains of primary metaphors don’t. Of course, this is just what
exponents of conceptual metaphor theory would say, and we will discuss
the implications for a particular understanding of the embodiment of cog-
nition below.
3.2. Evolution in historical time
As illustrated in 2.2., discourse metaphors evolve over historical time in

social interaction. Some have a short life span, like FRANKENFOOD, some
stay on to become entrenched cultural models, like NATURE IS A BOOK or
DISEASES ARE INVADERS. What they share is that they “live” in the semio-
sphere (Lotman 1990). Individuals encounter them in discourse, take them
up, modify or reject them. They become part of situated discursive and
narrative practices.
Primary metaphors, by contrast, are hypothesised to be acquired as the
result of non-semiotic experience. This view of metaphors as fundamental
conceptual structures does not account for the flexible evolution of meta-
phors in use. It should be pointed out that there seem to be some differ-
ences between older versions of Conceptual Metaphor Theory (Lakoff and
Johnson 1980; Lakoff 1987, 1993), and newer elaborations (Grady 1997;
Lakoff and Johnson 1999; Grady 1999; Grady and Johnson 2003). In older
versions, social organisation and culture are mentioned as one relevant
aspect of experience:
In other words, what we call ‘direct physical experience’ is never merely a
matter of having a body of a certain sort: rather every experience takes place
within a vast background of cultural presuppositions. (Lakoff and Johnson
1980: 57)
In more recent elaborations, cognitive development and the acquisition of
primary metaphors are more explicitly modelled as an individual endeav-
our. Discourse data suggest that an account of individual metaphorical
reasoning needs to take into consideration the interactional negotiation of
perspectives and the entrenchment of projected perspectives in terms of
concept elaboration (Evans 2004).
3.3. Cultural component
Conceptual Metaphor Theory is predominantly interested in universal as-

pects of metaphor. Primary metaphors are more or less explicitly claimed
to be universal (Lakoff and Johnson 1999; Grady and Johnson 2003;
Özçalişkan 2003). Such a claim is not made with respect to complex meta-
phors, but the culture-specific component should be irrelevant, if, as
claimed in CMT, the metaphorical potential of complex metaphors is re-
stricted to the inferential patterns of the primary metaphors constraining

them.
We would argue that relatively rich images resulting from our cultural
experience and interaction(s) with the world lie at the heart of metaphorical
reasoning in discourse. A classic example of the sociocultural situatedness
of metaphorical reasoning is Chilton and Ilyin’s (1993) discussion of the
metaphor EUROPE IS A HOUSE in Russian and Western European discourse.
Chilton and Ilyin show that the differences in the stereotypes of a house in
Russia and in some Western European countries led to vastly different
inferences in envisaging European politics. The metaphor model FLORA
also shows that sociocultural values and traditions are not irrelevant to
metaphorical reasoning in the real world: the fact that FLORA-metaphors –
as opposed to ORGANISM-metaphors – usually throw a positive light on the
target domain in Russian and German discourse is explained best by the
romantic tradition in European culture, which entrenched an idyllic, and
generally speaking a positive picture of nature.
CMT is vague with respect to a point that is important in this context.
There does not seem to be a clear stance on whether in reasoning meta-
phorically, we carry out online-extensions within the source domain or not.
In other words, in talking of a well-founded theory, does the hypothesised
primary metaphor ORGANISATION IS PHYSICAL STRUCTURE become acti-
vated every time? Or is the primary metaphor regarded as the diachronic
starting-point of building-metaphors, without being necessarily accessed
when we reason about theories in terms of this metaphor (cf. Gibbs 1999)?
The first position would involve a strong grounding of reasoning in uni-
versal aspects of conceptualisation (because primary metaphors are mod-
elled as universal). The second position would mean that the actual dis-
course metaphors entrenched in a community are the tools of our reasoning
– that we reason in terms of the culture-specific information-chunks en-
trenched in symbols rather than breaking down these chunks into their
possibly universal pieces.
In summary, the major difference between primary metaphors and dis-
course metaphors lies in the type of source domain regarded as basic for
metaphorical activity in the two approaches. This issue will be discussed in
the context of notions of embodiment in the next section.
4. (How) are discourse metaphors embodied?
Embodiment is understood in a variety of ways within the cognitive sci-

ences (Ziemke 2003). In the approach most influential in metaphor theory,
three main aspects of the embodiment of cognition are distinguished: neu-
ral embodiment, embodiment on the phenomenological level, and embodi-
ment in the cognitive unconscious (Lakoff and Johnson 1999: 102ff.).
As has become evident in the above discussions, conceptual metaphor
theory prefers to treat metaphor as evidence for the grounding of cognition
in the cognitive unconscious. The entities regarded as “basic” in under-
standing metaphor as a cognitive phenomenon are hypothesised entities
located in the cognitive unconscious: image schemas and primary meta-
phors.8
We have tried to illustrate that discourse studies provide evidence for
the sociocultural situatedness of metaphorical reasoning. The basic entities
are source domains which are associated with rich images of (real or ficti-
tious) objects salient in the cultural Umwelt. Whereas the source domains
of primary metaphors are very abstract (as in PERSISTING IS BEING ERECT,
ORGANISATION IS PHYSICAL STRUCTURE), the source domains of discourse
metaphors are part of the interactional and cultural space: material objects
that can be touched and pointed at (e.g., A NATION-STATE IS A HOUSE, THE
STATE IS A MACHINE) or concepts that have a strong cultural image or value
attached to them, due to textual, semiotic traditions (as in SOCIETY IS AN
ORGAN, REVOLUTIONARY EVENTS ARE A STORM, BELONGING IS HAVING
ROOTS, GRATEFULNESS IS BEING IN DEBT). Discourse metaphors provide
evidence for the cognitive usefulness of the culturally accumulated knowl-
edge entrenched in (source domain) symbols.
Lakoff and Johnson (1999) point out that all three levels of embodiment
(the neural, the phenomenological and the cognitive unconscious) must be
kept in view if we want to account for the embodiment of language and
cognition. Surely many aspects of language and cognition are unconscious.
However, it seems to us that more attention to the phenomenological level
of embodiment and to sociocultural situatedness could make some hypo-
thetical assumptions about entities in this unconscious realm superfluous
(Zlatev 2002, 2007)
8. Cf. also the discussion of the “cognitive unconscious” in Zlatev (2007).

5. Metaphor, embodiment and dual grounding
The main point of this article in the context of this volume has been to
argue that an explanation of the functioning and dynamics of metaphor
needs to address not only the embodiment of cognition, but also the
empowerment of cognition through symbols (Tomasello 1999; Gentner
2003). The supplementation of the individualist view on cognition in Con-
ceptual Metaphor Theory with a socioculturally situated view is the objec-
tive of several of the articles in this volume. The ultimate goal is an ac-
count of the “dual grounding” (Sinha 1999) of human cognition in both
biology and culture – to account for the fact that human cognition, like all
animal cognition, is constrained by biology, but that it is, unlike other ani-
mal cognition, not bound by the skin (Bateson 1972). We have tried to
make a step into this direction by arguing that:
1. in discourse metaphors, knowledge associated with basic level catego-

ries is projected onto the target domain;
2. the conventionalisation of a particular projection into a discourse prac-
tice is a socio-cultural process;
3. discourse metaphors therefore provide evidence for the social situated-
ness as well as the phenomenological embodiment of metaphor.
If metaphorical thought fundamentally involves the images and feelings

embedded in our culture, then this highlights another aspect of embodi-
ment, one which seems fundamental to human cognition: embodiment as a
process, the process of incorporating the symbolically accumulated ideas
and values of our fellow men and ancestors (Bourdieu 1977; Tomasello
1999). A focus on embodiment might therefore usefully be supplemented
by a focus on enculturation. The term enculturation is normally used to
describe the adoption of the behaviour patterns of the surrounding culture
or the socialisation of children to the norms of their culture, but this term
could also be used to describe the adoption of certain metaphorical patterns
for thinking about the world, acting in the world, for imagining the past
and future and for framing current crises. As Clifford Geertz wrote in his
1973 collection of essays and ethnography, The Interpretation of Culture:
[C]ulture is best seen [...] as a set of control mechanisms – plans, recipes,
rules, instructions [...] – for the governing of behavior. [And] man is pre-
cisely the animal most desperately dependent upon such extragenetic, out-
side-the-skin control mechanisms, such as cultural programs, for ordering
behavior (p. 44). [While these ideas are not new], the results of recent re-
search have made them susceptible of more precise statement as well as
lending them a degree of empirical support they did not previously have. [...]
The control mechanism view of culture begins with the assumption that hu-
man thought is basically both social and public – that its natural habitat is
the house yard, the marketplace, and the town square. Thinking consists not
of ‘happenings in the head’ but in a traffic in what has been called by G. H.
Mead and others, significant symbols – words for the most part but also
gestures, drawings, musical sounds, mechanical devices like clocks, or natu-
ral objects like jewels – anything in fact that is disengaged from its mere
actuality and used to impose meaning on experience [...]. (Geertz 1973: 45)
To this list we would add discourse metaphors.
Acknowledgements
Brigitte Nerlich’s work on this article was supported by the Leverhulme

Trust and the ESRC (grant number: L144 25 0050). Iina Hellsten’s and
Jörg Zinken’s work on this article was supported by the British Academy.
References
Annas, George
1995 Reframing the debate on health care reform by replacing our meta-
phors. New England Journal of Medicine 332: 744–747.
Baranov, Anatolij and Jörg Zinken
2003 Die metaphorische Struktur des öffentlichen Diskurses in Russland
und Deutschland: Perestrojka- und Wende-Periode. In: Bernhard
Symanzik, Gerhard Birkfellner and Alfred Sproede (eds.), Metapher,
Bild und Figur, 93–121. Hamburg: Verlag Kovac.
2004 Blühende Länder, wuchernde Staaten. Das Metaphernmodell FLORA
im russischen und im deutschen Transformationsdiskurs. In:
Symanzik et al. (eds.), Sprache, Literatur, Politik. Ost- und
Südosteuropa im Wandel, 15–31. Hamburg: Verlag Kovac.
Bateson, Gregory
1972 Steps to an Ecology of Mind. Chicago: University of Chicago Press.
Bernárdez, Enrique
culture.
Bono, James J.
2001 Why metaphor? Towards a metaphoric of scientific practice. In:
Probing the Dynamics of Scientific Knowledge, 215–234. Bielefeld:
Transcript Verlag.
Bourdieu, Pierre
1977 Outline of a Theory of Practice. Cambridge, UK: Cambridge Univer-
sity Press.
Chilton, Paul
1996 The meaning of security. In: Francis A. Beer and Robert Hariman
(eds.), Post-Realism: The Rhetorical Turn in International Relations,
193–216. East Lansing: Michigan State University Press.
Chilton, Paul and Mikhail Ilyin
1993 Metaphor in political discourse. The case of the ‘Common European
House’. Discourse and Society 4 (1): 7–31.
Cohen, H. Floris
1994 The Scientific Revolution. Historiographical Inquiry. Chicago and
London: University of Chicago Press.
Cresswell, Tim
1997 Weeds, plagues and bodily secretion: A geographical analysis of
metaphors of displacement. Annals of the Association of American
Geographers 87 (2): 330–345.
Croft, William
2003 The role of domains in the interpretation of metaphors and metony-
mies. In: René Dirven and Ralf Pörings (eds.), Metaphor and Meton-
ymy in Comparison and Contrast, 161–205. Berlin: Mouton de
Gruyter.
Evans, Vyvyan
2004 The Structure of Time. Language, Meaning and Temporal Cognition.
Amsterdam/Philadelphia: John Benjamins.
Frank, Roslyn M.
Gentner, Dedre
2003 Why we’re so smart. In: Derdre Gentner and Susan Goldin–Meadow
(eds.), Language in Mind. Advances in the Study of Language and
Thought, 195–235. Cambridge, Mass: MIT.
Geertz, Clifford
1973 The Interpretation of Cultures. New York: Basic Books.
Gibbs, Raymond
world. In: Raymond W. Gibbs, Jr. and Gerard J. Steen (eds.), Meta-
phor in Cognitive Linguistics, 145–166. Amsterdam: John Ben-

