DEVELOPMENT AS A SIOCIAL PROCESS as a Siocial Process

CULTURAL DYNAMICS OF SOCIAL REPRESENTATION

Development as a Social Process Contributions of Gerard Duveen Edited by Serge Moscovici, Sandra Jovchelovitch and Brady Wagoner

Development as a Social Process

This volume discusses the interface between human development and socio-cultural processes by exploring the writings of Gerard Duveen, an internationally renowned figure, whose untimely death left a void in the fields of socio-developmental psychology, cultural psychology and research into social representations. Duveen’s original and comprehensive approach continues to offer fresh insight into core theoretical, methodological and empirical problems in contemporary psychology. In this collection the editors have carefully selected Duveen’s most significant papers to demonstrate the innovative nature of his contribution to developmental, social and cultural psychology. Divided into three sections, the book includes:

• • •

Duveen’s engagement with Jean Piaget the role of social life in human development and the making of cognition social representations and social identities

Introduced with chapters from Serge Moscovici, Sandra Jovchelovitch and Brady Wagoner, this book presents previously unpublished papers, as well as chapters available here in English for the first time. It will be essential reading for those studying high level developmental psychology, educational psychology, social psychology and cultural psychology. Serge Moscovici is Professor of Social Psychology at the School for Advanced Studies in the Social Sciences (EHESS), Paris, and founder of the European Laboratory of Social Psychology at the Maison de Sciences de l’Homme, Paris, France. Sandra Jovchelovitch is Professor of Social Psychology at the London School of Economics, UK, where she directs the Masters programme in Social and Cultural Psychology. Brady Wagoner is Associate Professor at Aalborg University, Denmark. He has received a number of prestigious academic awards, including the Sigmund Koch Award, Gates Cambridge Scholarship and the Jefferson Prize.

The series Cultural Dynamics of Social Representation is dedicated to bringing the scholarly reader new ways of representing human lives in the contemporary social sciences. It is a part of a new direction – cultural psychology — that has emerged at the intersection of developmental, dynamic and social psychologies, anthropology, education, and sociology. It aims to provide cutting-edge examinations of global social processes, which for every country are becoming increasingly multi-cultural; the world is becoming one ‘global village’, with the corresponding need to know how different parts of that ‘village’ function. Therefore, social sciences need new ways of considering how to study human lives in their globalizing contexts. The focus of this series is the social representation of people, communities, and – last but not least — the social sciences themselves. In this series Symbolic Transformation: The Mind in Movement through Culture and Society Edited by Brady Wagoner Trust and Conflict: Representation, Culture and Dialogue Edited by Ivana Marková and Alex Gillespie Social Representations in the ‘Social Arena’ Edited by Annamaria Silvana de Rosa Qualitative Mathematics for the Social Sciences: Mathematical Models for Research on Cultural Dynamics Edited by Lee Rudolph Development as a Social Process: Contributions of Gerard Duveen Edited by Serge Moscovici, Sandra Jovchelovitch and Brady Wagoner

Development as a Social Process Contributions of Gerard Duveen

Edited by Serge Moscovici, Sandra Jovchelovitch and Brady Wagoner

“Through a carefully selected set of essays, the editors have created a marvelous Symposium, conducted by the late Gerald Duveen, where he, Piaget, Moscovici, Vygotsky and Bartlett explore the need for a synthetic approach to the nature of human development. A challenging and rewarding reading experience that taught me a lot.” Professor Michael Cole, Director of Laboratory for Comparative Human Cognition, University of California San Diego, USA “Gerard Duveen was a remarkable scholar who developed an original conceptual approach linking together developmental, cultural and social psychology. In this volume the editors present Duveen’s critical engagement with Piaget’s and Moscovici’s theories and his original thought in advancing difficult concepts like decentration, social representations, identities, beliefs and doubts, among many others. The volume, building on intellectual scholarship of the highest standard, will be inspirational for researchers in human and social sciences.” Ivana Markova, Emeritus Professor in Psychology, University of Stirling, UK

First published 2013 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2013 S. Moscovici, S. Jovchelovitch and B. Wagoner The right of the editors to be identified as the authors of the editorial material, and of the author for his individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Development as a social process : contributions of Gerard Duveen / [edited by] Serge Moscovici, Sandra Jovchelovitch, Brady Wagoner. p. cm.—(Cultural dynamics of social representation) 1. Developmental psychology—Social aspects. 2. Social psychology. 3. Social representations. 4. Duveen, Gerard. I. Moscovici, Serge. II. Jovchelovitch, Sandra. III. Wagoner, Brady, 1980– BF713.D457 2013 155—dc23 2012038348 ISBN: 978–0–415–63459–5 (hbk) ISBN: 978–0–203–38797–9 (ebk) Typeset in Times New Roman by RefineCatch Limited, Bungay, Suffolk

Contents

Series editor’s foreword

ix

JAAN VALSINER

Preface Acknowledgements Introduction: The context and development of ideas

xi xiii xv

SANDRA JOVCHELOVITCH AND BRADY WAGONER

PART I

Piaget: A view from afar 1 Children’s understanding of friendship (1984)

1 3

2 The child’s re-construction of economics (1988)

27

3 Piaget ethnographer (2000)

42

4 Genesis and structure: Piaget and Moscovici (2001)

56

PART II

Development as decentration

65

5 Social life and the epistemic subject (1984)

67

6 Psychological development as a social process (1997)

90

7 Construction, belief, doubt (2002)

112

8 On interviews: A conversation with Carol Gilligian (2005)

124

9 The constructive role of asymmetry in social interaction (2008) – with Charis Psaltis

133

viii Contents PART III

Thinking through social representations

155

10 The significance of social identities (1986) – with Barbara Lloyd

157

11 Social representations as a genetic theory (1990) – with Barbara Lloyd

173

12 Representations, identities, resistance (2001)

182

13 Culture and social representations (2007)

196

14 Social actors and social groups: a return to heterogeneity in social psychology (2008)

217

Bibliography: The published papers of Gerard Duveen Index

222 227

Series editor’s foreword The idea that will live: Genetic Social Psychology

This book is an appropriate tribute to a gentle and humble scholar who slowly but systematically synthesised relevant ideas at the intersection of developmental and social psychologies, and whose life ended all too early. It is also a tribute to the kind of scholarship that is needed in psychology – thorough, systematic, and deep in theoretical insights. Psychology is in deep crisis in our days precisely because of its success of amassing large quantities of empirical evidence – rarely addressing the question what for? What is the value of such accumulation of empirical evidence for the generalising power of science? The life work by Gerard Duveen that is collected in this book gives a good illustration of what kind of scholarship could bring psychology out of its crisis of limited generalisation value. Gerard Duveen was a true scholar, and a trustable friend. As the contributions to this book amply demonstrate, he was impervious to the superficial fashions for one or another kind of fancy psychological label. Such fashions capture psychology at recent times – each one comes, gains popularity – only to be replaced by another similar fetish. This was not the world according to Gerard. For him, serious theory was worth deep analysis and constructive critique. Jean Piaget, Lev Vygotsky, Serge Moscovici and other thinkers who have made major contributions were the targets of his careful inquiries into their ideas. The introduction to this book by its editors illustrates beautifully how Gerard’s intellectual oeuvre was created – including the story of his insistence of being labelled – in the 21st century – a genetic social psychologist. The confusion it created and Gerard’s insistence upon that label needs to be viewed from a standpoint that focuses on the historical myopia of contemporary psychology. The notion of genetic as used by the predecessor of Jean Piaget and Lev Vygotsky – James Mark Baldwin – was a marker of thoroughly developmental theoretical quest (Valsiner, 2010). The term carries with it the heritage of serious theoretical inquiries of over a century. Gerard Duveen’s insistence on his own identity in the historical environment of Corpus Christi College deserves itself to be seen as a personal act to emphasise the historicity of our social sciences. Gerard was my co-creator of this book series – Cultural Dynamics of Social Representation – as we designed it together during my first stay in Cambridge in 2005. His interest in the theory of social representation was the key for his uniting

x

Series editor’s foreword

of social and developmental perspectives. We both felt that the focus on culture and the dynamic side of social representation needed a public arena for scholars to have freedom to express their new ideas. So the series was born. Even as his own commitments later did not allow him to participate in the editorial busywork of the series, his and his students’ support for the series was substantial for keeping the series developing well. The appearance of this memorial volume to Gerard Duveen is a relevant milestone for the series, and constitutes a concentrated presentation of the ideas of one of the most conscientious thinkers of the recent decades at the intersection of social and developmental psychologies. The message of genetic social psychology in this volume should live on, and transform the research traditions in the field of social representation. Jaan Valsiner São Paulo August, 2012

Reference Valsiner, J. (2010). A persistent innovator: James Mark Baldwin reconsidered. Introduction to J. M. Baldwin, Genetic theory of reality (pp. xv–lix). New Brunswick, N.J.: Transaction Publishers.

Preface Serge Moscovici

There is in every life something surprising. With regard to Gerard, it was his ability to work in and between different academic genres. Gerard was both an amazingly erudite theoretician and brilliant empirical researcher. As a theoretician, he worked with great care on some of the major unresolved theoretical problems of psychology with precise concepts and novel interpretations. As an empirical researcher, he conducted fascinating experiments on children inspired by Piaget’s work and ethnographies on classroom interaction. He could creatively handle theoretical, experimental and ethnographic genres, constructively developing each in dialogue with the others. Given that this ability to work in and between genres is indeed exceptionally rare in contemporary academia, it is worth reflecting on what personal characteristics made this possible. In Gerard’s case perhaps it was his ability to stand at a distance from a specialist role and thereby let himself become attracted and puzzled by the richness of social phenomena. He was a colleague of tremendous intellectual quality, always capable of expressing new views on social phenomena. Just as Piaget, Gerard was an authentic creator, a sensitive observer of social and psychological life and not only as a psychologist. His passion for the arts – in particular music and literature – framed the way in which he saw the world. Perhaps this gave him the tremendous intellectual quality of always being capable of expressing new interpretations, which combined with an immense intellectual curiosity and generosity to others. Gerard was attached to the theory of social representations in which he played a central role. His reading of Piaget and his admiration for Piaget’s expanded theoretical horizons guided his engagement with the field. He understood that Piaget had uncovered the child as an epistemic subject but that more was needed if we were to fully capture the child as a social actor. His work demonstrated that the epistemic life of the child evolves in a world of social representations, which circumscribe the development of cognition. He showed that the child is not only an epistemic but also a social and psychological subject: the knowledge of the child develops intertwined with the knowledge of society. And let us not forget his intellectual generosity. Gerard gave much to students and colleagues because he was an authentic listener. He taught in the Socratic tradition of sustained dialogue and was able to spend hours talking and reflecting

xii Preface on the concerns and preoccupations of those around him. His qualities as a person and as a scholar will live on in his writings and in the trajectories of so many students and friends whose lives he touched and inspired.

Acknowledgements

In putting together this collection of Gerard Duveen’s writings and working through the final structure of this book it became clear again how much Gerard’s scholarship was firmly grounded in the development of ideas as a social encounter. This is expressed in the support we received from colleagues and students of Gerard who worked with us to make this book happen. Steve Gaskell at the Psychology Workshop of the Institute of Social Psychology, LSE, carefully digitalised Gerard’s PhD and MSc Theses as well as papers that were unpublished or without an electronic format. Jacqueline Priego-Hernandez’s determination and energy guaranteed that all of Gerard’s publications were successfully brought together. Ivana Marková provided advice and constant support. Steffen Ernoe transcribed the ‘conversation with Carol Gilligian’ and assembled the book’s references. We are grateful to all for their invaluable help. Additionally, we would like to thank a number of publishers for allowing us to reproduce papers for this volume. The following is a list of papers reproduced in part or whole here: Duveen, G. (1984). From Social Cognition to the Cognition of Social Life: an essay in decentration. Unpublished PhD Thesis. University of Sussex. Duveen, G. (1988). The Child’s Re-construction of Economic Life. In K. Ekberg and P. E. Mjaavatn (Eds) Growing into a Modern World (pp. 177–199). Centre for Child Research, University of Trondheim, Norway. Duveen, G. (1997). Psychological Development as a Social Process. In L. Smith, J. Dockrell and P. Tomlinson (Eds) Piaget, Vygotsky and Beyond (pp. 67–90). London: Routledge. Duveen, G. (2000). Piaget Ethnographer. Social Science Information, 39, 79–97. Duveen, G. (2001). Representations, identities, resistance. In K. Deaux and G. Philogène (Eds) Representations of the Social (pp. 257–270). Oxford: Blackwell. Duveen, G. (2001). Genesis and structure: Piaget and Moscovici. In F. Buschini and N. Kalampalikis (Eds) Penser la vie, le social, la nature: Mélange en l’honneur de Serge Moscovici (pp. 163–173). Paris: Editions de la Maison des Sciences de l’Homme. Duveen, G. (2002). Construction, Belief, Doubt. Psychologie et Societé, 3, 139–155. Duveen, G. (2007). Culture and Social Representations. In J. Valsiner and A. Rosa (Eds) Cambridge Handbook of Socio-Cultural Psychology (pp. 543–559). Cambridge: Cambridge University Press.

xiv Acknowledgements Duveen, G. (2008). Social Actors and Social Groups: A return to heterogeneity in social psychology. Journal for the Theory of Social Behaviour, 38(4): 370–374. Duveen, G. and Lloyd, B. (1986). The Significance of Social Identities. British Journal of Social Psychology, 25, 219–230. Duveen, G. and Lloyd, B. (1990). Social Representations as a Genetic Theory. In Social Representations and the Development of Knowledge (pp. 1–10). Cambridge: Cambridge University Press. Duveen, G. and Psaltis, C. (2008). The constructive role of asymmetry in social interaction. In U. Mueller, J. I. M. Carpendale, N. Budwig & B. Sokol (Eds), Social life and social knowledge: Toward a process account of development. Mahwah, NJ: Lawrence Erlbaum.

Introduction The context and development of ideas Sandra Jovchelovitch and Brady Wagoner

The theoretical problematic In his doctoral thesis Gerard Duveen discussed extensively the relations between the epistemic subject and social life (Duveen, 1984). How is the subject of knowledge at one and the same time the subject of society? This was a problem that he pursued unrelentingly in his writing, teaching and scholarly conversations and, as we recently discovered, can be found at the heart of his early work. In a detailed and highly theoretically driven empirical exploration of children’s concepts of friendship Duveen demonstrates that in human development the separation between knowledge and everyday experience does not hold. He writes: . . . objects are encountered first of all outside the theoretical structures of psychology; they exist primarily as human reality, as aspects of being produced and maintained through human action. In this sense they are not concepts produced through the operation of psychological theories but rather are created within the context of the lived experience of everyday life; they are encountered first of all within the life space of individuals (the lebenswelt or what Sartre has termed le vecu). Indeed they are so closely interwoven into the texture of this context that in thinking about them it is difficult to detach them from this totality; they are implicated at every level: the conceptual, the affective and the practical (Duveen, 1984: 224) Duveen shows that the development of cognition is inseparable from the emotional structures of psychological development and the material structures of social, economic and cultural worlds: the purely epistemic subject stands as a fiction. He understood the inter-relations between these different domains. His contribution was focused in demonstrating that as children develop and construct knowledge they also grow into competent social actors. Human development is social and psychological; take one at the expense of the other and you have only half of the story and an incomplete theory. And as the good Piagetian that he was, Gerard knew that this development is also articulated with the biological maturation of

xvi Introduction the child, which intersects in complex and absolutely necessary ways with the development and consolidation of cognition (Duveen, 1997). In a collaboration with Psaltis and Perret-Clermont posthumously published (an apt legacy, in a paper written both with a former student and a long-standing colleague who provided important intellectual references) Gerard Duveen again engages in this project. With his co-authors, he re-examines the Piagetian effort of integrating the psychological and social domains, assesses the research programme that followed in its wake and considers how Piaget and Vygotsky could be fruitfully articulated (Psaltis, Duveen and Perret-Clermont, 2009). Developed in this final paper is the idea of operativity-in-context, which condenses many years of experimental and ethnographic research in Geneva, Sussex and Cambridge on how the construction of the mental is not only permeated by social interaction but dependent on it. As they show, it is only in context, never without it or out of it, that the psychological emerges. The paper also provides a sobering statement about how misleading the usual portrait of Piaget’s developmental theory as asocial is. As Nicolopoulou and Weintraub (2009) point out in their pertinent comments on this provoking paper, even a casual inspection of Piaget’s essays and books is sufficient to demonstrate that the social was not only a preoccupation but also a fundamental component of Piagetian psychology, with the essays collected in Sociological Studies (1965/1995) offering the most clear formulation. So from very early on in his academic writings to the very end of his life Duveen sustained a critical engagement with Piagetian psychology based on a dual commitment to think with and against Piaget (Duveen, 2000a; 2001b). Duveen never shed the inspiration and influence that Piaget exerted on him and never ceased to be critical of Piaget. The problem was not that Piaget was asocial, but how Piaget treated the social and how the excessive rationalism of his universal sequence of development obfuscated the role of contextual determinants, not to mention the role of context on Piaget’s own views. To address this problem Duveen turned to the possibility of integrating social and developmental psychology, something that he cherished throughout his career and that is fully expressed in his work on development (Duveen and Lloyd, 1990; Duveen, 1985, 1988, 1996; Lloyd and Duveen, 1992). As he wrote many times, children develop in a world of preceding social representations and deal with self-development amidst the tensions and the potentials of existing social identities. Both phenomena are part of the contextual matrix of development and need to be taken into account if we are to fully understand how a human infant becomes a person. As an ambivalent Piagetian Duveen found in the social psychology of Serge Moscovici, himself an admirer and a critic of Piaget, the concepts that helped him to make the jump from social cognition to the cognition of the social. Social representations became central for answering the fracture between the epistemic and the social psychological subject identified in Piaget. Carefully retaining what was important to retain, he went beyond Piaget to show that to know an object involves knowing an object that is always-already known and signified by a socio-cultural context, made relevant or discarded by a community, invested with values and larger social representations that frame what is knowable and thinkable in a given time and place.

Introduction xvii Social influence and social representations, as concepts and as phenomena, allowed him to explain the ways in which children come to understand society and in this process understand themselves and the object-world. The studies on children’s understanding of kinship and friendship, his later and widely recognised contribution with Barbara Lloyd on children’s understanding of gender, and our more recent project on how children in different social and cultural milieus construct public spheres, all tried to show that the construction of cognition by individual minds is always and at the same time a co-construction with other minds, which collectively, in social institutions, in political systems, in historical trajectories, in cultural communities, define at given times and contexts the boundaries of what is known. His ambivalent relation to Piaget also transpires in the manner he understood agency. As a Piagetian, Duveen knew that the agency of the child is crucial, indeed the child must see and believe herself an absolute innovator, a creator and constructor of all that is there waiting for her to remake. But in line with his teacher and early supervisor, Hans Furth, Duveen problematised this agency by introducing the dynamic of desire in the constitution of the epistemic subject. To know-my-object is also to want-my-object. This insight probed a long-term commitment to the deep psychology of Freud, which was never too far from Duveen’s thinking and way of seeing the world. In writing about the development of representation and culture in the pretend-play of young children he goes back to the links between knowledge and desire to show how the drive that moves the child to create and construct the world as innovator is not far from libidinal organisation (Duveen, 2000b). We discussed this problem extensively when the first author was formulating Knowledge in Context (2007) and it was from his insights on the ontogenesis of knowledge that she developed the idea of exploring what a system of knowledge wants to represent. It was awkward perhaps to pose this question to knowledge, but he helped to pursue the idea of seeing epistemic constructions as ontological constructions, ways of reading and investing the world, permanently traversed by the deep psychology of self-other relations. Knowledge statements are never too far away from statements about being; our epistemic constructions are inseparable from the multitude of voices with which we engage as we develop a sense of self and knowledge of the object-world. To know the world is to be in the world for the world. This, we concurred, was the radical insight of Moscovici (2008), Jodelet (1991) and Marková (2003), in one way or another offered by most sociocultural traditions in psychology: the psychology of cognitive development must deal with the fact that cognition is hot from the start and knowledge emerges and evolves intertwined with the deep psychology of self-other relations. What a delight and what a privilege it was to engage with such an interlocutor, benefit from his vast knowledge and erudition and listen to his examples flowing from fiction, music and developmental psychology alike. Decentration is yet another concept in which Gerard’s ambivalent engagement with Piaget becomes evident. His thesis sub-title was ‘An essay in decentration’. The notion of decentration has not found too strong a space in the prevailing impetus of modern psychology. While the Vygotskian concept of mediation eventually found a firm position in the theoretical landscape of the discipline, decentration,

xviii Introduction which is not too far away from mediation, remained slightly marginal and to some extent unknown. But Gerard knew that decentration is paramount to a societal psychology, indeed to all human psychology, for it refers back to what Rob Farr (1996) described as the inner core of all major psychological systems of a social kind: the relational and dialogical processes that allow the human self to know and understand itself by stepping out of itself and moving towards other people in action and cognition. George Herbert Mead’s theory is the locus classicus in the description of these processes with regards to the genesis of selves and societies; Piaget called them a mini Copernican Revolution; Winnicott referred to them as transitional phenomena; Bartlett (1932) referred to this process as ‘turning around on [one’s] own schema’ (p. 206); and more recently the term intersubjectivity and decentration is to be found even in theory of mind. Gerard showed throughout his work that this basic ontogenetic process, at the level of the developing self and the emerging epistemic capacities of the child, went hand in hand with societal structures. Larger histories are always dialectically related to micro, singular stories. His long-term thinking on human development, social identities and cultural representations helped us to see that, contrary to the illusions of much of psychology, human beings are ‘off-centre’, depending and relying on others, requiring connection and affiliation to produce themselves as social and psychological beings. In Gerard’s account of decentration we find inspiration to understand the primacy of intersubjectivity and dialogical communication in the making of human life as a distinctive form of life. It is this process that accounts for the genesis of the human mind as well as for the genesis of societies and cultures. Finally as a social psychologist Duveen knew that the social environment can be unforgiving. He understood the power of the social, the Durkheimian idea that the social is a fact for the psychic subject. As beginners we come into a social world that is already structured by social representations, power relations and socio-economic circumstances, and through processes of social influence this social world is ready to structure us, to make us into what we are. This tension beautifully comes to light in Gerard’s thesis and was a recurrent theme of his later work and his interest on how social influence shapes the knowing competences and the identity of the developing child (Duveen, 2001a, 2002, 2007). Piaget and Moscovici met again in the way Gerard theorised conformity and innovation in human development: while conformity to the mainstream collective is a powerful tendency of all human societies, transgression and minority influence are the counterpart, a battle which reminds us of the potentials embedded in the role of innovators and minorities, of individuals who are able to step back and challenge the chains of culture and what they impose on our ways of seeing the world and ourselves. This, Gerard taught us, should not be ‘disregarded as banal individualism because underneath it there is a struggle and a desire to see things in a new light, to discover and to illuminate parts of the world that get hidden by the blindness produced by the assumptions we take for granted and the cultural traditions we inherit’ (Jovchelovitch, 2007, p. 44). His writings on the development of representations of gender captured this permanent tension between tradition and innovation, conformity and rebellion, acceptance and resistance.

Introduction xix

Gerard’s life and personality In reading Gerard’s thesis we see again all that always inspired and impressed us in his scholarship as well as many of the issues discussed and reflected upon in our collaborative work over the years. There it is already clear that in all of his many dimensions Gerard Duveen was not a typical psychologist. His scholarship had a scope and sweep that an age of small specialisations can now only dream of recapturing; he thought and wrote about human development and social representations drawing on large vistas that came from psychology but also from his engagement with the continental social philosophy of the second half of the 20th century, in particular Sartre, Goldman and Wittgenstein. Gerard was a refined, outstanding thinker; a selective and profound writer – who always chose quality over quantity. Music was a great passion and he never completely gave up the idea that there is something superior in its language when it came to the hard task of understanding our human existence and the tribulations of our psychological lives. Very few people will have known Piaget more than he did and in reading his thesis we understand better how far and how deep Gerard’s relation to Piaget’s psychology went. Out of this deep engagement came the insight that epistemic structures alone cannot account for the overall process of cognition: a full account also needs to consider the societal and psychic processes that are essential to its formation. But perhaps even more important than these insights are the lessons he left in the manner of his critical engagement. All theoretical traditions contain a certain accumulated level of wisdom and knowledge, but as with all human knowledge they are brittle and serve us best as platforms that allow us to see further and hopefully jump to a different level. They are not dogmas to be defended or rejected but ideas, clusters of evidence, potential spaces from where we draw the conditions of our understanding today and, perhaps more importantly, the conditions for understanding the limitations of our understandings. From Gerard we learn the lesson to think with and against the intellectual traditions you cherish. In the engagement itself we find that very Gerardian skill of being able to produce a stance that mixed criticality and generosity, a human and professional talent that he showed as a friend, as a teacher and as a scholar. The second author recalls Gerard’s unique ability in conversation to turn polemical statements of his interlocutor into a form that could be constructively worked on. In his classes, it was not uncommon to hear him respond to immobile positions ‘yes and no . . .’ so as to make them more pliable. All the above comes together in Gerard’s position as a genetic social psychologist. When Gerard was made a Reader by the University of Cambridge, he decided to be a Reader in Genetic Social Psychology, something that gave him an almost child-like satisfaction. The first author had written about this and suggested to him that the word ‘genetic’ in Anglo-Saxon psychology has long lost its connection with the Piagetian and Vygotskyan vision of understanding the making of psychological structures through their genesis and developmental history. In an age ironically dominated by fascination with all things biological he might have ended up being seen as a psychologist of DNA! We laughed. But Gerard insisted in declaring

xx

Introduction

himself a genetic social psychologist emphasising again and again the problem of genesis and development. Development for him went far beyond the development of the child. The development of the child was one particular case of a much larger process of structuration and transformation that needed to be apprehended and explained at multiple levels. He was never too far away from the Piagetian conception that structures are systems of transformation; to see it so you need the right lenses and his lenses were the genetic method. In this sense Gerard was a historian of human psychological development, and studied a variety of its empirical instantiations from the development of the child, to the development of cultural representations, to the development of ideas, groups and social influence. Soon after Gerard’s death the first author and Marie-Claude Gervais sat to mourn our beloved friend and to write about his legacy. Perhaps out of an unconscious loyalty to Gerard the text never materialised. But as we used to do while with him, we talked and talked. We talked about his being an inspiring and influential teacher who always had time for listening and for continuing a conversation, who had taught in the great tradition of Socrates’ maieutics, a dialogical exercise between teacher and pupil for generating knowledge and understanding. We talked about his being one of the last of the flaneurs, always ready to perambulate and to walk around the city, with a book and a newspaper under his arm, order a double espresso and without any hurry light up his Gauloise without filter. We talked about his passion for music and for the arts, his love of Berg and Caetano Veloso, of Bach and Maria Betânia, of the angels that were always with him. Gerard had a self-effacing quality that stayed with him to the end of his life. And yet he was widely recognised by his friends and colleagues across the world as the genuine article. His early departure leaves us with a tremendous sense of loss, not only because his untimely death has deprived us of many more important scholarly contributions but also because few people could radiate such warmth and humanity. He was a wonderful friend, the best and most generous of colleagues, an inspirational voice. It is perhaps not accidental that one of the last papers Gerard wrote explored the diversity of relational bonds and communicative genres that bring people together (Duveen, 2008). He started this paper just before he learned that he was very ill and managed to conclude it shortly before his death. It is a brief paper that while clearly unfinished retains all the usual qualities of Gerard’s scholarship. There he proposes solidarity, sympathy and communion as distinctive relational forms. Gerard had always known about human solidarity and sympathy, but it was in illness that he experienced communion. How moved and how surprised he was when he realised how much he was loved by the friends, family, colleagues, girlfriends and students who packed the hospital wards and who flew from all over Europe to attend the party held by his college in the summer of 2008, and who finally came to the chapel of Corpus Christi college where his funeral took place in November of the same year. His illness and eminent death made him ever so more aware of the relations that bound him to others and in his death, as so many times before in his life, Gerard was again a teacher to us all. His lessons and his influence on us will live on.

Introduction xxi

A preview of this book This book aims to bring together the seminal work of Gerard Duveen into a coherent structure, centred on understanding the interface between psychological development and social-cultural processes. The unique way Duveen made links between developmental, social and cultural psychology offers fresh insights to thinkers in each of these sub-disciplines. These insights provide novel strategies for approaching core theoretical, methodological and empirical problems faced by contemporary psychology. The book contains publications on a range of topics from throughout Duveen’s career, together with previously unpublished material and writings formerly only available in French. These papers are organised into three parts, each highlighting a primary concern of his scholarship, namely (1) a critical engagement with Piaget’s legacy, (2) conceptualising psychological development in terms of decentration, and (3) situating psychological processes within society using the concepts of Social Identities and Social Representations. Part I. ‘Piaget: A View from Afar’ provides a selection of chapters that express the depth of Duveen’s engagement with Piaget and the solutions he offered to limitations discovered in Piaget’s work. Piaget remains a figure of central importance to the discipline of psychology but precisely what aspects of his work are important to remember is still an open question. Duveen’s knowledge of Piaget was extensive and he was perhaps one of the few who was able to fully situate Piaget’s work in the intellectual landscape of both developmental and social psychology as well as sociology. The chapters here include empirical studies, methodological discussions and theoretical contributions casting light on thorny issues of Piaget’s work, in particular the role of context and culture in the development of cognitive structures and the child’s understanding of the world. As we discussed above, Duveen thought with and against Piaget and the manner of his critical engagement continues to offer lessons about how to critically examine and work with key intellectual sources. The first two chapters in this section are studies of children’s representations of friendship and economic life, using methods such as interview and ethnography. The third chapter traces methods developed by Piaget over his long career, and in so doing makes an argument for characterising Piaget as a kind of ethnographer. Lastly, the final paper in this section explores theoretical commonalities between Piaget and Moscovici, with regard to their use of the concepts ‘genesis’ and ‘structure’. In Part II ‘Development as Decentration’, Duveen provides one of the most powerful elaborations on Piaget’s notion of ‘decentration’. As mentioned earlier, decentration was Piaget’s name for the process of knowing by which the self steps outside of itself towards the perspective of others. Thus, the concept explicitly links cognition to social interaction. This is aptly illustrated in Piaget and Inhelder’s (1948/1956) ‘three mountain task’ whereby the developing child must reconstruct the mountain scene from the perspective of others (dolls). Central to Duveen’s work was the role of social life in human development and the making of cognition. Chapters in this section discuss at length these processes, illustrating how Duveen’s critical engagement with Piaget developed into a research

xxii Introduction programme that systematically introduced social relationships and culture into the study of children’s development. The first paper, taken from a chapter of Duveen’s PhD thesis, is a theoretical reflection on the relationship between the social and epistemic subject. This is followed by a chapter that critically explores the contributions of Piaget, Vygotsky and Moscovici to an understanding of how development is a social process. The proceeding chapter then analyses the relationship between belief and knowledge and the conditions under which doubt arises. Chapter 9 returns to the question of research methodology by considering interviews as a forum whereby a researcher can decentre their own theoretical position. The final chapter of the section summarises a programme of research on sociocognitive conflict that Duveen had been involved in since the late 1990s. Part III ‘Thinking through Social Representations’ encompasses cultural and social psychological papers that grapple with the relations between the sociogenesis, ontogenesis and microgenesis of thought. Duveen teaches us that developmental psychology can only artificially separate itself from social psychology; to understand development it is necessary to also articulate a broader social theory, so as to situate the developing person in society. It was through his social psychological work that the problem of identities became explicit in his thinking, although it had already implicitly emerged in his struggles with the limitations of Piagetian psychology. The papers collected here provide an overview of how social representations and social identities came to be a pillar of his theory of human development and cognitive construction, while at the same time a contribution in their own right to these fields of psychology. The first paper describes how the individual-social dialectic can be fruitfully conceptualised with the concept of social identity, which is elucidated with examples from a variety of studies on children’s representations of gender. In Chapter 12 Duveen describes social representations as a genetic theory that can be applied to socio-, onto- and microgenetic levels of analysis. Another Chapter working out the relation between social representations and social identities follows this, which adds to the picture the crucial insight that resistance to identification can also occur. The final two chapters deal with social representations from a macro perspective. The first of these looks at relations between Social Representations Theory and Cultural Psychology, and describes research conducted on social representations of madness in India. The second reflects on the constitution of social groups through processes of communication, discussing the issue with particular reference to the second Part of Moscovici’s Psychoanalysis: Its Image and Its Public.

References Bartlett, F.C. (1932) Remembering: A Study in Experimental and Social Psychology. Cambridge: Cambridge University Press. Duveen, G. (1984) From Social Cognition to the Cognition of Social Life: An Essay in Decentration. Unpublished PhD Thesis. University of Sussex. Duveen, G. (1985) A Developmental Study of the Influence of Situation and Actor on Children’s Judgements about Friendship. Cahiers de Psychologie Cognitive, 5, 411.

Introduction xxiii Duveen, G. (1988) The Child’s Reconstruction of Economic Life. In K. Ekberg and P. E. Mjaavatn (Eds) Growing into a Modern World (pp. 177–199). Trondheim: Centre for Child Research, University of Trondheim, Norway. Duveen, G. (1996) The Development of Social Representations of Gender. Japanese Journal of Experimental Social Psychology, 35, 256–62. Duveen, G. (1997) Psychological Development as a Social Process. In L. Smith, J. Dockrell and P. Tomlinson (Eds) Piaget, Vygotsky and beyond (pp. 67–90). London: Routledge. Duveen, G. (2000a) Piaget Ethnographer. Social Science Information, 39, 79–97. Duveen, G. (2000b) La culture dans les jeux imaginaires de jeunes enfants (Culture in the Pretend Play of Young Children]. In D. Saadi-Mokrane (Ed) Sociétés et cultures enfantines (pp. 111–117). Lille: Editions du Conseil Scientifique de l’Université Charles-de-Gaulle– Lille 3. Duveen, G. (2001a) Representations, Identities, Resistance. In K. Deaux and G. Philogène (eds.) Representations of the Social: Bridging theoretical traditions (pp. 257–270). Oxford: Breakwell. Duveen, G. (2001b) Genesis and Structure: Piaget and Moscovici. In F. Buschini and N. Kalampalikis (Eds) Penser la vie, le social, la nature: Mélange en l’honneur de Serge Moscovici. Paris: Editions de la Maison des Sciences de l’Homme. (pp. 163–173). Duveen, G. (2002) Construction, Belief, Doubt. Psychologie et Societé, 3, 139–155. Duveen, G. (2007) Culture and Social Representations. In J. Valsiner and A. Rosa (Eds) Cambridge Handbook of Socio-Cultural Psychology (pp. 543–559). Cambridge: Cambridge University Press. Duveen, G. (2008) Social Actors and Social Groups: A Return to Heterogeneity in Social Psychology. Journal for the Theory of Social Behaviour, 38(4), 370–374. Duveen, G. and Lloyd, B. (1990) Introduction. In G. Duveen and B. Lloyd (Eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Farr, R.M. (1996) The Roots of Modern Social Psychology, Oxford: Blackwell. Jodelet, D. (1991) Madness and Social Representations. London: Harvester/Wheatsheaf. Jovchelovitch, S. (2007) Knowledge in Context: Representations, Community and Culture. London: Routledge. Lloyd, B. and Duveen, G. (1992) Gender Identities and Education. London: Harvester Wheatsheaf. Marková, I. (2003) Dialogicality and Social Representations: The Dynamics of Mind. Cambridge: Cambridge University Press. Moscovici, S. (1961/1976/2008) Psychoanalysis, Its Image and Its Public. Cambridge: Polity Press. Nicolopoulou, A. and Weintraub, J. (2009) Why Operativity-in-Context is Not Quite a Sociocultural Model. Commentary on Psaltis, Duveen and Perret-Clermont. Social Development, 52: 320–328. Piaget, J. and Inhelder, B. (1948/1956) The Child’s Conception of Space. London: Routledge & Kegan Paul. Psaltis, C., Duveen, G., and Perret-Clermont, A. (2009) The Social and the Psychological: Structure and Context in Intellectual Development. Human Development, 52, 291–312.

Part I

Piaget A view from afar

1

Children’s understanding of friendship Gerard Duveen

Introduction Conceptions of friendship as an object for psychological research seem to refer to a variety of issues and topics. A pilot study was undertaken in order to clarify some aspects of this conceptual development. One of the most sensitive techniques for exploring these issues is to allow the child to develop his ideas within the context of a semi-structured interview. Thus interviews were conducted on the basis of a series of questions covering the area of friends and friendships although each particular interview proceeded according to the child’s responses to these questions. The interviewer’s role was to enable the child to elucidate as clearly as possible his thinking about these issues rather than attempting to extract definitions of particular concepts from each and every child. The great advantage of this method is the freedom of expression which it gives the child, whose responses are not constrained to fit any set of categories predetermined by the interviewer’s questioning. This benefit is obtained, however, at the cost of certain problems. In the first place the concepts articulated in the interviews are not given clear and definite criteria (one hesitates to describe such criteria as ‘objective’), as would be the case for a standardized questionnaire. This objection is trivial in so far as it is precisely these criteria which the study aimed to explore. A more serious objection is that this method discriminates against the younger children in so far as it demands a verbal fluency which is beyond their capacity. The substance of this objection is that this method, related as it is to Piaget’s clinical interviews, shares with Piaget’s work the consistent risk of underestimating the actual capabilities of young children. In so far as the present study is intended as a pilot study it is not necessary to enter into a theoretical refutation of this objection. It will suffice to note that this study aims to explore conceptual development and that if its expression in language is slower to develop than at the level of action, or the demand characteristics of an interview inhibit a young child’s linguistic fluency, these do not invalidate an analysis of the interviews themselves.

4

Piaget: A view from afar

Subjects and method 21 children from a primary school in Lewes were interviewed, their ages ranged from 4 yrs 8 mths to 11 yrs 5 mths, with 3 or 4 children from each year being interviewed. There were 11 boys and 10 girls in the sample. Each child was seen individually by the interviewer for a session which lasted approximately half an hour. 3. The Interview Outline (i) Aspects of Friendship Do you know what a friend is? What can you tell me about being a friend? Do you have to be any kind of a special person to be a friend? How do you make friends with someone? How do you know if someone is your friend? How do you think someone knows if you are their friend? (ii) Being Friends Can you be a friend with anyone? A brother? A cousin? A parent? A teacher! If you’re a friend of someone do you have to be anything special for them? Or them for you? If a friend asks you to do something do you have to do it? What if you ask him/her to do something? If a parent or teacher asks you to do something do you have to do it? What would happen if you didn’t? How many friends do you have? Are any of them special friends? Why are they special? When you are with them what kinds of things do you do together? (iii) Development of Friendships If you make friends with someone is that forever? Why/Why not? Could you stop being friends with someone? How? Why? Could someone stop being friends with you? (iv) Contrast: the Category of Not-Friends What about people who aren’t your friends? What kind of people are they? Why aren’t they your friends? Do you play with them? Could you stop being friends with someone you didn’t like? (v) The Necessity of Friendship Do you have to have friends? What would happen if you didn’t have friends? (vi) Relations with Adults Could you be friends with a grown-up? Which ones? How do you make friends with a grown-up?

Children’s understanding of friendship 5 Could you stop being friends with a grown-up? Do you have to make friends with your mother/father/teacher? Could you stop being friends with them? How do grown-ups make friends with each other? Do they ever stop being friends with each other?

Analysis The analysis of material produced through interviews such as these usually begins with the attempt to identify different types of responses to the questions. Subsequently the conceptual structures underlying these different types are sought. From a cognitive developmental point of view a successful analysis yields a progressive logical sequence of conceptual structures, which may be regarded as stages. Each stage is assumed to have some overall characteristics which determine the type of responses to the various questions. Examination of individual protocols may, however, reveal inconsistencies within any one child’s answers (horizontal decalages) such that responses of varying types may be found for different questions. In the present case such a unified stage-theoretical analysis was not attempted. In part this was a consequence of the interviews being a pilot study, but in part also recognition that, although different types of responses were found, the inconsistencies in individual children’s answers were too great to be subordinated within such a stage analysis. No doubt one source of this problem was the difficulty which the interview presented for the younger children. In addition the style and phrasing of the questions was not directed towards such an analysis, the identification of stages is not an arbitrary process, but relies on asking questions which elicit such organisation. The present study had the more limited aim of exploring the child’s thinking about friendship as a preliminary to a more structured investigation. What did emerge from the analysis of the interviews were two broad orientations towards friendship; that is friendship was considered by the subjects either from the point of view simply of the actions which either partner might perform or, alternatively, from the point of view of a more complex conceptual model of social relationships between persons. These two orientations, which were apparent in responses to the whole range of questions in the interview schedule, are presented in some detail below. Following this discussion some of the themes and issues which emerged in the responses to particular sections of the interview schedule are reviewed (choice of friends, necessity of friends and relations with adults). Finally two particularly interesting examples from individual interview scripts are presented in some detail.

