From www.drugswell.com we present to you the ebook:
IN ORGANIZATIONAL BEHAVIOR
GIVE US HELP AND WE GIVE YOU MORE INFO firstname.lastname@example.org CONTACT US, FOR YOUR DONATION HERE (CLICK)
This page intentionally left blank
Charmine E. J. Hartel
Wilfred J. Zerbe
Neal M. Ashkanasy
LEA LAWRENCE ERLBAUM ASSOCIATES, PUBLISHERS 2005 Mahwah, New Jersey London
Copyright © 2005 by Lawrence Erlbaum Associates, Inc. All rights reserved. No part of this book may be reproduced in any form, by photostat, microform, retrieval system, or any other means, without the prior written permission of the publisher.
Lawrence Erlbaum Associates, Inc., Publishers 10 Industrial Avenue Mahwah, New Jersey 07430
Cover design by Kathryn Houghtaling Lacey
Library of Congress Cataloging-in-Publication Data
Emotions in Organizational Behavior, edited by Charmine Hartel, Wilfred J. Zerbe, and Neal Ashkanasy.
ISBN 0-8058-5098-8 (cloth : alk. paper).
Includes bibliographical references and index.
Copyright information for this volume can be obtained by contacting the Library of Congress.
Books published by Lawrence Erlbaum Associates are printed on acid-free paper, and their bindings are chosen for strength and durability.
Printed in the United States of America 1 0 98765432 1
Getting to a point like this where one writes a book requires much dedication, effort, and driving force. I have thought a lot about what is behind my driving motivation. I must say that it is not the task itself but the relationships I have with the people involved and those in my life. So I dedicate this book to those people whose names don't appear in any of the bylines but whose love or devotion fuels my passions and work. And I thank my children and partner, who have understood this and loved me more for it.
I would like to dedicate this book to my late parents, Maurice Ashkanasy and Heather Ashkanasy.
With thanks to the many people who have supported this work in countless ways: colleagues, friends, and family.
This page intentionally left blank
|PrefaceCharmine E. f. Hartel, Wilfred||J. Zerbe, and Neal M. Ashkanasy||xv|
|About the Editors and Contributors||xix|
|1||Organizational Behavior: An Emotions PerspectiveCharmine E. F. Hartel, Wilfred F. Zerbe, and Neal M. Ashkanasy||1|
|Part I:||Organizational Behavior and Emotions|
|2||Emotions: From "Ugly Duckling" Via "Invisible Asset" Toward an Ontological ReframingDorthe Eide||11|
|Part II:||The Individual Within the Organization|
|3||"You Wait Until You Get Home": Emotional Regions, Emotional Process Work, and the Role of Onstage and Offstage SupportMaree V. Boyle||45|
Yongmei Liu and Pamela L. Perrewe
Yuka Fujimoto, Charmine E. F. Hartel, and Debra Panipucci
6 A Bounded Emotionality Perspective on the Individual in the Organization 113
Neal M. Ashkanasy, Wilfred F. Zerbe, and Charmine E. F. Hartel7 Individual and Group Affect in Problem-Solving Workgroups 119
Matthew F. Grawitch and David C. Munz
Stefan Meisiek and Xin Yao
Mona White, Charmine E. F. Hartel, and Debra Panipucci
Neal M. Ashkanasy and Wilfred F. Zerbe
Robert S. Rubin, Vicki M. Staebler Tardino, Catherine S. Daus, and David C. Munz
|12||Toward Understanding Emotional Management at Work: A Quantitative Review of Emotional Labor ResearchJoyce E. Bono and Meredith A. Vey||213|
|13||The Interaction Effect of Emotional Intelligence and Emotional Labor on Job Satisfaction: A Test of Holland's Classification of OccupationsChi-Sum Wong, Ping-Man Wong, and Kenneth S. Law||235|
|14||The Relationship With Patients: "Emotional Labor" and Its Correlates in Hospital EmployeesVanda L. Zammuner and Cristina Galli||251|
|15||A Bounded Emotionality Perspective on Work CharacteristicsWilfred F. Zerbe and Charmine E. F. Hartel||287|
|Part V:||Organizational Change and Changing Organizations|
|16||Emotion Management to Facilitate Strategic Change and Innovation: How Emotional Balancing and Emotional Capability Work TogetherQuy Nguyen Huy||295|
|17||Managing Emotion: A New Role for Emergent Group LeadersAnthony T. Pescosolido||317|
|18||For Better or For Worse: Organizational Culture and EmotionsMichelle K. Pizer and Charmine E. F. Hartel||335|
|19||A Bounded Emotionality Perspective on Organizational Change and CultureNeal M. Ashkanasy and Charmine E. F. Hartel||355|
|20||What an Emotions Perspective of Organizational Behavior OffersCharmine E. F. Hartel, Neal M. Ashkanasy, and Wilfred F. Zerbe||359|
This page intentionally left blank
Foreword: Workplace Emotion:
Where We've Been, Where We're
Going, and Where We Ought to Be
Professor Russell Cropanzano
Workplace emotion has finally arrived! And none too soon for many of us. In the span of about a decade, emotion scholars have gone from lamenting the dearth of relevant research (Ashford & Humphrey, 1995; Pekrun & Frese, 1992) to celebrating a new explosion of interest (Ashkanasy, Hartel, & Daus, 2002; Brief & Weiss, 2002). There is more than just volume in this current cascade. Emotion researchers have begun to incorporate innovative research strategies, such as qualitative inquiry (Ayoko & Hartel, 2002; Rafaeli & Sutton, 1990; Sutton, 1991; Sutton & Rafaeli, 1988), experience sampling (Weiss, Nicolas, & Daus, 1999; Williams, Suls, Alliger, Learner, & Wan, 1991), and multidimensional scaling (Russell, Lewicka, & Nitt, 1989). Moreover, scholars have begun to hammer out historically thorny theoretical distinctions. For example, considerable work has distinguished moods from emotions (Weiss & Brief, 2002) and mapped the structure of affect (Cropanzano, Weiss, Hale, & Reb, 2003). Perhaps most exciting of all, workplace emotion has taken an interdisciplinary turn, incorporating work from sociology (Ashford & Humphrey, 1993; Hochschild, 1983), social psychology (Isen & Baron, 1991; Kelly & Barsade, 2001), clinical and counseling psychology (Hartel, Kibby, & Pizer, 2003), and personality (George, 1992, 1996; Judge & Larsen, 2001). This combination of conceptual openness and theoretical flexibility has become a hallmark of emotion research. In one form or another, all of these influences are manifest in this current book. Those of us with long-standing interest in emotion have seen a lot of significant developments in the course of our careers.
If you've read this far you must have at least a passing interest in workplace emotion. As such, you've made my first task easy—I probably don't need to convince you that the topic is, at least potentially, important. But why should you read on? By "on" I don't mean in the mundane sense of finishing this little foreword or even reading through the entire book (although I certainly hope that you do both!). Rather, I mean "on" in a more substantive sense, to refer to the workplace emotion literature as a whole. Why should you stick with it? You probably know that emotion research offers or could offer something now. But should you have confidence in its future? In preparing this foreword, I have thought long and hard about that future. My evaluation is optimistic, of course, but that optimism needs to be based on the potential for future accomplishment and not the dusty glory of past—even recently passed—achievements. It's no longer enough to shout that "Emotion matters!" because we've already convinced everyone (assuming that there was ever a large cadre of doubters). The task before us is to provide the shape and substance that will allow our literature to continue to grow.
Fortunately, we've reached an at least serviceable understanding of the major constructs—affect, mood, emotion, and well-being. With that small, but not insignificant, measure of theoretical cohesion, emotion research has elbowed its way to the table of organizational behavior. We are ready to begin our careers. That future will see chapters in undergraduate textbooks, doctoral seminars, quotes in major newspapers, and popular books on how feelings shape our experience at work (this last prediction has already come true). In broad outline, we know what our future looks like, but what will be the content of those books, quotations, seminars, and chapters? No doubt we will all have much to say, but right now I can't tell you what they will be.
Herein lies a special opportunity for you: In the years to come, you can be part of the intellectual adventure that will provide solutions for those questions. At this particular juncture of our history, the thrill does not come from what we know, but from what remains for us to learn. That's why I encourage you to read on—because the biggest questions have yet to be answered and because fascinating challenges await you and us. The remaining enigmas promise to keep emotion research vital and dynamic for many years to come.
I next argue that at least three challenges wait at the horizon of our understanding. I treat each of these as a tension or dialectic within our discipline. For me, the push-and-pull between superficially opposite-sounding ideas can lead to new insights. This is my personal special list of challenges, and I hope you enjoy it. Of course, it is only my own appraisal. You probably have your own favorite problems and your own ideas for solving them. This is as it should be. So we should take my small questions at face value—they're only questions—and not as anything else. Maybe someday I'll have an opportunity to hear your ideas!
My first challenge pertains to the relationship between power and subtlety. Emotion can be experienced in a holistic and all-encompassing fashion. We speak of being "carried away," "torn apart," "overwhelmed," or "giddy" with our feelings. These words carry communicative meaning only because they designate powerful shared human experiences. Yet despite this not infrequent sense of power and completeness, an emotion is a multifaceted—almost fragile—construct. Each emotion has different parts. Like children's Tinker Toys, at least some of these parts can be disassociated and rebuilt to form slight but important shades of meaning (Mascolo & Griffin, 1998; Mascolo & Harkins, 1998). Seemingly modest changes in how we interpret or analyze an event can alter the course of a torrent of feeling.
To illustrate this point, consider three closely related emotions: embarrassment, shame, and guilt. Each is similar, in that the terms refer to self-conscious evaluations of our own behavior. Additionally, each occurs when we do something that others (and probably we ourselves) perceive as wrong. Despite these basic similarities, there are subtle but important differences in the phenomenology of each. Shame is the most general. We experience the feeling of shame when we behave in such a way as to question our identity as a certain type of person. Guilt is like shame, only more narrowly defined. We experience guilt when we misbehave in a specific instance that does not directly confront our sense of self. That is, we are guilt-ridden when we fail to live up to our own moral standards. Unlike shame and guilt, embarrassment does not carry these moral connotations. We feel embarrassed when we do something silly or dumb, but not when we experience an ethical failure (for details and evidence, see Keltner & Anderson, 2000; Keltner & Buswell, 1997; Tangney, 1995; Tangney & Fischer, 1995).
Notice, of course, that the exact same event can provide any one of these three emotions, depending on how it is understood. Does it pertain to morality? Does it pose a challenge to one's identity? Different answers to these questions alter substantially our affective experience. What are we to make of this? Are emotions affective bulldozers that push aside all else? Or are they delicate will-o'-the-wisps that change their meaning—in some ill-defined psychological version of the Heisenberg Principle—upon close scrutiny of our experience? In some sense, emotions must be both subtle and powerful. Or, perhaps more precisely stated, from the arrangement of subtle events, powerful affective forces can be unleashed (Mascolo & Griffin, 1988; Mascolo & Harkins, 1998).
The second challenge pertains to the relative attention provided to negative and positive feelings. In an insightful paper, Ashford and Humphrey (1995) reminded us that people do more than think. They also feel. While none would gainsay this observation, we should add that people don't only feel bad (Seligman & Csikzentmihalyi, 2000). In fact, in one survey of psychological research, Myers and Diener (1995) determined that there were 17 publications on negative emotion for every one on positive well-being. To be sure, a complete understanding of the human experience requires an attention to the despair, loneliness, and anxiety that are sometimes a part of our lives. However, our theories of work behavior must make room for joy and love as well. Recognition of this possibility is likely to yield practical gains. For instance, several studies have found that psychological well-being is related positively to job performance (Cropanzano & Wright, 2001). Likewise, research by Moliner, Martinez-Tur, Peiro, Ramos, and Cropanzano (2003) found that positive feelings (operationalized as psychological engagement) and negative feelings (operationalized as burnout) contribute independently to the prediction of work outcomes.
The third challenge has to do with relationship between feeling and thinking. Emotion is not the antithesis of cognition. When workplace emotion was still struggling for acceptance, it might have made sense to draw sharp demarcations. Perhaps some felt that conceding an inch might open the door for a sort of intellectual imperialism. I can't speak with certainty about the past, but I do believe that we can be more open to integration in the future. Although emotions cannot be reduced to cognitions, research tells us that thinking is certainly involved. These influences are bidirectional, with affect influencing our thinking (e.g., Forgas & George, 2001; Isen & Baron, 1991) and our thoughts impacting how we feel (e.g., Reisenzein & Schoenpflug, 1992). The processes of cognition and emotion are heavily intertwined (Judge & Larsen, 2001)—so much so that it is difficult to clearly locate the boundary where one leaves off and the other begins. For instance, emotion involves affect, but it also requires that we label and interpret events (Mascolo & Griffin, 1988; Mascolo & Harkins, 1998; Pirola-Merlo, Hartel, Mann, & Hirst, 2002). Likewise, healthy human functioning depends on the close interplay of our thoughts and feelings (Goleman, 1998; Hartel et al., 2003). Judge and Larsen (2001) illustrated this point nicely by examining the case of Elliot. Elliot was a patient described by Damasio (1994). Brain damage caused Elliot to lose his capacity for affect, while retaining his cognitive capacity. This loss of feeling made Elliot less rational in his dealings with others. Among other things, he had trouble making evaluations and ranking priorities. Making a distinction between cognition and emotion is practically useful. These are big topics and taking them apart allows us to more closely scrutinize each. However, we should not lose sight of the underlying unity between these two sets of processes.
We emotion researchers have a lot to think about. However, as I have already argued the case for optimism is predicated on this road of unsolved mysteries. In the final analysis, I believe these challenges are part of an adventure. We have much to look forward to.
Read on and enjoy yourself ...
This edition was conceived and compiled to meet the need for a comprehensive book for practitioners, academics and students on the research on emotions in organizational behavior. The book is the first of its kind to incorporate organizational behavior and bounded emotionality and is the third in a series of books on emotions.
The editors' primary aim is to communicate the research presented at the biannual Emotions in Organizational Life conference to a wider audience. With the exception of the six chapters coauthored by the editors and two invited chapters, each chapter in this book embodies papers from the 2002 conference. As such, we would like to express our deep gratitude to John Basch and Cynthia Fisher of Bond University who contributed to the organization of the conference. We would also like to express our appreciation to the 29 conference paper reviewers who provided us with constructive reviews of the papers submitted to the conference. Based on these reviews, quality research from around the world was carefully selected for inclusion in this volume. Of the 40 papers received from scholars for the conference, 26 were accepted. In particular, we would like to thank the following for their time and diligence in providing quality reviews:
Yvonne A. Athanasaw
Julie Baker Sigal Barsade
Dr. Maree Boyle
Karen J. Crooker
James M. Diefendorff
Hillary A. Elfenbein
Cynthia D. Fisher
Alice C. F. Hsu
David Martinez Inigo
Anthony T. Pescosolido
Michelle K. Pizer
Lyndall M. Strazdins
The first book in this series, Emotions in the Workplace: Research, Theory, and Practice, provided a range of research regarding emotions within the work environment. The second book, Managing Emotions in the Workplace, provided research on the management of emotions. This edition, in contrast, looks at the range of research on emotions within an organizational behavior framework, organized in terms of the individual, interpersonal, and organizational levels. Particular emphasis has been placed on obtaining the leading research in the international sphere, and this book is intended to be as useful to the student of organizational behavior as to the managers of organizations.
Our choice of organizational behavior as the theme for this book reflects the growing acceptance of emotions in the mainstream of organizational studies. We believe it is the right time to update courses in the area of organizational behavior to include emotions and to bring together, for researchers and practitioners alike, the contributions emotions researchers are bringing to the understanding and management of organizational behavior.
Before taking you on this journey, we wish to thank those persons who have been especially important to bringing this compilation to you. Indeed, without their help, this edition would not have been possible. In particular, we are indebted to Debra Panipucci for her outstanding assistance in managing the administration of such an undertaking, and her invaluable contributions to the review and editing process. We are also deeply appreciative of the exceptional efforts of Victoria Strybosch throughout the review and editing process. Last but not least, we would like to thank editor Anne Duffy for believing in the value of such an endeavor. We are deeply grateful to her and the staff at Lawrence Erlbaum Associates for publishing and enabling this volume to reach people around the world.
This page intentionally left blank
Neal Ashkanasy is professor of management in the UQ Business School at the University of Queensland. He has a PhD (1989) in social and organizational psychology from the University of Queensland. His research in recent years has focused on the role of emotions in organizational life. He has published in journals such as the Academy of Management Review, Academy of Management Executive, journal of Management, and Journal of Organizational Behavior, and has edited the two earlier books on emotion. He is a past chair of the Managerial and Organizational Cognition Division of the Academy of Management.
Joyce E. Bono is an assistant professor in the Psychology Department at the University of Minnesota. She received her PhD in organizational behavior from the University of Iowa. The underlying motivation for Dr. Bono's research is discovering how individual differences and work relationships contribute to worker motivation and quality of work life, with a current focus on emotional experiences. Specifically, her recent projects examine the role of both employee personality and manager behaviors as antecedents of emotional labor. In addition to her interest in emotions as work, Dr. Bono conducts research on leadership, teaches organizational psychology, and works with both public and private organizations to improve leadership and management effectiveness.
Catherine Daus is an associate professor of psychology at Southern Illinois University Edwardsville. She obtained her PhD in 1994 from Purdue University in industrial/organizational psychology. Dr. Daus's current research focus is in the area of emotions in organizations, with particular emphasis on emotional labor, emotional intelligence, and customer service behaviors. She is also interested in diversity issues, stress and coping, and work attitudes—job satisfaction, in particular. Along with edited books, Dr. Daus has recently published in the Academy of Management Executive, Journal of Management, Journal of Quality Management, Organizational and Human Behavior Processes, and Journal of Occupational and Health Psychology.
Dorthe Eide is working as an assistant professor at the Bod0 Graduate School of Business in Norway. Currently she is finishing her doctoral dissertation on the role of emotions and social interactions in knowing and learning in organizations, proposing a broad situated-relational approach instead of seeing, knowing, and learning in practice as a cognitive process of individuals or taking place as social cognition. She has mainly been teaching and doing research within organizational studies at the college level, as well as doing some minor work within industries. Besides literature within organizational studies, she draws on philosophy, sociology, anthropology, and psychology when exploring and elaborating emotions in organizations.
Professor Charmine Hartel is Chair of Strategic Business in the Deakin Business School, Melbourne, Victoria, Australia. She received her BA in psychology with distinction from the University of Colorado and completed her master's degree and PhD in industrial and organizational psychology at Colorado State University. Her current research and consulting activities focus on emotions and patterns of relating at work; development of emotional intelligence, diversity, leadership, and team effectiveness. She is the recipient of several awards, including recognition in Who's Who, the Richard M. Suinn Commendation Award for Excellence in Research and the Advancement of Psychology, the Martin E. P. Seligman Applied Research Award, the Distinguished Leadership Award from the International Directory of Distinguished Leadership, and the Jacob E. Hautaluoma Distinguished Alumni Award. She is author of three books and over 40 refereed journal articles, which have appeared in journals
such as the Journal of Management, Academy of Management Review,Applied Psychology: An International Review, Leadership Quarterly, and Journal of Applied Psychology.
Dr. Yuka Fujimoto is currently researching diversity management issues with particular interest in the diversity effects on emotions. Her teaching focus is in the advanced human resource management (HRM). The key emerging concepts of strategic HRM are taught in this subject, and the concept of emotions is incorporated as the HR challenge in achieving business success. She has a wide range of industry experience in conducting research on emotional experience of workers. To date she has studied diversity effects on emotions across five industries in Australia, namely, banking and finance, construction, manufacturing, textile and clothing, and services, which add up to 15 multicultural organizations.
Cristina Galli is a PhD student in Cognitive Science at the University of Padova, Faculty of Psychology. She was awarded a 2-year research grant (2001-2002) from the University of Padova for the service jobs and emotional labor project. Her major research interest lies in the area of psychology of emotions, on topics such as emotions in the workplace, emotion regulation and its correlates in different work settings, concepts of emotion, and structure of the emotion lexicon in children, adolescents, and adults.
Matthew J. Grawitch is a graduate student at St. Louis University, where he is completing his PhD in organizational psychology. He received a BA and an MS in psychology from St. Louis University. His research interests focus on workplace affect at all levels within organizations, conducting applied research in both field and laboratory settings. He has published several articles examining the effect of mood on temporary workgroups, most notably with regard to group effectiveness and creativity. He also consults for public and private organizations in the areas of strategic planning and organizational assessment.
Dr. Kenneth S. Law is a professor at the Department of Management of Organizations, Hong Kong University of Science and Technology. His research interests include emotional intelligence, localization of human resources, meta-analysis, extrarole performance, and the application of structural equation modeling in behavioral research. In the past few years, he has worked on projects demonstrating the construct validity of emotional intelligence and its influences on life satisfaction and job outcomes.
Yongmei Liu is a doctoral student in the College of Business at Florida State University. Her current research interests include the social functions of emotion, and the role of emotion in team process and organizational change.
David C. Munz is professor of psychology at St. Louis University, where he is director of the psychology department's doctoral program in organizational psychology. He received a BS in psychology from the University of Cincinnati and a PhD in industrial and organizational psychology from the University of Oklahoma. His research interests include workplace affect at the individual, group, and organizational levels and the design and evaluation of workplace interventions at each of these levels. He has published or presented over 100 scientific articles and chapters on topics such as occupational stress, mood in workgroups, and affectivity's role in organizational assessment. He is a fellow of the American Psychological Association (APA) and a charter member of the American Psychological Society.
Debra Panipucci is a doctoral student in the Centre for Business Research, Deakin Business School. Her research interests include social influences and diversity in teams, with particular emphasis on the effects of perceived dissimilarity on intrateam behavior and the individual team member outcomes and feelings that result.
Pamela L. Perrewe is the Jim Moran Professor of Management in the College of Business at Florida State University. She received her bachelor's degree in psychology from Purdue University and her master's and PhD degrees in management from the University of Nebraska. Dr. Perrewe teaches a doctoral seminar in organizational behavior with an emphasis on emotions. She has focused her research interest in the areas of emotions, job stress, and personality. Dr. Perrewe has published over 70 book chapters and journal articles in outlets such as the Academy of Management Journal, Journal of Management, Journal of Applied Psychology, and Journal of Occupational Health Psychology.
Anthony T. Pescosolido is an assistant professor in the Department of Management within the Whittemore School of Business and Economics at the University of New Hampshire. His research interests include emergent leadership, the impact of emotions and emotional expression on group processes and productivity, and the group processes that lead to long-term group effectiveness. He has taught courses on organizational behavior, teamwork, and career development, and has facilitated workshops on emotional intelligence, teamwork, and leadership within a variety of manufacturing and service firms. His research has appeared in Human Relations, Leadership Quarterly, and Small Group Research.
Michelle K. Pizer is a graduate student in the Centre for Business Research, Deakin Business School, where she is completing her PhD in management. Her research focus combines an interest in the emotional experience of individuals with the world of work, directly reflecting her current work as a registered psychologist in private practice providing psychotherapy with a psychodynamic orientation and her previous management experience. She received a MBus from Swinburne University, where she conducted a case study investigating the differential experience of an organization's culture based on career success within the firm. Currently she is exploring, more specifically, the relationship between organizational culture and workplace emotions.
Robert S. Rubin is an assistant professor of management at DePaul University. He received his PhD in organizational psychology from St. Louis University and holds an MA in industrial/organizational psychology from Southern Illinois University Edwardsville. His research and teaching interests span human resources management and organizational behavior, including transformational leadership, managerial development, academic assessment centers, and emotions at work. His published work appears in outlets such as Human Resource Management Journal and Leadership Quarterly. In addition, Dr. Rubin has been a Human Resource/Organizational Development (HR/OD) consultant to a variety of industries including biotechnology, healthcare, and transportation.
Vicki M. Staebler Tardino, MA, is an organization development practitioner and, more recently, a doctoral candidate in organizational psychology at St. Louis University. For over 10 years she has consulted in internal and external capacities in business, academia, and government. In her work with organizations, emotion plays a role in many leadership and team development interventions. Her master's thesis examined perceptions of personal, social, and work-related consequences of emotional expression at work. Her more recent research interest is emotional labor. She has served on the Executive Committee of Gateway Industrial/Organizational Psychologists and is on the Board of the St. Louis Organization Development Network.
Meredith A. Vey earned her PhD in industrial and organizational psychology from the University of Minnesota. In addition to exploring the structure of work performance, Dr. Vey's research has focused on the motivational and affective determinants of performance, particularly in the areas of citizenship behavior and emotional labor. Her research interests also include human capital development and employee engagement as determinants of organizational performance. Dr. Vey is a research fellow at the Accenture Institute for High Performance Business in Cambridge, MA.
Mona White is a doctoral student in the Centre for Business Research, Deakin Business School. Her current research interests include the areas of business communication and negotiations, in particular, the differences and similarities in mass and personal communication within culture and cross-cultural negotiants between Australian and Chinese nationals in a business context. Having lived and worked in both China and Australia, her understanding and insight into each country's culture and people are invaluable to clients operating in these countries. She understands the types of problems and issues confronted by Australian companies entering the Chinese markets and the problems and issues confronted by Chinese companies entering the Australian market.
Dr. Chi-Sum Wong is a professor at the Department of Management, Chinese University of Hong Kong. His research interests include emotional intelligence, localization of human resources, career interests, job design, and application of structural equation modeling in behavioral research. In the past few years, he has worked on projects demonstrating the construct validity of emotional intelligence and its influences on life satisfaction and job outcomes.
Dr. Ping-Man Wong is the senior lecturer and head (2000–2003) of the Department of Educational Policy and Administration of the Hong Kong Institute of Education. As a school manager and former secondary school principal, Dr.
Wong is program leader of various leadership training programs for local principals, teachers, and students. He also has experiences in coordinating and training Sri Lanka principals and school inspectors of Cambodia in programs sponsored by the World Bank and the CAMSET. His other recent outputs related to emotion studies include a leadership training program for primary school teachers and pupils—an EQ perspective (2001), and emotional intelligence attitudes toward lives and attainment of educational goals (2001).
Xin Yao is a PhD candidate in organizational behavior/human resource management at the University of Washington Business School. Her research interests include emotional labor, impression management, humor, self theories, and social dilemmas. Her dissertation examines the psychological process of performing emotional labor in a customer service context. A lab study is adopted to investigate how external humorous stimuli and the way in which one carries out emotional labor may affect the extent to which one becomes psychologically exhausted. In addition, her curiosity about humor leads to a theoretical probing of its genesis and development during emotion sharing at the workplace.
Vanda L. Zammuner is a professor at the Psychology Faculty of the University of Padova (Italy) (Dipartimento di Psicologia dello Sviluppo e della Socializzazione). From 1978, she has given seminars, lecture cycles, and courses on topics related to social psychology. Currently she is giving courses on techniques of questionnaires and interviewing, and on psychology of emotions. She is a member of the Cognitive Science Doctoral Program at the University of Padova and of various national and international research societies, such as the European Association of Experimental Social Psychology (EAESP), the Psychonomic Society, International Society for Research on Emotions (ISRE; in the directive board 1994–1998). She has taught at and conducted research seminars at several European universities, including Bern, Amsterdam, Bielefeld, and Umea. She is the author of about 100 publications and presentations at both national and international meetings. Her research, often interdisciplinary in nature, is related to several sectors, including emotions, studied from different perspectives, and methodological issues, with special reference to cognitive processes implied in data collection methods. She regularly acts as referee for various international journals, such as Cognition and Emotion and European Journal of Experimental Psychology. Finally, she has organized international and national symposia, meetings, and conferences.
Wilfred J. Zerbe is associate dean (MBA Program) and professor of human resources and organizational dynamics in the Haskayne School of Business at the University of Calgary. His research interests focus on emotions in organizations, organizational research methods, service-sector management, and
leadership. His publications have appeared in books and journals including the
Academy of Management Review, Industrial and Labour Relations Review, Canadian Journal of Administrative Sciences, Journal of Business Research, Journal of Psychology, Journal of Services Marketing, and Journal of Research in Higher Education. He is also an active consultant and executive educator.
This page intentionally left blank
The winner of the inaugural Best Paper award at the first Emotions in Organizational Life Conference (EMONET) and valued colleague.
Dr. Jan Paterson, friend, colleague, mother, and partner, died 31 December 2002. She is survived by her husband of 35 years, Errol Paterson, one son, and numerous friends and admirers.
Jan's research passion was furthering the understanding and practice of change management processes and the tragic effects of such processes when the teachings of our field are ignored.
Jan's career in organizational psychology crossed over from a practitioner role in New Zealand to an academic in the psychology department at the University of Queensland before returning to New Zealand.
In amazing testimony of her commitment to the field, Jan produced three journal publications and one book chapter in the 7 months between the date she was diagnosed with cancer and her death. Jan continued to work as a re-engineering officer with Health Waikato right up to the date of her admittance to the hospital with a brain hemorrhage. During that time she worked closely with the director of nursing in an effort to assist the Waikato Nursing Staff with many major work changes. Although she deteriorated too much to complete the process, her input was accepted and put into practice with excellent results.
Jan was an individual of the highest integrity, a person who not only passionately valued social justice but also actively worked to create it. The quality of her character is captured well by one of the events held in celebration of her life. Jan was one of three trustees of the Noa Foundation, which provides the Maori people with business assistance. To acknowledge her assistance to the New Zealand Maori's, the Tainui Tribe sent a Whaka—a large war canoe containing about 50 warriors—20 km from Ngaruawahia to a point 500 meters upstream of the family house, where they turned and paddled down the Waikato River. As they approached her house on the down-river run, they raised their paddles in salute to her memory, keeping them raised until they passed the property. This was done with full ceremony, including a female elder standing on the bank chanting in Maori calling for Jan's spirit to witness the event. The scene was very moving, with a large crowd of neighbors, friends, and associates from the Maori community sharing in the event. The honor given was great, a recognition of nobility of character, as such salutes normally are reserved only for tribal chiefs.
Jan—We, your friends and colleagues, also raise our paddles to you, in memory of a person whose life is a legacy to what it means to live and work with integrity, generosity, and concern for a positive society.
—Charmine E. F. Hartel
This page intentionally left blank
Charmine E. J. Hartel
Wilfred J. Zerbe
Neal M. Ashkanasy
All organizations make at least some decisions that hurt their employees and cause some level of distress or disruption. For instance, jobs may need to be cut or high performance levels may be demanded. Without recognition of emotions, however, organizations cannot deal effectively and humanely with the pain caused (Frost, 2003). Organizations that understand this and attend to keeping their environments emotionally healthy will create less emotional pain and reap more productive work behavior. Moreover, organizations that have this understanding, and also value having a culture that is emotionally constructive, can create organizational behavior that contributes to employees' emotional health.
This opening chapter provides an overview of organizational behavior theory and research and the paradigms that have dominated the field to date. Running through a discussion of rational notions of organizational behavior, to concepts of bounded rationality and most recently the call for bounded emotionality perspectives, we identify for the reader what a bounded emotionality perspective adds to the understanding of organizations. We then provide an overview of the remaining chapters in the book and how they contribute to the book's objectives.
ORGANIZATIONAL BEHAVIOR: AN OVERVIEW
Organizational behavior is an ever-expanding, dynamic field with a variety of theories and models used to describe how and why people interact differently within the workplace. The aim of this book is to provide the reader with a realtime approach to understanding organizational behavior with a particular focus on emotions, to increase understanding and awareness of the role of emotions within the workplace.
For the most part, organizational research tends to rest on the assumptions that human beings are rational creatures and that emotions are irrational and unproductive (Styhre, Ingelgard, Beausang, Castenfors, Mulec, & Roth, 2002). Therefore, to act in emotional or unpredictable ways is "unacceptable" (Antonacopoulou & Gabriel, 2001). Subsequently, emotion is excluded from the majority of research. Prior to 1992, the study of emotions was, with a few notable exceptions, essentially next to nonexistent in the management literature. Textbooks in the field, for example, provided no coverage of the role of emotions in work settings, except with respect to job satisfaction—factors that have since been shown to be largely cognitive (e.g., see Meyer & Allen, 1991; Wright & Staw, 1999). This began to change with the publication of Pekrun and Frese's (1992) chapter titled "Emotions in Work and Achievement," followed in 1993 by publication of Fineman's edited book Emotion in Organizations (1993a).
The momentum rapidly accelerated in the years following 1993, with particular intensity following Ashforth and Humphrey's (1995) call to action. By 2000, this culminated in a veritable explosion of literature. Edited books by Ashkanasy, Hartel, and Zerbe (2000), Ashkanasy, Zerbe, and Hartel (2002), Fineman (2000), Lord, Klimoski, and Kanfer (2002), and Payne and Cooper (2001) emerged, and special issues of journals guest edited by Fisher and Ashkanasy (2000b), Fox (2000), Humphrey (2002), and Weiss (2001, 2002) were requisitioned. This proliferation of literature was recognized in 2002 by the editors of the Annual Review of Psychology, who invited their first review of this literature (Brief & Weiss, 2002). The present volume is the latest addition to this upsurge of interest.
FROM THE COGNITIVE REVOLUTION
TO THE EMOTIONAL REVOLUTION
In 1976, Simon founded the concept of bounded rationality, arguing that rational economic theory was flawed because it failed to take proper account of the limitations of human rational thought. A core part of this paradigm was the definition of "irrational" and "arational" decisions, with emotion in the former category. Essentially, Simon's work marked the beginning of the cognitive revolution in economics and management theory. As such, emotional responses were treated as weaker than those based on reason.
In direct opposition, Mumby and Putnam (1993) introduced the term bounded emotionality, intended to be a foil to Simon's (1976) concept of
1. ORGANIZATIONA L BEHAVIOR : AN EMOTION S PERSPECTIV E 3
"bounded rationality." Specifically, Mumby and Putnam reexamined bounded rationality and concluded that it failed to take sufficient account of the power of emotion. They suggested that "emotions constitute a way of knowing that differs from but complements traditional rationality" (p. 480). Thus, in effect, cognitive theory is flawed because it dismisses emotion as "irrational" and therefore beyond the pale of systematic analysis. As such, bounded emotionality formed "an alternative mode of organizing in which nurturance, caring, community, supportiveness, and interrelatedness are fused with individual responsibility to shape organizational experiences" (Mumby & Putnam, 1993, p. 474). It extends the idea of bounded rationality (and cognitive theory) to take in the idea that emotions can be analyzed systematically and should be included in models of organizational behavior and decision making.
Thus, just as bounded rationality signaled the beginning of the cognitive revolution (Ilgen, Major, & Tower, 1994), bounded emotionality marked the launch of the emotional revolution (Barsade, Brief, & Spataro, 2003).
Although emotion has always been a critical factor in organizational management (Mastenbroek, 2000), it was not until the emergence in the 1980s of critical management theory that the topic finally bubbled to the surface of management scholarship. Indeed, the idea of bounded emotionality emerged from the critical feminist perspective that, in turn, was presaged by Van Maanen and Kunda's (1989) seminal description of the organizational process of emotion management as well as Fineman's (1993a) seminal book in the area.
The emotional revolution was also reflected in the increasing interest in the study of emotions in the wider psychological and popular literature. The American Psychological Association launched a new journal, Emotion, in 2001, and books on emotions by Damasio (2003) and Goleman (1995) reached the best-seller lists. Ten years after Fineman's book, Barsade and her colleagues (2003) announced that the revolution was, indeed, a "paradigm shift."
AN EMOTIONS PERSPECTIVE
OF ORGANIZATIONAL BEHAVIOR
Emotions are linked to organizational behavior in a number of ways. Generally speaking, when employees are unhappy, it shows in organizational behavior, because unhappy people are disconnected from their work (Frost, 2003, p. 33). Failure to understand the emotional aspects of organizational behavior means that an organization is unlikely to realize its toxic actions such as unreasonable company policies, disruptive coworkers or clients, abusive managers, and poorly managed change (Frost, 2003, p. 14). Further, they are unlikely to recognize and support the efforts of those trying to change and deal with the con
sequences of the toxic behavior, and thus are unlikely to retain those who can support a healthy organizational culture. Taking an emotions perspective of organizational behavior, on the other hand, can help leaders increase the happiness of their workers and assist employees to achieve positive outcomes for themselves even when painful events are encountered. Not only does implementing an emotions perspective of organizational behavior provide bottom-line benefits to the organization, it holds the promise of making it more ethical and socially responsible, as heartfelt caring for the needs of others has been linked to moral acts (Josselson, 1992).
Specifically, organizational behavior is a study of individuals and groups in organizations, and the ways in which they display attitude, actions, and behaviors (Weiss, 1996). In order to understand these attitudes, actions, and behaviors, emotion-based theories emerged, such as emotional intelligence, affective events theory, and emotional labor, as discussed in various chapters within this book. Such integration of emotions and organizational behavior assists the development of understanding, implications, and solutions of managing emotions within the workplace.
As can be seen, organizations are complex entities. Many revolutions, theories, and models have eventuated in an attempt to better describe organizational behavior. For the most part, research has tended to be based on assumptions of rationality, excluding notions of emotionality. However, a reliance on one dimension of work experience, such as rationality, overlooks alternative forms of organizational experience. "By shifting rationality to include intersubjective understanding, community, and shared interests, insights into alternative forms of organizing are created" (Mumby & Putnam, 1993, p. 480).
In this volume, we present a broad range of the latest research in emotions in organizational behavior. In particular, the latest research is gathered from Australia, France, China, Sweden, the United States, and Italy. Further, the book introduces new theories, models, and recent advances in emotions, such as Fujimoto and Hartel's model of the types of emotions arising in interactions among culturally diverse workteams and how these are shaped by organizational policies and practices.
FRAMEWORK OF THE BOOK
The book is segmented into five parts and comprises a total of 20 chapters. Part I of Emotions in Organizational Behavior introduces the reader to the concept of emotions. Parts II to IV discuss the concept of emotions on three levels of analysis: individual, interpersonal, and organizational. Finally, as organizational behavior is dynamic and ever-changing, part V discusses the role of emotions in organizational change.
1. ORGANIZATIONA L BEHAVIOR: AN EMOTIONS PERSPECTIVE 5
Part I: Organizational Behavior and Emotions
Part I is a theoretical piece that illustrates the important role emotions play for individuals and groups and suggests that organizational behavior is enhanced through the careful management of emotions, knowledge, and learning. This chapter looks at why we are emotional through a discussion of the psychological development of emotions and an integration of this into the role of emotions in the workplace. The discussion is particularly useful for students and practitioners as it provides a sound basis for their knowledge in this field and leads into the subsequent chapters, which build on this knowledge basis.
Part II: The Individual Within the Organization
We then turn to the individual within the organization and how they manage their emotions within the workplace. Chapters 3 to 5 comprise empirical studies and build on part I by incorporating situational factors and the potential conflict arising from work roles. Together, these chapters demonstrate the importance of emotions to individual behavior and work experiences, providing insight on the organizational behavior strategies that can enhance behavior and experience.
In particular, chapter 3 discusses the emotional conflict that service providers experience when performing emotional process work. It details the link between coping with these emotions and the conflicting gender roles that societal rules impose upon us. Chapter 4 addresses the implications of emotion in employee counterproductive work behavior. Specifically, it discusses the effects of situational factors in the role of managing behavior within the workplace. It demonstrates that, if managed inappropriately, counterproductive work behavior results, which can have a detrimental effect on both the employee and the organizational setting.
Chapter 5 demonstrates that perceptual and attitudinal processes occur when diverse individuals interact, based on observable attributes. The individuals within the interaction have attitudes about these perceptions and respond in an affective way. The chapter provides important new insights on interactions between dissimilar persons and forwards a model of the affective response individuals have toward dissimilar team members that will be useful to both practitioners and researchers alike.
The final chapter in this section, written by the editors, integrates the contributions of each chapter with other state-of-the-art research to provide a bounded emotionality perspective on the individual in the organization.
Part III: The Interpersonal Within the Organization
Part III presents a combination of empirical and theoretical research on workgroup dynamics, communication and emotions. This section is particularly pertinent for practitioners who seek further understanding and new strategies for dealing with organizational change.
Chapter 7 presents a theoretical model incorporating group affect into problem-solving activities. Specifically, it reviews the literature surrounding individual and group level affect in workgroups and provides a detailed explanation of each stage within the model: problem identification; brainstorming; solution development; and implementation. It suggests that the feelings of individuals within the group and the collective feelings of the group are influenced by the problem-solving process. Further, as groups mature, their level of affect is shown to become more influential over individual affect. As such, group level affect is influenced by, and will in turn, influence individual members' affect.
In order for organizational behavior to be effective, students, practitioners, and scholars must be aware of the importance of different identities and negotiation skills. Chapter 8 addresses this need, presenting a picture of intergroup dynamics and the barriers to effective negotiation. Drawing on affective events theory (AET) and communication accommodation theory (CAT), the chapter elucidates cross-cultural communication and negotiation and demonstrates application of the model using a Chinese-Australian negotiation scenario.
Chapter 9 discusses the impact of communication on emotion by addressing the role of humor in emotional workplace events. Specifically, it discusses the social sharing of emotion (SSE) process and the impact that humor has on the workplace at both the primary and secondary stages of this process. The chapter suggests that humor can be used in the "storytelling" of emotional events to allow individuals to cope with the initial negative aspect of the event, in both a social and a psychological manner.
The final chapter in this section, written by the editors, integrates the contributions of the other chapters in the section with other state-of-the-art research to provide a bounded emotionality perspective on interpersonal behavior in organizations.
Part IV: Organizational Processes, Structure, and Design
Part IV focuses on organizational processes, structure, and design, and on how the organization impacts individual and group emotions. The chapters included in this section integrate job design with individual differences, such as personality and motivation. In particular, chapters 11 and 12 are theoretical pieces aimed to develop an understanding of the emotional dimension of job design and the impact it has on employees. These are followed by chapters 13 and 14, which empirically test the cause and effects of emotional demands on employees. Each of these chapters argues that individual differences such as emotional intelligence and personality and job characteristics such as level of autonomy, complexity, and responsibility impact on the emotional demands of
1. ORGANIZATIONA L BEHAVIOR : AN EMOTION S PERSPECTIV E 7
employees and their responses. In addition, they suggest that a variety of outcomes of emotional labor are possible, in particular, emotional exhaustion, depersonalization, personal accomplishment, physical complaints, role internalization, self-esteem, and job satisfaction.
Written by the editors, the final chapter in this section integrates the contributions of chapters 11 to 14 with other research to provide a bounded emotionality perspective on work characteristics.
Together, the chapters in this section introduce the reader to the demanding aspects of "emotional" jobs, that is, jobs that are designed to regulate emotional displays, the demands associated with emotional regulation itself, and behavioral responses to emotional regulation. The section also highlights how behavior in organizations is influenced to a large extent by organizational rules and norms, which both work through emotional responses and dictate the emotional characteristics of jobs.
Part V: Organizational Change and Changing Organizations
Part V concludes the book. It begins with a theoretical chapter on the notion of emotions and organizational change, which illustrates that in order to achieve effective continuous change, organizations need to develop emotional capabilities. They must understand the emotional consequences of their change actions and use these emotions to advance the change. Chapter 17 follows and proposes a new role for group leaders, the role of managing group emotion. The chapter shows that individual group members are perceived as leaders by others when they provide direction and guidance during times of ambiguity, particularly when this guidance involves the display of appropriate emotional responses to unsettling events. Such leadership resolves ambiguity, provides the group with the direction needed for action, and can increase group solidarity by creating both shared emotion and shared action within the group. The chapter provides important new insights on leadership, showing that the leadership in the management of group emotion is a leadership action that can be performed by different group members at different times.
Chapter 18 takes an emotions perspective of organizational culture, illustrating how it must be managed to gain employee commitment while avoiding a destructive, cultlike following. It introduces the concept of a healthy organizational culture, in terms of an organizational environment that induces positive emotions and reduces negative emotions. It stresses the importance of organizational culture on employees' work experiences and how it can influence them for better or for worse. Further, it suggests that the elements of culture can be shaped by the organization to enable social control. As such, organizations must be careful to construct a culture avoiding the destructive and
cultlike nature of commitment while enhancing the positive nature of commitment to the organization.
Chapter 19 provides a bounded emotionality perspective on organizational change and culture, and chapter 20 provides a conclusion to the book, drawing on the contributions of the book itself to present a comprehensive analysis of what an emotions perspective of organizational behavior offers.
Each chapter in this volume is designed to enhance practitioner knowledge and student interest in and understanding of how organizational behavior concepts translate into workplaces. The principles underlying the compilation of research in this book are both empirically and theoretically based to cover the scope of first-class research that has been performed by specialist scholars from across the world. Consequently, Emotions in Organizational Behavior is designed as a valuable learning facilitator for students and as a means for scholars and practitioners to incorporate the latest research into their knowledge of organizational behavior.
This page intentionally left blank
The ability to cope and act in capable ways with and through emotions has become vital as more people and organizations become involved in work intensive on social interactions, emotional labor, and/or changes. However, linking emotions to capable activity challenges traditional views on organizations and knowledge. This chapter argues that the shift from seeing emotions as an "ugly duckling" to an "invisible asset" is not sufficient in order to understand emotions in organizations; a more radical rethinking is needed. A situated-relational ontology is suggested where emotions are seen as one of the fundamental and inseparable parts of being human, and consequently also of human knowledge and action in organizations.
EMOTIONS AND KNOWLEDGE—
Knowledge and learning processes have gained increased legitimacy in society and organizations. They are claimed to be critical "invisible assets" and core processes in order to survive, innovate, and increase competitiveness (Drucker, 1993, 2001; Nonaka & Takeuchi, 1995; Quinn, 1992). As this claim becomes stronger, it becomes important—both for functional and for symbolic reasons—to explore critically what it means to know and work in organizations. If not, there is a high risk that we will only appreciate the kind of work and knowledge already understood, and thereby silence other voices of experience, knowledge, and opinions in organizational life and studies.
During the last decade the calls for rethinking emotions and knowledge increased, and at least two main alternative research schools have been identified. One view centers on knowledge processes in organizations where it is argued that knowledge is a critical resource, but where the idealized theoretical view of knowledge is argued to be out of touch with the meaning of knowledge in practice. The other school of thought within research, emotions in organizations, describes how emotions are present and constitute a vital resource in work. It also criticizes the traditional "rational" view for taking an overly dualistic view of cognition and emotions, as well as assuming that cognition tends to be rational whereas emotions tend not to be (e.g., Domagalski, 1999; Fineman, 1993a). There are examples where these two schools of thought overlap—for instance, both tacit knowledge and emotions are argued as invisible assets and/or human ways of knowing and being.
Rethinking Knowledge in Organizational Studies
The traditional view of emotions and knowledge has a common "root" in a
positivistic/realism (i.e., functionalistic, to use the term of Burrell & Morgan,
1979) paradigm that depicts emotions as the "sand in the machinery" or, in
other words, that emotions prohibit workplace effectiveness. Therefore emo
tions and knowledge came to be seen as two dualistic and incompatible phe
nomena. Scholars working within an interpretive-constructionist paradigm
have claimed that traditional perspectives on human knowledge are insuffi
cient (i.e., where knowledge is seen as only or mainly something explicit, "ob
jective," general, theoretical, and often technical). Instead, versatile concepts
of knowledge have been suggested (Bru
ner, 1990; Gherardi, 1999; Lave & Weng-The life—
and the love of everything—
er, 1991; Polanyi, 1958; Schon, 1983) that
you must feel your way toward
involve dimensions such as the personal,
—with the rough skin of your
relational, and social dimensions; the
hands—like a blind person learns tacit, narrative, and explicit dimensions; the face of her lover and the local embedded and more gen-through her fingertips. eral aspects. During the last decade, or-(Hans Borli, 1972, my translation) ganizational studies have recognized and focused on knowledge in practice, particularly through the subphenomena of tacit knowledge and narrative knowledge. Tacit knowledge means knowledge that is difficult to articulate fully in words; that is, most of what we know is tacit, and even the more explicit knowledge dwells in the tacit (see, e.g., Polanyi, 1966/1983). Narrative knowledge is largely discursive and expressive in words; it is a story-based way of understanding and makes meaning out of ongoing experiences in order to know oneself and the world (Bruner, 1990; Czarniawska, 1997; Polkinghorne, 1988). Both the tacit and the narrative/discursive are core processes in how people understand in everyday life, and are therefore central to organizational behavior and the concept of knowledge.
2. EMOTIONS: TOWARD AN ONTOLOGICAL REFRAMIN G 13
Even though our understanding of knowledge has developed and enlarged, science still seems to have problems with bringing emotions "back in" to organizational studies, particularly with regard to knowledge and processes related to intelligent/capable activity. This is a field that appears to be among the most dominated by cognition and/or traditional instrumental rationality. Is this the last "bastion" where emotions are the "ugly duckling"? When it comes to learning in organizations, emotions have recently been claimed to be the most promising and unexplored dimension (Hopfl & Linstead, 1997). Is this the case for knowledge in organizations as well? Yes, I will argue; the next vital step is to include emotions in studies of learning and knowledge in organizations.
Toward Integrating and Rethinking Emotions
Ashkanasy (1995, p. 2) argued that it is time to place "Cinderella" in the limelight and to include emotions in order to make organizational studies "more complete." In present working life it has become more apparent that human actions are often emotional because a growing number of occupations and trades are characterized by emotionally intensive work, especially due to increasing focus on service quality, customer orientation and rapid changes (Fineman, 1993a, 2000; Van Maanen & Kunda, 1989). Coping with one's own and others' emotions seems vital, but can also be difficult. The emotional dimension can be a smaller part of the service "product" or it can be a critical part of the core process of customer service (Forseth, 2001). However, indicating that emotions can be related positively to knowledge does not resonate well with the traditional and "rational" views on knowledge and organizations.
The main purpose of this chapter is to contribute in the rethinking discourse of emotions and knowledge. My argument is twofold: First, in order to grasp human emotions and knowing in practice, we need to engage in a more radical rethinking that includes an ontological choice. Here a broad situated-relational approach (within the interpretive-constructionist paradigm) is suggested. Second, stemming from this ontology is the basic assumption that being human involves emotions; therefore, human knowledge can involve emotions as well. By arguing that emotions have a fundamental role, I do not mean that other human aspects (e.g., cognition) should be omitted. My point is not to throw out one for the other but rather to aim for a broader and more integrated view on human being and human activity.
This chapter is divided into four sections: First, I briefly review traditional views of emotions in society and organizations. Second, more recent views on emotions in organizations and work are addressed, particularly the view of emotions as an invisible asset in emotionally intensive work. Third, an ontological reframing of emotions and knowledge is suggested and elaborated. Finally, implications are summarized.
EMOTIONS AS THE "UGLY DUCKLING"1
Emotions have been viewed as the "ugly duckling" within the triangular of three distinct faculties of affect/emotion, cognition (how people think, know, reason), and will (conation, motivation) (Forgas, 2000). During the Victorian period (about 1820-1920) the prevailing view developing was that emotions needed to be structured and controlled (May, 1983), whereas reason and science became the new promising force for industry and governmental activity. Since this first phase of industrialization, Western cultures and especially science have mainly celebrated so-called instrumental, rational, and/or cognitive views on humans and organizations, such as the machine view of organizations (see Morgan, 1986). Instead of the religious dogma in the Middle Ages, with one almighty God and truth, science became the new "hero" in the search for the truth. Personal aspects such as feelings and intuition, as well as other types of values and rationality (e.g., care and substantial rationality,2 practical sense/ judging), were ignored, suppressed, or in other ways treated as "sand in the machinery." Emotions were seen as something that characterized primitive creatures and cultures, including children, women, or artists, and therefore became as the antithesis of both scientific work and of knowledge (Bendelow & Williams, 1998; Carnall, 1995; Morgan, 1986).
Existential phenomenology replied by asking, "But what sort of world do we dwell in?" (Luijpen, 1962, p. 88). Human philosophers (e.g., Heidegger, 1927/1996; James, as cited in Forgas, 2000) recognized a close relationship in thinking, feeling, and behavior, and started to question and protest against the rationalism of their time (late part of the 1800s and beginning of 1900s). Something vital was missing and had gone wrong when natural science ruled the "playground" alone. Not even psychology assigned much importance to emotions in the two most dominant paradigms of behaviorism and cognitivism (Forgas, 2000).
Within organizational studies, Weick (1979) and Fineman (1993a) among others argued against the "machine" perspective of organizations and human beings. Bruner (1990) claimed that cognitivism still suffers from insufficiency due to the machine metaphor assumed when discussing learning and knowledge. Nonaka (1994) argued that the rational and hierarchical view of organizations and knowledge hinders the understanding and facilitation of knowledge creation and innovation. However, this "rational" view is not purely historical. Ashkanasy (1995, pp. 1-2) argued that research still seems to be "influenced by the Weberian belief that emotions and feelings are not proper
The term the ugly duckling is inspired by the fairy tale with the same name by the Danish poet
H. C. Andersen.
For example, care rationality has often been seen as female and emotional, opposed to the instrumental technical/economical rationality, which was seen as a male rationality and where emotions were excluded (Martinsen, 1989).
2. EMOTIONS : TOWARD AN ONTOLOGICA L REFRAMIN G 15
subjects for serious study." Today an increasing number of researchers have become aware of the importance of including emotions and more implicit aspects when studying organizational life and capable action/knowledge. One may argue that it is high time that management and organizational scholars start including the everyday emotional life in their studies and theories (Ashforth & Humphrey, 1995), and then studying the many sides of emotions and We are left with an image of an not only seeing them as irrational. actor who thinks a lot, plans, plots
and struggles to look the right part
In summary, I argue that the view of
at the right time. But we do not
cognition as something more important
hear this actor's anger, pain, em-than, and as something separable from, barrassment, disaffection or passion
other human aspects and meanings is and how such feeling relates to ac
definitely being challenged. The ques-tions — except when it forms part of the organizational script.
tion is, however, are we moving toward
(Fineman, 1993a, p. 14)
more fruitful alternatives?
EMOTIONS IN ORGANIZATIONS—
AN INVISIBLE ASSET?
In more recent writings on emotions in organizations, different perspectives of emotions are taken. The two most common and complementary conceptual views are the psychoanalytic and the social constructional perspectives (Fineman, 1997). The former sees emotions as mainly an inner individual and private process (see Calori, 1998), whereas the latter argues that emotions are mainly something relational, social, cultural, and thereby also public (e.g., Gergen, 1994; Sandelands & Boudens, 2000; Strati, 1998). Both views can study what, why, and how feelings are expressed and repressed. For example, this can be carried out in a study of the differences and similarities between groups (e.g., the front line vs. the back line). In addition to, or partly overlapping with, the two views, there is a rapidly growing school of thought—emotional intelligence—that addresses the role of emotions in capable action and in learning (see Goleman, 1998; Mayer & Salovey, 1997).
This section does not address the main recent views in depth; instead, I focus on emotionally intensive work and studies thereof, which tend to draw on one or more of the three lines. My purpose is to show problems with addressing emotions mainly as an invisible asset to be utilized and managed in organizations in instrumental ways. In this way, the examples can be seen as arguments against stopping the rethinking of emotions, with the idea that emotions are an invisible asset. In the third section I suggest and elaborate a third and more radical rethinking. However, first, what is emotionally intensive work and how can it be problematic?
Emotionally Intensive Work
Emotions have received increasing atten-
People working in customer service
tion and importance in organizations be-
roles find their employers specifying cause a large, if not the largest, part of the how they act and dress, what they workforce in Western countries work say and even what they should think and feel...
within service industries or service work.