jamins.
Grady, Joseph
1997 Foundations of Meaning: Primary Metaphors and Primary Scenes.
Department of Linguistics, U.C. Berkeley. Unpublished doctoral dis-
sertation.
1999 A typology of motivation for conceptual metaphor. Correlation vs.
resemblance. In: Raymond W. Gibbs, Jr. and Gerard J. Steen (eds,),
Metaphor in Cognitive Linguistics, 79–100. Amsterdam: John Ben-
jamins.
Grady, Joseph and Christopher Johnson
2003 Converging evidence for the notions of subscene and primary scene.
In: René Dirven and Ralf Pörings (eds.), Metaphor and Metonymy in
Comparison and Contrast, 533–554. Berlin: Mouton de Gruyter.
Hellsten, Iina
2000 Dolly: Scientific breakthrough or Frankenstein’s Monster?
Journalistic and scientific metaphors of cloning. Metaphor and
Symbol 15 (4): 213–221.
2002 The Politics of Metaphor: Biotechnology and Biodiversity in the
Media. Acta Universitatis Tamperensis; 876. Tampere: Tampere
University Press. http://acta.uta.fi/pdf/951-44-5380-8.pdf
2003 Focus on Metaphors: The case of ’Frankenfood’ on the web. Journal
of Computer-Mediated Communication 8 (4). http://jcmc.indiana.edu
/vol8/issue4/hellsten.html
2005 From sequencing to annotating: Extending the metaphor of the book
of life from genetics to genomics. New Genetics and Society 24 (3):
283–297.
Kay, Lily
2000 Who Wrote the Book of Life: A History of the Genetic Code. Stan-
ford: Stanford University Press.
Kristiansen, Gitte
Lakoff, George
1987 Women, Fire and Dangerous Things. What Categories Reveal about
the Mind. Chicago: Chicago University Press.
1993 The contemporary theory of metaphor. In: Andrew Ortony (ed.),
Metaphor and Thought, 2nd. edition, 202–251. Cambridge: CUP.
1999 Philosophy in the Flesh. The Embodied Mind and its Challenge to
Western Thought. Chicago: Basic Books.
1987 Foundations of Cognitive Grammar. Vol. 1: Theoretical Prerequi-
sites. Stanford, CAL: Stanford University Press.
Larson, Brendon
this vol. Entangled biological, linguistic and cultural origins of the war on
invasive species
Larson, Brendon, Brigitte Nerlich and Patrick Wallis
2005 Metaphors and biorisks: The war on invasive species and infectious
diseases. Science Communication 60: 2629–2639.
Lotman, Yurij
1990 Universe of the Mind. A Semiotic Theory of Culture. London: Tauris
and Co.
Mey, Jacob
2001 Pragmatics. An Introduction. Oxford: Blackwell.
Musolff, Andreas
2004 Metaphor and Political Discourse. Analogical Reasoning in Debates
about Europe. Basingstoke: Palgrave Macmillan.
Musolff, Andreas and Jörg Zinken (eds.)
in prep. Metaphor and Discourse. London, UK: Palgrave Macmillan.
Nerlich, Brigitte
2005 ‘A River Runs Through it’: How the discourse metaphor crossing the
Rubicon was exploited in the debate about embryonic stem cells in
Germany and (not) the UK. metaphorik.de 08/2005. http://www.
metaphorik.de/08/nerlich.pdf.
Nerlich, Brigitte and Iina Hellsten
2004 Genomics: Shifts in metaphorical landscape between 2000 and 2003.
The New Genetics and Society 23 (3): 255–268.
Özçalişkan, Şeyda
2003 Metaphorical motion in crosslinguistic perspective. A comparison of
English and Turkish. Metaphor and Symbol 18 (3): 189–228.
Riddell, Mary
2003 The bio-terrorism of nature. Can Sars be the catalyst that unites a
divided world against the common enemy of global disease? The Ob-
server [Sunday edition of The Guardian], Sunday, April 27. [Online
source: Guardian Unlimited]
Rosch, Eleanor, Mervis, Carolyn, Gray, Wayne, Johnson, David and Penny Boyes–
Braem
1976 Basic objects in natural categories. Cognitive Psychology 8: 382–
439.
Sinha, Chris
1999 Grounding, mapping and acts of meaning. In: Theo Janssen and
and Methodology, 223–255. Berlin: Mouton de Gruyter.
Temmerman, Rita
this vol. Sociocultural situatedness of terminology in the life sciences: The
history of splicing
Tomasello, Michael
Väliverronen, Esa and Iina Hellsten
2002 From ”Burning Library” to ”Green Medicine”. The role of
metaphors in communicating biodiversity. Science Communication
24 (2): 229–245.
Van Dijck, José
1998 Imagenation. Popular Images of Genetics. New York: New York
University Press.
Wallis, Patrick and Brigitte Nerlich
2005 Disease metaphors in new epidemics: The UK media framing of the
2003 SARS epidemic. Social Science & Medicine 60: 2629–2639.
White, Michael and Honesto Herrera
2003 Metaphor and ideology in press coverage. In: René Dirven, Roslyn
M. Frank and Martin Pütz (eds.). Cognitive Models in Language and
Thought: Ideology, Metaphors and Meanings, 277–326. Berlin/New
Ziemke, Tom
Annual Meeting of the Cognitive Science Society, 1305–1310. Law-
rence Erlbaum.
Zinken, Jörg
2002 Imagination im Diskurs. Zur Modellierung metaphorischer Kommu-
nikation und Kognition. Ph.D. thesis, University of Bielefeld, Ger-
many. http://bieson.ub.uni-bielefeld.de/volltexte/2004/483/ [accessed
02.02.07]
2004 Metaphor practices in the German Wende discourse. Journal of
Multilingual and Multicultural Development 25 (5/6): 424–436.
in press Discourse metaphors: The link between figurative language and
habitual analogies. Cognitive Linguistics.
2003 What is ‘cultural’ about conceptual metaphors? International Jour-
nal of Communication 13 (1/2): 5–29.
Zlatev, Jordan
1997 Situated Embodiment, Studies in the Emergence of Social Meaning.
Stockholm: Gotab Press.
2002 Mimesis: The ‘missing link’ between signals and symbols in phylog-
eny and ontogeny. In: Anneli Pajunen (ed.), Mimesis, Sign and the
Evolution of Language, 93–122. Turku, Finland: University of Turku
Press.
2007 Embodiment, language and mimesis. In: Tom Ziemke, Jordan Zlatev
and Roslyn M. Frank (eds.), Body, Language and Mind. Vol. 1. Em-
bodiment, 297–337. Berlin: Mouton de Gruyter.
Zybatow, Lew
1998 Metaphernwandel als Sprach- und Kulturwandel. Alte und neue Me-
taphorik im modernen Russisch, Ukrainisch, Polnisch und Bulga-
risch. Rostocker Beiträge zur Sprachwissenschaft 6: 251–274.
The relationship between metaphor, body and
culture1
Ning Yu
Abstract
This paper discusses the relationship between metaphor, body and culture. Cogni-
tive linguistics maintains that the mind is embodied. While abstract concepts are
mostly metaphorical, metaphors that structure them are largely derived from bodily
experience. Since human beings all share a basic body structure, and have many
common bodily experiences, it follows that different languages should have parallel
conceptual metaphors across their boundaries. The question asked in this paper is
what role culture plays in this theory. It is suggested that metaphor, body and cul-
ture may form a “circular triangle relationship” (Yu 2003 a). That is, conceptual
metaphors are usually grounded in bodily experiences; cultural models, however,
filter bodily experiences for specific target domains of conceptual metaphors; and
cultural models themselves are very often structured by conceptual metaphors. As
such, any one of the three constraining the next one will affect the third one as well.
The paper concludes with a reference to the hypothetical “Triangle Model” for the
relationship between language, culture, body and cognition (Yu 2001), which pro-
poses the embodiment and sociocultural situatedness of human language and cog-
nition.
Keywords: bodily experience, circular triangle relationship, conceptual metaphor,

cultural model, embodied and socioculturally situated cognition.
1. My sincere thanks go to René Dirven and Roslyn Frank for their very helpful
comments and suggestions on earlier versions of this paper. I am solely respon-
sible, however, for any deficiencies that remain. I also want to thank the Samuel
Roberts Noble Foundation for its support of my research at the University of
Oklahoma.
388 Ning Yu
1. Introduction
In this paper I discuss the relationship between metaphor, body and culture.
I want to point out at the outset that the term metaphor in the title of this
paper is used in a broad sense that includes both metaphor and metonymy
in the narrow sense of the terms. In actuality, “the distinction between
metaphor and metonymy is scalar, rather than discrete: they seem to be
points on a continuum of mapping processes” (Barcelona 2000 a: 16). It
has been noted that metonymy may be a more fundamental cognitive phe-
nomenon than metaphor (Panther and Radden 1999) and, in many cases,
metaphor may be motivated by metonymy (Barcelona 2000 b; Radden
2002). To put it differently, metonymy very often is the link between bod-
ily experience and metaphor in the mapping process from concrete experi-
ence to abstract concepts: bodily experience → metonymy → metaphor →
abstract concepts.
Cognitive linguistics maintains that the mind is embodied (e.g., Gibbs
1994; Johnson 1987, 1999; Lakoff 1987; Lakoff and Johnson 1980, 1999).
While abstract concepts are mostly metaphorical, metaphors that structure
them are, by and large, derived from bodily experience. Since human be-
ings all share a basic body structure, and have many common bodily expe-
riences, it follows that different languages should have parallel conceptual
metaphors across their boundaries. As Dirven (2002: 11) points out, the
cognitive theory of metaphor is “revolutionary” in that it is intimately
linked to two major claims: (i) the experientialist, bodily basis of metaphor
and metonymy, and (ii) the universalist basis for conceptual metaphors and
metonymies.
The question that I ask in this paper is what role culture plays in this
theory. Based on my studies of Chinese, sometimes in comparison with
English, I have suggested that metaphor, body and culture may form a “cir-
cular triangle relationship” (Yu 2003 a), as shown in Figure 1 (Yu 2003 a:
29).
metaphor
body culture
Figure 1. The “circular triangle” relationship between metaphor, body and culture
The relationship between metaphor, body and culture 389
That is, conceptual metaphors are usually grounded in bodily experiences;

cultural models, however, filter bodily experiences for specific target do-
mains of conceptual metaphors; and cultural models themselves are very
often structured by conceptual metaphors. As such, any one of the three
constraining the next one will affect the third one as well.2 Thus, culture,
by interpreting bodily experience, affects the formation of conceptual
metaphors; body, by grounding metaphorical mappings, affects cultural
understanding; and metaphor, by structuring cultural models, affects the
understanding of bodily experience.
2. Body that grounds metaphor
Our body plays a crucial role in our creation of meaning and its under-
standing, and our embodiment in and with the physical and cultural worlds
sets out the contours of what is meaningful to us and determines the ways
of our understanding (Gibbs 1994, 1999; Johnson 1987, 1999). It follows
that human meaning and human understanding are to a considerable extent
metaphorical, mapping from the concrete to the abstract. It also follows
that our body, with its experiences and functions, is a potentially universal
source domain for metaphorical mappings onto more abstract domains.
This is because humans, despite their racial or ethnical peculiarities, all
have the same basic body structure, and all share some common bodily
experiences and functions, which fundamentally define us as being human
(see, also, Yu 2001, 2003 b, c, 2004).
For instance, my comparative study of body-part terminology shows
that the terms for the face in Chinese and English have developed figura-
tive meanings along similar routes with similar stops, as shown in Table 1
(Yu 2001: 25). Thus, Chinese and English have the following metonymic
and metaphorical expressions that are similar in their literal and figurative
meanings.
2. As Dirven (personal communication) points out, there exists a possibility that

the direction of constraint may go the other way around: culture → metaphor →
body → culture. This is an interesting hypothesis worth exploring. In my de-
scriptive model (in Figure 1), however, such constraint is indirect, by way of
“circularity”, that is, “any one of the three constraining the next one will affect
the third one as well”.
390 Ning Yu
Table 1. Senses associated with the body part of face in English and Chinese
English Chinese
Relevant senses associated with the body part of face face lian mian
1. front of head from forehead to chin + + +

2. a look on the face as expressing emotion, character, etc. + + +
3. front, upper, outer, or most important surface of something + + +
4. outward appearance or aspect; apparent state or condition +
5. composure; courage; confidence; effrontery + + +
6. dignity; prestige + + +
7. have or turn the face or front towards or in a certain direction + +
8. meet confidently or defiantly; not shrink from; stand fronting + +
1. Chinese English
a. lao miankong (old face) old face
b. beng-lian (stretch-face) pull a long face
c. ban-lian (harden-face) straighten one’s face
d. lou-mian (show-face) show one’s face
e. dang-mian (to-face) to one’s face
f. mian-dui-mian (face-to-face) face to face
g. liang-mian (two-face) two-faced
h. diu-lian (lose-face) lose face
i. baoquan-mianzi (keep-face) save face
k. you-lian (have-face) have the face/cheek
It is suggested that the figurative extension of the senses of face in English

and its counterparts lian ‘face’ and mian ‘face’ in Chinese reflects the me-
tonymic and/or metaphoric understanding of the face as “highlight of ap-
pearance and look”, “indicator of emotion and character”, “focus of inter-
action and relationship”, and “locus of dignity and prestige”. The
commonality observed here, it is argued, is rooted in some biological facts
and functions of the face as part of our body: namely, the face is the most
distinctive part, on the interactive side, the front, of a person, which dis-
plays emotion, suggests character and conveys intention (see Yu 2001 for a
detailed discussion).
In a different study of figurative expressions containing the body-part

terms for the eyes, which is a subpart of the face (Yu 2004), I found that
Chinese and English share some conceptual metonymies and metaphors.
These include:
– PERCEPTUAL ORGAN STANDS FOR PERCEPTION (OR EYES STAND FOR

SEEING)
– SEEING IS TOUCHING (OR SEEING IS CONTACT BETWEEN THE EYE AND
THE TARGET)
– THINKING, KNOWING, OR UNDERSTANDING IS SEEING
For example, the Chinese compounds in (2) and the English idioms in (3)
respectively illustrate the linguistic instantiation of the conceptual meton-
ymy and metaphor: PERCEPTUAL ORGAN STANDS FOR PERCEPTION and
SEEING IS CONTACT BETWEEN THE EYE AND THE TARGET, as listed above.
2. a. chu-yan (touch-eye) ‘eye-catching; striking; conspicuous’

b. da-yan (beat-eye) ‘catch the eye; attract attention’
c. zha-yan (prick-eye) ‘dazzling; offending to the eye; loud; offen-
sively conspicuous’
d. ci-mu (thorn/stab-eye) ‘dazzling; offending to the eye’
e. duo-mu (seize-eye) ‘catch the eye; dazzle the eyes; be striking to
the eye’
3. a. catch sb.’s eye

b. take sb.’s eye
c. jump to the eye(s)
d. leap to the eye
e. strike the eye
f. hit sb. in the eye
Next, the Chinese and English sentences in (4) instantiate the third con-
ceptual metaphor THINKING, KNOWING, OR UNDERSTANDING IS SEEING
shared by these two languages.
4. a. Da-chu zhuo-yan, xiao-chu zhuo-shou.

large-place put to-eye small-place put to-hand
392 Ning Yu
‘Keep the general goal in sight (or bear one’s larger interests in
mind) while taking hold of the daily tasks’ (lit. Put one’s eyes to
large things, and put one’s hands to small things).
b. Ta you zhengzhi yan-guang.

he has political eye-light
‘He has political foresight.’
c. The scientists at the meeting all cast a skeptical eye on that

theory.
d. She is nothing but a slave in her husband’s eyes.
Example (4a) advocates that one should “think big” and “act small”. Only
when people bear the general goal in mind and handle the ordinary tasks
day in and day out can they actually succeed. In (4b), the “political fore-
sight” refers to the person’s mental ability to predict (i.e., know and under-
stand) and, perhaps, influence the political situation. The “eye” in (4c)
refers to the scientists’ mental attitude toward the theory. The idiom in
one’s eyes in (4d) means “in one’s opinion” or “in one’s mind”, since how
one “sees” determines how one “thinks”.
Figure 2. Metonymic and metaphoric mappings shared by Chinese and English