Two orientations to friendship The basic results of the analysis of the interview materials are presented here as a distinction between two orientations to friendship. Each of these is described as a set of ideas and attitudes. The limitations of the analysis noted above precluded

6


any attempt to expound them as structured conceptual systems, although these orientations can indeed be seen in responses to every section of the interview schedule. They do, therefore, appear to have some degree of generality. In addition it was also noted that the first orientation (friendship as action) was prevalent among, although not restricted to, the younger children; whilst the second (friendship as concept) was more common among the older children. There is thus some suggestion that these orientations also have a developmental character, and indeed some sense of logical progression can be seen in moving from the first to the second orientation. Although these orientations do show some of the characteristics of cognitive stages they are not yet descriptions of the structured conceptual systems (structures d’ensemble) through which stages are defined (what they lack above all else is an analysis of the structuring principles underlying each set of attitudes and ideas). They are intended, rather, as a first attempt at abstracting a general scheme from the material, at identifying what seem to be the significant conceptual distinctions in the different types of responses. (a) ‘Friendship as action’: In this orientation, prevalent but not exclusive to the younger children (5–7 or 8 years), the child is immersed in the practical activity of doing and being. To be friends means to play with or to do something with another. ‘Friends’ are the other children with whom the child is engaged in these practical pursuits; the other children at his worktable, or in the playground. Personal qualities are not mentioned in descriptions of friends or friendship formation. Indeed the process of making friends with other children does not seem to involve anything other than ‘being with’ other children; and similarly the breaking of friendships is the consequence of the interruption of this activity (whether through disputes or simply through the physical removal of the other). There is therefore a great sense of the fluidity of personal relationships in the descriptions given by the subjects. They do not report specific strategies for making friends, and although they may use terms such as ‘nice’, ‘kind’, ‘good’ or ‘like’ these refer to an undifferentiated positive evaluation of friends. There is no subtlety or ambiguity in the child’s descriptions and evaluations of personal relationships. It is almost as though there is here a conceptual absence, a noticeable lack of ideas. Consider this example: No. 9. A boy of 4 yrs 8 mths. Q. We’ll just have a talk about being friends. A. Some people aren’t friends in my class. Q. Aren’t they? Why not? A. ’Cos two of them always kicking people Q. Is that friendly? A. No Q. What’s a friendly thing to do? A. Be nice, sometimes they’re friends and sometimes they’re not Q. Do you know when they are and when they aren’t?

Children’s understanding of friendship 7 A. Q. A. Q. A. Q. A.

Yes ‘cos they kick people and be horrible Are you ever like that? No What can you tell me about being friends, what’s a friend? A friend you play with Can you stop being friends with someone? I do sometimes ’cos they have to go in their house

Yet these young children do not seem to be as completely indiscriminate in their practical social relationships as this suggests. It should be recalled that these interviews deal only with the conscious contents of the child’s thinking; at a practical level there may indeed be a greater sophistication in their social relations. Thus the conceptual absence can be seen as an expression of a vertical decalage, in the sense that the regulation of the child’s interactions with others is not yet separated from the activity itself so that the child’s conceptual apparatus does not match his practical capacity. (b) ‘Friendship as concept’ In this orientation (more common among the older children, 7 or 8 to 11 years) the child appears capable of reflecting upon himself and others as well as the relations between them. A friend is now someone of definite personal qualities, who is liked for these qualities, who may be called upon for help in times of distress and to whom a reciprocal obligation is recognised. The formation of friendships is understood as a process extending beyond the temporal horizon of immediate joint activity. Mixed feelings and ambiguities toward others appear to have replaced the undifferentiated responses of the first orientation. Referring to this orientation as ‘friendship as concept’ is intended to emphasise that the child’s social relations have now become a definite object of theoretical knowing for the child. He is aware of himself and others as independent subjects engaged in mutual relations. Here is an example: No. 1. A boy of 11 yrs 3 mths Q. What can you tell me about being a friend? When you say someone is a friend of yours what do you mean? A. Well, someone who’d help you when you’re in distress, if you fall over they’d pick you up, they’d play games with you Q. Do you have to be any kind of a special person to be a friend? A. No Q. Anyone can be a friend to anyone else? A. Yes Q. How do you know if someone is your friend? A. Well, you just know it. If they play friends with you, if they’re friends with you, you can play games with them, and they work with you, that’s virtually being a friend. Q. How do you go about making friends with someone?

8


A. Well virtually sticking to them, that’s how I find it, sticking to them, and working with them and playing games with them Q. How do you go about deciding which people you want to be friends with? A. You don’t really Q. You don’t decide that you want to be friends with this person or that person? A. No Q. How do you think someone knows if you’re their friend? A. I’m not sure, some people in the school would say it’s giving sweets to somebody, but that’s not really being friends Q. Why’s that? A. Well you can give away as many sweets as you like to a person but they may not like you Q. Can you know if someone likes you? A. Yes I think you can. They’d play with you, keep with you, walk home with you sometimes. This extract is particularly interesting as it shows so clearly some of the dilemmas involved in trying to analyse the interviews. In defining the term ‘friend’, for example, this child first of all argues that a friend is someone who will help you in times of distress. This definition goes beyond the perceptible and objective definitions based around the notion of ‘being-with’, yet all the same in elaborating his argument he has recourse to the same events as the children in the earlier extracts. Several times he refers to friendship as an engagement in joint activity, particularly when he is describing the process of friendship formation. Despite his use of more abstract categories in defining ‘friend’ his instantiations of the concept refer only to objective events; subjective qualities and dispositions are not used to elucidate the term. Yet in considering how someone can know if you are their friend the emphasis is on the attraction between people, to the exclusion of objective events. Here the friendship as concept orientation re-emerges, only to dissolve once more into objective events when he is asked what he means by ‘liking’ someone. Such vicissitudes in thinking show clearly the limitations on the analysis of this pilot study. Yet it is also clear that these children’s responses are not unstructured, since it is possible to identify some aspects of the conceptual systems underlying their responses. It may well be that the interviewing was itself not sufficiently sensitive to elaborate these structures. Equally some of the children interviewed may have been in the process of transition from one orientation to the other, and were consequently unable to articulate their ideas more clearly. Whatever the reasons these inconsistencies remain, they may even be horizontal decalages; their existence making it difficult to abstract any developmental trends from the data. These two different orientations do seem, however, to represent a priori a logical sequence. Contrasting them one can see a movement from a world which is comprehended in terms of objective events to an understanding based on subjective, psychological factors. This movement clearly parallels the change in the child’s understanding of his world which occurs with the elaboration of concrete operations and which has been noted in so many studies dealing with the child’s

Children’s understanding of friendship 9 social world. In the former the child’s thinking is based on the appearances of figurative knowing, whilst the latter emphasises the operative relations between theoretical objects. These pilot interviews do suggest that a similar process underlies the development of children’s thinking about friendship, even if this study remains inconclusive on this point.

Choice of friends The two orientations to friendship can also be traced through the discussions of friendship choice. Nearly all the children interviewed agreed that it was possible for anyone to be a friend to anyone else whilst also finding it difficult to offer any explanation of why one becomes friends with only certain others and not with everyone. Beyond this bare summary the responses can be distinguished between the two orientations. Within the friendship as action orientation the picture presented is of an extremely fluid social world in which the operation of chance meetings is the only determining factor. No doubt this is the case conceptually, where it is joint activity alone which constitutes the bond and where the maintenance of current activities or repeated contacts are sufficient to sustain it. There is little or no sense of any personal volition in the development of friendships, nor of any attraction between people prior to the formation of the bond. Even after the formation of the bond personal qualities are not dissociated from the relationship itself. One simply likes one’s friends and one’s friends are the people one likes. In this sense both ‘like’ and ‘friend’ are always associated with the same others, both terms form part of that single positive and evaluative complex noted in the previous section. To this set there corresponds the equal and opposite negative evaluations of dislike and not-friend. The other persons in the child’s world are grouped in terms of these two sets. Indifference as a conscious attitude towards other persons appears to make only a late appearance in the child’s development. If another child who is at first evaluated in terms of either of these sets takes on any of the characteristics of the opposite set he also takes on the totality of the other set so that the evaluation is completely reversed. The rapidity with which young children make and break friends lends support to this interpretation. Such sudden reversals of valorizations lead to the conclusion that these sets are not considered by the child to refer to stable dispositions of other persons, but rather to be descriptions of how things are at any given moment in time. In each situation the child is dealing with what he takes to be the totality of the other, yet this totality is not stable, it may change from one moment to the next from one situation to the next (in this sense the friendship as action orientation repeats the pre-operational child’s reliance on figurative knowing). Other children do not as yet appear as stable or durable psychological entities for the child, and his relations with them are determined primarily by the situation in which he encounters them. Considered from the orientation of friendship as concept these issues appear very differently. In the last extract in the preceding section for example when asked how he makes friends with someone the boy (subject No. 1) suggests that

10


he has somehow decided that he wants to be friends with someone and then ‘sticks’ to them. This implies an act of volition on his part which it would be difficult to see as the result of only a chance encounter. The opportunity to meet someone may be a necessary condition for the formation of a friendship but it can hardly be considered sufficient. As this child himself says in response to later questions one must also like the other person. At this point in the interview another inconsistency occurs, for in his very next statement after indicating his own volition in forming friends this child denies deciding which people he wants to be friends with! It appears as though his own attraction to others is a mystery for this child. Nevertheless his answers do give some indication of an attraction preceding the formation of a friendship which may motivate one or other party to induce a closer relation with the other. Somehow (presumably through observation, although the child is not clear on this point) there is an inference that the other possesses qualities which the child finds attractive. This represents an advance in the structuring of the child’s social world since he is becoming capable of attributing stable dispositions to other persons and of orienting himself towards them in time. The fluidity of the child’s social world with its reliance on the immediate situation is giving way to a world in which other persons possess characteristics which are set in a temporal context. Such attributions of stable dispositions to others require a coordination of the temporal relations of past, present and future.

The necessity of friends A clear indication of the importance of friends for the child can be seen in the responses to section (v) of the interview schedule. All of the children asserted the necessity for interacting with others. Life without friends would be lonely since “there’d be no-one to play with”. At each age and from each orientation to friendship a life alone was equated with a miserable life. Indeed befriending a lonely child was often mentioned as an instance of doing “something nice” which might form a friendship. The strength of the children’s responses to these questions reflects the crucial importance of the peer group for the child; an importance which is not restricted to the child’s emotional life but extends also to his cognitive development. For Piaget the confrontation with and accommodation to a world of thought external to his own was an essential factor in the development of operatory thinking in the child. Indeed he frequently drew a parallel between the development of logical operations at the level of individual thinking and the appearance of co-operations at the level of social interactions. From the child’s point of view the importance of the peer group lies, however, in the tangible benefits of friendship rather than recognition of the enigmatic influence of peer interactions on his own cognitive development.

Relations with adults So far this analysis has concentrated on the various conceptual differences which the interviews revealed in relations between peers. The strongest contrast,

Children’s understanding of friendship 11 however, is that which every child interviewed demonstrated between his relations with his peers and his relations with adults. The child’s social world is divided by the different moralities that pertain in his relations with his peers and with adults. This distinction corresponds precisely to Piaget’s (1932) contrast of autonomous with heteronomous morality. Piaget’s analysis of the development of moral judgements about relations with peers shows a change from heteronomy to autonomy. Yet these interviews clearly show that there is a continuity throughout the child’s development which relates to what actions are possible within these two spheres of action with his peers and with adults. It is in this sense that these two spheres of action can be contrasted as autonomous and heteronomous. Some developmental changes in this distinction were noted in the interviews, and in presenting these results the earliest conceptions are considered first followed by a commentary on the positions advanced by some of the older children interviewed. In interacting with his peers the young child projects himself in a sphere of action in which he is free to make his own decisions about how to conduct his relationships. Vis-à-vis his relations with adults the child is constrained by the hegemonic influence exerted over him by the adult. He recognises a general obligation to do what an adult requests him to do simply because the other is an adult and the child a child. It is only in his dealings with other children that the child finds the freedom of any room for manoeuvre, any possibility of coming to a decision for himself as to how he should respond to another’s request. To be sure this freedom is conditioned by the responsibility for accepting the consequences of his actions; but here too the split between his relations with his peers and with adults makes itself felt. The consequence of not doing what another child has asked him to do is limited to a break of friendly relations with the other child. This is not in itself a necessarily threatening consequence for the child, since young children seem to accept the fluidity of a social world in which friendships are made and broken frequently. They can, after all, make it up with the other child another time. If they express the necessity for having friends this is not the same as a necessity for maintaining their existing friendships. Refusing an adult’s request is more immediately threatening for the child. The dictatorial authority of the adult is backed up by the threat of punishment, frequently physical. The adult’s hegemonic power over the child is associated for the child with the adult’s ability to inflict punishment on the child. These conceptions are illustrated in the following extracts: No. 15. A girl of 8 yrs 3 mths Q. If you’re a friend of someone do you have to do anything special for them? A. No Q. Do you expect them to do anything for you? A. No Q. If your friend tells you or asks you to do something do you have to do it? A. No Q. Why not? No reply

12


Q. What if you ask them to do something, do you expect them to do something for you? A. Sometimes Q. What kinds of things do you expect them to do? A. Any sorts of things Q. If a parent asks you to do something would you do it? A. Yes Q. Would you have to do it? A. Yes Q. Or if a teacher asked you to do something? A. Yes Q. What would happen if you didn’t do something a parent asked you to do? A. Get smacked. Q. Or didn’t do what a teacher asked you to do? A. Get smacked. Q. Or if you didn’t do what a friend asked you to do? A. Just break friends. No. 7. A boy of 8 yrs 6 mths Q. If a friend asks you to do something do you have to do it? A. Sometimes Q. Which times do you have to do it? A. Well if someone asks me to play – do something like dig snow up with a toy tractor, I’d do that Q. And what if you ask them to do something, do you think they ought to do it? A. I don’t know Q. Well are there things you wouldn’t do if a friend asked you to? A. If one of my friends asked me to help fight I wouldn’t Q. Why not? A. Cos I’ll get hit Q. What if a parent asks you to do something, do you have to do it? A. Yes Q. Whatever it is? A. Yes Q. Or a teacher A. Yes Q. What would happen if your friend asks you to do something and you didn’t do it? A. He’d do it on his own Q. What would happen if a parent asked you to do something and you didn’t do it? A. They’d tell me to do it and if I didn’t I’d get sent up to bed Q. And if you didn’t do what a teacher said? A. Get sent up to see Mrs Pleasance (The Headmistress) and I’d get told off

Children’s understanding of friendship 13 This contrast between relations with peers and relations with adults was evident in all the interviews. Even amongst the older children these two areas were distinct in the freedom of action which the child considered appropriate to each area. The only difference was that a few of the older children no longer saw the adult’s authority as being based solely on their capacity to inflict punishment, instead they tended to argue that the child’s obligation to do as an adult asks was derived from a respect for the instrumental help given by the adult in the child’s own life. In this way they show evidence of a little more mutuality in their relations with adults, although this does not yet extend to a freedom from the constraints imposed by the adult. There remains the general obligation for the child of doing what an adult asks him to do, an obligation which, although it may be set in a more liberal context for the child, rests ultimately on the threat of punitive sanctions which the adult possesses. No. 3. A boy of 9 yrs 8 mths Q. If a friend tells you to do something do you have to do it? A. You don’t have to, but to be a friend it’s best to Q. And what if you ask him to do something would you expect him to do it? A. Not really. Well they just do it Q. What if a parent told you to do something? Would you have to do it? A. Yes you would Q. Why’s that? A. Because a parent does a lot of things for you, she goes to the butcher’s and gets some meat and cooks your dinner. It’s best to do things for your mother and father as well Q. What about a teacher, do you have to do what a teacher tells you? A. Yes I expect you do Q. Why’s that? A. Well if you want to learn, to learn things and get a good job when you grow up it’s best to do all the work you can when you’re at school Q. So you think you ought to do what a teacher asks you to? A. Yes Q. So why then don’t you have to do what a friend asks you to? A. Well, it depends how much you like them Q. So with friends you don’t have to but with teachers and parents you have to, is that right? A. Well, you don’t have to with parents but it’s best to and with teachers you almost have to but . . . Q. Not quite? A. Not quite No. 2. A girl of 11 yrs 5 mths Q. If a friend tells you or asks you to do something do you have to do it? A. No you don’t have to but if you were a friend you would

14


Q. Why would that be? A. ’Cos you’re their friend Q. And what if you ask him or her to do something for you would you expect them to do it? A. Well, if it’s a good thing, if it’s bad no Q. So it would depend on what kind of thing you asked them to do? A. Yes Q. If a parent told you to do something would you have to do it? A. Yes Q. Why’s that? A. Well because they’re the ones who brought you up and you have to do what they say Q. And what about a teacher, if a teacher told you to do something would you have to do it? A. Yes, ‘cos they’re in charge of you In these last extracts the expressions of obligations towards friends show some elements of the friendship as concept orientation in so far as they are based on the comprehension by self and other of the exchanges taking place between them. Such thinking can also be seen behind their descriptions of adult authority. The adult’s capacity for inflicting punitive sanctions is now given a reason, a justification. This contrasts with the earlier transcripts in which the adult’s action was seen as simply an unmediated response to the child’s refusal to comply with the adult’s request. Although aware of the punishment’s relation to their own antecedent action the younger children showed no evidence of expecting any justification for adult sanctions. For them the world of adults is a world in which they are subject to the hegemonic power of the adult and their awareness of this power is sufficient to activate them to discharge the obligations it imposes on them and to accept as legitimate the sanctions imposed on them when they transgress. For the older children the adult retains his unilateral authority and this again legitimises for the child the obligations and punishments imposed on him. However, in so far as this authority is now seen as having some justification by the child it has been mediated through his awareness of the adult as a subjectivity external to his own; the child now sees the adult’s actions as being motivated by the adult’s own values and ideas rather than as a reflex response to the child’s antecedent action. To be sure these children have not escaped from the adult’s authority altogether, but in attributing a justification to the adult the child is treating the adult as another subject and to this extent the change can be seen as a development towards a mutual reciprocity in their relations (the only ‘adults’ considered in these interviews were parents and teachers, whether or not every adult the child encounters holds the same authority is not clear, it may be restricted to those adults ‘in charge’ of the child). This discussion of the ways in which the child perceives the differences between his relations with his peers and his relations with adults has touched on some

Children’s understanding of friendship 15 questions relevant to the child’s comprehension of his social world as a whole. The issues of freedom and responsibility in relations with others, of the development of a sense of the other as being motivated by their own values and ideas of seeing the other’s actions as being mediated through their own psychological structure; all these questions refer also to the development of the child’s own sense of self to the development of subjectivity. Although such issues are also to be found in the other parts of the interview schedule they stand out most clearly in the area of the child’s relations with adults where the sharply contrasted realms of autonomy and heteronomy project these issues to the fore.

Individual cases At particular points in some of the interviews the children began to talk about their own problems and concern in a very direct way. They do so when the affective dynamics associated with their own unresolved problems are engaged by the interviewer so that the discussion moves from the hypothetical to the real. In this sense some of the children are provoked into articulating their thinking about their problems, which may have a particular interest from a theoretical point of view (Eleanor Duckworth makes a similar point concerning cognitive developmental testing, cf. Duckworth 1974). The two cases presented and commented on below illustrate clearly how some of the problems already described in the analysis of the interviews present themselves in the course of the child’s own life. (a) The first case touches on the problem of the development of a sense of self. The child is a girl of 9 yrs 0 mths, subject No 8. Elsewhere in the interview, this child seemed to be on the verge of ‘friends as concept’ orientation, by trying to say something more than the mere fact of being together. She did not manage to articulate what this ‘something else’ might be, but later in the interview came the following exchanges: Q. A. Q. A. Q. A. Q. A. Q. A. Q. A. Q. A.

Can you stop being friends with someone? Yes sometimes I do Why? If I stop telling jokes and then . . . I stop being friends What makes you stop being friends with someone? If I tell them to do something or I swear at them or shout at them And could someone stop being friends with you? Yes Why? If I hurt them they don’t be friends, accidentally and they think I really hurt them true Does that mean you’re never friends again? Sometimes it does and sometimes we’re friends again And how do you make friends again? I say sorry I didn’t mean it and then they start being friends again.

16


The first part of this extract is a good illustration of some of the problems encountered in this type of interviewing; the girl responds to the interviewer’s questions, but after her own fashion and not always in the sense in which the questions were intended. When asked why she can stop being friends with someone she speaks very precisely about breaking an established pattern of relating to another (“If I stop telling them jokes”), which although it is primarily an example of the friendship as action orientation is nevertheless phrased in such a way as to raise questions about her sense of her own volition in the breaking of friendships. In an attempt to clarify her sense of her own volition the interviewer asks her “What makes you stop being friends with someone?” The reply is rather unexpected for the girl answers not from a perspective based around notions of volition but on the contrary from a perspective in which actions are the primary causes (“If I tell them to do something or I swear at them or shout at them”). Here there is no prior sense of intention or volition behind these actions. There is no distinction between the meaning of an action and the action itself. In the very next exchange, however, where she is asked why other people can stop being friends with her she replies “If I hurt them then they don’t be friends, accidentally and they think I really hurt them true”. The first part of her reply continues the discourse which she has already established in which action and its meaning are not dissociated. She goes on, though, to qualify this phrase with another (“accidentally and they think I really hurt them true”). It is noticeable that in articulating this phrase her command of English is less grammatically competent; although retaining its expressive power. In distinguishing accidental from purposive interpretations of actions she is opening up a different perspective for her thinking since this distinction creates a space between action and its meaning (and it is presumably because she is articulating an idea which is new to her that her grammatical competence is affected). This distinction arises in the following way: Our subject has understood her own action as being accidental, whereas her friend has understood it to be purposeful. When her own action is reflected back to her through her friend’s interpretation of it the meaning has been transformed from accidental to purposeful. How is such a transformation possible? How can the girl account for this transformation in the meaning of her own action? She recognises, more or less clearly, the poles of this transformation in her own and her friend’s interpretations yet she seems to remain unclear about the transformation itself. From the way that she speaks it would seem that this problem remains unresolved for her and it is in this sense that the girl can be said to be grappling with a current problem in social relations. What she has not yet grasped is that the transformation in meaning of her own action has been effected through the psychological action of the other (one could speak of the other as a ‘transforming-agent’ indicating a transformation by the other of the meaning one’s own action has for oneself). To resolve her problem the girl has to develop some psychological insight into the other person. To do so requires a separation of action and meaning so that actions become susceptible to more than one interpretation. It requires further that actions and meanings be coordinated so that interpretations can be ascribed to self and other. If experiences

Children’s understanding of friendship 17 such as this are a source of insight into the other then they are also reciprocally constitutive of the subject’s own sense of self. What is emerging in this process is the reconstruction of relations between self and other at a theoretical (operatory) level from their preoperational comprehension. Self and other become differentiated as self attempts to comprehend the other’s transformations of the meaning of self’s own activity. Self and other need to become coordinated if self’s own understanding of the action is to be conserved across the face of its apparent transformation by the other. (In parenthesis here it can be noted that this represents one of the tremendous advantages which the child’s relations with peers have over their relations with adults; the hegemonic authority of adults would seem to preclude such developments, although the child’s relations with adults may also provide one level of security from which it becomes possible to explore other levels.) In the dialectic of self and other evident in this passage from the interview script this girl is clearly in the process of becoming conscious (what Piaget has called the process of ‘cognizance’ or ‘la prise de conscience’) of new ways of thinking about herself in relation with others. A growing differentiation between action and its interpretation is appearing which is also leading to their integration in new systems of thought. As this integration appears ‘self’ and ‘other’ are coming to be defined in terms of a new set of operative structures made possible by this girl’s ‘discovery’ or construction of the notion of psychological action (or intent) as something distinct from physical action. It is this notion of psychological action of which the girl is in the process of becoming cognizant. This passage also offers some insight into the description of ‘la prise de conscience’ given by Piaget. His thesis is that consciousness proceeds from the periphery to the centre (Piaget 1972, 1977). In using this vocabulary of ‘periphery’ and ‘centre’ Piaget wanted to emphasise that “knowledge does not proceed from the subject or from the object but from the interaction between the two” (Piaget 1977, p. 335). It is therefore the point of interaction between subject and object which is said to be peripheral to both, and cognizance “proceeds toward the central mechanisms of the subject’s action, whereas awareness of the object moves in the direction of it’s intrinsic properties (in this sense also central), and no longer in that of the superficial properties connected only with the subject’s actions” (ibid p. 335). Thus from this psychological point of view ‘subject’ and ‘object’ are always correlative, and, Piaget adds, “this correlation constitutes the basic law, both of the understanding of objects and of the conceptualization of actions” (ibid. p. 335). In the present case, too, the development which is in progress in the interview is proceeding from the periphery to the centre, that is from the point of interaction between self and other toward an operatory comprehension of both. The periphery here is the zone of uncertainty marked by the girl’s reply “If I hurt them they don’t be friends, accidentally and they think I really hurt them true”. In this reply the girl acknowledges an unresolved contradiction between her understanding of her action and that given to her by her friend. Her uncertainty is not concerned with either her own understanding of what happened nor with the interpretation given

18


to her by her friend but rather in her inability to satisfactorily coordinate the two points of view, that is in the interaction between the two. It is in this sense that the two points of view can be seen as contradictory rather than merely as opposites, for it is her expectation that there ought to be some coordination between her interpretation of what happened and her friend’s interpretation which generates her uncertainty. This situation is, then, decisively different from those arguments so typical of young children which culminate in the repeated assertions of “Tis”, “Tisn’t”, “Tis”, “Tisn’t”, etc. etc. In such arguments each child is merely wishing to impose his or her point of view on the other with no effort being made to accommodate the other person’s perception. In the present case it is the credence given to her friend’s point of view which underpins the girl’s uncertainty; in seeking to accommodate her friend’s perspective the girl has discovered the maladaptiveness of her own assimilatory processes since her friend’s perspective resists assimilation into her own understanding. Thus Piaget’s point remains that “consciousness is at first connected with an environmental situation blocking some activity; therefore with the reasons for this maladaptation and not with the activity itself, which does not give rise to reflection so long as it remains adapted. Consciousness thus proceeds from the periphery to the centre and not inversely” (Piaget 1972, p.135). The ‘periphery’ is therefore not a definite zone which exists permanently between subject and object; but rather a zone of uncertainty which appears in the course of development between subject and object as the object resists being assimilated into the subject’s existing structures. Periphery in this sense is a developmental term. In using the present example from these interviews to clarify the notion of ‘periphery’ it should also be clear that this example relates to a single moment of a developmental process and not to the process as a whole. This moment of uncertainty is clearly dependent on some previous development such that this girl has indeed given some credence to her friend’s interpretation of what happened—that is she has made some attempt to accommodate her friend’s perspective. Without this attempt there would be no uncertainty and the two interpretations would be left simply as opposites and not perceived as an unresolved contradiction, as they would have been at earlier stages of development. The appearance of a peripheral zone of uncertainty is clearly an important moment in the process of development itself, of the transformation from one cognitive structure to another; a process which Piaget has attempted to analyse in his theory of equilibration (Piaget 1978). As noted above the imbalance between assimilation and accommodation is an important source of developmental tension. The present example emphasises in relation to the socio-cognitive structures of self and other what has also been emphasised in recent Genevan research on the development of cognitive structures in the more traditional sense of relating to ‘physical objects’; this is the importance of social interaction in generating this imbalance between assimilation and accommodation (cf. Doise 1978, Perret-Clermont 1980). (b) In the second case the extract comes from an interview with a boy of 9 yrs 8 mths (subject No. 3, another extract from this interview was given in the section ‘Relations with Adults’)

Children’s understanding of friendship 19 Q. If you make friends with someone is that forever? A. No, because say, if he takes another friend and I don’t like that friend we’re still friends but we’re not best friends Q. Why not? A. Because there’s a person I don’t like which is his best friend, and a person that I like very much that he doesn’t like Q. And that would stop you two being friends? A. That would stop us being best friends Q. Why would that stop you being best friends? A. I don’t know Q. Can you think of an answer? A. No . . . he’s my best friend but not quite the person I really like best. I’ve got brothers, cousins, my mum and dad, people like that Q. But let’s go back, your best friend has a friend you don’t like and you have a friend that he doesn’t like A. Yes but them friends take it, they take . . . say I want to go out with my friend and Terry wants to go out with his friend and we both want to go out together, well the other, Terry’s friend takes Terry out to his house and my friend takes me out to his house; something like that, and we can’t keep together and we can’t talk and we can’t still be best friends Q. So you could stop being friends with someone. What kinds of things make you stop being friends with someone? A. We can’t play together, we can’t ride together on our bikes, we can’t sit next to each other in the hall which we usually do very much because we like them, and we can’t sit next to them in class because say my friend sits next to me and then Terry’s friend sits next to me and that stops Terry sitting next to me. This child articulates the problem he is engaged with in terms of Heider’s balance theory! The problem remains unresolved, but in thinking about it he is having to adjust himself to a reorganisation of his network of relations. In so doing he is, like the girl of case (a), reconstructing his thinking about social relations at an operatory level. His conceptualisations of relations with others are changing at the same time as he is clarifying the extent and range of his notions of ‘friend’ and ‘best friend’. These classifications are being coordinated with his understanding of the relations to which they refer. The reference to balance theory is not simply fortuitous, for in a real sense this is exactly the problem as the child conceives it, how is he going to be able to balance this set of relations (which is not to suggest that balance theory is the best or most felicitous of theories for dealing with problems of social relations, only that it can be used to describe this particular problem). In speaking about his problem the child articulates some of the points which have been discussed in preceding sections of this analysis. Firstly the whole problem is conceived by the child as an issue set in the context of continuing and maintaining relationships with others. It is no longer necessary for him to consider these others as friends only as long as they are jointly engaged in some activity. For him the question of

20


friendship has progressed beyond the simple recognition of a state of affairs so that ‘friendship’ is now seen as a relation which extends in time and space beyond the immediate and given situation. Secondly, this conception of relations with other people goes hand in hand with a recognition that other people do have stable dispositions which also extend beyond any particular situation. At heart, then, the child’s problem is not simply one of arranging his affairs so that he can maintain his relations with others taking into account the mutual incompatibilities which exist. His problem is really a structural one (which has both cognitive and affective components) of coordinating his situation in a way that respects each person’s understanding of the network of relations and their affective responses to it. It can be seen from the transcript that this child does not expect others to like him or accept him as a friend simply because they are together. Quite the opposite, he sees their likes and dislikes as determining factors in deciding whether or not it is possible for them to be together at all. Thus this case study shows a child for whom the psychological dispositions of others precede the engagement in friendly relations, as opposed to the attribution of a complex of descriptions (like/nice/ kind/friend) through engaging in shared activities. This is again the contrast between the two orientations, friendship as concept and friendship as action.

Conclusions This analysis has set out to offer an interpretation of what the children said in the interviews. It has continually asked the question how is it possible for the children to say these things, how must the child represent his world to himself for such answers to be possible? Obviously these interpretations (and indeed the questioning in the interviews) have been mediated through the analyst’s own conceptualizations. In so far as the starting point has been Piaget’s general theory the results can be read as an attempt to theorize about the interviews from this point of view so that some parameters can be clarified for further investigations. In the first place the analysis has emphasised that conceptions of friendship must be located within the child’s comprehension of the totality of his social world. Since friendship is one aspect of relations with others, the ideas children present about friendship reflect their more general conceptions of social relations. Discussions about friendship occur within the same field, and frequently refer to conceptions of self and other and the relations between them. Children’s theories about friendship are at the same time theories about the social world, about relations with others and about what kind of beings are implied in such relationships. The analysis identified two orientations towards friendship, which if they cannot be read as description of a sequence of stages can be seen as a rough guide to a developmental sequence. The earlier orientation of friendship as action is based on the child’s practical ‘doing’ and explained on the basis of objective appearances. The later friendship as concept orientation is an operatory or theoretical comprehension of the child’s activity and reasons on the basis of the transformations underlying the ‘objective appearances’. These descriptions repeat the

Children’s understanding of friendship 21 changes observed in other areas of social development, dealing with the child’s conceptualisations of other and people and relations between people. When the child’s thinking is based on ‘objective’ considerations other people appear to the child as undifferentiated with respect to their personality or character. At least no reference to such aspects appear when they talk of other people. Others impinge on the child only in so far as he interacts directly with them. They then become subsumed under a set of undifferentiated positive or negative evaluations, and the process of choosing friends comes down to little more than the random operations of chance. With the emergence of operatory thinking the child’s view of others becomes increasingly differentiated as other persons acquire a solidity which extends in time and space beyond the immediate situation (which is the limit of the earlier orientation’s horizons). This permanence finds its expression in the attributions of stable dispositions to others, of the development of psychological insights into others and in the development of subjectivity. Equally the term ‘friendship’ now refers to relationships which extend beyond the immediate situation. Personality and affective preferences are now seen as important influences in the evolution and maintenance of friendships, indeed they are seen as structuring the child’s interactions with others. In the two case studies an attempt was made to examine such psychological considerations by viewing them as operations which bear on transformations underlying contradictory appearances, that is by isolating what is conserved across such transformations. Associated with each orientation to friendship is the child’s sense of what specific practices are relevant to this social relationship, what duties and obligations it imposes on the child and what he has the right to expect of others. Such expectations of self and others mark out the boundaries of the child’s friendships in his social world. It is only in the later friendship as concept orientation that an action and its meaning become separated and coordinated. In the earlier orientation it seems that the meaning of an action is inherent in the action, that it is as ‘objectively’ visible as the action itself (just as Piaget describes the young child’s belief that the name ‘sun’ is inherent in the object which he sees in the sky). All of these issues relate to the child’s relations with his peers. In both orientations however a strong distinction is drawn between such relations and relations with adults (at least with those adults perceived as having authority over the child). The adults’ hegemonic influence imposes constraints on the child which mitigate against the formation of mutually reciprocal relations. With his peers on the other hand the child experiences a realm of freedom which allows him to accept the responsibility for the consequences of his own actions in relation to others. The child, therefore, is able to control his relations with his peers in a way which is not possible with adults. This description of two diverse orientations to friendship corresponds quite closely to results of developmental studies in the area of person perception. A number of studies have shown that children’s descriptions of other persons change around the age of 7–8 years from descriptions based on the activities of others to

22


descriptions founded on the psychological attributes or dispositions of other persons (this literature has been summarised in two recent reviews; by Shantz 1975 and by Chandler 1977). Although these studies have used a variety of techniques for eliciting children’s descriptions of others and a variety of theoretical models for interpreting the results, there is a unanimity in the description of this developmental change. A good example is the study by Livesley and Bromley (1973) in which 320 children aged between 7 and 15 were asked to write descriptions of themselves and of eight others who varied in age, sex and affective relationship. The children had first of all to provide an example for a category of person defined by the experimenters (e.g. ‘a girl I know very well and like is . . .’) and subsequently to write a description of this person. The analysis of these descriptions showed that the number and proportion of psychological statements (or dispositional constructs) increased significantly with age. However, the proportion of psychological statements increased significantly only between the 7 and 8 age groups and indeed the authors noted that “frequently the differences between 7 and 8 year old groups were greater than those between 8 and 15 year old groups” (Lives1ey and Brom1ey 1973, p. 147). The categories used by the 7 year olds in describing others tended to invoke material criteria such as appearance and possessions while the older children focused on values, beliefs and dispositions, that is on perceived regularities in behaviour across variations in space and time and in different situations. Amongst other things one important conclusion drawn by these authors from their study was the suggestion that “the eighth year is a critical period in the developmental psychology of person perception” (Livesley and Bromley 1973, p. 147). The significance of the results of such studies in person perception for the present study of friendship is threefold. In the first place they offer some external validity for the distinction made in this analysis between an orientation to friendship as action and an orientation to friendship as conception. Secondly they offer some circumstantial evidence in support of a developmental relationship between these two orientations; in the present analysis the ‘friendship as conception’ orientation also emerged around the age of 7–8 years. Both of these points establish empirical correspondencies between the results of the analysis of these interviews and other, related studies; they are thus essentially concerned with the process of verification. The third point, however, goes beyond empirical concerns and raises a theoretical problem which is important not simply for the analysis of friendship but also for the organisation and direction of further research. That such correspondencies could be found between the analysis of these interviews about friendship and studies in the development of person perception reinforces a point made in the analysis of the interviews themselves. This was the sense in which friendship could not be considered an independent ‘domain’ in the field of social cognition but seemed, rather, to be linked theoretically to other aspects of social cognition. Thus, for example, friends are also other people (and it would consequently have been more surprising if no correspondencies had emerged between thinking about friendship and studies of person perception) just as the dimension of friendship is an important boundary

Children’s understanding of friendship 23 marker in the field of the individual’s social relations with others. It follows that thinking about friendship (as with thinking about other ‘domains’ of social cognition) arises only on the basis of an underlying set of operative structures which are concerned with the understanding of social relations in general. If this is indeed the case and friendship is not an independent ‘domain’ but rather one aspect of some greater whole what justification can there be for organising further research around the notion of ‘friendship’? Surely the research strategy should be directed toward isolating and describing the greater whole of operative structures bearing on social relations; and on the relations between this greater whole and the parts it subsumes (friendship, person perception etc)? This question touches one of the major issues confronting every current research effort directed toward an analysis of the development of the child’s comprehension of social relations. To put it as plainly as possible the problem is that “social relations” as an object of knowledge remain illusive and ill-defined; existing theoretical analyses of social relations are all partial, no generally accepted analysis exists which could serve as an organizing model for structuring research. In Piaget’s work there is a clear relationship between objective, or rather scientific knowledge of an object and the psychogenesis of that knowledge in the child. Scientific knowledge is, for Piaget, the ‘highest form’ of knowledge, in the sense of being the most adequate or most valid knowledge about an object. Knowledge is, of course, always a process for Piaget rather than a fixed state, and as such scientific knowledge is always itself a historical product, and always in the condition of being open to further revision. What is important from the psychogenetic point of view is that scientific knowledge marks a relative end-point in the developmental construction of knowledge. Thus a kind of parallelism is established between the history of scientific knowledge and the psychogenesis of knowledge. It is not a perfect parallel, there are a number of examples in which a kind of knowledge appears earlier in psychogenesis whilst appearing later in the history of science (perhaps the best known of these is in the field of geometry where topological notions predate metrical notions in psychogenesis whilst in the history of geometry they appear after the development of metrical analysis). Nevertheless the point to be understood is the importance of scientific knowledge in providing a source from which research in psychogenesis may be organized and directed (cf Piaget 1972). In the field of social relations there is, from this point of view, a novel situation for genetic epistemology. In the various Genevan studies of cognitive development there is an implicit reliance on the fact that an analysis of the structure of (physical) objects was not only available but also, more importantly, existed as a body of verified and accepted scientific knowledge. The analysis of social life, on the contrary, has been a continual source of dispute for as long as men and women have concerned themselves with these issues, and this dispute shows no signs of any imminent resolution. If, as Piaget argues, development always proceeds towards objective knowledge the question here is how could objective knowledge of social relations be defined? It is as though through this project of psychogenetic research itself there is a need to analyze not only the structures of the subject in

24


relation to this object but the very structure of the object itself. Any research which is directed toward the analysis of the operative structures through which social relations are comprehended will, therefore, also be led to consider the structure of those social relations themselves. There is then a recursive quality to any such research; the researchers’ conceptions of the nature and structure of social relations become as much an issue in the research as any resulting descriptions of the structures through which social life is comprehended. This problem of recursiveness is not peculiar to the area of social cognition, it exists just as strongly in every area of psychology and indeed in every kind of scientific activity. However where there exists a body of verified and accepted scientific knowledge about an object this provides a certain degree of closure in the recursiveness in the study of the psychological structures through which that object is comprehended. The lack of such a body of scientific knowledge is what distinguishes the area of social cognition, leaving an open recursiveness from the point of view of studying the psychological structures through which social life is comprehended. Such general considerations will, of course, continue to exert their influence throughout this current research yet by themselves they do not resolve the problem of organising further research beyond the interview study. Some direction for research needs to be found, even if this cannot be marked out in reference to any ‘objective’ criteria derived from a knowledge of social life. The strategy which has generally been adopted by researchers in the field of social cognition has been to study various ‘domains’ of social thought as though they were independent of one another with the prospect that, eventually, a comparative analysis of the various ‘domains’ would enable some approach to be made toward describing the ‘whole’ of which these ‘domains’ form the constituent parts. This goal of theoretical integration, however, has not (or perhaps not yet) been reached, in spite of a number of attempts at synthesizing results from various ‘domains’ (e.g. Kohlberg 1969, Damon 1977, Selman, 1980). This state of affairs is all the more surprising in view of an extraordinary convergence of empirical evidence from many diverse studies of various aspects of social thinking. Time and again results have shown that around the age of 7–8 years the child’s thinking about social relations undergoes a transformation from ‘objective’ criteria to more ‘subjective’ criteria. This result has emerged not only in studies bearing on interpersonal relations (person perception, role-taking, moral judgment, friendship) but also in relation to the development of political and economic conceptions (cf. Connell 1971, Furth 1980). Despite this convergence no general model of the operative structures bearing on social relations has yet emerged. It raises the suspicion that the difficulties encountered in pursuing research in social cognition may well be located in the conceptual basis on which that research has been undertaken and the strategy with which it has been pursued. What conclusions, then, can be drawn from these reflections? It is not simply a question of acknowledging that research in this area needs to be concerned as much with theoretical issues as with practical ones. Nor is it necessarily the case that a new type of method is required, a new paradigm for studying social cognition. It is after all concerned with the development of children’s thinking about

Children’s understanding of friendship 25 social life and the methodological question remains the same, namely to find a way in which the child may be enabled to articulate his thinking. Rather it seems that the primary conclusion to be drawn concerns the way in which any practical research is situated within a theoretical context. Thus in analysing these interviews the notion of friendship emerged not as an independent structure in the child’s thinking but as one aspect of the child’s understanding of social relations in general. Notions of friendship appeared infused with other aspects of the child’s thinking about social relations; as well as the correspondencies already noted with research on person perception there were also examples of role-taking and of moral judgments being made as the children spoke about friendship. In this sense thinking about friendship needs to be placed in the context of the development of those operative structures which bear on the child’s comprehension of his social world. It seems clear that friendship cannot be analysed as an independent ‘domain’ of social thought, that to do so would be to hypostatize friendship by abstracting this notion from the context in which it is embedded. A more fruitful approach would seem to be one which is based on a recognition that all of these aspects of social thought are indeed inter-related and inter-dependent. The assumption here is that there are indeed some underlying, general conceptions of social relations and of the kind of beings implied in these relations which develop through childhood and that these conceptions may be accessed through a discussion of a variety of topic areas. No doubt each topic will approach these underlying conceptions from a particular perspective (each of them presents as it were, a different aspect of and on the underlying conceptions) and will thus highlight some issues while leaving others rather more opaque. Yet each of these topics is, as it were, circling around a central core.