Service staffs are paid as much for
The number of employees who are paid
their "emotional labor" as for their
to express positive emotions and atti
tudes such as commitment, sensitivity, (Guerrier, 1999, pp. 212 and 234) care, and hospitality (e.g., through smiling, greetings such as "welcome" and "have a nice day"), while not expressing negative emotions or other feelings and identities, are increasing. Such work is also termed emotional labor and involves emotional management, as one is supposed to control and manage one's feelings so they are appropriate when "on stage" with customers. Hospitality and frontline work in particular are often described as intensive regarding emotions; "the word 'hospitality' conjures up images of warm, smiling welcomes" (Guerrier, 1999, p. 211). In this type of work, emotions are argued an "invisible asset," that is, an unrecognized, intangible and/or central resource in the organizations:
We can think of emotion as a covert resource, like money, or knowledge, or physical labor, which companies need to get the job done. Real-time emotions are a large part of what managers manage and emotional labor is no small part of what trainers' train and supervisors supervise. It is a big part of white-collar "work." (Hochschild, 1993, p. xii)
Rather than biasing and neglecting emotions, I argue that emotions are more often seen as an invisible asset for the individual and particularly the organization both in practice and in the more recent theoretical schools of thought briefly introduced earlier. One main reason for this change in the view on emotions (from "ugly duckling" to "invisible asset") seems to be the global ideas of service management, quality, and customer orientation, and the necessity of these in successful business transactions. Such ideas are not only cognitive (i.e., mental models and structures, thinking, and information processing), they are also embedded with emotions, meanings, and values that direct and shape the internal and external activities of the involved organizational members. The service itself is highly intangible, displayed by and creating feelings and symbols. Learning and maintaining appropriate emotions through contextual situated "feeling rules" have therefore been argued to be a central component in service work, both in so-called low-skilled work (e.g., fast-food restaurants) and in occupations acquired through university degrees (e.g., doctors, teachers, and consultants).
This is a significant question, which is addressed and discussed in this chapter, and which is a main reason for continuing the rethinking of emotions. The movement in organizational theory and practice from viewing emotion as "ugly duckling" toward emotions as "invisible assets" involves significant ethical dilemma and work environment problems, which indicates that this approach to emotions has problematic limitations. I next show empirical examples of problematic aspects when emotions mainly are approached as an invisible asset to manage and utilize in the workplace.
Consequences of Emotionally Intensive Work
There are many consequences of emotionally intensive work. Based on recent empirical studies of service work, I show and discuss four arguments against viewing emotions mainly as something easily manageable or purely concerned with surface acting.
First, in-depth studies within the hospitality industry (Eide & Lindberg, 1997) show that customers can see through surface acting and subsequently experience the customer interaction asfalse and untrustworthy. Hotel customers, for example, describe such interactions as negative, especially if the conference host or receptionist continues to smile, instead of really being able and willing to help:
"I, and four other guys were just about to check into the hotel. Two girls were in
the reception. One of these girls put on a false 'pro' attitude which was perceived
by all of us as a very arrogant attitude, and it was very irritating."
"When I was about to arrange the course, the receptionist showed little ac
commodating attitude. I just got a feeling of insecurity towards the accomplish
ment of the whole course. I simply had to talk to the manager, and he provided
me with a person with a more accommodating attitude." (Eide & Lindberg, 1997,
Managers and various occupational groups of employees in hotels also point out the importance of being natural during interactions. For example, an illustration of such an argument told to us by a receptionist was, "To be service minded is exaggerated. One has to be natural. If you like to work with people, it will come naturally" (Eide & Lindberg, 1997, p. 5). We found in our study (Eide & Lindberg, 1997) that being able to help and interact with customers face to face is often complex and involves doing far more than just smiling or repeating simple phrases like "have a nice day" at the right time, as has also been found in Australian hotels (Faulkner & Patiar, 1997). Service or other relational work, therefore, is neither completely manageable by oneself or others, nor is it primarily a theatric surface performance as indicated by Goffmann (1959) and Pine and Gilmore (1998). Customers may distrust surface acting as they are able to "read between the lines," and customer interactions are ba
2. EMOTIONS : TOWARD AN ONTOLOGICA L REFRAMIN G 19
sically relational, that is, reciprocal and dialogical, not mainly monological and in the control of only one part. For this reason the work, including the emotions and knowledge, cannot and should not be standardized a priori. The dynamic, reciprocal, and complex customer interactions depend on the service worker's ability and willingness to see, judge, relate, and cope with different customers and situations (Eide & Lindberg, 1997). In short, being highly capable is far more than just wearing prescribed masks with a false or felt smile, repeating simple expressions, and using other prescribed simple "scripts" in static ways. The relational and complex nature of the interactions depends rather on another emotional side of the work, that is, the emotional attunement, sensitivity, and intuitions that guide the practical judgments and practical knowing (I return to this point later in the chapter).
Second, studies (Forseth, 2001; Hochschild, 1983) also showed that prescribed emotional rules (e.g., smiling) can make the work more alienating and blur personal and work life in a way that commercializes not only the feelings of customers, but also those of employees. Being service-minded can increase the pressure, cause stress, and develop into serious personal and work environment problems such as burnout. These problems are increasing because emotions are not only a large part of work life, they are more often argued as a critical "tool" for an increasing number of occupations. Forseth (in Myhr, 2001, p. 18, my translation3) comments that "it is not only that you have to use emotions. You also have to be clever when doing it."
When emotions are assumed to be "the key tool" that makes a difference in customer service, this "little extra" often means that one is able to get into the spirit of the work (i.e., to be committed and to identify with and live for the work and customers). Service workers with emotional burnout problems tend to cope by being more impersonal and cynical in interactions with customers, which reduces the quality of service created (Forseth, 2001). Furthermore, Forseth (2001) found that traditional care workers (i.e., in health and education sectors) tend to be more prepared for work where the care about others is central than the new groups within private and public sectors. The study included a group of approximately 1,000 employees from three industries: banking, nursery, and retail. Forseth (2001) found that 50% of the bank employees are more or less emotionally exhausted, as are 40% and 34% in the two latter industries. Also, the high number and level of radical changes in the industries increased the tensions, anxiety, and emotionality. In my view, Forseth's (2001) study indicates that the ability to cope with emotions is critical in at least three ways:
1. In the core production (i.e., interactions with the customers or other actors), emotions are a central part of how and what is being "produced." There
3"My translation" means that I have translated the quote from a Scandinavian language into English.
fore, emotions are also something an employee must be able to know, draw on, and cope with in a capable way.
The third reason for not addressing emotions as an invisible asset that is easily managed and utilized is the observed temporality and vulnerable side of interacting in capable ways. My ethnographically inspired, hermeneutical study of receptionists in four hotels (Eide, 2000, in press) shows that the assumed cycling from more explicit (e.g., prescribed rules) to internalized knowledge, feeling, and acting is not the only critical change that can take place. Highly capable receptionists who are dedicated to a service-care ideology can begin questioning their own practice and their continued desire to do it. They balance belief and practice on one side, and struggle with doubt and a possible change in dedication on the other. The mode(s) of practicing (i.e., identity, knowing, and doing) what they believe in, find meaningful, and experience daily in interactions with customers may gradually shift toward less capable practice.
The three reasons addressed thus far can be seen as a critical lens toward work being intensive on emotions since they mainly address problems. However, Hardy, Lawrence, and Phillips (1998, p. 81) asserted that the critical lens is too narrow and needs to be balanced with other views that do not "lead to a similar conclusion as that of the rational view of organizations: both conclude that emotions are (or should be) the private domain of individuals and should not be considered as organizational phenomena." Further, if only using a critical lens one runs the risk of overlooking the many positive aspects of such work and of emotions in organizations. A fourth reason for not addressing emotions as an invisible asset to be managed and utilized is that such an approach can ruin the meaning and positive sides of emotions in organizations. The limitations of only seeing the problems in service work were illustrated when reinterpreting the results of a Norwegian national work environment study con
2. EMOTIONS: TOWARD AN ONTOLOGICAL REFRAMIN G 21
ducted by Grimsmo (1996a, 1996b) in which he concluded that the hotel and restaurant industries have the worst work environment. However, the negative conclusion was mainly due to an assumption built into the survey (i.e., that a high number of customer interactions always caused stress and other negative problems). This assumption is, however, highly questionable because more in-depth studies of the hospitality industry have shown that receptionists and other front-line workers often argue the opposite, as illustrated in, "customers are different. It is through relations with customers that they [receptionists] are challenged, i.e. that they learn, develop, encounter variation, participation and social contact, and experience meaning" (Eide, Jensen, & Lilleby, 1996, p. 7, my translation). First, this shows that general models and assumptions may lead to incorrect conclusions because they do not grasp the fact that there can be considerable differences among individuals, occupations, and sociocultural contexts. Second, it shows that work involving a high degree of emotions and social interactions can be experienced as highly meaningful, learning, and fun, if the work is varied and accommodates autonomy, initiative, and personal style (i.e., being natural caring and relating rather than mainly prescribed, false, and controlled). These experiences also covary with a strong interest and dedication in meeting new people and caring for people. If such interest and dedication are not present, then the assumption that a high level of customer interaction is stressful and negative for the work environment is more accurate. Such lack of interest and dedication can be due to that the organization has hired the "wrong" person(s) for frontline work, or the person has become sour.
Emotions in Work—A First Conclusion
In summary, emotions are not only present, but are often central to being able and willing to do the work in capable ways, especially in work with a high degree of customer interactions. A fundamental question for managers within hospitality and service organizations more generally is how to facilitate and organize so that the customers feel welcome and cared for. Bureaucratic control and heavy standardization tend to be alienating, inflexible, and risky in regard to the interhuman tasks and competence, as the many subordinate ways of interacting cannot be dictated a priori by a manager or trainer in the same way as is possible with more technical tasks. Surface acting and emotional management are a part of everyday life, but this does not mean that emotions in work are purely a phenomenon to manipulate. On the contrary, studies show that emotionally intensive work is far more complex, paradoxical, and dependent on knowing, relating, and dedicating as a whole person. What modes of emotions and knowledge are we dealing with, and how can we understand the complex, dynamic, and subtle layers of emotions and knowledge ? These questions are the focus of the remainder of this chapter.
EMOTIONS AS PART OF A LARGER WHOLE:
AN ONTOLOGICAL REFRAMING
Rather than assuming the traditional dualisms of "cognitive" versus "emotions," or addressing emotions mainly in functionalistic ways, I propose to undertake a more radical rethinking, that is, an ontological reframing. By ontology I mean the more fundamental assumptions about the world and about human nature. Furthermore, I suggest that a situated-relational ontology be assumed. This ontological view represents a shift that tries to overcome and rethink some of the main dualisms in science, such as individual versus environment, mind versus body, thinking versus action, and cognition as rationality versus emotions as irrationality. This ontology includes a mixture of writers within the interpretive-constructionist paradigm. As I show, Aristotle inspires some of the vital writers contributing to this ontology. Aristotle argued in Greek antiquity that pathos, ethos, and logos were not distinctly separate (dual), as was assumed during the Enlightment, nor did he see emotions only as something private but rather as largely public and social. The main thesis argued by this ontological view is that human beings and their more fundamental conditions for actions are situated in within four main situators: as "a whole person" (i.e., with body, emotions, and cognition, not either-or), "in-relations" (i.e., relations with things, others, and self), "in-worlds" (i.e., in different physical, social, and cultural contexts), and "in-history" (i.e., a time involving the past, present, and future). This section begins by introducing key terms and assumptions in this ontology and gradually moves more specifically to the consequences of how to understand emotions and knowledge/knowing. In addition to seeing emotions as intertwined with cognition, and as having both personal and collective sides, I further argue that emotions are relational and part of a larger whole.
Emotions in Organizations as Relational and Part of a Larger Whole—An Introduction
The organizational psychologist Weick (1995, p. 39) suggested a relational view, claiming that emotions "do not grow within us but between us." The Norwegian psychiatrist Finn Skarderud (2002) was critical of the traditional dualism of body versus soul or body versus mind, and suggested that a broader integrated understanding is needed:
We feel. We think. We have a body. And we relate to other human beings. These four phenomena together constitute our mental life ... How our feelings and our thoughts develop has to do first of all with our meetings. It is about individuals we meet. And how we meet. And where and when we meet. Culture means that within us there are depositions of all the meetings that we have ever participated in. (Skarderud, 2002, p. 228, my translation)
2. EMOTIONS : TOWARD AN ONTOLOGICA L REFRAMIN G 23
Skarderud elaborated emotions, cognition, and body as vital and not excludable sides of being human and relating in the world; feelings were also argued as having a particular role in becoming a more whole integrated person:
Our feelings make us whole human beings ... help us move forward in life. The will (lust) drives us toward other humans ... The healthy shame makes sure that the excitement does not become too intense and that we get a grip on ourselves. The fear holds back the foolhardiness ... There is one factor that protects against many things: good self-esteem. (Skarderud, 2002, pp. 228 and 235, my translation)
Emotions were, here, seen as both positive and negative, not either-or; his view seems similar to my own in this regard. The Norwegian philosopher Arne Naess (1998) also assumed a relational approach when arguing for an integrated view of human beings by claiming that emotions should not be seen as something one "has." Rather, these emotions are part of being human and originate from the situated, relational being/becoming in the world:
Emotions are not things, that we own. They are some (-"thing") that originates in the meeting with ourselves-and-the-world.... And it is not so that we only have feelings, just as seldom as we have relations, but in inner circumstances we are emotions and relations. We cannot make ourselves external to ourselves. The thoughts, feelings and relations that we identify with are a part of us in a wider sense. The emotions that we have for a bus driver, for example, depend upon whether or not we perceive him or her as a subject—a co-human, or as an object—a part of the bus that gets us where we want. The goal must be to meet other co-creatures with almost our whole self. (Naess, 1998, p. 24, my translation)
Viewing emotions as relational in this way fits well with what I see as a broad situated-relational approach, because it does not exclude or largely downgrade the personal and deeper sides like some social constructionist writers (e.g., Gergen, 1994) seem to do. Emotions, I argue, should be situated not only within the specific relation and/or the sociocultural world, but also within at least the two further ontological situators—that is, within time, as well as within the whole bodily person.
A Broad Situated-Relational Ontology
Here, I concentrate on the core ideas suggested by Heidegger (1927/1996). For a broader and more in-depth elaboration of a situated-relational ontology, and the consequences for emotions, knowledge, and learning, see Eide (in press).
The following idea indicates the essence of the situated-relational ontology: "This with-like (mithaften) being-in-the-world, the world is always the one that I share with others ... Being-in is Being-with others" (Heidegger, 1927/ 1996, p. 118). This primordial view, also known as an ontological structure, highlights concepts such as situatedness, attunement, and fore-understanding. Heidegger also elaborated other fruitful concepts that are relevant both for understanding emotions and knowing, such as the terms care for self and authenticity. Meaning and interpretations are vital in an ontological sense, and this view is also shared by other approaches, such as constructivism within the interpretive paradigm. Constructivism depicts human nature as an individual and cognitive phenomenon. Heidegger (1927/1996) did not interpret human beings, meaning, and interpretation as purely cognitive.
The individual and cognitive sides are important in the ontological approach, but they are also part of a larger complexity because human nature and meaning are multidimensional and have their origins in the broader situatedness (Heidegger, 1927/1996). People develop and act by means of the situatedness (in body, in relations, in worlds, and in history/time), not as an autonomy-cognitive individual or a completely cultural determined "actor." Rather than viewing the individual as neutral or distanced, being situated means always being attuned one way or another and with a fore-understanding. It reaches beyond traditional dualism between rationality as thinking and irrationality as emotions (Guneriussen, 1996). Therefore, this approach is particularly fruitful for rethinking emotions and humanistic ways of knowing (including knowledge).
Before elaborating further on conceptualizations of human nature, I find it useful to emphasize the following: I share the notion that there are problematic sides with elaborating views of human beings in general terms, because every human is unique. Consequently, it is necessary to separate between analyses on an ontological level (i.e., discussion about humankind and worldview) and an ontic or existential level that involves discussion about the unique phenomenon or person in question. For instance, the situated-relational ontology proclaims attunement as an ontological notion (i.e., that involves every human). How attunement emerges ontically is never obvious a priori because it depends on the context, relations, and people involved. Simply stated, we can say generally that human beings are emotional beings, just as they are thinking beings at an ontological level, but what a person feels or thinks in a specific situated time and place or how a person expresses the feelings or thoughts must be studied empirically as it varies at an ontic level (Heidegger, 1927/1996).
Attunement and Understanding—Emotions and Knowledge as Blurred
Because human beings are situated, they find themselves attuned. But what does attunement mean? We know that it involves deeper emotions, and F10istad (1993) argued that it is a kind of state of being tuned/attuned. Lindberg
2. EMOTIONS : TOWARD AN ONTOLOGICA L REFRAMIN G 25
(2001) asserted that, due to the situatedness of human beings, attunement is
neither something inward, nor something determined from the outward, but
ontically depends on the in-world-existence, while finding oneself is primordially ontological. ... Dasein is indicated by situatedness, which again is indicated by attunements (and feelings). But every human is born with, and [has] developed, different possibilities for attunement and different ability to sense it.... To be "at-tuned" means to be tuned within the world a certain way. (Lindberg, 2001, pp. 96 and 99-100)
The attunement, as a deep emotion, depends on the individual's belonging to the world and interacting in the world; it is relational and does not only involve the inner individual. Attunement, according to Heidegger (1927/1996), is the primary way of discovering and understanding "the world, other individuals, and ourselves" (Lindberg, 2001, p. 97). Attunement, therefore, provides insight and a sense of a " 'being-there-feeling' which is necessary for our ability to grasp reality, act and think" (Lindberg, 2001, p. 101). Angst is a kind of attunement that can provide the break leading to movement and change. In short, Heidegger (1927/1996) elaborated attunement as a precondition that occurs before thinking, insight, and acting; I therefore elaborate further by focusing on attunement and its role for understanding and knowledge. Central in his elaboration is also the hermeneutical process and the situatedness in history, which means that any understanding follows from a fore-understanding (see Gadamer, as cited in Linge, 1976), and the shift between parts and the whole (e.g., that the explicit knowledge or that which we have in focus always dwells in the larger whole of the tacit understanding and knowledge; see Polanyi, 1958, 1966/1983, and Polanyi & Prosch, 1975; more discussion follows).
Both situatedness and attunement show the dynamic and infinite nature of being and understanding, because "the attunements can be destroyed and can change, show that a human is always already attuned" (Heidegger, as cited in F10istad, 1993, p. 196, my translation). F10istad (1993) questioned where attunements and feelings end, and where the cognitive sides of understanding begin. His answer is clarified next:
It is through the attunement that we primarily recognize and understand ourselves, the world and other humans ... the attunements and feelings are the primary source of our consciousness of things, ourselves, and our fellow beings. (F10istad, 1993, p. 196, my translation)
This notion can be exemplified by how one answers the question "How are you?" A reply can be based on the situated feeling of the situation as a human being within time and place. Every relation with, and understanding of, others, things, or self therefore is based on attunement and situatedness; attunement has a kind of cognitive function for summing up the situation. F10istad (1993) believed that it could be thus in many situations, but was critical of this concept as a general claim about understanding. Dreyfus (1991, p. 185), however, argued that Heidegger's (1927/1996) notion of understanding is not a cognitively distanced phenomenon. Alvesson and Skoldberg (1994) argued, when discussing Heidegger, that all "understanding right from the beginning ('always already') is colored by emotional attunements; therefore there is no purely cognitive or rational understanding" (Alvesson & Skoldberg, 1994, p. 136, my translation).
My own view is that the situatedness of the whole person means that our understanding always (and not only sometimes) is partly based on attunement and perhaps other kinds of emotions. Human knowledge is primarily about understanding4 where tacit knowledge and emotions are involved in the horizon of understanding. Active knowledge (i.e., knowing) is already incorporated with emotions; that is, it is not a purely cognitive knowledge or a purely passive, rational, or theoretical one, particularly not when it is included as a part of one's practice. Such understanding of the world, oneself, and others is a vital part of being, becoming and knowing.
In summary, the attunement based on the broad (the four types of) situatedness means that emotions are included at different levels and are intertwined with understanding and knowledge. Emotions are not only a matter of the surface level, the present, the isolated individual, or something irrational. For pedagogical purposes, I have made a simplified sketch—at an analytical level—that includes the key concepts and their relations in a situated-relational approach that focuses on emotional and cognitive aspects (Fig. 2.1).
The situatedness in body, time, relations, and world(s) involves three key ontological structures of being and becoming that have to do with our deep existence: "understanding" (i.e., a vague or clear understanding of our basic existential conditions, such as where we are), "thrown" (i.e., the person finds her/himself thrown into a situation that the person partly may have chosen, and that partly is determined), and "falling" (toward the world and future). These three states depend on the situatedness, and they interact continually and contribute to three states of being: a colored "horizon" (a historical, cultural-based fore-understanding), "attunements," and "approach" (with attention). These three states of being do not exist in isolation, and all three can bring insight that contributes to tacit feelings and knowing (knowledge) and also contributes to "external relational activities" in the world. Such external activity includes feelings being directly and indirectly expressed in interactions. Due to the relational nature of "being" in the world, situatedness can continually be altered, which can in turn alter the three key structures and consequently the three states of being, and so on.
44Polanyi (1958) argued strongly for the importance of understanding in knowing and knowledge.
fore ruined their music experience. This shows that the experience at the concert that normally should have resulted in great enthusiasm also depended on the broader situatedness and not only the isolated listener or the musicians playing.
Further Discussion of Heidegger's Ideas
Like Heidegger (1927/1996), Charles Taylor (1995) also argued that emotions delve deeply into who we are, what we experience, and what is significant to us. Therefore, emotions elevate our attention and contribute to a certain understanding and "movement." As emotions are one of the ontological characteristics of being human, emotions are not something we can "step in and out of"— or neglect (Taylor, 1995). Understanding emotion and cognition as ontological characteristics of human nature means that they are intertwined and interdependent. Present-day society is criticized for idealizing consciousness and action as something driven by the explicit, objective, and general; in doing this, he reclaimed the key points of interpretive writers:
The more well-known argument—forwarded by Wittgenstein, Heidegger, Michael Polanyi and others . . . Rules do not adapt themselves—no matter how extensive and detailed they are. Norms and ideals always demand new interpretations in new situations. (Taylor, 1995, p. 294, my translation)
Here the vital role of meaning, interpretation, and tacit knowledge in action is addressed. Furthermore, Taylor (1995) situated emotions in a larger whole, that is, as part of the self that is seen as
What we need to explain is people
relational in a dialogic way. Self-con
living their lives; the terms in which
sciousness does not develop as distanced
they cannot avoid living them can-
inner thinking. It develops when living in
not be removed . . . the world, as "we become self-conscious We cannot just leap outside of
through thinking, feeling and interact-these terms altogether, on grounds
ing" (Nyeng, 2000, p. 168, my transla-that their logic doesn't fit some model of "science."
tion). Self-understanding, identification,
(Taylor, quoted in Nyeng, 2000,
and development cannot develop without
first participating in a culture, and articulation is a vital part of it.
Humans experience genuine kinds of emotions that are characterized as "strong evaluations" and subject-related emotions that are fundamental when trying to understand human beings (Grimen, 1995). Animals also experience desires of the first order, whereas only humans can experience those of the second order, that is, judging whether our desires are desirable. It is the latter that characterizes strong evaluations/judgments (Grimen, 1995). Examples of genuine human emotions include pride, shame, love, sense of worth, morality, dignity, and evil (see Taylor, 1985, p. 102). For example, other animals do not
2. EMOTIONS: TOWARD AN ONTOLOGICAL REFRAMIN G 29
feel shame, machines of course do not feel at all; even so, machine metaphors have been so popular. These unique emotions are possibly due to the vital role of language in our lives and our situatedness in social practices. For this reason, even prearticulated emotions are dependent on language and are often interpreted emotions (Taylor, 1985). Human beings are described as a continually self-interpreting and narrating animal where meaning and emotions are fundamentally dynamic and intertwined:
This is an animal whose emotional life incorporates a sense of what is really important to him, of the shape of his aspirations, which asks to be understood, and which is never adequately understood. His understanding is explicated at any time in the language he uses to speak about himself, his goals, what he feels, and so on; and in shaping his sense of what is important, it also shapes what he feels.
(Taylor, 1985, p. 74)
Emotions can express meanings and understanding because strong judgments and values are anchored in emotions and struggling. This anchoring is why emotions motivate us and make us attentive5 (Grimen, 1995, p. 28). Kirkeby (2001, pp. 88-89) claimed that certain emotions, such as conscience, justice, and honesty, can express deep and strong judgments. Such judgments and emotions are seen as originating from active participation in social practices and from the integration of the three basic aspects of humanity, that is, of emotions, thought (reason), and will. The idea that emotions can carry and be a mode of expressing values and knowledge dates back to Aristotle (1998). Drawing on both Aristotle and Heidegger, Taylor (1995, p. 158) argued that practical judging and orientation take place in the gestalt of the unique human emotions. Meaning, emotion, and action originate in the relational sphere and not just in the inner and private spheres, as in the constructivist approach. This does not mean that human action is solely determined and "written" by outside sources, as that would be an externally orchestrated monologue and not be dialogical.
Taylor's (cited in Nyeng, 2000, p. 83, my translation) concept of emotions includes at least four subterms: affections, feelings, attunements, and sensations. The last concept—sensations—is a joint term for bodily expressions of all kinds of emotions, particularly those that follow strong affections. To only address this side of emotions is, however, a limited view of emotions, as that would merely measure physical changes during anger or reduce emotions to bodily conditions alone. According to a situated-relational approach, emotions are interpreted emotions with meanings. A feeling cannot be understood as an isolated phenomenon because it is embedded within a larger whole, as a "system" of feelings and other aspects contributing to a web of significance. What
5It is a well-known argument that attention is central to knowledge and learning. In my view, emotions can contribute both to increase and to decrease the attention.
are important when a person is angry are the circumstances—for example, did the person experience an injustice? Thereby, sensations and bodily expressions are not seen as the core of emotions. However, I argue that the body "can tell" by means of intuition and deeper emotions what is "right" for the person. Drawing on the notion of attunement, Taylor (cited in Nyeng, 2000) also argued that human beings are always attuned one way or another, but often are not conscious of it. This attunement means that a human is not neutral when confronting things and others. Anxiety is one example of attunement; it is a deep and fundamental emotion that threatens the existence and ontological security of the person, whereas fear is seen as a feeling directed toward something more specific (Giddens, 1991; Nyeng, 2000). The different terms can be described as follows—I quote at length:
In a way we can say that attunements and affections make up the two endpoints. Everything "in between" is feelings. Both endpoints are characterized by very limited distance and reflection. However, while the attunements are the emotional
fundaments that life exists upon, the affections represent the immediate manifestations of emotional "shaking," that which is strong, that puts us out of play, but which is usually over just as fast as it comes into being...I live in an immediate and all-absorbing here and now ... Feelings are processed experiences that are emotionally loaded—a processing that integrates emotional involvement and rea-son-based reflection and argumentation in a more distanced attitude to the object and power of emotions. (Nyeng, 2000, p. 85, my translation)
All these terms are seen as partly different but related sides of emotions. I share this view; it should be noted that, for example, Salovey and Mayer (1990,
p. 186) defined emotions far more narrowly and saw emotion as something shorter and more intense than mood. Mood, in my opinion, approaches attunement and is part of the larger concept of emotion.
The process of reflecting and articulating feelings makes it possible to grasp what it means to be human, and to "create an individualized and authentic self in the first person" (Nyeng, 2000, p. 90, my translation). When Taylor (1995) claimed that emotions are fundamental, it was because emotions are integrated into the existence as ontological phenomenon and the deeper, more implicit, and unarticulated are present and significant parts of being. When emotions are articulated, the implicit background for actions becomes more conscious: "To articulate a framework is to explicate what makes sense of our moral responses" (Taylor, cited in Gergen, 1994, p. 104). Life can become more human through reflexive awareness since we can reach more of the deeper values of life. Articulating brings emotions into language, and thereby more into focus and into a public space, although what is articulated is always only the tip of the iceberg (Grimen, 1995). Articulations can make us more aware of who we are, what are important to us, and it can also change how we understand and feel about a situation and ourselves (Taylor, 1985).
2. EMOTIONS : TOWARD AN ONTOLOGICA L REFRAMIN G 31
ticulation is not only a representation or translation into an articulated form. It can also have an intervening function:
Formulating how we feel, or coming to adopt a new formulation, can frequently change how we feel. When I come to see that my feeling of guilt was false, or my feeling of love self-deluded, the emotions themselves are different. (Taylor, 1985, p. 100)
By claiming that exerting falseness tends to change the emotion, Taylor (1985) also asserted that emotions can be constituted and shaped by our understanding and interpretations, and not just the other way around as Heidegger (1927/1996) seemed to suggest. Furthermore, articulations can also increase the emotions: "I can describe my emotions by describing my situation, and very often I must do so really to give the flavour of what I feel" (Taylor, 1985, p. 101). Articulation therefore influences how we interact with the world, how and who we are.
Does this mean that experiencing a beautiful rainbow or an exciting football match is not meaningful enough unless it is articulated? I do not interpret this claim on articulation as such; however, the point is that we need to be more aware of what is important to us. Taylor (1985) seemed to assign greater significance to the more explicit and verbally articulated aspects and to self-consciousness than Heidegger (1927/1996) did, and more than other proponents of his ideas such as Bourdieu (1990) and Dreyfus (1991, p. 4), who argued that "what is most important and meaningful in our lives is not, and should not be, accessible to critical reflection." However, critical reflection and articulation are not the same as we might narrate on what we feel without reflecting on it. As I understand it, there are also many situations where we can both know and feel with sufficient meaning without having to articulate it explicitly in words; sometimes articulating might even ruin the meaning.
Compared to Heidegger (1927/1996), Taylor (1995) also elaborated more about the social and public side of life, which I see as a vital contribution to the understanding of human beings and of emotions. There is blurred distinction between the experiences of (lived) emotions versus emotional cultures; I argue that both should be included when trying to study and understand emotions. A vital strength in Taylor's elaboration of emotions is that he does not erode the separation between the surface and the depths or between the private and the public. The differences are needed in development and communication, and not only in deep genuine relations (Nyeng, 2000). Differences shown in contrasting dialogues can contribute to deeper value experiences. It can separate the more instrumental values or superficial expressive values and emotions from deeper significant life values. This is a vital difference between Taylor's discussion (1995; cf. Nyeng, 2000) compared to social constructionist writers that tend to focus too much on the expectations and surface acting. At
the same time, Taylor criticized the consumer society as an instrumental, surface community that lacks focus on real meanings. He argued that today's fragmented and individualized selves only seek to satisfy instant pleasure and that this hampers meanings and relations. Strong values that are expressed as genuine emotions point toward goals to strive for and orient us so that the result is something that is right/good or better/worse.
A Knowledge Concept Where Emotions Are a Part
The situated-relational approach results in a different understanding of hu
man emotions compared to the "ugly duckling" or "invisible asset" views, that
the ontological change also has conse
quences for the meaning of understand-The Aristotelian description makes
us think of real practical insight and
ing, knowledge, and the grounds for ac
understanding as an integrated
tion have also been briefly indicated. How
activity that calls upon the whole to conceptualize human knowledge is soul...
further explored next. Aristotle (1998) Often relying on the intellectual
elaborated on a versatile (although hier-ability can actually be a hindrance for true ethical sensing by putting
archical) concept of knowledge and hu
itself in the way of, or undermining,
man beings as early as the Greek period
the emotions and sensitivity.
of antiquity; he was a great inspiration for (Nussbaum, 1995, p. 85, my trans-
the two main writers (Heidegger and lation)
Taylor) already discussed. Versatile views
of knowledge relate knowing to emotions and intuition. In other approaches,
intuition has often been seen as a primitive knowledge (see Simon, quoted in
Sinclair & Ashkanasy, 2001; cf. Cooper, 2001), or as no knowledge at all,
whereas others see intuition and even emotions as part of highly capable action
and knowledge (Schon, 1983; Sjostrand, 1997; Strati, 2000, 2001). Two exam
ples of the latter are illustrated:
That intuition has an unsolvable connection with emotions seems undeniable to me, this has nothing to do with an obscurity of intuitive representations, but has to do with its richness and unity. (Larsson, 1892/1920, p. 5, my translation)
Into every act of knowing there enters a tacit and passionate contribution of the person knowing what is being known, and that this coefficient is not mere imperfection, but a vital component of all knowledge. (Polanyi, 1958, p. 312)
Larsson and Polanyi share Aristotle's (1998) view that intuition and emotions can increase the ability to see, understand, and act in a powerful way due to its ability to connect a range of parts into a unit. Polanyi (1958, freely quoted) claimed that all knowledge, more or less, includes and depends on personal and tacit knowledge, as "we know more than we can tell." Polanyi
2. EMOTIONS: TOWARD AN ONTOLOGICAL REFRAMIN G 33
(1958) and Polanyi and Prosch (1975)
We know a person's face, and can elaborated how all knowing and knowl-recognize it among a thousand, in-edge include dwelling-in tacit elements deed among a million. Yet we usually cannot tell how we recognize
and tacit integration. The tacit integra-
a face we know. So most of this
tion takes place as a "from-to-attention,"
knowledge cannot be put into
constituted by the content in focus (i.e.,
words ... Werecognize the moods the focal attention) and the tacit content of the human face, without being
(background, i.e., the subsidiary atten-able to tell, except quite vaguely, by what signs we know it.
tion), "showing that all thought contains
(Polanyi, 1969, pp. 4 and 5)
components of which we are subsidiary
aware of in the focal content of our think
ing, and that all thought dwells in its subsidiaries, as if they were parts of our
body" (Polanyi, 1966/1983, p. x).
Trying to articulate the tacit into explicit more or less fails because much of our knowledge (e.g., as shown in the quotation about recognizing faces) is expressed as, and relates to, feelings and activities in action. These two kinds of attentions are vital parts, I argue, in any approach in practice. This contention occurs not only in the private life or in craftwork, but also in the practical research work:
Even when attempts are made to state them explicitly, what these explicit statements mean can be known only by the scientists in the particular field involved. There is much that cannot be made explicit because it lies at the level of feelings about fitness and in working attitudes that betray an essentially imaginative grasp of how things in that field may be expected to work or to be. (Polanyi & Prosch, 1975, p. 186)
Later Wittgenstein (1953/1992) also elaborated on understanding and knowledge by pointing to the limitations of explicit rules and the vital role of personal and social practice for knowledge. Learning to "see" and understand human affairs, and the use of language, is not a rule-based technique to apply in a fixed system such as mathematics. Only experienced individuals can draw upon the rules in highly capable actions, and then feelings inform and guide this use:
What one learns is not technique; one learns tojudge correctly. There are also rules, but they build no system, and only experienced individuals can apply them properly ... The genuineness of the expression cannot be proved; one must feel it. (Wittgenstein, 1953/1992, pp. 261-262, my translation)
It is thus when distinguishing faces, when reading a poem, or when knowing what word to use when talking, or how to move the body and the ski when skiing. Certain rules become a part of us, incorporated into the self-identity and what we know, in a bodily, emotional, and anonymous way. The "rules" locate us in a culture. We do not know them
When I read a poem, a story with
as explicit rules and cannot give full ac
emotions, then something takes
counts of them because such "rules" are
place within me that does not take tacit, dynamic, and social, and differ from place when I only scan the rows to the logic of formal rules that are general, be informed ...
The composition sounds different.
static, and explicit (Wittgenstein, 1953/
I carefully attend the tone. Some
1992). Polanyi (1958, 1966/1983) and
times a word has a false tone, is
Wittgenstein (1953/1992) are now vital
too strong or too weak.
sources for organizational scholars in re-(Wittgenstein, 1953/1992, p. 247,
thinking the discourse of knowledge in my translation)
Sketches Toward a Model
Based on the theoretical elaboration and empirical studies discussed so far, I suggest that interactional emotional work, which constitutes a large part of hospitality and service industry, cannot be fully managed by the manager or the employee, as social interactions often are reciprocal, complex, dynamic, and unforeseen—that is, they are basically relational and dialogic in nature and not monologic and static. Also, the other (the customer or colleague) and the particular relationship that originates between the involved parts have and should have a real influence on the process; therefore, the core production cannot be fully managed by the manager or the employee alone. How then does one understand and approach knowledge and learning? One main implication is that the experience-based practical judgment and understanding become more critical as the employee must be able and willing to act in capable ways on the spot. These ways of knowing draw on and dwell in the more emotional and intuitive ways of being and acting. How knowledge in practice is relational and emotional, and how emotions are relational, are illustrated in Fig. 2.2. My intention in using this model is not to propose an overall and universal theory of emotions in knowledge and work, but to propose an alternative understanding of emotions and knowledge according to the situated-relational approach. The sketch is based on the theoretical discussions and empirical interpretations of my hotel receptionist study (Eide, 2000; Eide & Lindberg, 1997), which can also be seen as a more practical illustration and elaboration of Fig. 2.1.
Person B is the individual focused on and followed in the figure. Like person B, persons A and C also participate in both situations characterized as "inaction" and "processing about-action." The four levels of being, knowledge, and emotions described in the left box exist both when the person is in situations characterized as "in-actions" and as "processing about-actions," because the latter situation is at the same time also a kind of "in-action" here and now. The approach, attention, talk, and other activities tend, however, to differ depending on what kind of main situation one participates in. Knowledge and
become largely negative. In some organizations you may sense the atmosphere upon entering the room, or you may gradually notice it during interactions with the employees, for example, through lack of hospitality, respect, or ability to solve problems. When the "soil is good," however, the interactions may proceed like "a smooth dance" even when they are complex.
Knowledge and emotions are described as multirelational in the figure, as they are seen as developing not only within one particular dyad but also through participation in various interactions among different subjects and across both interactions characterized as "in-action" and those characterized as "processing about-action." For receptionists, both "in-action" and "processing about-action" often took place as social interactions, but the "other" participating in the interaction varied. Experiencing breaks and dilemmas may challenge reality, self, what you/we do, and the knowledge and meaning; such challenging experiences primarily originate in interactions with others who are not receptionists, that is, most often employees from other sections and especially managers, and less often with customers or fellow receptionists. Even though the receptionists normally worked alone in the reception (because the organizations were small) and mainly met during work shifts and formal meetings, the "processing about-action" mainly took place in informal interactions with fellow receptionists. Such processing of experiences, tensions, and ideas often meant renegotiating and reconstructing personal, and perhaps also collective, practices—including knowledge and emotions. Past, present, and/or future experiences that received attention, often due to some kind of break and tension (e.g., how to cope with the angry boss, discussed later), were expressed in conversations, narratives, and other less oral ways. Together they produced outcomes (e.g., coping strategies, how to understand self/work) that later became part of the horizon of understanding, as well as of knowledge and emotions. Social processing about action therefore can be more vital than just being a social gathering or personal/social therapy. It can be a vital way to learn and develop the ability and will to know and feel within the community of practice and organization.
I next illustrate a slightly different side of how emotions and knowledge are relational and intertwined by pointing to the role of insecurity followed by inattention to the customer, versus security and attention at the customer. Attention to the customer and the ongoing situation has been shown to be vital for performing the work in a capable way (Eide & Lindberg, 1997). A further elaboration of knowledge and emotions, as well as empirical examples interpreted within a situated-relational approach, can be found in Eide (in press).
The Angry, Scolding Boss (An Example From Front Office Work)
To obtain a concrete illustration of how emotions and knowing can be intertwined in service work (from my hotel study), I have chosen a vignette, "The
2. EMOTIONS: TOWARD AN ONTOLOGICA L REFRAMIN G 37
Angry-Scolding Boss." Among other things, the vignette is related to the phenomenon where the characteristics and expectations of manager-employee interactions (MI) intersect and spill over into the characteristics and expectations of employee-customer interactions (CI).
One of the receptionists—we will call Kristin—describes how she finds it difficult to cope with and act in her work in a competent way due to the temper of the boss: "He gets angry so easily." One thing is the anger, she argues, but what is worse is the unpredictability and that others, even customers, may overhear:
"One never knows when it will come. I feel insecure. For example, sometimes I can lower the price and it is OK Other times when we do the same thing, we are scolded. Yelling and verbal snapping makes you nervous and it is strenuous. What makes it worse is that we are scolded when everybody can hear. Even worse, we are scolded when the customers are listening."
Being scolded, particularly when customers are listening, symbolizes a meaning, in her view, that deviates from how she and the colleagues understand and practice customer orientation. Kristin continues:
"I have experienced how difficult it is to suggest improvements or to talk about problems with the boss due to his unpredictability. One day I went in to his office and asked a question. When I went back to the reception he yelled at me, 'Are you stupid?' Guests were in the reception and heard. What sort of customer orientation is that? It makes me insecure about myself."
Being insecure is not a good condition when one needs to interact capably in customer interactions. Rather, it hampers one's ability to act in capable ways, partly because it takes the attention away from the customer, and thereby reduces the ability to sense, understand, and relate. In addition, it does not look good when customers overhear such reprimands either. This example shows how the receptionist's participation in customer interactions (CI) and interactions with the manager (MI) are intertwined; the latter creates breaks and tension not only for the receptionist but probably for the customer as well. A similar situation happened when the hotel became the local supplier for a national airline company. The boss argued that the contract was important to the hotel, routines had to be altered, and the involved customers were followed up very closely to prevent failure and to learn. Several small "crises" had already taken place and this had increased the insecurity and the frequency of the reprimands. One receptionist describes one of the "crises":
"One morning the crew guests were not awakened. That is critical. When the boss heard it, the receptionist ['Marit'] on duty was heavily scolded at the recep
tion desk. The scolding took place even before finding out what caused the mis
take. It turned out to be a technical mistake in the call terminal. The next day
Marit was crying when she came to work."
When Marit came to work the next day, Kristin was on duty, and they interacted (RI, receptionist interaction) during the work shift. Marit expressed through verbal and nonverbal ways what had happened and how she felt. She was particularly afraid that the boss might start scolding her or others when customers were listening. Together, while expressing, articulating, and processing the experiences and how they felt, the receptionists coconstructed how to make sense of and cope with the situation in order to go on with their work and life. The social construction that developed during the interaction was something like this: "Do not worry, it's the boss's problem. We just have to make the best of it and do all that we can for the guests." Indirectly they took responsibility and reconstructed themselves as heroes, and the boss as the stupid and irresponsible one. The social support and social relational constructions created by their camaraderie made the reconstruction possible. The example also shows how participation across interactions with three different kinds of others, that is, in the three different kinds of interactions (CI, MI and RI), can become a web and interplay of subprocesses in an integrated learning process where emotions are part of the learning process and result.
In the instance just mentioned, the event was at least temporarily solved in a positive way. It even seemed to strengthen the identity of the receptionists. However, Marit could have become so self-centered and uncertain that she could have become unable or unwilling to interact with the customers in a knowledgeable and learning manner. This could have occurred if all the vital attention that should be on the customer and their mutual interaction process was turned toward herself and/or the manager. She might have stopped really seeing and relating dialogically with the customer. The anxiety can also grow and spread to most or all the receptionists, and thereby start a more radical change in the personal and social practice in a broader sense (i.e., including how they feel, think, act, talk, identify, know, and learn). As part of this, the individual or the whole community of receptionists may start to question their roles and work. This doubt was not only expressed in the hotel with the "angry boss," but also less tension-filled examples were observed. For example, one argued, "Since the boss does not seem to care and regard me as an important and knowledgeable person, why should I care and regard my work and role as important?" Highly capable and dedicated receptionists expressed the tension and doubt about how long they were going to be willing to continue the employment, or be able and willing to work so hard and carry out their duties in a caring and competent way.
Being able and willing to care for others (e.g., guests, customers, users, and colleagues, and thereby the work) is relational. One basic question shows this
2. EMOTIONS: TOWARD AN ONTOLOGICA L REFRAMIN G 39
relational nature: Is an organization with managers that really care for their employees more likely to get employees that really care for their customers and the organization as a larger whole? There are several indications of this type of connection in my study. In the case just described, it was the fellow colleagues that really cared and contributed to the social negotiation concerning the care ideology and identity. In this way, individuals other than the manager can maintain this vital role as participators in the relational reconstruction of practice, emotions, and knowledge. The case, however, also shows that the manager is nevertheless significant to the practice of caring—in this case as a threat—both on a short- and a long-term basis: A doubt had been planted, and it gnawed and could enlarge if replanted.
My first line of argument in this discussion of emotions was to show the importance of discussing ontological assumptions (i.e., to rethink the view of human nature and the world) for understanding more about emotions and knowledge. Assuming that human beings are whole persons situated in time, body, and in relation, when being in the world, means that we are multidimensional and thereby emotional beings. This fundamental condition is, therefore, also present in human knowledge in practice. We cannot "step in and out" of ontological conditions, nor fully distance ourselves from, or manipulate, how we feel. I have also argued that the move from viewing emotions as an "ugly duckling" to mainly seeing them as an "invisible asset" to be managed and utilized instrumentally is still a cynical and limited view. It does not grasp properly the nature and role of emotions in human life. Emotions can be "good" and "bad," and they occur at both deeper and surface levels; however, they are a fundamental part of being human. I have suggested that a broad situated-relational approach can contribute to a more versatile, holistic, and relational understanding of emotions. Figure 2.3 summarizes the core points of the three main views of emotions explored in this chapter.
I hope I have demonstrated that the common dichotomy of rational versus emotional, knowledge versus feeling, somehow collapses within a situated-relational approach. This has conceptual implications; there is no pure rationality or cognition without emotions, and vice versa, in human organizational practice. This view on rationality has similarities with Domagalski's (1999, p. 838) description of rationality in organizational practice, "organizational actors quite rationally draw upon their emotions to evaluate their circumstances." Also, it comes close to Fineman's (1993a) description of good communication as dependent on the ability and will to read the situation of, and interact with, a constant mix of messages that are embedded with feelings. Practice, like leadership, is more about acting than planning; it is more about practicing skills
Traditional view —> Ugly duckling —> Suppress, control, secure functioning 'machinery'
Emotional labor —> Invisible asset —> Emotional performance and management: Tap/utilize, surface acting in accordance with some standards (expectations, norms, rules) To maximize productivity and service quality Dilemma 1: Increased risk for stress, burnout, and other emotional tensions, and reduced ability and will to cope and act in capable ways Dilemma 2: Increased risk of being perceived as static/not relational, false and with low capability/service quality
Situated-relational —> Ontologically structured: Being/becoming is multidimensional and attuned; i.e., emotions are part of being human and cannot be excluded —> Varies at ontic/empirical levels —> Emotions and knowledge:
FIG. 2.3. Three main ways of approaching emotions.
than about following scientific or other explicit methods and rules (Bryman, 1996; Mintzberg, 1990; Sjostrand, 1997). Rules cannot formulate practice because practice is more contextual, more relational, and messier (Bourdieu, 1990; Bruner, 1990; Wittgenstein, 1953/1992). Further, rules have no use without people using them; they are always depending on actors that know and act. Stretching emotions too far can decrease the ability and will to interact in capable ways. From this follows a first practical implication: Capable interaction cannot mainly be organized a priori through explicit knowledge (e.g., standardized rules, emotional or voice scripts), as capable action and knowledge depend on living human beings that are situated and that know and act. This involves tacit knowledge where understanding, attention, judging, emotions, and intuition are involved parts in a larger process. Thus, even highly capable employees can become less capable if becoming insecure and moving the attention to self or other persons, things, or routines that takes the attention away from what is the core process (e.g., the customer interaction).
A second practical implication of the earlier theoretical discussion on emotions, which also resonates well with my empirical findings in the hospitality study, is that the employees' desire to care and act in capable ways in customer interactions is "created," "reconstructed," or "reduced" during participation in and across interactions, including the interactions with managers. Questions such as "Why should I care when the manager does not seem to care for the guests and for me?" seem fundamental to receptionists and should be no
2. EMOTIONS: TOWARD AN ONTOLOGICAL REFRAMIN G 41
less vital for managers. Such questioning and doubt have consequences for the quality of service and for the work environment. But why is it so? My suggestion and conclusion is that care is relational, infinite, and vulnerable, not inborn or only about the individual sphere, for example, as fixed personality characters (e.g., as a fixed EQ rate). It involves, and is constituted by, emotions, identity, action, respect, and knowledge, which all are situated, relational and reciprocal subphenomena—not monologues. Care is not something one develops and then has and owns forever; rather, it is modes of being and doing overlapping with ability and will; all are dynamic and relational, and they are continually tested and reconstructed in relations in worlds. Being met with respect, appreciation, and care by important others seems to increase the likelihood that the person also respects, appreciates, and cares for others and the self. Being yelled at or in other ways not being interacted with in a way where one is respected, appreciated, and cared for creates doubts and questions the practice of the persons. This can then reduce the ability and will to know and care. Such a change in the dominating attunement of the person also alters the approach, attention, and carefulness, which also results in a change in knowledge in practice (e.g., carelessness).
Emotions have been irrationalized or neglected for too long; now emotions also are seen as something positive. This chapter shares this goal, but it also exposes an alternative to the more recent cynical view wherein emotions are seen as a new "gold mine" that mainly should be exploited and managed. Human beings as well as the quality will suffer under such instrumental control.
An earlier version of this text was part of a paper presented at EGOS 17th Colloquium, Sub Theme Group 27: Knowing as Desire, Lyon, 5-7 July 2001; I would like to acknowledge the comments from participants in this group. My colleagues Elisabet Ljunggren and Jean Jaminon contributed with inspiring ideas in the first phase of this text. Frank Lindberg, as well as the three blind reviewers and the editorial team, proposed significant comments on later versions. None of them, however, bears any responsibility for its contents. Finally, I acknowledge the financial support of Bod0 Graduate School of Business, Economic Research Fund, and Nordland College.
This page intentionally left blank
This page intentionally left blank
'You Wait Until You Get Home": Emotional Regions, Emotional Process Work, and the Role of Onstage and Offstage Support
Maree V. Boyle
This chapter explores the connections between emotionalized regions within organizations and the kinds of emotional process work that occur within these regions. A study of an emergency service organization over an 18-month period found that the performance of emotional process work is a vital stage in the overall performance of emotional labor within this industry. Interviews with emergency service workers also indicated that a substantial amount of emotional process work occurs within one of three emotional regions within the organization—the "offstage" (or nonwork) region. The organization in question, known here as the Department of Paramedical Services (DPS), relies heavily on informal "off stage" emotional support. Thoits's (1985) work on emotional process work and Goffman's (1959) work on regions are used to demonstrate how the individual management of emotion and the organizational ordering of emotional regions are intertwined closely. Organizational implications for the overreliance on offstage forms of support are also discussed briefly.
Although we don't get enough training in how to cope, the expectation is that you have to cope.... So you can do your job, you can let the tears flow but you still have to be efficient at what you are doing. At least until you get back to the station. But that doesn't usually happen. You wait until you get home ...
The principal aim of this chapter is to explore how the existence of emotional regions within organizations and the individual practice of emotional process work as part of emotional labor are intertwined. This discussion is based on qualitative field data collected over an 18-month period within an emergency
services organization specializing in prehospital emergency care. The organization in question, the Department of Paramedical Services (known hereafter as the DPS), could best be described as an emotion-laden organization, where emotionality is central to the raison d'etre of the organization.
The linkage between emotional regions within organizations and emotional process work is illustrated through a closer examination of the offstage (or nonwork) region within the DPS. Thoits's (1984, 1985, 1991) work on emotional process work and Goffman's (1959) work on regions will also help illustrate how, in this instance, emotional process work can be "privatized" and thus removed from the realm of organizational responsibility.
This chapter is divided into the following sections. First, I provide an overview of the research site and design. Second, I discuss the concepts of emotional regions, emotional labor, and emotional process work, and how these relate to the study discussed here. Third, I provide a detailed and richly narrative style account of the nature of emotional labor and how it is performed within the DPS. Within this section I also highlight the different approaches to the performance of emotional labor in front-stage regions, which includes illustrations of surface and deep acting, and the process of emotional "switching." Fourth, I illustrate the nature of offstage support within the DPS, and discuss how the overreliance on this kind of support may leave an organization such as the DPS vulnerable to the negative financial consequences of occupational stress.
EMOTIONAL CULTURES, REGIONS,
Emotional culture within organizations consists of three components: emotional vocabularies (Gerth & Mills, 1953; Gordon, 1981), emotional norms (Gordon, 1989; Hochschild, 1979, 1983; Scheff, 1979, 1990), and meanings of power and status (Kemper, 1978). Gordon (1990) also differentiated between institutional and impulsive orientations within emotional cultures. Institutional meanings of emotions are those given by organizational members when they are in full control of their emotions. Members effect achievement and maintenance of institutional norms, and in doing so continue to uphold and reproduce emotional culture (Gordon, 1989).
In formal organizations such as the DPS, the application of impulsive modes of emotional expression is considered either deviance or indicative of faulty socialization (Gordon, 1981; Thoits, 1989). However, permission to express impulsive emotion is granted to those with power and status, typically middle- and upper-class men of Anglo-Celtic origin (Hochschild, 1983; Pierce, 1999). These "status shields" also apply to relationships between clients and organizational members. Those with greater professional status are less likely to witness impulsive orientations to emotion than those with lower
3. "YOU WAIT UNTI L YOU GET HOME " 47
status (Hochschild, 1983). For example, patients are less likely to exhibit displays of extreme emotion in the presence of medical consultants than they are in front of clinical staff that are considered of lower status.
The concept of emotional region is derived from Goffman's (1959) dramaturgical perspective. Performance, which is a central component of an organization, is defined as "all the activity of a participant on a given occasion that serves to influence any of the other participants" (Goffman, 1959, p. 26). Performances are only successful when individuals can show that their actions are genuine, while sustaining simultaneously a "front" that is considered authentic (Goffman, 1959, p. 28).
Successful performance is also staged by teams "who share both the risk and discreditable information in a manner comparable to a secret society" (Goffman, 1959, p. 108, as cited in Manning, 1992). Teams are organized by "directors" who manage disputes and delegate responsibility. Teams also act in "front regions," which are defined as spaces within which they perform for their public (Goffman, 1959, pp. 102-114). Teams "rehearse, relax, and retreat" to "back regions," spaces hidden from publics' view when front region performances are "knowingly contradicted as a matter of course" (Goffman, 1959, pp. 110-114). Goffman's (1959) conceptualization of front and back regions is used here heuristically to develop further Fineman's (1993b) notion of the "emotional architecture" of organizational culture, in which he suggests within organizations physical spaces exist in which different feeling rules apply.
The concept of emotional culture builds on Gordon's (1981) original conceptualization, joining both Goffman's (1959) description of regional behavior and audience segregation and the differentiation perspective of organizational culture (Martin, 1992), which recognizes the importance of subcultures. Therefore, emotional culture can be observed within three "regions"—frontor onstage, backstage, and offstage. The front-stage sector is where emotional labor is performed, whereas the backstage sector is where interaction with organizational members happens and where emotional process work is likely to occur. In comparison, offstage regions are located outside the physical realm of the organization itself. As Hosking and Fineman (1990) asserted, differentiation between front-stage and backstage organizational emotionality helps us understand the nature and consequence of emotional labor, particularly within the context of emotional culture.
EMOTIONAL LABOR AND EMOTIONAL
Recent research on the links between the performance of emotional labor and emotional dissonance indicates that a complex array of factors has both positive and negative effects on the individual's emotional well-being in the workplace (Hartel, Hsu, & Boyle, 2002). These factors include the quality of the immediate workplace emotional climate in which the service encounter occurs (Ashforth & Humphrey, 1995), the influence of gendered cultural norms (Wharton, 1993), the degree of job control and routinization (Leidner, 1993; Van Maanen, 1991), and the quality of organizational responses to stress induced by emotional labor (Kunda & Van Maanen, 1999).
Early work on the conceptualization and operationalization of emotional labor created a clear distinction between emotion work and emotional labor (Hochschild, 1979, 1983). Expanding on Hochschild's (1983) definition of emotional labor previously discussed, Hochschild (1983) also argued that emotive dissonance was an inevitable consequence of emotional labor because it resulted in a transmutation of the private emotional region into the public commercial region. However, emotion as a process involves the appraisal of a series of affect-related events, which may involve the experience of discrete or private emotions such as sadness or envy. Although the context in which the appraisal and subsequent emotional regulation take place may change from a public to private one, the process of appraisal, attribution, and regulation of emotion is essentially the same (Weiss & Cropanzano, 1996).
