Dirk Hermans, Bernard Rimé, Batja Mesquita - Changing Emotions

Changing Emotions
Edited by
Dirk Hermans Bernard Rimé Batja Mesquita
Changing Emotions
The question ‘how far can emotions be changed?’ lies at the heart of innumerable
psychological interventions. Although often viewed as static, changes in the intensity,
quality, and complexity of emotion can occur from moment to moment, and also over
longer periods of time, often as a result of developmental, social, or cultural factors.
Changing Emotions highlights several recent developments in this intriguing

domain, and provides a comprehensive guide for understanding how and why
emotions change. The chapters are organized into five parts:
●● Lifespan Perspective
●● Learning Perspective
●● Social-Cultural Perspective
●● Emotional-Dynamics Perspective
●● Intervention Perspective
In each chapter, an internationally renowned scholar presents a concise review

of key findings from their own research perspective. The book will be of great
interest to researchers in the area of emotion and emotion regulation, as well as
related fields such as developmental psychology, educational psychology, social
and clinical psychology and psychotherapy. It may also be of interest to sociolo-
gists, philosophers and economists interested in learning more about emotions.
Dirk Hermans is a Professor and the Director of the Center for the Psychology
of Learning and Experimental Psychopathology at the University of Leuven,
Belgium. His work focuses on associative learning and fear, autobiographical
memory and the study of the automatic affective processing of stimuli.
Bernard Rimé is Emeritus and Invited Professor at the University of Louvain at

Louvain-la-Neuve, Belgium and is a Past President of the International Society
for Research on Emotion (ISRE) and the Belgian Psychological Society. His
research examines how emotional episodes stimulate interpersonal and collective
communication as well as the consequences of sharing emotions with others.
Batja Mesquita is a Professor and the Director of the Center for Social and
Cultural Psychology at the University of Leuven, Belgium. Her work focuses
on the cultural and social constituents of emotion and emotion regulation. She is
currently a Senior Editor for Psychological Science.
Changing Emotions
Edited By
Dirk Hermans, Bernard Rimé, and
Batja Mesquita
Proceedings of the Contactforum “Changing Emotions” (October 23, 2009)

supported by the Royal Flemish Academy of Belgium for Science and the Arts.
First published 2013
by Psychology Press
27 Church Road, Hove, East Sussex BN3 2FA
Simultaneously published in the USA and Canada
by Psychology Press
711 Third Avenue, New York, NY 10017
Psychology Press is an imprint of the Taylor & Francis Group, an informa business
© 2013 Psychology Press
The right of the editors to be identified as the authors of the editorial material, and of the authors for
their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright,
Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form
or by any electronic, mechanical, or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or retrieval system, without permission
in writing from the publishers.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this book has been requested
ISBN13: 978-1-84872-090-9 (hbk)
ISBN13: 978-1-84872-136-4 (pbk)
ISBN13: 978-0-203-07563-0 (ebk)
Typeset in Times by Fakenham Prepress Solutions, Fakenham, Norfolk NR21 8NN
Contents
Contributors ix
Preface xii
Dirk Hermans, Bernard Rimé, and Batja Mesquita
Part 1
Lifespan Perspective 1
1 How kids keep their cool 3

Young children’s use of cognitive strategies to regulate emotion
Linda J. Levine, Robin L. Kaplan, and Elizabeth L. Davis
2 Defining and regulating the self through emotion narratives 10

Robyn Fivush
3 Age-related changes in empathy-related responding 17

Nancy Eisenberg, Jennifer Betkowski, and Tracy L. Spinrad
4 Children’s expressive behavior in different cultural contexts 24

Linda A. Camras and Michael M. Shuster
5 Shifts in emotional experience and regulation across adulthood 31

Tammy English and Laura L. Carstensen
6 Changing the neural mechanism of emotion regulation in children

with behavior problems 37
Steven Woltering and Marc D. Lewis
vi Contents
part 2
Learning Perspective 45
7 Individual differences in the acquisition of fears 47

Susan Mineka
8 Extinction learning and its retrieval 53

Michelle G. Craske and Bram Vervliet
9 Mechanisms of extinction in emotional regulation 60

James Byron Nelson
10 Generalization as a basis for emotional change 67

Perceptual and non-perceptual processes
Dirk Hermans and Frank Baeyens
11 Learning mechanisms in the acquisition of disgust 74

Peter J. de Jong
12 Preclinical analysis of developmental transitions in the

extinction of learned fear 81
From infancy through adolescence to adulthood
Bridget L. Callaghan, Stella Li, Jee H. Kim, and Rick Richardson
part 3
Social-Cultural Perspective 89
13 Can socially sharing emotions change emotions? 91

Bernard Rimé
14 From group-based appraisals to group-based emotions 97

The role of communication and social sharing
Vincent Yzerbyt and Toon Kuppens
15 Emotion and emotion regulation 105

Robert W. Levenson
16 Emotional climate 113

How is it shaped, fostered, and changed?
Darío Páez, Agustín Espinosa, and Magdalena Bobowik
17 Dynamics of ideal affect 120

Jeanne L. Tsai
Contents vii
18 Emotional acculturation 127
Jozefien De Leersnyder, Batja Mesquita, and Heejung Kim
part 4
Emotional-Dynamics Perspective 135
19 Emotion regulation 137

Two souls in one breast?
Nico H. Frijda
20 Understanding emotion change requires an understanding of

emotion causation 144
Agnes Moors
21 Learning to self-generate positive emotions 151

Barbara L. Fredrickson
22 The role of control in emotion, emotion regulation, and empathy 157

Kevin Ochsner
23 What time can tell us 166

The temporal dynamics of emotion regulation
Ravi Thiruchselvam and James J. Gross
24 The duration of emotional episodes 174

Iven Van Mechelen, Philippe Verduyn, and Karen Brans
part 5
Intervention Perspective 181
25 Can expressive writing change emotions? 183

An oblique answer to the wrong question
James W. Pennebaker and Jason D. Ferrell
26 The powerful impact of mental imagery in changing emotion 187

Arnaud Pictet and Emily A. Holmes
27 Cognitive mechanisms involved in therapeutic change for depression 195

Reducing abstraction and increasing concreteness
Edward R. Watkins
28 A functional approach to the study of human emotion 202

The centrality of relational processes
Dermot Barnes-Holmes and Sean Hughes
viii Contents
29 Self-regulation as a mediator of change in psychotherapy 209
Timothy J. Strauman, Megan M. Klenk, and Kari M. Eddington
30 Mindfulness-based interventions 217

The dialectic of changing emotions by accepting them
Pierre Philippot and Alexandre Heeren
Postscript: Experimental rigor and clinical complexity 223

Les Greenberg
Index 230
Contributors
Frank Baeyens, University of Leuven, Belgium

Dermot Barnes-Holmes, National University of Ireland, Maynooth, Ireland
Jennifer Betkowski, Arizona State University, USA
Magdalena Bobowik, University of the Basque Country UPV/EHU, Spain
Karen Brans, University of Leuven, Belgium
James Byron Nelson, University of the Basque Country UPV/EHU, Spain
Bridget L. Callaghan, University of New South Wales, Australia
Linda A. Camras, DePaul University, USA
Laura L. Carstensen, Stanford University, USA
Michelle G. Craske, University of California, Los Angeles, USA
Elizabeth L. Davis, University of California, Riverside, USA
Peter J. de Jong, University of Groningen, the Netherlands
Jozefien De Leersnyder, University of Leuven, Belgium
Kari M. Eddington, University of North Carolina at Greensboro, USA
Nancy Eisenberg, Arizona State University, USA
Tammy English, Stanford University, USA
Agustín Espinosa, Pontificia Universidad Católica del Perú PUCP, Peru
Jason D. Ferrell, University of Texas at Austin, USA
Robyn Fivush, Emory University, USA
Barbara L. Fredrickson, University of North Carolina at Chapel Hill, USA
Nico H. Frijda, Amsterdam University, the Netherlands
Les Greenberg, York University in Toronto, Canada
x Contents
James J. Gross, Stanford University, USA
Alexandre Heeren, University of Louvain at Louvain-la-Neuve, Belgium
Dirk Hermans, University of Leuven, Belgium
Emily A. Holmes, MRC Cognition and Brain Science Unit, Cambridge, UK
Sean Hughes, National University of Ireland, Maynooth, Ireland
Robin L. Kaplan, University of California, Irvine, USA
Heejung Kim, University of California, Santa Barbara, USA
Jee H. Kim, University of New South Wales, Australia
Megan M. Klenk, Duke University, USA
Toon Kuppens, Cardiff University, UK
Robert W. Levenson, University of California, Berkeley, USA
Linda J. Levine, University of California, Irvine, USA
Marc D. Lewis, Radboud University Nijmegen, the Netherlands
Stella Li, University of New South Wales, Australia
Batja Mesquita, University of Leuven, Belgium
Susan Mineka, Northwestern University, USA
Agnes Moors, Ghent University, Belgium
Kevin Ochsner, Columbia University, USA
Darío Páez, University of the Basque Country UPV/EHU, Spain
James W. Pennebaker, University of Texas at Austin, USA
Pierre Philippot, University of Louvain at Louvain-la-Neuve, Belgium
Arnaud Pictet, University of Oxford, UK
Rick Richardson, University of New South Wales, Australia
Bernard Rimé, University of Louvain at Louvain-la-Neuve, Belgium
Michael M. Shuster, DePaul University, USA
Tracy L. Spinrad, Arizona State University, USA
Timothy J. Strauman, Duke University, USA
Ravi Thiruchselvam, Stanford University, USA
Jeanne L. Tsai, Stanford University, USA
Iven Van Mechelen, University of Leuven, Belgium
Contents xi
Philippe Verduyn, University of Leuven, Belgium
Bram Vervliet, University of Leuven, Belgium
Edward R. Watkins, University of Exeter, UK
Steven Woltering, University of Toronto, Canada
Vincent Yzerbyt, University of Louvain at Louvain-la-Neuve, Belgium
Preface
The central topic of this book is emotional change. This topic lies at the heart of
professional psychological intervention. Whether in school psychology, in organ-
izational psychology, in clinical psychology, or in psychotherapy, the important
question is “How and to what extent can emotions be changed?”
While fundamental research on emotion initially conceptualized emotions as
static entities, it has increasingly focused on how emotions develop and change
over time, or can be modified. The book brings together a wide range of theoretical
and empirical perspectives relevant to the question of emotional change. Some of
the contributions describe the change for a single interaction or episode, others
describe changes over longer periods of time; e.g., developmental changes from
infancy to adulthood, or even from early adulthood to old age. Different contribu-
tions also vary with regard to the emotional components included; for instance,
changes in experience, neurological changes, or changes in expression. Finally,
some contributions focus on the human capacity for emotional change, and others
on the processes that give rise to such change: examples of the latter are cognitive
emotion regulation or a particular social context.
The idea for the book was born during a one-day conference on ”Changing
Emotions” in Brussels in October 2009, under the auspices of the Royal Flemish
Academy of Belgium for Sciences and the Arts. The line-up of invited talks
presented a remarkable show-case of perspectives on emotion dynamics: devel-
opmental, social, cultural, therapeutic. By all standards, the conference was a
success. Intended as a local meeting, it attracted broad international attention,
with attendants from many European and even non-European countries.
Interactions during and after the conference made clear how much the field
was in need of a broader perspective on change processes. The study of how
emotions change and can be changed is part of various research traditions (e.g.,
basic emotion theories, social cognition, psychology of learning, affective neuro-
science, developmental psychology). Attention for emotional dynamics has long
been present in some domains (e.g., developmental psychology), whereas in other
domains it is newly emerging. However, there is little communication between
the different domains of expertise, and cross-fertilization is even more rare.
With this book, we intended to create a forum in which internationally
renowned scholars would be able to present core findings on the topic of
Preface xiii
Changing Emotions. The format of the book is unique: rather than asking a
restricted number of authors to write extensive chapters, we opted for an approach
in which a larger number of authors were invited to write a short contribution
(3,000 words) on a very specific research topic that was relevant to the study
of emotional change. As a result, the book consists of a series of 30 solid, brief,
high-quality, and highly readable chapters. Each chapter presents a line of study
in concrete language, with emphasis on the relevant empirical work. Each chapter
brings new insights into this intriguing area of emotion research. Each chapter
is concluded by a discussion of how the work has moved the field of emotional
change a scientific step forward, and by a discussion of remaining questions for
future research.
The chapters are thematically organized in five sections: lifespan perspective,
learning perspective, socio-cultural perspective, emotional dynamics perspective
and intervention perspective. Most of these sections include at least one contri-
bution that specifically targets psychobiological mechanisms. The chapters can be
read separately, or by perspective.
The book closes with some afterthoughts by Dr Les Greenberg, who laid
the foundation of Emotion Focused Therapy (EFT). Dr Greenberg is Professor
of Psychology at York University in Toronto (Ontario), and one of the world’s
leading authorities on working with emotions in psychotherapy. He kindly agreed
to present some thought-provoking reflections on the book and the themes that are
described in the chapters.
We are convinced that the chapters in this volume will be an inspiration for
many. Young as well as experienced researchers will find food for thought,
and may find ways of complementing their own expertise with insights from
neighboring fields. The book will spark new inspiration to clinicians and other
practitioners interested in changing emotions as part of their professional work.
To all others, students as well as those merely interested in human emotional life,
the book will provide answers to existing questions, but also raise new questions
and provide directions for further reading.
Dirk Hermans, Bernard Rimé, and Batja Mesquita

Leuven and Louvain-La-Neuve, July 2012
Part 1
Lifespan Perspective
1 How kids keep their cool
Young children’s use of cognitive strategies
to regulate emotion
Linda J. Levine and Robin L. Kaplan
University of California, Irvine
Elizabeth L. Davis
University of California, Riverside
Tina was mad that her friend monopolized the best crayons at preschool; sad that
her mother was too busy to look at her picture when she got home; scared of the
monster that was surely lurking under her bed as she prepared to go to sleep. In
short, it was an ordinary day. Upsetting events occur frequently and are impossible
to avoid at any age. Adults faced with upsetting events can often keep their cool
by drawing flexibly from an extensive toolbox of strategies for managing their
emotional responses. But what emotion regulatory tools do young children have the
knowledge and skills to use? This question is important because learning to manage
emotion is one of the central accomplishments of childhood. Better emotion
regulation skills predict fewer behavior problems, better peer relationships, and
higher academic achievement (e.g., Cole, Martin, and Dennis, 2004; Graziano,
Reavis, Keane, and Calkins, 2007). Moreover, as attested to by the growth of
school-based emotion education programs, there is promise that these skills can be
taught (Davis and Levine, in press; Rice, Levine, and Pizarro, 2007). Research on
the regulatory strengths and limitations that young children bring to such programs
may contribute to fulfilling this promise. This chapter reviews research on young
children’s ability to use a particular set of emotion regulation tools, cognitive strat-
egies, and factors that promote and hinder their acquisition of this ability.
Behavioral and cognitive strategies for regulating emotion

Emotion regulation refers to the processes people use to modify the type,
intensity, duration, or expression of emotion (Koole, 2009). To appreciate how
children learn to alter their emotional responses to events, it is useful to consider
how emotions are evoked in the first place. According to functionalist theories,
people experience emotions when they appraise events as relevant to their goals,
values, or wellbeing. The specific emotion experienced depends in part on further
appraisals such as whether the event is conducive to or thwarts their goals and
whether they can do anything about it (e.g., Scherer, 1999). Because emotions
4 Linda J. Levine, Robin L. Kaplan and Elizabeth L. Davis
depend on appraisals of the relations between events and goals, two broad classes
of strategies can be used to alter emotional experience. People can use behavioral
strategies to change external events so that the events conform to their goals, or
they can use cognitive strategies to change their goals or appraisals of events
(see Koole, 2009 for a more fine-grained analysis of types of emotion regulation
strategies).
Use of behavioral strategies to manage emotion remains relatively constant in
frequency across the lifespan (Heckhausen, Wrosch, and Schulz, 2010). Even infants
can avert their gaze from a stranger who makes them feel wary or increase the intensity
of their cries to elicit help from parents. In contrast, deliberate use of cognitive strat-
egies to manage emotion requires an understanding that goals, thoughts, and emotions
are interrelated and that changing goals and thoughts can lead to changes in emotional
experience. Limitations in children’s mental state knowledge raise questions about
when and how young children acquire the ability to use such strategies.
The development of young children’s use of cognitive strategies

A rudimentary understanding of the links between emotions and goals or desires
emerges early. Children first refer to emotions in spontaneous conversation around
two years of age. By the age of three, children can predict that people will feel
happy if they get something they want, and sad if they do not (Wellman, Phillips,
and Rodriguez, 2000). By four or five years of age, children can predict emotional
responses even when protagonists’ desires conflict with their own (Moore, Jarrold,
Russell, Lumb, Sapp, and MacCallum, 1995). They also demonstrate an under-
standing of the association between emotions and beliefs (e.g., Harris, Johnson,
Hutton, and Andrews, 1989). This understanding emerges as children get better at
inhibiting their own salient mental states (desires and beliefs) and come to appre-
ciate that people’s mental representations of the same events can differ.
Several studies suggest, however, that children do not understand that emotions
can be controlled by thoughts alone until middle childhood. For example, when
asked how a protagonist could stop a negative emotion, eight- and 12-year-olds
routinely described strategies to change mental states such as forgetting about an
aversive event but five-year-olds referred primarily to behavioral strategies for
changing the environment (e.g., Pons, Harris, and de Rosnay, 2004). Similarly,
when five- and eight-year-olds and adults were asked to explain how a story
protagonist’s emotion could change with no external cause, only two out of 20
five-year-olds indicated that cognitive strategies, such as merely thinking about
something happy or reappraising a negative situation, could change a person’s
emotional state without any external input (Flavell, Flavell, and Green, 2001).
Findings such as these have contributed to a growing consensus that, until middle
childhood, limited understanding of the relations between thinking and feeling
restricts children’s ability to generate cognitive strategies for regulating emotion
(e.g., Cole et al., 2004; Pons et al., 2004).
Factors other than lack of knowledge may contribute to these findings,
however. Children produce more sophisticated emotion regulation strategies
How kids keep their cool 5
when asked about situations with which they have had extensive prior experience
(Lagattuta, Wellman, and Flavell, 1997). Moreover, when possible, both children
and adults prefer to change troubling situations directly, so that situations
conform to their desires, rather than accept situations and change their goals or
appraisals instead (Heckhausen et al., 2010). With limited prompting, then, young
children may produce behavioral strategies even if they are capable of producing
cognitive strategies. The roots of children’s understanding of cognitive strategies
may be more apparent when children are asked about highly familiar situations
and given ample opportunity to display their knowledge. Studies that have taken
this approach reveal a more nuanced developmental progression from appreci-
ating the usefulness of cognitive strategies suggested by others, to being able
to produce such strategies with prompting, to being able to produce cognitive
strategies in appropriate situations without prompting.
Children as young as three possess a rudimentary understanding that remem-
bering and forgetting events can influence people’s emotions (Lagattuta et al.,
1997). Extending these findings, Dennis and Kelemen (2009) assessed preschool
children’s understanding of cognitive strategies using a task that limited the
verbal and memory demands on the child. Children watched as puppets described
and acted out different strategies for alleviating negative feelings. Children
then rated the effectiveness of each strategy. Three- and four-year-old children
rated distraction as more effective than rumination, suggesting that preschoolers
recognize the relative effectiveness of some cognitive regulatory strategies.
Davis, Levine, Lench, and Quas (2010) investigated five- and six-year-old
children’s ability to generate cognitive strategies like changing thoughts (e.g.,
deciding to think about something else) and changing goals (e.g., deciding to want
something else) to regulate negative emotions. Children were presented with familiar
hypothetical scenarios such as being unable to go outside to play or having to eat a
disliked food. They were asked twice to suggest strategies that protagonists could
use to make their sad or mad feelings go away (e.g., “If Billy couldn’t do [insert
child’s strategy], then what could he do to make his sad/mad feelings go away?”).
Children were also asked about strategies they had used in their own lives when
faced with situations that made them mad, sad, or afraid. Over half of the children
described at least one cognitive strategy to reduce negative emotion. Specifically,
children described changing thoughts by forgetting, using distraction, or going to
sleep (e.g., “Go to sleep, because when you sleep you don’t know if you have
good days or bad”) and reappraising negative outcomes as positive, temporary, or
unimportant (e.g., “He’ll eat the thing he doesn’t like knowing there’s something
else he likes coming on”). They also described changing goals or desires (e.g., “He
decided he didn’t want to go outside and play”). When relating their own personal
experiences, cognitive strategies were the most frequent type of strategy that
children reported. Moreover, the strategies children described were tailored to the
emotion they intended to regulate. Consistent with functional theories of emotion,
children were most likely to describe taking action when angry about obstacles
to their goals, changing thoughts when frightened by situations characterized by
uncertainty, and changing goals when saddened by irrevocable goal loss.
If children are capable of describing cognitive strategies by five or six years
of age, why has other research suggested that they do not typically do so until
age eight or ten? Babb, Levine, and Arseneault (2010) examined developmental
changes in coping flexibility. Six- to 11-year-olds responded to hypothetical
vignettes about problematic interactions with peers that shifted from control-
lable to uncontrollable over time. As situations became uncontrollable, older
children were much more likely than younger children to shift flexibly from
behavioral strategies directed toward changing external situations to cognitive
strategies directed toward changing their thoughts or goals. This age difference
in flexibility was mediated by younger children’s difficulty in gauging the extent
to which situations were controllable. Thus, even though young children are
capable of generating cognitive strategies, they may have difficulty recognizing
when it is adaptive to use cognitive strategies without prompting from adults.
In late childhood and adolescence, greater flexibility and further differentiation
of cognitive strategies is seen (Fields and Prinz, 1997; Garnefski, Kraaij, and
Spinhoven, 2001). In summary, recent studies have made use of familiar situa-
tions and have given children ample opportunity to display their knowledge. The
results have helped clarify young children’s competencies and limitations in the
use of cognitive strategies to regulate emotion.
Factors that help and hinder children’s use of cognitive strategies

Children’s growing understanding of the links among desires, beliefs, and
emotions provides them with a wider repertoire of emotion regulation strategies
to choose from when faced with upsetting events. But mere knowledge is not
enough when it comes to regulating emotions in daily life. Using a cognitive
strategy such as distraction or reappraisal requires children to preempt their
immediate emotional, cognitive, and behavioral impulses (e.g., to strike out
angrily and take the crayons) in the service of social or long range goals (e.g., to
maintain a friendship with the child monopolizing the crayons). These abilities
fall under the umbrella of executive functioning. Between the ages of three and
six, children make great strides in future planning, engaging in goal-oriented
behavior, and inhibiting impulsive responses. These executive functions depend
importantly on the frontal cortex, one of the last regions of the brain to reach
maturity. Of direct relevance to emotion regulation, these gains mean that, with
age, children are better able to divert their attention from emotion-eliciting events,
substitute long term for immediate goals, and reinterpret events in a positive
manner (Eisenberg, Spinrad, Fabes, Reiser, Cumberland, Shepard, et al., 2004).
All children show improvements in executive functioning with age, but there
are marked individual differences in children’s ability to implement cognitive
strategies to regulate emotion in their daily lives. Intrinsic factors such as
children’s temperamental reactivity and inhibitory control, and extrinsic factors
such as parenting, contribute to these differences. Children vary with respect
to the threshold for the elicitation of negative emotions and in the intensity and
persistence of their reactions. Intense negative emotion captures attention, leaving
fewer cognitive resources available to devote to cognitive regulation strategies.
Thus greater reactivity can make it more difficult for children to down-regulate
emotion (Buss and Goldsmith, 2007). Individual differences in inhibitory control
also contribute to children’s ability to regulate emotion. For example, four-year-
olds who could refrain from touching an attractive toy were also better able to hide
negative emotion when later given an undesirable toy (Carlson and Wang, 2007).
A component of children’s more general capacity for self-regulation or executive
control, this ability to inhibit a prepotent response continues to predict emotion
regulation across childhood. Young children who could not delay eating a treat
immediately in order to obtain a larger treat in the near future were rated in adoles-
cence as less capable of regulating frustration and stress (Mischel, Shoda, and
Peake, 1988). Thus, intrinsic differences in emotional reactivity and inhibitory
control affect children’s ability to use cognitive strategies to manage emotions.
Children’s social environment also contributes to their ability to alter emotions.
Children learn to use regulatory strategies by observing how parents, family
members, and peers manage their emotions. Although adults may use a wide
range of regulatory strategies, cognitive ones like reappraisal or distraction
are less readily observed than behavioral strategies. Because of this, the way
parents talk to their children about emotion plays an important role in shaping
children’s abilities. Parents who talk empathetically with children about their
emotional experiences, supporting children’s attempts to solve problems, engage
in distraction, and reappraise situations, have children who are more proficient
at inhibiting impulses and regulating emotion (Thompson and Goodvin, 2007).
The development of emotion regulation is compromised when stressors or
mental disorders such as depression impair parents’ ability to be responsive to
their children (e.g., Blandon, Calkins, Keane, and O’Brien, 2008). In summary,
children’s ability to grapple with strong emotions can be helped or hindered by
individual differences in children’s temperament and inhibitory control, and by
the extent to which parents promote children’s emotional understanding and
regulatory efforts.
Conclusions
Exploring young children’s ability to regulate their emotions is a critical area
of research inquiry with implications for a range of cognitive, behavioral, and
mental health outcomes (Cole et al., 2004). With age, children transition from
relying on adults for assistance with emotion regulation to assuming greater
responsibility for regulating their own emotions (Thompson and Goodvin, 2007).
In doing so, they are aided by a growing understanding of mental states which
provides them with new regulatory tools to choose from. Children as young as
three or four are able to recognize the effectiveness of cognitive strategies such
as forgetting and distraction. By five or six, children understand that changing
their thoughts and goals can lead to changes in their emotional experience.
Moreover, they can use this knowledge to generate a wide range of cognitive
strategies for regulating emotions including forgetting, distraction, reappraisal
and modifying goals. By middle childhood, children can retrieve these strategies
without prompting and they can shift flexibly from behavioral to cognitive strat-
egies as situations demand. Thus, by the time children begin formal schooling,
they are capable of both understanding and producing cognitive strategies but
need support to retrieve and enact these strategies in appropriate situations. These
research findings may prove useful to parents, practitioners, and school-based
programs dedicated to helping kids keep their cool. Further research is needed
to identify the conditions under which specific cognitive strategies (such as
distraction, reappraisal, and rumination) benefit or hinder children’s learning and
mental health (e.g., Davis and Levine, in press; Garnefski et al., 2001; Wright,
Banerjee, Hoek, Rieffe, and Novin, 2010).
References
Babb, K. A., Levine, L. J., and Arseneault, J. M. (2010). Shifting gears: Coping flexibility
in children with and without ADHD. International Journal of Behavioral Development,
34, 10–23.
Blandon, A. Y., Calkins, S. D., Keane, S. P., and O’Brien, M. (2008). Individual differ-
ences in trajectories of emotion regulation processes: The effects of maternal depressive
symptomatology and children’s physiological regulation. Developmental Psychology, 44,
1110–1123.
Buss, K. A., and Goldsmith, H. H. (2007). Biobehavioral approaches to early socioe-
motional development. In C. A. Brownell and C. B. Kopp (Eds), Transitions in early
socioemotional development: The toddler years. New York: Guilford.
Carlson, S. M., and Wang, T. (2007). Inhibitory control and emotion regulation in
preschool children. Cognitive Development, 22, 489–510.
Cole, P. M., Martin, S. E., and Dennis, T.A. (2004). Emotion regulation as a scien-
tific construct: Challenges and directions for child development research. Child
Development, 75, 317–333.
Davis, E. L., and Levine, L. J. (in press). Emotion regulation strategies that promote learning:
Reappraisal enhances children’s memory for educational information. Child Development.
Davis, E. L., Levine, L. J., Lench, H. C., and Quas, J. A. (2010). Metacognitive emotion
regulation: Children’s awareness that changing thoughts and goals can alleviate negative
emotions. Emotion, 10, 498–510.
Dennis, T. A., and Kelemen, D. A. (2009). Preschool children’s views on emotion
regulation: Functional associations and implications for social-emotional adjustment.
International Journal of Behavioral Development, 33, 243–252.
Eisenberg, N., Spinrad, T., Fabes, R., Reiser, M., Cumberland, A., Shepard, S., et al.
(2004). The relations of effortful control and impulsivity to children’s resiliency and
adjustment. Child Development, 75, 25–46.
Fields, L., and Prinz, R. J. (1997). Coping and adjustment during childhood and adoles-
cence. Clinical Psychology Review, 17, 937–976.
Flavell, J. H., Flavell, E. R., and Green, F. L. (2001). Development of children’s understanding
of connections between thinking and feeling. Psychological Science, 12, 430–432.
Garnefski, N., Kraaij, V., and Spinhoven, S. (2001). Negative life events, cognitive
emotion regulation and emotional problems. Personality and Individual Differences,
30, 1311–1327.
Harris, P. L., Johnson, C. N., Hutton, D., and Andrews, G. (1989). Young children’s theory
of mind and emotion. Cognition and Emotion, 3, 379–400.
Heckhausen, J., Wrosch, C., and Schulz, R. (2010). A motivational theory of lifespan
development. Psychological Review, 117, 32–60.
Koole, S. L. (2009). The psychology of emotion regulation: An integrative review.
Cognition and Emotion, 23, 4–41.
Lagattuta, K., Wellman, H., and Flavell. J. (1997). Preschoolers’ understanding of the link
between thinking and feelings. Child Development, 68, 1081–1084.
Mischel, W., Shoda, Y., and Peake, P. K. (1988). The nature of adolescent competencies
predicted by preschool delay of gratification. Journal of Personality and Social
Psychology, 54, 687–696.
Moore, C., Jarrold, C., Russell, J., Lumb, A., Sapp, F., and MacCallum, F. (1995).
Conflicting desire and the child’s theory of mind. Cognitive Development, 10, 467–482.
Pons, F., Harris, P. L., and de Rosnay, M. (2004). Emotion comprehension between 3 and
11 years: Developmental periods and hierarchical organization. European Journal of
Developmental Psychology, 1, 127–152.
Rice, J. A., Levine, L. J., and Pizarro, D. A. (2007). Just stop thinking about it: Effects of
emotional disengagement on children’s memory for educational material. Emotion, 7,
812–823.
Scherer, K. R. (1999). Appraisal theories. In T. Dalgleish, and M. Power (eds). Handbook
of Cognition and Emotion (pp. 637–663). Chichester: Wiley.
Thompson, R. A., and Goodvin, R. (2007). Taming the tempest in the teapot: Emotion
regulation in toddlers. In C. A. Brownell and C. B. Kopp (eds), Transitions in early
socioemotional development: The toddler years (pp. 320–344). New York: Guilford.
Wellman, H. M., Phillips, A. T., and Rodriguez, T. (2000). Young children’s understanding
of perception, desire, and emotion. Child Development, 71, 895–912.
Wright, M., Banerjee, R., Hoek, W., Rieffe, C., and Novin, S. (2010). Depression and
social anxiety in children: Differential links with coping strategies. Journal of Abnormal
Child Psychology, 38, 405–419.
2 Defining and regulating the self
through emotion narratives
Robyn Fivush
Emory University
A large part of who we are is defined through the stories we tell; autobiographical
narratives move beyond simple descriptions of what happened to include
reflection, evaluation, and interpretation, integrating our thoughts and feelings
into memory in ways that render the event personally meaningful. Indeed, a
large number of experiences we share with others through stories are about
everyday and highly emotional experiences that come to define who we are in
the world and in relation to others (Rimé, 2007). Thus, personal narratives are
a critical mechanism for creating a sense of an emotional self. As adults, self
and autobiographical memory reciprocally interact and define each other. In this
review, I focus on development, and explicate first how emotion narratives begin
to define the self as an emotional being, and second, on how children begin to use
emotion narratives to regulate self through restructuring and reappraising difficult
emotional experiences.
Throughout, I place these concepts in a sociocultural and feminist model
of autobiographical memory development (Fivush, 2010; Nelson and Fivush,
2004). Autobiographical narratives are culturally canonical linguistic forms
that shape the way we understand our experiences. Through sharing our experi-
ences with others through stories, and mutually constructing evaluative and
interpretive frameworks, these stories come to take on personal meaning and
significance, and, in this way, autobiographical narratives become self-defining.
Cultures define the shape of a life, and prescribe how a life should be lived;
through constructing personal narratives that place the individual within cultural
frameworks, individual lives take on meaning. This process occurs in local
social interactions, in which personal stories are told and retold, negotiated and
validated, and in this way, feminist issues of voice, or who has the authority to
tell the story in what way, become paramount.
Developmentally, children learn about who they are as individuals and
members of a culture through participating in parentally structured stories that
provide children with culturally appropriate frameworks for understanding and
evaluating personal experience. Moreover, some experiences and/or aspects of
experiences are validated through elaborated retellings that allow for voice. By
definition, however, if some interpretations are voiced, other interpretations are
silenced. From a feminist perspective on voice and silence, it becomes important
Defining and regulating the self through emotion narratives 11
to move beyond a description of whether emotion is voiced to examine how it is
voiced in order to gain a more nuanced understanding of how emotion defines
and regulates self. What emotions are validated and how emotional experience
is understood through narrative restructuring within parentally scaffolded narra-
tives becomes a site for examining how individuals come to understand their own
emotional experience within canonical cultural frames.
In the first section, I describe the developmental emergence of emotion narra-
tives, and show how children are learning to evaluate their emotional experiences
in particular ways through participating in parentally scaffolded narratives, and
how gender and culture provide frameworks for voicing and silencing emotion.
In the second section, I describe more specifically how children learn to regulate
their emotions through parentally scaffolded narratives. This section focuses
specifically on narratives of negative experiences, and explores how voicing these
kinds of experiences in particular ways may or may not help children learn to
regulate their emotion.
Defining the self through emotion narratives

I start with an excerpt from a conversation between a mother and her four-year-
old child about seeing a bear at a carnival:
Mother: Were you scared (of the bear)?

Child: No.
Mother: No? Not even a little scared?
Child: I’m not scared of bears.
(a few conversational turns later)
Child: It was scary.

Mother: You were scared of the bear?
Child: Yes.
Mother: That’s ok. It’s ok to be scared of bears.
In this excerpt, we see the glimmerings of the daughter’s concept of herself as

an emotional being. Not only does this child experience emotion, but the how
and why of these experiences defines who she is as a coherent self through time.
Whether or not this child was actually scared in this specific situation becomes
irrelevant; in reflecting, interpreting, and evaluating on this experience, the
mother helps her child construct a narrative of herself as someone who is scared
of bears. From a feminist perspective, the mother, who is in a position to claim
authority, provides a particular voicing of this experience, and thus facilitates
a particular construction of self-continuity through time. Moreover, because
personal narratives include information about thoughts and emotions, they also
provide a sense of emotional continuity over time. Although young preschoolers
do not yet understand that past emotions continue to motivate self in the present,
12 Robyn Fivush
by age five- to six-years old, children understand that past emotions are causally
related to present emotions and actions (see Fivush and Nelson, 2006, for a full
explication).
This developing sense of emotional continuity is scaffolded by parents; as in
the example above, children of mothers who talk more about emotions when
co-constructing the shared past have children who themselves come to integrate
more emotions into their own personal narratives, and show higher levels of
emotion understanding (see Fivush, Haden and Reese, 2006, for a review). In
essence, parents are helping their children to learn the forms and functions of
narrating the past; through participating in parentally structured reminiscing,
children are learning both how to tell a coherent narrative and what information
is appropriate to include. In turn, these narratives become a reconstructed filter
through which children understand their own personal experience.
Intriguingly, there are gender differences in how parents structure conversa-
tions about emotions with their preschool children (see Fivush et al., 2006, for
a review). Although there are many nuances to these findings, in general, with
daughters as compared to sons, mothers talk more about emotion and include
a wider variety of emotions. Specific to sadness, both mothers and fathers talk
more about the experience of sadness, and talk more about the causes and conse-
quences of sadness with daughters than with sons. There is also some suggestion
that mothers talk more about anger with sons than with daughters. It is possible
that these gender differences reflect behavioral differences in the experience
and expression of children’s emotions, but it is equally likely (in fact, it is most
likely transactional and reciprocal) that these co-constructed narratives are
helping children to construct a sense of self as an emotional being. In narrating
about past emotional experiences, parents and children are removed from the
heat of the moment and are better able to reflect on and interpret the experience.
For example, a fight with a sibling can be interpreted as an event that made the
child angry or an event that made the child sad. Further, parents can choose to
focus on particular experiences (e.g., times the child was sad) and not talk about
other experiences (e.g., times the child was angry). By selecting and interpreting
particular emotional experiences in particular ways, parents are helping their
children to construct a self that experiences certain kinds of emotions in particular
ways.
By focusing more on emotion, especially sadness, with daughters, parents are
teaching girls that emotion is an important part of their experience, that sadness
is prevalent, and that it is appropriate to share your emotional experiences with
others. In this way, girls are learning to voice emotion. In contrast, boys may be
learning that emotion is not as integral a part of experience, and/or that emotion is
not to be discussed in great depth, and that sadness is not an appropriate emotion
to have. In some sense, boys’ expression of emotion is being silenced. Obviously
this is an overly broad characterization, but by age four, girls narrate personal
experiences that are more richly studded with emotion language, and this gender
difference persists through the lifespan. As adults, females report thinking about
emotion more frequently than do males, valuing emotional experience more and
discussing emotional experience more with others than do males (see Fischer,
2000, for a review). Thus these early parentally structured narratives about past
emotional experiences may facilitate the development of a particular type of
emotion self-concept, as a self that experiences particular emotions in particular
ways.
There are also culture differences in parent-child narratives about emotional
experiences. Again, there are nuances, but the general findings are that Asian
parents talk less about emotion than do European parents (Wang, 2002). More
interesting, Asian mothers provide a different explanatory context for emotions than
do European mothers. Whereas European mothers discuss emotions as autonomous,
internal feelings that create and cause behaviors in the world, Asian parents are
more likely to talk about emotions as arising within social situations and as being
disruptive to the social and communal good. European children are learning
that emotions are part of an autonomously defined agentic self, whereas Asian
children are learning that emotions are to be controlled in order to be in harmony
with others.
Thus early parentally structured narratives about past emotional experiences
seem to be setting the stage for children learning how to understand and interpret
their emotional experience, and these interpretations define self as an emotional
being in the world. For females, especially in European cultures, emotions are
seen as central to self-definition and are an integral part of sharing self with
others, whereas for males, especially in Asian cultures, emotions are silenced,
in the sense that they are not validated as central to self-definition and are not
to be socially shared. I note here that research on emotion disclosure has found
that males self-report sharing emotional events with others to the same extent as
females (Rimé, 2007), but the argument here is about how emotions are discussed
and shared. Although males may tell others about the events that occurred at
the same rate as females, males may not share and elaborate on their emotional
reactions to these events to the same extent that females do (see Fisher, 2000, for
a review). The way in which emotions are shared with others is also critical for
self-regulation, to which I now turn.
Regulating the self through emotion narratives

Through telling and retelling experiences to and with others, we reshape and
reconstruct those experiences in order to create meaning. When sharing the
positive events in our lives, this sharing seems to serve the purpose of social
bonding, and of reliving positive emotions in ways that may be beneficial both
for relationship building and for creating positive mood that allows for healthier
emotional and cognitive functioning (Fredrickson and Joiner, 2002). However,
we also share the more stressful events of our lives, and these are the events that
we often struggle with in order to create sense and meaning.
There is substantial evidence that adults who are able to create more coherent
and emotionally expressive narratives about stressful experiences subsequently
show higher levels of both psychological and physical health (Pennebaker and
14 Robyn Fivush
Chung, 2007). Using an expressive writing paradigm, in which adults are asked
to narrate stressful experiences over several consecutive days, individuals who
use language indicative of narrative coherence (e.g., “because”, “therefore”),
cognitive processing (e.g., “understand”, “realize”), and emotional expression
(use of both negative and positive emotion words) subsequently show lower
levels of depression and anxiety, visit the doctor less, and show better immune
system functioning. The ability to restructure emotional experiences in a more
coherent, comprehensible, and emotionally balanced way allows the individual
to regulate emotional experiences in productive ways.
Yet children, who are still in the process of developing sophisticated emotion
regulation skills, may not benefit from expressive narratives (see Bohanek and
Fivush, 2010, for a review). Expressive writing studies with children between 8-
and 12-years of age show either no effects or actually increased levels of anxiety
following expressive writing, whereas one study with 14-year-olds showed the
adult-like expected benefits of expressive writing. We recently asked 14–16-year
olds to narrate highly stressful events in their lives, and, although females generally
expressed more emotion in their narratives than did males, those males who
expressed more emotion also showed higher levels of well-being whereas there
were no relations for females (Bohanek and Fivush, 2010). Thus, the development
of the ability to use narratives in the service of reappraisal and emotion regulation
may follow a protracted (and gendered) developmental course.
In line with the sociocultural model, research indicates that children learn
to regulate emotion through narrative restructuring by participating in paren-
tally scaffolded co-narrations. Mothers who co-narrate more coherent and
emotionally expressive narratives with their preschool children have children
who show higher levels of emotion understanding (see Fivush et al., 2006, for
a review). As children grow older, parents who scaffold narratives that include
more explanations and resolutions of emotional experiences have preadolescent
children who display higher levels of well-being, both in terms of fewer inter-
nalizing (anxiety, depression, withdrawal) and externalizing (acting out, anger,
aggression) behaviors. Critically, it is not simply the expression of emotion but
specifically parental scaffolding of emotional explanation, in terms of regulation
and resolution, which seems to be related to child well-being. Moreover, as
in early emotional self-definition, gender continues to play a modulating role;
maternal contributions to these co-constructed narratives accounts for more of the
variance in emotional well-being than either the children’s contributions or the
fathers’ contributions, and especially so for girls as compared to boys (see Fivush,
Bohanek, Marin and Duke, 2010, for a review). Clearly, supportive parenting
and open communication is generally facilitative of children’s emotional devel-
opment, but, importantly, maternal reminiscing about past emotional events
predicts unique variance in children’s developing emotional development over
and above other contexts (see Fivush et al., 2006).
The gender patterns highlight developing and evolving relations between
voice and silence in relation to emotion regulation. Whereas females may be
socialized to express more emotion, because parents focus on the attribution
and expression of emotion with daughters (and especially of sadness), girls
may develop an emotionally expressive style that leads to rumination rather
then regulation. There is good evidence that, at least beginning in adolescence,
girls begin to ruminate both when alone and when interacting with their friends
in ways that may be detrimental and lead to depressive symptoms (see Nolen-
Hoeksema et al., 2008, for a review). Thus in thinking about the development of
emotion narratives in relation to voice and silence, the way in which emotions
are voiced must be considered. Narratives that simply express and ruminate
on emotional experience may be detrimental whereas narratives that regulate
emotion through restructuring and reappraisal may have positive benefits
for emotional well-being. Thus the question changes from whether or not
individuals disclose emotional experience, to how they disclose this experience,
and whether they are able to create narratives of emotional experience that move
beyond simple expression to include cognitive restructuring and reappraisal that
allows for emotional resolution and regulation. This is a critical question for
future research.
Conclusions
Narratives are the way in which we create meaning from the emotional
experiences of our lives, and children learn to do this through participating in
parentally scaffolded reminiscing about the past. Parents who provide emotionally
expressive, and especially emotionally explanatory co-constructed narratives
with their preschool through preadolescent children have children who both learn
to tell more elaborated coherent narratives of their own emotional experiences,
and show higher levels of emotional well-being. The way in which this develop-
mental process unfolds by gender (and most likely by culture) provides critical
evidence that our emotional experiences are created in social interactions, and
these interactions facilitate the construction of culturally canonical narratives that
structure the voicing and silencing of particular aspects of emotional life in ways
that define and regulate self.
References
Bohanek, J. G., and Fivush, R. (in press). Personal narratives, well-being, and gender in
adolescence. Cognitive Development.
Fischer, A. H. (2000). (ed.) Gender and emotion: Social psychological perspectives. New
York, NY: Cambridge University Press.
Fivush, R. (2010). Speaking Silence: The social construction of voice and silence in
cultural and autobiographical narratives. Memory, 18, 88–98
Fivush, R., Bohanek, J. G., Marin, K., and Duke, M. (2010). Family narratives and
adolescent well-being. In K. C. McLean and M. Pasupathi (eds). Narrative Development
in Adolescence: Creating the Storied Self (pp. 45–64). New York, NY: Springer.
Fivush, R., Haden, C. A., and Reese, E. (2006). Elaborating on elaborations: Maternal
reminiscing style and children’s socioemotional outcome. Child Development, 77,
1568–1588.
16 Robyn Fivush
Fivush, R., and Nelson, K. (2006). Parent-child reminiscing locates the self in the past.
British Journal of Developmental psychology, 24, 235–251.
Fredrickson, B.L., and Joiner, T. (2002). Positive emotions trigger upward spirals toward
emotional well-being. Psychological Science, 13, 172–175.
Nelson, K., and Fivush, R. (2004). The emergence of autobiographical memory: A social
cultural developmental model. Psychological Review, 111, 486–511.
Nolen-Hoeksema, S., Wisco, B. E., and Lyubomirsky, S. (2008). Rethinking rumination.
Perspectives on Psychological Science, 3, 400–424.
Pennebaker, J. W. and Chung, C. K. (2007). Expressive writing, emotional upheavals, and
health. In H. S. Friedman and R. C. Silver (eds), Foundations of health psychology (pp.
263–284). New York, NY: Oxford University Press.
Rimé, B. (2007). The social sharing of emotion as an interface between individual and
collective processes in the construction of emotional climate. Journal of Social Issues,
63, 307–322.
Wang, Q. (2002). Did you have fun?: American and Chinese mother-child conversations
about shared emotional experiences. Cognitive Development, 16, 693–715.
3 Age-related changes in
empathy-related responding
Nancy Eisenberg, Jennifer Betkowski and
Tracy L. Spinrad
Arizona State University
Empathy is commonly defined as a response that stems from the apprehension

or comprehension of another’s emotional state or condition and is identical or
very similar to what another is feeling or would be expected to feel. An example
of empathy is if a girl views a sad boy and experiences sadness herself, even
though she is aware that the boy’s emotion is not her own emotion. Unlike
emotional contagion, empathizers must realize that another person is experi-
encing emotion (or would be expected to do so), not just themselves. Empathy
is generally believed to be based on cognitive processes such as identifying
another’s emotion, cognitively taking the role of the other (i.e., perspective
taking), or assessing information in memory that is relevant for understanding
another person’s emotion or situation. Although many investigators also
require the affective component to label a response as empathy, others label the
cognitive processes contributing to empathy as empathic accuracy or cognitive
empathy.
Empathy, when defined as having an affective as well as cognitive component
(i.e., as defined previously), is believed to frequently lead to sympathy or
personal distress. Sympathy is an affective response that can stem from empathy,
perspective taking, or other cognitive processing including retrieval of infor-
mation from memory. It consists of feelings of sorrow or concern for another
(rather than feeling the same emotion as another). Thus, if an observer feels
concern for someone who is sad or distressed, or for someone who is in a situation
likely to evoke sadness or distress, the observer is experiencing sympathy. Like
sympathy, personal distress also frequently stems from exposure to another’s state
or condition but is a self-focused, aversive emotional reaction to the vicarious
experience of another’s emotion (e.g., discomfort, anxiety; see Eisenberg and
Fabes, 1998). For example, if a girl feels distress when observing someone who
is sad because it makes her uncomfortable, she is viewed as experiencing personal
distress. In general, sympathy has been positively related to prosocial behaviors
performed for selfless reasons, whereas personal distress tends to be negatively
or unrelated to such prosocial behavior (Eisenberg and Fabes, 1998; Knafo et al.,
2008).
There has been considerable interest in understanding the emergence of
empathy and its related responding. Based on both theory and empirical
18 Nancy Eisenberg, Jennifer Betkowski and Tracy L. Spinrad
findings, it is believed that empathy-related responding is exhibited early in
life and continues to improve with age, especially in the toddler, preschool, and
elementary years.
Hoffman (2000) proposed a theoretical model including a series of phases,
delineating the development of empathy-related responding and prosocial
behavior. Specifically, rudimentary forms of empathy are evidenced by the
newborn’s reactive or contagious crying in response to the cries of other infants,
although researchers have questioned whether these findings support the notion
of rudimentary empathy or if infants may simply find a novel cry to be particu-
larly aversive (Eisenberg and Lennon, 1983). Although this issue has not been
resolved, it is clear that infants are responsive to others’ emotional signals.
According to Hoffman’s theory, beginning around the end of the first year of
life, infants typically seek comfort for themselves when exposed to another’s
distress. Because infants at this age cannot fully differentiate between their own
distress and that of another, they are likely to respond with personal distress. In
the second year of life, toddlers can begin to express concern (sympathy) for
another, rather than simply seeking comfort for themselves. However, toddlers’
prosocial behaviors may involve giving the other person what they themselves
find comforting (e.g., bringing a favorite teddy bear to a distressed adult) because
they have difficulty differentiating factors that affect their own versus others’
emotions.
Hoffman argued that as children develop cognitively, they are increasingly
capable of understanding another person’s needs and that others’ needs may
differ from their own. In early childhood, however, these empathic responses are
limited to another’s immediate distress. With increasing perspective-taking skills
and cognitive understanding of others’ emotional states, older children begin to
experience empathy towards people who are not physically present (e.g., if they
hear about someone in distress), and by later childhood (around 9 or 10 years
of age), the ability to experience empathy for another’s life condition or general
plight develops (Hoffman, 2000). Thus, the adolescent is viewed as capable of
comprehending and responding to the plight of an entire class of people, such
as the impoverished. In brief, Hoffman argued that with increasing cognitive
maturation, children are better able to respond with concern to others’ distress,
be it observed or presented symbolically (e.g., through the written word), and are
better able to empathize and sympathize with a broad range of people.
Others (e.g., Eisenberg and Fabes, 1998) have argued that the ability to regulate
vicariously induced affect is important for children to experience sympathy
rather than personal distress. This is because high levels of empathically induced
negative emotion are likely to be experienced as aversive, which results in children
focusing on their own aversive emotional state rather than on the emotions and
needs of the other person. Because self-regulation develops rapidly in the first
four years of life and continues to emerge more gradually across childhood and
adolescence, one would expect sympathy to increase with age.
Empirical research provides some support for Hoffman’s ideas. At six months
of age, children are rarely upset by others’ distress. At about a year of age, children
Age-related changes in empathy-related responding 19
may attend to others’ emotional displays but react with emotion relatively infre-
quently. By about 12–18 months of age, toddlers sometimes respond to another’s
distress with empathy and prosocial reactions, although such responses are
not very frequent and are directed mostly at people they know well (e.g., their
mother). Moreover, children of this age often respond with aggression, ignoring,
and distress reactions (Eisenberg et al., 1998; Zahn-Waxler et al., 2001).
The frequency of empathy and sympathy has been found to increase with age
in the first years of life (e.g., Knafo et al., 2008). For example, using longitu-
dinal data and observed measures of empathy and empathy-related responding,
Zahn-Waxler and colleagues (2001) studied toddlers’ reactions to an unfamiliar
female and to their mother feigning injuries at 14, 20, 24, and 36 months of age.
An increase in toddlers’ concern was found, and self-focused distress decreased
with age, particularly from 14 to 24 months. This is consistent with the finding
in a meta-analysis that there was a significant increase in prosocial behavior
within the infant and preschool periods (effect sizes = 0.24 and 0.33 for infant
and preschool, respectively; Eisenberg and Fabes, 1998; see Hay and Cook, 2007,
for the argument that prosocial behavior does not increase systematically in the
first two years of life). In a number of the studies in the meta-analysis, empathy/
sympathy was part of the measure of prosocial behavior. The authors qualified
their meta-analytic findings, however, by noting that they were typically based on
relatively small samples and cross-sectional data.
Consistent with Hoffman’s arguments about the importance of differentiating
one’s own and others’ emotions for young children’s empathy, at about 1.5–2
years of age, children who are able to recognize themselves in a mirror—a
behavior considered indicative of the development of a sense of self that is
separate from that of others—are more likely to exhibit empathy and prosocial
behavior than are their peers without such self-recognition. Moreover, in the
second year of life and the preschool years, children who respond with more
cognitive forms of empathy (e.g., hypothesis testing, inquisitiveness whereby the
child tries to understand others’ distress) appear especially likely to assist others
due to empathy (Knafo et al., 2008; see Eisenberg and Fabes, 1998).
Despite some increases in empathy and prosocial behavior, empathy during
early childhood appears to be relatively stable across time and context. For
example, Knafo et al. (2008) found rank-order (intraindividual) stability in
empathy across the second and third years of life. In addition, the researchers
showed evidence that children’s empathy toward the feigned distress of the
mother and an unfamiliar female experimenter were positively related. Moreover,
the indices (empathy and hypothesis testing, across the two victims) loaded on a
single factor at each age, supporting the existence of an overall empathic dispo-
sition in young children.
Findings on age-related changes in empathy-related responding past the early
years are somewhat less consistent than findings for the first few years of life,
likely due in part to the measures used to assess empathy and sympathy. Most
of the measures of empathy/sympathy used in elementary school involve self-
reports on questionnaires or in response to empathy-inducing stories. In studies
of empathy-related responding before approximately 1985, there was evidence
that self-reports of empathy increase in the school years until age 11; however,
findings were not very consistent after that age (see Lennon and Eisenberg, 1987).
Facial/gestural indices appeared to be either inversely related or unrelated to age
in the early school years, perhaps due to increases with age in the ability to mask
emotion. In many early studies of empathy, children were asked their feelings
in response to a series of very short vignettes about other children in emotion-
evoking contexts. Studies of this sort have been criticized for not being evocative
and for assessing the desire to behave in socially appropriate ways rather than
actual empathic responding (Eisenberg and Lennon, 1983). Therefore, it is wise
to be cautious when drawing conclusions from early studies using picture-story
methods.
The pattern of findings in more recent childhood studies using better measures
is somewhat clearer. Eisenberg and Fabes (1998) reported a meta-analysis of
age differences in empathy/sympathy (rather than prosocial behavior) in studies
published from 1983 to about 1995 and found an overall unweighted effect size
of 0.24 (favoring older children). Effect sizes for age-related increase in empathy
varied with the method of assessing empathy-related responding; they were larger
for observational (usually behavioral) and self-report indices than for nonverbal
(facial/physiological) or other-report measures (for which the effects sizes were
not significant).
Studies of empathy-related responding across adolescence are limited. In
studies before about 1985, findings were inconsistent (Lennon and Eisenberg,
1987). However, most early studies did not assess sympathy, which one might
expect to increase along with adolescent gains in emotional understanding. In
studies since the 1980s, researchers have found evidence for an increase in self-
reported empathy-related responding, especially sympathy, across the junior
and high school years (see Eisenberg, Morris, McDaniel, and Spinrad, 2009).
Unfortunately, there are few studies of age-changes in adolescence using other
types of measures in adolescence. Thus, we cannot be sure that sympathy actually
increases rather than adolescents’ self-perceptions of being sympathetic.
Studies of change in empathy-related responding across the adult lifespan
are more limited than studies with children. The majority of the research has
been cross-sectional and has found either lower empathy in older adults than
in younger persons or equivalence across ages. For instance, of two narrative
studies, one showed age equivalence and the other found higher empathy in
adolescents but no difference between midlife and older adults (see review in
Grühn et al., 2008). Four questionnaire studies showed lower empathy in older
adults, but two showed age equivalence (see review in Richter and Kunzmann,
2010). In a relatively recent study, Grühn and colleagues (2008) found lower self-
reported combined cognitive and affective empathy in older adults when grouped
for cross-sectional analyses by decades from ages 10–87.
Longitudinal studies are needed to disentangle cohort from developmental
effects. The few longitudinal studies suggest that empathy changes slightly
from adolescence to early adulthood and minimally if at all across adulthood.
Specifically, in a study of individuals starting in their 20s and 30s, Helson and
colleagues (2002) found that empathy (defined as interest and resourcefulness
in understanding others) declined over 40 years. Grühn and colleagues found no
average change over a 12-year period in 10–87-year olds even though they found
the aforementioned decline with age in cross-sectional analyses. They suggested
that cross-sectional findings reflect increased psychological thinking among
young persons in recent years.
Various factors may explain age differences in adults’ empathy. There is some
evidence that education and income partially explain cross-sectional age effects
(Schieman and Van Gundy, 2000, but not Grühn et al., 2008), perhaps because
education and income enhance social connection and social-cognitive abilities
but are lower in older cohorts. In addition, social factors may be important. For
instance, Schieman and Van Gundy (2000), using a measure of empathy that
included a strong affective (as well as cognitive) component, found that adjusting
for education and income, health, widowhood and retirement, mastery, introspec-
tiveness, and concern for social approval substantially reduced the age-empathy
association (by between 7% and 27% for subsets; 66% for all predictors). In the
same study, positive personal relationships and religiosity suppressed the decline
in empathy with age. In a longitudinal study, people high in positive affect and
self-acceptance and low in depressive symptoms and autonomy showed increases
in a measure of cognitive and affective empathy over 12 years, although these
effects were small and involved the measure used by Helson et al. (2002), which
probably assesses cognitive empathy more than affective empathy (Grühn et al.,
2008).
Specific contexts also might contribute to age-related patterns. Richter and
Kunzmann (2010) examined empathic accuracy (emotion recognition), emotional
congruence (i.e., feeling the emotion of a target person), and sympathy (i.e.,
reported and observed) in response to tapes of people discussing age-related
situations. Younger adults were more accurate in perceiving emotions for
divorce-related life transitions, and older adults had more emotional congruence
during discussions of social loss. Older adults had greater sympathy for both
types of situations. The authors concluded that older individuals’ low empathic
accuracy is limited to non-relevant situations and that older adults may show
superior emotional abilities in highly relevant situations. They also might exhibit
more sympathy—a construct not examined in most of the relevant studies.
The aforementioned findings in adulthood highlight the importance of context
as well as the need to examine patterns in cognitive and affective empathy (as
well as between empathy and sympathy). Some investigators have proposed
that cognitive skills associated with empathy such as emotion recognition and
perspective taking may decline along with general cognition into old age. On the
other hand, emotional aspects of empathy may improve with age due to increasing
emotion regulation and affective orientation (Grühn et al., 2008; Richter and
Kunzmann, 2010). Most studies show lower cognitive empathy in older adults,
although almost all findings are cross-sectional (see reviews in Bailey et al.,
2008, Richter and Kunzmann, 2010). In another study, 19–25-year-olds were
equivalent to 65–87-year olds in total or affective empathy, but significantly
higher in cognitive empathy (Bailey et al., 2008). In a longitudinal study,
perspective taking and empathic moral reasoning increased whereas personal
distress declined from age 17 to 20, but trends leveled off by the endpoint age of
26 (Eisenberg et al., 2005); sympathy did not change from the teens to mid-20s.
In contrast, Richter and Kunzmann (2010) found that 46–71-year-olds were more
sympathetic compared to 21–44-year-olds. Thus, cognitive capacities of empathy
(e.g., perspective taking) appear to be more susceptible to decline than are the
affective components of empathy-related responding.
In summary, in studies of adults, cohort differences are frequent and likely due
to education, social, and psychological influences. There might be some decline
in cognitive skills involved in empathy, but perhaps only for unfamiliar, less
personal situations. Older adults may have a superior ability to experience others’
emotions in more intimate, personally meaningful situations.
In studies with individuals of all ages, there is a need for replication
using a variety of measures and contexts. In addition, social and cognitive
factors associated with and underlying age-related changes in empathy-related
responding merit additional attention. Finally, the role of heritability in changes
in empathy-related responding merits attention (Knafo et al., 2008; see Eisenberg
and Fabes, 1998).
Acknowledgements
This research was supported by a grant from the National Institute of Mental
Health (MH060838) awarded to Nancy Eisenberg and Tracy L. Spinrad (PIs).
Jennifer A. Betkowski was supported by an NIMH training grant (T32 MH
018387, PI, Laurie Chassin).
References
Bailey, P. E., Henry, J. S., and Von Hippel, W. (2008). Empathy and social functioning in
late adulthood. Age and Mental Health, 12, (4), 499–503.
Eisenberg, N., Cumberland, A., Guthrie, I. K., Murphy, B. C., and Shepard, S. A. (2005).
Age changes in prosocial responding and moral reasoning in adolescence and early
adulthood. Journal of Research on Adolescence, 15, (3), 235–260.
Eisenberg, N., and Fabes, R. (1998). Prosocial development. In W. Damon (Editor-in-
Chief) and N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Social,
emotional, and personality development (5th edn, pp. 701–778). New York: Wiley.
Eisenberg, N., and Lennon, R. (1983). Gender differences in empathy and related capac-
ities. Psychological Bulletin, 94, 100–131.
Eisenberg, N., Morris, A. S., McDaniels, B., and Spinrad, T. L. (2009). Moral cognitions
and prosocial responding in adolescence. In I. Steinberg and R. Lerner (Eds), Handbook
of adolescent psychology (3rd edn; pp. 229–265). Hoboken, NJ: John Wiley & Sons.
Grühn, D., Rebucal, K., Diehl, M., Lumley, M., and Labouvie-Vief, G. (2008). Empathy
across the adult lifespan: Longitudinal and experience-sampling findings. Emotion, 8,
753–765.
Hay, D. F. and Cook, K. V. (2007). The transformation of prosocial behavior from infancy
to childhood. In C. A. Brownell and C. B. Kopp (eds), Socioemotional development in
the toddler years: Transitions and transformations. New York, NY: Guilford Press.
Helson, R., Jones, C., and Kwan, V. S. Y. (2002). Personality change over 40 years of
adulthood: Hierarchical linear modeling analyses of two longitudinal samples. Journal
of Personality and Social Psychology, 83, 752–766.
Hoffman, M. L. (2000). Empathy and moral development: Implications for caring and
justice. New York: Cambridge University Press.
Knafo, A., Zahn-Waxler, C., Van Hulle, C., Robinson, J. L., and Rhee, S. H. (2008). The
developmental origins of a disposition toward empathy: Genetic and environmental
contributions. Emotion, 8, (6), 737–752.
Lennon, R., and Eisenberg, N. (1987). Gender and age differences in empathy and sympathy.
In N. Eisenberg and J. Strayer (eds), Empathy and its development (pp. 195–217). New
York: Cambridge University Press.
Richter, D. and Kunzmann, U. (2010). Age differences in three facets of empathy:
Performance-based evidence. Psychology and Aging, November 8, 2010. Advance
online publication.
Schieman, S. and Van Gundy, K.V. (2000). The personal and social links between age and
self-reported empathy. Social Psychology Quarterly, 63, (2), 152–174.
Zahn-Waxler, C., Schiro, K., Robinson, J. L., Emde, R. N., and Schmitz, S. (2001).
Empathy and prosocial patterns in young MZ and DZ twins: Development and genetic
and environmental influences. In R. N. Emde and J. K. Hewitt (eds), Infancy to early
childhood (pp. 141–162). New York: Oxford University Press.
4 Children’s expressive behavior in
different cultural contexts
Linda A. Camras and Michael M. Shuster
DePaul University
Contemporary developmental theories of emotion emphasize the important role

of culturally-based socialization processes in influencing children’s affective
responses, including their emotional expression. However, in contrast to this
view, several influential scholars (Freedman, 1974; Kagan and Fox, 2006)
have argued that some cultural variations in emotion are rooted in race-based
biological differences. Accordingly, exposing individuals to socialization influ-
ences that differ from those characterizing their culture of origin should not
affect (or only minimally affect) these emotion responses. Current global trends
in adoption provide a unique opportunity to explore the influences of biology
and socialization on emotion by studying its malleability under circumstances of
cultural change. After briefly reviewing the relevant literature, we present a study
that explores these influences on emotional expressivity by comparing adopted
Chinese to nonadopted Mainland Chinese, Chinese American, and European
American children.
Culture, emotion, and emotional expressivity

Cultural differences in emotional expression have long been recognized by those
who study human expressive behavior (e.g., LaBarre, 1947). According to anthro-
pologists and cultural psychologists, such differences reflect values, attitudes,
goals, and consequently norms for emotional expression (e.g., Cole and Tan,
2007; Mesquita and Leu, 2007). Several conceptual frameworks for character-
izing culture have been proposed in recent decades (e.g., Hofstede, 2001; Markus
and Kitiyama, 1994). However, virtually all share the notion that an important
dimension of cultural variability involves the relative focus placed on the person
as an autonomous individual vs a participant in relationships among members of
a collective or group. These important distinctions are captured in Markus and
Kitiyama’s conceptualization of interdependence/independence as a dimension of
cultural variability, and within Hofstede’s broader framework as the dimension of
individualism/collectivism.
Although intra-regional variability exists, Asian societies (e.g., China, Japan,
Korea) are generally considered to be collectivistic or interdependent, while North
American and European cultures are considered individualistic or independent.
Children’s expressive behavior in different cultural contexts 25
Correspondingly, differences between Asians and Americans in their emotion-
related responding have been identified that are consistent with this distinction.
For example, Tsai, Knutson, and Fung (2006) found that Chinese individuals
seek and value lower intensities of positive emotion in comparison to Americans.
According to Tsai’s Affect Valuation Theory, lower intensities of emotion are
more appropriate for individuals who are oriented toward adjusting to existing
group values, goals, and norms, while higher intensities are more appropriate
for individuals who are oriented toward exerting influence over the group.
This difference in orientation is consistent with the more general individualism-
collectivism distinction.
With regard to emotional expression (as distinct from emotional experience),
individualistic cultures are thought to value such expression as a manifestation or
reflection of the importance placed on the individual, while collectivistic cultures
value emotional restraint in the service of promoting good relationships and
social harmony (Bond and Hwang, 1986). Consistent with this view, Matsumoto
et al. (2008) found that members of collectivistic cultures report being less
expressive overall than members of individualistic cultures. From the perspective
of many psychologists, such cultural differences in adult expressivity might result
from socialization influences that begin during infancy and childhood.
Emotion socialization
Several widely-recognized models of emotion socialization (e.g., Cole and Tan,
2007; Denham, Bassett, and Wyatt, 2007; Dunsmore and Halberstadt, 2009;
Eisenberg, Cumberland, and Spinrad, 1998; Holodynski and Friedlmeier, 2006)
share an emphasis on the roles of parent modeling of emotion, contingent responding
to children’s affect, and verbal explanations and discussion of emotion. Parental
modeling of emotions (i.e., parents’ own emotional behavior) is thought to provide
children with information on how and when emotions are expressed. Contingent
responding to children’s expressive behavior provides them with positive and
negative feedback about the effectiveness and/or appropriateness of their own
expressions in various situations. Similar information is provided through verbal
explanations or discussions that may take place during observations of others’
emotions or in the context of story-reading or story-telling. While most develop-
mental research focuses on socialization by parents, these processes may also take
place during children’s interactions with other adults or even with other children.
Variations in parental/adult socialization behaviors relevant to emotion are
thought to be present across cultures and influenced by cultural variations in
parental socialization goals (Cole and Tan, 2007). These socialization goals are
themselves strongly influenced by cultural standards and norms. For example,
American parents are more concerned with developing children’s self-esteem
than are Chinese parents, and consequently express less negative emotion when
recounting their children’s behavioral transgressions (Miller, Fung, and Mintz,
1996). At the same time, Chinese parents generally talk less about the child’s
emotion than about the emotions of others (Wang, 2001) and this also may
26 Linda A. Camras and Michael M. Shuster
communicate a message about the appropriateness of child emotional expressivity.
Additionally, parental contingent responding to children’s expressive behavior
(e.g., their attention to or tolerance of children’s expressions, their implicit or
explicit approval or disapproval) is highly influenced by cultural variations
in parental socialization goals. For example, in their study of two different
ethnic groups in Nepal, Cole, Tamang, and Shrestha (2006) found that Brahmin
parents were responsive to their children’s anger, while Tamang parents were
responsive to expressions of shame, reflecting their differential valuing of these
two emotions.
Biological explanations for cultural differences