The three metonymic and metaphorical mappings listed above can be pre-
sented schematically as in Figure 2 (Yu 2004: 680). At the lower level,
there are two mappings onto the same target domain, the perceptual experi-
ence of seeing. One is a metonymic mapping from the perceptual organ of
eyes; the other is a metaphoric mapping from the physical action of touch-
ing. At the upper level, the perceptual experience of seeing now serves as
the source domain, and is metaphorically mapped onto the mental function
of thinking, knowing, or understanding, the target domain. These map-
pings, metonymic and metaphoric, show how “lower” bodily experiences
work their way up to help conceptualize “higher” mental experiences, or
how the more abstract is understood in terms of the more concrete (John-
son 1987, 1999). Although imagination is involved, these metonymic and
metaphoric mappings are grounded in the biological functions of, and bod-
ily experiences with, the eyes as part of our body: namely, the eyes are our
organs of sight, in particular, and of cognition in general.
In short, the conceptual metonymies and metaphors shared by Chinese
and English seem to rest upon a common bodily basis that defines what is
human. The linguistic instances cited, be they in Chinese or English, mani-
fest the underlying conceptual metonymies and metaphors that are
grounded in the common human body structure and bodily experience.
They mean what they mean because we have the kind of body we have.
3. Culture that interprets body
While the body and bodily experiences are potentially universal source
domains for conceptual metaphors structuring abstract concepts, cultural
models set up specific perspectives from which certain parts of the body
and certain aspects of bodily experience are viewed as especially salient
and meaningful in the understanding of those abstract concepts (Gibbs
1999; Yu 2000, 2001, 2002, 2003 a, d). That is, cultural models have an
interpretative function in viewing the body and its role in grounding meta-
phor: They may interpret the same embodied experience differently and
attach different values to the same bodily experiences or the same parts of
the body. Thus, it is possible that, in different cultures and languages, dif-
ferent body parts or bodily experiences are selected to map onto and
structure the same abstract concepts and, conversely, the same body parts
or bodily experiences are selected to map onto and structure different ab-
stract concepts. The convergence and divergence of these kinds, therefore,
394 Ning Yu
give rise to varied conceptual metonymies and metaphors in different lan-

guages.
Table 2. Distribution of some conceptual metonymies and metaphors involving

hand, finger and palm in Chinese and English
Conceptual Metonymies and Metaphors Chinese English
Hand
THE HAND STANDS FOR THE PERSON + +
THE HAND STANDS FOR ATTITUDE + +
UNITY/COOPERATION IS JOINING HANDS + +
DISUNITY/SEPARATION IS PARTING HANDS + +
THE HAND STANDS FOR ACTION + +
ACTION IS DOING WITH THE HAND + +
THE HAND STANDS FOR ACTIVITY + +
THE HAND STANDS FOR SKILL + +
THE HAND STANDS FOR MEANS + +
THE HAND STANDS FOR MANNER + +
FREEDOM TO ACT IS HAVING HANDS FREE FOR ACTION + +
THE HAND STANDS FOR CONTROL + +
CONTROL IS HOLDING IN THE HAND + +
THE HAND STANDS FOR POSSESSION + +
POSSESSION IS HOLDING IN THE HAND + +
Finger
THE POINTING FINGER STANDS FOR TARGET + +
TARGET IS WHAT THE FINGER POINTS TO + +
THE POINTING FINGER STANDS FOR GUIDANCE/ DIRECTION +
GUIDANCE/DIRECTION IS POINTING WITH THE FINGER +
THE FINGER STANDS FOR ACTION +
ACTION IS DOING WITH THE FINGER +
Palm
THE PALM STANDS FOR CONTROL + +
CONTROL IS HOLDING IN THE PALM OF THE HAND + +
For instance, studies of metonymic and metaphoric expressions involving

body-part terms for hand, finger and palm in Chinese and English (see Yu
2000, 2003 c; Kövecses and Szabó 1996) seem to suggest intricate rela-
tionships, in terms of conceptual metonymies and metaphors involved,
between these two languages. These relationships are both symmetric and
asymmetric across the language boundary, as summarized in Table 2.3
Those indented ones can be seen as either specific or related cases of the
more general mapping above them.
As one of the defining characteristics of human beings, our hands are
one of our most important body parts with which we deal with the external
world. We do and hold things with our hands. Our bodily experiences with
the hands serve as the common experiential basis for many shared concep-
tual metonymies and metaphors structuring abstract concepts in both lan-
guages.
However, when it comes down to finger and palm, the subparts of the
hand, the two languages display more variations between them. Thus, the
metonymy THE POINTING FINGER STANDS FOR GUIDANCE OR DIRECTION
and metaphor GUIDANCE OR DIRECTION IS POINTING WITH THE FINGER exist
only in Chinese, as exemplified by the compounds in (5):4
3. This table is based on the findings of my research and others. The extent of its
validity remains to be tested by further studies. See Yu (2000, 2003 c) for de-
tails.
4. Two things are worth mentioning in passing. First, in the pair of conceptual
metonymy and metaphor, DIRECTION is used in the more abstract sense of “su-
pervision”. In Chinese, however, the compounds that express senses related to
supervision, guidance, and direction, as in (5), evoke the imagery of finger
pointing that can be either metonymic or metaphorical (see Geeraerts 2002 for a
detailed discussion of the interaction of metaphor and metonymy in composite
expressions). This is because, in these compounds, zhi ‘finger’, though used
verbally to mean “to point”, is originally the body-part noun for the finger. The
shift from noun to verb as in these compounds has conflated the manner (the
body part involved) into the action (pointing). It is interesting to note that such
metonymic or metaphorical imagery, which obviously has an experiential, bod-
ily basis, is culture-specific, manifesting “cultural conceptualization” and repre-
senting “cultural cognition” (Sharifian 2003). Secondly, here and elsewhere, I
give a metonymy (A STANDS FOR B) and a metaphor (A IS B) as two of a pair that
often interact and intertwine with each other (see Geeraerts 2002; Goossens
2002).
396 Ning Yu
5. a. zhi-dian (finger pointing-point) ‘give pointers, advice, directions;

show how; gossip about someone’s faults; find fault with’
b. zhi-chu (finger pointing-out) ‘point out; lay/put one’s finger on;
state briefly; show clearly; advise; indicate; pinpoint’
c. zhi-ming (finger pointing-light/bright) ‘show clearly; demon-
strate; point out’
d. zhi-yin (finger pointing-lead) ‘point the way; guide; show’
e. zhi-dao (finger pointing-guide) ‘guide; direct; supervise; advise;
coach’
f. zhi-bo (finger pointing-pluck/poke) ‘give pointers, advice; show
how; coach’
g. zhi-zheng (finger pointing-straight/right) ‘point out mistakes so
that they can be corrected; make a comment or criticism’
h. zhi-jiao (finger pointing-teach) ‘give advice or comments’
i. zhi-shou (finger pointing-instruct) ‘instruct’
j. zhi-hui (finger pointing-wave) ‘command; direct; conduct; com-
mander; director; conductor’
k. zhi-ling (finger pointing-order) ‘instruct; order; direct; directive;
command’
l. zhi-shi (finger pointing-show) ‘indicate; point out; instruct; di-
rective; instruction; indication’
On the other hand, the metonymy THE FINGER STANDS FOR ACTION and
metaphor ACTION IS DOING WITH THE FINGER solely exist in English. Thus,
the English idioms in (6) have no literal counterparts in Chinese.
6. a. get one’s fingers into something (“participate in something”)

b. have a finger in something (“take part in something; play a
role in something”)
c. have a finger in the pie (“concern oneself with or be connected
with the matter, especially officiously”)
d. have/stick a finger in every pie (“have a part in everything that is
going on; concern oneself with or be connected with many mat-
ters, especially in an unwelcome way”)
e. keep fingers on something (“take care of or handle something”)
f. get one’s fingers burnt (“suffer after a foolish act or mistake;
suffer for meddling or rashness”)
g. one’s fingers itch to do something (“one is longing or anxious to
do something”)
h. do something without lifting one’s finger (“do something with

least effort”)
i. do something with a wet finger (“do something with little effort”)
j. get/pull/take one’s fingers out (“begin work in earnest; hurry up”)
k. work one’s fingers to the bone (“work very hard”)
l. one’s fingers are (all) thumbs (“one is clumsy”)
It is interesting to note that Chinese assigns the role of “actual doer” exclu-
sively to the hand whereas English divides this role between the hand and
finger. Thus, as I have noticed earlier (Yu 2000), many English idioms
containing “finger” are matched by Chinese conventionalized expressions
involving “hand”. For instance, in English a thief’s “fingers are sticky”
whereas in Chinese a thief’s “hands are sticky”. In English one’s “fingers
itch” when one is anxious to do something while in Chinese one’s “hands
itch” under the same circumstances. In English, if one is said to have done
something “without lifting a finger”, that means the person has done it with
little or no effort. In Chinese, if something is done with ease, it is done
“with one’s hands drooping”. Also, it is “sticking one’s fingers into some-
thing” in English while it is “sticking one’s hands into something” in Chi-
nese. The preference of one over the other here, it seems, is conventional,
and convention is culture.
Languages differ not only in the validity of conceptual metonymies and
metaphors; they may also differ in the applicability of certain conceptual
metonymies and metaphors to the target domain concepts. For instance, in
both Chinese and English THE FINGER STANDS FOR TARGET and TARGET IS
WHAT THE FINGER POINTS TO form a valid pair of conceptual metonymy
and metaphor. However, in English they are only applicable to certain
kinds of negative targets, as manifested by idioms in (7).
7. a. put the finger on (“tell the police about [a criminal]; inform

against; identify as victim”)
b. put one’s finger on (“point with precision to [cause of trouble];
find; show [cause of trouble]”)
c. point a/the/one’s finger at (“criticize; censure; scold”)
d. shake/wag a/one’s finger at (“censure; scold; point out”)
e. give somebody the finger (“insult; mistreat”)
In contrast, the pair of metonymy and metaphor is applicable to a wider

range of targets in Chinese, both negative and positive, as shown in (8).
398 Ning Yu
8. a. zhi-kong (finger pointing-accuse) ‘accuse/charge (sb. of/with a

crime, etc.)’
b. zhi-gong (finger pointing-testify/confess) ‘testify; confess’
c. zhi-ren (finger pointing-recognize/confirm) ‘point out and con-
firm; identify’
d. zhi-chi (finger pointing-scold) ‘reprove; reprimand; denounce’
e. zhi-ze (finger pointing-reproach) ‘charge; censure; criticize; find
fault with’
f. zhi-zhai (finger pointing-blame) ‘pick faults and criticize; cen-
sure; blame’
g. zhi-wang (finger pointing-expect) ‘look to; count on; prospect;
hope’
h. zhi-kao (finger pointing-rely on) ‘depend on (for ones’ liveli-
hood); look to (for help); count on’
i. zhi-zhang (finger pointing-depend on) ‘(dial.) count on; rely on’
j. zhi-ding (finger pointing-decide) ‘appoint; assign; designate’
k. zhi-pai (finger pointing-dispatch) ‘appoint; name; designate’
l. zhi-shi (finger pointing-send) ‘instigate; incite; put someone up to
something’
It seems that the expressions in (7) and (8) are all based on a single physi-
cal act: pointing with one’s (index) finger. However, Chinese differs from
English in that it maps this common bodily gesture onto a wider range of
abstract concepts, such as accusation, intention, dependence, appointment
and assignment. When people identify a target, they point to it with their
(index) finger.
Moreover, languages also differ in the extent to which certain concep-
tual metonymies or metaphors are manifested linguistically. A telling ex-
ample is the conceptual metonymy and metaphor involving “palm”: THE
PALM STANDS FOR CONTROL and CONTROL IS HOLDING IN THE PALM OF
THE HAND. As shown in Table 2, this pair is present in both Chinese and
English. In English, however, it is manifested linguistically to a very lim-
ited extent. I only found the two idioms in (9), which are in fact two vari-
ants of the same one.
9. a. hold … in the palm of one’s hand

(“have complete control over …”)
b. have … in the palm of one’s hand
(“have complete control over …”)
In Chinese, on the other hand, numerous compounds contain “palm” as a

morpheme that carry the meaning of “control”. Those in (10) are some
examples.5
10. a. zhang-xin (palm-center) ‘the center/hollow of the palm; control;

influence’
b. mo-zhang (devil-palm) ‘(derogatory) devil’s clutches; evil hands’
c. zhang-kong (palm-control) ‘control’
d. zhang-wo (palm-hold) ‘have in hand; take in one’s control; grasp;
master; know well’
e. zhang-yin (palm holding-seal) ‘keep the seal; be in power’
f. zhang-quan (palm holding-power) ‘be in power; wield power;
exercise control’
g. zhang-shi (palm holding-affair) ‘be in charge of; administer’
h. zhang-guan (palm holding-administer) ‘be in charge of; admin-
ister’
i. zhu-zhang (manage-palm holding) ‘be in charge of; manage’
j. zhi-zhang (direct-palm holding) ‘be in charge of; direct’
It is worth noting that in both Chinese and English the concept of control is
also figuratively understood in terms of the hand, as indicated in Table 2.
As a matter of fact, THE HAND STANDS FOR CONTROL and CONTROL IS
HOLDING IN THE HAND are extensively manifested in English, as the exam-
ples in (11) show.
11. a. He’s got the matter in hand.

b. We have the situation well in hand.
c. His life was in my hand.
d. The meeting is getting out of hand.
e. We fell into enemy hands.
f. I suffered at his hands.
g. I’ll soon have him eating out of my hand!
h. Let’s leave it in his hands.
i. The child is in good hands.
5. Just as in the previous examples with zhi ‘finger’, zhang ‘palm’ is the body-part
term for the palm, but is also used as a verb to mean “hold (in control)”, as
some of the compounds in (10) clearly illustrate.
400 Ning Yu
j. The cabinet approved last week strengthened his hand for the
difficult tasks ahead.
That is, the body part related to the concept of control in English is pri-
marily the hand, rather than its subpart, the palm. In Chinese, on the other
hand, the concept of control is associated with both the palm and the hand
(Yu 2000, 2003 c).
Finally, languages can also differ in whether they explicitly use body-
part terminology to help construct and express certain abstract concepts.
Thus, it may be the case that the use of a body-part term is explicit in one
language but implicit in another (Yu 2000). For instance, in English point
out implies the use of (index) finger, but its Chinese equivalent zhi-chu
(finger pointing-out) ‘point out’ makes an explicit use of the body-part
term in its reference of a bodily action. This difference suggests that in
English the conceptual metaphor is GUIDANCE OR DIRECTION IS POINTING
(WITH THE FINGER), namely the involvement of the body part is implied,
whereas in Chinese it is GUIDANCE OR DIRECTION IS POINTING WITH THE
FINGER, where the involvement of the body part is specified. Nonetheless,
both the Chinese and English versions are grounded in the same bodily act
of pointing with one’s (index) finger. For further illustration, we can con-
sider the examples in (12) as implicit linguistic manifestation of the con-
ceptual metaphors CONTROL IS HOLDING IN THE HAND and CONTROL IS
HOLDING IN THE PALM OF THE HAND.
12. a. He tried to hold his temper.