Therefore, I propose that emotional process work is an integral part of emotional labor, and is an extension of the service provider-client interaction. In addition, I also propose that organizational response to this aspect of an employee's work influences significantly both the quality of the service outcome and the levels of individual employee stress fitness and emotional health.
Emotional Process Work
Emotional process work occurs before, during, and after a service encounter, and involves a number of strategies that enable the employee to maintain a normative emotional state. Thoits (1984, 1985) explained that when emotional management techniques fail and individuals are unable to deal satisfactorily with "deviant" or "outlaw" emotions such as disgust, extreme anger, or hatred, they then have to process this failure as a violation of feeling or expression norms. Thoits (1985) cited two conditions that she viewed as central to the prediction of emotion work failure: the persistence of deviant or outlaw emotions, and the absence of social support. Thoits (1985) explained that when individuals are committed to competent identity enhancement and are aware of a discrepancy between situational feelings and emotional norms, attempts at emotional process work follow, and self-attributions of deviance occur as a result of persistent failure to create an individual normative state. For example, if a paramedic felt extreme anger after attending a case such as child abuse, this would be considered the emotional norm for this particular situation. However, if the same officer felt nonchalant or disinterested about the same case, he might attempt to move his feelings closer to the emotional norm for this sit
3. "YOU WAIT UNTI L YOU GET HOME " 49
uation. If he were unable to do this, then he would be more likely to label his own emotions as deviant.
Therefore, Thoits's (1984, 1985, 1991) work has implications for how emergency sendee organizations confront the reality of work stress and the maintenance of appropriate emotional climates within the organization. Within an emergency service context, emotional process work occurs after a case has been completed and involves a variety of strategies that are designed to assist the officer to return to a normative emotional state. The parameters of a normative state are determined by both societal and organizational cultural norms, and are influenced by gender, national culture, and generational emotional norms. Emotional process work may be as simple as one officer acknowledging to another officer that the previous patient was rude or obnoxious, or it may involve many weeks of coping with a major traumatic event such as a plane crash. All officers "do" emotional process work, and the degree to which they accomplish successfully emotional normality varies according to level of experience, degree of social support, and ability to cope with the demands of emotional norms and feeling rules that the organization places on them.
RESEARCH SITE AND DESIGN
The DPS provides prehospital emergency care to subscribers to its service. It is a public-sector organization that has developed a culture emergent from a combination of both militaristic and not-for-profit influences. The DPS is a male-dominated organization, with over 90% of the on-road staff being men. As part of their duties as "caring" paramedics, the DPS on-road staff are expected to perform as emotionally complex individuals while simultaneously adhering to a strict hegemonically masculinist code of conduct. Officers are expected to display the "softer" emotions of compassion, empathy, and cheerfulness in public, while refraining from the expression of grief, remorse, or sadness in the company other officers. Although this expectation is not harsh in itself, it becomes untenable when the DPS relies heavily on the "privatizing" of emotional process work.
Using Thoits's (1991) work on social support as a basis, emotional process work is defined as the emotion work in which officers engage after emotional labor has been performed. Although this practice incorporates what Hochschild (1983) referred to as emotional management, in this particular context it is used to differentiate between the processes used while emotional labor is being performed, and those utilized to make sense of the interaction to which emotional labor is central.
In keeping with Hochschild's (1983) original definition, emotional labor is defined here as the appropriate level of display, feeling, and exchange that occurs between the service provider and the service recipient. Therefore, the
practice of emotional labor includes both individual emotion work and emotion management of others' feelings. In the DPS context, emotional labor is specifically defined as the management of the emotional interface between paramedic and patient, and/or persons located within the vicinity of the interaction with whom the officer needs to communicate with in order to successfully accomplish the task at hand.
This ethnographic style qualitative study of the emotional labor practices within ambulance work utilizes a triangulated approach that involves extensive observation of work routines and practices. Given that self-reports of intangible and unobservable feelings and inner emotion work may be difficult to validate through formal interviews only, I chose this observational methodological approach because I considered it the most appropriate way of accessing this kind of data (James, 1993). Document analysis of training and human resource materials, recruitment practices, and organizational mission statements was also performed.
My approach to ethnographic research is influenced by the classic anthropological approach, which requires the researcher to adopt the role of "professional stranger" (Agar, 1996). According to Van Maanen (1988), this kind of ethnographic study can be categorized as more of a "critical" than "realist" account of the culture of the DPS. Therefore, it does not focus exclusively on my personal experience as a fully immersed participant, but rather is a critical account of organizational emotionality within the workplace. Although I was physically and emotionally involved in particular cases, I did not wear a uniform, was not permitted to comfort or reassure patients, and was not fully accountable to the DPS as an employee or volunteer. My own experiences in the field did not involve "doing emotion" in the same way that ambulance officers did. This psychic distance from the actual work in which officers engaged is indicative of the well-documented dilemma field-workers face when they are restricted in their ability to gain unlimited and pure access to informants in the field (Hubbard, Backett-Milburn, & Kemmer, 2001).
Fieldwork was conducted within the DPS over an 18-month period. During that time I conducted 500 hours of observation, and attended and partially observed 110 cases. I observed cases with 50 on-road officers, 9 of whom were women. In addition to these observations, I conducted 30 in-depth interviews with officers across the 7 DPS geographical regions. I also attended training sessions, spent time within communication call centers, and held informal discussions with senior managers and counselors about DPS policies regarding posttraumatic stress disorder and stress debriefing.
An ethnography is a written representation of either a whole or parts of a culture, and carries serious intellectual and ethical responsibilities (Van Maanen, 1988). Therefore, every effort has been made to protect the identities of the paramedics who agreed to be interviewed for this study. The names of these interviewees have been changed to maintain anonymity. Care has been taken to
3. "YOU WAIT UNTI L YOU GET HOME " 51
choose names that do not correspond with those officers who were observed or interviewed. At no time during fieldwork were patients' names recorded.
ON-STAGE AND BACKSTAGE REGIONS: THE DPS
AND THE PERFORMANCE OF EMOTIONAL LABOR
Emotional labor is an integral part of a paramedical officer's work day. Several aspects of paramedic work impinge on how and why emotional labor is performed. They include the management of patients' emotions through surface and deep acting; the management of one's own emotions through emotional switching; generational and geographical differences between officers; and training to perform emotional labor. Hochschild (1983) differentiated between surface and deep acting, in that surface acting involves a superficial expression of the appropriate emotion. Deep acting involves the combination of the actual exhortation along with the expression of the emotion. Emotional switching is an emergent category from the data, and it involves the ability to change one's emotional demeanor quickly. In paramedic work this occurs frequently, particularly in the move from a highly emotionally charged job to one that is less taxing and emotionally demanding.
There are four emotional labor practices that are specific to paramedical work: the ability to manage the patients' and one's own anxiety; managing people who are out of their usual environment; being able to cope with a constant state of uncertainty; and the ability to simultaneously "care for" and "care about" others (James, 1993). This section focuses on emotions through surface and deep acting, as they are a critical component of a paramedical officer's position.
Most paramedical officers admit that they do a significant amount of surface acting. This involves smiling, teasing, and light humor, all of which should be taken at face value only. During the study's observations, there were several occasions when officers were friendly toward the patient but then complained bitterly about the patient's attitude or expectations after the patient had been transported. In these cases, patients complained about the length of time spent for the ambulance to arrive, which irritated the officers, because lengthy delays were often out of their control.
Although officers generally displayed a friendly demeanor in front of most older patients, several officers either ignored dementia patients, or treated quiet or depressed patients in a condescending manner. A more somber demeanor was observed on several night shifts, when officers treated many intoxicated, indigenous, or adolescent patients.
During observations, officers reported that they used surface acting techniques with almost every patient. Whether it be sharing a joke with a shy or fearful patient, smiling in the face of a belligerent one, or developing a "neutral" expression when having to coax an intoxicated person into the back of the ambulance, officers view acting as part of their work. However, they also distinguish between surface acting, that is, expressing positive or negative emotions even if they are not authentic, and deep acting, when officers are required to actually feel the emotion they are projecting. Officers define surface acting either as the act of expressing positive feelings, especially in circumstances where the situation does not lend itself to positive emotionality, or as the suppression of negative emotions, such as sadness or disgust. In both circumstances, the performance requires officers to give something that is either themselves, or denies parts of themselves.
There are many techniques officers use to perform surface acting successfully. Humor was cited as the major strategy used, and during observations it was the most common way of engaging with patient: As one senior officer stated:
If you can make someone laugh or smile you've got the battle half won. Treat them like an equal. Get their mind off their worries. Have a joke with them or whatever. If you get their mind off what's hurting them you're well on your way.
According to Francis (1994), humor as emotion management involves a sophisticated cultural performance, which strengthens and often restores normal feelings. Francis (1994) also stated that humor is used as an integrative device that effectively reduces (or redefines) an external threat. This bonding process often occurs at the expense of the excluded person or persons.
A good example of how humor is used to alleviate patients' and relatives' anxiety was when a man in his eighties was transported home from a hospital after a hip replacement operation. Officers needed to be able to gauge how much humor they can perform, and in this case, they used the opportunity to make light of a delicate situation. When the patient arrived home, the officers sensed that the relatives needed much reassurance, as they seemed particularly anxious about the difficulty the patient would experience walking up the front stairs. The officers were extremely jovial, and joked with both the patient and his relatives. One of the younger officers even took the crutches and ran up and down the sidewalk outside the house with them. The relatives were not offended by this, and even seemed to appreciate the mild joke at the patient's expense. This is an example where the officers judged correctly how much humor was acceptable, and even used the permission given by the relatives to engage in "horseplay."
However, there are occasions when humor either fails or is unappreciated. When a form of surface acting has failed, officers have the option of reverting
3. "YOU WAIT UNTI L YOU GET HOME " 53
to silence or changing the topic of conversation to strictly medical issues (Palmer, 1983). In the case of long trips from rural areas to metropolitan hospitals and vice versa, humor is one of many strategies officers used to quell irritable patients. As this officer of 6 years indicated, there is often a difference between the type of surface acting performed on a longer and shorter trips:
I think a lot of DPS officers joke a lot of things off and they've got standard lines that they can use to laugh it away and maybe try to get this person around through humor. But there's a bit of a difference there between city and country too because where I am I'll probably have a patient on board a maximum of twenty minutes. So for me, it's quite easy. I can put up with them and be nice to them for that period of time and then I can just refer them on. But with a country person having to put up with a whinger from the back of the bush to the city on a long trip would be extremely difficult. There are two in the car, and if by the end of the day you're starting to become weary you can swap over.... All you can do is what is within your power at the time. All you can do is try to be friendly, try to understand where they're coming from.
During the course of an average day shift, officers transport many nonurgent cases. These cases usually involve a low level of technical skill, although they often involve a considerable amount of surface acting. Many nonurgent patients are older frail women, sometimes suffering from dementia, and some officers expressed a significant amount of ambivalence toward these patients. Although all officers I observed transporting these patients expressed either positive emotions or affective neutrality, in many cases officers expressed dislike or discomfort about these cases to other officers after these patients had been transported. For example, one younger officer referred to nonurgent cases as "granny-busting." Others referred to these cases as "humpers" or "geries," which refer to older patients who have difficulty walking or communicating. As one officer of 20 years explains, this kind of emotional distancing is a vital part of the job for officers who are barely out of adolescence:
DPS officers dissociate very well. I mean, how can you not? The DPS is now made up of officers in their late teens to early thirties who have all got families and they pull kids out of pools and go to motor accidents. In a lot of cases they're treating elderly people around the same age as their parents with the same sortof complaints. If they don't dissociate themselves with that patient emotionally they get caught up in the web of themselves. And occasionally you do.
As one officer explained, emergency service workers are exposed to more emotionally distressing situations in the course of one shift than many people would experience in a year. This unsanitized experience of the social world is one that most officers come to accept as "part of the job." Officers also justify their tendency to distance themselves emotionally as part of the process of developing a veneer. The following quote is from an officer with 8 years of experience who explained that developing such a shield is important part of acting out the professional role of the paramedical officer:
A lot of people may be upset but you just harden because you do see so much of death and dying ... it's just a part of life. We're not isolated from it the way the general public is. Years ago everyone saw death and dying. There was a lot more blood around, we killed our own animals. It's not that we're necessarily special, it's just that we keep seeing what most people saw forty or fifty years ago. We also get to see a lot more inappropriate behavior and people trying to cope with things. If you don't have that veneer, if you don't have some hardness, you're not going to survive.
Therefore, surface acting is not only a strategy for managing patients' emotions; it is also considered an important method in managing one's own emotions and emotional conflicts. Many officers admit that developing an extreme focus at the scene of major trauma, what they describe as tunnel vision, is the most common method of suppressing or denying individual emotional responses. This form of affective neutrality is also considered necessary for both managing the anxiety of the patient, and appearing professional:
You become more job orientated. You see the patient as a job. You don't see them as a person. Once you get them settled down you might give them some reassurance. It's a professional thing. You have to close it out. It's very hard to explain.
Thus, officers' perception of professionalism involves successful accomplishment of an affectively neutral performance. In this respect, ambulance officers are not unique. What distinguishes DPS officers is the degree to which they find themselves in situations that call on them to engage in deep acting.
In the context of the DPS, deep acting involves the management of all aspects of an affective state, including the ability to change how one is feeling about an event or person. The most common deep acting technique officers referred to was when a patient was objectified. This was achieved through officers developing the image of the patient as a thing or object, and in doing so, temporarily suspending any emotional attachment they may have to the patient as a human being. Although most officers admit that this strategy should not be used frequently, many explained that high levels of stress and emotional exhaustion encourage the overuse of this strategy. One officer justified the use of this technique by highlighting that deep acting strategies are no different from those used by other health professionals:
3. "YOU WAIT UNTI L YOU GET HOME " 55
The analogy of a piece of meat was given to me early on in my DPS career and it's right. I mean, it's a terrible term. But the analogy I was given was that a surgeon treats a patient as a piece of meat and does his best possible job on it. Carves it like Michelangelo carved David and then puts it back. And that's the way DPS paramedics do their job. If they get emotionally tied to it then they can't effectively do the job.
Smith and Kleinman (1989) explained this tendency toward objectification by drawing comparisons between the socialization processes of military recruits and medical students. Both undergo a process of desensitization where emotions are blunted by sleep deprivation and changes in the way students perceive human bodies.
Although some officers overuse surface and deep acting strategies needed to achieve emotional distancing, others are prone to emotional attachment with patients, particularly those who remind an officer of a relative, child, or loved one. Many officers I spoke to explained that all officers are prone, at some time during their careers, to communicate inappropriate emotional messages to the relatives of the dying patient, because officers find it difficult to reconcile their lifesaving role with the very real possibility of losing the patient.
Most of the officers who had worked in the DPS for many years described circumstances where they had given false hope to a patient or relative because they were unsure or unable to perform the kind of emotional labor specific to that context. A common example was giving false hope to the relatives of burn victims based on the percentage and degree of burns to the body. Officers stated that they were most likely to fail in their efforts to engage in deep acting at the end of a series of night shifts, after a series of urgent cases in one shift, or when they failed to dissociate successfully.
Cases involving critically ill children were cited as the most risky in this instance. This is similar for the majority of emergency service personnel, most of whom cited cases involving the death of, or serious injury to, children as the most emotionally challenging (Palmer, 1983). The cases I observed involving children always evoked a different response from officers in that they appeared to work at a faster pace. One officer explained that apart from the fact that many officers had young children themselves, they also felt added pressure to always save the child, regardless of the severity of the case. This demands of officers a dramaturgical performance that is quite distinct from adult cases. The following quote from an officer with 8 years of experience illustrates the extreme emotional and physical pressure officers finds themselves under during the course of attempting to save a child's life:
When you walk into a scene where there's a child who is ill, if it's a SIDS or something, there are so many expectations. The parents, especially with SIDS, know the child's dead. You know the child's dead. But if the child's still warm you have to give them every opportunity.... You see, there's so much expectation on us when we go into a situation we can't turn around and go to pieces because then we're not doing our job.
The death of a child is probably one of the few instances where a strong emotional reaction to a case is legitimized. Although several younger officers expressed angst about this, most officers believed that that an increase in the acceptance of public displays of grief, if only for children, is a positive change within the emotional culture of the DPS.
In summary, the majority of officers interviewed and observed cited emotional labor as an important "tool" in the DPS officers' repertoire of skills. During the course of observation, the officers described a number of surface and deep acting techniques utilized in the performance of emotional labor. These included managing the emotions of patients through extortion and suppression of emotion. The specific techniques also included use of humor to manage anxiety, the avoidance of false hope, emotional distancing through developing tunnel vision, disparaging stereotypes, and the categorization of patients into deserving and undeserving cases. Officers also spoke of the process of hardening through developing an emotional veneer of affective neutrality and professionalism.
One of the most enjoyable and frustrating aspects of paramedical work is the unpredictability of each shift. An officer may be kept busy with many nonurgent cases in the morning, only to find that the afternoon is quiet. Then, just before the end of the shift, the officer may be called to several urgent cases. As such, the paramedic cannot predict exactly what he or she will be doing on any one shift.
Nonurgent cases make up the bulk of an ambulance officer's work, particularly on day shifts and during the early part of night shifts. Although an urgent case involves less technically skilled work than nonurgent cases, it does involve a considerable amount of emotional labor. However, there is no compulsion to perform emotional labor during nonurgent cases, and many officers express ambivalence regarding how much "skill" is involved in these cases. For example, some officers are very frank about their assessment of nonurgent cases as boring, dead end, and no better than "driving a taxi." Although it is highly likely that officers would attend an urgent case during every shift in urban regions, in provincial and rural areas officers may have to settle for weeks of nonurgent cases.
However, although some officers may complain about too little stress, others experience shifts that are almost dangerously frantic. The effects of too much or too little stress in emergency services have been documented by Mitchell (1984). A busy night shift may consist of up to 10 urgent cases, and each of these may be legitimate. The constant pressure to be emotionally "up" for the duration of a shift often takes its toll, and many officers argue that it
3. "YOU WAIT UNTIL YOU GET HOME" 57
takes up to 3 days to recover both physically and emotionally from this. One officer who has a military background described how being "up" for extended periods of time can extract an emotional price off-stage:
We did about six urgent cases in an eight-hour shift recently and each one needed to be urgent. I was up here all the time and I never actually got down. If you do a nine- or ten-hour shift, you just go all the time. You start to get up and up and up and you end up staying there. After fourteen hours, it's a long time to maintain that. When I get home, I just want to veg out.... But everyone else wants company [at home] and I find that hard to cope with at times. It's hard for them to understand, and it's not their fault.
The extraordinary amount of organizational change that has occurred within the DPS in recent years has meant significant changes in the volume of work that an officer is expected to handle. An emphasis on structural efficiency has led the DPS to undertake an efficiency drive, thus resulting in officers having to do more work but with less resources. These changes include a reduction in the number of officers rostered to cover weekend shifts and sick leave, and fewer breaks between cases.
The level of work intensification or "speed-ups" as described by Hochschild (1983) that DPS officers have experienced in recent years are similar to what Hochschild's flight attendants experienced during the 1980s. The DPS officers are now expected to attend more cases during a shift while still providing the same quality of service. This means that there is less time between cases in which to emotionally process the previous job. This workload increase is particularly problematic in urban regions. Officers spoke angrily about the lack of time available to "come down" between jobs, especially from a previously stressful urgent case. Many expressed regret at initially supporting organizational change because they felt it had failed to meet the practical needs of customer service.
Increased expectations to perform more efficiently with less organizational support have ramifications for how officers switch emotionally from one type of case to another. Although most officers make the switch successfully from a nonurgent to an urgent case, the reverse switch is often problematic. Problems arise when officers are sent back out on the road immediately following a difficult emergency code. Officers complained that this was a frequent occurrence, even when the officers involved need to change their clothing. This following quote illustrates how officers negotiate time for themselves in the face of ever-increasing demands to "always be ready":
Like you've done CPR and your car is a mess and they'll be saying on the way back, do you mind doing this job for us and you've got to get on the radio and tell these people, no. I'm not doing it. I'm coming back, and having a shower and I'm going to get changed. So, the way I work it out now is I go and do a case, I do the paper work and I don't call free until I've done that, and if we're run all day long, I will stop and I will have my break.... Whether they know it or not, they'll lose me for 30 minutes. I don't like doing that because there might be an emergency that might be pending, they might be looking for you. But that's the only way I can pace my day.
One officer expressed the view that management often placed unreasonable expectations on officers regarding emotional switching from one type of code to another, particularly in cases where the patient died:
It used to affect me and I used to wonder why I couldn't be demobilized for a time, because it felt quite cold.... It felt quite cold with regards to administration and management, that they could be so callous as to give me another three geriatrics after I've finished a cot death case or something like that. I started to do things to protect myself, to overcome the emotional states, so it doesn't worry me anymore.
What this officer referred to when discussing "overcoming emotional states" is the emotional process work that occurs between codes either by oneself or with a partner.
Coping and Emotional Labor
Despite the fact that officers now are required to have more academic credentials to gain permanent employment, formal training for emotional labor is still a rare occurrence. At the shop floor level, emotional labor may be considered "the best tool to have." However, in terms of resource allocation for training, technical skills are still given primacy. This has implications for how the DPS approaches the whole issue of emotionality. Younger officers who are being trained as technological experts may be better equipped to cope with the intensity of trauma work because they have developed the technical competence needed to accomplish successfully their work. However, the reality is that a vast majority of the workday is comprised of mundane, nonurgent work. It is during these cases that the bulk of the emotional labor is performed. Conversely, it is after major trauma cases that the bulk of emotional process work is conducted. Even after 100 years, the DPS has yet to successfully prepare officers for this.
This narrative of emotional labor within the DPS also illustrates the degree to which the DPS officers switch constantly from frontstage to backstage regions. If emotion were the only major component of the switch between front-and backstage, officers would find this process straightforward and unproblematic (Lois, 2001; Spitzer & Neely, 1992). However, a major part of the switch from one region to the other is the ability to change from what is essentially a "feminine" demeanor in the front-stage region to one that is conven
3. "YOU WAIT UNTIL YOU GET HOME" 59
tionally masculine in the backstage region. Generally speaking, backstage regions include base stations, truck stops, gas stations, and any place where the officer is not "onstage" and performing emotionally for a public.
In the case of the DPS, backstage emotional process work involves a range of practices, including rapid speech patterns, the telling of "war" stories, and the use of black humor. In extreme cases, backstage processing may involve formal critical incident debriefing, which involvesa counselor alerting officers to signs and symptoms of posttraumatic stress. However, many officers find this level of support inadequate for processing much of what an officer experiences throughout a working shift because of the gendered constraints of male-dominated team dynamics, which influence what is acceptable in backstage areas. Explanations for this phenomena are varied and complex and beyond the scope of this chapter (for a full review, see Boyle, 2002). However, officers do receive considerable emotional support from offstage regions, as discussed in detail next.
THE ROLE OF OFFSTAGE SUPPORT IN THE DPS
To state that spouses provide a significant amount of emotional support to DPS officers is something of an understatement. The expectation that the family, particularly the spouse, will provide offstage emotional support is based on the assumption that a certain kind of emotional gender asymmetry exists within heterosexual relationships. As Duncombe and Marsden (1993, 1995) explained, this gender asymmetry is an indicator of a form of emotional power, which is situated within the wider context of continuing gender inequalities of societal resources and power. Although the central feature of men's lives may be the rational, affectively neutral workplace, "women carry the emotional responsibility for the private sphere" (Duncombe & Marsden, 1995, p. 150). This includes the performance of emotion work that sustains the relationship itself.
During earlier DPS administrations, spouses not only provided emotional support, but were also intensely involved in providing ancillary labor to help run the organization. Spouses assisted in raising funds, provided secretarial assistance, and sometimes acted as surrogate counselors, especially in rural and remote areas. Thus, there was an expectation that wives would be as "wedded" to the organization as their husbands (Finch, 1983).
Although this reliance on women's ancillary labor may have been exploitative in principle and often in practice, the benefit of co-opting women into the service meant that the whole family enjoyed considerable status within the community. Conversely, the local DPS officer was "owned" by the community, which meant that officers and their families were expected to adhere to strict moral and behavioral codes. This applied particularly to senior DPS officers, whose role was to care for the ambulance "family," as well as the whole community.
As community role models and representatives, officers stationed in rural and remote areas often found themselves with no emotional support outside their family. Several rural officers interviewed believed that community members viewed them as "above" feelings or emotion. This hero status meant that officers were expected to perform in a superhuman fashion. It was expected that emotional support either was received from their immediate families, or was something that male DPS officers did not need. Thus, officers were only ever the givers of public emotional support, never the receivers. This situation was even more difficult for young officers who had no significant other or spouse living in the community.
Although most officers who work in urban and provincial centers do not experience this constant blurring of the boundaries between work and home, there is still an implicit assumption by the DPS that officers will receive most of their emotional support offstage. In this study the majority of officers interviewed and/or observed were male and married with children. With the exception of one officer, all officers admitted that they would find it difficult to do their job without the support of their spouse. They also admitted that the emotional stress of ambulance work often placed significant pressure on their relationships. In some cases, they cited this as the main reason for marital breakdown and subsequent divorce.
Officers also believe difficulties arise when there is a tension between balancing how much they disclose to family members about work, and the amount of process work an officer chooses to engage in individually. As one officer illustrated, "protecting" family members from the unpleasant and gruesome parts of the job is in itself an extension of the workday. Therefore, emotional labor continues after hours in the form of shielding the family from the extreme aspects of ambulance work. A station manager explained that "shielding" his wife and children from the more unpleasant aspects of his job was important:
I think my wife has to be fairly unique to be able to put up with some of the stuff I bring home. But I don't bring everything home to her. There are certain things that I won't discuss with her. I told her not to look at my textbooks. The pictures that are in them are just too graphic. I find it distressful to look at them myself, particularly if you've got kids the same age as the kids in the pictures in these books.
Officers also reported that spouses become frustrated when an officer chooses to be selective about how much disclosure occurs at home after a shift. As an officer of over 20 years' standing explained, there is a thin line between deciding what the officer thinks is distressing for the spouse, and guessing what the spouse will want to hear:
3. "YOU WAIT UNTI L YOU GET HOME" 61
She can say, I can understand the nature of your job but she can't identify with the job. If I go home and tell her what I've seen, she's got to now deal with that problem. And there are very few women in that context who can deal with that situation. No matter how much they say, I want you to tell me. You've got to deal with a lot yourself and talk to people you can trust but still not totally keep your family out of it. That's a mistake. You're going to fall down in a big way if you do that. You've got to do a balancing act, you have to say, I can tell you this much. I can tell you this happened, but I can't tell you that happened.
Several officers reported that they negotiated "routines" with their spouses for dealing with difficult shifts. It is not surprising then, that the longest serving officers were the ones who were willing to discuss these routines and their relative successes or failures. The success of these routines in defusing anger or frustration depended on the spouse being committed to doing this kind of emotional process work. For example, an ambulance "couple" that has been married for over 20 years has developed a routine whereby the officer "signals" to his spouse if he needs emotional space when he comes home:
Over time my wife had learned that if I walk in and my mannerisms are such, then she'll just walk out the other door. She'll just go away and leave me alone for an hour or two. Until I've processed it and dealt with it and put it where it's supposed to go. Then I'll walk out and we'll start from there.
Learning when to "go away" is an expectation with which many spouses have extreme difficulty coping, particularly younger couples with children. One young officer stated that his spouse "hounded" him at the end of every shift to tell her everything that had happened during the day. When he refused to comply, she became frustrated and this often led to marital conflict. The officer's main problem with this situation was that he had little private space in which he could process the day's events and emotions.
By their own admission, many officers became intolerant and irritable at what they perceived as the triviality of life outside the DPS. Requests from spouses to behave like "average" men outside of work and within the family environment, together with shift work and the ensuing irregular sleeping and eating patterns, led to conflict about issues of support. One officer who had recently divorced after 15 years of marriage admitted that his inability to cope with the emotional switch from work to home was one of the major contributing factors to the breakdown of his marriage:
Why am I the Mr. Fixit for everyone? ... I've fixed up fifteen patients' problems today, and then I come home and she's got a migraine and she wants me to cook dinner, and I think, no! I've clocked off. You cook dinner. You get very intolerant.
In addition to providing direct emotional support, spouses often have to deal with the officer frequently being absent from social gatherings. This can lead to tension within the extended family, as the spouse has to do the emotion work of allaying fears as to why the officer is not in attendance. Although the absent father is not all that unusual at school functions, for an officer involved in a traditional heterosexual relationship, continual absences from family events may cause problems as the family may view the marital relationship as dysfunctional or abnormal.
Although there is no evidence to suggest that DPS officers have a higher than average divorce rate, there is some validity in suggesting that officers who have young children are under considerable offstage pressure compared with officers who have served for longer periods and whose families are older and more established. It is argued here that younger, married, and rural officers have the greatest need to develop sound emotional process skills, for it is these groups that may be at greatest risk of emotional exhaustion if their offstage support is inadequate or deficient. Offstage emotional process work may involve something as simple as a short amount of time and space away from family and friends, or it may be as complex as a close relationship with a religious minister or health professional such as a psychologist to help them deal with their emotional conflict (Boyle & Healy, 2003).
This chapter attempted to discuss the emotional conflict experienced during emotional process work, and the role of offstage support. Specifically, it examined a study involving the DPS, an emergency service organization where emotions are a central part of the job.
There are a number of reasons why the DPS has not fully recognized and compensated the emotional component of a paramedic's work. The DPS was, and still is, a "harsh" cultural entity, consisting of a paradoxical mix of cultural values and practices, such as display of "softer" emotions to the public (i.e., compassion, empathy, and cheerfulness) and an expectation that officers will not display emotional expressions of grief, remorse, or sadness. In other words, officers were expected to keep up a constant "masculine" demeanor of self-control and stoicism, while simultaneously presenting a "caring" and "feminine" demeanor to patients.
In terms of emotional expression, this expectation is not harsh. This expectation becomes problematic when the organization deems that only a narrow range of feelings can be expressed publicly. Officers who express emotions outside the acceptable range risk harassment and ostracism.
Despite this, officers were generally positive about performing emotional labor. This is in contrast to Hochschild's (1983) findings, where the flight attendants spoke of emotional labor as a burden that often led to a lack of connection with others. Of course, there are specific reasons why there is this dif
3. "YOU WAIT UNTI L YOU GET HOME" 63
ference between the two groups, and some of them have little to do with emotional labor itself. For example, flight attendants are prone to suffer dislocation because of the travel involved. Spending many nights away from home, constant disruption of personal routines, and the stress of flying all combine to create a sense of alienation. In contrast, ambulance officers do not experience the same level of disruption or dislocation, despite the similarities of performing emotional labor.
DPS advertising campaigns explicitly promote men as carers. Many of the posters and television advertisements show male officers in caring poses such as holding the hand of a child or an older person. Indeed, the fact that the DPS has a baby carseat rental service attests to their commitment to be seen as a caring organization. However, men are not usually associated with the act of caring in public. Therefore, officers have to reconcile this traditionally feminine act with the fact that they are still men. Men benefit emotionally from the traditionally patriarchal arrangements that separate home and work. In an orthodox Marxist sense, this means that men are able to sell their labor power because they are able to reproduce emotionally within the private sphere (Milkie & Peltola, 1999). However, this arrangement is contingent on women being available to engage in emotion work at home. The DPS is heavily dependent on this public/private arrangement as a means of providing informal and offstage emotional support to male officers.
As the DPS culture is strongly adherent to hegemonically masculine practices and ideologies (see Connell, 1995), it would be fair to assume that the majority of male DPS officers are heterosexual and practice a conventionally heterosexual lifestyle. Apart from student officers, all officers interviewed or observed were married with children. In most cases. the male officers were the principal earners in the family. Many officers commented on the level of emotional support they received from their spouses, and most believed they could not do their work efficiently if this support was unavailable.
According to some officers, when they are experiencing marital problems, their work performance diminishes. If they are unable to secure emotional support from elsewhere, they are likely to be more vulnerable to emotional stress at work. The DPS then is almost totally dependent on officers receiving emotional support from outside of the organization. The distinction between work and home then becomes blurred as spouses, partners, and occasionally good friends and relatives assume the role of employee assistance counselor. They are, by default, "hidden" employees of the DPS.
Therefore, the DPS is ill-equipped to cope with a rise in the number of men and women who lack offstage emotional support. The individualistic and masculinist cultural framework within which the DPS operates is unable to provide the kind of emotional support needed by most officers, because as a masculinist culture, it does not recognize that men need public as well as private forms of emotional support. The culture is not tolerant of the existence of
an emotionally complex paramedic, because to do so would require the acknowledgment of multiple masculinities (see Connell, 1995).
DPS officers are allowed to "step out of" the traditional hegemonic frame within the confines of performing emotional labor. However, emotional labor does not terminate when patients leave the hospital. It has process characteristics and extends beyond the boundaries of the narrowly defined service provider-client relationship (Dodier & Camus, 1998, as an example). Failure by the DPS to acknowledge the need for officers to have the time, space, and permission to engage in emotional process work as a valid part of ambulance work and often results in increased levels of emotive dissonance and emotional exhaustion (Thoits, 1991). This does not always occur as a result of the interaction with the patient, but can occur when the officer realizes that the organization objectifies his or her emotional self. This happens when the DPS fails to acknowledge that part of its responsibility to its organizational members is to take into consideration the time needed to replenish the mind, body, and spirit when developing operational procedures and policies (Boyle & Healy, 2003).
Thus, the DPS has privatized this part of its responsibility to officers by asserting that emotional process work is something that should occur primarily in the private sphere of the home and family. Given the state of flux that familial arrangements are in, this strategy holds considerable risk for the organization in question. The DPS has based its strategy on the assumption that most officers are involved in orthodox hegemonic heterosexual relationships where the female spouse sees it as her role to manage the emotions of her male partner.
If support cannot be found within the private sphere, and if traditional male friendships prove to be inadequate in terms of providing emotional support, then this can lead to officers neglecting to engage in emotional process work.
The quote at the beginning of this chapter illustrates nicely the connections between front-stage expectations, and back- and offstage realities. The front-stage emotional culture deems that officers "cope" under all circumstances, regardless of the amount of emotional pressure placed on officers. As representatives of the DPS, the community expects officers to perform at their emotional peak with patients—yet from an organizational perspective, maintaining emotional health is an individual responsibility. Officers have to usually wait until they get home to deal with emotion and receive support, which indicates the lack of recognition classifying emotional process work as actual "work."
Although the DPS is an organization literally saturated with emotion, emotional support for DPS officers is mainly viewed by the organization as something that needs to primarily occur off-site and not the responsibility of the organization. Although emotional process work occurs during organizational
3. "YOU WAIT UNTI L YOU GET HOME" 65
time, particularly in backstage regions, a closer inspection of the structure and culture of the organization indicate a serious level of emotional denial. Anecdotal evidence suggests that employees appear more likely to engage in acts of resistance if an organization fails to either recognize or legitimate the consequences of emotional labor for front-line employees. In turn, this has implications for the overall level of individual as well organizational well-being.
This page intentionally left blank
The Role of Emotion in Employee Counterproductive Work Behavior: Integrating the Psychoevolutionary and Constructivist Perspective
Pamela L. Perrewé
We argue that many forms of counterproductive work behavior (CWB) can be understood better through an examination of individuals' emotional adaptive efforts trying to survive in the organizational setting. An emotion-based, process model of CWB is developed based on two emotion theories: the psychoevolutionary theory and the constructivist view of emotion. We view emotion as a chain of cognitive, emotional, and behavioral reactions to stimuli events, and discuss several "problem points" in the emotion process that may lead ultimately to CWB. The chapter concludes with a discussion of organizational implications and suggestions for future research and practice.
Workplace aggression and violation, employee theft, sabotage, and many other forms of employee counterproductive work behaviors (CWB) have drawn increasing theoretical attention among management researchers in recent years (Spector & Fox, 2002). In the literature, CWB has been given different names and definitions, including deviant workplace behavior (e.g., Hollinger, 1986), antisocial behavior (e.g., Robinson & O'Leary-Kelly, 1998), and dysfunctional behavior (e.g., Collins & Griffin, 1998). Although some have examined CWB in general, others have focused on specific types of behaviors such as aggression and violence (e.g., Fox & Spector, 1999; Greenberg & Barling, 1999), employee theft (e.g., Greenberg, 1990), and sabotage (e.g., Giacalone & Knouse, 1990). There are also differences in researchers' emphases on either individual differences or situational factors as primary antecedents of CWB (for a more complete and detailed review, see Spector & Fox, 2002). Some theoretical perspectives that have been used include social learning theory (Bandura, 1977), justice theory (Greenberg, 1990), and attribution theory (Martinko, Gundlach, & Douglas, 2001; Martinko & Zellars, 1998).
It is worth noting that, despite the increasing volume of research on this topic, there has been relatively little effort exploring the role of emotion in employees' conducting CWB. Imagine an extreme situation in which a person has no feelings of anger. Facing a colleague's insulting language or action, he or she might simply inquire why the colleague would have done so, and then rationally find out where the problem lies and try to solve it. Chances are that the insulting behavior will be stopped and the problem solved without any major influences on the subsequent interaction between the two (although in some instances the apparent calmness of the target may even further irritate the perpetrator, such as when he or she interprets the calmness as arrogant or pretending). Unfortunately, in daily organizational life, insulting behavior often leads to strong feelings within the target, such as guilt, shame, and anger (Gabriel, 1998). Negative emotions such as anger often motivate retaliations such as attacking back in the same manner, which is often how CWB is induced in organizations (cf. Folger & Skarlicki, 1998). It is not hard to imagine that employees who conduct such violent behaviors as shooting a supervisor or physically attacking a coworker must have undergone some strong inner emotional experiences. Of course, it is fair to say that not all CWBs have to do with emotions. For example, checking personal e-mails during working hours or work avoidance often have more to do with poor work ethics than emotions. However, it is still quite possible for emotions to play a significant role. For example, intentionally slowing the pace of work may well involve strong emotions such as anger toward the supervisor.
Our review of the literature revealed some research addressing the role of emotion in CWB. Spector and Fox (2002) argued that emotions play a central role in CWB. Specifically, they proposed that negative emotions are likely to lead to CWB, especially when employees are exposed continuously to emotion-arousing events. Martinko and colleagues (2001) argued that feelings of guilt, shame, anger, and frustration mediate the relationship between individual attribution of perceived disequilibrium in the workplace and subsequent CWB. In their empirical studies, Fox and colleagues (Fox & Spector, 1999; Fox, Spector, & Miles, 2001) found that negative emotions serve as a mediator through which employee's experience of situational constraints induces CWB. This line of research, however, left unexplored exactly how the emotional mechanism functions as is involved in CWB. In addressing the gap in the current literature, we draw on two influential theories of emotion and propose a process model of CWB. We believe to identify the emotional antecedents of CWB will help not only to understand it better, but also to control it effectively, and, further, to enhance individual and organizational well-being.
4. COUNTERPRODUCTIV E WOR K BEHAVIOR 69
CWB has been defined as behaviors that have a direct and visible negative effect on either organizations or their members (e.g., Spector & Fox, 2002). With a few exceptions (e.g., Martinko et al., 2001), passive forms of CWB (those directed to self such as drug use and depression) have been ignored largely. However, it seems appropriate to include such behaviors in the CWB category because they tend to influence negatively productivity and will, over time, have a corruptive influence on organizational climate (Robinson & O'Leary-Kelly, 1998). It is reasonable to expect that such less overt forms of CWB could at times become equally harmful to both the organization and the individual. Therefore, for the purpose of this chapter, we adopt a broad definition of CWB to include all employee behaviors, intended or unintended, that are directed at the organization, other organization members, clients and customers, or self, and that have the potential to hurt organizational productivity and well-being. Thus, according to our definition, behaviors such as expressing fake emotions all the time, or consistently working overtime without attending to health problem, may also belong within the CWB category.
The rest of the chapter is organized as the following: First, we introduce the psychoevolutionary theory (Plutchik, 1980) and the constructivist view of emotion (Averill, 1980). We then discuss how these two theories can be integrated meaningfully and applied in an organizational setting. Next, we present a process model, drawn from the psychoevolutionary theory, which explains CWB from an emotion-based perspective. The purpose of this chapter is to provide a theoretical framework within which CWB is understood better.
Although research on emotions in organizations has been a newly immerged area of interests to most organization researchers (Ashforth & Humphrey, 1995; Fineman, 2000), systematic research on human emotions can be traced back to at least a century ago when William James (1884) asked, "What is an emotion?" Since then, numerous emotion theories have developed in the literature. Ashforth and Humphrey (1995) identified two streams of research in the field of emotions that take very different approaches, the naturalist perspective and the social constructivist perspective. According to them, the naturalist perspective takes a biological perspective, examining emotion based on the physical reactions and individual genetic attributes; the social constructionist perspective posits that emotions are learned, and they take on socially defined meanings and rules in the process of learning. Both approaches have developed a useful framework in explaining emotions, but do so with very different foci. Many researchers (e.g., Fineman, 2000; Lazarus, 1991; Porter & Samovar, 1998), however, believe that a sound theory of emotions should look to the interplay of the two because neither of the two provides a complete understanding of human emotions. Following this school of thought, we built our theory based on the belief that the two poles of theorizing on emotion are not necessarily contradictory rather they compliment each other.
Psychoevolutionary Theory of Emotion
Psychoevolutionary theory of emotion, developed by Plutchik (1980), is one of the biological oriented theories of emotions. Building on Darwin's early work on the expression of emotions of human beings and animals, Plutchik (1980) posits that emotions serve as a basic function of survival by helping to organize animal and human behavior in a way that is appropriate to the immediate environment. Emotions are described as having a purpose, either unconscious or deliberate. Plutchik (1980) proposed that the purpose of emotion is adaptation, and that emotions are activated in an individual when issues of survival are raised in fact or by implication. He argued that the effect of the emotional state is to create an interaction between the individual and the event that precipitates the emotion, which usually serves to reduce the disequilibrium and reestablish a sense of equilibrium (Plutchik, 1980, 1989).
From an evolution point of view, Plutchik (1980) argued that emotion should not be treated as simply a feeling state. Rather, it is understood best when viewed as a chain of complex and loosely connected cognitive, emotional, and behavioral responses (i.e., emotional chain) to a stimulus event. As a major part of his theoretical development, Plutchik (1980) proposed that the occurrence of certain stimuli events is followed by cognition, which in turn will be followed by introspective feelings; the feeling state is only one step in the emotional chain and will serve to motivate certain goal-directed behaviors; when successfully conducted, such behaviors will lead to desired consequences (Plutchik, 1980). For example, when a man encounters a dangerous animal, the cognition of danger will motivate the inner feeling of fear, which prompts him to run and, consequently, protects the person from being attacked.
Plutchik (1980) argued further that there are a number of problem points in the emotional chain which disrupt the smooth flow of the emotional chain. First, the initial cognition may be in error so that the threat is misperceived or misinterpreted. Second, the feeling generated by the cognition can be blocked, modified, or distorted. Third, appropriate reactions to feelings may (or may not) occur due to either environmental or internal restraints, or both. Finally, depending on whether appropriate and effective behavior occurs, the goal or purpose of the emotional chain may (or may not) be served. If the desired purpose or goal of the behaviors is not fulfilled, the unaccomplished goal will serve as new stimulus to motivate further actions to reduce the imbalance (Plutchik, 1980). For example, we discussed previously the man encountering the dangerous animals; if he runs away but cannot find any protection, he may begin to cry. Here, crying is an emotional expressive behavior, motivated by the felt emotion of fear, probably with the purpose of release the inner tension.
4. COUNTERPRODUCTIV E WOR K BEHAVIOR 71
Plutchik's (1980) approach to studying emotions provides a useful way from which CWB can be examined. Later, we discuss this framework and note that many forms of CWB have an emotion element and are people's conscious or unconscious efforts to adapt to the environmental stimuli. These efforts turn out to be counterproductive only when either the organization's or the individual's well-being is (or will be) harmed. However, before this is discussed, we modify the framework to ensure that it is suitable for our discussion and application in the organizational setting.
One element to be added to the framework before the psychoevolutionary perspective can be applied readily to the organizational setting is the influence of the social context. Focusing on the biological side of emotion and trying to explain human and animal behaviors within a general framework, Plutchik (1980) deliberately limited his discussion regarding the influence of social environment on human behavior only. Although he did mention briefly the mechanisms of emotional regulation in his theory, he did not go so far as to incorporate them into the general framework. For example, he gave a description of how a little boy who gets angry with his mother may seek other ways (such as kicking his dog) to release the anger instead of expressing his feelings to the mother. Plutchik (1989) also talked about the role of social institutions in controlling the expression of emotions—for example, how anger and aggression are channeled into socially acceptable pathways through the social institutions of sports and war.
In summary, from the psychoevolutionary theory perspective, emotions are survival mechanisms based on evolutionary adaptations. Emotions are best described as a complex chain of events with stabilizing feedback loops that produce some kind of behavioral homeostasis (Plutchik, 2001).
The Constructivist View of Emotion
Human behaviors can become much more complex when it comes to the social setting. That is predominantly why researchers from social psychology and sociology have critiqued the biology-oriented examination of human emotions. Thoits (1989) argued that the strict biological explanation of emotions would lose its meaning and significance when one looks at the complex social context within which emotions take place. This is consistent with several other perspectives for theorizing in emotions (e.g., Hochschild, 1983; Kemper, 1984; Shott, 1979; Zurcher, 1982), one of which is the constructivist view of emotion (Averill, 1980) is one of the examples.
From a sociological viewpoint, Averill (1980) argued that human emotions are socially constructed and inherit important social meanings. He viewed emotions as transient roles that individuals take in the drama depicted by the society within which individuals are involved. Thus, the social norms provide shared expectations about appropriate behaviors and exert powerful influences on individuals' actual behaviors. From this perspective, emotions are best described as roles, which are both socially constructed and individually enacted. Individuals interpret actively their own emotions based on the explicit or implicit rules and norms provided by societies. Further, individuals act based on these interpretations, and expect others to interpret the emotions using similar rules accordingly. Thus, emotions also serve as signals of appropriate actions under certain social contexts (Averill, 1980; Zurcher, 1982).
Averill (1980) emphasizes the passion characteristic of emotion, in contrast to emotion as an action. In other words, emotions are not actions that people take. Rather, people don't usually have control over their emotions, and indeed, sometimes they are overcome by them.
In his research, Averill (1980) also discussed the possible influence of conflict between personal norms (i.e., individual expectancies about one's own behavior) and social norms (i.e., shared expectancies about appropriate behavior) in shaping individual emotional responses and experiences. He argued that behaviors that are consistent with personal norms but not social norms are liable to be deemed inappropriate, and that well-socialized individual norms reflect social norms. Thus, it is reasonable to expect that individuals will seek actively emotional cues that are appropriate in certain social contexts in an effort to match their personal norms with social norms.
In summary, it is evident that from the social psychological perspective, emotions are built on one or more biological systems of behaviors, and the meaning of emotion, or its functional significance, is embedded primarily in the sociocultural system (Parkinson, 1996; Thoits, 1989). Thus, it is necessary to consider the broader social context within which the emotions take place to fully understand individuals' emotional behaviors.
APPLYING A SOCIAL PSYCHOEVOLUTIONARY
PERSPECTIVE TO THE ORGANIZATIONAL SETTING
Upon closer examination, it is evident that, despite the different emphasis of the two approaches discussed previously, they have interesting similarities. First, both the psychoevolutionary theory of emotion and the constructivist view argue that emotion is not a simple feeling state. Specifically, both theories agree that cognitive and behavioral elements in emotions. In the psycho-evolutionary theory of emotion, this is evident from the argument of emotions as a chain of cognition, feelings, and behaviors. From the constructivist view of emotion, this is evident in Averill's (1980) argument that emotions are interpretations and roles. This is a critical understanding. On one hand, this underscores the importance of understanding emotions within the context rather than isolated feeling states. On the other hand, it also implies that emotions
4. COUNTERPRODUCTIV E WORK BEHAVIOR 73
are important in the understanding of cognitions and behaviors (Fisher & Ashkanasy, 2000a; Fineman, 2000; Weiss & Cropanzano, 1996).
Second, both theories agree that emotion has a survival value, although they may differ on the meaning of survival. It is evident that the psychoevolutionary theory of emotion emphasizes the individual's physical and psychological experiences, whereas the constructivist view of emotion emphasizes the social facet of human emotions and describes the individual as an active role taker in the social context. At the same time, the psychoevolutionary theory of emotion looks at survival at its basic biological level, whereas the constructivist view of emotion adds to survival a social meaning, namely, the survival of self or the self identity. When examining individual behaviors in contemporary organizations, however, the second meaning of survival becomes more salient. For example, dying from hunger may not be as significant a concern for an employee as losing a job from which he obtains a sense of belonging and meaning. The latter case and other concerns, such as belonging to certain social groups (e.g., middle-class, professionals), having certain reputations, and being loved and well accepted, are usually the meanings of survival in the modern social setting. In fact, Plutchik (1980) acknowledged identity as a fundamental existential crisis for all organisms and acknowledged it as a more complex issue when applied to human behaviors.
Therefore, it appears that two of the most diverse theories of emotions within the literature share common ground and tend to differ primarily in focus and emphasis. Thus, we propose that by integrating the psychoevolutionary theory and the constructivist view of emotion, a social psychoevolutionary perspective can be applied readily to the organizational setting.
We put the psychoevolutionary theory in the psychodynamic context that the constructivist perspective implies, and developed a model of CWB from the emotion perspective (see Fig. 4.1). In daily organizational life, individuals are exposed to work events or other stimuli events occurring outside the workplace (i.e., life events). Upon encountering the stimuli events, individuals perceive information selectively in ways prescribed by situational factors and individual characteristics; in this process some information may be lost during the perception and interpretation process, and some other information may be misinterpreted (Motowidlo, 1986; Weick, 1995). These perceptions influence further the feelings of individuals, and a disequilibrium feeling may result, which often leads to emotional tension.
The direct link between stimulus events and feelings indicates that it is also possible for stimuli to lead directly to inner feelings (Zajonc, 1984). It is important to note that there is still an unresolved debate as to whether cognitive appraisal is a necessary precondition for elicitation of emotion (Lazarus,
FIG. 4.1. An emotion-based process model of counterproductive work behavior.
1984), or whether there are circumstances under which affect may precede cognition (Zajonc, 1984). Although many argue that cognition is a necessary precondition (Lazarus, 1991), or an intertwining part (Solomon, 2000), of emotion, some have shown that emotions could be induced without the involvement of cognition (Zajonc, 1980). It has also been found that emotion can influence cognition and behaviors (Staw & Barsade, 1993). For example, research has found that emotional states may affect an individual's cognition ability (Isen & Means, 1983) and social orientation (Staw & Barsade, 1993; Staw, Sutton, & Pelled, 1994). Taking all this literature into consideration, we depict the relationship as non-recursive to show that cognition and feelings may influence each other in a cycling manner.
To rebuild the emotional balance result from stimuli events or cognitive appiaisal, individuals actively involve themselves in behaviors (including emotional expression) that are intended for certain desired effects. In certain situations, discussed later in this chapter, the behaviors can be counterproductive in nature. The behavioral reactions to inner feelings can then lead to consequences, which may (or may not) be the desired effects or goals. Depending on whether the purpose is satisfied, another episode of emotional experiences may (or may not) follow.
The arrows (pointing left) from "perceived effect" back to the stimulus event, cognition, feelings, and behavior represent situations in which the de
4. COUNTERPRODUCTIV E WORK BEHAVIOR 75
sired effects are not fulfilled. These unfulfilled goals may serve as new stimuli events that invoke a entire new chain of emotional reactions, or induce new perceptions, feelings, or behavior reactions that may lead to CWB. Theoretically, these feedback loops in the emotional chain only end when the desired goals (either the original or the adjusted ones) are achieved. To better illustrate the point, an example is in order. Imagine an employee who failed to get an expected promotion (stimulus event). Imagine further that she interpreted it as unfair (cognition) and felt angry (feeling). In a department meeting, therefore, she "caused a scene" by publicly offending her supervisor (behavioral reaction) to try to release her emotional tension (desired effect). Rather than regaining her emotional balance, however, she may instead be punished by the organization (undesired effect). The punishment then may serve as a new stimulus event that intensifies the worker's perception of unfairness (new cognition), which induces her subsequent feelings of frustration (new feeling) and sabotage (new behavioral reaction).
It is important to note that the entire process of the emotional chain is influenced by organizational, situational, and individual factors (cf. Averill, 1980; Parkinson, 1996; Thoits, 1989; Hochschild, 1983). For example, organizations often have explicit requirements or implicit expectations for employees' emotions in the forms of feeling rules, defined by Hochschild (1983) as the social guidelines for individuals as to what emotions to have in certain situations. Situational factors, such as being in a football game or a department meeting, also provide cues to individuals' emotional expressions and behaviors, where individuals consciously get themselves "psyched up" or "calmed down" (Zurcher, 1982). Overall, the social influences both enrich and constrain individuals' emotional experience. They enrich emotional experiences by providing the tacit knowledge on how things should be perceived, giving cues that indicate which emotion and behavior are socially appropriate. At the same time, they constrain emotional expression by imposing social rules that individuals must follow in order to survive. In addition, individual differences in such factors as attribution style and emotional stability (Martinko & Zellars, 1998) also influence how stimuli events are interpreted and reacted to. In all, the conceptual model posited that the emotional chains that are invoked by stimuli events and shaped by the organizational, situational, and individual factors serve as critical antecedent of CWB.
According to the psychoevolutionary theory of emotion, the adaptive function of emotion depends on the smooth flow of the entire emotional chain. In situations where links between two constructs are distorted, the entire process from stimulus to behavior and subsequent effect becomes problematic. Unfortunately, in modern organizational settings, there are many factors such as
intense competition and large-scale environmental change (Cascio, 1995) that may hinder or block the adaptation process, as is discussed in detail in later sections.
At each of the linking point in the proposed model two possibilities coexist, both of which may lead to CWB. The first occurs where the link is hindered, blocked, or distorted. For example, when a feeling of anger toward a customer cannot be expressed due to organizational policies, it is then redirected toward organizational property or coworkers. The second is when the chain functions smoothly but in a way that harms the organization. For example, an outburst of extreme anger in the case just mentioned may be due to the accurate appraisal of the situation (e.g., abuse by the customer) and be effective in helping the person regain emotional balance (i.e., fulfill the goal of emotions). However, such expression of anger clearly has negative implications for organizational outcomes (e.g., customer satisfaction and retention). Thus, even adaptive emotions may lead to CWB.
A Taxonomy of CWB Based on Emotion
To understand better how emotional adaptation and maladaptation affect CWB, we have developed a taxonomy of CWB. It should be noted that the emotional chain might be adaptive and functional from the perspective of individual, but not necessarily be so from the social perspective of the society, which is, in our case, the organization.
As indicated in Fig. 4.2, based on the notion of emotional adaptation, CWB can be classified into four different categories according to its individual and social implications. We propose that the conventionally defined CWBs, such as abuse of others, threats, work avoidance, and sabotage (Fox et al., 2001), are behaviors that are adaptive from the individual's perspective but maladaptive from the social perspective. However, there are three other forms of CWB that are understudied. First are those CWBs that are maladaptive to both individual and society, including self-destruction, drug use, and depression (as shown in the bottom left quadrant in Fig. 4.2). Second are those that are maladaptive individually but seem to be adaptive socially at the surface level (see the bottom right quadrant in Fig. 4.2). Such behaviors include passive emotional regulation behaviors such as suppression of negative emotional expressions (Gross, 1999) and surface acting when performing emotional labor (Grandey, 2000; Hochschild, 1983). This type of behavior is maladaptive in the sense that it may be harmful for the physical and psychological well-being (Grandey, 2000; Gross, 1998b; Hochschild, 1983), as well as the cognitive ability, of individuals (Gross & Levenson, 1997; Richards & Gross, 2000). Finally, CWBs that are adaptive both to the individual and the immediate social groups (see upper left quadrant in Fig. 4.2). This includes deviant behaviors (e.g., stealing) attempting to adhere to certain organizational cultures or group norms. Viewing CWB
4. COUNTERPRODUCTIVE WORK BEHAVIOR
Emotional adaptation from the social perspective
FIG. 4.2. A taxonomy of CWB as employee emotional adaptation behaviors.
from the emotional adaptation perspective highlights that these three forms of CWB should be emphasized in organizational research since they could be harmful equally to the individual and the organizational well-being.