Consistent with the literature on cultural differences in adult expressive behavior,
differences in infants’ and children’s expressivity also have been documented.
Such studies have consistently found Chinese or Chinese American babies to be
less facially expressive than European American infants. For example, Camras
et al. (1998) found that 11-month-old Chinese infants both smiled and cried less
than same-age American babies. Examining even younger infants, Freedman
(1974) reported that Chinese American neonates demonstrated less reactivity
and distress than European American neonates during infant testing procedures.
Freedman attributed these findings for newborns to innate differences between
the ethnic groups. Similarly, Kagan and Fox (2006) have suggested that Chinese
and European American infants differ in emotional reactivity due to inherent
differences in amygdala functioning. Extending this line of reasoning, one might
argue that differences in Chinese vs American children’s and adults’ expressivity
are based largely on such innate differences rather than the socialization practices
described above. Nonetheless, expressivity in older Chinese and European
American children might yet be influenced by cultural socialization. In fact, one
recent model of culture-gene co-evolution proposes that social values and practices
may develop in order to support (and sometimes compensate for) genetically-
based characteristics and behaviors of culture members (see Chiao and Blizinsky,
2010).
Emotional expressivity in adopted Chinese children

Studying adopted Chinese children provides an opportunity to partially disen-
tangle possible race-based inherent differences in emotional expressivity from
the influence of cultural socialization. These children experience a change in
both their family and cultural environments that might affect their expressive
behavior. Herein, we briefly describe a study that compared facial expressivity
in three-year-old adopted Chinese and nonadopted European American, Chinese
American, and Mainland Chinese girls (Camras, Chen, Bakeman, Norris, and
Cain, 2006). We proposed that, irrespective of any biological differences that may
exist, cultural socialization would influence the Chinese children’s expressive
behavior. More specifically, we proposed that the adopted Chinese children would
be more expressive than nonadopted Mainland Chinese or Chinese American
children because they are being raised by European American parents. However,
we expected them to be less expressive than nonadopted European American
children because of their early experience with Chinese caregivers and/or innate
differences in expressivity as proposed by Kagan and Freedman. We also hypoth-
esized that Chinese American children would fall between the European American
and Mainland Chinese children in expressivity because of possible acculturation
by their parents (see De Leersnyder, Mesquita and Kim, this volume) and also
children’s direct exposure to socialization influences from outside the family
(e.g., observation of expressivity in European American children and adults).
Lastly, we hypothesized that the expressivity of children would be directly related
to their mothers’ self-reported socialization attitudes and behaviors.
The adopted Chinese girls had been abandoned shortly after birth and had resided
in a Chinese orphanage for approximately one year (on average). They were subse-
quently adopted into American families, and had lived with their families for at
least 18 months. Thus these children had experienced expressive behavior produced
by caregivers from two different cultural environments. The Chinese American
girls were living in immigrant families; their parents themselves had been raised
in China but were now living in the U.S. Thus to some extent, these children also
were potentially exposed to culturally divergent forms of expressive behavior.
All children were videotaped while viewing color slides of positive and
negative emotionally evocative stimuli (e.g., sleeping kitten, girl eating a worm).
Facial expressions were coded using a modified version of Ekman, Friesen, and
Hager’s (2002) Facial Action Coding System. Scores for several specific expres-
sions (e.g., smiles, disgust-related expressions) and a composite score for overall
expressivity were generated. In order to explore maternal behaviors and attitudes
that might influence their children’s emotional expressivity, mothers completed
questionnaires assessing their parenting styles, tendency to themselves display
emotion, and identification with Asian culture (for the non-Mainland Chinese
groups).
Consistent with past studies of emotional expressivity (Freedman, 1974;
Camras et al., 1998; Kagan, Kearsley, and Zelazo, 1978), European American
children scored significantly higher than Mainland Chinese children for smiles,
disgust-related expressions, and overall expressivity. As predicted, the adopted
Chinese children fell between these two groups. Similarly, Chinese American
children tended to be more expressive than Mainland Chinese children, but less
expressive than European American or adopted Chinese children (although the
differences were not always significant).
Differences in socialization-related maternal behaviors and attitudes also were
found. Mainland Chinese and Chinese American mothers reported greater strictness
and aggravation than European American mothers, while European American and
Chinese American mothers reported greater positive expressivity. Not surpris-
ingly, Chinese American mothers identified more with Chinese culture than the
European American mothers of the adopted Chinese children, who identified
themselves more with Chinese culture than the other European American mothers.
As we had predicted, the maternal variables were related to children’s emotional
expressivity. Mothers’ strictness and identification with Asian culture were related
to lesser overall expressivity in children while maternal positive expressivity was
related to greater child expressivity. Of particular interest, many items on the
maternal strictness subscale referred explicitly to encouraging emotional restraint
(e.g., “I teach my child to keep control of her feelings at all times.”). Thus mothers
who reported discouraging their child’s emotional expression had children who
were indeed less expressive in our study. Hierarchical regression analyses yielded
no group differences beyond those that could be predicted by the maternal social-
ization-related variables. Thus we interpret our findings as supporting the position
that culturally-based socialization significantly influences children’s expressive
behavior. Correspondingly, we propose that changes in the culturally-based
socialization practices to which children are exposed (as occurred for our adopted
Chinese children) will produce changes in expressive behavior.
Extending the model of emotion socialization

Our view is largely consistent with widely-recognized developmental models of
emotion socialization as described above. However, going beyond these models,
we also propose that emotion socialization (particularly parental modeling) may
engage an additional process that is currently receiving considerable attention
in the adult emotion literature, i.e., emotion embodiment. When individuals
embody another’s emotional expression (i.e., by mimicking their expression either
overtly or internally via internalized sensory-motor activation), they are thought
to experience sensations corresponding with the emotion associated with the
mimicked expression (i.e., emotion contagion, Hatfield, Cacioppo, and Rapson,
1994). In terms of development, such contagion via embodiment may occur if
infants or children mimic a caregiver’s expressive behavior during the course of
social referencing, i.e., when the child looks to adults for information regarding the
appropriate affective response in an emotionally-ambiguous situation. However, it
may also occur during affective mirroring that takes place during other types of
social interactions (see Holodynski and Friedlmeyer, 2006). Over time, associations
between the emotion situation, the embodied expressions, and their subsequently-
produced feeling states may contribute to the development of knowledge regarding
environmental contingencies associated with both the expression and its corre-
sponding emotion. Consistent with this proposal, Shuster and Camras (unpublished)
found that children were more likely to choose the correct facial expression for an
emotion-eliciting situation when they mimicked that expression prior to making
their choice.
Given the potential role of mimicry in the development of knowledge regarding
situation-appropriate emotion responding, cross-cultural variations in parental
expressivity may influence cultural and individual differences in the degree of
expressiveness manifested by a child or adult. It is therefore important for future
research to further examine cross-cultural differences in parental expressivity and
affective mirroring across different emotional situations.
Concluding remarks
In summary, our observations of expressive variability across groups of Chinese
children experiencing different (and sometimes changing) cultural and family
environments suggest that emotional expressivity is highly malleable. However,
our study had limitations that point to several other important directions for future
research. First, we were unable to completely disentangle the potential influences of
biology and cultural socialization because our adopted children had spent approxi-
mately ten months in Chinese orphanages and/or foster care before joining their
American families. Unfortunately, this limitation would be difficult to overcome in
studies of Chinese children because government policies in China preclude interna-
tional adoption shortly after birth. However, future investigations involving children
adopted from other cultures (e.g., South American) might feasibly overcome this
difficulty. A second limitation was our lack of longitudinal data. Future research that
directly documents developmental changes in adopted children’s expressivity would
provide even stronger evidence for the influence of cultural socialization. A third
limitation was our reliance on questionnaire measures to assess maternal sociali-
zation. While such measures are commonly used and have proved valuable, we
hope that future investigations of expressive socialization across cultures will more
often include direct observations. Despite these limitations, the current research was
successful in showing that facial expressivity is a flexible system that is responsive
to aspects of the family environment that differ across cultures. Our hope is that
this study will pave the way for subsequent investigations that elaborate on the
processes involved in the cultural socialization of emotional expressivity.
References
Bond, M., and Hwang, K. (1986). The social psychology of the Chinese people. In M.
Bond (ed.), The psychology of the Chinese people (pp. 213–266). Hong Kong: Oxford
University Press.
Camras, L. A., Chen, Y., Bakeman, R., Norris, K., and Cain, T. R. (2006). Culture,
ethnicity, and children’s facial expressions: A study of European American, Mainland
Chinese, Chinese American, and adopted Chinese girls. Emotion, 6, (1), 103–114.
Camras, L. A., Oster, H., Campos, J., Campos, R., Ujiie, T., Miyake, K., et al. (1998).
Production of emotional facial expressions in European American, Japanese and
Chinese infants. Developmental Psychology, 34, 616–628.
Chiao, J.Y. and Blizinsky, K. (2010). Culture-gene coevolution of individualism-
collectivism and the serotonin transporter gene (5-HTTLPR). Proceedings of the Royal
Society B: Biological Sciences, 277, 529–537.
Cole, P. M., Tamang, B. L., and Shrestha, S. (2006). Cultural variations in the socialization
of young children’s anger and shame. Child Development, 77, 1237−1251.
Cole, P. M., and Tan, P. Z. (2007). Emotion socialization from a cultural perspective. In
J. E. Grusec and P. D. Hastings (eds), Handbook of socialization: Theory and research.
(pp. 516–542). New York, NY: Guilford Press.
De Leersnyder, J., Mesquita, B., and Kim, H. (2013) Emotional acculturation. In D.
Hermans, B. Rimé, and B. Mesquita (Eds.), Changing Emotions. Hove: Psychological
Press.
Denham, S. A., Bassett, H. H., and Wyatt, T. (2007). The socialization of emotional
competence. In J. E. Grusec and P. D. Hastings (eds), Handbook of socialization: Theory
and research. (pp. 614–637). New York, NY: Guilford Press.
Dunsmore, J. C., and Halberstadt, A. G. (2009). The dynamic cultural context of emotion
socialization. In J. A. Mancini and K. A. Roberto (eds), Pathways of human development:
Explorations of change. (pp. 171–190). Lanham, MD: Lexington Books/Rowman and
Littlefield.
Eisenberg, N., Cumberland, A., and Spinrad, T. (1998). Parental socialization of emotion.
Psychological Inquiry, 9, 241–273.
Ekman, P., Friesen, W. V., and Hager, J. (2002). The Facial Action Coding System, 2nd ed.
Salt Lake City, UT: Research Nexus.
Freedman, D. G. (1974). Human Infancy: An evolutionary perspective. New York: Halsted
Press.
Hatfield, E., Cacioppo, J. T., and Rapson, R. L. (1994). Emotional contagion. New York;
Paris: Cambridge University Press Editions de la Maison des Sciences de l’Homme.
Hofstede, G. (2001). Culture’s consequences: Comparing values, behaviors, institutions
and organizations across nations. Thousand Oaks, CA: Sage Publishers.
Holodynski, M., and Friedlmeier, W. (2006). Development of emotions and emotion
regulation. New York, NY: Springer Science + Business Media.
Kagan, J., and Fox, N. (2006). Biology, culture, and temperamental biases. In W. Damon
(Series ed.) and N. Eisenberg (Vol. ed.), Handbook of child psychology: Vol. 3. Social,
emotional and personality development, 6th ed. New York: Wiley.
Kagan, J., Kearsley, R., and Zelazo, P. (1978). Infancy: Its place in human development.
Cambridge, MA: Harvard University Press.
LaBarre, W. (1947). The cultural basis of emotions and gestures. Journal of Personality,
16, 49–68.
Markus, H. R., and Kitayama, S. (1994). The cultural construction of self and emotion:
Implications for social behavior. In S. Kitayama and H. R. Markus (eds), Emotion and
culture: Empirical studies of mutual influence (pp. 89–130). Washington, DC: American
Psychological Association
Matsumoto, D., Yoo, S., Fontaine, J., Anguss-Wong, A., Arriola, M., Ataca, B., and Grossi,
E. (2008). Mapping expressive differences around the world: The relationship between
emotional display rules and individualism versus collectivism. Journal of Cross-
Cultural Psychology, 39, 55–74.
Mesquita, B. and Leu, J. (2007). The cultural psychology of emotion. In S. Kitayama and
D. Cohen (eds) Handbook of cultural psychology. (pp. 734–759). New York: Guilford
Press.
Miller, P., Fung, H., and Mintz, J. (1996). Self-construction through narrative practices: A
Chinese and American comparison of early socialization. Ethos, 24, 237–280.
Shuster, M., and Camras, L. A., (in preparation). Emotion Embodiment and Facial
Expression Recognition in Adopted Chinese and Non-adopted European American Girls.
Tsai, L., Knutson, B., and Fung, H. H. (2006). Cultural variation in affect valuation.
Journal of Personality and Social Psychology, 90, (2), 288–307.
Wang, Q. (2001). Did you have fun? American and Chinese mother–child conversations
about shared emotional experiences. Cognitive Development, 16, 693–715.
5 Shifts in emotional experience and
regulation across adulthood
Tammy English and Laura L. Carstensen
Stanford University
Aging is associated with decline in many psychological processes, such as

executive functioning and working memory, yet recent research suggests that
emotional functioning is relatively well preserved. In fact, in some ways,
emotional experience and regulation appears to improve. Older people are slower
to anger and more likely to forgive, manage social relationships with less conflict,
and report greater satisfaction with family and close friends than do younger
people (Luong, Charles, and Fingerman, 2011). Importantly, older adults are not
only more skilled at regulating emotional states, but they also appear to be more
motivated to do so (Carstensen, 2006). In this chapter, we provide an overview
of age differences in emotional experience and regulation during adulthood, and
discuss possible mechanisms that may underlie such age-related shifts.
Aging and emotional experience

There is growing evidence that emotional experience in older adults is relatively
positive (Scheibe and Carstensen, 2010), at least until very close to death
(Gerstorf et al., 2010). Findings from scores of studies based on questionnaires,
surveys, and clinical interviews converge. Older people reliably report greater
happiness, less depression, and more satisfaction with their lives than younger
and middle aged people. Given well-documented age-related losses in many
aspects of physical and cognitive functioning, such findings were initially met
with skepticism. A number of explanations were offered to account for positive
reports aside from actual improvements in functioning. They ranged from the
serendipitous benefits of reduced biological sensitivity to cognitive decline to
outright denial (for review, see Scheibe and Carstensen, 2010). Until recently,
these alternatives could not be ruled out; and because the vast majority of findings
about emotion and aging have been based on cross-sectional age comparisons,
another explanation entirely independent of age, namely cohort effects, remained
a viable alternative. That is, experiences of today’s older generations, like world
wars and the Great Depression, may have produced uniquely resilient cohorts that
will not be observed in the future.
Because cross-sectional comparisons cannot rule out cohort effects, studies
that focused on within-individual change were critically needed. One early
32 Tammy English and Laura L. Carstensen
study based on questionnaire responses to the Bradburn Affect Balance Scale
administered between 1971 and 1994 observed similar reductions in negativity
in four different birth cohorts over time (Charles, Reynolds and Gatz, 2001).
However, such global responses to questionnaires lack nuanced descriptions
of emotional life. Findings from a very recent longitudinal study—which was
conducted for more than a decade and was based on experience sampling of
daily emotions—also revealed that the patterns observed in cross-sectional
studies play out within individuals (Carstensen, et al., 2011). The reliance on
experience sampling of momentary emotion obviated the possibility that memory
lapses or biases like cognitive dissonance account for age differences. The study
was based on a sample spanning the entire adult age range. Participants carried
electronic pagers and were sampled at random times seven times a day for one
week. Using a burst design, participants were similarly revisited at five-year
intervals. Results were then modeled using growth curve analyses. Reports of
negative emotions decreased over time while the frequency of positive emotions
remained stable, resulting in a more positive balance of emotions over time.
Just as in cross-sectional studies, the intensity of emotional experience did not
differ by age; when experienced, emotions were felt just as strongly in older and
younger people. Moreover, the decline in the frequency of negative emotions
began early in adulthood, long before potential reductions in biological sensitivity
or cognitive deficits. All told, these findings suggest that age-related emotional
benefits may be reflective of developmental gains (rather than loses in biological
or cognitive functioning).
Importantly, Carstensen et al. (2011) found that older adults’ emotional
experience is not best characterized by a happy-go-lucky quality, but rather by
a reduction in negative emotions and a tendency to experience more mixed and
less polarized emotional states. Aging was also associated with greater emotional
stability (i.e., less emotional liability). That is, negative emotions in older adults
were less likely to linger or persist over time. Thus, there seems to be a shift
toward maintaining an emotional equilibrium.
Overall, the pattern of age-related changes in emotion (decreased frequency,
but not intensity, of negative emotion and increased emotional stability) suggests
that improvements in experience may be associated with age-related improve-
ments in emotion regulation. In the next section, we outline evidence of changes
in emotion regulation across adulthood.
Aging and emotion regulation

The ability to regulate emotions improves considerably in early life. Increasingly,
evidence suggests that gains may also continue across adulthood. Older adults
report that they can better control strong emotions, especially the inner experience
of emotion (Gross, et al., 1997), and there is a shift towards the use of more
effective emotion regulation strategies (Blanchard-Fields, Stein, and Watson,
2004). When instructed to regulate emotion in the laboratory, older adults are
equally or more effective than younger adults (Kunzmann, Kupperbusch, and
Shifts in emotional experience and regulation across adulthood 33
Levenson, 2005; Phillips, Henry, Hosie, and Milne, 2008). In addition, there
is some evidence that down-regulation of negative emotion requires fewer
cognitive resources for older adults than younger adults, suggesting an age-related
improvement in emotion regulation efficiency. For instance, one study found that
emotional suppression in younger adults led to reduced memory for pictures
presented during the regulation period compared to when they were told to just
respond naturally, whereas for older adults suppression did not interfere with
memory (Emery and Hess, 2011). Similarly, another study showed that working
memory performance was disrupted by down-regulation of negative emotion for
younger adults, but not older adults (Scheibe and Blanchard-Fields, 2009).
Older adults also appear to rely heavily on antecedent emotion regulation by
constructing social environments that are predictable and emotionally satisfying
(Carstensen, Gross, and Fung, 1997). Antecedent regulation is arguably the most
effective form of emotion regulation in that unwanted emotions are avoided
altogether. Social networks are smaller overall in old age, but they contain similar
numbers of very close social partners as younger adults’ networks. It appears
that social networks are selectively pruned over time. Social partners who do not
provide emotional satisfaction and meaning are eliminated from networks. There
is also evidence from experience-sampling in daily life that older adults are better
at avoiding arguments and other daily stressors (Charles, et al., 2010).
In sum, there is growing cross-sectional evidence that emotion regulation
improves across adulthood. However, longitudinal studies are still needed.
Possible mechanisms
How can we understand these age-related changes in emotional experience and
emotion regulation? What are the mechanisms that drive changes? One possible
explanation concerns biological reactivity. Older adults may be better able to
keep their emotions under control because they are less physiologically reactive
(i.e., emotions become less intense and therefore easier to regulate; Cacioppo,
Berntson, Bechara, Tranel, and Hawkley, 2011). A related argument is that older
adults use less effortful emotion regulation strategies in an attempt to conserve
increasingly limited cognitive and physical resources (Heckhausen and Schulz,
1995; Labouvie-Vief, 2003). However, it is worth noting that evidence that
emotions are experienced just as intensely in later adulthood as they are earlier in
life speaks against a biological interpretation, and the finding that emotional well-
being improves across adulthood, even when taking into account measures of
cognitive ability, suggests that cognitive decline cannot fully explain age-related
changes in emotion. Also, as mentioned previously, age-related improvements in
emotion processes often occur earlier in adulthood than do age-related declines
in biological and cognitive processes.
Another possible explanation of this age-related shift toward improved
emotional well-being and regulation focuses on life experience (Blanchard-
Fields, 2007). Experience and knowledge about emotions play an important role
in emotion regulation. Older adults may be better at realizing emotional goals
because they have had more practice regulating their emotions, and have more
knowledge about who and what makes them feel positive or negative. That is,
after years of experience, older adults may have learned which strategies are most
effective, how to better use each strategy, and when it is best to use each strategy
(i.e., how to flexible apply strategies in an appropriate manner).
One life-span theory of motivation, socioemotional selectivity theory,
maintains that emotion-related goals are increasingly prioritized across adulthood
(Carstensen, 2006). Theoretically, motivational shifts occur because time horizons
grow shorter with age. Faced with constraints on time, cognitive and social
resources are divested from goals related to future possibilities and instead
invested in emotional balance, meaning, and satisfaction. A number of studies
show that older adults favor emotionally meaningful experience over expanding
horizons and exploring novel experiences (for a review, see Carstensen, 2006).
Motivation directs attention and cognitive resources to goal-relevant infor-
mation. Thus, age-related motivational shifts also influence cognitive processing.
Indeed, older adults show goal-consistent information processing that enables
successful regulation of emotion and the maintenance of well-being. Specifically,
there is evidence of age-related changes in attention and memory that support a
more positive affective profile. Compared to younger adults, older adults attend to
and recall positive material relatively more than negative material, a phenomenon
referred to as the positivity effect (Mather and Carstensen, 2005). Even at a neural
level, older adults seem to differentially process emotional material based on
valence: older adults have greater amygdala activation when viewing positive as
opposed to negative images, whereas younger adults display similar activations
regardless of valence (Mather et al., 2004). There is some initial evidence linking
this age-related positivity effect to emotion regulation. Specifically, Isaacowitz et
al. (2008) found that older adults showed positivity in attention (looking towards
positive faces and away from negative faces) when they were in a negative mood,
whereas younger adults had mood-congruent gaze, looking more at negative
faces when in a negative mood.
Consistent with a motivational account, the positivity effect seems to be
strongest in those who have high levels of executive function. Mather and Knight
(2005) found that greater cognitive control in older adults was associated with
disproportional memory for positive over negative information. When they placed
experimental constraints on cognitive resources using a dual attention task, the
positivity effect was reversed. In addition, consistent with socioemotional selec-
tivity theory, the age-related positivity effect disappears in experiments that
explicitly redirect motivation by stressing non-emotional goals (e.g., Löckenhoff,
and Carstensen, 2007).
Conclusions
Emotional lives seem to improve across adulthood—negative emotions are experi-
enced less frequently, while positive emotions are maintained at levels observed
in youth. Older people appear to place greater value on emotionally meaningful
Shifts in emotional experience and regulation across adulthood 35
experience and relatedly appear to regulate emotional states more effectively than
younger people. Though biological and cognitive changes that come with age
may contribute in some ways to these shifts in emotion, the empirical evidence
is most consistent with motivational explanations. Socioemotional selectivity
theory maintains that time horizons change motivation in ways that benefit
emotional experience.
References
Blanchard-Fields, F. (2007). Everyday problem solving and emotion—An adult develop-
mental perspective. Current Directions in Psychological Science, 16, 26–31.
Blanchard-Fields, F., Stein, R., and Watson, T. L. (2004). Age differences in emotion-
regulation strategies in handling everyday problems. Journals of Gerontology Series B:
Psychological Sciences and Social Sciences, 59, 261–269.
Cacioppo, J. T., Berntson, G. G., Bechara, A., Tranel, D., and Hawkley, L. C. (2011).
Could an aging brain contribute to subjective well being? The value added by a social
neuroscience perspective. In A. Tadorov, S. T. Fiske and D. Prentice (eds), Social neuro-
science: Toward understanding the underpinnings of the social mind (pp. 249–262).
New York: Oxford University Press.
Carstensen, L. L. (2006). The influence of a sense of time on human development. Science,
312, 1913–1915.
Carstensen, L. L., Gross, J. and Fung, H. (1997). The social context of emotional
experience. In M. P. Lawton, and K. W. Schaie (eds). Annual Review of Geriatrics and
Gerontology (pp. 325–352). New York: Springer.
Carstensen, L. L., Turan, B., Scheibe, S., Ram, N., Ersner-Hershfield, H., Samanez-Larkin,
G. R., Brooks, K. P., and Nesselroade, J. R. (2010). Emotional experience improves with
Age: evidence based on over 10 years of experience sampling. Psychology and Aging,
26, 21–33.
Charles, S. T., Luong, G., Almeida, D. M., Ryff, C., Sturm, M., and Love, G. (2010). Fewer
ups and downs: Daily stressors mediate age differences in negative affect. Journals of
Gerontology: Series B. Psychological Sciences and Social Sciences, 65, 279–286.
Charles, S. T., Reynolds, C. A., and Gatz, M. (2001). Age-related differences and change in
positive and negative affect over 23 years. Journal of Personality & Social Psychology,
80, 136–151.
Craik, F. I. M., and Salthouse, T. A. (2008). The handbook of aging and cognition (3rd
edn). New York, NY: Psychology Press.
Emery, L., and Hess, T. M. (2011). Cognitive consequences of expressive regulation in
older adults. Psychology and Aging, 26, 388–396.
Gerstorf, D., Ram, N., Mayraz, G., Hidajat, M., Lindenberger, U., Wagner, G. G., and
Schupp, J. (2010). Late-life decline in well-being across adulthood in Germany, the UK,
and the US: Something is seriously wrong at the end of life. Psychology and Aging, 25,
477–485.
Gross, J. J., Carstensen, L. L., Pasupathi, M., Tsai, J., Gotestam-Skorpen, C., and Hsu, A.
Y. C. (1997). Emotion and aging: Experience, expression, and control. Psychology and
Aging, 12, 590–599.
Heckhausen, J., and Schulz, R. (1995). A life-span theory of control. Psychological
Review, 102, 284–304.
Isaaccowitz, D. M., Toner, K., Goren, D., and Wilson, H. R. (2008). Looking while unhappy:
Mood-congruent gaze in young adults, positive gaze in older adults. Psychological
Science, 19, 848–853.
Kunzmann, U., Kupperbusch, C. S., and Levenson, R. W. (2005). Behavioral inhibition and
amplification during emotional arousal: A comparison of two age groups. Psychology
and Aging, 20, 144–158.
Labouvie-Vief, G. (2003). Dynamic integration: Affect, cognition, and the self in
adulthood. Current Directions in Psychological Science, 12, 201–206.
Löckenhoff, C., and Carstensen, L. L. (2007). Aging, emotion, and health-related decision
strategies: Motivational manipulations can reduce age differences. Psychology and
Aging, 22, 134–146.
Luong, G., Charles, S. T., and Fingerman, K. L. (2011). Better with age: Social relation-
ships across adulthood. Journal of Social and Personal Relationships, 28, 9–23.
Mather, M., Canli, T., English, T., Whitfield, S., Wais, P., Ochsner, K.,Gabrieli, J. D., and
Carstensen, L. L. (2004). Amygdala responses to emotionally valenced stimuli in older
and younger adults. Psychological Science, 15, 259–263.
Mather, M., and Carstensen, L. L. (2005). Aging and motivated cognition: The positivity
effect in attention and memory. Trends in Cognitive Sciences, 9, 496–502.
Mather, M., and Knight, M. (2005). Goal-directed memory: The role of cognitive control
in older adults’ emotional memory. Psychology and Aging, 20, 554–570.
Phillips, L. H., Henry, J. D., Hosie, J. A., and Milne, A. B. (2008). Effective regulation
of the experience and expression of negative affect in old age. Journal of Gerontology:
Psychological Sciences, 63B, 138–145.
Scheibe, S., and Blanchard-Fields, F. (2009). Effects of emotion regulation on cognitive
performance: What is costly for young adults is not so costly for older adults.
Psychology and Aging, 24, 217–223.
Scheibe, S. and Carstensen, L.L. (2010). Emotional Aging: Recent Findings and Future
Trends. Journals of Gerontology: Psychological Sciences, 65, 135–144
6 Changing the neural mechanism of
emotion regulation in children with
behavior problems
Steven Woltering
University of Toronto
Marc D. Lewis
Radboud University Nijmegen
Externalizing behavior problems are characterized by high levels of aggression

and delinquency. These problems entail serious costs for the person, the immediate
social environment, as well as social institutions and society as a whole, and they
are thought to be persistent across the lifespan if left untreated (Caspi, Moffit,
Newman, and Silva, 1996). Attempts to change the trajectory of externalizing
problems through intervention programs that start in childhood seem promising
but, unfortunately, they show a lot of variability in their outcomes. Studies
generally reveal that about 40 percent of aggressive children do not show any
significant improvement when treatment has ended.
In order to improve treatment efficacy, many researchers call for a better
understanding of what makes some children, and not others, respond to treatment.
Cognitive mechanisms that children use to regulate emotions may play an
important role in determining treatment outcomes. Emotion regulation allows
individuals to control their emotional impulses, appraisals, and expressions.
An effective strategy of emotion regulation, therefore, helps individuals display
emotional responses with appropriate intensity and form, at the correct time and
place, and modulate or inhibit inappropriate responses. Externalizing problem
behaviors may be due to a lack of effective emotion regulation with respect to
anger, and that is why they are characterized by excessive and disproportionate
aggressive responses. Not surprisingly, the development of effective emotion
regulation is thought to be crucial for adaptive socialization. Ineffective emotion
regulation, on the other hand, can lead to the development of externalizing
pathologies when aggressive responses become habitual and uncontrolled. These
behavioral patterns are often highly resilient and may interfere with successful
treatment outcomes once these pathologies have emerged.
Recently, with the discovery of neural mechanisms and neural markers of
psychological processes, clinical-developmental psychology has increasingly
been informed by neuroscientific research. The current chapter will discuss what
we have learned in our lab about changes in the neural mechanisms of emotion
regulation with treatment aimed at reducing children’s aggressive behavior. We
38 Steven Woltering and Marc D. Lewis
will first ask what networks in the brain are associated with emotion regulation.
Second, we will ask if successful treatment aimed at improving emotion
regulation yields changes in those hypothesized neural systems. Finally, we will
speculate on how these results may improve clinical-development theory and
stimulate future research, by postulating a hypothesis that underlying internal-
izing problems play a key role in the development of externalizing behavior
problems.
Emotion regulation and its neural correlates

Emotion regulation is an umbrella term for a suite of cognitive strategies, such
as inhibitory control and error-monitoring, which are used to modulate emotions.
Generally, two types of emotion regulation can be distinguished: reactive
and deliberate regulation. Reactive regulation is fast and stimulus-driven, and
it includes implicit evaluations of objects or events that can be aversive or
rewarding. Possibly as a result of such appraisals, reactive regulation includes
the rapid execution of automatic behavioral tendencies, such as avoidance or,
in the case of externalizing behavior, approach. An effective reactive regulation,
for example, could inhibit physical or verbal aggression toward others when it
gets crowded upon entering the school bus. By contrast, a child who impulsively
insults others shows ineffective, socially non-adaptive, reactive control. In the
latter example, this behavior is likely to be at odds with long-term goals, such
as establishing reliable friendships or avoiding punishment for harming other
people. Reactive regulation can rely on attentional biases of threat perception
which appraise the situation in a negative light and generate a rapid fight-or-flight
response. With reactive regulation, the line between what is regulation and what is
emotion blurs somewhat, because processes of emotional arousal and emotional
regulation can be seen as overlapping or adjacent (Kappas, 2011). Deliberate
emotion regulation, on the other hand, can be seen as less automatic than reactive
regulation. It is characterized as slower, reflective, and more sensitive to strategy.
Deliberate strategies, such as reappraisal and planning, may reduce the state of
frustration and provide a sense of control. For example, a child who feels anxious
at school because of a disappointing grade may reduce this feeling by realizing
that other smart children also had large difficulties. This way, the child will not
feel as a failure, and may maintain his commitment to academic achievement.
Both reactive and deliberate types of regulation can occur within the same
person and at the same moment, and the effectiveness of these coping strategies
can differ according to the situation and emotional state. It is possible that the
coordination between reactive and deliberate regulation is crucial for effective
emotion regulation. For example, an initially weak reactive regulatory response
can be countered by strong deliberate regulation. However, little is known about
the coordination of these processes, and future research could better ground these
speculations.
Each type of regulation has been associated with a particular neural network
(Phillips, Ladouceur, and Drevets, 2008; Ray and Zald, 2011). The anterior
Changing the neural mechanism of emotion regulation in children 39
cingulate cortex (ACC) is a structure at the midline of the brain, which is associated
with reactive as well as deliberate emotion regulation. The ACC has typically
been divided into a more “affective” ventral and a more “cognitive” dorsal
component. The ventral ACC has strong connections with other ventral prefrontal
regions, such as the orbitofrontal cortex, and subcortical regions involved in
rapid affective processing such as the amygdala, hypothalamus, and brain stem.
Because these networks mediate rapid, impulsive, and visceral regulation, they
are thought to be associated with reactive emotion regulation. The dorsal ACC has
strong connections to lateral prefrontal regions implicated in working memory,
decision making, error monitoring, and response control, such as the dorsolateral
prefrontal cortex (PFC). These networks support “higher order” reasoning and are
therefore thought to mediate deliberate emotion regulation. Because of its dorsal
and ventral divisions, the ACC has been dubbed a “hub” of emotion regulation,
involved in the coordination of deliberate and reactive types of regulation.
Changes in emotion regulation with treatment of externalizing

behavioral problems
Intervention studies investigating brain regions associated with emotion regulation
have been rare, especially with child populations, because such studies are
time-consuming and challenging to conduct. However, intervention studies
are particularly valuable for understanding mechanisms of change. Instead of
studying components of emotion regulation through experimental manipulation,
treatment supposedly changes social behaviors that rely on emotion regulation
capacities. Thus, changes in observed social behaviors can be inferred to tap
changes in emotion regulation, and brain activity in regions hypothesized to
mediate emotion regulation can then be examined in relation to those changes.
We will next review a recent intervention study from our lab that investigated the
neural correlates of emotion regulation.
The study (Woltering, Granic, Lamm and Lewis, 2011) tested seventy-one
8- to 12-year-old children and their families before and after they had completed
a 14-week treatment program. The treatment program was aimed to increase
children’s self-regulatory capacity. It consisted of Cognitive Behavioral Therapy
(CBT) and Parent Management Training (PMT) and was conducted by community
agencies. CBT targets effective regulation of emotion and impulses through strat-
egies such as cognitive restructuring, problem solving, role-playing, social and token
reinforcements, and generalization activities. PMT promotes positive parenting
practices such as skill encouragement, problem solving, and monitoring, as well as
the replacement of coercive or lax discipline strategies with mild sanctions targeting
misbehavior. Treatment success was assessed by various parent and clinician
reports, such as the Child and Adolescent Functional Assessment Scale and the
Child Behavior Checklist (CBCL). Children were included in the study when their
scores revealed clinical levels of externalizing behavior problems on the CBCL. We
note that, although the criteria were aimed at externalizing behavior problems, it was
remarkable that the sample appeared largely comorbid for internalizing problems.
Since it is difficult to measure broad psychological constructs such as emotion
regulation directly, the cognitive process of inhibitory control, which is thought to
be a component of emotion regulation, was measured by means of a go/nogo task.
Children were fitted with an electroencephalography (EEG) net and instructed to
press a button as fast as possible when a letter appeared on the screen, but to inhibit
their response when a letter was repeated. To ensure engagement with the task,
children were told that they needed to obtain a lot of points to receive a desirable
prize. Points were given for correct responses and subtracted when mistakes were
made. A dynamically adjusted algorithm ensured that the difficulty-level of the
task remained challenging for each child. Neural correlates of inhibitory control
were then investigated by examining event related potentials (ERPs—averaged
fluctuations in electrophysiological activity) and source models estimating where
in the brain this activity takes place.
ERP components called the N2, occurring 200–400 ms after the nogo stimulus,
and the frontal P3, occurring between 300 and 900 ms after the nogo stimulus,
have been related to inhibition and attentional control specifically (Falkenstein,
Hoormann, and Hohnsbein, 1999) and self-regulation in general (Lewis et
al., 2008; Cappadocia, Desrocher, Peppler, and Schroeder, 2009). Converging
evidence points to a source or generator of N2 and frontal P3 activity in medial
prefrontal regions such as the dorsal and ventral ACC (Bekker, Kenemans, and
Verbaten, 2005). Activity in these regions, and other regions, such as the lateral
PFC (indicative of reappraisal, a specific regulatory strategy viewed as deliberate)
as well as para-amygdalar regions in the temporal lobe (suggestive of reactive
regulation) were investigated. We note that although it is unlikely that amygdala
activity can be detected directly through EEG technology, para-amygdalar activity
is likely to reflect it. We predicted that these components, and their underlying
sources, would change for those children who improved with treatment compared
to those who did not. Larger N2 magnitudes, specifically, had been associated
with comorbid anxious/aggressive behavior problems in previous work, and a
decrease in magnitudes of this component was hypothesized to reflect a reduction
in the effort required to maintain inhibitory control in stressful situations.
The findings showed differences in the N2 and the frontal P3 between the
nonclinical and clinical groups. The N2, which showed larger magnitudes for the
clinical group, was found to be highly sensitive to treatment-based improvement:
improvers showed a sharp reduction in activation whereas activation for
non-improvers stayed the same from pre- to post-treatment. Improvers reached
levels of activity similar to those of the normal controls, whereas this was not
the case for the non-improvers. Additional analysis revealed that N2 magnitudes
correlated with the degree of improvement, lending further credence to the
assumption that these neural indicators are related to the capacity for emotion
regulation.
These ERP results were consistent with those of the source analysis. Similar
to Lewis et al. (2008), and as expected from N2 localization studies, reduced
activation was found in ventromedial PFC (indicative of the ventral ACC and
OFC) regions for improvers only. In addition, a reduction in activation was
also found in limbic regions such as the bilateral anterior medial temporal
lobe. These results suggest changes in fronto-limbic systems associated with
reactive emotion regulation. However, in contrast to Lewis et al. (2008), the
dorsomedial PFC (dorsal ACC region) also showed a significant reduction in
activation. The reduction in dorsal activation for improvers with treatment may
indicate a reduction in deliberate control, which seems counterintuitive, since we
would expect deliberate regulation to be bolstered with treatment. It is possible,
however, that the dorsal ACC plays a supportive role, and that a decrease in
this particular system could reflect a reduced need of “deliberate” support when
ventral systems are doing their job more effectively in the first place. Bilateral
regions in the dorsolateral as well as ventrolateral PFC regions, implicated in
reappraisal strategies, did not show changes with treatment.
We concluded that children with externalizing problems mostly demon-
strated an overactivity in neural circuits indicative of a reactive style of emotion
regulation. The activation decreases in these systems for improvers could reflect a
reduction in the rigid, reactive style of emotion regulation which is characteristic
of children with behavior problems.
The anxiety hypothesis of aggression

How could these results increase our understanding of children’s externalizing
problem behavior and stimulate future research directions?
We concluded that a reduction in this ventrally mediated fronto-limbic activity
in improvers could indicate that this neural system of reactive regulation, and
the socio-emotional behaviors it mediates, has normalized. Intervention studies
with a neuroimaging component, conducted with populations diagnosed with
externalizing problems, are rare, so it is difficult to draw comparisons. However,
it is remarkable that treatment studies using brain imaging with internalizing
populations have shown reductions in similar brain systems. The overactivity in
those fronto-limbic systems in our clinical sample of externalizing children seems
to be in line with the results of imaging studies in anxious populations. Both
research programs find overactive limbic circuits to be associated with rapid,
threat-focused attentional biases (Bishop, 2007; Ressler and Mayberg, 2007). In
this light, it’s interesting to note that an extremely large proportion of children in
our sample showed high (clinical) levels of internalizing problems. This comor-
bidity was reported to be a general phenomenon in a recent review by Bubier and
Drabick (2009), and it may be key in interpreting our neural results as well. These
findings support the idea that ventrally mediated fronto-limbic overactivity in our
children could underlie their internalizing problem behavior and could explain
the tension these children bring to social situations—a tension that eventually
manifests in aggressive outbursts.
In our work, we have fostered a hypothesis of aggression based on comorbid
anxiety that can serve as the basis for further research. Similar to Dollard (1939),
the hypothesis proposes that aggression occurs because blocked goals would
lead to a state of frustration which could lead to aggression in order to resolve
this uncomfortable state. In our hypothesis, we propose that anxiety leads to an
increased risk of frustration that can underlie externalizing behavior problems
in a subset of children. Anxious individuals are known to cling to predictable
behavioral patterns and thoughts that provide them with a sense of control. This
perceived control can become so important that it becomes a goal in itself. But
since social life is full of events that are beyond one’s own control, such as other
people’s opinions about you, rapidly shifting group dynamics, and unexpected
mood swings, it is easy to feel a loss of control, particularly for anxious children.
Together with a negative threat-focused bias, anxious children could have a
greater propensity for becoming frustrated because more events are perceived as
threatening, blocking goals of safety and control. Frustration can then build up
and may result in aggression (see Figure 1, for illustration of the model). Whether
aggression occurs may depend on a child’s learned behavioral response style, the
intensity of the frustration, and a child’s self-regulatory strength. The overactivity
in ventrally mediated fronto-limbic systems could thus be seen as a neural marker
of this ineffective reactive regulation. The response patterns developed due to
these negative biases could then not only manifest themselves in maladaptive
avoidance behavior (e.g., the development of anxious behavior problems), but
also in maladaptive approach behavior (e.g., the development of aggressive
behavior problems).
Trait
State
Response
Anxiety Frustration Reactive

aggression
Figure 1 An anxiety hypothesis of aggression: a path through which anxiety can lead
to externalizing problem behavior.
The quest for many clinical neuroscientists is to find neural markers that relate to
the syndrome of interest. Neural markers can inform diagnosis, suggest prognosis,
and help decide on the most effective type of treatment, and they can also increase
understanding of neurocognitive mechanisms that translate into general models
of psychological functioning. The implications of our hypothesis for the devel-
opment and treatment of externalizing problem behavior may prove to be valuable.
Interventions could improve their effectiveness by directly targeting anxious
behavior for a significant subset of children. Whether we continue to validate this
hypothesis, or turn our attention to other approaches, we hope to have shown how
translational research that applies neuroscientific principles to clinical problems can
contribute to the discussion of hypotheses that benefit child-clinical practice.
References
Bekker, E. M., Kenemans, J. L., and Verbaten, M. N. (2005). Source analysis of the N2 in
a cued Go/NoGo Task. Brain Research Cognitive Brain Research, 22, 221–231.
Bishop, S. J. (2007). Neurocognitive mechanisms of anxiety: an integrative account.
Trends in Cognitive Sciences, 11, 307–316.
Bubier, J. L., and Drabick, D. A. (2009). Co-occurring anxiety and disruptive behavior
disorders: the roles of anxious symptoms, reactive aggression, and shared risk processes.
Clinical Psychology Review, 29, 658–669.
Cappadocia, M. C., Desrocher, M., Pepler, D., and Schroeder, J. H. (2009). Contextualizing
the neurobiology of conduct disorder in an emotion dysregulation framework. Clinical
Psychology Review, 29, 506–518.
Caspi, A., Moffitt, T. E., Newman, D. L., and Silva, P. A. (1996). Behavioral observations
at age 3 years predict adult psychiatric disorders: Longitudinal evidence from a birth
cohort. Archives of General Psychiatry, 53, 1033–1039.
Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., and Sears, R. R. (1939). Frustration
and Aggression. New Haven, CT: Yale University Press.
Falkenstein, M., Hoormann, J., and Hohnsbein, J. (1999). ERP components in Go/Nogo
tasks and their relation to inhibition. Acta Psychologica, 101, 267–291.
Kappas, A. (2011). Emotion and Regulation are One! Emotion Review, 3, 17–25.
Lewis, M. D., Granic, I., Lamm, C., Zelazo, P. D., Stieben, J., Todd, R. M., Moadab, I.,
and Pepler, D. (2008). Changes in the neural bases of emotion regulation associated
with clinical improvement in children with behavior problems. Development and
Psychopathology, 20, 913–939.
Phillips, M. L., Ladouceur, C. D., and Drevets, W. C. (2008). A neural model of voluntary
and automatic emotion regulation: implications for understanding the pathophysiology
and neurodevelopment of bipolar disorder. Molecular Psychiatry, 13, 829, 833–857.
Ray, R., and Zald, D. H. (2011). Anatomical insights into the interaction of emotion and
cognition in the prefrontal cortex. Neuroscience & Biobehavioral Reviews, 36, 479–501.
Ressler, K. J., and Mayberg, H. S. (2007). Targeting abnormal neural circuits in mood
and anxiety disorders: from the laboratory to the clinic. Nature Neuroscience, 10,
1116–1124.
Wiers, R. W., Eberl, C., Rinck, M., Becker, E. S., and Lindenmeyer, J. (2011). Retraining
Automatic Action Tendencies Changes Alcoholic Patients’ Approach Bias for Alcohol
and Improves Treatment Outcome. Psychological Science, 22, 490–497.
Woltering, S., Granic, I., Lamm, C., Lewis, M. D. (2011). Neural changes associated with
treatment outcome in children with externalizing problems. Biological Psychiatry, 70,
873–879.
Part 2
Learning Perspective
7 Individual differences in the
acquisition of fears
Susan Mineka
Northwestern University
Fear is a highly adaptive emotion that often signals potential or actual danger
in humans and many other species. Fear also serves as a central motive state,
sometimes motivating escape or avoidance behaviors. Not surprisingly many
have argued that fear has been central to mammalian evolution because it is a
product of natural selection and therefore shaped and constrained by evolutionary
contingencies (e.g., Öhman and Mineka, 2001). Some sources of fear are innate
(especially in subhuman animals), but it has long been known that many objects or
situations that we fear are based on learning. Indeed, since at least the 1920s it has
been known that many fears are based on a fundamental form of learning known
as classical or Pavlovian conditioning. In classical conditioning, the conditioned
stimulus (CS) is paired with an unconditioned aversive stimulus (US) one or more
times. Generally the CS gradually acquires the capacity to elicit a conditioned
defensive response that may resemble the unconditioned response, but may also
differ from it in significant ways such as being a compensatory response (e.g.,
Öhman and Mineka, 2001). Another important point about classical conditioning
relevant here is that once acquired through conditioning fears are not “forgotten”
simply with the passage of time. Instead, for CRs to gradually diminish there
must be a number of extinction trials in which the CSs are presented without any
USs (Mackintosh, 1974).
Direct classical conditioning is not the only pathway to the development of
fears and phobias. Indeed, humans and certain other species can readily acquire
fears either vicariously simply through observing a conspecific behaving fearfully
in the presence of some object or situation or, in humans, through verbal instruc-
tions about the danger posed by some object or situation. Mild and transient
fears have been conditioned this way in many studies in laboratory settings in
both adult humans and young children (see Askew and Field, 2008, for a recent
review). Moreover, strong and persistent phobic-like fears have been acquired
vicariously in numerous experiments conducted in lab-reared rhesus monkeys
by Mineka and colleagues (see Cook and Mineka, 1991, for a review). Indeed
lab-reared monkeys who were not initially afraid of snakes quickly acquired an
intense phobic-like fear of snakes (as indexed by three different measures) after
a total of only 24 minutes of exposure to a wild-reared monkey showing a strong
phobic-like fear to a live boa constrictor and toy snakes. Indeed, the level of fear
48 Susan Mineka
observed in the lab-reared monkeys following observational conditioning was
nearly as intense as that of the model monkeys who had acquired their phobic-
like fears in the wild in India several decades earlier. Another indication of the
robustness of the vicariously acquired fear was that, in one experiment, monkeys
who had acquired the fear vicariously successfully served as models for other
unrelated observer monkeys who then also acquired it vicariously. Moreover,
there were no signs of diminution of fear of the snakes over a three month
follow-up period.
When fears are acquired through direct or vicarious conditioning, or through
instructional/verbal learning there is a wide variation in how quickly they
are acquired. Speed of acquisition is, for example, partly a function of the
aversiveness of the US (cf. Mackintosh, 1974) with more intensely aversive USs
generally being associated with more rapid and robust conditioning. Another
important factor that can contribute to the speed of acquisition is the “belong-
ingness of the CS and US”. That is, some combinations of CSs and USs condition
especially well together relative to other combinations of CSs and USs (e.g.,
Hamm et al., 1989). In the early 1970s, Seligman had proposed the relevance of
work on belongingness (much of it done in animals) to the understanding of fears
and phobias (1971) in his classic paper on “phobias and preparedness”. Phobias
are generally seen as very intense and persistent fears that the person realizes are
excessive or unreasonable, and that are triggered by the presence of a specific
object or situation. When a person encounters a phobic stimulus, they often
show an immediate fear response that may resemble a panic attack except for the
presence of a clear external trigger for fear.
According to the preparedness theory of fears and phobias, we are evolution-
arily prepared to acquire fears and phobias more readily to certain objects or
situations that may once have represented a threat during our early evolutionary
history (Öhman and Mineka, 2001; Seligman, 1971). A large number of studies
conducted by Öhman and his colleagues with human participants and several
studies of lab-reared rhesus monkeys conducted by Mineka and colleagues have
provided strong support for many aspects of this theory, as reviewed by Öhman
and Mineka (2001). In dozens of human studies from Öhman’s laboratory, condi-
tioning to prepared or fear-relevant CSs (e.g., pictures of snakes or spiders) is
found to be more robust than when the CSs are unprepared or fear-irrelevant
(e.g., flowers or mushrooms). Robustness has most often been demonstrated by
participants’ showing greater resistance to extinction with prepared CSs than is
seen with unprepared or fear-irrelevant CSs. In addition, however, robustness of
prepared fears has also been demonstrated by findings that strong conditioning
can occur after only one CS−US pairing using prepared or fear-relevant CSs, but
not when using unprepared or fear-irrelevant CSs (Öhman and Mineka, 2001).
In Mineka and Cook’s experiments on preparedness, lab-reared rhesus monkeys
served as observers who watched model monkeys on spliced videotapes behaving
fearfully either with toy snakes or with brightly colored artificial flowers. (A prior
study had demonstrated that observer monkeys could acquire the fear vicariously
simply through watching a fearful monkey behaving fearfully with snakes on a
Individual differences in the acquisition of fears 49
videotape.) Only the observers who watched models behaving fearfully with toy
snakes acquired a fear of snakes; by contrast, the observers who watched models
behaving fearfully with brightly colored flowers acquired no fear of flowers. This
is in spite of the fact that the fear performance that the model monkeys showed to
toy snakes and to flowers was identical (which was accomplished through video
editing).
Thus both monkeys and humans seem selectively to associate certain fear-
relevant stimuli with threat or danger. So individual differences in who acquires
fears depend in part upon which stimuli are paired with threat. Moreover, the
lab-reared monkeys had no prior exposure to any of the stimuli involved (e.g.,
snakes or flowers) before participating in these experiments. Thus the monkey
results support the evolutionarily-based preparedness hypothesis even more
strongly than do the human experiments. For example, human participants (unlike
the lab-reared monkeys) might show superior conditioning in the laboratory to
snakes or spiders because of ontogenetic factors, such as preexisting negative
associations to snakes or spiders, rather than because of evolutionary factors.
The monkey experiments also have the advantage of demonstrating that selective
associations occur not only with mild and transient conditioning, as seen in the
human experiments, but also with intense and long-lasting phobic-like fears.
Nevertheless it should be acknowledged that preparedness theory has generated a
good deal of criticism and controversy (e.g., Davey, 1995), which was countered
by Öhman and Mineka (2001) in their extensive review of relevant evidence.
One common but seriously misconceived assumption about conditioning
models of fear acquisition is that people should be able to recall their direct,
vicarious, or instructional experiences. There are many reasons to doubt the
validity and accuracy of retrospective recall for such situations, and accord-
ingly recently developed non-associative accounts of the acquisition of fears
and phobias seem to have little merit (e.g., Mineka and Sutton, 2006). Another
common but seriously misconceived assumption about conditioning models is
that everyone exposed to the same direct or vicarious conditioning experience
should acquire comparable levels of fear. This is also clearly not the case. Instead
there is a host of genetic and temperamental, as well as experiential, variables that
strongly influence the speed and strength of conditioning in any given individual.
Among experiential variables, one is simply the familiarity of the CS to the
person prior to a conditioning experience. As illustrated by the well-known and
carefully studied phenomenon of latent inhibition, pre-exposure to a CS prior to
conditioning reduces the amount of fear that is subsequently conditioned relative
to what is observed with a truly novel CS. For example, in an illustrative study,
Davey (1989) asserted that children who reported having had first painless dental
treatments with a dentist were less likely to develop a dental phobia if they were
subsequently traumatized during a dental visit than were children without as many
earlier benign experiences with a dentist. Cook and Mineka (1991) also reviewed
evidence showing that protective or immunization experiences from pre-exposure
to a nonfearful model monkey behaving nonfearfully with snakes further reduced
the strength of conditioning when the observer later witnessed a fearful model
50 Susan Mineka
behaving fearfully with snakes. Indeed, there was no significant effect of condi-
tioning for 6 out of 8 monkeys that first had the immunization experiences.
Egliston and Rapee (2007) found somewhat parallel results regarding protective
effects of initial positive modeling by a mother against subsequent maternal
fearful modeling in one–two-year-old human children. If the children had first
watched their mothers show positive modeling with a fear-relevant stimulus, they
were later protected against the effects of fearful modeling by the mother, relative
to children without the initial positive maternal modeling.
Another important experiential variable during conditioning that strongly
affects the amount of fear that is conditioned in an aversive situation is the degree
of control the individual has over either the onset or the offset of the US. In
traditional classical conditioning experiments, the organism is a passive recipient
of the CSs and USs that are delivered because these are all controlled by an
experimenter. But in our everyday lives, when conditioning experiences typically
occur, the person often has control over some aspect of the situation, such as
controlling when the US will terminate (e.g., by escaping from it). Being able to
control the offset of the US paired with any CS has a major effect of attenuating
the amount of fear that is conditioned. Indeed, in one illustrative experiment with
rats, Mineka, Cook, and Miller (1984) found that rats that had experienced tones
paired with escapable shocks later showed only approximately half the levels of
conditioned fear seen in rats that had experienced tones paired with exactly the
same amount and number of shocks that were inescapable/uncontrollable. Thus
the dynamics of classical conditioning are dramatically affected by the control-
lability of the US.
Individual differences in what happens following a conditioning experience
also affect the amount of fear that is maintained over time because fear memories
are somewhat malleable. For example, Rescorla (1974) reported that if rats are
first conditioned to show a mild fear (by pairing a CS with a mild US), their fear
level will later increase if they are subsequently exposed to more intense USs (not
paired with the CS)—a phenomenon known as the “inflation effect”. Similarly,
in humans, White and Davey (1989) demonstrated an inflation effect when US
re-evaluation occurred following conditioning by delivering more intense USs
(alone) than had been used during conditioning. That is, the participants in the
inflation group later showed increased levels of fear, even though they had not
had any pairings of the CS with the more intense USs. Thus whether or not
someone has an inflation experience affects the amount of fear they maintain into
the future.
In addition to experiential variables on which individuals may differ, there are
also temperamental and personality variables that affect the acquisition of fear.
For example, prospective studies initiated in the 1980s determined that toddlers
with a behaviorally inhibited temperament (i.e., shy, timid, easily distressed)
assessed at 21 months were at higher risk for developing multiple specific phobias
by age seven or eight than were uninhibited children (32 per cent versus 5 per
cent) (Biederman et al., 1990), although it is not yet clear if these phobias were
acquired through conditioning. Personality variables such as neuroticism and trait
Individual differences in the acquisition of fears 51
anxiety also affect the speed and strength of conditioning in laboratory studies
(see Oehlberg and Mineka, 2011, for a review).
An exciting recent study (Lonsdorf et al., 2009) using sophisticated molecular
genetic techniques has also shown that several specific genetic polymorphisms of
the 5-HTTLPR and COMT genes have strong effects on either fear acquisition
or fear extinction. Normal college students first underwent DNA extraction from
blood for genotyping, and then underwent a discriminative fear conditioning
procedure using facial stimuli as CS+s and CS−s. Startle potentiation was the
primary index of fear. Carriers of one or two short alleles (s or ss) of the 5HTT
gene showed significantly stronger fear potentiation then did the l/l homozygous
carriers and this pattern persisted in extinction. In contrast, the two polymor-
phisms of the COMTVal158Met gene had no effect during acquisition. However,
the results during extinction were very different. Specifically, those with the
homozygous met/met COMT Val158Met polymorphism showed much greater
CS+ fear potentiation than those who were COMT Val-allele carriers who showed
no fear at all (i.e., no resistance to extinction). The participants who showed the
most pronounced startle responding in extinction were those with at least one
short s allele of the 5 HTT gene and two COMT met alleles. The authors suggest
that such individuals “are likely to expand their sets of fear- and anxiety-evoking
stimuli through facilitated fear conditioning and poor extinction” (p. 204). These
exciting results obviously need to be replicated before strong conclusions can be
drawn, but they may provide a precise mechanism helping to account for why
only a subset of people develop fears of many different stimuli, and why such
fears may be so persistent in only a subset of them.
In conclusion, although fears can be innate, they are frequently acquired through
direct or vicarious conditioning, or instructional learning. Importantly, however,
not all individuals who have the same learning experiences will acquire or retain
the same levels of fears. To the contrary, the acquisition of fears is strongly influ-
enced by a host of individual difference variables. Some of these variables are
genetic and temperamental differences over which we have little or no control,
but a myriad of other such variables reflect wide individual differences in life
experiences that people have had prior to, during, or following conditioning. Due
to space limitations, what has been reviewed here is only an illustrative set of
examples of such individual differences, but they should be sufficient to convince
readers that there is what I once called a “frightful complexity of the origins of
fears.”
References
Askew, C., and Field, A. P. (2008). The vicarious learning pathway to fear 40 years on.
Clinical Psychology Review, 28, 1249–1265.
Biederman, J., Rosenbaum, J. F., Hirshfeld, D. R., Faraone, S. V., Bolduc, E. A., Gersten,
M., et al. (1990). Psychiatric correlates of behavioral inhibition in young children of
parents with and without psychiatric disorders. Archives of General Psychiatry, 47,
21–26.
52 Susan Mineka
Cook, M., and Mineka, S. (1991). Selective associations in the origins of phobic fears
and their implications for behavior therapy. In P. Martin (ed.), Handbook of Behavior
Therapy and Psychological Science: An integrative approach (pp. 413–434). New York:
Pergamon Press.
Davey, G. (1995). Preparedness of phobias: Specific evolved associations or a generalized
expectancy bias? Behavioral and Brain Sciences, 18, 289–325.
Egliston, K., and Rapee, R. (2007). Inhibition of fear acquisition in toddlers following
positive modeling by their mothers. Behaviour Research and Therapy, 45, 1871–1882.
Hamm, A., Vaitl, D., and Lang, P. J. (1989). Fear conditioning, meaning and belongingness:
A selective association analysis. Journal of Abnormal Psychology, 98, 395–406.
Lonsdorf, T. B., Weike, A. I., Nikamo, P., Schalling, M., Hamm, A. O., and Öhman, A.
(2009). Genetic gating of human fear learning and extinction: Possible implications
for gene-environment interaction in anxiety disorder. Psychological Science, 20, (2),
198–206.
Mackintosh, N. (1974) The psychology of animal learning. London: Academic Press.
Mineka, S., Cook, M., and Miller, S. (1984). Fear conditioned with escapable and
inescapable shock: Effects of a feedback stimulus. Journal of Experimental Psychology:
Animal Behavior Processes, 10, 307–323.
Mineka, S., and Sutton, S. (2006). Contemporary learning theory perspectives on the
etiology of fears and phobias. In M. Craske, D. Hermans, and D. Vansteenwegen (eds).
Fear and Learning: From Basic Proccesses to Clinical Implications (pp. 75–97). APA
Books.
Oehlberg, K., and Mineka, S. (2011). Fear conditioning and attention to threat: An
integrative approach to understanding the etiology of anxiety disorders. In T. Schactman
and S. Reilly (eds), Associative Learning and Conditioning: Human and Animal
Applications (pp. 44–78). Oxford University Press.
Öhman, A., and Mineka, S. (2001). Fears, phobias, and preparedness: Toward an evolved
module of fear and fear learning. Psychological Review, 108, (3), 483–522.
Rescorla, R. A. (1974). Effect of inflation of the unconditioned stimulus value following
Conditioning. Journal of Comparative and Physiological Psychology, 86, (1), 101–106.
Seligman, M. E. P. (1971). Phobias and preparedness. Behavior Therapy, 2, 307–332.
White, K., and Davey, G. (1989). Sensory preconditioning and UCS inflation in human
‘fear’ conditioning. Behaviour Research and Therapy, 27, 161–166.
8 Extinction learning and its retrieval
Michelle G. Craske
University of California, Los Angeles
Bram Vervliet
University of Leuven
Fear conditioning involves learning a relationship between a neutral stimulus,

the conditional stimulus (CS), and an innately aversive stimulus (the uncondi-
tional stimulus, US), such that the CS elicits a conditional response (CR). Fear
conditioning has been applied as an etiological model for phobias and anxiety
disorders. Fear extinction involves the decay of CRs with repeated presenta-
tions of the CS without the US. Deficits in fear extinction are believed to
contribute to the persistence of phobias and anxiety disorders, and the process
of fear extinction parallels the clinical practice of exposure therapy for anxiety
disorders. Significant advances have been made in the behavioral and neurobio-
logical mechanisms underlying extinction, their relevance to phobias and anxiety
disorders, and their translation to exposure therapy.
The specific line of research covered in this article pertains to the theory
that extinction involves the formation of a new inhibitory association between
representations of the CS and the US, which actively suppresses the US memory.
Thus, the amount of fear expressed after completion of extinction training
is dependent on which set of associations is retrieved, the initial excitatory
association or the newly formed inhibitory association (Bouton, 2002). This
retrieval model of extinction is described in detail by Nelson (in this volume).
Herein, we review the neurobiological evidence, and strategies for enhancing
the formation and retrieval of inhibitory associations, and their translation to
exposure therapy for anxious individuals who appear to show deficits in inhib-
itory learning.
Retrieval model of extinction

In the retrieval model of extinction, inhibitory associations compete with original
excitatory associations and serve to deactivate the US memory. Also, the inhib-
itory association is dependent on both the CS and the context in which the CS
is presented, whereas the initial excitatory association is independent of context.
Thus, the context can be seen as a way of disambiguating the current meaning of
the conditioned-and-extinguished CS (i.e., the excitatory “danger” or inhibitory
“safety” meaning). Additionally, the context may serve as a retrieval cue for the
extinction memory that might otherwise be forgotten.
54 Michelle G. Craske and Bram Vervliet
Evidence for the fragility of fear extinction is found in four behavioral
paradigms that lead to a return of conditional fear, even after complete extinction
of that fear. That is, each paradigm shows a continuing effect of the original
excitatory association after extinction (see Nelson, in this volume, for a detailed
description of each paradigm). They include: spontaneous recovery, which trans-
lates to a return of fear when a previously feared stimulus is encountered for the
first time after a lengthy interval since completion of exposure therapy; context
renewal, which translates to a return of fear when the previously feared stimulus
is encountered in a context that is distinctly different from the exposure treatment
context; reinstatement, which translates to return of fear due to independent
adverse events following exposure therapy; and rapid reacquisition, which trans-
lates to easy and rapid reacquisition of fear with re-traumatization, as may occur
in combat situations or other dangerous environments.
Neurobiology of fear extinction

The brain-behavior associations underlying fear conditioning and extinction
have been well mapped. The amygdala plays a primary role in fear conditioning.
In animals, amygdala lesions interfere with fear conditioning, and in healthy
humans, fMRI studies show elevated amygdala activation during conditioning
(see Shin and Liberzon, 2010, for a review). Conditional fear is presumed to be
mediated by the transmission of sensory information about the CS and US to
the amygdala and the subsequent control of fear reactions via projections from
the amygdala to hypothalamic and brainstem regions that regulate behavioral,
endocrine, and autonomic responses (LeDoux, 2000). Other regions associated
with fear conditioning are (1) the hippocampus, which is involved in processing
contextual cues of conditioning, (2) the insular cortex, which is involved in inter-
oception and awareness of and sensitivity to visceral activity, and (3) the dorsal
and rostral ACC, which appear to have a role in anticipation of the CS and US.
The ventromedial prefrontal cortex (vmPFC) is believed to mediate extinction,
and the hippocampus to modulate extinction by providing information regarding
safe versus dangerous contexts. In animals, extinction and extinction retest are
impaired by lesions of the vmPFC and enhanced by electrical stimulation of the
vmPFC. Human neuroimaging studies have shown increased vmPFC activation
during extinction and at extinction retest (e.g., Milad et al., 2009). Also,
hippocampal lesions can reduce the context-dependence of extinction in rodents,
and hippocampal activation occurs during extinction in humans. Furthermore,
vmPFC and hippocampal activity are positively correlated during extinction and
extinction retest, suggesting their interaction may be particularly important for
effective extinction (Shin and Liberzon, 2010). Finally, activity in the vmPFC
and hippocampus has been found to correlate with strength of extinction retest,
as measured using skin conductance response in human samples, as has the
thickness of the vmPFC in one study. It is theorized that the PFC exerts inhibitory
control over the amygdala through vmPFC activation of inhibitory lateral nucleus
interneurons or inhibitory projections to the central nucleus of the amygdala. That
Extinction learning and its retrieval 55
is, the PFC serves as the neurobiological basis for inhibitory learning. Also, it is
presumed that the hippocampus creates a unique representation of the context in
which extinction took place, and modulates the effect of the PFC on the amygdala
(Quirk and Mueller, 2008).
Deficits in extinction learning in anxiety disorders

A meta-analysis of behavioral studies showed that individuals with a variety
of anxiety disorders exhibit stronger responding to the CS+ not only during
conditioning, but also during extinction relative to controls (Lissek et al.,
2005). Furthermore, in differential conditioning paradigms, they showed elevated
responding to the CS− (the stimulus never paired with the US) during condi-
tioning and extinction relative to controls. Since the meta-analysis, additional
studies have confirmed either delayed extinction (i.e., elevated responding to the
CS+ during extinction training) or deficits in extinction at retest (i.e., elevated
responding to the CS+ at extinction retest), as well as elevated responding to the
CS−, in anxiety disorders. One explanation of these findings is that individuals
with anxiety disorders show both elevated excitatory learning and impaired
inhibitory learning (Lissek et al., 2005). Impaired inhibitory (or safety) learning
to either a stimulus that does not signal threat (CS−) or to a stimulus that no
longer signals threat (CS+ during extinction) has been theorized to be central to
the pathology of anxiety disorders. In accord, individuals with anxiety disorders
show deficits in direct tests of inhibition, such as the transfer of safety learning to
an excitatory stimulus in summation tests (Jovanovic et al., 2010). Consequently,
one might expect anxious individuals to show stronger evidence for spontaneous
recovery, context renewal, reinstatement, and rapid reacquisition than healthy
controls, although no studies have directly made these comparisons.
In addition, at the neural level, one would expect anxious individuals to show
deficits in vmPFC during extinction. Indeed, although limited to only a few
studies, there is evidence for decreased orbitofrontal and medial PFC (including
vmPFC/subgenual ACC) activation during extinction and at extinction retest
in adults with posttraumatic stress disorder and high trait anxiety relative to
increased activity in those areas in healthy controls (e.g., Milad et al., 2009).
Enhancing the formation and retrieval of extinction learning