b. He’s got a good hold of his subject.
c. Grasp your chances while you can.
d. He is in the grasp of a wicked man.
e. The people regained power from the grasp of the dictator.
f. An anarchic fervor gripped the campus.
g. He kept a firm grip on his children.
h. Don’t get into the grip of moneylenders.
i. Teachers should loosen their grip on the curriculum.
j. She felt herself in the grip of sadness she could not understand.
Here, both verbal and nominal uses of English words as hold, grasp and
grip in the sense of “control/possession” may also imply the use of the
hand to hold, to grasp, or to grip. They are grounded in the bodily experi-
ence of holding in (the palm of) the hand. In my 1995 and 1998 studies I
pointed out that Chinese seems to have more conventionalized expressions,
in the form of compounds and idioms, involving explicit use of body-part

terminology than does English, and my subsequent studies (Yu 2000, 2001,
2002, 2003 a, c, d, 2004) have reinforced that observation.
The examples presented in this section should illustrate how different
interpretations of the same or similar bodily experiences (such as doing
things with one’s hands/fingers, holding things in [the palm of] one’s
hands, or pointing with one’s finger) by different cultures can lead to
variations in linguistic expression. The linguistic expressions of bodily-
based metonymies and metaphors instantiate conceptualizations at the cul-
tural level of cognition (Sharifian 2003).
4. Metaphor that constitutes culture
In the previous two sections, I showed how metaphor is grounded in bodily

experience and how bodily experience is interpreted by culture to affect the
outcome of metaphorical mappings. Meanwhile, cultural models as shared
understandings of the world in a culture may be metaphorically constructed
themselves. This has a consequence in the role of cultural models in inter-
preting and selecting bodily experiences as source domains for metaphors.
In Yu (1998), I argued that the theories of yin-yang and five elements of
Chinese philosophy and medicine actually shape the way Chinese culture
sees the world. As part of the shared understandings of Chinese culture that
constitute Chinese cultural models, they have shaped the selection of meta-
phors in the Chinese language in a significant way (e.g., ANGER IS GAS IN A
HEATED CONTAINER in Chinese vs. ANGER IS FLUID IN A HEATED CON-
TAINER in English). However, the question is “whether shared understand-
ings of a culture or cultural models can themselves be free of metaphor, or
whether they can be structured by metaphor to a certain extent” (Yu 1998:
81). I argued:
The theories of yin-yang and five elements are in essence theories of catego-
rization and conceptualization. They categorize and conceptualize things in
certain relations in terms of other things in similar relations. Put differently,
they understand one thing in terms of another of a different kind. And that is
metaphorical in the contemporary theory of metaphor. […] The theories of
yin-yang and five elements, as I tend to believe, are giant metaphors that
constitute metaphorical ways of categorizing and conceptualizing the world
for people who accept the theories or metaphors. If this is true, then the role
played by metaphor in culturally shared understanding of the world would
consequently be major as well. […] If culture is reflection and pattern of
402 Ning Yu
thinking and understanding, and if thinking and understanding can be and

are sometimes inevitably metaphorical, then culture and metaphor would
also fall into a relation of mutual promotion or restraint, depending on how
this relation is interpreted. That is, culture plays a role in shaping metaphor
and, in return, metaphor plays a role in constituting culture. (Yu 1998: 81–
82)
In one of my recent studies (Yu 2003 a), I presented a case in which a cul-
turally constructed metaphorical understanding of an internal organ, the
gallbladder, forms the base of the cultural model for the concept of cour-
age. According to the theory of internal organs in traditional Chinese medi-
cine, which is largely based on the theories of yin-yang and five elements
of ancient Chinese philosophy, the gallbladder, metaphorically conceptu-
alized as the “Office/Organ of Justice”, has the function of making judg-
ments and decisions in mental processes and activities, and it also deter-
mines one’s degree of courage. This culture-specific understanding of the
gallbladder leads to a pair of conceptual metaphors that in part constitutes
the Chinese cultural model for courage:
– GALLBLADDER IS THE CONTAINER OF COURAGE

– COURAGE IS QI (GASEOUS VITAL ENERGY) IN GALLBLADDER
This pair of conceptual metaphors, based on the CONTAINER image schema,

entails the following mappings or correspondences between the source and
target domains:
Source domain Target domain

physical container of courage → gallbladder
gaseous energy of qi in the container → courage
capacity of the container → amount of courage
degree of internal pressure of the container → degree of courage
Thus, Chinese has the following compound words:6
13. a. dan-zi (gall-SUFFIX) ‘courage; guts; nerve’

b. dan-qi (gall-qi [gaseous vital energy]) ‘courage’
c. dan-li (gall-strength) ‘courage and boldness’
d. dan-zhuang (gall-strong) ‘bold; fearless; courageous’
6. Readers are referred to Yu (2003 a) for a detailed discussion of the linguistic

evidence, comprised of both compounds and idioms, and its cultural context.
e. dan-xu (gall-weak/void) ‘afraid, scared; timid’

f. dan-liang (gall-capacity) ‘courage; guts; pluck; spunk’
g. dan-da (gall-big) ‘bold; audacious’
h. dan-xiao (gall-small) ‘timid; cowardly’
i. dan-xiao gui (gall-small devil) ‘coward’
j. luo-dan (drop/fall-gall) ‘extremely scared’
k. sang-dan (lose-gall) ‘be terror-stricken; be smitten with fear’
All these compounds contain dan ‘gallbladder,’ but are related to courage.
As can be seen, courage is respectively connected to the gallbladder itself
(13a) and its gaseous vital energy qi (13b). People’s courage has to do with
the strength or internal pressure of their gallbladder (13c). The strength of
the gallbladder depends on how much qi it contains. If one’s gallbladder is
full of qi, it is “strong”, that is, this person is bold, fearless, or courageous
(13d). Conversely, if one’s gallbladder is “void” of qi and “weak”, this
person is then afraid, scared, or timid (13e). Courage is also related to the
capacity of the gallbladder as its container (13f). If one’s gallbladder is
“big”, this person is bold and audacious (13g). If, on the other hand, one’s
gallbladder is “small”, this person is timid or cowardly (13h). A coward, in
Chinese, is called a “gall-small devil” (13i). Sometimes, the emotion of
fear can be so intense that it “snaps the base of the gallbladder” and makes
it drop off its stem” in a complete “loss” (13j and 13k).
In short, here is a case in which an abstract concept (courage) is under-
stood in part via a pair of conceptual metaphors grounded in the body, but
shaped by a culture-specific metaphorical understanding of an internal
organ (gallbladder) inside the body. In this case, the metaphorical under-
standing of the gallbladder actually defines one aspect of Chinese culture,
and the conceptual metaphors, GALLBLADDER IS THE CONTAINER OF
COURAGE and COURAGE IS QI IN GALLBLADDER, are partly constitutive of
the Chinese cultural model for the concept of courage.
5. Conclusion
In this paper I have argued that the human body and bodily experience are
a potentially universal source domain for metaphors creating the potential
for structuring abstract concepts. However, cultural models, which can be
metaphorically constructed, set up specific perspectives from which certain
aspects of bodily experience or certain parts of the body are viewed as
404 Ning Yu
especially salient and meaningful in the structuring and understanding of

those abstract concepts. While embodied experience serves as the experi-
ential basis for conceptual metaphors in all languages, in each language,
however, only a portion of this basis is actually focused upon and selected
to be mapped onto various target domains and manifested in linguistic
expressions, as is determined by the cultural models associated with that
language. For any two languages, the portions selected by their cultural
models may partially overlap each other, whereas overlapping is defined as
parallel conceptual mappings from the same source domains to the same
target domains. The portions that overlap between these two languages
account for their commonalities in conceptual metaphors. On the other
hand, the portions that do not overlap, that is, conceptual mappings from
the same source domains to different target domains, from different source
domains to the same target domains, or from different source domains to
different target domains, constitute their differences in conceptual meta-
phors.
To conclude this paper on the relationship between metaphor, body and
culture, I would like to refer back to the hypothetical “Triangle Model” that
I (Yu 2001) have previously proposed to describe the more general rela-
tionship between language, culture, body and cognition. This hypothetical
model is represented schematically, in the shape of an upside-down trian-
gle, by the diagram in Figure 3 (Yu 2001: 30). I believe that it captures the
dynamic and intricate nature of a multidimensional relationship, and the
simultaneously embodied and socioculturally situated nature of human
language and cognition.
Figure 3. Triangle Model for relationship between language, culture, body and
cognition.
This triangle-shaped diagram is interpreted as follows. A stands for the

bodily basis, which consists of our basic knowledge about the structure and
function of our body. Line BC represents the level of language, with the
distance between B and C representing the difference between two lan-
guages. By the same token, line DE represents the level of culture (includ-
ing social and physical environment), with the distance between D and E
representing the difference between two cultures. The distance between D
and E is a variable, depending on how different or similar the two cultures
are. The cultural distance between D and E affects the corresponding lin-
guistic distance between B and C. No matter how far apart D and E may be,
they always come down, respectively through B and C, and meet at A. That
is, cultures and languages are all wired to the very essence of humanness –
the human body, more so with languages than cultures as represented by
the different distances. Thus, line AF has a double function. First, it sets
the boundary between the two languages and cultures. Second, it also rep-
resents the commonality between these two languages and cultures, arising
from the common structure and function of human body. What this means
is that, however different two languages and cultures may be, they should
always have a shared dimension that extends from point A to point F. It is
impossible for them to be separated because they are all tied together by
the humanness that exists in the common human body. Outlined above is
the relationship between language, culture and body while cognition is the
totality of the relationships between all the points and all the lines in Figure
3. That is to say, in summary, our language and cognition are at the same
time embodied and socioculturally situated.
References
Barcelona, Antonio
2000 a Introduction: the cognitive theory of metaphor and metonymy. In:
Antonio Barcelona (ed.), Metaphor and Metonymy at the Cross-
roads: A Cognitive Perspective, 1–28. Berlin: Mouton de Gruyter.
2000 b On the plausibility of claiming a metonymic motivation for concep-
tual metaphor. In: Antonio Barcelona (ed.), Metaphor and Metonymy
at the Crossroads: A Cognitive Perspective, 31–58. Berlin: Mouton
de Gruyter.
Dirven, René
2002 Introduction. In: René Dirven and Ralf Pörings (eds.), 1–38.
406 Ning Yu
Dirven, René and Ralf Pörings (eds.)

2002 Metaphor and Metonymy in Comparison and Contrast. Berlin:
Mouton de Gruyter.
Geeraerts, Dirk
2002 The interaction of metaphor and metonymy in composite expressions.
In: René Dirven and Ralf Pörings (eds.), 435–465.
Gibbs, Raymond W.
1994 The Poetics of Mind: Figurative Thought, Language and Under-
standing. Cambridge: Cambridge University Press.
world. In: Raymond W. Gibbs and Gerard J. Steen (eds.), Metaphor
in Cognitive Linguistics, 145–166. Amsterdam: John Benjamins.
Goossens, Louis
2002 Metaphtonymy: The interaction of metaphor and metonymy in ex-
pressions for linguistic action. In: René Dirven and Ralf Pörings
(eds.), 349–377.
Johnson, Mark
1999 Embodied reason. In: Gail Weiss and Honi F. Haber (eds.), Perspec-
tives on Embodiment: The Intersections of Nature and Culture, 81–
102. New York: Routledge.
Kövecses, Zoltán and Péter Szabó
1996 Idioms: A view from cognitive semantics. Applied Linguistics 17:
326–355.
Lakoff, George
Lakoff, George and Mark Johnson.
1999 Philosophy in the Flesh: The Embodied Mind and Its Challenge to
Panther, Klaus-Uwe and Günter Radden (eds.)
1999 Metonymy in Language and Thought. Amsterdam: John Benjamins.
Radden, Günter
2002 How metonymic are metaphors? In: René Dirven and Ralf Pörings
(eds.), 407–434.
Sharifian, Farzad
2003 On cultural conceptualizations. Journal of Cognition and Culture 3,
187–207.
Yu, Ning
1995 Metaphorical expressions of anger and happiness in English and
Chinese. Metaphor and Symbolic Activity 10: 59–92.
1998 The Contemporary Theory of Metaphor: A Perspective from Chi-
nese. Amsterdam: John Benjamins.
2000 Figurative uses of finger and palm in Chinese and English. Metaphor
and Symbol 15: 159–175.
2001 What does our face mean to us? Pragmatics and Cognition 9: 1–36.
2002 Body and emotion: Body parts in Chinese expression of emotion. In:
Nick J. Enfield and Anna Wierzbicka (eds.), special issue on “The
Body in Description of Emotion: Cross-linguistic Studies”. Prag-
matics and Cognition 10: 333–358.
2003 a Metaphor, body and culture: The Chinese understanding of gallblad-
der and courage. Metaphor and Symbol 18: 13–31.
2003 b Synesthetic metaphor: A cognitive perspective. Journal of Literary
Semantics 32: 19–34.
2003 c The bodily dimension of meaning in Chinese: What do we do and
mean with “hands”? In: Eugene H. Casad and Gary B. Palmer (eds.),
Cognitive Linguistics and Non-Indo-European Languages, 330–354.
2003 d Chinese metaphors of thinking. In: Gary B. Palmer, Cliff Goddard
and Penny Lee (eds.), special issue on “Talking about Thinking
across Languages”. Cognitive Linguistics 14: 141–165.
2004 The eyes for sight and mind. Journal of Pragmatics 36: 663–686.
Idealized cultural models: The group as a variable
in the development of cognitive schemata
Gitte Kristiansen
Abstract
The thesis that human cognition is embodied finds general acceptance in Cognitive
Science. The term, however, has been interpreted in a number of different ways.
One line of research has successfully explored the bodily basis of human cognition
in terms of universal cognitive operations and schemas; meaning is embodied in the
sense that Lakoff has characterized it (1987: 267): “in terms of our collective bio-
logical capacities and our physical and social experiences as beings functioning in
our environment.” Emphasis, however, seems to be on the former factors, at the
expense of social experiences. In this paper we examine the role of “cultural” and
“social” factors in the sense of group-specific mechanisms. It is argued that in order
to account in precise terms for the ways in which Cognitive Models emerge, a
framework which predominantly stresses facets of subjective and universal cogni-
tion is insufficient, as it fails to account for the fact that conceptual structure, far
from being universal, varies from culture to culture. In this respect it is suggested
that a distinction should be established between deontic and epistemic social sche-
mata. The paper also critically revises a short selection of classic (but still highly
productive) empirical research in the field of Social Psychology with the group as
its object of study. Social cognition, social categorization and successful social
functioning thus constitute the main topics of this paper.
Keywords: deontic schemata, epistemic schemata, situated embodiment, social