In the preceding subsection, we discuss situations where a linking point in the emotional chain becomes problematic, which, we propose, is how CWB is induced. It is important to understand that although CWB can be induced when only one of the linking points become problematic (i.e., each problem point can lead directly to CWB), it is through the mechanism of the entire emotional chain that the antecedents of CWB function. Thus, we cannot understand fully the process of the influence without examining the complex chain of emotional adaptation. In other words, viewing CWB from this perspective helps to better understand CWB.
Stimuli and Cognition
The organizational context has a strong influence over individual behaviors. Events occurring daily in the workplace can serve as important antecedents of strong emotional and behavioral reactions (Weiss & Cropanzano, 1996). Injustice events are likely to induce CWB (Greenberg & Barling, 1999). For example, supervisor's emotional abuse of subordinates has been found to be associated with pervasiveness of fear and breakdown of employees (Harlos & Pinder, 2000). Many extreme cases of workplace aggression and violence also seem to be direct responses to workplace injustice (Cropanzano & Greenberg, 1997). We argue that organizational injustice should be related positively to CWB through cognition and/or feelings.
However, it is usually not the objective event itself that serves as the immediate cause of CWB. In the organizational setting, people actively construct their own realities based on the limited information readily available, and through the socialization and collective sense-making with other organizational members (Weick, 1979). Thus, it is important to examine individuals' cognitive appraisal of a stimulus event beyond its objective attributes. Both individual differences (e.g., attribution style) and contextual factors (e.g., environmental uncertainty) influence individuals' cognitions. We discuss both factors in detail next.
Attribution Style. Attribution style is a traitlike individual characteristic that directs the individual's attention when one makes causal reasoning. Attribution styles influence individuals' appraisals as to their relationship to the situation. For example, individuals who have an external attribution style tend to attribute success or failure to the environment; in contrast, those with an internal attribution style tend to attribute success or failure to themselves. There is evidence that attribution style influences the relationship between organizational frustration and CWB, such that, in reaction to frustration, individuals who tend to make external attribution are more likely to sabotage than their internal counterpart (Storms & Spector, 1987). It has also been proposed that individuals who have an external attribution style are more likely to exhibit aggressive and violent behaviors as a result of aversive outcomes than employees who tend to make internal attributions (Martinko & Zellars, 1998). Thus, it is reasonable to expect that individuals' attribution style will influence their cognitive appraisal of events.
Environmental Uncertainty. The modern work environment is characterized by constant changes, which has resulted in additional pressures both on organizations and individuals (Cascio, 1995; Greenberg & Barling, 1999). On the one hand, changes bring about a high level of uncertainty, and therefore the need to process more information within a constrained time limit. On the other hand, information gathering and processing become problematic due to the limited cognitive capability of individuals (Simon, 1997). For example, research indicates that during threatening situations, individuals, groups, and organizations tend to become more rigid and rely on less information for decision making (Staw, Sandelands, & Button, 1981). In addition, there are situations where management feels it is necessary to withhold information from the employees for a certain period of time, which makes it more difficult for em
4. COUNTERPRODUCTIV E WOR K BEHAVIO R 79
ployees to fully comprehend the actual situation at the time the stimulus event occurs. Situations simultaneously involving information overload and a lack of information may increase the possibility of misinterpreting certain events, which can generate further frustration and stress among employees (Spector, 1997). Increasing stress within the workplace is associated with CWBs, such as theft (Greenberg, 1990), interpersonal aggression, and sabotage (Chen & Spector, 1992). However, misinterpretation or distorted perception is also possible in less stressful situations. For example, role overload and role ambiguity may influence individuals' feelings of control over situations and affect further how they perceive situations (Perrewe & Ganster, 1989). Thus, we expect that individual differences (e.g., attribution style) and situational factors (e.g., uncertainty) affect the relationship between the stimulus event and the cognition such that the misinterpretation or distorted perception of work or life events can occur and lead to CWB. As an example, individuals with an external, pessimistic attribution style are more likely to interpret stimuli events negatively. Further, situations that are characterized with uncertainty and stress are likely to be related positively to employees' misinterpretation of stimuli events.
Cognition and Feeling
Even with the stimuli being perceived as fully and as functionally as possible, the cognition may not lead to appropriate feelings that are both adaptive to individuals and beneficial to organizations. The reasons are twofold. First, the adaptive feelings may not be elicited from the perception. Second, when the adaptive feeling is elicited within an employee, it may not be beneficial to the organization. We discuss each scenario next.
There are a number of factors that will influence people's elicitation of emotions. In the extreme case, individuals may lack certain kinds of emotions, such as love or fear. For example, Damasio (1994) reported a subject who lost his ability to get in touch with his own emotions. More commonly, being in certain moods may also predispose individuals to feel certain emotions and not others. For example, an individual in a bad mood may be irritated more easily than one who is in a good mood. From a social norm perspective, strong display and feeling rules may influence and, over time, guide individuals to learn not only what emotions to express but what emotions to experience (Hochschild, 1983; Scherer, 1986; Zurcher, 1982). For example, people may initially find it appropriate socially to express sadness when attending a funeral, and over time the perception of a funeral setting will actually generate feelings of sadness inside the person. In a work setting, when a supervisor reprimands publicly an employee for a misdeed, even if he or she believes that the comments are justified, the simple fact of being criticized in public may generate a feeling of shame or humiliation. Feelings of shame may motivate further retaliation behaviors toward the supervisor. Thus, we expect that individual factors (e.g., disability) and social and organizational norms affect the relationship between cognition and feelings such that an accurate perception may still lead to an inability or unwillingness to elicit or express appropriate feelings, which may further result in CWB.
A more typical case in organizations is when individually and psychologically adaptive emotions are expressed with a potentially negative implication for organization, which is especially true when the emotions are ones commonly viewed as negative (e.g., anger). There are several situational and individual psychological factors that may induce employee negative feelings.
Injustice. Perceived unfairness is an important situational factor that induces negative emotions (Spector & Fox, 2002). Two types of justice have been frequently discussed, procedural and distributive justice. Procedural justice refers to the degree to which procedures are perceived as fair in decision making and resource allocation. Distributive justice reflects the perceived fairness of the rewards employees receive for their performance inputs. Perceptions of injustice are associated with counterproductive behaviors such as employee theft, withdrawal, aggression, and other forms of CWB (Fox et al., 2001; Greenberg, 1990; Greenberg & Barling, 1999). Research illustrates that distributive justice also induces employee theft in striving to restore a balance between their rewards and contributions to a job (Greenberg, 1990, 1993). As suggested by O'Leary-Kelly, Griffin, and Glew (1996), workplace violence may ensue when employees perceive valued outcomes (e.g., promotions, compensation) as having been distributed unfairly. In contrast, research also illustrates that fair procedures can minimize the dissatisfaction resulting from poor outcomes (Greenberg, 1990) and can promote employee citizenship behavior (Organ & Ryan, 1995).
Violation of Psychological Contract. Psychological contracts refer to employees' beliefs about the reciprocal obligations between themselves and their organization (Rousseau, 1989). It is argued that that the content of psychological contracts is not always clear and may become difficult for organization to fulfill, as when organizations undergo dramatic changes such as corporate restructuring, large scale downsizing, and increased reliance on temporary workers (McLean Parks & Kidder, 1994; Morrison & Robinson, 1997). Although such changes are becoming increasingly necessary for organizations (Cascio, 1995), it may result in various levels of employees' perceptions of violation of psychological contracts by the organization (McLean Parks & Schmedemann, 1994; Robinson, Kraatz, & Rousseau, 1994). Similar to distributive injustice, violation of psychological contracts may also lead to the employee engaging in behaviors such as theft or sabotage in order to "get even" (Morrison & Robinson, 1997).
4. COUNTERPRODUCTIV E WOR K BEHAVIO R 81
Loss of Control. Perceived control is a critical concept in coping with stress. Research has found consistently that high level of perceived control leads to better task performance and a lower level of felt stress appraised (Averill, 1973; Langer, 1975; Thompson, 1981; Thompson, Armstrong, & Thomas, 1998). Unfortunately, within organizations, changes such as technology modernization and large-scale downsizing often introduce considerably high levels of job insecurity, therefore resulting in feeling of loss of control for employees (Greenberg & Barling, 1999). Possible consequences of a loss of control include feelings of powerlessness, loss of identity, anxiety, and stress, which may result in aggression or other forms of CWB in an effort to regain feelings of control. Thus, we argue that perceptions of injustice, violation of psychological contract, and loss of control will affect the relationship between cognition and feelings. Specifically, when feelings of injustice, a violation of a psychological contract, and loss of control are high, perceptions of an event may lead to more negative feelings, which will induce CWB.
Feeling and Behavior
Felt emotions bear strong influence on subsequent behaviors (Weiss & Cropanzano, 1996). Although positive feelings may bind people together (Kemper, 1984) and facilitate interpersonal relationship (Fredrickson, 1998), negative feelings such as anger and sadness tend to pull people apart (Kemper, 1984). Research indicates that negative emotions are related to both organization-and person-targeted CWB (Fox et al., 2001). It is worthwhile noting that the negative emotions may be adaptive reactions to negative stimulus event in the strict physiological and psychoevolutionary sense (e.g., anger may serve as warnings that intend to avoid harmful behaviors of others from happening next time); however, such individually functional behaviors may become socially harmful and dysfunctional in a social setting, as demonstrated by anger-induced aggression (Folger & Skarlicki, 1998). Thus, we argue that the hedonic tone of felt emotion affects the feeling-behavior relationship such that negative emotions are positively related to conduction of CWB.
Interestingly, individuals' good intentions to behave in a socially desirable way can prove equally to be harmful strategically for organizations. For example, suppression of negative emotions has been found to cause stress among flight attendants because of the discrepancies between felt emotion and expressed emotions (Hochschild, 1983). (Boyle discusses this study in detail in a chapter in this book.) We examine next the influences of social and organizational norms, as well as feeling rules, on the linkage between feeling and behavior.
Social and Organizational Norms. An individual's behavior is a function of social influences (Bandura, 1977). Certain organizational culture factors such as organizational values and integrity influence the extent to which employees engage in CWB (Boye & Jones, 1997). Workgroup context also has a significant influence on antisocial behaviors of individual employees (O' Leary-Kelly et al., 1996; Robinson & O' Leary-Kelly, 1998). Further, individuals' antisocial behaviors become stronger as their group experiences enrich and they begin to become socialized with the deviant norms (Robinson & O' Leary-Kelly, 1998). In fact, in some organizations employee theft was so institutionalized that it served as a symbol indicating the employee was well socialized within the organization (Altheide, Adler, Adler, & Altheide, 1978).
Social norms also influence an individual's selection of targets to whom they express particular emotions. For example, anger toward a supervisor tends to be viewed as highly inappropriate behavior at work. Thus, people who feel angry with their supervisors may resort to targeting their anger toward coworkers or subordinates, or even organizational property, which are all potentially counterproductive.
Feeling Rules. In certain organizations or occupations, there are strong feeling rules, and emotional labor is deemed part of the work role. Often negative emotions are not deemed appropriate to be expressed at work (cf. Schaubroeck & Jones, 2000). For example, Mundy (1998) found that crying remains as a more serious violation of office etiquette than stereotypical manly reactions such as displayed anger and shouting. Often, employees are trained and required to suppress their negative feelings, which can lead to emotional dissonance (Hochschild, 1983; Morris & Feldman, 1997). The negative emotional energy also accumulates inside the individual and could lead one to resort to CWB to regain emotional balance. Thus, we argue that social and organizational norms and feeling rules will affect the relationship between feelings and behavior such that employees may not behaviorally express their felt emotions, which induces CWB.
Behavior and Effect
Even if adaptive behavior is conducted properly, the desired effect or purpose may not be fulfilled. For example, an employee's complaint regarding his unfair wage may not be responded to immediately, or an expression of anger toward a coworker may not stop the coworker's insulting behavior. Therefore, the fulfillment of the desired effects of employees depends significantly on organizational and situational factors, such as resources availability and task structures. First, organizations (or the behavior target) may lack the resources to which they can draw on to react to the emotional adaptation requirement. For example, facing increasing employee dissatisfaction because of a noisy working environment, an organization may not be able to provide a better one simply due to financial constraints. When a lack of resources interferes with employees' hygiene and/or career development needs, it may lead to frustration among employees (Spector,
4. COUNTERPRODUCTIV E WOR K BEHAVIO R 83
1978, 1996), which gives rise to CWB such as aggression, hostility, sabotage, and theft (Chen & Spector, 1992). Second, task structures such as level of task interdependency also affect the degree to which a desired effect is obtainable. Organizations as cooperative systems are characterized by interdependency (Thompson, 1967) and interlocking behavior (Weick, 1979). Thompson (1967) suggested that there are three types of interdependency (pooled, serial, and reciprocal), which range from connected loosely to closely tied relationships among individual behaviors. In the case of serial and reciprocal interdependency, one person's goals (desirable effects) depend heavily on the way others behave. When one person conducts a behavior that influences negatively the interdependent relationship, a negative pattern of interlocking behaviors may be formed and influence further subsequent interpersonal exchange. Thus, when task interdependency is high, interpersonal conflicts become more likely (Earley & Northcraft, 1989). Based on this research, we argue that resource constraints and task structures affect the relationship between behavior and outcomes such that even appropriate behaviors may fail to induce the desired effect.
As long as the gap between the desired and actual effect exist, the tension will act as new stimulus event, which induces another round of cognitive, emotional, and behavioral reactions. In this case, an escalation of CWB may occur. Thus, an unfulfilled desired effect will serve as a new stimulus event, or lead directly to negative cognition and/or negative emotions, which will induce another round of emotional reactions leading to same or other forms of CWB.
Complex Combinations of the Problem Points
It is important to understand that situations in which only one link becomes problematic will be the simplest scenario. In reality, multiple links in the process may be hindered, blocked, or distorted, which will make the situation more complex. Thus, the already mentioned problem points are only the simplest illustration of the mechanism through which the emotional chain leads to CWB. It is also possible for two or more linkages to become problem points and function together. For example, a stimulus event can be misinterpreted and, simultaneously, an inappropriate negative feeling is generated. When these complex combinations of the problem points occur, which is probably closer to the real situations occurring daily in the organizational life, the whole emotional adaptation mechanism and the consequences of the emotional chain become less predictable and CWBs become more likely.
In this chapter, we proposed an emotion-based process model of CWB. The central message is that many forms of CWB can be understood better by examining the emotional adaptation efforts made by employees. Integrating the social constructivist view of emotion with the framework provided by the psychoevolutionary of emotion, we proposed that CWB occurs when one or multiple linking points in the emotional chain become problematic. Organizational, situational, and individual factors play significant roles in CWB because they influence each of the linking point in the emotional chain.
The organizational implications for conceptualizing CWB as being emotion based are numerous. First, the emotional chain perspective indicates that many factors leading to CWB are actually under the control of employers. Second, control at an early stage of the emotional chain may be more effective and fruitful than at a later stage. Third, this approach also suggests that a healthy work environment that allows for a wider range of emotional expression is very important both for the employee and the organization. Research has suggested that employees appreciate being allowed to talk through their emotional experiences. In an interesting piece of research, Vince and Broussine (1996) used illustrations to access the complex emotions that managers experience during an organizational change process. They found that being able to express and discuss personal feelings about change helped middle managers both to release their emotional tension and to promote a better understanding between top management and middle managers. It is possible that a caring, emotionally honest work environment may help to reduce CWB. Finally, organizations may consider offering employee training on emotion, such as emotional control and regulation, in order to help employees gain a better understanding of their own emotions and enhance their emotional health (Jordan, Ashkanasy, Hartel, & Hooper, 2002).
The model has a few limitations. First, it is possible that the proposed model may not be able to explain all occurrences of CWB. For example, in the case of theft, it may be simply an ethical problem rather than an emotional issue. However, the model is useful for understanding a major portion of CWB. Besides, based on previous conceptual and empirical research (e.g., Spector & Fox, 2002), it is expected that the model will explain a significant portion of variance for various forms of CWB.
Second, we have focused primarily on the effects of situational factors rather than individual differences. We acknowledge that individual differences play important roles in CWB. For example, factors such as locus of control and emotional stability are likely to influence the way people respond to their work and life events, and therefore the possibility for individuals to involve in CWB (Martinko & Zellars, 1998; Martinko et al., 2001). Despite this, O' Leary-Kelley and colleagues (1996) suggested that too much of an emphasis on individual differences as antecedent factors could be misguiding for organizational practices. Consistent with their thoughts, we have focused mainly on situational factors and discussed individual difference factors only where it is necessary.
4. COUNTERPRODUCTIV E WORK BEHAVIOR 85
Despite the limitations, we believe the model nonetheless contributes to the literature with a better understanding of CWB, and highlights many opportunities for future research. First, empirical investigations are needed to explore whether individual difference factors play a more prominent role within the model. For example, it will be interesting to see whether factors such as emotional self-monitoring or deep acting (i.e., to align inner feeling with the social requirement) will help in an individual's emotional adaptation process in a way that is beneficial both to self and to the organization. Research indicates that when employees successfully changed their mood states to align with the organizational context, they may actually experience more positive emotions at work (Grandey, 2000), which is beneficial for both individuals and the organization.
It is also interesting to determine whether the current model explains both CWB and OCB. Spector and Fox (2002) showed that emotion can serve as a link in explaining both CWB and employees' voluntary contributions that go beyond specified task performance or the psychological contract with the employer (i.e., organizational citizenship behavior, OCB). It is reasonable to expect that emotional chains characterized by favorable cognitive appraisal of stimuli events and positive feelings are likely to induce employee extrarole behaviors, such as helping, with the desired effect of maintaining positive emotions and/or positive social interactions with coworkers (cf. Staw et al., 1994).
Finally, the model is yet to be empirically tested. As mentioned earlier, emotion may not explain all forms of CWB. Through empirical studies, specific forms of CWB that have an underlying emotional mechanism may be identified. Due to the process-focused nature of the model, a qualitative research design, such as critical event technique (Flanagan, 1954) and experience sampling methodology (cf. Fisher, 2000).
To sum up, this chapter contributes to the literature with an emotion-based process model of CWB based on psychoevolutionary theory and a constructivist view of emotion. It was argued that CWB is understood better by highlighting the emotional antecedents in the emotional chain. Contextual factors (organizational and family/social factors) can generate events in the form of uncertainty and stress. Individuals differ as to their cognitive ability and awareness of their own emotional states, which, under stressful situations, leads to distorted cognition and feelings of self-identity being threatened, and, in turn, leads to self-protection. Such intention of self-protection under stressful situations may lead to maladaptive behaviors in the forms of CWB. It is worth noting that emotions involved in one emotional chain could be very complex, such as feelings of anger and joy, or shame and pride, or fear and excitement, all at the same time. Further, positive feelings may also lead to CWB, as people attempt to maintain a good mood. Organizational structure or culture may not be supportive enough for positive emotional expression, which gives rise to emotional tension. Emotional chains that fail to generate the desired consequences may lead to further increased tension/emotional energy and, possibly, maladaptive behavior.
We thank Mark Martinko, Robert Plutchik, Paul Spector, and three anonymous reviewers for their insightful comments in developing this chapter.
Charmine E. J. Hartel
Globally, the increasing diversity in the workforce means that the effectiveness of contemporary organizations rests largely on their ability to be open to the differences associated with different cultures. Prejudice, or the negative affective response to perceived dissimilar others, is a significant barrier that organizations must deal with to realize this goal. This chapter makes a unique contribution to the area of diversity and emotions management by considering how cognitive and affective reactions to dissimilar others impacts on individual, group, and organizational outcomes such as hope, intergroup anxiety, trust, perceived fairness, deviant behavior, and task performance. More importantly, this chapter outlines how organizational human resource management practices can alter the consequences of dissimilarity in the workforce and manage more effectively the impact that this has on organizational behavior.
Due to an increase in workforce diversity in recent years, prejudice, or the negative affective response to others perceived as dissimilar, has become an important strategic issue in organizations (Cressey, 2001; Dass & Parker, 1996; Dibben, 2001). Indeed, a number of organizational scholars have foreshadowed that the extent of the effectiveness of contemporary organizations will be determined largely by their ability to be open to dissimilar races, ethnic groups, and the values associated with different cultures (e.g., Dass & Parker, 1996).
Although scholars recognize the need to understand better the effects of diversity on organizational culture and productivity (Eagly & Wood, 1999), discourse in workforce diversity research has evaded largely the issues of stereotyping, prejudice, and discrimination (Linnehan & Konrad, 1999). Further, affect-based theories are adopted rarely in empirical studies of diversity within organizations, despite the well-established affective basis for prejudice (cf. Ashkanasy, Hartel, & Daus, 2002).
In this chapter, we underscore the importance of taking an emotions perspective on the study and management of diversity. In particular, we show how cognitive and affective reactions to dissimilar others impact on individual, group and organizational outcomes such as hope, intergroup anxiety, trust, perceived fairness, deviant behavior and task performance. We also demonstrate how the characteristics of an organization's climate and human resource (HR) policies and practices influence the likelihood of individuals having negative affective reactions (prejudice) to dissimilar others, and the negative emotional consequences of poor relationships.
THE VALUE OF DIVERSITY MANAGEMENT
Studies of both organizational and national diversity have demonstrated that a major deterrent to realizing the positive potential cultural diversity has to offer is prejudice between people whose race or ethnicity is dissimilar to that of dominating employee groups in positions of power and influence (e.g., Bochner & Hesketh, 1994; Fullerton, 1987). Through the perception of similarity/ dissimilarity, observable differences (such as age, gender, and race) are subject to categorization processes (Brickson, 2000; Graves & Powell, 1995; Hartel & Fujimoto, 1999; Linnehan & Konrad, 1999). These processes activate stereotypes and beliefs about certain people. Prejudice is a "social emotion triggered by beliefs about a group's characteristics and relationship to one's own group ... negative beliefs lead to negative emotions" (Linnehan & Konrad, 1999, p. 403). The costs of this prejudice include psychological pain, physical suffering, economic costs, lost opportunities, and denial of the rights to life, liberty, and hope (Snyder & Miene, 1994). Creating openness to perceived differences through managing diversity is proposed to combat these negative experiences. Thus, managing cultural diversity not only improves the operational aspect of the organization, it improves the emotional experience of employees' work life, which is reflected in such things as hope, confidence, satisfaction, and reduced levels of stress, anxiety, and tensions in the workplace (Fujimoto & Hartel, 2002).
In the past, governments have fostered organizations' acceptance of diversity through equity policies such as Equal Employment Opportunity (EEO) and Affirmative Action (AA). However, such legal and regulatory interventions have been viewed as ambiguous and ineffective in improving the quality of
5. EMOTIONAL EXPERIENC E OF I -C WORKGROUP S 89
work life of employees (Gilbert & Ivancevich, 2000). It is, therefore, important for organizations to go beyond simply satisfying legal mandates, to formulate and implement diversity-open policies and practices that best suit their organization's demographic composition. In recognition of this need, there is an expanding literature base on managing diversity with particular reference to the development of management policies and practices that recognize and value diversity in the workplace (Roosevelt, 1996). In this chapter, we are concerned with identifying diversity management policies and practices that thwart employees' tendency to feel dislike for others perceived as dissimilar and improve the work life of those typically discriminated against—the minority-group members in the organization. We tackle this issue by focusing on the triggers of prejudice in diverse workgroups.
Full utilization of diversity can only be achieved when prejudices are overcome. As a result, those who overcome prejudices will contribute their strengths as well as accepting the dissimilar talents brought by perceived dissimilar others. The present research aims to incorporate the concept of self with the cultural orientation of collectivism and individualism to identify the key outcomes of differences in cultural orientation for the individual, the group, and the organization. In particular, we look at the group dynamics, individuals' positive and negative emotional experience of work, affectively driven attitudes, and affectively and attitudinally driven negative and positive employee behaviors. An overview of the model developed herein is provided in Fig. 5.1.
FIG. 5.1. Moderators and consequences of affective response (prejudice) toward dissimilar others in workgroups comprised of individualists and collectivists.
THE PERCEPTION OF DISSIMILARITY
AND ITS EFFECT
Across the cultures of the world, the most important dimension of cultural difference is the relative emphasis on individualism versus collectivism (Triandis, 1990). This dissimilarity reflects individual differences in racioethnicity, values, and views of the self (Hofstede, 1980; Probst, Carnevale, & Triandis, 1999; Triandis, 1980). An important clue to identifying effective ways in managing diversity for positive outcomes should be provided by ascertaining the key triggers of negative emotional responses to the perception of cultural dissimilarity, such as the expression of prejudice.
For the purpose of our research, we define prejudice based on that by Stephan and Stephan (1993) as the low affection associated with perceived dissimilar others. It comprises of emotional reaction like hatred or affection as well as evaluative reactions such as dislike or approval (Stephan, 1999). This definition of prejudice highlights that diversity is an important potential trigger of affective events, which, more often than not, are negative.
Perceived Dissimilarity Based on Race
Race, which individuals often use to infer one's values and self-orientation, is among the most significant stereotype-evoking characteristics (Greenhaus, Parasuraman, & Wormley, 1990; Milliken & Martins, 1996). The term race is defined as a definite combination of physical characteristics associated with national origin or as a group of people sharing the same culture (e.g., Western vs. Eastern nations) (Cox & Nkomo, 1993; Oxford Dictionary Thesaurus, 2001).
Although race is a powerful factor in stereotype formation and prejudice, there is hope for relationships between collectivists and individualists as research shows stereotypes based on such observable characteristics usually operate only when relationships are superficial. However, even if race doesn't provoke the negative stereotype expected, the inevitable difference in values, which are a distinguishing feature of both collectivism and individualism, threatens the realization of harmonious relationships.
Perceived Dissimilarity Based on Values
Values are defined as a nonobservable higher order concept thought to provide a structure for organizing attitudes (Hogg & Vaughan, 1998). Value differences arise from the independent orientation of individualists, which contrasts with the interpersonal orientation of collectivists. Individualists emphasize an "I" consciousness through independent values such as primary concern for personal goals and immediate family (Hofstede, 2001). Collectivists, on the other
5. EMOTIONAL EXPERIENC E OF I-C WORKGROUP S 91
hand, express a "WE" consciousness through interdependent values such as cohesive in-groups, mutual obligations, and concern for one's groups with unquestioning loyalty (Hofstede, 2001).
Values influence the individualist—collectivist interaction (ICI) after social categorization based on observable differences because as members interact with one another, stereotypes are replaced subsequently by a deeper level knowledge of the psychological features of one another (Harrison, Price, & Bell, 1998). In other words, as relationships deepen, people are less affected by stereotypes associated with observable characteristics such as race, and more affected by unobservable characteristics such as values. Consequently, this impacts the way in which we interact with others and, thus, our organizational behavior.
Although values refer to implicit differences and are not easily detected, Harrison et al. (1998) argue that actual dissimilarity in values has great potential for influencing organizational outcomes. For example, individualists may define a good employee as one who states explicitly individual goals and stands up for his or her rights, whereas collectivists may define a good employee as one who follows collective norms and maintains social harmony in the group (Chen & DiTomaso, 1996). Therefore, in individualist cultures, dissimilarity in values may result in unfair performance appraisals based on evaluating collectivist employees as too submissive, lacking confidence, and lacking initiative (Chen & DiTomaso, 1996). As a consequence, these values represent a major obstacle to effective ICI interactions.
In summary, values shape attitudes and behaviors through shaping an individual's self-concept (Marsella, Devos, & Hsu, 1985; Wagner & Moch, 1986). This self-concept refers to the processes of individuals' thoughts, perspectives, feelings, and desires, which, in turn, affect their interaction, and emotion, with others (Deschamps & Devos, 1998). Thus, we argue that actual dissimilarity in values will tend to be a negative affective event, eliciting negative affect, negative group dynamics, and negative work-related behaviors. The overarching proposition we discuss in this chapter is, therefore:
Differences in values associated with the independent orientation of individualists and the interpersonal orientation of collectivists will be correlated positively with negative group dynamics, and negative work-related behaviors such as prejudice.
THE CATEGORIZATION PROCESS
Categorization processes are evoked during ICIs based on the perception of differences. Self-categorization theory states that people tend to classify themselves and others into various social categories (Turner & Oakes, 1989). In-group and out-group distinctions are formed on the basis of categorizations, dictating the extent to which another person is included in an interaction (Brewer, 1979; Dunbar, Saiz, Stela, & Saez, 2000; Miller, 1998; Mor Barak, 2000; Moreland, 1985).
The attraction to those perceived as similar, and the subsequent in-group/ out-group distinctions, activate negative stereotypes and prejudices that cause group members to make biased attributions (Allport, 1954; Hewstone & Ward, 1985; Jackson, Stone, & Alvarez, 1993; Taylor & Jaggi, 1974). Therefore, based on social identity theory, which states that prejudice results from the need for a positive social identity with an in-group (Tajfel, 1981), both individualists and collectivists may possess prejudices against each other due to the definition of their own in-group.
The negative affective and behavioral effects of ICI can be explained by a human's inclination to be attracted to those who hold similar attitudes and opinions (Byrne, 1971) and, thus, form part of their in-group. Individualists tend to define their in-group as individuals or groups who are in agreement with them on personally important issues and values (Rokeach, 1960). Conversely, collectivists tend to define the in-group based on social memberships related to them or who are concerned for their welfare such as family, friends, or people from similar racial backgrounds (Gudykunst, Yoon, & Nishida, 1987; Triandis, 1972, 1990).
The cause and type of prejudice depend on a person's values. Individualists may possess prejudices against collectivists based on both racioethnicity and values as they place greater emphasis on defining the in-group according to similar values and beliefs than do collectivists (Triandis, 1990). As such, individualists' negative affect or prejudices toward collectivists' dissimilar values may present the greatest challenge to obtaining positive employee behaviors. For example, research indicates that individualists feel smaller social distance from collectivists whose values are similar to their own than from European Americans whose values are dissimilar to their own (Rokeach & Mezei, 1966).
Collectivists, in contrast, may hold prejudice against individualists as a result of intergroup anxiety. Intergroup anxiety refers to concerns for negative outcomes, such as rejection and disapproval, arising from in-group members' perceived dissimilarity toward out-group members (Stephan, Ybarra, Martinez, Schwarzwald, & Tur-Kaspa, 1998). Collectivists experience intergroup anxiety as they place greater emphasis on interdependence such that the self is almost defined entirely in the context of significant others (Stipek, 1998). As such, research reveals that collectivists tend to place a stronger emphasis on
5. EMOTIONA L EXPERIENC E OF I -C WORKGROUP S 93
in-group/out-group membership than do individualists (Gudykunst et al., 1987), thus eliciting their in-group favoritism within the ICI setting.
The numerical status of group members may also impact on in-group/outgroup distinctions and behavior toward those perceived as dissimilar. Although cultural minority group members tend to discriminate less than majority group members, indicating their recognition of the superiority of members of the majority group (Moscovici & Paicheler, 1978), research indicates that minority-group membership, and a socially disadvantaged status, activate intergroup comparison and increase competitive behavior toward the majority-group members of individualists (Espinoza & Garza, 1985). In addition, studies have shown that minority members display higher levels of intergroup differentiation and in-group favoritism than do majority members (Gerard & Hoyt, 1974).
Key Outcomes for the Individual, the Group, and the Organization
Although the major difference between individualists and collectivists is in value orientations, we must not forget that the members of each group also tend to be dissimilar in racioethnicity. Research states that such observable differences often lead to prejudices and negative short-term effects because of the stereotypes evoked (Harrison et al., 1998; Pelled, 1996). For example, supervisors tend to categorize subordinates as either in- or out-group members early in their relationship when little information exchange has occurred between the two (Tsui, Egan, & Porter, 1994). This out-group status impacts heavily on the work experience of these employees. Minority-group members tend to perceive less support and feel less attraction and commitment, which, in turn, result in higher rates of absenteeism and turnover (Cummings, Zhou, & Oldham, 1993; Greenhaus, Prasuraman, & Wormley, 1990; Pelled, 1996; Tsui, Egan, & O'Reilly, 1992; Tsui & O'Reilly, 1989). These negative findings underscore the costs involved with ineffective diversity management.
Prejudice has a further detrimental effect on minority members' emotional experience of work and attitudes. It tends to increase their propensity to engage in negative group dynamics such as self-segregation from individualists, and in negative employee behaviors such as turnover and absenteeism (e.g., Kanter, 1977).
The Mediational Role of Self-Representation on ICIs
The previous discussion on prejudice and social identity suggests that the effectiveness of ICI is influenced highly by how individuals within ICI contexts represent themselves in relation to others (Brewer & Gardner, 1996; Brickson, 2000; Markus & Kitayama, 1991). In particular, individualists tend to define the "self" using an independent perspective, whereas collectivists take an interdependent view (Markus & Kitayama, 1991). The independent self refers to a self-concept that is differentiated from all others, whereas the interdependent self refers to a self-concept that reflects assimilation to others or significant social groups. The independent self of individualists and the interdependent self of collectivists explain the tendency for individualists to use the individual as the unit of analysis for social behavior, and for collectivists to use the group (Triandis, 1998). Thus, collectivists view more from an intergroup perspective than do individualists, whereas individualists view more from an interpersonal perspective than do collectivists, as demonstrated in the previous discussion of the "I" and "WE" perspective.
The implication of the different perspectives of collectivists and individualists for ICIs is that collectivists tend to create close groups that possess a stable relationship derived from long-term group membership, whereas individualists tend to create open groups that are flexible in changing their memberships (Triandis, Dunnette, & Hough, 1994; Ziller, 1965). Therefore, we propose that the different self-representations adopted by individualists and collectivists (ICs) play a mediational role in the emotional quality of ICI. Further, the limited understanding of the effect of self-representation on ICIs pose a serious hindrance to this area of research (Brewer & Gardner, 1996; Brickson, 2000). This chapter attempts to address this research gap with the effect of self-representations on ICIs explained in detail next.
Dissimilar Self-Representation Explains Negative ICI Effects
Following from the previous discussion, a number of differences exist between individualist and collectivist group members, which may result in negative ICI effects. For instance, when an individualist is in a group where the majority of members are collectivists, he or she may perceive negative stereotypes deriving from the dissimilar collectivist value of high group identity and in-group favoritism. Further, collectivists expect to gain group benefits from the group, whereas individualists expect to gain personal benefits from the group. As such, the individualists' desire to transcend the group means they are no more likely to favor people within the group than from outside the group, as long as there is a personal benefit in the relationship. Therefore, the individualists may perceive that their personal benefit is thwarted by their need to be unconditionally loyal to the group, thus creating new emotion. The self-reliant orientation of the individualist may also be perceived as selfishness (Bellah, Madsen, Sulivan, Swindler, & Tipton, 1986) and thus pose another hindrance to effective ICI.
Similarly, collectivists are expected to find working with individualists difficult because their core principle is to work cooperatively with in-group mem
5. EMOTIONA L EXPERIENC E OF I -C WORKGROUP S 95
bers and to not seek personal benefits. As such, when a collectivist is in a group where the majority of members are individualists, the collectivist may feel that his or her group benefits are thwarted because of high independence within the in-group. Further, the collectivists within such a group are expected to perceive negative stereotypes deriving from the dissimilar individualist value of high personal identity and to possess intergroup anxiety and fear of abandonment from the in-group.
EFFECT OF THE INDEPENDENT
SELF-REPRESENTATION OF INDIVIDUALISTS
Although individualists may exhibit prejudice toward those with dissimilar values to themselves, they are also expected to have an open group boundary, exhibiting openness to those with similar values, regardless of group membership. Individualists exhibit more prejudice toward a person when the person has a negatively valued characteristic than do collectivists (Crandall et al., 2001). However, research indicates repeatedly that individualists demonstrate openness when collectivists share their values (Rokeach & Mezei, 1966). Therefore, when ICs meet for the first time, individualists are expected to be more open compared to collectivists as long as collectivists agree with individualists on issues that lead to their achievement.
The tendency for individualists to have an open group boundary appears to lead to a cooperative orientation toward groups; however, individualists' orientation to independence means they may also have a greater concern with attaining high status and distinctiveness (Triandis et al., 1994). Consequently, although collectivists emphasize unconditional relatedness to their in-group, individualists emphasize conditional relatedness that calculates carefully the costs and benefits of their relationships with others (Kim, Triandis, Kagitcibasi, Choi, & Yoon, 1994; Triandis, 1998). Research has found that Western participants tend to define themselves as a unique individual to a group when their group membership has a negative value for their self-image. However, when group memberships have a positive value for their self-image, they tend to emphasize their group membership more than individual attributes (Simon, Pantaleo, & Mummendey, 1995). Individualists, therefore, are expected to possess interpersonal prejudices against collectivists, negatively affecting relationships between individualists and collectivists when there is division in values. The emotional experience of individuals in workgroups comprised of people with different cultural orientations is likely to be characterized by the presence of more hassles than workgroups homogeneous in orientation unless care is taken in the formation and management of such groups.
EFFECT OF THE INTERDEPENDENT
SELF-REPRESENTATION OF COLLECTIVISTS
Like the individualist group within the ICI context, the strong intergroup perspectives of collectivists may lead to intergroup prejudices that affect negatively the IC relationship (Perdue, Dovidio, Gurtman, & Tyler, 1990). There is an ethnocentric quality to the collectivist cultures' intergroup orientation in that they expect their perception of the in-group to be valid universally, and everything is scaled and rated with reference to the in-group (Triandis, 1990). Moreover, the salience of the numeric minority membership and socially disadvantaged status of collectivists also accentuate their intergroup prejudices. Research shows that intergroup prejudices by collectivists affect negatively the behavioral integration of individualists with collectivists. When cross-cultural groups are formed for the first time, members from collectivist cultures behave less cooperatively than members of individualist cultures (Gabrenya & Barba, 1987). In other words, collectivists with intergroup perspectives see individualists as out-group members and act less cooperatively toward individualists.
Based on the foregoing review, it is hypothesized that:
H1. The more individualist a person, the more the person leans toward in
terpersonal prejudice, and the more collectivist a person, the more the
person leans toward intergroup prejudice.
H2. The interpersonal prejudice of individualists and the intergroup prej
udice of collectivists will be associated negatively with an individual's
Negative outcomes arise from in-group members' attraction to those perceived as similar, which, in turn, leads to the exclusion, or cultural alienation, of others who are perceived as dissimilar (Zenger & Lawrence, 1989). In particular, collectivists may perceive threats to their self-representation, which may lead to feelings of intergroup anxiety. The intergroup anxiety perceived by these group members, in turn, leads to prejudice toward the individualist members (Britt et al., 1996). As a consequence, prejudices held by individualists toward collectivists, and vice versa, facilitate collectivists' feelings of cultural alienation within individualist groups or organizations. Subsequently, cultural alienation caused by perceived discrimination may affect negatively the relationship between ICI and employees' emotional experience of work. Organizational policies and practices need, therefore, to foster emotionally constructive interactions between employees holding different cultural orientations.
H3. Higher levels of cultural alienation in collectivists will be associated with lower levels of positive emotional experience of work and higher levels of negative emotional experience of work.
5. EMOTIONA L EXPERIENC E OF I-C WORKGROUP S 97
Fostering Emotionally Constructive
Although the literature is replete with cross-cultural studies of individualism and collectivism, little information is available on the factors that foster emotionally constructive ICI within organizations. In fact, although the differences in values associated with individualists and collectivists may provide an important source of negative affective events in diverse workgroups, there is research suggesting that these differences could also be a source of positive affective events. For example, research shows that members who are dissimilar in values and beliefs can develop more creative and better alternatives in problem solving than similar members (McLead & Lobel, 1992). The question that is raised, therefore, is, "What determines whether culturally-based value dif
ferences result in negative versus positive affective events?"
We suggest two reasons, which have diversity management implications. First, we contend that individuals' orientation to cross-cultural experience, defined by the variables of bicultural self, open-self, and intercultural experience, determines one's affective predisposition to cultural difference. Second, we contend that a diversity climate of openness is associated with cultural differences triggering positive, rather than negative, affective reactions. Each is discussed in detail next.
The Moderating Role of an Individual's Affective
Predisposition to Cultural Difference on ICIs
Although cultural values are very influential in shaping the self-representation of individuals (Marsella et al., 1985; Triandis, 1989), there are other factors that contribute. This chapter describes the individual characteristics which moderate the process of perceived dissimilarity and ICI. In particular, we propose that individuals may possess both individualist and collectivist cultural orientations regardless of their cultural origins (Singelis, 1994; Yamada & Singelis, 1999) through their level of openness and experience, and that this impacts on their cognitive and affective reactions to dissimilar others.
Bicultural Self. Our discussion so far has been in reference to separate independent and interdependent orientations. However, Singelis (1994) and Yamada and Singelis (1999) found that the independent and interdependent self could coexist in individuals regardless of one's culture. In agreement, Triandis (1989) suggested that both allocentrics (interdependent self) and idiocentrics (independent self) exist within a culture. Similarly, in an empirical study of stress-coping behavior among American and East Asian students, Cross and Markus (1991) found that East Asian students developed an independent self-representation similar to their American counterparts. However,
their study also found that the East Asian students still had a more developed interdependent self-representation than did American students, meaning that one's self-concept can become bicultural. We propose that there are individuals who hold a balance of both individualist and collectivist values (i.e., have a bicultural self-concept) and that these impact positively on their cognitive and affective reactions to perceived dissimilarity.
The self-representation that demonstrates this phenomenon of both high interdependence and independence has been termed the bicultural self (Cross & Markus, 1991; Yamada & Singelis, 1999). Individuals with a bicultural self are expected to demonstrate reduced prejudice associated with collectivist and individualist cultures.
|H4a.||The bicultural score will be correlated negatively with interpersonal|
|prejudice in individualists. Similarly, the bicultural score will be|
|correlated negatively with intergroup prejudice in collectivists.|
|H5a.||Consequently, bicultural self will be associated positively with em|
|ployees' positive emotional experience of work and related inversely|
|to employees' negative emotional experience of work.|
We propose two factors that lead to a bicultural self: individual's openness (open self) and intercultural experience. These are discussed next.
The interpersonal prejudice and intergroup prejudice sometimes observed in ICIs indicate that ICs hold negative stereotypes toward each other's group (Ashmore & del Boca, 1981). Negative cultural stereotypes are automatically activated in the presence of the member of a stereotyped group regardless of one's prejudice level (Devine, 1989). Research findings indicate that high- and low-prejudice Whites categorized Blacks as poor and aggressive (Devine, 1989). The low-prejudice person, however, controlled their automatically activated stereotypes in considering Black-White interactions whereas high-prejudice persons did not (Devine, 1989). Research findings also indicate that low-prejudice participants are more accurate than high-prejudice participants in estimating their partner's attitudes (Scodel & Mussen, 1953).
The individual with low prejudice is termed here as "the open self." Those with an open self are expected to accept and try to understand self-representations dissimilar to their own, which enables them to develop a bicultural self. The open self may encompass a more positive meaning of self in dealing with perceived dissimilar others. For example, those with an open self will not only be open to those with dissimilar self-representations, but they will also be open to others perceived as dissimilar (such as in gender or knowledge). Therefore, it is hypothesized that:
5. EMOTIONAL EXPERIENCE OF I -C WORKGROUPS
|H4b.||The openness score will be correlated negatively with interpersonal|
|prejudice in individualists. Similarly, the openness score will be cor|
|related negatively with intergroup prejudice in collectivists.|
|H5b.||Consequently,||an open self will be associated positively with||em|
|ployees' positive emotional experience of work and related inversely|
|to employees' negative emotional experience of work.|
An individual's intercultural experience may also shape their self-representation so that it comprises both an interdependent and an independent self-view (Singelis, 1994; Yamada & Singelis, 1999). Intercultural experience may arise from spending time abroad in another culture, from daily interaction within a culturally diverse community, or through having parents from different cultures (Singelis, 1994; Yamada & Singelis, 1999). Research shows that individuals with intercultural experience demonstrate cultural flexibility (Bhawuk & Brislin, 1992). That is:
H4c. The intercultural experience score will be correlated negatively with
interpersonal prejudice in individualists. Similarly, the intercultural
score will be correlated negatively with intergroup prejudice in
collectivists. H5c. Consequently, intercultural experience will be associated positively with employees' positive emotional experience of work and related inversely to employees' negative emotional experience of work.
In summary, positive key outcomes during ICIs are predicted to be impeded largely by members possessing prejudices against perceived dissimilar others. ICIs are therefore expected to achieve positive outcomes when prejudices among dissimilar members are overcome. Further, the paradoxical effect of ICIs also indicates the need for a proactive role by organizations. In light of this, organizations need to develop techniques and procedures to facilitate increasing openness for actual individual differences and the ability to perceive the world from the view of actual dissimilar others (Byrne, 1971). The main way organizations can achieve this is through their human resource policies and practices, which are discussed next.
The Moderating Role of Human Resource Management Policies and Practices in ICI
Although organizations recognize the importance of diversity management, there is still a gap between recognition and action (Noble, 1994). Research shows that although they are leading the way in diversity management, only one-third of U.S. employers had any sort of diversity policy (Carrell & Mann, 1993). Moreover, about half of them were related to EEO/AA requirements (Carrell & Mann, 1993), which are the predecessors to diversity and are not capable of reducing prejudice and obtaining the full utilization of ICs.
In addition to the lack of diversity management, the United States has focused previously on individualistic human resources management (HRM), deemphasizing collectivist HRM, which is insufficient for achieving positive emotional experiences and behaviors in a diverse workforce (Cascio, 1995). As we have shown, organizations will be competitively disadvantaged if they do not realize the importance of, and act to facilitate, the relationships among ICs.
Fostering Diversity Openness
The contextual features of an organization are crucial to the development of effective ICIs, as they influence the self-representation of individuals and determine the effectiveness of relationships that emerge between members (Brickson, 2000). HRM practices typified by individualistic values emphasize personal achievement through clearly defined task responsibilities and reward of individual performance rather than groups (Deutsch, Katz, & Jensen, 1968). The down side for organizations is that such HRM practices encourage employees to focus on a concern for one's own welfare over that of the group's (Brickson, 2000), demotivating them from pursuing other's welfare (Batson, 1998). Exclusive reliance on such individualistic HRM practices then may exacerbate the effect of interpersonal prejudices of individualists. In a similar vein, collectivistic HRM practices such as equality in reward allocation promote individuals' group achievement rather than personal achievement (Ramamoorthy & Carroll, 1998), which motivates individuals to enhance their group's welfare over the individual's welfare (Brickson, 2000; Sherif, 1966). Consequently, we propose that a combination of individualistic and collectivistic policies and practices within an organization, through the development of diversity openness, will have the most positive influence on key affective, attitudinal, and behavioral outcomes of IC workgroups than either could on their own.
HRM policies and practices that foster openness to both an individual's culture and dissimilarity will shape employee attitudes and behaviors and reinforce an organizational culture of diversity openness, thereby enhancing the organization's ability to facilitate effective utilization of ICI (Hartel & Fujimoto, 1999; Kossek & Lobel, 1996). Organizations are more likely to be viewed as valuing differences if: Top management are committed to diversity management; resources are provided (human, financial, technical) for diversity-oriented programs; ongoing cultural diversity training is provided; effective ICI performance is recognized; selection criterion such as "the ability to work with diverse members" is included; constructive conflict is fostered; and mi
5. EMOTIONAL EXPERIENC E OF I - C WORKGROUP S 101
nority members are supported. These organizational policies and practices establish a culture of dissimilarity openness, which is expected to reduce the prejudice level of IC members. Specifically, it is proposed that HRM policies and practices are crucial contributors to ICs' emotional experiences, group dynamics, attitudes, and behaviors.
H6. In organizations with IC workgroups, higher levels of diversity-oriented human resource management policies and practices will be associated with lower levels of prejudice.
H7. In organizations with IC workgroups, higher levels of diversity-
oriented human resource management policies and practices will be
associated with lower levels of cultural alienation.
To date, there is no published measure of assessing an organization's diversity openness available. Therefore, it was necessary to undertake a study to identify a scoring system, which would enable independent judges to rate the extent to which an organization's HR policies and practices are diversity oriented. Fifty structured interviews with persons representing different stakeholder perspectives (e.g., HR managers, union representatives, equity officers, diversity trainers and consultants, and members of ICI groups) were conducted over a 5-month period in Australia. Also, where feasible, the organization's documentation of their HR strategy and policies were analyzed with the aim of identifying differentiating features related to diversity.
The interviews were transcribed verbatim, and the resulting documents were sorted by abstracts into 347 cards. The cards were analyzed using two stages of an unforced Q-sort methodology. In the first stage, five diversity experts were asked to sort the cards into any category of their choosing without any foreknowledge of the research objectives. Their categories were then grouped into prime categories, represented by 99 cards. In the second stage, the 99 cards were converted into survey format and 32 participants were asked to rate each of the 99 items using a 7-point Likert-type scale ranging from very diversity closed to very diversity open. A principal components factor analysis with varimax rotation suggested a stable factor structure of four factors with 47 of the 99 items loading adequately and simply on to those four factors. Next, 19 participants with more than 1 year of work experience in an organization with more than 500 employees were asked to think about their organization and indicate the extent to which they agreed or disagreed with each of the statements. The principal components factor analysis with varimax rotation suggested a stable two-factor structure comprised of 21 items. Factor 1 reflected the level of diversity openness in the organization, and Factor 2 reflected an
ongoing recognition and support for minority members. An additional 15 items were added from the organization's documentation of HRM policies and practices and a literature review of diversity management. The resultant set of 36 items formed the diversity orientation survey used in Study 2.
Study 2 aimed to assess empirically the model of the moderators and consequences of affective response (prejudice) toward dissimilar others in workgroups comprised of individualists and collectivists (see Fig. 5.1). Employee surveys were administered electronically and in person to 359 employees across 14 organizations. The supervisor of the participant's workgroup completed Section 2 of the survey, which asked them to evaluate their workgroup's organizational citizenship behavior, task performance, and prevalence of deviant behaviors.
The final survey, the diversity orientation survey, was developed in Study 1 and was administered to at least two HR managers in organizations with more than 400 employees. Included in this survey were individualistic and collectivist HRM policies and practices questionnaires to assess the cultural orientation of the organization.
Individualistic organizations were sampled randomly by selecting organizations from the Business Who's Who database of Australia. Collectivistic organizations were selected randomly from Japanese firms listed by the Japanese Embassy of Australia and by contacting organizations that had participated in Study 1. Approximately 50 organizations with more than 100 employees were approached to participate. The final sample comprised of 14 organizations, of which 7 were Japanese, 3 were Australian, 2 were European, and 2 were American.
Value Orientations. Individualists and collectivists were identified using the 17-item, 6-point Likert-type questionnaire developed by Triandis (1986). The scale shows strong convergent validity (r = .73). For individualism, a composite score of both self-reliance with hedonism and separation from in-group was computed by taking the mean of the items in the scale. Applying the formula reported in Triandis (1986), individualism = (0.07 X self-reliance with hedonism) + (0.11 x separation from in-group). For collectivism, a composite score of both family integrity and interdependence and sociability was computed by taking the mean of the items in the scale. Applying the formula reported in Triandis (1986), collectivists = (0.57 x family integrity) + (0.25 X interdependence and sociability).
5. EMOTIONA L EXPERIENC E OF I -C WORKGROUP S 103
Interpersonal and Intergroup Prejudices. Singelis's (1994) independent and interdependent self-construal scale was adapted to measure the prejudice dimension of interpersonal prejudice and intergroup prejudices. The reliability of the measure was reported as Cronbach alpha = .73 (Singelis, 1994). Participants were asked to indicate on a 7-point Likert scale the degree of like and dislike felt toward examples representing the collectivist interdependent self and the individualist independent self. Mean scores were used in all analyses involving these constructs.
Cultural Alienated Self. Cultural-alienated self was measured with the 7item, 7-point Likert scale adapted from Jessor and Jessor (1977). The response scale ranged from strongly agree to strongly disagree. The cultural-alienated self score was created by computing the mean of the items in the scale.
Bicultural Self. Singelis's (1994) independent and interdependent self-construal 19-item, 7-point Likert scale ranging from strongly disagree to strongly agree was adapted to measure the bicultural self. As a bicultural person is expected to have a liking toward both an interdependent and an independent person, unlike the prejudice measure already described, the interdependent and independent scales were not reverse coded so that the maximum rating (7) would be like. The bicultural self score was measured as the average score of the interdependence score and interdependence score.
Open Self. Diversity openness was measured with the 4-item, 7-point Likert scale (ranging from strongly disagree to strongly agree) adapted from Fujimoto, Hartel, Hartel, and Baker (2000). The items were derived through a review of the diversity literature and represent the following aspects of openness to perceived dissimilarity: (a) expectations toward perceived dissimilarity; (b) in-group/out-group classification; (c) causal attribution toward failure or success; and (d) openness-closeness to perceived dissimilarity. The scale has a reported reliability of .78. A predictive validity of .21 was shown for the scale in a study of cross-cultural service performance.
Diversity Climate of Openness. Diversity climate of openness was measured with a 36-item, 7-point Likert scale (1 = strongly agree to 7 = strongly disagree). It comprised two scales, namely, diversity-open HRM policies and practices and combined individualist and collectivist HRM policies and practices. Items for the diversity-open HRM policies and practices scale were derived in Study 1.
Three of Ramamoorthy and Carroll's (1998) HRM practices preferences scales were modified to measure combined individualist and collectivist HR policies and practices. The progressive HRM scale and equity in reward alloca
tion scale assessed individualist HRM practices, and the equality in reward allocation scale assessed collectivist practices. Participants were asked to indicate agreement with each statement with respect to HRM within their organization. Responses were given on a 7-point Likert-type scale ranging from strongly agree to strongly disagree. The average score on the progressive HRM scale, the equity in reward allocation scale, and the equality in reward allocation scale was used to represent the level of combined individualist and collectivist HRM policies and practices in the organization.
Emotional Experience of Work. Individual experiences at work were categorized into negative and positive emotions (Fujimoto & Hartel, 2002). The negative emotion category comprised situational anxiety, stress, depression, job-related tension, and lack of supervisory support. The positive emotion category comprised hope, confidence in skills and knowledge, and organizational-based self-esteem (Fujimoto & Hartel, 2002). The average score was created for each emotion category. Depression, situational anxiety, and stress were measured by the 7-item, 4-point Likert-type Depression Anxiety Stress Scales (DASS) (Lovibond & Lovibond, 1995), which has high reported internal consistency (Brown, Chorpita, Korotitsch, & Barlow, 1997). Job-related tension was measured with the 13-item, 5-point Likert-type scale developed by Kahn, Wolfe, Quinn, Snoek, and Rosenthal (1964). Reliability of the measure was reported as Cronbach alpha = .83. Lack of supervisory support was measured with the 4-item, 7-point Likert scale developed by Pearce, Sommer, Morris, and Frideger (1992). The reported internal consistency reliability of the scale ranged from .88 to .95. Feelings of lack of supervisory support were measured by computing the mean of the reverse-coded supervisory support items. Organization-based self-esteem was measured with the 10-item, 7-point Likerttype scale developed by Pierce, Gardner, Cumming, and Dunham (1989). The average reliability of the measure was reported as Cronbach's alpha of .91. Confidence in skills and knowledge was measured with the 3-item, 7-point Likert scale developed by Wagner and Morse (1975). The internal consistency reliability of the measure was reported as .96. Hope was measured by taking the mean of the 4-item, 4-point Likert-type scale developed by Snyder et al. (1991). Reliability of the measure was reported as Cronbach alpha ranging from .74 to .84.