Significant advances are being made in ways to enhance the formation of
inhibitory associations throughout extinction and their retrieval at retest. These
advances have direct relevance for exposure therapy for individuals with anxiety
disorders, who appear to be in particular need of such enhancement strategies
given their deficits in inhibitory learning.
An obvious approach for enhancing the formation of inhibitory learning is to
simply increase the number of extinction trials. In rodent samples, “massive”
numbers of extinction trials (i.e., 100 times the number of acquisition trials) have
been shown to offset context renewal effects compared to a moderate number
of extinction trials (20 times the number of acquisition trials). However, human
conditioning studies are absent, and the one clinical analog study failed to find
effects perhaps due to the limited range in number of exposure trials tested.
Another possibility is to conduct “super-extinction” involving multiple condi-
tioned excitors throughout extinction training that provide stronger evidence
for disconfirmation of “danger” expectancies than a single conditioned excitor.
Animal research has shown positive benefits in terms of reducing context renewal
although this has not been found in human conditioning studies, perhaps because
humans treat the compound presentation of two stimuli during extinction as a
different stimulus than each stimulus presented separately at extinction retest.
In “deepened extinction” (Rescorla, 2006), multiple fear conditional stimuli are
first extinguished separately before being combined during extinction. In animal
studies, deepened extinction decreases spontaneous recovery and reinstatement of
fear. Initial human laboratory research in this area is showing similar results. Also,
for some time, exposure protocols for panic disorder have included deepened
extinction in the form of independent exposure to feared interoceptive stimuli and
to feared external situations that are combined in subsequent exposures.
Elimination of safety signals or safety behaviors (e.g., presence of another
person, superstitious objects, medications) during extinction is another option, since
they protect from extinction in rodent samples. Several studies have shown that use
or availability of safety signals during exposure therapy for specific fears weaken
the overall outcomes. However, recent data have presented contradictory findings.
Based on evidence for fear extinction to be weakened by antagonists of the
glutamate receptors in the amygdala, Walker et al. (2002) tested and demonstrated
that drug agonists of the same receptors, and in particular, d-cycloserine, enhance
extinction in animal studies. Later studies showed that d-cycloserine is mainly
involved in the consolidation of newly formed extinction memories, because
post-extinction administration of the drug produced the same effect (up to four
hours). Others have shown the d-cycloserine prevents reinstatement, although
it does not influence renewal and rapid reacquisition in rodents. Furthermore,
in excellent clinical translation work, d-cycloserine has been shown to improve
outcomes at follow-up, after exposure therapy, for several anxiety disorders. On
the other hand, other experimental data show that d-cycloserine may augment
fear under some conditions in both rodent and human samples (Vervliet, 2008).
A number of options for enhancing retrieval of the extinction memory have
been tested. One option is to include retrieval cues during extinction training to be
used in other contexts once extinction is over. This has been shown to be effective
in animal studies. The positive effects of retrieval cues on context renewal have
been demonstrated in human conditioning studies as well, although they appear to
acquire an inhibitory value and became a safety signal. In clinical analog samples,
the effects of a retrieval cue upon context renewal were very weak in one study,
although instructions to “mentally reinstate”, or to recall, what was learned during
exposure had more robust effects in reducing context renewal in another study.
Another option is to provide multiple contexts in which extinction takes place.
This approach has been shown to offset context renewal in rodent samples, and in
a clinical analog study of exposure therapy. On the other hand, one conditioning
study in rodents and another in humans failed to demonstrate detectable benefits
of multiple contexts throughout extinction on context renewal.
Further options include increasing stimulus variability throughout exposure,
since varying the to-be-learned enhances the retention of newly learned infor-
mation. Variability is believed to enhance the storage capacity of newly learned
information, pair the information to be learned with more retrieval cues, or
generate a rule that captures the invariance among tasks. Variability in terms
of the length of time between exposure sessions or the stimuli used during
exposure was demonstrated to lessen spontaneous recovery in three studies of
fearful individuals, while a fourth study did not. Emotional variability throughout
exposure also predicted less spontaneous recovery in two studies of fearful
individuals. Conceivably, emotional state serves as a retrieval cue or aid in
generalization.
Weakening the fear memory

Memories go through a typical consolidation phase after they are formed. Hence,
studies have been conducted to assess whether extinction conducted immedi-
ately after acquisition interrupts the consolidation of the acquisition memory. In
support, one rodent study showed that immediate extinction (ten minutes–one
hour) led to little or no return of fear, in terms of renewal, reinstatement, or
spontaneous recovery, relative to extinction 24–72 hours later. However, another
study reported the opposite results, and the discrepancy appears attributable to the
level of fear at the start of extinction. If fear is low prior to immediate extinction,
then the effects of immediate extinction are stronger. In other words, immediate
extinction may be most powerful for mild fears. Also, two human conditioning
studies have yielded inconsistent or null findings and have failed to assess the
long-term effects of immediate extinction. Thus, immediate extinction does not
appear to be a robust phenomonen at this time.
Another recent (re-)discovery is that retrieving already stored memories
induces a process of reconsolidation (Nader et al., 2000), since the memory
is written into long-term memory again, requiring de novo neurochemical
processes. Thus, it may be possible to change memories during the recon-
solidation time frame. Propranolol, a beta blocker, has been shown to block the
reconsolidation of memories, and Debiec and Ledoux (2004) found that infusions
of propranaolol blocked the reconsolidation of a previously formed CS−US
memory, and led to erasure of the fear response and resistance to reinstatement
effects. This suggests that propranolol upon retrieval may be a useful clinical tool,
and indeed, two fear conditioning studies in healthy humans have replicated the
effects.
Finally, Monfils et al. (2009) used an alternative strategy to make use of
reconsolidation mechanisms to weaken fear memories, hypothesizing that novel
information presented during the reconsolidation window may be incorporated
into the memory and change it. Thus, extinction during a reconsolidation window
may weaken the fear memory itself, and indeed this is what their results showed
in terms of the four main markers (spontaneous recovery, renewal, reinstatement,
and rapid reacquisition) in a rodent sample. This effect has since been demon-
strated for spontaneous recovery and reinstatement in healthy human samples and
tied to underlying cell processes. On the other hand, an opposite pattern of results
was reported in a rodent sample by another investigating team. Furthermore, it
will be difficult to translate this paradigm into clinical practice, where timing of
exposure to reminders of feared stimuli is difficult to control.
Summary
In sum, there are some promising methods for enhancing retrieval of extinction
learning and weakening the fear memory. Each has significant implications for
enhancing the outcomes from exposure therapy for anxiety disorders. However,
much more research is needed, since the critical determinants are often not
known, and either there are failures as well as successes in each case, or the
results are based on single studies and warrant replication.
References
Bouton, M. E. (2002). Context, ambiguity, and unlearning: sources of relapse after behav-
ioral Extinction. Biological Psychiatry, 52, 976–986.
Debiec, J., and LeDoux, J. E. (2004). Disruption of reconsolidation but not consolidation
of auditory fear conditioning by noradrenergic blockade in the amygdale. Neuroscience,
1, (29), 267–272.
Hermans, D., Dirikx, T., Vansteenwegen, D., Baeyens, F., Van den Bergh, O., and Eelen,
P. (2005). Reinstatement of fear responses in human aversive conditioning. Behaviour
Research and Therapy, 43, 533–551.
Jovanovic, T., Norrholm, S. D., Blanding, N. Q., Davis, M., Duncan, E., Bradley, B., et al.
(2010). Impaired fear inhibition is a biomarker of PTSD but not depression. Depression
and Anxiety, 27, 244–251.
LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23,
155–184.
Lissek, S., Powers, A. S., McClure, E. B., Phelps, E. A., Woldehawariat, G., Grillon,
C., et al. (2005). Classical fear conditioning in the anxiety disorders: a meta-analysis.
Behaviour Research and Therapy, 43, 1391–1424.
Milad, M. R., Pitman, R. K., Ellis, C. B., Gold, A. L., Shin, L. M., Lasko, N. B., et al.
(2009). Neurobiological basis of failure to recall extinction memory in posttraumatic
stress disorder. Biological Psychiatry, 66, 1075–1082.
Monfils, M. H., Cowansage, K. K., Klann, E., LeDoux, J. E. (2009). Extinction-
reconsolidation boundaries: key to persistent attenuation of fear memories. Science,
324, 951–955.
Nader, K., Schafe, G. E., and LeDoux, J. E. (2000). Fear memories require protein
synthesis in the amygdale for reconsolidation after retrieval. Nature, 406, 722–726.
Norberg, M. M., Krystal, J. H., Tolin, D. F. (2008). A Meta-Analysis of D-Cycloserine
and the Facilitation of Fear Extinction and Exposure Therapy. Biological Psychiatry,
63, 1118–1126.
Quirk, G. J., and Mueller, D. (2008). Neural mechanisms of extinction learning and
retrieval. Neuropsychopharmacology, 33, 56–72.
Rescorla, R. A. (2006). Deepened extinction from compound stimulus presentation.
Journal of Experimental Psychology: Animal Behavior Processes, 32, 135–144.
Shin, L. M., and Liberzon, I. (2010). The neurocircuitry of fear, stress, and anxiety
disorders. Neuropsychopharmacology, 35, 169–191.
Vervliet, B. (2008). Learning and memory in conditioned fear extinction: effects of
D-cycloserine. Acta Psychologica, 127, 601–613.
Walker, D. L., Rattiner, L. M., and Davis, M. (2002). Group II metabotropic glutamate
receptors within the amygdala regulate fear as assessed with potentiated startle in rats.
Behavioral Neuroscience, 116, 1075–1083.
9 Mechanisms of extinction in
emotional regulation
James Byron Nelson
University of the Basque Country UPV/EHU
Associative-learning processes endow stimuli with the ability to elicit powerful

emotions that both produce behaviors in their own right, and motivate behaviors
that manage those emotions (see Nelson and Bouton, 2002). The relationships
that establish associations between stimuli, however, are not static and the
learning process must be able to adapt to changing contingencies. Just as condi-
tioning is a way in which adaptation occurs through associative learning that
allows important events to be expected, extinction is a way in which we learn that
ordinarily expected events no longer occur.
A person that has a traffic accident on the way to work will likely experience
anxiety on that route in the future. Such anxiety may motivate him or her to
take alternate routes to control the anxiety. In the laboratory, a tone might signal
an imminent shock, eliciting fear in a rat. Given the opportunity to escape
the signal and avoid the shock, the rat will undoubtedly do so. These initially
neutral stimuli are conditioned stimuli (CSs) that acquire their ability to elicit
and modulate emotions through their pairing with inherently significant stimuli
(unconditioned stimuli or USs). With exposure to the travel route (in the absence
of further mishap) anxiety should subside. With exposure to the tone alone, the
rat’s fear will be reduced. In both situations, the CS now occurs without the
US, which procedurally defines extinction. Theoretically, we can also say that a
process of “extinction” took place. This chapter discusses processes that might
be implied by the latter use of the term. The discussion will be devoted, in large
part, to the observation that extinction leaves the original learning more or less
intact, and does not involve unlearning. Thus, the discussion will omit theories
that are important, but do not necessarily address the representational content of
what could be learned during extinction or do not provide a mechanism through
which initial learning can remain intact following extinction (e.g., Capaldi, 1967;
Gallistel and Gibbon, 2000).
Pavlov suggested that the representation of the CS undergoes modification
during extinction such that it would be unable to activate the representation of the
US. Support for this idea can be found in an experiment on “negative-patterning”
in pigeons by Robbins (1990). Pigeons first learned a discrimination where two
visual CSs were paired with a food US when presented alone (i.e., CS1−US, CS2−
US) but not when presented together sequentially (i.e., CS1→CS2 – No US). Birds
Mechanisms of extinction in emotional regulation 61
first mastered the discrimination showing strong responding to either CS alone,
but not when presented sequentially. Next the birds underwent extinction with
CS1. The ability of CS1 and CS2 to jointly control non-responding when presented
together was believed to be a function of the internal representations of these
stimuli. Therefore, if the reduction in responding to CS1 during extinction was
due to a change in the internal representation evoked by CS1, the reduction would
paradoxically weaken CS1’s ability to control non-responding in the CS1→CS2
compound; exactly the result obtained. Extinction of CS1 produced a decrease in
responding to CS1, and an increase in responding to the CS1→ CS2 compound.
The CS−focused approach to extinction proposed by Pavlov links extinction to
processes associated with habituation, opening a variety of additional theoretical
avenues by which extinction might be better understood. Such a case is presented
in detail by McSweeney and Swindle (2002).
Other theories can be portrayed as involving “interference”. One idea repre-
sented today in the work of Wagner (1981), assumes that “excitatory” associations
are learned between the CS and US during conditioning so that the CS can evoke
an internal representation of the US. During extinction, an “inhibitory” association
is formed that interferes with the ability of the excitatory association to elicit the
US representation, leaving the CS functionally neutral. This idea makes many
of the same predictions as the subtly distinct one put forth by Jerzy Konorski,
where excitatory associations are formed during extinction between the CS and a
representation of the absence of the US, or “No-US” representation. This No-US
representation suppresses the activation of the US representation. In this theory, the
interference is not between associations competing to activate a single represen-
tation, but between different activated representations. In both, the original learning
remains intact during extinction upon which new learning is superimposed.
Research on “conditioned inhibition” is relevant to this latter interpretation
of extinction as conditioned inhibition is assumed to be fundamentally the
same as that inhibition acquired during extinction. In a conditioned inhibition
procedure, one stimulus signals a US (e.g., CS1−US) but not when combined
with another stimulus (CS1CS2−No US). In these designs, the organism treats
CS2 as if it predicts the absence of the US. CS2 will suppress the emotional
reactions elicited both by CS1, and other CSs conditioned with the same US.
There are examples where an extinguished stimulus also shares these properties
(e.g., Calton, Mitchell, and Schachtman, 1996). Inhibitors appear to elicit a
representation with its own emotional characteristics. Pigeons, for example,
will actively avoid an inhibitor for food. Additionally, an inhibitor for shock
seems to be motivationally similar to an excitor for food, and vice-versa (see
Dickinson and Dearing, 1979). These observations are fundamental in appre-
ciating that extinction is an active process with its own significance. Extinction
elicits emotional states associated with the absence of the US, such as frustration
when an expected favorable US does not occur, or relief when an expected
aversive US is omitted. Extinction involves learning about the absence of the
US; a representation with its own emotional properties which can motivate
behavior.
62 James Byron Nelson
A third form of interference occurs, not at the representational level, but
at the response level. Evidence for such a mechanism comes predominately
from studies of instrumental conditioning where responses are maintained by
their consequences. Instrumental conditioning is a mechanism through which
particular behaviors are learned that can help to manage emotions, such as
changing the route to work in the example above. When an instrumental response
(e.g., lever pressing) is extinguished in the presence of a stimulus, that stimulus
can later suppress that particular response more than some other response (e.g.,
chain pulling) that was maintained by the same outcome. If the stimulus were to
be associated with only the absence of the outcome during extinction, it should
suppress either response equally. Such an effect is consistent with the idea that
extinction involves more than a mechanism that interferes with the represen-
tation of the outcome, but also involves direct suppression of an outcome-guided
response (see Rescorla, 2001).
The interference-type theories discussed above provide mechanisms through
which initial learning is prevented from being expressed, yet leave that learning
more or less intact. That characteristic is important for any theory of extinction
because extinguished conditioned responding can recover, as demonstrated by
“spontaneous recovery”, “reinstatement”, and “renewal” effects. Spontaneous
recovery occurs when a period of time passes without exposure to the CS
following extinction. When the CS is encountered after that interval the condi-
tioned response evident prior to extinction is typically observed. Reinstatement
is a procedure that involves unsignaled re-exposure to the US. After extinction, a
single unsignaled exposure to the US is sufficient to cause a restoration of condi-
tioned responding to the CS.
Renewal is arguably the most important of the three effects mentioned
above as it can provide an explanation of the other two, as shall be discussed
shortly. Renewal refers to the recovery of an extinguished conditioned response
that happens when the CS is presented in a context other than that in which
extinction occurred. The term can also refer to the experimental paradigms that
produce “renewal”, or the set of theoretical constructs typically used to explain
it. Procedurally renewal comes in three forms. It can occur in an “ABA” design
where initial learning occurs in one context, “A”, extinction occurs in a different
context, “B”, and testing occurs back in the training context. Renewal is also
observed when initial learning, extinction, and testing all occur in different
contexts (ABC) or initial learning and extinction occur in the same contexts,
but testing is in a different context (AAB). Examples can be found referenced in
Nelson et al. (2011).
Renewal has been explained by assuming that inhibition learned during
extinction is gated by the context in which the learning occurs. Bouton (2004)
provides an excellent source for the origins and a clear exposition of that, and
the following, ideas. The language used in Bouton’s account describes the
phenomenon as an interference effect that is linked to memory retrieval, and
retrieval failure. Extinction involves new learning whose retrieval interferes with
old learning. New learning that is established after initial learning is assumed
to be more contextually dependent for retrieval and expression than the initial
learning. Thus, when tested in the context where the new learning was acquired,
performance associated with that new learning (e.g., extinction) interferes with
that acquired in the original learning experience. When tested outside the context
where the new learning occurred, the new information is not fully retrieved and
performance associated with the original learning is expressed with less inter-
ference from the newer learning.
The explanation offered above allows for reinstatement and spontaneous
recovery to likewise be explained as other examples of renewal. Many different
things can serve as contexts, including stimuli as varied as physical boxes or
rooms to drug states. Conceptualizing the passage of time as a gradually changing
context allows spontaneous recovery to be explained as an example of renewal.
Reinstatement can likewise be explained as a case of the context-specificity of
extinction. Unsignaled presentations of the US serve to condition context-US
associations, which were not a feature of the context during extinction. In that
way, the context of testing is made different from the context of extinction, and
the recovery observed to the CS can be treated as a renewal effect.
The exact process by which contextual control emerges during extinction is
presently unknown. Available evidence shows that contextual control appears
not only with extinction, but also in cases where a CS is associated with multiple
outcomes. After a CS is associated with one outcome, learning about the CS and
a new outcome comes under the control of the context regardless of whether the
new learning involves an inhibitory process (as in extinction) or not. Acquisition
of contextual control also appears related to the acquisition of interfering infor-
mation. That is, a stimulus may be associated with multiple outcomes across
phases as discussed at the start of this paragraph, but if those outcomes do
not produce behaviors that interfere with each other, contextual control is not
observed. Juan Rosas et al. (see Rosas, Callejas-Aguilera, Ramos-Álvarez, and
Abad, 2006) draw on an existing assumption that when a stimulus is associated
with different outcomes, attention is directed to contexts to resolve the resulting
ambiguity. Their new assumption is that once attention is directed to a context,
everything learned afterwards in that context will be context-dependent. Thus,
they assume that extinction evokes attention to contexts, and contextual control is
a direct function of that attention. Evidence for that suggestion is presently mixed
(see Nelson, Lombas, and Leon, 2011).
Extinction is not only an important phenomenon for basic research; it has
significant applied relevance (see Craske and Vervliet, this volume). Exposure-
based therapies involve extinction to remove unwanted emotions and behaviors,
and, not surprisingly, the success of these therapies is compromised by relapse.
Understanding the ways in which behavior is changed through extinction could
suggest important ways to improve treatments. Some ways that have been
successful involved presenting stimuli at test that were also present in the
extinction context. Other procedures have sought to enhance the generalization of
extinction by extinguishing the relevant stimuli in multiple contexts, meeting with
both success and failure (see Bouton, Garcia-Gutierrez, Zilski, and Moody, 2006).
When investigating ways to reduce recovery from extinction, it is important to
consider the many mechanisms through which such performance could appear
to recover. The renewal effect demonstrates that performance produced by an
extinction procedure is clearly affected by changes in context. It is worth consid-
ering, however, that the reports of the effect in the literature may over-represent the
extent to which extinction learning has been demonstrated as being contextually
controlled. There are many ways that an extinguished response could appear to
recover that do not require the learning about the CS during extinction to be context
specific. For instance, in an ABA design, some portion of the initial learning might
not transfer to Context B. Thus, although responding would be shown to decrease
in that context, the decrease might not entirely be due to new learning during
extinction, but rather due to a generalization decrement. Recovery in Context A
would simply be recovery from the generalization decrement that occurred when the
CS was placed in Context B. Alternatively, extinction could transfer well to the test
context, yet responding might be observed through simple associative summation
between any context-US associations present in the conditioning context and
residual associations remaining to the CS. Though the boundary conditions of such
summation are not completely understood, contexts and CSs can summate in both
animal and human methodologies (see Leon, Abad, and Rosas, 2011).
Ruling out such mechanisms to conclude that extinction learning is context
specific, within a particular method, relies on evaluating the associative status
of the context, or testing in a context associated neither with conditioning nor
extinction as in an ABC design. Interestingly, ABC renewal is reported to be
smaller than ABA renewal. Presumably associations accrued to Context A may
contribute to the recovery observed, thus such evidence suggests that ABA
renewal might overestimate the true context-specificity of extinction. It is also
possible that the extinction context itself could become inhibitory. Should the
context itself signal the absence of the US directly, the US would not be expected
when the CS is presented. Therefore, no extinction would need to take place with
the CS (the so-called “protection from extinction effect”).
Recovery has, of course, been observed when context-US associations are
ruled out (see Nelson et al., 2011, for discussion). Nevertheless, those observa-
tions do not guarantee that any recovery observed in a renewal paradigm, with
any method or species, will be due to context-specific extinction learning as can
be implied by the term “renewal”. Treatments developed that might affect one
mechanism may not be expected to affect all mechanisms through which perfor-
mance eliminated by an extinction manipulation can recover.
Extinction involves new learning with its own representational and emotional
properties, and that learning has been shown to be largely context-dependent.
Performance associated with extinction procedures is lost when the extinguished
stimulus is presented outside the context where extinction occurred. Contextual
control is a fundamental part of the extinction process, though the contri-
bution of other mechanisms that could contribute to an observed recovery is
seldom assessed. As a primary characteristic of extinction, contextual control is
unlikely to be easily removed. An understanding of the processes responsible for
extinction, its context-specificity, and the variety of mechanisms that can change
performance associated with extinction when contextual information changes, are
essential to understanding the origin and expression of emotion.
Acknowledgements
Juan Rosas and Mark Bouton provided invaluable comments on earlier versions
of this chapter. Portions of the research presented here, and preparation of the
chapter, were supported by a Ramón y Cajal fellowship (RYC2005-447 funded
in part by the Fondo Social Europeo), and Grants IT-276-07 from the Basque
Ministry of Science, as well as both PSI2008-00412/PSIC and PSI2011-24231
from the Spanish Ministry of Science and Innovation.
References
Bouton, M. E. (2004). Context and behavioral processes in extinction. Learning and
Memory, 11, 485–494.
Bouton, M. E., García-Gutiérrez, A., Zilski, J., and Moody, E. W. (2006). Extinction
in multiple contexts does not necessarily make extinction less vulnerable to relapse.
Behaviour Research and Therapy, 44, 983–994.
Calton, J. L., Mitchell, K. G., and Schachtman, T. R. (1996). Conditioned inhibition
produced by extinction of a conditioned stimulus. Learning and Motivation, 27,
335–361.
Capaldi, E. J. (1967). A sequential hypothesis of instrumental learning. In K. W. Spence
and J. T. Spence (eds). Psychology of learning and motivation (Vol. 1, pp 67–156). New
York: Academic Press.
Dickinson, T., and Dearing, M. F. (1979) Appetitive-aversive interactions and inhib-
itory processes. In A. Dickenson and R. Boakes (eds). Mechanisms of learning and
motivation: A memorial volume to Jerzy Konorski.
Gallistel, C. R., and Gibbon, J. (2000). Time, rate, and conditioning. Psychological
Review, 107, 289–344.
Leon, S. P., Abad, J. F., and Rosas, J. M. (2011). Context-outcome associations mediate
context-switch effects in a human predictive learning task. Learning and Motivation,
42, 84–98.
McSweeney, F. K., and Swindel, S. (2002). Common processes may contribute to
extinction and habituation. The Journal of General Psychology, 129, 364–400.
Nelson, J. B., and Bouton, M. E. (2002). Extinction, inhibition, and emotional intelligence.
In L. Feldman-Barrett and P. Salovey (eds). The wisdom in feeling: Psychological
processes in emotional intelligence (pp. 60–85). New York: Guilford Press.
Nelson, J. B., Lombas, S., and León, S. P, (2011). Appetitive conditioning is not context
specific when learned during extinction. Learning & Behavior, 39, 87–94.
Nelson, J. B., Sanjuán, M. C., Vadillo-Ruiz, S., Pérez, J., and León, S. P. (2011).
Experimental renewal in human subjects. Journal of Experimental Psychology: Animal
Behavior Processes, 37, 58–70.
Rescorla, R. A. (2001). Experimental Extinction. In Mowrer, R. R., and Klein, S. B. (eds).
Handbook of Contemporary Learning Theories. Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Robbins, S. J. (1990). Mechanisms underlying spontaneous recovery in autoshaping.
Journal of experimental Psychology: Animal Behavior Processes, 16, 235–249.
Rosas, J. M., Callejas-Aguilera, J. E., Ramos-Álvarez, M. M., and Abad, M. J. F. (2006).
Revision of retrieval theory of forgetting: What does make information context-
specific? International Journal of Psychology and Psychological Therapy, 6, 147–166.
Wagner, A. R. (1981). SOP: A model of automatic memory processing in animal behavior.
In N. E. Spear and R. R. Miller (eds). Information processing in animals: Memory
mechanisms (pp. 5–47). Hillsdale, NJ: Erlbaum.
10 Generalization as a basis for
emotional change
Perceptual and non-perceptual processes
Dirk Hermans and Frank Baeyens
Changing emotions from a learning perspective

When emotions are viewed from a learning perspective, there are two issues
that deserve attention. The first is that emotions should be viewed as “behavior”,
which is defined as “any meaningful response to a meaningful stimulus”,
distinguishing it from more basic reflexes. It will be clear that this definition of
behavior encompasses a broad range of reactions, including motor responses
like running away, screaming, or fighting, psychophysiological responses such
as sweating hands, increases in respiratory frequency or heart rate, and cognitive
responses such as remembering, evaluating, or ruminating. From this perspective,
changing emotions can be viewed as changing behaviors.
A second premise is that changes in behavior (and, hence, in emotions) are
most commonly caused by learning experiences. As a matter of fact, learning
refers to a relatively enduring change in responding that results from one or more
experiences. If responding at moment t+1 differs from responding at moment t
(i.e., behavioral change), and in the interval between t and t+1 a certain event
(E) took place, one could hypothesize that this change in responding is caused
by experience E. To test for the causal nature of this interpretation, a control
situation is needed in which E is not presented between t and t+1. If no similar
changes in responding are observed under these latter conditions, this supports
the hypothesis that the behavioral change is caused by E (and not, for instance,
by other factors such as physical maturation or tiredness).
Procedures like the one outlined here have traditionally been used in human
classical conditioning research. One type of “learning experience” that has often
been investigated is the contingent presentation of two stimuli. An example is the
repeated presentation of a neutral stimulus (conditioned stimulus, CS; e.g., a tone)
that is contingently followed by an aversive stimulus (unconditioned stimulus,
US; e.g., a shock). After several trials that constitute the learning experience,
changes in responding towards the CS are typically observed, such as increased
skin conductance responding, higher levels of self-reported fear, and increased
escape or avoidance responses. These changes are not observed when this
particular learning experience is not presented (control condition). This procedure
is known as “fear conditioning” and is a royal highway to study the acquisition
68 Dirk Hermans and Frank Baeyens
of fear. Even though other affective/emotional phenomena have been investigated
from a learning perspective (e.g., anxiety, disgust and more general positive/
negative affective reactions), we will focus on changes in fear responding. An
example of a learning perspective on disgust is presented in Chapter 11 (by Peter
de Jong).
In the example described above, the contingent co-occurrence of a neutral and a
threatening event resulted in increases in fear responding to the originally neutral
stimulus. Other learning experiences have also been employed to study changes
in fear responses. Examples include CS−only presentations after acquisition
(extinction), US-only presentations with increased intensity after acquisition
(US-inflation), or the contingent presentation of the CS with a US of opposite
valence (as compared to the US used in acquisition; i.e., counterconditioning). All
these procedures have been successfully used to study preconditions and mecha-
nisms of (changes in) fear (e.g., increase, decrease, and reversal of the emotion).
Generalization as a basis for emotional change

The learning account allows one to study how certain emotional responses are
acquired, transformed, inhibited, or expressed. An intriguing phenomenon that
has been relatively underinvestigated in this context concerns the generalization
of (fear) responses. This phenomenon has not only been observed in the lab, but is
actually a core characteristic of most anxiety disorders, for which the fear condi-
tioning procedure is a laboratory equivalent (e.g., Hermans, Craske, Mineka and
Lovibond, 2006). Indeed, even though learning experiences are highly individual
and usually related to specific, identifiable stimuli or events, an obvious, but
nevertheless essential characteristic of anxiety disorders is that the focus of fear is
usually not one specific object, person or situation, but rather a “class” of objects,
persons, or situations. For instance, a dog phobic person usually does not fear one
particular dog (e.g., the neighbors’ terrier by whom he was bitten), but the whole
class of dogs, or at least a subclass of them (e.g., all terriers). The severity of
inappropriate fear responses—and corresponding maladaptive escape/avoidance
behaviors—is in large part determined by this factor of “generalization”. As a
matter of fact, if one only feared the particular and unique dog (CS) that was
involved in the bite incident (US), the impact of the “phobia” would be minimal.
It is the very fact that fear responses generalize to certain classes of stimuli or
situations that largely determines the severity of the anxiety problem.
More formally, generalization refers to those situations in which a learning
experience is presented for CS1 (e.g., cotermination with a shock US), but in
which the behavioral change—from t to t+1—is not only observed for CS1,
but also (often to a lesser extent) for other CSs (CS2, CS3, CS4 . . .). Even
though extensive forms of generalization can be the basis for the development
and expansion of anxiety problems, generalization is usually functional in that
meaningful learning experiences should not be repeated for each individual
stimulus or event. On the other hand, it is equally essential that generalization
does not proceed indiscriminately and without limits. This raises the question
Generalization as a basis for emotional change 69
of how generalization and discrimination interrelate. And, more fundamentally,
it raises the question about the core processes through which generalization
proceeds.
Generalization through perceptual and non-perceptual dimensions

One dimension that functions as a vehicle for generalization of acquired (fear)
responses is the perceptual similarity dimension. For example, in a recent fear
generalization study from our lab (Vervliet, Kindt, Vansteenwegen and Hermans,
2010), a yellow triangle was used as the CS+ and perceptually similar geometrical
figures were used as generalization stimuli (GS; blue triangle; yellow square).
Similarly, in a study by Lissek and colleagues (2008), a small circle (CS+) was
followed by an electric shock, while presentations of a large circle remained
unreinforced (CS−) (or vice versa). After this acquisition phase, a generalization
test followed during which the CS+ and CS− were presented again, as well as a
series of eight generalization stimuli (GSs) that varied in size between the CS+
and CS− (gradually smaller/larger circles). These eight circles of intermediary
size created a continuum of perceptual similarity from CS+ to CS−. A psycho-
physiological index of acquired fear demonstrated continuous increases in fear
responding as the presented GSs became more similar to the CS+.
Studies that investigate generalization across a perceptual dimension provide
a relevant laboratory model for those clinical situations in which fear gener-
alizes to perceptually similar stimuli or events (e.g., the class of spiders in
spider phobia). There are, however, many situations in which fear generalizes
across non-perceptual dimensions. An example is the obsessive client who is
afraid of causing the death of their child and avoids an ever-increasing range of
“dangerous” stimuli, such as herbicides, knives, or unstubbed cigarettes. These
stimuli do not (necessarily) share perceptual features, but rather a symbolic/
semantic relation (of being “dangerous”). In other situations, the stimuli that are
part of the generalization process do not share perceptual or symbolic/semantic
connotations, but are rather linked by a merely functional relation. For instance,
fear of an escalator (CS1), due to the experience of a panic attacks (US) while
on an escalator in a warehouse (CS3), might generalize to queuing at the counter
(CS2), by the mere fact that both escalators and checkout queues are often
encountered in the particular context of warehouses (CS1−CS3 and CS2−CS3
relations, followed by CS1−US experience).
At least two different research traditions in human/animal learning are highly
relevant for an experimental study of symbolic/semantic or functional generali-
zation, and both have recently demonstrated particular relevance for the study of
generalized fear acquisition. The first pertains to (human) “matching-to-sample”
based equivalence class learning and the emergence of other types of relational
frames (see below). The second involves both animal and human acquired equiva-
lence learning based on common outcomes or antecedents of cues (“many-to-one”
or “one-to-many” relational learning). In terms of understanding generalization
through non-perceptual dimensions, the importance of both domains of study
relies on the fact that they offer a venue for an experimental induction and study
of functionally equivalent classes of stimuli: that is, sets of completely arbitrary
stimuli, bearing no perceptual or a priori semantic connotations, that nevertheless
may support a within-class generalization (transfer of function; see below).
Non-perceptual generalization: equivalence class learning through

matching-to-sample
One procedure to create classes of functionally equivalent stimuli is through
“matching-to-sample”. In such a procedure, participants are typically asked
to match comparison stimuli to sample stimuli in conditional discrimination
training. For example, in the presence of one arbitrary stimulus A1 (sample),
choosing another stimulus B1 rather than B2 (comparisons) is reinforced.
Then, choosing C1 and not C2 (comparisons) in the presence of B1 (sample) is
reinforced. When human participants are explicitly taught to relate A1 to B1 and
B1 to C1 (versus A2 to B2, B2 to C2), they will also typically relate A1 to A1
(“reflexive” responding), B1 to A1 and C1 to B1 (“symmetrical” responding), A1
to C1 (“transitive” responding), and finally C1 to A1 (“equivalent” responding),
without any explicit training of these derived relations being necessary. When
there is positive evidence for these “emergent” relations, an “equivalence class”
containing three members (A1, B1, C1) is said to have been established (Sidman,
1994).
It has been argued that equivalence class formation may be at the origin
of symbol formation in humans (the written word “dog”, the sound of the
word “dog”, the sight of a real dog, and a picture of dog can be thought of as
members of one particular equivalence class), and that a further development
of qualitatively different types of relational responding (involving relations of,
for example, “opposition”, “difference”, “more/less than”, “before/after”, rather
than of “sameness” as implied in equivalence classes) may be at the origin
of semantics. One interesting property of equivalence classes is that a behav-
ioural function (for example, a conditioned reward or a discriminative function)
explicitly trained to one member of the class (A1), may automatically transfer
(or be transformed) to all others members belonging to the class (B1 and C1) but
not to the members of another class (A2, B2, C2).
An example of how fear transfers (generalizes) within a class of stimuli is
presented in a study by Dougher et al. (1994), who used the matching-to-sample
procedure to teach participants two four-member equivalence classes. In a first
phase, participants were presented with trials on which an arbitrary geometrical
“sample” was presented on the top half of the screen (e.g., A1). In the bottom
half, three other geometrical symbols were presented (e.g., B1, B2, B3) of which
participants were asked to select one (B1) by pressing a button. Only one of
the three possible choices was reinforced by the feedback “correct”. Through
trial and error, participants learned six conditional stimulus relations (A1-B1,
A1-C1, A1-D1, A2-B2, A2-C2 and A2-D2). The third set of stimuli (A3, B3,
C3, D3) served only as incorrect comparisons during match-to-sample trials. All
participants successfully passed subsequent tests for symmetry and equivalence,
thus demonstrating the formation of two four-member equivalence classes (A1,
B1, C1, D1; A2, B2, C2, D2). In a subsequent phase, one member of the first
class (B1; CS+) was contingently followed by an unpleasant electrocutaneous
stimulus (US). A member of the second class (B2; CS−) was never followed
by the US. In a final phase (test for transfer) skin conductance responses were
measured for all members of the two equivalence classes. Results showed that the
conditioned responses “transferred” (generalized) from B1 to the other members
of that class (C1, D1), whereas this was not the case for the second class (C2,
D2). These findings are of theoretical interest because they demonstrate gener-
alization of conditioned fear responses to stimuli that were only symbolically
associated to the CS+. From a clinical perspective, these findings suggest a route
by which stimuli can acquire fear-inducing properties, (a) without ever having
entered into a direct relation with an aversive event, and (b) based on associations
(e.g., the derived relation B1-C1) which were never actually encountered. This
might provide an explanation for the fact that patients sometimes have difficulty
retrieving discrete conditioning experiences for anxiety provoking stimuli or
events (Tierney and Bracken, 1998).
More recently, this finding was extended in a fascinating study in which
the relational frame of “smaller than” was learned using a matching-to-sample
procedure (Dougher et al., 2007). In basic terms, participants learned—for an
arbitrary set of abstract stimuli—that A is “smaller” than B, and B is “smaller”
than C. Note that this acquired semantic meaning was independent of the actual
physical appearance of these stimuli. In a second phase, B was then coupled
with a mild electric shock. In the test phase, skin conductance responses were
larger for B than for A, but interestingly, also larger for C than for B. This is
an intriguing finding, because it shows (a) that fear responses can generalize
to other stimuli within a semantic class (a stimulus set to which the relational
frame “smaller/larger than” has been applied), and (b) that the nature of this
generalization was such that fear responses were actually larger for one of the
generalization stimuli as compared to the actual CS+ (transformation of function
rather than mere “transfer”).
Given that avoidance is a core response of the fear emotion, we also want to
refer to interesting work on the transfer/transformation of avoidance responding
within equivalence classes/relational frames (e.g., Dymond et al., 2008).
Non-perceptual generalization: equivalence by mapping on to

similar outcomes/antecedents
A second way to create classes of functionally equivalent stimuli is by mapping
them on to the same outcome (or antecedent). For instance, in a “forward”
acquired equivalence task, an organism/animal learns that two or more stimuli
are equivalent in terms of being mapped on to the same outcomes (or responses)
(“many to one” relation): two antecedent stimuli, A1 and A2, are first followed by
a common outcome (B1), while another antecedent stimulus, A3, is not. When, in
a second “transfer stage”, a new behavioural function is explicitly trained to A1,
this function will generalize to A2, but not to A3. Alternatively, in a “backward”
equivalence task, an organism/animal may learn that two or more stimuli are
equivalent in terms of being preceded by the same antecedents (“one to many”
relation).
In some recent studies, it has been shown that the potential to elicit Pavlovian
conditioned fear may be exactly one of those behavioural functions that can
be transferred/generalized amongst members of a functionally equivalent set
of stimuli, established along either of these pathways. Honey and Hall (1989)
provided one of the first clear demonstrations in rodents of transfer of condi-
tioned fear as a consequence of acquired equivalence based on common outcomes
(“many-to-one”). In a similar rat study by Johns and Williams (1998), a food
stimulus, either preceded, followed, or preceded and followed each of the
auditory stimuli A and B, whereas auditory stimulus C was not food-associated.
Next, B was counterconditioned with an aversive foot-shock until it suppressed
an appetitively motivated behaviour. When A and C were tested for generalized
suppression, A (treated like B) evoked more suppression than C (treated differ-
ently than B).
Discussion
Acquired emotional responses often generalize to other stimuli/events that were
not part of the original learning experience. Generalization is often helpful. For
instance, we do not have to “experience” a negative outcome for each stimulus
or event that is somewhat dissimilar from the original learning situation. In other
cases, generalization is part of a pathogenic process such as is often seen in
anxiety disorders.
Two of the theoretically and clinically most important questions at this moment
concern the dimensions through which generalization occurs, and the factors that
impact the generalization process. In many cases both the CS+ and the generali-
zation stimuli share a perceptual relationship. Nevertheless, generalization does
not indiscriminately occur to all perceptually related stimuli in a similar fashion.
For instance, verbal instructions might draw attention to certain subsets of the
perceptual features that then guide the generalization process, while this is no
longer the case for other perceptual features of the same stimulus (Vervliet et al.,
2010). Studies like this one are informative for understanding clinical reality.
In other cases, generalization seems to occur for stimuli that hold no perceptual
relationship to one another, and recent studies have started to shine a light on
this phenomenon. Generalization has been shown to proceed within classes of
arbitrary stimuli that share a functional relationship (e.g., because they were once
related with the same outcome) or that are part of an equivalence class. Much
more research is needed in this area, but the domain holds a strong promise in
understanding emotional changes for stimuli that were not directly involved in
a “learning experience”, and for which no relationship is immediately apparent
with stimuli that were involved in such conditioning experiences.
Acknowledgements
The writing of this chapter was supported by KU Leuven Center of Excellence on
Generalization Research (GRIP*TT; PF/10/005).
References
Dougher, M. J., Augustson, E., Markham, M. R., Greenway, D. E., and Wulfert, E. (1994).
The transfer of respondent eliciting and extinction functions through stimulus equiva-
lence classes. Journal of the Experimental Analysis of Behavior, 62, 331–351.
Dougher, M. J., Hamilton, D. A., Fink, B. C., and Harrington, J. (2007). Transformation
of the discriminative and eliciting functions of generalized relational stimuli. Journal of
the Experimental Analysis of Behavior, 88, 179–197.
Dymond, S., Roche, B., Forsyth, J. P., Whelan, R. and Rhoden, J. (2008). Derived
avoidance learning: Transformation of avoidance response functions in accordance with
same and opposite relational frames. The Psychological Record, 58, 269–286.
Hermans, D., Craske, M. G., Mineka, S., and Lovibond, P. F. (2006). Extinction in human
fear conditioning. Biological Psychiatry, 60, 361–368.
Honey, R. C., and Hall, G. (1989). Acquired equivalence and distinctiveness of cues.
Journal of Experimental Psychology: Animal Behavior Processes, 15, 338–346.
Johns, K. W., and Williams, D. A. (1998). Acquired equivalence learning with antecedent
and consequent unconditioned stimuli. Journal of Experimental Psychology: Animal
Behavior Processes, 24, 3–14.
Lissek, S., Biggs, A. L., Rabin, S., Cornwell, B. R., Alvarez, R. P., Pine, D. S., and
Grillon, C. (2008). Generalization of conditioned fear-potentiated startle in humans:
Experimental validation and clinical relevance. Behaviour Research and Therapy, 46,
678–687.
Sidman, M. (1994). Equivalence relations and behavior: A research story. Boston:
Authors Cooperative.
Tierney, K. J., and Bracken, M. (1998). Stimulus equivalence and behavior therapy. In
W. T. O’Donohue (ed.). Learning and behavior therapy. Needham Heights, MA: Allyn
and Bacon.
Vervliet, B., Kindt, M., Vansteenwegen, D., and Hermans, D. (2010). Fear generalization
in humans: Impact of verbal instructions. Behaviour Research and Therapy, 48, 38–43.
11 Learning mechanisms in the
acquisition of disgust
Peter J. de Jong
University of Groningen
Although there is considerable variation in the concrete stimuli that people find
disgusting, these stimuli share several common features in that they are somehow
implicated in the risk of transmission of infectious disease. This seems to imply
that particular stimuli acquire a disgusting status more easily than others. Variation
is also substantial in people’s disgust propensity (DP). The repeated finding that
high DP runs in families (e.g., de Jong, Andrea and Muris, 1997) supports the
view that there is a systematic influence of parents on the development of disgust
in their offspring (although it may also reflect a genetic influence). Another salient
feature of disgust is that people find it usually very difficult to articulate why a
particular stimulus is disgusting; it is just a dirty animal, nasty food-item; ugly
looking stimulus, etc. (e.g., Rozin and Fallon, 1987). As a final critical feature of
disgust, physical contact with a disgusting stimulus may “contaminate” an origi-
nally neutral stimulus (e.g., a plate may become disgusting when it has been in
contact with decayed meat) (Rozin and Fallon, 1987).
In an attempt to provide a parsimonious explanation for these features of
disgust, it has been proposed that disgust might be considered as an adaptive
psychological mechanism that drives behavioral avoidance of infectious disease
(Oaten, Stevenson, and Case, 2009). The costs of infection constitute an
important selection pressure, and indeed disgust induced avoidance of potential
contaminants might represent an important behavioral disease avoidance (BDA)
mechanism that complements the physiological immune systems. This framework
seems also helpful to delineate what type of learning mechanisms might be
involved in the acquisition of disgust and will therefore be taken as the starting
point for the remainder of this contribution.
Following the BDA-framework, one would predict that in the evolutionary
history some types of stimuli, such as the smell or taste of decaying flesh or
faeces, might have acquired primary disgusting properties as a warning signal
that evolved from the association between these stimuli and previous disease
threats. In line with this, disgust is readily identifiable at birth, and particular
tastes and smells can readily elicit the typical disgust response in terms of facial
expression and avoidance tendencies. In addition to these apparently innate
reflexive responses to particular gustatory stimuli shaped by past threats, it would
be important for a disease avoidance mechanism to also be sensitive to signals of
Learning mechanisms in the acquisition of disgust 75
current threats. For optimizing survival, this new information would need to be
easily learned and relatively resistant to extinction. Three types of learning effects
seem especially relevant in this context of disgust learning: (i) social referencing,
(ii) evaluative conditioning (EC), and (iii) contagion learning. These learning
effects will be successively discussed below.
Social referencing
Social referencing refers to the process of emotional communication where the
perception of another person’s interpretation of a stimulus is used to form one’s
own appraisal of that stimulus. It can thus be considered as a specific form of
observational learning or modeling. From the proposed evolutionary perspective
it would be important that disgust responses toward disease-relevant stimuli can
be acquired early in childhood to promote the timely avoidance of hazardous
pathogens. Thus it seems reasonable to assume that parents play a critical role
in this process. Social referencing might be an important candidate to explain
how children learn from their parents that particular stimuli are disgusting. This
process seems especially powerful when it concerns novel stimuli, and when the
“model” is a significant other such as a parent. The common observation that
the originally positive reactions in young children toward faeces and other body
products readily disappear after toilet training seems a case in point to illustrate
the potentially strong influence of the parents’ responses on the acquisition of
disgust in their offspring (e.g., Rozin and Fallon, 1987). Although, it seems
intuitively plausible that the emotional communication via facial expression and
vocalization might play an important role in the child’s acquisition of disgust,
empirical evidence for this hypothesis is extremely scarce.
Perhaps the most convincing evidence for the role of parental behavior in the
acquisition of disgust comes from a recent study of Stevenson, Oaten, Case,
Repacholi, and Wagland (2010). Children from various age groups as well as
their primary caregiver were exposed to disgust-relevant behavioral approach
tests, involving various types of stimuli such as a candy in a potty, a filthy sock,
and mealworms. They were asked to evaluate the items and subsequently invited
to actually approach the items (e.g., they were asked to remove a lid of a jar
containing mealworms, invited to touch a mealworm, and asked whether they
would like one placed on their hand). First, the child participated in the tasks
with only the experimenter present. During the second stage of the study, the
parent joined the child, and was then presented with the same disgust modules
as were previously presented to the child. The performance of both the child and
the parent was filmed, and the emotional expressions as well as the approach/
avoidance behaviors were coded. In line with parents entraining disgust in young
children, parents of younger children showed stronger avoidance and more
intense expressions of disgust than parents of older children. Consistent with the
alleged parent-child transmission of disgust, the level of disgust responsivity of
the children was closely associated with the responsivity of their parent during
the second stage of the experiment.
76 Peter J. de Jong
Clearly, this pattern of findings is in line with the hypothesis that children
acquire disgust via parental transmission. However, as these findings were
correlational in nature, it requires complementary experimental research to test
the alleged causality of the relationship between the parents’ and the children’s
disgust responsivity. One approach could be to test whether intentional parental
expressions of disgust upon uncovering a novel/unknown stimulus are effective
in eliciting enhanced disgust responsivity in the child when confronted with this
stimulus during a subsequent stage (cf. Gerull and Rapee, 2002).
Evaluative conditioning
Another type of learning procedure that has been proposed to be involved in the
acquisition of disgust is known as evaluative conditioning (EC). EC refers to a
change in the valence of a stimulus that is due to the pairing of that stimulus with
another positive or negative stimulus. Thus this type of learning does not rely on
other people’s emotional (e.g., disgust) responding to particular stimuli, but depends
on the (number of) co-occurrences of an initial neutral stimulus and a stimulus with
a strong valence (e.g., primary disgust elicitors). This type of associative learning
might be especially important for people’s ability to quickly adapt to recent health
threats. That is, to quickly adapt to new health threats, it would be helpful for the
BDA mechanism if novel or initially neutral stimuli would rapidly acquire a disgust
eliciting status when these stimuli are contingently present in time and space with
the earlier acquired “primary” disgust elicitors. For example, when an unknown
stimulus smells like decaying meat, it would be adaptive if this stimulus would elicit
disgust as a mechanism to support avoidance of this potentially infectious object.
In the first study to test this conditioning pathway under controlled condi-
tions, participants were presented with 54 neutral pictorial Conditional Stimuli
(CS) (e.g., an umbrella) that were either followed by a disgusting Unconditional
Stimulus (US) picture (dirty toilet, decaying dog, or faeces), a pleasant picture,
or a neutral picture (Schienle, Stark, and Vaitl, 2001). Although the USs reliably
elicited disgust responses as indexed by self-reports and facial electromyo-
graphical (EMG) activity of the levator muscles, the 18 CS disgust pairings did
not result in enhanced disgust responding towards the CS. Thus the results
provided no straightforward evidence for the role of EC in the acquisition of
disgust. One explanation for the absence of a conditioning effect might be that
the design was too complex. In line with this, only few participants were able
to indicate post-experimentally which CS was paired with what US during the
experiment. Finally, the current neutral CS pictures might have been suboptimal.
From the BDA- framework, one would predict that especially disease-relevant
stimuli would be sensitive to acquire a disgust-evoking quality. Initially neutral
food items or body products might therefore be more powerful stimuli to
determine the relevance of conditioning in the acquisition of disgust than disease-
irrelevant (inanimate) stimuli such as an umbrella.
To test further the role of EC, a subsequent experiment used a similar (but less
complex) differential conditioning paradigm; comprising only one CS that was
never paired with a disgusting US (CS−) and one CS stimulus that was always
paired with a disgusting US (CS+) (Olatunji, Forsyth, and Cherian, 2007). The CS
stimuli were neutral verbal stimuli (cloud, prize, guest, or board). To counteract
habituation to the US there were 12 rather than three different disgusting US
stimuli. In this experiment, the disgusting USs were pictures representing bodily
mutilation, whereas the neutral “USs” depicted commonplace (inanimate) neutral
items. This study covered acquisition (12 CS+ UCS trials) as well as extinction
(8 CS+, 8 CS−) to see whether disgust would indeed not only be easily acquired,
but also difficult to be extinguished. Results clearly showed that the CS+ acquired
disgusting properties over the 12 paired trials. During the subsequent extinction
procedure, participants continued to evaluate the CS+ as more disgusting than
the CS−. Thus, although also this experiment used CSs that are probably subop-
timal, the results clearly showed that affective learning did occur. Interestingly,
the conditioning procedure also resulted in heightened fear ratings for the CS+.
Yet, unlike the disgust ratings, the fear ratings readily declined during extinction.
Together, this pattern not only confirmed that disgust can be transferred through
associative learning, but also that disgust is relatively resistant to extinction. In
line with this, a series of treatment studies focusing on individuals suffering from
dysfunctionally enhanced disgust responsivity (e.g., spider phobia, obsessive
compulsive disorder) showed that exposure in vivo is relatively ineffective in
reducing disgust (e.g., Olatunji, Wolitzky-Taylor, Willems, Lohr, and Armstrong
2009). The relative insensitivity to CS−only experiences may be functional from
the perspective that the very source of infection is often rather unclear and usually
impossible to detect without additional aids (e.g., microscope; immune tests).
Thus in the absence of explicit positive information indicating that the stimulus
is now actually safe, the CS may retain its disgusting properties (better safe than
sorry).
However, it should be noted that the conditioning study of Olatunji et al.
(2007) exclusively relied on explicit self-report measures. Thus, on the basis of
this study, it cannot be ruled out that the findings were influenced by demand.
In addition, it remained to be tested whether the conditioning effects were
restricted to subjective self-reports or could also be found at the behavioral level.
Therefore, a more recent conditioning study also included a more objective
behavioral measure of disgust (visual avoidance) in addition to self-reported
disgust (Mason and Richardson, 2010). In this differential conditioning study, one
of two neutral faces (CS+) was paired eight times with disgusting images (a man
vomiting in a toilet, faeces in a toilet bowl). To test the influence of extinction
on visual avoidance, half of the participants were tested immediately following
acquisition, whereas the other half were tested following extinction. For both
groups of participants, the disgust rating for the CS+ clearly increased from the
pre- to the post-experimental rating. Thus, the EC procedure was again effective
in conditioning disgust, whereas this acquired disgust appeared largely unaffected
by the presentation of (seven) extinction trials. Importantly, this finding was not
only evident at the self-report level, but also in the index of visual avoidance.
For both groups, viewing time for the CS+ was systematically lower than for the
78 Peter J. de Jong
concurrently presented CS− and very similar to the relative viewing time for the
disgusting US compared to the neutral “US” (umbrella and table with chairs).
Interestingly, this pattern was especially pronounced in individuals with
habitually enhanced trait DP. This finding suggests that people with enhanced DP
may not only be susceptible for relatively enduring dysfunctional disgust respon-
sivity via enhanced avoidance, but also via a relative insensitivity to extinction
of acquired disgust. In line with this, it has been shown that phobic individuals
with heightened levels of DP profit less from exposure treatment than phobic
individuals with relatively low levels of DP (de Jong et al., 1997).
The finding that the conditioned disgust responses are relatively insensitive to
extinction (i.e., presenting the CS without the US) is consistent with the more
general finding that evaluatively conditioned responses are relatively resistant
to extinction (e.g., Kerkhof, Vansteenwegen, Baeyens and Crombez, 2011). To
the extent that (dysfunctional) disgust responsivity is acquired via EC, proce-
dures aiming at modifying individuals’ disgust responding may be improved by
directly addressing the valence instead of the predictive value of the disgust-
eliciting stimulus, for example via counter-conditioning procedures. Germane to
this suggestion, a recent study showed that extinction trials were ineffective in
fully eliminating previously acquired conditioned preferences (in a picture-taste
paradigm), whereas a counterconditioning treatment did succeed in doing this.
Attesting to the robustness of this counterconditioning procedure, a follow-up
assessment revealed that this effect persisted after a seven-day delay period
(Kerkhof, et al. 2011).
Contagion learning
A final pathway to acquire disgust is via contagion learning: when a disgusting
stimulus makes physical contact with another previously neutral stimulus, this
new stimulus immediately acquires a disgust label. This type of learning can be
conceptualized as a specific form of associative learning in which the CS and US
are not only contingently present in time and space, but are also making physical
contact. Following such a procedure, the CS may elicit disgust because it acquired
signaling properties (i.e., it may signal potential contamination). From the BDA
perspective, this type of learning effect seems highly adaptive as it makes the
BDA mechanism sensitive to the spread of pathogenic contamination from
surface to surface and may thus sustain avoidance of physical contact with stimuli
that may contain pathogens. Although it is beyond dispute that this type of disgust
conditioning takes place (Rozin and Fallon, 1987), and may be highly relevant
for understanding certain psychopathologies such as OCD (Tolin, Worhunsky,
and Maltby, 2004), it has not yet been subjected to systematic research. From
a functional perspective, one would predict that contagion-based disgust would
only be a temporary phenomenon and item specific. Disgust acquired via physical
contact seems not likely to generalize to other exemplars of the same category
of items (if one part of the sandwich falls on the pavement, people will usually
continue eating the remainder without any hesitation), and is likely to dissipate
when it will be cleaned (Tolin et al., 2004). Yet, in apparent contrast with this, it
has been described that contagion-based disgust might be quite persistent (e.g.,
Rozin and Fallon, 1987). For example, a shirt that has been in contact with dog
faeces might still elicit some disgust although it may no longer show any trace
of the faeces after being thoroughly cleaned. Possibly, this may be explained by
EC. When a disgusting stimulus makes physical contact with an initial neutral
stimulus, this may not only give rise to contagion learning, but also to affective
learning (EC), which is known to be relatively resistant to extinction. Clearly,
future empirical studies are required to determine the boundary conditions of
contagion-based disgust learning and how this type of learning relates to evalu-
ative conditioning.
Recapitulation
To conclude, thus far only very few studies specifically focused on the learning
mechanisms involved in the (un)learning of disgust. Yet a series of recent studies
provided supportive evidence for at least three types of learning procedures that
may be involved in the acquisition and maintenance of disgust. The preliminary
results support the view that social referencing plays an important role in the acqui-
sition of primary disgust via parent-child transmission, whereas EC contributes
to the further elaboration of enduring disgust responsivity towards new stimuli,
and contagion learning may contribute to temporary disgust of stimuli that have
been in contact with enduring disgusting stimuli. Future research is necessary to
determine whether these learning effects reflect only differences in procedures or
also reflect different underlying mechanisms. In addition, it would be important
for future research to test the alleged causal influence of social referencing in the
acquisition of disgust in early childhood, and to test further the characteristics
and boundary conditions for both EC and contagion learning in the (un)learning
of secondary disgust. Insight into the learning laws of disgust may not only be
of theoretical relevance, but may also have important implications for health
prevention strategies and the treatment of disgust-based psychopathology (e.g.,
de Jong, van Lankveld, Elgersma, and Borg, 2010).
References
Gerull, F. C., and Rapee, R. M. (2002). Mother knows best: effects of maternal modelling
on the acquisition of fear and avoidance behaviour in toddlers. Behaviour Research and
Therapy, 40, 279–287.
de Jong, P. J., Andrea, H., and Muris, P. (1997). Spider phobia in children: Disgust and fear
before and after treatment. Behaviour Research and Therapy, 35, 559–562.
de Jong, P. J., van Lankveld, J., Elgersma, H. J. and Borg, C. (2010). Disgust and sexual
problems: Theoretical conceptualization and case illustrations. International Journal of
Cognitive Therapy, 3, 24–41.
Kerkhof, I., Vansteenwegen, D., Baeyens, F. and Hermans, D. (2011). Counterconditioning:
An effective technique for changing conditioned preferences. Experimental Psychology,
58, 31–38.
80 Peter J. de Jong
Mason, E. C., and Richardson R. (2010). Looking beyond fear: The extinction of other
emotions implicated in anxiety disorders. Journal of Anxiety Disorders, 24, 63–70.
Oaten, M., Stevenson, R. J., and Case, T. I. (2009). Disgust as a disease-avoidance
mechanism. Psychological Bulletin, 135, 303–321.
Olatunji, B. O., Forsyth, J. P., and Cherian, A. (2007). Evaluative conditioning of disgust:
A sticky form of relational learning that is resistant to extinction. Journal of Anxiety
Disorders, 21, 820–834.
Olatunji, B. O., Wolitzky-Taylor, K. B., Willems, J., Lohr, J. M., and Armstrong, T. (2009).
Differential habituation of fear and disgust during repeated exposure to threat-relevant
stimuli in contamination-based OCD: An analogue study. Journal of Anxiety Disorders,
32, 118–123.
Rozin, P. and Fallon, A. E. (1987). A perspective on disgust. Psychological Review, 94,
23–41.
Schienle, A., Stark, R., and Vaitl, D. (2001). Evaluative conditioning: A possible expla-
nation for the acquisition of disgust responses. Learning and Motivation, 32, 65–83.
Stevenson, R. J., Oaten, M. J., Case, I. C., Repacholi, B. M., and Wagland, P. (2010).
Children’s response to adult disgust elicitors: Development and acquisition.
Developmental Psychology, 46, 165–177.
Tolin, D. F., Worhunsky, P., Maltby, N. (2004). Sympathetic magic in contamination-
related OCD. Journal of Behavior Therapy and Experimental Psychiatry, 35, 193–205.
12 Preclinical analysis of developmental
transitions in the extinction of learned
fear
From infancy through adolescence to
adulthood
Bridget L. Callaghan, Stella Li, Jee H. Kim
and Rick Richardson
University of New South Wales
Animals learn predictive relationships between stimuli in their environment. For

example, learning that a stimulus predicts an aversive outcome can allow the
animal to make preparatory responses to reduce the impact of that stimulus, if
not avoid it entirely. As a result of this learning, the predictive cue will elicit fear
when subsequently encountered and this fear motivates the preparatory/avoidance
behavior in the animal. However, stimulus relationships in the environment often
change so animals need to learn about such changes as well. That is, a stimulus
that predicted an aversive outcome in the past may no longer reliably predict that
outcome. In that situation, the predictive cue should no longer elicit fear. Both of
these types of emotional learning can be modeled in the laboratory using Pavlovian
conditioning procedures. In a typical study of this type, animals are first trained
to fear an initially neutral conditioned stimulus (CS, e.g., light) by pairing it with
an aversive unconditioned stimulus (US, e.g., footshock). This association is
quickly learned and the CS begins to elicit species-specific defence responses (e.g.,
freezing). This conditioned fear, however, can be diminished through the process
of extinction, where the CS is repeatedly presented without the aversive outcome.
Studies of learned fear, and its extinction, have been shown to have important
implications for the treatment of anxiety disorders, which are amongst the most
debilitating of mental illnesses affecting our society today (Quirk and Mueller,
2008). These disorders are commonly treated with exposure-based therapy,
which is explicitly based on the process of fear extinction. As noted by McNally
(2007), exposure-based therapies are an undeniable success story in psychology.
However, they do not work for all (e.g., some patients drop out before treatment
is complete, and a substantial proportion of those who do complete treatment
relapse). McNally suggested that we have reached a “therapeutic impasse,” and
that improvements in exposure-based therapies require a better understanding of
the neurobiological mechanisms of fear inhibition.
Although we have made substantial gains in our understanding of the neural
mechanisms involved in fear extinction in the last decade (e.g., Quirk and
82 Bridget L. Callaghan, Stella Li, Jee H. Kim and Rick Richardson
Mueller, 2008), there remains much to be determined. As just one example, recent
preclinical research from our laboratory has shown that extinction is a far more
dynamic process across the lifespan than previously appreciated. Specifically,
while rats of all ages exhibit extinction of learned fear, the neural circuits and
neurotransmitters mediating that inhibition of fear differ depending upon the
animal’s age.
Extinction in the adult rat

It is widely accepted that fear extinction is not simply erasure of the original
fear memory, but instead involves the formation of a new inhibitory (CS−no US)
memory. Evidence that the original CS−US memory remains intact following
extinction comes from behavioral studies showing that the fear often returns
following extinction training. This relapse can be induced by a change in context
(renewal), time (spontaneous recovery), or application of stress (reinstatement).
Another indicator that extinction reflects new learning is that N-methyl-D-
aspartate receptors (NMDAr), which are critical for memory consolidation,
are also recruited during extinction learning. For instance, NMDAr blockade
prevents the formation of a long-term extinction memory (Quirk and Mueller,
2008). Conversely, extinction of fear is enhanced through systemic or intra-
amygdala administration of D-cycloserine (DCS), an NMDAr partial agonist
(Davis et al., 2006).
Not only are the behavioral correlates and neurotransmitter systems involved in
extinction well-documented in the adult animal, but so too is the neural circuitry.
Electrophysiological, immunohistochemical, and lesion studies have shown that
fear extinction in the adult recruits the medial prefrontal cortex (mPFC) and
amygdala. Specifically, following extinction training, it is thought that the mPFC
modulates fear expression through projections to the intercalated cells in the
amygdala. Importantly, it has been shown that a similar neural circuit mediates
extinction learning in humans (see Schiller, Levy, Niv, LeDoux and Phelps, 2008).
Qualitative differences in fear extinction during infancy

Although the neural circuitry supporting extinction is relatively well established
in adult rats, until recently it was unknown whether the same circuitry supported
extinction early in development. Given the large structural and functional changes
that take place in the developing brain, it would not be surprising that the neural
mechanisms supporting extinction would be different in infant and adult animals.
Indeed, this possibility was supported by the results of several recent behavioral
studies showing that the consequences of fear extinction were fundamentally
different in infant compared to older rats. In particular, those studies showed
that, when levels of fear acquisition were equal across ages, rats that were 24
postnatal days (P) old exhibited fear relapse following extinction (i.e., renewal
and reinstatement; just like adult rats), while younger rats (P16–17) do not (for
review see Kim and Richardson, 2010). In other words, although rats of both ages
Preclinical analysis of developmental transitions 83
exhibited equal fear conditioning, and equal within session extinction learning,
the propensity to exhibit fear relapse following extinction was dependent on age;
extinction is relapse-resistant in infant rats, but becomes relapse-prone following
weaning. Whether this developmental difference in sensitivity to relapse means
that extinction early in life is due to erasure of the original fear memory rather
than the formation of a competing CS−no US memory is yet to be conclusively
determined. Whatever the case, if a similar developmental dissociation can be
found in humans, it could lead to a major advance in the treatment of anxiety
disorders. Specifically, it is possible that extinction applied early in life will be
more effective than a similar treatment applied later in development. Regardless
of what is found in humans, these findings in the developing rat provide a novel
insight into the mechanisms mediating extinction by showing that this apparently
simple process changes in fundamental ways as the animal matures.
In addition to the behavioral differences noted above, we have also found that
extinction early in life is supported by fundamentally different neural circuitry
than is extinction later in life. For example, NMDA receptors are not involved
in extinction occurring early in life (e.g., at P17), while they are involved
when extinction occurs later in life (e.g., P24; Langton, Kim, Nicholas, and
Richardson, 2007). Specifically, consistent with findings in adult rats, P24 rats
treated with the NMDAr antagonist MK-801 exhibited poor extinction retention
compared to saline-injected rats, showing that NMDAr function is critical for
extinction learning at this age. However, injection of MK-801 (across a range of
doses) did not impair extinction retention in rats extinguished at 17 days of age.
Importantly, these findings were not due to a lack of NMDAr functioning at P16
because MK-801 administered before conditioning did impair the acquisition
of fear learning in rats this age. Thus, it appears that while NMDA receptors
are functional and are necessary for some forms of learning at P16, they are not
yet integrated into the circuitry mediating the extinction of fear. A final point to
be noted here is that such findings may have important implications clinically,
suggesting that pharmacological adjuncts that may effectively enhance exposure
therapy in adults (e.g., DCS) may not be very effective adjuncts earlier in life.
The neural structures involved in extinction also are different across devel-
opment. As discussed earlier, extinction in adult animals involves an interaction
between the mPFC and intercalated cells in the amygdala. However, in infant rats
we reported that extinction does not involve the mPFC. Specifically, extinction
training did not lead to increased neuronal activity (measured by quantifying
phosphorylated mitogen-activated protein kinase; pMAPK) in the infralimbic
(IL) cortex of the mPFC in P17 rats (see Figure 2), and temporary inactivation of
the IL, by infusion of the GABAergic agonist muscimol, did not affect extinction
retention in rats this age (Kim, Hamlin, and Richardson, 2009). In contrast, rats
extinguished at P24 exhibit heightened neuronal activity in the IL (see Figure 2),
and temporary inactivation of this structure impaired extinction retention in rats
this age (just like in adults).
While the involvement of NMDA receptors and mPFC in extinction changes
as the animal ages, the amygdala appears to be involved in extinction of fear
75
pMARK-IR neuron count
50
25
0
P17 P24 P35 P35 P70
Double
Age (days)
Figure 2 Mean difference scores of pMAPK-labelled neurons in the IL of extinguished

rats above levels seen in non-extinguished rats at P17, P24, P35 and P70.
Group ‘P35 double’ received twice the amount of extinction as all other groups.
across development. Specifically, extinction retention is impaired in both P24 and

P17 animals if the amygdala is inactivated (via infusion of the GABAa agonist
muscimol) before extinction training. Additionally, activity in the basolateral
amygdala (BLA; measured through pMAPK) is increased following extinction
training in rats of both ages (see Kim and Richardson, 2010 for review). Hence,
it has been suggested that early-life extinction recruits a very simple neural
circuit contained within the amygdala, resulting in a relapse-resistant form of
extinction learning. Across development additional neural structures and neuro-
transmitters are incorporated into the circuitry underlying extinction of learned
fear, integrating inputs from the mPFC and NMDAr activity to support a more
flexible form of extinction learning.
Quantitative differences in fear extinction during adolescence