cognition, social group, social stereotypes.
1. Embodiment: From the individual to the species?
Half in jest and half in earnest we could conveniently begin by observing

that if on closer examination most categories turn out to be complex and
asymmetrical, Cognitive Linguistics itself is not an exception. Different
410 Gitte Kristiansen
ways of “doing CL” have undoubtedly emerged, or are in the process of

emerging (for an overview see Dirven 2004). The thesis that human cogni-
tion is embodied, for instance, finds general acceptance, but the term em-
bodiment itself has been interpreted in numerous ways (cf. Chrisley and
Ziemke 2002).
In Cognitive Linguistics, one line of research (e.g. Rosch 1975, 1978;
Johnson 1987; Lakoff 1987; Lakoff and Johnson 1999) has successfully
explored the bodily basis of human cognition in terms of supposedly uni-
versal cognitive operations and schemas, the source of which is our shared
neurobiological system and bodily experience. Meaning is embodied in the
sense that it is characterized by Lakoff (1987: 267): “in terms of our col-
lective biological capacities and our physical and social experiences as
beings functioning in our environment”. Emphasis, however, seems to be
on the former factors, at the expense of social experiences. Johnson (1987)
argued that preconceptual structure arises from preconceptual experience,
consisting of basic-level structure (gestalt perception, our capacity for
bodily movement and ability to form rich mental images) and kinesthetic
image-schematic structure (simple structures that constantly recur in our
everyday bodily experience, such as containers, paths, up-down, part-
whole). Abstract conceptual structure then arises either by metaphorical
projection from the domain of the physical to abstract domains or by the
projection from basic-level categories to superordinate and subordinate
categories (cf. Rosch et al. 1976). In this respect, the following claims (La-
koff 1987: 268) were also made:
– Since bodily experience is constant experience of the real world that

mostly involves successful functioning, stringent real-world constraints
are placed on conceptual structure. This avoids subjectivism.
– Since image schemas are common to all human beings, as are the prin-
ciples that determine basic-level concepts, total relativism is ruled out,
though limited relativism is permitted.
The question, which is still of great present interest, is the extent to which
absence of subjectivism and total relativism entails universal cognition.
Sinha and Jensen de López (2000) and Bernárdez (2002), for instance,
have examined the social and physical conditions involved in the acquisi-
tion of spatial concepts in Native American languages and call for a less
universal approach to the notion of image-schema (see Tyler 1995; Yu
1998 for further examples of culture-specific realizations).
Idealized cultural models 411
It seems fairly reasonable to question the apparent ease with which facts
about the individual – or one particular group of individuals – often be-
come extrapolated to the human species as a whole. The leap from “indi-
vidual” to “humanity” is a giant one, indeed. For a start, the general human
capacity of categorization and conceptualization on the one hand and lin-
guistic encoding, production and decoding on the other hand must surely
be kept separate from the specific cognitive schemata and linguistic in-
stantiations we encounter in different languages and linguistic varieties (in
terms of subcategorizations, or specific instantiations, of a language). Gen-
eralizations should be “handled with care”, as the facts and findings about
a particular language and a particular culture (no matter how much devel-
oped, prestigious or dominant both happen to be) do not necessarily char-
acterize other languages and cultures. Though “language” is common to
“humankind”, “languages” pertain to subcategorizations of “human” – and
“linguistic varieties” to further subordinate levels of (human and lectal)
abstraction. While language as a universal faculty pertains to the human
species as such, languages and varieties pertain to human subspecies, or
social groups, the term I shall use in this paper to denote instantiations of
the more general schema [HUMAN]. Using a biological metaphor1, we may
say that language is a species and languages and varieties constitute sub-
species.
It is thus suggested that in addition to the wealth of otherwise brilliant
work carried out on those aspects of cognition which are common to our
species, it would be fruitful also to turn our attention to more intermediate
levels of abstraction. As Sinha and Jensen de López (2000) have pointed
out, in line with the work of Vygotsky ([1930] 1978), social factors seem
to play an important role in the acquisition of language and cognitive
schemas:
Every function in the child´s cultural development appears twice: first, on
the social level, and later, on the individual level; first, between people (in-
terpsychological), and then inside the child (intrapsychological). This ap-
plies equally to voluntary attention, to logical memory, and to the formation
of concepts. (Vygotsky 1978: 57)
1. See Dirven and Polzenhagen (2004) for an overview of the ways in which
metaphors have been used by linguists to conceptualize language and language
evolution. The authors rightly criticize the fact that even linguists on occasions
fail to distinguish between non-metaphorical and metaphorical inference pat-
terns.
Cognition is thus not only embodied in the sense of individual percep-

tion and interaction with a perhaps mainly physical environment, but also
collective and situated; perception and action only acquire meaningfulness
(i.e. become delimited and value-laden) when the individual engages in
socio-cultural practices, interacting repeatedly with both objects and peo-
ple in ways which are understood as successful within a given social group.
In phenomenology (Merleau-Ponty 1945), too, human cognition is viewed
as an accumulated product of factors (i.e. perception, intentionality, action
and emotion) which should not be viewed separately and which are force-
fully enacted through collective interaction.
By social factors or social level I shall not refer to interaction between
individuals perceived as such. Much fruitful research has already been
carried out on language acquisition and the transmission of concepts from
adults to children in terms of inter-individual processes. By intermediate
level of abstraction, I shall refer specifically to the group. There are im-
portant cognitive and psychological stages between the individual and the
species to which it belongs: human is a superordinate category subject to
multiple subcategorizations. Human beings not only constitute groups, but
also categorize other human beings into groups according to a wide range
of social, cultural, ethnic or ideological parameters. Moreover, when refer-
ence is made to “social” or “cultural” factors, these are quite often syn-
onymous with “group-specific” factors. To a large extent the values, con-
cepts and linguistic expressions parents pass onto their children are those
of their own social background and environment. The family probably
constitutes a basic group itself, the one the child first interacts with and
soon compares with other groups. Once at school, group membership (and
peer pressure) are relevant dimensions in every child’s life. Then the
child’s awareness of social networks and categories at wider levels of ab-
straction gradually increases. Teenager, mother, pensioner, bachelor, so-
cialist, Muslim and Boy Scout are all culture-specific concepts related to
human subcategorization. Nations2 are social groups (and social concepts)
writ large.
We are thus going to centre on those aspects of concept-formation and
categorization which we may label as group-dependent; either because a)
the group plays a fundamental role, or so we believe, in concept-formation,
concept-maintenance and concept-transmission or because b) the group
2. In a multicultural civil nation state such as the United States of America, “feel-
ing American” is surely not just the privilege of white Anglo-Americans.
itself constitutes the object of categorization – and subsequent concept-

formation. Both perspectives could perhaps be subsumed under the more
general term social cognition, since we in either case enter as different and
yet at the same time closely related realms as those of categorization and
concept-formation. However, there are also a series of differences to take
into account. The distinction between a) and b) can suitably be illustrated
by means of the following examples, both of which (Lakoff 1987) have
been classified as social stereotypes:
1. Good mothers are housewives

2. The Japanese are industrious
What Western housewife-mothers3 and industrious Japanese have in com-

mon is that both seem to act in terms of a cognitive reference-point con-
struction: an abstract, mental representation, the image which first springs
to a given native speaker’s mind (i.e. a case of folk perception). What I
have in mind is thus similar to a “[…] ‘core meaning’ which consists of the
`clearest cases´ (best examples) of the category, ‘surrounded’ by other
category members of decreasing similarity to that core meaning” as Rosch
(1973: 112) once defined the prototype.4 And yet they do not seem to rep-
resent exactly the same kind of concepts: while the Japanese are catego-
rized and conceptualized as an outgroup, Western mothers – when con-
ceptualized from a Western perspective – constitute an ingroup.
In order to work adequately around such differences, let us establish the
distinction between deontic and epistemic schemata. Deontic schemata are
those we live by (unless we choose to live as amoral or immoral creatures).
Deontic schemata affect the ingroup (Western mothers, and in this par-
ticular case most other mothers in a still basically patriarchal world) in
direct and pervasive ways. While epistemic schemata characterize out-
groups (industrious Japanese) and remain descriptive, deontic schemata are
predominantly prescriptive; they constitute systems of social rules and
normative behaviour that guide individual action and act as category-
3 As Sego (2003) explains, in North American culture, the term “home-maker” is

gradually replacing the now pejorative, old-fashioned term “house-wife”.
4. For the present purposes, let us start out by using “prototype” and “stereotype”
as covering more or less the same notion: a central, abstract image which acts in
accordance with the principle of cognitive economy, irregardless of whether
“central images” in such domains constitute the cause or effect of category
structure.
builders. Failure to comply with the ingroup’s social conventions, social

norms or morals (the difference is rather ill-defined and probably best un-
derstood in terms of a continuum) affect moral emotions such as shame,
guilt and remorse.
A bird’s eye view with respect to “our own” cultural conventions, moral
codes and normative behaviour is thus often quite desirable. Shweder, Ma-
hapatra and Miller (1987), for instance, once asked children from North
America and India to rate how morally transgressive it would be for a wife
to play cards while her husband cooked rice. In order to ask such a ques-
tion, the answer must of course already be known or at least “sensed” to
some extent, but this observation apart, the results showed that the Indian
children rated such an action as morally wrong, while the North American
children considered it a rather trivial affair. Similarly, eating chicken after
the death of their father was rated as a serious moral transgression by the
Indian children.
In this paper the discussion centres on epistemic schemata.5 In this re-
spect I critically reexamine a short selection of classic, but still highly pro-
ductive, empirical research in the field of Social Psychology with the group
as its object of study.
2. Social stereotypes as epistemic schemata
In Social Psychology, needless to say, group is a key word and there is a

breath-taking amount of research on social cognition which might well be
of interest to cognitive linguists. Social Identity Theory (henceforth SIT),
for instance, was developed in Bristol in the 1970s by Henri Tajfel, John C.
Turner and their associates.6 Tajfel centred his research on issues such as
group formation, group interaction, intergroup conflicts, social categoriza-
tion and social stereotypes.
5. A discussion of deontic schemata would centre on social norms. Taken together,

the two perspectives call for a reconsideration of ICMs (Lakoff 1987) in terms
of construals which relate to social groups and to social identities.
6. A research group, including John C. Turner, S. Alexander Haslam, Penelope J.
Oakes and Craig McGarty, has for several decades carried out extensive re-
search in the field of SIT and Self-Categorization Theory (Turner et al. 1987) at
the Australian National University.
SIT was based on the four general principles of categorization, identifi-

cation, comparison and differentiation. Let us briefly examine these no-
tions step by step.
2.1. Categorization, identification, comparison and differentiation
Tajfel (1978: 63) defined social identity as “that part of an individual’s

self-concept which derives from his knowledge of his membership of a
social group (or groups) together with the value and emotional significance
attached to that membership”. Accordingly, social identity can “only be
defined through the effects of social categorizations segmenting an indi-
vidual’s social environment into his group and others” (Tajfel 1978: 67).
Group, then, denotes a cognitive entity, a category. Tajfel and Turner
(1979: 40) defined social categorizations as:
[...] cognitive tools that segment, classify, and order the social environment,
and thus enable the individual to undertake many forms of social actions.
But they do not merely systematize the social world; they also provide a
system of orientation for self-reference: they create and define the individ-
ual’s place in society. Social groups, understood in this sense, provide their
members with an identification of themselves in social terms. These identifi-
cations are to a very large extent relational and comparative: they define the
individual as similar to or different from, as “better” or “worse” than, mem-
bers of other groups. […] It is in a strictly limited sense, arising from these
considerations, that we use the term social identity.
If social identifications provide individuals with an identification of them-
selves in social terms, our multiple social identities are ultimately as im-
portant to the self as our personal identity (the perception of oneself as a
unique individual). Social identifications are defined as relational and
comparative. This idea was largely based on Leon Festinger’s (1954) no-
tion of social comparison.
Tajfel and Turner (1979: 40–41) based Social Identity Theory on the
following general assumptions:
1. Individuals strive to achieve or to maintain positive social identity.
2. Positive social identity is based to a large extent on favourable compari-
sons that can be made between the in-group and some relevant out-
groups: the in-group must be perceived as positively differentiated or
distinct from the relevant out-groups.
3. When social identity is unsatisfactory, individuals will strive either to

leave their existing group and join some more positively distinct group
and/or to make their existing group more positively distinct.
The basic hypothesis, then, is that pressures to evaluate one’s own group
positively through in-group/out-group comparisons lead social groups to at-
tempt to differentiate themselves from each other. [...] The aim of differen-
tiation is to maintain or achieve superiority over an out-group on some di-
mensions. Any such act, therefore, is essentially competitive. This
competition requires a situation of mutual comparison and differentiation on
a shared value dimension.
Shared value dimensions often constitute continuous dimensions, some-
thing we can be to a certain degree; wealthy-poor, extremist-moderate,
progressive-conservative, religious-irreligious, dark-fair, etc. (cf. Zadeh´s
1965 fuzzy set theory), and only (Tajfel 1981: 258) in comparison to other
people:
The characteristics of one’s group as a whole (such as its status, its richness
or poverty, its skin colour or its ability to reach its aims) achieve most of
their significance in relation to perceived differences from other groups and
the value connotation of these differences. For example, economic depriva-
tion acquires its importance in social attitudes, intentions and actions mainly
when it becomes “relative deprivation”; [...] the definition of a group (na-
tional, racial or any other) makes no sense unless there are other groups
around. A group becomes a group in the sense of being perceived as having
common characteristics or a common fate mainly because other groups are
present in the environment.7
7. When Tajfel affirmed that “a group becomes a group in the sense of being per-
ceived as having common characteristics or a common fate mainly because
other groups are present in the environment”, we have reasons to believe that
what he had in mind was not a series of “common components” as in compo-
nential analysis, but rather the fact that as an outcome of the process of catego-
rization the members of a given category are perceived, for purposes of general
understanding, as interchangeable and sharing a series of relevant features; we
know, in fact, that Tajfel (1981: 147) was familiar with (and acknowledged)
Wittgenstein’s notion of family resemblance.
2.2. Social categories and social stereotypes
Following Allport (1954), Tajfel viewed stereotypes as the outcome of the