Results are grouped together by hypothesis for ease of presentation. Regression analyses were conducted to test the seven hypotheses.
5. EMOTIONAL EXPERIENCE OF I -C WORKGROUP S 105
H1. The more individualist a person, the more the person leans toward interpersonal prejudice, and the more collectivist a person, the more the person leans toward intergroup prejudice.
This hypothesis was supported (F(l,342) = 17.33, p < .0001, R2Adj = .05, B = .22). This indicates that the more individualist a person, the more the person leans toward interpersonal prejudice. Similarly, the more collectivist a person, the more the person leans toward intergroup prejudice.
H2. The interpersonal prejudice of individualists and the intergroup prejudice of collectivists will be associated negatively with an individual's emotional experience.
Analyses showed some support for the second hypothesis. Specifically, support was found for a relationship between negative emotion and intergroup prejudice (F(l,355) = 8.9053, p = .003, R2 = .024, (3 = .21351) and between positive emotion and interpersonal prejudice (F(l,351) = 20.3075, p = .000,
= .004, B = -.29094). Although not significant, the directions of relationships were in the predicted direction for negative emotion and interpersonal prejudice (F(l,354) = 1.0084, p = .316, R2 = .003, B = .061389) and for positive emotion and intergroup prejudice (F(l,352) = 3.0328, p = .082,R2 = .006, B = –.13656).
H3. Higher levels of cultural alienation in collectivists will be associated with lower levels of positive emotional experience of work and higher levels of negative emotional experience of work.
Hypothesis 3 was supported. Negative emotion was higher with higher cultural alienation for collectivists (F(1,196) = 10,3,9,p = .0015, R2Adj = .05, B = .18), and positive emotion was lower with higher cultural alienation (F(l,196) = 10.4,p = .0015, R2 = .05, B = –.18). Unexpectedly, this was also true of individualists (F(1,196) = 24.6, p < .0001, R2Adj = .11, B = .24), and there was no difference in the slopes between the two groups.
H4a. The bicultural score will be correlated negatively with interpersonal prejudice in individualists. Similarly, the bicultural score will be correlated negatively with intergroup prejudice in collectivists.
This hypothesis was not supported (p > .1) for either individualists (F(l,144) = 2.89, ns) or collectivists (F(l,196) = 3.09, ns).
H4b. The openness score will be correlated negatively with interpersonal prejudice in individualists. Similarly, the openness score will be correlated negatively with intergroup prejudice in collectivists.
Support was found for this hypothesis but for individualists only (F(l,124) = 7.51,p = .0071, R2 = .057091, B = –.56). However, the relationship between interdependent prejudice and openness for collectivists was in the predicted direction (F(l,175) = .0373, ns, B = –.03).
H4c. The intercultural experience score will be correlated negatively with interpersonal prejudice in individualists. Similarly, the intercultural score will be correlated negatively with intergroup prejudice in collectivists.
This hypothesis was not supported for individualists (F(l,134) = .0110, ns) or for collectivists (F(l,180) = .0043, ns).
H5a. Bicultural self will be associated positively with employees' positive emotional experience of work and related inversely to employees' negative emotional experience of work.
The relationship between negative emotion and bicultural self was significant but in the opposite direction to that predicted (F(l,342) = 18.3752, p < .0001, R2 = .05, B = .18). Although not significant for the relationship between positive emotion and bicultural self, the relationship was in the predicted direction (F(l,342) = 3.2014, p = .0745, B = .09).
H5b. Open self will be associated positively with employees' positive emotional experience of work and related inversely to employees' negative emotional experience of work.
Although the predicted relationships were not significant, they were in the predicted direction (F(l,312) = .0012, ns, B = –.0008 for negative emotion and F(l,311) = .1143, ns, B = .008).
H5c. Consequently, intercultural experience will be associated positively with employees' positive emotional experience of work and related inversely to employees' negative emotional experience of work.
Although the predicted relationships were not significant, they were in the predicted direction for negative emotion (F(l,320) = 1.1161, ns, B = –.01064) and for positive emotion (F(l,319) = 2.5776, p = .1094, B = .008).
5. EMOTIONA L EXPERIENC E OF I -C WORKGROUP S 107
H6. In organizations with IC workgroups, higher levels of diversity-oriented human resource management policies and practices will be associated with lower levels of prejudice.
Although this hypothesis was only supported for interpersonal prejudices, the relationship was in the predicted direction for intergroup prejudices. Breakdown of the significant relationships showed a negative correlation between interpersonal prejudice and collectivist HR practices (F( 1,337) = 10.1031, p = .0016, R2 = .03, B = -.096), between interpersonal prejudice and diversity climate (F(l,337) = 4.27, p = .0396, R2 = .013, B = –.10302), between interpersonal prejudice and equity in reward systems (F(l,337) = 8.9578, p = .003, R2 = .0259, B = –.07647), and between interpersonal prejudice and progressive HR practices (F(l,337) = 3.9829, p = .011878, R2 = .01168, B = –.11878).
H7. In organizations with IC workgroups, higher levels of diversity-oriented human resource management policies and practices will be associated with lower levels of cultural alienation.
General support was found for this hypothesis, with three of the relationships being significant and the remaining relationship being in the predicted direction. Breakdown of the significant relationships showed a negative correlation between interpersonal prejudice and collectivist HR practices (F(1,334) = 7.8509, p = .0054, R2 = .023, B = –.11143), between interpersonal prejudice and diversity climate (F(l,334) = 13.9921,p = .0002, R2 = .04, B = –.24145), and between interpersonal prejudice and progressive HR practices (F(l,334) = 15.8732, p < .0001, R2 = .045368, B = –.30707).
This chapter had two aims. The first was to show how cognitive and affective reactions to dissimilar others impact on individual, group, and organizational outcomes such as hope, intergroup anxiety, trust, perceived fairness, deviant behavior, and task performance. The second was to demonstrate how the characteristics of an organization's climate and human resource (HR) policies and practices influence the likelihood of individuals having negative affective reactions (prejudice) to dissimilar others and the negative emotional consequences of poor relationships.
The Cognitive and Affective Reactions to Dissimilar Others
A major deterrent to realizing the positive potential cultural diversity has to offer is the affective and cognitive responses between dissimilar people. These responses are shaped largely by the values the individual holds (Marsella et al.,
1985; Wagner & Moch, 1986). This chapter presented a model of individualist-collectivist interactions (ICIs), which incorporated an individual's values with the person's intercultural experience, linking them to key outcomes for the individual, the group and the organization.
Specifically, we found that individuals are more inclined to exhibit a certain type of prejudice based on their own cultural values. For instance, people who hold individualist values are more likely to lean toward forms of interpersonal prejudice, whereas people with more of a collectivist nature are more likely to hold intergroup prejudice. However, we did not find evidence in this study of a relationship between bicultural values and prejudice levels.
The direction of relationships conformed to the hypothesis that openness and intercultural experience are inversely related to a person's tendency to be prejudiced. However, the relationship was significant only for individualists on the openness construct.
Emotional Experience From Perceived Dissimilarity. This chapter examined the affective consequences of diversity, which has been identified as an important factor overlooked largely in diversity research (Ayoko & Hartel, 2000; Fujimoto et al., 2000; Linnehan & Konrad, 1999). We found a significant relationship between interpersonal prejudice and the positive emotional experiences of hope, confidence in skills and knowledge, and organizational based self-esteem, and between intergroup prejudice and the negative emotional experiences of situational anxiety, stress, depression, job-related tension, and lack of supervisory support.
Further, when individuals feel culturally alienated they are more likely to experience negative emotions, and less likely to experience organizational-based self-esteem, confidence in skills and knowledge, and hope. This occurs regardless of whether a person places greater emphasis on self or group achievement.
Although the prediction that an open self and intercultural experience would be associated with positive emotional experience and inversely associated with negative emotional experience was not upheld, the predictions were in the hypothesized direction. The same was true for the hypothesized relationship between bicultural values and positive emotions. An unexpected finding, however, was the significant positive relationship between negative emotional experience and bicultural self. This finding may reflect the presence of a third variable operating, such as the presence of a poor climate for diversity openness, which would lie in contradiction to the values of a bicultural person. Future studies should assess the effect of the diversity climate on the relationship between bicultural self and emotional experience.
This chapter makes a significant contribution to both diversity research and research into emotions. It has tackled the difficult issues of stereotyping, prejudice and discrimination, which are frequently ignored in the literature
5. EMOTIONAL EXPERIENC E OF I -C WORKGROUP S 109
(Linnehan & Konrad, 1999), yet have been shown to create psychological pain, physical suffering, economic costs, lost opportunities, and denial of the rights to life, liberty, and hope (Snyder & Miene, 1994). We, ourselves, have found that they have a powerful impact on the emotional experiences of individuals in the workplace. Through gaining understanding of how stereotypes and prejudice impact on emotions, we can better understand our own and other's emotions and increase the capabilities of individuals, groups, and organizations to manage constructively emotions (Shih et al., 1999).
The Influence of an Organization's Climate and Human Resource (HR) Policies
The second purpose of this chapter was to examine how the characteristics of an organization's climate and human resource (HR) policies and practices influence the likelihood of individuals having negative affective reactions (prejudice) to dissimilar others and experiencing the negative emotional consequences of poor relationships. One of the key ways in which organizations can influence employee openness to perceived differences and the ability to perceive the world from the view of actual dissimilar others is through their human resource management policies and practices. In fact, there is an expanding literature base on managing diversity with particular reference to the development of management policies and practices that recognize and value diversity in the workplace (Roosevelt, 1996). This research area posits that group boundaries can be manipulated to reduce intergroup bias and conflict through diversity practices and interventions (Dovidio et al., 1997). Yet little research exists on how these practices and interventions impact on the cognitions and affect of employees, particularly when one party is dominated by another. This chapter attempted to address this gap through looking at the orientation of HR policies and practices to individualist, collectivist and diversity values.
Specifically, we argued that HRM policies and practices themselves reflect cultural value orientations. Furthermore, we demonstrated that HRM policies and practices that motivate individuals to enhance their group's welfare over the individual's welfare (Brickson, 2000; Sherif, 1966) can exacerbate the tendency for individualists to feel interpersonal prejudices toward collectivists. Subsequently, we proposed that the HRM policies and practices with the most positive influence on key affective, attitudinal, and behavioral outcomes of IC workgroups would comprise both individualist and collectivist values.
We found that organizations with HR policies and practices that were diversity oriented, that is, comprised of both individualist and collectivist values, showed lower levels of interpersonal prejudice and cultural alienation. This has significant consequences for our earlier findings. It suggests that when workgroups are comprised of both individualists and collectivists, organiza
tions need to integrate both individualist and collectivist policies and practices into their HRM in order to overcome the inclination of individualists working in an individualist country to feel interpersonal prejudice toward collectivist team members. Further, we found that individuals who feel culturally alienated, regardless of which value orientation they lean toward, are more likely to experience situational anxiety, stress, depression, job-related tension, and lack of supervisory support. However, organizations can combat this with diversity-oriented policies and practices and can increase the individual's feelings of organizational-based self-esteem, confidence in skills and knowledge, and hope.
In order for organizations to effectively manage ICIs, we must develop a better understanding of the ICI process and its relationship to individual, group and organizational outcomes. It is crucial that we gain a better understanding of the organizational capabilities and practices required to manage culturally diverse groups and minimize negative emotional experiences while improving positive emotional experiences at work. A number of scholars have identified the need for research to examine the effect of organizational culture and human resource management initiatives for diversity management (e.g., Linnehan & Konrad, 1999). It is only through increasing knowledge and awareness of how organizational policies and practices impact on the cognitive and affective reactions of their employees to dissimilar others, that effective management of diversity will be realized.
Taking a bounded-emotionality perspective of organizations, we demonstrated how diversity is an environmental variable that can be a source of negative or positive affective events, depending on the characteristics of the individuals and diversity management practices present in an organization. Using this vantage point, we also demonstrated how the culturally shaped individual difference of cultural orientation results in particular prejudice tendencies and influences the emotional experience of individuals at work. These outcomes, as such, may be moderated by an organization's HR policies and practices.
Prejudice refers to the negative affective response expressed toward a person who is perceived as dissimilar. This chapter determined that the different ways in which ICs perceive the self lead to a different basis for prejudice in ICs. Further, propositions were offered and supported with respect to how prejudice might impact on group dynamics and employees' attitudes, emotional experience, and behaviors at work. Together, the findings suggest that organizations comprised of ICs are best served by a diversity-open climate, progressive HR policies and practices, and both collectivistic and individualistic HR practices. The findings provide strong evidence that an organization's
5. EMOTIONA L EXPERIENC E OF I -C WORKGROUP S 111
HR policies and practices can moderate group dynamics, individuals' negative emotional and positive emotional experience of work, employment attitudes, positive employee behaviors, and withdrawal behavior. In the 21st century, understanding diverse individuals' perspectives, values, and affective events will help organizations move closer to achieving work environments that promote constructive and healthy emotional, cognitive and behavioral reactions at work. This chapter makes a step toward that understanding.
This page intentionally left blank
Neal M. Ashkanasy
Wilfred J. Zerbe
Charmine E. J. Hartel
Traditionally, books on organizational behavior proceed in a linear fashion from the individual, to groups, and then to the organization as a whole. This volume is no exception. Organizations exist only because of the people within them. Consequently, understanding organizations first of all requires understanding the people who populate them, and especially their needs, drives, and capabilities. It is all the more surprising therefore to find that organizational behavior scholars, and behavioral science researchers in general, were so slow to appreciate the centrality of emotion in organizations (see Ashforth & Humphrey, 1995). Indeed, even today, many scholars are reluctant to accept an emotions-oriented explanation of motivation and behavior (e.g., Becker, 2003). Interestingly, and as Weiss and Brief (2002) pointed out, early organizational behavior scholars were deeply interested in the role of feelings and emotion, but this research seems to have withered with the rise of behaviorism in the 1940s and 1950s.
From the perspective of the present volume, the major breakthrough in bringing the study of emotions to the individual level of analysis was Weiss and Cropanzano's (1996) introduction of affective events theory (AET). This theory emphasized the dynamic and interactive association of emotion, behavior, and performance in organizations. More recently, Ashkanasy (2003a) proposed a "multi-level theory of emotions" (p. 3), which puts AET at the core of the model and extends its discussion to five levels: within-person, between-person, interpersonal dyads, groups, and organization. In a follow-up article, Ashkanasy (2003b) emphasized that the integrating factor across the five levels of analysis is the neurobiological basis of emotion. In this respect, human beings are bound to behave in ways predetermined by their neurobiological architecture. In the introduction to our first book (Ashkanasy, Hartel, & Zerbe, 2000), we made a similar point—that understanding emotions in organizational settings requires first an understanding of the mechanisms underlying the generation and control of emotion.
The chapters in this section do not deal specifically with the neurobiology of emotion, but drive home a corollary—that emotions constitute the essence of behavior in organizations: a message that lies at the core of the present volume. This point is especially salient in Maree Boyle's analysis of the performance of emotional work within an emergency services organization in chapter
3. In Boyle's rich account, which takes a dramaturgical perspective (Goffman, 1959), members of the Department of Paramedical Services (DPS) engage in emotional work in three "emotional regions." But it is not only in the obviously emotion-laden aspects of emergency service work that emotion work is required, but "offstage"—in the work areas behind the scenes, and especially after work—that Boyle found the most interesting manifestations of emotion work. Here, workmates, family members, and the wider community provide essential emotional support to stressed DPS officers, most of whom are men. Yet, perhaps not surprisingly, the department seems to be unable or unwilling to recognize the importance of support. Consistent with the rational organization model prevalent in the public service, there seems to be an expectation that, somehow, departmental officers are heroes—that they are able to control their emotions at work, and to deal with the emotional consequences afterward. Thus, the burden of this role is placed on the officers' informal networks. Amazingly, this seems by and large to have worked. But the point here is not so much the issue of the organization's responsibility toward the welfare of its employees, but the finding that emotion is so central to the performance of work by the employees, both on and off duty. In this instance, even the department's indifference is subsumed in an organizational environment where emotion and emotion management lie at the very core of organizational effectiveness.
In chapter 4, Yongmei Liu and Pamela Perrewe delve more deeply into the murky underside of organizational behavior—which they refer to as "employee counterproductive work behavior" (CWB). Taking an approach based on ideas of emotional adaptation, Liu and Perrewe seek a less superficial understanding of CWB than is available through the traditional behavioral and cognitive models of organizational behavior. But CWB, which often involves acts of anger driven by deep feelings of hurt, is inherently emotion charged. Liu and Perrewe bring together two streams of thought—psychoevolutionary theory (Plutchik, 1980) and the constructivist view of emotion (Averill, 1980)—and integrate them in an intriguing and original model of the genesis of CWB. In this model, CWB is not some form of irrational behavioral response to a per
6. A BOUNDE D EMOTIONALIT Y PERSPECTIV E 115
ceived stimulus, but is a natural response to environmental pressures, and is derived from a basic need for survival in the organizational context. As with Boyle's examples, emotion is not something that can be tacked on or managed in the conventional linear sense. Instead, emotion forms the basic kernel underlying particular behaviors—in this case, CWB.
Chapter 5 is the third chapter in this section. Here Yuka Fujimoto, Char-mine Hartel, and Debra Panipucci delve into understanding the emotional dimensions of diversity and that most fundamental of human instincts: individualism versus collectivism. Their results demonstrate that individual attitudes in respect of collectivism-individualism and diversity are related to emotional experiences. This research raises some intriguing questions about the interactions of diversity and emotion (see also Ashkanasy, Hartel, & Daus, 2002b), and the way that organizational members' feelings about their work are conditioned by the way they feel about their fellow workers and the work environment. Viewed through the lens of Liu and Perrewe's model of CWB, it would seem that the same processes of psychoevolution and constructivism are operating here too, in that organizational members react instinctively to perceived threats in their working environment. Therefore, when a threat is seen to derive from the makeup of the work group, emotional responses related to basic survival are triggered.
In summary of the chapters in this section: All bring different perspectives to the topic of emotions at the individual level of organizational analysis, but at the same time all bring a consistent theme that emotion lies at the core of organizational behavior. This holds true irrespective of whether organizational behavior is altruistic (Boyle) or counterproductive (Liu and Perrewe). It also holds true with respect to attitudes toward diversity and working together. As Ashkanasy (2003b) stressed, the unifying thread is that all human beings are ultimately dependent on their neurobiology for the way they think, feel, and behave.
The implication of this conclusion is obvious in many ways, but apparently not so obvious for many years to scholars working in the field of organizational behavior research. Despite the early interest in emotions, as noted by Weiss and Brief (2001), scholars of organizational behavior in the following years concentrated their efforts on developing sophisticated models that were seriously flawed in their failure to account for the emotional underpinnings of individual behavior. The chapters in this section of the book help to make a start on redressing this shortcoming.
This page intentionally left blank
This page intentionally left blank
Matthew J. Grawitch
David C. Munz
A large volume of literature suggests that positive affect can facilitate individual problem solving. The effect of positive affect stems from its influence on processes related to perception, cognition, and motivation. More recent research has begun to examine the relationship between affect and different stages of group problem solving. In this chapter, we propose a theoretical framework to understand the relationship between affect and group performance at various stages of problem solving that synthesizes individual- and group-level research. This framework, the Group Affect Problem Solving (GAPS) model, shows relationships between individual affective processes and group affective norms at the problem identification, brainstorming, solution development, and implementation stages of the group problem-solving process. Additionally, we offer propositions underlying the GAPS model to be examined empirically in future research. Research issues and practical implications are also discussed.
Modern organizations seek continuously to maximize their human capital. Often, employees, working within problem-solving groups (e.g., cross-functional, quality improvement, or reengineering teams), generate ideas or solutions to the challenges faced by their organization and may manage the implementation of those solutions (Uhl-Bien & Graen, 1998). Therefore, it is important to understand individual- and group-level constructs that influence outcomes at various stages of the problem-solving process.
Affect at the individual and group levels contributes to our understanding of group problem-solving performance, because of its pervasive effect on motivation and information processing (Gendolla, 2001). Affect is defined as "mental
7. AFFECT AND GROU P PROBLE M SOLVIN G 121
solving process typically used in organizations. The model also contains two key elements that influence group problem solving: individual affective processes and group affective norms. In order to explicate the model and the rationale supporting it, we explore both the individual and group affect constructs in detail. Then, we discuss each link in the GAPS model using current empirical research to support and hypothesize specific relationships. Additionally, we propose and discuss the implications of context, with an emphasis on organization and group factors that may influence model relationships. Finally, we present practical implications and research opportunities to bridge the gap between theory, research, and practice.
THE STRUCTURE OF INDIVIDUAL AFFECT
Affect at the individual level encompasses a wide variety of psychological phenomena, including emotions and moods. Affective experiences provide information about the state of the individual in transaction with the environment. These affective experiences motivate behavior in any given situation (Forgas, 2002; Forgas & Vargas, 2000; Fredrickson, 2001; Gendolla, 2001; Nemanick & Munz, 1997; Parkinson et al., 1996). For example, research suggests that individuals experiencing positive affect are motivated to engage in proactive or prosocial behaviors (e.g., helping behaviors), whereas individuals experiencing negative affect are motivated to engage in avoidance or withdrawal behaviors (e.g., distraction; Parkinson et al., 1996). Until recently, affect was discussed primarily in terms of a circumplex composed of two conceptually and statistically independent dimensions: activation and hedonic tone (Russell, 1980). All affective experiences could be mapped along these two dimensions. The activation dimension relates to the level of physiological arousal inherent in the experience, and the hedonic tone dimension refers to the pleasantness or unpleasantness of the experience. Thayer (1967, 1978) further defined dimensions related to energy and calmness, and Watson and colleagues (Watson & Clark, 1997; Watson, Clark, & Tellegen, 1988) defined dimensions of positive and negative affect. Additionally, recent efforts have integrated different conceptualizations of affect into a unified affect circumplex as depicted in Fig. 7.2 (Yik, Russell, & Feldman Barrett, 1999).
THE FUNCTION OF INDIVIDUAL AFFECT
The affective circumplex provides a theoretical conceptualization of affect from a structural perspective, but it does not provide a conceptualization of the functionality of affect (Carver, 2001). Carver argued that although affect may be understood in structural terms, those structural components do not explain
FIG. 7.2. An integration of several common affect constructs: a, dimensions identified by Russell (1980); b, dimensions identified by Thayer (1967, 1978); c, dimensions identified by Watson, Clark, and Tellegen (1988). Adapted from Yik, Russell, and Feldman Barrett (1999).
the function of affect. Instead, Carver proposed that affect is best understood as two separate functional systems: an approach system and an avoidance system, both of which have received significant theoretical and physiological support (Carver, 2001; Carver & Scheier, 1999) and identify closely with social psychological theories of motivation and self-discrepancy (i.e., promotion/prevention focus; Higgins, 1998).
The Approach System
The approach system is activated when the goal is to obtain or achieve a positive outcome (Carver, 2001), motivating individuals to seek growth and accomplishment (Higgins, 1998). Therefore, approach processes motivate the individual toward positive outcomes, and the individual evaluates the progress made in achieving those outcomes or goals through continuously operating self-regulatory mechanisms (Carver & Scheier, 1999). That is, as approach processes inform the individual that exceptional progress is being made toward a positive outcome, high positive affect results, in the form of positive mood, elation, or eagerness (Carver, 2001). However, when the approach system informs the individual that inadequate progress is being made toward a positive outcome, then low positive affect results, in the form of sadness, de
7. AFFECT AND GROU P PROBLE M SOLVING 123
pression, or tiredness (Carver, 2001). As Carver argued, the experience of positive affect also initiates the approach system, causing individuals to emphasize or focus on the obtainment or achievement of a positive outcome (Carver, 2001). Figure 7.3 provides a graphical presentation of the approach system.
The Avoidance System
The avoidance system is activated when the goal is to avoid or minimize a negative outcome (Carver, 2001), motivating individuals to seek safety and security (Higgins, 1998). Avoidance processes motivate the individual away from negative outcomes, and the individual evaluates the progress made in avoiding those outcomes through continuously operating self-regulatory mechanisms (Carver & Scheier, 1999). In other words, as avoidance processes inform the individual that exceptional progress is being made to move away from a negative outcome, low negative affect results, in the form of relief, calmness, or relaxation (Carver, 2001). However, when the avoidance system informs the individual that insufficient progress is being made to move away from a negative outcome, then high negative affect results, in the form of anxiety, negative mood, fear, or irritability (Carver, 2001). As Carver argued, the experience of negative affect also initiates the avoidance system, causing individuals to emphasize or focus on the avoidance of negative outcomes. Figure 7.3 provides a graphical presentation of the avoidance system.
IMPLICATIONS OF THE FUNCTIONAL PERSPECTIVE
Because of its emphasis on goals and goal achievement, the functional perspective offers implications for understanding the relationships indicated by the GAPS model. Approach and avoidance processes that result from and pro-
FIG. 7.3. Graphical representation of approach and avoidance mood outcomes. Adapted from Carver (2001).
duce affect influence how that affect is regulated and directed toward the development of effective group norms and performance. In essence, the functional perspective purports an ongoing gap analysis, in which an individual evaluates progress made toward positive goals and standards and away from undesired outcomes (also known as control theory; Carver & Scheier, 2002). High positive affect results from a decrease in the gap between positive goals and the individual's current state, whereas low positive affect results from an increase in the gap between positive goals and the individual's current state. Alternatively, high negative affect results from a decrease in the gap between negative outcomes and the individual's current state, whereas low positive affect results from an increase in the gap between negative outcomes and the in-dividual's current state.
The functional theory's emphasis on goals and standards aligns suggests a transactional approach to affect (Carver, 2001). A transactional approach accounts for the importance of individual characteristics (e.g., personality), environmental conditions (i.e., events in the external world), and the mutual influence each exerts to create the momentary states that individuals experience (Carver & Scheier, 2002). Because the functional theory proposes that individuals set goals and standards to be achieved (approach standards) or avoided (avoidance standards), the progress that is continuously evaluated includes characteristics of the person (e.g., differences in approach/avoidance emphasis, different goals), environment conditions (e.g., aspects of the environment that facilitate or impede progress), and mutual influences between the person and the environment (e.g., the type of goals an individual sets causes him or her to seek out specific environmental events that influence the progress toward those goals). These self-evaluations then result in a particular affective state.
Within the functional paradigm affect serves two important roles: (a) an informational role and (b) a motivational role (Gendolla, 2001). The GAPS model indicates that both roles influence the outcomes of problem-solving workgroups. Affect influences how information is perceived, synthesized, and cultivated within the group (the informational role), and it influences the desire of workgroup members to engage in specific types of behavior (the motivational role).
The Informational Role
The informational role of affect occurs in the influence on perception and information processing. Positive and negative affect occur due to separate self-evaluative processes, and the influence on perception and information processing tends to be congruent with the type of affect. High positive affect resulting from the approach system signals that sufficient progress is being made toward positive outcomes. Therefore, less attention is required in critically
7. AFFECT AND GROU P PROBLE M SOLVING 125
evaluating environmental events and characteristics, allowing individuals to perceive and process information at a more global level (Gasper & Clore, 2002). Conversely, high negative affect resulting from the avoidance system signals that insufficient progress is being made away from negative outcomes. Consequently, more attention is required in critically evaluating environmental events and characteristics, causing individuals to perceive and process more localized information (Gasper & Clore, 2002). Essentially, positive affect allows individuals to focus on the forest, whereas negative affect causes individuals to focus only on the trees.
The difference in focus that results from the experience of positive and negative affect also influences cognitive flexibility. Individuals experiencing high positive affect exhibit greater cognitive flexibility than individuals experiencing other types of affect (e.g., low positive, high negative affect; Isen, 2000). Because high positive affect signals alignment with positive goals and outcomes, the individual redirects cognitive resources away from intense environmental scanning and toward other more relevant tasks, exploring a wider variety of behavioral options when making decisions and increasing the efficiency with which information is synthesized (Fredrickson, 2001; Isen, 2000), positive outcomes especially relevant to our later discussion of the GAPS model.
Because of the self-referent nature of affect (i.e., affect resulting from self-evaluations of one's own progress), information and appraisals congruent with the experienced affective state are often primed. For example, positive affect has been found to prime information (Doherty, 1998; see also Isen, 2000) and appraisals (Munz, Benaka, & Walters, 2000) with a positive valence, whereas negative affect has been found to prime information (Doherty, 1998) and appraisals (Munz et al., 2000) with a negative valence. This priming filters individual experiences through either a positive or negative lens, leading to perceptions and cognitions congruent with the current affective state.
The Motivational Role
The motivational role of affect was discussed to a degree earlier. What is of importance to the GAPS model is how affect manifests itself in an individual's motivation to engage is certain types of behaviors. The functional theory proposes that individuals strive to approach positive outcomes and avoid negative outcomes, but the affect that results from the continuous self-evaluations must be managed (i.e., affect regulation), which can occur consciously or unconsciously (Larsen, 2000; Mayer, 2000; Parkinson & Totterdell, 1999; Parkinson et al., 1996; Thayer, Newman, & McClain, 1994). Individuals may engage in a variety of affect management strategies, from active mood management (e.g., relaxation techniques) to withdrawal-avoidance techniques (e.g., avoiding other people; Thayer et al., 1994). The extent to which individuals actively manage affect depends on their ability to recognize and monitor their affective states (Swinkels & Giuliano, 1995) and use that information as input to select appropriate cognitive and behavioral mechanisms for affect regulation (Martin, Ward, Achee, & Wyer, 1993). Therefore, affect management is likely to result in different regulation strategies when the approach system has been initiated than when the avoidance system has been initiated (Carver, 2001).
Because positive affect generally results from approach processes, in which the goal is to obtain or achieve positive outcomes, previous research has suggested that positive affect is related positively to positive outcomes (e.g., satisfaction, commitment, prosocial behaviors; Brief & Weiss, 2002; Munz, Huelsman, Konold, & McKinney, 1996). Alternatively, because negative affect generally results from avoidance processes, in which the goal is to avoid or evade negative outcomes, this same research has suggested that negative affect is related positively to negative outcomes (e.g., turnover, turnover intentions, burnout; Brief & Weiss, 2002; Munz et al., 1996; Parkinson et al, 1996). In spite of this, negative affect provides information that the interface between the individual and the environment is suboptimal and in need of improvement. As a result, it can serve to motivate behavior toward improving transactions with the environment, resulting in positive outcomes (e.g., process improvements) if appropriate actions are taken to reframe, remove, or improve the conditions that reinforce negative affect (George & Zhou, 2002). The differential predictive validity of positive and negative affect provides more evidence that affect operates through separate underlying motivational processes reflecting an approach-avoidance paradigm (Carver, 2001).
Affect, through the approach and avoidance transactional systems, has a pervasive effect on an individual's functioning, influencing perceptions, information processing, motivation, and behavior. In workgroups, affect has important implications for both the individuals that operate within that group and the group itself. Consequently, before discussing the GAPS model in depth, we must first understand how affect manifests itself in a group, and how individual-level affect develops into group affect constructs, which then reinforce individual-level affect.
FROM INDIVIDUAL TO GROUP AFFECT
Individual affective processes have consequences for workgroups, due to their effects on the individual members of the group and the group's affective norms (George, 1990). Group affective norms refer to those norms that develop within groups and govern the experience, management, and use of affect within the group. The GAPS model highlights three important affective norms that can develop within workgroups: group atmosphere, affect management, and decision frames. In subsequent sections, we discuss how these group af
7. AFFECT AND GROU P PROBLE M SOLVIN G 127
fective norms develop from individual affective processes and how they influence the GAPS model.
Group atmosphere refers to the general atmosphere or climate of the group (Bierhoff & Muller, 1999). A positive group atmosphere is one that is open, honest, and constructive, resulting from the consistent experience of positive affect by group members. That is, a positive group atmosphere is one in which approach processes play a significant role in helping the group to develop positive interactions among members (e.g., conflict has positive consequences and should be managed effectively). A negative group atmosphere is one that is tense and often destructive, resulting from the consistent experience of negative affect by group members. In other words, a negative group atmosphere is one in which avoidance processes play a significant role in shaping how group members relate to each other (e.g., conflict has negative consequences and should be avoided). The group atmosphere results from processes of mutual influence at a conscious level, whereby group members influence each other through explicit means (George, 1990). For example, if group members convey positive attitudes toward the group, those attitudes are likely to result in increased positive affect within the group (Grawitch, Munz, Elliott, & Mathis, 2001).
Additionally, group members may influence the affect experienced by other group members unconsciously, a process known as affect contagion (Doherty, 1998; Hatfield, Cacioppo, & Rapson, 1994; Neumann & Strack, 2000). Affect contagion occurs through unconscious mimicry processes, in which group members begin to mimic the affective expressions of other group members (e.g., facial expressions, postures, voice tone; Hatfield et al., 1994; Laird et al., 1994). Due to cognitive and physiological feedback mechanisms, this mimicry leads to affective convergence within the group (Hatfield et al., 1994). Over time, these processes result in consistent levels of affect within the group setting (also known as affective tone; George, 1990), which influences the group atmosphere (Bierhoff & Muller, 1999). Groups that develop a negative atmosphere engage in fewer prosocial behaviors and are less satisfied with the group (Bierhoff & Muller, 1999; George, 1990). Workgroups that cultivate consistent positive affect among their members create an open, honest, and constructive group environment, a key element in group problem-solving success.
Affect Management Norms
Affect management norms develop as group members reinforce each other's af
fect regulation strategies and refer to those processes sanctioned by the group
as acceptable means of affect regulation (Marks, Mathieu, & Zaccaro, 2001).
As members of the group begin to sanction implicitly and explicitly certain affect regulation behaviors, especially as those behaviors relate to successes and failures experienced by the group, affect management norms develop (Marks et al., 2001). These norms can have significant consequences on the ability of group members to manage both positive and negative affect effectively, most notably as that affect results from conflict, feedback, and evaluation. There is a direct relationship between affect management norms and affective outcomes, such as cohesion and satisfaction with the group (Marks et al., 2001), and work improvement processes (George & Zhou, 2002). Of importance to the GAPS model is that affect management norms instituted explicitly and implicitly by the group determine the type of affect that is accepted, promoted, and regulated actively by the group, creating an affective group norm. For example, if the group has few strategies for decreasing negative affect when it occurs (e.g., when conflict arises), then negative affect will have a stronger influence on the group. Although affect management norms do not influence performance outcomes directly, they serve to reinforce the experience and regulation of affect within the workgroup and may contribute to the group atmosphere (Bierhoff & Muller, 1999).
Decision frames represent the way in which workgroups frame potential opportunities and actions. Decision frames refer to the extent to which the group seeks to approach potentially positive outcomes or avoid potentially negative outcomes (Higgins, 1998). Research has suggested that positive and negative affect influence the decision frames of individuals (Higgins, 1998; Isen, 2000). Additionally, it is evident that a decision frame produces a response bias congruent with that frame (Friedman & Forster, 2001). For instance, at the individual level, positive affect promotes approach processes, resulting in increased motivation to take calculated risks that may bring potentially greater benefits, thus maximizing desired outcomes (promotion-focused frame). Negative affect promotes avoidance processes, resulting in increased motivation to accept low-cost, low-risk decisions, thus maintaining security and avoiding negative outcomes (prevention-focused frame). At the group level the promotion-focused frame leads to more calculated risks with the expectation of greater benefits for the group, whereas the prevention-focused frame leads to decisions to engage in less risky, more conservative alternatives (Levine, Higgins, & Choi, 2000). Therefore, promotion-focused frames consider the problem and the solution in terms of maximizing positive outcomes, whereas prevention-focused frames consider the problem and the solution in terms of minimizing negative outcomes. Little research has examined the relationship between affect and decision frames at the group level directly. However, if members of a group experience positive or negative affect consistently, result
7. AFFECT AND GROU P PROBLE M SOLVIN G 129
ing in a positive or negative atmosphere, the group may be more likely to adopt decision frames consistent with that atmosphere. Groups that have developed a positive group atmosphere may be more likely to use promotion-focused decision frames, whereas groups that have developed a negative group atmosphere may be more likely to use prevention-focused decision frames. Consistent positive or negative affect within a group may result in a consistent group-level frame that can affect how problems are solved and the subsequent actions that result.
THE GAPS MODEL
The previous sections defined individual affective processes and group affective norms. In this section we focus on the problem-solving process in order to describe the GAPS model more fully. The GAPS model is a theoretical model that integrates affect and group problem-solving research at each stage of the problem-solving process. First, we describe the four broad problem-solving stages and present key indicators and outcome measures for each stage. Then, we propose relationships between the components of the model in an attempt to synthesize previous research at the individual and group levels and suggest effects of affect on group problem-solving outcomes. Throughout our discussion we present five important propositions or key assumptions underlying the GAPS model.
Problem-Solving Stages Defined
The GAPS model presents four general problem-solving stages used by groups in organizations. The first stage of the problem-solving process is the problem identification stage. This stage emphasizes the adequate description of the problem and its seriousness (Orlitzky & Hirokawa, 2001). Of additional consequence is the extent to which the problem is defined as an opportunity to grow and achieve (promotion-focused frame) or as an opportunity to seek safety and security (prevention-focused frame; Levine et al., 2000). In many instances, management engages in problem identification and then assigns the identified problem to a project, reengineering, or problem-solving group to solve (Uhl-Bien & Graen, 1998).
There are at least two criteria for evaluating problem identification effectiveness: the decision frame applied to the problem and the relative importance of the problem. Because of the difference in emphasis using a promo-tion-focused versus a prevention-focused frame, promotion-focused frames lead to an increase in the types and number of ideas that are considered in both the brainstorming and solution development stages (Levine et al., 2000). A second important criterion at this stage refers to the relative importance of the identified problem. Serious problems are more likely to energize the problem-solving group to devote the time and energy required to develop effective solutions than are less serious problems (Orlitzky & Hirokawa, 2001). When selecting from a variety of potential problems on which to direct the effort of a group, an emphasis should be placed on solving more serious problems. Table
7.1 presents ways to measure the problem frame and the relative importance of
Proposed Indicators and Possible Measures
of Group Problem-Solving Criteria
Indicators Possible Measures
Problem Identification Stage
Problem frame Ratings of promotion-focused and prevention-focused qualities of the problem statement Rankings of problem statements from most to least promotion focused and most to least prevention focused Relative importance Rankings of potential problems from most to least important by relevant stakeholders (e.g., employees, management, customers) Ratings of problem importance by relevant stakeholders
Quantity of ideas (fluency) Count of the number of ideas generated Originality of ideas Ratings of the uniqueness of each idea
Potential implementation cost Ratings of potential solution cost by management Rankings of solutions from least to most cost Potential benefit Ratings of potential solution benefit by management Rankings of solutions from least to most beneficial Flexibility Cross-functional management ratings to assess the degree to which solution appeals to multiple stakeholder groups Management ratings of the complexity and multifaceted nature of the solution Elaboration Ratings or rankings of the number of ideas generated during brainstorming integrated into the solution Ratings or rankings to evaluate the completeness of a given solution (e.g., clarity)
Efficiency Estimates of time, budget Ratio of performance to time Quality Ratings of quality dimensions, either inductively or deductively derived
Note. The indicators and measures presented here represent some criteria for evaluating outcomes at the various stages of problem solving. They by no means provide an exhaustive list.
7. AFFEC T AND GROU P PROBLE M SOLVIN G 131
the problem in the problem identification stage through the use of ratings and rankings.
The second stage of group problem solving is the brainstorming stage. In this stage the group generates as many ideas as possible to be used in solving the problem, without evaluating the merit of those ideas (Kramer, Fleming, & Mannis, 2001; Offner, Kramer, & Winter, 1996). Therefore, the goal of brainstorming is to generate a list of possible ideas for solving the problem before proceeding onto the next stage.
Evaluating brainstorming effectiveness can include two important criteria: the quantity and the originality of ideas. The quantity of ideas generated by a brainstorming group is commonly referred to as fluency (Paulus, 2000). The larger the number of ideas, the more likely the group is to develop a multifaceted solution that accounts for multiple viewpoints (Paulus, 2000). Although generating a large number of ideas may be important at later stages of problem solving, generating original ideas is also important (Paulus, 2000). If a group generates novel or original ideas, then the group is more likely to develop a creative solution (Unsworth, 2001). Although some ideas may not seem feasible at the time, they provide creative input into the later stages of problem solving. Table 7.1 presents ways to measure fluency and originality in brainstorming through the use of frequencies and ratings.
The third stage of problem solving requires that the ideas generated during brainstorming be evaluated and synthesized into a solution or recommendation. This stage may result in the creation of multiple alternative solutions to be evaluated (Orlitzky & Hirokawa, 2001), or it may result in a single recommended solution. In some problem-solving groups, solution development may culminate only in recommendations to be evaluated by management. A group may also revisit various problem-solving stages as it receives feedback from various sources (e.g., management, the environment; Marks et al., 2001).
Solutions developed at this stage can be evaluated using at least four criteria: cost, potential benefit, flexibility, and elaboration. Cost refers to the estimated amount of resources required to effectively implement the solution, estimated as time, money, or manpower (Sundstrom, DeMeuse, & Futrell, 1990). Potential benefits may include increases in organizational goal attainment (e.g., increased return-on-investment), organizational capabilities (e.g., the establishment of in-house training facilities), or softer organizational measures (e.g., employee participation or commitment; Levine et al., 2000). Flexibility refers to the extent to which a solution considers multiple perspectives or contains a multifaceted approach (Paulus, 2000). A solution that addresses and integrates the views of multiple stakeholder groups has a high degree of flexibility. Thus, a new product solution that contains a customer-focused component (e.g., seek customer input), an employee-focused component (e.g., training in sales), and a marketing component (e.g., advertising) would be a more flexible solution than one containing only an employeefocused component. Finally, elaboration refers to the ability of the group to synthesize and integrate the ideas generated during brainstorming and its ability to articulate that solution (Paulus, 2000). The ability to integrate a variety of ideas generated during brainstorming into one solution and to articulate that solution becomes important if the solution is to be implemented effectively and have the desired outcomes. Table 7.1 presents ways to measure cost, potential benefit, flexibility, and elaboration through ranking and rating methods.
The final stage of problem solving involves the implementation of the solution. This stage is often the most complicated, time-consuming, and difficult (Floyd & Lane, 2000; Katzenbach & Smith, 1994). In some organizations, individuals outside the problem-solving group may be responsible for this final stage. Implementation requires knowledge of the system within which it is to occur, the ability to develop contingencies in the event of setbacks, the flexibility to experiment with alternate approaches, and the support (e.g., time, expertise, finances) required to institutionalize change (Uhl-Bien & Graen, 1998).
At this stage, at least two evaluation criteria are important. The first concerns the efficiency with which implementation occurred. Efficiency can refer to the amount of time required to implement the proposed solution, or it may refer to the degree to which the activities occurred under budget. The second evaluation criterion is quality. Quality implementation processes are those that allocate the appropriate resources to ensure that it occurs fully and effectively (Katzenbach & Smith, 1994; Sundstrom et al., 1990). For example, if an organization chooses to institute quality control processes, it must allocate the appropriate type and amount of resources (e.g., technology, training) to ensure that those quality control processes are implemented effectively. Table 7.1 presents ways of measuring implementation effectiveness.
As discussed, the various stages of problem solving follow a logical progression from problem identification to implementation. Proposition 1 of the GAPS model suggests that outputs at one stage serve as inputs into subsequent stages of the group problem-solving process. That is, the results of problem identification influence the ideas generated during brainstorming, which serve as the input into the solution, which serves as the input into implementation. Proposition 1 argues that when evaluating a solution, whether in research or applied settings, the number and originality of ideas generated during brainstorming will influence the quality of the solution generated by the group. Therefore, if groups do not generate enough creative ideas during brainstorming, the resulting solution will be negatively affected.
Although the group problem-solving stages are key elements in the GAPS model, the stages are influenced by and influence affect. The GAPS model proposes two specific bidirectional relationships between affect and group problem-solving performance. The bidirectional nature of these relationships reflects the transactional nature of the GAPS model. Proposition 2 suggests that both individual affective processes and group affective norms influence
7. AFFECT AND GROU P PROBLE M SOLVING 133
and are influenced by the problem-solving process. Proposition 2 suggests that not only do individual affective processes and group affective norms serve as input into the problem-solving process, but they can also be considered outcomes of the process. Outcomes at each stage of the problem-solving process serve as feedback to both the individual and the group about performance, the processes used to achieve performance, and the norms that have developed within the group. Therefore, positive performance at any one stage may result in increased positive affect that will influence the next stage in the process (George & Zhou, 2002). Negative performance, on the other hand, may result in negative affect, which may signal a need to adjust processes or norms within the group (George & Zhou, 2002). If the negative affect is not managed appropriately, it will become reinforced, creating a negative group atmosphere.
In the next subsection, we discuss in more detail the relationship between individual affective processes and performance at each stage of the problem-solving process. Then, we discuss the relationship between group affective norms and performance at each stage of the process. These proposed relationships were derived through the synthesis of both individual- and group-level research, as well as some hypothesized relationships.
Relationship Between Individual Affective Processes and Performance
Based on previous literature at both the individual and group levels, individual affective processes may influence performance at all four stages of the problem-solving process. With regard to problem identification, research has suggested that affect influences an individual's orientation toward approach or avoidance (Carver, 2001; Erez & Isen, 2002). Positive affect facilitates an orientation toward approach, and negative affect facilitates an orientation toward avoidance. Consequently, during problem identification, positive affect is likely to foster a motivation among individual group members to perceive the problem as an opportunity to approach potential positive outcomes, whereas negative affect is likely to foster a motivation among individual group members to perceive the problem as an opportunity to avoid potential negative outcomes. Similarly, positive affect promotes an orientation toward more serious or important problems and away from less important problems. Grawitch, Munz, Elliott, and Mathis (2003) induced positive and neutral affect within workgroup members and asked them to identify important problems within their organization. Grawitch et al. (2003) obtained peer rankings of the identified problems and determined that when group members were induced to experience positive affect, their groups identified more important problems than did groups whose members were induced to experience neutral affect. The results suggested that positive affect orients group members toward more important problem domains and away from less important domains.
With regard to brainstorming, positive affect increases both the overall number of ideas and the number of original ideas that are generated by individuals (Isen, 2000). Additionally, this research has suggested that negative affect has no effect on individual-level idea generation. Isen (2000) argued that increased cognitive flexibility results from the experience of positive affect, which leads to the generation of a greater number of ideas. Although there have been few studies conducted at the group level, our research suggests that when group members are induced to experience positive affect, there is little or no effect on fluency (i.e., the absolute number of ideas generated; Grawitch et al., 2003; Grawitch, Munz, & Kramer, 2003). Instead, when asked to brainstorm ideas to solve a particular problem, induced positive affect increases the originality of the ideas (Grawitch et al., 2003). In comparison, induced negative affect has no effect on group brainstorming outcomes as compared to neutral affect (Grawitch et al., 2003).
In terms of solution development, no research has examined the relationship between individual-level affective processes and performance at the group level. Instead, using individual-level research as a source of hypotheses, we propose four sets of relationships between affect and performance at this stage of problem solving. First, positive affect promotes expanded cognitive functioning within individuals, allowing them to synthesize information more effectively (Isen, 2000; Isen & Means, 1983) and to consider a wider variety of options before making a decision (Fredrickson, 2001). Therefore, we hypothesize that positive affect permits group members to more efficiently synthesize and integrate the ideas they generated during brainstorming, increasing solution elaboration. Second, because positive affect allows individuals to see the broader system and process information at a systems level (Gasper & Clore, 2002), we hypothesize that positive affect permits group members to consider a systems perspective and integrate the concerns of multiple stakeholders, increasing solution flexibility. Third, positive affect increases the motivation to approach positive outcomes, increasing the desire for growth and achievement (Carver, 2001; Higgins, 1998). Hence, we hypothesize that positive affect motivates group members to focus on maximizing positive outcomes, leading to solutions that possess a higher potential benefit, which may also increase the potential implementation cost. Finally, because negative affect motivates individuals away from potential threats and toward safety and security (Carver, 2001; Higgins, 1998), we hypothesize that negative affect leads to solutions that possess less potential risk (and benefit), but also less potential cost. These four hypotheses require focused empirical attention in future research.
The influence of individual affective processes on implementation may be somewhat more complex than the influence on problem identification, brainstorming, and solution development. In many organizations one group generates the solution, and a second group implements it (Katzenbach & Smith, 1994). The effectiveness of that implementation may be influenced by
7. AFFECT AND GROU P PROBLE M SOLVING 135
the quality of the solution that was developed. If the solution is complex and requires broad participation, implementation processes may require the addition of more group members or the development of subgroups (Katzenbach & Smith, 1994). Therefore, implementation may be viewed as a series of activities, resulting in assessments of each activity required for effective implementation.
Our research suggests that positive affect has an effect on a group's ability to implement its own solution. Grawitch et al. (2003) found that groups whose members were induced to experience positive affect received higher overall performance scores and were more efficient in their implementation processes than were control groups. Consequently, it may be that positive affect allows groups responsible for implementation to more effectively (a) determine how to implement a solution, and (b) anticipate system-wide effects and plan contingencies.
Relationship Between Group Affective Norms and Performance
Although the direct relationship between affect and group performance at the various stages of problem solving result from individual affective processes, group affective norms also play a role in problem-solving performance. Proposition 2 of the GAPS model suggests that individual affective processes influence the development of group affective norms, which in turn influence the individual affective processes operating within the group setting. This proposition is based on the premise that group members influence each other's affective experiences, which in turn influence the affective norms of the workgroup (George, 1990). As group members begin to sanction others' affective reactions and affect management strategies, those affective reactions and strategies influence the group's affective norms. For example, if group members sanction venting as an affect management process either explicitly (e.g., members encourage others to blow off steam when upset) or implicitly (e.g., members listen attentively as another member vents), then venting becomes an accepted means of regulating negative affect. Additionally, if the group does not develop effective strategies for regulating negative affect when it occurs, the group is likely to develop a more negative group atmosphere. As affect regulation and group atmosphere norms develop, they begin to influence the affective experiences of members of the workgroup, creating a consistent level of affect within the group (George, 1990).
For group affective norms to influence the problem-solving process, they must have time to develop. Proposition 3 of the GAPS model directly follows Proposition 2 and states that as groups mature, group affective norms become more influential and individual affective processes become less influential. In the GAPS model, this is represented by the decreasing density (thickness) of the arrows stemming from individual affective processes and the increasing density (thickness) of the arrows corresponding to group affective norms. This is not to say that individual affective processes disappear as groups mature. Instead, group norms drive individual processes (Tuckman, 1965). That is, individual affective processes become more a product of group affective norms than individual-level factors. Furthermore, as stable, long-term groups revisit the problem-solving process (e.g., a long-term group focuses on solving a new problem), the influence of group affective norms may remain consistent.
There has been little published research examining the influence of group affective norms on problem-solving performance at the group level. When discussing the relationship between group atmosphere and decision frames to problem-solving performance, we are discussing hypothesized relationships based on our review and understanding of the literature.
The Importance of Group Atmosphere to Performance. The group atmosphere has the ability to influence all four stages of the group problem-solving process. A positive group atmosphere leads to a more open and constructive environment, whereas a negative group atmosphere leads to a critical, harsh environment (Bierhoff & Muller, 1999). Based on the work of Bierhoff and Muller (1999), we hypothesize that during the problem identification, brainstorming, and solution development stages, the group atmosphere determines how forthcoming and open group members are in presenting their ideas. In the problem identification stage, a positive group atmosphere allows group members the opportunity to discuss organizational problems more openly, resulting in the identification of more serious problems that require immediate attention. Because of its destructive qualities, a negative group atmosphere does not permit group members to discuss potential problems openly, resulting in the identification of less difficult and more superficial problems.
In the brainstorming stage, we propose that a positive group atmosphere leads individuals to provide as many ideas as possible, without fearing criticism from other group members, but that a negative group atmosphere leads group members to censor their ideas for fear that other members will criticize those ideas. The fear of criticism in idea-generating groups is also known as evaluation apprehension and may cause groups to generate fewer ideas than they would without the presence of evaluation apprehension (Diehl & Stroebe, 1991; Kramer et al., 2001; Offner et al., 1996). Our position is that a positive group atmosphere results in more ideas overall and more original ideas than a negative group atmosphere.
In the solution development stage, we hypothesize that the group atmosphere has a similar influence. A positive atmosphere fosters a more open and constructive environment in which to evaluate ideas (Bierhoff & Muller, 1999) and reach the optimal solution. Because of the open and constructive nature of the positive group atmosphere, the resulting solution may possess high flexi
7. AFFECT AND GROU P PROBLE M SOLVING 137
bility, elaboration, and an optimized cost-benefit ratio. Conversely, a negative group atmosphere leads to destructive criticism, causing group members to censor themselves to avoid confrontation. This may result in a solution that has low flexibility (limited in addressing the views of multiple stakeholder groups) and low elaboration (addressing only a subset of the organizational system rather than the total system).
Finally, in the implementation stage, a positive group atmosphere should foster teamwork and cooperation among group members because of the mutual respect that is created by that atmosphere (Katzenbach & Smith, 1994). Alternatively, a negative group atmosphere should undermine the cooperation and teamwork required for effective implementation. As a result, a positive group atmosphere may result in greater implementation efficiency and quality, whereas a negative atmosphere may inhibit both efficiency and quality.
The Importance of Decision Frames to Performance. The group atmosphere influences group problem-solving effectiveness through the constructive or destructive nature of group interactions, but the decision frame may influence problem-solving effectiveness through more cognitive means. The decision frame refers to the extent to which the group seeks to approach (or avoid) potentially positive (or negative) outcomes and serves as a focal point for ideas and solutions (Levine et al., 2000). At the problem identification stage, the decision frame influences the ways in which the group identifies and defines the problem. A promotion-focused (or prevention-focused) frame motivates the group to attain positive (or avoid negative) outcomes. Therefore, the decision frame used by the group in the problem identification stage serves as a source of approach or avoidance motivation throughout the problem-solving process.
In the brainstorming stage, the way the problem is framed influences the types of ideas that are generated (Levine et al., 2000). For example, a group considering the issue of customer satisfaction might frame the problem as "How do we reduce customer complaints?" and consequently, conservative ideas to improve the status quo may result (e.g., lower defect rates, better sales training) that are directed at avoiding negative outcomes (i.e., customer complaints). If the group frames the problem as "How do we increase customer satisfaction?" many of the same ideas may result, but additional innovative ideas may also surface, especially ideas that may change the status quo (e.g., expand our product lines, offer additional services), directed at attaining positive outcomes (i.e., customer satisfaction). Therefore, we hypothesize that at the brainstorming stage, groups using a promotion-focused frame produce more ideas and more original ideas than groups using a prevention-focused frame.
The expanded problem conceptualization is also likely to influence the solution developed by the group. Groups using a promotion-focused frame adopt riskier solutions than those using a prevention-focused frame (Levine et al., 2000). These promotion-focused solutions may possess higher potential benefits for the organization (or a lower cost-benefit ratio) and greater flexibility. When implementing the solution, a promotion-focused frame, because it emphasizes the maximization of achievement and growth, is more likely to lead to experimentation than a prevention-focused frame, which emphasizes the maximization of safety and security (Floyd & Lane, 2000; Levine et al., 2000).
As discussed, both individual and group affect play a role in the various stages of group problem solving. Many of the relationships just discussed are hypothesized relationships, because of the lack of empirical research. When considering the application of the GAPS model to organizations, an additional issue that must be addressed is context.