While the transition from infancy to the juvenile stage of development involves
a qualitative change in the mechanisms underlying fear extinction, quanti-
tative differences in extinction are observed at another developmental stage.
Adolescence is a period of heightened emotional reactivity and vulnerability to
poor outcomes (e.g., suicide, anxiety, and depression). It is believed that these
effects during adolescence are related to the nonlinear developmental trajectory
of the mPFC observed in both rats and humans. Specifically, the mPFC slowly
develops to a peak in volume and synaptic density during pre-adolescent
childhood, both of which then decrease dramatically during adolescence, prior to
exhibiting a small increase and stabilization by early adulthood (for more infor-
mation, see Kim, Li, and Richardson, 2011). Considering the documented role of
the mPFC in fear extinction, at least in juvenile and adult animals, we hypoth-
esized that the dramatic changes occurring in the mPFC during adolescence may
act as a “natural lesion”, leading to impairments in extinction retention during
this period of development. In other words, we predicted that the same underlying
system would mediate extinction in the adolescent as in juvenile and adult rats,
but that this system would be relatively impaired in the adolescent.
In two studies, we found that pre-adolescent (P24), adolescent (P35), and adult
(P70) rats did not differ in terms of either fear acquisition (following three white
noise CS−shock US pairings), or in rate of within session loss of fear (i.e., the
reduction of fear observed during the CS−alone trials). However, when tested the
next day, adolescent rats showed essentially complete return of fear compared to
P24 and P70 rats (Kim et al., 2011; McCallum, Kim, and Richardson, 2010). That
is, adolescent rats failed to exhibit extinction retention over the 24-hour interval.
This is exactly the same pattern of results observed in adult rats with lesions of
the mPFC (Quirk and Mueller, 2008). Adolescent rats also failed to show signifi-
cantly elevated levels of pMAPK-immunoreactive neurons in the IL following
extinction compared to no-extinction control rats, whereas P24 and P70 rats did
(Kim et al., 2011; see Figure 2).
Interestingly, adolescent rats sucessfully exhibited long-term extinction when
NMDA neurotransmission was enhanced by a systemic post-extinction injection of
DCS (McCallum et al., 2010). Further, doubling the amount of extinction training
also facilitated long-term extinction in adolescent rats, and led to enhanced IL
activity in these rats (Kim et al., 2011; see Figure 2). Taken together, these findings
suggest that extinction processes in adolescent rats are “quantitatively” different from
pre-adolescent and adult rats, due to the attenuated post-extinction neural activity in
the mPFC. That is, the same neural mechanisms are involved in extinction during
adolescence compared to pre-adolescence and adulthood but adolescent rats require
more extinction training or neural activity to achieve a similar level of long-term
extinction to the other ages. This idea is consistent with previous neuroimaging and/
or cognitive control studies showing that adolescents can display adult-like behavior,
but they require repeated training and need to recruit prefrontal regions more than
adults (for more information, see Kim et al., 2011).
The quantitative changes in the processes mediating extinction during adoles-
cence highlight that it is not a period of distinct linear development of the PFC,
but rather is a period of non-linear “maturation”. That is, developmental changes
in cognition during adolescence are distinct from those that occur earlier in
infancy and childhood, as adolescent changes involve improvements in existing
capacities rather than the acquisition of new abilities (for more information, see
Kim et al., 2011). It is not yet clear which biological mechanisms are responsible
for the observed transition at this period of development. Nevertheless, it is
likely that gonadal hormones are involved as adolescence is a period of sexual
maturation.
Environmental alterations of the developmental transitions between

extinction systems early in life
The neural and behavioral processes underlying extinction are clearly dynamic
across development, with both qualitative and quantitative changes occurring as
the animal matures. Our understanding of these developmental transitions is far
from complete, and we do not yet have insights into the mechanisms that mediate
the various transitions in the processes underlying extinction that occur across
the course of development. However, we have recently found that the age at
which these transitions occur can be influenced by external factors. For example,
Callaghan and Richardson (2011) separated infant rats from their mother for
three hours per day across P2–14 to create a stressful rearing environment. They
found that while non-stressed P17 rats exhibited the typical relapse-resistant form
of extinction learning, stressed P17 rats displayed adult-like relapse phenomena
(i.e., renewal and reinstatement of extinguished fear). Hence, stress accelerated
the transition into the adult-like, relapse-prone extinction system. Similarly,
renewal of extinguished fear was evident in P17 rats following early-life (P1–5)
injections of the neurotrophic factor fibroblast growth factor-2 (FGF2; Graham
and Richardson, 2010), perhaps suggesting that FGF2, and indeed maternal-
separation, was able to speed up the development of the brain resulting in more
“adult-like” extinction behavior. These findings raise the interesting possibility
that the neural circuitry of extinction is also reactive to external and environ-
mental manipulations. Specifically, we predict that the circuitry underlying
extinction (in terms of structures and neurotransmitters) in infancy becomes more
“adult-like” following early life stress or FGF2 administration.
Conclusions
Infancy and adolescence are unique periods of development, and recent research
has shown that the process of extinction is different at each of these stages
compared to adulthood. Early in life, extinction does not involve the mPFC
or NMDA receptors, and is resistant to relapse. As the animal develops from
infancy to preadolescence, additional neural structures and neurotransmitter
systems become integrated into the extinction circuit resulting in a fundamental
change in the process of extinction. Then, the waxing and waning of cortical
structures which occur during adolescence lead to impaired extinction retention.
However, the effectiveness of extinction during adolescence can be enhanced if
NMDAr activity is increased or if the extinction training procedure more actively
recruits the mPFC. Finally, environmental factors appear to influence when these
transitions in extinction learning occur. Together, these developmental changes
in extinction learning present a dynamic picture of emotion regulation across
the lifespan. Until recently, little attention had been given to extinction across
development and most researchers would have had a relatively static conception
of extinction across the lifespan. The challenge now is to integrate these devel-
opmental findings into the wider human and animal literature, and to explore
the possibility that extinction in adulthood might also be more plastic than once
thought. The malleability of extinction across the lifespan raises the exciting
possibility that we may be able to enhance the efficacy of anxiety treatments by
tailoring such treatments to an individual’s developmental stage and early-life
experience.
Acknowledgements
Our work described in this chapter was supported by grants from the Australian
Research Council (to RR). BLC was supported by an Australian Postgraduate
Award during preparation of this chapter.
References
Callaghan, B. L., and Richardson, R. (2011). Maternal separation results in early emergence
of adult-life fear and extinction learning in infant rats. Behavioral Neuroscience, 125,
20–28.
Davis, M., Ressler, K., Rothbaum, B. O., and Richardson, R. (2006). Effects of
D-cycloserine on extinction: Translation from preclinical to clinical work. Biological
Psychiatry, 60, 369–375.
Graham, B. M., and Richardson, R. (2010). Early life exposure to fibroblast growth
factor-2 facilitates context-dependent long-term memory in developing rats. Behavioral
Neuroscience, 124, 337–345.
Kim, J. H., Hamlin, A. S., and Richardson, R. (2009). Fear extinction across development:
The involvement of the medial prefrontal cortex as assessed by temporary inactivation
and immunohistochemistry. The Journal of Neuroscience, 29, 10802–10808.
Kim, J. H., Li, S., and Richardson, R. (2011). Immunohistochemical analyses of long-term
extinction of conditioned fear in adolescent rats. Cerebral Cortex, 21, 530–538.
Kim, J. H., and Richardson, R. (2010). New findings on extinction of conditioned fear
early in development: Theoretical and clinical implications. Biological Psychiatry, 67,
297–303.
Langton, J. M., Kim, J. H., Nicholas, J., and Richardson, R. (2007). The effect of NMDA
receptor antagonist MK-801 on the acquisition and extinction of learned fear in the
developing rat. Learning & Memory, 14, 665–668.
McCallum, J., Kim, J. H., and Richardson, R. (2010). Impaired extinction retention in
adolescent rats: Effects of D-cycloserine. Neuropsychopharmacology, 35, 2134–2142.
McNally, R. J. (2007). Mechanisms of exposure therapy: How neuroscience can improve
psychological treatments for anxiety disorders. Clinical Psychology Review, 27,
750–759.
Quirk, G. J., and Mueller, D. (2008). Neural mechanisms of extinction learning and
retrieval. Neuropsychopharmacology, 33, 56–72.
Schiller, D., Levy, I., Niv, Y., LeDoux, J. E., and Phelps, E. A. (2008). From fear to
safety and back: Reversal of fear in the human brain. The Journal of Neuroscience, 28,
11517–11525.
Part 3
Social-Cultural Perspective
13 Can socially sharing emotions change
emotions?
Bernard Rimé
University of Louvain at Louvain-la-Neuve
Immediately after an emotion, the episode recurs in the subject’s working

memory thus increasing thoughts of this experience. As time passes, related
emotional memories become less frequent and their impact on current adaptation
becomes progressively negligible. At the final stage of evolution, the memory of
an emotional episode becomes “dormant”. Recall will occur only if appropriate
cues are met in the current context. More intense emotions generate slower
extinction slopes. Emotional memories can also fail to reach a dormant stage and
are then said to be “unrecovered”: the subject still “carries on” the past episode
and does not feel “in peace” with it (Rimé, 2009).
The immediate aftermath of an emotional experience is also characterized
by the social sharing process. When individuals experience an emotion, they
tell about it in the next hours and days, they do so several times, with several
successive targets within their social network. The social sharing of emotion
occurs for 80–95 per cent of emotional episodes. More intense emotional experi-
ences are shared more repetitively, with a larger number of targets, and for a
longer period (for a review, Rimé, 2009). Collective emotional events entail
similar effects. To illustrate, both in New York in September 2001 and in Madrid
in March 2004, more than 80 per cent of respondents shared with others what
they felt about the terrorist attacks (e.g., Rimé, Paez, Basabe, and Martinez,
2010). Pennebaker and Harber (1993) delineated three stages in the psychosocial
responses to a collective emotional event. First, intense emotional reactions and
social manifestations (e.g., social sharing of emotion, emotional communion,
solidarity, mutual help) characterize an emergency stage which lasts for about
one month. Next, a plateau occurs for another month. Mental rumination is
maintained at high levels, but social sharing vanishes. Finally, some two months
after the event, an adaptation stage takes place. Manifestations of solidarity
decline, spontaneous bonding decreases, rumination diminishes to a low level.
Data collected following March 11 terrorist attacks in Madrid clearly supported
this model (Rimé et al., 2010).
The question that arises is whether the decline of the emotional impact is a
result of the social sharing of the emotional episode. In other words, is recovery
favored by the extent to which an episode is talked out? We thus undertook
studies in which we examined the level of emotional recovery of an emotional
92 Bernard Rimé
episode as a function of the extent to which this episode had been socially
shared. Emotional recovery, or the decline of the emotional impact of an episode,
was defined as the difference between the intensity of the emotion elicited by
the episode when it happened and the intensity of emotion it still elicits. Some
of the studies compared emotional episodes which were shared and episodes
which had been kept secret (Finkenauer and Rimé, 1998). They showed that
when re-accessed, nonshared memories were no more emotionally arousing
than shared ones. In other studies, participants were followed up after a specific
emotional event (Rimé, Finkenauer, Luminet, Zech, and Philippot, 1998). Both
the extent to which they had shared the episode and their degree of emotional
recovery were assessed in this period. None of these studies yielded positive
correlations between those two variables. Finally, in several studies, participants
recalled a negative emotional episode and then shared it with an experimenter
(Zech and Rimé, 2005). In experimental conditions, they emphasized either the
factual aspects of the episode, or their related feelings. In a control group, they
talked about a nonemotional topic. The residual emotional impact of the shared
event was assessed first immediately after, then again some days later, and finally,
in some studies, two months later. None of the experimental studies yielded
significant effects of sharing type on indices of emotional impact. Altogether, the
findings failed to support the view that emotional sharing contributes to emotional
recovery.
These various null results fit other observations, such as those collected from
psychological debriefing. In this group intervention technique, which is often
implemented immediately after disasters, victims are asked to share with other
victims their emotional experience with a view to preventing stress disorders.
However, meta-analytic reviews consistently concluded negatively for the efficacy
of this technique in reducing trauma-related symptoms (e.g., Van Emmerik,
Kamphuis, Hulsbosch, and Emmelkamp, 2002). Adverse effects were even found.
Empirical studies of collective emotional expression also led to negative conclu-
sions. Many data concur to conclude that participation does not reduce and can
even reinforce negative affect. Weiss and Richards (1997) followed-up bereaved
partners of men who died from AIDS. They assessed whether these participants
did or did not take part in mourning rituals. Twelve months after the death of their
partner, no significant relations were observed between participation and indicators
of emotional recovery related to depression, grief symptoms, or mental rumination.
In a retrospective study of survivors of the Guatemalan genocide, Martin-Beristain,
Paez, and Gonzalez (2000) found that the extent of participation in mourning rituals
was actually associated with higher levels of negative emotions and uncorrelated
with emotional recovery. In post-genocide Rwanda, participation in transitional
justice rituals was found to enhance rather than decrease negative emotions both
among the victims and the accused (e.g., Rimé, Kanyangara, Yzerbyt, and Paez,
2011). Data collected one, three and eight weeks after the terrorist act perpetrated in
Madrid in 2004 revealed that higher levels of sharing during the first week predicted
higher event-related emotional arousal and mental rumination three weeks later,
and again eight weeks later (Rimé et al., 2010).
Can socially sharing emotions change emotions? 93
What are the reasons why talking out and sharing an emotional experience
fail to reduce the impact of this experience? Answering this question requires us
to look at the constituents of an emotional impact. Emotions are triggered when
meanings elicited by supervening events conflict with the person’s systems of
expectations (e.g., Carver and Scheier, 2002). A complex network of the episode
is then formed in long-term memory with cognitions of three types: (1) the initial
appraisal, or meaning, of the encoded emotional situation, (2) the representation
of the goals that were blocked in this situation, and (3) the representations (i.e.,
expectations, schemas, models, world views) that were disconfirmed in the
situation. Such cognitions have the power to sustain the signaling function of
emotions: the network then recurrently re-accesses the working memory and
re-captures focal attention. For getting rid of this emotional impact of a past
episode, and thus for achieving emotional recovery, turning off each of these
various memory-sustaining cognitions is necessary. This requires achieving
complex cognitive tasks involving (1) assimilation of the episode through
reframing or reappraisal of the event, (2) abandonment of the frustrated goals and
reorganization of one’s hierarchy of motives, and (3) accommodation of models
and schemas.
Can the social sharing of emotion contribute to these cognitive tasks? Several
arguments lead us to answer this question negatively. Most of the spontaneous
sharing develops immediately after the episode. In 60 per cent of cases, it is initiated
on the day the episode happened, and it vanishes after a day or two. In this early
period, people are generally not open to cognitive changes. Rather, they concen-
trate on the unattained goal with invigoration and repetitive efforts (e.g., Martin
and Tesser, 1989). They refuse to abandon their frustrated goals, do not consider
modifying their hierarchy of motives, stick to their existing schemas, do not want
to change their representations, stand by their initial appraisal of the emotional
situation, and do not feel ready to reframe it nor to change their perspective. In
addition, their listeners are not likely to stimulate the cognitive work. Simplistic
interventions usually prevail among nonvictims (e.g., Burleson, 1985). Thus, when
more intense emotional experiences are shared, listeners’ responses become less
verbal and increasingly nonverbal (Christophe and Rimé, 1997). In sum, sponta-
neous emotional sharing is unlikely to fuel the cognitive route leading to emotional
recovery. Of course, nothing precludes listeners from providing responses proper
to stimulate the sharing cognitive processing. But the investigation of spontaneous
sharing situations suggests that this is far from being common. Ample opportunity
is then left to professional intervention whose work is precisely to focus their clients
on the cognitive processing of emotional episodes. Data showed that when this is
done, social sharing can bring emotional recovery (Nils and Rimé, 2012).
Should it then be concluded that the emotional expression, as it naturally
develops in social situations, is simply a vain process? An affirmative response
to this question would be absurd considering the spectacular urge with which
individuals and communities engage in social sharing as soon as they undergo an
emotional episode. Hence it is important to consider the other face of studies that
failed evidencing recovery effects of emotional sharing. Thus, in experimental
94 Bernard Rimé
studies conducted by Zech and Rimé (2005), compared to participants in control
conditions, those who had shared an emotional episode rated the sharing session
as more emotionally relieving, more cognitively helpful, and more interpersonally
beneficial. Similarly, in spite of their inconclusive results regarding emotional
recovery, psychological debriefing procedures commonly yield clear benefits
among participants who generally express gratefulness for the support, validation,
and understanding they found. Abundant data in the same direction resulted from
the study of collective emotional sharing. In Weiss and Richards’ (1997) study,
bereaved partners who took part in mourning rituals showed a superior quality
of social functioning on a variety of indicators compared to those who had not.
In the Guatemala study conducted by Martin-Beristain et al. (2000), survivors
who participated in mourning rituals scored higher than those who did not, on
various indicators of good social functioning. They evidenced higher scores on
reconstruction of social support and on altruistic behaviour. They also had lower
scores for helplessness and disengagement. In the post-genocide Rwanda study,
participation in transitional justice rituals enhanced social integration by reducing
perceived outgroup homogeneity, decreasing ingroup self-categorization, and
increasing positive stereotypes among both victim and prisoner participants
(Kanyangara et al., 2007; Rimé et al., 2011). After the terrorist attacks in
Spain, participation in protest demonstrations was predictive of positive affect,
of received social support, of social integration and well-being, as well as of
posttraumatic growth, when assessed in later weeks (Rimé et al., 2010).
Thus, despite its lack of effect on emotional recovery, there is abundant
observation of positive effects resulting from the social sharing of emotional
experiences. Where do these effects come from? A typical interpersonal dynamic
develops in sharing situations (Christophe and Rimé, 1997). Targets evidenced a
marked interest for shared emotional episodes, their emotions varied in intensity
as a function of the intensity of the shared episode, and they reacted to higher
intensity episodes by reducing their verbal responses and by enhancing their
nonverbal ones (body contact, taking into the arms, or kissing). Thus, Person
A who experienced an emotion shares it with B who, by manifesting a strong
interest, stimulates sharing so that A expresses emotions more and more. This
arouses emotions in B and this reciprocal stimulation ends up in an emotional
communion. Empathy leads B to help and support A which enhances B’s liking
for A. The latter being the focus of B’s attention, interest, empathy, and support,
also experiences enhanced liking for A. In sum, sharing emotion has the potential
to bring interacting persons closer to one another. As sharing is most predominant
among intimates (Rimé et al., 1998), it is instrumental in maintaining, refreshing,
and strengthening important social bonds.
Given that emotional sharing arouses emotion and given that emotion elicits
social sharing, those who were exposed to the sharing of an emotion thereafter
strongly incline in sharing what they heard with third persons. In other words,
a process of secondary social sharing develops (Christophe and Rimé, 1997).
Curci and Bellelli (2004) demonstrated that three-quarters of emotional episodes
that are personally confided to someone are then shared by this target with other
Can socially sharing emotions change emotions? 95
people. Further, as targets of secondary sharing also experience emotion when
listening, they also incline to tertiary sharing. Episodes heard in a secondary
sharing are indeed shared again with one new listener for one third of participants,
and with several new listeners for another third (Rimé and Christophe, 1997).
Emotional episodes thus propagate very easily across social networks. The fact
that such a process actually occurs in real life was nicely confirmed in a field
observation of 33 college students (Harber and Cohen, 2005). Their emotional
responses to a visit of a morgue predicted how many people they told (primary
sharing), how many people their friends told (secondary sharing), and how many
people their friends’ friends told (tertiary sharing). Within ten days, nearly 900
people heard about this event through these cascading levels of social sharing.
Social consequences of emotions can be enhanced dramatically when events
strike collectively (e.g., collective loss, victory, defeat, success, failure, catas-
trophe, accident, common threat, etc.). In private emotional experiences, sharing
originates in one single source. It thus propagates essentially in an exocentric
direction, and the propagation wave extinguishes as the social distance from
the original source increases. By contrast, in collective emotional events, there
are as many sharing sources as there are members in the concerned community.
Consequently, the sharing diffuses in as many directions and thus generates
innumerable propagation waves. Within this process, each moment of social
sharing reactivates emotions in each of the involved interactants, thus reloading
the propagation flow and further feeding a chain reaction.
From the above observations, it appears that in social sharing situations,
individuals’ consciousnesses echo one another. The shared episode and the
expression of associated emotions by a member of the social group have the
power to vividly elicit analogous feelings in people around them, so that a
reciprocal stimulation of emotion follows. Such a circular process is particularly
propitious to eliciting a state of emotional communion among participants. As
a result, the salience of their self is lowered, and elements of their common
identity are enhanced. They thus end up experiencing unity and similarity. It was
abundantly demonstrated that mere self-disclosure already brings up such effects,
enhancing reciprocal liking, and bringing interactants closer to one another
(Collins and Miller, 1994). Such a process is proper to enhance participants
positive affects, with particular consequences in terms of feelings of solidarity
and prosocial orientation. In collective situations, the social sharing of emotion
can thus easily boost participants’ feelings of group belonging and of social
integration. By the same token, shared beliefs and collective representations are
brought to the foreground, thus consolidating participants’ faith in their cultural
beliefs and confidence in collective action. As a consequence, participants will
be able to return to their individual life endowed with feelings of strength, with
enhanced trust in life, and with feelings of self-confidence. Such a rationale was
already proposed by Durkheim (1912) a century ago. It still offers a very likely
model of the way through which, without bringing emotional recovery, sponta-
neous social sharing of emotion can buffer the temporary destabilization that any
negative emotional experience entails.
96 Bernard Rimé
References
Burleson, B. R. (1985). The production of comforting messages: Social-cognitive founda-
tions. Journal of language and social psychology, 4, 253–312.
Carver, C. S., and Scheier, M. F. (2001). On the self-regulation of behaviour. Cambridge:
Cambridge University Press.
Christophe, V. and Rimé, B. (1997). Exposure to the social sharing of emotion: Emotional
impact, listener responses and the secondary social sharing. European Journal of Social
Collins, N. L., and Miller, L. C. (1994). Self-Disclosure and liking: A meta-analytic
review. Psychological Bulletin, 116, 457–475.
Curci, A., and Bellelli, G. (2004). Cognitive and social consequences of exposure to
emotional narratives: Two studies on secondary social sharing of emotions. Cognition
and Emotion, 18, 881–900.
Durkheim, E. (1912). Les formes élémentaires de la vie religieuse [The Elementary Forms
of Religious Life]. Paris: Alcan.
Finkenauer, C. and Rimé, B. (1998). Socially shared emotional experiences vs. emotional
experiences kept secret: Differential characteristics and consequences. Journal of Social
and Clinical Psychology, 17, 295–318.
Harber, K. D. and Cohen, D. J. (2005). The emotional broadcaster theory of social sharing.
Journal of Language and Social Psychology, 24, 382–400.
Martin, L. L., and Tesser A. (1989). Toward a motivational and structural theory of
ruminative thought. In J. S. Uleman and J. A. Bargh (Eds), Unintended Thought
(pp. 306–326). New York: Guilford.
Nils, F. and Rimé, B. (2012). Beyond the myth of venting: Social sharing modes determine
the benefits of emotional disclosure. European Journal of Social Psychology. 42,
672–681.
Pennebaker, J. W. and Harber, K. (1993). A social stage model of collective coping: The
Loma Prieta earthquake and Persian Gulf war. Journal of Social Issues, 49, 125–145.
Rimé, B. (2009). Emotion elicits the social sharing of emotion: Theory and empirical
review. Emotion Review, 1, 60–85.
Rimé, B., Finkenauer, C., Luminet, O., Zech, E., and Philippot, P. (1998). Social sharing
of emotion: New evidence and new questions. In W. Stroebe and M. Hewstone (eds.),
European review of social psychology (Vol. 9, pp. 145–189). Chichester, UK: John
Wiley & Sons Ltd.
Rimé, B., Kanyangara, P., Yzerbyt, V. and Paez, D. (2011). The impact of Gacaca tribunals
in Rwanda: Psychosocial effects of participation in a Truth and Reconciliation process
after a genocide. European Journal of Social Psychology, June 5, 2011.
Rimé, B., Paez, D., Basabe, N., and Martinez, F. (2010). Social Sharing of Emotion,
Posttraumatic Growth and Emotional Climate: Follow-up of Spanish citizen’s response
to the collective trauma of March 11th terrorist attacks in Madrid. European Journal of
Social Psychology, 40, 1029–1045.
Van Emmerik, A. A., Kamphuis, J. H., Hulsbosch, A. M., and Emmelkamp, P. M. (2002).
Single session debriefing after psychological trauma: a metaanalysis. The Lancet, 360,
766–771.
Zech, E. and Rimé, B. (2005) Is it talking about an emotional experience helpful? Effects
on emotional recovery and perceived benefits. Clinical Psychology and Psychotherapy,
12, 270–287.
14 From group-based appraisals to
group-based emotions
The role of communication and social
sharing
Vincent Yzerbyt
University of Louvain at Louvain-la-Neuve
Toon Kuppens
Cardiff University
Right after the infamous September 11, 2001 terrorist attacks, some political
leaders declared that “This enemy attacked not just our people, but all freedom-
loving people everywhere in the world. . . . The freedom-loving nations of the
world stand by our side” (G. Bush) or that “We are all Americans” (T. Blair).
The work presented in this chapter shows that statements such as these likely
(re)define the social landscape with consequences that are far from trivial. This
occurs because the salience of social identity makes group members appraise the
world from the perspective of the group rather than the individual, which results
in emotions felt on behalf of the group instead of the person. Across four studies,
we focus on these so-called group-based emotions, as well as other reactions
such as group-based appraisals and group identification. Specifically, we argue
that communication among ingroup members can make group identity salient,
which in turn shapes group-based emotions.
In a seminal chapter, Smith (1993) combined two lines of work to account
for the complexity of emotional reactions in intergroup contexts (Yzerbyt and
Demoulin, 2010). His first source, appraisal theories of emotions (Scherer,
Schorr, and Johnstone, 2001), suggests that the way individuals react to events
is predicted by their emotional reactions, themselves resulting from a cognitive
process of appraisal. Because appraisal theories of emotion deal with idiosyn-
cratic reactions of individuals and do not apply to intergroup reactions, Smith
called upon a second perspective, Self-Categorization Theory (SCT; Turner,
Hogg, Oakes, Reicher, and Wetherell, 1987). SCT proposes that, when people
find themselves in an intergroup context, they shift psychologically from an
individual to a social level of identity. Combining these frameworks, inter-
group emotion theory (IET; Smith, 1993) holds that, to the extent that their
social identity is salient, people appraise surrounding events not so much with
regard to their own personal concerns, but rather with respect to those of their
group.
98 Vincent Yzerbyt and Toon Kuppens
Our research program extends Smith’s insight by putting a much stronger
emphasis on the self-categorization mechanism of IET (Yzerbyt and Kuppens,
2009). Because people belong to several groups, and each of these groups can be
salient at a particular moment, we focus on the specific way people categorize
themselves in a group. To illustrate, one of our early experiments (Gordijn,
Wigboldus and Yzerbyt, 2001) examined the emotional reactions of students from
the University of Amsterdam as they learned about students from the University
of Leiden suffering an unfair decision imposed by their professors. We varied
participants’ salient group membership so they would categorize themselves
either in the same or in a different category as the victims. Specifically, we
informed some participants that the study examined the reactions of students and
professors, thereby stressing the joint membership of participants and victims in
the student category, or focused on the reactions of students from different univer-
sities, thus stressing the different categories between participants and victims. As
predicted, participants categorized in the same group as the victims reported more
anger than those categorized in a different group.
Interestingly enough, studies on group-based emotions always asked partici-
pants to report their emotional reactions in isolation (Yzerbyt and Demoulin,
2010). This procedure rests on SCT’s assumption that, in an intergroup compar-
ative context, people depersonalize and function as members of a social group.
Interestingly, however, the social landscape with which people are confronted
often emerges as a spontaneous by-product of their interactions. Many reactions
have at one point or another been communicated to and discussed with others.
Indeed, about two thirds of informal conversations are about social topics
(Dunbar, Marriott, and Duncan, 1997) and emotional topics in particular are
subject to social sharing (Rimé, 2009). One may thus wonder whether group-
based emotions could emerge from social communication, even in the absence
of explicit reminders of social identity. There are several reasons to contemplate
such a possibility. First, communication concerning emotionally relevant topics
forges interpersonal bonds (Peters and Kashima, 2007) and increases group
cohesion (Espitalier, Tcherkassof, and Delmas, 2003). Second, communicating
about emotion-laden events leads to a shared perspective because of emotion
contagion and social appraisals (Manstead and Fischer, 2001), and thus increases
group homogeneity. It seems therefore plausible that social sharing leads to a
more group-based perspective (or social identity salience). We propose that this
group-based point of view manifests itself in group-based appraisals and, in turn,
in group-based emotions.
Several research efforts (for a recent illustration, see L. G. E. Smith and
Postmes, 2011) already investigated the effect of social interactions on group-
based cognition (especially stereotypes). The evidence suggests that social
communication affects the content of people’s views about outgroups, although
these effects are limited to interactions that occur in an intergroup context.
Building upon these studies, we hypothesized that group interaction and commu-
nication should foster the emergence of group-based emotions provided the
intergroup context is salient during the interaction. Moreover, we conjectured
From group-based appraisals to group-based emotions 99
that such group-based emotions would rest on the emergence of group-based,
as opposed to individual, appraisals of the situation. In this chapter, we briefly
review four studies conducted to test these ideas.
Communication and sharing breeds group-based emotions

In two experiments (Yzerbyt, Kuppens, and Mathieu, 2011), we checked whether
the mere discussion between people about a group-relevant issue could trigger
group-based emotions. Whenever group members join to focus on some event
that is at the heart of the social definition of the setting, they likely start reacting
as group members, much like when they are induced to think of themselves as
group members before reporting their reactions in isolation.
In a first experiment (Yzerbyt et al., 2011, Experiment 1), senior high-school
students learned that they would take part in a study about some decisions
regarding students’ access to universities in Belgium, and that they would
discuss with another student before answering questions about this topic. Before
proceeding, some participants were assigned to the group member condition: they
read that the study was about comparing their reactions as future college students
to those of politicians. Other participants were assigned to one of two individual
conditions: they learned that the study concerned their reactions as individuals.
After all participants had conveyed their experience with college administration
and their personal choice of a major, they individually read a (fake) article stating
that the French-speaking authorities wanted to implement a Dutch exam before
acceptance to college in order to reduce the number of non-Belgian students in
Belgian French-speaking universities. Student representatives were claiming the
unfairness of this sudden decision because of its dramatic consequences for foreign
students (i.e., few would succeed) and called for mobilization. Participants in the
group member condition and half of participants in the individual condition were
invited to discuss the content of the article with another student. The remaining half
of the individual condition participants were asked to discuss their experience with
college administration and their subject choice. Although subject choice in higher
education concerns all students, it is not a topic that affects students as a group.
We expected the discussion to facilitate the emergence of group-based
emotions only if it concerned the group-relevant event. That is, participants
should report more anger (the emotion most relevant to our scenario) when
the discussion concerned the controversial policy than when another topic was
discussed. Because the discussion should make partners endorse the same group
perspective about the policy, we did not expect an impact of whether partici-
pants were led to see themselves as group members or individuals before this
discussion. Importantly, we also predicted a different appraisal of the (un)fairness
of the policy depending on whether or not the discussion concerned the policy. As
expected, participants perceived more injustice and, in turn, were angrier when
they had talked about the group-relevant topic rather than the other topic. There
was no effect of whether participants’ group identity (as future college students)
or personal identity had been made salient at the beginning of the study.
To be sure, one would expect participants categorizing themselves as group
members to approach the situation in terms of their group concerns even before the
interaction. In contrast, for participants first categorizing themselves as individuals,
a social take on the event would emerge only after the discussion. We tested this
hypothesis in a follow-up experiment (Yzerbyt et al., 2011, Experiment 2) using
the same procedure albeit with one important exception: after reading the instruc-
tions and the fake article, but before the discussion, participants answered questions
pertaining to their emotional reactions and to the emotional reactions that they
attributed to the other future college students. Again, instructions asked partici-
pants to answer as future college students in the group member condition and as
individuals in the individual conditions. We also checked whether we replicated the
pattern of Experiment 1 by asking these questions anew after the discussion.
As predicted, even before the discussion, participants in the group member
condition reported more anger for themselves and attributed more anger to other
future college students than participants in the two individual conditions. This
confirms that a group-based interpretation reading of the situation was already
operating among these participants even without the interaction and corroborates
earlier work (Yzerbyt and Kuppens, 2009). Importantly, the discussion had the
same impact as in Experiment 1: only those participants discussing the actual
content of the article expressed more anger and attributed more anger to other
future college students. In sum, explicitly emphasizing the social identity (the
group member condition before the discussion in Experiment 2) or letting people
discuss a certain group-relevant issue (the individual conditions where partici-
pants discussed the relevant social event in Experiments 1 and 2) had similar
consequences on group-based emotions.
Several additional findings confirm the crucial role of social identity salience
in accounting for the impact of the group discussion. First, in Experiment 1,
participants found the decision more unjust in the relevant than in the irrelevant
discussion condition. Confirming that perceived injustice is an important appraisal
for anger, the effect of the manipulation (relevant vs irrelevant discussion)
on anger was mediated by the appraisal of injustice. Participants thus shared
the victims’ cognitive perspective. Because participants were not themselves
affected by the policy, we are certain that this appraisal of injustice is a group-
based appraisal; that is, a consequence of viewing the world through a group
lens. Secondly, in Experiment 2, we also measured participants’ identification
with future college students. Identification was stronger after the group-relevant
than after the irrelevant discussion, showing that the group of future college
students indeed became more important to our participants. Finally, participants
not only reported feeling angrier after discussing the group-relevant topic than
after discussing an irrelevant topic, but they also thought that other future college
students would feel angrier. Together, these three pieces of evidence suggest that
the impact of communicating about the relevant social event on the emergence of
group-based emotions is due to participants taking a group perspective.
More evidence for group-based appraisals.
If the above efforts underscore the role of social identity salience and group-
appraisals in the surfacing of group-based emotions, two additional experiments
differentiate even better between individual and group-based appraisal. This
time, we relied on a thought-listing procedure to measure people’s evaluation of
the situation because we wanted to know people’s spontaneous thoughts instead
of forcing responses on pre-formulated items. We used this open-ended measure
both in a first experiment where we explicitly manipulated social identity,
and in a second where this was done more subtly by way of a discussion on a
group-relevant event. Our hypothesis was that group-based appraisals (based on
the thought-listing content) would mediate the relation between social identity
salience and group-based emotions.
In our third experiment (Kuppens, Yzerbyt, Dandache, Fischer, and van der
Schalk, 2011, Experiment 1), we presented Belgian university students with a
fake article discussing an inequitable decision of the rector of another Belgian
university: English was allegedly imposed as the sole teaching language in
master programs. The students of this university were described as opposing this
decision and as planning to act against it. Again, the decision did in no way affect
our students, but only those of that other university.
As before, we manipulated the social identity of our participants. We told
participants in the “student” condition that we were interested in the opinion of
students and professors, and they also completed a series of questions tapping
their identification with the group of students. Participants in the “control”
condition learned that we investigated their opinion as unique individuals, and
they answered items measuring how they saw themselves as unique individuals.
After reading the article, participants were given four minutes to list every
thought that had crossed their mind while reading. All thoughts relevant to the
newspaper article were coded according to whether they expressed a favorable vs.
unfavorable opinion toward the rector’s decision and whether they mentioned the
word “student”. Participants also indicated the extent to which they experienced
anger, sadness, happiness, and fear in response to the article. We predicted that
participants in the student identity condition would list more thoughts related to
their student identity and that they would also feel emotions (primarily anger) on
behalf of the other students who were the victims of the unfair decision. In fact,
we expected the group-based character of the thoughts to mediate the effect of
social identity salience on group-based emotions.
Consistent with previous studies, making the student identity salient generated
more anger and less sadness. Because these reactions are rooted in the student
social identity and differ from those of control participants, they reflect group-
based emotions. Our unobtrusive measure of participants’ appraisals of the event
also revealed that those in the student identity condition had more thoughts that
mentioned students and simultaneously expressed an unfavorable opinion about
the rector’s decision. The proportion of such thoughts was related to anger and
sadness, and partially mediated the effect of social identity salience on anger and
sadness. In contrast, thoughts that were unfavorable but did not mention students
were in no way related to anger or sadness. That is, only those unfavorable
thoughts mentioning the relevant ingroup membership independently predicted
anger and sadness. The unfair decision had to be seen in an intergroup context in
which participants belong to the same group as the victims in order to elicit anger.
This pattern provides a powerful demonstration of how group-based appraisals
affect emotions.
A fourth experiment (Kuppens et al., 2011, Experiment 2) combined the
thought-listing measure with the discussion paradigm used in Experiments 1
and 2. The specific goal of the experiment was to check whether the content of
the group discussion would be linked to the emotions. We used a thought-listing
procedure to assess each individual’s account of the discussion. This time, we
also measured a simple yet noteworthy behavioral signature, namely whether
participants would give us their personal email address so as to be kept informed
about the issue.
Our predictions were borne out. The discussion on the relevant topic led to
stronger indignation and, in turn, this indignation led participants to provide their
email address. Also, the proportion of thoughts that expressed an unfavorable
opinion and also mentioned the relevant ingroup or outgroup was related to more
intense indignation and to marginally more anger. Unfavorable thoughts that did
not mention the ingroup or outgroup were not related to emotions. In other words,
the negative evaluation of the issue only leads to stronger emotions if the salience
of the relevant group membership makes it an issue that affects the social self of
the participant. These findings provide another striking demonstration that social
communication facilitates a group-based perspective, and that this engenders
emotional reactions to group concerns.
Conclusions
The empirical efforts evoked in this chapter show that a discussion of a group-
relevant event has very similar effects on group-based appraisals and group-based
emotions as explicit manipulations of social identity salience. Importantly, the
social interactions in small groups constitute a much more ecologically valid
way of manipulating social identity salience. First, in real-life situations, social
identity is seldom made salient in an explicit way. Second, the group discussion
paradigm also provides a more ecologically valid approach to the content of
group-based appraisals and emotions. Because emotional thoughts and experi-
ences are very likely to be shared with close others (Rimé, 2009), the content of
existing group-based appraisals and emotions has often been influenced by social
communication among group members. This naturally occurring process is better
approached in a dynamic situation such as a group discussion than when partici-
pants answer to a questionnaire in isolation.
Our emphasis on group-based appraisals does not mean that this is the only
or even the most important process leading to group-based emotions. In fact, we
have little information on the dynamic process of how group-based emotions arise
during episodes of group interactions. Emotional contagion, social appraisal, and
outright persuasion within discussion groups all likely influenced participants’
post-discussion emotions. Finding empirical evidence regarding these processes
will thus be an important task for future research.
Over the years, our program of research has demonstrated that people’s emotional
reactions to surrounding events are way more malleable than they would like to
admit. Depending on the way people (are led to) see themselves, the same events
may be evaluated and reacted upon very differently. Striking as our introductory
quotes may be, many other examples confirm that the social and political implica-
tions of our findings cannot be overestimated (Yzerbyt, 2006) but they also suggest
that more work investigating the impact of social sharing and communication among
group members on the emergence of group-based appraisals and emotions is needed.
Acknowledgements
This chapter was written with the support of grant ARC 06/11-337 from the
Communauté française de Belgique and grant FRFC 2.4531.06 from the Belgian
Fund for Scientific Research FRS/FNRS.
References
Dunbar, R. I. M., Marriott, A., and Duncan, N. D. C. (1997). Human conversational
behaviour. Human Nature, 8, 231–246.
Espitalier, M., Tcherkassof, A., and Delmas, F. (2003). Partage social des émotions et
cohésion de groupe. In J.-M. Colletta and A. Tcherkassof (eds). Les émotions. Cognition,
langage et développement (pp. 91–94). Sprimont, Belgium: Mardaga.
Gordijn, E. H., Wigboldus, D., and Yzerbyt, V. Y. (2001). Emotional consequences of
categorizing victims of negative outgroup behavior as ingroup or outgroup. Group
Processes & Intergroup Relations, 4, (4), 317–326.
Kuppens, T., Yzerbyt, V. Y., Dandache, S., Fischer, A. H., and van der Schalk, J. (2011).
Social identity salience shapes group-based emotions through group-based appraisals.
Manuscript submitted for publication.
Manstead, A. S. R., and Fischer, A. H. (2001). Social appraisal: The social world as object
of and influence on appraisal processes. In K. R. Scherer, A. Schorr and T. Johnstone
(eds). Appraisal processes in emotion. Theory, methods, research (pp. 221–232).
Oxford: Oxford University Press.
Peters, K., and Kashima, Y. (2007). From social talk to social action: Shaping the social
triad with emotion sharing. Journal of Personality and Social Psychology, 93, (5),
780–797.
Rimé, B. (2009). Emotion elicits the social sharing of emotion: Theory and empirical
review. Emotion Review, 1, (1), 60–85.
Scherer, K. R., Schorr, A., and Johnstone, T. (2001). Appraisal processes in emotion:
Theory, methods, research. Oxford: Oxford University Press.
Smith, E. R. (1993). Social identity and social emotions: Toward new conceptualiza-
tions of prejudice. In D. M. Mackie and D. L. Hamilton (eds). Affect, cognition, and
stereotyping: Interactive processes in group perception (pp. 297–315). San Diego, CA:
Academic Press.
Smith, L. G. E., and Postmes, T. (2011). The power of talk: Developing discriminatory
group norms through discussion. British Journal of Social Psychology, 50, 193–215.
Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., and Wetherell, M. S. (1987).
Rediscovering the social group: A self-categorization theory. Oxford: Basil
Blackwell.
Yzerbyt, V. Y. (2006). From subtle cues to profound influences: The impact of changing
identities on emotions and behaviors. In P. A. M. van Lange (ed.). Bridging social
psychology: Benefits of transdisciplinary approaches. (pp. 391–396). Mahwah: Erlbaum.
Yzerbyt, V. Y., and Demoulin, S. (2010). Intergroup relations. In D. T. Gilbert, S. T. Fiske
and G. Lindzey (eds). The handbook of social psychology (5th edn, pp. 1024–1083).
Hoboken, NJ: Wiley.
Yzerbyt, V. Y., and Kuppens, T. (2009). Group-based emotions: The social heart in the
individual head. In S. Otten, K. Sassenberg and T. Kessler (eds). Intergroup relations.
The role of motivation and emotion (pp. 143–161). Hove: Psychology Press.
Yzerbyt, V. Y., Kuppens, T., and Mathieu, B. (2011). When talking makes you feel like
a group: Social interaction leads to group-based emotions. Manuscript submitted for
publication.
15 Emotion and emotion regulation
Robert W. Levenson
University of California, Berkeley
A wife expresses affection toward her husband as he walks toward a waiting

taxi. He responds with affection, which leads to her expressing sadness in antici-
pation of the upcoming separation. In social situations, human emotions rarely
occur in temporal isolation. Rather, in our interactions with others, emotions
beget other emotions. The resulting chains of emotions can become long and
variegated, as interactants express and regulate emotions, creating complex
patterns of activation and soothing. Ironically, in both laboratory and survey
research, emotion has typically been measured in ways that cause the inter-
personal, sequential, and temporal information to be lost. In this chapter, I will
review several ways in which we have attempted to preserve this information in
our work, and will consider the value these approaches have had for providing
a deeper understanding of the role that emotion plays in intimate relationships.
Background
In the 1980s, John Gottman and I developed a new observational paradigm for
studying marital interaction in which couples engaged in several unrehearsed
15 minute conversations about marital conflicts and other topics (Levenson and
Gottman, 1983). During the conversations, we obtained continuous measures
of emotional behavior (coded from video and audio recordings) and emotional
physiology (sampled from cardiovascular, electrodermal, and somatic systems
thought to be important for emotion). To obtain a continuous measure of
subjective emotional experience we developed a video recall methodology. In this
procedure, each spouse separately viewed the videotape of their interaction and
used a “rating dial” to rate continuously the valence of her or his own emotions as
experienced during the actual interaction (Levenson and Gottman, 1983).
This paradigm for studying couples interactions represented a marked departure
from the prevailing methods of the day (largely questionnaire studies) in terms
of observing actual emotional behavior, studying two interacting individuals, and
sampling multiple streams of continuous emotional data. Because of this, new
ways to reduce, quantify, and analyze these data were needed.
106 Robert W. Levenson
Emotion dynamics in couples interactions
In typical laboratory studies of emotion, stimuli are standardized and the timing
of stimulus and response are the same for all participants. In our new approach to
studying emotion in couples, conversations were unrehearsed and idiosyncratic.
Accordingly, emotional moments were different in kind and timing for each
couple. Because of this, conventional data averaging techniques, which are useful
for identifying common signals amidst random noise in classic experiments,
are not very helpful. Figure 3 illustrates the problem. The top panel depicts a
husband’s second-by-second rating of the valence (1 = very negative, 5 = neutral,
9 = very positive) of his emotional experience during a five minute period of
sitting silently across from his wife, followed by a 15 minute discussion of a
marital problem. This panel illustrates the ebbs and flows of emotion that are
typical of these kinds of data. The bottom panel depicts the result of applying
second-by-second averaging across 151 of these ratings obtained from husbands
in a study of long-term marriages (Levenson, Carstensen, and Gottman, 1994).
The resultant average is fairly featureless, reflecting the fact that there are few
features in common across participants when these kinds of data are obtained in
this manner.
One husband
1
3
Rating
5
7
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
All husbands (N = 151)

1
3
Rating
5
7
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Figure 3 Upper panel: One husband’s rating of valence (1 = very negative, 5 = neutral,
9 = very positive) of own emotional experience during discussion with
spouse of marital problem (5 minute pre-interaction period followed by 15
minute discussion). Bottom panel: Rating data averaged across 151 husbands.
All data are from a study of long-term marriages (Levenson, Carstensen, &
Gottman, 1994)
Emotion and emotion regulation 107
Measuring the dynamics of couples emotion and emotion regulation
Physiological linkage
Physiological data are typically recorded continuously and thus are compelling
candidates for characterizing temporal dynamics. However, these data are plagued
by autocorrelations (i.e., intrinsic cyclicities) that can inflate simple Pearson
correlations for reasons unrelated to the psychological phenomena of interest. A
good example is found in heart rate, which naturally rises and falls with breathing
(this respiratory sinus arrhythmia is produced by action of the vagal nerve acting
on the pacemaker cells located in the sino-atrial node of the heart). Given that
most individuals breathe at somewhat similar rates, the resultant similarities
in rises and falls of heart rate can inflate correlations both when heart rate is
measured in two interacting individuals as well as when measured on multiple
occasions in the same individual.
One solution to this problem is to examine streams of physiological data for
regular cycles of change and then remove these cyclicities prior to computing
correlations. Time-series analysis provides these kinds of tools. Once autocor-
relations are removed, the relatedness between the residual data streams can be
calculated. To accomplish this in our work, we utilized a bivariate time series
analysis (Gottman, 1981), which characterized the degree of relatedness or
“linkage” between streams of physiological data and proved useful both across
and within individuals.
Emotional “reliving”. One of the first applications of this technique came as
we were searching for a way to establish the validity of the video recall method
for obtaining continuous self-ratings of subjective emotional experience during
couples interaction. Most of us have had the experience of seeing ourselves on
videotape and know how emotionally powerful this can be. Thus, we thought that
watching oneself in a video of an affectively-laden marital interaction would cause
the viewer to “relive” the experience emotionally. Moreover, if this happened, the
mirrored emotions occurring while viewing the taped interaction should produce
patterns of attendant physiological activation that were similar to those that
occurred during the original interaction. When bivariate-time series analyses were
applied to the physiological data obtained during discussions of a marital conflict
and while watching the video recording of these discussions, they did in fact reveal
significant levels of physiological linkage (Gottman and Levenson, 1985).
Linkage and marital satisfaction. If physiology provides a window on to
the emotional quality of marital interaction, then it is interesting to think about
the conditions under which spouses’ autonomic and somatic physiology might
become synchronous or linked. One possibility is that linked physiology is a
marker of “closeness”, which on the surface sounds like it might be a marker
of a satisfied marriage. However, if linkage is matched with a metaphor such as
“enmeshment”, then it might not be such a positive sign.
To investigate the association between physiological linkage and marital
satisfaction, we again used bivariate time series analysis, this time assessing
the degree of relatedness between streams of physiological data obtained from
husbands and wives (corrected for autocorrelations) as they discussed an area
of current conflict in their relationship. We found that the higher levels of physi-
ological linkage were associated with higher levels of marital dissatisfaction
(Levenson and Gottman, 1983). In this context, we interpreted high levels of
physiological linkage as indicating the kind of emotional enmeshment that
occurs when distressed couples become trapped in repetitive cycles of escalating
negative emotion.
Linkage and empathic accuracy. Having used the video recall method to
obtain continuous reports of spouses’ own emotion, we thought we could also use
this approach to assess an important aspect of empathy, the ability to recognize
another person’s emotions. Most previous studies of emotion recognition had
subjects identify the emotions shown in static photographs of facial expressions.
What was needed was a way to assess ability to track emotions in others as they
unfolded over time in dynamic social contexts.
We (Levenson and Ruef, 1992) developed a new measure of dynamic emotion
recognition by modifying our video recall procedure. In this new variant, subjects
watched videotaped interactions between married couples and used the rating dial
to provide continuous ratings of how they thought one of the spouses was feeling
(we again used ratings of valence). Accuracy was operationalized as the similarity
between the subject’s rating and the rating the target spouse had provided of his
or her own emotions during the interaction. In addition to being based on an
objective criterion of accuracy, this method had considerable ecological validity
insofar as subjects were making continuous emotion judgments from information
that was dynamic, multimodal (face, voice, posture, content), and interpersonal1.
While our subjects were rating the emotions of the targets on the videotapes, we
also obtained continuous measures of their physiological responses. We hypoth-
esized that raters who were particularly empathic would be sufficiently attuned
to the emotions of the target that they would experience the target’s emotions
themselves. The similar emotions experienced by raters and targets at similar
points in time should produce similar patterns of physiological response, thus
creating physiological linkage between raters and targets. The results supported
this hypothesis, with high accuracy in rating the negative emotions of others
associated with high levels of physiological linkage (again assessed by bivariate
time-series analysis) between raters and targets (Levenson and Ruef, 1992).
Positive emotion and physiological “soothing”

We have long been interested in the role that certain positive emotions play in
providing us with an antidote that helps “undo” the physiological activation
1 At about the same time as we were developing our measure, Ickes et al., (1990) were developing
a similar measure of what they called “empathic accuracy” in which they had subjects rate the
thoughts and feelings of others.
associated with negative emotions (Levenson, 1988). In our first experimental
test of this notion, we found that positive emotions such as contentment and
enjoyment do in fact shorten the duration of physiological arousal produced by
negative emotions such as fear and sadness (Fredrickson and Levenson, 1998).
In this work we studied individual subjects whose emotions were stimulated by
films, quantifying the duration of physiological response as the time taken for
physiological activation to return to pre-emotional levels.
More recently, we began studying the soothing effect of positive emotions
during couples interactions. For this, we (Yuan, McCarthy, Holley, and
Levenson, 2010) developed a new way to quantify the dynamics of emotion
regulation using a technique based on sequential analysis (Bakeman and
Gottman, 1986). In this approach, we operationalized soothing “events” as
moments in a dyadic interaction where a spouse transitions from being in a state
of high physiological activation (defined as a certain number of physiological
systems exceeding a given threshold for a certain period of time) to being in
a state of low physiological activation (defined as all measured physiological
systems staying below a given threshold for a certain period of time). When
discussing an area of conflict in their relationship, we found that most couples
show at least one of these soothing moments. Examining the observational
coding of emotional behavior that occurred during these soothing moments,
we found that the ratio of positive to negative emotional behaviors was more
emotionally positive than during comparison non-soothing moments (Yuan,
McCarthy, Holley, and Levenson, 2010). This finding supports the notion
that positive emotions are associated with emotion down-regulation during
couples interactions, and points the way toward additional studies focusing on
specific positive emotions and on cross- and co-regulation of emotion between
spouses.
Characterizing the temporal dynamics of emotional behavior, physiology, and

experience
Finding ways to take into account physiology, behavior, and subjective experience
presents significant challenges for research on emotion dynamics. This problem
was pursued by the late Loren McCarter, an immensely talented and creative
scientist who began working on ways to quantify emotion dynamics while a
graduate student in my laboratory. The goal of this work was to identify emotion
cycles in each spouse, which were defined as the transitions between positive and
negative valence in emotional experience and emotional behavior, and between
low activation and high activation in emotional physiology.
I will briefly outline the methodology McCarter developed; additional details
are available on request. The data used to identify emotion cycles were spouses’
second-by-second averages of their own emotional experience (from the rating
dial), emotional behavior (from behavioral coding collapsed into five point
positive-neutral-negative scores), and physiology (using cardiac interbeat interval
scores corrected for somatic activity by regressing the cardiac interbeat interval
time-series on the somatic activity time-series) obtained during the discussion
of a marital problem. Time-series analysis requires that data meet criteria for
stationarity (consistency of mean over time); thus, spouse’s time series that did
not meet these criteria were corrected by fitting a linear trend and using the
residual scores. To remove high frequency noise and oscillations, a natural cubic
splines smoothing method was applied to all of the time series (experience,
behavior, and physiology). In this method, data were taken in “knots” N points
at a time (19 points were used) and a smoothed line was fitted to each knot using
a cubic polynomial function to connect the lines. The resulting data were then
processed as follows: (a) each time series was examined for cycles using a differ-
encing method that focused on identifying minima and maxima; (b) lead and lag
relationships across the separate emotion experience, behavior, and physiology
time-series were examined within spouses using lag-correlations to indentify the
temporal offset that produced the maximum correlation; and (c) each spouse’s
experience, behavior, and physiology time-series were combined into a single
composite time series using a principle components analysis and the resultant
composite was examined for cycles in the same way that had been done for the
individual time series. These analyses revealed a number of interesting findings
concerning emotion dynamics during these couples interactions .
Three-minute cycles. Examining the composite of physiology, behavior, and
experience for each spouse revealed that, on average, spouses went through a
cycle of increasing and decreasing emotional negativity/physiological activation
every three minutes, or five times during the 15 minute conversation. These
cycles reflect the ebb and flow of emotional arousal and soothing/regulation that
we described earlier when characterizing the emotional dynamics that occur
when couples discuss a marital problem. We believe the cycles occur because few
couples can stay fully engaged with this kind of difficult material for the entire 15
minutes, and thus go through cycles of arousal and soothing that are facilitated
by numerous behaviors including positive emotion (Yuan, McCarthy, Holley, and
Levenson, 2010), staying on and going off topic, and engaging and disengaging.
Amplitude of emotion cycles decreases with age. Comparing cycles between
older couples (in their 60s and married at least 35 years) and middle age couples
(in their 40s and married at least 15 years) revealed that older couples had
smaller amplitude cycles (maxima minus minima), but no age differences in cycle
duration. Some of this reduction in amplitude may result from habituation to or
lessening of the impact of the severity of longstanding marital problems over
time. However, these smaller amplitude cycles may also reflect older couples
having greater emotion regulatory skills that allow them to deal with difficult
marital problems without reaching emotional extremes. Consistent with this is a
recent finding from our group that older individuals are better at utilizing emotion
regulation strategies that involve positive reframing than middle-aged or young
individuals (Shiota and Levenson, 2009)
Emotion cycles begin with visceral changes. Examining the pattern of lead
and lag relationships across the 312 spouses in the sample, physiology (cardiac
interbeat interval corrected for somatic activity) was found to be the leading
indicator of each emotion cycle. These cycles that began in physiology were
followed three seconds later by cycles in emotional behavior and two seconds
after that by cycles in emotional experience. Although there are clearly caveats
that must be registered (e.g., there is undoubtedly delay between the “actual” start
of subjective experience and registration of experience on the rating dial), this
suggests a bottom-up organization of emotion dynamics during couples interac-
tions, with physiology occurring first, followed by behavior, and then finally by
emotional experience. Such a model is consistent with the peripheralist views
espoused most famously in the James-Lange theory of emotion and with the
notion that emotional experience is constructed post-facto from the processing of
visceral and somatic information (Levenson, 1999).
Summary and concluding thoughts

Emotions are by nature dynamic rather than steady-state. Biological, behavioral,
and phenomenological aspects of emotion onset progressively in organized ways,
are sustained for finite periods of time, and then offset before cycles start anew.
Emotions rarely occur without the involvement of emotion regulatory processes.
Patterns of activation accompanied by patterns of regulation further add to the
cyclical qualities of emotion. In addition, because most human emotion occurs in
social, interpersonal contexts, co-activation and co-regulation of emotion across
individuals adds additional complexities and richness to emotion dynamics.
Affective science, especially as it branches into the realms of functional versus
dysfunctional emotions, normal versus abnormal emotional behaviors, and the
nature of individual and group differences will be enriched greatly by studying
emotions in ways that allow emotion dynamics to emerge fully. These efforts
will also be well-served by utilizing methods for data reduction and analysis that
preserve temporal and sequential information, and by adopting research designs
and considering research questions that do justice to the rich complexities of
emotion dynamics.
References
Bakeman, R., and Gottman, J. M. (1986). Observing interaction: An introduction to
sequential analysis. New York: Cambridge University Press.
Fredrickson, B. L., and Levenson, R. W. (1998). Positive emotions speed recovery from
the cardiovascular sequelae of negative emotions. Cognition & Emotion, 12, 191–220.
Gottman, J. M. (1981). Time-series analysis: A comprehensive introduction for social
scientists. New York: Cambridge University Press.
Gottman, J. M., and Levenson, R. W. (1985). A valid procedure for obtaining self-report of
affect in marital interaction. Journal of Consulting & Clinical Psychology, 53, 151–160.
Ickes, W., Stinson, L., Bissonnette, V., and Garcia, S. (1990). Naturalistic social cognition:
Empathic accuracy in mixed-sex dyads. Journal of Personality and Social Psychology,
59, 730–742.
Levenson, R. W. (1988). Emotion and the autonomic nervous system: A prospectus for
research on autonomic specificity. In H. L. Wagner (ed.), Social psychophysiology
and emotion: Theory and clinical applications. (pp. 17–42). Chichester: John Wiley &
Sons.
—(1999). The intrapersonal functions of emotion. Cognition and Emotion, 13, 481–504.
Levenson, R. W., Carstensen, L. L., and Gottman, J. M. (1994). Influence of age and
gender on affect, physiology, and their interrelations: A study of long-term marriages.
Journal of Personality & Social Psychology, 67, 56–68.
Levenson, R. W., and Gottman, J. M. (1983). Marital interaction: Physiological linkage
and affective exchange. Journal of Personality & Social Psychology, 45, 587–597.
Levenson, R. W., and Ruef, A. M. (1992). Empathy: A physiological substrate. Journal of
Personality & Social Psychology, 63, 234–246.
Shiota, M. N., and Levenson, R. W. (2009). Effects of aging on experimentally instructed
detached reappraisal, positive reappraisal, and emotional behavior suppression.
Yuan, J. W., McCarthy, M., Holley, S. R., and Levenson, R.W. (2010). Physiological
down-regulation and positive emotion in marital interaction. Emotion, 10, 467–474.
16 Emotional climate
How is it shaped, fostered, and changed?
Darío Páez
Agustín Espinosa
Pontificia Universidad Católica del Perú
PUCP
Magdalena Bobowik
During Christmas, positive feelings are fostered through sharing, and embedded
in symbols (e.g., Santa Claus). This symbolic communication generates an
affective field. Individuals may react to such celebrations with a spectrum of
emotions from excitement and joy to nervousness, anxiety, and loneliness.
Nevertheless, this common affective field constantly fosters hope and joy. The
emotional climate (EC) of Christmas influences personal and group dynamics.
Indeed, being satisfied with Christmas rituals predicts one month later higher
levels of well-being and a more positive perception of family climate (Páez,
Bobowik, Bilbao, Campos, and Basabe, 2011).
This chapter aims to analyze how EC is different from other emotion-related
processes, and to identify the mechanisms that shape and change it. First, we briefly
define EC and how it affects behavior, beliefs, and personal emotions. We then focus
on processes which may affect EC. Objective social indicators, values, collective
identity and memory are factors related to formation and maintenance of EC. Shared
historical experiences of collective violence, as well as dealing with them, can
provoke change in EC.
Emotional climate: what does it mean and what purpose does it

serve?
ECs refer to predominant collective emotions perceived as shared by members of
social groups, such as national communities or ethnic minorities. The term also
reflects how an individual thinks that most of the people feel about their ingroup’s
situation. Unlike emotional atmospheres, which depend on group members
focusing on a particular event, ECs involve relationships among them. ECs can
be expressed as perceptions of collective and interpersonal feelings: fear, used
114 Darío Páez, Augustí Espinosa and Magdalena Bobowik
by dictatorships to ensure order; trust, essential to the formation of social capital;
security, provided by adequate attention to human rights; or anger and despair,
aroused by pervasive corruption.
The key to understanding ECs are the dominant emotions perceived in others
(De Rivera and Páez, 2007). ECs are socially constructed, yet simultaneously
objective, because they exist independently of an individual’s personal feelings
(De Rivera and Páez, 2007; De Rivera, 1992). That is, emotions which people sense
in others have distinct consequences in contrast to own emotional experience. For
instance, individuals tend to feel more positive than negative intergroup emotions,
particularly toward the ingroup. However, at times, members of a low-status
group perceive more negative than positive emotions toward their ingroup. This
perception is often shared by the outgroup. Indeed, research has revealed that
Basques perceive hostility towards their ingroup, and this perception is shared by
Andalusians—see Table 1 (Techio, Zubieta, Páez et al., 2011).
Importantly, EC serves as a psychosocial context that influences behavior
(Bar-Tal, Halperin and De Rivera, 2007). The way people behaved after the
terrorist attack on Madrid in March 2004 was associated with their perception of
the EC. Even controlled for personal emotions, perception of negative EC predicted
avoidance of outgroups (e.g., Muslims). In turn, perception of positive EC explained
altruistic behavior (De Rivera and Páez, 2007). EC also acts as a context that influ-
ences social beliefs. Research has demonstrated that the stronger the perception
of positive EC after the March 2004 bombing, the higher the perception of inter-
personal and collective positive reactions to trauma (e.g., post-traumatic growth
as index of positive social beliefs). Finally, EC influences personal emotions:
perceiving positive EC one week after an event predicts individual positive affect
three weeks after the event (Rimé, Páez, Basabe, and Martínez, 2009).
Emotional climate: how does it change?

Social development and social status. ECs are influenced by diverse macro-social
factors. Páez et al. (1997) and Basabe et al. (2002) found a direct association
between social development and EC balance (quoted in De Rivera and Páez,
2007). An analysis of the relation between human development index (HDI) and
Table 1 Means of perceived emotions in others toward regional in-group and

out-group in Spain
Basque Participants Andalusian Participants

Perceived emotions in In-group or Out-group or In-group or Out-group or
others towards groups Basques Andalusians Andalusians Basques
Disgust 2.87 1.98 2.13 3.15
Anger 2.89 1.82 2.05 3.03
Contempt 2.96 2.11 2.32 3.42
Source: Techio et al. (2011).
Emotional climate 115
EC in Argentina, Brazil, Chile, Peru, and Spain (Table 2) revealed a prevalent
perception of negative EC across nations in Latin America (overall mean of
positive climate: M = 2.98, SD =. 61; negative climate: M = 4.47, SD =. 78),
while in Spain the EC was more positive (Techio et al., 2011). Similarly, EC is
related to social status and social class: positive EC tends to be more common in
upper than in lower socioeconomic classes (De Rivera and Páez, 2007).
Social identity and internalized values. Other studies suggest that EC could be
affected by more stable psychological aspects, such as collective identity. One context
in which people are immersed is national culture, based on a collective identity and
the acceptance of predominant values in the social environment. Espinosa (2011)
demonstrated that strong collective self-esteem and national identification were
related to higher perception of positive and lower perception of negative EC.
Similarly, positive national self-stereotyped beliefs (e.g., the perception of Peruvians
as reliable, patriotic, and supportive) were associated with positive EC, whereas the
perception of Peruvians as unreliable was associated with a negative climate.
Culture expressed in values also plays an important role in relation to EC. At
a collective level, research has demonstrated that nations with egalitarian and
individualistic values exhibit better ECs (De Rivera and Páez, 2007). This occurs
because in these contexts there are fewer stressful events and more equity, social
support, autonomy, and mastery. All of these factors tend to promote well-being
and positive collective affect (De Rivera and Páez, 2007). At an individual level,
studies have also confirmed that individualistic and egalitarian values are linked
to positive affect because they enhance positive emotions, a sense of mastery,
and life satisfaction (Espinosa, 2011). In contrast to the case of collective-level
studies, studies at the individual level found a positive association between
collectivistic values and positive affect. Tradition and conformity values tend
to bring meaning and increase satisfaction with the social environment. In line
with this, research has found that collectivistic values are related to positive
EC. However, negative EC has also been associated with conservative values,
Table 2 Means and Standard Deviations of EC according to nation and its development
Spain Chile Argentina Brazil Peru

(n = 122) (n = 299) (n = 101) (n = 149) (n = 200)
IDH Rank 1 2 3 4 5
Position
M SD M SD M SD M SD M SD
Security 3.94 .69 3.15 .77 2.46 .70 2.62 .75 3.08 .70
Confidence 3.38 .70 3.17 .80 2.86 .69 2.87 .84 2.87 .73
Hopelessness 2.84 1.13 3.87 1.29 4.78 1.19 4.81 1.17 4.51 1.08
Fear 2.91 1.20 3.28 1.31 3.0 1.28 4.14 1.33 3.43 1.16
Anger 4.10 1.32 5.21 1.14 5.63 1.10 5.77 1.06 5.44 1.07
Sources: Techio et al., (2011), Espinosa (2011). Scale: 1–7.
suggesting that collectivist values emphasizing tradition and conformity erode
EC by constraining people (Espinosa, 2011).
Shared experiences. Also socio-economic and political events can change and shape
EC. For instance, Pinochet’s coup in Chile in 1973, in which hundreds people were
killed when he seized power, gave rise to a powerful affective field. Some people
were delighted by the success of the coup, and others quite relieved. Yet there was
also an overall climate of fear (De Rivera, 1992). People were perceived to be afraid
because it was dangerous to say certain things in public, and an unexpected visit
was more likely to lead to fear than pleasant anticipation. Social distrust affected
emotional relationships. People could not speak about relatives who had disap-
peared, or publicly state their political opinions. The fear created social isolation. This
prevented people from knowing how others thought, and prevented the organization
of political opposition against the regime. Other examples confirmed by research
are numerous: overcrowding within prisons in Colombia negatively impacted EC;
massacres created a negative climate in Mayan communities in Guatemala, where a
policy of impunity impeded efforts to restore a climate of trust (De Rivera and Páez,
2007); and victims of collective violence in the Basque Country reported lower
positive and higher negative EC compared with controls (Figure 4).
However, psychosocial processes reflected in the behavior of ordinary people
also influence ECs. The continuity of this behavior appears to maintain the climate.
Thus, it persists beyond the objective conditions that were originally involved. For
example, the climate of fear in Chile continued for a long time after the massive
repression of the two first years of the dictatorship (De Rivera and Páez, 2007).
3.5 3.28 3.20

3 2.65
2.5 2.26
2
1.5
0.94
1
0.5
0
-0.5 Victims Control
-1 -0.63
Positive EC Negative EC Difference
Figure 4 Positive and negative EC in terrorism victims in Spain

(Source: Techio et al. 2011)
Collective memory. Another cultural element that drives EC is a group’s social
representation of its past. Thus, EC will be positive or negative depending on
how the past is perceived, as a function of collective memory, which may evoke
extreme success and pride, or failure, guilt, and shame for the ingroup. For
example, experimentally induced salience of collective violence in the past (such
as the Spanish Civil War or ETA’s violence in Spain) involves poorer percep-
tions of EC (Techio et al., 2011). In Espinosa’s (2011) study, positive EC was
similarly associated with positive appraisal of historical characters and events.
Nevertheless, it was in fact the assessment of Peru’s social situation that predicted
positive EC more than collective identity, values, and collective memory. Such
findings suggest that social situation is a more powerful factor in explaining EC
than distal processes such as collective memory.
Social sharing. Social sharing related to collective events is very common, and
fosters the transmission of feelings and the construction of EC. A longitudinal
study conducted one week after the terrorist attacks in Madrid in 2004 showed
that higher levels of sharing predicted higher event-related negative emotions and
rumination, contributing to the maintenance of negative EC. However, talking
about a collective trauma also plays a role in the creation of positive EC through
the improvement of cognitive resources and social integration (Rimé et al., 2009).
Exposure to information about collective violence in the mass media helps to
shape EC (De Rivera and Páez, 2007).
Collective coping through rituals. Collective forms of coping such as rituals

also influence EC. Rituals are cultural devices that allow people to focus on
common topics and share emotions; they increase bonding, reinforce social
representations, and facilitate change and the improvement of affects and cogni-
tions. A ritual can turn negative emotions and beliefs into positive ones (Rimé et
al., 2009). Secular rituals or demonstrations are collective gatherings in public
spaces aimed at transmitting a symbolic message involving both expressive and
instrumental goals. For instance, in the aftermath of the Madrid terrorist attacks,
demonstrations expressed political conflicts between left- and right-wing political
ideologies, and played a role in the expression of anti-war attitudes among
Spaniards. Participation in demonstrations predicted positive EC two months
after the attacks. In addition, an indirect effect was found: participation reinforced
the perceived social integration after three weeks and post-traumatic growth in
response to trauma. Both, in turn, led to positive EC two months later.
Participation in transitional justice rituals, such as the Gacaca (or popular
trials) in Rwanda, also affected perceived EC. Gacaca elicited an emotional
communion among participants and fostered intense emotional manifestations
and re-evocations of the genocide. Among participating victims, positive EC was
rated as higher than among their controls before Gacaca, probably because of the
hopes and positive expectations that characterized participating victims in the
period preceding the trial. Their positive perception of the climate decreased after
the trial, but remained higher than among victims in the control group, suggesting
that their hopes did not entirely vanish with the trial. Perpetrators, for their part,
perceived the EC as less positive than their controls before Gacaca, probably
because they expected to be punished. After Gacaca, the perception of positive
emotions in the social climate was markedly increased among perpetrators who
participated. These findings are consistent with those of other studies, which
found that active perpetrators showed a more positive attitude toward transitional
rituals, especially when they did not receive hard punishment—as occurred in
the Gacaca trials. Furthermore, both victims and perpetrators increased positive
out-group stereotypes and individualized perceptions of out-group members,
factors that fuel reconciliation and positive EC (Martin-Beristain, Páez, Rimé,
and Kanyangara, 2010).
Institutional reparatory acts. Finally, institutional reparatory actions also influence

EC. When exposed to past collective misdeeds by one’s group, people tend to
believe that victims deserved their fate, devaluing them and justifying ingroup
actions. In contrast, an official apology reframes past misdeeds positively,
showing that at least some symbolic action can be performed. This helps people
to reject beliefs that the world is just (and that victims deserved their fate), and
can be a source of collective pride that helps people accept the past misdeeds of
their group. An experimental study showed that the salience of past collective
violence in Spain at the same time as the Law of Historical Memory (an institu-
tional reparatory behavior) reinforces positive EC and agreement with reparatory
behaviors in comparison with a control condition in which people were only
reminded of collective violence. In a similar study on the effects of an apology
issued by the Basque Parliament addressing the victims of collective violence in
the Basque Country, participants exposed to the apology, as contrasted with the
control condition, reported higher levels of shame and agreed more with repara-
tions. As concerns positive EC, it was higher in the experimental than the control
condition for participants with low identification with Basques, confirming that
apologies have positive effects on social cohesion, especially among vicarious
victims (Bobowik, Bilbao, and Momoitio, 2010). Finally, in line with the findings
of previous studies, the Belgian prime minister’s official apology for past colonial
misdeeds in the former Belgian Congo, together with the salience of outgroup
suffering, increased the perception of a positive EC, which was significantly
higher than in the condition where past misdeeds were merely remembered
(Lastrego and Licata, 2010). In addition, Belgians who were exposed to the
apology, in contrast to a control group, reported a less positive view of the
past. This less positive view of the past mediated the effect of the apology on
agreement with reparation and positive intergroup attitudes. This result suggests
that apologies help to increase collective moral emotions and to construct a more
integrative social representation of the past. In a reframed narrative, victims’
experience matters, and they are entitled to dignity. Such a narrative, which both
parties can live with, helps to improve EC, reflecting how current EC is related to
social representations of the past.
Conclusions
As in the case of differences in individual affect, direct regulation of negative EC
is more difficult and is determined mainly by collective stress, while regulation of
positive EC is more successful and is mainly associated with social support and
positive reappraisal. Indeed, our review of recent research suggests that experi-
ences of social problems, conflict, and collective violence boost negative EC. In
turn, improvements in social problems, participation in altruistic behaviors and
demonstrations, institutional reparatory rituals, community empowerment and the
construction of inclusive collective memories may augment positive EC. These
tools therefore constitute the vehicles for directly reinforcing positive climate and
indirectly attenuating the perception of negative climate.
Acknowledgements
Preparation of this chapter was facilitated by Grants Psi2011-26315 from the
Spanish Ministry of Scientific Research and UFI 11/04 from the University of the
Basque Country UPV/EHU.
References
Bar-Tal, D., Halperin, E. and De Rivera, J. (2007). Collective emotions in conflict situa-
tions: societal implications. Journal of Social Issues, 63, 441–459.
Bobowik, M., Bilbao, M. A., and Momoitio, J. (2010). Psychosocial effects of forgiveness
petition and ‘self-criticism’ by the Basque Government and Parliament directed to the
victims of collective violence. Revista de Psicología Social, 25, 1, 87–100.
De Rivera, J. (1992). Emotional Climate: Social Structure and Emotional Dynamics. In K.
T. Strongman (ed.). International Review of Studies on Emotion (Vol. 2, pp. 197–218).
New York: John Wiley & Sons.
De Rivera, J. and Páez, D., (2007). Emotional climate, human security and culture of
peace. Journal of Social Issues, 63, 233–253.
Espinosa, A. (2011). Psychological, Social and Cultural Correlates of National Identity in
Peru. Doctoral Dissertation. San Sebastian, Spain. University of the Basque Country.
Lastrego, S. and Licata, L. (2010). Should a country’s leaders apologize for its past misdeeds?
An analysis of the effects of both public apologies from a Belgian official and perception of
Congolese victims continued suffering. Revista de Psicología Social, 25, 1, 61–72.
Martin-Beristain, C., Páez, D., Rimé. B., and Kanyangara, P. (2010). Psychosocial effects of
participation in rituals of transitional justice. Revista de Psicología Social, 25, 1, 47–60.
Páez, D., Bobowik, M., Bilbao, M. A., Campos, M., and Basabe, N., (2011). Merry
Christmas and Happy New year! The impact of Christmas rituals on subjective well-
being and family’s emotional climate. Revista de Psicología Social, 26, (3), 373–386.
Rimé, B., Páez, D., Basabe, N, and Martinez, F. (2009). Social sharing of emotion, post-
traumatic growth, and emotional climate: Follow-up of Spanish citizens’ response to the
collective trauma of March 11th terrorist attacks in Madrid. European Journal of Social
Techio, E., Zubieta, E., Páez, D., et al. (2011). Emotional Climate and Collective Violence:
the state of art and measures. In D. Páez, C. Martín-Beristain, J. L. Gonzalez and J.
De Rivera (eds). Overcoming Collective Violence and Building a Culture of Peace
(pp. 99–148). Madrid: Fundamentos.
17 Dynamics of ideal affect
Jeanne L. Tsai
Stanford University
“The word ‘romance,’ according to the dictionary, means excitement,

adventure, and something extremely real. Romance should last a lifetime.”
Billy Graham, Christian evangelist
“I learned that the richness of life is found in adventure. . . . It develops

self-reliance and independence. Life then teems with excitement. There is
stagnation only in security.”
William Orville Douglas, former Supreme Court justice
“To find your own way is to follow your bliss. This involves analysis,
watching yourself and seeing where real deep bliss is—not the quick little
excitement, but the real deep, life-filling bliss.”
Joseph Campbell, mythologist and philosopher
“I just want to be happy. Normally for me that means I would be doing

something exciting. I just want to be entertained . . . I just like excitement.”
European American college student
“My ideal state is to be quiet, serene, happy and positive.”