References Chandler, M.J. (1977). Social Cognition: A selective review of current research. In W.F. Overton & J.M. Gallagher (eds.), Knowledge and Development, vol. 1. New York: Plenum Press. Connell, R.W. (1971). The Child’s Construction of Politics. Victoria: Melbourne University Press. Damon, W. (1977). The Social World of the Child. San Francisco: Jossey-Bass. Doise, W. (1978). Groups and Individuals. Cambridge: Cambridge University Press. Duckworth, E. (1974). The Having of Wonderful Ideas. In M. Schwebel and J. Raph (eds.), Piaget in the Classroom. London: Routledge. Furth, H.G. (1980). The World of Grown-ups. New York: Elsevier. Kohlberg, L. (1969). Stage and Sequence: The cognitive-developmental approach to socialization. In D.A. Goslin (ed.), Handbook of Socialization Theory and Research. Chicago: Rand McNally. Lives1ey, W.J. and Brom1ey, D.B. (1973). Person Perception in Childhood and Adolescence. London: Wiley. Perret-Clermont, A.-N. (1980). Social Interaction and Cognitive Development in Children. London: Academic Press. Piaget, J. (1932). The Moral Judgment of the Child. London: Routledge & Kegan Paul.

26


Piaget, J. (1972). Insights and Illusions of Philosophy. New York: World Publishing Company. Piaget, J. (1977). The Grasp of Consciousness. London: Routledge. Piaget, J. (1978). The Development of Thought. Oxford: Basil Blackwell. Selman, R.L. (1980). The Growth of the Interpersonal Understanding: Developmental and Clinical Analyses. New York: Academic Press. Shantz, C.U. (1975). The Development of Social Cognition. In E.M. Hetherington (ed.), Review of Child Development Research, Vol. 5. Chicago: University of Chicago Press.

2

The child’s re-construction of economics Gerard Duveen

1 Introduction It may appear somewhat strange at first to consider children’s understanding of economics in an ethnographic perspective. Most ethnographic research with children has concentrated on the analysis of themes associated with face-to-face interaction. Even when collective concepts such as friendship have been investigated ethnographically, they have been examined in so far as they figure within the nexus of face-to-face interaction. From this point of view the sphere of economics seems remote from the social world of children’s everyday lives, and an unlikely theme for the expression of children’s spontaneous attempts at understanding the world in which they live. Rather, a knowledge of economics appears to be the product of direct or indirect instruction, and hence as something elaborated through the course of schooling. One aim of this paper is to argue that such is not in fact the case, but that, on the contrary, from a very early age children are actively engaged in re-constructing an understanding of such economic aspects of society as the buying and selling which they see going on around them in shops, the sources and functions of money, the nature of work and the productive process, such institutions as the bank, and the conditions of wealth and poverty. Perhaps the best way to begin this discussion is to give you an in vivo example of young children’s engagement with these themes. This anecdote was told to me by the parents of a young boy aged 2 years 9 months who knew of my interest in this topic. One morning Jack was somewhat unhappily watching his father get ready to go to work when he announced that: “Mens go to work, mummys don’t”. This assertion of a sexual division of labour was actually an affront not only to the liberal conscience and ideology of his parents, but also to their practice. His mother does in fact go out to work, and, somewhat outraged by her son’s precocious sexism, she began to explain to him that things were not really like this, that she herself went out to work, that when she went to the study and sat at the typewriter that, too, was work, and that when she and his father cooked and cleaned the house that was also a kind of work. Jack, however, remained unmoved by these assertions of liberal ideology. His response to his mother’s patient explanations was to note that: “If no-one goes to work then it’s Saturday”.

28


There is much of interest in this anecdote. Notice, for instance, how Jack formulates his knowledge: “Mens go to work, mummys don’t” so that while the male gender is a substantive presence, the female gender has been completely absorbed into the social role of mothering. It is not women who don’t work but mummys. Jack’s association of men with the world of work and women with the domestic life illustrates a theme which also emerged in our own research with young children which I shall describe later. There is also something else in Jack’s expression relevant to my argument. One kind of explanation for the transmission of representations of social life would assign a dominant influence to familial sources. Children, in these terms, could be expected to reproduce the ideas and values of their parents. But Jack has done precisely the opposite and asserted a correlation between gender distinctions and the social division of labour which contradicts the values asserted and practised by his parents. He seems to be responding to influences which come from outside the family, but which are also powerful enough to negate the active influence of his parents. Indeed, when his mother remonstrated with him about his representations of economic life he remained unmoved by her attempt at exercising some social influence. His response was tangential, as though he was not able to absorb the information she was presenting to him. Against his parents’ minority views Jack upholds the dominant view of the majority. Even very young children, it seems, are in touch with aspects of society from which they seem, at first sight, to be remote. A consideration of children’s understanding of economics serves, then, as a reminder that children’s social worlds extend across a broader range than just their immediate face-to-face interactions. But this theme also possesses another significance for ethnographic research in so far as it focuses attention on the influence of collective representations on the elaboration of children’s understanding of social life. Again, this is a theme to which I shall return in considering our own research with young children.

2 Research on children’s understanding of economics In preparing this paper I have assumed that research on children’s understanding of economics is not widely known. I shall, therefore, say a few words about this research before going on to describe our own work. Two distinct foci can be identified in this research on children’s reconstruction of economics. Firstly, there are studies which have attempted to explicate the developmental course of children’s understanding of economic relations. The psychological model informing these studies has been that of cognitive development, Piagetian or otherwise. By and large this work has concentrated on the emergence of children’s understanding of transactions of various kinds. Since most developmental psychologists are also ethnographers of a kind (at least they have a naive appreciation of the ethnography of their culture) they have usually begun their investigations by considering those transactions which are ethnographically closest to children’s experiences, buying and selling in shops. From this point they have extended the range of their inquiries

The child’s re-construction of economics 29 to consider children’s understanding of the chains of distribution and production which result in the availability of goods in the shops. Within this framework there have also been particular studies of children’s understanding of work, money, the bank, and the notion of profit. The investigation of this broad network of economic relations has yielded some clear descriptions of the development of children’s understanding of a range of economic actors and the relations between them. The first systematic studies in this area were those undertaken by Anselm Strauss (1952, 1954) in the early 1950’s. He interviewed children from preschool ages through to adolescence about such issues as buying and selling, and the sources and uses of money. (Strauss was the first to describe the belief of young children that change given by shopkeepers provides a source of money for people, a belief which has been noted in every subsequent similar study.) Although some other studies were published in the later 1950’s and 1960’s (e.g. Danziger 1958; Sutton 1962), it was only with the more recent work of Hans Furth (Furth 1980; Furth, Baur and Smith 1976) and Gustav Jahoda (1979, 1981, 1983, 1984a, 1984b) that this theme received any sustained attention. The most extensive series of studies undertaken so far has been the work of two Italian psychologists Anna Emilia Berti and Anna Sylvia Bombi, whose book on The Child’s Construction of Economics will shortly be published in an English translation. In recent years there has been a rapid expansion of work in this area, so that there is now quite an extensive bibliography which extends beyond research on urban Western children to include studies of children living in different social and cultural environments where economic relations may have a different social organisation, or a different significance in daily life (reviews of this literature have recently been published by Stacey 1982, 1985; and by Berti and Bombi 1988). The second focus for research has been children’s images of economic realities in so far as these are revealed through their judgements, ratings or estimations of such things as relative rates of pay for different occupations, or the distribution of wealth through society. For this group of studies, which has been less numerous than the first, the developmental perspective has been supplemented as a psychological model with an interest in processes of social influence, of the communication of knowledge about society through the medium of social representations. These studies have not been directly concerned with how children acquire a knowledge of economics in the sense of a more or less objective understanding of economic relations. Children, especially when they are young, have little direct contact with economic activities. What contact they do have with the world of economics comes, rather, through the social representatons of economics which circulate within their milieux. The issue for this group of studies, then, is to examine how children come to reconstruct these social representations. Economic socialisation, therefore, can be seen from this point of view as a suitable arena in which to raise questions about the social transmission of knowledge. The contrast between these two foci is not simply between research which seeks to give an exclusively developmental account of economic socialisation and research which recognises social influences in this process. Within the first group there are studies which have examined social influences, both at a macro level

30


(variations in development as a function of social position; e.g. Jahoda 1983; Hong Kwan & Stacey 1981), and a micro level (the influence of social interaction on the development of economic concepts; e.g. Berti & Bombi 1988). At the same time studies within the second group have also considered developmental issues (e.g. Emler & Dickinson 1985; Kourilsky 1981; and also the research reported in Section 4 below). Nevertheless an important contrast can be drawn between these two groups of studies. The first group emphasises the cognitive aspects of economic representations, being largely concerned with questions about the structural organisation of children’s understanding of economics. The second group are characterised by a dominant interest in the evaluative, judgemental and attitudinal aspects of economic representations. The contrast between these two approaches can also be characterised in terms of the different levels of explanation which Doise (1986) identifies in social psychological theory. The four levels which Doise identifies are: (i) the intrapersonal, concerned with intra-personal processes; (ii) the inter-personal and situational, concerned with the immediate effects of the dynamics of inter-personal relations within given situations irrespective of the social position of the participants; (iii) the positional level, concerned with the effects of subjects’ social position on their behaviour; and (iv) the ideological level, which concerns the influence of collective systems of belief and representation. Cognitive-developmental theory is articulated almost exclusively at the first of these levels, that of intrapersonal processes, and thus the first group of studies of economic socialisation are primarily concerned with analyses at Doise’s Level 1, though some studies have also examined the influence of social position on intra-personal development (Level 3, or, more specifically, articulations of Level 3 with Level 1). In contrast, the importance of social representations for the second group of studies indicates a focus on analyses at Doise’s Level 4, the ideological level, and articulations of this level of analysis with intra-personal and positional levels. The following two sections present examples of these two groups of studies. In Section 3 I present a very brief synthesis of some of the main lines emerging from Berti and Bombi’s work as an example of the first group of studies. Section 4 presents some data drawn from my own research with Maureen Shields as an example of the second group.1

3 The development of economic understanding In their Book Berti and Bombi (1988) present a developmental analysis of the differentiation of economic relations as a distinct domain of social life through childhood and into adolescence. A brief sketch of their analysis will provide a good example of the kind of work typical of the first group of studies described above. Their analysis identifies four distinct developmental levels2, which they identify as follows: 1. Conceptions of the Pre-Operatory Period (3–6 years). Pre-school children’s knowledge of economics is largely restricted to the identification of regularities

The child’s re-construction of economics 31 within those situations of which they have first-hand experience. Their representations of economics consist, therefore, of isolated, unco-ordinated bits of information. By the age of 4 or 5 most children have elaborated a kind of “script” to describe and connect the sequence of actions which occurs in shops: asking the shopkeeper for goods and receiving them, followed by the giving and receiving of money. But this script is not generalised to every type of buying and selling. Not all goods come to be recognised as such, only those which can be incorporated within this “shop script” are construed as capable of being bought and sold. Other things, such as houses, land or cows, are not recognised as commodities, though children do recognise that such things can be “owned”. However, at this stage ownership is generally identified according to the criterion of use so that something is held to belong to someone because they use it, however transitory that use may be. Thus, a bus may be said to belong to the passengers because they get on it. The relationship between work and payment at this stage has more the character of convention or ritual than any clear understanding of remuneration. One works and one gets paid, rather than one receives money because one works. As well as identifying work as a source of money, these children also speak about a number of other sources as though they were equivalent. The bank, for instance, distributes money to anyone who asks for it, or people can acquire money through the change given to them in shops. Complementing this notion of the sources of money is a very limited understanding of the notion of wealth. Whoever has money is thought to be “rich”, and since anybody can obtain money from these various sources everybody can be “rich”, even if the experience is only momentary, lasting only as long as they actually remain in possession of this money. Children at this age have practically no ideas about production. They believe that shopkeepers themselves make the goods they sell, or get them from other shopkeepers, who get them from others, and so on in an endless regression. Overall, the ideas of this early stage can be described as pre-economic, since the recognition that goods and money derive from extra-domestic sources is associated with the conviction that money and goods are available without restriction. Within this system only two kinds of roles are identified: “distributors” of goods, services and money on the one hand, and “customers” or “consumers” on the other, exchanges between these two roles are construed in a script-like form, without any understanding of a correspondence in value between goods or the work done and payments made for them. 2. Conceptions of the Intuitive Level (6–7 years). A second stage can be identified when children begin to construct rules which allow them to create qualitative correspondences between the prices of objects and the amounts paid to buy them, or between work done and its remuneration. Children are now able to discriminate the values of various types of money, and to order things in a series from the least to the most expensive. Variations in prices between things are related to their most striking characteristics. Each price is made to correspond to a particular

32


denomination of notes or coins, so that buying and selling represents an exchange of equivalents, rather than being merely a ritual. However, since this correspondence is not yet based on any extensive quantification children are still not able to interpret the function of change correctly. A distinction between work and other ways of obtaining money also begins to emerge at this stage. Where children can identify a “consumer” of work, they construe the activity as something undertaken for their benefit and for which they pay. In other cases, however, they continue to construe payment for work as a kind of ritual. An understanding of work as employment continues to present difficulties to children at this stage, largely because they do not recognise that the means of production can belong to anybody other than the people who use them. Possession of the appropriate tools is seen as the principle means of access to particular jobs. In spite of these limitations the notion of work shows some elaboration from the earlier stage. Goods as well as money have their origin in work. In addition, only someone who works can be “rich”. Poverty, by contrast, remains a marginal phenomenon; only the old, the sick and the lazy are poor because they do not work and, hence, cannot earn any money. One other advance over the earlier stage is that children are now aware of the sphere of production as well as that of distribution. In many cases, however, they attribute both activities to a single “producer-distributor” figure. 3. Conceptions of the Concrete Operatory Period (7–10 years). It is at this stage that the pre-economic ideas of the earlier levels are replaced by a wider and more articulated understanding of buying and selling on the one hand, and of work on the other. This is reflected in the way in which a number of the themes from earlier stages are elaborated at this level. The establishment of quantitative correspondences between the price of something and the money used to pay for it means that children are now able to grasp the function of change correctly. Work comes to be seen as the sole source of money, and there is a finer discrimination of economic strata. Some people are said to be paid more than others because they work longer hours, or their work is more tiring. Children at this level also have a clearer understanding of production. Factories are where goods are made, and children now also recognise hierarchies of authority within the factory. It is the boss who owns both the factory and the tools, who pays the workers and to whom people must apply for a job. The sphere of buying and selling is no longer limited to the exchange of goods for money in shops, but extended to include non-transportable goods such as houses or land. The shopkeeper is also clearly differentiated from the producer, intermediate commercial figures, such as transporters and wholesalers, begin to appear. However, children do not yet understand that the price of goods is based on the costs of production (including the cost of labour). They continue to believe that price is a characteristic of the goods themselves, related to their phenomenal attributes. Hence, children at this stage also believe that the price of something remains invariant through the course of a series of commercial transactions.

The child’s re-construction of economics 33 In spite of the developments at this level, children’s understanding remains limited by the fact that it consists of two as yet unco-ordinated areas. There is the system of work which comprises bosses and workers, and in which payments are made in recompense for activities undertaken. There is also the system of buying and selling, composed of consumers, shopkeepers, middlemen and producers (at least in so far as they also sell their goods). In this system payment is made in exchange for goods. Thus the relations of exchange which constitute the connecting links within each system are different in the two areas. Money is exchanged for work in one case and for goods in the other. Children do not yet know how to find any common denominator between these two systems of exchange. 4. Conceptions of the Formal Operatory Period (11–14 years). It is at this stage that children begin to co-ordinate the different systems of exchange within a single framework. They recognise that the owner of a factory pays the workers with money received from the sale of the factory’s products. The price of goods is now construed as reflecting the costs of materials, labour and the various intermediate commercial transactions between production and consumption. In addition, children also recognise that prices include profit margins for each of the various economic figures. Children’s vision of economic stratification also broadens through the recognition that different rates of pay for different jobs are not related only to the time and effort expended in different jobs, or their different social utilities, but also to the prestige of different occupations. Thus children come to have some sense of social groups, or classes, as being structural features in the socio-economic organisation of society. At this level children also understand the functions of various institutions more clearly. They know, for instance, that the money which banks use to make loans comes from deposits, and that the interest from loans constitutes the source of income from which the bank obtains the money to pay interest on deposits. The functions of public institutions also become clearer. Words such as “Council”, “Government” or “State” are no longer simply assimilated to a very general idea of someone who commands resources or provides services for other people; they now denote institutions which provide collective services.

4 The social transmission of economic representations Economic representations do not develop within a conceptual void. Rather children acquire an understanding of economics within the context of their representations of social life. A principal aspect of these representations concerns the theme of gender, and the study reported here concerns the influence of gender on the development of young children’s economic understanding. In this context gender refers both to an ideology, that is to those social representations of gender current in the community within which the child develops, as well as to social categories, that is the child’s social position as male or female. This study, therefore, concerns the

34


articulations of Doise’s Level 4 (ideology) with Level 3 (social position) and Level 1 (intra-personal). In acquiring economic representations children are gaining access to a set of social representations held by the community in which they live. This acquisition can be construed, therefore, as a process of social transmission, and it should be possible to examine the influence of different levels of analysis on this development. Specifically, this study sought to explore the influence of intra-personal, developmental processes (Level 1) and social categorisation effects (Level 3 – social position) on the acquisition of economic representations in preschool children. To achieve this aim, the representations of these children are compared with the corresponding adult representations. The study is limited in so far as data were only collected from children, so that we have had to rely on other sources for a sense of the corresponding adult representations. 4.1 Subjects and method All of the data reported below are drawn from a series of interviews undertaken with children aged 3 to 5 years from four inner London nursery and primary schools. Altogether 110 children were interviewed, divided into three age groups with mean ages of roughly 3½, 4½ and 5½ years. The age and gender distribution of the subjects is shown in Table 2.1. The interviews were conducted in quiet areas adjacent to the classrooms and consisted of a series of questions built around a variety of tasks. The tasks are described in the following sections with the corresponding data. 4.2 Relationship between child and adult representations 4.2.1 Similarities The simplest relationship to envisage is one of similarity, where children’s representations resemble those of the adult community and where, at least across the age range considered in this study, similarity is not subject to developmental variation. 4 . 2 . 1 . 1 M A L E A ND F E MAL E ACCE S S T O OCCUP AT IO N A L RO LES

An example of such a relation can be seen in the distinction which children made between domestic and occupational settings in relation to male and female access Table 2.1 Age and gender distribution of subjects Age group Gender

3½ Girls

Boys

4½ Girls

Boys

5½ Girls

Boys

N Mean age

10 3.8

16 3.7

23 4.6

20 4.6

22 5.5

19 5.6

The child’s re-construction of economics 35 to social roles. For this task children were shown five photographs of men in occupational roles (doctor, police, shopwork, farmer and factory work) and three photographs of women engaged in domestic activities (feeding a baby, sweeping, washing up). For each photograph children were asked “What is this person doing?” and then “Can a lady do this job?” or “Can a man do this as well?” as appropriate. Responses were divided into two categories. Either children said that the opposite gender could perform the task – equal access – or they maintained a restriction on access to one gender only – restricted access. The same analytic frame was used to sum children’s responses to produce an occupational access and a domestic access score. In each case children were given scores of 1 if they made equal access judgements to each of the figures in the appropriate setting, or a score of 0 if they made a restrictive access judgement to any of the figures. These scores were then compared in a repeated measures analysis of variance with age group and gender as between-subject variables. The only significant effect was for the access scores which showed children of all age groups and both genders more readily granted equal access in domestic settings (mean = 0.51) than in occupational settings (mean = 0.35; F = 7.46; df = 1, 103; .01>p). Thus, even children as young as three year olds distinguish between occupational and domestic settings in terms of the access to roles which they judge to be appropriate to men and women in these settings. Associating men with the world of work and women with the domestic sphere is, of course, a pervasive feature of social representations of gender, so that it may not be altogether surprising that the distinction is reproduced by such young children, nevertheless, it is worth noting that there is, apparently, neither developmental nor social variation associated with the transmission of this distinction. A closer examination of children’s judgements about the occupational roles in this task emphasises the similarity with adult conceptions. Responses that only men could undertake these roles were not distributed evenly across all five occupations. Table 2.2 lists the rank order of “men only” attributions for these five occupations; these rankings did not vary as a function of age and gender. Table 2.2 Rank order of male restricted attributions to occupations Rank

Occupation

% “men only” attributions

% women in the labour force*

1 2 3 4 5

Farmer Factory worker Police Doctor Shop worker

43.0 41.5 30.3 21.3 17.6

1.3 5.0 9.8 63.5 59.9

* Data taken from Murgatroyd (1982), who derived her figures from the 1979 Labour Force Survey, which groups occupations into much wider categories than the photographic stimuli. In particular the entry for ‘doctor’ is taken from Murgatroyd’s category of ‘professional and related in welfare and health’. Other estimates for the proportion of women doctors are much lower. Mackie & Pattullo (1977), for example, quote a figure of 27%.

36


What is particularly interesting about this rank order is that it corresponds in large measure to the ranking of these occupations in terms of the actual proportion of women in the labour force in these occupations (cf. the second column in Table 2.2). Of course these data do not refer to the representations of female participation in the economy held by the adult community, but to the actual number of women in these occupational categories as estimated by government survey. Nevertheless, when one considers those public representations available to young children, in comics or books or t.v. etc., it remains the case that women are rarely portrayed as being farmers or factory workers, while they may be observed more frequently in the roles of shop workers or doctors. In short, children by the age of three appear to have already reproduced within their representation of economic life the marking of occupations for gender. It is this sense of the straightforward reproduction of adult representations by even very young children which is characteristic of this kind of similarity relations between adult and child representations, as though there were no mediation in the process of transmission between the one and the other. Theoretically these examples would correspond to a view of the transmission process as simple reproduction, that is, that children’s understanding reproduces adult representations independently of any developmental process or social variation. 4.2.2 Differences More complex examples of transmission are those in which children’s representations differ significantly from those of adults. In these cases it can be assumed that there is some developmental process occurring within the age range considered in this study. One might expect that children’s representations of economic life would progressively come to resemble those of adults. This view is essentially a simple developmental extension of the notion of transmission as simple reproduction. It construes the development of economic representations as the accretion of items of economic knowledge by the child and would predict a simple developmental function in which, with age, children’s representations progressively come to resemble those of adults. In point of fact the data collected in this research provided no examples of such straightforward developmental effects. There were age effects in the analysis of the data, but these were either main effects in which children’s representations became less veridical as they became older, or interaction terms in which the effect for age was modulated by variations due to gender. 4. 2 . 2 . 1 D E V E L O P ME NT AL RE P RODUCT I ON – RE L A TIV E PA Y

An example in which a main effect shows children’s judgements becoming less veridical with age is a task concerned with the relative pay of men and women for doing the same job. Pairs of photographs showing men and women in six occupations (doctor, police, shopkeeper, farmer, factory worker and teacher) were shown

The child’s re-construction of economics 37 to children who were asked whether the man and woman were paid the same or if one was paid more than the other. Responses were assigned to four categories: (i) Don’t know, or not paid; (ii) men paid more; (iii) women paid more; and (iv) both men and women paid the same. Responses across all six occupations were summed within these categories to produce a score of between 0 and 6 for each type of response. A two-way (Age × Gender) analysis of variance of the ‘same’ responses produced only an age main effect (F = 7.59; df = 2, 104; p <. 001) which showed that the proportion of ‘same’ judgements increased with age (mean score at 3 years = 1.31; at 4 years = 2.42; at 5 years = 3.32). We have discussed elsewhere (Duveen & Shields 1984, 1985) the interpretation of this result as a reflection of children’s increasing usage of schema of equality which may become tempered in older children as inequalities become legitimated (cf. Leahy 1983). The relevance of this result for the present discussion is that it indicates that children’s representations of economic life are mediated through developmental processes. Theoretically this interpretation corresponds to a view of transmission as a process of developmental reproduction. 4 . 2 . 2 . 2 S O C I A L R E PRODUCT I ON

As well as developmental variations in children’s representations a number of social variations were also observed as a function of gender. In this context gender refers both to children’s own gender as well as to the gender marking of social roles and activities, indeed significant effects were frequently due to the interaction of these two parameters. One example of this type of effect can be seen in the analysis of the ‘men paid more’ and ‘women paid more’ responses to the relative pay task described in the previous section. Scores in these two categories were analysed as a repeated measures variable in a three-way ANOVA with age and child’s gender as the between-subjects variables. The results showed both a significant responsecategory main effect as well as a response-category by child’s gender interaction. The main effect (F = 25.19; df = 1, 104; p <. 001) showed that men were generally judged to be paid more (mean = 1.68) than were women (mean = 0.63). The interaction term (F = 10.05; df = 1, 104; p <.01) shows that this difference is much more exaggerated in boys than it is in girls (see the mean scores in Table 2.3), thus providing an example of the influence of social position in the transmission of economic representations. Table 2.3 Pay differential responses broken down by gender

Girls Boys

Mean numer of ‘men more’ responses

Mean number of ‘women more’ responses

1.24 2.13

0.84 0.42

38


A second example occurs in a task concerned with role multiplicity. Children were shown photographs of four occupational roles, two illustrated by male figures (doctor and policeman) and two illustrated by female figures (bus conductress and factory worker). For each photograph the child was asked if the figure had a home; where they ate and slept; whether they could be a parent; and if they undertook domestic chores. Responses were separated into two categories. If the child said that the figure had a home, ate and slept at home, could be a parent and did housework, this was judged to demonstrate a recognition of role multiplicity and awarded a score of 1. If the child made any exceptions to these points their responses were construed as lacking a sense of role multiplicity and given a score of 0. Responses were summed within gender categories so that each child was assigned two scores ranging from 0 to 2, one for the female figures and one for the male figures. A three-way ANOVA, Age by Gender by Figures, with repeated measures on the Figures factor (Female Figures and Male Figures), produced a significant main effect for Figures and a Figures by Gender interaction. The main effect (F = 46.86; df = 1, 101; p <.001) indicated that female figures were more frequently construed as occupying multiple roles (mean = 1.25) than were male figures (mean = 0.68). Again this effect varied as a function of the child’s own gender. In this case, however, the effect was more strongly marked among girls than among boys (see mean scores in Table 2.4). In both these cases the representations expressed by children vary as a function of their membership of a gender group. This gender differentation indicates one way in which the transmission of economic representations is a socio-psychological process, for what these children are asserting in their judgements is a particular social identity centered on gender. Elsewhere (Duveen & Lloyd 1986) we have described the development of social gender identities as the process through which children come to situate themselves in relations to social representations of gender. It is within this framework that economic notions are also being elaborated.

5 Conclusions The results reported in the previous section illustrate very clearly the influence of social representations of gender on the transmission of economic concepts (an ideological, Level 4 effect in Doise’s term). In itself this ideological influence is unremarkable given that the articulation of gender and occupation remains firmly

Table 2.4 Mean role multiplicity scores

Girls Boys

Female figures

Male figures

1.52 1.00

0.73 0.64

The child’s re-construction of economics 39 anchored within the social representations of adults. What is more interesting are the two other types of influences shown by these results. On the one hand there are developmental processes which limit the cognitive possibilities open to the child for construing economic life. On the other hand their membership of social groups, specifically gender groups, also acts as a limit on the emergence of economic representations. The first set of influences correspond to intra-personal level of analysis (Doise’s Level 1), while the second set correspond to influences arising from social position (Doise’s Level 3). Taking account of both of these influences becomes possible if we construe the development of social representations of economic life as the acquisition of particular social identities (in Doise’s terms, what is necessary here is a theory capable of articulating the ideological Level 4 with the influences of social position – Level 3 – and intra-personal processes – Level 1). Thus it is not simply differences between social groups which need to be considered in relation to variations in children’s representations of economic life. In the terms of the perspective outlined here it is variations in social identities which are crucial. A particular issue where further research is necessary is the identification of those criteria which lead to differentiated social identities in the sphere of economic life. For example, in Gustav Jahoda’s study of the development of economic concepts in Zimbabwean children (Jahoda 1983), their precocity over European children was associated with their closer engagement with the activities of setting prices in real markets. A European child’s experience of buying and selling is typically limited to the perspective of being a consumer in well regulated shops where prices appear as a fixed attribute of goods for sale. These two sets of circumstances generate two distinct social identities. It may well be that cross-national comparisons are not associated with differentiated social identities if the actual social groups investigated share a more or less similar relation to the field of economic activity. Conversely, comparisons of divergent social groups within nationalities may prove to be of interest. The complexity here arises from the fact that in analyses of the effects of social position (Doise’s Level 3) the definition of independent variables cannot be taken for granted. As Emler and Dickinson (1985) note, for instance, the variable of social class defined through parental occupation alone may not in fact specify distinctive social representations. In our own research categorical differentiation in the development of economic representations was established on the criterion of gender. Social representations of gender present the categories of ‘male’ and ‘female’ as set in distinct relations to economic life as well as establishing male and female as distinctive social identities. The co-ordination of these two distinctions underlies the interactions between the child’s own gender and the social representations of gender as these are instantiated in the social marking of roles, occupations and activities. Social position variables must themselves be defined sociopsychologically in terms of distinct social identities, rather than assuming that social status variables automatically define distinctive social positions. One final comment concerns the distinction between the two groups of studies of economic socialisation and their characterisation in terms of the way in which each of them presents a particular articulation of the levels of explanation

40


identified by Doise. This characterisation has helped to clarify the multiplicity of influences which need to be considered in the analysis of economic socialisation. From a cognitive-developmental perspective what is unusual in considering economic socialisation as a process of social transmission is that not all children elaborate similar conceptions. A cognitive-developmental perspective is a good basis for grasping communalities in the development of children’s understanding, but an insufficient one for investigating variations between children which arise as a function of social position. Trying to account for both developmental and social variations within a unified theoretical framework is the source of much of the complexity surrounding this issue.

Notes 1 This work was supported by a grant from the Social Science research Council (Grant no, HR C 0023 0027). 2 These levels reflect a Piagetian analysis of cognitive development, though these authors have also been influenced by Case’s (1985) reformulations of Piaget.

References Berti, A.E. and Bombi, A.S. (1988) The Child’s Construction of Economics. Cambridge: Cambridge University Press. (Translated by Gerard Duveen from A.E. Berti and A.S. Bombi, Il mondo economico nell bambino. Florence: La Nuova Italia.) Case, R. (1985) Intellectual Development from Birth to Adulthood. New York: Academic Press. Danziger, K. (1958) Children’s earliest conceptions of economic relationships (Australia). Journal of Social Psychology, 47, 231–240. Doise, W. (1986) Levels of Explanation in Social Psychology. Cambridge: Cambridge University Press. Duveen, G.M. & Lloyd, B.B. (1986) The Significance of Social Identities. British Journal of Social Psychology, 25, 219–230. Duveen, G.M. & Shields, M.M. (1984) The influence of gender on the development of young children’s representations of work roles. Paper presented to the First European Conference on Developmental Psychology, Groningen. Duveen, G.M. & Shields, M.M. (1985) Children’s ideas about work, wages and social rank. Poster for the International Society for the Study of Behavioural Development meeting at Tours. Emler, N. and Dickinson, J. (1985) Children’s representations of economic inequalities: The effect of social class. British Journal of Developmental Psychology, 3, 191–198. Furth, H. (1980) The World of Grown-ups. New York: Elsevier. Furth, H., Baur, M. and Smith, J.E. (1976) Children’s conception of social institutions: A Piagetian framework. Human Development, 19, 351–374. Hong Kwan, T. and Stacey, B. (1981) The understanding of socio-economic concepts in Malaysian Chinese school children. Child Study Journal, 11, 33–49. Jahoda, G. (1979) The construction of economic reality by some Glaswegian children. European Journal of Social Psychology, 9, 115–127. Jahoda, G. (1981) The development of thinking about economic institutions: The bank. Cahiers de Psychologie Cognitive, 1, 55–73.

The child’s re-construction of economics 41 Jahoda, G. (1983) European ‘lag’ in the development of an economic concept: A study in Zimbabwe. British Journal of Developmental Psychology, 1, 113–120. Jahoda, G. (1984a) The development of thinking about socio-economic systems. In H. Tajfel (Ed.) The Social Dimension. Cambridge: Cambridge University Press. Jahoda, G. (1984b) Levels of social and logico-mathematical thinking: their nature and interrelations. In W. Doise and A. Palmonari (Eds.) Social Interaction in Individual Development. Cambridge: Cambridge University Press. Kourilsky, M. (1981) Economic socialization of children: Attitude toward the distribution of rewards. Journal of Social Psychology, 115, 45–57. Leahy, R. (1983) The development of the conception of social class. In R. Leahy (Ed.) The Child’s Construction of Social Inequality. New York: Academic Press. Mackie, L. & Pattullo, P. (1977) Women at Work. London: Tavistock. Murgatroyd, L. (1982) Gender and occupational stratification. Sociological Review, 30, 574–602. Stacey, B. (1982) Economic socialisation in the pre-adult years. British Journal of Social Psychology, 21, 159–173. Stacey, B. (1985) Economic socialisation. Annual Review of Political Science. Norwood, N.J.: Ablex. Strauss, A. (1952) The development and transformation of monetary meaning in the child. American Sociological Review, 17, 275–286. Strauss, A. (1954) The development of conceptions of rules in children. Child Development, 25, 193–208. Suttón, R.S. (1962) Behavior in the attainment of economic concepts. Journal of Psychology, 5, 37–46.

3

Piaget ethnographer1 Gerard Duveen

1 Introduction It might at first sight appear odd to introduce a discussion of Jean Piaget into a consideration of qualitative methods, and perhaps even odder to characterise his work as ethnographic. However, my intention in this essay is to do precisely this and to suggest that Piaget’s work provides a paradigmatic example of qualitative, or more broadly interpretive, research (an example, and not the example, since qualitative methods have always embraced a pluralism). While this essay focusses on the example of Piaget, it is also the case that the renewed interest in qualitative methods in psychology could be further enriched by sustained attempts to recover the significant models offered by earlier generations of psychologists. Even within the field of developmental psychology, however, Piaget’s contribution to a qualitative methodology has been eclipsed. This has been the case even where such a contribution might have been expected to have a continuing resonance, for example, the recent collection on Ethnography and Human Development (Jessor, Colby and Shweder, 1996) does not include any reference to Piaget. Two reasons for this eclipse can be identified fairly easily. Firstly, the rise of Vygotskyian or “socio-cultural” approaches have emphasised the significance of concrete contexts for the analysis of human development, a perspective which has been seen in contradistinction to what has been viewed as the abstract and decontextualised account found in Piaget’s analysis of cognitive development. As I have argued elsewhere (Duveen, 1997), such a way of contrasting Piagetian and Vygotskyian perspectives is misleading, since it undervalues Piaget’s engagement with social processes and overvalues the extent of Vygotsky’s account of the influence of social life on human development. However, for the present essay it is sufficient to note that there has been a general tendency for the Vygotskyian perspectives to displace Piagetian perspectives in recent developmental psychology. The second reason for the eclipse of Piaget’s methodological contribution is more germane to the theme of this paper, since it refers to Piaget’s methods themselves which have been the specific object of severe critical attention. The most notable critics have been experimentalists who have tended to suggest that what might be considered the most characteristically Piagetian aspects of

Piaget ethnographer 43 children’s developing thinking are, in fact, no more than artifacts of his inadequate methods, and that more stringent experimental controls have revealed greater consistency and continuity in the developmental process than Piaget’s insistence on discontinuities and clearly demarcated cognitive stages. Such a view rests upon a profound misapprehension about Piaget’s method. Indeed it is a misapprehension which is at the centre of my concern, since it relates to the logic of Piaget’s method and it is this logic which I suggest can be characterised as ethnographic. Ethnographic as a term to characterise a research program has come to be associated with what Cliff Geertz, quoting Gilbert Ryle, described as “thick description” (Geertz, 1973; see also Duveen and Lloyd, 1993). On the one hand the idea of “thick description” points towards the effort at capturing the intentional structures of the social actors, groups and institutions being described, while on the other it also highlights the interpretive activities of the ethnographer themself. In this way the ethnographic perspective establishes a dialectic between observation and interpretation, between the material collected and the categories employed in rendering it comprehensible. I shall return to the character of ethnography later in the paper, for the moment this brief formulation is sufficient to justify examining Piaget’s work under this rubric. In few other psychological researchers does this dialectic between material and its interpretation stand out so clearly as it does in Piaget’s work, not only in the sense that one can trace this dialectic at work throughout his long career, but also in the way in which in each of his studies he is concerned to present his work to the reader in such a way that he can persuade the reader of the correctness of his interpretation of the material. This “persuasive stance” which Piaget adopts in his writing reflects a sensitivity to the ethnographic situation in which it is not appeals to the weight of external or objective facts which can sustain or justify a particular interpretation, but rather the ethnographer’s ability to convince a reader that the system of categories which they propose are adequate to grasp and render intelligible the social action under investigation. I should add that Piaget never thought that he had left the field of science by pursuing his research through this method. Quite the opposite in fact, he always insisted on the scientific character of his work. In this sense he does not pose a contrast between quantitative and qualitative methods as the touchstone for separating science from some other form of inquiry. Quantitative research too must pass through the medium of interpretation if it is to render its material intelligible. Rather, the operative contrast here is between a view of science as an interpretive endeavour and a purely positivist view.