The Importance of Context in Understanding GAPS Relationships
Problem-solving groups operate within a specific context that includes both organizational and group factors (Sundstrom et al., 1990). These factors may alter the influence of affect at various stages of the problem-solving process. As a result, contextual factors can have implications for the application of the GAPS model.
Two organizational context factors that are especially relevant to the GAPS model are performance management systems and organizational culture. These two factors provide a strong normative influence on behavior. Performance management systems influence the extent to which groups are rewarded for group performance. The system used to reward performance can foster higher levels of interdependency among group members, making processes of mutual affective influence more possible. Further, the extent to which the organization rewards and reinforces group performance outcomes, especially solution quality and implementation effectiveness, can serve as a continuous source of feedback (George & Zhou, 2002), creating stronger relationships between individual affect and group performance.
Organizational culture is another contextual factor that can influence relationships in the GAPS model. First, cultural norms about the expression of affect in the workplace may influence the development of group processes to effectively manage that affect. Many organizations assume that affective expression is incongruent with the desired tone of the workplace. Consequently, employees are often discouraged from expressing their genuine affective states, when those affective states violate organizational norms. Instead, employees are often compelled to express only those affective states that are
7. AFFECT AND GROU P PROBLE M SOLVING 139
sanctioned by the organization (i.e., emotional labor; Fisher & Ashkanasy, 2000a; see also previous chapters of this book). If individuals within a workgroup are implicitly or explicitly discouraged from affective expression in the workplace, then the group will have difficulty developing effective affect management norms.
Additionally, culture can be a determinant of initial affective processes that individuals bring with them to a new group setting. For example, if an organization regularly uses workgroups to identify solutions but then never actually implements those solutions, the organization may establish norms that participation in such groups is a waste of time (Schein, 1990). In those cases negative affect may pervade problem-solving groups initially, and over time, this negative affect may create a negative group atmosphere and few positive outcomes.
Group contextual factors can also influence the GAPS model. Although many group-level factors can influence the development of affect in workgroups, we discuss two important factors: (a) the duration of the group's existence, and
(b) the functional nature of group composition.
The duration of a group's existence has been proposed to play an important role in the development of group affective norms. Grawitch et al. (2003) argued that there are differences in the expected duration of a group's existence from the moment the group is designed or conceived. That is, some groups are designed to be long-term workgroups, chartered to complete on-going or recurring tasks over a prolonged period of time or to complete large-scale projects within the organization. Many of these groups can be thought of as long-term workgroups. These groups accomplish the daily business of the organization, and they are generally responsible for maintaining or increasing some desired performance standard. Alternatively, other groups tend to be short-term in nature, or temporary workgroups. These workgroups are brought together for a finite period of time within which they are chartered to complete one or more given tasks (Uhl-Bien & Graen, 1998), such as launching a new product. Once the assigned task is completed, temporary workgroups disband, although some may develop into long-term groups. Often, short-term temporary workgroups do not have the time required to develop group processes, reducing or negating the relative influence of group affective norms on problem-solving performance.
The functional nature of group composition can also influence the GAPS model. Some groups are composed of members from the same functional unit (e.g., a group of marketing analysts). These workgroups are organized according to "similarity in the skills, expertise, and resource use of members" (Uhl-Bien & Graen, 1998, p. 341). They have open lines of communication and coor
dination, permitting members to better interact with each other because of their common background and training. Conversely, other groups are cross-functional. They are composed of a variety of individuals from different functional areas within the organization and often bring together individuals with dissimilar, but complementary, skills, educations, and backgrounds (Katzenbach & Smith, 1994).
Because they come from different functional units, members of cross-functional groups often represent different subcultures within the organization, bringing potentially conflicting normative influences to the group setting. This may increase the amount of time required for cross-functional groups to develop effective norms for regulating and using affect (Uhl-Bien & Graen, 1998). Because cross-functional groups often have difficulty communicating and resolving conflict (Uhl-Bien & Graen, 1998), they also may experience greater levels of negative affect in the early stages, which may lead to a negative group atmosphere or prevention-focused decision frame. Because of the difficulties that exist in cross-functional groups, we propose that functional groups develop more effective affect regulation norms in significantly less time than cross-functional groups.
The organizational and group context has implications for the relationships in the GAPS model. However, the model itself suggests several application and research directions. That is, although the GAPS model does require future research, some of the relationships discussed could be capitalized on by problem-solving groups in organizations. In the final subsections we focus on both practical and research implications.
Practical Implications of the GAPS Model
Although the GAPS model is a theoretical model and requires extensive future research, there are some practical implications that can be drawn from its foundational research. First, drawing on previous individual-level research, the GAPS model indicates that both positive and negative affect possess functional utility and therefore must be capitalized on in the problem-solving context (George & Zhou, 2003). Positive affect offers the benefits of expanded cognitive processing, which can lead to more creative solutions (e.g., Fredrickson, 2001; Isen, 2000). Consequently, groups should maximize, to the extent possible, a consistent level of positive affect within the group, including the establishment of a positive group atmosphere, effective means of regulating affect, and promotion-focused decision frames.
The experience of negative affect also provides functional utility within the problem-solving context, as long as negative affective experiences do not become a regular occurrence within the workgroup. Previous individual-level research suggests that negative affect can serve as information regarding the effectiveness of processes to achieve performance (George & Zhou, 2002).
7. AFFECT AND GROUP PROBLE M SOLVING 141
Therefore, occasional negative affective experiences within the group setting may point to improvements required in some area of group functioning. Groups should attempt actively to manage negative affect when it surfaces to manage and improve the group proactively and effectively (Marks et al., 2001). If the informative value of negative affect is avoided or only managed passively, then negative affective experiences within the workgroup may become more consistent, resulting in a negative group atmosphere (Bierhoff & Muller, 1999). Interventions can be designed to: (a) educate workgroup members on the benefits of both positive and negative affect, (b) train groups in active affect management strategies, and (c) facilitate the development of positive individual processes and group norms.
Second, the GAPS model specifies that problem-solving groups function best in contexts that facilitate group effectiveness. Contextual factors (e.g., performance management systems, culture) that diminish the capacity of problem-solving groups to achieve optimal performance should be minimized. Alignment of organizational context factors facilitates optimal levels of performance in a variety of workgroups, especially problem-solving workgroups (Sundstrom et al., 1990). For example, if a problem-solving workgroup is expected to function as a cohesive unit, there should be some incentives for group performance, rather than an emphasis only on individual performance. Key organization members must examine the expectations they have about problem-solving groups (e.g., how information is shared, commitment to the group and group performance), and promote organizational context factors that enhance the desired results.
Opportunities for Future Research
There are four broad opportunities for future research implicated by the GAPS model. First, an opportunity exists to test the theoretical links in the model. For example, research has yet to examine fully how group affective norms influence the various stages of problem solving, and until recently, little research has focused on the relationship between the functional aspects of affect and problem solving at the group level (Grawitch et al., 2003). Second, future research should test the four propositions that we set forth based on the model. For example, future research may determine whether the quality of outcomes at one stage of group problem solving (e.g., brainstorming) influences the quality of outcomes at the next stage of problem solving (e.g., solution development). Or, future research may test directly the proposition that individual affective processes influence the development of group affective norms, which reinforce and maintain individual affect. Third, by testing the links in the model and the propositions of the model, future research will refine and explicate more fully the ways in which affect (both positive and negative) influences group problem solving. Finally, the specific mechanisms through which affect influences group problem solving need to be examined. Although we have suggested several potential explanations for the relationship between affect and group problem-solving performance (e.g., informational and motivational effects), future research should more closely examine the mechanisms through which affect influences group problem solving at each stage of the process. For example, future research may determine that positive affect increases brainstorming performance due to its effects on the cognitive flexibility of group members rather than through an increase in the approach motivation of group members.
The purpose of this chapter was to discuss the implications of affect on group problem-solving processes and performance using a transactional model. The GAPS model provided us a means of examining functional affective processes that occur at the individual level and the opportunity to explore and hypothesize additional relationships between group-level affective norms and group problem-solving performance. We discussed in some detail the relationship between individual affective processes and group affective norms, and we proposed a reinforcing system in which individual affective processes influence the development of group affective norms which reinforce individual affective processes. We offered a series of hypotheses and propositions related to the GAPS model. Finally, we explored both the practical and research implications of the model. Although it is a theoretical model that has received little directed empirical attention, we believe the GAPS model provides a starting point for examining the relationships of both individual and group affect with group problem-solving performance.
The authors thank Tim Huelsman, Tom Kramer, and Bob Rubin for their helpful comments on previous drafts and presentations of this chapter.
The purpose of this theoretical chapter is to suggest a model of humor in the social sharing of emotion at the workplace. With this in view we first briefly review humor research with special reference to organization science. We then introduce the theory of social sharing of emotion (SSE), that is, the narrative framing of emotional work events in connection with the production of humor at the workplace. With the help of this theory we develop a model to clarify the process and the relevant antecedents and outcomes of humor in SSE at the workplace. We discuss in some detail the factors in our model in two phases of SSE. We contend that humor is a frequent product of SSE due to the specific social nature of the workplace. We conclude by suggesting future research directions pertaining to our model.
Emotional events are often molded in a humorous way in the course of frequent retelling after the event. An instructive example, in terms of our concerns in this theoretical chapter, appears in Linstead (1985) with an analysis of a particular case. A worker on a fruit-pie machine was left with a mutilated hand after an accident on the job. His amputated finger could not be found, and 4,000 pies were discarded. In the days and weeks that followed, his coworkers reinvoked the emotional event among themselves. Linstead identified comments, remarks, discussions, and jokes in which the emotional event was reframed in a humorous way. One coworker recalled, "I can't remember 'em all now, but when it happened there were loads of jokes about it. It sounds terrible, dun' it, but tha'd hear a new 'un every break. I wish I could remember 'em" (Linstead, 1985, p. 753). One such joke was, " 'They were going to get some 'finger-hunter' stickers made and pack them 4000 pies' (promotion boxes usually bore a sticker marked 'Bargain Hunter')" (Linstead, 1985, p. 754). As Linstead explained, humor was added to the emotional event by the workers in order to maintain their own status, competence, and independence, and to cope with their particular work environment. In the end the event was well recognized and well established in the sensemaking of the workgroup. What is striking about this example is the molding of a negative emotional event through the injection of humor. In addition to the social effects described by Linstead, humor seems to offer a means whereby the individual can talk about strong emotional events in the social sphere of work.
Repeated storytelling produced well-established narratives, and the development and enactment of such narratives are by no means uncommon in organizational settings (Boje, 1991; Czarniavska, 1997). What emerge are caricatures reflecting the culture and beliefs of an organization, often in the shape of insights arising in organizational life and conveyed in a fairly simple and humorous, although profound, manner (Hatch, 1997; Hatch & Ehrlich, 1993). It is the initiation and development of such narratives and the humor they contain that interest us in this chapter. Research on narratives and the way they develop in the stories has largely neglected the humorous element or has taken it for granted. We are intrigued by the way a humorous view of emotional events—what we could call "humor" in a broad sense—inserts itself into the everyday interactions of the members of an organization.
In recent years emotion, together with cognition, has become recognized as a persistent factor that can help us to understand organizational behavior in areas such as leadership, job satisfaction, employee well-being, and so on (Ashforth & Humphrey, 1995). From this growing interest, theories began to emerge about emotional labor, emotional intelligence, emotional events at the workplace, and various other related concepts (Ashkanasy, Hartel, & Daus, 2002b; Weiss & Cropanzano, 1996). In this chapter we explore the progression whereby emotional events become humorous narratives as a result of interpersonal and intrapersonal processes during the social sharing of those events.
Before going into detail about the way humor may be created and shared in narratives, we need to define and qualify some of the terms to be used in this chapter. Emotion as we understand it is short-lived, of considerable intensity, and directed toward an object, an individual, or a collective. It is regarded as a reaction to the experiencing of an event (Frijda, 1993). Mood, on the other hand, is seen as less intense, longer lasting, and less focused on a specific target. In this chapter, we adopt the concept of emotional events as described by Weiss and Cropanzano (1996). Event refers to a change in an individual's present experience, and our emphasis is on the emotional consequences of the event. For our basic framework we draw on social psychological research, but also make occasional forays into ethnographical or sociological terrain, where
8. NONSENS E MAKES SENSE 145
making sense of what has occurred takes precedence over the demonstration of benefits or drawbacks (Weick, 1995).
In the following pages, we first briefly review humor research, paying particular attention to organization science. We then introduce the theory of the social sharing of emotion (SSE), that is, the narrative framing of emotional work events (Rime, 1995; Rime, Finkenauer, Luminet, Zech, & Philippot, 1998; Rime, Philippot, Boca, & Mesquita, 1992), and its connection with the production of humor at the workplace. With the help of this theory we develop a model to clarify the process of humor production and the relevant antecedents and outcomes of humor in SSE at the workplace. We then discuss in detail the factors in our model in two phases of SSE. We contend that humor is a frequent product of SSE at the workplace. We conclude by suggesting future research directions pertaining to our model.
The study of humor goes back more than 2,000 years. Interest in the topic has spread across disciplines such as philosophy, sociology, psychology, psychiatry, communications, and anthropology (Roeckelein, 2002). Despite centuries of widespread interest in the subject, the study of humor remains fragmented. This lack of concentration is particularly noticeable in management and organizational research. In a review of humor studies within the field of management, Duncan, Smeltzer, and Leap (1990) commented that this kind of research was in its infancy compared with its status in other disciplines such as sociology and that its scattered and sporadic nature is also reflected in the definitions that researchers have used in their studies.
The definitions of humor given by Roeckelein (2002) are concerned first and foremost with explaining of bodily fluids, second with temperament, and only after that with what we generally think of as "humor" today. This reflects the origin of humor in the philosophy of Ancient Greece. The Greek physician Hippocrates (460-370 BC) believed that an imbalance among the four bodily fluids or "humors" (blood, phlegm, black bile, and yellow bile) resulted in illness and pain. Galen (200-130 BC), the Greek physician and philosopher, later proposed that four basic "temperaments" (sanguine, phlegmatic, choleric/bilious, and melancholic) reflect the state of the four bodily humors more accurately. These Greek theories thus make a connection between physiological and psychological conditions. In medieval times humor was frequently considered to be indicative of ignorance and foolishness (Roeckelein, 2002). In modern colloquial language, however, the word humor is described in many different ways—for example, as a stimulus, a response, or a disposition. A sense of humor has become a very desirable trait and is regarded as part of a healthy personality. (It should be noted that these observations, especially that concerning the historical status of humor, may be particularly appropriate to Western society; Roeckelein, 2002.)
According to Martin (2000), humor has cognitive, emotional, behavioral, social and psychophysiological aspects. Lefcourt and Martin (1986) classified various theories of humor into three categories: arousal theories (e.g., Freud, 1928); incongruity theories (e.g., Kant, 1790/1914); and superiority theories (e.g., Hobbes, 1650/1994). The basic idea underlying arousal theories is that certain types of mirthful experiences can reduce tension and negative emotional energy that have built up over time. In particular, humor helps release energy associated with negative emotions and has been assigned a unique value due to its "liberating" and "elevating" effects (Freud, 1928). Freud was writing essentially about the self-deprecating and derisive kind of humor that transforms a negatively affective situation into a laughable one. He also claimed that the fundamental quality of humor consists in the belittling of a painful reality. In other words, humor occurs when a serious situation is viewed with a playful eye, so that it appears less grave. Since Freud's time, almost all theories of humor include the idea that a changed cognition in the experience of humor generates pleasure and laughter. According to incongruity theories, people perceive humor in the unexpected juxtaposition of two disparate ideas, concepts, or situations. There has been considerable debate about whether a solution to such incongruity is necessary to the individual's experience of humor (e.g., Nerhardt, 1976; Suls, 1972). Regardless of this debate, incongruity theories clearly indicate the importance of the perspective angle adopted in the humor-appreciation process. Finally, superiority theories can be traced back to the days of Plato (428-348 BC) and Aristotle (384-322 BC). But it was Thomas Hobbes (1650/1994) who introduced this approach into the existing theory, maintaining that laughter was born of a sense of superiority over the inferior quality of others or of oneself in the past. Further, this type of humor usually conveys a sense of aggression and hostility (Hobbes, 1650/1994). However, after reviewing research of the positive effect of humor on well-being (e.g., Levine, 1977; Mishkinsky, 1977), Lefcourt and Martin (1986) contended that even in the form of expressing superiority, humor does not signal an intention to put other people down, as much as a desire to raise their own self-esteem and personal efficacy.
In summary, different theories of humor converge in the common proposition that humor provides individuals with a chance to view a particular situation from a different angle, thus creating a sense of relief and control. Studies have recorded positive physiological effects stemming from humor, such as raising the threshold for physical discomfort (Zillmann & Stocking, 1976). Similar studies have elaborated also on the connections between humor/ laughter and human anatomical/cerebral functions and development, and confirmed that humor is related positively to physical and emotional well-being (see Roeckelein, 2002).
8. NONSENS E MAKES SENSE 147
With this in mind we have defined humor as the ludicrous or absurdly incongruous that is intended to induce laughter or amusement, often displayed in exaggeration or eccentricity (cf. Roeckelein, 2002). Humor can thus be reflected in objects or the manifestations of humor (e.g., cartoons or toys), as well as in different forms of communication and mental representation (stories, jokes, or songs). Our focus, however, is on verbal accounts in face-to-face communication, as this is the most personal and common occasion for the spontaneous production of humor.
When it comes to different kinds of humor, we adapt Fry's (1963) classification that differentiates between "canned," "situational," and "practical" jokes. Fry's classification is based on the joke's position within the context in which it is told. In this chapter we do not equate "humor" with "joke" because the latter is a form of humor, and is thus a narrower concept. Instead we extend and apply Fry's (1963) framework, in that we distinguish between canned, situational, and practical humor (see Fig. 8.1).
If humor is presented with little obvious relationship to any ongoing human interaction, it qualifies as canned humor. For example, people may recall and share a humorous story from a magazine or book in the middle of a conversation, in order to amuse and entertain. Situational humor originates in the ongoing interpersonal process. Situational humor is thus spontaneous and created on the spot, whereas canned jokes are known before they are told, often demonstrating the joke-teller's conscious intention to induce a sense of mirth and laughter in the audience. In contrast, practical humor is both intentional and spontaneous in that the joker intends the humor, but its content is not predetermined; rather, it develops as the interaction unfolds (cf. Fry, 1963).
FIG. 8.1. Illustration of the relationship between situational, practical, and canned humor.
HUMOR AT THE WORKPLACE
Humor at the workplace has not received much attention among organizational scholars (exceptions include Collinson, 1988, 2002; Duncan, 1982; Duncan & Feisal, 1989; Duncan et al, 1990; Hatch, 1997; Hatch & Ehrlich, 1993). Because of the mirth that humor arouses and the lighthearted aura it often conveys, researchers tend not to take it very seriously. It is assumed that in organizations humor is a peripheral phenomenon, unrelated or only remotely related to such things as profitability, efficiency, staff turnover, job satisfaction, and so on, all of which clearly impinge on organizational effectiveness. Although in certain situations humor can be regarded as "good" and useful, few areas of organizational life actually call for it. Moreover, humor is an elusive concept that escapes precise definition; its widespread use and imprecise nature make it difficult to tie it down in models and theories. In reality, however, organization members come across humor almost daily in their work life. It can even be pretty pervasive, particularly in various types of organizational communication—live conversations, e-mails, written material, and so on—or attaching itself to certain objects or events.
Interest in humor, joking, and laughter at the workplace began to grow in the late 1950s and early 1960s (Duncan et al., 1990). The early researchers tended toward an ethnographic and sociological stance. Roy (1960), for example, reported his personal experience as a participant-observer in a small group of factory machine operatives. In an exploratory analysis he found that informal interactions, including humorous exchanges like "horseplay," helped to improve job satisfaction. In the few publications that began to call attention to humor at the workplace from the 1980s onward, humor was often defined in terms of stimulus, that is, something that can be interpreted as humorous, and response, usually laughter. Malone (1980, p. 357), for example, adopted Chapman and Foot's (1976) description of humor as "a process initiated by a humorous stimulus, such as a joke or cartoon, and terminating with some response indicative of experienced pleasure, such as laughter." The study of humor at the workplace has focused primarily on joking behavior in a group context (e.g., Duncan, 1982; Duncan et al., 1990). Here, humor has also been closely linked with communication in a group context and used interchangeably with "joke" (Duncan, 1982; Duncan & Feisal, 1989; Duncan et al., 1990). In addition, humor has been investigated as an artifact within the culture of an organization (Dandridge, Mitroff, & Joyce, 1980; Vinton, 1989).
Researchers generally seek to suggest ways in which management can foster humor with a view to improving the effectiveness of their organizations, although the feasibility of such an enterprise has not gone unchallenged (Collin-son, 2002).
Although sociological, psychological, and ethnographic perspectives on humor have been adopted in previous management research, only rarely have the
8. NONSENS E MAKES SENSE 149
more theory-based psychological mechanisms been explored. Specifically, we believe that humorous accounts are created as individuals share emotional events with other members of their organizations. These events may then be spread further as one member passes them on to another, who in turn does the same, and so on. The humor contained in the original account is thus relayed, and probably modified by all these others as the sharing proceeds. Some of the humorous stories eventually become part of the organization's narrative corpus, its lore, reflecting certain characteristics of the organizational culture and introducing these to newcomers, as well as reverberating among the veteran members of the organization in a process of repetitive sharing. The emotional event thus provides the content and background for the production of humor. Hence, in this chapter we focus on the role of emotion in the creation and development of humorous narrated accounts.
SOCIAL SHARING OF EMOTION
The SSE Process
The social sharing of emotion (SSE) is conceptualized as an interpersonal process that occurs following the experience of some life event of emotional significance to an individual (Rime et al., 1998). This line of research emerged from an interest in traumatic events, such as serious accidents, and the sudden death of a spouse (Rime et al., 1992). People often have an urge to share their feelings and thoughts with others after the occurrence of some major negative life event. Rime et al. (1992, 1998) argued that this phenomenon is not exclusive to major emotional events, but that it applies to daily emotional experiences as well. They identified five arguments for their own conjectures as to why social sharing occurs. Briefly, these are:
Early studies of SSE used a procedure involving the recall of critical incidents (Rime, Mesquita, Philippot, & Boca, 1991). Experimental studies involving the viewing of film excerpts were also conducted to address the question of memory bias and to corroborate the connection identified between emotional intensity and SSE in correlation studies (Luminet, Bouts, Delie, Manstead, & Rime, 2000). Other studies were conducted for different age groups and across cultures and in relation to personality (see Rime et al., 1998). The general conclusion was that SSE is a universal phenomenon that can be observed in both men and women, young and old, and across cultures. Most people share their experiences with someone else, if possible almost immediately after their occurrence, and they may continue sharing them with several receivers on more than one occasion (Rime et al., 1991). A general positive relationship was also found between the intensity of the emotion and the frequency of social sharing (Luminet et al., 2000).
SSE at the Individual and Collective Level
To understand how SSE connects the individual and the collective levels it is important to make a distinction between two steps in social sharing. Primary social sharing of emotion (PSSE) is conceptualized as the basic form of sharing in which individuals repeatedly share their own experience with different receivers (Rime et al., 1991). Secondary social sharing of emotion (SSSE) arises from the fact that hearing about an emotional event is itself an emotional event for those with whom the events have been shared. The receivers in PSSE may then share the heard story with other people, in much the same way they pass on rumors or gossip (Christophe & Rime, 1997). We use initiator and receiver to refer to the two parties involved in sharing. The initiator is the one who shares the emotional event, and the receiver is the one with whom the experience is shared. In PSSE, the initiator is the one who experienced the emotional event and produces the humor. In SSSE, the initiator is the one who presents or reproduces the humorous slant they received in an earlier sharing. PSSE refers to "repeated reproduction" on the part of the individual who actually experienced the emotional event, whereas SSSE is "serial reproduction" of vicarious experience that comes with the relaying of information from A to B to C, and so on. PSSE focuses on the behavior of a single person who tells the same story repeatedly to one or more recipients. SSSE, on the other hand, focuses on a chain of individuals connected by the sharing of the same emotional event. Narratives, rumors, and gossip are common examples of the kind of thing reproduced in the SSSE.
One of the theoretical assumptions in explaining why SSE occurs is that people have a need to organize their emotional experiences cognitively (Rime et al., 1998). Hence, sharing serves to process and complete emotional memory. Studies of shared and kept-secret memories (Finkenauer & Rime, 1998)
8. NONSENSE MAKES SENSE 151
found that memories of unshared emotional events were associated with a more extensive search for meaning and a greater effort to get some order in one's mind about what had happened. In essence this is what Weick (1995) called sensemaking. The underlying idea of sensemaking is that reality is an ongoing accomplishment that emerges from efforts to create order and make retrospective sense of what occurs. One source and stimulant for sensemaking consists of the emotions that are elicited by significant interruptions of the everyday life of individuals (Weick, 1995, p. 46). In PSSE, individual sensemaking appears in the initiator as an explicit effort to tell a coherent story about an emotional event to the recipient and, implicitly, to themselves. The listener makes sense of the story as well, and the sensemaking activity persists throughout the process of SSSE (Rime et al., 1998).
Consequently, the iterative process attributed to PSSE and SSSE leads to collective sensemaking, that is, the social construction of what is taken for granted as part of a shared reality (Berger & Luckmann, 1966; Weick, 1995). Collective sensemaking is inherent in social sharing, namely, in the "construction and dissemination of social knowledge on emotion" (Rime et al., 1998). This is built largely on the receivers' processing of information and the subsequent SSSE. Receivers retain information about the shared event in light of their own emotion schemata (Shaver, Schwartz, Kirson, & O'Connor, 1987). These schemata contain simplified and easily accessible prior knowledge about particular emotions and related situations (cf. Rime, 1995; Rime, Philip-pot, & Cisamolo, 1990). The receiver highlights salient information in reference to his or her own schemata and organize it in ways that are compatible with his or her existing knowledge and expectations (Bartlett, 1932). In SSSE receivers retrieve from memory certain patterns from prior sensemaking, add details from the shared emotional event that fit their own knowledge, and pass it on to new receivers. During this recurring process of selecting, filtering, storing, and retrieving, pieces of emotional memories are added to the collective's reservoir of emotion knowledge (Shaver et al., 1987). And during this construction and reconstruction of the emotional event, diverse social schemata maintained by the various individuals involved and linked to the emotions in focus become aligned with one or a limited number of stereotypical conclusions. These are frequently expressed in a humorous story. For example, at Siemens AG there is a story about Werner von Siemens, the founder of the company and an immensely rich man, visiting the production plant one day. His contemporaries knew von Siemens as a brilliant engineer, but he was also feared as obsessed by accuracy and intolerant to ineffectiveness. He picked up a small piece of metal from the otherwise clean floor and showed it to an engineer. "Do you know what this is?" he asked. "A piece of metal," replied the intimidated engineer. "No," replied von Siemens, "this is my money" (Feldenkirchen, 1996). Werner von Siemens died in 1892, but the story is still told today. No one knows who started telling this story or how it has been transformed since it was first told, but somehow it became established as a humorous narrative in the Siemens culture, implying that waste should always be avoided. The event, probably in a form very different from its original one, remains as representing the collective's way of making sense of the relevant circumstances and emotions.
HUMOR IN SSE AT THE WORKPLACE
A large part of people's lives revolve around their work, the organizations they work for, and the people they work with. We can thus expect that a good deal of SSE is connected with events occurring in working life. Further, because colleagues at work know about the work processes involved and are implicitly interested in work events in the organization, they are also frequently the initiators and receivers of social sharing. As such SSE proves to be a common phenomenon in organizations (Meisiek, 2002; Rime, personal communication, 2 March 2003).
Earlier field studies on humor at the workplace confirmed that humor is a frequent mode for communication at all functional levels of the organization (Collinson, 1988; Hatch, 1997; Hatch & Ehrlich, 1993; Linstead, 1985; Roy, 1960), which leads us to believe that it is presumably also an important ingredient in the process of SSE at the workplace and that it contributes to general sensemaking in organizations. This ties up with another observation, namely, that humor creates in those involveda sense of relief and of being in control.
The workplace has often been described as a place where the expression of strong emotion is widely regarded as inappropriate; unless it is officially sanctioned as part of the work role (Hochschild, 1983). Organizations do not expect their workforce to give way to emotion, either positive or negative, for fear that they might do something to harm the organization's policies or its culture (Ashforth & Humphrey, 1995). Consequently, organizations develop and maintain display rules as part of their culture, indicating the appropriate range, intensity, duration, and even the target of emotional expressions in certain given situations (Ekman & Friesen, 1969; Hochschild, 1983). It has been suggested, too, that this could lead to the repression of emotional displays that are regarded as inappropriate (Hochschild, 1983). However, as SSE theory tells us, emotion does demand articulation for sensemaking purposes (Rime et al., 1998). Humor—which is generally regarded as an appropriate and beneficial ingredient in organizations (Collinson, 2002)—can thus be regarded as a useful mode for invoking emotional events in the complex social setting of the workplace. SSE also allows the person who shares the emotional event a chance to back away from the story, for example by adding, "I was just kidding. ..." Moreover, humor offers a way of describing an emotional event that doesn't demand any direct empathy on the part of a receiver, which is particu
8. NONSENS E MAKES SENSE 153
larly important when confiding in receivers that are not intimates. And apart from all this, emotional events that have been transformed into humorous stories are pleasant for people to hear and are more likely to be well received.
As regards the basic SSE process, an important observation is that people tend to share emotional experiences with close others, such as parents, family members, spouses, or close friends (Rime et al., 1998). In an organizational setting, however, intimates may not be very numerous, such as their coworkers, which means that employees have to resort to sharing emotional experiences with people with whom they are unlikely to have such a close relationship. Sharing may then occur because of some common experience or mutual understanding of the background to the work events, or because employees are connected through some aspect of organizational life like their department or their functional area.A work setting thus provides both opportunities and constraints when it comes to the sharing of emotional experiences—which ones to share, how to share them, and where. The actual emotional events shared at work may of course be private as well as work related, but the latter can be expected to have more impact on subsequent organizational processes.
We propose next that emotional events can be transformed into practical humor and then into canned humor by way of PSSE and SSSE. Eventually they may crystallize into narratives that reflect the culture, history, and characteristics of the organization in question and of its employees. This proposition is addressed in this chapter by exploring the critical factors of the relationship between initiator and receiver, the emotions associated with an event, the organizational context, and the individual characteristics that encourage initiators to adopt a humorous approach in presenting a story to their receivers. Drawing on research into SSE and humor in organizations, Fig. 8.2 demonstrates the PSSE and SSSE process.
The process begins with the need felt by individuals to share an event that carries emotional significance for themselves. In the course of PSSE the initiator continues to process the emotional event cognitively and in doing so generates meaning. When an initiator launches the sharing with a view to presenting the emotional event in a humorous way, it may become possible to generate practical humor as the sharing evolves. In the course of this sharing the initiator and the receiver both have an opportunity to make sense of the emotional event. The initiator also gets a chance to go through the process over and over again and to revisit the meanings he or she has constructed and perhaps reconstructed over time. The receiver, on the other hand, may proceed to share the emotional event with other individuals within the organizations. The receiver retains the memory not only of what had happened to the initiator, but also the humor created during the PSSE. If the original humor is successfully conveyed, it is very likely to be retained in the organizational system, as repeated SSSE occurs and evolves as a humorous aspect of the narrative. However, although the emotional event may eventually become an example of
8. NONSENSE MAKES SENSE 155
canned humor, the original story is likely to change in the process of the secondary sharing. As in the primary sharing process, the initiator and receiver both make sense of the emotional event in the course of sharing it, while also extending their understanding of the organization, the individuals concerned, and the particular situation involved. These memories are retained, and they help people who are exposed to them to make sense of their environment in the present and the future. It is also probable that several receivers may be able to share the same emotional events and their accompanying humor on several occasions. This labyrinth of mutual exchanges reinforces, and possibly alters or corrects, the way people remember and make sense of the emotional event, which in turn facilitates collective sensemaking in the organization. It is also likely that people are sharing is not a single emotional event but a whole emotional episode consisting of a series of emotional events around a common theme (Weiss & Cropanzano, 1996). Collective sensemaking can thus also be based on a chain of related events.
We next look at PSSE and SSSE separately, in order to discuss the specific circumstances under which humor relating to a shared emotional event may be produced. In other words we look at the antecedents of practical humor creation in PSSE and of canned humor production in SSSE.
HUMOR IN PSSE
Relationship Between Initiator and Receiver
Although most of us find at least some coworkers with whom we feel reasonably comfortable to "be ourselves" and whom we trust with our emotional experiences, we are rather more alert to the needs of self-presentation at the workplace than we are elsewhere (Bolina, 1999). Our motives for impression management may vary, depending on our goals and values. For example, if people want to rise quickly in the organizational hierarchy, they will seek to present themselves as competent and conscientious workers (Wayne & Ferris, 1990). They will then engage in various behaviors that convey such an image. Although it is usually assumed that humor helps to build up a desirable impression, because it is commonly interpreted as a favorable quality, it can backfire if used inappropriately. Thus an initiator has to consider the listener: Who is this person with whom they are sharing an experience, and does the relationship really allow for the expression of humor? Two attributes of the relation may be significant here: the distance/closeness between initiator and receiver, and their relative status in the organization.
Distance/closeness is regarded as a fundamental dimension of interpersonal behavior (e.g., Brown & Levinson, 1987; Spencer-Oatey, 1997). It is related to friendliness and affect or the degree of association. Hays's (1984) conceptualization of close and casual friendships can be used to tap this dimension (Spencer-Oatey, 1997). There are four behavioral contents that distinguish between close and distant relations: companionship, consideration, self-disclosure, and affection. Self-disclosure is related most closely to SSE. It pertains to disclosing or discussing information about oneself, to the exchange of ideas, facts, opinions, or confidences on any topic. This aspect of a close relationship corresponds to the SSE findings that emotional experiences are more frequently disclosed to intimate others rather than to strangers or to people of less importance to the focal individual at that particular stage in his or her life—an observation also suggesting that there is more freedom for self-disclosure in close relationships. However, because close relationships at work are rare, and because humor signals informality and reduces social distance (Graham, 1995; Sykes, 1966), closeness may be enhanced in primary sharing at the workplace by the introduction of humor into the exchange.
Proposition 1: Due to their relative distance between coworkers, people
are more likely to produce humor when they share an emotional event at
the workplace than when sharing events with their personal intimates.
The relative status of the initiator and the receiver is another important aspect. Studies have shown that low-status individuals tend not to joke very much with those of higher status (e.g., Vinton, 1989). Hence, a big difference in status hampers the production of humor, and thus perhaps of SSE as well. In a study of humor in workgroups Duncan and Feisal (1989) reported that "solid citizens," that is to say, members of a group who possessed no formal authority and who were socially close to the group's rank-and-file members, were well received by the others in the group when it came to joking behavior. They could joke about the others and were among the favorite targets for other people's jokes, which suggests that they were liked and enjoyed special joking privileges. Lundberg (1969) found that people of similar status enjoyed themselves more when spending time exclusively with one another.
Proposition 2: People are more likely to produce humor when sharing an
emotional event with someone of equal or lower organizational status
than when sharing with someone of higher status.
Whether or not humor can be made out of an event, and the type of humor thus created may be affected by the intensity, the valence, and the type of the emotions associated with the event in question.
As noted earlier, SSE increases with the intensity of the emotional event. This observation has been corroborated by the results of studies of emotional
8. NONSENS E MAKES SENSE 157
secrecy in which Finkenauer and Rime (1998) found that intensity and valence were not necessarily precursors of emotional secrecy. To the present authors' knowledge, no research has been done on the relationship between the intensity or valence of an emotion and the content, format, or quality of its sharing. However, it is reasonable to assume that the more intensive an emotion is, the more of its original perspective and framework will be retained in its social sharing compared with those of less intensive emotions. At the workplace, however, the expression of intense emotions may be hampered by the organizational display rules mentioned earlier (Hochschild, 1983). Thus, intense emotional events may be quickly converted into humorous narratives, to adapt them to the organizational context. This relationship between the intensity of emotion and the production of humor is also addressed in arousal theory (see Lefcourt & Martin, 1986). The basic idea underlying arousal theories is that, in the course of time, laughter can reduce tension and the buildup of energy (Freud, 1928). Thus a very intensive emotional experience at the workplace may be shared sooner rather than later and with lots of humorous comments. Further, the really intense emotions that may normally require a good deal of time and several iterations before they can be viewed in a different light (Rime et al., 1998) will reappear sooner as an event in the production of practical humor in PSSE at the workplace. The valence of the emotions concerned may also influence the timing of the humor production—in that positive emotions delay the onset of this production for longer than negative emotions do (Linstead, 1985).
Proposition 3a: People are more likely to produce humor in PSSE at the
workplace if they are sharing events of higher emotional intensity than
when they are sharing events of less emotional intensity.
Proposition 3b: People are more likely to produce humor in PSSE at the
workplace when they are sharing events of negative valence when they
are sharing events of positive value.
Finkenauer and Rime (1998) also found that emotional secrecy is correlated positively with events eliciting negative social emotions such as shame and guilt. An immediate implication of this in relation to the production of humor is that such emotions are probably less likely to become a subject of humor to begin with. And even if they were, the humor may be unintended, due to an underlying desire to avoid revealing socially inappropriate behaviors or thoughts.
Proposition 4: People are less likely to produce humor in PSSE at the
workplace in relation to events that elicit negative social emotions such
as shame and guilt.
Context also affects the behavior of organizational members. Here we consider three types of contextual factors: uncertainty and ambiguity, organizational culture, and the physical environment.
Any uncertain, paradoxical, and/or ambiguous elements in an organization also have an impact on humor in PSSE at the workplace. Uncertainty often develops in the course of radical changes such as mergers and acquisitions, and it arouses strong emotions (Huy, 1999). The prevalence of emotional events in periods of organizational change encourages the primary sharing, thus also promoting the production of humor as a coping and sensemaking strategy. Hatch and Ehrlich (1993) also contended that when an organization is infused with paradoxes and ambiguity is particularly marked, humorous comments will flourish. Hatch (1997) identified irony in particular as a sign that employees are confronting contradictions. Observations like these suggest that an uncertain and ambiguous organizational environment provides a fertile ground for the production of humor in PSSE.
Proposition 5: People are more likely to produce humor in PSSE at the
workplace if the organizational environment is uncertain, paradoxical,
Culture and language serve as a sense-giving backdrop for humor at the workplace. A joke, even one made with the best intentions, may not always be appreciated, due to differences in cultural values regarding the target or the style of the humor. In much the same way, ethnicity or language may have a significant impact on humorous behavior at work. Duncan (1982) asked, "Have you ever sat through a comedy in a foreign country and discovered that 'everybody is laughing but me'? Or, even worse, 'no one is laughing but me'?" In our model we consider organizational cultures favorable to humor. A strong humor culture encourages humor at the workplace. Some entrepreneurs such as the founder of Southwest Airlines recognize and promote a pro-humor culture (Quick, 1992). In this context, it has been proposed that humor can help to maintain an organizational culture (Linsted, 1985), and that humor can be an artifact or manifestation of a culture (Dandridge, 1986). In a culture of humor, organizational members believe that humor offers a desirable approach to daily activities. People are more conscious of humor and will act on this belief by creating humor in PSSE as they interact with others in the organization. In an organizational culture of humor, however, joking behavior may also be used as a cover for strong emotions. Buroway (1979) discovered that racial prejudice between Black and Caucasian workers was articulated in jokes on the shop floor. Because the production process demanded a degree of cooperation between workers, overt racial hostility had to be kept to a minimum and was therefore diluted by humor.
8. NONSENS E MAKES SENSE 159
Proposition 6: People are more likely to produce humor in PSSE when
the culture of their organization encourages humor at the workplace.
To our knowledge, researchers have not yet examined the relationship between the physical environment and humor at the workplace. But the physical environment may in fact contribute to the production of humor in PSSE, thus also enhancing a culture of humor. Davis (1984) identified three key elements in the physical environment: physical structure, physical stimuli, and symbolic artifacts. All three of these can affect people's perception of their environment and the way they interact with each other, which in turn will either encourage or discourage the production of humor.
People are more likely to interact with others who are physically close. Thus, workplaces whose physical structure leaves employees isolated or with little formal or informal contact with one another will restricts SSE and the chances it offers for producing humor. If on the other hand people are given more opportunities for meeting, they will have more time to develop relationships providing opportunities for humorous interaction (Davis, 1984). Within the physical structure, behavior is also affected by physical stimuli: the color scheme in the office (visual), the sound of the phone bell (audio), or the smell of the coffee brewer (olfactory). Most of the literature on physical stimuli focuses on the way the stimuli are manipulated in order to reduce distractions and to induce desired behaviors (Davis, 1984). Finally, symbolic artifacts are aspects of the physical environment that can direct interpretations of the social setting in a certain way. Office signs, carpeting, or type and style of furnishing, for example, can convey information about the organization and the people who work there (Davis, 1984). A workplace that is designed to be informal and enjoyable sends an implicit message to its occupants that a playful attitude is encouraged there, so that people feel more comfortable about making humorous comments, telling funny stories and laughing together with their colleagues. We believe that certain physical stimuli and symbolic artifacts may facilitate the production of humor.
Proposition 7: People are more likely to produce humor in PSSE if the physical environment at their workplace is favorable to interaction among the employees, providing pleasant stimuli and symbolic artifacts that encourage humor.
Although the initiator may be inspired or stimulated by the receiver or by some factors of the situation, to imbue the narrative with a humorous quality, certain innate personal qualities help to also determine the possible importance of external conditions (Ross & Nisbett, 1991). In our model we consider two individual characteristics: sense of humor and gender.
A sense of humor is a distinct characteristic attributable to the individual. It makes a person more prone to convert emotional events into humorous stories, as well as making it easier for them to do so. A sense of humor can be regarded as a personality trait. The assumption is that there are stable differences between individual people as regards the extent to which they perceive, enjoy, and create humor (Lefcourt & Martin, 1986). However, a sense of humor is a multidimensional construct to which various meanings have been attached (Eysenck, 1972; Moody, 1978). Eysenck (1972), for example, suggested three subdivisions for the sense of humor concept, namely, the conformist, the quantitative, and the productive types. The first two refer to the extent to which an individual agrees with most other people about what is humorous, and to the frequency and immediacy of their laughter and smiles. Only the third type refers to the probability of a particular person perceiving a situation as humorous and amusing, and thus deriving humor from it (cf. Lefcourt & Martin, 1986, p. 22). Most sense of humor studies have started from individual differences in the appreciation of humor, such as a preference for particular types of humor (e.g., Eysenck, 1942, 1943; O'Connell, 1960; Redlich, Levine, & Sohler, 1951). We concentrate on the third meaning to emphasize the productive aspect of a sense of humor. In the course of PSSE, a strong sense of humor will predispose the initiator to convert an emotional event into a humorous story together with the receiver.
Proposition 8: An individual with a strong sense of humor is more likely
than other people to produce humor when they share emotional events
with others at work.
Studies have found gender differences in the appreciation and production of different types of humor (Collinson, 2002; Duncan et al., 1990; Lefcourt, 2001). Women in Western society seem to display a stronger tendency to appreciate and produce self-deprecating humor, whereas men prefer wit and jokes characterized by competition and aggression and often directed at others (Crawford & Gressley, 1991; Levine, 1976; Zillman & Stocking, 1976). In the workplace it can be expected that women may produce self-deprecating humor as a defense mechanism against stressful and disturbing events (Lefcourt & Martin, 1986). Self-deprecating humor allows people to laugh at themselves and to take things less seriously (Freud, 1928). Males at the workplace, on the other hand, tend to produce aggressive and competitive humor in connection with workplace events in which some coworker has behaved or responded in a foolish way (Collinson, 1988).
Proposition 9a: Females tend to produce self-deprecating humor in PSSE when sharing stressful work events.
8. NONSENS E MAKES SENSE 161
Proposition 9b: Males tend to produce aggressive and competitive hu
mor in PSSE when sharing work events that involve a colleague's be
HUMOR IN SSSE
SSSE is similar to PSSE in its processes and many of its elements. Much of what has already been discussed is thus also valid for SSSE (see Fig. 8.2). However, because the initiator of SSSE has not experienced the shared emotional event personally, we can expect some differences in the nature of this kind of humor sharing. SSSE illustrates the way humorous descriptions of emotional events are developed to become well-established organizational narratives.
Sense of humor is again important to the initiator's reproduction of the humorous account (Lefcourt & Martin, 1986). However, because the initiator in SSSE has been a receiver in an earlier round of primary or secondary social sharing, the appreciation of humor is more likely to be relevant here than in its production. Only when a receiver has proved able to appreciate the humor shared earlier, as a receiver, will the person be able to share the humor that accompanies the description with a third party.
Proposition 10: People with a greater sense of humor are more likely to
appreciate and to reproduce humor in SSSE at the workplace.
Similarly,gender differences in the appreciation of humor directly affect the retention of the humor first produced in the PSSE. A female receiver may not appreciate competitive and aggressive humor, so it is not very likely that she will then share it with someone else in a humorous form, which means that the practical humor generated in PSSE will not be preserved. On this part, a male receiver may not appreciate self-deprecating humor, which will thus be lost when the male receiver goes on to share the emotional event with someone else. The relation between the gender of the initiator, the type of humor, and the gender of the receiver may affect the choice of humorous mode: A male employee may avoid sharing aggressive humor with a female coworker, and a female employee may avoid sharing self-deprecating humor with her male colleagues (Zillman & Stocking, 1976).
Proposition 1 la: Women are more likely then men to appreciate and sub
sequently reproduce self-deprecating humor in SSSE at the workplace.
Proposition 11b: Men are more likely than women to appreciate and sub
sequently reproduce aggressive and other-directed humor in SSSE at
It can be posited that the relative impact of relationship and context may be different in SSSE. Relationships may be less important because the initiator is not the one who actually experienced the original emotional event. Over time the weight shifts toward contextual cues such as uncertainty and ambiguity, organizational culture, and the physical environment that are more likely to trigger the humorous elements in SSSE at the workplace. People may also engage in SSSE because they enjoy the humor and simply want to reproduce it.
Proposition 12: The influence of relational factors on the production of
humor in SSSE is less than it is in PSSE, while the influence of contex
tual factors becomes greater.
In sum, the production of humor in PSSE may be a direct outcome of experienced emotional events, while the reproduction of humor in SSSE may be a direct outcome of an earlier sharing of experiences and of perceived features in the emotional events. The two stages of SSE occur in a work setting as follows. Employees experience emotional events at the workplace and share these with their coworkers, because the emotions associated with the events demand articulation for sensemaking purposes. As the employees proceed to share the experience, there are various factors—their work-based relationship with the receiver, the emotions associated with the experience, the organizational context, or their individual characteristics—that incline them to give a humorous spin to the narrative. Those receivers who appreciate the humor may then share the story socially with someone else in the organization. The receivers may also change the original story a little or give it a different humorous spin in order to adapt it to another conversational context. In this way some stories will withstand the trials of time and become well-known features of the organizational culture. They convey important aspects of organizational life in a lighthearted way that people can easily remember.
The purpose of this chapter has been to suggest a model of humor in the social sharing of emotion at the workplace. Building on prior humor research and the theory of SSE, and taking into account the specific social situation of the workplace, we have discussed the part played by humor in SSE in organizations on the one hand and, on the other, the part played by SSE in the development of humorous narratives in the organization. Our model unfolds in three steps: SSE is characterized as a theory for explaining why people share their emotional experiences in PSSE and SSSE at the workplace; PSSE is characterized as relying to a certain extent on the production of humor in organizations; sto
8. NONSENS E MAKES SENSE 163
ries from PSSE are transformed into organizational narratives through an iterative process of SSSE.
In organizational behavior research, humor has been widely regarded as an antecedent of emotion. It has been suggested that humor arouses emotions: joy, for example, in case the humor has been appreciated, or anger if it has been resented. Humor, thus, has been conceptualized as beneficial for the well-being and job satisfaction of the employees. This perspective seems to ignore the "ongoing" aspect of the presence of emotion in organizations. Consequently, our model aims to embed humor in the ebb and flow of emotional experiences at work, which means humor may be an outcome as well as an antecedent of emotions in organizations. Given the individual's need to articulate emotional events for sensemaking purposes and the collective's need for narratives as a "sensemaking currency" (Boje, 1991), as well as the specific social conditions at the workplace that imply relatively distant personal relationships, humor emerges as a powerful ingredient in communication at the workplace.
Sharing can occur via a telephone call, an e-mail, or a written note, but we assume that its main medium is face-to-face communication (Rime et al., 1998). Further, although we have not directly addressed the many forms of sharing that occur—comic strips on the wall of a cubicle, ironic comments posted on a bulletin board, or artwork displayed in an office, for instance—all these and many more can be regarded as possible means of conveying affective reactions to certain events. Such things also serve to help individual people to comprehend the culture of their own organization, to learn about its norms and other people's attitude towards it. Situational humor as expressed in "horseplay" has not been considered in this chapter, as it has been identified primarily as a way of counteracting monotony or boredom and is irrelevant to our discussion on the impact of humor in organizations.
An entire emotional event can be presented as a joke, but under certain conditions it is only possible to take up certain aspects of a story, which cannot then be regarded as a complete story with a beginning and an end or as located in its own particular setting. The emotional event on which the account is based may have happened long ago, or just before it was told the first time. Or an experience may be hinted at, discussed, reshaped as a joke, and so on, without any attempt to present it as a complete story. It is not altogether clear how far such behavior can be regarded as an instance of SSE, but in the light of the theory just presented, it seems reasonable to regard it as covered by the model.
The intricacies of producing and retelling instances of practical humor make it unlikely that humor produced in the course of SSE will survive for long. Few such stories are likely to enter the corpus of organizational narratives, reflecting common grounds for the sensemaking of an event and its emotional outcomes. But when they do, they may be used to socialize and educate new employees, or as a way of strengthening bonds among the veteran employees through the frequent reiteration of emotional events. Humorous accounts of this kind can also venture beyond the borders of the organization, allowing outsiders a glimpse of various aspects of it. Humor can increase the cohesive nature of a group, or offer individual people a way of being able to talk about emotional events at work. At the same time, of course, humor serves to entertain, to strike a responsive chord in others, or to attract attention.
It seems to us that the social sharing of emotion, on which we base our ideas about the production of humor at the workplace, has not yet received much scholarly attention in organizations. We could perhaps say that it has been addressed indirectly in theories on informal communication. Certain findings from social psychology, however, encourage its use in seeking to explain organizational phenomena.
The model described in this chapter is based largely on research findings and theories developed in Western societies. Although SSE has been tested in both individualistic and collectivist cultures, theories of humor and related speculations have their roots in the individualistic world. Readers of this chapter should be cautioned about applying the same model in a collectivist context. Although we believe that the basic elements of humor—production, retention, and transformation—are likely to hold across cultures, the relation between the antecedents and production of humor may be affected by the presence of other social norms, values, or concepts of self. Social norms may dictate the amount of humor that can appropriately be shared, for example, or the kind of humor most suitable for sharing. We thus suggest that parallel studies should be conducted in both the individualistic and the collectivist type of culture, an undertaking that would contribute at the same time to cross-cultural humor research.
Although we have presented separate sets of antecedents for the production of humor in SSE, we do not deny that there may be interactive effects between them. Hence, certain factors in one category may interact with some factors in another, in this way affecting the production and reproduction of humor. At this stage of theory development we have concentrated on identifying certain factors and speculating about their possible general relationship with the phenomenon under study. Further refinement of the theory would necessarily have to build on empirical findings. A research program based on the present model could start by surveying and examining the correlation between all the antecedent factors proposed here and the creation of humor. Specific factors could then be focused and their relationship with the creation of humor tested. For example, the relation between the gender of the initiator, sense of humor, type of emotion, and the type of humor most likely to be created could be investigated by using a diary. It would be interesting to examine the survival of humorous stories produced in relation to an emotional event in relation to various characteristics of the initiator and receiver.
8. NONSENS E MAKES SENSE 165
In brief, the intention of this chapter has been to introduce a general framework for addressing humor in SSE at the workplace and to propose some directions for research. Despite the limitations of the present approach, we suggest that humor and emotion research in organizational behavior seem to be more closely related than has been assumed in previous research. Further exploration of this relationship promises to extend the sensemaking paradigm and to contribute to social psychological research on emotions in organizations.
This page intentionally left blank
Charmine E. J. Hartel
This chapter attempts to bridge the gaps of strategic business negotiation, communication, and emotion in a cross-cultural context. In particular, we argue that miscommunications are "boundary-crossing mishaps," which are affected by the negotiator's understanding of the respective cultures (and cultural backgrounds) of the parties, negotiation skill, cultural differences, emotional awareness and regulation, negative affect, and discrepancy in convergence–divergence between the interactants. This chapter concludes that when too many hassles or mishaps occur, negotiation breaks down and the need to understand cross-cultural communication through alternative theoretical models arises.
We begin this chapter with a short case illustrating the effect of emotions and behaviors in cross-cultural negotiations.
In 2001, an Australian fertilizing company (AusFert1) employed a Chinese national in Australia to explore the possibility of exporting to the Chinese market. The company was already successfully exporting to American, Middle Eastern, and several European markets. Realizing that these markets are mature, the need for the company to grow into a new, possibly infant, market was recognized. After much consideration, the company decided on China.
Historically, the Chinese government had never allowed fertilizer importation and was just beginning to consider the possibilities of developing this avenue. AusFert's importation process began with obtaining an import permit in
Not the company's real name.
China. This was a complex process as it required the establishment of a Chinese importing company (ChinFert2), which imported fertilizer from Australia. AusFert, however, was concerned that ChinFert might import other brands once the license was obtained, so it agreed to enter into a joint venture with an importing company in China to eliminate this possibility. AusFert drafted a contract with its Chinese alliance and proceeded to set a date for fertilizer to be purchased, as well as a quantity estimation for the next 3 years.
It wasn't long before problems began to surface. First, the shipment did not arrive on the due date, and AusFert began to realize that the possibility of receiving an order was remote. Aware of cultural sensitivity, AusFert approached an outside contact for advice. As soon as AusFert did this, however, communication with ChinFert ceased as ChinFert was insulted that the Australians were seeking outside advice. As a result, the Chinese began ignoring the contract, did not initiate any progress to obtain a permit, and did not engage in negotiation with AusFert. In addition, the Chinese negotiator was uncontactable both in Australia and in China.
Both parties believed that the other was to blame for the miscommunication, although it is evident that it was simply a matter of cultural differences and lack of understanding. According to AusFert, there was no explanation as to why the Chinese negotiator refused to communicate with them. After failing to establish any communication with the Chinese negotiator through "normal" channels, AusFert decided that communication would be established with the negotiator's parents.
Another problem that arose in the negotiation process was that the Chinese national, who lacked English-language skills, used an interpreter for all negotiations. However, the interpreter for her was a Chinese from Hong Kong and was excluded by AusFert, as it was assumed that "he was up to no good with her from day one." At the time of the communication breakdown, the interpreter was contacted directly by ChinFert and, like his client, refused to communicate. AusFert could not understand that the signing of the contract by AusFert's contact did not bind the contact to the delivery of orders on a certain date. The only option to complete the task was to discover the real reason for the communication breakdown.
By the time communication was established with the Chinese negotiator, the negative emotions were so strong that the Chinese national believed AusFert had no understanding of how the Chinese system worked and had completely discredited her effort and ability. She did not see fit to engage in any immediate communication. Her feelings also led her to believe that there was a conspiracy to undermine her role when ChinFert communicated with certain officials in Beijing without her knowledge. The effects of such negative
2Not the company's real name.
9. UNDERSTANDIN G CROSS-CULTURAL NEGOTIATION 169
emotion resulted in both parties viewing the other as divisive, untrustworthy, uncooperative, and dishonest. ChinFert felt that, on this basis, no relationship existed and therefore there was no further need for communication. AusFert, however, felt that despite everything, there was a binding contract that needed to be honored.