Hong Kong college student
Whether you are an eminent spiritual leader, outspoken Supreme Court justice,
influential philosopher, or typical college student, chances are you have some
notion about which feelings you would like to feel. As illustrated by the above
quotes, however, people vary in the specific feelings that they believe are good,
moral, and virtuous. Whereas Billy Graham states that excitement “should last
a lifetime,” Joseph Campbell believes that genuine bliss is more than “quick
little excitement.” What explains these differences in how people ideally want
to feel? Affect Valuation Theory (AVT) posits that much of what we learn about
our feelings comes from our cultures—those historically derived and socially
transmitted ideas that are instantiated in artifacts, practices, and institutions
(Kroeber and Kluckhohn, 1952). And whether we realize it or not, these cultural
prescriptions influence how we act in the world: how we represent ourselves,
Dynamics of ideal affect 121
how we perceive others, what choices and decisions we make, and what we think
comprises success, health, and happiness.
But are these affective ideals dynamic, and if so, what are the conditions under
which they change versus remain the same? Prior to answering these questions,
we briefly review AVT and the empirical work testing its predictions. We then
discuss four sources of change in ideal affect (daily, acculturative, cultural, and
emotional). We conclude with directions for future research.
Affect Valuation Theory

AVT is a theoretical framework that attempts to integrate affective values into
working models of emotion (Tsai, 2007). The first premise of AVT is that how
people ideally want to feel (“ideal affect”) differs from how they actually feel
(“actual affect”). “Actual affect” may be an immediate response to an event
(state), or an aggregate of responses to different events (trait). The vast majority
of the literature on emotion and other affective phenomena has focused on
actual affect. In contrast, relatively little work has focused on ideal affect.
Ideal affect is a goal or desired state, again either in response to a specific
event, or aggregated over time. Although there are times when people feel
exactly how they want to feel, our data suggest that across a number of diverse
cultural contexts, on average, most people want to feel more positive than
negative, and want to feel more positive and less negative than they actually
feel (Tsai, Knutson, and Fung, 2006). These findings hold whether we examine
people’s global ratings of how they actually and ideally want to feel over the
course of a typical week, or whether we examine people’s momentary ratings of
actual and ideal affect over the course of a day (Tsai, Sims, Wang, and Thomas,
2012; Tsai, 2007).
The second premise of AVT is that, although culture shapes both actual
and ideal affect, culture shapes ideal affect more than actual affect. In other
words, our cultures teach us which feelings to strive for and which feelings to
avoid. However, our ability to actually achieve these ideals may depend more
on a host of other factors, including our affective predispositions (i.e., temper-
ament), immediate circumstances, and regulatory skills and abilities. Conversely,
although culture shapes how we actually feel, decades of studies have observed
a stronger link between temperament and actual affect (specifically between
neuroticism and the experience of high arousal negative states, and between
extraversion and the experience of high arousal positive states). Therefore, we
predict that temperament may shape people’s actual affect more than their ideal
affect.
We have tested these hypotheses using a variety of methods (standardized
instruments, open-ended questions, experimental paradigms) and a variety of
samples (college students, preschoolers, community adults) in North American
and East Asian contexts. Consistently, we have observed that North American
cultural contexts value excitement, enthusiasm, and other high arousal positive
states more and calm, peacefulness, and other low arousal positive states less
122 Jeanne L. Tsai
than Chinese contexts do (see Tsai, 2007). Moreover, we have found that these
differences hold after controlling for temperament and for any differences in
actual affect. Finally, whereas cultural values account for a greater percentage of
variance in ideal than actual affect, temperamental factors account for a greater
percentage of variance in actual than ideal affect (Tsai et al., 2006).
These cultural differences in ideal affect are reflected in, and transmitted and
reinforced by, widely distributed “artifacts” and products. For example, North
American women’s magazines and children’s storybooks contain more excited
and fewer calm smiles than Chinese magazines and children’s storybooks (Chim,
Moon, and Tsai, 2009; Tsai, Louie, Chen, and Uchida, 2007). Similarly, Christian
self-help books encourage their readers to experience excitement states more
and calm states less than Buddhist self-help books, and these same patterns are
reflected in classical texts, such as the Gospels of the New Testament and the
Dhammapada (Tsai, Miao, and Seppala, 2007).
The third premise of AVT is that ideal affect has a variety of behavioral conse-
quences. For instance, the more people value excitement states, the more they
prefer exciting (vs. calm) music; engage in physically rigorous activities; choose
exciting (vs. calm) gums, lotions, and beverages; positively rate physicians who
promote energetic (vs. relaxed) lifestyles; and perceive people with excited (vs.
calm) expressions as friendly (Moon, Chim, Tsai, Ho, and Fung, 2011; Sims, Tsai,
Thomas, and Goldstein, 2012; Tsai, Knutson, and Rothman, 2012). Thus, cultural
differences in a number of behaviors are due at least in part to cultural differences
in ideal affect.
The dynamics of ideal affect

The findings described above are consistent patterns that we have observed over
a variety of studies. And yet, both within and between individuals, we also find
significant variability in ideal affect. We discuss the sources of this variability
next.
Daily Change. Over the course of a day, we feel a variety of ways. We may
feel tired in the morning, alert after drinking a cup of coffee, interested while
reading the paper, stressed while reading email, and excited about seeing our
friends. Similarly, our ideal affect may also change over the course of a day. We
may typically want to feel excited and passionate about what we do, but during a
difficult exchange with a colleague, we may want to feel more calm. But despite
these changes, we predict that momentary ideal affect should be less variable than
momentary actual affect: participants should retain an idea of how they want to
feel that holds across situations. Therefore, changes in ideal affect should only
occur under special circumstances. In contrast, actual momentary affect may be
more susceptible to the whims of the situation.
Consistent with these predictions, in two experience sampling studies,
college students and community adults varied in their actual and ideal affect,
and the variability in their ideal affect was on average smaller than that of
their actual affect (Tsai et al., 2012). These findings held for both European
American and Chinese American community adults. However, we also found
interesting cultural differences: Chinese American reports of momentary ideal
affect were more variable than those of European Americans. These findings
held after controlling for variability in momentary actual affect (European
American reports of momentary actual affect were more variable than Chinese
American reports of momentary actual affect), suggesting that how Chinese
American adults want to feel may be more context-dependent than how
European American adults want to feel. This finding is consistent with other
work demonstrating that East Asian contexts encourage greater situational
malleability than mainstream American contexts (Leu, Mesquita, et al., 2010;
Oishi, Diener, Scollon, and Biswas-Diener, 2004).
What specific situational factors might account for daily change in momentary
ideal affect? In previously published research (Tsai, Miao, Seppala, Fung,
and Yeung, 2007), we demonstrated that ideal affect varied as a function of
individuals’ interpersonal goals. When people wanted to influence others (i.e.,
change others’ behaviors to be consistent with their own), they valued high
arousal positive states more and low arousal positive states less than when they
wanted to adjust to others (i.e., change their own behaviors to be consistent with
others’). Although there are cultural differences in the importance placed on
influencing vs. adjusting to others, we predict that within cultures, individuals
may also vary in terms of whether they are leading a group (influence goal) or
listening to their partners (adjustment goal). Thus, across cultures, daily changes
in interpersonal goals may result in daily changes in ideal affect.
Acculturative Change. Changes in ideal affect should also occur as a function
of enculturation (e.g., a young child learning how to behave in a specific situation)
and of acculturation (e.g., a recent Chinese immigrant learning how to live in
American culture). In Tsai, Louie, et al., (2007), we demonstrated that short-term
exposure (during the experimental session) to exciting vs. calm stories immedi-
ately altered children’s affective ideals. In Koopmann-Holm, Sze, Ochs, and Tsai
(in press) we demonstrated that longer-term engagement in a cultural practice
(i.e., participating in an eight-week meditation class) increased the value placed
on calm states, but did not affect the actual experience of calm states. Importantly,
these changes were not due to expectancy or demand effects. Similarly, we have
observed that the more oriented to American culture Chinese Americans are, the
more they value excitement states, suggesting that as individuals become more
exposed to and engaged with American culture, they value excitement more.
It is also possible that bicultural Chinese Americans are exposed to Chinese and
American situations in the same day (e.g., being at home with Chinese parents
vs. at school with European American friends). As a result, Chinese Americans’
ideal affect may change depending on whether they are in a Chinese or American
context, and whether their Chinese or American values are more or less salient
(Perunovic, Heller, and Rafaeli, 2007). This may explain why the ideal affect of
Chinese Americans and other East Asian Americans at times resembles that of
European Americans, at times looks more like that of East Asians, and at times
falls in between the two groups.
124 Jeanne L. Tsai
Sociocultural Change. The value placed on excitement in US culture likely
stems from the immigrant origins of the United States: people who left their
homelands in the hopes for a better life were people who could anticipate gains
and rewards, even in the face of incredible uncertainty and adversity. Immigrants
had the goal of influencing their circumstances, or changing their circumstances
so that they would be in line with their goals and desires for themselves and their
families (Kitayama, Ishii, Imada, Takemura, and Ramaswamy, 2006). Thus, the
characteristics of the founding fathers may have created an excitement culture,
although the specific expression of this culture may change over time. In contrast,
Chinese cultural ideas and practices are grounded in Confucian, Taoist, and
Buddhist traditions, which emphasize fitting in, adjusting to existing roles and
hierarchies, and respecting elders and cultural traditions. These cultural ideas and
practices laid the foundation for a calm culture, although the specific expression
of this culture may also change with time. These different historical traditions
explain why across studies, we consistently find that Northern American cultures
value excitement states more and calm states less than Chinese and other East
Asian cultures.
However, despite this cultural stability, cultures also change in ways that may
alter the degree to which specific states are valued. In addition to consistent
cultural differences in the relative value placed on excitement vs. calm states,
we also find that the absolute levels of ideal affect vary across studies. For
example, in some studies European Americans value excitement more than
calm, whereas in others, we see the reverse. Regardless, we always find that
European Americans value excitement more and calm less than their Hong Kong
Chinese and other East Asian counterparts. What factors might account for these
changes in absolute levels of ideal affect? One possibility is the occurrence of
significant national events that may change citizens’ concerns. For example, the
value that European Americans placed on calm states increased after September
11, 2001. Since then, there have been persistent threats on American security.
With the constant reminders of the possibility of terrorist attacks, Americans’
anxiety levels have increased, resulting in a greater value placed on calm states.
Similarly, changes are occurring in Chinese contexts that could alter individuals’
ideal affect: in certain corners of China, rapid industrialization and an infusion of
wealth might increase the value placed on influence (relative to adjustment), and
therefore, increase the value placed on excitement (vs. calm).
Emotional Change. In addition to examining how ideal affect may itself
change, we have been interested in how ideal affect affects the unfolding of an
emotional event. We have demonstrated that ideal affect and actual affect differ
from each other. But to what degree does ideal affect influence how someone
responds to an emotional event, especially one that is chosen, like riding a
rollercoaster. As others have argued, our responses to an emotional event include
much more than just how we feel at one moment. For example, we first anticipate
the event (“How will it make me feel?”), then we actually experience the event
(“How does it make me feel?”), and then we recall the event (“How did it make
me feel?”). Consistent with Robinson and Clore’s accessibility model (Robinson
and Clore, 2002), we predicted that people’s ideal affect would influence the more
reflective aspects of an emotional event (i.e., the anticipation and recall parts of
the episode) more than the experiential aspects of an emotional event (the actual
experience). We have found evidence in support of this prediction, as have others
(Scollon, Howard, Caldwell, and Ito, 2009). Specifically, in a study in Hong
Kong examining people’s experience of exciting and calm amusement park rides,
we have observed that ideal affect is associated with anticipated and recalled
reports of affect more than online reports of affect (Chim, Moon, Tsai, Ho, and
Fung, 2011). These findings may explain why cultural differences in emotional
experience are greater for retrospective vs. online reports of affect (Oishi, 2002) .
Future directions
To further examine the dynamics of ideal affect, we are currently using a variety
of methods including fMRI, which will allow us to examine the temporal course
of ideal affect with greater precision. We are also hoping to capitalize on current
technologies like the iPhone, which will allow us not only to examine affect
online, but also to link people’s desired affect with their concurrent and subse-
quent choice of activities, music, and videos. We are interested in conducting
longitudinal studies to examine how ideal affect changes over the life span.
Finally, although most of our work has focused on ideal affect, we have also
explored individual and cultural variation in avoided affect (the affective states
people want to avoid) and should affect (the affective states people think they
ought to feel) (Chim, Tsai, Zhu, and Zhang, 2011; Koopmann-Holm and Tsai,
2011). In future work, we hope to examine how these affective constructs interact
with each other to influence people’s emotions and behavior across cultures.
Together, these studies should give us a richer, fuller, and more dynamic under-
standing of the cultural shaping of ideal affect.
References
Chim, L., Moon, A., and Tsai, J. L. (2009). Beauty is in the culture of the beholden:
The occurrence and perception of American and Chinese smiles in magazines. Paper
presented at the 10th Annual Meeting of the Society of Personality and Social
Psychology.
Chim, L., Moon, A., Tsai, J. L., Ho, Y. W., and Fung, H. H. (2012). Riding the emotional
roller coaster: The ‘roll’ of ideal affect in emotional experience. Paper presented at the
Stanford-Berkeley Social/Personality/Affective Science Talks.
Chim, L., Tsai, J. L., Zhu, L., and Zhang, X. (2011). Cultural differences in the importance
of ought affect for mental health. Paper presented at the 12th Annual Meeting of the
Society of Personality and Social Psychology.
Kitayama, S., Ishii, K., Imada, T., Takemura, K., Ramaswamy, J, (2006). Voluntary
settlement and the spirit of independence: Evidence from Japan’s ‘northern frontier’.
Journal of Personality and Social Psychology, 91, (3), 369–384.
Koopmann-Holm, B., Sze, J., Ochs, C., and Tsai, J. L. (in press). Buddhist-inspired
meditation increases the value of calm. Emotion.
126 Jeanne L. Tsai
Koopmann-Holm, B., and Tsai, J. L. (2011). Cultural differences in avoided affect: A
comparison of American and German contexts. Poster presented at the 12th Annual
Meeting of the Society for Personality and Social Psychology, San Antonio, TX.
Kroeber, A. L., and Kluckhohn, C. (1952). Culture: A critical review of concepts and
definitions. Papers of the Peabody Museum of Archeaology and Ethnology (Vol. 47).
Cambridge, MA: Harvard University Press.
Leu, J., Mesquita, B., Ellsworth, P. C., Zhang, Z. Y., Yuan, H. J., Buchtel, E., et al.
(2010). Situational differences in dialectical emotion: Boundary conditions in a cultural
comparison of North Americans and East Asians. Cognition and Emotion, 24, (3),
419–435.
Moon, A., Chim, L., Tsai, J. L., Ho, Y. W., and Fung, H. H. (2011). The influence of
cultural differences in ideal affect on self-presentation and other-perception of Facebook
profiles. Paper presented at the 12th Annual Meeting of the Society of Personality and
Social Psychology.
Oishi, S. (2002). The experiencing and remembering of well-being: A cross-cultural
analysis. Personality and Social Psychology Bulletin, 28, 1398–1406
Oishi, S., Diener, E., Scollon, C. N., and Biswas-Diener, R. (2004). Cross-Situational
Consistency of Affective Experiences Across Cultures. Journal of Personality & Social
Psychology, 86, (3), 460–472.
Perunovic, W. Q. E., Heller, D., and Rafaeli, E. (2007). Within-person changes in the
structure of emotion: The role of cultural identification and language. Psychological
Science, 18, 607–613.
Robinson, M. D., and Clore, G. L. (2002). Belief and feeling: Evidence for an accessibility
model of emotional self-report. Psychological Bulletin, 128, (6), 934–960.
Scollon, C. N., Howard, A. H., Caldwell, A. E., and Ito, S. (2009). The role of ideal affect
in the experience and memory of emotions. Journal of Happiness Studies, 10, (3),
257–269.
Sims, T., Tsai, J. L., Koopman-Holm, B., Thomas, E. and Goldstein, M. (2012). Valuing
excitement shapes medical choices. Manuscript under review.
Tsai, J. L. (2007). Ideal affect: Cultural causes and behavioral consequences. Perspectives
on Psychological Science, 2, 242–259.
Tsai, J. L., Knutson, B., and Fung, H. H. (2006). Cultural variation in affect valuation.
Journal of Personality and Social Psychology, 90, 288–307.
Tsai, J. L., Knutson, B., and Rothman, A. (2011). The pursuit of ideal affect. In L.
Cheriassky (chair). The motivational properties of anticipatory affect. Symposium
conducted at the annual meeting of the Society of Personality and Social Psychology,
San Antonio, TX.
Tsai, J. L., Louie, J., Chen, E. E., and Uchida, Y. (2007). Learning what feelings to desire:
Socialization of ideal affect through children’s storybooks. Personality and Social
Psychology Bulletin, 33, 17–30.
Tsai, J. L., Miao, F. F., and Seppala, E. (2007). Good feelings in Christianity and
Buddhism: Religious differences in ideal affect. Personality and Social Psychology
Bulletin, 33, 409–421.
Tsai, J. L., Miao, F. F., Seppala, E., Fung, H. H., and Yeung, D. (2007). Influence and
adjustment goals: Sources of cultural differences in ideal affect. Journal of Personality
and Social Psychology, 92, 1102–1117.
Tsai, J. L., Sims, T., Wang, I., and Thomas, E. (2012). Ideal affect in daily life. Manuscript
in preparation.
18 Emotional acculturation
Jozefien De Leersnyder, and Batja Mesquita
Heejung Kim
University of California, Santa Barbara
Do emotions acculturate when people move from one culture to the next? We
conceive of “emotional acculturation” as the process by which immigrants come
to share the host culture’s most prevalent patterns of emotional experiences. In
this chapter we will discuss the first evidence that emotional acculturation takes
place, and provide details on the dynamics of this process. We will also highlight
how the finding of emotional acculturation speaks to mechanisms of emotional
change generally.
Emotional patterns and emotional similarity

Each culture is characterized by a distinct pattern of emotional experiences:
emotions that are consistent with the prevalent cultural ideas and practices are
experienced at higher frequency and intensity, while emotions that are incon-
sistent are rather rare (e.g., Markus and Kitayama, 1994; Mesquita, 2003;
Mesquita and Leu, 2007). Individuals engaging in the same cultural context will
tend to share the same patterns of emotions. Conversely, the emotions of people
engaging in different cultural contexts tend to diverge.
What does it mean exactly when the patterns of emotions diverge? Let’s take
Ayşe, a Turkish girl, and Ann, a Belgian girl. Both girls encounter the situation
that a colleague at work claims a prestigious task for herself, while this task had
been assigned to the girls as well. In some ways Ann and Ayşe’s emotional experi-
ences are similar: they are both angry. Yet, Ayşe’s angry feelings are accompanied
by shame and guilt, because the situation prompts her to think about the different
ways in which she herself may have contributed to the incident. Ann would likely
report high intensity anger, but low intensity for all other emotions. The patterns
of emotions that the girls experience may be seen as reflecting their respective
take on the emotional event. While Ann focuses on the way her colleague blocks
her goals, Ayşe is more focused on the relational aspects of the situation (and how
she herself might have contributed to them). The differences in emotional patterns
can be understood from differences in the dominant meanings of each cultural
context. Ann’s anger is consistent with the dominant Belgian values of autonomy
and independence. In contrast, Ayşe’s feelings of shame and guilt would be
consistent with the dominant cultural ideas in Turkish contexts that emphasize
128 Jozefien De Leersnyder, Batja Mesquita, and Heejung Kim
social harmony. In each case, these meanings fit within the dominant ideas of
the cultural context in which she has spent most of her life. Different patterns of
emotions reflect different cultural practices and beliefs; and by the same logic,
two people’s emotional patterns will converge to the extent that their world views
are similar.
Emotional acculturation: adoption of the host culture’s emotional

pattern
Suppose Ayşe moved to Belgium: would her emotional responses become
more like Ann’s? Would Ayşe’s exposure to the new cultural context change
her emotional patterns? Several recent studies suggest that people’s emotional
patterns do change when they change cultures (De Leersnyder, Mesquita, and
Kim, 2011). We conducted two studies in which we tested these ideas, involving
Korean immigrants in the US and Turkish immigrants in Belgium. In both
studies, we found that an immigrant’s emotional concordance (similarity) with
the host culture’s emotional pattern was commensurate to his or her engagement
in that new culture.
We selected these two immigrant groups because they were maximally
different: on average, Korean immigrants in the US tend to be educated and
middle class, whereas the majority of Turkish immigrants in Belgium received
little education and are working class. Convergent findings for Korean and
Turkish immigrant groups would thus bolster our confidence in the phenomenon
of emotional acculturation.
We did not ask immigrants to what extent their emotions were similar to
those of the host culture. Rather, we had respondents from both immigrant and
host culture groups rate their emotional experiences on the Emotional Patterns
Questionnaire (EPQ) that was developed for the purpose of measuring emotional
acculturation. In the EPQ, participants first report a recently encountered situation
that elicited one of four types of emotion, as defined by both valence (positive,
negative) and social engagement (socially engaging emotion, socially disen-
gaging emotion) in one type of relationship context (either family or work/
school). Then, they rate the intensity of their feelings during that situation on a
set of 17 emotion scales (1: totally not – 7: extremely); these 17 emotion scales
were confirmed to be structurally equivalent across cultures. To measure people’s
emotional concordance, we first calculated the average host group pattern of
emotions for each type of situation. We then correlated, by situation type, each
person’s emotional pattern to the corresponding average emotional pattern of the
host culture. We coined these correlations the person’s “emotional concordance
score”.
There are several indications that emotional concordance is a meaningful
measure of acculturation. First, the mean emotional concordance with the (Belgian
or European American) host culture pattern was highest for the host groups
themselves, and lowest for first generation immigrants; the concordance scores
of second or later generation immigrants fell neatly in between (De Leersnyder
Emotional acculturation 129
et al., 2011; for the Belgian data, see Figure 5, left panel). Furthermore, both
Korean and Turkish immigrants were more emotionally concordant to the
respective host group patterns to the extent that they spent a greater part of their
life in the host culture. Finally, the number of social contacts that Korean and
Turkish immigrants had with members of their respective host groups (measured
as the degree to which they used the host culture language across many different
social contexts) was predictive of their level of concordance. Therefore, across
immigrant groups and host cultures, a person’s exposure to mainstream culture
predicts convergence to the mainstream patterns of emotional experience.
A qualification should be made: across cultural contexts, and for immigrants
and host group members alike, we found consistent evidence that emotional
concordance was higher in positive than in negative situations. The reason may
be that the emotional complexity of positive situations is lower: the same set of
emotion items co-varied more in positive than in negative situations. Therefore,
it may be easier to acquire the new culture’s emotional patterns in positive than
in negative situations; an idea that is consistent with developmental evidence
showing that children master emotion knowledge in the positive domain before
they do in the negative domain (Doost, Moradi, Taghavi, Yule, and Dalgeish,
1999).
Interestingly, while emotional concordance is commensurate with objective
measures of acculturation (e.g., age of immigration), it is unrelated to the
commonly used acculturation scales. The latter scales measure the immigrant’s
willingness to adopt the host culture’s values and traditions (e.g., Ryder, Alden,
and Paulhus, 2000). An immigrant may welcome a new culture without being
emotionally concordant to it (e.g., new immigrants) or, conversely, be reluctant
to adopt the mainstream culture’s values and yet be emotionally concordant
0.8
0.7
0.6
0.5
Turkish students
0.4 Turkish first generation
Turkish second generation
0.3
Belgians
0.2
0.1
0
Belgian emotional pattern Turkish emotional pattern
Figure 5 Mean emotional concordance scores to the Belgian and Turkish average
emotional patterns, matched for type of situation.
(e.g., second generation). This means that the extent to which an immigrant
shares the meanings and practices of the host culture, as reflected by emotional
acculturation is independent of that immigrant’s desire to be part of the host
culture. Emotional change is contingent on the immigrant’s implicit acceptance,
but not on his/her explicit endorsement of the new culture.
Emotional acculturation: maintenance of the heritage culture’s

emotional patterns.
Now suppose that Ayşe moved to Belgium, and that her emotional responses
acculturated. Under what circumstances would Ayşe’s emotional patterns also be
concordant with the Turkish emotional patterns? What would predict her mainte-
nance of the Turkish emotional patterns? And would Ayşe’s emotional patterns be
more Belgian in some contexts, and more Turkish in other?
To answer these questions, we expanded the Turkish-Belgian data with a large
sample of native Turkish college students in Turkey. The Turkish respondents
completed the same Emotional Patterns Questionnaire that we used for the
Belgian samples. We first established the structural equivalence between the
emotion ratings in the native Turkish group in Turkey and the Turkish immigrant
groups in Belgium; the majority of emotions were again found to be struc-
turally equivalent. We then calculated the common heritage culture’s emotional
patterns by averaging the ratings of the Turkish college students by situation. The
emotional concordance to this Turkish average emotional pattern was established
for every Turkish immigrant in Belgium who had participated in our earlier
studies. The results show that, at the group level, mean concordance scores
with the Turkish emotional pattern are highest for Turkish students themselves
and lowest for second generation immigrants to Belgium, with first generation
immigrants’ concordance scores in the middle (De Leersnyder, Mesquita, and
Kim, 2012; see Figure 5, right panel). Emotional concordance was yet again
higher in positive than in negative situations.
Furthermore, we found that Turkish immigrants were more concordant to the
Turkish emotional patterns to the extent that they spent a greater part of their
life in Turkey and immigrated to Belgium at an older age. Furthermore, Turkish
immigrants’ maintenance of Turkish emotional patterns was predicted by the
number of daily social interactions in which Turkish was the language spoken.
This means that immigrants who engaged in Turkish cultural contexts tended to be
more emotionally similar to the average Turkish emotional patterns. Immigrants’
explicit attitudes towards the maintenance of Turkish values and traditions did
not predict concordance with the heritage emotional patterns; these findings are
parallel to the findings on mainstream culture acculturation of emotions.
As a final step, we calculated the emotional concordance of native Belgians
living in Belgium to the Turkish emotional pattern (Figure 5, right panel). As
would have been expected on the basis of their exposure to Turkish culture, the
Belgian group has lower emotional concordance with the Turkish emotional
patterns than any of the other groups. This finding suggests that emotional
concordance is not merely a matter of education. The Turkish and the Belgian
samples were more similar in education than either of these groups were with the
Turkish immigrant groups—especially the first generation immigrants—and yet,
their emotional patterns were the most divergent.
Moreover, we did the reverse as well (Figure 5, left panel), and found that
Turkish college students in Turkey were less emotionally concordant with
the Belgian emotional pattern than any of the immigrant groups. The lower
concordance of monocultural respondents with the other culture’s emotional
pattern is evidence that immigrants’ lower emotional concordance with the host
country’s emotional patterns is not due to unreliable data. Rather, in Figure 5, the
ascending concordance with the Belgian emotional pattern, and the descending
concordance with the Turkish emotional pattern, seem to reflect the exposure of
the various groups to these two cultures respectively.
Our research has also provided some first insights into the conditions under
which the one or the other culture is foregrounded. In family contexts or at home,
Turkish first generation immigrants were far more concordant to the Turkish
emotional patterns than to the Belgian ones, whereas Turkish second generation
immigrants were equally concordant to both culture’s emotional patterns. In
work or school contexts, Turkish second generation immigrants are far more
concordant to the Belgian emotional patterns than to the Turkish ones, whereas
Turkish first generation immigrants were equally concordant to both culture’s
emotional patterns. These results may mean that given contexts signal to the
immigrant the relevance of one type of emotional pattern over another: interac-
tions with Belgian colleagues at work signal the relevance of a Belgian pattern,
whereas interactions with Turkish family members at home signal applicability
of the Turkish pattern of emotional experience. Or put differently, immigrants’
emotional patterns may be maximally congruent with the pattern of emotions
that is most prevalent in that social context. Further research is needed to fully
understand how and when these contextual effects occur.
Emotional acculturation: a model for continued emotional change

Why would immigrants’ emotional concordance increase after they have spent
more time with majority members? Let’s go back to the example of Ayşe at the
beginning of the chapter, and let’s assume that the incident with the colleague took
place in a Belgian context. If Ayşe had experienced the situation in the Turkish
way—that is, if she had felt shame or guilt in addition to anger and, therefore,
had tempered her anger—she might not have been taken seriously. In the eyes of
her colleagues, her behavior would have been harder to understand than if she
responded with full-fledged anger—which is the Belgian pattern of experiencing
the situation. In other words, in a Belgian context, the Belgian emotional pattern
will serve Ayşe better than the Turkish emotional pattern, and is thus reinforced.
Moreover, repeated exposure to Belgian contexts will make the Belgian way
of emotional experience chronically accessible, at least in certain contexts. We
predict that, over time, and depending on the immigrants’ level of contact with
majority members, immigrants’ emotional experiences will converge with those
of members of the host culture in ways that maximize the fit of immigrants.
Our model of emotional acculturation can be taken as a model of emotional
change generally. It suggests that people’s emotional patterns shift in response to
changes in their socio-cultural context. Social circumstances may change, even
for non-immigrants: it is not uncommon to move between cities or neighbor-
hoods, to change jobs, or to start new relationships. We suggest that each of these
changes may stand for a shift in reinforcement structure or affordance, and thus
to (small) and, at times, incremental changes in emotional patterns. Consistent
with this idea, longitudinal research has yielded increased emotional concordance
in dyads and groups after they have spent time together (Anderson, Keltner,
and John, 2003; Totterdell, 2000). More generally, emotional change may thus
be a function of an individual’s socio-cultural contexts, and particularly their
reinforcement structures.
Our research on acculturation, therefore, suggests a model of emotional
change that is not unlike some learning models (e.g., Bouton, 2010). Each new
interaction or experience affords new ways of emotional responding. Changes
in emotional patterns throughout the life span are answers to changing require-
ments of a person’s (new) social environments. Consecutive interactions with
others may thus produce successive (not end-point-oriented) changes in emotions
(Saarni, 2008) that enhance their functionality to the particular social context.
References
Anderson, C., Keltner, D., and John, O. P. (2003). Emotional convergence between people
over Time. Journal of Personality and Social Psychology, 84, 1054–1068.
Berry, J. W., and Sam, D. L. (1997). Acculturation and adaptation. In J. W. Berry, M.
H. Segall, and C. Kagitcibasi (eds), Handbook of cross-cultural psychology (2nd edn,
Vol. 3, pp. 291–226). Needham Heights, MA: Allyn and Bacon.
Bouton, M. E. (2010). The multiple forms of ‘context’ in associative learning theory. In B.
Mesquita, L. Feldman Barrett, and E. S. Smith (eds), The mind in context (pp. 233–258).
New York, NY: Guilford Press.
De Leersnyder, J., Mesquita, B., and Kim, H. S. (2012, January). Emotional acculturation:
immigrants’ emotional fit to host and heritage cultural contexts. Poster session at the
Society for Personality and Social Psychology Thirteenth Annual Meeting. San Diego,
CA.
De Leersnyder, J., Mesquita, B., and Kim, H. (2011). Where do my emotions belong?
A study of immigrants’ emotional acculturation. Personality and Social Psychology
Bulletin,. 37, (4), 451–463.
Doost, H. T. N., Moradi, A. R., Taghavi, M. R., Yule, W., and Dalgeish, T. (1999). The
development of a corpus of emotional words produced by children and adolescents.
Personality and Individual Differences, 16, 322–340.
Markus, H. R., and Kitayama, S. (1994). The cultural construction of self and emotion:
Implications for social behavior. In S. Kitayama and H. R. Markus (eds), Emotion and
culture: Empirical studies of mutual influence (pp. 89–130). Washington, DC: American
Psychological Association.
Mesquita, B., and Leu, J. (2007). The cultural psychology of emotions. In S. Kitayama
and D. Cohen (eds), Handbook for cultural psychology (pp. 734–759). New York, NY:
Guilford.
Ryder, A., Alden, L., and Paulhus, D. (2000). Is acculturation unidimensional or bi-
dimensional? A head-to-head comparison in the prediction of personality, self-identity,
and adjustment. Journal of Personality and Social Psychology, 79, 49–65.
Saarni, C. (2008). The interface of emotional development with social context. In
M. Lewis, J. M. Haviland-Jones, and L. F. Barrett (eds), Handbook of emotions
(pp. 332–347). New York, NY: Guilford.
Tiedens, L. (2001). Anger and advancement versus sadness and subjugation: The effect
of negative emotion expressions on social status conferral. Journal of Personality and
Social Psychology, 80, 86–94.
Totterdell, P. (2000). Catching moods and hitting runs. Journal of Applied Psychology,
85, 848–859. Figure 1. Mean emotional concordance scores to the Belgian and Turkish
average emotional patterns, matched for type of situation.
Part 4
Emotional-Dynamics
Perspective
19 Emotion regulation
Two souls in one breast?
Nico H. Frijda
Amsterdam University
People and many animals manifest emotion regulation. The term denotes varia-
tions in strength or kind of emotional response which, supposedly, are not
primarily due to variations in strength or importance of the eliciting event.
A major problem in discussing emotion regulation is to distinguish emotion
and regulation (Gross, Sheppes, and Ury, 2011). What is it that is regulated, and
what is it that regulates? Are there two souls in one breast—an emotional mind
overlaid by an executive controller, an ego and an id?
I will present a perspective on emotion regulation under development, that is
a joint endeavour by Batja Mesquita and myself, and will argue the following
points.
First, emotions primarily are motive states, including states of motivational
loss. This will be argued in the section on “Emotion and Motivation”
Second, emotional events frequently simultaneously elicit multiple emotions
(or the foresight of such emotions). This will be argued in the second section
Third, interaction of multiple emotions may result in emotion.
Emotion and motivation

What are “emotions”? The word “emotion” refers to multicomponential response
patterns that include as components appraisal: the process of detecting emotional
meaning in information from an event; affect: pleasure or pain; arousal: physio-
logical reaction; motive state: readiness to act or not to act in a particular way;
and feeling: conscious awareness of any or all of the other four.
The component of motive state is the core of what emotions are for: to maintain
or modify a current person-object relationship. Motive states mobilize doing
things to profit from or deal with events that are appraised as relevant to the
individual’s interests and concerns (Frijda, 2007). Fear seeks to obtain safety by
escape from danger. Anger seeks to cause others to modify their actions so as to
leave one and one’s interests alone. Love incites togetherness and, in the event,
reproduction. Or the motive state reflects the appraised inability to diminish some
ongoing misery.
Emotions are motive states. That is: they represent “wanting” (Berridge, 1999).
They seek to achieve something that at that moment does not obtain, or that
138 Nico H. Frijda
one desires to continue. Or they manifest the definitive failure to achieve that
something.
Action readiness. I call these momentary motive states in emotions “states of
action readiness”, to emphasize that these motive states are not pure wishes. They
are states of wanting to change or continue a subject-object relationship. Modes
of action readiness include acceptance, attending, approach, affiliation, desire,
avoidance, submission, exuberance, hypoactivation, inhibition, and helplessness
(Davitz, 1969; Frijda 2007).
Wanting and striving (or explicitly not wanting or striving) are major character-
istics of emotional responding and experience. They manifest characteristics that
I call control precedence (Frijda, 2007). Emotional urges, thoughts, and actions
tend to persist in spite of obstacles, interruptions, and counteractions by the
target, for as long as the aims of the motive states have not been reached. Modes
of action readiness are inferences from the functional equivalence of expressive
behaviors manifest under similar circumstances (Frijda, 2007). They all further
the same motivational aim. When angry, one may adopt an intimidating threat
posture, utter threatening sounds, or attack the antagonist, which all share the
ability to frighten or hurt that antagonist.
States of action readiness are not just felt urges. They are activated repre-
sentations, or “mental structures” in the sense described by Jackendoff (2007):
structures of information in the mind/brain, that may (or may not) select and
activate action schemas. The structures contain a motivational aim, usually an
object or target, and an envisaged action outcome (Frijda and Parrott, 2011).
The mental structures form the informational content of the neural dispositions
activated in executing an action, imagining an action, and when perceiving
someone else executing a familiar action (Jeannerod, 2001).
Action readiness triggered by appraisal. Action readiness is elicited by
events or thoughts, as these are perceived and appraised (Frijda, 2007; Ellsworth
and Scherer, 2003). Appraisals generally emerge automatically, and can automati-
cally influence response activation. One need not be aware of the objects or
appraisals to incite affect, modifications of action readiness, and physiological
reactions (Bargh and Williams, 2007; Lambie and Marcel, 2002). Impressions
may make us happy or sad without us knowing why. Different states of
action readiness result from different appraisal structures. That at least is the
core hypothesis of appraisal theory, for which there is ample evidence (e.g.,
Ellsworth and Scherer, 2003; Frijda, 2007), and which seeks to account for the
coherences found between kinds of event (e.g., threats) and kinds of emotions
(e.g., fear).
Emotional events tend to elicit multiple emotions

But there is an important hitch in these relationships between appraisal and
action readiness. Emotional events rarely occur in isolation. They occur
within some context. They may impinge upon multiple concerns, and thereby
have multiple meanings. An event may represent a loss as well as a gain, a
Emotion regulation 139
challenge as well as a threat, touch upon achievement failure as well as loss of
self esteem.
That is to say that one and the same event may simultaneously give rise to
several different appraisals profiles. By consequence, it may elicit multiple
emotions at the same time.
“Multiple emotions” means that several modes of action readiness are evoked
at the same time. One is inclined to approach as well as to avoid; one is inclined
to hostility, and to implement tenderness.
Such a constellation is, in fact, quite common. Being treated unkindly by
one’s spouse may evoke anger. But one’s very anger may also risk hurting one’s
beloved spouse. At least, it threatens to spoil one’s evening together.
Multiple emotions may tend to cause emotion regulation

The why of emotion regulation. Emotion multiplicity, I propose, forms the major
cause of emotion regulation. It causes emotional response to a given event to
be different from what at least one of the part-appraisals could be expected
to evoke, had it occurred in isolation, because the multiple modes of action
readiness interact. They have to, since they share the same output channels. They
also often have to because the states of action readiness may be contradictory or
incompatible.
Analysis of the reasons for emotion regulation supports this perspective. Anger
is often held back by fear of retaliation, and by consideration for the target. By
contrast, anger emerges in full vigour when considered justified. One sticks to
one emotion; one makes no room for multiplicity. Likewise, risky behaviours
abound in adolescents, because novel powerful positive emotions elicited by
novel concerns turn fear of risk into appraisal of challenge
All this supports this conclusion: the motivations for emotion regulation are
themselves emotional. They often are as emotional as the regulated emotion
itself. This can be said more precisely. Emotions are regulated to the extent that
one cares about the implications of having an unregulated emotion.
This conception of emotion regulation abolishes the problematic distinction
between processes of emotion and processes of emotion regulation. The processes
that cause regulation—response attenuation, effortful response suppression,
avoiding to think of long-term consequences—are emotional processes. This
perspective fits with models of how prepotent instrumental responses are
controlled, as in go/no-go tasks, Stroop tasks, and antisaccade tasks: all “processes
in which we select the best response among the competing responses” (Curtis and
D’Esposito, 2009: 72).
The how of emotion regulation. What happens upon arousal of multiple
emotions depends on the strength relationships of the action tendencies involved,
and on whether their aims are compatible or incompatible. One may overshadow
the other, or weaken it. The processes involved can operate in an automatic,
non-deliberate way, and not guided by foresight. Major mechanisms that influence
those strength relationships include restriction of attentional range, and focusing
140 Nico H. Frijda
attention elsewhere. Emotions tend to restrict that range (Easterbrook, 1959).
Anxiety may trigger automatic inhibitory processes, like avoiding exploration of
cognitive contents (Derakshan, Eysenck, and Myers, 2007). By contrast, mecha-
nisms of focusing attention let that attention dwell on some content, facilitating
associational flow.
Another elementary mechanism involved in regulation is the capacity for not
acting and preparing to do so (Curtis and D’Esposito, 2009). One stiffens one’s
muscles, or attends elsewhere. This option permits postponing executing prepared
actions until the subject and the target are of the right nature or at the right
location or distance.
The actual emergence of multiple emotions can be forestalled by the acqui-
sition of action facilities that can satisfy multiple aims. Preparing not to act is
one of them, and it is in fact an important social skill. Knowing how to attenuate
one’s impact on others is another one. A careful approach satisfies the aims of
being careful as well as of approaching. Tactful or respectful speech satisfies the
aims of exerting control over other people’s actions as well as not hurting them.
Such multiple purpose action schemas belong to what is learned in socialization
(Campos, Frankel, and Camras, 2004), and they can be acquired on the spot by
sensitivity to the relevant appraisals.
Much emotion regulation thus appears to proceed without effort and without
regulatory intent. It results from automatic information processes for handling
multiple appraisals, and from interaction of simultaneous states of action
readiness. It uses mental and physical actions that are largely part of the individ-
ual’s standard action repertoires, as well as of one’s repertoire of well-learned
action modes that each represents some compromise or synthesis.
Emotion conflict
The domain of emotional reactions is, however, replete with contingencies that do
not allow such automatic handling of multiple emotions. There are various kinds.
In one, the urges of the multiple emotions are incompatible: one wants to delight
in drinking, and also to remain clear of mind and in good health. One wants both
equally badly, which leads to emotional conflict. There is an impasse with no
obvious route for escape.
In another kind, a strong urge for action can find no cue for any useful action. A
threat may be coming from anywhere: as when one is the target of aerial bombing,
subjected to an earthquake or tsunami, or living under a dictatorial regime. There
exist drastic automatic processes: transitions to altered consciousness. One gets
into a state of disorientation, numbness, or depersonalization: feeling as if what
happens does not really take place or happens to someone else (Hilgard, 1977).
Such changes are common when one’s car skids, and during torture (Frijda,
2010).
Other conflicts that have no automatic resolution are those between having to
face, and escaping from, powerful negative emotions. They are exemplified by the
conflict between seeking to keep one’s mouth shut during torture and betraying
one’s friends. Kuhl and Koole (2004) distinguished them as “self-maintenance”
from simple self-control.
Facing irresolvable motive state conflict has a further general issue: the
emergence of reflective conscious awareness. Conscious awareness of conflict,
according to Morsella’s Supramodular Interaction Theory (SIT; Morsella, 2005),
is the mechanism for integrating information from different supramodular
response systems, such as different high-level concerns. It leads to some form of
reflective stepping back, and planning deliberate emotion control.
Deliberate emotion regulation

Deliberate emotion regulation occurs when there is no emotion, but major possible
aversive event outcomes are known to the subject. This happens when emotional
appraisal is powerless because the aversive outcomes are far away in the future,
and event impact thus obeys the principles of time discounting (Ainslie, 2003).
Mere knowledge, however, still allows events and their implications to be
rationally deemed important. Both contingencies—that in which irresolvable
emotional conflict forces deliberation, and that in which it comes from mere
knowledge of possible aversive impact—form the domain to which the currently
dominant dual process conception of emotion regulation applies (e.g., Gross
and Thompson, 2007). “Executive functions” (Zelazo and Cunningham, 2007)
are engaged, in line with dual process conceptions of cognition, motivation, and
behaviour determination in general (e.g., Strack and Deutsch, 2004).
Conclusions
My main point: the present perspective largely denies the specificity of regulation
processes. Emotion processes produce the phenomena of emotion regulation:
inhibition, attenuation, enhancement, emotion transformation. The processes of
emotion and emotion regulation do not differ. The phenomena of regulation stem
from the interaction of concurrent multiple emotions.
The present perspective emphatically widens the view of what determines
emotion regulation. It does not emphasize the role of social norms, cultural
prescriptions, and interpersonal processes. It suggests that regulation results from
any interaction between concurrent multiple emotions, and thus from any multi-
plicity of meaning of emotion-arousing events.
The present perspective helps to lay to rest what De Waal (2005) has called
the “veneer theory” of human morality, according to which human reflectivity
overlays primitive animal mentality. Human morality employs evolutionary
developments of emotions and emotional sensitivities that make use of emotional
endowments in other primates.
In the title of this paper, I ask: “Emotion regulation: two souls in one breast?”
My answer is: “no”. But there can be as many souls in one breast as there can be
simultaneous motive states.
142 Nico H. Frijda
References
Ainslie, G. (2001). Breakdown of will. Cambridge: Cambridge University Press.
Bargh, J. A. and Williams, L. E. (2007). The nonconscious regulation of emotion. In J.
J. Gross (ed.), Handbook of emotion regulation (pp. 429–445). New York: Guilford
Press.
Baumeister, R. F., Muraven, M., and Tice, D. (2000). Ego depletion: A resource model
of volition, self-regulation, and controlled processing. Social Cognition, 18, 130–150.
Baumeister, R. F., Zell, A. I., and Tice, D. (2007). How emotions facilitate and impair
self-regulation. In J. J. Gross (ed.). Handbook of emotion regulation (pp. 408–428). New
York: Guilford Press.
Berridge, K. C. (1999). Pleasure, pain, desire, and dread: Hidden core processes of
emotion. In D. Kahneman, E. Diener, and N. Schwarz (eds), Foundations of hedonic
psychology: Scientific perspectives on enjoyment and suffering. (pp. 525–557). New
York: Sage.
Bühler, K. (1934). Ausdruckstheorie. Jena: Fisher.
Campos, J. J., Frankel, C. B., and Camras, L. (2004). On the nature of emotion regulation.
Child Development, 75, 377–394.
Curtis, C. E. and D’Esposito, M. (2009). The inhibition of unwanted actions. In
E. Morsella, J. A. Bargh and P. M. Gollwitzer (eds), Oxford Handbook of Human Action
(pp. 72–87). Oxford: Oxford University Press.
Davitz, J. R. (1969). The language of emotion. New York: Academic Press.
De Waal, F. B. M. (2005). Our inner ape: A leading primatologist explains why we are who
we are. London: Granta Books.
Derakshan, N., Eysenck, M. B., and Myers, L. B. (2007). Emotional information
processing in repressors. Cognition and Emotion, 21, 1585–1614.
Easterbrook, J. A. (1959). The effects of emotion on cue utilization and the organization
of behavior. Psychological Review, 66, 183–201.
Ellsworth, P. C. and Scherer, K. R. (2003). Appraisal processes in emotion. In
R. Davidson, K. R. Scherer, and H. H. Goldsmith (eds), Handbook of the affective
sciences (pp. 572–596). Mahwah: Erlbaum.
Frijda, N. H. (2007). The laws of emotion. Mahwah: Erlbaum.
—(2010). Not passion’s slave. Emotion Review, 2, 68–75.
Frijda, N. H. and Parrott, W. G. (2011). Basic emotions or ur-emotions? Emotion Review,
3, 406–415.
Gross, J. J. and Thompson, R. A. (2007). Emotion regulation: Conceptual foundations. In
J. J. Gross (ed.), Handbook of emotion regulation. New York: Guilford Press.
Gross, J. J., Sheppes, G. and Urry, H. L. (2011). Emotion generation and emotion
regulation: A distinction we should make (carefully). Cognition and Emotion, 25,
765–781.
Hilgard, E. R. (1977). Divided consciousness: Multiple controls in human thought and
action. New York, Wiley.
Jackendoff, R. (2007). Language, consciousness, culture: Essays on mental structure.
Cambridge, MA: MIT Press.
Jeannerod, M. (2001). Neural simulation of action: A unifying mechanism for motor
cognition. NeuroImage, 14, S103–S109.
Kuhl, J. and Koole. S. L. (2004). Workings of the will: A functional approach. In
J. Greenberg, S. L. Koole, and T. Pyszczynski (eds), Handbook of experimental
existential psychology (pp. 411–430). New York: Guilford.
Lambie, J. A. and Marcel, A. (2002). Consciousness and emotion experience: A theoretical
framework. Psychological Review, 109, 219–259.
Morsella, E. (2005). The function of phenomenal states: Supramodal interaction theory.
Psychological Review, 112, 1000–1021.
Strack, F. and Deutsch. R. (2004). Reflective and impulsive determinants of social
behavior. Personality and Social Psychology Review, 8, 220–247.
Zelazo, P. D. and Cunningham, W. A. (2007). Executive function: Mechanisms underlying
emotion regulation. In J. J. Gross (ed.), Handbook of emotion regulation (pp. 135–158).
New York: Guilford Press.
20 Understanding emotion change
requires an understanding of emotion
causation
Agnes Moors
Ghent University
One aim of this chapter is to argue that research about the effectiveness of
strategies for emotion change should be conducted in a (more) theory-driven
manner. Another aim is to present a two-stage model for thinking about controlled
emotion change. The chapter is structured as follows: After delineating the
concepts of emotion and emotion change, I clarify the relation between emotion
causation and controlled emotion change, and I present a two-stage model for
controlled emotion change. Next, I consider the predictions of two theories of
emotion causation about the two stages of emotion change.
Delineating concepts
Many emotion theorists agree that emotions are collections of components.
These are changes in organismic subsystems. Examples are (a) a cognitive
component with changes in appraisal, (b) a motivational component with
changes in action tendencies (e.g., the tendency to increase contact), (c) a somatic
component with changes in (central and peripheral) physiological activity, (d) a
motor component with changes in gross behavior and facial and vocal expres-
sions, and (e) a subjective component with changes in experience or feelings.
Theories of emotion causation have hypotheses about how these changes come
about.
Emotion change refers to cases in which one or more components in
the emotional episode change with regard to their intensity (from strong to
weak or vice versa) or quality (e.g., from an appraisal of low to high coping
potential, from the tendency to fight to the tendency to flee, from a feeling of
sadness to a feeling of happiness). Given that components already consist of
changes, a change in a component comes down to a change in a change, or a
second change after a first change. Processes involved in the first change may
to some extent also be involved in the second change. In other words, theories
of emotion causation can be useful for understanding certain cases of emotion
change.
Understanding emotion change requires an understanding 145
Two-stage model of controlled emotion change
Emotion change can be controlled or spontaneous. It is controlled when it is
caused by the goal to change the emotion. It is spontaneous or uncontrolled when
it is not caused by this goal. I focus on controlled emotion change1. I consider
cases in which people want to change their emotion. Given that emotion is a
multi-componential phenomenon, it is possible to split the goal to change the
emotion into the more specific goals to change the components. A person may
want to change her feelings, her behavior, or her physiological responses. In the
present chapter, I focus on the case in which a person has the goal to change her
feelings. This goal may be achieved by applying various strategies.
The literature contains various strategies for emotion change. I propose organ-
izing strategies according to their targets and I consider as targets each of the
components in the emotional episode plus the stimulus2. Examples of strategies
that target the stimulus are problem solving (trying to change the stimulus),
and withdrawal or distraction (moving away from the stimulus in reality or in
spirit). A strategy that targets appraisal of the stimulus is called reappraisal. This
consists of changing the values of appraisal variables like goal relevance (e.g.,
by changing goal priorities) and coping potential (e.g., from low to high). Other
strategies target action tendencies or expressive behavior such as acting happy or
suppressing a grimace. Still other strategies target physiological responses (e.g.,
taking a pill). A final set of strategies targets feelings directly (e.g., suppression
or denial of angry feelings).
A strategy can be considered as a type of goal, more in particular, a goal that
is subordinate to the overarching goal to change one of the components. If a
person has the overarching goal to change the feeling component, she may turn to
various strategies to achieve this goal. She can try to change her feelings directly,
but she can also try to change them indirectly by changing the stimulus, her
appraisal, her behavior, or her physiological state. In sum, in many cases, emotion
change will follow two steps: one step in which the person tries to change the
component figuring in the chosen strategy and another step in which this change
influences the component that the person ultimately wants to change.
1 The notion of controlled emotion change has overlap with the notion of emotion regulation.
Emotion regulation is broader because it includes attempts to continue an emotion in addition to
attempts to change it. The present paper focuses on emotion change (consistent with the title of this
volume) but most insights apply to emotion regulation as well. I further wish to note that controlled
emotion change can be conscious or unconscious: It can be caused by a conscious or unconscious
goal.
2 My proposal to organize strategies according to components is similar to Gross and Thompson’s
(2007) proposal to organize strategies according to the stage in the emotion process. Yet, my
proposal does not presuppose that the components follow a fixed order. It is thus suitable to accom-
modate predictions from various emotion theories (not only those that assume a fixed order).
1
146 Agnes Moors
strategy stimulus
to change change
stimulus
strategy to appraisal
change change
appraisal
feeling
goal to change
change strategy to behavior
feelings change change
behavior
strategy to physiological
change response change
physiological
responses
Figure 6 Direct and indirect routes for changing feelings
Effectiveness of strategies for emotion change: theoretical predictions

Based on the two-stage model of emotion change sketched above, the effec-
tiveness of strategies for emotion change is (often) dependent on two elements.
First, it is dependent on the direct controllability of the component that is targeted
in the strategy (grey arrows in Figure 6). Second, it is dependent on whether a
change in the component that is targeted in the strategy has an influence on the
component that the person ultimately wants to change (black arrows in Figure
6). Different theories of emotion causation have different assumptions about
which components can influence each other. Thus, they are likely to make
different predictions about which strategies will be effective. To exemplify this, I
consider the predictions of two theories of emotion causation: appraisal theories
and network theories. After a brief description of these theories, I examine their
assumptions about the components that influence the feeling component (cf.
black arrows). After that, I examine their assumptions about the direct control-
lability of the components targeted in various strategies (cf. grey arrows).
Appraisal theories and network theories

According to appraisal theories (e.g., Lazarus, 1991; Scherer, 1984), stimuli
are evaluated on a number of appraisal values, such as goal relevance, goal
congruence, and coping potential. For example, encountering a robber may be
appraised as goal relevant, goal-incongruent, and difficult to cope with. Appraisal
is presented as a typical cause of the other components: Changes in appraisal
drive changes in action tendencies. These are manifested in physiological
changes that prepare and support changes in actual behavior. Aspects of all these
components are reflected in the content of the feeling component. Contemporary
appraisal theories allow that later components feed back to earlier ones. This is
called recurrence.
Many appraisal theorists endorse a dual-mode view, arguing that appraisal
comes in two forms: one is rule-based; the other is associative. In the rule-based
variant, appraisal values are computed for each appraisal variable and these are
combined into a pattern. In the associative variant, the stimulus activates a previ-
ously computed appraisal pattern that was stored in memory. It is often assumed
(but not tested) that rule-based appraisal is flexible but non-automatic and
preferably operates on verbal-like representations whereas associative appraisal
is more rigid but automatic and preferably operates on image-like representations.
Network theories of emotion (e.g., Bower, 1981; Leventhal, 1984) do not put
forward a particular causal order for the components in the emotional episode.
When an emotion occurs, a representation of the emotion is stored in memory as a
network. This network consists of representations (nodes) of stimuli and compo-
nents. Each specific emotion (e.g., anger, sadness, fear) has a separate network.
Activation of an emotion network is the principal cause of emotions. Activation
can happen via different routes: via stimuli that are identical or similar to stored
ones or via any of the components that are part of the network. For example,
anger can be elicited by stimuli that elicited anger in the past, but also by fighting
or frowning. Activation of one node in the network spreads to other nodes. When
activation in a node reaches a threshold, it triggers the actual occurrence of the
corresponding component.
Network theories also endorse a dual-mode view. They consider the activation
of associations as the primary cause of emotions, but they leave room for rule-
based computation (often situated in the appraisal node). Importantly, both
network theorists and appraisal theorists have argued that the associative process
is the preferred route for emotion causation. It is not always clear why. One
proposal is that associative processes are automatic and therefore difficult to stop.
Another proposal is that associative processes operate on image-like representa-
tions, which are more vivid than verbal-like ones (Frijda, 1988).
Components that influence the feeling component (black arrows)

In appraisal theories, the process responsible for emotion elicitation is appraisal.
Therefore, the strategy of choice should be reappraisal. Strategies that target
the stimulus should also be effective because they influence appraisal. But what
about strategies that target the somatic and motor components, such as taking
a pill or putting on a happy face? Both components are also reflected in the
feeling component. Hence, a change in these components can be expected to
also color feelings. Unsolved questions are whether these other components
add to the quality or only to the intensity of feelings, and if they add to the
quality, whether artificially changing these other components can override the
quality of the feelings as determined by appraisal. Suppose a person encounters
148 Agnes Moors
a goal-incongruent event but puts on a happy face. Is this enough to turn a sad
feeling into a happy one?
As mentioned, network theories do not favor a particular causal order of
components. An emotional network can be activated via different routes: via
stimuli or via any of the components. Activation spreads from one node to all
the other nodes. As such, a change in any component can produce a change in
feelings. These changes go beyond changes in intensity, given that there is a
separate network for each specific emotion. For example, if a person changes her
facial expression from sad to happy, this should activate the happiness network
including the node for happy feelings. Activation of the new emotion probably
has to compete with activation of the old emotion. Putting on a happy face may
not lead to happy feelings if the sadness network is activated more strongly.
Direct controllability of components (grey arrows)

(1) Directly changing one’s physiological responses is virtually impossible, but
they can be changed indirectly via behavior (e.g., taking a pill, doing relaxation
exercises) (2) Changing behavior seems easy, but may be difficult when one has
a strong emotion. According to appraisal theory, emotional behavior is caused by
an action tendency, which is a kind of goal. For example, fighting is caused by
the goal to fight. The goal to change one’s emotional behavior has to compete
with the goal that is currently active. If a person wants to hug instead of fight,
the goal to hug must be strong enough to override the goal to fight. (3) Changing
one’s feelings directly seems difficult. This makes sense to appraisal theories
because they portray feelings as the reflection of all the other components in
consciousness. Choosing one’s feelings is as difficult as choosing any other
content of consciousness. (4) What about changing appraisals? As mentioned,
appraisal can be rule-based or associative, and the associative variant is supposed
to be automatic. The associative variant may thus be difficult to control. Add to
this that associative appraisal is the preferred route for emotion causation and thus
the preferred route for emotion change. Hence, appraisal theories do not expect
reappraisal to be always effective or to be easy.
Effectiveness of strategies for emotion change: empirical data

Below, I discuss empirical data for the effectiveness of two strategies for changing
feelings: changing facial expressions and reappraisal.
Strategy of changing facial expressions

In a study by Gross and Levenson,(1997) participants watched a sad, amusing,
and neutral film. One group was instructed to suppress their facial expressions
whereas another group was not. Facial expressions were weaker in the first
than the second group, but in the first group, they were stronger during sad and
amusing films than during neutral films. This indicates that facial expressions
can be changed only partially. Suppressing facial expressions had some effect
on positive but not on negative feelings. It may be noted that most studies that
do report changes in feelings due to changes in expressions (Strack, Martin, and
Stepper, 1988) show changes in intensity, not quality, and it is often not certain
that the changes circumvented reappraisal.
Strategy of reappraisal
Several laboratory studies show that reappraisal has an influence on feelings. In a
typical study (e.g., Ochsner, Bunge, Gross, and Gabrieli, 2002) participants watch
pictures of problematic situations (e.g., a sick child). One group is instructed to
reappraise the stimuli by imagining that the problem will be solved (e.g., the
child will recover) whereas the other group is not. The first group reports less
negative feelings than the second group. On the other hand, there is the clinical
observation that reappraisal does not always work, or often does not work. For
example, spider phobics often claim they know that spiders are not dangerous, but
they nevertheless feel afraid. The dual process assumptions in appraisal theories
and network theories provide hints to solve this seeming contradiction.
Both theories hold that emotions are preferably elicited by an associative
process. To the question why this is the case, I mentioned one proposal that
refers to the automatic nature of associative processes and another proposal that
refers to the image-like format of the representations in associative processes.
I apply both proposals to the contradictory findings, starting with the first
proposal. In the clinical case, the spider is appraised as dangerous on the basis
of an associative process. New information about the spider, that she is safe,
must initially be processed via a rule-based process. Associative processes are
said to be more automatic than rule-based ones. This means that they are faster
to influence feelings than rule-based ones. Thus, the old information that spiders
are dangerous is faster to influence feelings than the new information that spiders
are safe. In the laboratory study with the pictures, participants do not have
strong prior associations between the pictures and certain appraisals. Therefore,
the appraisals they make do not have to compete with automatic appraisals to
influence feelings. Based on the first proposal, a therapy for spider phobics
consists in installing a new association between spider and the appraisal “safe”
and to train it so much that it becomes more automatic than the old association
between spider and “dangerous”.
The second proposal is that rationally acknowledging that the spider is not
dangerous is not enough; the person has to experience it—as in exposure therapy.
This proposal assigns a crucial role to the format of the representations involved.
Representations must be in image-like, not verbal-like format. A related proposal
is that it is necessary to change specific instead of abstract knowledge (Moberly
and Watkins, 2006).
150 Agnes Moors
Conclusion
Theories of emotion causation can improve our understanding of emotion change.
They can guide research and help make sense of mixed results about the effec-
tiveness of strategies for emotion change. Turning it around, research on the
effectiveness of strategies for emotion change can improve our understanding of
emotion causation. It can help evaluate these theories. Caution is due, however. If
one wants to use results about the effectiveness of strategies for emotion change
to draw conclusions about the plausibility of these theories, one has to take into
account not only the hypothesized influences among components (black arrows),
but also the assumptions that theories endorse about the direct controllability of
the components targeted in the strategies (grey arrows).
Acknowledgements
Preparation of this chapter was supported by Methusalem Grant BOF09/01M00209
of Ghent University.
References
Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129–148.
Frijda, N. H. (1988). The laws of emotion. American Psychologist, 43, 349–358.
Gross, J. J., and Thompson, R. A. (2007). Emotion regulation: Conceptual foundations. In
J. J. Gross (ed.), Handbook of emotion regulation (pp. 3–24). New York, NY: Guilford
Press.
Gross, J. J., and Levenson, R. W. (1997). Hiding feelings: The acute effects of inhibiting
positive and negative emotions. Journal of Abnormal Psychology, 106, 95–103.
Lazarus, R. S. (1991). Emotion and adaptation. New York: Oxford University Press.
Leventhal, H. (1984). A perceptual-motor theory of emotion. In L. Berkowitz (ed.),
Advances in experimental social psychology (Vol. 17, pp. 117–182). New York:
Academic Press.
Moors, A. (2009). Theories of emotion causation: A review. Cognition and Emotion, 23,
625–662.
Moberly, N., and Watkins, E. (2006). Processing mode influences the relationship between
trait rumination and emotional vulnerability. Behavior Therapy, 37, 281–291.
Ochsner, K. N., Bunge, S. A., Gross, J. J., and Gabrieli, J. D. (2002). Rethinking
feelings: An FMRI study of the cognitive regulation of emotion. Jounal of Cognitive
Scherer, K. R. (2005). What are emotions? And how can they be measured? Social Science
Information, 44, 695–729.
Strack, F., Martin, L. L., and Stepper, S. (1988). Inhibiting and facilitating conditions
of human smile: A nonobtrusive test of the facial feedback hypothesis. Journal of
Personality and Social Psychology, 54, 768–777.
21 Learning to self-generate
positive emotions
Barbara L. Fredrickson
University of North Carolina at Chapel Hill
People’s daily lives can brim with emotional upsets. Unforeseen obstacles,
incivilities, rebuffs, insults, and arguments abound. These and other upsets often
ignite the pain of anger, anxiety, or sadness, with their attendant downward spirals
and destructive behaviors.
Although the link between negative events and negative emotions can seem
automatic and altogether inescapable, people do have enormous choice in how
they respond to the slings and arrows of daily life. What does it take for people to
experience these and other upsets without inner turmoil or outer destructiveness?
Is it possible?
Indeed it is possible. What it takes is the ability to regulate attention and
cognition in the service of self-generating inner states that are more open and
optimistic—states that till the soil for positive emotions to take root, ranging
from serenity and inspiration, to joy, gratitude, and more. This chapter, grounded
in empirical evidence, outlines why these skills are important, and how they can
be learned.
The beneficial correlates of positive emotions