2 The development of the clinical interview As is well known Piaget came to the study of child psychology through his concern with developing a biological epistemology. His training and formation in natural history and zoology had given him a developmental or genetic perspective so that it seemed altogether consistent to pursue his interest through a study of how knowledge developed in the human child. But as well as providing a broad

44


framework through which he could pose theoretical questions and structure a research program, Piaget’s background also provided him with a methodological approach. Central to this approach was observation and the interpretation of the material this brought to light. Even in his early work in natural history as a teenager in Neuchâtel it was observation and interpretation which were the keys for classifying fossils and other fauna and flora. Again in his doctoral work on molluscs he relied on close observation and classification of different forms and their adaptation to different environments which was the basis for his research. In fact this emphasis on adaptation provides an early glimpse of a second theme which became important for Piaget’s methodological approach to child psychology, namely the sense that it is through studying adaptive activities that underlying structures can be analysed most clearly. Piaget’s first professional encounters with children during the time he spent in Paris in the laboratory of Simon had convinced him that the development of thought in children was a progressive process of changing forms of structure and organisation. If children reason differently from adults, this is not simply a question of ignorance, but rather a consequence of a different kind of mentality, so that for the child psychologist the issue becomes one of investigating and analysing these changing forms of mentality which lead to adult thinking. Here too, of course, the biological perspective he brought with him is evidently present. Living things do not develop from chaos to order, but rather from one form of structural organisation to another. So for Piaget the methodological question when he began to study the children’s thinking in Geneva in the 1920’s was to find a technique which could address questions of structure and organisation as a way of investigating the child’s developing mentality. And he turned first of all to a method he knew from his own training, that of observation. His early works, in fact, seem an extension of his studies of natural history in which children are observed in their everyday habitats (schools) and these observations sorted and classified to reveal the developing structural organisation of their thinking. However, as he noted in his introduction to The Child’s Conception of the World (1926/19292) pure observation is limited in some important ways. Children do not always explicate their understanding explicitly, and it can be hard to distinguish their real convictions from playful attitudes. Thus he began to interview children as a means of exploring their understanding of the world. This was the beginning of what has become known as the “clinical method”, which by the time he began his studies of the development of operations in young children in the late 1930’s had assumed the characteristic form through which it has become familiar. It is interesting to follow the emergence of the clinical method in Piaget’s research (cf Vinh-Bang, 1966; Ginsburg, 1997), both because there is much to be gained by examining this development (a kind of recursive application of Piaget’s own dictum that if we want to understand something we need to understand how it has developed), and because this development shows the extraordinary dialectic between research and theory in the formation of Piaget’s work. One can in fact identify five steps, or stages, in the development of the clinical method, and an

Piaget ethnographer 45 overview of these steps is presented in Table 3.1. These five steps in fact overlap with other more general accounts of the development of Piaget’s work. Montangero and Maurice-Naville (1997), for example, distinguish four main periods in Piaget’s career. A first period (1920’s to early 1930’s) focussed on questions of child mentality and the gradual socialization of thought during which the methods used were observations and the early, verbal, forms of clinical interview (steps 1–3 in Table 3.1). The second period (mid 1930’s to 1945) concerned the beginnings of knowledge, and relied largely on a return to observational methods (step 4 in Table 3.1), while the third period (end of the 1930’s to the end of the 1950’s) focussed on the structural analysis of the formation of the categories of thought and utilised the clinical interview in its mature form (step 5 in Table 3.1). Piaget’s later works on the primacy of operatory structures and general developmental mechanisms did not introduce any new methodological elements into his research. 2.1 Observation In these first works (to which he gave the collective title Studies in Child Logic) Piaget’s method consisted of close observation of children in the classrooms of the school run by the Institut Jean-Jacques Rousseau. These careful transcriptions of children’s speech and conversation were analysed both through linguistic categories (to address questions of the child’s developing grasp of logic) and through a set of functional categories which attempted to analyse the changing character of children’s communicative activities. As Ginsburg (1997), amongst others, has noted these functional categories owed much to the early influence of Freud and psychoanalysis on Piaget’s own developmental scheme, although he characteristically forged his own structural description of a sequence of stages in children’s communicative development running from an initial autistic stage,

Table 3.1 The Development of Piaget’s Research Methods 1. Observation The Language and Thought of the Child (1923/1926) Judgement and Reasoning in the Child (1924/1928) 2. First ‘Clinical’ Interviews (Verbal) The Child’s Representation of the World (1926/1929) The Child’s Conception of Physical Causality (1927/1930) 3. Participant Observation and Clinical Interview The Moral Judgement of the Child (1932/1932) 4. Observations The Origins of Intelligence in the Child (1936/1952) The Construction of Reality in the Child (1937/ 1954) Play, Dreams and Imitation (1945/1951) 5. Clinical Interviews Number (1941/1952), Time (1946/1969), Space (1948/1956), Logic (1955/1964), etc

46


through a transitional period of egocentrism to the sociocentrism achieved in middle childhood. 2.2 First ‘clinical’ interviews (verbal) As I noted earlier, while recognising the strengths of pure observation as a technique, Piaget also came to realise its limitations. In order to expand his investigations to consider a more systematic approach to children’s developing representations of the world he began to use interviews focussed on specific themes which reflected both something of what he had observed in children’s spontaneous observations and his own emerging account of the categories employed by children in constructing their representations. In describing his interviews as “clinical” Piaget drew on a model of the interview techniques employed in psychiatric assessment. Just as the psychiatrist used the interview to diagnose the condition of the patient, the child psychologist could employ an interview to “diagnose” the structure and pattern of children’s mental life. To achieve this the child was presented with a problem and invited to respond, this response itself becoming the point of origin for a more extended conversation in which the child’s thinking was systematically explored by the interviewer through questioning and cross-questioning. In this way the children became (unconscious) informants on their own mental life, and Piaget, as the interviewer, sought to gather as much information as possible through carefully crafted questions adapted to the productions of each individual child. Thus, while the themes and problems remained constant for all the children interviewed, the specific character of the conversation and the questions posed by the interviewer varied from child to child, with each interview taking concrete shape around the responses of the particular child. Piaget was very aware of the problems and difficulties involved in interpreting such interview material, problems which he addressed in his introduction to The Child’s Conception of the World (1926/1929), one of his most overtly methodologically focussed chapters. Here he tried to classify children’s contributions to these discussions by separating responses which were a direct product of the interviewer’s influence on the child, or in which the child engaged only in romance or phantasy, from those responses which reflected the child’s real convictions in the sense of their underlying beliefs about the world as they understood it and represented it. While such distinctions might be analytically clear in a theoretical sense, they are not always easy to apply in specific cases. And indeed there are many instances in the analytic chapters of these books where Piaget addresses just this question in discussing particular responses of individual children. This is obviously a key interpretive moment in such research studies, but its significance should not be overestimated. One could challenge one or another of Piaget’s judgements about whether a particular response was “romancing” by the child or whether it reflected a “liberated” or “spontaneous” conviction (and in general one can note that the extensive excerpts of the interviews given by Piaget in his written text provide the basis for doing so). However, the theoretical structure of the

Piaget ethnographer 47 analyses of children’s representations does not depend on such individual judgements, but rather on the coherence of Piaget’s own system of categories for interpreting the interview material as a whole. There is, of course, a kind of limiting case here. If Piaget’s judgements were always wrong, or systematically biased in some way, then this would indeed seem to challenge not so much the coherence of his interpretive categories, but their applicability to the material he had collected. The categories which Piaget employed in his interpretation focussed on general characterisations of children’s representations in terms of realism, animism and artificialism. In each case he analysed children’s development from some form of non-differentiation (broadly characteristic of the young child’s egocentric thinking) to a more mature differentiation; between self and the external world in the case of realism, animate and inanimate nature in the case of animism, and between human activity and natural causes in the case of artificialism (cf the discussion in Chapman, 1988, pp. 48–50). In establishing these categories, Piaget also drew on the work of Lévy-Bruhl (1926), suggesting a parallel between his account of the “participations” in the magical thinking of so-called primitives and the characteristics of young children’s thinking. For Piaget, though, while the representations which he analysed in young children were no less constitutive of a specific mentality than they were for the French anthropologist, they also needed to be seen from a developmental point of view. Unlike the primitive, the child also went on to develop a different mentality which reflected the emergence of logical thinking and its systematic application. Thus while Piaget drew on Lévy-Bruhl’s descriptions of primitive mentality, in his own analysis he set these forms of thought in a different context, so that his interpretations of the interview material he collected analysed children’s thinking as developing through a series of structurally differentiated stages which led to the emergence of mature adult thought. 2.3 Participant observation and clinical interview Piaget’s work on The Moral Judgement of the Child (1932/1932) represents something of an intermediate stage in the development of the clinical interview. On the one hand it continues the general form established in the earlier works, but it also introduces some new elements into the method while not yet reaching the final form visible in the later stages. In fact the first of the studies reported in the book is where Piaget emerges most clearly as a true ethnographer of children’s worlds through his use of participant observation. In this study he analysed children’s developing grasp of the rules of the game by sitting down with small groups of boys and asking them to teach him to play marbles. The conversations which ensued were again punctuated by Piaget probing children’s representations through carefully crafted questions which enabled him to produce separate analyses of children’s knowledge of rules and of their practice in relation to rules. From the methodological point of view, the novel element in this study is the integration of observations of action within the context of the interview, so that what children do provides as much material for interpretation as what they say.

48


In the subsequent chapters of the book Piaget’s further studies of other aspects of children’s moral judgements also introduced new elements into his research technique. While he continued to employ the same kind of clinical interview, the stimulus material he employed was now no longer a simple question to the child as it had been in his studies of the representation of the world, but rather a dilemma presented to the child through short vignettes. In some ways this development can be seen as the extension of an element which was already present in the earlier interviews. In questioning the child about their representation of the world Piaget had always employed a form of counter-questioning in which he presented the child with a contrary case to their initial argument, a practice which helped to focus on the characteristics of children’s thinking as it emerged in an active engagement with a problem rather than merely as a static reflection. Now, through the use of these vignettes this procedure came to structure the form of the interview as a whole. In asking the child to consider whether a boy who broke one cup accidentally was naughtier than one who broke several cups while doing something he should not have been doing, Piaget could not only follow the twists and turns of children’s thinking as they pondered the question, he could also intervene through his own questions both to clarify what the child meant at any point, and to introduce further aspects of the situation for them to consider. While the interpretive categories which Piaget employed in analysing this material certainly introduced themes specific to the question of moral judgement, they remain within the general developmental framework which he had articulated in his earlier books. That is, they analyse the development of moral judgement from an early egocentric position to a later sociocentric position, but are not yet analysed in terms of the structural descriptions of operations which characterised Piaget’s later work. Thus, while there are many reasons why this remains one of the most fruitful of Piaget’s books, from the methodological point of view it presents a kind of transitional phase in the development of the clinical method. 2.4 Observations Following his early work on the development of children’s thinking, Piaget devoted considerable attention to his studies of sensorimotor development through infancy, and the transition to operatory forms of knowing. In studying infancy, of course, the verbal clinical interviews he had devised were not a feasible form of research strategy, and he returned to the use of observation as his principal technique. As is well known, the works on sensorimotor development drew on the observations made by Piaget and his wife on their own three children. However, in returning to observation as his technique Piaget also introduced into his work a more active element which makes a kind of parallel in these observational studies to the clinical interviews of his studies on children’s thinking. While some of the observations he records are straightforward records of his children’s activity, many of them are also records of interventions from Piaget or his wife. A re-arrangement of the things around the child would produce a new situation in which the child’s reaction could be observed, and successive re-arrangements

Piaget ethnographer 49 functioned in a similar way to the questioning of the clinical interview, testing out the limits of the child’s developing coordinations at each stage. Thus, for example, the familiar instances of the children coming to search for a vanished object (Piaget, 1937/1954; Obs. 34–38, pp. 44–50) are all records of reactions to interventions of this type. As well as illustrating the influence of the “clinical” method on his observational practice, these studies also indicate Piaget’s continuing ethnographic engagement with his material. Again these observations are analysed in terms of the set of interpretational categories which Piaget introduced through his analysis of sensorimotor development, primarily in terms of action schemes and their various coordinations. Indeed, it may be that the absence of verbal elements from these studies of infant development helps to make the relation between observation and interpretation even clearer. When he arranged for a doll or a watch to disappear from the child’s view, Piaget was interested in the child’s reaction to this disappearance, and the point where the children began an active search for the disappeared object marked an important moment in their development. But Piaget does not simply record this as an objective feature of the child’s activity, rather these reactions are interpreted in terms of his general theory of sensorimotor development. The infant who loses interest in a disappeared object has not yet developed the coordination of schemes which would allow them to search for it. What Piaget described as the “object concept” has not yet begun to emerge in the child. In this way the data furnished by observation are rendered intelligible through interpretation. Or one could say that these observations only become data through being interpreted. 2.5 Clinical interviews Following his studies of sensorimotor development Piaget returned to the study of intellectual development in early and middle childhood (and later into adolescence as well), but he did so using both a revised theory and a revised method. The revised theory became what can be considered as most characteristically Piagetian, since it considers intellectual functioning from the perspective of operations, and describes intellectual development in terms of a sequence of stages, from the preoperational through the development of concrete operations to the final emergence of formal operations in adolescence. The first fruit of this work was his study with Alina Szeminska of The Child’s Conception of Number (1941/1952)3, but this was closely followed by Piaget’s own presentations of the logic of operations (1942; 1949). The revised method extended the earlier form of the clinical interview by structuring it entirely around practical problems. In his Foreword to the book Piaget (1941/1952) explains the reasons for adapting the clinical method in this way. The earlier work had considered the child’s verbal and conceptual constructions, but the aim of this work (and the series of works of which it formed the first instalment) was to go beyond the purely verbal forms of reasoning to the analysis of underlying structures, or as Piaget puts it the aim was to “trace the development of the operations which give

50


rise to number . . . operations which . . . lead from intuitive and egocentric prelogic to rational co-ordination that is both deductive and inductive” (1941/1952, p. vii). He goes on to say that while he intends to continue to use the procedure of “free conversation” with the child, “our investigation of sensory-motor intelligence has, however, shown us the necessity for actual manipulation of objects” (1941/1952, p. vii). Thus the final form of the clinical interview incorporates aspects of the innovations in observational procedures prompted by earlier forms of the clinical interview. With these works Piaget arrived at a research procedure which he retained throughout the remaining course of his active career which lasted up till his death in 1980. In these studies the child is first presented with a practical problem, and when they have given an answer they are asked to give a justification for it. Subsequently, the interviewer starts to offer counter-arguments to the child, and frequently to re-arrange the physical material to produce new forms of the problem. Throughout this conversation the interviewer is always responding to the child’s answers, questioning them, seeking justifications, presenting counterarguments. At each step the interviewer’s intervention is guided by the child’s thought and action, such that the interviewer really has to frame a hypothesis as to what the child means by what they say and do, and then frame their next intervention in terms of this hypothesis. Thus, as in the earlier forms of the clinical interview, the conversation with each child takes its own individual course although the corpus of interviews is thematically coherent since they all relate to the same problem. The advance in this revised method is that the child is no longer, as Piaget puts it, “thinking in the void” but rather “talking about actions he has just performed” (1941/1952, p. vii). All of these elements can be seen in a very familiar example, the problem of the conservation of the amount of water, which appears as Chapter 1 (The conservation of continuous quantities) of the book on number. While there is not space in this paper to analyse this example in detail, some significant points can be noted not only in relation to the method used but also about the narrative style through which Piaget presents his work to the reader. The chapter is divided into sections and this general scheme is presented in Table 3.2 where page numbers are given for both the English and (7th) French editions. The introduction serves to locate the problem under study and to present a hypothesis to be examined about the relation between arithmetical notions and psychological structures. As Piaget makes clear (p.4) his study aims to demonstrate that arithmetical notions acquire their structure because of the emergence of conservation, so that from the outset the reader is led along a particular path. This pathway is then reinforced in the following section which has the intriguing title “Technique and general results”, where he not only describes the general form of the problem presented to the children and the nature of the interview, but also gives a first overview of his findings in terms of broad descriptions of the developmental levels he found (a form of research reporting which, one suspects, few contemporary journal editors would allow their authors to adopt). These levels are then analysed in some detail in the following three sections through the

Piaget ethnographer 51 Table 3.2 Chapter 1 – Conservation of Continuous Quantities

Introduction § 1 Technique and general results § 2 Stage I: Absence of conservation § 3 Stage II: Intermediary reactions § 4 Stage III: Necessary conservation

English Edition

French Edition

pp 3–4 pp 4–5 pp 5–13 pp 13–17 pp17–24

pp 16–17 pp 17–19 pp 19–28 pp 28–33 pp 33–42

presentation and discussion of many examples of interview protocols. In these discussions Piaget is concerned to demonstrate why specific forms of answers can be allocated to particular stages in the development of conservation. The most extended of these discussions is the final section, where children’s responses are analysed in terms of the logic of groupings through which Piaget defined the structure of concrete operations (it is this section which has been curtailed in the English edition). From a methodological point of view it is important to note the role of justification and counter-argument in these interviews. As in his earlier work Piaget is above all concerned to analyse the child’s convictions about the problem, which come through most powerfully when the child resists suggestions which come from the adult interviewer. It is through this resistance that a sense of the structure and organisation at each level of children’s development emerges most clearly (cf Duveen, 2000). This is apparent even among the younger children who are convinced that the amount of water is not conserved across various transformations, they too resist the adult’s attempt to suggest otherwise. Among the older, conserving, children this resistance to counter-argument is also an important marker of their grasp of the logical necessity of conservation which for Piaget is a key criterion for the identification of this stage. In this study (and the countless other examples to be found in his research) one can see Piaget actively striving to present a persuasive case that his interpretive categories (here the concepts of operations and their structure) are not only adequate for grasping the empirical material generated by the interview, but also the most economical way of accounting for the totality of the data by providing a developmental scheme in which each stage can be identified and defined. Persuasion, of course, implies that this research report needs to be considered as part of a dialogue, and one can in fact identify different levels of dialogue with different interlocutors at work in this text. The interview material itself consists of dialogues with children, but there are also multiple dialogues present in the report. First of all there is Piaget’s own internal dialogue, which itself can be seen as operating at two levels, first in the engagement of his interpretive scheme with the interview material and second in his own attempt to persuade himself of the adequacy of his interpretation. Then, of course, the report itself can be read as a persuasive dialogue with the reader.

52


3 Ethnographic engagements and disengagements I have suggested that Piaget’s research can be seen as ethnographic because of the open dialogue it sustains between empirical material and interpretive categories, a dialogue which certainly enables him to offer a “thick description” of the development of children’s thinking in terms of different levels of structural organisation. Thus Piaget’s research method follows what we can describe as an ethnographic logic, and what this means can be appreciated more clearly by contrasting Piaget’s research method with those employed by some of his critics (a different but related attempt to describe contrasting logics of research has been given by Basil Bernstein (1996) in terms of the different languages of description at work in the research process). In the 1970’s and 1980’s a number of research studies appeared which purported to demonstrate significant weaknesses in Piaget’s work (Donaldson, 1978, is a notable example, and a good overview and analysis of such studies is given by Wood (1988)). Common to these studies was the suggestion that Piaget had seriously underestimated the intellectual competence of young children (ie those he had described as pre-operational), and that he had done so as a consequence of his research methods. Various aspects of his clinical interviews were held to be at fault. Donaldson (1978), for example, suggested that the problems he presented to children were too abstract and failed to make “human sense” to the child, so she and her colleagues sought to reframe these problems in forms which they considered to be more accessible to children. Others, such as Rose and Blank (1974), suggested that the nature of the interview itself incorporated forms of dialogue which introduced a systematic bias in the results. Collectively these studies presented a body of evidence which was held to demonstrate that children were capable of engaging in the forms of thought which Piaget attributed to middle childhood at much earlier ages. It is not my intention here to review the weaknesses of these studies as critiques of Piagetian research. Others (eg Gold, 1987; Wood, 1988) have pointed out how, in the process of “revising” the tasks, these studies have generally also significantly changed the problem presented to children. What concerns me here, however, is the logic of research at work in these critical studies which have generally been framed as experimental studies. However, they also differ from Piaget’s work in a crucial respect. In these critical studies children are presented with a problem (in carefully controlled experimental settings) and asked for a solution. There is not, however, any attempt to pursue the conversation with the child any further, they are not asked to give any justification for their answer, nor are they exposed to any counter-argument from the interviewer. In short these studies operate with a method which is far removed from that of Piaget’s clinical interview, and as Leslie Smith (1992) has argued, this difference carries important consequences since the inferences which a researcher can make on the basis of children’s judgements alone are different from those which can be made on the basis of more extended interviews. Donaldson (1978) herself has argued that one should not engage the child in this kind of argument, since young children are not

Piaget ethnographer 53 capable of explicating their own thought processes. Thus for her it is the child’s judgement alone which is a reliable source of data. The difference between Piaget and his critics is not merely one of style nor simply an argument over the precise age at which children become capable of employing particular forms of thought (and in any case, age itself was never taken as independent index of development by Piaget). Rather, this difference relates to conflicting logics of research. I described above how the dialogue of the clinical interview is central to Piaget’s research method. Within this dialogue the interviewer is always attempting to grasp and understand the child’s comprehension of the situation. Interpretation by the interviewer and the exploration of that interpretation through further conversation is at the heart of the clinical interview, which is itself embedded within a broader interpretive framework. For the critical studies the research situation is constructed rather differently. Here the child is presented with a problem and asked for a judgement, and it is assumed that the meaning of this judgement can be interpreted directly from the logic of the experimental situation itself. If the child says “the same” it is assumed that they have understood that something has been conserved, if they say “different” a conservation is denied. A great deal of ingenuity has been expended in constructing different experimental situations which can generate more “same” judgements at earlier ages. The contrast with Piaget is very stark. For him the child’s initial response is only a beginning. If they say there is the same amount this is the point of departure for a conversation about what they mean by “the same”; and it is clear from the protocols he presents that children’s initial judgements are often modified and changed through the course of the interview itself. But this fluidity in children’s thinking and the conversation in which it is embedded are central to his method which is concerned with identifying the underlying structures which organise children’s thinking. We can summarise these contrasting logics of research in the following way. Piaget’s clinical interview adopts an open position about how the child understands a problem, which provides a space for the interviewer to interpret the child’s responses as well as a space for a dialogue with the child. This material is then further interpreted by Piaget in his analysis of the interviews as a whole. Consequently his logic consists of a dialectic between material and its interpretation. For his critics the situation is rather a closed one in which the meaning of the child’s response is immediately given by the structure of the experimental design or setting itself. In this position the “data” is always an objective and unmediated feature of the experiment itself. Such experimental logic derives from a positivist view of scientific research, and one can see in these studies precisely that fetishism of data which is such a characteristic of scientific positivism and which marks a disengagement from ethnographic concerns. This contrast provides a fitting point at which to conclude this paper. While it may have seemed strange at the outset to consider Piaget as an ethnographer of children’s development, I hope to have been successful in suggesting that he can indeed be seen in this light on the basis of an analysis of the logic of his research method. Further, taken as a whole his work can be viewed as an important

54


contribution to current debates about qualitative methods, providing as it does such a clear example of how interpretive methods can lead to powerful theoretical models.

Notes 1 Earlier versions of this paper were given at the meeting of German language developmental psychologists in Vienna in 1997 and at the joint IACCP and ITC European meeting in Graz in 1999. I am grateful to the organisers and participants in both these meetings for their constructive comments, and particularly to Uwe Flick. I have also benefited from discussions with Jacques Vonèche, Anastasie Tryphon and Rosita Hadded Zubel at the Archives Jean Piaget in Geneva. 2 Where two dates are given in a reference the first is that of the original French publication, the second that of the English translation. 3 The English edition of this text suffers from some significant deletions of portions of the French original. While the translator notes that “with the author’s permission . . . we have omitted the logistic algorism introduced by the author in Chapters III and X” (p. ix) in fact the omissions are more extensive than this. Almost every chapter of the French original concludes with a section where the research study reported in the chapter is interpreted in terms of the logic of operations, and little of this interpretation remains in the English text. It’s not clear why these deletions were made, and at this distance of time almost impossible to trace. Certainly the Archives Jean Piaget in Geneva holds no correspondence on this matter. The most likely explanation is an economic one, that the publishers wished to shorten the length of the text.

References N.B. Where two dates are given in a reference the first is that of the original French publication, the second that of the English translation. Bernstein, B. (1996) Research and Languages of Description. In B. Bernstein Pedagogy, Symbolic Control and Identity. Pp. 134–144. London: Taylor & Francis. Chapman, M. (1988) Constructive Evolution. Cambridge: Cambridge University Press. Donaldson, M. (1978) Children’s Minds. London: Fontana. Duveen, G. (1997) Psychological development as a social process. In L. Smith, P. Tomlinson and J. Dockerell (Eds), Piaget, Vygotsky and Beyond. London: Routledge. Duveen, G. (2000) Representations, Identities, Resistance. In K. Deaux and G. Philogène (Eds) Social Representations: Introductions and Explorations. Oxford: Basil Blackwell. Duveen, G. and Lloyd, B. (1993) An ethnographic approach to social representations. In G. Breakwell and D. Canter (Eds) Empirical Approaches to Social Representations. Pp 90–109. Oxford: OUP. Geertz, C. (1974). The Interpretation of Cultures. New York: Basic Books. Ginsburg, H. P. (1997) Entering the Child’s Mind. Cambridge: Cambridge University Press. Gold, R. (1987) The Description of Cognitive Development. Oxford: Oxford University Press. Jessor, R., Colby, A. and Shweder, R. A. (1996) (eds) Ethnography and Human Development. Chicago: University of Chicago Press. Lévy-Bruhl, L. (1926) How Natives Think. London: Allen & Unwin.

Piaget ethnographer 55 Montangero, J. and Maurice-Naville, D. (1997) Piaget or the Advance of Knowledge. Mahwah, NJ.: Lawrence Erlbaum Associates. Piaget, J. (1923/1926) The Language and Thought of the Child. London: Routledge & Kegan Paul. Piaget, J. (1924/1928) Judgement and Reasoning in the Child. London: Routledge & Kegan Paul. Piaget, J. (1926/1929) The Child’s Representation of the World. London: Routledge & Kegan Paul. Piaget, J. (1927/1930) The Child’s Conception of Physical Causality. London: Routledge & Kegan Paul. Piaget, J. (1932/1932) The Moral Judgement of the Child. London: Routledge & Kegan Paul. Piaget, J. (1936/1952) The Origins of Intelligence in the Child. London: Routledge & Kegan Paul. Piaget, J. (1937/1954) The Construction of Reality in the Child. London: Routledge & Kegan Paul. Piaget, J. (1945/1951) Play, Dreams and Imitation. London: Routledge & Kegan Paul. Piaget, J. (1941/1952) The Child’s Conception of Number. London: Routledge & Kegan Paul. Piaget, J. (1942) Classes, relations et nombres. Paris: Vrin. Piaget, J. (1946/1969) The Child’s Conception of Time. London: Routledge & Kegan Paul. Piaget, J. (1949) Traité de logique. Paris: Dunod. Piaget, J. and Inhelder, B. (1948/1956) The Child’s Conception of Space. London: Routledge & Kegan Paul. Piaget, J. and Inhelder, B. (1955/1964) The Early Growth of Logic in the Child. London: Routledge & Kegan Paul. Rose, S. and Blank, M. (1974) The potency of context in children’s cognition. Child Development, 45, 499–502. Smith, L. (1992) Judgements and Justifications: Criteria for the attribution of children’s knowledge in Piagetian research. British Journal of Developmental Psychology, 10, 1–23. Wood, D. (1988) How Children Think and Learn. Oxford: Basil Blackwell. Vinh-Bang (1966) La Méthode Clinique et la Recherche en Psychologie de l’Enfant. In F. Bresson and M. de Montmollin (Eds) Psychologie et Epistémologie Génétiques: Thèmes Piagétiens. Pp. 67–81. Paris: Dunod.

4

Genesis and structure Piaget and Moscovici Gerard Duveen

In his recent interview with Ivana Marková, Serge Moscovici (2000) recalls that in reading Piaget’s child psychology he “had the impression of discovering what social psychology might be”, an impression which was so strong that he describes it also as a “vision” of a social psychology which retains its vitality for him even today, more than forty years after this first encounter. Such an emphatic acknowledgement of the significance of the work of Jean Piaget for the development of his own social psychology might surprise many readers of Moscovici, even those familiar with his œuvre. While one can find many references to Piaget’s work in Moscovici’s own writings, surely the psychology of the developing child is so far removed from the concerns of social psychology as to make the idea of the one providing a vision for the other altogether improbable? Improbable or not, there is little reason to doubt that something decisive for the development of his social psychology emerged from Moscovici’s encounter with Jean Piaget. As he himself expressed it, notwithstanding the differences which might seem to be apparent between developmental and social psychology, there is in fact a deep continuity connecting them, a continuity which flows from the sense that both approach questions of the transformations of knowledge. As he puts it, the former explores these transformations through time, the latter through space (Moscovici 1990). In fact, looking back now across the body of Moscovici’s work on social psychology, one can see how the figure of Piaget has often appeared at significant points in these writings. Let me give just one example from one of his best known texts, the essay on The Phenomenon of Social Representations (Moscovici 1984a, reprinted in 2000). Certainly this has been a central text in the construction of the theory of social representations. For Anglo-Saxon audiences in particular, who remain without an English translation of La Psychanalyse (1961), this essay has become the classical reference for the theory. In the second section of this essay, entitled What is a Thinking Society?, Moscovici introduces the theme of social representations, partly in relation to Durkheim of course, but also through a key reference to Piaget. Let me quote this passage in full: Similarly, one knew that social representations occurred in societies, but nobody worried about their structure, or about their inner dynamics. Social psychology, on the other hand, is and must be preoccupied solely with both

Genesis and structure 57 the structure and the dynamics of representations. For us, it is summed up in the difficulty of penetrating the interior to discover the inner mechanisms and vitality of social representations in the greatest possible detail; that is in ‘splitting representations’, just as atoms and genes have been split. The first step in this direction was taken by Piaget when he studied the child’s representation of the world and his enquiry remains, to this day, exemplary. So what I propose to do is to consider as a phenomenon what was previously seen as a concept (Moscovici 1984: 16–17, and 2000: 30). As you can see, this passage ends with a phrase which has often been quoted and which serves to justify the title of the essay itself. It is a passage which draws together many themes which have been significant in Moscovici’s work, from the conviction that social psychology as a science is concerned with the analysis of representations, to the sense that this science must be pursued in the same manner as those other great sciences of the twentieth century, physics and biology, by constructing its object of inquiry precisely in such a way that it can be split into its component parts whose interrelations and dynamics become the focus for scientific activity. (Indeed, one can note that the idea of splitting may be considered as a significant themata of modern science, see Moscovici and Vignaux 1994, translated in Moscovici 2000). And, for Moscovici, it is Piaget who presented an “exemplary” instance of “splitting representations” in his early studies of the development of the child’s representation of the world. One might quibble with Moscovici’s judgement that Piaget’s work was the “first step” in this direction (Freud for one might be thought to have already taken a step in this direction a generation earlier than Piaget, but one can also think of others who had contributed to this movement of thought) but reading this passage again now one cannot doubt the significance which Piaget’s work had for Moscovici’s own development of his characteristic social psychology. But what exactly was the significance of Piaget for Moscovici? The answer to this question is really a tale of two cities, Paris and Geneva. As Moscovici recalls in his interview with Ivana Marková, it was in Paris in the 1950s that he encountered Piaget, both the man and his work, when Piaget was appointed as the successor to Merleau-Ponty in the Chair of Child Psychology at the Sorbonne. In that interview Moscovici recounts that he found Piaget a rather isolated figure in Paris at that time, but one who made a strong impression on the young social psychologist (see Moscovici 2000: 250; and one should add that this impression seems to have been mutual; certainly when Piaget later sought to introduce social psychology into the curriculum at the University of Geneva it was Moscovici he invited to come and lecture). Although Moscovici does not recount this in particular, it seems probable that one point of sympathy he would have found was Piaget’s rigorous constructivist epistemology (behaviourism seems never to have held any attraction for either Piaget or Moscovici). As Lucien Goldmann, another Romanian Jewish intellectual, had pointed out, Piaget’s work was always dialectical in its form and consistent with a Marxist perspective (Goldmann 1959), something which would surely have appealed also to the young Moscovici. At this

58


point in his life, Piaget was engaged in what Montangero and Maurice-Naville (1997) describe as his third period in which he was concerned with the use of structural analysis to study the formation of the ‘categories’ of thought. A central issue here was the problem of the transformation from preoperational to concrete operational thinking, and it is the issue of the transformation of knowledge which is at the centre of Moscovici’s enthusiasm for Piaget. Transformation as a concept for understanding development implies that change is always from one structured form to another, never simply from chaos to order. Piaget’s appreciation of this theme in the study of children’s development in fact takes us back once again to Paris, but this time to a much earlier period, to the Paris of the years after the end of the First World War. It was to Paris that the young Piaget came in 1919 to spend two years as a post-doctoral researcher in the laboratory which had been established by Binet, and which was directed after his death by his colleague Simon (Piaget 1952). Indeed this period in Paris seems to have been the setting for Piaget’s first professional encounters with children. His interest in establishing a biological epistemology was already clear for him, but at that time it remained rather abstract and without a precise empirical focus. Through his contact with Simon he came to see that it would be possible to study the development of knowledge in the child, something which would give concrete form in the realm of human knowledge to his abstract thoughts, which had been generated through his biological and zoological studies. The specific task which Simon had suggested to Piaget was the standardization, with a French population, of the intelligence test items which Cyril Burt had recently introduced in England. These items presented children with elementary logical problems of the kind which have continued to feature in intelligence tests. For instance, the child might be presented with the syllogism: “If Mary is taller than Jane, and Jane is taller than Susan, who is taller, Mary or Susan?” The only aspect of the child’s response which was relevant for the standardization of such items was whether they gave the correct answer or not: “correct” of course, being defined in terms of the ordinary logic of adult discourse. Piaget, however, found something more interesting in these encounters than merely the repetition of the same question over and over again (Piaget 1952). First, he saw clearly that when children gave “wrong” answers to such questions, their responses were not simply random or chance efforts. Rather, their answers were organised and structured: it was simply that the logic of this structure was not the same as the logic of adult discourse. Already, therefore, these early encounters with children suggested that transformation from one structure to another was a central category for understanding development. But Piaget also introduced something else into the situation. He began to question the children, to engage them in “conversations patterned after psychiatric questioning, with the aim of discovering something about the reasoning process underlying their right, but especially their wrong answers” (Piaget 1952: 244). In this way, Piaget’s Parisian experiences also laid the groundwork for the emergence of his “clinical interview” technique (Duveen 2000a). Piaget’s reference to the model of the psychiatric interview is itself of considerable interest.