general, cognitive process of categorization, but unlike Allport, he chose to
adopt a positive perspective,8 rather than representing a distorted image
that contributes to intergroup conflicts and prejudice, an “exaggerated be-
lief associated with a category” as Allport (1954: 191) defined them,
stereotypes (Tajfel 1969: 82–83) are necessary and functional mental con-
structs:
Stereotypes arise from a process of categorization. They introduce simplicity
and order where there is complexity and nearly random variation. They can
help us cope only if fuzzy differences between groups are transmuted into
clear ones, or new differences created where none exist. [...] in each relevant
situation we shall achieve as much stereotyped simplification as we can
without doing unnecessary violence to the fact. [...] When a classification is
correlated with a continuous dimension, there will be a tendency to exagger-
ate the differences on that dimension between items which fall into different
classes, and to minimize these differences within each of the classes.
Categorization as a cognitive process thus involves accentuation of intra-
group similarities and accentuation of intergroup differences on relevant
continuous dimension. Let us try to schematize this process:
Figure 1. Stereotype-formation according to SIT
8. This positive perspective has also recently (see Kristiansen 2003) been applied
to linguistic stereotyping
The outcome is a series of apparently distinct and homogeneous categories

in which stereotyped, simplified content is ascribed to all the members.
The exaggeration commonly associated with social stereotypes is thus
viewed as a natural by-product of the process of accentuation of intragroup
similarities and intergroup differences, itself a categorization effect. Exag-
geration apart, stereotypes are moreover generally considered to be impre-
cise. This we all realize; what we do not realize is how often we actually
make use of them for purposes of rapid, efficient identification and char-
acterization. Imprecision, too, is viewed in SIT as no more than a categori-
zation effect: stereotypes comprise that which is common (in the eyes of
the perceiver) to a whole group, and categorization inevitably leads us to a
certain level of abstraction. Self-categorization theory (Turner et al. 1987)
proposes that when a particular social identity is salient and we perceive
ourselves as members of a given group, a flexible process of depersonal-
isation is carried out that enable us to view ourselves as interchangeable in
terms of attitudes and beliefs with other members of that particular group.
Generalizing thus entails impreciseness.
Social categorization accordingly effects ingroups and outgroups.
These, however, differ considerably with respect to the ways in which their
members are viewed; there is a general tendency to see members of out-
groups as far more similar to each other than members of our own in-
groups. This phenomenon is known as the outgroup homogeneity effect. It
is probably as difficult to view ourselves epistemically as it is to realize
that the models we live by are context-embedded and culture-specific. That
cognition is culturally situated is accordingly easier to acknowledge when
we examine other cultures than our own; moreover, the “stranger” (in the
sense of “really different”) the other culture, the easier it seems to be to
recognize its presence.
Let us now try to reinterpret Tajfel’s account of stereotype formation.
For a start, it is not incompatible with the two general principles proposed
by Rosch (1978) for the formation of categories: cognitive economy and
perceived world structure. As Rosch (1978: 29) pointed out with reference
to natural objects, “what attributes will be perceived given the ability to
perceive them is undoubtedly determined by many factors having to do
with the functional needs of the knower interacting with the physical and
social environment”. The components of social stereotypes are relative,
too, in part existing in the real world and in part in the eyes of the per-
ceiver. Tajfel and Turner (1979: 43–46) claim that if the social identity of a
given group is perceived as inadequate, a series of either group or indivi-
dualistic strategies are set into motion to achieve positive distinctiveness.

Group strategies include (a) comparing the ingroup to the outgroup on
some new dimension, (b) changing the values assigned to the attributes of
the group, so that comparisons which were previously negative are now
perceived as positive (e.g. “black is beautiful”; the dimension of skin co-
lour remains the same, but the values assigned to it change) and (c) avoi-
ding the use of the high-status outgroup as a comparative frame of refe-
rence.
Stereotypes are also variable in another sense; they change with time,
but older stereotypes may remain, stored in our minds, if still useful. Most,
but not necessarily all Westerners can activate at least three social stereo-
types relative to the Japanese people and society: Imperial Japan, World
War II Japan and present-day Japanese society.
The following question then inevitably arises: provided that the dimen-
sions, attributes and values that intervene to shape social stereotypes are
relative and subject to change, to what extent do stereotypes in the sense of
central images then merely act in terms of categorization effects? It seems
plausible to assign a more active role to such constructs; instead of just
passively reflecting social categories they contribute actively to building
them. As Rosch claims in a recent article (1999: 72):
Concepts and categories do not represent the world in the mind; they are a
participating part of the mind-world whole of which the sense of mind (of
having a mind that is seeing or thinking) is one pole, and the objects of mind
(such as visible objects, sounds, thoughts, emotions, and so on) are the other
pole. Concepts – red, chair, afraid, yummy, armadillo, and all the rest – in-
extricably bind, in many different functioning ways, that sense of being or
having a mind to the sense of the objects of mind.
While both Rosch and Tajfel initially assigned a rather passive role to the
perceiver, they eventually both opted for a far more active role.
Last, but not least, Tajfel’s approach is perfectly compatible with the
concept of social situatedness (Lindblom and Ziemke 2003), according to
which “the development of individual intelligence requires a social (and
cultural) embedding”. In the case of social categories and social stereo-
types, these are surely not effected just as the result of a series of auto-
matic, out-of-context, universal cognitive processes, but when real people
interact with other people in a specific historical moment and a specific
physical environment, with a series of real resources at their disposal.
While such construals are spread within a social group, they are shared
only to a degree. As Sharifian (2003; this volume) reports, the elements of
cultural schemas are not shared by all members of a cultural network, but
rather distributed across the minds of its members. It is furthermore not by
virtue of the knowledge of (or the belief in) only one schema that one be-
comes a member of a cultural group, as two people can share more ele-
ments from one cultural schema and less from another. It is the overall
degree of how much a person draws on various cultural schemas that
makes an individual more or less representative of a cultural group. Sche-
mas thus thrive within groups and the group emerges as such, shaped and
brought into existence by relatively shared beliefs, values and norms. In
similar ways, the group also determines and is determined by speech pat-
terns such as “dialects”, “accents” and “styles”. In a previous paper (Kris-
tiansen 2003) I discussed the link between linguistic and social stereotypes,
and argued that salient speech patterns both metonymically reflect and
more actively build the social groups that brought them about in the first
place. Sound change, including that which in comparative and generative
phonology was thought of in terms of automatic, regular phonetic laws
(e.g. /p, t, k/ in language X became /b, d, g/ in language Y), may accord-
ingly be conceived as systematic ways of rendering a group distinctive, of
achieving psycholinguistic distinctiveness in Speech Accommodation ter-
minology. Social conceptions such as “self, other, us, them” are relevant
also for such seemingly different research areas as linguistic varieties and
phonology.
3. The “cognitive” and “social” functions of social stereotypes
The general claim is that whereas social categorization and social stereo-
typing constitute universal cognitive operations, the delimitations, attrib-
utes, values and dimensions involved are group-dependent. Summing up,
Tajfel attempted to specify both the individual and the collective functions
served by stereotypes. He identified (Tajfel 1981: 143–161) two individual
and three group-level functions:
The individual functions of stereotypes comprise a cognitive function
and a motivational function. The cognitive function renders a complex
social world systematic and manageable through the general process of
categorization (accentuation of intracategory similarities and intercategory
differences). The motivational function represents and defends the values
of the individual through the social values associated with social categori-
zations, as opposed to the neutral values assigned to most categorizations

of natural objects.
The group-related functions of stereotypes include social causality, so-
cial justification and social differentiation. Social causality is when group
ideologies and beliefs are created and maintained for explanatory purposes.
The plague, for instance, was “explained” in terms of “causes” such as
these: in 1639 the Scots were accused of having poisoned the wells of
Newcastle and in 1577 Catholic sorcery was held responsible for an out-
break of gaol fever in Oxford. Social justification is when collective action
is justified by means of stereotypical beliefs about other groups. To illus-
trate this function, Tajfel (1981: 156) quotes Kiernan (1972: 24) on the
European attitudes in the imperial age:
The idea of Europe’s “mission” dawned early, but was taken up seriously in
the nineteenth century. Turkey, China, and the rest would some day be pros-
perous, wrote Winwood Reade, one of the most sympathetic Westerners.
“But those people will never begin to advance […] until they enjoy the
rights of men; and these they will never obtain except by means of European
conquest.”
The use of quite similar stereotypical ideologies, one might add, still seem
to play a major role in justifying collective military action.
Finally, the principle of social differentiation is a dynamic process
which can only be understood against the background of relations between
social groups and the social comparisons they make in the context of these
relations. The creation or maintenance of differentiation, or of a ‘positive
distinctiveness’ of one’s own group from others which are relevant to the
group’s self-image seems to be, judging from the accounts of social anthro-
pologists, a widespread phenomenon in many cultures”. (Tajfel 1981: 157).
Spears et al.9 (1997), however, assign an even more central role to the
group than Tajfel and his associates did. They criticize the fact that SIT
views individual needs (i.e. adaptation to cognitive overload, positive self-
esteem) as the underlying motives with respect to the behaviour of group
members and claim that the psychological processes involved in stereo-
9. Russell Spears is a professor in experimental social psychology at the Univer-

sity of Amsterdam, Penelope J. Oakes a professor in psychology at the Austra-
lian National University, Naomi Ellemers a professor in social psychology at
Leiden University, and S. Alexander Haslam a senior lecturer in psychology at
the Australian National University.
typing only emerge when people interact as social beings, guided by col-
lective rather than individual goals.
As Tajfel himself (1981: 146) stated the problem, the cognitive empha-
sis on what, in the early 1980s, was a recent revival of interest in the study
of stereotypes, constituted “but one instance of a much more general trend
of work and thought in social psychology”. This trend, Tajfel argued, was
based on the following two assumptions:
The first is that the analysis of individual processes, be they cognitive or
motivational, is necessary and also (very often) sufficient for the under-
standing of most of the social behaviour and interactions. The second as-
sumption follows from the first: such an analysis need not take into account
theoretically the interaction between social behaviour and its social context.
The latter is seen as providing classes of situation in which the general indi-
vidual laws are displayed. Alternatively, the social context is conceived as
providing classes of stimuli which “impinge” upon social interactions, i.e.
they selectively activate certain individual “mechanisms” or modes of func-
tioning which are already fully in existence. These “individualistic” views
have recently been contested in a number of publications (e.g. Doise, 1987b;
Moscovici, 1972; Perret-Clermont, 1980; Stroebe, 1979; Tajfel, 1978a; see
chapters 2 and 3) and therefore the details of the argument will not be re-
hearsed here once again. It will be enough to say that, in the case of social
stereotypes, ”social context” refers to the fact that stereotypes held in com-
mon by large numbers of people are derived from, and structured by the re-
lations between large-scale social groups or entities. The functioning and
use of stereotypes result from an intimate interaction between this contextual
structuring and their role in the adaptation of individuals to their social envi-
ronments. (Tajfel 1981: 146)
In spite of the criticism of Spears et al., Tajfel had little doubt about the
most adequate sequence of individual and group functions in stereotyping:
It seems that, if we wish to understand what happens, the analytic sequence
should start from the group functions and then relate the individual functions
to them. As we argued […], an individual uses stereotypes as an aid in the
cognitive structuring of his social environment (and thus as a guide for ac-
tion in appropriate circumstances) and also for the protection of his system
of values. In a sense, these are the structural constants of the sociopsy-
chological situation, it is the framework within which the input of the so-
cially-derived influence and information must be adapted, modified and rec-
reated. No doubt, individual differences in personality, motivation, previous
experiences, etc. will play an important part in the immense variety of ways
in which these adaptations and re-creations are shaped. It remains equally
true, however, that – as we argued at the beginning of this chapter – a

stereotype does not become a social stereotype until and unless it is widely
shared within a social entity. As long as individuals share a common social
affiliation which is important to them (and perceive themselves as sharing
it), the selection of the criteria for division between ingroups and outgroups
and the kind of characteristics attributed to each will be directly determined
by those cultural traditions, group interests, social upheavals and social dif-
ferentiations which are perceived as being common to the group as a whole.
(Tajfel 1981: 158)
3.1. Illusory correlation: the power of frames
As Hewstone and Giles (1997: 274–278) once pointed out, while recent
European (and, we might add, Australian) work on the process of stereo-
typing has emphasized its social functions, the North American work has
concentrated on the cognitive processes either leading to or stemming from
stereotyping. The cognitive processes issuing in or deriving from stereo-
typing include illusory correlation and causal attribution. In social psy-
chology there is an extensive amount of research on both notions.
The term illusory correlation refers to a false perception of a relation-
ship where none exists, “an erroneous inference about the relationship
between two categories of events” (Hamilton and Gifford 1976: 392). As
such, it is a cognitive bias, a counterfactual kind of reasoning, similar to
the cognitive shortcut effected by the process of categorization. Work on
illusory correlation began with the work of Loren J. Chapman (1967),10
who showed subjects a series of word pairs projected onto a screen. One
word was projected on the left of the screen and one on the right. The word
pairs were constructed by combining each of four words from one list with
each of three words from a second list. Each left-hand word had a high-
strength associate among the right-hand words, and though each possible
word pair was shown an equal number of times, subjects consistently over-
estimated the frequency of co-occurrence of the words which already had
an associative relationship (such as bacon-eggs, lion-tiger). Two words
from each list were furthermore considerably longer than the rest of the
words and also the frequency of co-occurrence of these two words was
systematically overestimated. The long words were perceptually distinctive
10. See also Chapman and Chapman (1967) for a study on systematic errors in the
report of co-occurrence of diagnostic test signs with patients’ symptoms.
because all the other words were short, and thus statistically infrequent in
comparison with the short words. Chapman thought that in this case it was
the co-occurrence of distinctive stimuli that resulted in an overestimation
of the frequency with which such events occurred together. Illusory corre-
lation thus seems to take place (a) between already associated stimuli and
(b) between two distinctive stimuli.11
Hamilton and Gifford (1976) were the first scholars to apply illusory
correlation to intergroup judgments and test the possibility that stereotypic
judgments can be acquired on the basis of purely cognitive, information-
processing mechanisms. The underlying rationale was as follows: interac-
tion with minority groups (e.g. blacks in the case of the typical white sub-
urbanite) is a relatively infrequent occurrence, and as such a distinctive
event. Non-normative, undesirable behaviour is statistically less frequent
than desirable behaviour, and also distinctive. As the frequency of co-
occurrence of distinctive events are overestimated, distinctive behaviour on
the part of members of minority groups will be overestimated. In their
study, Hamilton and Gifford labelled their groups Group A and Group B, to
avoid establishing reference to previously formed associations. Subjects
were exposed to 39 statements describing 26 behaviours about members of
group A (the majority group) and 13 behaviours about members of group B
(the minority group). The ratio of positive to negative types of behaviour
for each group was the same: 18 positive and 8 negative statements about
members of group A versus 9 positive and 4 negative statements about
members of group B. The 27 moderately desirable and 12 moderately un-
desirable behaviour descriptions (i.e. “visit a sick friend in the hospital”)
had been selected through a previous test involving 95 common, everyday
behaviours. The hypothesis, which was confirmed, was that subjects would
overattribute the number of undesirable behaviours to members of group B.
Only one-third of the undesirable statements described members of group
B, but over half of them were attributed to this group by the subjects as
shown in Table 1.
The differential perception of minority and majority groups, Hamilton
and Gifford argue, can thus result solely from the cognitive mechanisms
that process information about statistically infrequent co-occurring events.
To Hewstone and Giles (1997), however, a purely cognitive approach is
insufficient, as it fails to consider not only the functions of stereotypes,
11. See McGarty et al (1993) for an overview of memory-based explanations of