People who experience positive emotions more frequently than others seem to
have it made. Scientific evidence documents that they are more resilient to life’s
adversities, more socially connected, and more successful in their personal and
work lives. They’re also healthier, with lower rates of hypertension and cardio-
vascular disease. They even fall prey to fewer viruses, like the common cold.
Plus, they live longer, up to ten years longer than those who experience the least
positive emotions (for an accessible review, see Fredrickson, 2009).
To the extent that people assume that trait positive affect is an immutable,
biological given, encountering research findings like these can be disheart-
ening. Still, there is reason for hope. Although heritable, evidence suggests that
only about 50 per cent of individual differences in trait affect are genetically-
determined, with the remaining 50 per cent reflecting a combination of life
circumstances and daily habits (Lybomirsky, Sheldon and Schkade, 2005), which
can and do change, especially with age. Indeed, age-related changes in emotions
are striking, with younger adults favoring negatively-valenced information and
152 Barbara L. Fredrickson
older adults favoring the positive (see Chapter 5). Key to this wisdom of old age
appears to be the ability to flexibly self-generate positive emotions.
More than a decade of empirical work on the broaden-and-build theory of
positive emotions (Fredrickson, 1998, 2001, 2009) casts positive emotions as key
engines of personal growth and resilience, not simply as the pleasant products
of them. In this chapter, I describe how positive emotions can widen one’s
perspective on life and build personal resources like mindfulness and the ability
to connect with others. Through incremental broaden-and-build processes, micro-
moments of positive emotions grow into stable affective dispositions that foster
people’s physical health and render their lives more satisfying.
Positive emotions as means, not ends
Positive emotions open minds

The first tenet of the broaden-and-build theory is that positive emotions
expand people’s awareness, temporarily allowing them to take in more of their
surrounding contextual information than they do during neutral or negative states
(Fredrickson, 1998, 2001). This momentary cognitive effect of positive emotions
has been demonstrated in a wide range of tightly controlled experiments carried
out in multiple laboratories. For instance, experimentally induced positive
emotions have been shown to broaden the scope of people’s visual attention in
behavioral tests (Fredrickson and Branigan, 2005), including tests that measure
fine-grained behavioral responses using milliseconds of reaction times (Rowe,
Hirsch, and Anderson, 2007) and eye-tracking technology (Wadlinger and
Isaacowitz, 2006). Moreover, experiments with brain imaging (e.g., fMRI) reveal
that positive emotions expand people’s field of view at very early perceptual
encoding stages (Schmitz, De Rosa and Anderson, 2009). Positive emotions,
then, quite literally widen people’s outlook on the world around them. (See Gable
and Harmon-Jones, 2008, for a contrasting view for approach-motivated positive
affect.)
Although the expansion of awareness that comes with positive emotions
is as subtle and as short-lived as the emotion itself, it accounts for positivity-
related increases in creativity (Rowe et al., 2007), and may well account for
the documented benefits of positive emotions for autobiographical memory,
integrative decision-making, test and work performance, coping and resilience,
interpersonal trust, social connection, teamwork, and negotiation ability (for
a review, see Fredrickson, 2009). In short, open and flexible awareness and
thinking are core attributes of positive emotional states.
Positive emotions transform lives

The second tenet of the broaden-and-build theory is that, over time, the momentary
states of expanded awareness sparked by positive emotions accumulate and
compound to build durable personal and social resources that ultimately reshape
Learning to self-generate positive emotions 153
people’s lives for the better (Fredrickson, 1998, 2001, 2009). This means that
people who learn skills to self-generate positive emotions—which in turn allow
them to increase their daily diets of positive emotions—build resources and
resilience that help to minimize future suffering and cultivate future health
and well-being. Recent randomized controlled trials have tested the effects of
learning loving-kindness meditation (LKM) as a means to self-generate positive
emotions more frequently (for more information on this ancient Buddhist
mind-training practice, see Salzberg, 1997). Results indicate that, relative to
a monitoring waitlist control group, LKM practice reliably elevates positive
emotions (Fredrickson, Cohn, Coffey, Pek and Finkel, 2008; see next section for
more details). Most importantly, however, the upward shift in positive emotions
evident in people practicing loving-kindness meditation also increases their
personal resources, including their mindfulness, their environmental mastery,
their positive relations with others, and their self-reported health. In turn, these
increased resources account for reduced depressive symptoms and improved life
satisfaction (Fredrickson et al., 2008). As LKM increases daily positive emotions,
it has also been shown to increase cardiac vagal tone (Kok et al., in press), a
marker of both physical health and behavioral flexibility (Thayer and Sternberg,
2006). This nascent research on the long-range health and psychological benefits
of cultivating positive emotional states provides a compelling rationale to
consider the value of learning to self-generate positive emotions in daily life.
Evidence that trait positive affect can change

Although the randomized controlled trials of LKM, described above (Fredrickson
et al., 2008; Kok et al., in press), were designed as tests of the broaden-and-
build theory, the resulting data can also be examined as evidence that people
can elevate their own levels of trait positive affect. Participants in these studies
reported on nine distinct positive emotions each day for nine weeks (i.e.,
amusement, awe, contentment, gratitude, hope, interest, joy, love, and pride). For
each of these distinct emotions, statistical analyses revealed a pattern of steady
increase, indicated by significant Experimental Condition X Week interactions
(Fredrickson et al., 2008, Table 2). Underscoring the importance of individual
effort to self-generate more frequent positive emotions, the data also revealed
that above and beyond the effects of random assignment to Experimental
Condition, the amount of time study participants devoted to meditation practice
each day also significantly predicted their experience of daily positive emotions
(Fredrickson et al., 2008, Table 3). Most strikingly, further analyses revealed that
the dose-response relationship between time spent meditating and its positive
emotion yield tripled over the 9-week study, a pattern of results that resoundingly
defies a pattern of hedonic adaptation.
Another way to probe the effects of learning to self-generate positive emotions
via LKM is to examine emotions experienced during an ordinary work day a
few weeks after the completion of LKM training. Using the Day Reconstruction
Method to do so, we found evidence for four independent effects: first, that
meditation produces positive emotions during meditation practice; second, that
these positive emotions persist after meditation episodes end; third, that, over
time, repeated LKM practice produces a cumulative increase in positive emotions
that appears on subsequent days, whether or not individuals meditate on that day;
and fourth, that the biggest LKM-related boosts in positive emotion occur when
people are interacting with others.
We also retested these same participants 15 months later and learned that 35%
of them continued to mediate regularly. Compared to those who did not continue
meditating, the daily positive emotions remained elevated among those who
continued to meditate (Cohn and Fredrickson, 2010).
In a second randomized controlled trial of LKM, we obtained similar evidence
for people’s ability to elevate their own trait positive affect, together with
evidence suggesting how these effects may emerge. Specifically, we discovered
that the positive emotions produced by LKM practice also increase resting levels
of vagal tone (Kok et al., in press). Given that past research has linked high vagal
tone to superior attention and emotion regulation (Thayer and Sternberg, 2006),
change in this trait-like autonomic resource may support people’s efforts to self-
generate positive emotions in daily life.
Taken as a whole, the pervasiveness and durability of these effects suggest that
enduring upward shifts in positive feelings are indeed possible, especially among
those who invest more time in their meditation practice (Fredrickson et al., 2008;
Cohn and Fredrickson, 2010) and who show increases in cardiac vagal tone (Kok
et al., in press).
Upward spirals counter downward spirals

Because both positive and negative emotions alter people’s attention, thinking,
motivation, and behavior, they also trigger self-perpetuating dynamics—or
spirals—that can either drag people down or buoy them up. To illustrate, the
negative emotions of anger, stress, or sadness each narrow people’s attention
and reinforce emotion-consistent appraisal patterns (e.g., blame, threat, or loss,
respectively) that initiate further bouts of anger, stress, or sadness, with attendant
social friction or isolation. These cycles perpetuate themselves to produce the
downward spirals all too familiar to clinical psychologists.
The broaden-and-build theory holds that positive emotions create opposing
upward spiral dynamics, in which the broadened awareness that accompanies
positive emotions allows people to step back or “decenter” from stressful circum-
stances and appraise them in a more positive light, which in turn can trigger
further experiences of positive emotions. As this upward spiral unfolds, it creates
resilience, well-being, and greater opportunities for social connection. A number
of prospective studies have now documented this upward spiral dynamic and
my collaborators and I have recently outlined how upward spirals might drive
neuroplasticity in ways that can be productively applied within psychotherapy
(Garland, et al., 2010).
Learning to self-generate positive emotions 155
Using the broaden-and-build theory to change emotional habits
The broaden-and-build theory originated to explain how positive emotions were
shaped by the forces of natural selection. The key is that, over time and through
repeated experiences, these fleeting pleasant states augmented our human ancestors’
resources for survival. Although the theory has been tested primarily in healthy
populations with typical life stressors, more recently, clinical scientists have created
clinical applications of the theory, targeting a range of psychological disorders
characterized by emotion dysfunctions and deficits, such as depression, anxiety,
and schizophrenia (for a review, see Garland et al., 2010). For instance, promising
results have emerged from a pilot test that used LKM to unlock more frequent
self-generated positive emotions as a means to treat the negative symptoms of
schizophrenia, which include anhedonia (diminished pleasure), avolition (dimin-
ished motivation), asociality (diminished desire for interpersonal relationships),
alogia (diminished speech), and blunted affect (diminished expression of emotion)
(Johnson, et al., 2011). Likewise, initial evidence suggests that depression and
anxiety disorders can be successfully treated using adaptations of cognitive behav-
ioral therapy that more explicitly cultivate positive emotions, either through mental
imagery (Tarrier, 2009) or positive reappraisals (Garland, Gaylord and Park, 2009).
Closing sentiments
When people come to understand how positive emotions work—how they open
minds, transform futures, and create uplifting upward spiral dynamics—they are
more likely to see the wisdom of cultivating these heartfelt momentary experi-
ences more frequently. Seen from the perspective of the broaden-and-build
theory, unlocking more momentary experiences of positive emotions is not
simply the end-goal of a desire to feel good, but rather, doing so is an important
vehicle for reshaping people’s abiding levels of resilience, health, and well-being,
as well as a host of other resources and personality traits that make life more
satisfying and meaningful. In short, trait affect can change: people can learn to
self-generate more frequent positive emotions, which can have sweeping reper-
cussions throughout their lives as a whole.
References
Cohn, M. A., and Fredrickson, B. L. (2010). In search of durable positive psychology inter-
ventions: Predictors and consequences of long-term positive behavior change. Journal
of Positive Psychology, 5, 355–366.
Fredrickson, B. L. (1998). What good are positive emotions? Review of General
—(2001). The role of positive emotions in positive psychology: The broaden-and-build
theory of positive emotions. American Psychologist, 56, 218–226.
—(2009). Positivity. New York: Three Rivers Press.
Fredrickson, B. L., and Branigan, C. (2005). Positive emotions broaden the scope of
attention and thought-action repertoires. Cognition and Emotion, 19, 313–332.
Fredrickson, B. L., Cohn, M. A., Coffey, K. A., Pek, J. and Finkel, S. M. (2008). Open
hearts build lives: Positive emotions, induced through loving-kindness meditation, build
consequential personal resources. Journal of Personality and Social Psychology, 95,
1045–1062.
Gable, P. A., and Harmon-Jones, E. (2008). Approach-motivated positive affect reduces
breadth of attention. Psychological Science, 19, 476–482.
Garland, E. L., Fredrickson, B. L., Kring, A. M., Johnson, D. P., Meyer, P. S., and Penn, D.
L. (2010). Upward spirals of positive emotions counter downward spirals of negativity:
Insights from the broaden-and-build theory and affective neuroscience on the treatment
of emotion dysfunctions and deficits in psychopathology. Clinical Psychology Review,
30, 849–864.
Garland, E. L., Gaylord, S., and Park, J. (2009). The role of mindfulness in positive
reappraisal. Explore, 5, 37–44.
Johnson, D. P., Penn, D. L., Fredrickson, B. L., Kring, A. M., Meyer, P. S., and Brantley,
M. (2011). A pilot study of loving-kindness meditation for the negative symptoms of
schizophrenia. Schizophrenia Research, 129, 137–140.
Kok, B. E., Coffey, K. A., Cohn, M. A., Catalino, L. I., Vacharkulksemsuk, T., Algoe, S. B.,
Brantley, M., and Fredrickson, B. L. (in press). How positive emotions build physical
health: Perceived positive social connections account for the upward spiral between
positive emotions and vagal tone. Psychological Science.
Lybomirsky, S., Sheldon, K. M., and Schkade, D. (2005). Pursuing happiness: The archi-
tecture of sustainable change. Review of General Psychology, 9, 111–131.
Rowe, G., Hirsh, J. B., and Anderson, A. K. (2007). Positive affect increases the breadth
of attentional selection. Proceedings of the National Academy of Sciences USA, 104,
383–388.
Salzberg, S. (1997). Lovingkindness: The revolutionary art of happiness. Shambhala
Publications.
Schmitz, T. W., De Rosa, E., and Anderson, A. K. (2009). Opposing influences of affective
state valence on visual cortical encoding. Journal of Neuroscience, 29, 7199–7207.
Tarrier, N. (2010). Broad minded affective coping (BMAC): A ‘positive’ CBT approach
to facilitating positive emotions. International Journal of Cognitive Therapy, 3, 64–76.
Thayer, J. F. and Sternberg, E. (2006). Beyond heart rate variability: Vagal regulation of
allostatic systems. Annals of New York Academy of Sciences, 1088, 361–372.
Wadlinger, H. A., and Isaacowitz, D. M. (2006). Positive mood broadens visual attention
to positive stimuli. Motivation and Emotion, 30, 87–99.
22 The role of control in emotion,
emotion regulation, and empathy
Kevin Ochsner
Columbia University
The ability to control the contents of our mind, and how those contents lead
to behavior, is required in virtually every sphere of life. In this chapter I focus
on two that involve emotion: on one hand, the use of control to regulate the
experience and expression of our emotions—thereby enabling us to change what
we feel—and on the other hand, the use of controlled processes to help us make
sense of the emotions of others—thereby enabling us to change our perceptions
of what others feel.
The starting premise is that neuroscience data can usefully inform knowledge
of the mechanisms underlying these two uses of control. With that in mind, the
chapter is divided into four parts. In the first, I sketch a current conception of
how we exert control over our behavior and the neural systems that make this
possible. In the second and third sections I apply these conceptions first to the
study of emotion regulation, and second to the study of empathy. The final section
considers the implications of this work for various areas of psychology.
Of two minds
Among the first questions to arise with respect to how we control our selves
are: why is it so hard? And how does it work when it’s effective? While these
questions have been of long-standing interest to both lay and scientific audiences,
contemporary psychology and neuroscience have begun to offer an intriguing
two-part answer.
The first part concerns the basic psychological and neural processes that govern
our behavior. Decades of work in experimental psychology and neuroscience
have made clear that we are “of two minds” for pretty much everything we do,
and especially when attempting to control our selves (Chaiken and Trope, 1999;
Lieberman, Gaunt, Gilbert, and Trope, 2002; Loewenstein, Weber, Hsee, and
Welch, 2001; Metcalfe and Mischel, 1999; Sloman, 1996).
On one hand, we are able to go on “autopilot,” acting on the basis of a set of
relatively, if not completely, automatic patterns of thought, feeling, and action.
These relatively automatic patterns can be complex and adaptively executed
(i.e., they are able to deal with some types of obstacles thrown in their way), but
are often relatively circumscribed in the sense that they pertain to a specific set
158 Kevin Ochsner
of cues and situations. The brain systems supporting these automatic patterns
are both cortical and subcortical, and importantly include: (1) the amygdala, an
almond shaped and sized cluster of subcortical nuclei important for detecting,
encoding, and triggering responses to affectively arousing and especially poten-
tially threatening stimuli (Ochsner and Gross, 2007; Phelps, 2006), and (2)
the striatum, a larger set of subcortical nuclei, important for laying down and
executing sequences of thought, affect, and action (Kober et al., 2008; Schultz,
Tremblay, and Hollerman, 2000). Together, these two subcortical structures,
and allied subcortical and (typically posterior) cortical systems enable us (a) to
rapidly identify goal-relevant, and therefore affectively salient, stimuli and events
and (b) to start responding to them. These evolutionarily older and relatively
automatic systems guide our behavior much of the time, as our default mode
of being in the world is to go with our habitual ways of thinking, feeling, and
acting.
On the other hand, we are able to consciously monitor, set goals for, and exert
control over our thoughts, feelings, and actions. This takes conscious and delib-
erate effort and has been shown to depend on regions of prefrontal cortex, one
of the evolutionarily newest and most highly developed portions of the human
brain (Miller, Freedman, and Wallis, 2002; Ochsner and Gross, 2005; Wager and
Smith, 2003). Our abilities to inhibit pre-potent, but potentially inappropriate
thoughts, feelings, and actions—in favor of more context-appropriate ones—
has been shown to depend critically upon a number of distinct prefrontal regions,
each of which may implement important control-related processes (Aron and
Poldrack, 2006; Badre and D’Esposito, 2007; Badre and Wagner, 2007). Our
control capacity is limited, however, as we have only a certain amount of
resources available to devote to whatever responses need to be shaped, guided,
or altered.
While this two-system view of the mind and brain is surely an oversimplifi-
cation, it has permeated virtually every area of psychological and neuroscience
research because it has great explanatory power. The basic notion is that we can
explain what we think, feel, and do in terms of interactions between the response
tendencies quickly queued up by the automatic system and the extent to which
we are motivated, and have the resources, to use the controlled system to modify
them (Chaiken and Trope, 1999; Lieberman et al., 2002; Loewenstein et al., 2001;
Metcalfe and Mischel, 1999; Sloman, 1996).
The second part of the answer to the question of how we can control our minds
has to do with the nature of the processes carried out by both the automatic and
controlled systems. At a fundamental level, each system is interpreting infor-
mation in the environment in a way that makes sense based on our prior history.
While much of our histories are similar, there are individual variations. As such,
for each person, in each situation, each system is making its own kind of inter-
pretation of stimulus inputs, and together they guide you to subjectively construe
the meaning of what’s going on (Kosslyn et al., 2002; Ochsner, 2007a, 2007b).
Or put another way, each type of system has a “belief” about what it is perceiving
and promotes actions on the basis of that belief.
The role of control in emotion, emotion regulation, and empathy 159
Putting these two parts together offers an answer as to why we can, and in
some cases, cannot or do not, properly control our emotional selves. If most of
the time we are on “autopilot”, then the automatic system(s) will simply queue up
emotional responses that make sense based on the way they interpret the current
situation. In many circumstances, however, these interpretations might not be
the most useful or appropriate. In the section below, we unpack the mechanisms
used in two such situations—one where our own emotions might need to be
regulated, and another in which our perceptions of another’s emotions might
need to be regulated—to understand how a core set of controlled processes may
be deployed to exert conscious, deliberate, top-down control over our interpreta-
tions, appraisals, and/or construals of our emotional world.
Emotion and emotion regulation

Against this backdrop, it is relatively easy to see how one can use top-down forms
of cognitive control to change the way one appraises the meaning of emotionally
evocative stimuli, and thereby change one’s emotional response. For the past
decade, this has been the focus of research in my laboratory, as well as many
other laboratories around the world. In general, we and others have found that
one’s initial emotional appraisal of the situation—guided by brain systems like
the amygdala and striatum, described above—can be modified through the use
of lateral and medial prefrontal systems that support the use of various kinds of
cognitive control processes (Ochsner and Gross, 2005, 2008).
While it is clear that prefrontal systems can modulate subcortical systems in
such a way that they increase, decrease, or maintain their activity in accordance
with regulatory goals, exactly how they achieve these effects is not yet clear.
Some of the prefrontal systems that are engaged by reappraisal—typically those
on the ventral and orbital surface of the frontal lobes—have direct intercon-
nections with the amygdala or striatum, and through these connections, may
directly influence their activity (Cavada, Company, Tejedor, Cruz-Rizzolo, and
Reinoso-Suarez, 2000; Ghashghaei, Hilgetag, and Barbas, 2007). But many of
the more dorsal prefrontal regions engaged by reappraisal do not have direct
interconnections with the systems that trigger emotional responses. Instead, they
are interconnected with parietal and temporal regions that represent the location,
size, shape, and general perceptual characteristics of the stimuli that elicit our
emotions (Barbas, 1992; Barbas, Ghashghaei, Dombrowski, and Rempel-Clower,
1999; Goldman-Rakic, 1992). Together these frontal-posterior networks are
thought to support specific higher-level control abilities like the retrieval of infor-
mation from semantic memory, working memory, and selective attention.
These anatomical and functional facts suggest an alternate route by which
reappraisal may exert its emotion modulatory effects: through the use of memory
and attention systems—as well as language, which also depends on frontal-
posterior networks—we can generate and maintain appraisals of emotionally
evocative stimuli that are different than the ones initially generated bottom-up by
subcortical appraisal systems. The idea is that reappraisal uses prefrontal systems
160 Kevin Ochsner
typically used just for selective attention or memory to control the activation of
spatial and object representations that comprise a new “percept” that is sent to
subcortical emotion systems. The appraisals for these new “perceptual” inputs
compete with the initial bottom-up appraisals of external stimuli. With sustained
effort and attention, these inputs force a new appraisal of stimuli from the
top-down.
While plausible and consistent with current data, this account has yet to be
directly tested. That being said, it has interesting implications for understanding
the development and breakdown of emotion regulatory abilities that we will
consider in the final section of the paper. In the next section, we consider other
ways that control can play another essential role in our emotional lives—in this
case, aiding our empathic perception of other’s emotions.
Empathy
“Empathy” is an umbrella term that refers to a constellation of related abilities. In
psychological research, three are typically enumerated: first, the tendency to take
on or share the feelings of others; second the ability to cognitively understand
those feelings; and third the tendency to act pro-socially on the basis of those
feelings (Decety and Batson, 2007; Zaki, Bolger, and Ochsner, 2008; Zaki and
Ochsner, 2009).
In recent years, neuroscience research has begun to focus on the first two of
these empathic abilities. The first—the ability to share the feelings of others
(and their internal states more generally)—is thought to depend on premotor and
sensory systems, including those for the perception of physical pain (Decety and
Batson, 2007; Gallese, Keysers, and Rizzolatti, 2004). These systems are engaged
relatively automatically both during first person sensory experience and during
the third person observation of someone else having the same kind of experience.
For example, frontal and parietal premotor systems (commonly referred to as the
“mirror system”) are engaged during both action execution and observation (Gallese
et al., 2004; Keysers, Kaas, and Gazzola, 2011). Similarly, regions of the cingulate
and insular cortex that received ascending spinal information about painful stimuli
are engaged both during the direct experience of pain and when one sees or knows
that someone else is in pain (Decety, 2009). The activation of systems for motor
planning, pain, or affect more generally when you’re observing others is thought to
provide a relatively automatic and intuitive basis for understanding their behavioral
intentions or affective state. The second ability—to cognitively understand the
feelings of others—is thought to depend on a network of regions centered around
the dorsal portion of the medial prefrontal cortex (and including the precuneus,
superior temporal sulcus, and temporal poles) (Frith and Frith, 2006; Mitchell,
2009; Olsson and Ochsner, 2008). These systems are engaged when one explicitly
reasons or makes attributions about mental states, including emotions, whether
they’re one’s own or someone else’s, current feelings or dispositional tendencies.
Most neuroscience research on empathy has focused on the use of one or the
other of these two types of systems when one is passively perceiving or making
simple judgments about another’s emotional states, and, as a consequence, hasn’t
explored the questions of when or how control processes may be important for
empathy (Zaki and Ochsner, 2009). Social cognition research suggests a poten-
tially critical role for control in two types of situations: where the behavior of
a social target is ambiguous, or where one is motivated to modify an initial
impression on the basis of situational or contextual information (Chaiken and
Trope, 1999). In either case, prefrontal control processes may be important for
top-down appraisals that integrate with or modify bottom-up appraisals to help us
identify the emotions of others.
Recently, we have investigated these two types of situations and found support
for this idea. In one experiment, we asked participants to watch videos of targets
talking about emotional events from their personal lives (Zaki et al., 2008; Zaki,
Weber, Bolger, and Ochsner, 2009). Participants were asked to continuously
rate the emotions experienced by targets, who themselves had provided ratings
of their own emotions. Correlating these two ratings provided a measure of
the accuracy with which participants empathically understood the emotions of
targets. Importantly, targets in the videos provided multimodal (i.e., verbal and
nonverbal), dynamic, and often subtle cues to their emotions—a situation that
social cognition research would suggest should require the use of control systems
to properly contextualize the meaning of each individual cue (e.g., realizing that
a neutral face when talking about something sad might not mean that you don’t
have any feelings about it). We found that overall levels of accuracy tracked
with activity in premotor systems involved in experience sharing, elements of
the medial prefrontal network involved in explicit attributions, and additional
prefrontal regions implicated in cognitive control.
While this provided initial evidence for our hypothesis, there also were
intriguing individual differences in the extent to which individuals engaged each
type of network, with some relying more on the experience sharing systems and
others relying more on the systems for mental state attribution. We suspected that
prefrontal control systems arbitrated the interactions between these two types of
systems as participants figured out when they should rely on each type of emotion
cue. Because this study was not designed to directly address this question, we
conducted another study designed specifically to tackle it. In this study, partici-
pants viewed short silent video clips drawn from videos in the first study where
targets were feeling strong positive or negative emotion (Zaki, Hennigan, Weber,
and Ochsner, 2010). These clips were paired with captions that implied the targets
were talking about topics that were either positive (e.g., a party) or negative (e.g.,
their dog died). When the two types of cues were in conflict (e.g., a video with
nonverbal cues to positive emotion paired with a negative caption), we predicted
that participants would need to engage control systems to figure out how to shift
their attention towards, and rely on, one type of cue or the other. That’s exactly
what we found: On one hand, as participants’ judgments reflected greater reliance
on the nonverbal cues presented in the video, activation increased in premotor
systems that support sharing of the intentions implied by targets actions. On the
other hand, as participants’ judgments reflected greater reliance on the contextual
162 Kevin Ochsner
cues provided in the captions, activity increased in a medial prefrontal region
that supports making explicit attributions. But most importantly, the extent to
which a participant showed one or the other pattern of activation was predicted
by functional connectivity with prefrontal and cingulate control systems, which
seemed to “direct traffic” by shifting activation from one system to the other and
shift judgments of target emotions accordingly.
Together, these data suggest new ways in which we can study the role of
control systems in empathic understanding, empathic experience sharing, and
the perception of others’ emotions more generally. By studying perceptions of
emotion in artificial contexts—e.g., by presenting only static and/or posed facial
expression—prior work generally failed to show evidence of control system
involvement (Zaki and Ochsner, 2009; Zaki and Ochsner, 2011). Our work, and
some emerging work from other labs as well, suggests that a key to studying the
role of control in empathy is to examine naturalistic contexts where control will
be needed to direct attention to the various types of cues and targets present, and
arbitrate between them.
Implications and future directions

If we are to make progress in our understanding of the mechanisms underlying
our various emotional abilities, we will need to have theories that move beyond
behavior-level descriptions of phenomena and move deeper into process- and
neural-level descriptions of underlying mechanisms (Ochsner, 2007b; Ochsner
and Gross, forthcoming). The goal of this chapter has been to give a taste of the
research in two domains where this type of multi-level approach to understanding
emotion is taking place. The first concerned our ability to exert control over
and change our emotional responses. The second concerned our ability to exert
control over and guide our empathic understanding of other people’s emotions.
In both cases, a domain-general set of prefrontal control systems was used to
influence activation in different types of subcortical and posterior cortical systems
that represent different types of domain-specific information related to one’s own
emotional response or the emotional responses exhibited by others. By biasing
processing in systems related to triggering emotions, sharing the experiences of
others, and/or making explicit attributions about them, prefrontal control systems
enable us to shape and change our first person experience of our own emotions
and our third person experience of others emotions.
This basic framework for understanding the role of control in emotion may
be applied to understanding the full range of normal to abnormal emotional
experience and expression. Within the normal range, it could be applied to under-
standing the development of emotion regulation and empathy from early in life
to young adulthood, as well as how both change as we age, in terms of changes
in the nature of and interactions between the systems described above. Outside
the normal range of emotion this framework could be useful in examining
emotion dysfunction in various kinds of clinical populations. For example,
it could be applied to understand whether and how individuals with major
depressive or borderline disorders have a dearth of positive and an abundance of
negative emotion because they have problems with emotional reactivity, emotion
regulation, or both (Johnstone, van Reekum, Urry, Kalin, and Davidson, 2007;
Koenigsberg et al., 2009). It could be similarly applied to understanding how the
deficits in empathy and emotion perception shown in individuals with Autism
Spectrum Disorders arise from problems with systems for experience sharing,
mental state attribution, or the use of control to shift between and integrate their
processing (Zaki and Ochsner, 2009; Zaki and Ochsner, 2011).
Hopefully, some of the benefits of this type of multi-level, integrative
approach to understanding emotion—and its relationship to control—have
been highlighted by this chapter. Not only do neuroscience-informed theories
of emotion regulation and empathy—and other affective phenomena more
generally—cut across multiple levels of analysis, they also speak to and connect
with more domains of research, and as a consequence may be more robust
and enduring (Ochsner, 2007b). That being said, the day it still young for this
approach to understanding emotion, and the framework we present here is at best
a crude sketch in need of correction, revision, and expansion. So while we are
enthusiastic about the promise of the approach and the framework it has built, we
fully expect that future research will bring significant and welcome changes to
both.
References
Aron, A. R., and Poldrack, R. A. (2006). Cortical and subcortical contributions to Stop
signal response inhibition: role of the subthalamic nucleus. The Journal of Neuroscience,
26, (9), 2424–2433.
Badre, D., and D’Esposito, M. (2007). Functional magnetic resonance imaging evidence for
a hierarchical organization of the prefrontal cortex. Journal of Cognitive Neuroscience,
19, 2082–2099.
Badre, D., and Wagner, A. D. (2007). Left ventrolateral prefrontal cortex and the cognitive
control of memory. Neuropsychologia, 45, (13), 2883–2901.
Barbas, H. (1992). Architecture and cortical connections of the prefrontal cortex in the
rhesus monkey. Adv Neurol, 57, 91–115.
Barbas, H., Ghashghaei, H., Dombrowski, S. M., and Rempel-Clower, N. L. (1999).
Medial prefrontal cortices are unified by common connections with superior temporal
cortices and distinguished by input from memory-related areas in the rhesus monkey.
J Comp Neurol, 410, (3), 343–367.
Cavada, C., Company, T., Tejedor, J., Cruz-Rizzolo, R. J., and Reinoso-Suarez, F. (2000).
The anatomical connections of the macaque monkey orbitofrontal cortex. A review.
Cereb Cortex, 10, (3), 220–242.
Chaiken, S. E., and Trope, Y. (eds) (1999). Dual-process theories in social psychology.
New York, NY: Guilford Press.
Decety, J. (2009). Empathy, sympathy and the perception of pain. Pain, 145, (3), 365–366.
Decety, J., and Batson, C. D. (2007). Social neuroscience approaches to interpersonal
sensitivity. Soc Neurosci, 2, (3–4), 151–157.
Frith, C. D., and Frith, U. (2006). How we predict what other people are going to do. Brain
Res, 1079, (1), 36–46.
164 Kevin Ochsner
Gallese, V., Keysers, C., and Rizzolatti, G. (2004). A unifying view of the basis of social
cognition. Trends Cogn Sci, 8 (9), 396–403.
Ghashghaei, H. T., Hilgetag, C. C., and Barbas, H. (2007). Sequence of information
processing for emotions based on the anatomic dialogue between prefrontal cortex and
amygdala. Neuroimage, 34, (3), 905–923.
Goldman-Rakic, P. S. (1992). Working memory and the mind. Sci Am, 267, (3), 110–117.
Johnstone, T., van Reekum, C. M., Urry, H. L., Kalin, N. H., and Davidson, R. J. (2007).
Failure to regulate: counterproductive recruitment of top-down prefrontal-subcortical
circuitry in major depression. J Neurosci, 27, (33), 8877–8884.
Keysers, C., Kaas, J. H., and Gazzola, V. (2011). Somatosensation in social perception. Nat
Rev Neurosci, 11, (6), 417–428.
Kober, H., Barrett, L. F., Joseph, J., Bliss-Moreau, E., Lindquist, K., and Wager, T. D.
(2008). Functional grouping and cortical-subcortical interactions in emotion: a meta-
analysis of neuroimaging studies. Neuroimage, 42, (2), 998–1031.
Koenigsberg, H. W., Fan, J., Ochsner, K. N., Liu, X., Guise, K. G., Pizzarello, S., et al.
(2009). Neural correlates of the use of psychological distancing to regulate responses
to negative social cues: a study of patients with borderline personality disorder. Biol
Psychiatry, 66, (9), 854–863.
Kosslyn, S. M., Cacioppo, J. T., Davidson, R. J., Hugdahl, K., Lovallo, W. R., Spiegel, D.,
et al. (2002). Bridging psychology and biology. The analysis of individuals in groups.
Am Psychol, 57, (5), 341–351.
Lieberman, M. D., Gaunt, R., Gilbert, D. T., and Trope, Y. (2002). Reflexion and reflection:
A social cognitive neuroscience approach to attributional inference. In M. P. Zanna
(ed.), Advances in experimental social psychology, Vol. 34 (pp. 199–249). San Diego,
CA: Academic Press, Inc.
Loewenstein, G. F., Weber, E. U., Hsee, C. K., and Welch, N. (2001). Risk as feelings.
Psychol Bull, 127, (2), 267–286.
Metcalfe, J., and Mischel, W. (1999). A hot/cool-system analysis of delay of gratification:
dynamics of willpower. Psychol Rev, 106, (1), 3–19.
Miller, E. K., Freedman, D. J., and Wallis, J. D. (2002). The prefrontal cortex: categories,
concepts and cognition. Philos Trans R Soc Lond B Biol Sci, 357, (1424), 1123–1136.
Mitchell, J. P. (2009). Inferences about mental states. Philos Trans R Soc Lond B Biol Sci,
364, (1521), 1309–1316.
Ochsner, K. N. (2007a). How thinking controls feeling: A social cognitive neuroscience
approach. In E. H. Jones, and P. Winkielman (eds), Social Neuroscience: Integrating
biologicaland psychological explanations of behavior (pp. 106–136). New York, NY:
Guilford Press.
—(2007b). Social Cognitive Neuroscience: Historical Development, Core Principles,
and Future Promise. In A. Kruglanksi and E. T. Higgins (eds), Social Psychology: A
Handbook of Basic Principles (2nd edn, pp. 39–66). New York: Guilford Press.
Ochsner, K. N., and Gross, J. J. (2005). The cognitive control of emotion. Trends in
Cognitive Sciences, 9, (5), 242–249.
—(2007). The neural architecture of emotion regulation. In J. J. Gross, and R. H.
Thompson (eds), The handbook of emotion regulation (pp. 87–109). New York:
Guilford Press.
—(2008). Cognitive emotion regulation: Insights from social cognitive and affective
neuroscience. Currents Directions in Psychological Science, 17, (1), 153–158.
—(forthcoming). The neural bases of emotion and emotion regulation: A valuation
perspective.
Olsson, A., and Ochsner, K. N. (2008). The role of social cognition in emotion. Trends in
Cognitive Sciences, 12, (2), 65–71.
Phelps, E. A. (2006). Emotion and cognition: insights from studies of the human amygdala.
Annu Rev Psychol, 57, 27–53.
Schultz, W., Tremblay, L., and Hollerman, J. R. (2000). Reward processing in primate
orbitofrontal cortex and basal ganglia. Cereb Cortex, 10 (3), 272–284.
Sloman, S. A. (1996). The empirical case for two systems of reasoning. Psychological
Bulletin, 119, (1), 3–22.
Wager, T. D., and Smith, E. E. (2003). Neuroimaging studies of working memory: a meta-
analysis. Cogn Affect Behav Neurosci, 3, (4), 255–274.
Zaki, J., Bolger, N., and Ochsner, K. N. (2008). It takes two: the interpersonal nature of
empathic accuracy. Psychol Sci, 19, (4), 399–404.
Zaki, J., Hennigan, K., Weber, J., and Ochsner, K. N. (2010). Social cognitive conflict
resolution: contributions of domain-general and domain-specific neural systems. The
Journal of Neuroscience, 30, (25), 8481–8488.
Zaki, J., and Ochsner, K. N. (2009). The need for a cognitive neuroscience of naturalistic
social cognition. Ann N Y Acad Sci, 1167, 16–30.
Zaki, J., and Ochsner, K. N. (2011). Reintegrating the study of accuracy into social
cognition research. Psychological Inquiry, 22, (3), 159–182.
Zaki, J., Weber, J., Bolger, N., and Ochsner, K. N. (2009). The neural bases of empathic
accuracy. Proc Natl Acad Sci U S A, 106, (27), 11382–11387.
23 What time can tell us
The temporal dynamics of emotion
regulation
Ravi Thiruchselvam and James J. Gross
Stanford University
Imagine a world without any type of emotional control. Provocations would

escalate into bare-knuckled fist fights. Roads and highways would be clogged
with honking horns and screaming drivers. Moments of hilarity would spiral out
of control, leading to chaos. Small setbacks would trigger paralyzing cycles of
sadness, culminating in depression. It is not hard to see from this hypothetical
world that the ability to regulate emotions when they run awry—as they all too
often do—is one of the most critical of human capacities (Gross, 2007).
Commensurate with its significance in everyday life, emotion regulation has
become the focus of intense research over the past decade. Both theory and
emerging empirical findings have begun to point to the importance of under-
standing how regulatory processes impact the time-course of emotion generation.
In this chapter, we consider why the investigation of temporal dynamics is
playing such a key role in our growing efforts to understand emotion regulation.
We begin by outlining the process model of emotion regulation (Gross, 1998)—a
framework that is premised on the notion that a central feature of a regulation strategy
is when it impacts emotion generation. We then describe one recent empirical test of
this model, and delineate some of the specific ways by which the analysis of temporal
dynamics might further inform our understanding of emotion regulation processes.
The process model of emotion regulation

The process model of emotion regulation (Gross, 1998) is based on the notion that
emotion generation unfolds through a temporally-extended sequence of stages.
In this sequence, an emotional stimulus (or situation) first compels our attention
towards it. We then evaluate its meaning on various dimensions (valence,
goal relevance, coping potential, etc.) in order to appraise its affective signifi-
cance. This pattern of appraisals gives rise to loosely coordinated multi-system
responses (across subjective experience, autonomic physiology, and expressive
behavior). Together, this situation-attention-appraisal-response sequence consti-
tutes one iteration in the unfolding of an emotional response.
Using this timeline as a basic framework, the process model proposes that
different emotion regulation strategies act on separate stages in this temporal
trajectory. For instance, distraction—a strategy that involves diverting attention away
What time can tell us 167
from the emotional stimulus—is believed to act on the early attentional deployment
stage. Reappraisal, another powerful strategy, involves altering the meaning of an
emotional event, and it is thought to operate later on in the emotion-generative
process, at the appraisal stage. Expressive suppression, by contrast, targets the final
response stage by inhibiting the overt display of expressive behavior.
In our view, emotions unfold and gather strength through a successive recursion
of multiple situation-attention-appraisal-response iterations (Gross, 2007). Thus,
it is possible to apply a particular regulation strategy at various points in the
unfolding of an emotional response. For instance, distraction or reappraisal can
be applied on the first few emotion-generative iterations, when the emotion has
not gathered much force and therefore intensity is relatively low, or in the later
iterations when intensity is high (Sheppes and Gross, 2011). Both distraction and
reappraisal can therefore be engaged after an emotion has been set into motion. In
other cases, it is sometimes possible to pre-emptively plan to apply a strategy in
the context of an upcoming situation if the individual anticipates the situation to
elicit emotion. For instance, an arachnophobic who expects an image of a spider
to appear in an upcoming film scene can begin to apply distraction when the scene
begins, thereby precluding emotion from being generated.
As noted earlier, one crucial prediction made by the process model is that
distraction and reappraisal intervene at separate stages in emotion generation.
In a recent study (Thiruchselvam et al., 2011), we tested this key prediction
by capitalizing on the excellent temporal resolution afforded by EEG/ERP
methods. Subjects were exposed to emotionally arousing pictures, and were
asked to generate mental imagery of neutral geometric shapes (distraction) or to
re-construe the picture’s underlying meaning in a neutral manner (reappraisal).
To examine whether these two strategies act on separate stages of emotion
generation, we focused specifically on an ERP component known as the late
positive potential (LPP). The LPP is known to be enhanced when individuals are
exposed to emotionally arousing stimuli compared to neutral stimuli (Cuthbert et
al., 2000), and crucially, it has been shown in a number of studies to be sensitive
to the way that an emotional stimulus’ meaning is evaluated (becoming attenuated
when an emotional stimulus is evaluated in a neutral manner; Foti and Hajcak,
2008). Thus, a reduction of the LPP from its point of onset (approximately
300ms) would reflect restricted evaluative processing of the affective meaning
of the stimulus. By contrast, an attenuation of the LPP beginning at later stages
would signify that some processing of the stimulus’ affective meaning has
occurred.
Our results showed that, although both strategies powerfully reduced the
LPP, there were important timing differences in their relative effects. In support
of the process model, we found that distraction attenuated the LPP from its
point of onset (approximately 300ms), suggesting that under distraction, no
appraisal of the stimulus’ affective meaning had occurred. Reappraisal, however,
attenuated the LPP later in its temporal trajectory (at about 1500ms), suggesting
that reappraisal involves evaluating the affective significance of the emotional
stimulus. These results are presented in Figure 7.
168 Ravi Thiruchselvam and James J. Gross
A
-5
Amplitude (µV)
10
0 1000 2000 3000 4000 5000
Time (ms)
B
-5
Neutral-View
Negative-Watch
Amplitude (µV)
0
Negative-Distract
Negative-Reappraise
10
300 500 700 900 1100 1300 1500 1700
Time (ms)
Figure 7 ERPs by Instruction Type during picture presentation of the regulation task.
The LPP by Instruction Type during the 300–1700ms time window (the
range between the thick vertical superimposed bars) is shown as a separate
panel for clarity. Note that the y-axis is reversed (positive voltage is plotted
downwards) as per convention. This figure is from Thiruchselvam et al.
(2011).
The temporal dynamics of emotion regulation

In addition to addressing questions about the nature of different emotion
regulation processes, an analysis of temporal dynamics can inform our under-
standing of: i) the longer-term consequences of using different strategies, ii) the
effectiveness of different strategies under varying levels of emotional intensity,
and iii) how regulatory processes may be rendered dysfunctional in certain forms
of psychopathology. Below, we elaborate on these points.
1) Longer-term consequences of employing different strategies

The differential impact of distinct strategies on the emotion generative trajectory
during regulation may have certain consequences that extend beyond the regulatory
episode, influencing the processing of the stimulus when it is later encountered.
For instance, as distraction intervenes early in the emotion-generative process—
thereby preventing the processing of the stimulus’ underlying meaning—it
should lead individuals to evaluate the stimulus as more novel upon subsequent
re-exposures, relative to a stimulus that was previously attended to and evaluated.
Since novel emotional events elicit stronger emotional responses than events that
have been previously encountered and attended to (Wilson, and Gilbert, 2008),
stimuli with a distraction-history should elicit greater emotional responses upon
re-exposure than those with a history of simple viewing.
By contrast, insofar as reappraisal intervenes later in the emotion-generative
trajectory by enabling one to evaluate a stimulus’ affective meaning, stimuli with
a reappraisal-history should not have this detrimental effect upon re-exposure.
In fact, as reappraisal involves altering the meaning of the stimulus, the
modified (more neutral) appraisal may become activated upon re-exposure,
biasing appraisals towards a neutral direction. Thus, stimuli with a reappraisal-
history might elicit weaker emotional responses upon re-exposure compared to
those with a history of simple viewing, a prediction that is supported by recent
findings (MacNamara, Ochsner, and Hajcak, 2010).
To test this prediction, we recently compared the effects of distraction and
reappraisal on the subsequent processing of emotional stimuli (Thiruchselvam et
al., 2011). In an initial regulation phase, images were simply viewed or regulated
using distraction or reappraisal. Thirty minutes later, all images were presented
again in a passive viewing context. Results showed that, upon re-exposure, images
with a distraction-history elicited a larger LPP (signifying greater emotional
arousal) than those with a history of simple-viewing. In contrast, images with
a reappraisal-history elicited weaker LPPs than images with a simple viewing-
history, although this latter effect was only present for a specific temporal window
(800–1400ms) of the LPP.
This finding is interesting because it demonstrates that the temporal properties
of a regulatory process may lead to counter-intuitive consequences that may not
be apparent during the regulatory episode. During regulation, distraction power-
fully attenuated emotion generation at a very early stage. This property may
make distraction particularly rewarding to use, leading individuals to rely on it
frequently. However, the very capacity to decrease an emotional response quickly
may come at a cost over the long-term by eliciting greater emotional responses
upon re-exposure to the stimulus, compared to simply attending to it.
2) Consequences for employing different strategies under varying

emotional intensity
Temporal dynamics are also critical because they may have important conse-
quences for the effectiveness of different regulation strategies under varying
levels of emotional intensity. Sheppes and Gross (2011) have recently elaborated
on the process model (Gross, 1998) by highlighting how emotion-generative
processes and emotion-regulatory processes compete at two major stages of
processing.
At an early stage of information processing, incoming perceptual information
compete for entry into selective attention. The resolution of competition between
conflicting inputs at the early selection stage is proposed to involve minimal
cognitive resources, since the stimulus has not yet been fully represented in
working memory. Information that manages to pass through the selective
attention filter then enters the later stage of processing, in which it is afforded
more elaborate semantic analysis. At the late stage of processing, different
semantic representations (i.e., an emotionally-laden evaluation of a stimulus’
meaning versus a neutral evaluation of the stimulus’ meaning) compete for
dominance. Resolution at this late selection stage, however, is posited to require
much more cognitive resources, since resolution must now occur between
conflicting representations that have been fully developed in working memory
(Sheppes, and Gross, 2011).
In Sheppes’ and Gross’ (2011) model, distraction is proposed to be an early
selection strategy, in that it acts as a filter at the selective attention stage by deter-
mining whether an emotional stimulus is afforded any later meaning evaluation
(i.e., semantic analysis). Reappraisal, however, is posited to be a late selection
strategy which acts at the semantic analysis stage by trying to override one
semantic representation of the stimulus (i.e., the more emotional representation)
with another one (i.e., the more neutral representation).
The notion that distraction and reappraisal act at separate temporal stages
in information processing, and therefore differ in the levels of cognitive effort
required for implementation, should have important consequences for which
strategy is best under varying levels of emotional intensity. Specifically, emotional
intensity should not influence the effectiveness of early selection strategies, such
as distraction, because these strategies block incoming emotional information
with little cognitive effort. However, emotional intensity should heavily affect
the success of late selection strategies such as reappraisal, since modifying
semantic representations requires much more cognitive effort. When emotions
are highly intense, distraction should fare much better than reappraisal, since
distraction blocks emotionally-laden information from being represented in
working memory with minimal cognitive effort. When emotions are less intense,
however, both distraction and reappraisal could be effective.
In direct support of this prediction, Sheppes and Meiran (2007) showed in
a series of studies that manipulations of emotional intensity did not affect the
success of distraction, but did affect the success of reappraisal. Specifically,
distraction attenuated sadness irrespective of the level of development (i.e.,
intensity) of the emotion. By contrast, although reappraisal reduced sadness
when the emotion was not well-developed, it became much less effective when
regulating highly developed, intense levels of sadness.
3) Implications for understanding emotion (dys)regulation in psychopathology

Several forms of psychopathology are marked by profound impairments in
emotion regulation. Such impairments may be due in large part to alterations
in the temporal properties of emotion-generative processes. These alterations
may render the use of certain emotion regulation strategies difficult, leading
individuals to rely on ineffective forms of regulation instead.
For instance, a number of studies have shown that anxiety involves accelerated
evaluation of threatening information (Larson et al., 2006; Mogg, and Bradley,
1999). The rapidity of threat evaluations in anxiety may substantially hinder
efforts to reappraise the threatening stimulus, as individuals may not be able to
recruit reappraisal mechanisms before a fully developed appraisal pattern of the
stimulus has been established.
This may lead individuals to rely on ineffective—and even potentially detri-
mental—forms of emotion regulation, such as expressive suppression, which
are accessible further down the emotion-generative trajectory. As a regulation
strategy, expressive suppression fails to attenuate negative emotional experience
and comes with physiological costs by increasing sympathetic nervous system
activation (Gross, 1998). It also depletes cognitive resources, contributing to
working memory deficits (Johns, Inzlicht, and Schmader, 2008) and poor memory
for the emotion-eliciting situation (Richards, 2004). Since expressive suppression
is employed as emotion response tendencies are elicited (i.e., after an evaluation
of the stimulus has been made), it may easily become the “default” strategy
when rapid meaning-evaluations of emotional stimuli in affective disorders
preclude the use of more adaptive strategies such as reappraisal. Indeed, a number
of studies have found an association between suppression use and affective
disorders (Gross, and John, 2003; Kashdan, and Steger, 2006; Levitt et al., 2004).
Conclusion and future directions

In this chapter, we have argued that the temporal properties of emotion regulatory
processes can have critical implications for our understanding of their underlying
mechanisms, the consequences of employing them, and their dysfunction in
psychopathology. The considerations outlined in this chapter serve to highlight
only part of the importance of temporal dynamics in understanding emotion
regulation. There are a number of important questions in this field which warrant
further research.
First, we have focused in this chapter on two core regulatory processes,
distraction and reappraisal. This is by no means an exhaustive list, however,
as individuals often use other regulatory strategies to control emotions. For
instance, labeling one’s emotions—a strategy termed affect labeling—has been
shown to reduce emotional responding (Lieberman et al., 2007). Moreover,
emerging research on mindfulness meditation suggests that it can have important
consequences for emotion regulation as well (Brown, Ryan, and Creswell, 2007;
Slagter et al., 2007). Further research is needed to understand how such strategies
may alter the temporal properties of emotion generation and regulation.
From our perspective, other pressing questions in this domain include the
following. Does training in emotion regulation processes (i.e., attentional
redeployment, cognitive reappraisal, etc.) alter how they impact the temporal
trajectory of emotion generation? Do changes in the temporal properties of
regulatory processes actually influence their overall effectiveness in modulating
emotion? How do situational factors (existing cognitive load, regulatory goals,
etc.) influence the temporal dynamics of emotion regulation? Questions such as
these are critical to explore in light of an emerging view which, as highlighted
in this chapter, increasingly suggests that time can tell us something meaningful
about emotion regulation.
References
Brown, K. W., Ryan, R. M., and Creswell, J. D. (2007). Mindfulness: Theoretical founda-
tions and evidence for its salutary effects. Psychological Inquiry, 18, 211–237.
Cuthbert, B. N., Schupp, H. T., Bradley, M. M., Birbaumer, N., and Lang, P. J. (2000).
Brain potentials in affective picture processing: covariation with autonomic arousal and
affective report. Biological Psychology, 52, 95–111.
Foti, D., and Hajcak, G. (2008). Deconstructing reappraisal: Descriptions preceding
arousing pictures modulate the subsequent neural response. Journal of Cognitive
—(1998). Antecedent- and response-focused emotion regulation: Divergent conse-
quences for experience, expression, and physiology. Journal of Personality and Social
—(ed.) (2007). Handbook of emotion regulation. New York, NY: Guilford Press.
Gross, J. J., and John, O. P. (2003). Individual differences in two emotion regulation
processes: Implications for affect, relationships, and well-being. Journal of Personality
and Social Psychology, 85, 348–362.
Johns, M. J., Inzlicht, M., and Schmader, T. (2008). Stereotype threat and executive resource
depletion: Examining the influence of emotion regulation. Journal of Experimental
Psychology: General, 137, 691–705.
Kashdan, T. B., and Steger, M. F. (2006). Expanding the topography of social anxiety:
An experience-sampling assessment of positive emotions, positive events, and emotion
suppression. Psychological Science, 17, 120–128.
Larson, C. L., Schaefer, H. S., Siegle, G. J., Jackson, C. A. B., Anderle, M. J., and
Davidson, R. J. (2006). Fear is fast in phobics: Amygdala activation in response to fear-
relevant stimuli. Biological Psychiatry, 60, 410–417.
Levitt, J. T., Brown, T. A., Orsillo, S. M., and Barlow, D. H. (2004). The effects of acceptance
versus suppression of emotion on subjective and psychophysiological response to carbon
dioxide challenge in patients with panic disorder. Behavior Therapy, 35, 747–766.
Lieberman, M. D., Eisenberger, N. I., Crockett, M. J., Tom, S. M., Pfeifer, J. H., and Way,
B. M. (2007). Putting feelings into words: Affect labeling disrupts amygdala activity to
affective stimuli. Psychological Science, 18, 421–428.
MacNamara, A., Ochsner, K. N. and Hajcak, G. (2011). Previously reappraised: the lasting
effect of description type on picture-elicited electrocortical activity. Social, Cognitive,
and Affective Neuroscience, 6, 348–358.
Mogg, K., and Bradley, B. P. (1999). Some methodological issues in assessing attentional
biases for threatening faces in anxiety: A replication study using a modified version of
the probe detection task. Behaviour Research and Therapy, 37, 595–604.
Richards, J. M. (2004). The cognitive consequences of concealing feelings. Current
Directions in Psychological Science, 13, 131–134.
Sheppes, G., and Gross, J.J. (2011). Is timing everything? Temporal considerations in
emotion regulation. Personality and Social Psychology Review, 15, 319–331.
Sheppes, G., and Meiran, N. (2007). Better late than never? On the dynamics of on-line
regulation of sadness using distraction and cognitive reappraisal. Personality and Social
Slagter, H. A., Lutz, A., Greischar, L. L., Francis, A. D., Nieuwenhuis, S., Davis, J. M.
and Davidson, R. J. (2007). Mental training affects use of limited brain resources. PLoS
Biology, 5, (6).
Thiruchselvam, R., Blechert, J., Sheppes, G., Rydstrom, A., and Gross, J. J. (2011). The
temporal dynamics of emotion regulation: An EEG study of distraction and reappraisal.
Biological Psychology, 87, 84–92.
Wilson, T. D., and Gilbert, D. T. (2008). Explaining away: A model of affective adaptation.
Perspectives on Psychological Science, 3, 370–386.
24 The duration of emotional episodes
Iven Van Mechelen, Philippe Verduyn and
Karen Brans
Emotions are not just momentary states, but processes that evolve and unfold
over time (Sonnemans and Frijda, 1994). A full account of them therefore inevi-
tably requires an in-depth understanding of their temporal dynamics. This is the
major challenge for the research domain that was named by Davidson (1998)
affective chronometry. In the present chapter, we will focus on one key aspect of
the time dynamics of emotions: emotion duration.
Up to 15 years ago, emotion duration has been a largely underinvestigated
topic, apart from pioneering work by Frijda and colleagues (see, e.g., Frijda,
Mesquita, Sonnemans and Van Goozen, 1991). More recently, however, we
witnessed several attempts to contribute solid empirical evidence on the topic,
and especially on factors that account for a significant amount of variability in
emotion duration.
We will present a brief review of the most important theoretical and empirical
findings on determinants of emotion duration. Subsequently, we will briefly
discuss the possible process basis underlying the operation of these determinants.
We will start, however, with a discussion of the very concept of emotion duration.
The concept of emotion duration

Emotion duration refers to the duration of emotion episodes. For a given episode,
the duration can be simply defined as the amount of time between its beginning
and its end point. When looking more closely at this definition, however, it
appears to include several aspects, each of which requires some further clarifi-
cation. In this clarification, quite different choices can be made, leading to quite
different specifications of the concept of emotion duration.
A first aspect pertains to the concept of emotion. As argued by Frijda “‘An’
emotion (. . .) is a slippery notion” (Frijda, 2007, p. 195) and “the same emotion
words are being used for psychological events of vastly different kinds and
durations” (Frijda et al., 1991: 221). At this point, a careful distinction between
the concept of emotion and a few neighboring concepts may be especially
important, as this may be very consequential for implied differences in duration.
Important neighboring concepts in this regard include moods and sentiments.
Frijda et al. (1991) define moods as more or less continuous feeling states that,
The duration of emotional episodes 175
unlike emotions, do not bear upon a particular object. Frijda (2007) further
defines sentiments as “dipositions to respond emotionally to a particular object”
(p. 192); those dispositions are further considered by him as attitude-like entities
that figure in complex, bidirectional relations with emotions (Frijda et al., 1991).
A second aspect pertains to the concept of emotion episode. Perhaps the most
straightforward definition of an emotion episode is that of the period that elapses
between the onset point of an emotional experience and the moment at which this
experience is no longer felt (see, e.g., Verduyn, Delvaux, Van Coillie, Tuerlinckx,
and Van Mechelen, 2009). Across this simple definition, Frijda (2007) proposes
a much more complex alternative in which emotions are considered “coherent
narratives”. The coherence stems from a continuously felt impact of the emotion-
eliciting event or series of events (Frijda et al., 1991). Frijda’s definition is further
complicated by the inclusion of a number of paradoxical (and perhaps even
contradictory) elements, in that he emphasizes on the one hand the continuity
of emotional episodes, and on the other hand the fact that they may span quite
a few interruptions (Frijda, 2007). Also, on the one hand he links the coherence
of an emotional episode to the continuity of its associated core relational theme,
whereas on the other hand he emphasizes the possible multiplicity of emotions
within a single emotion episode (Frijda et al., 1991).
A third aspect pertains to the concept of the beginning of an emotional episode.
In general, this is not the subject of much discussions, also taking into account
the broad agreement on the view that emotions are elicited by particular events
or series of events (notwithstanding the fact that these events may not necessarily
exactly coincide with the beginning of the episodes in question—see Frijda,
2007: 182).
A fourth aspect pertains to the concept of the end point of an emotional episode,
for which the story is quite different from that for its counterpart, the beginning.
With regard to it, Sonnemans and Frijda (1994) talk about the moment at which
“the emotion episode is closed”. A straightforward specification of closure could
refer to the moment at which the emotion is no longer felt (see, e.g., Verduyn
et al., 2009), that is, the moment at which the experienced emotion intensity
is again zero. A slightly more subtle specification may include a reference to
a return to some baseline level, with such a baseline level in some cases being
linked to the emotional intensity experienced by some control group (see, e.g.,
Sbarra, 2006). One may wonder, however, whether such measures of closure
(which are being used in the majority of studies on emotion duration) do not
capture (possibly temporary) emotional relief rather than full-blown emotional
recovery (Brans, Van Mechelen, Verduyn, and Rimé, 2010). Linking up with this
question, the issue of an episode’s end point may be rendered more difficult by
requiring closure to imply that, after it, no relapse in emotional intensity may
occur. As an example, one may again refer to Sbarra (2006) who operationalizes
closure in terms of a number of consecutive occurrences of zero-level emotional
intensity. An interesting, yet even more involved, perspective is opened when the
end point issue is looked at from the viewpoint of psychological closure. Lack of
such closure could, for instance, be linked to emotion determining thoughts and
176 Iven Van Mechelen, Philippe Verduyn and Karen Brans
behavior, which may still be the case when that emotion is no longer felt. More
specifically, Beike and Wirth-Beaumont (2005) discuss psychological closure
from the angle of autobiographical memory. For them, closure immediately
relates to decreased emotional detail in the constructed autobiographical memory
representation of the emotion-eliciting event. To support this claim, Beike and
Wirth-Beaumont (2005) provide empirical evidence that the amount of emotional
detail in memory accounts for a significant amount of variance in subjective sense
of closure reported by participants, with less detail being associated with a greater
sense of closure. (Otherwise, the authors hypothesize that even adjustment to
traumatic events essentially relies on reduction of emotional detail in memory.)
Determinants of emotion duration

Empirical research has consistently documented variation in the duration of
emotional episodes, with several studies reporting durations ranging from a few
seconds to several hours (and even longer) (see, e.g., Gilboa and Revelle, 1994;
Sonnemans and Frijda, 1994; Verduyn et al., 2009; Verduyn, Van Mechelen and
Tuerlinckx, 2011). Variation has been found both within and between individuals.
Also, some evidence for cross-cultural differences has been reported (although
within-culture differences have been found to be more sizeable) (Scherer and
Wallbott, 1994). An obvious question is then how this variance in duration could
come about?
Potential determinants of emotion duration can be grouped into three
categories: (1) characteristics of the emotion-eliciting event (stimulus, situation),
(2) characteristics involving the subject who experiences the emotion, and (3)
characteristics of the emotion itself. Below we will successively summarize the
most important theoretical and empirical research results for each of these three
categories.
1. Characteristics of the emotion-eliciting event

On a theoretical level, Frijda (2007) conjectures that emotion duration is directly
related to the duration of the emotion-eliciting event and its aftermath. Somewhat
related to this, Verduyn et al. (2009) reported fairly consistent empirical evidence
that physical reappearances of the source that elicited the emotion under study
increased the length of the emotional episode involved. Verduyn and colleagues
(2009, 2011) further also reported consistent evidence that the more important the
emotion-eliciting event is, the longer the emotion will last.
2. Characteristics involving the subject who experiences the emotion

This category first of all comprises dispositional characteristics. In this regard,
Gilboa and Revelle (1994) hypothesized that neuroticism is correlated with
protracted negative affectivity, whereas extraversion goes with protracted positive
affectivity. In a vignette study, however, this hypothesis was only partially
confirmed, in that neurotics were found to display longer emotional reactions to
mildly negative as compared to mildly positive events; also longer durations of
negative emotional responses have been found for people who scored simultane-
ously high on neuroticism and introversion; yet, apart from this, neurotics were
primarily found to report longer emotion durations, irrespective of valence. From
his part, Hemenover (2003) studied rate of affect change rather than duration
in the strict sense. In line with the hypotheses of Gilboa and Revelle (1994),
he reported that extraverts and emotionally stable persons display slow rates of
positive affect decay and rapid rates of negative effect decay, whereas introverts
and neurotics display an opposite pattern of decay. Verduyn et al. (2009) further
examined the possible role of all Big Five factors; yet, they did not find consistent
results at this point. Taking all findings together, the safest summary seems to be
that the results on the relation between the Big Five and emotion duration are not
entirely conclusive. On a much more specific level, Sbarra (2006) studied the
duration of sadness and anger episodes after the dissolution of a serious romantic
relationship. The dispositional characteristic that turned out to be relevant in his
study was attachment security, in that secure attachment was associated with
faster anger and sadness recovery.
A second kind of characteristics involving the subject who experiences the
emotion pertains to the subject’s past or learning history. Frijda (2007) conjec-
tures in this regard that durations may be curtailed by habituation, with this
habituation probably proceeding faster in response to pleasant as compared
to adverse circumstances. To the best of our knowledge, however, empirical
evidence in support of this conjecture is anecdotic in nature only.
A third and most important kind of characteristics within this category pertains
to actions undertaken by the subject who experiences the emotion under study.
This kind comprises both covert actions (with, for example, a possibly important
role being reserved for rumination: see, e.g., Gilboa and Revelle, 1994), and their
overt counterparts (including not least interpersonal interactions: see, e.g., Frijda,
2007). Conceptually, several of the actions in question are being studied in the
burgeoning field of emotion regulation (e.g., Gross, 2007). With regard to the
impact of covert actions on emotion duration, the most solid empirical evidence
has been contributed by Verduyn and colleagues (2009, 2011). These authors
initially found that mental reappearances of an emotion-eliciting stimulus extend
the duration of emotional experiences. Subsequently, they refined this finding by
also taking the contents of the stimulus-related cognitions into account: When
these have the same valence as the emotion in question (as is often the case), their
occurrence leads to a prolonged duration; when their valence is opposite to that of
the emotion, however, their occurrence leads to a shortening. Otherwise, the latter
also holds for stimulus-unrelated cognitions (which may, for example, figure
in attempts at distraction). With regard to overt actions, Brans, Van Mechelen,
Verduyn, and Rimé (2010) contributed evidence that, for negative emotions,
social sharing leads to a decrease in emotion duration (or at least to a temporary
relief).
3. Characteristics of the emotion itself
Within this category, one may first consider characteristics of the emotion as
a whole. This comprises the nature of the emotion. Frijda (2007) conjectures
that emotions may have a built-in time course; hence, one may expect between-
emotion differences in duration. In line with this, Scherer and Wallbott (1994)
report evidence for the following rank order of duration:
fear = disgust = shame ≤ anger < guilt < joy < sadness.
Verduyn and colleagues (2009, 2011) report independent evidence for (parts of)
the same rank order.
Beyond the nature or content of the emotion, one may consider formal charac-
teristics of the emotion episode other than its duration. Those include overall
intensity, initial intensity or intensity at onset, peak amplitude, and rise time to
peak (see, e.g., Frijda, 2008; Davidson, 1998). Among these, primarily intensity
has been investigated in relation to duration. Yet, results are somewhat mixed at
this point, in that Sonnemans and Frijda (1994) report a weak relation between
intensity and duration, whereas Verduyn and colleagues (2009, 2011) found
consistent evidence in three studies for a positive relation between intensity and
duration. Part of this ambiguity could be due to the complexity of the notion of
emotion intensity, and of differences in the ways intensity has been empirically
dealt with in the different studies, with Sonnemans and Frijda (1994) focusing
on overall felt intensity, and Verduyn and colleagues (2009, 2011) on intensity at
onset.
A second subtype of determinants within the category of emotion character-
istics relates to the multi-componential nature of emotions. Emotions, indeed, do
involve quite different response systems, each of which implying quite different
latencies and decay times (Frijda et al., 1991). Component processes at different
levels can therefore be expected to have quite different durations. At this point,
Frijda (2007) observes that responses at hierarchically lower levels (such as,
e.g., facial expressions) are on much smaller time scales than their hierarchically
higher counterparts (such as, e.g., emotional engagement and arousal of action
readiness). There is still a need, however, for systematic empirical investigations
of the relationship between the hierarchical level and the duration of emotion
component responses.
Process basis of emotion duration