Genesis and structure 59 Immediately before setting off for Paris, he had spent some time in Zurich, where, as well as becoming more interested in Freud and psychoanalysis (in part through attending lectures by Pfister and Jung; see Harris 1997, for an interesting discussion of Piaget’s early relation to psychoanalysis and its significance for his years in Paris), he also attended Bleuler’s psychiatric clinic. His intuition in Paris was to see that just as the psychiatrist attempted to diagnose the condition of his patients through a structured interview, so it should be possible to investigate the different mentality of the young child through a similarly organised interview. This was an extraordinary intuition. Of course, at that time research methods in psychology were not nearly so carefully framed and organised as they are today. Nevertheless, this innovation shows something of the extraordinary creative fertility of the young Piaget’s approach to his research. When he returned from Paris to Geneva to work in the Maison des Petites of the Institut Jean-Jacques Rousseau this clinical technique soon came to replace the observation of natural conversation as a more direct method for exploring the development of children’s thinking. Piaget was very aware of the problems and difficulties involved in interpreting such interview material, problems which he addressed in his introduction to The Child’s Conception of the World (1926/1929). This book is the text Moscovici referred to as an “exemplary” study of social representations, and the introduction is one of Piaget’s most overtly methodologically focussed chapters. Here he tried to classify children’s contributions to these discussions by separating responses which were a direct product of the interviewer’s influence on the child, or in which the child engaged only in romance or fantasy, from those responses which reflected the child’s real convictions in the sense of their underlying beliefs about the world as they understood it and represented it. While such distinctions might be analytically clear in a theoretical sense, they are not always easy to apply in specific cases. And indeed there are many instances in the analytic chapters of these books where Piaget addresses just this question in discussing particular responses of individual children. This is obviously a key interpretive moment in such research studies, but its significance should not be overestimated. One could challenge one or another of Piaget’s judgements about whether a particular response was “romancing” by the child or whether it reflected a “liberated” or “spontaneous” conviction (and in general one can note that the extensive excerpts of the interviews given by Piaget in his written text provide the reader with the basis for doing so). However, the theoretical structure of the analyses of children’s representations does not depend on such individual judgements, but rather on the coherence of Piaget’s own system of categories for interpreting the interview material as a whole. There is, of course, a kind of limiting case here. If Piaget’s judgements were always wrong, or systematically biased in some way, then this would indeed seem to challenge not so much the coherence of his interpretive categories, but their applicability to the material he had collected. In these early works, the categories which Piaget employed in his interpretation focussed on general characterizations of children’s representations in terms of realism, animism and artificialism. In each case he analysed children’s development from some form of non-differentiation (broadly characteristic of the young

60


child’s egocentric thinking) to a more mature differentiation; between self and the external world in the case of realism, animate and inanimate nature in the case of animism, and between human activity and natural causes in the case of artificialism (see the discussion in Chapman 1988: 48–50). In establishing these categories, Piaget also drew on the work of Lévy-Bruhl (1926), suggesting a parallel between his account of the “participations” in the magical thinking of so-called primitives and the characteristics of young children’s thinking. For Piaget, though, while the representations which he analysed in young children were no less constitutive of a specific mentality than they were for the French anthropologist, they also needed to be seen from a developmental point of view. Unlike the primitive, the child also went on to develop a different mentality which reflected the emergence of logical thinking and its systematic application. Thus while Piaget drew on Lévy-Bruhl’s descriptions of primitive mentality, in his own analysis he set these forms of thought in a different context, so that his interpretations of the interview material he collected analysed children’s thinking as developing through a series of structurally differentiated stages which led to the emergence of mature adult thought. Over the course of his long, productive life, the interpretive framework through which Piaget analysed cognitive development changed considerably. His early reliance on Lévy-Bruhl and others (there are also traces in Piaget’s early work of Freudian concepts describing the characteristics of unconscious thought being used to express his analysis of the mentality of young children) gave way to a conceptual system all of his own. These early analogies between the mentality of the young child and the “primitive” mentality described by Lévy-Bruhl, or the unconscious mentality described by Freud, came to be replaced in Piaget’s later works by his own concepts of schemes, functions and operations. This change certainly enabled Piaget to offer a more coherent and consistent account of development itself, as a series of progressive decentrations made possible through the construction of cognitive structures of increasing complexity. Indeed, one could say that Piaget’s theory was always in a constant state of change, or as he himself put it, he “always considered himself one of the chief ‘revisionists of Piaget’ ” (Piaget 1970: 703). Yet within the constant flux of Piaget’s theorizing, some themes remained constant. One central theme was his attachment to a biological epistemology, to a sense that human knowledge is a special case of biological adaptation (a Piagetian themata par excellence). A second theme concerns the idea of transformation itself as a relation between structures. To grasp how one structure can be transformed into another Piaget saw that it was necessary to distinguish different aspects, or moments, in a structure in such a way that their relations, balanced or conflictual, could explain both the stability of structures and their transformations. Hence his distinction between assimilation and accommodation, the functional invariants as he described them, as moments in the structural process of adaptation (indeed, for Piaget genesis and structure are interdependent categories). Without such a distinction, structures would remain either fixed or changeable only through the action of the external environment. Epistemologically this would

Genesis and structure 61 mean accepting either a priorism or empiricism, both of which conflicted with Piaget’s commitment to an epistemological interactionism founded on the correlative relations between subject and object. As he puts it: “no subject without action on the objects, and no objects without a structuration1 contributed by the subject” (Piaget 1969: 157). For Piaget, no other position was consistent with the idea of a self-regulating organism which was central to his vision of human knowledge as a form of biological adaptation. It seems clear that for Piaget himself these two themata, knowledge as a form of biological adaptation and transformation as a consequence of a distinction between different moments in a structure, were intimately connected, and, indeed, it often seems that he took the second to be either a consequence of, or implicit in, the first. However, it nevertheless seems necessary to insist on distinguishing these two themata. Even if it were the case that the second was in some way implicit in the first, this would not mean that the second could not be taken independently of the first. In other words, a sense of transformation as the product of internal relations between different moments of a structure has no necessary relation to a commitment to the kind of biological epistemology which Piaget sought to construct. Indeed, as much other work has shown (not least Lucien Goldmann’s genetic structuralism) transformation can be quite independent of a biological epistemology. I have reviewed these aspects of Piaget’s work at some length because it seems to me that to grasp his significance for Moscovici we need to understand both how they are similar and how they are different. There is something in the architectonics of both thinkers which is undoubtedly similar, primarily in this idea of transformation as the product of the relation between different moments of structure, assimilation, and accommodation, for Piaget, whereas for Moscovici the categories are anchoring and objectification, or innovation and conformity. It is not, then, that Moscovici simply adapted Piagetian psychology, but rather that he found within Piaget’s work some conceptual and theoretical elements or ideas which he could work with (or against). And as his recent essays show (Moscovici 2000) he also saw Piaget as standing within a certain tradition of the social sciences, a tradition in which he could also position his own work in relation not only to Piaget, but also to Durkheim and Lévy-Bruhl, as well as Vygotsky. But there is also something creative and imaginative in Moscovici which sets his work on a different path. This is his concern with processes of social change. Indeed, it could be said that what is most characteristic of Moscovici’s social-psychological imagination is the attempt to thematize change. Or, rather, since many other psychologies – Piaget’s included – have also addressed questions of change, it is the way in which change is thematized in Moscovici’s social psychology which is distinctive. Piaget, of course, offers an analysis of development as a series of transformations, in which one can see the construction of ever more complex forms of thought as the most elementary action schemes are eventually transformed into formal operations. In this sequence structural change is located within functional constancy, that is, both schemes and operations play the same functional role in the regulation of the organism’s interactions with its environment. But it is also change which is articulated in a specific direction. Indeed, the very notion of development

62


in Piaget’s thought implies a movement towards a more adapted organisation of the exchanges between organism and environment. The child as an epistemic subject may have a characteristic mentality, but it is a mentality identified, defined, and conceived in terms of a movement towards another mentality, that of the adult. Indeed, one could say that Piaget’s analysis is premised on a knowledge of the sciences of logic, mathematics and physics, a premise which implies that development is a movement which always imposes a specific direction on the successive transformations of structures. For Moscovici, by contrast, change in the structures of representations is not regulated by some pre-established end-point in the same way. Whatever direction can be assigned to the transformations of representations emerges from the constraints exercised on construction by thematic considerations, questions of relevance and the forms of the social relations involved. In other words, structural change in the organisation of representations is not conceived in terms of a triumphal march towards the high plateau of scientific understanding. Indeed, as his study of La Psychanalyse (1961) indicates, one of his abiding interests is precisely the opposite movement, of the changes engendered by the transformations of scientific thought into common sense. More generally, the transformations of social representations are not regulated by some external set of predefined, universal criteria, but reflect rather the necessities imposed by the contingencies of the situations in which knowledge and understanding are elaborated and communicated by different social groups. To understand more clearly the transformations of knowledge effected by social representations we need to recognise the dynamic character of representations. Representations may be conceived as particular or specific organisations of ideas, values, and practices, but as structures they are also always embedded within a specific context of communication, both face-to-face and mass-mediated. Hence representations are always structures whose organisation depends on a balance of communicative influences. Again a comparison with aspects of the architectonics of Piaget’s theory can be helpful here. For Piaget, genesis and structure are intimately related, since a structure is always the relatively enduring organisation of a function (just as a function could not operate except through being organised in some structure). But organisation implies that structures always achieve a measure of closure, which for Piaget is effected by a certain logic, so that different levels of structural organisation are identified precisely by the different logics which serve to hold the structure together and provide its defining characteristics. Similarly, social representations also require a form of closure if they are to become organised structures. In this case, however, closure is effected by the balance of influence processes in any given situation. Social influence is always directed toward sustaining or transforming representations. While this relationship between representation and influence may not have been always clear, or even explicitly articulated, in Moscovici’s various work, it is at the centre of what I have referred to as his “social-psychological imagination” (Duveen 2000b; also Duveen 1998). The exercise of social influence always, of course, operates within the context of themata, relevance, and social relations, a context which itself indicates something

Genesis and structure 63 of the complexity involved in the transformations as knowledge circulates within and between societies. In this sense too, the closure of representational structures effected through social influence processes means that they are rarely, if ever, constituted around a single logic. Rather, it is more frequently the case that representations embody a form of co-existence between ideas, values, and practices drawn from inconsistent or even contradictory elements. Such a state of cognitive polyphasia can be seen, for example, in the way in which traditional and Western psychiatric ideas are combined in contemporary representations of madness in India (Wagner et al. 2000). This idea of cognitive polyphasia offers an important point of contrast with Piaget’s insistence on the rule of logic: a contrast which is most visible in Piaget’s discussion of the difference between cooperation and constraint as forms of social relations which imply distinct transformations of knowledge, social transmission in the case of constraint as against rational reconstruction in cooperation (Piaget 1932). On the basis of such a distinction between forms of knowledge generated in different social relations Piaget himself might have moved towards a notion of cognitive polyphasia, but he remained a committed monophasic. For him, logic – and the science which was its fruit – displaced magic; this is the nature of cognitive development, and thus between one phase and another there could only be a radical discontinuity. For Moscovici, the idea of cognitive polyphasia is associated with the persistence of magical forms of thought in the everyday world of common sense (Moscovici 1992). These few brief remarks by no means exhaust the rich source for reflection provided by a consideration of Moscovici’s relation to Piaget. They have, I hope, gone some way towards both recognising the significance of Piaget for the development of Moscovici’s own thinking, and suggesting the benefit of exploring this relationship more systematically. The “splitting” of representations effected by the theoretical impulse in Moscovici’s work is different from that of Piaget. Where Piaget’s work, centered on the idea of the epistemic subject, revealed the different organisation of the mentality of child and adult and the developmental transformation of one to the other, Moscovici’s work renders visible and intelligible the complex dynamics of genesis and structure at work in the phenomenon of social representations.

Note 1 The symposium from which these comments are taken was held in 1968, so that Piaget’s use of this term ‘structuration’ predates its more familiar usage in recent years in the work of Anthony Giddens. While there are obviously similarities in the meanings given to the term by these two auditors, it is by no means clear that this implies an identity in the use of the term.

References Chapman, M. (1988) Constructive Evolution. Cambridge: Cambridge University Press. Duveen, G. (1998) The Psychosocial Production of Knowledge: Social Representations and Psychologic. Culture and Psychology, 4: 455–472.

64


Duveen, G. (2000a) Piaget Ethnographer. Social Science Information, 39: 79–97. Duveen, G. (2000b) La culture dans les jeux imaginaires de jeunes enfants (Culture in the pretend play of young children]. In D. Saadi-Mokrane (Ed) Sociétés et cultures enfantines (pp. 111–117). Lille: Editions du Conseil Scientifi que de l’Université Charles-de-Gaulle–Lille 3. Goldmann, L. (1959) Recherches Dialectiques. Paris: Gallimard. Harris, P. (1997) Piaget in Paris: From Autism to Logic. Human Development, 40: 109–123. Lévy-Bruhl, L. (1926) How Natives Think. London: Allen & Unwin. Montangero, J. and Maurice-Naville, D. (1997) Piaget or the Advance of Knowledge. Mahwah, NJ: Lawrence Erlbaum Associates. Moscovici, S. (1961) La psychanalyse, son image et son public. Paris: Presses Universitaires de France. Moscovici, S. (1984) The Phenomenon of Social Representation. In R. Farr and S. Moscovici (Eds). Social Representations. Cambridge: Cambridge University Press. Moscovici, S. (1990) Social Psychology and Developmental Psychology: Extending the Conversation. In G. Duveen and B. Lloyd (Eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Moscovici, S. (1992) La Nouvelle Pensée Magique. Bulletin de Psychologie, XLV, 405: 301–324. Moscovici, S. (2000) Social Representations: Explorations in Social Psychology (Edited by G. Duveen). Cambridge: Polity Press. Moscovici, S. and Marková, I. (2000) Ideas and their Development: A Dialogue between Serge Moscovici and Ivana Marková. In G. Duveen (Ed) Social Representations: Explorations in Social Psychology. Cambridge: Polity Press. Moscovici, S. and Vignaux, G. (1994) Le concept de thêmata. In C. Guimelli (Ed) Structures et transformations des representations sociales. Neuchâtel, Delachaux et Niestlé: 25–72. English translation in Moscovici, S. (2000) Social Representations: Explorations in social psychology. (Edited by G. Duveen) Cambridge: Polity Press. Piaget, J. (1926/1929) The Child’s Representation of the World. London: Routledge & Kegan Paul. Piaget, J. (1932) The Moral Judgment of the Child. London: Routledge & Kegan Paul. Piaget, J. (1952) Autobiography. In E.G. Boring, H. S. Langfeld, H. Werner, and R.M. Yerkes (Eds) A History of Psychology in Autobiography, Vol IV (pp. 237–256). Worcester, MA: Clark University Press. Piaget, J. (1969) Afterthoughts. Following the Discussion of ’The Gaps in Empiricism’ by J. Piaget and B. Inhelder. In A. Koestler and J.R. Smythies (Eds) Beyond Rectionism: New Perspectives in the Life Sciences (pp. 157–160). London: Hitchinson. Piaget, J. (1970) Piaget’s Theory. In P.H. Mussen (Ed) Manual of Child Psychology, Vol 1 (pp. 703–732). New York: John Wiley & Sons. Wagner, W., Duveen, G., Verma, J. and Themel, M. (2000) I have some faith and at the same time I don’t believe. Cognitive Polyphasia and Cultural Change. Journal of Community and Applied Psychology: Special Issue: Health, Community and Development, 10: 301–314.

Part II


5

Social life and the epistemic subject Gerard Duveen

Objects are encountered first of all outside the theoretical structures of psychology; they exist primarily as human realities, as aspects of being produced and maintained through human action. In this sense they are not concepts produced through the operation of psychological theories but rather are created within the context of the lived experience of everyday life; they are encountered first of all within the life space of individuals (the lebenswelt or what Sartre has termed le vecu). Indeed they are so closely interwoven into the texture of this context that in thinking about them it is difficult to detach them from this totality; they are implicated at every level: the conceptual, the affective and the practical. The relation of psychological theory to lived experiences has been problematic throughout the history of psychology. Concepts which come to psychology from lived experience have, nonetheless, to be formulated in terms of psychological theory and the relations between these two domains appear to be in continuous tension, a tension which is both necessary and uncomfortable. On the one hand there is some immediate and direct quality of lived experience which seems to be irreducible to any structured set of theoretical concepts. On the other hand psychologists in their theorising attempt to transform lived experience into just such a set of structured concepts. Yet a psychological theory which fails to resonate with the timbres of lived experience appears to be abstracted to the point of emptiness, to have lost its meaningfulness as a theory about human subjects. The naivety with which psychological theory is projected into the ‘real’ world becomes clearest in an essay by Hartup (1978) on “Children and their friends”. A simple process is apparently envisaged. At its origin is the psychologist who, through his research and theory, is seen to have access to two areas of knowledge. On the one hand he has access to an objective analysis of the structure and function of peer relations while on the other hand he has some cognisance of how these relations are regulated by school rules. Perceiving that these two areas of knowledge are not very well co-ordinated the psychologist can intervene with the teachers in order to persuade them to change their social policy. The essential point here is that the school itself exists within a social context; it is not an independent institution freely created and maintained by its staff and pupils.

68


The school exists as one institution within a given culture, a given society, and its internal structure will reflect the structure of the wider whole of which it is a part and in which it has a particular function. There are relationships of power and authority embedded in the social relations within which the school is one institution. Such relationships exert a pervasive influence on the internal structure of the school, influences which do indeed prevent the teachers from acting as purely rational beings. The school system in general has shown resistance to change its internal structure in accordance with the resistance to change of the power relationships embedded within social relations in general. It is in neglecting the influence of this social and ideological context that Hartup’s argument can be considered naive. Although everyday life appears to our consciousness as a psychological phenomenon, we are ourselves embedded within the social and ideological relations of our own culture. We are not purely rational beings but rather suffer the influences of our own culture. Such extra interpersonal relations do exist and exert a powerful influence on the cognition of social life. The burden of this argument is that psychology has no immediate and direct relation to lived experience. The relationship is mediated through the culture in which that lived experience is itself embedded. Consequently there is a need for psychology to move away from a naïve sense of the world and of social life toward a more critical attitude which rejects the idea that the concepts of psychology are simply related to realities which exist in some external world. Rather, they need to be understood as theoretical structures which are elaborated through a process of construction in just the same way as the world which is known by the child is not simply an external reality to which the child somehow becomes connected but a product of the subject’s own construction. The way to approach this question lay in the analysis of the relations between the epistemic subject and social life.

2 The epistemic subject In the perspective of genetic epistemology it is the subject which, through the coordination of its own interacting actions with the world constructs an understanding of the world. Yet some clarification is necessary of the status of this subject which is at the centre of the constructive process. It is a subject of knowing, yes, but Piaget is careful to distinguish it not only from previous philosophical conceptions of the knowing subject (the Cartesian cogito or Kant’s transcendental ego) but also and more importantly for this discussion, from the psychological subject identified with the ego. There is the ‘psychological subject’, centred in the conscious ego whose functional role is incontestable but which is not the origin of any structure of general knowledge; but there is also the epistemic subject or that which is common to all subjects at the same level of development, whose cognitive structures derive from the most general mechanisms for the co-ordination of actions. (Piaget 1966, p. 308)

Social life and the epistemic subject 69 The subject which is at the centre of epistemological construction is, according to Piaget, not identified with the conscious ego. It is this mysterious ‘epistemic subject’ which is the terrain for the perspective of genetic epistemology. Piaget further makes clear that this subject is an analytical abstraction and not some particular or individualised or embodied subject. To quote a phrase from Lucien Goldmann it is a “transindividual subject” and Goldmann further suggests that this subject “has the function of rendering the facts we propose to study intelligible and comprehensible” (Goldmann 1976, p. 92). The epistemic subject is a theoretical construction necessitated by the operation of the theory itself. Another way of expressing this point is to emphasise that Piaget’s theory is not a theory about the growth and development of a particular individual or individual experience but it is instead concerned with the development of forms of knowledge common to all individuals. To achieve this it necessarily operates at a level of abstraction beyond the concrete particulars of individual experience. Thus Piaget’s analyses of the progressive coordinations of actions and the formation of cognitive structures are related to the epistemic subject and not to the conscious ego or the self, which are themselves structures whose development can be examined in relation to the development of the epistemic subject. This conception of the epistemic subject is not reducible to a kind of lowest common denominator of individual experience. To be sure it can only be approached through an analysis of individual activity, but this analysis must be carried further than the identification of common elements or patterns of individual activity. Each developmental level must be identified as a particular structural organization and situated within the progressive logic of development as a whole. Once again the epistemic subject is encountered here as an analytical abstraction, for it is only in terms of this abstraction that the process of development can be analysed as a whole. Of course some such abstraction is necessary for any developmental theory. At the most immediate level every developmental theory is an abstraction from the analysis of individuals at various levels or ages or stages. Even if it remains unrecognised by the theory there is an abstract subject at the centre of every developmental theory (marked by the use of the term ‘the child’). What distinguishes Piaget’s theory is the conscious articulation of this abstraction at the heart of the theory itself. Without it the theory would collapse, for it is also associated with the crucial separation of the world of appearances from the underlying structures which make these appearances comprehensible. This separation of appearance from underlying structure is present at every moment of Piaget’s theory; it distinguishes particular contents from general forms of knowing (cf. Furth 1978, p. 230). It stands at the origin of a series of oppositions whose articulation is, to a large degree, the core of the theory of cognitive development itself (assimilation/accommodation; scheme/schema; operative/ figurative). Piaget notes that in the course of development . . . sooner or later reality comes to be seen as consisting of a system of transformations underlying the appearance of things (Piaget and Inhelder 1971, p. xiii)

70


Piaget makes clear that the epistemic subject is common to all individuals; it is in fact that structure which makes possible individual thought. In this sense it may be compared to Kant’s argument that it is the forms of our intuition (which for Kant refers to the universe described by space and time) and the categories of our judgements which make our thinking that of a human subject. If it were otherwise our world would not be what it is for us, it would not be a human world. The separation of appearance and structure and the identification of the epistemic subject with the level of structure are also implicated in the relationship which Piaget draws between biology and epistemology. Cognitive processes are considered to be biological products, the results of evolution, with a clear functional role for the organism: Cognitive processes seem to be, at one and the same time, the outcome of organic auto-regulation reflecting its essential mechanisms, and the most highly differentiated organs of this regulation at the core of interactions with the environment so much so that, in the case of man, these processes are being extended to the universe itself. (Piaget,1971 p. 26) The epistemic subject is, therefore, to be considered as a species specific evolutionary product, which from a human stand point, emphasises again the sense in which it expresses what is common to well-formed members of the species. It cannot be considered as an individuated subject. The epistemic subject and the psychological subject belong to different levels (structure and appearance) yet it is only through the appearances of the psychological subject that the structure of the epistemic subject can be approached. Piaget’s empirical method is grounded in the recognition of this relationship. His interview technique (sometimes described as clinical and sometimes as critical) is essentially a means for allowing the structure to appear. The questioning of the subject (who is here a particular individual and who often remains unconscious of the structure which is finding its articulation through his performance) is designed to probe the limits of the subject’s thought, to examine what thinking is possible for the subject. Thus the performances of individuals can be analysed to reveal the structure which makes these performances possible (and here the separation of structure and appearance is similar to the distinction between competence and performance found in cognitive psychology and linguistics). This method operates by respecting the separation between appearance and structure, psychological subject and epistemic subject. These same remarks also pertain to the observational methods which Piaget employed in his studies of sensori-motor development. Indeed it is just this characteristic which forms the methodological continuity between the observational studies of infants and the clinical interviews of children. Thus as well as operating within the content of the theory the separation of appearance and structure is important to the construction of the theory. To posit a relationship between the epistemic subject and social life may appear problematic to the extent that the epistemic subject is indeed a centre of logical

Social life and the epistemic subject 71 activity, an activity which is traditionally conceived as being independent of social action. However, it is this very assumption which needs to be questioned and the starting point for this discussion is the idea introduced above of the epistemic subject as a transindividual subject. What exactly is intended by this formulation, a transindividual subject? For Goldmann the problem is one of establishing some basic parameters and categories for sociological analysis. In denying the “individual status of the subject of thought and action as a self-evident truth” (Goldmann 1971, p. 71), he wants to emphasise that the “complete subject of the action and, implicitly, the structure of consciousness can be comprehended only by starting from the fact that men act together – that there is a division of labour” (Goldmann 1976, p. 97). He gives a clear and precise example of three removal men moving a piano. It is totally impossible to understand what is happening he argues, if we: . . . suppose that one of them has the status of subject, since this would mean that we assimilate the two others to the piano as objects of the thought and action of the first. I would add that to see things this way means breaking the link between the consciousnesses of the individual considered as subject and the moving of the piano (which is not his individual action), this giving to his thought a contemplative status with regard to the piano moving. Clearly the only way of understanding the facts and restoring the link between consciousness and action is to admit that the three removal men make up together the subject of action which has the piano as its object and the moving as its result. (Goldmann 1971 p. 72) Thus for Goldmann a comprehension of the action of social groups is only possible in terms of these two interdependent concepts – a transindividual (collective) subject and a corresponding division of labour. The idea of the individual consciousness as the sole focus of thought and action becomes impossible. There are indeed phenomena which Goldmann recognises as those of an individual subject, they are “everything which Freud designated as being of the order of the libido” (Goldmann 1976, p. 97). Yet so long as individual consciousness is understood as being the sole locus of thought and action then such simple facts as these three men moving a piano will remain incomprehensible. Comprehension only becomes possible in relation to the idea of a transindividual, collective subject. The idea of such a collective subject has long been accepted in sociology in association with an understanding of the division of labour in co-ordinating the action of social groups. In psychology, however it has only made very rare appearances, even in social psychology. No doubt this is in part because psychology has considered itself as being concerned with precisely those actions which seem to be achieved by an individual acting alone and in isolation. Yet such a position is increasingly becoming untenable. Where once there was the perspective of the individual subject achieving some psychological development there is

72


a growing recognition of the importance of the coordination of actions among individuals; that there is in effect some psychological counterpart to the division of labour. A good example would be the study of the acquisition of language which is no longer seen from the Chomskian perspective of an innate language acquisition device enabling the child to extract linguistic rules from the speech to which he is exposed. Rather this acquisition has been set in the context of the interaction of the infant with those others around him in such a way that the acquisition is predicated on the collective action of the infant with these others (cf. Bruner 1974; Bates, 1976). An earlier use of this idea can be seen in Vygotsky’s argument that a child’s level of intellectual development needed to be assessed not simply in terms of what the child could achieve alone but also in relation to his performance in co-operation with adults. This led Vygotsky to the concept of the “zone of proximal development” (Vygotsky 1962, p. 103). Perhaps the best example for the present discussion is the Genevan work on social interaction and cognitive development (cf. Doise 1978, Perret-Clermont 1980). These studies have emphasised the importance of the external co-ordination of actions with others for the child’s construction of internal co-ordinations of actions, that is, for the construction of operations. It is these operations which are themselves the structures of the epistemic subject, and the movement from external to internal co-ordinations of actions would seem to suggest that this development is itself the product of collective action. To follow Goldmann once again in this case and insist that the subject “has the function of rendering the facts we propose to study intelligible and comprehensible” (Goldmann 1976, p. 92) means to insist that from this developmental genetic point of view the epistemic subject must be considered as a transindividual subject.1 To speak of the epistemic subject as a transindividual subject emphasises again something noted earlier, that this subject is not an individuated subject. One of the important limitations of genetic epistemology is precisely that it does not claim to offer any comprehension of the particular individual, of the person as a psychological subject. This limitation is an essential feature of the epistemic subject. It operates as a boundary for the area of social life with which this perspective is concerned. It does not deny the reality of other, non-cognitive processes. Piaget himself frequently stressed that every psychological action was both a cognitive and an affective action. In the individual, psychological subject, both aspects appear fused together. In seeking to deal with the former genetic epistemology has necessarily recognised its lack of comprehension of the latter. From this perspective indeed recognition of the reality of affective processes is almost as far as its comprehension of them extends.2 The question must be posed, then, as to whether such a limitation is injurious or even fatal for the project of a genetic epistemology. Surely to admit the existence of psychological processes and structures which remain unknowable introduces a dualism which threatens to render the conception of the epistemic subject a purely ideal fiction, inseparably divorced from those concrete persons through whom it is assumed to operate?

Social life and the epistemic subject 73 In so far as this criticism amounts to saying that genetic epistemology does not furnish a psychology of the whole person it is surely correct. Yet, paradoxically, this critical statement also expresses one of the great strengths of this perspective, namely the clear sense which it has of its own limitations, of where the limits of this perspective lie. The epistemic subject is an analytical abstraction, yes, but an abstraction from what? It is abstracted from the mass of activities which are the observable behaviours of children. It claims to abstract some sense by offering a framework for interpreting these events. A child of nine months reaches under a handkerchief which has hidden from him the watch he was grasping for. A month ago he simply let the watch disappear; now he acts toward this absent object. The construction of the object concept as an important stage in the development of the epistemic subject offers an understanding of this change. Why is the child interested in the watch? Or at a late stage when presented with different coloured shapes why begin to sort them by collecting together the red ones and not the yellow or the blue ones? To these questions genetic epistemology offers no answers. An altogether different frame of reference would be required to answer these questions: an affective one. The epistemic subject is concerned with the child’s relations with the world of objects and the means by and the manner in which these relations develop. In doing so it is of course claiming for these relations some autonomy relative to the totality of the child’s relations with the world. It deals with this world only as epistemic objects; it is not concerned with what particular kind of object it may be. For the truth of the matter is that if the epistemic subject is an analytical abstraction then the objects of its knowing are just as equally analytical abstractions. Thus when Piaget describes the development of the object concept he is completely indifferent: as to whether this object is a person (the proverbial ‘mother’) or a thing (the equally proverbial ‘watch’). Both are epistemic objects, and both are equally assimilated to the child’s schemes or knowing objects at whatever stage of development these may be. Piaget himself recognises that “just as people doubtless constitute the first permanent objects recognised by the baby, so also they are very probably the first objectified sources of causality” (Piaget 1955, p. 318). This phrase which almost casually identifies people as “doubtless” the first permanent object comes at the end of a long chapter in which he analyzes the sensori-motor development of causality. Indeed the very casualness with which the identification is made is revealing. For Piaget the importance of the object does not lie in whether or not it is a person. The same casualness in identifying people and things as possible objects appears in another passage where Piaget is explicitly stating what he means by an ‘object’: I call an object a polysensory complex which can, consequently, be simultaneously seen, heard, touched etc; but a polysensory complex which, in the eyes of the subject, continues to exist in a durable way beyond all perceptual contact. The criterion for an object is, then, its permanence at a time when it has disappeared from the sensory field. As long as there is perception, while the infant sees, hears or touches people or things, we cannot know if there is

74

Development as decentration an ‘object’ or not: there are simply what we call perceptual tableaux (I don’t speak of images because this would be confused with the mental image) – the child sees the perceptual tableaux which are alive, which move, which have all the same characteristics as perception at every stage. But a world of objects signifies something more: it signifies that something continues to exist after the disappearance of the perceptual tableaux. On the other hand a world without objects signifies that when a perceptual tableau disappears, that is to say leaves every perceptual field, then it no longer corresponds to anything which could be localisable. (Piaget 1954, p. 59; my translation, emphasis added)

This passage clearly expresses the sense in which the object has, for Piaget, an epistemological significance which is untouched by any affective valorisation which this thing may have for the subject. For something to become an object means only that it is conceptualized by a subject in a particular way. The only values recognised in this epistemological discourse are those which are implicit in the operation of the schemes themselves (something is, for the child, something to be seen, to be touched etc); affective valorisations belong to a different order. If the epistemic subject is limited to mere recognition without comprehension of affective processes then the epistemic object is equally limited to being an object invested only with a bare minimum of affective interest. This limitation of the epistemological perspective is not peculiar to genetic epistemology but extends throughout cognitive psychology, even where it goes unrecognised or ignored. To stay for a moment with the same developmental process, a convergence has often been noted between Piaget’s analysis of the development of object concept and the appearance and development of infant attachments. In speaking of ‘attachments’ there is the implication that in some way affective processes are being considered since they concern the child’s relations with other people. Yet even here there is little beyond a bare recognition of affective processes, with little comprehension of them. There is certainly some assertion that the child’s bond to an adult provides some security for the child’s explorations of the world around him; but beyond this only some generalisations about the importance of such early bonding for the child’s later ‘adjustment’. The comprehension of affective processes seems to be equally limited in cognitive accounts of attachment. The relations between these two orders, the epistemic and the affective, remain obscure. One crucial difference between them concerns the function of memory. For the epistemic subject at any level few traces remain of earlier stages, however necessary they may have been for the achievement of the present level. Thus the child who asserts the conservation of quantity will deny that two months ago they asserted an inequality of quantities. The cognitive memory does not conserve its own developmental processes (cf. Piaget and Inhelder 1973). For affective processes this is not the case. From Freud onwards psychoanalysts have insisted on the persistence of earlier affective relations in later affective processes. This distinction seems to drive a wedge between these complementary facets of psychological

Social life and the epistemic subject 75 activity, preventing their being jointly comprehended in a single frame of reference; no matter how much the influence of affective processes on cognitive development may be recognised it remains recognition and not comprehension. I have tried to argue that this limitation is rooted in the conception of the epistemic subject itself. As a transindividual subject it cannot also be an individuated subject. The distinctive concern of genetic epistemology is with the development of the epistemic subject and it is this concern which establishes, as it were, its theoretical terrain or frame of reference. Of course, like every theoretical perspective it may be distinguished just as much by what it obscures from vision as by what it brings to light; every perspective, since it is grounded in some point of view, is limited by its own understanding of what it regards. Such limiting effects of the frame of reference become more important (in the sense of being more pertinent) when what once stood on the fringes of the perspective becomes assimilated into the central theoretical structure. This problem is confounded because so much of the recent research on ‘social cognition’ has originated from within the Anglo-American tradition of psychological thought. Despite their recognition of Piaget as a major influence some essential aspects of the Genevan project have not been clearly perceived by the Anglo-American tradition. Particular results and even theoretical formulations continue to be extracted in isolation from the wider theoretical structures on which their significance depends and in terms of which their meaning can be understood. So long as research within the Anglo-American tradition remained concerned with apparently the same substantive topics as those addressed by the Genevan research (the development of logic, of number, of space, etc.) the issues of conflicting theoretical perspectives appeared as largely abstract questions, somewhat removed from the practicality of empirical research. However, in concerning itself with ‘social’ cognition the Anglo-American tradition has moved into an area which was not only relatively unstructured in its own terms but which also lay outside what it had taken to be the Piagetian position. In structuring this area cognitive/information processing theories have had to fall back on their own presuppositions so that the structuring qualities of its own theoretical perspective have become more prominent. Questions of conflicting theoretical perspectives no longer appear, therefore, as abstract issues but as being rather more intimately concerned with practical research. These cognitive/information processing theories are faced with the problem of structuring for themselves the relationship between social life and their conception of the subject. From the point of view of the subject the most important development has been their acceptance of a fundamental division between ‘social’ and ‘non-social’ cognition, a problem which is considered in detail below. From the point of view of the conception of social life the problem has been, if anything, even more acute, posing as it does the relationship between psychology and sociology. Typically social relations have figured in this tradition as a set of objective categories external to psychological structures and processes themselves. The ‘social variables’ are the archetypal independent variables for the analysis of variance model which has not only dominated the empirical investigations of this

76


tradition but has also had a pervasive influence on its theorising. Thus variations in group membership, whether social class or cross-cultural, are seen as variables constraining the distribution of psychological ‘structures and processes’. Indeed the only thing which distinguishes these ‘social variables’ from that other archetypal independent variable, ‘age’, is the assumption that age differences are developmental, that the child at any age is progressing towards another age level. Even this difference can sometimes appear to be illusory; where a similar developmental hierarchy is imputed to social or cultural differences (the concept of ‘deprivation’ seems to imply just such a sense of thwarted progress towards the ideal). This rather rigid separation of objective social factors from subjective psychological structures has begun to break down in the field of ‘social cognition’ through studies of social interaction (especially those concerning infant development). In restructuring the relationship between social life and their conception of the subject of psychology social relations are beginning to be seen as penetrating the psychological subject itself. The relationship has moved away from the rigid separation of subjective from objective without yet achieving its own balance; it has become a problematic relation. At present the problem is largely restricted to a consideration of the child’s relations with those others (peers and adults) immediately surrounding him; although a more general set of sociological considerations seem to loom on the horizon. If the relation of the psychological to the social has been problematic for the Anglo-American tradition there is also some justification for their perception of a dislocation between the Genevan work and the analysis of social life. But it is only a partial justification; for although Piaget and his collaborators did not present a systematic account of how the development of the epistemic subject was situated within social life this remained a constant theme in their work, ever present yet lacking a coherent analysis (although it must be added that Piaget’s major text on social relations, the collection of Etudes Socioloqiques, (1977), remains untranslated into English*). Nevertheless from the outset the Genevan enterprise did conceive of some relationship between psychological structures and social relations (cf. Piaget 1932). Indeed it would be quite impossible to give an adequate account of genetic epistemology without some comprehension of how the epistemic subject is situated within social life. For genetic epistemology social relations cannot be reduced simply to a set of objective categories. Their relationship to psychological structures is not a purely external one. It is only by analysing this relationship that some systematic approach may be developed towards the issues of ‘social cognition’.

3 ‘Social’ and ‘non-social’ cognition It is now possible to comprehend exactly what is intended by using the term ‘social’ as an adjective to qualify the noun ‘cognition’. In general the intention has

*

Editors’ note: Piaget’s Sociological Studies are now translated, see Piaget, 1995.

Social life and the epistemic subject 77 been to distinguish ‘social cognition’ from some other form of cognition which has been labelled either as ‘non-social’ or ‘physical’ cognition. According to this distinction cognition is ‘social’ when it concerns people and ‘non-social’ when it concerns other things. It is taken to be apodictic that cognition about people differs in some way from cognition about things. Yet clearly, from the argument of the previous section, any such distinction between ‘social’ and ‘non-social’ cognition cannot be construed in epistemological terms; the epistemic object is not divided between people and things. At least this is the case for the Piagetian position. However it is precisely this position which is criticised by those authors who subscribe to a fundamental division between ‘social’ and ‘non-social’ cognition. Their argument is generally that while Piaget’s analysis may be adequate for the child’s comprehension of the physical environment it fails to take account of some fundamental features of the social environment. Typically it is suggested that people show a greater degree of variability or of indeterminacy in their behaviour, than do physical objects, and that the operations described by Piaget are insufficient for the comprehension of this greater complexity (cf. Glick, 1978). When these arguments are examined more closely, however, it seems that the insufficiencies are to be located not so much within Piaget’s position as within these authors’ representations of Piaget’s position. What is at issue is whether the cognitive structures pertinent to the cognition of social life differ essentially from those which bear on physical objects. It will be recalled that in speaking of the epistemic subject as a transindividual subject it was also stressed that in its construction, it is itself a social product. Interactions with other people as much as with physical objects form the prerequisites for the construction of operations. In this sense even those structures which are seen to be exclusively concerned with the physical environment are themselves, in part, the product of social interactions. The present argument turns on whether or not these same structures are sufficient for the cognition of social life or whether some other types of cognitive structure are necessary. The representation of Piaget’s position seems to rely on a criticism of the attempts to extrapolate the ‘three-mountain test’ from being concerned with roletaking in the sense of spatial perspectives to role-taking as the central process in ‘social’ cognition. Yet this extrapolation and the experimental work to which it gave rise was not undertaken by Piaget! It was, rather, the product of American commentators on Piaget, having its source primarily in the work of Flavell (cf. Flavell et al. 1968). If these attempts to ground ‘social’ cognition on roletaking as the fundamental human social phenomenon have not been entirely successful this should not be attributed to a failure in Piaget’s theory. To be sure he always argued for the importance of the child’s exposure to the thinking of other children, of the child’s confrontation with a world of thought external to his own as a necessary moment in the dissolution of the young child’s egocentrism. This is not however the same as the ability to identify the other person’s point of view, which is the ‘operational’ definition given to role-taking in the American experimental work. The Genevan studies on social interaction and cognitive development have emphasised the importance of cognitive conflicts in the

78


development of operations (cf. Perret-Clermont 1980) which are not simply reducible to one child identifying the point of view of another. In these experiments pairs of young children frequently achieved collective solutions to problems which went beyond the initial capacity of either child (cf. PerretClermont 1980; Mugny and Doise 1978). In these social interactions the essential mechanism was not ‘role-taking’ but rather the coordination of actions achieved by the subjects acting collectively. Much of the criticism of Piaget is in fact the criticism of a representation of Piaget which is itself a departure from Piaget. Two points are essential to that position; firstly the view that the cognitive operations described by Piaget are in some way limited in their application to physical objects, and secondly that in some way ‘social’ cognition is different from ‘physical’ cognition in so far as it necessitates a comprehension of ‘unobservable internal causes of behaviour’. Neither of these points corresponds to a Piagetian position. The second point is the more straightforward to comment upon. The implication is that ‘physical’ objects behave under the constraint of external forces which thereby distinguishes them from ‘social’ objects whose motivating forces are ‘internal’. Yet this is not the case for Piaget’s analysis of cognitive operations, which are precisely directed at comprehending transformations which remain ‘invisible’ to the subject. It has already been emphasised that ‘number’, for instance, is not a given feature of any collection of objects but rather a cognitive construction. The same might also be said of any of the other conservations studied by Piaget. Weight, for example, is not conserved simply because of the inertia it presents to perception. The conservation of weight, and thereby the establishment of weight as a metrical order, is the product of cognitive operation and not an ‘objective’ feature of things. The point here is that every conservation is focussed on some feature of an object or collection of objects which remains constant across an apparent transformation. As Piaget notes: . . . the object is merely an instant cut from the continuous flux of causal relations. Sooner or later reality comes to be seen as consisting of a system of transformations beneath the appearance of things. (Piaget and Inhelder 1971, p. xiii) It might be objected here that while these conservations might be unobservable they should not be construed as ‘causes of behaviour’ and that it is this sense of the causes of behaviour which is the important differential between ‘physical’ and ‘social’ cognition. Yet this difference is equally illusory. That things fall downwards under the influence of gravity may well be understood as the operation of some external force on the thing by an adult, but it is not necessarily apparent to the child. From the point of view of the ‘social’ the child’s experience of the world is frequently that his own behaviour is constrained by the external influence of adults. To sum up it is clear that for Piaget all operations are aimed at comprehending ‘hidden’ features in the behaviour of an object, ‘physical’ or ‘social’ whether these features are held to be ‘internal’ or ‘external’ to the object.