The preceding actual case demonstrates how cultural differences can impact on felt emotions, the ability to detect others' emotions in a timely manner to avoid breakdown in negotiation and relationships, and the understanding of what steps need to be taken to repair the bad feelings the other party has. This chapter contributes to the understanding of the negotiation process by developing a cross-cultural negotiation model based on an integration of affective events theory (AET) and communication accommodation theory (CAT).
Within today's global marketplace, it is inevitable that business interactions will continue to cross cultural boundaries. These boundaries are no longer purely physical but also include culture, religion, regulation, values and beliefs, social groups and circles, and unfamiliar emotions and communication methods (e.g., Ayoko, Hartel, Fisher, & Fujimoto, 2003; Phatak & Habib, 1996). Boundary crossing, which is the process of either uniting or separating people, is mediated through negotiation. This chapter aims to bridge the boundaries of strategic business negotiation, communication, and emotion in a cross-cultural context. Each of these three boundaries is discussed briefly, followed by a review of cross-cultural communication literature.
People use boundaries to achieve integration and separation from others, as a way of maintaining comfort zones (Petronio, Ellemers, Giles, & Gallois, 1998). They exist at the level of cultures as well as at the individual level. As such, during negotiation, people have a boundary they place between themselves and the other party and they evaluate the extent to which they need to tighten or loosen it. This boundary-crossing adjustment may be initiated by one or both parties to build the relationship. When boundaries are unattended to, miscommunication may occur (Coupland, Wiemann, & Giles, 1991, as cited in Petronio et al., 1998).
Miscommunication may occur when boundaries are too tight or when they are too loose. When they are too tight, the parties involved are not receptive to each other's definition of an affective event (Petronio et al., 1998). When boundaries are too loose, communication may cease altogether as the boundary is not salient and the parties are unsure of the appropriate behavior (Dayringer, 1998). Miscommunication will generally cause the tightening up
of boundaries in all cultures when it is detected (Willemyns, Gallois, Callan, & Pittam, 1997).
The boundary-crossing process is achieved through negotiation. Negotiation is a frequent part of everyday life; thus, it is largely an unconscious process (Ferraro, 2002). Chang (2002) identified five negotiation style preferences: accommodation, collaboration, and withdrawal (cooperative), and competition and consultation (instrumental). Further, the dual concern model introduced by Rubin, Pruitt, and Kim (1994) portrayed negotiation style as a combination of high/low concern for the self and high/low concern for others.
The addition of differing cultures in the negotiation process adds further complexities. This is because the style of our negotiations is conditioned culturally, both consciously and unconsciously (Dayringer, 1998). Culture impacts on the way in which the parties think, feel, and behave (Casse & Deol, 1985). In terms of cross-cultural negotiation, Adler, Brahm, and Graham (1992) showed that the cooperative approach is more successful for both Westerners and Easterners when negotiating with each other. We assume this approach in our further discussion.
As shown, the process of boundary crossing in cross-cultural negotiations is dependent on the culture of the parties involved. In particular, cultural differences impact negotiation, with some cultures preferring to tighten the walls at the beginning and work toward loosening up when trust is built between parties, and other cultures preferring to start with very open boundaries and tighten these boundaries only when they feel threatened. Our focus, the Sino-Australian negotiation setting, involves these two opposing approaches to boundary crossing.
From a business perspective, the issue of boundary crossings involved in Sino-Australian negotiations is an important one. This is because the low success rate of foreign investments in China has been attributed, in part, to miscommunications (cf. Pye, 1982).
Both Hofstede (1980) and Trompenaars (1993) studied cultural differences between Westerners and the Chinese in their attempt to frame the behavioral patterns of the Chinese to ensure successful communication with their Western counterparts. Their research showed that certain patterns of behavior are evident within cultures. These patterns have been widely used within cross-cultural research and practice to train negotiators. Yet we argue that more is needed.
The Chinese culture is notably among the least understood and studied markets within the business arena (Zhang & Neelankavil, 1997). Therefore, this chapter attempts to bridge this gap through developing a theoretical framework for analyzing cross-cultural negotiation processes within Sino-Australian negotiations. Further, the framework developed represents a bounded-emotionality perspective of the negotiation process. This link between negotiation and emotion is discussed next.
9. UNDERSTANDIN G CROSS-CULTURA L NEGOTIATIO N 171
EMOTION AND NEGOTIATION
Notwithstanding the fact that skill is an undeniable critical element in successful negotiations, negotiation is also affected, to a large extent, by emotion. Although not widely researched, emotional or affective reactions such as distress or anger are often experienced during the negotiation process (George, Jones, & Gonzalez, 1998), partly because of external and internal conflicts arising from the process of attitudinal structuring (George et al., 1998). When miscommunication occurs, it is likely to have a negative effect on negotiators, rather than a positive effect. This negative effect is likely to trigger the negotiators' subconscious, which may result in the negotiator displaying a defense mode or initiating the negotiation toward a negative emotion (Dayringer, 1998). Subsequently, the relationship between emotion and negotiation involves the impact of feelings or moods that people experience during cross-cultural negotiations (George et al., 1998).
Conflicts that take place during negotiations have an inherent affective component (Pondy, 1967), and distress or anger emotions are often experienced during the negotiation process (George et al., 1998). Further, these negative emotional reactions may continue to develop at an accelerated and accumulative pace—a pattern also known as a "negative spiral" (George et al., 1998). Fueled with negative emotions, negotiators may experience a negative spiral at different speeds, whereby there is little chance for that spiral to be broken by either party. Therefore, the negotiation result can be unsuccessful and can even terminate the possibility of future negotiations.
As with negotiation styles, culture impacts on the emotions experienced during the negotiation process. Studies have shown cultural differences in a range of emotional experiences including intensity, duration, and control of emotion (Matsumoto, 1993). Studies also conclude that cultures differ when judging other emotions (Matsumoto, 1993). As such, incorrect emotional judgments due to cultural differences will affect the outcome of a negotiation.
Behavior during the negotiation process is also affected by the emotions evoked. George and colleagues (1998) identified that an individual's behavior is as affected by the person's moods and emotions as by his or her attitudes and values. They suggested that there are two major emotional feelings during negotiation: positive and negative. However, we suggest that there may also be degrees of emotional uncertainty, with some factors during the negotiation process increasing the propensity for one's emotion to move in a positive direction, and others increasing the propensity to move in a negative direction. In other words, although the process of negotiation may begin positively, an "uncertain feeling" can emerge when either or both parties are not sure if the negotiation is heading in the agreed direction.
Toward the conclusion of the negotiation, negative feelings may emerge if the result is not satisfactory, and likewise, positive feelings may emerge if the result is satisfactory. Again, we argue for a third feeling, which appears between positive and negative, namely, the uncertain feeling.
The preceding discussion of cross-cultural negotiation shows that there are many points in a negotiation where an affective event can occur. According to affective events theory (AET), an emotional response flows from a particular type of event labeled as an affective event. These affective events are shaped by the environment in which interactions occur. The emotions that flow from affective events are in part a consequence of the event and in part a consequence of individual factors that shape the interpretation of the event. These emotions, in turn, impact on individuals' attitudes and behaviors.
INFLUENCE OF AUSTRALIAN AND CHINESE
CULTURE ON NEGOTIATION
Cultural factors impact on whether negotiations begin with positive, negative, or uncertain feelings. For example, observations of Chinese-Australian negotiations suggest that Chinese negotiators have the tendency to start negotiating in a state akin to the uncertain feeling whereas Australians tend to start with a positive feeling (cf. Breth & Jin, 1991; Tung, 1991). Further, this uncertain feeling may also occur after negotiations have begun and neither party is certain of the direction of the negotiation, leading to a negotiation standstill. When negotiating with the Chinese, this type of ending may terminate the negotiation process or close avenues for future negotiations (Engholm, 1989; Pye, 1990).
In cross-cultural negotiation, cultural values play an important role in reaching an effective agreement. For example, Australians and the Chinese differ in their definitions of value. Therefore, the Chinese would not use the phrase "too expensive" because it suggests that one has not enough money to purchase the item, which negates the achievement of saving face. However, the Australian negotiator would have difficulty in understanding why the Chinese do not inform them of the prices that are too expensive and thus "save face." Pye (1982) supported this contention, commenting that Sino-American business negotiation misunderstandings are due largely to cultural differences and that joint-venture negotiations often intensify this problem further. Cultural difference in values, in effect, fuel the possibility of miscommunication. Cultural understanding is, therefore, an important moderator of the types of emotions flowing from misunderstandings in a cross-cultural negotiation and is depicted in our model (see Fig. 9.1).
Hall's (1976) theory of high- and low-context cultures is also helpful in explaining cross-cultural communication behaviors. In high-context cultures, such as in China, nonverbal or situational cues bear considerable importance (Stone, 2002), whereas the emphasis on time and formal documentation bears
FIG. 9.1. Theoretical model of cross-cultural negotiation.
high importance in low-context cultures such as Australia. When negotiating between these two cultural groups, the group from the high-context culture requires a clearer interpretation of its negotiator's nonverbal signals. Failure to do so decreases the chance of a successful negotiation.
Culture also determines the negotiation strategy used by the parties involved. For instance, Chinese negotiation teams are often constructed with multiple layers of powerful people. This strategy is used to shift power from one person to the other, in the attempt to confuse other negotiators through role play. This shift may also be adjusted when a member of the team is entering a negative emotive state. Further, the Chinese society is a high power distance society (Hofstede, 1980), which means that subordinate persons converge more to the dominant persons rather than vice versa (Thakerar, Giles, & Cheshire, 1982). These features of Chinese negotiating norms contrast markedly with the features of Australian negotiations, making Sino-Australian negotiations undoubtedly difficult tasks (Shenkar & Simcha, 1987). It is no wonder that despite numerous attempts to develop successful negotiation relationships between the two cultures, studies in this area have not yet adequately addressed the dynamic character of the cultures involved. We next attempt to address this gap by first discussing culture and emotion, and then applying CAT to the Australian-Chinese negotiation process.
CULTURE AND EMOTION
Research has shown that cultures exert considerable influence over emotion through behavioral norms (Ekman, 1994; Friesen, 1972; Matsumoto, 1993). Awareness of this means that negotiators working in the cross-cultural context will not take the process of communication for granted but instead pay more attention to the processes employed and their potential emotional consequences.
There are several types of emotional effects that may influence the Sino-Australian negotiation process. The most pertinent one is a postdecision emotion that occurs when a decision is reached with strong emotions attached. This creates problems in negotiation, for instance, as our earlier scenario demonstrated; because the Chinese society is based on trust, as opposed to Western society's basis in legalities, the concept of being bound by a contract is less important in Chinese culture than abiding in accordance with one's emotions. In other words, the moral obligation was perceived by the Chinese as insignificant compared to one's emotion, especially when revenge was an added component.
Although great differences exist in expression and behavior, research suggests that the experience of basic moods and emotions tend to be universal (Ekman, 1994; Scherer, 1994, as cited in George et al., 1998). For example, crying, anger, and frustration are negative feelings that we have all experienced.
9. UNDERSTANDIN G CROSS-CULTURAL NEGOTIATIO N 175
Difficulties arise during cross-cultural negotiation regarding the conversion of negative feelings to positive feelings. First, these feelings may occur when there is a misunderstanding. Moreover, the negative feelings are not detected by the other party. Second, the process of converting negative feelings to positive feelings is difficult, although not impossible. They may be overcome if the negotiator adopts the concept of "no hurt feelings." This suggests that no one parry is at fault, and that business should continue as planned. Evidently, negative emotions can be controlled in this situation as the cause of the misunderstanding is not of paramount importance.
Further difficulties arise due to the terminology used in negotiations. For instance, there are often terms used by Australians that have no direct translation in the Chinese language, such as "to hold a grudge" and "no hard feelings." This indicates that the Chinese culture is lacking a process to convey verbally negative feelings. The only term used in China that is similar to this is qing biejianjue, which translates as "please do not see it as odd." When it is used, it suggests that the Chinese person caused the misunderstanding. This confirms further Hofstede's (1980) power distance theory, which states that in Eastern cultures, the less powerful party is more likely to submit to the higher power party.
COMMUNICATION ACCOMMODATION THEORY
IN SINO-AUSTRALIAN NEGOTIATIONS
In this section we extend communication accommodation theory (CAT) (Coupland et al., 1991; Giles, 1973) to detect positive and negative emotions in the communication process and in cross-cultural negotiations. This is achieved by drawing on George, Jones, and Gonzalez's (1998) model, which describes the affect of negotiators in cross-cultural negotiations. These authors identified negative emotional spirals as the major cause of unsuccessful negotiations. Further, unless the negotiators break out of the negative spiral, miscommunications or negotiation breakdowns are likely to result.
CAT is designed to examine the efficiency of communication (Petronio et al., 1998). In particular, it focuses on the convergence and divergence involved in speech communication. CAT is applied broadly in a number of areas, including convergence and divergence, cross-boundary communications, nonverbal communications, miscommunication, and integration of intergroup and outer-group communications (Petronio et al., 1998).
There are six levels of CAT, relevant to understanding problems in negotiation:
The application of the CAT framework to Australian-Chinese negotiations can facilitate further the link between culture, emotion, and negotiation. Extending it to include emotions allows us to describe how a given negotiation process resolves.
Level 1 of CAT
The first level of CAT is concerned with the meaning and discourse, or cultural and language differences, in the use of the English language. Although English is becoming the international negotiation language, it is not used in the same way by everyone, and this can create a barrier to communication (Crystal, 2000). In other words, although Chinese and Australians speak English, they may not necessarily use the language in the same way (Crystal, 2000). When this issue is not recognized, miscommunication may arise, resulting in negotiation problems that may not be discovered until greater damage has surfaced.
Level 2 of CAT
The second level of CAT is primarily concerned with the minor misunderstandings likely to accumulate in cross-cultural negotiations, especially with the Chinese. When negotiators come from different origins, the potential for misunderstandings is even greater. Culture provides a perceptual filter that influences the way we interpret events (Adler & Rodman, 2000). Parties of different cultures begin the negotiation process testing many communication methods until a mutually understood method is established. This process is a fertile ground for misunderstandings, albeit mostly minor ones, and the accumulation of these misunderstandings causes negative emotions (George, Jones, & Gonzalez, 1998). For example, Chinese negotiators do not want to cause their counterparts hurt feelings and do not have any contingency plans for resolving such problems when they do occur. Although Australians are capable of saying "no hard feelings" or "no worries, mate" and continuing with the negotiation process, Chinese negotiators will not be able to forget the
9. UNDERSTANDIN G CROSS-CULTURA L NEGOTIATIO N 177
"hurt feelings." The accumulation of these hurt feelings throughout the process can become a large issue at a later stage of the negotiation, which is evident in the scenario presented at the beginning of this chapter.
Level 3 of CAT
The third level of CAT shows that individual differences have enormous potential to cause miscommunications. These personal differences influence people's communication pattern. Unless communication parties consciously are making allowances for such differences, miscommunication may occur.
Level 4 of CAT
The fourth level of CAT addresses the cultural differences which pose a barrier for negotiators in achieving the full awareness of misunderstandings. When negotiation occurs between high- and low-context culture groups, the interpretation of the negotiators' goals may differ dramatically from their actual goals. This is because the low-context group presents its evidence before its conclusions whereas the high-context group presents its conclusions before its evidence (Hall, 1976). These two contrasting ways of presenting information may cause some negotiators to mistake unimportant evidence as their counterpart's goals, thus missing the counterpart's actual goals, and vice versa. Unless the negotiators are trained or aware of such critical cultural differences, the miscommunication may not be detected and thus successful negotiation may be impaired.
Level 5 of CAT
Once boundaries are created, groups will remain within their boundary to avoid uncertainty (Petronio et al., 1998). Within Level 5 of CAT, these boundaries are used to keep nongroup members out and other group members in. According to Coupland and colleagues (1991), this involves regulating, aging, blurring, and coping with boundaries that often are fuzzy.
As Petronio and colleagues (1998) observed, "We fit in our environment by drawing lines around those things that are important to us, and then control them through rules." Yet we also recognize that to fit successfully within the environment, we must have enough flexibility within these boundaries to allow a degree of integration between ourselves and the world within which we live. Such boundary negotiation evokes emotions in ourselves and others, and requires emotional awareness and regulation (see Fig. 9.1).
Level 6 of CAT
The Australian culture is less structured and lower on Hofstede's (1986) power distance scale in comparison to the Chinese. In negotiation processes between these two cultures, Chinese negotiators are more likely to have an established team structure (cf. Breth & Jin, 1991). Further, Level 6 of CAT illustrates that Chinese negotiators have a power distance structure that facilitates decision making from the person with the most authority. When negotiating with one or two Australians, it is often difficult for the Chinese negotiator to determine how the power structure should be maintained and how to keep the balance of power for both parties (cf. Breth & Jin, 1991).
In summary, the CAT model predicts that speakers will be evaluated positively or negatively: Converging speakers will be evaluated positively, whereas divergers and maintainers will be evaluated as hostile and unfriendly (Hornsey & Gallois, 1998). When the style of negotiation is converted from convergence to divergence, it is also converted from positive to negative. At this stage, we propose that a negative emotion is likely to play a stronger role than a positive one (see Fig. 9.1), and it is at this point that the negotiation is likely to be guided toward a negative outcome (George et al., 1998). The first indication of this problem is the presence of ignoring or reducing (Dayringer, 1998) behaviors not coded among the nonverbal behaviors reported in Jones and colleagues' (1999) research. Therefore, these need to be added to CAT when applying it to the cross-cultural negotiation context.
The application of the CAT framework to Australian-Chinese negotiations can facilitate further the link between culture, emotion, and negotiation.
EXTENDING CAT FOR THE CROSS-CULTURAL
INTEGRATING AFFECTIVE EVENTS THEORY (AET)
The integration of AET in cross-cultural negotiations is beneficial because CAT in its current conceptualization fails to consider the role of affective events or emotional responses and their subsequent attitudinal and behavioral consequences.
Conversion and Diversion
Do all negotiators begin as converging speakers and convert to diverging throughout the negotiation (i.e., when they are ignored by their counterpart) ? When negotiation occurs between people of the same culture, this changing process should occur at the same time due to a shared mental model of the negotiation process (cf. George et al., 1998). When the negotiation is between
9. UNDERSTANDIN G CROSS-CULTURA L NEGOTIATIO N 179
people of different cultural backgrounds (e.g., Australian and Chinese), the process may not occur at the same rate because of cultural differences and expectations (Eiteman, 1990). This could also precipitate misunderstanding as parties with less experience in cross-cultural communication will not be able to interpret this change (see Fig. 9.1). The hostility and the unfriendliness will be stimulated further by the change in the emotion and may possibly become a mixture of confusion and frustration, which again is vulnerable to misinterpretation or misunderstanding. This begins to form a circle of misunderstandings. Unless this circle is broken, the end result will be a total communication breakdown (see Fig. 9.1). Figure 9.1 demonstrates the many factors that contribute to misunderstanding, such as culture, emotions, and the negotiation context. Unless this circle of misunderstanding is broken, the negotiation process is unlikely to move on to the next stage of boundary crossing.
In applying AET, a negative misunderstanding result is likely to identify the boundary crossing as a hassle while a positive result is likely to identify the boundary crossing as uplift. At this stage of the negotiation, emotion levels are likely to change, and the flow of convergence or divergence from this point will, indeed, mark the end result of any negotiation process (see Fig. 9.1).
The likelihood of negative emotions arising in negotiations may be due simply to misunderstandings and conflicts. According to Trompenaars (1995), emotions and cultures are combined. Some cultures are affective (i.e., individualists), whereas other cultures are neutral (i.e., collectivists). People from affective cultures are likely to respond in a direct way whereas those from neutral cultures are likely to respond in an indirect way (Trompenaars, 1995). Thus, people from affective cultures express their emotions more than people from neutral cultures. There is also a tendency for those with norms of emotional neutrality to dismiss anger (Trompenaars, 1995), whereas those from affective cultures are likely to consider it a salient event. Thus, the level of affectivity of a culture is likely to determine whether an event is perceived as affective or not.
The display of any emotion has a strong impact on communication, especially during negotiation. For example, Chinese negotiators are less likely to display their negative feelings during negotiation. Even when the negotiation is terminated, the Chinese are often reluctant to provide any feedback in order to avoid the unnecessary display of emotions, whereas this characteristic is unlikely for Australian negotiators. Therefore, it is evident that emotional displays differ across cultures and contribute to miscommunication.
What Do Negotiators Affect?
According to George et al. (1998), there are three key sets of determinants that negotiators affect during cross-cultural negotiations: individual differences, cross-cultural differences, and contextual factors. These key determinants suggest that individual differences are likely to be affected by an individual's character, feelings, thoughts, and behaviors. These differences may exist both within and across cultures.
George and her colleagues (1998) identified further two major affective dispositions in the cross-cultural negotiation process: negative and positive affect. The diversity literature suggests that the former will negatively affect a negotiation process as well as the negotiating parties' decisions, whereas the latter will bring positiveness to the negotiation parties' decisions as well as a positive negotiation outcome (George et al., 1998). The second dimension of these individual differences is how preexisting cultural knowledge should influence one's decision-making process. In cross-cultural negotiations, an individual is prone to be guided by the existing schemas and to rely on them to make decisions (George et al., 1998). This proves the value of pretraining negotiators for a given cross-cultural context.
There are also three broad types of cross-cultural differences: differences in internalized cultural values and norms; cross-cultural differences in emotional expression; and cross-cultural differences in linguistic style (George et al., 1998). These are discussed next.
Differences in Internalized Cultural Values and Norms. It is believed that these internalized values and norms are likely to influence the moods and emotions negotiators experience. In cross-cultural negotiations, each party brings its own set of values and norms, and these are likely to differ between parties. Negative affect and distress is likely to result when these differences cause conflicts between the negotiation parties. The party that comes from an affective culture has a tendency for deeper distress, whereas the party from the neutral culture has a tendency to discount the emotional expressions of the other.
Cross-Cultural Differences in Emotional Expression. The cross-cultural differences in emotional expression suggest that different cultures express their emotions differently. Reserved negotiators express their feelings less, and this may be interpreted by their counterparts as uncertainty, thus causing misunderstanding, which potentially sways the negotiation toward a negative outcome. This example is prevalent in the Chinese culture. Australians, on the other hand, easily express their emotions. This is often interpreted by the Chinese as overenthusiastic and therefore is seen as signifying undeclared hidden benefits that the Australians may be gaining over the Chinese.
Cross-Cultural Differences in Linguistic Style. George and colleagues (1998) proposed further that cross-cultural differences in linguistic styles dampen negotiators' positive moods and emotions, and contribute to negative moods and emotions during negotiation. The differences in linguistic styles also create uncertainty for negotiators (George et al., 1998). In a typical Australian-Chinese
9. UNDERSTANDIN G CROSS-CULTURA L NEGOTIATIO N 181
negotiation, the Chinese will fully engage the interpreter (generally the interpreter is a colleague of the negotiation team) and have lengthy discussions with the person in Chinese during the negotiation. However, Australians tend to be uncomfortable with this method of negotiation. As Ting-Toomey and Oetzel (2001) noted, the use of "code-switching" by a bilingual person is a tactic engaged in to increase or decrease distance toward his or her native culture by switching languages (Billikopf-Encina, 2000). In business negotiations, this process of code-switching assists the previously discussed boundary-building process through the increase or decrease of culture distance.
The Chinese are also likely to engage in conversation when Australian negotiators are speaking, as they trust their interpreter to translate words and intentions, and would rather utilize the time to form strategies. Again, Australians are uncomfortable with this Chinese characteristic. The more frequently these behaviors occur, the more likely it is that Australians become uncertain and consequently develop negative feelings toward the Chinese.
Although cross-cultural differences in negotiation are increasingly becoming a topic of interest (Tse, Francis, & Walls, 1994, as cited in George et al., 1998), unfortunately, the empirical literature on cross-cultural negotiations is very limited (Fisher, 1990; Pruitt & Carneval, 1993, as cited in Pearson & Stephan, 1998). Past studies suggest that people use different negotiation approaches due to differences in their perceptions of the decision-making situation (George et al., 1998). Zhao (2000) suggested that this difference is due largely to the value difference in cultures. As such, cultural differences and the inability to bridge cultural gaps often lead to business negotiation failures (Swierczek, 1990). Without cultural knowledge, negotiation instruction on collaboration should not commence (Shell, 2001).
This chapter has illustrated that some misunderstandings are fertile ground for major miscommunications in the cross-cultural negotiation context. These misunderstandings, which may evoke negative emotions, can be cumulative, building to a point where negotiation fails. This challenge indicates that the selection of negotiators is extremely important. They must possess cultural understanding, negotiation skills, and be able to (a) anticipate and shift convergence-divergence in concert with the other parties, (b) moderate their affective cultural background appropriately (see Fig. 9.1), and (c) appropriately adjust their linguistic styles on the basic understanding of both cultures to adjust boundaries favorably. Negotiators should also given the opportunity to plan the negotiation process.
When negotiation personnel are selected, training in cross-cultural negotiation should take place using case studies of other negotiators in the given negotiation context. Further, negotiators need to understand the impact of high-and low-context contexts within Hofstede's interpretation of power distance (Hall, 1976) and that the fundamental differences in cultural communication styles may result in serious miscommunication. To avoid this occurring in the Sino-Australian negotiation context, Australian negotiators should be assessed according to their cultural backgrounds and, if necessary, be provided with the necessary training to work with their negotiating counterparts. The Chinese, on the other hand, should also emphasize the need for monitoring the negotiation process to detect misunderstandings and act on them as they occur. If negative emotion is detected by either party, the negotiation process should be interrupted and, if necessary, negotiators may be replaced.
We have argued that cross-cultural negotiations may involve crossing boundaries and that miscommunications can be viewed as "boundary-crossing mishaps." Such mishaps are affected by negotiators' understanding of the respective cultures of the parties, negotiation skill, affective cultural background, cultural differences, emotional awareness and regulation, positive and negative affect, and discrepancy in convergence-divergence between the interactants. When too many hassles or mishaps occur, the negotiation breaks down as depicted in Fig. 9.1. As we have shown, it is often the accumulation of many minor misunderstandings that breaks the negotiation process and the possibility of interacting successfully across cultures.
In this chapter, we integrated the CAT model and AET to develop a model that captures the role of emotion in Australian-Chinese cross-cultural negotiations. The model predicts that converging negotiators will be evaluated positively, whereas divergers and maintainers will be evaluated as hostile and unfriendly. This likelihood is increased by the differences between the two cultures. For example, certain nonverbal communications exhibited by the Chinese may be interpreted as divergence and therefore negativity is sensed in the process. Therefore, this may result in the beginning of a negative spiral (see Fig. 9.1).
Misunderstanding in negotiations act as affective events, which evoke perceptions of boundary crossing as a hassle rather than an uplift. The result is a negative effect on emotions during negotiation, which leads to a diverging communication style and increases negative attitudes toward the negotiation. The negative emotional spiral puts the negotiation decision at risk. What underlies successful negotiation, therefore, is cultural understanding and negotiation skill coupled with emotional awareness and regulation.
Neal M. Ashkanasy
Wilfred J. Zerbe
The experience and expression of emotion is an essentially individual phenomenon (see chap. 6); however it is tightly bound within a social context. For example, the effects of emotion often become much more pointed when people are in groups. Interactions with others are a primary source of emotions; our relations with others serve to provoke, to help transfer, and to constrain emotions. These kinds of effects are reflected in the intragroup and intergroup situations that are the subject of the chapters in this section. Indeed, De Dreu, West, Fischer, and MacCurtain (2001) argued that group processes play a special role in developing emotion at this level of analysis. They described group interactions as a process of emotional incubation, where group members come to their group with a set of feelings and emotional expectations that are tested and then developed or diminished, depending on other group members' reactions. The resulting group-level emotional tenor then infuses the whole group, creating what De Rivera (1992, p. 197) referred to as group "emotional climate" (see also Barsade, 2002; Kelly & Barsade, 2001).
In the first chapter in this section, Matthew Grawitch and David Munz tackle the difficult issue of the role of individual affect in problem-solving workgroups. They present a model where individual and group affect norms influence each stage of ostensibly rational problem solving. The model, which they call GAPS (Group Affect in Problem Solving), includes four stages: problem identification; brainstorming; solution development; and implementation. At each of these stages both individual and group affect are influenced by the problem-solving process, with group affect becoming more influential over individual members' affect as the group matures. Distinctive features of the model are that it covers both positive and negative emotion, and includes reciprocal influences between the individual and the group problem solving process. Although yet to be tested, the model that these authors propose is the first to provide a comprehensive explanation of the nexus between affect and problem-solving in groups.
In chapter 8, Stefan Meisiek and Xin Yao address the manner in which emotions are promulgated in groups, through an original theory of social sharing of emotion (SSE) based on humor. In this model, humor is not something light and easily dismissable—as it so often is, outside the model. Instead, it is an essential tool that people at work use to deal with everyday affective issues, and to share experiences. In this model, humor is not just related to positive affect— it also provides a powerful medium for employees to deal with the negative things that happen to them, including strongly negative issues such as the death of a friend or colleague. Meisiek and Yao provide a thoughtful set of propositions concerning both the antecedents and the consequences of humor, including identifying the special roles of "canned", "practical", and "situational" humor, and the role of gender.
In the final chapter in this section, Mona White, Charmine Hartel, and Debra Panipucci consider intergroup issues, with a focus on what they refer to as "boundary crossing." Dealing specifically with cross-cultural communication in the context of strategic business negotiation, they identify in particular miscommunications in this context as "boundary-crossing mishaps." They argue that these mishaps are affected by numerous factors, including negotiators' understanding of the respective cultures of the parties, negotiation skill, affective cultural background of the parties, cultural differences, emotional awareness and regulation, negative affect, and discrepancy in convergence—divergence between the parties. Based on Weiss and Cropanzano's (1996) affective events theory, they propose that negotiation ultimately breaks down because of an accumulation of these mishaps. The theory is applied in the specific context of Chinese-Australian negotiation.
The three chapters in this section provide but a glimpse, however, of the topic of emotion in groups and teams. Indeed, we believe that this is perhaps the most underdeveloped arena for future research in the role of emotions in organizational behavior. From the perspective of bounded emotionality, the three chapters are especially convincing. It is clear from these authors' writing that emotion plays a central role in the way teams function, both endogenously and exogenously. Perhaps the most intriguing of the contributions, however, is Meisiek and Yao's. Here we see that humor, so often dismissed as unimportant, may be an important component of social cohesion in groups. In their model, the emotional relief of humor serves as the quintessential ingredient of group belongingness, and also as a release for negative events that beset group members from time to time. The practical and research possibilities of these ideas are, to our knowledge unexplored to date.
10. A BOUNDED EMOTIONALITY PERSPECTIVE 185
Mumby and Putnam's (1992) formulation proposed bounded emotionality as an alternative to bounded rationality. For them, bounded emotionality represented an emancipating organizing principle, based on a feminist reading of organizations, that stood in opposition to those underlying traditional, male-centered, rationally based organizations. It is not that they saw traditional organizations as unemotional, but rather that in traditional organizations emotions are managed for instrumental means. Indeed, they saw the concept of emotional labor as part of the organizational control inherent to traditional organizations. They proposed instead that organizations "be identified with work feelings rather than emotional labor" (p. 477): work feelings that are spontaneous and emergent rather than organizationally controlled, based on interrelatedness rather than used as commodities, meaning-centered rather than organizationally ascribed, and based on practical rather than instrumental reasoning. The chapters in this section are centered in traditional organizations, and although they adopt a tolerant view of the role of emotions in organizations, they implicitly adopt a more or less instrumental view of organizational emotions. Emotions thus serve as means to ends, such as effective problem solving, group harmony, and the outcomes of negotiations. On the one hand, then, bounded emotionality suggests that at the very least we need to acknowledge our assumptions as to the presence of instrumentality among the subjects of our research. Mumby and Putnam would advocate that ideally we would be open to alternative ways of conceiving organizations and become advocates for less instrumental forms of organizing (or forms that have different instrumentalities, i.e., aimed at achieving community rather than profit).
On the other hand it may be that bounded emotionality represents merely an alternative means of control, as Martin, Knopoff, and Beckman (1998) suggested. In their study of the Body Shop, an organization that was characterized by many of the features of a "feminist" organization, they found evidence for markers of bounded emotionality, but also found that organizationally and interpersonally ascribed emotional labor was alive and well, and that at times the expectation on employees to be free with their emotions was equally coercive. In short, and as the chapters in this section make plain, the interplay of individual, organizational, and interpersonal influences on emotions creates a complicated puzzle, for which we are only just beginning to make sense of the nature of the elements, much less the dynamics that connect them.
This page intentionally left blank
This page intentionally left blank
Robert S. Rubin
Vicki M. Staebler Tardino
Catherine S. Daus
David C. Munz
A close examination of the emotional labor literature reveals a myriad of conceptualizations contributing to a current state of theoretical disorientation. The purpose of this chapter is to clarify the nature of the construct and specify a more unified research model. To maximize theoretical clarity, the present model separates behaviors, emotional states, and situational demands. In addition, we suggest that the construct of perceived emotional dissonance is an important mediator between situational demands and emotional labor behavior, thereby disentangling the emotional state from behavior.
Recently, the construct of emotional labor has garnered a great deal of scholarly interest in the organizational literature (cf., Ashforth & Humphrey, 1993; Ashforth & Tomiuk, 2000; Brotheridge & Grandey, 2002; Brotheridge & Lee, 1998, 2002; Fisher & Ashkanasy, 2000a; Grandey, 2000, 2001, 2003; Kruml & Geddes, 2000; Mann, 1999; Pugliesi, 1999; Schaubroeck & Jones, 2000; Tews & Glomb, 2000; Wharton & Erickson, 1993). Along with the enthusiasm generated by such interest has come the rapid development of multiple theoretical frameworks. Although some aspects of these frameworks converge, many are quite theoretically discrepant. Consequently, these various conceptualizations have created confusion and have left researchers in a conceptual quandary regarding how best to understand the construct.
More specifically, current conceptualizations of emotional labor have confused or often treated as equivalent, two theoretically distinct constructs: One construct is a subjective state (experience) resulting from organizational requirements/norms regarding emotional expression, and the other is the motivated behavior undertaken to manage or regulate the subjective state. We propose, as have others (Mann, 1999; Wakefield, 1989), that it is conceptually necessary to separate emotional states from behavior. Thus, we contend that separating the state (as experienced) from behavior (broadly defined) will increase the theoretical clarity and ultimate utility of the emotional labor construct.
OF THE EMOTIONAL LABOR CONSTRUCT
Although diverse in nature, the previous conceptualizations of emotional labor can be roughly collapsed into three categories: (a) those that conceive of emotional labor as an emotional state arising from organizational or social norms and requirements; (b) those that suggest emotional labor consists of the behaviors undertaken to manage an implied or explicit emotional state (Note: "behavior" implies action and/or reaction, some of which may be unobservable, such as self-talk or cognitive reappraisals); and (c) those that consist of states, behavior, and/or situational factors. Most common, however, are conceptualizations that combine two or more of these categories within a single construct. For example, Morris and Feldman (1996) defined emotional labor as consisting of five dimensions, four of which are situational factors (frequency, duration, variety, and intensity of emotional display) and one of which is an individual state (emotional dissonance). Others have proposed similar definitions (Morris & Feldman, 1997; Rafaeli & Sutton, 1989; Sutton, 1991). Recently, Schaubroeck and Jones (2000) commented that the emotional state of dissonance (similar to Morris & Feldman, 1997) was the most similar in concept to their construct; however, their definition seems to also include situational factors ("in order to meet the demands of the job"), emotional states ("the experience of workers"), and behavior ("to suppress negative and express positive"). Mann (1999) defined the construct as having three dimensions: (a) expectations or rules about emotional display, (b) emotional suppression, and (c) emotional faking. As she implied, the first dimension represents a situational component and the second and third dimensions represent behavioral components. She remarked, "The last two dimensions incorporate the conflict idea, since emotions can only be suppressed or faked because of a discrepancy between expected and felt emotion" (p. 354). Thus, Mann's (1999) conception involves a behavioral reaction to a state of conflict.
One of the earliest conceptualizations of emotional labor (in the organizational literature) by Ashforth and Humphrey (1993) narrowed the scope considerably by defining emotional labor as "the act of displaying appropriate emotion" (p. 90). They added, "We prefer to focus on behavior rather than on the presumed emotions underlying behavior." Further, they explained that
11. PERCEIVED DISSONANCE AND EMOTIONAL LABOR 191
their definition "decouples the experience of emotion from the expression of emotion" (p. 90). Thus, they made a clear distinction of identifying emotional labor as a behavior versus an emotional state.
Although these formulations of emotional labor are certainly relevant and highly related to the construct, the collective consequence of these conceptualizations has been theoretical and practical confusion. Alicia Grandey (2000, 2003) developed a comprehensive model of emotional labor, defining emotional labor as an act rather than an emotional state and specifying situational factors and outcome variables. In particular, Grandey conceptualized emotional labor as the emotion regulation process enacted in response to situational cues such as service interaction expectations (frequency, duration, variety, display rules) and emotional events. Emotional regulation is undertaken to meet organizational demands felt by individuals, and can be achieved through "surface acting" (regulating expression) and "deep acting" (regulating feelings) (Hochschild, 1983). Thus, Grandey has separated the situational factors, such as display rules, from other parts of the emotion regulation process. We modify and extend her conceptualization by specifying the nature of an emotion state—perceived emotional dissonance—and expand the notion of emotional labor as one possible behavioral response to the state of perceived emotional dissonance.
TOWARD A RECONCEPTUALIZATION
OF EMOTIONAL LABOR
We employed Wakefield's (1989) framework of levels of explanation to disentangle the emotional labor construct. His model makes distinctions between distal explanations of behavior (e.g., primary traits such as the extraversion and intelligence) and more proximal levels of explanation (e.g., specific psychological states such as emotion and cognition). Pertinent to our discussion is the distinction between what he termed the intentional state, or what is going on within the person, and the actual behavior being performed, called the motivated act. Again, motivated acts are broadly defined such that the behaviors employed are not always visible to others, such as cognitive reappraisal. Wakefield used the term motivated act to make the distinction between unmotivated behaviors/actions such as automatic responses, and motivated or volitional behaviors/actions. He explained that because of its proximity to behavior, "intentional states form the reasons for actions, and therefore, explain behavior" (p. 335). Thus, intentional states are the best predictors of motivated acts within a given context. Current models of emotional labor have confused and/or merged the intentional state experienced by an individual with the motivated act or behavioral response to that intentional state. Wakefield's framework suggests that a given emotional state should influence behavior, but is necessarily distinct from the behavior.
Certainly we are not the first to argue that confusing the state and behavior may be theoretically unfavorable. Although we used Wakefield's model to guide our thinking delineating the emotional state from the motivated act, others have made similar distinctions in studying emotion constructs (cf. Weiss & Cropanzano, 1996). For example, Mann addressed the challenges presented by current emotional labor theory, stating, "The emphasis [of previous theory] thus ranges from the internal effort on the one hand, to the external behavioral display on the other. But which part is 'emotional labor'?" (p. 348). To date, the distinction between states, behaviors, and situational factors remains unspecified in the emotional labor literature.
We believe that an additional construct is necessary to conceptually clarify the literature and provide the critical link between the situational demands and state conditions, and the outcome, behavior or act. This alternate and intervening construct is needed to disentangle the response to the organizational requirements before emotional labor is undertaken. One such construct that has been examined in the literature is that of dissonance. We conceive of this dissonance as a perceived emotional state representing the dissonance between
felt emotion, and emotion that is perceived to be required and contend that it arises from situational demands combined with individual differences. Emotional labor in the form of surface acting or deep acting, then, is one possible motivated response to a psychological condition of high perceived dissonance. In Fig. 11.1, we offer our proposed model, and we next discuss the theoretical conceptualizations of each section. We begin with a discussion of dissonance, emotional labor, and their relationship, and then move to a discussion of situational factors, individual differences, and outcomes.
The Affective State of Perceived Emotional Dissonance
As mentioned earlier, we conceive of perceived dissonance as a disconnect between emotions that are genuinely felt and those that are required for the situation (Middleton, 1989). We place a high importance on this state of dissonance because we believe perceived dissonance to be a necessary precursor to emotional labor. A relatively recent but well-supported theory of job satisfaction, affective events theory (AET; Weiss & Cropanzano, 1996), underscores the critical nature of affective states as precursors to outcomes such as job attitudes and behaviors. Similarly, we consider perceived dissonance to be the affective state necessary for the behavioral response of emotional labor. There have, however, been a number of dissonance constructs (e.g., emotive disso
FIG. 11.1. Proposed research model of perceived emotional dissonance and emotional labor.
nance, emotional deviance, emotional dissonance) utilized in emotional labor research (Kruml & Geddes, 2000; Morris & Feldman, 1997; Zerbe, 2000). Our review of the literature shows generally two types of dissonance: (a) dissonance experienced prior to the behavior of emotional labor, and (b) dissonance experienced after this behavior has occurred. Zerbe (2000) attempted to provide some clarity surrounding the dissonance construct by distinguishing between these two types. Specifically, he described one type of dissonance as, "the degree of mismatch between felt emotions and displayed emotions, or faking" (p. 202). He described the second type of dissonance as "the mismatch between expressed emotions and local norms" (p. 202). Although both forms of dissonance are valid and intuitively appealing, only a mismatch between felt emotion and normative requirements for emotional display captures a purely affective state. Dissonance between felt and displayed emotion (or faking) requires a behavioral response to an affective state. Note that this type of dissonance (faking) is the "labor" of emotional labor and reflects a process whereby
one experiences an affective state and is motivated to deal with it.
Thus, in order to separate clearly emotional states from motivated behavior, we define dissonance as a purely affective state occurring prior to emotional labor. Morris and Feldman (1996) remarked that "It should require little emotional labor to sell products one genuinely believes in" (p. 992). Thus all things being equal, the larger the gap between genuine felt emotion and those required for the situation, the greater perceived dissonance. Again, note that dissonance is not a product of conflict between felt emotions and objective organizational requirements per se (i.e., written rules or supervisor instructions); rather, it is a discrepancy between felt emotions and what the individual perceives to be the required display. Much like the well-known literature on cognitive dissonance, emotional dissonance is likely to create psychological discomfort that individuals are motivated to reduce. We believe that emotional labor is the motivated behavior that individuals employ in an attempt to reconcile their feeling of emotional dissonance.
Emotional Labor: A Motivated Behavioral Response
to Perceived Dissonance
We argue that individuals typically engage in one of two strategies for dealing with a situation of perceived dissonance. These strategies are similar to Hochschild's (1983) and Grandey's (2000) mechanisms for managing emotional demands, but do not include the appraisal of the affective state (perceived dissonance). The first behavioral strategy for managing perceived dissonance, "surface acting," is accomplished when an individual simulates emotion that is appropriate for the situation. An example of surface acting would be when a customer service representative smiles and acts warmly in the face of an abrasive and disrespectful customer who churns up feelings of anger and resent
11. PERCEIVE D DISSONANC E AND EMOTIONAL LABOR 195
ment. The expression of positive emotion is fitting for the customer-service context, but negative emotion is being experienced.
A second strategy for managing perceived dissonance is through the process of "deep acting," in which the individual endeavors to experience the emotion that is appropriate for the situation. In this case, the individual is actually changing the feeling experienced prior to the behavior by changing the underlying causes, or the mental construction of the feeling state (e.g., cognitive reappraisal). Changing the feeling can occur by evoking or suppressing an emotion, or through trained imagination (Hochschild, 1983). For example, a ticket counter agent may imagine himself successfully completing transactions and noticing the smiles on customers' faces. As Grandey (2003) implied, deep acting may require advanced emotional regulatory skills.
In sum, Hochschild claimed that surface acting can be likened to compliance, while deep acting strategy can be likened to conversion. Both strategies can be adaptive (i.e., allow for productive performance) in terms of dealing with perceived dissonance. Over the long term, deep acting should be more adaptive in that it allows the individual to internalize a role, and thus reduce the level of perceived dissonance (as shown by the feedback loop in Fig. 11.1). Regardless of adaptability, both strategies are based on the notion of acting—that is, manufacturing a substitute for the emotion one genuinely wants to display. Thus, surface and deep acting may temper perceived dissonance, but are unlikely to eliminate completely the source of the motivation to act, or perceived dissonance.
Extending Hochschild's typology, Ashforth and Humphrey (1993) noted that there are instances in which the sanctioned emotion corresponds to the felt emotion, as is ordinarily the case in medical emergencies. Ashforth and Humphrey (1993) and Tews and Glomb (2000) classified this match between genuine emotion and the situational requirements as a third and important way of accomplishing emotional labor. Yet, note that we propose that if one does not perceive dissonance, one will not be motivated to engage in behaviors that might be described as emotional labor. Regardless of the organization's normative requirements, if one does not experience dissonance, there will be no motivation to act, to "put on a face", and thus, no emotional labor—the emotional display will be genuine. If an employee is required by the organization to act pleasantly, and he or she does not perceive this to be emotionally taxing, there will be no necessity for him or her to "act" pleasantly. In fact, a match between organizational requirements and felt emotion may even facilitate or emotionally energize an individual to display particular emotions and would most likely appear to be effortless. Taken together then, perceived dissonance and emotional labor represent the entire emotion regulation process as Grandey (2000) discussed; however, our framework separates situational factors, emotional states, and behaviors: Perceived dissonance is the perceptual state of appraisal of the situational factors, and emotional labor is the behavioral strategy chosen to manage the dissonant state.
As noted, we view emotional labor as one potential and typical response to perceived dissonance. There are then, multiple other strategies for reconciling the state of perceived emotional dissonance, including withdrawing from the task or interaction completely (e.g., hanging up on a customer); becoming hostile or aggressive (e.g., raising one's voice); becoming violent (e.g., physically attacking a customer); or using passive-aggressive behavior (e.g., acting polite during the interaction and then sabotaging a customer's order afterward). These behaviors, however, are unlikely to be viewed by the organization as constructive and representative of appropriate protocol or policy and therefore, are likely to be punished or at least heavily discouraged. Because both emotional labor strategies (surface acting and deep acting) are attempts to produce organizationally desired and rewarded behavior, they are likely to be the strategy of choice utilized by individuals attempting to reduce perceived emotional dissonance, while reaping organizational benefits and rewards (e.g., promotions). Thus, depending on the situational requirements, the actual behavioral display could be positive or negative (e.g., A debt collector must be intimidating and downright hostile at times to be successful). Although this chapter does not directly address at length the numerous possible responses beyond labor, we do present some of them in the model to emphasize their potential importance in future research. Thus, Fig. 11.1 is illustrative, not exhaustive.
Situational Demands of Perceived Dissonance
As stated earlier, our premise is that much of the early work on emotional labor is descriptive of an emotional state influenced by a variety of factors. Given the proposed model, our attempt here is to identify situational demands that impact perceived dissonance and thus indirectly impact emotional labor. We discuss two categories of situational demands we believe largely contribute to the perception of dissonance and subsequent emotional labor. As the model depicts, these situational demands are comprised of (a) job-specific demands, including characteristics of the job, and (b) emotion-specific demands, including emotional characteristics of the job, display rule norms, and a special case of display rules, customer expectations. Interestingly, these variables have all appeared in prior research on emotional labor; however, many have actually been labeled as dimensions of the emotional labor construct (e.g., display rules and frequency of emotional display), and not as situational factors. Our view is that these important variables work together to create the critical environmental stimuli that may lead to an appraisal of perceived dissonance, and subsequent emotional labor acts.
Job Characteristics. Humphrey (2000) suggested that the critical psychological states resulting from job characteristics (Hackman & Oldham, 1975) clearly impact affective conditions like job satisfaction and thus they should
11. PERCEIVE D DISSONANC E AND EMOTIONA L LABO R 197
strongly impact secondary work behavior ("behavior that is influenced by the work task, but not actually part of the task," p. 239) including emotional displays. He further argued that job characteristics influence the social context of the work environment. Thus, job characteristics are powerful contextual factors that contribute to an individual's affective state. In our terms, the affective state of perceived emotional dissonance should similarly be impacted by job characteristics.
To date, research on emotional labor and job characteristics has focused primarily on job autonomy. For example, Morris and Feldman (1997) reported a negative relationship between autonomy and emotional dissonance. Pugliesi (1999) explored the relationship between emotional labor and certain characteristics of the job among university employees. She found negative correlations between emotional labor and job characteristics such as control (autonomy) and complexity (skill required, level of challenge, variety and intrinsic interest) of the job. Other research yielded similar results (Rafaeli & Sutton, 1989; Wharton, 1993). Yet beyond skill variety and autonomy, very little research exists linking other job characteristics to emotions. However, given the effects found for autonomy, we argue that other job characteristics might impact perceived dissonance similarly. For example, individuals who experience a high degree of meaningfulness via skill variety, task identity, and task significance are likely to experience a more positive psychological state (Humphrey, 2000). This state should positively influence individuals' emotional appraisal of their work situation and lead to lower perceived emotional dissonance, or at least make it easier to maintain a positive emotional state. Similarly, high responsibility (autonomy) and knowledge of results (feedback) from the job should lead to a lesser degree of perceived emotional dissonance.
Display Rule Norms. Display rules can be conceptualized as societal, occupational, and organizational norms that provide structure for service transactions (Ashforth & Humphrey, 1993; Ekman, 1973a; Rafaeli & Sutton, 1989). Display rules serve to regulate the type of sanctioned expression in a given situation, as well as the degree to which it is expressed. In an organizational service setting, display rules are often highly explicit to ensure consistency of service. Additionally, display rules are found in all types of organizational communication, including handbooks, performance appraisals, and new employee orientation classes. For example, at the Ritz-Carlton hotels, employees are required to memorize service credo cards, which explicitly state what is meant by service, and which all employees carry on their person at all times. One service "basic" reads, "Smile—We are on stage. Always maintain positive eye contact. Use proper vocabulary with our guests (e.g., "Good morning", "Certainly" and "My pleasure")." Service employees are "on stage" in the sense that these display rules require employees to become actors (Grandey, 2001), acting out organizationally required emotions that may not be congruent with their own personal emotions.
Empirical research demonstrating a link between display rules and emotional labor has provided mixed results. For instance, Hochschild (1983) found that explicit display rules for flight attendants heavily impacted their behavior on the job. In contrast, Morris and Feldman (1997) found that the explicitness of display rules was negatively related to the frequency of emotional labor. These differences are most likely due to the discrepancies in conceptualizations of the construct. As we mentioned, it is important to disentangle the stimulus (e.g., the organizational context) from an individual's internal state (e.g., the felt experience). Our model suggests that an incongruence between the desired emotional expression required by the organization and the felt emotion of the individual creates an experience of perceived dissonance. Therefore, highly explicit display rules would exacerbate the importance of sanctioned emotional expression and create a larger gap between required and felt emotion. Thus, the less latitude employees have (i.e., the more explicit the display rules) in expressing their own felt emotions in lieu of organizationally sanctioned emotion, the greater is the possible perceived dissonance.
Customer Expectations. Schneider and Bowen (1999) asserted, "By concentrating on [customer] needs, we have an opportunity to delve more deeply into the customer's internal states to offer managers insights about how to create customer delight, as well as how to avoid customer outrage" (p. 38). In addition, they highlighted multiple needs customers expect to receive from service transactions that contribute to positive organizational outcomes, including keeping promises and commitments, help when needed, and friendliness, honesty, and politeness. Similarly, Zeithamal, Parasuraman, and Berry (1990) found that customer evaluations are a function of 10 dimensions, including responsiveness, courtesy, credibility, access, communication, and understanding (in Ashforth & Humphrey, 1993, p. 91). Clearly, customer needs and criteria for good service are laden with emotional content (e.g., "Greet every customer with a smile"). Thus, customer expectations are a special and highly salient type of display rule.
For example, a customer can easily interpret a simple sigh exuded by a customer service agent to mean "I don't care about you—I want to go home," when in fact the agent may simply be tired. Recently, Pugh (2001) found a positive relationship between the display of positive emotion by bank employees and customer positive affect. Similarly, Daus (2001) found converging evidence from a laboratory experiment where expression of positive emotion contributed to higher evaluations of performance for waiters/waitresses. This research confirms what organizations have intuitively known for years—positive employee displays satisfy customers. Organizations are increasingly advertising and educating their customers about this relationship, and customers now approach service exchanges with clear expectations for what good service should look like (e.g., Delta Airline's slogan, "We love to fly and it shows").
11. PERCEIVED DISSONANCE AND EMOTIONA L LABOR 199
Thus, for some organizations, customers will have incredibly high service expectations, including expectations for how employees will display emotions. These expectations often are built into the fabric of the organization and serve as yet another source of information about how employees should behave in customer service interactions. Yet the paradox is that this may be taxing to employees because organizations that promise customers consistent top-notch service thereby increase the emotional requirements of their employees. We therefore propose that customers' expectations will be positively associated with the perceived dissonance of employees.
Emotional Job Characteristics. Service-role jobs vary considerably across industries and job categories. Logically, these jobs also differ in their emotional characteristics, or the specific type and level of emotional expression inherent in the job itself. These specific emotional characteristics of the job may lead to variance in how an individual actually feels on the job. Therefore, we include them in our model as predictors of perceived dissonance. Three emotional job characteristics that have appeared in prior research are the frequency of emotional display or interaction, duration of emotion display or interaction, and variety of emotions required. Again, these have previously been presented as dimensions of the emotional labor construct (Hochschild, 1983; Morris & Feldman, 1996, 1997). We separate them from the behavior of emotional labor as Grandey (2000) did, and extend her work to conceptualize them as precursors to perceived dissonance.
Frequency of emotional display or interaction can be defined as the rate of emotional displays or interactions between service providers and clients (Grandey, 2000; Hochschild, 1983; Morris & Feldman, 1997). Some jobs inherently require higher frequencies of emotional displays than others. For example, a cashier at a busy fast-food restaurant may be required to serve hundreds of people in one shift, whereas a concierge at a small hotel may only service a handful of people in a day. In their study of supermarket cashiers, Rafaeli and Sutton (1990) found that during particularly busy hours, cashiers displayed fewer positive emotions. Other research has shown both positive (Brotheridge & Lee, 1998) and negative (Grandey, 1999) relationships between frequency and emotional labor. Note that these studies did not measure the impact of frequency on the emotional state, but rather the impact of frequency on behavior, which may explain the inconsistent findings. The state should impact behavior more strongly than contextual circumstances (Wakefield, 1989). Therefore, for any particular job, the greater the frequency of emotional display required, the more likely it is that an individual will experience a higher degree of perceived emotional dissonance.
Duration of emotional display or interaction is generally referred to as duration. Service interactions that are relatively short tend to follow a scripted interaction schema (Rafaeli, 1989) and require less emotional effort. As the duration increases, so does the level of emotional effort required to go beyond the social script in order to comply with organizational demands. Back to our previous example: The cashier at the fast-food restaurant may on average spend about 3 or 4 minutes with a customer, whereas the concierge at the hotel may spend upward of 10 to 15 minutes working with a customer planning entire outings. Thus, the concierge's job requires him or her to engage in positive emotional displays for more extended periods. Long durations could lead to a feeling of heightened perceived dissonance if in fact the discrepancy between what the job requires from the concierge and how he or she feels is too great. Thus, we propose that the longer the duration of interaction, potentially the greater is the amount of perceived dissonance.