A plethora of theoretical concepts and models has been invoked in theoretical
and empirical work on emotion duration to account for the process basis through
which variability in duration comes about. Yet, taking everything together, the
whole issue has been dealt with in a rather fragmented way.
Psychological concepts and models that have been mentioned include abilities
(such as the ability to distinguish between affects: see Hemenover, 2003),
concepts from learning theory such as reinstatement (which was invoked by
Verduyn et al. (2009) to account for effects of stimulus reappearances) and from
cognitive social learning theories (such as self-efficacy expectations with regard
to mood regulation: see Hemenover, 2003), metaphorical models such as extin-
guishing and stirring up an emotional fire (Verduyn et al., 2011), and concepts
borrowed from attachment theory (Sbarra, 2006).
Finally, also a few biological concepts and models have been advanced to
account for the process basis of emotion duration. At this point, Gilboa and
Revelle (1994) refer to modulation of arousal theory to account for the relations
between emotion duration and dispositional characteristics such as neuroticism
and extraversion. On a more fundamental level, on the basis of a series of
laboratory studies, Davidson (1998) suggests that the prefrontal cortex (in inter-
action with the amygdala) plays a key role in the modulation of the time course
of emotional responding, and in particular of emotion duration.
Concluding remarks
A true account of emotions is impossible without a proper account of their time-
dynamic characteristics, including emotion duration. For a long time, the emotion
domain primarily paid lip service to the duration issue, with sound empirical
findings being few in number. Some recent studies, however, open new perspec-
tives on this topic. Promising aspects of them include more refined distinctions
between determinants of emotion duration on different levels (viz., on the level
of the subject, of the emotion episode as a whole, and of events that occur within
that episode), the use of more valid ways of online data collection, and the use of
more suitable data-analytic methods such as survival analysis (Brans et al., 2010;
Sbarra, 2006; Verduyn et al., 2009, 2011).
Obviously, the studies reviewed in this chapter constitute a mere beginning
of a systematic empirical investigation into the issue of emotion duration, with
many questions at this moment being left unanswered. Ideally such an empirical
endeavor should go hand in hand with the building of a solid theoretical
framework. The latter is not trivial, especially in view of the challenges that go
with the concept of emotion duration.
As to this concept, beyond major tasks for the emotion domain in general (such
as arriving at a suitable definition of the concept of an emotion, and accounting in
an appropriate way for the multi-componential nature of emotions, with compo-
nents belonging to different, multilayered response systems), a particular challenge
is the definition of what constitutes the end point of an emotion episode. This issue
is far from trivial, in part also because it is interwoven with normative concerns,
including when an emotion episode can be considered to be fully processed (in
a “healthy” sense). Another intriguing factor at this point pertains to the way in
which the emotion episode is ended. As shown by Sonnemans and Frijda (1994),
emotion episodes can end because of quite different causes. Future research on
emotion duration will inevitably need to take this factor better into account, as it
may act as an important moderator in any explanatory model of emotion duration.
Acknowledgements
The research reported in this chapter was supported in part by the Research Fund
of KU Leuven (GOA/10/002).
References
Beike, D. R., and Wirth-Beaumont. (2005). Psychological closure as a memory
phenomenon. Memory, 13, 574–593.
Brans, K., Van Mechelen, I., Verduyn, P., and Rimé, B. (2010, April). The effect of social
sharing on the duration of emotional experiences. Paper presented at the Third European
Conference on Emotion, Villeneuve d’Ascq, France.
Davidson, R. J. (1998). Affective style and affective disorders: Perspectives from affective
neuroscience. Cognition & Emotion, 12, 307–330.
Frijda, N. H. (2007). The laws of emotion. Mahwah, NJ: Erlbaum.
Frijda, N. H., Mesquita, B., Sonnemans, J., and Van Goozen, S. (1991). The duration
of affective phenomena or emotions, sentiments and passions. In K. T. Strongman,
International review of studies on emotion (pp. 187–225). Chichester: John Wiley &
Sons.
Gilboa, E., and Revelle, W. (1994). Personality and the structure of affective responses. In
S. Van Goozen, N. E. Van De Poll, and J. A. Sargent (eds), Essays on current issues in
the field of emotion theory (pp. 134–159). Hillsdale, NJ: Erlbaum.
Gross, J. J. (ed.). (2007). Handbook of emotion regulation. New York: Guilford Press.
Hemenover, S. H. (2003). Individual differences in rate of affect change: Studies in
affective chronometry. Journal of Personality and Social Psychology, 85, 121–131.
Sbarra, D. A. (2006). Predicting the onset of emotional recovery following nonmarital
relationship dissolution: Survival analyses of sadness and anger. Personality and Social
Scherer, K. R., and Wallbott, H. G. (1994). Evidence for universality and cultural
variation of differential emotion response patterning. Journal of Personality and Social
Sonnemans, J., and Frijda, N. H. (1994). The structure of subjective emotional intensity.
Cognition & Emotion, 8, 329–350.
Verduyn, P., Delvaux, E., Van Coillie, H., Tuerlinckx, F., and Van Mechelen, I. (2009).
Predicting the duration of emotional experience: Two experience sampling studies.
Emotion, 9, 83–91.
Verduyn, P., Van Mechelen, I., and Tuerlinckx, F. (2011). The relation between event
processing and the duration of emotional experience. Emotion, 11, 20–28.
Part 5
Intervention Perspective
25 Can expressive writing change
emotions?
An oblique answer to the wrong question
James W. Pennebaker and Jason D. Ferrell
University of Texas at Austin
It seems like such a simple question. There is good evidence that expressive
writing methods bring about relatively long-term changes to physical health,
mental health, and, yes, emotions. Unfortunately, the question implies that we
know what we are talking about when we use the term “emotion.” The purpose
of this paper is to challenge the ways scientists think about emotion by drawing
on recent findings in neuroscience, language, and expressive writing.
The problem of emotion and emotion regulation

Emotions do not stand alone. Charles Darwin, William James, Walter Cannon,
Stanley Schachter, and most modern neuroscientists have acknowledged that
emotions are connected to eliciting events. More recent models emphasize that
the perceptions of eliciting events involve the fusing of cognitive and emotional
systems in ways that cannot be separated (Scherer, 2009). Striking examples
have been reported by Antonio Damasio (1994) wherein patients with temporary
frontal lobe damage lost the ability to experience emotion. While playing a
laboratory-based poker game, the emotion loss actually hampered the ability
to play effectively. Although a logical game, Damasio demonstrated that logic
requires emotional responses to judge a risky versus a safe bet.
Emotions and cognitions are ultimately all part of the same neural system. This
is apparent in our research with people who have endured a traumatic or other
upsetting experience. In the months after losing a spouse or being fired from a
job, people continue to obsess about their experience while feeling the emotional
reactions when the thoughts recur. In short, emotional states continue over time
by merely thinking about them.
If a man is still angry about losing his job six months afterwards, can we “fix”
the problem by training him to regulate his emotions? If he reports feeling angry
much of the time, should we be initiating therapy, giving drugs, or encouraging
psychosurgery to reduce the anger? If emotions and the cognitive representations
of an experience are a single system, such a strategy makes no sense. To regulate
emotions, we also need to regulate the experience. That is, if we are seeking
long-term change in emotions, we must regulate the whole system.
184 James W. Pennebaker and Jason D. Ferrell
Expressive writing, emotions, and thoughts
The inherent connection between thoughts and emotions is apparent in work
that has been done using expressive writing. In the typical study, participants
are randomly assigned to write about a traumatic experience or a non-emotional
control topic for three or four sessions, each lasting 15–30 minutes (Frattaroli,
2006; Pennebaker and Chung, 2011). Those assigned to the trauma condition are
asked to explore their deepest thoughts and emotions about the most upsetting
experience in their lives on each day of writing. Control participants are generally
asked to describe non-emotional objects or events such as a tree or how they have
used their time in the last 24 hours.
On the surface, the expressive writing paradigm sounds innocuous. In fact,
participants in the trauma condition usually write intensely personal and powerful
narratives. After each day’s writing, participants usually report feeling emotionally
drained, sad, and sometimes angry. In post-writing interviews, many tell us that
between the writing sessions they tend to think about, and even dream about, their
writing topics.
The success of the expressive writing paradigm has been in its ability to
influence people’s physical and mental health. Since the publication of the first
study in 1986, over 200 articles have been published using the expressive writing
paradigm. Those assigned to the trauma conditions typically evidence better
physical health in the weeks and months after writing, as measured by visits
to physicians, markers of immune function, days absent from work, and other
behavioral and biological measures. The writing paradigm produces consistent
but very modest effects in objective health outcomes over several months.
Meta-analyses report that the effect sizes for behavioral, biological, physical, and
psychological outcomes average d = 0.15 (e.g., Frattaroli, 2006).
Expressive writing produces long-term improvements in emotional states as
well. People assigned to expressive writing conditions have more positive moods
than those assigned to write about non-emotional topics 2–3 months following
writing sessions. People who write expressively have reduced depressive symptoms
compared to participants who write non-emotionally one month following
expressive writing sessions (Sloan and Marx, 2004). Indeed, replication of the
expressive writing paradigm in randomized experimental studies provides evidence
of long-term improvements in emotions in people that write expressively compared
to people who don’t (for meta-analyses, see Frattaroli, 2006; Smyth, 1998).
Expressive writing can change both emotions and physical health over several
weeks and months. But what are the active mechanisms? A large number of
experiments have attempted to answer this question over the last two decades.
Sadly, the answer is not straightforward. Rather, there appears to be a cascade of
inter-related effects. Here is a brief summary of what we currently know:
●● For expressive writing to work, people must write about both their thoughts
and feelings about an upsetting experience. Writing just about the events or
just about their emotions is not sufficient to bring about change.
Can expressive writing change emotions? 185
●● Expressive writing must involve the building of a narrative that allows the
authors to stand back and adopt a broad perspective (e.g., Smyth, True, and
Souto, 2001).
●● Expressive writing allows people to move through the traumatic experience
in a way where they don’t think about it as much. Various studies indicate
that people subsequently have better working memory (e.g., Klein and Boals,
2001) which may partially explain why they later perform better at school
(e.g., Lumley and Provenzano, 2003).
●● Expressive writing brings about subtle but measurable changes in the
ways people socialize with others. In the months after writing, they tend to
talk with others more, laugh more, and are more socially engaged (e.g., Kim,
2008).
Expressive writing as a way of reconstructing an emotional event

In reading people’s emotional essays, very rarely do people ever talk in great
depth about their emotional states. Rather, they write about the emotional event
itself and how it affected their lives and the people around them. Their stories
reveal personal struggles about how they have tried to make sense of the event.
Yes, they often mention the emotions that have come from living through their
upheavals, but the emotions themselves are generally not the problem. The
problem has been the events and their memories.
In the months after participating in our studies, we have often asked people to
describe how the writing studies affected them. The reports are overwhelmingly
positive. Interestingly, the most common responses are cognitive. Examples
include: “It helped me think about what I felt during those times. I never realized
how it affected me before” or “I was able to put things together in ways I couldn’t
do back then.”
When people write, they are rethinking the event itself. As a number of
our computer-based language analyses have demonstrated, those people who
construct or reconstruct what happened are the ones who benefit the most
from writing (e.g., Pennebaker, 2011). For the expressive writer, perception
of the emotional event changes during the cognitive process of rethinking and
reconstructing. The intimate links among perception, sensation, and cognition
presupposes that any change changes the whole system. If cognitions and percep-
tions change, the sensations, emotions, and feelings associated with the event also
change.
Summary
In answer to the title of this paper, yes, expressive writing can change emotions.
But ultimately, the question exposes a misguided view of emotion. Expressive
writing changes the event itself—the way it is thought about, the way it is
organized, and even the way it is remembered. As the event is reshaped in the
person’s mind, the corresponding emotions have to change as well.
186 James W. Pennebaker and Jason D. Ferrell
Can expressive writing change emotions? The more appropriate question is:
can expressive writing change the memory of traumatic experiences? We think
that future research will provide an unequivocal answer in the affirmative.
Acknowledgements
Preparation of this manuscript was aided by funding from the Army Research
Institute (W91 WAW-07-C-0029) and the National Science Foundation
(NSCC-0904822).
References
Damasio, A. R. (1994). Descartes’ error: Emotion, reason, and the human brain. New
York: G. P. Putnam.
Frattaroli. J. (2006). Experimental disclosure and its moderators: A meta-analysis.
Psychological Bulletin, 132, 823–865.
Kim, Y. (2008). Effects of expressive writing among bilinguals: Exploring psychological
well-being and social behaviour. British Journal of Health Psychology, 13, 43–47.
Klein, K., and Boals, A. (2001). Expressive writing can increase working memory
capacity. Journal of Experimental Psychology: General, 130, 520–533.
Lumley, M. A., and Provenzano, K. M. (2003). Stress management through written
emotional disclosure improves academic performance among college students with
physical symptoms. Journal of Educational Psychology, 95, 641–649.
Pennebaker, J. W. (2011). The secret life of pronouns: How our words reveal who we are.
New York: Bloomsbury Press.
Pennebaker, J. W., and Chung, C. K. (2011). Expressive writing: Connections to mental
and physical health. In H. S. Friedman (ed.), Oxford handbook of health psychology.
New York, Oxford University Press.
Scherer, K. R. (2009). The dynamic architecture of emotion: Evidence for the component
process model. Cognition and Emotion, 23, 1307–1351.
Sloan, D. M., and Marx, B. P. (2004). A closer examination of the structured written
disclosure procedure. Journal of Consulting & Clinical Psychology, 72, 165–175.
Smyth, J. M. (1998). Written emotional expression: Effect sizes, outcome types, and
moderating variables. Journal of Consulting and Clinical Psychology, 66, 174–184.
Smyth, J. M., True, N., and Souto, J. (2001). Effects of writing about traumatic experiences:
The necessity for narrative structuring. Journal of Social and Clinical Psychology, 20,
161–172.
26 The powerful impact of mental
imagery in changing emotion
Arnaud Pictet
University of Oxford
Emily A. Holmes
MRC Cognition and Brain Science Unit,
Cambridge
Intrusive mental images powerfully impact on our patients’ emotional state. Such
images may be of distressing memories from the past or feared futures. Within
the field of experimental psychopathology a series of experiments have tested the
assumption that imagery has a special relationship with emotion. Studies using
a cognitive bias modification scenarios paradigm and a picture-word paradigm
will be discussed. Compared to verbal processing, mental imagery was found to
have a more powerful impact on changing emotion for both negative and positive
emotional states. Further examining the role of mental imagery perspective
indicates emotion is only magnified by field (first person) perspective but not
observer (third person) perspective imagery. Clinical implications for changing
emotion include asking about imagery in assessment, reducing the impact of
negative imagery, and finding ways to promote positive/adaptive imagery.
What is mental imagery?

Mental images inhabit our everyday mental life. They can take a range of forms,
from the sudden recollection of past events popping to mind unexpectedly, to
deliberately imagining the future. Images can be fuzzy and fragmented, and
vanish in the blink of an eye. But they can also be vivid and detailed, and seem
clear like actual vision. Mental images are sensory-like experiences in the
absence of direct perceptual stimulation. They can occur in any sensory modality
and have therefore been described as “seeing with the mind’s eye”, “hearing with
the mind’s ear”, and so on (Kosslyn, Ganis, and Thompson, 2001).
Mental imagery in psychopathology

The hallmark clinical examples of mental images are the flashbacks associated
with Posttraumatic Stress Disorder (PTSD). Flashbacks comprise vivid and
distressing images of a traumatic event which intrude into the mind unbidden. In
social phobia, distorted images of the self (e.g., blushing badly in a social situation)
188 Arnaud Pictet and Emily A. Holmes
can be mistaken as an accurate reflection of reality. People with depression may
have insufficient positive imagery, or those with bipolar disorder an excess of
imagined goals when in mania. This led us to suggest that imagery might act as an
“emotional amplifier” in psychopathology, amplifying both negative and positive
emotional states (Holmes, Geddes, Colom, and Goodwin, 2008).
Evidence for the powerful impact of mental imagery on changing

emotion
It has long been assumed that mental imagery has a strong impact on emotion.
However, in the absence of experiments to test this effect, such claims were
dismissed as “clinical anecdote”. A series of experiments will now be discussed,
testing the hypothesis that mental imagery has a more powerful emotional impact
than verbal processing of the same material.
Experimental evidence from a “cognitive bias modification” paradigm using

scenarios
How could we invite people to generate mental imagery in the laboratory? One
possibility was to use a computerized procedure called “cognitive bias modifi-
cation for interpretation” (CBM-I). In CBM-I, participants were encouraged
to generate mental images of auditorily presented scenarios that were initially
ambiguous but then ended up being resolved in either a negative or benign way.
The sentence is ambiguous until the final word provided. An example was: You
have gone to Australia. You are swimming in the sea and see a fin snaking toward
you and realise it is a . . .... A negative resolution would be the word shark, or a
benign resolution dolphin. In an imagery CBM-I condition, participants are given
general training in using mental imagery, for example via a lemon exercise. In this
exercise, participants imagine holding a lemon, looking at it skin, cutting it, and
having the juice squirting into their eye. Then, in the CBM phase, they imagine
approximately 100 scenarios unfold to the outcome. A typical trial consisted of
10–13 seconds of auditory scenario followed by an additional 2-second gap, and
then participants were asked to rate the vividness of their mental image.
Do instructions to use imagery have a more powerful impact on negative emotion
than verbal language? In experiment 1 of Holmes and Mathews (2005), participants
were randomly assigned to one of two negative conditions. In the imagery condition,
they were asked to imagine the CBM-I scenarios through their own eyes, as if they
were actively involved in the situation. In the verbal condition, they listened to
the same scenarios while thinking about the words and meaning. Anxiety pre and
post the procedure was assessed using the State-Trait Anxiety Inventory (STAI:
see Holmes, Lang, and Shah, 2009, for full references of the measures). Anxiety
increases were greater in the imagery (+ 7.3) than verbal (+ 1.9) condition. Findings
were replicated in a second experiment in Holmes and Mathews (2005), with anxiety
increasing more in the negative imagery (+ 5.67) versus verbal (+ 1.55) condition.
Does imagery also have a more powerful impact on positive emotion? Holmes,
The powerful impact of mental imagery in changing emotion 189
Mathews, Dalgleish and Mackintosh (2006) developed CBM-I scenarios resolved
with overtly positive outcomes. An example was: It’s your birthday, and your
partner reaches over to you with a present. You open it and feel incredibly
happy. As before, participants were allocated to imagery or verbal conditions.
Positive affect was assessed using the positive affect subscales of the Positive
and Negative Affect Schedule (PANAS). As predicted, the imagery group showed
greater increases in positive affect (+ 7.15) compared to the verbal group (-6.08).
The corresponding (but opposite) pattern was found for negative affect.
Findings supported the hypothesis that mental imagery had the greater effect
on positive mood. However, we had expected verbal processing to be “weaker”
rather than producing reverse effects. We shall later consider why verbally
thinking about positive scenarios might have a mood worsening effect. First, we
turn to whether the emotion results might transfer to other tasks.
Does the impact of mental imagery on positive emotion extend beyond

immediate state emotion effects and transfer to later test of emotional
vulnerability?
This question holds relevance for depression, as depressive relapse has been
associated with greater reactivity to even small negative mood challenges.
Holmes, Lang and Shah (2009) used a positive CBM-I procedure with imagery
versus verbal conditions. They added a negative mood induction at the end of the
experiment, which involved reading negative statements (e.g., “I’m discouraged
and unhappy about myself”) while listening to depressing music.
The Holmes et al. (2006) findings were replicated, in that mental imagery led to
a greater mood improvement (PANAS = +4.95) than verbal instructions (PANAS
= -2.90). Following the negative mood induction, mood worsened to a lesser
extent in the imagery (PANAS = -4.90) than verbal group (PANAS = -10.90).
This suggested that positive mental imagery was relatively protective against the
mood deteriorating effect of a stressor.
Why might verbally thinking about positive scenarios have a mood worsening
effect?
Two studies discussed above (Holmes, et al., 2009 experiment 1; Holmes, et al.,
2006) provided the unexpected result that verbal processing of positive material
had a paradoxical negative impact on positive emotion. Why might this be?
One possibility was that verbal instructions may prompt participants to make
unfavorable verbal comparisons between themselves and the overtly positive
scenarios. By contrast, mental images can take longer to generate and thus
making comparisons may be more effortful. To test this hypothesis, Holmes et al.
(2009, experiment 2) created two new verbal conditions which aimed to either
increase or decrease the amount of comparisons being made.
In the “verbal comparisons condition”, the original verbal instructions were
supplemented with the request to actively compare each scenario with how
things were in reality for the participant. In the “verbal reduced comparisons
condition”, the instruction to “focus on the meaning” was removed and the time
available to make comparisons reduced. As predicted, the “verbal comparisons
condition” led to greater increases in anxiety (STAI = + 3.5) than those in the
two other conditions (STAI = -1.55 in the “verbal reduced-comparisons” and =
-2.70 in the imagery condition). This supported the hypothesis that unfavorable
self-comparisons may have contributed to findings of mood deterioration in the
original verbal condition.
Further experimental evidence for the impact of imagery on emotion: using a

picture-word cue paradigm
To date, our overarching conclusion about the emotional power of mental imagery
might be challenged by the fact that the CBM-I stimuli used were all verbally based.
As all participants had to listen to the event descriptions before going on to produce
a mental image or to focus on their verbal meaning, it could be argued that the differ-
ences observed may merely have resulted from the recruitment of an additional
representational system in the imagery condition. In other words, the greater emotional
impact of the imagery condition relative to the verbal condition could have been due to
more processing resources being engaged rather than to the effect of imagery per se.
Holmes et al. (2008) therefore developed an experimental paradigm drawing
on the evaluative learning literature that combined both perceptual (pictures) and
verbal (words) cues. In this picture-word paradigm, participants were asked to
combine pictures with an associated caption using either an image (i.e., “imagine
the combination of the next picture and word”) or a verbal statement (i.e.,
“make a sentence combining the next picture and the word”), within alternating
experimental blocks of negative or positive combinations. An example included a
picture showing the view from a high bridge, with the word “leap” in the negative
and “view” in the positive condition.
Imagery had a mood amplifying effect relative to verbal condition for both the
negative and positive stimuli (negative blocks, STAI = +7.06 vs. +2.43; positive
blocks STAI = -3.81 vs. +1.0). A similar pattern was shown on liking ratings for a
subsample of pictures that were displayed alone before and after the picture-word
combination task (akin to an evaluative conditioning effect). Overall, findings
provided convergent evidence to the CBM-I studies for the greater impact of
imagery compared to verbal processing on emotion.
The impact of mental imagery perspective on emotion

A caveat in the broad conclusion about imagery and emotion on which we are
converging, is that not all images have an equal impact. Mental images can
be viewed either from a “field” perspective, whereby one visualizes the event
through one’s own eyes as if actively involved, or an “observer” perspective,
whereby one is watching oneself from the outside and experiencing the event
as a bystander. Observer perspective images are thought to be less emotional. In
the previous CBM-I studies, field perspective instructions were used. Would the
emotion effects be removed by observer perspective imagery?
Holmes, Coughtrey and Connor (2008) used a positive CBM-I paradigm
contrasting field perspective, observer perspective or verbal instructions. As
predicted, field perspective enhanced positive affect more (PANAS = + 4) than
imagining them from an observer perspective (-4.8) or thinking verbally (-4.7).
This suggests that field perspective imagery needs to be harnessed in order to
enhance positive affect.
A heuristic model for the powerful impact of mental imagery versus

verbal representations on emotion
Potential explanations about why mental imagery has a particularly powerful
impact on emotion have been discussed by Holmes and Mathews (2010). As
illustrated in Figure 8, one compelling hypothesis is that mental imagery has
direct access to emotional systems in the brain because of its similarity with actual
perception (see also De Houwer and Hermans, 1994). Compared to real perception,
imagining emotional stimuli has been found to activate similar neural activity
and to produce a similar physiological response. Sensory stimuli and thus images
may have privileged access to emotional systems. This is thought to apply to both
involuntary mental images (e.g., PTSD flashbacks) and deliberately generated
images (e.g., voluntarily imagining a snake). While intrusive imagery is automati-
cally elicited (bottom up) when a sensory cue partially matches a representation in
episodic memory, voluntary imagery is generated via top-down control processes.
In contrast to imagery (whether involuntary or deliberately generated) verbal
language may link to other knowledge in semantic memory but shows little overlap
with actual perception (for further detail, see Holmes and Mathews, 2010)
Steps from the lab towards clinical translation about positive

imagery: an example of promoting positive imagery in depressed
mood
Does the impact of positive mental imagery extend to a dysphoric sample and
transfer to behavior?
Depressed mood has been associated with a lack of positive imagery (Holmes,
Lang, Moulds, and Steele, 2008). Could generating mental imagery from overtly
positive picture-word cues also improve positive mood in a dysphoric sample?
Mentally imagining one’s own future behavior can increase the likelihood of that
behavior being later enacted in real life (for a review, see Holmes and Mathews,
2010), so could the effect of generating positive imagery transfer to increased
action (see Figure 8)?
Pictet, Coughtrey, Mathews and Holmes (2011) administered a picture-word
paradigm (similar to that described earlier) to individuals with mildly depressed
mood (mean = 17 on the Beck Depression Inventory-II). Participants were
Top down
control
processes
Select from
Verbal representation (Re) Constructed

Autobiographic and
of emotional meaning image of emotional
semantic memory
instance
knowledge base
Matching
Little overlap with Bottom up sensory cue Processing overlaps

processing of with perceived events
perceived events
Direct contact with

Associates: action emotional systems
Contact with
readiness, believability,
other semantic
attitude to self etc....
knowledge
Figure 8 The construction of imagery versus verbal representations, related to their

relative impact on emotion (taken from Holmes & Mathews, 2010).
assigned to either a positive, negative, or “neutral” condition. Results showed

clear and predicted mood effects (increasing in the positive condition and
decreasing in the negative relative to neutral). Thus, participants with dysphoria
were able to benefit in mood improvements from the positive imagery procedure.
To investigate any transfer to later performance, a fishing game task was used.
This toy consisted of colored plastic fish moving around in a circle, opening and
closing their mouths to reveal a magnet. Participants were instructed to catch as
many fish as possible in 2.5 minutes, by “hooking” them using a magnet attached
to the end of a plastic fishing rod. The fishing game task was chosen as a simple
behavioral performance measure assumed to tap behavior negatively associated
with depressed mood, such as approach motivation and persistence. As predicted,
participants in the positive condition caught significantly more fish than those in
the neutral condition. Although preliminary, our findings suggest the possibility
that positive imagery may be used in depression not only to increase positive
affect, but also to boost engagement in goal-oriented behaviors.
Clinical case series and a small clinical study

Blackwell and Holmes (2010) used a single case series approach with seven
patients with major depressive disorder. Each received seven sessions of positive
imagery CBM-I administrated once daily. Overall, there were large effect sizes
for improvement in depressive symptoms (as assessed by the Beck Depression
Inventory-II). Improvements extended to a follow-up two weeks later. These
results provided initial support for the potential of promoting a habit to generate
positive imagery in improving depressed mood. In the absence of a comparison
condition however, it was not possible to rule out “non-specific” factors as well
as spontaneous recovery.
In a next study, we compared the impact of repeated sessions of positive imagery
CBM-I versus a control condition in a sample of currently depressed patients (Lang,
Blackwell, Harmer, Davison, and Holmes, 2012). The positive condition again led
to relative improvements in depressive symptoms from pre- to post-treatment
compared to the control condition, maintained at the two-week follow-up.
Although encouraging, translational research about promoting the tendency to
generate positive imagery in depression is still in its infancy. Results need to be
replicated with a larger sample size. However, initial findings suggest that experi-
mental psychopathology paradigms may hold promise in treatment innovation for
depression. Encouraging patients to habitually generate vivid positive/adaptive
images about the future might provide an adjunct to existing treatments to
promote positive emotional change.
Conclusion and clinical implications

The long held clinical assumption that imagery has a strong impact on changing
emotion has been supported by experimental psychopathology research. Mental
imagery has a powerful impact on amplifying emotion; that is, exacerbating
negative emotional states or conversely increasing positive emotions. By contrast,
verbal processing has been found to produce less emotional change, or at worst,
mood deterioration when confronted with positive information. It was noted that
field rather than observer perspective imagery was required.
What are the potential clinical implications? In the context of traditional CBT
being dominated by “talking” interventions, these experimental findings shed
light on the potential of using mental imagery to access and modify emotional
dysfunction (Hackmann, Bennett-Levy, and Holmes, 2011). Clinical implications
thus include asking about imagery in assessment across disorders, reducing the
impact of negative imagery, and promoting more positive/adaptive imagery. In
terms of future treatment innovation using emerging technologies, examples of
computerized interventions designed to promote positive future-oriented imagery
in the context of depression might be possible.
Acknowledgements
Emily A. Holmes is supported by the Wellcome Trust Clinical Fellowship
(WT088217), The Lupina Foundation, The Medical Research Council, and the
National Institute for Health Research (NIHR) Oxford Biomedical Research Centre
based at Oxford University Hospitals Trust Oxford University. The views expressed
are those of the author(s) and not necessarily those of the NHS, the NIHR or the
Department of Health. Arnaud Pictet is supported by a grant from the Swiss National
Science Foundation (140104). The authors would like to thank the other members of
the Experimental Psychopathology and Cognitive Therapy (EPaCT) research group.
References
De Houwer, J., and Hermans, D. (1994). Differences in the affective processing of words
and pictures. Cognition & Emotion, 8, (1), 1–20.
Hackmann, A., Bennett-Levy, J., and Holmes, E. A. (2011). Oxford guide to imagery in
cognitive therapy. Oxford: Oxford University Press.
Hackmann, A., Clark, D. M., and McManus, F. (2000). Recurrent images and early
memories in social phobia. Behaviour Research and Therapy, 38, (6), 601–610.
Holmes, E. A., Coughtrey, A. E., and Connor, A. (2008). Looking at or through rose-tinted
glasses? Imagery perspective and positive mood. Emotion, 8 (6), 875–879.
Holmes, E. A., Geddes, J. R., Colom, F., and Goodwin, G. M. (2008). Mental imagery
as an emotional amplifier: Application to bipolar disorder. Behav Res Ther, 46, (12),
1251–1258.
Holmes, E. A., Lang, T. J., Moulds, M. L., and Steele, A. M. (2008). Prospective and
positive mental imagery deficits in dysphoria. Behaviour Research and Therapy, 46,
(8), 976–981.
Holmes, E. A., Lang, T. J., and Shah, D. M. (2009). Developing interpretation bias modifi-
cation as a ‘cognitive vaccine’ for depressed mood – Imagining positive events makes
you feel better than thinking about them verbally. Journal of Abnormal Psychology,
118, (1), 76–88.
Holmes, E. A., and Mathews, A. (2005). Mental imagery and emotion: A special
relationship? Emotion, 5, (4), 489–497.
—(2010). Mental imagery in emotion and emotional disorders. Clin Psychol Rev, 30, (3),
349–362.
Holmes, E. A., Mathews, A., Dalgleish, T., and Mackintosh, B. (2006). Positive inter-
pretation training: effects of mental imagery versus verbal training on positive mood.
Behavior Therapy, 37, (3), 237–247.
Holmes, E. A., Mathews, A., Mackintosh, B., and Dalgleish, T. (2008). The causal effect of
mental imagery on emotion assessed using picture-word cues. Emotion, 8, (3), 395–409.
Kim, S. E., Kim, J. W., Kim, J. J., Jeong, B. S., Choi, E. A., Jeong, Y. G., Ki, S. W. (2007).
The neural mechanism of imagining facial affective expression. Brain Research, 1145,
128–137.
Kosslyn, S. M., Ganis, G., and Thompson, W. L. (2001). Neural foundations of imagery.
Nature Reviews Neuroscience, 2, (9), 635–642.
Lang, T. J., Blackwell, S. E., Harmer, C. J., Davison, P., and Holmes, E. A. (2012).
Cognitive bias modification using mental imagery for depression: developing a novel
computerized intervention to change negative thinking styles. European Journal of
Personality, 26. 145–157.
Nigro, G., and Neisser, U. (1983). Point of view in personal memories. Cognitive
Psychology, 15, (4), 467–482.
Pictet, A., Coughtrey, A. E., Mathews, A., and Holmes, E. A. (2011). Fishing for happiness:
The effects of positive imagery on interpretation bias and a behavioral task. Behaviour
Research and Therapy, 49, (12), 885–891.
27 Cognitive mechanisms involved in
therapeutic change for depression
Reducing abstraction and increasing
concreteness
Edward R. Watkins
University of Exeter
An extensive literature has provided support for the hypothesis that the way that
individuals construe or represent personally relevant events has implications
for changing emotions. Indeed, this hypothesis underlies a significant tranche
of cognition-emotion theory as well as therapeutic approaches, most notably
cognitive-behavior therapy (CBT e.g., Beck, Rush, Shaw and Emery, 1979).
This chapter focuses on programmatic research into one aspect of cognitive
processing—the specificity and concreteness of mental representations—and its
impact on emotions, with particular emphasis on depression. The main thesis of
the chapter is that representation of negative emotional events in a more concrete
and specific way tends to result in less emotional reactivity and quicker recovery
from upsetting events, but that depressed individuals instead tend to process
such events in a more abstract and general way, contributing to their emotional
vulnerability.
Abstract versus concrete processing: theory and evidence

Watkins (2008, 2011) elaborated existing control theory accounts (Carver and
Scheier, 1990) to propose that personally relevant events and actions can be
represented at different levels of abstraction and specificity, with these different
levels of abstraction having implications for self-regulation and emotions. This
theory proposes that goals, events, and actions can be mentally represented
within a hierarchical framework that ranges from more abstract to more concrete
levels of identification. A more abstract level involves general, superordinate,
and decontextualized mental representations that convey the meaning of goals,
events, and actions, “why” an action is performed, and its purpose, ends, and
consequences. A more concrete level involves specific and subordinate mental
representations that include contextual details of goals, events, and actions, and
the specific “how” details of the way an action or event occurs. These opera-
tionalizations of abstraction versus concreteness are complex and potentially
incorporate several dimensions whose relative importance and independence
remains to be determined (e.g., how vs. why, verbal vs. imagery; see Watkins,
2011, for further discussion). Control theory proposes that identifications at the
196 Edward R. Watkins
most abstract levels may represent a global sense of idealized self (i.e., a decon-
textualized, superordinate meaning capturing the essence of the self), which in
turn sets the broad principles that organize goals and behavioral standards across
multiple situations (e.g., to be an honest person), whereas the reference values at
the more concrete levels represent the specific actions and behavioral programs
and sequences necessary to implement the principles in a particular situation.
Control theory accounts propose that an individual can only focus attention on
one level in the hierarchy at any moment in time resulting in that particular level
(or mode) being functionally and operationally superordinate at that moment,
thereby determining the prepotent processing style, i.e., a current focus on higher
levels in the hierarchy results in an abstract processing style.
Within this account, the potential advantages and disadvantages of an abstract
versus concrete level of processing center around their relative sensitivity to
contextual and situational detail. Relative to a concrete level of identification, an
abstract identification insulates an individual from the specific context, making
the individual less distractible, less impulsive, and enabling more consistency
and stability of goal pursuit across time, but also making the individual less
responsive to the environment and to any situational change, and providing fewer
specific guides to action and problem-solving because of their distance from the
mechanics of action. Moreover, representing a difficulty at a more abstract level
is hypothesized to have greater emotional impact because it will be related to a
central element of the self whereas a more concrete representation will relate to
more peripheral representations and less personally significant goals, such as
behaviors. Thus, it is hypothesized that processing a negative outcome (or poor
goal progress) at a more abstract level of goal/action identification may intensify
its negative emotional impact (Watkins, 2008, 2011).
Consistent with this account, there is experimental evidence that manipulating
specificity and abstraction influences emotional reactivity to a subsequent stressful
task. For example, participants who practiced focusing on imagined emotional
scenarios in a specific and concrete way (“focus on how it happened, and imagine in
your mind as vividly and concretely as possible a ‘movie’ of how this event unfolded”)
demonstrated smaller decreases in self-reported positive affect and smaller increases
in negative affect following a subsequent failure on an insoluble anagram task,
compared to participants who practiced more abstract processing (“think about why
it happened, and analyze the causes, meanings, and implications of this event”) when
focusing on the same emotional scenarios (Watkins, Moulds, and Moberly, 2008).
Similarly, Watkins (2004) randomly allocated participants to expressive
writing about a previously induced failure in either an abstract way (e.g., “Why
did you feel this way?”) or a concrete way (e.g., “How did you feel moment-by-
moment?”). Processing style influenced emotional recovery from the failure: At
higher levels of trait rumination, levels of negative mood 12 hours after the failure
were greater, but only in individuals who wrote abstractly, and not in individuals
who wrote concretely.
These different processing styles have also been implicated in rumination,
which itself is a major cognitive process implicated in the onset and maintenance
Cognitive mechanisms involved in therapeutic change for depression 197
of depression (Watkins, 2008). There is increasing experimental evidence that
there are distinct forms of rumination with distinct functions and consequences: a
helpful style characterized by concrete and specific thinking versus a pathological
style characterized by abstract, global thinking (Watkins, 2008).
The impact of specific versus overgeneral processing on emotion has also been
investigated with respect to autobiographical memory. Experimental studies have
demonstrated that manipulating memory specificity causally influences emotional
response. For example, voluntarily recalling an emotional event in specific detail
produces less emotional response than recalling it at a more general level (e.g.,
Philippot, Baeyens, and Douilliez, 2006). Likewise, practice at recalling specific,
contextualized autobiographical memories reduces the negative experience to a
subsequent stressful task relative to practice at recalling general, decontextualized
memories (Raes, Hermans, Williams, and Eelen, 2006).
Abstraction, specificity, and depression

The theory and empirical findings noted above lead to the prediction that an increased
tendency to adopt a more abstract processing style in response to difficulties would
be associated with greater emotional reactivity, more sustained negative affect, and
increased psychopathology. Consistent with this, Watkins (2011) reviewed evidence
suggesting that mood and anxiety disorders involve more abstract, general processing
of negative events. There is robust evidence that depression is characterized by
an increased tendency away from specificity and towards abstract, overgeneral
thinking. For example, individuals with depression demonstrate increased overgen-
eralization, in which a general rule or conclusion is drawn on the basis of isolated
incidents and applied across the board to related and unrelated situations (Beck et
al., 1979; Carver, 1998), i.e., an abstract, decontextualized identification focused on
a global implication. In addition, individuals who are currently depressed or have a
history of depression tend to find it harder to recall specific personal memories to
cue words than controls, instead recalling more categoric or overgeneral memories
(Williams et al., 2007). Individuals with depression also report increased abstract
rumination. Importantly, overgeneralization, overgeneral autobiographical memory,
and rumination all prospectively predict subsequent levels of depression (Carver,
1998; Watkins, 2008; Williams et al., 2007).
Specificity in therapy
A further implication of this theory and associated experimental findings is that
increasing concreteness and specificity should be beneficial within treatments for
depression. This was recognised early in the development of CBT, as illustrated
by the following quotation from the seminal manual for depression (Beck et al.,
1979: 79):
When the therapist poses a question or makes a comment, it should be based

on a definite rationale derived from the framework of cognitive therapy; it
should be phrased to elicit concrete information. We have found that specific
questions most clearly help to delineate the patient’s problem area. General,
abstract, vague questions often produce similarly vague responses removed
from the “hard” data of cognitions. In addition, vague phrases tend to confuse
or unsettle the patient and are more easily interpreted in a negative manner
by the patient. For example, if the therapist wants to know about the patient’s
ideation, he would ask questions such as, “What are you thinking right now?”
or “How do you feel?” Specific direct questions facilitate specific direct
responses. A comment such as, “I wonder what you have been thinking” is
less likely to elicit specific cognitions that a more concrete question such as,
“Try to recall just what words and pictures went through your mind right at
the time of that event”.
Consistent with this advice, there is evidence from clinical trials that increasing
specificity of thinking is a potential mechanism-of-action by which CBT reduces
depressive symptoms. Concrete treatment techniques within CBT, such as asking
for specific examples of difficult events, predict subsequent symptom reduction
when assessed early in CBT, whereas more abstract techniques do not (DeRubeis
and Feeley, 1990; Feeley et al., 1999). Likewise, patient improvement by
midpoint of therapy in the use of situational analysis, which involves generating a
specific description of the context relevant to a particular problem and generating
specific goal-oriented behaviors, predicts reduced depression at the end of a CBT
intervention (Manber et al., 2003). In addition, CBT and mindfulness-based CBT
have been demonstrated to increase the specificity of autobiographical memory
recall (McBride et al., 2007; Williams et al., 2000).
If this specificity-as-mechanism hypothesis is correct, then a treatment inter-
vention that specifically and exclusively focuses on increasing specific and
concrete thinking should be effective at reducing depressed symptoms. A recent
study provided a proof-of-principle test of this hypothesis by randomizing
dysphoric participants to an active intervention designed to increase specificity
(concreteness training), a bogus training condition that lacked elements to
increase specificity but was matched for treatment rationale, therapist contact, and
other non-specific factors, or a waiting list control (Watkins, Baeyens, and Read,
2009). The concreteness training consisted of explicit instructions to actively
engage in being specific (e.g., focusing on the specific sensory details of an event,
on what makes each event specific, unique and distinctive, and on the process
of how the event and behaviors unfolded) when imagining emotional events,
both standard vignettes and personal autobiographical memories. These instruc-
tions were derived from the experimental materials used in Watkins et al. (2008)
described above. Participants in the concreteness training condition practiced
this 30-minute exercise every day for a week, using an audio-recording of the
exercise. The bogus training condition consisted of repeated daily practice on a
computerized task that presented short written descriptions of social situations
that remained ambiguous in meaning until the final word, presented as a fragment
to be completed, which always resolved the meaning in a plausibly concrete way.
Whilst involving materials that had face validity for influencing specificity and
sharing the same explanation as concreteness training concerning the value of
becoming more specific, the bogus training differed from concreteness training
in not involving participants in (i) actively generating a vivid and sequential
description of the event (“How it happened?”), and (ii) focus on personal autobio-
graphical descriptions, and therefore was not expected to directly alter the degree
of concrete processing.
Consistent with the hypothesis that increased specificity of thinking may
be a mechanism-of-action responsible for symptom reduction, Watkins et al.
(2009) found that the concreteness training condition produced greater symptom
reduction on the Hamilton Rating Scale for Depression than both the bogus
training and waiting-list controls. Moreover, the concreteness training condition
resulted in more specific descriptions of problems than the other two conditions
and significantly greater reductions in rumination than the waiting-list control.
Thus, these findings provide proof-of-principle that increased specificity of
processing can reduce depressive symptoms and, as such, are consistent with
the hypothesis that CBT may work, at least in part, by increasing specificity of
processing.
This concreteness training approach has subsequently been developed into a
guided self-help treatment (using a booklet and audio CD) that involves daily
practice for at least six weeks, a two-hour face-to-face coaching session, and
up to three 30-minute follow-up phone calls to motivate and guide patients.
This intervention was investigated in a Phase II randomized controlled trial in
patients with a diagnosis of major depression recruited from primary care. One
hundred and twenty one patients were randomly allocated to treatment-as-usual
alone (TAU, including clinical management and antidepressant medication),
TAU plus the guided self-help concreteness training, or to TAU plus an active
attention control treatment (relaxation guided self-help), matched in structure,
rationale, and therapist contact to the concreteness training (Watkins et al.,
2012). Relative to TAU, TAU plus concreteness training significantly reduced
symptoms of depression and rumination, indicating that a treatment exclusively
focused on increasing specificity and concreteness can be efficacious. There was
no differential treatment effect of concreteness training versus relaxation training
on depressive symptoms, which raises the possibility that the treatment benefits
were due to non-specific effects such as a supportive therapist and providing
an alternative coping response to difficulties. Nonetheless, the concreteness
training condition reduced rumination and a negative overgeneral attributional
style significantly more than relaxation, suggesting that there may be a distinct
cognitive mechanism and cognitive benefit unique to concreteness training.
Further evidence for the potential value of increasing concreteness and speci-
ficity is provided by preliminary positive findings from trials utilizing training in
recall of specific autobiographical memory as a treatment for depression (Raes
et al., 2009; Serrano et al., 2004). Together, these approaches can be viewed
as a more explicit elaboration of an element within CBT, namely, encouraging
patients to describe situations in specific and concrete detail. As such, it is
possible that some of the benefits observed for full CBT come from the effects
of the concreteness training that implicitly occurs during CBT. One avenue for
future research is a dismantling study of CBT in which the specificity element
is compared to other elements of CBT such as thought challenging. If the speci-
ficity-as-mechanism hypothesis is further supported, it would suggest the value of
CBT becoming even more explicitly focused on making both therapist and patient
more specific.
Conclusion
There is now convergent evidence indicating the role of abstraction and specificity
in emotion change, especially in depression. Cross-sectional and prospective
longitudinal studies indicate that people with depression tend to have less specific
and more abstract representations of personally relevant events. Experimental
studies in undergraduate samples have shown that manipulating the abstraction
and specificity of mental representations influences emotional change. Change in
specificity is associated with symptom improvement in process-outcome studies
of CBT for depression. Pilot controlled trials have demonstrated that directly
training individuals to become more concrete leads to symptom reduction.
Together, this body of evidence suggests that reducing abstraction and increasing
specificity may be an important cognitive mechanism in therapeutic change for
depression.
Nonetheless, it is clear that this program of research is still in its initial stages,
with the findings to date providing necessary but not sufficient evidence to
support these hypotheses, such that we need to be tentative about these conclu-
sions. However, hopefully this research program indicates the potential value of
experimental research into identified vulnerability processes (e.g., specificity of
representations) which then leads into more targeted clinical interventions. This
program of research also indicates the potential value of developing interventions
that are focused on specific putative mechanisms as means to further clarify our
understanding of how therapy works and, thereby, to improve the efficacy of
treatments.
Acknowledgements
This chapter presents research funded by a NARSAD Young Investigator Award, a
UK Medical Research Council Experimental Medicine Grant, and two Wellcome
Trust Project Grant (GR065809 and GR080099), all held by Dr Edward Watkins.
References
Beck, A. T., Rush, A. J., Shaw. B. F., and Emery, G. (1979). Cognitive therapy of
depression. New York: Guilford Press.
Carver, C. S. (1998). Generalization, adverse events, and development of depressive
symptoms. Journal of Personality, 66, 607–619.
Carver, C. S. and Scheier, M. F. (1990). Origins and Functions of Positive and Negative
Affect – A Control-Process View. Psychological Review, 97, 19–35.
DeRubeis, R. J., and Feeley, M. (1990). Determinants of change in cognitive therapy for
depression. Cognitive Therapy and Research, 14, 469–482.
Feeley, M., DeRubeis, R. J., and Gelfand, L. A. (1999). The temporal relation of adherence
and alliance to symptom change in cognitive therapy for depression. Journal of
Consulting and Clinical Psychology, 67, 578–582.
MacBride, C., Segal, Z., Kennedy, S., and Gemar, M. (2007). Changes in autobiographical
memory specificity following cognitive behavior therapy and pharmacotherapy for
major depression. Psychopathology, 40, 147–152.
Manber, R., Arnow, B., Blasey, C., Vivian, D., McCullough, J. P., Blalock, J.A., et al.
(2003). Patient’s therapeutic skill acquisition and response to psychotherapy, alone or in
combination with medication. Psychological Medicine, 33, 693–702.
Philippot, P., Baeyens, C., and Douilliez, C. (2006). Specifying emotional information:
Regulation of emotional intensity via executive processes. Emotion, 6, 560–571.
Raes, F., Hermans, D., Williams, J. M. G., and Eelen, P. (2006). Reduced autobiographical
memory specificity and affect regulation. Cognition & Emotion, 20, 402–429.
Raes, F., Williams, J. M. G., and Hermans, D. (2009). Reducing cognitive vulnerability
to depression: A preliminary evaluation of MEmory Specificity Training (MEST) in
inpatients with depressive complaints. Journal of Behavior Therapy and Experimental
Serrano, J. P., Latorre, J. M., Gatz, M., and Montanes, J. (2004). Life review therapy using
autobiographical retrieval practice for older adults with depressive symptomatology.
Watkins, E. (2004). Adaptive and maladaptive ruminative self-focus during emotional
processing. Behaviour Research and Therapy, 42, 1037–1052.
—(2008). Constructive and unconstructive repetitive thought. Psychological Bulletin, 134,
163–206.
Watkins, E. R. (2011). Dysregulation in Level of Goal and Action Identification Across
Psychological Disorders. Clinical Psychology Review, 31, 260–278.
Watkins, E. R., Baeyens, C. B., and Read, R. (2009). Concreteness training reduces
dysphoria: proof-of-principle for repeated cognitive bias modification in depression.
Journal of Abnormal Psychology, 118, 55–65.
Watkins, E. R., Moulds, M., and Moberly, N. J. (2008). Processing mode causally influ-
ences emotional reactivity: Distinct effects of abstract versus concrete construal on
emotional response. Emotion, 8, 364–378.
Watkins, E. R., Taylor, R. S., Byng, R., Baeyens, C., Read, R., Pearson, K., and Watson, L.
(2012). Guided self-help concreteness training as an intervention for major depression
in primary care: a randomized controlled trial. Psychological Medicine.
Williams, J. M. G., Barnhofer, T., Crane, C., Hermans, D., Raes, F., Watkins, E. et al.
(2007). Autobiographical memory specificity and emotional disorder. Psychological
Bulletin, 133, 122–148.
Williams, J. M. G., Teasdale, J. D., Segal, Z. V., and Soulsby, J. (2000). Mindfulness-based
cognitive therapy reduces overgeneral memory in formerly depressed patients. Journal
of Abnormal Psychology, 109, 150–155.
28 A functional approach to the study of
human emotion
The centrality of relational processes
Dermot Barnes-Holmes and Sean Hughes
National University of Ireland Maynooth
Introduction
How stimuli come to acquire, maintain, and change their emotional properties
or functions has long captured the empirical and theoretical attention of learning
psychologists. For cognitive researchers, this question has often involved examining
the role that associative learning processes play in the development, persistence and
generalisation of fear, phobias, likes and dislikes, anxiety, avoidance, and disgust
(see other chapters in the current volume for a detailed overview). Some associative
models also incorporate indirect learning processes (e.g., vicarious conditioning) or
propositions and expectancies to account for human emotional learning. At its core,
this work is often guided by a mechanistic, mediational approach to psychological
science—the goal of which is to understand the mental processes or representations
that give rise to and guide behavior (e.g., associations). At the same time, a second
group of researchers have drawn upon an alternative philosophically framework
known as functional contextualism (referred to hereafter as the “functional approach”)
in order to understand many of these same phenomena. Unlike their cognitive
colleagues, functionally orientated researchers explain emotional behaviour exclu-
sively in terms of the interactions that occur between people in and with their past
and present environments without making reference to any mental constructs.
The present chapter aims to introduce readers to the functional approach and
explain how it is currently being applied to the study of human emotion. To provide
a context and rationale for the procedures and theory to follow, we begin by briefly
outlining the core assumptions of this framework as well as highlighting its points
of departure from modern cognitive psychology. We then draw on Relational Frame
Theory (RFT; Hayes, Barnes-Holmes, and Roche, 2001) to illustrate how both the
emergence of, and changes in, emotional responding can be understood from this
perspective. Finally, we end by considering how this work has provided the foundation
for a “third-wave” behavioral therapy termed Acceptance and Commitment Therapy
(ACT; Hayes, Luoma, Bond, Masuda and Lillis, 2006; Gaudiano, 2011).
The functional approach: an overview

The functional approach to psychology is guided by three primary goals: to
understand, predict, and influence behavior by identifying and then directly
A functional approach to the study of human emotion 203
manipulating features of the individual’s environment. The aim here is not
to build ever-increasing “lists” of behavioral effects, but rather abstract out
the environmental causes of behavior into general “behavioral principles”
that apply across a wide range of different situations (e.g., the principles of
reinforcement, punishment, and stimulus generalization). Functional theories
are established when researchers explain a range of different behavioral
interactions by drawing on a set of inter-related behavioral principles. Such
theories are not driven by mediating mental constructs—instead they describe
and seek to explain observed regularities in the functional relations between
environmental events and behavior. The validity and utility of any functional
account is therefore based on three pragmatic criteria—its ability to predict and
influence behaviour with scope (explain a comprehensive range of behaviours
across a variety of situations), precision (applying a restricted set of principles
to any event) and depth (cohere across analytical levels and domains such as
biology, psychology, and anthropology). Finally, and contrary to popular belief,
functional researchers are interested in all types of behavior that occur both
inside and outside the skin. Publicly observable as well as private behaviors
(e.g., thinking, feeling, imagining, and remembering) can both be subjected to
a detailed functional analysis, and researchers have refined such an approach
for several decades now.
An example of a functional theory that seeks to explain a wide variety of public
and private behaviors using only a handful of interrelated principles is RFT (Hayes
et al., 2001). This account suggests that humans are capable of learning in ways
that differ markedly from non-humans. Specifically, our ability and tendency to
relate stimuli bi-directionality (see below) allows for the emergence of complex
untrained relations that cannot be traced to a history of direct training or learning.
According to RFT, this form of relational learning emerges early on in our devel-
opment through interactions with the verbal community and is a defining element
of both human language and cognition. In the following section, we argue that a
systematic analysis of these relational abilities may prove essential for explaining
how a diverse range of emotional properties can be acquired and changed without
any prior training or instruction.
Relational frame theory and the emergence of human emotional

responding
To illustrate a functional approach to the study of complex and “novel” human
behavior, consider the following experimental procedure known as Matching-
To-Sample (MTS). Participants are shown a computer screen that contains an
abstract shape or nonsense word at the top of the screen and three other arbitrary
shapes or nonsense words at the bottom of the screen1. Through trial and error
1 Although visual stimuli are often used, derived relational responding has also been demonstrated
with auditory, olfactory, haptic and gustatory modalities.
204 Dermot Barnes-Holmes and Sean Hughes
learning they are taught to select one “comparison” stimulus (termed B here
for convenience) when presented with a “sample” stimulus (A) at the top of the
screen (i.e., match A to B). If verbally able humans are then presented with B
as the sample stimulus and A as one of three different stimuli at the bottom of
the screen, they will generally choose the latter despite having never received
any reinforcement or feedback for doing so. If choosing B in the presence of A
is explicitly taught, participants will spontaneously derive the reverse relation
(e.g., select A in the presence of B). When the same participants are then taught a
second stimulus relation—such as choosing C in the presence of B—the number
of relations that they derive increases. For example, if a person is taught to match
A to B and B to C they may subsequently match B to A and C to B (symmetry
relations) as well as A to C and C to A (equivalence relations). This spontaneous
emergence of a specific set of novel, untrained and bi-directional relations among
stimuli is termed stimulus equivalence (Sidman, 1994).
The importance of derived stimulus relating to emotion researchers rests on
one final behavioral process termed transfer of function. This refers to the finding
that when a set of relations are established among stimuli, the psychological or
emotional functions of one stimulus may alter the emotional functions of the
other related stimuli in the absence of any training or instruction. Imagine, for
example, that an equivalence relation is established among stimuli (A, B, C, D)
and then, using a Pavlovian conditioning procedure stimulus B is paired with an
electrical shock. When participants subsequently encounter the C and D stimuli
they typically report fear and produce signs of physiological arousal even though
these stimuli were never directly associated with an emotional event (e.g.,
Auguston and Dougher, 1997). This effect has now been replicated across a range
of different emotions including anxiety (Smyth, Barnes-Holmes and Forsyth,
2006), fear (Valverde, Luciano and Barnes-Holmes, 2009), mood states (Barnes-
Holmes, Barnes-Holmes, Smeets and Luciano, 2004), and sexual arousal (Roche
and Barnes, 1997). However, as we shall see in the next section, the transfer of
psychological functions is not restricted to equivalence relations but also occurs
when more complex stimulus relations are involved.
Relational frame theory and maladaptive human

emotional responding
According to RFT, equivalence is the most common and fundamental way stimuli
can be related—but only one of a large number of different types of stimulus
relations. Just as people can respond to stimuli as being equivalent, they may
also relate them as opposite, different, greater, or less than one another. Critically,
when increasingly complex relations between and among stimuli are involved,
psychological or emotional functions are not simply transferred but rather trans-
formed according to the way in which stimuli are related. For example, if A is
opposite to B and A is then paired with shock, the fear arousing functions of
A will not necessarily transfer to B. Rather the emotional functions of B may
come to be transformed in-line with the relation—in this instance by acquiring
reinforcing functions. As such, humans can find stimuli directly paired with
unpleasant events as pleasurable or reinforcing when the relation established is
one of opposition (Whelan and Barnes-Holmes, 2004) or more/less than (Whelan,
Barnes-Holmes and Dymond, 2006)2.
To date, the transformation of functions has been replicated for both fear and
sexual responses. To demonstrate the former, Dougher et al. (2007) provided half
of their participants with training to establish three nonsense symbols (A, B, C)
as meaning “smallest”, “middle”, and “largest” (the other half did not receive
such training). Thereafter, the B stimulus was paired with an electric shock for
all participants. Results showed that for those who received relational training,
the fear established for the B stimulus was indirectly acquired by both A and C—
but in a non-equivalent way. Specifically, the A stimulus elicited a smaller fear
response than B whereas C (despite having never been associated with shock)
elicited a larger fear response than either A or B. Participants who did not receive
any relational training showed higher skin conductance changes to B relative to
A and C. These findings suggest that when stimuli are related in increasingly
complex ways humans can come to fear relatively harmless stimuli more than
those that were directly associated with aversive events. As noted above, human
sexual arousal functions have also been shown to transform based on complex
stimulus relations involving “same” and “opposite” (Roche and Barnes, 1997).
The derived transformation of functions also has a number of theoretical
and clinical implications relevant to the study of maladaptive emotions. On the
one hand, many associative learning accounts frame emotional disorders as a
product of direct aversive conditioning with treatments based on the assumption
that extinction (i.e., repeatedly non-reinforced exposure to the feared object or
event) will successfully alleviate these problematic behaviors (for a discussion
see Mineka and Zinbarg, 2006). Yet many people suffering from emotional
problems do not necessarily have a history of direct conditioning with respect to
the stimuli that they are anxious about. Derived stimulus relating may offer one
explanation for how maladaptive emotional responses may emerge for stimuli
that have never been directly associated with emotional events in the past. On
the other hand, “the clinical significance of a stimulus is not always its particular
emotional function (e.g., fear) but rather the extent to which it engenders
2 Functional researchers typically generate complex relations between stimuli using contextual cues
that are themselves established using the Matching-To-Sample procedure mentioned previously.
Specifically, a nonsense word or symbol (i.e., contextual cue), a sample stimulus and three or more
comparison stimuli are presented on a computer screen. If the to-be-trained “OPPOSITE” cue is
presented, choosing the comparison stimulus (e.g., large square) that is furthest removed from the
sample (e.g., small square) along some physical dimension is reinforced. On alternate trials the
to-be-trained “SAME” cue is presented and choosing the comparison which is physically identical
to the sample is reinforced. Participants are trained in this way across a variety of situations (e.g.,
big and small circles, thick and thin lines, few and many dots) until they respond appropriately to
novel samples and comparisons in the presence of the SAME and OPPOSITE cues in the absence
of reinforcement.
avoidance behaviour” (Auguston and Dougher, 1997, p.183). For instance, people
often do not seek treatment for phobias because they are afraid of a particular
object/event but rather due to the deleterious effects that avoiding such stimuli/
situations has on their everyday life (see Hayes, Strosahl, and Wilson, 1999).
According to RFT, if emotional functions can be indirectly acquired by stimuli
then people may also attempt to avoid those stimuli even though they have never
previously been associated with aversive events. Consistent with this assumption,
avoidance responding—like emotional functions—has been found to transfer
through stimulus relations. Recall that in the Auguston and Dougher (1997) study
two equivalence relations were established (A1-B1-C1-D1 and A2-B2-C2-D2).
Thereafter, B1 was paired with an electrical shock while B2 was presented in
the absence of a shock. In a second study the same procedure was employed,
but participants were also taught they could avoid being shocked by repeatedly
pressing a button in the presence of B. Following training, participants pressed
the key in the presence of not only B but also the A and C stimuli from the first
relation, but did not press the button for any of stimuli from the second relation.
Dymond et al. (2008) demonstrated similar results for the derived transformation
of avoidance through “same-opposite” relations using picture stimuli.
Finally, derived stimulus relating not only allows for the transformation of
emotional functions from stimulus to stimulus but also for their extinction.
A number of studies have now shown that when the emotional functions of
one stimulus are extinguished, the emotional properties of related stimuli are
typically extinguished as well (e.g., Dougher et al., 1994; Roche and Barnes,
1997). Interestingly however, in a recent study, Luciano et al. (2011) failed
to report evidence for the direct or derived extinction of fear (measured using
skin conductance) or avoidance functions in two separate experiments. In
a third experiment, however, while no extinction or reduction of fear was
obtained, avoidance of the feared stimuli was eliminated following an analogue
protocol based on Acceptance and Commitment Therapy. Given that fear,
avoidance, and behavioral extinction play an important role in the aetiology and
maintenance of psychopathology, it is to this functionally based psychotherapy
that we now turn.
Acceptance and commitment therapy

ACT is founded on a core theoretical premise of RFT (that language and
cognition are inherently relational in nature) and strongly connects to the basic
research program outlined above on derived stimulus relating (for a review see
Gaudiano, 2011). Whereas traditional cognitive-based therapies attribute psycho-
pathology to maladaptive or biased patterns of information processing which
require modification for healthy psychological functioning, ACT proponents
argue that psychological problems often result from “psychological inflex-
ibility”. This concept broadly refers to the excessive regulation of behavior
by derived relations, rather than direct reinforcement contingencies acces-
sible to the person in their wider environment (termed “cognitive fusion”) and
the attempted avoidance of private behaviors such as thoughts, feelings, and
memories, even when doing so results in psychological duress (termed “experi-
ential avoidance”). Cognitive fusion and experiential avoidance are suggested
to give rise to and maintain a restricted behavioral repertoire, which in turn
generates in a wide range of psychopathological behaviors. Instead of altering
information processing biases using techniques such as rational deliberation and
behavioral experiments, ACT aims to “change the psychological functions of
private events and the individual’s relationship to them through strategies such as
mindfulness, acceptance, or cognitive defusion” (Hayes et al., 2006, p.6) so that
self-defined values can be achieved, even when symptomatic behavior remains.
In other words, and contrary to most traditional psychotherapeutic approaches,
symptom reduction—while certainly welcomed—is not the primary goal of
ACT. Rather, it is enabling the person to behave in a value or goal consistent
manner. To illustrate this, consider the Luciano et al. study mentioned previously.
Although participants still responded fearfully towards the directly trained as
well as derived stimuli, they did not attempt to avoid those stimuli after receiving
an ACT intervention that targeted avoidance but not fear. Critically, a rapidly
growing literature of correlational, mediation and experimental psychopathology
studies, as well as randomized control trials, support ACT’s utility across a range
of domains. Moreover, a number of studies have directly implicated the psycho-
logical processes hypothesized by ACT as central to positive treatment outcomes
(see Gaudiano, 2011).
Conclusion
A wealth of research indicates that stimuli can come to acquire, maintain, and
change their emotional properties through respondent learning. In this instance,
the psychological functions of stimuli are established or modified through
direct contingencies and/or on the basis of physical similarity between stimuli
(generalization). At the same time, a rich and active empirical program has
recently emerged from the functional tradition that indicates emotional properties
can also be established, transformed and extinguished through the largely
human ability to spontaneously derive relations between and among stimuli
in a bi-directional fashion. This progress at the theoretical and basic research
levels has provided the foundation for an empirically-supported treatment
protocol in the form of ACT. According to this approach, derived stimulus
relating and the transformation of function through those relations can often
undermine the person’s ability to respond appropriately to changing environ-
mental contingencies and result in a wide range of psychological problems.
Consequentially, ACT proponents attempt to promote psychological flexibility
through the use of mindfulness, acceptance, commitment, and behavior change
processes.
References
Auguston, E. M. and Dougher, M. J. (1997). The transfer of avoidance evoking functions
through stimulus equivalence classes. Journal of Behaviour Therapy and Experimental
Barnes-Holmes, Y., Barnes-Holmes, D., Smeets, P., and Luiano, C. (2004). A derived
transfer of mood functions through equivalence relations. The Psychological Record,
54, (1), 95–113.
De Houwer, J. (2011). Why the cognitive approach to psychology would profit from a
functional approach and vice versa. Perspectives on Psychological Science, 6, 202–209.
Dougher, M., Hamilton, D., Fink, B., and Harrington, J. (2007). Transformation of the
discriminative and eliciting functions of generalized relational stimuli. Journal of the
Experimental Analysis of Behavior, 88, (2), 179–197.
Dugdale, N., and Lowe, F. (2000). Testing for symmetry in the conditional discriminations
of language – trained chimpanzees. Journal of the Experimental Analysis of Behavior,
73, 5–22.
Dymond, S., Roche, B., Forsyth, J. P., Whelen, R., and Rhoden, J. (2008). Derived
avoidance learning: Transformation of avoidance response functions in accordance with
the relational frames of same and opposite. The Psychological Record, 58, 271–288.
Gaudiano, B. A. (2011). A review of acceptance and commitment therapy (ACT) and
recommendations for continued scientific advancement. The Scientific Review of Mental
Health Practice, 8, 5–22.
Hayes, S. C., Barnes-Holmes, D., and Roche, B. (eds). (2001). Relational Frame Theory:
A Post -Skinnerian account of human language and cognition. New York: Plenum Press.
Hayes, S. C., Luoma, J. B., Bond, F. W., Masuda, A., and Lillis, J. (2006). Acceptance
and commitment therapy: Model, processes and outcomes. Behaviour Research and
Therapy, 44, 1–25.
Luciano, C., Valdivia-Salas, S., Ruiz, F., Valverde, M., Barnes-Holmes, D., et al., (2011).
Altering direct and derived avoidance responding. The role of extinguishing directly
established aversive eliciting function and the role of verbally motivating protocols.
Manuscript submitted for publication.
Mineka, S., and Zinbarg, R. (2006). A contemporary learning theory perspective on the
etiology of anxiety disorders: It’s not what you thought it was. American Psychologist,
61, (1), 10–26.
Roche, B., and Barnes, D. (1997). A transformation of respondently conditioned stimulus
functions in accordance with arbitrarily applicable relations. Journal of the Experimental
Analysis of Behavior, 67, 275–301.
Sidman, M. (1994). Equivalence relations and behavior: A research story. Boston:
Authors Cooperative.
Valverde, M. R., Luciano, C., and Barnes-Holmes, D. (2009). Transfer of aversive
respondent elicitation in accordance with equivalence relations. Journal of the
Experimental Analysis of Behavior, 92, (1), 85–111.
Whelan, R., and Barnes-Holmes, D. (2004). The transformation of consequential functions
in accordance with the relational frames of same and opposite. Journal of the
Experimental Analysis of Behavior, 82, 177–195.
Whelan, R., Barnes-Holmes, D., and Dymond, S. (2006). The transformation of conse-
quential functions in accordance with the relational frames of more-than and less-than.
Journal of the Experimental Analysis of Behavior, 86, (3), 317–335.
29 Self-regulation as a mediator of
change in psychotherapy
Timothy J. Strauman and Megan M. Klenk
Duke University
Kari M. Eddington
University of North Carolina at Greensboro
Affective science has enormous relevance for psychotherapy specifically, and for
behavior change more generally (Strauman, Eddington, and McCrudden, 2008).
First, affective science provides a sound basis for understanding how psycho-
therapy works. Specifically, affective science is not only compatible with effective
psychotherapeutic practice, but many of the specific and nonspecific factors
implicated in treatment efficacy have their basis in emotional processes. Second,
affective science can help to improve the outcomes from current psychotherapies.
Research on psychotherapy outcome and process has struggled to identify the
mechanisms or active ingredients of therapy that mediate change, with the unfor-
tunate result that in many disorders it is not yet possible to identify a priori which
treatment will be most effective for a particular individual. Affective science has
much to offer in this regard, specifically for developing a more microanalytic
understanding of how psychosocial interventions work. Third, affective science
can provide theory-based targets for change in therapy and innovative methods for
assessing changes. To the extent that theories in affective science are relevant to
understanding how psychotherapy works, they also should be useful in developing
and testing hypotheses regarding which psychological interventions we should
be trying to accomplish. Our own research, examining the role of self-regulatory
cognitive and motivational processes in vulnerability to and treatment of mood
disorders, represents one example of a theory-based translational model (see
below). And fourth, affective science provides conceptually sound and empirically
supported bases for treatment matching, which we define as a systematic a priori
process by which a treatment is selected as most likely to provide maximum benefit
for a particular individual (Beutler, 1991). We view affective science broadly, as
an invaluable tool for understanding not only emotions themselves, but also what
gives rise to emotions, the dynamic interplay between emotion and cognition,
how emotional responses can become maladaptive, and how to help the individual
return to adaptive functioning.
Theories of emotionally relevant psychological processes, such as self-
regulation, can provide rich, well-validated conceptual frameworks for
exploring how psychological interventions can change emotional experience
210 Timothy J. Strauman, Megan M. Klenk and Kari M. Eddington
and physiology. Our research has used regulatory focus theory (Higgins, 1998)
to examine the role of beliefs about the self in vulnerability to mood and anxiety
disorders. Actual-self discrepancies from people’s hopes and aspirations (ideals)
cause dysphoric affect that may contribute to depression, whereas actual-self
discrepancies from people’s beliefs about their duties and obligations (oughts)
cause anxious affect that may contribute to anxiety disorders. An actual-ideal
self-discrepancy is a promotion failure that can lead to diminished ability to
pursue and attain one’s ideals and chronic dysphoric affect, increasing risk for
major depressive disorder (MDD). In contrast, an actual-ought self-discrepancy is
a prevention failure that can lead to exaggerated emphasis on responsibilities and
obligations and hypervigilance, increasing risk for anxiety disorders. However,
not everyone with high levels of promotion goal pursuit failure becomes clini-
cally depressed, indicating that other factors are critical in the link between
self-regulation and depression. Recent findings in cognitive neuroscience and
in imaging genetics offer complementary perspectives on how chronic failure
in personal goal pursuit could, for a particular subset of individuals, lead to
depression. In this chapter, we briefly consider a new risk phenotype model for
depression and its implications for changing emotions and restoring adaptive
functioning.
Regulatory focus and depression