Social life and the epistemic subject 79 It is the first point, however, which is the nub of the argument: are the cognitive operations described by Piaget tied exclusively to the comprehension of physical objects? From within the terms of genetic epistemology this criticism is easily rebutted. The object is always an epistemological construction and as such is neither a ‘social’ nor a ‘physical’ thing. It is important to distinguish the epistemic object from the physical or material thing (whatever kind of thing this might be, a chair or a person). As an epistemic object something exists for the child in so far as it is assimilated to whatever operative structures are available to the child. Thus it may be said that the epistemic object develops as the child grows up (whereas clearly the material thing does not). For example a collection of stones remains a collection of stones whether they are seen by a child of three or a child of twelve. What the child may know about these stones will however vary between these ages. They may vary from being ‘a lot of stones’ to being ‘eight stones’; or from ‘things that make a splash when dropped into water’ to ‘things which displace an equal volume of water’. In short it is necessary to distinguish between the subject’s knowledge of an object and the material thing itself. The stones do not change qua stones as the child’s understanding of them develops. For Piaget the object exists as the subject’s cognitive construction, as an object of thought: it is not to be confused with the physical reality of something. In this sense genetic epistemology does not recognise any fundamental distinction between ‘social’ and ‘physical’ (fundamental from the epistemological point of view that is). The cognitive operations therefore cannot be considered as being tied exclusively to the comprehension of ‘physical’ objects. One can consider an everyday thing such as a chair. It certainly is a physical object constructed of wood, glue, cloth etc. But it is also a social fact. The existence of a chair presupposes a society which has produced it, as well as being a material thing the chair is also a commodity, something to be exchanged, something which has a use-value and an exchange value. It further has a cultural value in the sense of being made in a certain fashion, in a certain style. These are features of chairs which are never considered within theories which have emerged from the cognitive psychology/information processing paradigm.3 The chair is to be considered only as a ‘physical’ object. But even here, those features which are held to be the clearest expression of the chair as a ‘physical’ object, its size, shape, colour, spatial disposition etc, even these are themselves social facts. The history of chair design shows clearly the cultural influence on these ‘physical’ properties of a chair. A chair is not a natural object but a cultural product whose every feature has been mediated through the culture which produced it. The bias towards the materiality of things evident in the cognitive/ information processing paradigm can be clearly seen in the sense which the word ‘object’ has for these theories. In such theories the object remains the material thing and the child becomes cognizant of it in so far as features of that object are encoded into a representation; thus the object exists for the child as a cognitive representation of a material thing. Piaget, however, draws a distinction between two senses of the term representation. There is the wider sense in which representation is more or less equivalent with thought in general and there is re-presentation in the narrower,

80


more literal sense of making present something which is absent (it is this latter sense which Piaget refers to as the symbolic function. cf. Piaget 1951; Furth 1969). Thus Piaget distinguishes between the knowledge of an object and its re-presentation. The cognitive activity which produces knowledge of the object does not necessarily result in a representation, although the action of the symbolic function may make much representation possible. The object which is known by the child is, therefore, neither identical with the thing itself nor is it a representation of that thing: it is an epistemic object. Piaget has, of course, often been accused of holding to an idealist epistemology,4 although, as Goldmann (1969) points out, his position stands within the tradition of dialectical philosophy, that is with the Hegelian and Marxist tradition. In this interactionist perspective an object exists for a subject only in so far as it is known by that subject, such knowledge being derived from the subject’s action on the world, from the coordinations of those actions and from collective co-ordinations between the actions of the subject with the actions of others. It would seem, therefore, that the charge made against Piaget (i.e. that the cognitive operations which he describes are specifically bound to the world of ‘physical’ objects) actually says rather more about the inadequacies of information processing theories than it does about Piaget. The lack of any clear distinction between the material thing and the epistemic object means that for these theories the object known by the child is indeed forcefully bound to the materiality of the physical object, since it is this material object which is held to be represented in the child’s cognitive structure. For Piaget the known object is an object of thought and not the material object. The cognitive activity which produces the knowledge of an object is structured in terms of operations which are themselves instruments of knowing and as such specific to neither the ‘physical’ nor the ‘social’ world. They are a means to knowledge rather than knowledge itself (just as logic may be described as a means to knowledge rather than being itself knowledge of the world). The use of the term ‘the cognition of social life’ rather than ‘social cognition’ flows from the conclusion that there are no grounds for distinguishing ‘physical’ and ‘social’ cognitive operations. Rejecting any such distinction does not, however, entail the assertion that people do not differ from things! What is being argued is that the means for knowing people or things are the same in either case, the cognitive operations. If the ‘social’ and the ‘physical’ cannot be distinguished in terms of operations it is nonetheless true that the products of these operations (that is, the knowledge of the world produced by the operations) does include a distinction between people and things. What is crucial here is the distinction between the operations themselves and the knowledge which they produce. It is largely a failure to appreciate this distinction which has led to attempts at separating ‘social’ from ‘physical’ cognitive operations. Distinguishing operations from their products is another expression of the distinction between form and content. While it is true that these two aspects of cognitive structure are only progressively differentiated in the course of development this is not equivalent to suggesting that diverse contents are derived from diverse forms.

Social life and the epistemic subject 81 There is one area in which people do differ from things, namely in their capacity for social interaction (Damon, 1981). It is right to emphasise that interactions between people are different from interactions between people and things. In pursuing his analysis through an examination of social exchanges Damon has also clearly been influenced by Piaget (1968, 1995) whose own analysis of social exchanges is constructed in terms of valorisation and the exchange of values (cf. also Kitchener 1981). Social life constitutes a field of valorisations, and it is the influence of these valorisations which serves to distinguish the pure, or formal structures of the epistemic subject from the structures implicated in the cognition of social life. In this respect the distinction between operations and their products will have to be examined further in relation to the cognition of social life and the role of valorisations. The essential question here is to consider what kind of structures are in fact implicated in the cognition of social life. Having rejected the opposition of ‘social’ to ‘physical’ cognition, and with it the term ‘social cognition’ in favour of ‘the cognition of social life’, it remains to consider what sense may be attributed to the word ‘social’. The arguments of this section have suggested that any knowledge which derives from the developing structures of the epistemic subject will have a ‘social’ character. In this sense all knowledge may be considered a social construction. It may be objected here that, since all knowledge is derived in some way from the operation of the epistemic subject, the word ‘social’ has become so overextended that lacking any possible contrasting term, it has ceased to signify anything specific. There is some basis to this criticism, but only in so far as ‘social’ is used in this most general of senses. But even here its use is not entirely without foundation, for if it fails to find any contrasting term within the context of these theoretical arguments it nevertheless serves the purpose of contrasting these arguments with others which, in one way or another, continue to deny any social basis to knowledge, or refuse to acknow1edge the implications of this basis. In this most general of senses, then, the term ‘social’ has more of a polemical character than substantive significance. There is also, however, a narrower sense of the word ‘social’ in which it signifies precisely those phenomena which are produced only through the relations which are established between people either as individuals or as groups. At the more global level of social relations between classes, or between cultures, it is possible to examine the way in which these social relations affect or constrain the distribution of cognitive structures. These are the problems which have been studied under the rubric of the ‘sociology of knowledge’. They may concern the distribution of cognitive operations themselves (although this always raises problems of the form in which these operations are being assessed; cf. Perret-Clermont (1980), who considers the question from the point of view of achievement at school; and Seagrim and Lendon (1980) who consider the problem in the context of a specific non-western culture, that of the Australian Aborigines). The distinction between the more general and narrower sense of the term ‘social’ can be expressed in the following way: the cognitive operations of the epistemic subject are undoubtedly social in their construction, in the sense that they would be impossible without social interaction. Nevertheless once these

82


co-ordinations have been achieved and interiorised it is possible for the individual to use them without recourse to social interaction, and the knowledge thus produced will be ‘social’ only in the most general sense. On the other hand there are phenomena which are the direct products of interactions between people which are always ‘social’ in the narrower sense of the term. It is these phenomena which form the objects of the cognition of social life.

4 Structures in the cognition of social life The argument of the previous section entailed the rejection of any fundamental epistemological separation of what is ‘social’ from what is ‘physical’. The basis for that rejection lay in the identification of a specifically epistemological level in the analysis of any cognitive activity. At this level, where considerations are restricted to the relations between the epistemic subject and the epistemic object, neither subject nor object is specified as a particular individual. The identification of such an epistemological level is not in itself sufficient, however, to establish the terms within which it is possible to describe those structures which may be discerned in the cognition of social life. There is more involved in this process than epistemological considerations. It was also argued that social life is constituted as a field of values so that pursuing an analysis of the cognition of social life entails taking account of processes of valorisation as well as epistemological processes. If epistemological relations are a necessary aspect of any cognitive activity they do not, as Piaget recognised, provide a sufficient analysis for the psychological subject, that is, for what Apostel (1982) calls the ‘full subject’. It is pertinent to ask what relationship exists between the epistemic subject and the psychological subject. The notion of the epistemic subject is essential to psychological analysis, that is, to the understanding of the behaviour or conduct of any psychological subject. Given such a model the activity of the epistemic subject can be discerned within every type of conduct of the psychological subject, whether that conduct consists of tackling a problem about numbers or a problem concerning people. The activity of the psychological subject only becomes comprehensible through a reference to the epistemic subject. As an example, it is useful to consider that in thinking about people important differences arise which are due to the gender of the persons being thought about. Gender refers to something more than the biological determination of a person’s sex; it refers fundamentally to the articulation of sexual differentiations within the social relations of a culture. Thus gender may be considered as a prime example of a process of valorisation; to be perceived as male or female is to be allotted a particular value within the value system of a culture. The relations of the epistemic subject to the valorisations associated with gender are twofold. On the one hand the separation of persons into the categories of male and female is a process which is made possible through the operations of the epistemic subject, just as any system of classification is dependent on logical operations (and the categories of male and female appear to be one of the first classification systems which develops in children). On the other hand, the articulation of these categories with the value

Social life and the epistemic subject 83 system of our culture produces the effect noted above, that conceptions of friendship are not generalizable across genders. In this respect it may be said that the valorisations associated with gender differences function so as to restrict the generality of conceptions about persons; in the cognition of social life there are many aspects in which conceptions of persons differentiate between male and female. This brief example, as well as demonstrating some of the complexities involved in looking for structures in the cognition of social life, illustrates one way in which comprehending the conduct of the psychological subject depends on a reference to the epistemic subject. In this example the process of valorisation is an important mediation between the epistemic subject and the psychological subject. Representations of gender arise through the interactions between the epistemic subject and the value systems present in the social life of a culture. This state of affairs can be clarified, perhaps, by considering it from the point of view of the psychological subject. The object which is phenomenologically present to the psychological subject is a product of epistemic operations, experimental knowledge and valorisations. These valorisations are themselves of two distinct types: there are those which result from processes of social influence and those which result from what may be called personal influences (such influences deriving from “everything which Freud designated as being of the order of the libido” – Goldmann 1976, p. 97 – what distinguishes these psycho-analytic influences is their mode of operation, a mode which Freud called ‘primary process’). Each of these components may provide a distinct level of analysis, but they are not, generally sneaking, separated in the subject’s phenomenological apprehension of an object. There is in this sense a synthesising function of the mind which gives to these appearances a sense of wholeness, to borrow a phrase from Sartre this may be called a totalising function (cf. Sartre 1976), and these phenomenological presences may be described as totalisations. It may be instructive to consider some reflections on the object as a phenomenological totalisation. From this phenomenological, or existential point of view an object is always constituted in the context of the totality of a person’s world, it occurs within the lifespace of that person. It is only as an epistemological abstraction that an object can be considered neutral (or ‘objective’). In giving a place to the object in their world human beings give the object a value, valorisation is as constant a process as object construction. The object is always constituted existentially as an object-with-a-value (or one might choose to say that an object is always constituted ‘subjectively’). No object or better no knowing-of-an-object, is indifferent or neutral because no object can be known as a pure platonic form. Form is always an abstraction; everything presents itself as concrete, as content. Our knowledge of the world is given primarily as existential valorisation – Sartre’s formula “Existence precedes essence” may be paraphrased to read “Existence precedes epistemology”. We do not first know an object and then take up an attitude toward it. On the contrary our attitude is given in our existential construction of the object. Epistemology is always fused with valorisation at the existential level of lived experience. An object always has a value for the subject since it is

84


never known as it is in-itself but only as-it-is-for-the-subject (as for-itself rather than in-itself in Sartre’s terms). However sketchy they may be, these few reflections on the phenomenology of the psychological subject nonetheless emphasise that every conception of social life has both a cognitive and a valorised aspect. The object, as it is given for the psychological subject is always a valorised conception so that cognition and valorisation are complementary aspects of the psychological – as distinct from epistemological – construction of the object. Such complementarity recalls Piaget’s (1954) insistence that every psychological action has both cognitive and affective aspects, which, although they may appear inseparable from the point of view of the psychological subject may nevertheless be separated through the theoretical work of analysis. Should valorisation, then, be considered as belonging to the affective order in Piaget’s terms? To answer this question it is necessary to recognise a certain duality in the notion of values; a duality because values belong to both the cognitive and to the affective orders. In so far as values function so as to make a psychological action more or less likely, or even to make such actions possible at all, then they must be considered as constituting one element of what Piaget terms the ‘energetics’ of psychological action, and thus to belong to the affective order. In distinguishing affectivity from cognition Piaget argues that while the cognitive aspect of psychological action may be described in terms of structure (that is a closed totality with transformation rules etc.) there are no affective structures as such. Yet such a distinction takes account only of the epistemological structures of cognition, it does not deal with those cognitive structures which include descriptive knowledge of the world, that is, with cognitive formations in which epistemological structures are fused with particular contents. The definition of structure which Piaget uses to characterise the cognitive aspect of psychological action is a definition of content-free (and hence value free) structures. It does not apply to cognitive structures in which the logico-mathematical structuring elements are present together with particular content. Thus Piaget writes: We recall that affectivity cannot modify structures; it intervenes constantly in the contents. That is the interest (affective) which, for example, makes the child’s choice of objects to seriate; it is again affectivity which facilitates the success of the operation of classification, or makes it less difficult. But the rule of seriation remains unchanged. It is understandable that, in so far as the structure of operations is not very distinct from their content, there may be confusion. (Piaget 1954, p. 10, my translation) There may indeed be confusion when we are concerned with the cognition of social life where the separation of form from content is certainly less than distinct. We are presently engaged in a discussion of just this confusion; for, to return once again to the example of the representation of gender, we have to deal with aspects of mental life which cannot be reduced to pure form (logical structure) nor to pure

Social life and the epistemic subject 85 content (they are structured aspects of mental life) nor yet consigned to the affective order (they have a cognitive component). Wittgenstein speaks about “the greater ‘purity’ of objects which don’t affect the senses, numbers for instance” (Wittgenstein 1980, p. 26) and one feels that Piaget would have good reason to agree with this observation. Nevertheless it seems clear that in considering the cognition of social life we must be concerned with hybrid structures in which form and content are not separable to the same degree that one can separate the number of objects in a group from the type or types of object in a group. If it is possible to speak of structures in the cognition of social life then it is on the grounds that any such structures will modify the closed totalities which form the structures of the epistemic subject. But if a wider definition of cognitive structure is accepted what then becomes of the distinction between affectivity and cognition described above? Again this question may be considered within the specific context of representations of gender. The analysis of gender representations suggests that conceptions of social life which exist within the culture which forms the environment of the psychological subject are capable of being internalised, through a process of valorisation, so that they function as an energetic principle for the psychological subject. The transformation of cognitive structure into affective process, while it may take place within a psychological subject, is above all a relationship between a social group and an individual psychological subject. At this point it is possible to distinguish two perspectives on the cognition of social life, one of which may be described as a ‘vertical’ dimension and the other as a ‘horizontal’ dimension. The vertical dimension refers to the conceptual structure of the psychological subject. The greater part of this discussion of structures in the cognition of social life has focussed on this vertical dimension by implicitly taking as its focus the need to account for the conceptualisations of social life as they appear in the individual psychological subject. Although that discussion assumed that such conceptualisations would not be idiosyncratic to each individual but shared between psychological subjects it did not envisage any social process corresponding to this ‘sharing’ of conceptual structures. Pursuing the analysis of the process of valorisation has resulted in a comprehension of its mediating role between the processes in the social life of groups and conceptual structures in the psychological structure. It is the processes in the social life of groups which form the horizontal dimension, and a consideration of representations of gender showed that this horizontal dimension exerted influences on the individual conceptualisations of the vertical dimension which could not be analysed as a product of the activity of the psychological subject acting alone. This horizontal dimension has been the province of social psychological research, as it appears, for instance, in the work of Tajfel (e.g. Tajfel 1978; Tajfel and Forgas 1981). Tajfel presents a theory which concerns the cognition of values, that is, of how the valorisation of social categories affects the psychological structures and actions of individuals. For Tajfel this process is itself embedded within the relations between groups in a society such that the inter-group relations affect the differential valorisations made by particular groups. Tajfel’s argument is that there are actions of individuals whose comprehension only becomes possible in terms of the relation

86


between groups in society; such intergroup phenomena generally taking the form of ingroup favouritism and discrimination against the outgroup. Distinguishing these two dimensions is of considerable importance since developmental psychologists have tended to work exclusively within the vertical dimension to the neglect of a consideration of the social processes of the horizontal dimension. Developmental studies of social cognition have tended to search for forms in the conceptualization of social life to the detriment of a concern with content. This predilection for the analysis of form as distinct from content has a history in developmental psychology as a prevailing concern for the analysis of cognitive structure. The emergence of the cognitive paradigm from the dominant influence of behaviourism centered almost exclusively on a concern with logico-mathematical structures; a concern which was easily translatable into the terms of information theory. If Piaget’s work had an influence on the emergence of the cognitive paradigm in American developmental psychology it was above all due to the clear focus which his work gave to logico-mathematical structures. For Piaget though, such structures refer to the epistemic subject which he was careful to distinguish from the psychological subject. The cognitive paradigm made no such distinction choosing to ignore this aspect of Piaget’s work in favour of an abstract conception of ‘the child’ which is both epistemic and psychological subject combined. The result has been a developmental theory constructed in terms of an individual child coming to understand the world around him through the development of cognitive skills or strategies. Lacking any distinction between the epistemic subject and the psychological subject the cognitive paradigm, when faced with problems of the cognition of social life identified an epistemic difference between classes of objects such that the world of relations with things was separated from the world of relations between people. On the other hand, social psychological studies of the horizontal dimension have shown consistent concern for content as well as form even where it has produced formal statements of theoretical positions (as in attribution theory, for example). Tajfel’s work on the interrelationships between cognition, values and group processes in the examination of phenomena of social categorisation has already been noted. Another example is the work of Moscovici. What is common to these different authors is the proposition that in many important respects the actions of individual psychological subjects cannot be comprehended except through the intervention of social processes in the cognitive structures of individuals. As I have argued above, the analysis of the cognition of social life requires the comprehension of such social influences if conceptions of social life are themselves to be understood as social realities, as the products of social processes. Conceptions of social life have content as well as form. The contrast between these two dimensions may perhaps be stated by saying that the cognitive structures analysed in the vertical dimension have a purely contemplative relation to social life itself. When the functioning of these structures is investigated in a horizontal dimension the structures appear always as valorised conceptions of social life. The analysis of the cognition of social life

Social life and the epistemic subject 87 may be said to be concerned with the point of intersection between these two dimensions, the vertical dimension of cognitive structure and the horizontal dimension of social process. For Moscovici this point of intersection is defined in terms of the concept of social representation. Although Piaget does distinguish the epistemic subject from the psychological subject he has not pursued in any detail an analysis of the relations between them. His discussion of the relations between cognition and affectivity (Piaget 1954) is really limited to a discussion of the relations between the epistemic subject and affectivity. Although he does discuss valorisations in both the vertical dimension (‘energetics’) and the horizontal dimension (his analysis of social exchange) he does not discuss the interactions at the point of intersection between these dimensions (the point where socio-cognitive structures are transformed into affective processes). A consequence of this has been that those studies of the cognition of social life which have used a more or less Piagetian perspective (e.g. Damon 1977; Furth 1980; Selman 1980; Youniss 1980) have tended to adopt an ‘individualistic’ tone. In these studies the focus is on the conceptual structures through which social life is represented by children in interviews. None of these authors however, have found anything remarkable in the fact that all their subjects develop a similar set of ideas. Their silence on this point gives the impression that they consider the development of conceptions of social life to be a ‘natural’ process. In proposing to deal with conceptions of social life as socio-psychological realities it is proposed to change this focus in order, as Moscovici puts it, “to see a phenomenon in what hitherto was viewed as a concept” (Moscovici, 1981, p. 184).

Notes 1 This recent Genevan work offers a necessary corrective to Piaget’s own discussion of equilibriation, that is the process whereby a given cognitive structure is transformed into a more developed structure (Piaget 1978). Piaget describes this process from within the confines of a single subject. In spite of his complex argument it remains difficult to comprehend how and why this subject comes to recognise the insufficiency of its present structure. Of course once this insufficiency is recognised then the problem of establishing a new set of relations is fairly straightforward. These recent Genevan studies suggest that such an insufficiency arises in the course of the child’s interactions with other children, that is in the course of attempting to co-ordinate his own actions with the actions of others. Again this point emphasises the transindividual nature of the epistemic subject. 2 Even in Piaget’s most extended treatment of affective relations (Piaget 1954, partial English translation 1981) they are dealt with as generalities; for all its undoubted interest the analysis which he offers would not suffice for the analysis of particular subjects, that is it does not make possible any ‘Piagetian’ psycho-analysis. 3 Recent studies of cognitive processes have, of course, considered the significance of the use of objects in the perception of them (cf. Rosch 1977, 1978). In these studies ‘use’ refers primarily to possible interactions between people and things considered solely as physical objects (thus a chair is something which can be sat on). This physical ‘function’ is then considered as a feature of the object relevant to the cognition of that object. In the present discussion more general questions are raised about which features of an object are relevant to its cognition. 4 Piaget himself recalls a discussion with some Soviet psychologists in which he was asked: “Do you believe that the object exists before knowledge?” I replied: “As a

88


psychologist I know nothing of this, for I only know the object in acting on it and I can say nothing about it before this action”. Rubinstein then proposed the conciliatory formula: ‘For us the object is a part of the world, do you believe that the world exists before knowledge?’ I then said (and not with reference to the subject): “This is another matter. In order to act on the object it is necessary for there to be an organism and this organism is also part of the world. I therefore evidently believe that the world exists before all knowledge, but that we only divide it up into individual objects through our actions and as a result of an interaction between the organism and the environment” (Piaget 1972, footnote to p. 203). This response was evidently sufficient to satisfy the Soviet psychologists that Piaget was no idealist!

References Bates, E. (1976). Language and Context: The Acquisition of Pragmatics. London: Academic Press. Bruner, J. (1974/5). From communication to language: A psychological perspective. Cognition, 3, 255–287. Damon, W. (1977). The Social World of the Child. San Francisco: Jossey-Bass. Damon, W. (1981). Exploring children’s social cognition on two fronts. In J.H. Flavell & L. Ross (eds.), Social Cognition Development. Cambridge: Cambridge University Press. Doise, W. (1978). Groups and Individuals. Cambridge: Cambridge University Press Flavell, J.H., Botkin, P., Fry, C., Wright, J., and Jarvis, P. (1968). The Development of Role Taking and Communication Skills in Children. London: Wiley. Furth, H.G. (1969). Piaget and Knowledge: Theoretical Foundations. Chicago: University of Chicago Press. Furth, H.G. (1978). Young children’s understanding of society. In H. McGruk (ed.), Issues in Childhood Social Development. London: Methuen. Furth, H.G. (1980). The World of Grown-ups. New York: Elsevier. Glick, J. (1978). Cognition and social cognition: An introduction. In J. Glick and K.A. Clarke-Stewart (eds.), The Development of Social Understanding. New York: Gardner Press. Goldmann, L. (1969). The Human Sciences and Philosophy. London: Jonathan Cape. Goldmann, L. (1971). Reflections on history and class-consciousness. In I. Meszaros (ed.), Aspects of History and Class-Consciousness. London: Routledge. Goldmann, L. (1976). Cultural Creation in Modern Society. Saint Louis: Telos Press. Hartup, W.W. (1978). Children and their friends. In H. McGurk (ed.), Issues in Childhood Social Development. London: Methuen. Kitchener, R.F. (1981). Piaget’s social psychology. Journal for the Theory of Social Behaviour, 11, 253–277. Moscovici, S. (1981). On social representation. In J. Forgas (ed.), Social Cognition. London: Academic Press. Mugny, G. and Doise, W. (1978). Socio-cognitive conflict and structuration of individual and cognitive performances. European Journal of Social Psychology, 8, 181–192. Perret-Clermont, A.-N. (1980). Social Interaction and Cognitive Development in Children. London: Academic Press. Piaget, J. (1932). The Moral Judgment of the Child. London: Routledge & Kegan Paul. Piaget, J. (1951). Play, Dreams and Imitation. London: Routledge & Kegan Paul. Piaget, J. (1954). Les relations entre l’affectivité et l’intelligence dan le developpement mental de l’enfant. Paris: Centre de Documentation Universitaire.

Social life and the epistemic subject 89 Piaget, J. (1955). The Construction of Reality in the Child. London: Routledge & Kegan Paul. Piaget, J. (1966). Mathematical epistemology and psychology. In E. W. Beth & J. Piaget (eds.), Mathematical Epistemology and Psychology. Dordrecht, Holland: Reidel. Piaget, J. (1968) [1971]. Structuralism. London: Routledge & Kegan Paul. Piaget, J. (1971). Biology and Knowledge. Edinburgh: Edinburgh University Press. Piaget, J. (1972). Insights and Illusions of Philosophy. New York: World Publishing Company. Piaget, J. (1978). The Development of Thought. Oxford: Basil Blackwell. Piaget, J. (1995). Sociological Studies. London: Routledge. Piaget, J. and Inhelder, B. (1971). The Mental Imagery of the Child. New York: Basic Books. Piaget, J. and Inhelder, B. (1973). Memory and Intelligence. London: Routledge & Kegan Paul. Rosch, E. (1977). Human categorization. In N. Warren (ed.), Advances in Cross-cultural Psychology (Vol. 1). London: Academic Press. Rosch, E. (1978). Principles of categorization. In E. Rosch & B.B. Lloyd (eds.), Cognition and Categorization. Hillsdale, NJ: Erlbaum. Sartre, J.P. (1976). Critique of Dialectical Reason. London: New Left Books. Seagrim, G.N. and Lendon, R.J. (1980). Furnishing the Mind. London: Academic Press. Selman, R.L. (1980). The Growth of the Interpersonal Understanding: Developmental and Clinical Analyses. New York: Academic Press. Tajfel, H. (1978). Social categorization, social identity and social comparison. In H. Tajfel (ed.), Differentiation between Social Groups. London: Academic Press. Tajfel, H. and Forgas, J. (1981). Social categorization: cognitions, values and groups. In J. Forgas (ed.), Social Cognition. London: Academic Press. Vygotsky, L.S. (1962). Thought and Language. Cambridge, MA: M.I.T. Press. Wittgenstein, L. (1980). Culture and Value. Oxford: Basil Blackwell. Younis, J. (1980). Parents and Peers in Social Development. Chicago: Chicago University Press.

6

Psychological development as a social process Gerard Duveen

Introduction The debate around the contributions of Piaget and Vygotsky for contemporary developmental psychology is usually constructed around an opposition between individual and social perspectives on the process of development, with Piaget construed as the individualist as against the social theorising of Vygotsky. There are good grounds, however, for resisting such a characterisation of these theorists. As I shall argue in this paper, Piaget has a stronger grasp of social processes than this view admits, while Vygotsky’s grasp of social processes is limited in some important ways which this view does not recognise. While a review of some of the key themes in the classic contributions of these two authors will certainly entail a consideration of social aspects of development, I also want to argue that both of them ignore some central issues in considering psychological development as a social process. To make this argument it is also necessary to begin to reconnect developmental psychology with social psychology. There is something really very strange in the way these two endeavours have become so radically disconnected. Strange, because both of these traditions are fundamentally addressing similar concerns. As Moscovici has noted (1990, p. 169), they have a: common point of departure and are animated by such close underlying issues. Their different traditions and methods allow a deep similarity to show through and tie them together. It is as if social psychology and developmental psychology were concerned with the same thing, the former in space and the latter in time, the first by way of the exterior and the second by way of the interior. If the problem for developmentalists is to understand how the child develops as a social actor, social psychologists forget at their peril that every social actor has a developmental history whose influence cannot be ignored. In actual fact, although the designations ‘social’ and ‘developmental’ have come to signify distinct categories in the world of psychology, from the perspective of a constructivist epistemology the distinction necessarily collapses. Both Moscovici and Piaget share a common epistemological stance. The world that we

Psychological development as a social process 91 know is the world as we have constructed it through our psychological operations. Piaget elaborated and defended his constructivist – or genetic – position against the claims of both a priorism and empiricism. Knowledge is neither the product of inherent characteristics of the mind, nor simply the reflection of environmental influences. Rather, for Piaget, knowledge develops in children through their interaction with the environment, in the course of which they first come to co-ordinate their own actions and then to abstract more general operations from these co-ordinations. Similarly Moscovici (1972) elaborated and defended what he described as systematic social psychology against the claims of social psychological theories based on a priori or empiricist epistemologies. In a later work Moscovici (1976a) came to characterise his position as exemplifying a genetic social psychology, and the very use of this word genetic, so redolent with Piagetian overtones, ought to alert us to the harmony between these two authors. And it is in their common genetic approach that the harmony is heard most clearly. A central assumption of genetic theories in psychology is that to understand something we have to understand the process through which it has been produced, that is we have to grasp its developmental construction. This assumption has been most clearly articulated in the classic texts of Piaget and Vygotsky, but it has also been evident in the theory of social representations. If we want to understand social representations we have to understand the processes through which they are produced and transformed. For Moscovici this means not only the processes of anchoring and objectification,1 but also, as Doise (1993) has recently reminded us, the processes of propagation, propaganda and diffusion. In commenting on a collection of developmental studies of social representations, Moscovici (1990) suggests that there are two approaches to analysing the processes through which representations are constructed. The first, which he suggests is the more typical in social psychology, he describes as Bartlett’s way, since it proposes to analyse the construction and reconstruction of representations as they pass from one social group to another. The second approach he characterises as Vygotsky’s way (while noting that it is also Piaget’s way, though, as we shall see, while they may share some characteristics Piaget’s way is not always the same as Vygotsky’s way), and it seeks to understand representations by analysing their successive transformations through the different phases of childhood and adolescence. While Moscovici’s comment suggests that these are two ways to achieve the same ends, there are nevertheless differences between Bartlett’s way and Vygotsky’s way. Bartlett’s way has led to investigations of the processes through which representations are organised and the means through which they are communicated in society (Moscovici, 1976b), and, within this general approach, further reflections on the representations themselves has led to studies of the social psychological structures of the groups producing the representations (e.g. Jodelet, 1991). Some of the same concerns can also be observed in studies which have followed Vygotsky’s way (though there is nothing here which compares to the richness of Jodelet’s analysis of the underlying social psychology of her French villagers struggling to come to terms with the presence of the mad in their everyday lives), though the major focus for these studies has been the emergence

92


of the child as a social actor. For the most part Vygotsky’s way has been seen as an approach to a particular problem – if the child is born into a world which is already structured in terms of social representations, how does the child become a participating member of these communities? The emphasis, then, in studying social representations through Vygotsky’s way has been rather different to that which has been evident from Bartlett’s way. We could say that Vygotsky’s way has been seen as an appropriate approach to questions of the ontogenesis of social representations, whereas Bartlett’s way has been more generally adopted for the study of their sociogenesis (cf. Duveen and Lloyd, 1990). Thus, a focus on ontogenetic issues has generated a concern with a different, if complementary, set of questions from those which are evident in sociogenetic studies. Primarily, of course, because ontogenetic studies have had to negotiate their way around the classical edifices of developmental psychology with their central assumption that the mental capabilities of children are themselves changing as the child develops. What contribution, if any, does the child’s developing mental organisation exert on their acquisition of social representations? It is tempting to see the various responses to this question as illustrating two fundamental approaches, which either follow Piaget in emphasising the constitutive role of the child’s own emerging mental structures, or share Vygotsky’s emphasis on the internalisation of collective sign systems. Before considering the contribution which the work of these authors can make to an understanding of the development of social representations, I want to outline the approach to this theme which has emerged from the work which Barbara Lloyd and I have undertaken over the past few years into the development of social representations of gender in young children (Duveen and Lloyd, 1986, 1990, 1993; Lloyd and Duveen, 1989, 1990, 1992). In this approach we have used Moscovici’s theory of social representations as a starting point for considering psychological development as a social process. This approach provides a social psychological perspective of children as developing social actors, with a complementary emphasis on the symbolic aspect of their developing knowledge. It is this perspective which I use in the later sections of this chapter as a point of reference for discussing the contributions of Piaget and Vygotsky.

The development of social representations Children are born into a world which is already structured by the social representations of their communities, which ensure that they take their place within systematic sets of social practices and social relations. Contemporary approaches in developmental psychology demonstrate an increasing awareness of this situation, and nowhere is such a perspective more important than in relation to gender development, precisely because gender is such a powerful and ubiquitous dimension of the social world into which the child is born. Indeed, it seems inappropriate here to speak of ‘the child’ as though some shared set of characteristics could serve to identify the object of study. While Piaget could speak with some authority of ‘the child’ as an epistemic subject, and use this conceptualisation as a

Psychological development as a social process 93 central strategic notion for pursuing an analysis of the development of intelligence, such a generalisation cannot be sustained in relation to gender, where it is difference which is more salient. The force of categorisation is so strong in the representations of gender which circulate around children, that they always appear as girls or boys developing particular social identities. From their earliest beginnings (which thanks to modern technology now frequently means while they are still in the womb) children are construed as gendered beings by those around them, who consequently act towards them in the light of this construction. If children are born into a world which is structured by social representations of gender, and through which they are construed, this does not mean that they are born with the competence to be independent social actors in this world. Initially children figure as the objects of other people’s representations of gender, and only gradually do they come to internalise these representations, and, as they do so, they also come to identify a position for themselves within the world structured by these representations. Representations of gender provide an important framework through which children acquire an identity which enables them to situate themselves in the social world. This relationship between representations and identities is not specific to the field of gender. Wherever representations are internalised they are linked to a process of identity formation, although the consequences of the identity are not always the same. For example, we do not usually think of children’s internalisation of representations of mathematics as being linked to specific social identities, but this can indeed be the case. When the form of mathematics which children internalise is linked to their identity as the member of a marginalised social group this can lead to a disruptive relationship in their schooling, and it is only when we see the consequences of difficulties and failures in school that the sense in which representations of mathematics also express a social identity becomes apparent (de Abreu, 1993, 1995). If the relationship between representation and identity is usually opaque in the field of mathematics, it can nevertheless become clear in some contexts. The pervasiveness of variations and differences associated with gender ensures that the relationship between representations and identities is clear across a very wide range of contexts. That this should be so is due to the significance of gender as a dimension of power in the social world. Representations are always constructive, they constitute the world as it is known, and the identities which they sustain secure a place for the individual within this world. As representations are internalised they come to express an individual’s relationship to the world which they know and to situate them within this world. It is this dual operation of defining the world and locating a place within it which gives representations their symbolic value. Moscovici (1973, p. xiii) alludes to this when he speaks of social representations establishing ‘an order which will enable individuals to orient themselves in their material and social world and to master it’. Once established in this way the order which is provided by representations takes on a fixed and objective character; it assumes a stability which can guarantee security for the individuals who find their place within this world. This aspect of cognitive activity has rarely been the explicit focus of discussion, although Shweder (1990) makes a similar point in outlining

94


his idea of a cultural psychology when he speaks of cognition as functioning to reduce human existential uncertainty. Another example can be drawn from Mugny and Carugati’s (1989) study of social representations of intelligence. They found that those groups of respondents such as students who were removed from the world of children had less clearly structured and organised representations of intelligence than groups such as parents and teachers who have to face an everyday reality in which differences between children require some explanatory framework. These authors talk about such social groups having a different ‘social stake’ in these representations, an idea which is close to what I have in mind when speaking of the symbolic aspects of representations. It is perhaps no accident that it is through reflecting on social representations of gender that I have come to emphasise their symbolic functions, for representations of gender, precisely because they refer to such a central dimension of social power and organisation, also carry central consequences for our definitions of self. We cannot think of ourselves as neutral in the field of gender, in some way or another we always think of ourselves as male or female, and these social identities arise in the course of internalising social representations of gender. Other representational fields may carry fewer existential consequences, in which case their symbolic value would also be reduced, as with Mugny and Carugati’s students. As a social psychological process, the construction of an identity is a way of organising meanings which enable a person to position him or herself as a social actor. An identity provides a means of organising experience which contributes towards the definition of self, but it does so by locating the self within the collective world. For the newborn child social gender identities are at first external; they are extended to the child through the practice of others. What we see in the development of social representations of gender is a grasp of consciousness, as children develop a reflexive awareness of the meaning of the social act of assignment to a gender group.

Piaget’s way There is not the space in this paper to offer a systematic appraisal of the relevance of Piaget’s work for a developmental psychology of social representations. I have already alluded to the significance of his constructivist epistemology, and there are numerous other annotations which one could offer in this respect. Jovchelovitch (1995), for example, has emphasised the importance of Piaget’s analysis of the development of representation and of decentering as contributions to a theory of social representations. One could also add the evident importance which Piaget’s analysis of children’s thinking held for the development of Moscovici’s (1976b) analysis of the characteristics of social representations (and this reference to Piaget’s work brings with it the influence of Lévy-Bruhl, who was also an important source for both Bartlett and Luria; indeed it may be in Lévy-Bruhl that one can see a clear common origin for the two ways which Moscovici describes). No doubt a systematic analysis would bring forward many other points of confluence between Piaget and Moscovici. And yet for some authors Piaget has

Psychological development as a social process 95 become a highly controversial figure from the perspective of social representations. Emler (1986; Emler, Ohana and Dickinson, 1990) in particular has attacked Piagetian work on two central grounds. First that it articulates the child as an autonomous individual engaged in the construction of knowledge through processes which are only marginally affected by social influences. Secondly that Piagetian theory presents a restricted image of the child as a solver of primarily logical problems. Emler can of course find evidence in Piaget’s writings with which to condemn him for committing these sins, but this is based on too simplistic a reading of Piaget. Or perhaps we should say that in spite of the extraordinary theoretical integrity which sustained Piaget’s work over the course of a very long productive life, there are nevertheless lacunae in this work which open up precisely around his articulation of the social character of knowledge. More than anywhere else it is in this arena that one finds not a single Piagetian theory but a multiplicity of Piagetian texts. There is on the one hand the Piaget who wrote The Moral Judgement of the Child (1932) and the Sociological Studies (1995) (and perhaps one should also add Play, Dreams and Imitation, 1951, to this list), where the social character of knowledge and the social processes of its constitution are of central concern. On the other hand there are numerous texts (particularly those of his later years devoted to an explication of developmental theory based on his conception of equilibration) where Piaget appears to elide any distinction between biological and social forces.2 In short, the social is an unstable element in Piaget’s analysis of the development of knowledge. Indeed one could go further and suggest that it is precisely this instability which has contributed to the construction of an image of Piaget’s work as an ‘asocial’ theory, an image which is peculiarly deeply entrenched in the Anglo-Saxon world (one could conceive of a study of the social representations of Piaget’s work, which would illustrate the way in which his theory is transformed as it becomes anchored in other psychological approaches). Emler’s reading of Piaget is, I think, too closely linked to this image of Piaget, and fails to appreciate the more radical elements in his work which resist the interpretation which Emler seeks to sustain. In the first place, Piaget’s focus is not on the individual but on what he terms the knowing or epistemic subject, which he distinguishes from the psychological subject. ‘There is’, he writes (Piaget, 1966, p. 308), the ‘psychological subject’ centred in the conscious ego whose functional role is incontestable, but which is not the origin of any structure of general knowledge; but there is also the ‘epistemic subject’ or that which is common to all subjects at the same level of development, whose cognitive structures derive from the most general mechanisms of the co-ordination of actions. One might legitimately accuse Piaget of a transcendentalism which obscures a view of cognitive structures as social and cultural formations (e.g. Buck-Morss, 1975), but there is no trace of a theoretical individualism in his work. The question of focus is paramount in appreciating Piaget’s work. His central question was always ‘How is knowledge possible?’, and from this point of view sensorimotor

96


co-ordinations and mental operations are not in themselves knowledge but the grounds for knowledge, the processes through which knowledge of the world is constructed. Thus from Piaget’s point of view variations in the content of knowledge are less important than the uniformity of form which holds across social and cultural contexts. One can then see the point which a critical reading of Piaget needs to recognise, namely that what Piaget construes as universal forms of cognitive organisation are themselves particular social representations. In this sense to Piaget’s question of ‘How is knowledge possible?’ we need to add the further question ‘For whom is knowledge possible?’, so that we can introduce a third term into Piaget’s binary distinction. As well as the psychological and epistemic subjects, we need to consider the social psychological subject (cf. Duveen and Lloyd, 1986) for whom knowledge is not the product of an abstract universal but the expression of a social identity. To say this is not to deny the logical character of knowledge, but, rather, to assert that the use of a logic in constructing knowledge locates the subject in a social world where that logic is held to be legitimate. Emler’s second objection is that Piaget characterises the child primarily as a problem-solver. He cites his own studies on income inequality as evidence that such a view is in error. Far from solving problems, the children he interviewed responded as though they had access to ready made solutions which required little or no cognitive elaboration but which reflected primarily the child’s social position. Thus he found little variation with age in children’s judgements of relative income, but significant variations with social class (Emler and Dickinson, 1985).3 Emler construes such results as providing evidence against a Piagetian view of the development of social knowledge, since there is little indication of any reconstructive activity on the part of children who seem merely to be repeating the common knowledge of their social class. But again this argument reflects only a very partial reading of Piaget, and a reading which ignores what is still one of the most important contributions which Piaget has to offer a developmental psychology of social representations. In The Moral Judgement of the Child (1932) Piaget makes a fundamental distinction between two forms of acquiring social knowledge. On the one hand there is knowledge which he describes as the product of social transmission, where it is the authority of a dominant or privileged figure which is the source of knowledge. As against this Piaget also argues that there is knowledge which is acquired through cognitive elaboration in a process of reconstruction. The former, he argues, takes place in heteronomous relations where the asymmetry of power exercises a constraint on the less powerful. The latter, by contrast, can only occur in autonomous relations between equal partners, where each has the freedom to engage in argument and debate. Emler, it seems to me, has failed to take account of this distinction. The results he presents could be considered as a fine example of acquisition through social transmission. But there is also evidence from other sources which shows children engaged in a more productive cognitive elaboration of social knowledge. In this respect one could cite the Genevan research on social interaction and cognitive development (Doise and Mugny, 1984) as an exemplary

Psychological development as a social process 97 instance, where collective engagement in a problem was found to lead to the construction of more complex solutions than either partner was able to find when working independently. Perhaps, though, such an example would be discounted by Emler as being too explicitly oriented towards logic and problem-solving. A different source of evidence for the reconstruction of social knowledge in symmetrical relations can be found in studies of pretend play among young children, in which the social world is literally reconstructed to the extent that children have understood it (cf. Duveen and Lloyd, 1988; Furth, 1992). Further evidence is offered by Corsaro’s (1990) study of the constructive elaboration of social rules amongst peers in the world of the nursery. To illustrate this point, consider these two excerpts from our video recordings of children at play during their first year at school (Lloyd and Duveen, 1992). The first extract contains a long domestic pretend episode which features Mummy, Daddy, their bed, their babies and marriage. Oscar asserts his role and responsibilities as Daddy in turns 1, 4 and 5. Sally initiates a scene shift from domesticity to courtship with princesses in turn 17. Adults are portrayed as actively searching for partners who presumably, upon marriage, become mummies and daddies. Physical contact between a boy and girl inevitably results in marriage, as Sally proclaims in turn 27. Betty’s somewhat ambiguous comments in turn 28 leave little doubt that the Princess’s marriage involves procreation. These episodes provide us with a glimpse of children’s understanding of family life. It is based upon a view of adult sex group membership which offers little role choice or within-group variability. Sexuality is heterosexual and procreative. Extract 1 Oscar is Daddy, Rachel the Mummy and Betty the Baby. There have been prior turns about eating and mending things. 1. Oscar: 2. Sally: 3. Betty: 4. Oscar: 5. Oscar:

(On mattress.) I’m the Daddy. (To Rachel.) Dinner’s not ready, so just wait. (To Oscar.) And you hear my crying a-ha-aah-aah. (Goes to Betty who is still crying.) Be quiet, baby, be quiet. I put that by your bed in case you wanted some dinner. (He goes back to bed.) 6. Rachel: (Joins Oscar.) No – aah! 7. Oscar: That’s my bed. It’s my bed. (They both lie on it.) 8. (Break, followed by) 9. Sally: This is Mummy’s and Daddy’s bed. (Lying down on bed. Oscar goes over to her.) 10. (Character shift and Sally becomes Mummy.) 11. Rachel: (Rolling on bed.) Hello Mum. 12. Sally: No! Get off Mummy’s and Daddy’s bed. You’re being a very naughty girl today. 13. Betty: (Crawling over to them.) Googa googa.