Certain jobs require a greater variety of emotions to comply with organizational requirements. For example, some occupations "often require frequent changes of emotions that are displayed: positive emotions to build enthusiasm, negative emotions to support discipline and neutrality of emotions to demonstrate fairness and professionalism" (Morris & Feldman, 1996, p. 992). In the theatrical arts, actors often talk about "range" in reference to the ability to play a wide variety of characters and across multiple genres (e.g., slapstick comedy vs. Shakespearean theater). Similarly, service employees are often called on to express a wide range of emotions in order to successfully perform on the job. Some of these emotional expressions may even contradict one another. For example, an employee might force a smile while communicating a tough policy issue to a customer in a firm voice. Few researchers have explored this idea empirically, but there are similar theoretical conceptions (Grandey, 2000; Wharton & Erickson, 1993). Our proposition is that jobs requiring a wide range of emotions and/or multiple emotions to express simultaneously have a greater impact on one's emotional state than jobs requiring only a few emotions to be expressed. Much like cognitive load, perceived dissonance could increase under such circumstances.
The situational demands just outlined seem to contribute to an individual's overall affective state. Clearly, a single property may not produce a significant change in the emotional state; however, the cumulative effect contributes to the experienced state. Thus, it should be acknowledged that although the main effect of any single antecedent may impact perceived dissonance (positively or negatively), the potential interactions are likely to yield more complex types of relationships and may strengthen, weaken or nullify the main effects.1
The Direct and Moderating Effect of Individual Differences
In this model, perceived dissonance mediates the relationship between situational demands and emotional labor. All things being equal, a number of indi-
We thank Blake Ashforth for bringing this important point to our attention.
11. PERCEIVE D DISSONANC E AND EMOTIONAL LABOR 201
vidual difference variables could potentially influence the amount of perceived dissonance experienced, thereby impacting subsequent emotional labor and concomitant outcomes. In other words, some individuals might never experience a high degree of perceived dissonance whereas others might be plagued by it, even given identical antecedent conditions. It follows logically then that individual differences may also differentially affect the emotional labor process. In all, individual difference variables might influence the perceived dissonance-labor relationship. Directly, individual differences may impact the amount of perceived dissonance one experiences and the type and level of emotional labor undertaken. Individual differences may also moderate the relationship between situational demands and perceived dissonance, as well as the relationship between perceived emotional dissonance and emotional labor. We next use examples of individual differences from the extant literature to illustrate some of these potential pathways of influence.
Role Internalization. Role internalization or identification denotes the degree to which individuals identify with the job (or job values). Ashforth and Humphrey (1993) asserted that role internalization is a potential outcome of compliance with display rules over time. Morris and Feldman (1997) confirmed that assertion, demonstrating that longer duration of emotional labor results in greater internalization of role expectations. We do not disagree with this perspective; however, there may be a case for role internalization as an individual difference variable directly influencing perceived dissonance. Specifically, individuals who have internalized their career role as a "customer service" employee are likely to experience less perceived dissonance than individuals who have career identities in opposition to service. For example, an individual who has worked for multiple years in one service company may accept a job in a similar service company, multiple times within the same industry. Although each job is different, he or she is likely to incur similar role expectations on the new job and may have already internalized the standard industry display expectations. For instance, the task requirements of depart-ment-store sales clerks may be substantially different in terms of procedures, but rather similar in terms of customer service. Thus, the individual begins to view "customer service agent" as a career descriptor, much as an accountant (or other professional categories) might, and internalizes or transforms into the role. For these individuals, emotional expression is not a chore but rather second nature; it is what they do. Recently, Ashforth and Tomiuk (2000) further delved into this construct empirically, and proposed that role internalization leads to an experience of "deep authenticity" whereby "one's emotional expression or display is consistent with the display rules of a specific identity that one has internalized (or wants to internalize) as a reflection of self—re
gardless of whether the expression genuinely reflects one's currentfeelings" (p. 195). According to these authors, deep authenticity should lead to either no emotional dissonance (our construct of perceived dissonance) or, at worst, low emotional dissonance (when the temporary affective state is incongruent with normative requirements). Therefore, we propose that role internalization should buffer the effect of situational demands, as well as directly impact perceived dissonance, leading to less perceived dissonance.
Emotional Intelligence. Another construct that would seem intuitively to influence directly perceived dissonance is emotional intelligence. Emotional intelligence focuses on an individual's ability to both interpret emotional information (both internal and external) and manage it in an adaptive manner. Most importantly for the discussion of emotional labor, emotional intelligence (through mood maintenance) is an ability that allows people to better regulate their emotions compared to others (Ciarrochi, Chan, & Caputi, 2000; Mayer, Caruso, & Salovey, 2000; Mayer & Salovey, 1997). Mayer and colleagues (Mayer et al., 2000; Mayer & Salovey, 1997; Salovey & Mayer, 1990) argued that emotional intelligence has four important components. First, it involves the ability to perceive and appraise emotion, both verbally and nonverbally. For example, emotionally intelligent individuals are better at accurately identifying various emotional expressions on individuals' faces (Ekman, Friesen, & Ancoli, 1980), as well as at appraising emotional content in art or music, than nonemotionally intelligent individuals. This ability would seem vital for jobs where reading emotions of customers and clients is a critical precursor to interacting with them effectively.
Emotional intelligence also consists of the ability to assimilate emotions into perceptual and cognitive processes. Thus, emotionally intelligent individuals are better able to be flexible when planning, to generate more creative solutions, and to redirect attention and motivation (e.g., directing feelings away from trivial problems toward those of higher priority) (Mayer et al., 2000; Mayer & Salovey, 1997). Thus, the emotionally intelligent customer service agent should be better able to maintain focus during particularly busy times, allowing for creative and emotionally appropriate responses to impatient customers.
Third, emotional intelligence consists of the ability to reason about and understand emotions. A person with this higher level ability would have the capacity to understand, for example, how emotions progress from one to another (Mayer, 2001; Mayer et al., 2000). It would certainly seem to be an asset for a customer service agent to be able to recognize when and how frustration in a customer could lead to anger, which could progress to rage.
Finally, emotional intelligence involves the ability of regulation or management of emotion in oneself and in others. This is perhaps the key ability involved in emotional labor. The central mechanism for regulating emotions is mood manipulation. In the self, individuals are generally motivated to maintain or prolong pleasant moods and curtail unpleasant moods—"mood mainte
11. PERCEIVE D DISSONANCE AND EMOTIONAL LABO R 203
nance" and "mood repair," respectively (Isen, 1985). An emotionally intelligent customer service representative who recognizes her own and her customers' moods may be able to utilize the information to better serve her customers. For example, if she knew she was in a bad mood, she might try extra hard to avoid displaying this to her customers (Berkowitz, Jaffee, Jo, & Troccoli, 2000). As such, this ability inherently involves emotional expression that encompasses empathy (Ciarrochi et al., 2000; Fox & Spector, 2000; Hogan, 1969; Mayer et al., 2000; Mayer, DiPaolo, & Salovey, 1990). Empathetic responses lead to genuineness and appropriate behaviors, whereas a lack of empathy leads to ill-mannered behaviors. Empathy, then, is critically important in jobs that tend to require emotional labor.
Highly emotionally intelligent individuals are likely to experience less perceived emotional dissonance, based on their ability to effectively interpret and regulate their own and others' emotional states. Emotional intelligence may directly and indirectly influence the type of emotional labor undertaken and skill at emotional labor. In fact, recent research (Jordan, Ashkanasy, & Hartel, 2000) indeed demonstrated a significant moderating relationship of emotional intelligence on emotional labor. Thus, we propose that emotional intelligence abilities will directly affect perceived emotional dissonance, as well as moderate the perceived dissonance-emotional labor relationship.
Positive and Negative Affectivity. Positive and negative affectivity (PA/ NA) have particular relevance to our model. According to Watson (2000), "Affective traits/temperaments provide a 'set point' that is characteristic for each individual but differs across individuals. Mood varies around this set point" (p. 18). Given that this set point is variable across individuals, PA/NA may directly influence the amount of perceived dissonance an individual experiences. A person who is higher in positive affectivity may perceive less dissonance across situations. Generally, persons high in positive affectivity perceive things more positively, and persons high on negative affectivity generally perceive things more negatively. Recent research has focused on the role of positive and negative affectivity as a moderator of situation-response relationships (Munz et al., 1996; Williams, Gavin, & Williams, 1996). Specifically, PA and NA are assumed to substantively affect other variables and their relationships through perceptual, temperamental, and instrumental processes. For example, to the degree that employees with high NA perceive more stringent display rules (perceptual), have stronger negative reactions to internal and external stimuli (temperamental), and behave to evoke environmental reactions (instrumental), then NA will moderate the perceived dissonance-labor relationship. Similarly, Abraham (1999) suggested that individuals high in negative affectivity tended to experience a conflict between expressed (positive) and felt (negative) emotions in organizations that demand the expression of positive emotions. She found that negative affectivity moderated the relationship between emotional
dissonance and job satisfaction, whereas it mediated the relationship with emotional exhaustion.
In all, we believe that individual differences augment the understanding of the perceived dissonance-emotional labor process. The three individual differences highlighted in our model were used to illustrate the importance of individual difference variables in this process. Clearly, there are a wide range of individual difference variables that might be intuitively hypothesized (e.g., self-monitoring, public vs. private self). Future research is critical to establish the relevance and role of individual differences and their direct and/or indirect effects.
Individual and Organizational Outcomes
There is little doubt that emotions play out in various ways and are deeply embedded in organizational life. As such, research on emotions has been concerned with important individual and organizational constructs (e.g., job satisfaction, stress, burnout, organizational justice) that have strong direct and indirect effects on organizational success (Nord & Fox, 1995).
Our model suggests that there are a number of significant individual and organizational consequences of high perceived dissonance and subsequent emotional labor. It is important to note that in the short term, organizations may clearly benefit from individuals engaging in emotional labor; however, over time individuals may become increasingly emotionally exhausted, which could lead to a family of deleterious behaviors (e.g., burnout and performance) or, in extreme cases, a choice to employ a less appropriate strategy (e.g., workplace violence).
In other words, we recognize that the next subsection paints a rather gloomy scenario such that engaging in emotional labor in the long term leads to primarily deleterious outcomes and that all service employees will incur negative outcomes. This is not entirely our intention. Clearly, there are satisfied, authentic, lifelong employees who exist in service roles. Indeed, there are some positive effects that may emerge in the short term. However, based on our definition, the lifelong service employee is unlikely to be experiencing significant and prolonged periods of emotional dissonance. Individuals who are engaging in emotional labor behavior do so because they labor to reconcile a dissonant state. Thus, if one does not experience this dissonance, one will not be required to act incongruously with one's own genuine feelings and thus it is not laborious to act. As mentioned earlier, individual differences or situational demands that buffer or eliminate these effects or an interaction of the two may contribute to low perceived emotional dissonance.
Our discussion of outcomes covers three general categories, some of which are interrelated and include job-related attitudes and behaviors, physical and psychological well-being, and withdrawal behaviors. The following is not in
11. PERCEIVE D DISSONANCE AND EMOTIONAL LABOR 205
tended to exhaust every potential direct or indirect outcome of emotion management or emotions at work, but to focus on those outcomes most closely related to emotional labor as identified in the extant literature. Others have provided broader reviews beyond the scope of this chapter (cf. Mastenbroek, 2000b).
Job-Related Attitudes and Behaviors. A number of job-related attitude and behavior outcomes of emotional labor have been offered in the literature (Ashforth & Humphrey, 1993; Grandey, 2000; Morris & Feldman, 1996, Pugliesi, 1999). Many have suggested that emotional labor negatively impacts job attitudes such as job satisfaction and organizational commitment (Grandey, 2000; Hochschild, 1983). Some, however, point to potential benefits, arguing that emotional labor allows individuals to gain a sense of control over their work environment (Wharton, 1993) or allows for increased enjoyment (Ashforth & Humphrey, 1993). To date, there have been mixed findings regarding job attitudes, which may be attributable to the various operationalizations of emotional labor (Grandey, 2000). Given that many job attitudes are highly affective in nature (Weiss & Cropanzano, 1996), we contend that perceived dissonance and subsequent labor may lead to negative job attitudes (all things being equal). Thus, an individual who experiences a great deal of perceived dissonance and engages in emotional labor to reconcile the dissonance might feel like he or she is not being "true" to him- or herself. In the service role, where employees often are asked to "put the customer first," one's own feelings about one's job situation are secondary. As Hochschild (1983) suggested, this continuously felt inconsistency is likely to lead to a negative appraisal of one's job circumstances. For example, Hochschild (1983) related accounts of flight attendants who resented the notion of the "customer is always right," especially when the customer was clearly not right. Other research has supported this. Pugliesi (1999) found that among university staff, emotional labor decreased job satisfaction. Therefore, we propose that long-term emotional labor will decrease job satisfaction and would have a similar impact on other job attitudes such as organizational commitment and turnover intention. Indirectly, then, emotional labor might lead to other negative consequences, such as turnover, as a result of these negative attitudes (Griffeth, Horn, & Gaertner, 2000). Further, as evidenced by Hochschild, we propose that long-term emotional labor will similarly increase employee resentment or cynicism, which has also been shown to have detrimental effects in organizations (Bommer & Rubin, 2001; Reichers, Wanous, & Austin, 1997).
Individuals experiencing a high degree of perceived emotional dissonance and engaging in emotional labor might feel as if their emotional resources are depleted or overloaded. When struggling to draw on their own emotional resources to "put on a happy face," employees' willingness to display extrarole behaviors such as coming to the aid of a coworker or resolving interpersonal conflicts (Podsakoff, MacKenzie, Paine, & Bachrach, 2000) should decrease. This idea is consistent with the emotional regulation literature indicating that individual performance tapers off when one is forced to spend resources regulating one's own emotions (Gross & Levenson, 1997; Muraven, Tice, & Baumeister, 1998). In general, individuals who experience high perceived dissonance and respond via surface acting or deep acting will be less likely to go the "extra mile" on the job. Again, we depart from other researchers such that the relationships just described hinge on an individual experiencing perceived dissonance and choosing to engage in emotional labor. For individuals who have internalized their role to a great degree and do not experience perceived dissonance, we would not expect our model to predict attitudes or behaviors. We do not mean to suggest that a lack of perceived dissonance and subsequent emotional labor will lead to satisfaction and other positive behaviors. There are, of course, multiple pathways to job attitudes and behaviors. Our focus here is on how emotional labor might produce these outcomes.
Physical and Psychological Well-Being. As Grandey (2000) remarked, "Emotional labor may result in good organizational performance, but may have consequences for the employee's health" (p. 107). Long-term emotional labor efforts may lead to an increase in stress and reported physical and psychological strain. Hochschild (1983) suggested that long-term emotional labor would lead to a host of dysfunctional behaviors that impact individual well-being, including excessive drinking, drug use, and sexual dysfunctions. For example, Schaubroeck and Jones (2000) showed that individuals who perceived a greater demand to express positive emotions (i.e., contributing to dissonance) on the job reported a greater number of somatic symptoms. Grandey (2000) also suggested that emotional labor might lead to increased individual stress over time. Stress has a long history of study in management demonstrating its direct, negative relationship with job attitudes (e.g., job satisfaction), as well as long-term physical (e.g., heart disease) and mental health (e.g., depression) consequences (Matteson & Ivancevich, 1987). Yet here too there are mixed findings. For instance, Zerbe (2000) found that emotive dissonance (similar to our conception of perceived dissonance) was not associated with employee psychological well-being (as measured by dimensions of burnout). However, Pugliesi (1999) found that emotional labor was strongly and positively related to job stress and increased psychological distress. Again, these differences are possibly due in part to the range of operationalizations of emotional labor.
One important potential consequence of long-term emotional labor is that of burnout, which has been characterized by emotional exhaustion, depersonalization, and diminished personal accomplishment (Jackson, Schwab & Schuler, 1986; Maslach, 1982) primarily arising from stress. Burnout is a common syndrome found most often in helping professions such as nursing, social
11. PERCEIVE D DISSONANC E AND EMOTIONAL LABOR 207
work, and customer service and has been shown to have deleterious effects on both the individual and the organization (Cordes & Dougherty, 1993). There is growing evidence showing a positive relationship between emotional labor and burnout, often captured via the emotional exhaustion dimension of burnout. Hochschild (1983) argued that individuals who engage in deep acting strategies should experience more exhaustion because they become emotionally involved with the customers they serve. Brotheridge and Grandey (2002) did not find support for this hypothesis, showing a positive association between surface acting and emotional exhaustion but no relationship between deep acting and emotional exhaustion. Similarly, others found little support for Hochschild's contention (Brotheridge & Lee, 1998; Kruml & Geddes, 2000). Kruml and Geddes (2000) showed that individuals experience more stress by faking emotion (rather than genuinely expressing what they feel), thereby demonstrating a potential link to burnout. Morris and Feldman (1997) found a positive relationship between emotional labor and emotional exhaustion among debt collectors, military recruiters, and nurses. Thus, it seems that emotional labor has a potentially dangerous cascading effect (beginning with stress) on individuals who engage in it long-term.
Suppression of emotion has also been shown to have harmful effects on health, as well as psychological well-being and adjustment (Balswick, 1988; Berry & Pennebaker, 1998; Brotheridge & Grandey, 2002; Feshback, 1986; Gross, 1998b; Labott & Martin, 1987). In contrast, inhibition of negative emotional expression in the context of the workplace, such as avoiding an emotional outburst, generally speaking has been conventionally regarded as a positive characteristic. In other words, organizations generally want their employees to suppress negative emotional expressions. Yet again, research has demonstrated several negative individual and organizational consequences of inhibition of negative emotions, including decrements in problem solving, personal growth, and job satisfaction (French & Bell, 1990). Also, Gross (1998b) and Gross and Levenson (1997) addressed the effects of inhibition and suppression of emotion on the individual. Although there have been mixed findings, the data seem to suggest that faking emotion, regardless of the strategy employed, is harmful to the individual. Thus, individuals who experience high perceived dissonance may be in a "catch 22," whereby expressing emotions that are incongruent with display rules leads to negative outcomes (e.g., poor performance ratings), yet conforming to display rules (via emotional labor) leads to other negative outcomes (e.g., burnout).
Given the logic just presented, the relationship between emotional labor and performance is not altogether straightforward. On one hand, there is evidence that performing emotional labor strategies (either through surface acting or deep acting) leads to good customer service interactions (Daus, 2001; Pugh, 2001). In other words, the employee is able to successfully cope with the perceived dissonance he or she experiences by "acting" out his or her role in accordance with the display rules for the job and thus performs well (i.e., the customer is satisfied). On the other hand, insincerity of emotional display may be detectable by customers (Rafaeli & Sutton, 1987), leading the customer to believe the employee is "faking" the smile or simply going through the motions. Further, emotional labor is related to burnout, and there is evidence building that burnout, particularly emotional exhaustion, is negatively related to performance (Cordes & Dougherty, 1993; Wright & Bonett, 1997). Grandey (2000) suggested that the type of emotional labor strategy employed (i.e., surface acting or deep acting) will differentiate between good and bad performance. For instance, she asserted that surface acting, whereby the employee fakes the emotional display, would lead to poor performance based on customers' sensitivity to genuine expression. Deep acting, whereby the employee truly adopts the required emotional display as his or her own, should lead to good performance. We agree with this perspective but propose that a temporal element is important to consider. Long-term deep acting in which the person is able to truly feel and express the emotion required may lead to poor performance over time. A Broadway actor performing his or her 550th musical show is more likely to experience burnout than when he or she did the first 50 shows. If he or she has truly internalized the role (i.e., "become" the character), however, and less emotional energy is expended, the actor may be able to avoid complete burnout.
Withdrawal Behaviors. Individuals who experience sustained perceived dissonance over time in service roles and respond via surfacing acting or deep acting could begin to withdraw as a way of escaping the persistent emotional burden they perceive. Hulin and colleagues conceptualized withdrawal behavior as a family of adaptive behaviors that help individuals avoid the work circumstances (Hulin, 1990; Hulin, Roznowski, & Hachiya, 1985). Their model (Hulin et al., 1985) categorized withdrawal behaviors into four families: (a) behavioral intentions to increase non-job outcomes, such as using work time for personal tasks; (b) behavioral intentions to reduce job inputs, such as taking long breaks or talking with coworkers about non-business-related topics; (c) behavioral intentions to reduce work role inclusion, such as absenteeism, tardiness, or quitting; and (d) behavioral intentions to change the work role such as unionization or transfer attempts. Thus, in the short run, individuals may take refuge from the labor of emotional labor in the form of taking longer breaks or leaving a workstation unattended; however, in the long run this might lead to more significant behaviors such as turnover or requests for transfers. This argument seems to warrant merit. For instance, Grandey (1999) found a positive relationship between turnover intent and emotional labor, which may suggest that the high turnover rates observed in service industries (e.g., hospitality and call centers) may be in part due to emotional labor.
11. PERCEIVE D DISSONANCE AND EMOTIONAL LABOR 209
Continuous regulation of emotion may lead an employee to believe he or she is working in the wrong environment, and thus the employee becomes motivated to find a better match (Grandey, 2000), or even burns out before he or she recognizes the poor match.
As stated, emotional labor may impact emotional exhaustion and job satisfaction, which in turn might influence withdrawal behaviors (Hulin, 1990). Some research has supported this notion. For example, Abraham (1999) found emotional dissonance resulted in withdrawal intention indirectly through job dissatisfaction, whereas others have found that emotional exhaustion predicted turnover through job satisfaction (Babakus, Cravens, Johnston, & Moncrief, 1999). Wright and Cropanzano (1998) found that emotional exhaustion led to turnover among 52 social welfare workers. Given the logic just described, we propose a positive relationship between emotional labor and withdrawal.
In an increasingly customer-driven and "branded-service" market, the topic of emotional labor is likely to garner increased attention by academicians and practitioners alike. To date, the conceptual status of the construct is, at best, "muddy." Empirical research based on a variety of differing conceptualizations has contributed to this confusion. Our intent here was to propose a research model that might provide some conceptual utility and ultimately make a contribution to the empirical literature. Specifically, we attempted to make four important contributions to the literature and our understanding of the construct:
Hartel and Zerbe (2000) noted, "No theoretical framework has been put forward that identifies the crucial contextual features of organizations upon which emotional experience varies" (p. 216). The present model may directly address some of their expressed concern. Specifically, disentangling the potential situational demands from the state and behavior should provide more clarity regarding the selection of variables that are true dimensions of the construct from those that exist outside the construct. However, the present model does not in any way attempt to explain the dynamic complexity of emotional labor. In Fig. 11.1 we include a feedback loop between perceived dissonance and emotional labor to highlight the notion that this is, in all likelihood, a highly dynamic process. For example, engaging in deep acting may reduce the amount of perceived dissonance one experiences and therefore in turn may influence one's emotional labor strategy in the future. Thus, although we still believe that a dissonant state must exist for an individual to engage in emotionally laborious acts, we recognize that this relationship may be more complex than the rather linear model we present. However, in order to "see the forest through the trees," we felt some simplification was required to maximize theoretical clarity.
Future Research. Because it was beyond the scope of this theoretical chapter, we have not discussed how to measure perceived dissonance and/or emotional labor. Recently, there have been great advancements made in developing measures (e.g., Tews & Glomb, 2004), which have highlighted both the complexity and challenge of operationalization in this area. The model presented here may be useful in propelling, or at least supplementing, current measurement development efforts in the near future. Regardless of whether the reader agrees with the viability of our perspective, measurement and subsequent empirical work in this area would benefit greatly from clear explication of the theoretical underpinnings of the construct.
Additionally, because our intent was to present a reconceptualization of the emotional labor construct, we by no means claim to have tapped the entire domain of potential factors that surround emotion management in the workplace. Other authors have called for examination of broader issues in relation to emotional labor. For example, Ashforth and Humphrey (1993) suggested a number of emotion variables, on multiple levels, that have been recognized as critical elements in motivation, leadership, and group dynamics that may influence the emotional labor process. In addition, they called for future research that focused on other professional roles and settings beyond service. Grandey (2000) discussed the importance of examining coworker and supervisor support as a moderator of emotional labor. Further, Emotions in the Workplace (2000), edited by Ashkanasy, Hartel, and Zerbe, and Emotion in Organizations (2000, 2nd ed.), edited by Stephen Fineman, cover a broad range of emotion issues in workplace research. Clearly, there are a number of moderators and mediators that would fit nicely with the proposed model. Undoubtedly, there is still a great deal of work to be done in this area. In short, we hope the described
11. PERCEIVE D DISSONANCE AND EMOTIONAL LABOR 211
model realized some of our goals and will provide utility for those involved in emotions in organizations research.
The authors thank Blake Ashforth, Theresa Glomb, and Ron Humphrey for their helpful comments on previous drafts and presentations of this chapter. A version of this chapter was presented at the 17th Annual Conference of the Society for Industrial and Organizational Psychology, Toronto, Canada.
This page intentionally left blank
Joyce E. Bono
Meredith A. Vey
In this chapter, we cumulate the results of empirical research on the predictors and outcomes of emotional labor. After a discussion of the definitions and operationalizations of emotional labor, we use meta-analytic techniques to obtain estimates of the associations between emotional labor and organizational display rules, burnout, physical symptoms, personality, and job characteristics. We discuss these finding and their implications for future research on emotional management in organizations, with a focus on the ubiquity of emotional regulation, the role of personality, and ways that organizations can buffer the negative effects of emotional labor on their employees. We conclude with a brief discussion of the implications of emotional labor research for employee selection, job performance, and organizational success.
A customer service representative who loses his or her temper with a customer would be considered "unprofessional;" as would a funeral director who is perky and bubbly with grieving clients. Behavior in organizations is profoundly influenced by organizational norms and rules. Emotional behavior is no exception. Organizational rules and norms for emotional behavior are communicated to employees through both formal means, such as selection, training, evaluation, and incentive systems, and informal means, such as social influence and pressures. Many organizations encourage employees to exhibit only a narrow range of emotions while at work, such as expressing only cheerfulness when interacting with customers, or suppressing their irritation with a difficult coworker in the name of professionalism. However, as human beings, we can experience a wide range of emotions in a given workday. In order to comply with organizational requirements, many employees must suppress their true emotions or manipulate their emotional expressions. Employees experiencing discordance between felt and required emotions can suppress their genuine emotion, pretend to feel the required emotion (surface acting), or change their emotions to match their organization's display rules (deep acting; Ashforth & Humphrey, 1993; Brotheridge & Lee, 2002; Grandey, 2000). This emotional regulation at work was termed emotional labor by Hochschild (1983). In the 20 years since Hochschild's (1983) study, emotional labor researchers have focused their energy and attention on further defining the emotional labor construct, exploring possible operationalizations of emotional labor, and identifying possible antecedents and outcomes of emotional labor.
This chapter has two main objectives. The first objective is to quantitatively review the emotional labor literature. Toward this end, the history of the emotional labor construct and its definitions and operationalizations is reviewed. Also, relationships between emotional labor and its proposed predictors and outcomes are quantitatively synthesized, using Hunter and Schmidt's (1990) "bare-bones" meta-analytic techniques. After this descriptive and quantitative review of the literature, we turn to the future of emotional labor research. Our second objective is to use what we have learned from the quantitative review to advance emotional labor research. We identify several areas of concern with current emotional labor research and a number of areas ripe with possibilities for future research.
THE EVOLUTION OF EMOTIONAL LABOR
Reichers and Schneider (1990) proposed that scientific constructs evolve following a predictable sequence of three developmental stages: (a) concept introduction/elaboration, (b) concept evaluation/augmentation, and (c) concept consolidation/accommodation. During the initial concept introduction/elaboration stage of construct development, researchers and theorists work toward legitimizing a new construct or a construct recently borrowed from another field. Papers are written to disseminate information about the concept and to present data, with the construct as an independent or dependent variable, as evidence that the construct reflects a legitimate phenomenon. If the concept survives the first stage, the second stage of concept evaluation/augmentation begins. In this state, critiques of early papers and disagreements over how to properly define and operationalized the concept arise. Hopefully, these criticisms and debates lead to augmentation of the concept, through the development of improved measures, the introduction of possible moderating and mediating variables, and the utilization of advanced methodologies. The last stage of construct development, according to Reichers and Schneider, occurs when debate and controversy over the construct has ebbed. By this last stage, one or two key definitions and operationalizations have garnered common accep
12. UNDERSTANDIN G EMOTIONA L MANAGEMEN T AT WOR K 215
tance, and there is consensus as to the predictors and outcomes of the construct. During this stage, comprehensive meta-analyses are conducted and research efforts focused on the concept abate.
The construct of emotional labor is a relatively new one. Introduced by Hochschild (1983) approximately 20 years ago, emotional labor has been the topic of numerous book chapters, journal articles, and conference proceedings. Emotional labor researchers and scholars have focused their energy and attention on further defining the emotional labor construct, by identifying antecedents, emotional regulation strategies, and outcomes of emotional labor. Many of the journal articles and book chapters written on emotional labor have been theoretical in nature, proposing and revising definitions, operationalizations, process models, predictors, and outcomes of emotional labor. To date, fewer than 30 journal articles have been published on the topic, with fewer than half of those articles presenting data.
The construct of emotional labor seems to have survived its infancy, however. The legitimacy of emotional labor as a construct has been enhanced by research on organizational display rules, emotional management at work, and the possible deleterious outcomes of emotional labor. Findings from both qualitative and quantitative studies of employees from bill collection agencies, restaurants, universities, hospitals, airlines, and even Disneyland suggest that some organizations and jobs prohibit the free expression of emotions (Abraham, 1998; Adelmann, 1995; Brotheridge & Grandey, 2002; Cropanzano, Weiss, & Elias, 2004; Diefendorff & Richard, 2003; Morris & Feldman, 1997; Rafaeli & Sutton, 1989; Sutton, 1991; Van Maanen & Kunda, 1989). In some service jobs, employees are expected to express positive emotions to promote goodwill and spending. In contrast, bill collectors may be required to display negative emotions (anger, irritability) in an effort to gain debtor compliance (Sutton, 1991). Ekman (1973b) used the term display rules to refer to social norms regarding the appropriate experience and display of emotions. Perhaps more importantly, research has also revealed that employees do regulate their emotions to conform to organizational display rules by suppressing the emotions they genuinely feel and faking emotions they do not feel (Brotheridge & Lee, 2002; Erickson & Ritter, 2001; Glomb, Miner, & Tews, 2002; Grandey, 2003). Finally, researchers have couched the need to learn more about emotional labor in terms of preventing the possible negative effects of such emotion work. Individuals holding jobs involving emotional labor have reported both work-related stress and emotional exhaustion (Brotheridge & Grandey, 2002; Brotheridge & Lee, 2002; Erickson & Ritter, 2001; Grandey, 2003; Holman, Chissick, & Totterdell, 2002; Kruml & Geddes, 2000; Morris & Feldman, 1997).
Clearly, a number of researchers have deemed emotional labor to be a topic worthy of research and inquiry. Thus, the construct of emotional labor has survived the first stage of concept evolution, the concept introduction/elaboration phase. Few researchers and theorists would argue with the idea that employees in many jobs feel required to regulate their emotional expressions at work. Agreement on how to define and operationalized emotional labor, however, has not been reached. The emotional labor literature is thick with discussion of how to best capture the phenomenon of emotional labor through definition and operationalization. Thus, emotional labor now seems to be struggling through the throes of construct adolescence or concept evaluation/augmentation.
DEFINITIONS AND OPERATIONALIZATIONS
In Table 12.1, we provide a list of the definitions of emotional labor that have emerged over the last decade. Although researchers and theorists are in general agreement as to the importance of increasing our understanding of emotional labor, subtle differences in definitions and operationalizations of emotional labor have led to disagreement and confusion (Cropanzano et al., 2004;
Recent Definitions of Emotional Labor
Hochschild (1983) "The management of feeling to create a publicly observable facial or bodily display; emotional labor is sold for a wage and therefore has exchange value."
James (1989) "The labour involved in dealing with other people's feelings, a core component of which is the regulation of emotions" (p. 15). Ashforth and "The act of displaying the appropriate emotion (i.e., conforming with Humphrey (1993) a display rule" (p. 90). Morris and Feldman "The effort, planning, and control needed to express organizationally (1996) desired emotion during interpersonal transactions" (p. 987).
Pugliesi (1999) "The performance of various forms of emotion work in the context of paid employment. . . . Emotion work (or emotional management) refers to various efforts to manage emotional states and displays"
(p. 126). Zapf, Vogt, Seifert, "The emotional regulation required of the employees in the display of and Isic (1999) organizationally desired emotions" (p. 371).
Grandey (2000) "May involve enhancing, faking, or suppressing emotions to modify the emotional expression ... in response to display rules for the organization or job" (p. 95).
Kruml and Geddes "What employees perform when they are required to feel, or at least (2000) project the appearance of certain emotions in order to produce, for insurance, 'excellent customer service' " (p. 177). Glomb, Miner, and "The effort expended in expressing appropriate emotions and not ex-Tews (2002) pressing inappropriate emotions on the job, as defined by role requirements" (p. 4). Zammuner and "Emotion regulation that occurs within work contexts." Galli (2003) Diefendorff and "The management of emotions as part of the work role." Richard (2003)
12. UNDERSTANDIN G EMOTIONAL MANAGEMENT AT WORK 217
Grandey, 2000). Hochschild (1979, 1983) is credited with first conceptualizing emotional labor as emotional management at work by defining emotional labor as "the management of feeling to create a publicly observable facial and body display" (Hochschild, 1983, p. 7). She went on to note that although emotional management can occur at work or in a private context, emotional labor refers to emotional management that is "sold for a wage" (p. 7). Furthermore, she stated that emotional labor "requires one to induce or suppress feeling in order to sustain the outward countenance that produces the proper state of mind in others." Thus flight attendants must display a calm and friendly demeanor so that airline passengers feel welcome and safe. According to this emotional labor as emotional management viewpoint, employees conform to organizational display rules by manipulating their emotional displays and maybe even emotional states, if their genuine emotions are discordant with organizational display rules. A key element to the emotional labor as emotional management approach is dissonance. In this approach, emotional labor, or the management of emotions, takes place when the individual must express emotions that he or she does not actually feel.
According to Hochschild (1983), employees can manipulate their emotions by utilizing three main emotional management strategies: suppressing genuine emotion, surface acting, and deep acting. The suppression of genuine emotions might involve hiding one's irritation with a customer or coworker. Surface acting involves faking or pretending to feel the emotions required by organizational display rules, and deep acting describes the act of trying to actually change one's emotional state to match the organizationally required expressions. A customer service representative who plasters on a fake smile to comply with the organization's cheerful display rules while experiencing negative emotions on the inside would be surface acting. A customer service representative who actually tries to actually change his or her negative emotional state to a positive one in order to genuinely feel and experience cheerfulness would be deep acting.
Taking a more behavioral approach to emotional labor, Ashforth and Humphreys (1993) defined emotional labor as "the act of displaying the appropriate emotion (i.e., conforming with a display rule") (p. 90). The aim of this approach is to focus on observable behavior, not internal feelings or emotional states. Thus, emotional labor can be operationalized as the expression or display of certain prescribed emotions. For Ashforth and Humphreys, it is the observable emotional behavior in response to organizational display rules, rather than any internal dissonance between experienced and expressed emotion, that is key. The advantage of this approach is its focus on observable behaviors. However, by focusing only on behavior, rather than emotional management, we are left without a theoretical link between emotional labor and proposed outcomes, such as stress or burnout. Why would we expect that smiling would be stressful, unless the smile was not a genuine expression of happiness?
In a third approach to emotional labor, Morris and Feldman (1996) defined emotional labor as "the effort, planning, and control needed to express organizational desired emotion during interpersonal transactions" (p. 987). Their definition is quite similar to Hochschild's (1983), in that both are focused on emotional management and the expression of prescribed emotions. Morris and Feldman expanded Hochschild's definition in two ways. First, they suggested that emotional labor occurs even when employees are asked to display emotions they genuinely feel, as the appropriate expression of the genuinely felt emotion may be prescribed by the organization. For example, happy Waimart employees must be certain to express their happiness with appropriate smiles or greetings, which may require effort for employees who might naturally express their happiness in a more exuberant fashion (e.g., singing or hugging). Morris and Feldman also expanded the emotional labor concept by identifying four dimensions that affect the experience of emotional labor: the frequency of interpersonal interactions, attentiveness (the intensity of emotions expressed and duration of the interactions), variety of emotions required by the interactions, and emotional dissonance (discrepancy between felt and expressed emotions). Thus, the nature of a job, or the nature of the interactions required in a job, plays a role in defining emotional labor. Morris and Feldman's approach contributed to our understanding of emotional labor by pointing out that emotional labor is not a dichotomous variable (an employee is performing emotional labor or not). They focus our attention on the factors that influence the degree or intensity of the emotional labor. Some researchers have taken this perspective to an extreme, by operationalizing emotional labor as job demand or job characteristics (Wharton, 1993), thus confusing the antecedents of emotional labor with emotional labor.
As is clear in Table 12.1 (with the exception of Ashforth & Humphreys, 1993), there is general agreement that emotional labor involves managing emotions at work in order to conform to norms and expectations for emotional displays. In our examination of empirical studies, however, we find less agreement on the appropriate way to measure or operationalize emotional labor. Operationalizations of emotional labor found in the empirical studies identified for this review are presented in Table 12.2. The majority of studies have indeed operationalized emotional labor as emotional management or regulation. These studies explore deep and surface acting, emotional suppression, emotive effort, and dissonance. However, other Operationalizations have been used alone or in combination with emotional management. As seen in Table 12.2, some researchers have operationalized emotional labor as the existence of and/or compliance with display rules. However, the mere existence of display rules does not necessarily mean employees are engaging in emotional labor. Other researchers have operationalized emotional labor or in terms of role requirements, such as Hochschild's classification of jobs or Morris and Feldman's (1996) frequency, intensity, duration, and variety of interactions re
12. UNDERSTANDIN G EMOTIONA L MANAGEMEN T AT WOR K 219
Operationalizations of Emotional Labor
Emotional Labor Operationalization
Emotional Display Rule Existence Role Management and Compliance Requirements
Wharton (1993) X Adelmann (1995) X X X
Morris and Feldman (1997) X X
Abraham (1998) X
Pugliesi (1999) X X X
Zapf et al. (1999) X X X
Kruml and Geddes (2000) X
Schaubroeck and Jones (2000) X
Zerbe (2000) X
Erickson and Ritter (2001) X
Brotheridge and Grandey (2002) X X
Brotheridge and Lee (2002) X X X
Davies and Billings (2002) X X
Glomb, Miner, and Tews (2002) X
Holman et al. (2002) X
Grandey (2001) X
Zammuner and Galli (2004) X
Grandey (2002) X
Glomb and Tews (2004) X X
Note. X indicates which Operationalization of emotional labor was used in the article. Articles are listed in chronological order by publication date.
quired in the job. Although these job characteristics may be important antecedents of emotional labor, they do not define the psychological processes involved in emotional labor. One of the advantages of using meta-analytic techniques in conducting our review is that we will be able to shed light on the issue of whether the way in which emotional labor is operationalized is associated with the results found in a particular study. Thus, in the next section, we summarize what is known about the predictors and outcomes of emotional management at work.
A QUANTITATIVE REVIEW: WHAT DO WE KNOW
ABOUT EMOTIONAL LABOR?
Purpose and Method
The purpose of our quantitative review is not to prematurely rush emotional labor into the final stage of Reichers and Schneider's (1990) model—concept consolidation/accommodation. The topic is not mature enough for a comprehensive meta-analysis to definitively estimate the associations between emotional labor and its antecedents and consequences. Instead, this quantitative review is intended to guide and inspire further emotional labor research through its construct evaluation/augmentation stage, by examining what we know about emotional labor thus far. Through this quantitative review, we can examine which predictors and outcomes seem to have clear relationships with emotional labor, which do not, and which have yet to be empirically investigated. These initial "bare-bones" meta-analyses simply summarize the early research on predictors and outcomes of emotional labor across studies, removing the statistical effects of sampling error. Furthermore, they allow us to compare results across operationalizations.
We located emotional labor studies in three ways; searching PsychInfo, examining references and citations in articles and book chapters, and requesting working papers or those in press from authors who had published empirical work on emotional labor. We identified 18 articles (listed in Table 12.2 in the emotional management column) with data linking the emotional labor with any of its hypothesized predictors and outcomes. Only studies reporting correlations (or the information necessary to calculate correlations) were included in our analyses. Studies that reported only significant correlations were excluded from the analyses, unless all the correlations relevant to our study were significant. If a study included more than one correlation from the same sample (e.g. if a study contained multiple measures of dissonance), the correlations were averaged.
Using Hunter and Schmidt (1990) "bare-bones" meta-analytic techniques, we estimated sample-weighted mean correlations across studies and a standard deviation of those estimates for each of the relationship for which we found at least two empirical studies (or two samples). We also computed 80% credibility intervals. Credibility intervals estimate variability of the individual correlations in the population of studies. Thus, a credibility interval that does not include zero indicates that at least 90% of the correlations reported are greater than zero (or less than zero, in the case of a negative correlation).
In accordance with our review of emotional labor definitions, our analyses are confined to emotional labor operationalizations that conform to the emotional labor as emotional management approach. Four operationalizations met this criterion: emotional dissonance, deep acting, surface acting, and emotional labor performance. Most emotional dissonance measures of emotional labor asked employees to self-report dissonance between their emotional displays and their actual emotional states. For the most part, deep acting scales tap the extent to which individuals attempt to change their internal emotional states to match that of the required emotional display. In contrast, surface acting scales measure the extent to which an individual attempts tofake the expected emotion. Measures of emotional labor performance assess how well individuals regulate their emotions when dealing with customers (Diefendorff & Richard, 2003; Grandey,
12. UNDERSTANDIN G EMOTIONA L MANAGEMEN T AT WORK 221
2002, 2003). These emotional performance ratings consist of supervisor and coworker ratings of employees' emotional regulation when dealing with customers; they are not merely ratings of emotional expression.
Predictors of Emotional Labor
Although researchers have been active in exploring the negative outcomes of emotional labor, somewhat less attention has been paid to the predictors, or antecedents, of such emotional regulation. We found two or more studies that examined the following predictors: organizational display rules, job characteristics (control, autonomy, routinization), social support (coworker and supervisor), and individual differences (gender, positive and negative affectivity, and self-monitoring). When enough studies were available, we examined the associations between emotional labor and predictors (and outcomes), based on how emotional labor was measured (dissonance, deep and surface acting, or emotional performance). It is important to note that although the overall estimates in Tables 12.3 and 12.4 represent data from independent samples (we averaged multiple estimates from a single sample where necessary), the analyses by measurement type (e.g., dissonance or surface acting) are not independent of each other or of the overall estimate (though the data within each analysis does represent data from independent samples). That is, a correlation used in the deep acting analysis will also be used in the overall analysis (labeled
All in Tables 12.3 and 12.4).
The existence and explicitness of organizational rules with respect to emotional behavior (display rules) has been the most frequently explored predictor of emotional labor. Indeed, as shown in Table 12.3, a positive association between emotional labor and display rules was found (rcorr = .15). However, the credibility interval included zero, indicating that in more than 10% of the studies, the association between display rules and emotional labor was negative. Examination of the various emotional labor operationalizations is informative. Positive associations and credibility intervals that did not include zero were found for display rules, deep and surface acting, and emotional performance
( corr -32, .26, and .07, respectively). The association between display rules and emotional dissonance was also positive (rcorr = .12); however, the credibility interval included zero. Thus, employees who perceive explicit display rules are more likely to engage in emotional regulation, via surface and deep acting (though these associations are not large), but do not necessarily experience more emotional dissonance.
When examining job characteristics, we found that when employees perceive autonomy in their jobs, they are less likely to report experiencing emotional dissonance (rcorr = -.20). Similarly, the more routine the job, the greater is the emotional dissonance (rcorr = .16). This may be because individuals who can limit or shorten interactions with customers or coworkers or manage the
Predictors of Emotional Labor
n k corr SDT 80% CV
Emotional dissonance 1,417 5 .12 .14 -.07 to .30 Deep acting 521 3 .32 .07 .23 to .41 Surface acting 521 3 .26 .12 .10 to .41 Emotional performance 437 3 .07 .00 .07 to .07 All 2,090 9 .15 .13 -.02 to .32
Job Control and Autonomy Emotional dissonance 1,764 6 -.20 .06 -.28 to -.12
Job Routineness Emotional dissonance 1,159 3 .16 .00 .16 to .16
Supervisor and Coworker Support Deep acting 583 2 .13 .11 -.02 to .27 Surface acting 583 2 -.14 .03 -.17 to -.10 All 583 2 -.03 .10 -.16 to .10
Positive Affectivity (PA) All 501 2 -.13 .06 -.21 to -.05
Negative Affectivity (NA) All 501 2 .19 .08 .09 to .28
Self-monitoring Deep acting 386 2 .10 .03 .06 to .14 Surface acting 386 2 .26 .00 .26 to .26
Gender Deep acting 501 2 -.06 .00 -.06 to -.06 Surface acting 501 2 -.01 .00 -.01 to -.01 All 501 2 -.01 .00 -.01 to -.01
Note, k, Number of independent samples; rcorr, n-weighted mean correlation; CV, credibility interval calculated using formulas in Whitener (1990). There is no All category for self-monitoring because some nonsignificant correlations between self-monitoring and emotional labor variables in this study (Zammuner & Galli, chap. 15, this book) were not reported. Thus, an overall emotional labor result would be inflated.
timing of these interactions (due to greater job control) may not need to engage in emotional labor as frequently. Employees with a high degree of job control may be able to change tasks or otherwise limit interactions with customers if they are not feeling the required emotion. However, it should be noted that Abraham (1998), in a study that was not included in Table 12.3 results because it did not report zero-order correlations, did find a positive association between job autonomy and emotional dissonance. We cannot tell
12. UNDERSTANDING EMOTIONAL MANAGEMENT AT WORK 223
Outcomes of Emotional Labor
n k SDr 80% CV
Emotional exhaustion Emotional dissonance 2,376 11 .30 .06 .22 to .39 Deep acting 751 3 .14 .09 .03 to .26 Surface acting 955 7 .36 .10 .24 to .49 All 2,948 14 .28 .08 .18 to .38
Depersonalization Emotional dissonance 916 4 .24 .13 .07 to .42 All 1,152 5 .23 .12 .08 to .39
Personal accomplishment Emotional dissonance 916 4 .06 .12 -.09 to .21 All 1,152 5 .02 .12 -.13 to .18
Emotional dissonance 916 4 .36 .00 .36 to .36
Emotional dissonance 2,172 7 -.37 .12 -.52 to -.21 Deep acting 515 2 -.03 .11 -.17 to .11 Surface acting 515 2 -.34 .06 -.41 to -.26 Emotional performance 320 2 .10 .00 .10 to .10 All 2,492 9 -.30 .18 -.53 to -.06
Emotional dissonance 508 3 -.05 .04 -.10 to .00
All 798 2 -.32 .13 -.49 to -.15
Emotional dissonance 408 1 .35 — —
Emotional dissonance 408 1 .35 — —
Emotional dissonance 408 1 .02 — —
Note. k, Number of independent samples; rcorr, n-weighted mean correlation; CV, credibility interval calculated using formulas in Whitener (1990).
whether this is an anomaly of Abraham's study or sample, or whether including the other variables in the regression explains her results.
No clear conclusions can be reached about the relationship between social support and work and emotional labor. In the two studies that reported zero-order correlations between social support and deep and surface acting, we find that employees who report surface acting do not report strong social support. Although the relationship is in the opposite direction for deep acting (those who report deep acting report having strong social support), both associations are quite small and the credibility interval for deep acting includes zero. Abraham (1998) also examined social support as a moderating variable between emotional labor and job satisfaction. Her results are discussed in the outcomes section.
As predictors of emotional labor, individual differences have received less attention than have situational characteristics such as organizational display rules or job characteristics. We found two studies each that examined positive and negative affectivity, self-monitoring, and gender. Positive affectivity was negatively related to emotional labor (rcorr = –.13), and negative affectivity was positively related to emotional labor (rcorr = .19). Self-monitoring was positively associated with both deep and surface acting (rcorr = .10 and .26, respectively). Gender was unrelated to emotional labor. Thus, the few studies located suggest that individual differences do influence emotional labor. More research is needed to shed further light on the nature of the relationship between individual differences and emotional labor.
Outcomes of Emotional Labor
Many emotional labor scholars and researchers have hypothesized that the strain of emotional regulation negatively affects employee physical and psychological well-being (see Grandy, 2000, for a process model). We found at least two empirical studies addressing each of the following outcomes of emotional labor: emotional exhaustion, depersonalization, personal accomplishment (these three are components of Maslach and Jackson's  burnout syndrome), physical complaints (e.g., headaches), role internalization, self-esteem, and job satisfaction. We also found a single study that examined turnover intentions, role conflict, and role ambiguity. In Table 12.4, we present estimates of the relationship between emotional labor and these outcomes. Overall, the associations between emotional labor and its outcomes are somewhat larger and less variable than those found between emotional labor and the predictors.
Results in Table 12.4 support the notion that emotional labor is associated with poor physical and psychological health. With respect to Maslach and Jack-son's (1981) burnout syndrome, we found that irrespective of the operationalization, emotional labor was associated with emotional exhaustion (rcorr = .30) and depersonalization (rcorr = .23) but not personal accomplishment
( corr .02). Employees who reported engaging in emotional labor also reported more physical complaints (rcorr = .36).
With respect to job attitudes, we found a less clear picture. Employees engaged in emotional labor reported lower role internalization (rcorr = –.32), indicating that they were less likely to identify with and be committed to their jobs.
12. UNDERSTANDIN G EMOTIONAL MANAGEMENT AT WORK 225
In general, the association between emotional labor and job satisfaction was also negative (rcorr = –.30 when all operationali2ations were combined). However, the specific operationalization of emotional labor mattered for job satisfaction. Our results indicate that both emotional dissonance and surface acting were associated with decreased job satisfaction. However, for deep acting, the correlation was small (rcorr = –.03) and the credibility interval included zero. Furthermore, both studies that examined emotional performance found a positive link between emotional labor and job satisfaction. However, in this study reports of emotional management were provided by colleagues, who may have focused more on the expressions of emotion than the regulation of emotion. Only one study (Zerbe, 2000) examined the relationship between emotional dissonance and turnover (r = .35), role ambiguity (r = .35), and role conflict (r = .02). More research is needed to clarify the nature of emotional labor's impact on job attitudes.
There were a few additional studies focused on outcomes of emotional labor that are not included in Table 12.4 because they did not report correlations between emotional labor and either a predictor or outcome. These studies linked predictors (display rules) and outcomes (job stress and job satisfaction) of emotional labor, without measuring emotional labor directly. Results of these studies are consistent with a model in which display rules lead to emotional labor (dissonance or acting), which in turn leads to outcomes such as job stress or dissatisfaction. Pugliesi (1999) found that, across multiple measures, display rules were positively associated with both job stress (mean r = .40) and psychological distress (mean r = .37) and negatively associated with job satisfaction (mean r = .28). Schaubroeck and Jones (2000) found that display rules (demands to express positive and suppress negative emotions) were correlated with physical symptoms (mean r = .36). When they examined organizational identification and job involvement, however, they found small positive associations with demands to express positive emotions (r = .14 and .13, respectively) and slightly larger negative associations with demands to suppress negative emotions (r = -.23 for organizational identification and r = –.18 for job involvement).
Wharton (1993) operationalized emotional labor as having one of the jobs Hochschild (1983) identified as involving high emotional labor demands. Therefore, we did not include her data in our Table 12.4 analyses. Using this broad, rough measure of emotional labor, she found that emotional labor was positively associated with job satisfaction but did not find an association between emotional labor jobs and emotional exhaustion. Her results should be interpreted with caution, as it is likely that the emotional labor demands of jobs vary by organization.
Finally, we identified three studies examining outcomes of emotional labor that reported only the results of multiple regressions or path analysis (Abraham, 1998; Adelmann, 1995; Kruml & Geddes, 2000). As they did not report zero-order correlations, these studies were not included in Table 12.4. Moreover, some of their findings are consistent with Table 12.4 results and some are not. For example, Kruml and Geddes (2000) found that dissonance was positively associated with emotional exhaustion and depersonalization but deep acting was negatively associated with these outcomes. However, these results cannot be clearly interpreted as they represent partial correlations. Thus, they represent the association between the variables of interest after controlling for all other variables in the equation. Abraham (1998) reported that dissonance was positively association with emotional exhaustion (consistent with Table 12.4 results). She also found that dissonance was positively associated with job satisfaction, but the relationship was moderated by social support. If social support was high, emotional labor and job satisfaction were positively related. However, when social support was low, emotional labor and job satisfaction were negatively related. Finally, Adelmann (1995) found a positive association between emotional labor and job attitudes (satisfaction and commitment). Again, it is important to keep in mind these are partial correlations and zero-order correlations were not reported in this study. Because of the small number of studies in our analyses, the possibility that the differences we find in the emotional labor-job satisfaction link across studies and samples is as likely to be due to sampling error as they are to be due to true moderating effects of the way emotional labor was measured.
WHAT HAVE WE LEARNED AND HOW CAN
WE USE THIS INFORMATION?
Perhaps the most striking conclusion that can be reached from our analysis is that the variables studied by emotional labor researchers do not have strong associations with perceptions of emotional labor. Indeed the strongest association we found for predictors was r = .32 (display rules and deep acting) and the average correlation between emotional labor and the predictors was r = .13. The results for outcomes are a bit stronger with a high of r = .36 (emotional dissonance and physical complaints) and an average of r = .22. Nonetheless, the combined results of these empirical studies can serve as a guide for future research.
With respect to the predictors of emotional labor, the most obvious hypotheses (and the most empirically tested predictor hypothesis) is that when organizations demand that employees express certain emotions and suppress others, employees will either change their emotions, suppress emotions they genuinely experience, or fake emotions they do not feel. Hence, we find it somewhat surprising that the relationship between organizational display rules and reports of emotional labor is modest at best (rcorr = .32 and .26 for
12. UNDERSTANDIN G EMOTIONA L MANAGEMEN T AT WOR K 227
deep and surface acting, respectively,rcorr = .12 for emotional dissonance, and rcorr = .15 for all measures of emotional labor combined). The whole premise underlying theory and research on emotional labor is that when organizations demand certain emotions, employees perform emotional labor, which may be damaging to physiological and psychological health. Yet we do not find a strong association between organizational demands and employees' perceptions of emotional labor.
There are several possible explanations for this finding, including poor measurement of these variables. It would be plausible to assess organizational display rules by asking a representative of the employees' organization (e.g., human resources [HR] professional) about formal guidelines for emotional expression in order to obtain more "objective" measurement. However, display rules (or social norms for emotional expression) may exist completely independent of such formal guidelines. Thus, is seems appropriate to use self-reports of display rules and emotional labor. Nonetheless, we encourage research focused on the construct validity of both display rules and emotional labor.
If measurement is not the problem, do our findings imply that employees rarely accept or comply with emotional demands made by organizations? Alternatively, do our results suggest that emotional management at work is widespread and ubiquitous and thus occurs in the presence and in the absence of organizational display rules? We suspect the latter may be true. Goffman (1959), in his now classic book on the presentation of the self, argued that individuals routinely adapt their self-presentations (including emotional displays) to the norms, expectations, and demands of the social situations they encounter. Providing support for this notion is research by Gross, Feldman-Barrett, and Richards (1998, as cited in Gross, 1998b), who reported that 9 of 10 undergraduate students reported altering their emotions about once a day (and could provide examples of when they did so). Moreover, Tomkins (1984) reported that adult emotions are almost always regulated. Indeed, although emotional labor researchers rarely link their work to the impression management or social influence literature, we know that individuals in organizations regularly and voluntarily use impression management techniques (such as ingratiation, which includes the faking and suppression of emotions) to obtain jobs, raises, promotions, and better performance evaluations. A recent study (Higgins, Judge, & Ferris, 2003) documented the success of these efforts, especially during employment interviews. Moreover, in an experience sampling study, in which they assessed emotional labor several times a day for about 2 weeks, Glomb et al. (2002) found a general trend in the aggregate data toward emotional suppression at work. That is, employees tended to report feeling emotions more often than those emotions were expressed. Thus, it may be that emotional regulation at work is a simple, routine reality of work life.