illusory correlation.
apart from that of simplifying information processing, but also the reasons
why certain minorities are singled out for discrimination, different minori-
ties are liked or disliked with varying intensity, and certain dimensions and
attributes are brought into play while others are not.
Table 1. Results of Experiment 1 for Attributions of Group Membership. Adapted

from Hamilton and Gifford 1976: 397.
(a) Distribution of stimulus sentences (b) Attributions of group membership

means
Group Group assigned by subject
Behaviors A B Behaviors A B
Desirable 18 9 Desirable 17.52 9.48
Undesirable 8 4 Undesirable 5.79 6.21
3.2. Causal attribution
In illusory correlation, when two unrelated variables are perceived as re-

lated, it is often the case that coincidence is interpreted as causation. Once
social frames have been set up, it is easier to make the facts fit the frames
than modify the frames to fit the facts. Stereotyped frames become ex-
planatory in themselves, and no other dimensions of causality are brought
in to complete the picture. As Hewstone and Giles (1997: 276) phrased it,
“a set of processes conspire to ensure that outgroup members are ‘damned
if they do and damned if they don’t’”. Hamilton and Rose (1980) also
tested the role of illusory correlation in the maintenance of social stereo-
types. They found that stereotypic expectations influenced subjects’ judg-
ments of how frequently various attributes characterized group members in
the stimulus sentences: subjects perceived a relationship between variables
(traits and group membership) if it confirmed a stereotype, even if no evi-
dence was presented to support it.
3.3. The group as a polarizer of attitudes
The group, moreover, seems to act as a polarizer of not only opinions and
attitudes, but also of judgments in general. In 1969, Serge Moscovici and
Marisa Zavalloni challenged two widely held assumptions: (a) that group
judgments are less extreme than individual judgments and (b) that the risky
shift phenomenon is a content-bound exception to the averaging tendency
of the group. Risky shift studies generally implement the following proce-
dure: first, subjects are told to make a series of choices (dilemmas between
various alternatives) on an individual basis, each choice representing vari-
ous degrees of risk. Second, subjects are made to form groups that are re-
quired to select one level of risk which is unanimously acceptable to all
members of the group for each problem, and, third, the subjects are sepa-
rated once again and asked to state at what level they are willing to take
risks. The results are that groups are generally riskier than individuals, and
that individual postconsensus ratings largely correspond to the group con-
sensus ratings.
Moscovici and Zavalloni argued that risky shifts occur in any domain
where normative commitment has an influence on group behaviour. Let us
briefly examine the antecedents of the research in question. As Moscovici
and Zavalloni report, F. H. Allport (1924, 1962), Sherif (1935) and Kogan
and Wallach (1966) all obtained similar results: group judgments represent
the average of prior individual judgments. Kogan and Wallach, for in-
stance, asked a group of subjects to make individual judgments and then,
once constituted as a group, to achieve consensus of each prior judgment.
The group consensus reflected the average position of the individuals that
comprised the group. However, James Stoner´s (1961) discovery of the
risky shift challenged these findings: Stoner found that “when discussing
problems concerning possible loss of money, prestige or self-satisfaction,
groups tend to prefer a riskier alternative than one which would have re-
sulted from a compromise between the choices of the individuals compris-
ing these groups” (Moscovici and Zavalloni 1969: 126). The group seems
to act as a polarizer in the sense that it accepts higher levels of risk than do
the individuals who make up the group. As Moscovici and Zavalloni recall,
two different models have been proposed in order to explain risky shifts: a
social facilitation model and a normative model. The former has to do with
diffusion of responsibility: in group interaction there is a sense of shared
responsibility and a loss of personal responsibility for the consequences of
the decisions made, which, according to Wallach and Kogan (1965) result
from the affective bonds formed in group discussion. Bateson (1966) and
Flanders and Thistletwaite (1967), however, demonstrated that group dis-
cussion12 seems not to be the decisive factor in risky shifts: their studies
revealed that risky shifts occurred without group discussion, when indi-
viduals either familiarized themselves with some choice-dilemma items or
prepared arguments for future group debate.
The normative model thus seems more plausible: according to this
model groups are riskier because individuals who are more daring than
their peers are rewarded. In group discussions, the process of social com-
parison between individuals thus forces individuals to maintain their image
as risk takers to conform to the values and attitudes in the society. The
group as a whole is then led to shift towards a more risky position. When
the individual prepares himself for interaction with others, Moscovici and
Zavalloni claim (1969: 129), a group-anchored frame of reference is acti-
vated although the individual is still alone.
4. Conclusions
In this paper I have briefly drawn attention to a series of empirical studies

in the field of social psychology that all point in the same direction; the
group indeed seems to have quite a decisive and forceful influence. Acting
almost as a kind of hidden hand that guides individual action, the group
creates and maintains the social norms that govern so much of our behav-
iour. It defines the individual’s place in society, acts as a polarizer of atti-
tudes and judgments and intervenes in processes of causal attribution of
events. At the same time, individual and pragmatic choice-making among
the many meaningful variants at our disposal sets up social identities and
intervenes actively in the contrual of Cultural Cognitive Models, or ICMs.
To what extent, then, can it be justified to consistently extrapolate facts
about the individual directly to the human species? The group, it is
claimed, is psychologically very real to us; a cognitive construal which is
likely to be more basic than more abstract, superordinate categorizations
such as [HUMAN] – and in many aspects more basic than the individual
itself. It is curious, furthermore, to observe the correlations between “indi-
vidual” and “cognitive” on the one hand, and “group” and “social” or
“cultural” on the other hand. Moreover, when individual motives are be-
lieved to underlie a mental operation or schema, the term “cognitive” tends
12. Cf. Myers (1982) for an account of the polarizing effects of social interaction in
general.
to be used as a descriptor, but when it is group-related factors that effect

mental operations or schemas, the terms “social” or “cultural” are pre-
ferred, as if socially derived schemata established in the mind of the indi-
vidual were less cognitive in nature – or as if the term “cognitive”
prototypically evokes features such as “mental”, “individual” and “auto-
matic”.
This paper has focused on those aspects of meaning construal which, or
so we believe, cannot be attributed to a relationship between the individual
and the physical surrounding world. Consequently, it is not sufficient to
think of abstract conceptual structure as something which arises either by
metaphorical projection from the domain of the physical to abstract do-
mains or by the projection from basic-level categories to superordinate and
subordinate categories (cf. section 1). Abstract conceptual structure is also,
perhaps crucially, a social phenomenon; something which emerges as the
result of situated embodiment, i. e. when real people interact with other
people in a specific historical moment and a specific physical environment,
and with a series of real resources at their disposal.
Finally, it is also clear that the very same controversy that is currently
an issue in Cognitive Linguistics finds its counterpart in longstanding de-
bates in many other research fields, Social Psychology and Linguistics in
general included: are “social” (i.e. group-related and language-external)
factors decisive in processes of linguistic change or merely triggers of
change? Are the real, underlying causes “cognitive” in nature? The ques-
tion is whether in Cognitive Linguistics the term “cognitive” will remain as
a notion which can be viewed in strict opposition to the term “social”.
References
Allport, Gordon W.
1954 The Nature of Prejudice. Cambridge, MA: Addison Wesley.
Allport, Floyd H.
1924 Social Psychology. Boston: Houghton Mifflin.
1962 A structuronomic conception of behaviour. Individual and collective:
I. Structural theory and the master problem of social psychology.
Journal of Abnormal and Social Psychology 64: 3–30.
Bateson, Nicholas
1966 Familiarization, group discussion and risk-taking. Journal of Expe-
rimental Social Psychology 2: 119–129.
Bernárdez Sanchis, Enrique

2002 Lenguaje y cultura en la cognición espacial del cha´palaachi. Plenary
lecture delivered at the III Biannual Conference of Spanish Cognitive
Linguistics Association, Valencia, May 2002.
Chapman, Loren J.
1967 Illusory correlation in observational report. Journal of Verbal
Learning and Verbal Behavior 6: 151–155.
Chapman, Loren J. and Jean P. Chapman
1967 Genesis of popular but erroneous psychodiagnostic observations.
Journal of Abnormal Psychology 72 (3): 193–204.
2002 Embodiment. Encyclopedia of Cognitive Science 1102–1108. Mac-
Millan Publishers.
Dirven, René
2004 Major Strands in Cognitive Linguistics. Essen: LAUD 2004, Paper
No 634.
Dirven, René and Frank Polzenhagen
2004 Rationalist or Romantic Model in Language Policy and Globalisa-
tion. Essen: LAUD 2004, Paper No 622.
Festinger, Leon
1954 A theory of social comparison processes. Human Relations 7: 117–
140.
Flanders, James P. and Donald L. Thistlethwaite
1967 Effects of familiarization and group discussion upon risk taking.
Journal of Personality and Social Psychology 5 (1): 91–97.
Hamilton, David L. and Robert K. Gifford
1976 Illusory correlation in interpersonal perception: A cognitive basis of
stereotypic judgments. Journal of Experimental Social Psychology
12: 392–407.
Hamilton, David L. and Terrence L. Rose
1980 Illusory correlation and the maintenance of stereotypic beliefs. Jour-
nal of Personality and Social Psychology 39 (5): 832–845.
Hewstone, Miles and Howard Giles
1997 Social groups and social stereotypes. In: Nikolas Coupland and
Adam Jaworski (eds.), Sociolinguistics. A Reader and Coursebook,
270–283. London: MacMillan.
Johnson, Mark
Kiernan, Victor G.
1972 The Lords of Human Kind: European Attitudes to the Outside World
in the Imperial Age. Harmondsworth: Penguin Books.
Kogan, Nathan and Michael A. Wallach