98


14. Sally: Will you get that cover, Baby? 15. Betty: (Hands cover to Sally.) 16. (Going to bed, Baby crying, tap dripping – relevant?) . . . 17. Sally: I’m the Princess. (Sits down on her bed.) 18. Betty: Pretend, pretend, I want, pretend you was a bit beautiful and I was . . . 19. Oscar: I’m the Prince. 20. Betty: . . . beautiful. We were both beautiful. Decide who you want to marry. 21. Oscar: I’m the Prince. 22. Betty: You can searching for a beautiful woman. 23. (Oscar pretends to dip between Sally and Betty. Meanwhile Rachel is pulling at his trouser leg.) 24. Rachel: I caught him! I caught him! I caught him! I caught him! I caught him! I caught him! 25. (Sally wins the dip and Oscar goes and puts his arms around her, sits beside her on the bed, puts his head on her shoulder.) 26. Betty: [. . .] 27. Sally: He marries me. 28. Betty: [. . .?] Yes, Yes, and I have to be the . . ? . . Princess, but with you little girl. Pretend you got married and you, you had a . . ? . . grown up. And I had to go to school. The second extract illustrates once again children’s belief that physical/sexual contact between sex group members needs to be validated through marriage. Oscar is chased for some time by the girls but once he is kissed by Christine, perhaps somewhat to his surprise, he proclaims in turn 16 that he is going to marry her. Children create a simple world in which physical contact between sex group members is construed as sexual and involves marriage. In this world actions have direct and predictable consequences. Extract 2 1. Edith: 2. Christine: 3. Edith: 4. Lulu: 5. Edith: 6. Lulu: 7. Edith: 8. Edith: 9. Joan: 10. Edith: 11. Lulu:

. . . and Lulu kiss, uhm, Oscar. Go on. I’m not playing now. Go away, then. No, you kiss Oscar and I kiss Darren. I know. Look. You (Joan) kiss him Darren. And I’ll kiss Oscar. Joan kiss Oscar. Joan kiss Darren, and Oscar kiss . . . you! (Starts for Darren, who runs.) Hey! Come here. (Grabs Lulu and moves her towards Oscar, not unwillingly.) No, kiss! Kiss her on the lips. Kiss her on the lips. Come on! No way!

Psychological development as a social process 99 12. Edith: Go on. Kiss her. Kiss her. 13. Christine: (Makes a dash for Oscar.) I kissed him. 14. Oscar: I kissed HER! 15. Edith: Oooh! 16. Oscar: (Points at Christine.) I’m going to marry her. 17. Edith: (With Lulu, no longer struggling, very close.) Kiss her. 18. Oscar: I’m going to marry her. 19. Sally: (Also closing in on Oscar.) 20. Oscar: All right. (But which one should he kiss?) 21. Sally: Kiss me. (They kiss.) (All laugh. Oscar throws himself back on sofa.) These examples illustrate the reconstructive processes at work in children’s acquisition of social knowledge, which can be contrasted with aspects of social knowledge which appear to result from social transmission. In arguing that from the perspective of social representations knowledge is only acquired through social transmission, Emler not only gives a partial view of Piaget’s argument, but he also restricts the scope of social representations. If we consider for a moment other studies of economic knowledge, there is evidence of a constructive aspect in children’s elaboration of such concepts as wages, prices and profit (e.g. Berti and Bombi, 1988). These notions are clearly not acquired simply by a process of social transmission, but does that mean that the idea of profit, for instance, is not a social representation? Clearly not. From Piaget, then, we can draw an account of different processes at work in the acquisition of social knowledge. He himself associates social transmission with children’s acquisition of the collective representations of their society (and in using that term he makes explicit reference to Durkheim), while reconstruction is for him a function of engagement in cognitive activity. But if we read Piaget’s argument in the light of the critical perspective outlined above, what he construes as cognitive activity also needs to be understood as social representation. And from this point of view it is interesting to return to Moscovici’s contrast of social representations with Durkheim’s analysis of collective representations. For Moscovici Durkheim’s conception is too static, and his major reason for preferring the adjective ‘social’ is to emphasise the dynamic aspects of representations. He wishes to focus on social knowledge in the process of formation and transformation, rather than social knowledge as received wisdom (or what Sartre referred to as the practico-inert). There must, of course, be a relationship between social representations and collective representations, for even collective representations must at one time have emerged from a more dynamic elaboration (just as the practico-inert is no more than the accumulated products of past human praxis). In childhood much of what appears to be the static common sense of the adult world is subject to a more dynamic form of elaboration. And where children are engaged in such elaboration we can learn from Piaget to expect that they will articulate representations which reflect their cognitive development rather than being the immediate reflection of adult thought. Piaget also offers us a further lesson, that

100


cognitive elaborations occur in the context of symmetrical relations where thought is not constrained by hegemonic power. In the life of children it is in their relations with other children that this condition is most frequently found, so that the constructive cognitive elaborations in the development of social representations will be most clearly seen in analyses of children’s interactions with their peers in situations where they have the freedom to invent. If we take Emler’s arguments to their most extreme conclusion they imply not only that we should abandon Piaget, but that we should also abandon a commitment to any idea of cognitive development as an autonomous process. I have tried to argue that these positions do not hold, and that there is much to be gained from continuing a constructive dialogue with Piagetian theory. The constructive moment of cognitive elaboration is one of the most important elements in Piaget’s genetic psychology, and adopting the perspective of social representations does not mean abandoning a notion of cognitive development, but rather seeing that the structures which emerge are themselves social and cultural products. This may make us see cognitive development as a moment of relative autonomy (to adopt a phrase of Althusser’s), but it also serves to remind us that between the ‘thinking society’ of adults and the emergence of the child as a social actor there is a process of construction which needs to be addressed.

Vygotsky’s way Within the corpus of Vygotsky’s work the concept of internalisation appears to be at once both the most enigmatic and provocative of his contributions. Enigmatic because although internalisation is identified as a law of psychology, the ‘general genetic law of cultural development’, the process is described only through the briefest sketch of its contours which rarely provide a sense of detail in the way in which the process operates. What is provocative in Vygotsky’s formulation is that it suggests a simple solution to what is a complex problem, the relations between social and cognitive processes, so that it seems to dissolve the mysteries of the socialisation of cognitive functions and thereby to offer a resolution of some of the deepest problems of developmental psychology. Since the emergence of Vygotsky’s work in the Western world, his account of the process of internalisation has exerted an ever-growing influence in developmental psychology. It is not, perhaps, too difficult to identify at least some of the reasons why this should be so. Certainly, a primary reason lies in Vygotsky’s insistence that internalisation is a social process, both in the sense that it takes place in the interactions between people and in the sense that what is internalised is a semiotic system which is itself a collective or social product. This emphasis on social processes was seen as an alternative to what was considered to be the irremediably individualist perspective offered by Piagetian theory (although, as I have argued above, this is a particularly myopic view of Piaget’s contribution). In particular, cross-cultural evidence appeared to expose serious difficulties in the Piagetian assumption of universal forms in development, while a Vygotskian perspective seemed to offer a means for understanding cultures within the terms

Psychological development as a social process 101 of their own semiotic productions (although again, this view of Vygotsky fails to grasp his own evolutionary perspective on culture and society). In short, Vygotsky’s work seemed to offer a perspective from which social relations and cognitive processes could be brought into a single productive framework. And yet doubts remain about the extent to which the Vygotskian framework is able to achieve this resolution. Serge Moscovici, for one, has commented (1990, p. 179) that the Vygotskian formula is ‘too good to be true’, since it assumes a direct relationship between social practices and individual functioning. The suspicion here is that the lack of attention to any mediating structures between these two levels limits the usefulness of Vygotsky’s work for social psychology. It is this suspicion which I want to explore in this paper. Vygotsky’s idea of internalisation finds its most succinct expression in what he termed the ‘general genetic law of cultural development’ (1981, p. 163): Any function in the child’s cultural development appears twice, or on two planes. First it appears on the social plane, and then on the psychological plane. First it appears between people as an interpsychological category, and then within the child as an intrapsychological category. . . . Social relations or relations among people genetically underlie all higher functions and their relationships. Internalisation is the process through which functions which are first established in the child’s external relations with others are re-constructed internally. The connecting thread in this process is given by Vygotsky’s analysis of sign functions. In his characteristic metaphor he speaks of these functions ‘turning inward’, so that what once could be sustained only in interaction with others takes on the character of an internal function which the child can achieve independently. This transformation may be the ‘result of a long series of developmental events’ (1978, p. 57), but it creates a profound and radical restructuring of the child’s activity. A prototypical example which Vygotsky (1978) gives of internalisation is the development of pointing in the young child. When the young child tries to grasp an object which is out of reach, this gesture is initially directed at the object itself. It is through the actions of others that the child’s gesture is made meaningful. It is the intervention of another (an adult or older sibling) who interprets the gesture and gives the child the object which changes the situation. Gradually, the action of grasping is replaced by the action of pointing, which, as Vygotsky notes, is directed not at the object but at a person. The form of the action itself changes; from the outstretched fingers of the whole hand to the use of a single finger to indicate the desired object. The emergence of pointing not only serves to establish relations with others, but also transforms the child’s own sphere of activity by bringing within its horizons objects which cannot be directly grasped. The constitutive role of others in internalisation led Vygotsky to formulate another key notion, the idea of a zone of proximal development, which he defined (1978, p. 86) as: ‘the distance between the child’s actual developmental level as determined by independent problem solving and the level of potential development

102


as determined through problem solving under adult guidance or in collaboration with more capable peers’. This image of the child developing through their interaction with adults or peers with greater expertise has had a powerful impact in psychological and educational thinking, being elaborated in terms of ‘scaffolding’ (Wood, Bruner and Ross, 1976), ‘guided participation’ (e.g. Rogoff, 1990) or as the ‘construction zone’(Newman, Griffin and Cole, 1989). In all of these guises Vygotsky’s notion is the principal idea. It is through their participation in culturally patterned activities that children internalise the semiotic systems of their culture, and thus emerge as competent cultural subjects. The power of Vygotsky’s idea, as Moscovici recognised, is its simplicity. The child is born into a world which is already culturally structured, and this culture is mediated for the child through the activities of those around them. In this way the child comes to take their place within this culture, and eventually, through their own activities to become a mediator of the culture to successive generations. The process through which the child develops is social, and the products of this development are internalised sign systems, which, like all sign systems, are collective in character. Internalisation, then, stands at the centre of Vygotsky’s argument that psychological activities are social processes. However, as powerful as Vygotsky’s idea may be, there are nevertheless limits to its usefulness. Wertsch (1991, p. 46), for instance, points to weaknesses in Vygotsky’s approach to the way ‘historical, cultural and institutional settings are tied to various forms of mediated action’. In responding to these issues, Wertsch has sought to extend Vygotsky’s semiotics by including Bakhtin’s notion of ‘voice’. Other writers have emphasised the significance of social identities as mediators between the interpersonal and the intrapersonal (e.g. Goodnow, 1990; Litowitz, 1993; Duveen, 1994). Although both of these responses to Vygotsky are derived from a consideration of limitations in his grasp of the structural influences at work in the process of internalisation, they draw on different traditions to articulate this concern. One way in which to examine the limits of Vygotsky’s theory is to consider the body of work constructed around the notion of an apprenticeship model of cultural learning. This has become a major form in which Vygotskian ideas have been elaborated in contemporary developmental psychology (e.g. Rogoff and Lave, 1984; Lave, 1990; Rogoff, 1990). The metaphor of apprenticeship is employed to give concrete expression to Vygotsky’s idea of a zone of proximal development. A novice acquires cultural knowledge through participation in social practices under the guidance of an expert. Guided participation serves to structure and organise the task for the novice, so that eventually they are able to take independent responsibility for the task. Litowitz (1993, p. 185) identifies the principal characteristics of this model as: (1) cultural knowledge is transferred not from one person (adult) to another (child) but from two persons (the dyad) to one (the child); (2) the transmission is accomplished through semiotic means; and (3) the nonknower demonstrates equality in the dyad by becoming equally responsible for solving problems and accomplishing tasks.

Psychological development as a social process 103 As Litowitz makes clear, an important element in this model is a notion of equality between the partners based on mutual respect for the initial differences in expertise. The model emphasises a sense of intersubjectivity based on mutual engagement in a joint activity which subordinates differences between the partners. This model has been employed with considerable success in the analysis of some specific processes of cultural learning, for example in Greenfield’s (1984) study of young women learning to weave in Zinacantán, Mexico, and Lave’s (1990) study of apprentice tailors among the Vai and Gola in Liberia. In these analyses one can trace the relations between the cultural patterning of a particular practice, the mediating activities of the expert and the gradual acquisition of independent skill by the novice. It is the engagement of expert and novice in a joint activity which provides the arena for transmission of a particular skill which is embedded in a cultural tradition of practice. And yet the very power of the apprenticeship model to render this situation theoretically accessible points to the limits of the applicability of the model. Reflecting on these analyses one can in fact identify a series of conditions which limit the model, or set boundaries to the range of its applicability. 1

2

3

4

The apprenticeship model applies only to well-structured social practices. The model has been most successfully employed in situations where what is being acquired is a practical skill which depends on mastery of particular coordinations of actions, and it is the regulation of this coordination which is the focus of the expert’s guidance of the novice’s participation. Yet not all cultural knowledge has this concrete quality. The social practices which are being transmitted need to be highly valued by the community so that they are seen as a legitimate expression of the culture. In the examples of the Mexican weavers or the Liberian tailors, it is clear that these practices have an economic value for the community as well as practical and symbolic value. The model assumes a community of interest between novice and expert, so that no conflict arises between them as to the value of what is being learned. For an apprenticeship to be established both expert and novice need to see themselves as being engaged in a joint activity which, whatever their different roles may be, is meaningful to both parties. The novice must be willing to learn and the expert ready to guide their participation. In circumstances where such a community of interest does not exist, where there is instead dissent or resistance, it is difficult to see how an effective apprenticeship could be established. Some of the most persistent problems in secondary education, for example, seem to arise precisely because the students do not recognise any community of interest with their teachers. In championing the ‘practice of understanding’ as against the ‘culture of acquisition’ Lave (1990) herself discusses this problem. The model does not account for individual variability in the acquisition of cultural knowledge. Some young Zinacantan women will become better weavers than others, and some young Liberian men better tailors than others.

104

5

Development as decentration But all will have passed through the same process of apprenticeship. The model itself offers no means for discussing such diversity, which, at the extreme, means that it provides no explanation for failure in the transmission of cultural knowledge. The model offers an account of how specific social practices are transmitted from experts to novices, and in this sense is focussed exclusively on the reproduction of existing social relations. One can legitimately ask how social change can be brought about through a process of apprenticeship.

If we consider these five conditions as a whole, it is possible to recognise a common theme running through them. This is that the model treats social life as though it were in some important way undifferentiated. It is as though children progressively internalise the collective practices of their community through interaction with competent adults or more knowledgeable peers in the absence of any complications arising from the effects of any differential valorisation. Indeed, a certain homogeneity is the most striking characteristic of culture and of the child as they figure within the Vygotskian perspective. Culture as it appears in the Vygotskian formula is an ensemble of signs embedded in social practices, but, importantly, there is no significant diversity of values within this image of culture. It is as though social life were constructed without the differences of value and perspective, of power and conflict which structure social phenomena and which have been at the centre of social psychological concerns. And as a corollary to this one-dimensional culture, the child who develops within it is a subject unmarked by the vicissitudes of social life, a subject who is constituted only as a reproduction of the culture within which they live. The Vygotskian model establishes a dynamic which moves from the practices established as interpersonal relations through internalisation to the intrapersonal achievement of this practice. But what exactly should be understood by the term ‘practice’ here? In the examples I quoted earlier, as in most research bearing on the theme of apprenticeship, the practices which are being internalised are established first as dyadic structures. While this may be appropriate for the learning of weaving or tailoring, it is not the case that social practices are limited to dyadic situations. There is, of course, no clear reason why Vyotsky’s model should be limited to dyadic interactions. Both the examples he gives and his formulation of the zone of proximal development focus on dyads, but these instances illustrate his thinking rather than define his terms. In principle, internalisation should apply to the acquisition of any social representations where these are articulated through interpersonal relations. If, for example, we consider representations of gender as a set of social practices, it is of course the case that at times the child is a participant in dyadic interactions with adults who structure the child’s activity in terms of these representations. Yet gender is also a more diffuse phenomenon which surrounds the child through a variety of semiotic media: the toys which children play with carry gender markings; the social roles articulated in comics, picture books and in television programmes are also marked for gender; and in the collective institutions, such as the nurseries and schools which children attend,

Psychological development as a social process 105 representations of gender structure complex patterns of interaction. In all of these media the categories of gender are articulated as a collective semiotic system which effect a fundamental division in the social world. Becoming a part of this world means both internalising the social representation of gender and establishing an identity with one category or the other. Further evidence for the significance of social identities as mediating structures comes from our recent study of the development of gender identities through the first year of schooling (Lloyd and Duveen, 1992). An initial ethnography undertaken in the reception classes of two infant schools4 enabled us to investigate the way in which gender figures as a structural element in the organisation of the classroom (cf. Duveen and Lloyd, 1993). This structure is not limited to dyadic interactions, but is rather characteristic of the pattern of interactions throughout the classroom, which will include dyads, but also extend to larger scale interactions. Children bring with them into school the gender identities which they have acquired in their pre-school years. As they enter the reception class, however, they encounter representations of gender which are embedded in a set of social practices which are at first unfamiliar, but which are linked to the process of education itself. In short, as they enter school children encounter a set of representations of gender embedded in a novel institutional context, representations which exert a powerful influence on their subsequent educational careers. Sometimes these representations are indeed presented to children through dyadic interactions with their teacher or with a peer, but more frequently they are presented to them through larger scale social interactions. In our study we used our ethnographic account of gender as a dimension of classroom organisation to construct systematic observation schedules and a series of interview based tasks which explored children’s knowledge of gender marking in the classroom. The extent to which their activity was organised around a dimension of gender was clear from a number of indices related to patterns of peer associations. At the beginning of the school year boys were observed more frequently in single sex groups and in groups with more boys than girls, while girls were observed more frequently in groups with more girls than boys and in groups with even numbers of girls and boys. Further, single sex groups of boys were larger than single sex groups of girls. These patterns were observed in settings where children’s activity was not immediately directed by the teacher. When the teacher did organise children’s activity the differences between girls and boys in the gender composition of groups was less apparent, particularly for single sex groups. By the end of the first year these patterns had changed somewhat. Girls’ participation in single sex groups in peer organised contexts had increased (though still not to a level which matched that of boys), while in teacher organised contexts there were fewer observations of girls in single sex groups. As well as influencing patterns of peer association, children’s sex group membership also influenced their choice of materials for play, with boys being observed in active and constructive play more frequently, and girls more frequently in creative and role play. So far I have described differences between girls and boys in the gender composition of groups which form in the classroom in terms only of a contrast between

106


sex groups. In these contrasts the data have been aggregated for boys and for girls. Yet anyone who spends time in a reception class will also notice that there are variations within sex groups: not all boys are alike and neither are all girls. Different types of masculine and feminine identities can be observed developing within sex groups. For example, when we looked at individual patterns of association in single sex groups it was apparent that some children tended to associate particularly with some specific friends, while other children interacted with a much wider circle of peers. This variation was more noticeable among girls than among boys, to the extent that we observed some small groups of girls, who practically interacted only with each other, forming small isolated groupings in which they had little contact with other children in the classroom, girls or boys. These observations suggested to us that it was important to investigate varieties of gender identity which emerged within sex groups,5 and to do so we constructed an index of gender identity based on the proportion of time individual children were observed in single sex groups.6 Using this measure of gender identity we found a series of differences within sex groups in the ways that children made use of elements of the material culture of the classroom. For example, girls with low gender identity (that is girls who were observed less frequently in single sex groups) made as much use of activity play materials as did the boys, while girls with high gender identity (who were observed more frequently in single sex groups) made much less use of these materials. Some differences emerged when we compared children’s activity at the beginning and end of their first year at school, as was the case for the use of directed play materials in peer organised contexts (these are materials usually associated with teacher organised activities such as books, reading schemes, writing materials, etc.). In the autumn term all children made much the same use of these materials, but in the summer term high gender identity girls made twice as much use of these materials, while low gender identity girls and all the boys continued to use these materials at the same rate as they had in the autumn term. Finally, some differences were also specific to particular schools. In the autumn term all girls made much the same use of constructive play materials, although not as much as did the boys who used them between two and three times as much. By the summer term, however, not only had differences emerged among girls, but these differences varied also between schools. In both of the classrooms we observed, low gender identity girls continued to use these materials at the same rate. It was among the high gender identity girls that differences emerged. In one classroom their use of these materials declined dramatically in the summer term, while in the other classroom it increased to the level at which boys were using these materials. These data are complicated because in the latter school the teacher set aside time when these materials were only available to girls, so that the greater use of them by girls in single sex groups can be seen as the consequence of the teacher’s intervention. But the decline in the use of activity play materials by high gender identity girls observed in the other classroom was not related to any action by the teacher. Overall these results indicate that variations of gender identity emerge within sex groups through the course of the first year of schooling, particularly among

Psychological development as a social process 107 girls, and in relation to the local culture of gender within a school. Further indications of the significance of these emerging gender identities came from the results for some of our interview based measures. For example, in one task children were asked to identify the odd-one-out from a set of three figure drawings. Each figure showed a child playing with a toy, and each triad was composed of one figure of a boy playing with a masculine marked toy, one figure of a girl playing with a feminine marked toy and one figure of a child playing with a toy marked for the opposite gender. Children were asked to pick the odd-one-out for six triads, in three of which the third figure showed a girl playing with a masculine marked toy, and in three it showed a boy playing with a feminine marked toy. Children’s judgements for each triad could therefore focus on either the actor in the figure (e.g. picking out a boy in a triad with two girls), or the gender marking of toys (a feminine marked toy in a triad where the other two figures showed children playing with a masculine marked toy), or the figure in which there was a mismatch between the gender of the actor and the gender marking of the toy they were shown as playing with. The mismatch choices are more cognitively complex because they demand a coordination between the two dimensions of actor and toy, and overall it is perhaps not surprising that children did not make a great many such choices. In general, too, there was little change in children’s performance from the autumn term to the summer term. However, in one class the high gender identity girls showed a notable increase in the number of mismatch choices from the autumn term to the summer term. These were the same girls whose use of activity play materials (generally a masculine marked activity) declined quite sharply across the school year, suggesting that something in the local gender culture of this classroom is making the gender marking of persons and objects particularly salient. One might look at the results from these studies in reception classes and suggest that in general they follow a Vygotskian line from social practice to intraindividual accomplishment, even if it means displacing the Vygotskian formula from dyadic interaction to social practice on a broader scale. However, this generality obscures an important point, namely that variations of gender identity within sex groups emerges through the course of the year. If we accept that there is a representation of gender structuring the social practices in a classroom, why should these practices lead to differentiation within sex groups? Why should one group of girls acquire a different set of practices from another? The Vygotskian formula does not explain how interindividual practices can lead to different intraindividual formations. To do so we need to consider the role of developing social (in this case gender) identities in mediating the transition from interindividual to intraindividual practices.

Conclusions These reflections on the contributions of Piaget and Vygotsky to a developmental psychology of social representations lead to an interesting comparison between them. From Piaget there is an important lesson to be drawn from his distinction between two modes in the acquisition of social knowledge. There is acquisition through social transmission, characterised by an asymmetry of power so that

108


knowledge is reproduced because of the influence and prestige of its source. But there is also acquisition through reconstruction in symmetrical social relations between peers. As always in Piaget it is conflict which is at the heart of this process, but for conflict to be productive it has to be situated in a context where thought is unconstrained by hegemonic influences, so that thinkers have the freedom to invent and construct. Vygotsky’s work really only addresses the first of these modes. It describes the acquisition of social knowledge through social transmission in asymmetrical social relations. Here knowledge truly is power, for it is the possession of expertise and the prestige that it brings which distinguishes the expert from the novice. This conclusion may seem somewhat paradoxical, since the images of Piaget and Vygotsky which circulate in the contemporary world of psychology have more generally associated Vygotsky as a theorist of cognition as a social process and Piaget as a theorist of individual cognition. But, as I have tried to demonstrate, this paradox is more apparent than real. If my comments on Vygotsky have been sharper than those on Piaget it is because I have sought to emphasise this contrast. And yet neither of these authors really provides a ready made model in which to pursue an analysis of the child as a developing social actor. Piaget’s analysis always returns to a focus on the ‘epistemic subject’, and gives only a limited acknowledgement of the significance of the social representations which structure the collective world within which children develop. If Vygotsky’s theory can be said to focus on the child as a ‘cultural subject’ it does so by reducing culture to a set of signs which function as cognitive tools and excluding the sense in which signs also express the values of particular social groups. In their different ways, both Piaget and Vygotsky present a view of the developing child centered on an idea of the child as a single, unified subject, ‘epistemic’ for Piaget, ‘cultural’ for Vygotsky. In each case the child appears to live in a world which is marked by a homogeneity of meanings, so that in neither case is there a clear recognition that the social world is a world of differences and contrasts. To consider the developing child as an emerging social actor means to construe the child as a social psychological subject. From this point of view the social is neither simply an influence which can accelerate or retard the child’s development, nor is it simply a body of knowledge about the social world (in terms of rules, norms, etc.). What children know and believe also serves a symbolic function, since it provides a primary means through which children are able to locate themselves in the social world. It is in this sense that we can consider children’s developing representations as expressions of their social identities. In many instances social identities act to promote children’s development, since they provide the source for the community of interests which facilitates learning in the zone of proximal development. But there are also instances where children’s social identities can be a source of conflict for them. This was the case for the Brazilian children studied by de Abreu, and also in our studies of the development of social representations of gender in primary classrooms. It is in these instances where defending a social identity leads children to resist the influences of others that a sense of psychological development as a social process emerges most clearly.

Psychological development as a social process 109

Notes 1 For Moscovici the process of social representation is always concerned with the way in which the unfamiliar is made familiar, and he distinguishes analytically between two complementary aspects of this process of construction – anchoring refers to the way in which the unfamiliar is located within the context of existing representations, while through objectification representations are projected into the world as concrete objects (cf. Moscovici, 1981; Duveen and Lloyd, 1990). 2 Equilibration is a complex concept in Piaget’s account. While he suggests that it is as much a social as a biological process (e.g. Piaget, 1970; and Furth, 1980, uses the example of developing social knowledge to explicate the idea of equilibration) this generates problems for a social psychological perspective. The equilibration of cognitive structures results in the construction of necessary knowledge, and it is by no means clear that we can characterise all social knowledge in this way. Divisions within society result in different representations, and it is these differences which are the focus of social psychological concerns. However, as I make clear in the following pages, Piaget himself makes an important distinction between two forms of social knowledge. 3 These results have been challenged by Burgard, Cheyne and Jahoda (1989) on the basis of their attempt to replicate this study. However, my purpose here is not to attempt to resolve this dispute, but rather to concentrate on the way in which Emler’s use of his evidence reflects his reading of Piaget. 4 In England, where the research I shall describe was undertaken, children enter school at the beginning of the year in which they will be five. Thus the reception class (which is the term used for the first year of school) contains children between four and five years of age. 5 This is a theme which has received very little attention in developmental research, to the extent that some authors have suggested the terms sex and gender should be used interchangeably. Maccoby and Jacklin (1987), for example, have argued that since research designs usually employ a contrast between girls and boys there are no grounds for distinguishing sex and gender. While they may be right to point to a weakness in the logic of such research designs, their argument loses its force when research designs encompass contrasts within sex groups as well as between sex groups. 6 This is far from ideal as a measure of gender identity, since it uses a single dimension to index a much more complex reality. However, even this rather rough measure served to demonstrate that it is possible to distinguish important variations within sex groups among young children.

References Berti, A. E. and Bombi, A. S. (1988). The Child’s Construction of Economics. Cambridge: Cambridge University Press. Buck-Morss, S. (1975). Socio-economic bias in Piaget’s theory and its implication for cross-culture studies. Human Development, 18, 35–49. Burgard, P., Cheyne, W. and Jahoda, G. (1989). Children’s representations of economic inequality: a replication. British Journal of Developmental Psychology, 7, 275–87. Corsaro, W. (1990) The underlife of the nursery school: young children’s social representations of adult rules. In G. Duveen and B. Lloyd (eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. de Abreu, G. (1993). The Relationship Between Home and School Mathematics in a Farming Community in Rural Brazil. PhD thesis, University of Cambridge. de Abreu, G. (1995). Understanding how children experience the relationship between home and school mathematics. Mind, Culture and Activity, 2, 119–42.

110


Doise, W. (1993). Debating social representations. In G. Breakwell and D. Canter (eds) Empirical Approaches to Social Representations. Oxford: Oxford University Press. Doise, W. and Mugny, G. (1984). The Social Development of the Intellect. Oxford: Pergamon. Duveen, G. (1994). Crianças enquanto atores sociais. In S. Jovchelovitch and P. Guareschi (eds) Textos em representaçōes sociais. Petropolis, Brazil: Vozes. Duveen, G. and Lloyd, B. (1986). The significance of social identities. British Journal of Social Psychology, 25, 219–30. Duveen, G. and Lloyd, B. (1988). Gender as an influence in the development of scripted pretend play. British Journal of Developmental Psychology, 6, 89–95. Duveen, G. and Lloyd, B. (1990). Introduction. In G. Duveen and B. Lloyd (eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Duveen, G. and Lloyd, B. (1993). An ethnographic approach to social representations. In G. Breakwell and D. Canter (eds) Empirical Approaches to Social Representations. Oxford: Oxford University Press. Emler, N. (1986). The relative significance of social identities: a comment on Duveen and Lloyd. British Journal of Social Psychology, 25, 231–2. Emler, N. and Dickinson, J. (1985). Children’s representations of economic inequalities: the effects of social class. British Journal of Developmental Psychology, 3, 191–8. Emler, N., Ohana, J. and Dickinson, J. (1990). Children’s representations of social relations. In G. Duveen and B. Lloyd (eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Furth, H. G. (1980). The World of Grown Ups. New York: Elsevier. Furth, H. G. (1992). The developmental origin of human societies. In H. Beilin and P. Pufall (eds) Piaget’s Theory: Prospects and Possibilities. Hillsdale, NJ: Lawrence Erlbaum Associates. Goodnow, J. (1990). The socialization of cognition: what’s involved? In J. Stigler, R. Shweder and G. Herdt (eds) Cultural Psychology. Cambridge: Cambridge University Press. Greenfield, P. (1984). A theory of the teacher in the learning activities of everyday life. In B. Rogoff and J. Lave (eds) Everyday Cognition. Cambridge, MA: Harvard University Press. Jodelet, D. (1991). Madness and Social Representations. London: Harvester Wheatsheaf. Jovchelovitch, S. (1995). Social representations in and of the public sphere: towards a theoretical articulation. Journal for the Theory of Social Behaviour, 25, 81–102. Lave, J. (1990). The culture of acquisition and the practice of understanding. In J. W. Stigler, R. A. Shweder and G. Herdt (eds) Cultural Psychology. Cambridge: Cambridge University Press. Litowitz, B. (1993). Deconstruction in the zone of proximal development. In E. Forman, N. Minnick and C. Stone (eds) Contexts for Learning. Oxford: Oxford University Press. Lloyd, B. and Duveen, G. (1989). The reconstruction of social knowledge in the transition from sensorimotor to conceptual activity: the gender system. In A. Gellatly, D. Rogers and J. Sloboda (eds) Cognition and Social Worlds. Oxford: Oxford University Press. Lloyd, B. and Duveen, G. (1990). A semiotic analysis of the development of social representations of gender. In G. Duveen and B. Lloyd (eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Lloyd, B. and Duveen, G. (1992). Gender Identities and Education: The Impact of Starting School. London: Harvester Wheatsheaf.

Psychological development as a social process 111 Maccoby, E. and Jacklin, C. N. (1987). Gender segregation in childhood. In H. Reese (ed.) Advances in Child Development and Behaviour, vol. 20, London: Academic Press, 239–87. Moscovici, S. (1972). Society and theory in social psychology. In J. Israel and H. Tajfel (eds) The Context of Social Psychology. London: Academic Press. Moscovici, S. (1973). Foreword to C. Herzlich, Health and Illness. London: Academic Press. Moscovici, S. (1976a). Social Influence and Social Control. London: Academic Press. Moscovici, S. (1976b). La psychanalyse, son image et son public. Paris: Presses Universitaires de France. Moscovici, S. (1981). On social representation. In J. Forgas (ed.) Social Cognition. London: Academic Press. Moscovici, S. (1990). Social psychology and developmental psychology: extending the conversation. In G. Duveen and B. Lloyd (eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Mugny, G. and Carugati, F. (1989). Social Representations of Intelligence. Cambridge: Cambridge University Press. Newman, D., Griffin, P. and Cole, M. (1989). The Construction Zone. Cambridge: Cambridge University Press. Piaget, J. (1932). The Moral Judgement of the Child. London: Routledge and Kegan Paul. Piaget, J. (1951). Play, Dreams and Imitation in Childhood. London: Routledge and Kegan Paul. Piaget, J. (1966). Part Two of E. W. Beth and J. Piaget, Mathematical Epistemology and Psychology. Dordrecht: D. Reidel. Piaget, J. (1970). Piaget’s theory. In P. Mussen (ed.) Manual of Child Psychology, vol. 1. New York: J. Wiley and Sons. Piaget, J. (1995). Sociological Studies. London: Routledge. Rogoff, B. (1990). Apprenticeship in Thinking. Oxford: Oxford University Press. Rogoff, B. and Lave, J. (eds) (1984). Everyday Cognition. Cambridge, MA: Harvard University Press. Shweder, R. (1990). Cultural psychology: what is it? In J. W. Stigler, R. A. Shweder and G. Herdt (eds) Cultural Psychology. Cambridge: Cambridge University Press Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard University Press. Vygotsky, L. S. (1981). The genesis of higher mental functions. In J. Wertsch (ed.) The Concept of Activity in Soviet Psychology. Armonk, N.Y.: M. E. Sharpe. Wertsch, J. (1991). Voices of the Mind. London: Harvester Wheatsheaf. Wood, D., Bruner, J. and Ross, G. (1976). The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17, 89–100.

7

Construction, belief, doubt Gerard Duveen

Introduction – Epistemology from the outside and from the inside In the context of a symposium on beliefs it seems important to introduce also a discussion of doubt. Questions of doubt reflect something crucial about the notion of belief itself, and in particular the character of certainty which attaches itself to beliefs. Further, a focus on doubt, or more precisely on the emergence of doubt within the framework of belief, can also illuminate some central aspects of belief. Firstly a consideration of doubt can throw some light on the relation between knowledge and belief, and, secondly, on the character of the social relations which sustain each of these forms of understanding. My point of departure for this discussion is to return to the epistemology of representations in order to consider how we can envisage the emergence of doubt within a constructivist perspective. I want to begin by emphasising a distinction between two ways of talking about epistemology in relation to social representations. On the one hand as social psychologists we work within a tradition of the social sciences in which we find a rich heritage of epistemological reflection which frames the way we ask questions, the way in which we investigate phenomena, the ways in which we structure our research, and also the forms of reasoning we employ, both in terms of our general theories and the forms of explanation we offer for particular phenomena. On the other hand, when (still as social psychologists) we turn our attention away from the basis of our own understanding towards the features of the objects we study, then we can see that we can also speak about how social representations themselves contain epistemological arguments. That is, just as much as representations can be said to comprise identity functions or ideological functions, so too they comprise epistemic functions. For the most part in this paper I shall be concerned with epistemology in the form of the epistemic functions of social representations. But it seemed important at the outset to make this preliminary distinction, in part because in some recent contributions (Gervais, 1997; Wagner, 1998) a lack of attention to this difference has led to some confusion between levels of analysis, so that in these discussions of the epistemology of social representations it is not always clear whether

Construction, belief, doubt 113 these authors are talking about the epistemology of the theory of social representations itself, or discussing the nature of the epistemic functions of representations. It seems to me that this vision of either strong or weak forms of constructionism is itself misguided and misplaced. I don’t see from the point of view of construction how one can separate a strong and a weak form. Things are either constructed or not constructed. But we can talk about the ways in which representations themselves construct realities in a stronger or a weaker form. Things can be brought more or less closer to us as an existential phenomena in certain ways. In this way, we can explore and expand the discussion of epistemic functions of representations. Before pursuing this discussion, I should like to say briefly what I take to be the epistemological basis of the theory itself. As I have argued before (Duveen & Lloyd, 1990), the theory is a constructivist theory which treats subject and object of knowledge as correlative and co-constitutive. From this perspective the world which is known is the product of the set of socio-psychological structures through which it has been constructed. The realities we apprehend, and hence also the fields of action which they open up for us, are constructed realities. Let me be as clear as possible here: I am not suggesting a resurrection of an old fashioned idealism in which the totality of the world is a product of consciousness. Rather, I am simply arguing that our knowledge of the world is limited by the structures we have available for apprehending it. Thus I am also arguing against a crude materialism which asserts that there is a reality over and above our knowledge of the world. Such a position seems to imply the contradictory proposition that we have access to a knowledge beyond human knowledge. In defending a radical constructivism, I am thus suggesting a dialectic of human knowledge. Wittgenstein was no doubt right to suggest that « the world is all that is the case », but what is the case at one place in time is not necessarily what is the case at another. Partly there are general questions of culture which structure what may be the case for a particular society, but, importantly, communication and action also lead to the development of new knowledge, new forms of understanding. In other words, while it may be a persistent feature of human knowledge to assume that it is complete – in the sense that our knowledge is identical with reality – this is in fact an epistemological fallacy which appears to be almost as universal as the human condition itself. In fact all that can be said is that human knowledge is identical with the world as it is known, which does not imply that our knowledge of the world is complete. Indeed, it seems improbable that human knowledge could ever be complete. Our attempts at understanding are always limited, and our knowledge always provisional. As we engage in different kinds of projects we encounter points of obscurity and resistance, around which new forms of communication and action may generate new forms of knowledge. And with new forms of knowledge, the realities which we know and in which we move are constructed in new ways. If social representations can be considered a constructivist social psychology, I should also add two further qualifications, both of which seem to me to be implicit in a constructivist position, but which remain points of controversy

114


within the wider field of social psychology. The first is that this constructivism is social – knowledge always emerges from the dynamics of social life, from communication and practice. And what distinguishes a constructivist position from the dominant paradigm of social cognition is precisely this, that the psychological structures through which we apprehend the world are themselves considered as social products, the consequence of inhabiting a particular time and place, of living within a particular set of social relations. The second is that this constructivism is genetic, that is the social-psychological structures through which the world is known are themselves the products of a construction which can be described at different levels, the sociogenetic, the ontogenetic and the microgenetic (Duveen & Lloyd, 1990).