Individual differences in regulatory focus develop primarily from differential
exposure to positive-outcome-focused vs. negative-outcome-focused parenting,
with temperament manifesting an indirect influence by sensitizing or buffering
children to the reward and punishment contingencies from their interactions
with parents. Affective responses to ideal or ought goal pursuit feedback are
moderated by individual differences in strength of orientation to promotion and
prevention goals. People’s fundamental needs for promotion and prevention have
significant implications relating to the behaviors, evaluations, and experiences
that characterize our social identities, the nature of our social perceptions and
appraisals, and our reactions to social feedback.
Although regulatory focus was originally conceptualized at the psychological
level of analysis, EEG and FMRI studies have begun to identify its neuro-
biological correlates. These findings provide an important conceptual bridge to
understanding how individual differences in regulatory focus, as a cognitive/
motivational orientation, might be implicated in vulnerability to psychopathology.
Eddington, Dolcos, Cabeza, Krishnan, and Strauman (2007) found that priming
of promotion goals was associated with activation in left orbitofrontal cortex
(LOFC), whereas priming of prevention goals was associated with activation
in right OFC. In addition, individual differences in promotion focus correlated
significantly with magnitude of LOFC response to promotion goal priming. Thus,
individuals whose socialization history shaped a dominant orientation toward
promotion goals—that is, a characteristic world view in which social interactions
are typically viewed as opportunities to “make good things happen”—manifested
Self-regulation as a mediator of change in psychotherapy 211
a characteristic neural activation pattern when incidentally exposed to social
stimuli relevant to their own promotion goals.
The Eddington et al. study was the first to use FMRI to link idiographically-
assessed personal goal priming with changes in OFC activation—a region
implicated in representing the hedonic value of primary as well as abstract
(secondary) reinforcers in decision making, and in performance monitoring
(Ramnani and Owen, 2004). Furthermore, the left OFC activation following
promotion goal priming was detected while participants were performing a task
unrelated to goal pursuit, suggesting that promotion and prevention goals, as
highly accessible knowledge structures, can be activated implicitly and still have
a pervasive influence on the interpretation of social stimuli.
Chronic perceived failure to attain promotion versus prevention goals is
associated with symptoms of depression vs. anxiety respectively. Strauman
(2002) proposed a model in which individuals whose socialization histories
led them to acquire a strong orientation toward promotion goals were more
vulnerable to depression when exposed to chronic failure feedback. Whereas
under normal circumstances these individuals self-regulate effectively toward
their goals and respond to acute failure feedback adaptively, they would become
vulnerable to depression if they experience chronic or catastrophic goal pursuit
failure. If promotion goal priming is associated with left OFC activation, and
individuals experiencing chronic failure to attain promotion goals are vulnerable
to depression, then among such individuals depression should be characterized by
a dysfunction in left OFC activation following promotion goal priming. Given the
evidence for a bidirectional functional link between OFC and limbic structures
regulating affect and motivation, changes in OFC activation driven by perceived
failure in self-regulation could feed back directly to the amygdala and related
limbic structures.
Consistent with this hypothesized pathway, Eddington, Dolcos, McLean,
Cabeza, Krishnan, and Strauman (2009) found that clinically depressed
individuals showed attenuated left OFC activation following promotion goal
priming. Depressed participants who also reported significant levels of anxiety
symptoms likewise manifested a significant increase in right OFC activation
following prevention goal priming. This combination of left OFC hypoactivation
(representing a deficit in promotion goal pursuit motivation) with right OFC
hyperactivation (representing a vigilant and ruminative response to cues for
prevention goals) was consistent with the high degree of comorbidity between
unipolar depression and generalized anxiety disorder.
In addition to the prediction that promotion goal pursuit failure will be
associated with dysphoric affect and (if sufficiently chronic) depressive symptoms,
there are other characteristics of the promotion goal system that help to explain
how, under specific conditions, certain individuals might become vulnerable to
depression when they experience chronic promotion goal failure. Individuals
with a characteristic promotion orientation will be more likely to construe goals
in promotion (gain/non-gain) terms and will be more committed to goals that are
framed in promotion terms. For these individuals, there will be more opportunity
for promotion success (“gain”), but also more opportunities for promotion failure
(“non-gain”). Also, failure to attain a specific promotion goal increases the
importance of promotion goals in general as well as the individual’s motivation
to pursue them, because promotion goals are more interconnected than prevention
goals. And in addition, even a small actual-ideal discrepancy (promotion failure)
is highly motivationally significant, due to the greater degree of interconnect-
edness among promotion goals. Thus, for individuals with a dominant promotion
orientation, promotion failure feedback takes precedence in social information
processing and drives affective reactions to subsequent goal pursuit. Finally,
promotion goals are also more substitutable than prevention goals, but only if
the individual can effectively disengage from pursuing one promotion goal and
engage in a substitute goal. Otherwise, individuals will become trapped by unsuc-
cessful pursuit of a goal they are failing to attain and experience more intense and
chronic dysphoric affect.
Although the data are consistent with this hypothesized pathway to depression,
social-cognitive mechanisms alone cannot provide a sufficient account either for
normal variability in affective and motivational responses to goal pursuit failure
or for vulnerability to depression associated with maladaptive self-regulation.
Nonetheless, cross-disciplinary research linking genetic, neurobiological, and
behavioral levels of analysis is beginning to identify how variability in brain
function contributes to individual differences in complex behavioral traits and
how such diatheses interact with environmental factors to precipitate psychi-
atric disorders (e.g., Hariri, 2009). For example, research indicates that genetic
and neurobiological factors interact with socialization patterns to influence the
development of individual differences in self-regulation of goal pursuit. In turn,
the cognitive-motivational systems postulated by RFT are largely responsible for
affective responses to socially-embedded goal pursuit feedback, and therefore
represent potential sources of vulnerability to mood disorders via their inter-
actions with neural mechanisms that underlie incentive motivation, approach
behavior, and affective responses to goal-relevant feedback.
We hypothesize that a combination of three contributory factors—individual
differences in regulatory focus, genetically determined individual differences in
goal-specific information processing and the intensity and duration of affective
responses to goal-relevant feedback, and the experience of chronic failure to attain
a particular kind of personal goal—creates a self-regulation pathway to depression.
A critical feature of the proposed phenotype is the conditional nature of its relation
to depressive vulnerability. Because individual differences in regulatory focus
involve both costs and benefits, vulnerability to depression will depend upon the
individual’s social context. Under circumstances of chronic perceived promotion
goal pursuit failure, such individuals would be more vulnerable to intense,
prolonged dysphoric affect, would perseverate on failure feedback, would have
greater difficulty objectively evaluating their progress and, if needed, disengaging
from promotion goal pursuit, and could ultimately “spiral down” into a state of
diminished appetitive motivation, anhedonia, and hopelessness.
Self-regulation as a mediator of change in psychotherapy
A psychotherapy focused on self-regulation has the potential to be applied
broadly to a range of individuals experiencing depression, but also (via studies
of potential mechanisms of action) to identify self-regulation-based risk pheno-
types toward which preventive or therapeutic interventions could be targeted.
Self-system therapy (SST; Vieth, Strauman, Kolden, Woods, and Klein, 2003)
is a brief, structured therapy for depression designed for individuals for whom
problematic self-regulation is a prominent clinical feature. The primary objec-
tives of SST include education about depression, re-initiation of goal-directed
behavior that is relevant to the individual’s promotion (ideal) goals in particular,
systematic self-evaluation, identification of targets for change, and instantiating
change and/or compensatory strategies to reduce distress and restore adaptive
self-regulation.
How might a treatment such as SST work? Efficacious psychotherapies share
a number of active components, especially the so-called “universal” aspects such
as the working relationship between patient and therapist. In addition to such
components, therapeutic change in SST is hypothesized to occur via several
specific mechanisms drawn from RFT and from basic research in social cognition
and affective science:
●● Changing the availability and accessibility of goals. SST can promote change
by helping the patient modify the set of goals used in the process of self-
regulation. For instance, SST may help the patient acquire goals that are more
adaptive. Having more appropriate goals should lead to increased success
in goal pursuit. “Accessibility” refers to the likelihood that a particular goal
representation will be used in self-regulation. The greater the accessibility of
a goal, the greater influence it will have on self-evaluation. SST is designed
to increase the accessibility of adaptive goals and decrease the accessibility of
maladaptive ones.
●● Changing the importance and affective significance of goals. SST also seeks
to modify the emotional significance of goals and the consequences of goal
failure. The therapist may encourage a patient to question the “fit” of a goal for
current circumstances, help the patient recognize situations where particular
goals are more or less relevant, or explore the consequences of pursuing a
particular goal.
●● Changing patterns of goal-directed behavior. By teaching interpersonal skills,
helping patients to deal more effectively with challenging situations, and
increasing opportunities for success in attaining promotion goals, SST can
help to change how individuals engage with the social world more effectively
to become the kind of person they would like to be.
The primary therapeutic techniques of SST represent methods for exploring the
patient’s goals and her/his ways of pursuing them. Each is related to techniques
used in other efficacious psychotherapies. Self-in-Context Assessment (SCA),
adapted from the ‘interpersonal inventory’ technique of interpersonal therapy,
occurs early in treatment. SCA applies the developmental postulates of RFT,
which hypothesize that dominant regulatory orientations and characteristic self-
beliefs develop from early patterns of parent/child contingencies. The purpose
of SCA is to generate an initial “data base” from which the therapist and patient
can develop hypotheses regarding the patient’s problems in self-regulation. The
therapist and patient assess the relationships in which the patient learned that
being a particular kind of person was good or bad through the experience of
positive or negative emotions for behaving (or not behaving) in particular ways.
Psychological Situation Analysis, which occurs during the middle of treatment,
involves examining current or past interpersonal encounters to illuminate the
patient’s experiences of the interactions, the goal(s) that were operative, the strat-
egies the patient used to pursue them, and the outcomes and affective states that
resulted. The therapist and patient work to identify the patient’s modal psycho-
logical situations and her/his characteristic self-regulatory style. Self-Belief
Analysis (SBA) also takes place during the middle of treatment. The purpose of
SBA is to identify and examine the origins, content, and functions of the patient’s
beliefs about her/himself in relation to others, and to determine how these beliefs
may contribute to the patient’s symptoms. SBA parallels the analysis of automatic
thoughts and core beliefs in cognitive therapy (CT); however, whereas CT targets
the negative cognitive triad and underlying depressogenic schemas, SST focuses
on the role of goals in maladaptive self-evaluation.
Strauman at al. (2006) conducted a randomized trial comparing SST and CT
in patients meeting DSM-IV criteria for major depressive disorder or dysthymic
disorder. Two hypotheses were tested: that SST would be more effective for
individuals whose depressive symptoms were associated with self-regulation
failure, and that improvement within SST would be correlated with change in
self-regulation-derived affect. The overall efficacy of SST was equivalent to
that of CT. More importantly, an intent-to-treat analysis supported the prediction
that SST would be more efficacious than CT for patients with self-regulatory
dysfunction. Patients with poor self-regulation who received SST and patients
without substantial problems in self-regulation who received CT showed signifi-
cantly greater clinical improvement than patients with self-regulation problems
assigned to CT or patients without self-regulation problems assigned to SST.
We interpret these results as evidence for theory-based treatment matching:
patients with prominent self-regulation failure who received SST fared better
than similar patients randomized to CT. Furthermore, using a childhood memory
retrieval paradigm in which personal goals were used as incidental recall cues
for childhood experiences, we observed that SST was associated with signifi-
cantly greater reduction in goal-related negative affect (and greater increase in
goal-related positive affect) than was CT—important evidence regarding SST’s
mechanisms of action.
There is much more work to be done exploring how intervening with self-
regulatory dysfunction could reduce distress and improve well-being. In addition
to randomized trials in clinical populations, an alternative approach—more
translational in nature—is to design and test “microinterventions” that target
specific mechanisms of vulnerability. Such tests not only set the stage for larger-
scale treatment research, but also challenge the underlying theoretical model
itself. For example, RFT suggests a number of novel strategies for behavioral
intervention with the self-regulation risk phenotype. One such strategy is based
on the notion of engagement strength—the intensity with which an individual’s
regulatory system is activated in the context of goal pursuit. If depression is
maintained in part by an inability to discontinue pursuing particular promotion
goals for which there is currently no chance for success, then teaching the patient
to reduce (rather than increase) promotion engagement strength in response to
failure feedback would have the paradoxical effect of reducing dysphoric affect.
Another novel strategy can be derived from the concept of regulatory fit, the
match between the type of goal being pursued and the means used to pursue it. As
above, if depression is maintained in part by inability to discontinue promotion
goal pursuit, then helping the patient learn to intentionally disrupt promotion fit
in response to failure feedback (for example, by pursuing the troublesome goal
using a prevention-based strategy instead of a promotion-based strategy) should
also lead to a paradoxical reduction in dysphoric affect.
We believe there is significant potential for extending a self-regulation based
approach to behavior and affect change beyond our initial emphasis on mood
disorders. For example, depressive/anxious comorbidity can be conceptualized in
terms of regulatory focus, and different RFT-based interventions could be used to
minimize dysphoric vs. anxious symptoms (Klenk, Strauman, and Higgins, 2011).
Likewise, by identifying individuals whose personal history and self-regulatory
tendencies resemble the hypothesized risk phenotype, preventive strategies could
be implemented in order to reduce the likelihood of an initial depressive episode.
And of course, these interventions also constitute tests of the underlying theory
itself, which in turn facilitates the ongoing translational exchange between basic
and clinical science. We look forward to further developments in the application
of self-regulation theory to psychological interventions.
References
Beutler, L. E. (1991). Have all won and must all have prizes? Revisiting Luborsky et al.’s
verdict. Journal of Consulting and Clinical Psychology, 59, 226–232.
Eddington, K. M., Dolcos, F., Cabeza, R., Krishnan, K. R. R., and Strauman, T. J. (2007).
Neural correlates of promotion and prevention goal activation: An fMRI study using an
idiographic approach. Journal of Cognitive Neuroscience, 19, 1152–1162.
Eddington, K. M., Dolcos, F., McLean, A. N., Cabeza, R., Krishnan, K. R. R., and
Strauman, T .J. (2009). Neural correlates of idiographic goal priming in depression:
Goal-specific dysfunctions in the orbitofrontal cortex. Social Cognitive and Affective
Hariri, A. R. (2009). The neurobiology of individual differences in complex behavioral
traits. Annual Review of Neuroscience, 32, 225–247.
Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational
principle. Advances in Experimental Psychology, 46, 1–46.
Klenk, M. M., Strauman, T. J., and Higgins, E. T. (2011). Regulatory focus and anxiety:
A self-regulatory model of GAD-depression comorbidity. Personality and Individual
Differences, 50, 935–943.
Ramnani, N., and Owen, A. M. (2004). Anterior prefrontal cortex: Insights into functioning
from anatomy and neuroimaging. Nature Reviews Neuroscience, 5, 184–189.
Strauman, T. J. (2002). Self-regulation and depression. Self and Identity, 1, 151–157.
Strauman, T. J., Eddington, K. M., and McCrudden, M. C. (2008). Affective science
and psychotherapy: In search of synergy. In J. Rottenberg and S. L. Johnson (eds),
Emotion and psychopathology: Bridging affective and clinical science, (pp. 285–303).
Washington: American Psychological Association Press.
Strauman, T. J., Vieth, A. Z., Merrill, K. A., Kolden, G. G., Woods, T. E., Klein, M. H.,
Schneider, K. L., and Kwapil, L. (2006). Self-system therapy as an intervention for self-
regulatory dysfunction in depression: A randomized comparison with cognitive therapy.
Journal of Consulting and Clinical Psychology, 74, 367–376.
Vieth, A., Strauman, T. J., Kolden, G., Woods, T., and Klein, M. H. (2003). Self-System
Therapy: A theory-based psychotherapy for depression. Clinical Psychology: Science
and Practice, 10, 245–268.
30 Mindfulness-based interventions
The dialectic of changing emotions by
accepting them
Pierre Philippot and Alexandre Heeren
University of Louvain at
Louvain-la-Neuve
The notion of emotional change is undergoing a dramatic shift in clinical

psychology. Until recently, the focus was on either controlling emotional responses
(e.g., through relaxation or breathing retraining) or altering emotional appraisal
(e.g., through cognitive restructuring). Presently, a consensus is growing that it is
not so much the emotion itself that needs to be changed, but rather how people
react to their own emotion. The underlying notion is that clinical distress would
not result from primary emotional responses (e.g., fear of a perceived threat), but
rather from secondary reaction, triggered by the primary emotional responses
(e.g., fear of fear). The issue of emotional change in clinical psychology has thus
redirected its aim from controlling/altering emotion to modifying one’s secondary
reaction to emotion.
Emotional acceptance as the heart of emotional change

In many clinical conditions, the most common maladaptive secondary
reaction is emotional avoidance or suppression. This notion is at the heart
of Barlow’s model of the persistence of psychological distress (e.g., Moses
and Barlow, 2006). In that perspective, when anticipating a painful emotional
situation, individuals attempt to escape from emotional experience by a wide
diversity of means, from blatant behavioral flight to subtle forms of cognitive
avoidance, such as rumination or abstract and overgeneral thinking. However,
such attempts of suppression or avoidance are, most of the time, doomed to
failure. Indeed, if one can successfully avoid external events or situations, one
cannot escape from internal events, such as emotion. Emotion is happening
in the heart of the very stuff that constitutes us (goals, values, and our body);
therefore, they cannot be avoided. The failure of emotional suppression results
in even more distress, as the emotion reappears or continues to manifests
itself in uncontrolled way, such as in intrusions or emotion outbursts. This
eventually generates a vicious circle: anticipation of a distressing situation
triggers emotional suppression; failure of suppression generates distress, which
in turns calls for more attempts of suppression. For instance, after experi-
encing a trauma, an individual might anticipate that any confrontation with
218 Pierre Philippot and Alexandre Heeren
the traumatic situation or anything that can evoke it would generate distress
that is anticipated as so intense that the individual cannot tolerate experiencing
it. Hence, he/she will try by any means to escape from any trauma reminder
(avoiding certain places, avoiding watching television—scenes might be
reminders of the trauma, avoiding reading newspapers or listening to the news
on the radio, etc.). However, a large clinical literature has shown that these
suppression attempts fail and that trauma reminders quickly reappear under
the form of intrusions or nightmares. The latter are experienced with much
distress, which makes the anticipation of the confrontation to trauma reminders
even less tolerable, and increases attempts of suppression. A vicious circle thus
develops, spiraling to disordered emotion.
In this perspective, the sensible emotion regulation strategy is emotional
acceptance: withholding all attempts to escape or to suppress the aversive
emotion and focusing all resources on allowing the emotional processes to
unfold; on observing, exploring, and experiencing these processes through the
basic givens of moment to moment experience, i.e., body sensations, sensory
inputs, action tendencies, and thoughts and mental images. In other words,
following a dialectic move, this new perspective poses that emotional change
(synthesis) is best achieved by accepting emotion (antithesis) rather than
attempting to control or suppress it (thesis). Concretely, clients are encouraged
to expose themselves to the aversive emotional experiences in which their
continuing distress is rooted. During exposure, they have to develop as best as
they can the active acceptance attitude described above, which implies habitu-
ation and desensitization processes. To refer to the clinical example proposed
previously, trauma survivors are encouraged to confront trauma reminders and
memories, while abstaining to control or to lessen the emotion elicited, but
rather by focusing on what they feel and experience during the re-evocation:
body sensations, sensory inputs, action tendencies, and thoughts and mental
images.
Mindfulness-based interventions as means to promote acceptance

It is in that context that Mindfulness-based interventions (MBIs) have known
a dramatic development, as a means to develop the very mental attitude
of acceptance toward experience, attitude that promotes emotional change
as described above. Mindfulness has been defined as a state of awareness
resulting from “paying attention in a particular way: on purpose, in the present
moment, and nonjudgmentally” (Kabat-Zinn, 1994: 4) (note that the definition
of mindfulness is still the object of debate). It is thus the results of two psycho-
logical processes: voluntary allocation of attention to ongoing experience
and disengagement from automatically aroused judgments, preferences, action
tendencies. It also promotes decentring from thoughts and feelings—i.e., viewing
as mental events and increases self-compassion. It is a skill that can be developed
through repeated exercises. Originally, Kabat-Zinn (1982) has proposed a
structured program, articulated in eight weekly sessions, specifically aimed at
Mindfulness-based interventions 219
developing mindfulness capacities: the Mindfulness-Based Stress Reduction
program (MBSR). This program constitutes the initial basis on which many
other clinical interventions have been developed (for a review, see Baer, 2003). It
mostly consists of the practice of meditative-like exercises in which participants
are trained to focus their attention on present moment experience (body sensa-
tions, sensory inputs, and thoughts and mental images), to notice each time their
attention has been taken away from present moment experience, and to redirect
their attention to it.
There is evidence that short-term programs like MBSR have a significant,
although moderate, clinical impact in terms of symptoms reduction (Baer, 2003;
Grossman, Niemann, Schmidt, and Walach, 2004). These benefits include a
reduction of psychopathological symptoms as well as a decrease of somatic
complaints and a better health status in a wide diversity of conditions (e.g.,
psoriasis, tinnitus). However, the processes underlying these effects still need
to be specified. We have proposed that MBIs impact on different psycho-
logical processes (Philippot and Segal, 2010). At a molar level, MBIs train the
capacity to observe one’s flow of emotional experience moment after moment,
withholding automatic reaction to these experiences. This training also entails the
capacity to be reflexively aware of emotion by noticing bodily changes, thoughts,
and mental images automatically popping into awareness. By doing so, MBIs
diminish emotional avoidance and increase the tolerance to aversive states. At
the more molecular level, MBIs train several important skills and processes, such
as the capacities (a) to withdraw from or inhibit automatic prepotent responses of
avoidance from, or suppression of, emotion, (b) to tolerate emotional discomfort,
(c) to maintain attention on ongoing experience. In more technical terms, MBIs
should improve executive resources, especially those related to attention mainte-
nance, to disengagement of automatic attention capture, and to flexibility by
inhibiting prepotent automatic responses to reallocate resources to strategic
responses.
Although intuitively appealing, the rationale developed above needs to be
empirically tested. Further, if MBIs indeed improve these different cognitive
abilities, it should be specified whether improvement in a given cognitive
function specifically effect upon a given outcome or whether the effect is unspe-
cific. In the following sections, we will review evidence addressing two main
research questions: do MBIs increase executive capacities; and if so, which ones?
Does executive improvement mediate the effect of MBIs on symptom reduction?
Review of the cognitive mediators of mindfulness-based interventions

One of the first indications that mindfulness training might benefit executive
capacities was reported by Jha, Stanley, Kiyonaga, Wong and Gelfrand (2010).
They investigated two military cohorts facing a high-stress predeployment
interval, one of the cohorts was allocated to eight-week MBI. In this experi-
mental cohort, Jha et al. (2010) observed that mindfulness practice was related
to increased working memory capacity (estimated with the operation span task).
Further, the increase in working memory capacity mediated the beneficial effect
of MBI on the alleviation of negative affect. However, no differences were
observed between the control and MBI groups, limiting the interpretation that can
be drawn from these results.
Another study (Moore and Malinowski, 2009) compared a group of experienced
mindfulness meditators with a meditation naïve group on measures of cognitive
flexibility (Stroop interference) and attentional performance (d2 concentration
test). Overall, they observed that meditation practice was related to better perfor-
mance in both attention and cognitive flexibility. Still, the correlational nature of
these data leaves opened many possibilities for their interpretation, including a
self-selection bias.
We attempted to overcome these methodological limitations by using quasi-
experimental designs applied to MBI sessions for people suffering from various
psychological difficulties related to mood regulation, anxiety, pain, or chronic
illness. Our first study on cognitive mediators of MBI investigated whether the
clinical benefits of MBI are mediated by a reduction in maladaptive rumination,
and by an increase in adaptive rumination (Heeren and Philippot, 2011). Indeed,
maladaptive rumination is related to executive depletion or deficits (Philippot
and Brutoux, 2007), while constructive repetitive thinking requires executive
resources (Watkins, 2008). Rumination might thus index executive capacity
depletion: Individuals lacking of executive resources would be more prone to
rumination as they do not have the proper resources to disengage from this
rather automatic process. Compared to waiting-list controls, participants who
benefited of the eight-session MBI manifested a reduction of psychopatho-
logical symptoms. Of more interest for the present question, it was also found
that MBI reduced maladaptive rumination and increased adaptive rumination.
Further analyses indicated that MBI clinical effects were mediated by changes in
rumination style. Specifically, we observed that both the reduction in maladaptive
rumination and the increase in adaptive rumination partially but significantly
mediated psychopathological symptoms reduction. This pattern of observation
supports the notion that individuals trained with MBI present a more constructive
repetitive thinking style, and that this change in rumination is partly responsible
for the improvement of their psychopathological condition. Our rationale was
that more functional repetitive thinking should index better executive capacities.
The results of that study are thus suggestive that MBIs indeed increase executive
functioning. However, this evidence is only indirect.
In a next study (Heeren, Van Broeck and Philippot, 2009), we attempted to
replicate the finding that MBI increase autobiographical memory specificity
(as already reported by Williams, Teasdale, Segal, and Soulsby, 2000), and to
extend it by investigating whether this change is mediated by an improvement
of executive capacity. Deficits in autobiographical memory specificity are known
to be a central cognitive marker of mood disorders and a stable predictor of
depression relapse. As such, it constitutes one of the most reliable proxy for mood
vulnerability. Before and after an eight-session MBI, we administered an autobio-
graphical memory task specifically targeting autobiographical specificity (AMT).
Mindfulness-based interventions 221
We also recorded several cognitive parameters, namely capacities for inhibition
or flexibility, either in the cognitive or the motor domains. We expect cognitive
ability to be affected by MBI, but not motor abilities, as the former, but not the
latter, are trained in the MBI. Specifically, we administered a cognitive inhibition
task (Hayling task), a motor inhibition task (GoStop paradigm), a cognitive
flexibility task (verbal fluency task), and a motor flexibility task (trail making
test). Compared to matched controls, MBI participants showed increased autobio-
graphical memory specificity and decreased overgenerality, replicating previous
findings. MBI participants also manifested improved capacities in cognitive
flexibility and in inhibition of cognitive prepotent responses. As expected, all
significant effects concerned cognitive control tasks, while no differences were
observed for motor tasks. Further, mediational analyses indicated that changes
in cognitive flexibility partially mediated the reduction in overgeneral memories
observed after the MBI. Taken as a whole, this pattern of results indicates that
MBIs result in improve executive function, at least as regards inhibition and flexi-
bility applied to verbal material. Further, this executive capacity improvement
impacts upon autobiographical memory specificity, a proxy of mood vulnerability.
Obviously, what is missing are integrative studies combining the manipu-
lation of mindfulness training, and the observation of consecutive effects on
executive functioning and psychopathological variables. In this perspective, we
are presently assessing psychopathological symptoms and rumination style as
well as attentional and executive functioning (using the Attention Network Task)
before and after an eight-session MBI. The same assessments are conducted in
a matched control group. Such data should shed further light on the mediating
impact of cognitive improvements in MBIs.
Conclusions
While MBIs are eliciting a vivid interest in the Western world for at least a
decade, still little is known about the psychological processes that, on the one
hand, are affected by such interventions and, on the other hand, are mediating
the clinical outcome of MBIs. In this chapter, we reviewed the few and very
first studies that investigated the cognitive mediators of MBIs. Although scarce
or indirect, incipient empirical evidence supports the notion that MBIs improve
several aspects of executive and attentional functioning, which in turns impacts
upon psychopathological symptoms. More specifically, our data suggest that
MBIs develop the capacity to disengage attention from automatically activated
thoughts and feelings, such as worries or depressive rumination. Further, MBIs
train the capacity to re-engage attention on ongoing experience. This intensive
cognitive training (45 minutes of daily exercise during eight weeks) seems to
result in at least two beneficial outcomes. First, participants’ executive capacities
are improved, and they are thus likely to choose and implement more adaptive
behaviors. Second, by reducing rumination and other dysfunctional repetitive
modes of thinking, MBIs exonerate resources that were otherwise captured
by these maladaptive processes. Hence, with better executive abilities and
more cognitive resources, the individuals are better able to adopt an active and
constructive mode of acceptance when confronted to painful emotions.
Acknowledgements
Alexandre Heeren is a Post-doctoral research fellow at the Belgian National

Fund for Scientific Research (FNRS)
References
Baer, R. A. (2003). Mindfulness training as a clinical intervention: A conceptual and
empirical review. Clinical Psychology: Science and Practice, 11, 230–241.
Grossman, P., Niemann, L., Schmidt, S., and Walach, H. (2004). Mindfulness-based stress
reduction and health benefits: A meta-analysis. Journal of Psychosomatic Research, 57,
35–43.
Heeren, A., and Philippot, P. (2011). Changes in ruminative thinking mediate the clinical
benefits of mindfulness: Preliminary findings. Mindfulness, 2, 8–13.
Heeren, A., Van Broeck, N., and Philippot, P. (2009). The effects of mindfulness on
executive processes and autobiographical memory specificity. Behaviour Research and
Therapy, 47, 403–409.
Jha, A. P., Stanley, E. A., Kiyonaga A., Wong., L., and Gelfand, L. (2010). Examining the
protective effects of mindfulness training on working memory capacity and affective
experience. Emotion, 10, 54–64.
Kabat-Zinn, J. (1982). An outpatient program in behavioral medicine for chronic pain
patients based on the practice of mindfulness meditation: Theoretical considerations and
preliminary results. General Hospital Psychiatry, 4, 33–47.
—(1994). Wherever you go, there you are: Mindfulness meditation in everyday life. New
York: Hyperion.
Moore, A., and Malinowski, P. (2009). Meditation, mindfulness and cognitive flexibility.
Consciousness & Cognition, 18, 176–186.
Moses, E. B., and Barlow, D. H. (2006). A new unified treatment approach for emotional
disorders based on emotion science. Current Directions in Psychological Science, 15,
146–150.
Philippot, P., and Brutoux, F. (2007). Induced Rumination Dampens Executive Processes in
Dysphoric Young Adults. Journal of Behaviour Therapy and Experimental Psychiatry,
39, 219–227.
Philippot, P. and Segal, Z. (2009). Mindfulness based psychological interventions:
Developing emotional awareness for better being. Journal of Consciousness Studies,
16, 285–306.
Watkins, E. R. (2008). Constructive and unconstructive repetitive thought. Psychological
Bulletin, 134, 163–206.
Williams, J. M. G., Teasdale, J. D., Segal, Z. V. and Soulsby, J. (2000). Mindful
meditation reduces overgeneral autobiographical memory in depressed patient. Journal
of Abnormal Psychology, 109, 150–155.
Postscript
Experimental rigor and clinical complexity
Les Greenberg
The field of research on emotional change has grown tremendously in the last
decades. At the time of our first book on Emotion in Psychotherapy (Greenberg
and Safran, 1987) there was hardly any research on emotional change, especially
not experimental research on changing emotions. In this book, we proposed that
psychological dysfunction was based on disorder of the function of the emotion
system and that it was possible to work with emotion directly in psychotherapy
to promote change. This current volume on experimental research on changing
emotions is a timely major contribution to this topic, and most crucial to under-
standing human functioning in general and to the processes in psychotherapy. The
editors are to be complemented on pulling together so much within one cover, by
having succinct chapters on many topics.
The title of this volume “Changing emotions” carries a double meaning
highly relevant to psychotherapy and clinical work. The title implies both the
processes by which emotion can be changed by different means (such as thera-
peutic intervention) and the dynamic nature of emotion as it changes moment
by moment (as it does in psychotherapeutic dialogue). In our first book on
emotion in psychotherapy, we proposed that people were dynamic self-organizing
systems that changed moment by moment, and we subsequently developed a
more elaborate dialectical constructivist view of functioning that takes emotion
as the fundamental datum of human experience, but recognizes the importance
of meaning making and narrative, and ultimately views emotion and cognition
as inextricably intertwined (Greenberg, 2010; Greenberg and Pascual-Leone,
1995). In this view, emotions are seen as process of immediately informing us
of what is significant to us, and as providing a disposition to act, and healthy
adaptation involves using this rapidly provided information and action tendency
and then deciding by means of more deliberate processing what finally to do.
This dialectical constructivist view of human nature, in which people are seen
as being in a constant process of making sense of their automatic emotional
experience, is supported by many of the studies and viewpoints discussed in
this volume which emphasize the dynamic nature of emotion and the interaction
between felt experience, meaning making, and social influence (see, for example,
Camras and Shuster, Chapter 4; Tsai, Rimé, Frijda, Frederickson, and Ochsner,
Chapter 17).
224 Les Greenberg
This view also is supported by psychotherapy process research which indicates
that certain types of therapeutically facilitated emotional awareness and arousal,
when expressed in supportive relational contexts, in conjunction with some sort
of conscious cognitive processing of the emotional experience, is important for
therapeutic change for certain classes of people and problems (Greenberg and
Pascual-Leone, 2007). In addition, making sense of aroused emotion has been
shown to be a fundamental change process in emotion-focused therapy, and one
that predicts outcomes in the treatment of depression and complex trauma (for a
review, see Greenberg, 2010).
This postscript is an interesting opportunity to reflect on the two literatures,
of experimental psychology and psychotherapy research, to provide a type of
bifocal perspective on changing emotion. An issue of major significance for both
areas appears to be the importance of generating theory and research to help
understand to what extent automatic emotion processes can be changed through
deliberate conscious cognitive processes of self-control, and to what extent they
can only be changed through more implicit processes based on new emotional
and/or relational experiences. Addressing this question also needs always to keep
in mind to what extent experimental studies on changing emotion is studying
the same process as is studied in changing emotion in psychotherapy and in life.
Casting a therapist’s eye on the experimental studies of emotional change in
the different sections on learning, development, dynamics, and intervention, it
appears that studying emotional change in the lab is often a very different process
from studying the process of emotional change in therapy. In therapy, people are
often helped first to arrive at painful, often previously avoided, emotions, and
then to leave these emotions by a variety of therapeutic processes (Greenberg
2002, 2010).
The trade-off between experimental control and ecological validity is particu-
larly relevant to the study of such a basic process as emotion, and although many
of the chapters seem highly relevant to informing us about real world emotional
experience, it would also be important to blend the study of emotion in the lab with
the study of the more intense emotions experienced in real-life situations such as
the emotions in trauma, loss, humiliation, and triumph, and the joy, tranquility,
and awe that are dealt with in changing emotion in life and in psychotherapy. So
questions of the ecological validity of what many experimental psychologists
study in the lab always need to be kept in mind. It is important then to not only
assume continuity or discontinuity between in the lab emotions and life emotions,
but to investigate which of the laboratory findings address basic emotion processes
that are found both in life and in therapeutic change, and which are based on higher
forms of cognitive- affective process that can be produced in the lab? In addition,
as some of the chapters highlight, emotion occurs most importantly in relational
contexts and needs to be studied as such. Affective science, and affective neuro-
science, will be greatly enhanced by studying the dynamics of emotions as they
arise and change in the context of important life events and relationships.
Further, whereas much emotion research has focused on emotion regulation,
if we are seeking long-term change in emotions, we must regulate the whole
Postscript 225
system, and then emotion regulation is not distinct from emotion generation.
From a clinical and therapeutic perspective, emotion also has been shown to be
both adaptive and maladaptive. In therapy, emotions thus at times need to be
accessed, accepted, and used as guides, and at other times regulated and modified.
This appears to be because emotion seems to serve both epistemological and
hedonic functions, on the one hand informing us of our reactions to situations,
and on the other hand being a major source of pleasure and pain. Which of these
it serves may depend on its intensity. The role of the cognitive processing of
emotion in therapy also has been found to be twofold; either to help make sense
of the emotion, or to help regulate it. All this complexity needs ultimately to be
encompassed by experimental studies.
To add to this, an important clinical distinction, made in EFT theory, is that
between primary and secondary emotion on the one hand, and adaptive and
maladaptive emotion on the other (Greenberg, 2002). Primary emotions are
defined phenomenologically, as a person’s core, first, immediate, gut response to
a situation, such as sadness at loss or fear at threat, whereas secondary emotions
are defined as responses to preceding emotional reactions (e.g., anger at hurt,
hopelessness covering suppressed anger). Secondary emotions can also be
secondary to more cognitive processes (e.g., anxiety in response to catastrophic
thinking). Primary adaptive emotions are responses to the situation that serve
a person’s goals, needs, and concerns, whereas primary maladaptive emotions
are core painful emotions that are more a reflection of past unresolved issues
and unmet needs (e,g. fear in response to intimacy or the shame of inadequacy).
Measures of these constructs have been developed and their predictive validity
demonstrated (Hermann and Greenberg 2008), but more fundamental research
on these more complex emotional processes would be helpful. Can one’s first
automatic emotion, in the clinically relevant sense described here, be distin-
guished from more complexly processed secondary emotions?
In addition, basic research on the principles of emotional change described
below that have been articulated within the therapeutic context would be
enormously helpful. Six major principles of emotional change that appear
important in therapy (Greenberg, 2010) are awareness; expression; regulation;
reflection; transformation; and corrective emotional experience. Awareness of
what one feels appears to provide access to the adaptive information and the
action tendency in the emotion, and reconnects people to their motivation to meet
the needs and goals embedded in the emotions. An additional aspect of awareness
is acceptance of emotion which appears to reduce distress. Expression is a further
process that involves acceptance. It involves saying or showing what one feels
through the use of words or actions. Expressing emotion in therapy involves
overcoming avoidance to experience and expressing previously constricted
primary emotions, not to get rid of them, but to help move and inform people.
Regulation, on the other hand, involves managing emotional intensity, often of
secondary or primary maladaptive emotions. Reflection on emotional experience
helps people make sense of their experience, and promotes it’s assimilation into
their ongoing self-narratives.
226 Les Greenberg
The most novel and important principle for psychotherapy is the transformation
of emotion by emotion. This applies most specifically to transforming primary
maladaptive emotions such as fear, and the sadness of lonely abandonment and
shame (Greenberg, 2002, 2010). Often these withdrawal emotions are trans-
formed by accessing the adaptive approach emotions of empowering anger that
sets boundaries and helps overcomes obstacles, and the contact seeking sadness
of grief that both promotes compassion for the self, and reaches out to others
for contact/comfort. This principle of emotional change thus suggests that a
maladaptive emotional state can be undone by activating another more adaptive
emotional state. This involves the client first experiencing the maladaptive
emotion, in order to make it accessible to transformation. Just talking about
emotion is not helpful; rather it is in experiencing it and then in doing something
different that the possibility of change lies. This process of changing emotion
with emotion goes beyond ideas of catharsis, exposure, completion, letting go,
habituation, or detachment, in that the maladaptive emotion is not purged, nor is it
simply attenuated by the person feeling it. Rather, another more adaptive emotion
is used to transform or undo it. Transformation also does not simply involve
exposure to feared and avoided internal or external cues, but rather involves
changing emotions that are too often felt too much.
Some experimental research on changing negative emotion with positive
emotion already exists (Fredrickson, 1998), and this shows that one emotion
undoes rather than replaces the first emotion, to create a new emotional state by
some form of synthesis. In addition, research on the process of memory recon-
solidation (Nadel and Bohbot, 2001), in which memories can be updated by the
incorporation of new experience in the present, holds great promise for under-
standing the changing of emotion memories. Introducing new present experience
into currently activated memories of past events has been shown to lead to
memory transformation by the assimilation of new material into past memories
(Nadel and Bohbot, 2001). Memories activated in the present are restructured by
the new experience of both being in the context of a safe relationship, and by the
co-activation of more adaptive emotional responses to the remembered incident,
and the addition of new adult resources and understanding to help cope with the
old situation. The memories are reconsolidated in a new way by incorporating
these new elements. As memory reconsolidation only occurs once a memory is
activated, it follows that emotional memories have to be activated in therapy in
order to be able to change them, and that if within minutes of this a new emotion
is experienced, it will be incorporated into the memory and can change the
experience of the original memory
A final way of changing an emotion is to have a corrective emotional
experience in the world that changes an old feeling. New lived experiences with
another person (often the therapist) are especially important in providing an inter-
personal corrective emotional experience. For example, having one’s shame met
with acceptance and compassion rather than rejection engenders a new feeling
and leads to a transformation of the shame.
Postscript 227
Methods for accessing new emotions to change old emotions
A number of ways of helping the client in therapy access new emotions to change
old emotions have been outlined and studied (Greenberg, 2002). First empathy
often helps client access new feelings. Therapists can help the client access new
subdominant emotions occurring in the present by shifting attention to these
emotions that are currently being expressed non-verbally, but are only “on the
periphery” of a client’s awareness. When, however, no other emotion is present,
focusing on the unmet need—when a person is in a maladaptive emotion state—is
a key method of accessing a new emotion. Because emotion is generated by the
automatic appraisal in relation to a need when a need is raised to awareness, the
emotion system automatically generates a new emotional response to the unmet
need and thereby mobilizes a new, more adaptive emotion. These new feelings
were felt in the original situation but not expressed, or are felt now as an adaptive
response to the old situation. For example, accessing adaptive anger at violation
by a past perpetrator helps change maladaptive fear in a trauma victim. When the
tendency to run away in fear is transformed by anger’s tendency to thrust forward,
a new relational position of holding the abuser accountable for wrongdoing is
formed (Greenberg and Malcolm, 2002).
Other methods of accessing new emotion involve using enactment and
imagery to evoke new emotions, or simply remembering a time an emotion
was felt. These often help clients access that emotion, as does changing how
the client views things. Once accessed, these new emotional resources begin to
undo the psycho-affective motor program previously determining the person’s
mode of processing. In this view, enduring emotional change of maladaptive
emotional responses occurs, not through a process of insight, nor through
cognitive restructuring, but by generating new responses to old situations and
incorporating these into memory. Experimental testing of this hypothesis is
needed.
Emotion sequences in emotional change

Based on both clinical theory and practice, a model for the sequence involved in
the transformation of “bad feelings” has been proposed and tested (Greenberg
and Paivio, 1997; Pascual-Leone and Greenberg, 2007; Hermann and Greenberg
2008). This process involves moving from secondary emotions, through primary
maladaptive emotions, to primary adaptive emotions. A refined model of this
core change process has been validated in clients in therapy who successfully
resolved initial states of high emotional distress and low levels of meaning to
arrive at states of high meaning and low emotional distress (Pascual-Leone and
Greenberg, 2007)
This process of transformation of distressed feelings begins with attending
to the aroused bad feelings (I feel bad), and through exploration of secondary
states (I feel hopeless), arriving at activated core maladaptive emotion schematic
memories or self-organizations, based on core painful feelings of fear of
228 Les Greenberg
abandonment accompanied by sadness, or the shame of worthlessness (I can’t
survive alone, or I’m worthless). At this point in the transformation process,
acknowledging previously unmet needs (I deserved to be comforted/validated)
leads to the expression of new adaptive experiences of adaptive grieving for
what wasn’t, self-soothing, and /or empowering anger at violation. These newly
emerging adaptive feelings facilitate a sense of self-acceptance and agency
The movement depicted in this process, from secondary emotions through
primary maladaptive emotion, through needs to primary adaptive emotion repre-
sents a core change process in EFT. Transformation of core maladaptive states
occurs at a critical point when these states are differentiated into entitlement to
adaptive needs which serves to challenge the negative evaluations embedded in
the maladaptive emotion. Experiencing that one deserves to have one’s needs met
acts to refute core negative evaluations about the self.
Conclusion
Increasing complexity in the study of emotional change in experimental studies
in the future will add the rigour needed to sort out which hypotheses developed
from clinical experience and theorizing are reliable and which need to be revised.
Meanwhile, the advances in research represented in this book on changes in
emotion are to be applauded for carrying forward the study of emotion in
psychology.
References
Fredrickson, B. L., Mancuso, R. A., Branigan, C., and Tugade, M. M. (2000). The undoing
effect of positive emotions. Motivation and emotion, 24, 237–258.
Greenberg, L. S. (2002). Emotion-focused therapy: Coaching clients to work through their
feelings. Washington, DC: American Psychological Association.
—(2010). Emotion-focused therapy: Theory and Practice. Washington, DC: American
Psychological Association.
Greenberg, L. S., and Pascual-Leone, J. (1995). A dialectical constructivist approach to
experiential change. In R. A. Neimeyer and M. J. Mahoney (eds), Constructivism in
psychotherapy, (pp. 169–191). Washington, DC: American Psychological Association.
—(2001). A dialectical constructivist view of the creation of personal meaning. Journal of
Constructivist Psychology, 14, 165–186.
—(2006). Emotion in psychotherapy: A practice-friendly research review. Journal of
Clinical Psychology: In Session
Greenberg, L. S., and Malclom, W. (2002). Resolving unfinished business: Relating
process to outcome. Journal of Consulting and Clinical Psychology, 70, 406–416.
Greenberg, L. S., and Paivio, S. C. (1997). Working with emotions in psychotherapy. New
York, NY: Guilford Press.
Greenberg, L. S., and Safran, J. D. (1987). Emotion in psychotherapy: Affect, cognition,
and the process of change. New York, NY: Guilford Press.
Greenberg, L. S., and Watson, J. C. (2006). Emotion-focused therapy for depression.
Washington, DC: American Psychological Association.
Postscript 229
Hermann, I. and Greenberg, L. (2008). Emotion types and sequences in Emotion-focused
therapy. European Psychotherapy, 7, 41–60
Nadel, L., Bohbot, V. (2001). Consolidation of memory. Hippocampus, 11, 5660.
Pascual-Leone, A., and Greenberg, L. (2007). Emotional Processing in Experiential
Therapy: Why ‘the only way out is through’. Journal of Consulting and Clinical
Index
abstract processing 195–8, 200; and Beck, A. T. 195, 197

depression 198, 200 behavior 24–9, 157–8; children’s 24–9,
Acceptance and Commitment Therapy 37–42; definition of 67; emotional
(ACT) 202, 206–7; see also therapies 109–11, 148; parent 25–6, 75–6;
acculturation 127–32 problem 37–42; prosocial 19–20;
acquired equivalence learning 69 see also behavioral principles and
ACT (Acceptance and Commitment behavioral strategies
Therapy) 202, 206–7; see also therapies behavioral principles 203
action readiness 138 behavioral strategies 3–8
actual affect 121–4 Beike, D. R. 176
actual-ideal self discrepancies 210–12 Bekker, E. M. 40
actual-ought self discrepancies 210–11 Berridge, K. C. 137
adaptive emotions 225, 228 Beutler, L. E. 209
Affect Valuation Theory (AVT) 25, 120–5 Biederman, J. 50
affective science 209 Bishop, S. J. 41
aggression 41–2 Blackwell 192
aging 31–5 Blanchard-Fields, F. 32, 33
Ainslie, G. 141 Blandon, A. Y. 7
Anderson, C. 132 Bobowik, M. 118
anxiety 41–2, 55, 68, 171 Bohanek, J. G. 14
appraisal theories 97, 138, 146–9 Bond, M. 25
Aron, A. R. 158 Bouton, M. E. 53, 62, 63, 65, 132
Askew, C. 47 Bower, G. H. 147
Auguston, E. M. 204, 206 Brans, K. 175, 177, 179
autobiographical memory 10, 197, 220–1 broaden-and-build theory 152–5; see also
autobiographical narratives 10; see also positive under emotion
narratives Brown, K. W. 171
AVT (Affect Valuation Theory) 25, 120–5 Bubier, J. L. 41
awareness 225 Burleson, B. R. 93
Buss, K. A. 7
Babb, K. A. 6
Badre, D. 158 Cacioppo, J. T. 33
Baer, R. A. 219 Callaghan, B. L. 86
Bailey, P. E. 21, 22 Calton, J. L. 61
Bakeman, R. 109 Campbell, Joseph 120
Bar-Tal, D. 114 Campos, J. J. 140
Barbas, H. 159 Camras, L. A. 26, 27, 223
Bargh, J. A. 138 Capaldi, E. J. 60
Barnes-Holmes, D. 204 Cappadocia, M. C. 40
Index 231
Carlson, S. M. 7 Damasio, A. R. 183
Carstensen, L. L. 31, 32, 33, 34 Davey, G. 49
Carver, C. S. 93, 195, 197 Davidson, R. J. 174, 178, 179
Caspi, A. 37 Davis, E. L. 3, 5, 8
Cavada, C. 159 Davis, M. 82
CBM-I (Cognitive bias modification for Davitz, J. R. 138
interpretation) 188–9 De Houwer, J. 191
CBT (Cognitive Behavioral Therapy) 39, De Leersnyder, J. 27, 128, 130
197–200; see also therapies De Rivera, J. 114, 115, 116, 117
Chaiken, S. E. 157, 158, 161 De Waal, F. B. M. 141
Charles, S. T. 32, 33 Debiec, J. 57
Chiao, J. Y. 26 Decety, J. 160
Chim, L. 122, 125 deliberate emotion regulation 38–9, 141;
Christophe, V. 93, 94, 95 see also emotion regulation
classical conditioning 47–51, 67–8; see Denham, S. A. 25
also conditioning Dennis, T. A. 5
closure 175–6; see also episodes under depression 189, 197–200, 210–5
emotion Derakshan, N. 140
Cognitive Behavioral Therapy (CBT) 39, DeRubeis, R. J. 198
197–200; see also therapies dialectical constructivism 223–4
Cognitive bias modification for dialectics 218
interpretation (CBM-I) 188–9 Dickinson, T. 61
cognitive processing 195–200 disease avoidance mechanisms 74–5; see
cognitive strategies 4–5; and children 4–8, also disgust
37; see also emotional regulation disgust 74–9
Cognitive Therapy (CT) 214; see also disgust propensity (DP) 74, 78
therapies distraction 145, 166–70
Cohn, M. A. 154 Dollard, J. 41
Cole, P. M. 3, 4, 7, 24, 25, 26 Doost, H. T. N. 129
collective coping 117–18 Dougher, M. J. 70, 71, 205, 206
collective emotional events 91 Douglas, William Orville 120
collective identity 115–6; see also group Dunbar, R. I. M. 98
identity and social identity and social Dunsmore, J. C. 25
sharing Durkheim, E. 95
collective memory 117 Dymond, S. 71, 206
Collins, N. L. 95
concordance 128–31 Easterbrook, J. A. 140
concrete processing 195–200; and EC (emotional climate) 113–19
depression 197–200 Eddington, K. M. 210, 211
conditioned inhibition 61 EFT (Emotion Focused Therapy) 225,
conditioning 53–4, 60–2; classical 228; see also therapies
47–51, 67–8; evaluative 76–9; and fear Egliston, K. 50
53–4; instrumental 62; Pavlovian 47, Eisenberg, N. 6, 17, 18, 19, 20, 22, 25
72, 81 Ekman, P. 27
contagion learning 78–9 Ellsworth, P. C. 138
control precedence 138 Emery, L. 33
Cook, M. 47, 49 emotion 37–8, 93, 137, 144, 174–5;
corrective emotional experience 225–6 adaptive 225, 228; causation 144–50;
couples interaction 107–11 change strategies 145–50; conflict
CT (Cognitive Therapy) 214; see also 140–1; control 4, 7 157–63 (see also
therapies emotional regulation); duration 174–9;
Curci, A. 94 embodiment 28; episodes 91–3, 175–6,
Curtis, C. E. 139, 140 178–9; maladaptive 204–5, 225–6, 228;
Cuthbert, B. N. 167 multiple 138–40; narratives 10–15;
232 Index
positive 108–9,151–5, primary 225; Fredrickson, B. L. 13, 109, 151, 152, 153,
secondary 225; socialization 25–9 154, 226
Emotion Focused Therapy (EFT) 225, 228 Freedman, D. G. 24, 26, 27
see also therapies Frijda, N. H. 137, 138, 140, 147, 174, 175,
emotional acceptance 217–18 see also 176, 177, 178
Mindfulness-based interventions Frith, C. D. 160
emotional acculturation 127–32 functional approach 203–7
emotional concordance 128–31 functional contextualism 202; see also
emotional climate (EC) 113–19 functional approach
emotional events 91, 124–5, 138–9,
185–6 Gable, P. A. 152
emotional experience 10–15, 31–5, 91–5, Gacaca 117–18
127–32, 225–6; see also social sharing Gallese, V. 160
and couples interaction Gallistel, C. R. 60
emotional expression 24–9 Garland, E. L. 154, 155
emotional patterns 127–31 Garnefski, N. 6, 8
emotional recovery 92–6 Gaudiano, B. A. 202, 206, 207
emotional regulation 3–4, 137–41, generalization 68–73
159–60; and aging 31–4; dynamics Gerstorf, D. 31
107–11; and expressive writing 183–6; Gerull, F. C. 76
neural mechanisms 37–42; and temporal Ghashghaei, H. T. 159
dynamics 166–72 Gilboa, E. 176, 177, 179
emotional suppression 217–18 goals 210–5
empathy 17–22, 108, 160–2; see also Goldman-Rakic, P. S. 159
social sharing Gordijn, E. H. 98
engagement strength 215 Gottman, J. 105, 107
equivalence class learning 71 Graham, B. M. 86
Espinosa, A. 115, 116 Graham, Billy 120
Espitalier, M. 98 Greenberg, L. S. 223, 224, 225, 226, 227
evaluative conditioning (EC) 76–9; see Gross, J. J. 32, 137, 141, 145, 148, 166,
also conditioning 167, 169, 171, 177
expression 225; and cultures 25; Grossman, P. 219
facial 148–9; see also emotional group-based appraisals 97–102
expression group-based emotions 97–103
expressive suppression 171; see also group belonging 95–6
emotional regulation group identity 97; see also collective
expressive writing 14, 183–6 identity and social identity and social
extinction 53–8, 60–5; fear 51, 53–8, sharing
81–6 Grühn, D. 20, 21
Falkenstein, M. 40 Hackmann, A. 193

fear 47–51; acquisition of 47–51; Hamm, A. 48
conditioning 53–8, 68; extinction 51, Harber, K. D. 95
53–8, 81–6; generalization 68–73, 81–7; Hariri, A. R. 212
neurobiology 54–6, 82–6; responses 68; Harris, P. L. 4
transfers 70–1 Hatfield, E. 28
Feeley, M. 198 Hay, D. F. 19
Fields, L. 6 Hayes, S. C. 202, 203, 206, 207
Finkenauer, C. 92 Heckhausen, J. 4, 5, 33
Fischer, A. H. 13 Heeren, A. 220
Fivush, R. 10, 12, 14 Helson, R. 21
Flavell, J. H. 4 Hemenover, S. H. 177, 178, 179
Foti, D. 167 Hermann, I. 225, 227
Frattaroli, J. 184 Hermans, D. 68
Index 233
Higgins, E. T. 210 LaBarre, W. 24
Hilgard, E. R. 140 Labouvie-Vief, G. 33
Hoffman, M. L. 18 Lagattuta, K. 5
Hofstede, G. 24 Lambie, J. A. 138
Holmes, E. A. 188, 189, 190, 191, 192 Lang, T. J. 193
Holodynski, M. 25, 28 Langton, J. M. 83
Honey, R. C. 72 Larson, C. L. 171
Lastrego, S. 118
Ickes, W. 108n1 Lazarus, R. S. 146
ideal affect 120–5 learning experiences 67–9
identity: collective 115–6; group 97; social LeDoux, J. E. 54
100–2, 115–6; see also social sharing Lennon, R. 20
IET (intergroup emotion theory) 97–103 Leon, S. P. 64
inflation effect 50 Leu, J. 123
inhibitory associations 53–6, 61; see also Levenson, R. W. 105, 106, 108, 109, 111
extinction Leventhal, H. 147
institutional reparatory acts 118 Levitt, J. T. 171
instrumental conditioning 62; see also Lewis, M. D. 40, 41
conditioning Lieberman, M. D. 157, 158, 171
intergroup emotional theory (IET) 97–103 Lissek, S. 55, 69
Isaacowitz, D. M. 34 LKM (loving-kindness meditation) 153–4
Löckenhoff, C. 34
Jackendoff, R. 138 Loewenstein, G. F. 157, 158
Jeannerod, M. 138 Lonsdorf, T. B. 51
Jha, A. P. 219 loving-kindness meditation (LKM)
Johns, K. W. 72 153–4
Johns, M. J. 171 Luciano, C. 206, 207
Johnson, D. P. 155 Lumley, M. A. 185
Johnstone, T. 163 Luong, G. 31
de Jong 68, 74, 78, 79 Lybomirsky, S. 151
Jovanovic, T. 55
McBride, C. 198
Kabat-Zinn, J. 218 McCallum, J. 85
Kagan, J. 24, 26, 27 Mackintosh, N. 47, 48
Kappas, A. 38 McNally, R. J. 81
Kashdan, T. B. 171 MacNamara, A. 169
Kerkhof, I. 78 McSweeney, F. K. 61
Keysers, C. 160 maladaptive emotions 205–6, 225–6, 228
Kim, J. H. 82, 83, 84, 85 Manber, R. 198
Kim, Y. 185 Manstead, A. S. R. 98
Kitayama, S. 124 Markus, H. R. 24, 127
Klein, K. 185 Martin, L. L. 93
Klenk, M. M. 215 Martin-Beristain, C. 92, 94, 118
Knafo, A. 17, 19, 22 Mason, E. C. 77
Kober, H. 158 Mather, M. 34
Koenigsberg, H. W. 163 Matsumoto, D. 25
Kok, B. E. 153, 154 MBIs (Mindfulness-based interventions)
Koole, S. L. 3, 4 218–22
Koopmann-Holm, B. 123, 125 MBSR (Mindfulness-Based Stress
Kosslyn, S. M. 158, 187 Reduction program) 219
Kroeber, A. L. 120 meditation 153–4
Kuhl, J. 141 memory 82, 93, 226; autobiographical 10,
Kunzmann, U. 32 197, 220–1; collective 117; fear 57–8,
Kuppens, T. 101, 102 82
234 Index
mental imagery 187–93; and depression Páez, D. 113, 114
191–3; experiments in 188–92; and parent: behavior 25–6, 75–6; influence
field perspective 190–1; and observer 12–15
perspective 190–1; and perception Parent Management Training (PMT) 39
190–1 Pascual-Leone, A. 227
mental representations 195–200; and Pavlovian conditioning 47, 72, 81, 204;
abstract processing 195–8, 200; and see also conditioning
concrete processing 195–200; and Pennebaker, J. W. 13, 91, 184, 185
depression 197–200 perceptual similarity dimension 69
Mesquita, B. 24, 127, 137 personal narratives 10–12; see also
Metcalfe, J. 157, 158 narratives
Milad, M. R. 54, 55 Perunovic, W. Q. E. 123
Miller, E. K. 158 Peters, K. 98
Miller, P. 25 Phelps, E. A. 158
mimicry 28 Philippot, P. 197, 219, 220
mindfulness 218 Phillips, L. H. 33
Mindfulness-based interventions (MBIs) Phillips, M. L. 38
218–22 phobias 47–51, 205–6; see also fear
Mindfulness-Based Stress Reduction physiological linkage 107–8
program (MBSR) 219 physiological soothing 108–10
Mineka, S. 47, 48, 49, 50, 205 Pictet, A. 191, 193
Mischel, W. 7 PMT (Parent Management Training) 39
Mitchell, J. P. 160 Pons, F. 4
Moberly, N. 149 positivity effect 34
Mogg, K. 171 prevention failure 210–11
Monfils, M. H. 57 primary emotions 225, 228
Moon, A. 122 processing: abstract 195–8, 200; concrete
Moore, A. 220 195–200; and depression 197–200
Moore, C. 4 promotion failure 210–12
Morsella, E. 141 prosocial behavior 19–20; see also
Moses, E. B. 217 behavior
motive states 137–8 Psychological Situation Analysis 214
multiple emotions 138–41
Quirk, G. J. 55, 81, 82, 85
Nadel, L. 226
Nader, K. 57 Raes, F. 197, 199
narratives: autobiographical 10; emotion Ramnani, N. 211
10–15; personal 10–12; writing 14, Ray, R. 38
183–6 reappraisal 145, 149, 159, 167–70; see
Nelson, J. B. 53, 54, 60, 62, 63, 64 also emotion change
Nelson, K. 10 reflection 225
network theories 147–9 see also emotion regulation 225
causation regulatory fit 215
neural mechanisms 37–42, 81–7, 158 regulatory focus theory 210–12
Nils, F. 93 Relational Frame Theory (RFT) 202–6
Nolen-Hoeksema, S. 15 Rescorla, R. A. 50, 56, 62
Ressler, K.J. 41
Oaten, M. J. 74 RFT (Relational Frame Theory) 202–6
Ochsner, K. N. 149, 158, 159, 162, 163 Rice, J. A. 3
Oehlberg, K. 51 Richards, J. M. 171
Öhman, A. 47, 48. 49 Richter, D. 20, 21, 22
Oishi, S. 123, 125 Rimé, B. 10, 13, 91, 92, 94, 95, 98, 102,
Olatunji, B. O. 77 114, 117
Olsson, A. 160 rituals 117–18
Index 235
Robbins, S. J. 60 Sonnemans, J. 174, 175, 176, 178, 179
Robinson, M. D. 124 SST (self-system therapy) 213–14; see
Roche, B. 204, 205, 206 also therapies
Rosas, J. M. 63, 65 Stevenson, R. J. 75
Rowe, G. 152 stimulus equivalence 204
Rozin, P. 74, 75, 78, 79 Strack, F. 141, 149
Ryder, A. 129 Strauman, T. J. 209, 211, 214
sympathy 17–22
Saarni, C. 132
Salzberg, S. 153 Tarrier, N. 155
SBA (Self-Belief Analysis) 214 Techio, E. 114, 115, 116, 117
Sbarra, D. A. 175, 177, 179, temporal dynamics 109–11, 166–72,
SCA (Self-in-Context Assessment) 174–9; see also emotional regulation
213–14 tertiary sharing 95; see also social
Scheibe, S. 31, 33 sharing
Scherer, K. R. 3, 97, 146, 176, 178, 183 Thayer, J. F. 153, 154
Schieman, S. 21 therapies 39, 63, 81, 197–200, 202, 206–7,
Schienle, A. 76 213–14, 224–8
Schiller, D. 82 therapy, principles of 225–6
Schmitz, T. W. 152 Thiruchselvam, R. 167, 168, 169
Schultz, W. 158 Thompson, R. A. 7
Scollon, C. N. 125 Tierney, K. J. 71
SCT (Self-Categorization Theory) 97–8 Tolin, D. F. 78, 79
secondary emotions 225, 228 Totterdell, P. 132
secondary social sharing 94–5; see also transformation 226–8
social sharing transformation of function 204–6
Self-Belief Analysis (SBA) 214 Tsai, J. L. 25, 121, 122, 123, 223
Self-Categorization Theory (SCT) 97–8 Turner, J. C. 97
self-definition 10–15
Self-in-Context Assessment (SCA) 213–14 Valverde, M. R. 204
self-regulation 13- 15, 209–15 Van Emmerik, A. A. 92
self-system therapy (SST) 213–14; see Verduyn, P. 175, 176, 177, 178, 179
also therapies Vervliet, B. 56, 69, 72
Seligman, M. E. P. 48 video recall methodology 105, 107–8
Serrano, J. P. 199 Vieth, A. Z. 213
Sheppes, G. 167, 169, 170
Shin, L. M. 54 Wadlinger, H. A. 152
Shiota, M. N. 110 Wager, T. D. 158
Shuster, M. 28 Wagner, A. R. 61
Sidman, M. 70, 204 Walker, D. L. 56
Sims, T. 122 Wang, Q. 13, 25,
Slagter, H. A. 171 Watkins, E. R. 195, 196, 197, 198, 199,
Sloan, D. M. 184 200, 220
Sloman, S. A. 157, 158 Wellman, H. M. 4
Smith, E. R. 97 Whelan, R. 205
Smith, L. G. E. 98 White, K. 50
Smyth, J. M. 184, 185 Williams, J. M. G. 197, 198, 220
social identity 100–2, 115–6; see also Wilson, T. D. 169
group identity and social sharing Woltering, S. 39
social identity salience 100–2 Wright, M. 8
social referencing 75–6 writing 14, 183–6
social sharing 91–5, 97–103, 117
socioemotional selectivity theory 34 Yuan, J. W. 109, 110
236 Index
Yzerbyt, V. Y. 97, 98, 99, 100, 103 Zech, E. 92, 94
Zelazo, P. D. 141
Zahn-Waxler, C. 19
Zaki, J. 160, 161, 162, 163

Dirk Hermans, Bernard Rimé, Batja Mesquita - Changing Emotions

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dirk Hermans, Bernard Rimé, Batja Mesquita - Changing Emotions

Uploaded by

Copyright:

Available Formats

Changing Emotions

Changing Emotions highlights several recent developments in this intriguing

In each chapter, an internationally renowned scholar presents a concise review

Bernard Rimé is Emeritus and Invited Professor at the University of Louvain at

Proceedings of the Contactforum “Changing Emotions” (October 23, 2009)

1 How kids keep their cool 3

2 Defining and regulating the self through emotion narratives 10

3 Age-related changes in empathy-related responding 17

4 Children’s expressive behavior in different cultural contexts 24

5 Shifts in emotional experience and regulation across adulthood 31

6 Changing the neural mechanism of emotion regulation in children

7 Individual differences in the acquisition of fears 47

8 Extinction learning and its retrieval 53

9 Mechanisms of extinction in emotional regulation 60

10 Generalization as a basis for emotional change 67

11 Learning mechanisms in the acquisition of disgust 74

12 Preclinical analysis of developmental transitions in the

13 Can socially sharing emotions change emotions? 91

14 From group-based appraisals to group-based emotions 97

15 Emotion and emotion regulation 105

16 Emotional climate 113

17 Dynamics of ideal affect 120

19 Emotion regulation 137

20 Understanding emotion change requires an understanding of

21 Learning to self-generate positive emotions 151

22 The role of control in emotion, emotion regulation, and empathy 157

23 What time can tell us 166

24 The duration of emotional episodes 174

25 Can expressive writing change emotions? 183

26 The powerful impact of mental imagery in changing emotion 187

27 Cognitive mechanisms involved in therapeutic change for depression 195

28 A functional approach to the study of human emotion 202

30 Mindfulness-based interventions 217

Postscript: Experimental rigor and clinical complexity 223

Frank Baeyens, University of Leuven, Belgium

Dirk Hermans, Bernard Rimé, and Batja Mesquita

Behavioral and cognitive strategies for regulating emotion

The development of young children’s use of cognitive strategies

Factors that help and hinder children’s use of cognitive strategies

Defining the self through emotion narratives

Mother: Were you scared (of the bear)?

(a few conversational turns later)

Child: It was scary.

In this excerpt, we see the glimmerings of the daughter’s concept of herself as

Regulating the self through emotion narratives

Empathy is commonly defined as a response that stems from the apprehension

Contemporary developmental theories of emotion emphasize the important role

Culture, emotion, and emotional expressivity

Biological explanations for cultural differences

Emotional expressivity in adopted Chinese children

Extending the model of emotion socialization

Aging is associated with decline in many psychological processes, such as

Aging and emotional experience

Aging and emotion regulation

Externalizing behavior problems are characterized by high levels of aggression

Emotion regulation and its neural correlates

Changes in emotion regulation with treatment of externalizing

The anxiety hypothesis of aggression

Anxiety Frustration Reactive

Fear conditioning involves learning a relationship between a neutral stimulus,

Retrieval model of extinction

Neurobiology of fear extinction

Deficits in extinction learning in anxiety disorders

Enhancing the formation and retrieval of extinction learning