1966 Modification of a judgmental style through group interaction. Jour-
nal of Personality and Social Psychology 4: 165–174.
Kristiansen, Gitte
2003 How to do things with allophones: Linguistic stereotypes as cognitive
reference points in social cognition. In: René Dirven, Roslyn M.
Thought. Ideology, Metaphors and Meanings, 69–120. Berlin/New
Lakoff, George
1987 Women, Fire and Dangerous Things: What Categories Reveal About
1999 Philosophy in the Flesh. The Embodied Mind and its Challenge to
McGarthy, Craig, S. Alexander Haslam, John C. Turner and Penelope J. Oakes
1993 Illusory correlation as accentuation of actual intercategory difference:
Evidence for the effect with minimal stimulus information. European
Journal of Social Psychology 23: 391–410.
1945 Phénoménologie de la Perception. Paris: Gallimard.
Moscovici, Serge and Marisa Zavalloni
1969 The group as a polarizer of attitudes. Journal of Personality and
Social Psychology 12: 125–135.
Myers, David G.
1982 Polarizing effects of social interaction. In: Hermann Brandstatter,
James H. Davis and Gisela Stocker-Kreichgauer (eds.), Group Deci-
sion-making, 125–161. London: Academic Press.
Rosch, Eleanor
1973 On the internal structure of perceptual and semantic categories. In:
Timothy E. Moore (ed.), Cognitive Development and the Acquisition
of Language, 111–144. New York: Academic Press.
1975 Cognitive reference points. Cognitive Psychology 7: 532–547.
1978 Principles of categorization. In: Eleanor Rosch and Barbara B. Lloyd
(eds.), Cognition and Categorization, 27-48. Hillsdale, N J: Law-
rence Erlbaum.
1999 Reclaiming concepts. Journal of Consciousness Studies 6 (11): 61–
77.
Rosch, Eleanor, Carolyn Mervis, Wayne Gray, David Johnson and Penny Boyes-
Braem
1976 Basic Objects in Natural Categories. Cognitive Psychology 8: 382–
439.
Sego, Lewis P.
2003 “Housewife” – no longer a valuefree cultural model? In: René Dir-
ven, Roslyn M. Frank and Martin Pütz (eds.), Cognitive Models in
Language and Thought. Ideology, Metaphors and Meanings, 229–
243. Berlin/New York: Mouton de Gruyter.
Sharifian, Farzad
2003 On cultural conceptualisations. Journal of Cognition and Culture 3
(3): 187–207.
Sherif, Muzafer
1935 A study on some social factors in perception. Archives of Psychology
27: 1–60.
Shweder, Richard A., Manamohan Mahapatra and Joan G. Miller
1987 Culture and moral development. In: Jerome Kagan and Sharon Lamb
(eds.), The Emergence of Morality in Young Children, 1–83. Chi-
cago: Chicago University Press.
tive Linguistics 11 (1): 17–41.
Spears, Russell, Penelope J. Oakes, Naomi Ellemers and S. Alexander Haslam
(eds.)
1997 The Social Psychology of Stereotyping and Group Life. Oxford:
Basil Blackwell.
Stoner, James A. F.
1961 A comparison of individual and group decisions involving risk. Un-
published Master’s thesis, School of Industrial Management, Massa-
chusetts Institute of Technology.
Tajfel, Henri
1969 Cognitive aspects of prejudice. Journal of Social Issues 25: 79–97.
1978 Social categorization, social identity and social comparison. In: Henri
Tajfel (ed.), Differentiation between Social Groups, 61–76. London:
Academic Press.
1981 Human Groups and Social Categories. Cambridge: Cambridge Uni-
versity Press.
Tajfel, Henri and John C. Turner
1979 An integrative theory of intergroup conflict. In: William G. Austin
and Stephen Worchel (eds.), The Social Psychology of Intergroup
Relations, 33–47. Monterey, CA: Brooks/Cole.
Turner, John C., Michael A. Hogg, Penelope J. Oakes, Stephen D. Reicher and
Margaret S. Wetherell
1987 Rediscovering the Social Group: A Self-categorization Theory. Ox-
ford: Basil Blackwell.
Tyler, Stephen
1995 The semantics of time and space. American Anthropologist 97: 567–
592.
Vygotsky, Lev S.
esses. Cambridge, MA: Harvard University Press. Edited by Michael
Cole, Vera John-Steiner, Sylvia Scribner and Ellen Souberman.
Original texts issued in 1930 and 1960.
Wallach, Michael A. and Nathan Kogan
1965 The roles of information, discussion and consensus in group risk
taking. Journal of Experimental Social Psychology 1: 1–19.
Yu, Ning
1998 The Contemporary Theory of Metaphor. A Perspective from Chinese.
Amsterdam: John Benjamins.
Zadeh, Lotfi A.
1965 Fuzzy sets. Information and Control 8: 338–353.
Index
abduction 139, 265, 277, 279, 280, 165, 166, 211, 253, 297, 326,
282, 283 356, 360, 384, 409, 429
activity 9, 12, 23, 33, 69, 81, 102, competition 10, 12, 169, 171, 172,
104, 105, 109–111, 124, 130, 181, 184, 185, 188, 190, 194,
131, 137, 144, 148–153, 155– 248, 292, 304, 308, 319, 416
160, 162, 165, 171, 188, 221, complex adaptive systems 10, 105,
241, 242, 249, 253, 333, 350, 109, 115, 128, 131, 161, 215,
355, 369, 377, 380, 394 216, 239, 240–242, 244, 323,
affect 27, 53, 58, 59, 70, 73, 173, 356, 381
202, 273, 285, 387, 389, 401, 413 conceptual metaphor/-s 12, 28, 31,
anthropology of the body 15, 77, 78, 32, 85, 93, 97, 100, 101, 122,
91, 163 125, 169, 173, 181, 303, 306–
309, 319, 324, 363, 364, 367,
backformation 265, 268, 278, 280– 374, 375, 378, 382, 384, 387–
282, 284, 286 389, 391, 393, 400, 402–405
bodily experience 12, 23, 24, 40, 61, conceptual metaphor theory 363,
63, 78, 83, 100, 110, 125, 266, 375, 378
303, 310, 387–389, 393, 395, constructivist perspective 53, 55
400, 401, 403, 410 CONTAINER image schema 169,
body politic 91, 301–303, 316, 319 173–175, 185, 402
corpus-based analysis 301, 308,
chain of being 301–304 309, 311, 312
circular triangle relationship 12, cultural cognition 107–109, 111–
387, 388, 404, 405 116, 118, 119, 121–123, 128,
Cognitive Linguistics 1–3, 5, 6, 14– 129, 165, 260, 328, 329, 359, 395
18, 21, 27, 29, 37, 42, 43, 45, 50, cultural conceptualisations 109, 113,
51, 62, 63, 77, 78, 81, 94, 102, 118, 121, 124–127, 129, 134,
105–111, 116, 117, 119, 132, 165, 329, 355, 359, 431
135, 136, 140, 145, 149, 150, cultural model/-s 4, 7, 8, 12, 31, 85,
157–160, 162, 165, 166, 172– 98, 109, 119–123, 125, 127, 131–
174, 191, 246, 297–300, 322, 133, 136, 157, 163, 179, 243,
328, 363, 382, 384, 406, 407, 257, 303, 370, 376, 382, 387,
409, 410, 428, 429, 431 389, 393, 401–403, 409, 431
cognitive modelling 327, 330, 332, cultural phenomenology 77, 78, 95,
341, 343, 344, 346, 351, 355 102–104
Cognitive Science 5, 7, 13, 21, 36,
41, 48, 49, 51, 76, 108, 130, 132, deontic schemata 409, 413, 414
133, 135, 149, 154, 160–163, discourse analysis 363–365, 374
434 Index
discourse metaphor formation 10, 231, 238, 240, 241, 243, 246–
215–220, 224, 225, 231, 241– 249, 255, 256, 259, 261, 301,
247, 249, 251 304–308, 310, 314, 324, 326,
discourse metaphors 153, 157, 173, 328, 333, 342, 350, 364, 376, 411
252, 304, 306, 318, 363–365, evolutionary biology 22, 169, 261
367–370, 372–380 evolutionary linguistics 216, 239,
discourse studies 327–329, 365, 241, 245
375, 378 experience 7, 11, 13, 21–23, 25, 27,
distributed embodied cognition 77, 28, 31, 34, 38, 39, 41–47, 49, 53,
80, 96–98 54, 59, 62, 65, 69, 70, 72, 78, 81,
distributed emergent cultural 82, 85, 88, 89, 91, 92, 94, 95, 97,
cognition 327, 328, 351 98, 100–103, 105, 106, 114, 116–
118, 123–125, 128, 142, 147,
embodied and socioculturally 148, 154, 155, 174, 186, 240,
situated cognition 387, 393, 395, 266, 284, 327, 329, 332, 337,
401, 404, 405 340, 343, 344, 347, 348, 353,
embodied versus disembodied 354, 364, 366, 374, 376, 377,
cognition 21–24, 33, 43, 45, 47 380, 388, 389, 393, 401, 403, 410
embodiment 1–3, 5, 7, 8, 11–13, Experiential Realism 21, 25
15–17, 21–26, 31, 35–39, 45, 47,
48, 50, 51–56, 58–63, 65–72, 75, FOXP2 10, 197, 199, 207, 208, 211,
77–81, 83, 90–93, 95, 96, 100– 212
105, 107, 108, 110, 137, 145,
147–150, 157, 159, 162–166, genetic dysphasia 197, 210
174, 253, 259, 265, 266, 268, genetic misattributions 197, 205
270, 282, 284, 286, 288, 290,
297, 301–303, 326, 327, 348, habitus 84, 93, 98, 100, 102, 137,
360, 363, 364, 366, 373, 375, 147, 148, 153–157, 163, 164
377–379, 383, 384, 387, 389, heart of Europe 301, 311–317
410, 431 heterogeneously distributed
emergence 42, 71, 73, 75, 110, 122, cognition 109, 112–114, 118,
161, 164, 201, 216, 240, 261, 128, 129, 355
268, 278–280, 283, 287, 291, 369
emergent cognition 109, 111–115, image schemas 3, 8, 15, 77, 81–90,
120–122 93, 99, 102–106, 174, 175, 266,
emotion 22, 53, 58, 70, 85, 88, 90, 269, 378, 410
97, 136, 145, 273, 310, 390, 403, innateness (nativism) 197, 200, 201
407, 412 intersubjectivity 6, 53, 58, 59, 71,
enunciation 53, 72 73, 74
epistemic schemata 409, 413, 414 invasion biology 169, 172, 173,
evolution 3, 7, 10, 11, 42, 135, 181– 175, 176, 179, 180, 184, 185,
183, 203, 216, 217, 224–226, 187, 192
Index 435
invasive species 9, 169–172, 185, race 216, 217, 222, 224, 225, 227,
193, 194, 258, 365, 383 229–237, 244, 253
radial extension 265, 275, 278
lexical semantics 14, 259, 265, 267, reanalysis 265, 268, 276–280, 282,
268, 278, 284, 285, 288, 299 283, 285, 287
lexicalisation 327, 329, 331, 353, replication 301, 305–307, 318
373 retrojection 77, 98–102
linguistic organicism 216, 221, 222, rhetoric 169, 173, 194, 260
230
schema 3, 18, 40, 47, 61, 70, 78, 81,
meme 247, 249, 250, 301, 305–308, 82, 84, 86–94, 103, 109, 117,
325 119, 120, 125, 127, 132, 173–
Merleau-Ponty 16, 24, 50, 53, 57, 175, 177, 268–270, 272, 274,
61, 70, 75, 93, 105, 107, 149, 282, 285, 287, 290, 304, 352,
163, 412, 430 366, 411, 420, 427
metaphor 2, 4, 7–12, 14, 28, 31, 32, schematicity 86, 363, 364, 371
34, 59, 85, 90, 94, 99–101, 105, semantic change 14, 254, 259, 265,
106, 119, 120, 124, 136, 140, 275, 357
144, 152, 166, 169, 172, 177, semiosis 17, 51, 53, 55–58, 64, 65,
178, 180, 183, 184, 186, 187, 67, 69, 70, 72, 135, 254
189, 191, 193, 194, 201, 203, Semiotics 53, 55, 62, 64, 74
215–223, 225, 230, 235, 237– situated cognition 2, 3, 5–7, 11, 114,
239, 241, 243–253, 255, 256, 137, 151
262, 299, 301, 304–308, 310, situated embodiment 4, 5, 13, 103,
314, 318, 319, 322–324, 327, 149, 266, 267, 276, 409, 428
330, 331, 340, 346, 348, 350, situated meaning 53, 64, 276
352, 354–356, 363–379, 381– social cognition 13, 154, 268, 409,
383, 387–389, 391, 393, 395– 413, 414, 430
398, 401, 404–407, 411 Social Embodiment 21, 24–27, 37,
militarism 169, 172, 181, 185–188 39, 40, 47
mind/body dualism 21, 22, 25, 34 social group 13, 119, 137, 155, 158,
239, 409, 411, 412, 414–416,
neologisms 327, 328, 333 419, 421, 422
social stereotypes 409, 413, 414,
over 59, 265–300 417–420, 422, 425, 429
sociocognitive terminology 327, 332
Peirce 22, 53, 55–58, 64, 66, 74, 75, sociocultural situatedness 1–5, 7, 8,
283 10, 12, 13, 16, 21, 107, 169, 172,
phenomenology 4, 23, 39, 53, 56– 215–218, 242, 245, 247, 252,
58, 91, 92, 94, 96, 105, 412 255, 327, 328, 354, 363, 364,
polysemy 265, 287, 300, 327, 341 366, 374, 377, 378, 387
436 Index
socioculturally situated cognition subject 36, 39, 53, 57, 68, 69, 71–
77, 79, 85, 102, 103 74, 87, 92, 103, 107, 115, 124,
species 9, 40, 42, 44, 140, 149, 169, 221, 235, 237, 240, 242, 244,
170–172, 174–181, 184–190, 245, 302, 305, 333, 348, 366,
194, 195, 206, 216, 217, 224– 400, 412, 419, 425
237, 244, 246, 255, 258, 261, subjectivity 6, 38, 53, 58, 59, 70–72
409, 411, 412, 427 synergic cognition 9, 137, 151, 153,
splicing 11, 261, 327–334, 336– 159
354, 356, 384
usage-based model 142, 265, 298

Body, Language and Mind Volume 2: Sociocultural Situatedness

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Body, Language and Mind Volume 2: Sociocultural Situatedness

Încărcat de

Drepturi de autor:

Formate disponibile

Body, Language and Mind

Volume 2: Sociocultural Situatedness

Library of Congress Cataloging-in-Publication Data

Body, language, and mind. Volume 1, Embodiment / edited by Tom

Bibliographic information published by the Deutsche Nationalbibliothek

List of contributors vii

Introduction: Sociocultural situatedness

Section A: The dynamics of cultural categorization

An interview with Mark Johnson and Tim Rohrer:

Beyond the body: Towards a full embodied semiosis

Properties of cultural embodiment:

Distributed, emergent cultural cognition,

Collective cognition and individual activity:

Section B: The sociocultural situatedness of scientific discourse

Entangled biological, cultural and linguistic

The language-organism-species analogy:

Section C: Sociocultural situatedness in lexical and usage-based

Toward a socially situated, functionally embodied

The embodiment of Europe: How do metaphors evolve?

Sociocultural situatedness of terminology in

Section D: Exploring the sociocultural situatedness of language

The relationship between metaphor, body and culture

Idealized cultural models: The group as a variable

Enrique Bernárdez is Professor of English linguistics at the Complutense

Robson de Souza Bittencourt is a PhD student in Linguistics at the Eng-

Roslyn M. Frank is Professor Emeritus in the Department of Spanish and

Iina Hellsten is a Research Fellow at the Royal Netherlands Academy of

Joseph Hilferty, a San Francisco Peninsula native, graduated from San

Michael Kimmel (PhD Vienna University 2002, MA 1995) is a researcher

Gitte Kristiansen is Assistant Professor in linguistics at the Department of

Brendon Larson received an M.Sc. in evolutionary ecology from the Uni-

Andreas Musolff is Professor of German Language at Durham University

Brigitte Nerlich is a Principal Research Officer at the Institute for the

tory of semantics and pragmatics, cognitive semantics, figurative language,

Roberta Pires de Oliveira is a professor of semantics at Universidade

Kurt Queller (PhD, Stanford, 1994) teaches linguistics and languages

Gender-inclusivity in the town ordinances of medieval Douai”, with Ellen

Farzad Sharifian, PhD, is a Senior Lecturer in the School of Languages,

Rita Temmerman is the coordinator of Centrum voor Vaktaal en Com-

Fabra. He is the author of The Dissolution of Mind (Rodopi, 2002). Some

Patrizia Violi is Professor of Semiotics at the University of Bologna and

Ning Yu is an Associate Professor at the University of Oklahoma (USA).

Tom Ziemke is Professor of Cognitive Science in the School of Humani-

Jörg Zinken is a Senior Lecturer in the Department of Psychology at the

Whereas Volume 1 concentrates on the concept of embodiment, understood

exploring the complex relationship and influence of language on culture,

The historical roots of the concept that we refer to as sociocultural situat-

question by Zlatev (1997) and criticized for portraying the “embodied

embodiment, situated cognition and sociocultural situatedness have been

represents a loose-knit family of approaches to understanding the mind,

sent volume brings together valuable contributions from a cross-

4. Overview of the various sociocultural perspectives in this

In the present volume, the concept of sociocultural situatedness is ap-

4.1. The dynamics of cultural categorization

The volume opens with an examination of the dynamics of cultural catego-

The contribution of Roberta Pires de Oliveira and Robson de Souza

edge, it is not necessary for every single member of the community to

4.2. Sociocultural situatedness of scientific discourse

discourse of science, most particularly, the uncritical use of the metapho-

4.3. Sociocultural situatedness in lexical and usage-based approaches to

The contributions composing Section 3 deal with sociocultural situatedness

ness and particularly actual usage as the source of people’s automatic

4.4. Exploring the sociocultural situatedness of culture and cognition