The paradox of doubt From this constructivist position, doubt appears as a rather paradoxical feature of mental life. Axiomatic for a constructionist perspective in social psychology is the epistemological assumption that the known world is the product of the psychological operations through which it is constructed. This axiom is related to a second one – namely that such constructions are both the product of and stabilised by a set of social relations, that is, in some form or another, constructions are linked to specific forms of social life. How, then, might we come to doubt the very knowledge which we have constructed? We can see the nature of this paradox most clearly if we consider ontogenetic situations, and especially what has been described as the problem of transitions between stages in children’s development. Take, for instance, the pre-operational child, that is, the child who believes that if water is poured from one shaped glass to another then the amount of water is no longer the same. Why should they ever change their position? Why should they ever come to believe that such a way of understanding is wrong? If we hold to our constructivist view, we have to say that the pre-operational child constructs a knowledge in which there is more water in one glass than in the other. So what is it that enters into that child’s world that shifts and changes their construction in some way? What is it that opens a space, which creates, as it were, a doubt about their current understanding and which can lead to some new form of knowledge emerging? Well, the pre-operational child of course does not encounter the world simply as an isolated individual. The pre-operational child is forever engaged in interactions and exchanges with others. And in fact it’s through such interactions and exchanges with others that doubt emerges in the understanding of the preoperational child. This was something we saw very clearly in the research that came from Geneva in the 1970’s and 1980’s in the work of Willem Doise, Gabriel Mugny and Anne-Nelly Perret-Clermont, in which they showed the ways in which social interactions between children could be a source of cognitive developments (Doise & Mugny, 1984; Perret-Clermont, 1980). Children who are confronted with a conflict of perspectives as they work on problems jointly with other children are able to draw something productive from this experience. In these

Construction, belief, doubt 115 studies, in fact, it was not only pre-operational or intermediate children who made progress as a consequence of engaging in problems with children who were conservers, since in some of these studies it also emerged that children at the same developmental level, working together on a problem, were also sometimes able to produce solutions to the problem in their joint activity which went beyond anything either child was able to reach tackling the problem on their own. In this sense doubt appears as a function of the communications and the exchanges that take place in children’s relations with others. In parenthesis here we can contrast the ways in which Willem Doise and his colleagues talked about these situations with the kinds of explanations that Piaget himself had given for this problem of transition. In fact, I think that of all of Piaget’s work, these are some of the least persuasive pages that he ever wrote (see, for example, his discussion of equilibrium in Piaget, 1967). In addressing the issue of how it was possible for the child who believed there was more water when it was poured from one glass to the other might come to change that belief, Piaget too saw it as a question of conflicting perspectives and how the child might resolve conflicting perspectives. But he lacked (surprisingly perhaps, given his other writings on social relations and cognitive development, see Piaget, 1932, 1995) the means for seeing how that conflict of perspectives could in fact be engendered within the child. In his arguments here Piaget is most explicitly limited to considering the child as a single centre of epistemic agency engaged with the world. He talks about the child simply oscillating between one point of view and another, but it’s unclear why the child should begin to oscillate from one point of view to another. He seems oblivious to the child’s relations with others as a source not only through which different perspectives might be introduced, but also, as an arena in which a coordination of perspectives might be achieved.

Constraint and cooperation But Piaget also has some more interesting things to teach us perhaps about the nature of beliefs, doubt and knowledge, especially if we go back to some of Piaget’s earliest work – to the work on the moral judgment of the child and to the points where Piaget was most deeply engaged with knowledge as a social process as much as it was a psychological process (Piaget, 1932, 1995). In these works, Piaget in fact makes a distinction not simply between two forms of social relations, but also between the forms of understanding that those two different forms of social relation can sustain. In his work on the moral judgment of the child, he distinguishes between co-operation and constraint, that is between relations which are symmetric in terms of the power distributions involved and relations which are asymmetric. For Piaget, in relations which are asymmetric, where there is constraint, there is only the possibility of belief, or what he describes as knowledge acquired through social transmission where we accept an idea simply because of the prestige of its source. On the other hand in co-operation, ideas are the product of some

116


productive engagement, there is an elaboration, there’s a construction rather than simply the acceptance. It is in symmetrical relations that there is the possibility of constructing operational thought. When we look at Piaget’s distinction now, we see it also as a distinction that we could express today in the language of social influence processes. Thus where Piaget talks about cooperation that has a form of symmetry, he is also talking about the ways in which engagement in forms of social practice and communication lead to some form of conversion or innovation where one pattern of thought is transformed into another through some kind of elaboration. On the other hand where there is constraint, there is an asymmetry of power. We accept knowledge or a pattern of belief because of the prestige of the source and here we can talk about forms of compliance. If I can perhaps stretch our use of the terms knowledge and belief a little, we can make a kind of table (see figure below) where we can start with Piaget’s distinction between constraint and co-operation which refers to asymmetric and symmetric social relations. We can link these to forms of compliance on the one hand and to forms of conversion or innovation on the other. And at least in one of the ways in which we have been using the contrast between beliefs and knowledge, but I think only one of the ways, we can also talk about the former as a realm of beliefs, and the latter as realms of knowledge. CONSTRAINT

COOPERATION

ASYMMETRY

SYMMETRY

COMPLIANCE

CONVERSION/INNOVATION

BELIEF

KNOWLEDGE DOUBT

Certainly this is partly the way in which Moscovici (1998) contrasts representations based on beliefs with representations based on knowledge. Those based on beliefs form very closed and contained totalities, whereas those based on knowledge are more open to the world of experience and hence are also more open to change.

From the communication of doubt to new forms of understanding If we think back to the pre-operational child, we need to understand how they can move from the state of belief into a state of knowledge in some way. One of the important functions of doubt here is that it precisely offers a path that leads from beliefs to knowledge. In this sense the emergence of doubt is what can set the young child on a path towards new constructions. Doubt arises through communication and is always in some sense the mark of difference. It is others who bring doubt into our lives even if the other is no more than the other of our own internal

Construction, belief, doubt 117 dialogues at some point. And this sense in which doubt arises through communication with the other also suggests that we need to attend to how it is that a certain kind of structure, a certain organisation of social psychological operations includes within itself also a certain kind of defensiveness. It holds within itself that which is recognised and that which is accepted and expels and puts to the outside that which is rejected. This is the classic Freudian account of the nature of negation (a paper I was drawn back to in a recent seminar in Cambridge by a colleague, David Hillman). In Freud’s paper on negation, although he does not actually use the term belief, one could say that this is what he is discussing. He describes very clearly the dynamic through which we try to hold all that is good within the self and to expel all that is bad. Beliefs are held within the structure of the self, whereas what is contrary, what attacks that belief, is pushed to the outside. We see here again that what is doubtful or what may be contrary to our beliefs is expelled. It is rejected and belongs to the other. When we come to entertain doubts the expelled or the rejected returns to haunt us somehow. If it is not exactly the return of the repressed, it is the return of the expelled. Doubt in this sense expresses a lacuna in our ways of thinking. It is the point where the operations of closure, that totalising effect or the totalising function effected by a belief, no longer holds. It loses its power, it loses its force and it exposes us to some necessity of engaging in some effort, some representational work to recuperate our sense of stability, our grasp of the world. In our own work, we have begun to explore some of the ways in which doubt emerges within the patterns of communication by looking at conversations between children as they try to solve problems together (Leman and Duveen, 1996, 1999). In these studies we have adapted the methods devised in the work in Geneva, so that we have brought together children who have different forms of knowledge to work on a common problem. In one of our studies (Leman and Duveen, 1999) for instance, we have brought together children who initially gave different kinds of responses to some of Piaget’s classic moral judgement stories, either autonomous or heteronymous responses, and followed their conversation as they try to resolve the problem jointly. In this study we have also manipulated the gender of the children so that the pairs of children reflect not only differences in epistemic status, but also differences in social status. It will probably come as no surprise to you to discover that a girl who holds the autonomous position has a great deal of difficulty convincing a boy who is in the heteronymous position that autonomy is some kind of advance. This difficulty is marked in a number of ways. The kinds of conversations they have are not only much more extended in time, but also the quality and quantity of arguments they bring forward are much more elaborate and much more sophisticated. But what is equally interesting though is that in those conversations the boys also begin to engage in a much stronger form of defence of their position. Even if for the most part they finally accept the autonomous position, they also start to explicate something about why they believe in the heteronymous position. In such conversations one really can see a process of conversion at work. Precisely because there is a

118


clash of perspectives that draws the different points of view together. From a strategical point of view one could say that what the girls are seeking to do is to establish doubt in the boys. What the girls do is they attack the boys by focusing centrally on the boys’ own arguments. When the boy says « Oh it’s this boy who is naughtier because he broke more cups », the girls try and get them to explain why they think breaking more cups is what makes one child naughtier and then challenge precisely that. The girls are not aggressive with the boys, they make their interventions in ways which are not threatening. As Sandra Jovchelovitch commented, what the girls are doing is bringing the boys an apple, and when the boys bite into it, it becomes a new knowledge for them. In these conversations one can also see that one of the key functions of beliefs in fact is to provide or to furnish a stock of explanations which are available and which ensure that the operation of closure can be effected. In fact both beliefs in the sense of something which is constrained and knowledge in the sense of something which is constructed are patterns of understanding which act as repertoires of possible explanations. It is by having explanations always ready that closure is always possible. At its most dramatic one can see this in forms of psychosis where people have endless explanations, so that whatever happens nothing remains unexplained. Doubt, then, creates a lacuna in people’s ways of understanding. A lacuna that in some way is intolerable for us also means that doubt engenders a state of dissonance. It is a state of dissonance which can be resolved along several different pathways. Firstly, and perhaps most surprisingly, sometimes the disjunction between one way of thinking and another can just be accepted. We simply find a way to live with it in a state which Moscovici has described as one of cognitive polyphasia. A dramatic example of this can be seen in our work in India (Wagner et al., 1999, 2000), where one of the women we spoke to, in describing the ways in which she envisaged both traditional forms of understanding of mental illness and contemporary medicine, said about the traditional forms: « I have some faith but I don’t believe ». This seems to me a way of living with the dissonance, of accepting it, of simply making it liveable somehow. A second kind of response is what we can call compliance – where there is a shift to another system of belief in which the lacuna can be incorporated so that somehow the space that has opened can be sutured by another kind of belief. The child might, for instance, shift from believing that the amount of water is different, to believing that the amount of water is the same but without constructing the operations which sustain this new form of understanding as a necessary knowledge. Or from believing that a child is naughtier because they’ve broken more things, the child might come to accept that a child is naughtier because of the intention that they had, but again without re-constructing the basis for this new position. Lastly one can resolve that dissonance through some kind of conversion by shifting towards the constructions of new forms of knowledge. Here the representations really do become, as Moscovici says, representations based on

Construction, belief, doubt 119 knowledge because they are open to the lessons of experience in some way. But this also, I think, means that in talking about belief and knowledge, we are often not talking about two different things, but about the ways in which forms of understanding are constructed and sustained within particular ways of engaging within the social life or within people’s general orientations towards the world. In drawing up this kind of scheme, I think what I have described is not two sharply different social relations that one might see in practice, but rather, two ideal types. Wherever you see children engaged in practice, in realities, what you actually see is some mixture of these two types of relation. In fact I think that it is very typical of most of our social relations that they do not fall under the heading of being either simply relations of constraint or simply relations of co-operation. Just as most of our understanding is not either simply belief or knowledge, but also always some kind of polyphasic construction in which there are elements of both.

Discussion By way of discussion, I have included here my responses to questions and comments made during the meeting. A first question came from Xenia Chryssochoou who asked about the sense in which the development of knowledge in children can be considered as a form of innovation. One of the lessons we can learn from Piaget is that children do not simply receive an understanding of mathematics or logic from the outside. Rather, they reconstruct it. And in this sense it is an innovation for them. It may not be an innovation in the history of mathematics to understand that two and two can be four, but children have to reconstruct this knowledge if they are going to possess mathematics in a way that they can use it flexibly, that it can become a tool for them. If the only way in which they acquire this understanding is through some form of constraint, then that understanding is always going to be very fixed and immobile for them. It is never going to be a very flexible tool that they can adapt in different ways, because its use will always be limited to the kinds of circumstance in which it has been acquired. It will simply lack flexibility somehow. Even coming to understand things like the fact that the world turns round the sun rather than the sun going around the earth is an innovation for young children. If we look back also to Piaget’s early work on the child’s representation of the world, what strikes one is how much of the child’s representation, or the preoperational child’s representation of the world, actually also corresponds to various kinds of archaic and medieval cosmologies. Such ideas endured centuries, and even if we now take it to be common sense that the earth goes round the sun, it was not that long since the earth was considered flat and everything else turned around it. A second set of issues was raised by Nikos Kalampalikis, who commented on the significance of mythologies for the patterning of beliefs, Xenia Chryssochoou, who asked about the equation of children’s thinking, adult thinking and magical

120


thinking which seemed to be at the heart of the model, and by Alain Clémence, who emphasised the importance of a relational instability as a source of doubt in the conversations between children. I took the example of children in part because it’s what I have been working on, but mostly because in looking at the problem of how children’s thought changes and develops one can see precisely that there is a kind of paradox here from the point of view of a constructivist perspective. There is a sense in which of course in making an equation between children’s thinking, adult thinking and magical thinking, we are on the whole contemplating circumstances in which thought is constrained in some ways. This is precisely what we mean by representations based on belief, that they operate through being rather closed, thick patterns of thought. One can see this in the child, but one also can see it in a lot of everyday adult thinking, and it is very evident in magical thought. To say that it is closed does not mean to say that it cannot adapt to new situations. What do we do when faced with the unknown? Most of us have some kind of magical response to it, as can be seen in Gina Philogène’s discussion of anticipations of the year 2000. When the child begins to establish that in fact not only is the amount of water the same but that it is necessarily true that it’s the same, they’ve emerged into some new pattern of thought. I am sure there are instances which we could find within the world of adults where we could see similar transitions. This perhaps brings me back to Alain Clémence’s comments as well. There is some sense in which the first kind of instability which emerges in our relations with others is a relational instability. But it’s what one does with that relational instability which is the important thing. One can respond to it simply in terms of some form of compliance to whichever is the more powerful figure in that relation. Or one can take that relational instability and work towards some innovation from it. In some of our current research (Psaltis and Duveen, 2002, which is not about moral judgement but about the conservation of the amount of water) as well as looking at the conversations between children we have also begun to look at the outcome. In this work we are beginning to see that we can link features which characterise the conversation as a form of compliance or conversion to whether or not children show progress on a post test. This is very preliminary still, but there is something interesting about these kinds of engagements. They are social relations and precisely because of that the course of them is not very predictable. There are different opportunities, different possibilities that always emerge so that it is difficult to predict which pathway a particular conversation will follow. Sometimes one sees a path where influence operates through the form of compliance and at other times a path where influence operates through a form of conversion. Both possibilities exist, but which might be realised in practice is more difficult to predict. In part this is because the influences of other patterns of belief are also important, especially in these studies where the representations of gender are also engaged in the conversations. In the study using Piaget’s moral judgement stories (Leman and Duveen, 1999), for instance, when it is the boy who is in

Construction, belief, doubt 121 the autonomous position and the girl in the heteronymous position, one often sees very direct forms of compliance in which the conversations are very brief. The boy more or less just announces at the outset « well it’s this one who’s the naughtier » and the girl says « OK ». End of conversation. In these conversations one never gets to a position where the arguments from different perspectives are exposed. Where it’s the girl who’s in the autonomous position she has to work harder. Where the autonomous boy can assert a kind of power that constrains his partner, the autonomous girl has in some way or another to argue her way around her partner. And it is that argument that exposes shortcomings and weaknesses which can lead to some kind of reorganisation for her partner. Sandra Jovchelovitch commented: I see this model as a model that could be very easily transposed to other situations that are not necessarily situations related to how children get entangled with doubt. When you talk about belief and knowledge you talk about forms of representing. So from this kind of model you can work very comfortably towards other social situations, in putting questions about how traditions are established based on the criss-crossing between symmetric or asymmetric relationships. And instead of the apple being offered by a girl, today you have thousands of apples being brought to you by different people who come with different belief systems and introduce doubt into our takenfor-granted. That’s exactly what I would think. This model is established through the concrete example of what happens between children, but it is not just a description of this situation. It expresses something more general.

References Doise, W.; Mugny, G. 1984. The Social Development of the Intellect, Oxford, Pergamon. Duveen, G.; Lloyd, B. 1990. « Introduction », in G. Duveen and B. Lloyd (Eds), Social Representations and the Development of Knowledge, Cambridge, Cambridge University Press. Gervais, M.-C. 1997. Social Representations of Nature: The Case of the Braer Oil Spill in Shetland, Unpublished PhD thesis, Department of Social Psychology, London School of Economics. Leman, P.J.; Duveen, G. 1996. « Developmental differences in children’s understanding of epistemic authority », European Journal of Social Psychology, 26, pp. 683–702. Leman, P.J.; Duveen, G. 1999. « Representations of authority and children’s moral reasoning », European Journal of Social Psychology, 29, pp. 557–575. Moscovici, S. 1998. « The history and actuality of social representations », in U. Flick (Ed.), The Psychology of the Social, Cambridge, Cambridge University Press. Perret-Clermont, A.-N. 1980. Social Interaction and Cognitive Development in Children, London, Academic Press. Philogène, G. 2002. « Systems of beliefs and the future: The anticipation of things to come », Psychologie et societé. Piaget, J. 1967. « Genesis and structure in the psychology of intelligence », in J. Piaget, Six Psychological Studies, London, University of London Press.

122


Piaget, J. 1932. The Moral Judgement of the Child, London, Routledge & Kegan Paul. Piaget, J. 1995. Sociological Studies, London, Routledge. Psaltis, C.; Duveen, G. 2002. « Peer interaction, gender identity and cognitive development », Paper presented to the British Psychological Society, Developmental Psychology Section, Annual Conference, University of Sussex. Wagner, W. 1998. « Social representations and beyond: Brute facts, symbolic coping and domesticated worlds », Culture and Psychology, 4, pp. 297–329. Wagner, W.; Duveen, G.; Themel, M.; Verma, J. 1999. « The modernization of tradition: Thinking about madness in Patna, India », Culture and Psychology, 5, pp. 413–445. Wagner, W.; Duveen, G.; Verma, J.; Themel, M. 2000. « I have some faith and at the same time I don’t believe », Cognitive polyphasia and cultural change, Journal of Community and Applied Psychology, Special Issue: « Health, Community and Development », 10, pp. 301–314.

Abstract Axiomatic for a constructionist perspective in social psychology is the epistemological assumption that the known world is the product of the psychological operations through which it is constructed. Although some recent discussions have sought to distinguish between « strong » and « weak » forms of construction, such a distinction seems to subvert the very assumption it was intended to clarify. And yet rejecting such a distinction also seems to generate contradictions for a consistent constructionism – for if the known world is the constructed world it is difficult to grasp how such constructions might change, how some type of dynamic might be introduced capable of transforming the structure itself. One contribution towards resolving this dilemma is Moscovici’s suggestion of distinguishing between types of construction, that is, between beliefs, resistant to change and impermeable to experience, and knowledge, open to revision through experience. These different types of construction can also be seen as corresponding to different forms of social relations, constraint and cooperation respectively. But if these distinctions can be said to characterise different forms of construction, they do not necessarily distinguish between contents, that is the objects of the constructed representations. As studies of social influence processes have indicated, the emergence of doubt is a crucial element in the transformation of representations from the closed forms of belief into the more open forms of knowledge. A focus on the social psychology of doubt can, then, illuminate our understanding of belief.

Key-Words

Construction, constraint, cooperation, influence, doubt. Acknowledgements: I’ve taken the liberty of including some of the comments on my presentation made at the workshop in Paris, since they helped to give a shape

Construction, belief, doubt 123 and a form to my contribution. As well as being a way of acknowledging my gratitude to the participants at the meeting, it is also a way of recording something of the spontaneity and productivity which was such a notable feature of our discussions. I am also grateful to Joy Labern in Cambridge for transcribing my talk.

8

On interviews A conversation with Carol Gilligian

GD:

CG:

Well, first of all it’s a pleasure to have Carol back with us again and a pleasure I think we can look forward to in regular interval over the next couple of years. The second thing is that this is actually the continuation of a conversation, really, that began the last time you were here in October, when we discovered, almost by chance, that we had similar interests in relation to interviewing. And we thought it would be interesting to share this in some way. I think one point where we might begin is with some thing that was said by Jonathan Potter at a seminar he gave at the LSE a couple of weeks ago, where he announced to everybody that the interview was dead as a research tool. This is a kind of characteristically provocative statement. He wanted this to mean that we should give up trying to look for data in sources where the researcher has manipulated things in some way and we should look for freely occurring and spontaneous conversation rather than structured interviews of any kind. But I think both you and I have a conviction that he is profoundly wrong about this, and that interviews are actually or interviews should be a key and focal source of how we start to explore social psychological processes. Maybe, one place that we can begin is to go back to where we first started talking about interviewing, which was for me largely focused on Piaget and for you also some of Freud’s works. But perhaps where we can begin is just with something about where that interest in interviewing came from. I don’t know if you wanted to say something about what you saw in Freud’s studies on hysteria? I would be happy to take this up at the start. But also, since you started with the interview is dead and what we should do is to observe so-called real conversation, I would like to just say that does not in any sense address the problem. Because the minute you observe what you call natural, normal conversation, you of course change the conversation. So the idea that you could, somehow, make an observation or be present and see the thing in itself, it is just absolutely not true. So, it really takes us back to the question of why would you interview? What would lead you to interview another person? Both Gerard and I realized in a very accidental conversation that we had a very similar insight into the nature of interviewing and, in a sense, why it is key to a certain kind of psychological investigation, and what makes it an

On interviews 125 intrinsically empirical inquiry. Of course, we have a similar background in Piaget’s méthode clinique, that is his clinical method, and also to Freud. So I think I’ll just start by laying out what was the insight of Piaget, in a brilliant, very short, essay on method that introduces his book The Child’s Construction of Reality. He wrote this book in the 1920’s. It begins with a discussion of method that is as sharply relevant today as it was at the time. He says, if what you want to know or discover is how another person constructs reality you cannot use testing and you cannot use observation. Because if you use testing or the usual methods of assessment – that really have been regarded as the standard methods in psychology and are usually given the name science and seen as objective – what you are basically doing is you are mapping somebody else’s thinking or psychic functioning on your map of the world, because you have constructed the test. So you cannot discover, by testing, how another person constructs the world, how they would establish the test if they were making the test. And with observation you can observe but again you are interpreting completely apart from the other person. I mean you can see what you see and you can interpret it in a variety of ways. But what you don’t know is if their action or behavior or responses have a meaning to them that is very different from the meaning you would have attributed to it. So he said, if what you want to know is how another person constructs the world, what you have to do is what he calls his méthode clinique, that is you have to interview, but you have to interview in a particular way. Because you have to interview, basically, out of a deep sense of curiosity about how the other person is constructing. I mean what is the logic or construction that holds their thinking together? So if someone says something that is nonsensical to you that is what is really interesting, because they are obviously operating according to some structure or logic that is completely at odds with one’s own. When Gerard and I were talking at lunch today, Gerard was recollecting that when Piaget went to Paris and worked with Binet and Simon on the development of the intelligence test, what fascinated him was not so much the usual questions, ‘if Sally is taller than John and John taller than Erika, then who is the tallest one’, kind of thing. But instead he was interested in errors that children made. Because to the child who makes the error that seems like the right answer. So according to what is his logic, an answer that seems to you to be wrong, seems right to him. This is, of course, tremendously relevant to all the discussions today about cultural differences and so forth. But it identifies a particular attitude and strategy of interviewing which is not passive. But where in the process of the interview, and this is what we will really come to, you, the inquirer, seeking to understand how the other person discovers the world, have to ask a series of questions till you come to understand from that person the logic of their thinking rather than how that person’s thinking might fit into your own logic. It is a very different way of thinking about the interview. And for me that essay connects directly with Freud’s work on studies on hysteria and particularly the case of Elizabeth von R., which he says was the first full length analysis that he ever did and I read

126

GD:

Development as decentration that case as an essay on method, because what Freud describes in that case study is his discovery of a method of coming to understand, basically, what is the structure of hysteria. What is the connection between the symptoms and the event that lead up to the symptoms? I mean how you can understand a logic of dissociation, that is a failure of the usual kind of connection. So that something gets expressed in the symptom – in the case of Elizabeth in a paralysis and pains in her legs – and what is the meaning of that. And he describes a method of discovering the logic of the psyche that is a different logic from a binary logic or formal logic. And that is the background, really, to our discussion of method, because what Freud and Piaget have in common in both of these tests is what is usually set antithetical to objectivity and neutrality on the part of the researcher, which is they very actively engage in the process in the interview itself, testing out ‘hypotheses’ in the course of the actual interview. This is more Gerard’s language but I’ve found it just immensely helpful. Do you want to take it here? Yes. Perhaps I’ll take a step backwards and just remind us of the origin of Piaget’s méthode clinique because I think that is quite important and revealing. He came across this when he spent most of a year at the Berghoff clinic after his own ph.d. The Berghoff clinic in Zurich was a centre of psychoanalytic work. And one of the things that impressed him there was watching the clinicians interviewing new patients. Someone is brought into the clinic and it is the clinician’s task to uncover what is their problem. And they were doing this by using a very structured process of interviewing them, of questioning, reasoning, of trying to talk to a person whom they had never met before, who had been brought to the clinic because they seem to be suffering from some kind of mental disturbance. And it was the clinician’s responsibility to find, at least in a provisional sense, some kind of diagnosis. And the interviews were, in some sense, a process of discovery. And it was this that Piaget put to use. He went on to Paris from Zurich and in talking to children started working with standardized IQ items. As you said, the first thing that struck him was that when children make errors on these kinds of tests they don’t make random errors. Their errors are always organized and structured. They believe they’re getting the right answer. Intellectual development for Piaget is always a process that goes from one organization or structure to another. You don’t go from some random or confused stage to some stage of structure. So it became important to him then to find ways of understanding the mentality of the young child, and this is were he began what was in some ways a quite revolutionary process. He began talking to the child. And the connection you make with Freud sounds quite interesting because I think one of the things that Piaget and Freud shared is that they have an attitude toward their interlocutor – whether their interlocutor is a child or a patient – that what this person has to say is meaningful. And I think in psychiatry Freud is quite a radical in this sense of taking seriously what the patient has to say, that the patient is in some ways disclosing their construction of the world through what they have to say, just as the child is. You have to find a way to hear what

On interviews 127

CG:

that construction is. The point of the méthode clinique for Piaget – as it emerged over the 1920’s and as he began to use it more systematically – was that he might start by asking the child a question, about almost anything. In The Child’s Construction of Reality Piaget’s questions to the child are often very odd in some ways. For example, ‘where is the name of the sun?’ or ‘Is the name of the sun in the sun?’ Whatever children might say to him, he begins the process of questioning, of arguing and of always making an effort to grasp the mentality that produces the child’s responses. So at every point he is framing for himself a hypothesis about what that mentality is and looking for a question that allows him to test that hypothesis. And that is what gives shape and structure to the interview. There is one sense in which I think Piaget’s interviews take a different route to Freud’s. What interests him the most in talking to children is to reach a point where, in a sense, the child reveals a conviction about the world, a fundamental belief that is unshaken no matter what kind of influence Piaget might bring to bear on the child. They just stick and say ‘this is how things are’. It is those kinds of convictions that Piaget finds most interesting and revealing, and often they come out of a process of questioning. The complementary between the two is interested in the structure of conviction and belief and knowledge. And Freud becomes interested because he starts with hysteria. The puzzle of hysteria is how can you explain, for example, why a person who has no obvious neurological problem can’t see? Or speak? Or walk? Or whatever? How can you explain this? And that leads him to a very different kind of logic, which can be associative logic of the psyche, where it is not one or the other – both are true. And he’s fascinated therefore between what people know and what they know but can’t say that they know, which is of course dissociation at the height of hysteria. And Freud would bring what he would call unconscious knowledge to consciousness, so that it can be reflected on. So in that sense, I mean, it’s different from Piaget but both seem to me equally valid. I was trained initially when I was working with Lawrence Kohlberg in Piagetian interviewing. As a critical psychology graduate student, I had learned the psychodynamic and Freudian type approaches. I had learned in that early work that I did as a research assistant using the Piagetian method that you had to interview somebody until you so clearly understood the logic of their thinking that you could actually answer a question in their logic. So it really was almost an exercise in ‘negative capability’. You had to get out of your own head and into the category of somebody else’s thought. You had to make sense of what they were thinking. When I began interviewing women for my research, I saw that there was something in women’s responses or some women’s responses to hypothetical dilemmas that simply didn’t fit in to the categories of existing theory. I mean they were making “a mistake” that could be scored at the low stages of development and these were very bright women. It makes you think, ‘well, why don’t they see the logic of the right answer’, which in some sense seems obvious once you can see it. But the Piagetian question would have been: ‘for

128

GD:

CG:

Development as decentration them it wasn’t a mistake. So what was the category of their thinking that made that answer seem right to them?’ And it was that question that led to the writing of In a Different Voice. Which is literally not a woman’s voice but a different construction of reality. It was a different construction of what constitutes a modern problem. How you, you know, construct the self in relationship to others and what constitutes an ethical problem in that relationship. That was the technique of interviewing. You know it’s always true that when you’re doing research that the discrepant data are the most interesting data, I mean the data that don’t fit, because they tell you what can’t be explained or understood in terms of the prevailing theory. And interviewing as a method, to me, it’s the method par excellence for any kind of discovery research that requires you to enter another person’s way of seeing the world, understanding the world or experiencing the world. If you’re going to work with discrepant data you almost have to work that way, because it’s very discrepancy means it cannot be assimilated in to the construction of reality which is reflected in theory. And the only way you can see the construction of reality in a particular theoretical approach is to see what doesn’t fit into it. Otherwise the theory looks like reality, and in that sense it’s like culture. If you’re completely within one culture, you don’t see it as culture until you step outside into another culture and suddenly you can see that was the construction. So writing a book called In a Different Voice, one of the deeper implications of the book in the title was that what was said to be psychology, objective, neutral and so forth, was in fact a voice but you could not hear that voice unless you identified a different voice. That takes us back also to something about the process of interviewing. It seems to me there is the sense of continuity of structuring a hypothesis about how your interlocutor is understanding the world. There is an interpretive moment. And what I think is true what is clear in Piaget and Freud is also clear in A Different Voice. As an interviewer you’re always looking for some kind of interpretive framework, which allows you to grasp the mentality of your interlocutor. That interpretive framework can be something that is malleable and develops as you interview more and more people. If you read the early Piaget, you can see that interpretive framework taking shape. He picks up ideas from Lévy-Bruhl, from Freud, from other sources. He’s looking for a way of characterizing the mentality of young children in distinction to the mentality of older children and adults. He’s looking for an interpretive framework, and I think in your work, the book begins in a sense from this question that’s raised about the interpretive framework that comes from Kohlberg’s theory, which seems to be excluding from consideration or denying the value of what was being said by the women you were speaking to. First of all: Kohlberg’s theory interested me at the time because it was explicitly a theory of moral development. It made explicit a set of values that was inherent in all psychological theory. It was so in line with the surrounding culture that it was like water to a fish. This was the premise of separateness

On interviews 129 and the value on autonomy and independence, which was equated with development. So, in essence, psychologists were reading culture as nature. And it was that aspect of Kohlberg’s theory that spoke about the values and made values explicit that was implicit in the other work, which was very interesting to me and very revealing. But I have to say – and this is just a piece of history that I think is important – that what first caught my attention, what first led me to actually start to study people’s responses to real, rather than hypothetical dilemmas, had nothing to do with women. It was because I was teaching at the time. I was teaching a section of a large lecture course that Kohlberg was running on morals in political choice, and it was the time of the Vietnam War. With my students in my section we were talking about, you know, hypothetical dilemmas and objections to the draft and the war and so forth. I noticed that the men in my section didn’t want to talk about the draft dilemma. Because they were operating within the framework of Kohlberg’s theory, which was also within the framework that absolutely correlated with not only western psychology but western philosophy in the whole country and tradition and so forth. And they knew that if they spoke about what they were actually thinking about, which was in considering whether to object to the draft or to go and so forth. But they were also thinking about if other actions would affect their families and the people whom they were close to and their relationships, and they knew that within the framework of these theories to think about relationships, rather than abstract principles of morality, was at best low stage and a sign of a lack of moral maturity. They didn’t want to misrepresent themselves. So they actually didn’t want to talk about it. It was at that point that I thought that these theories, in representing – because I was really into gender at the time – human experience or men’s experience were not adequately representing men’s experience. And that was when I first started to do my research. I was going to study these Harvard students facing the Vietnam draft to the point where they become seniors and actually face the draft. I wanted to see what they did and how they thought about it. And then President Nixon ended the draft, which was the end of that [laughing]. This is why I studied hypothetical dilemmas [laughing] and the Supreme Court legalized abortion in Roe vs. Wade. This is 1973. And so my study was going to be of how people think about real dilemmas switched – and I was completely unaware of this – from a sample of men to a sample of women. And so the issue then became very focused by the disparity between women’s capacity to articulate a construction of the world that started from a premise that people’s lives are connected. That may not seem so radical today as it did in 1973 or 74. When you start from that premise, then your thinking doesn’t fit into the categories of theories that structure on premises of separateness. And the reason I called my book In a Different Voice and not In a Woman’s Voice is that women were articulating aspects of human experience that were also known to and for men, but were less readily articulated by men as they were seen as antithetical to certain notions of masculinity. And so that was how this work happened. But it was almost classical in following that

130

GD:

CG: GD: CG:

Development as decentration approach to inquiries related to the question ‘why interview?’ and I think I want to just come back to your question that you started with ‘Is the interview dead?’ or should you just observe real conversation, which of course you can never do, because everything stops being normal conversation the minute you start recording it. I think the interview is not some kind of facsimile of conversation it is a particular kind of conversation taken after a particular purpose. The purpose has to do with inquiry and the nature of the inquiry is to discover another person’s construction of the world. And again, you cannot passively observe another person’s construction of the world. You can only, in some sense, come to understand it by engaging with it. And this very much changes how you think about the role of the interviewer in the interview. I would add another trait to this argument because in discovering another person’s construction of the world it’s not the world as immediately the other person sees it. It is the world as you see the other person constructing it. So in a sense what you’re doing through the interview or through the interview as a method is trying to establish a theory of your own that gives some place to the categories, the constructions, that you can identify over the course of the interview. So Piaget is not trying to simply be like a five-year-old child. He is trying to understand how the child’s construction of the world makes sense to him within some broader account of the development of intelligence. We might say about Freud that he’s not simply trying to see how the world looks to hysterics but to find some interpretive framework that makes sense of those hysterical symptoms within his own set of categories. He can think about the hysteric’s construction of the world in terms of the conflict that is present in their construction and so on. That’s interesting. I think you talk about the word ‘empirical’ and I talk about the word ‘relational’. Okay [laughing] Because I think it is also what he learns from the hysteric about the psyche and how hysterical processes illuminate aspects of psychological functioning in a way that shines a certain light on them that one might not see so clearly in other instances. And I think for Piaget, in The Moral Judgment of the Child, how the child’s conception of the rules of the game, of the games that children play, illuminates something in general about rules and the nature and function of rules in human society and how there are different ways of constructing rules that have very real implication for the difference between an authoritarian and democratic societies. Whether you see rules as coming from on high or whether you see rules as made by people to agreements in relationships. You know it’s interesting I haven’t talked so much about Piaget for a long time but it’s something to me that is actually so alive and so exciting about that kind of research and it is, I think, essentially relational but not in the usual way that word is used, because it is a process of inquiry that requires the active engagement of two people with one another. And all the assumptions about objectivity and neutrality in research are, in particular on the part

On interviews 131

GD: CG: GD:

CG: GD:

CG: GD:

of the researcher in the interviews, have come to seem to me basically misguided and highly problematic, because in essence this means that the researcher either brings interpretive framework and holds it throughout the interview – you’re back to assessment – or else doesn’t interview and goes away and all by himself/herself decides what it means with no potential to check that out empirically in the relationship with the subjects or participants. So if you talk about the word ‘empirical’, I can talk about ‘relational’. Yeah. I’m happy to talk about the word empirical, because what you’re emphasizing there is always a dynamic form. Yes, exactly. It’s a dynamic that operates because you do have two interlocutors that collaborate in some important way. Rob Farr always used to talk about interview and rephrase it. So, he called it ‘Inter-view’. Inter hyphen view. View was probably the wrong word. View is probably the wrong word. But it was a re-contextualization that makes one aware of that dynamic of exchange that was the half of the interview. The empirical part of it is really quite interesting, because the empirical part is itself continually in the process of reorganization and restructuring, because the empirical part is not static. You ask the child a question; they say something. You have to make some interpretation of the meaning of this something. And following the question it seeks to establish whether your interpretation is well founded or not. So the empirical is constantly being reworked somehow. It’s not a given and objective sense data that is there. It is something that you are very actively producing. You’re structuring things in some way and that structuring of something is leading to what you might call the empirical part of the interview, but the empirical part doesn’t in itself contain the interpretable aspect that you are searching for. This is something that you bring to it as a researcher. You want to talk about Einstein’s letter. Yes. When we were talking earlier we connected this one thing that we both found interesting. It was the sense that if you look back to Piaget and to Freud, part of their claim to scientific legitimacy is an epistemological claim based around their methods of investigation. For Piaget, certainly, the interview was always a hypothetico-deductive method. It was a scientific method. It was about the generation of hypotheses and the testing of hypotheses. But instead of being framed as an experiment, in the ordinary sense of the word, it was contained within this dynamic form of the interview. We might even call it a micro-genetic method in some respects. So for Piaget the interview is continually moving between a hypothesis and the testing of that hypothesis. Did you see a new hypothesis and seek to test it. And one thing I came across that struck me as very connected to this was a letter from Einstein to a very old friend of his Mikhail Solovine, in which Einstein tries to speak about his own vision of what the scientific method is and it’s extraordinarily close to that movement through the clinical interview. In some sense, we begin with some empirical data and an intuition about a theory that might explain this. This

132

Development as decentration intuition generates a set of propositions, of hypotheses, to suggest a new set of questions. And those questions find an answer in another set of observations, and either they are consistent with it or inconsistent with it. There’s a sequence there. Einstein obviously thinks of this as a sequence that takes place over extended periods of time, with a number of different experimental interventions. But the point that Einstein comes back to is that between the empirical data and the theoretical hypothesis there is no logical connection. He talks about these as being intuitions or only a psychological connection. And in this sense, for Einstein too, science has this interpretive aspect to it, because it’s an empirical science but not an empiricist’s science. And I think that is a point that is shared with both Piaget and with Freud.

9

The constructive role of asymmetry in social interaction Gerard Duveen and Charis Psaltis

Introduction: social interaction and cognitive development In his account of the development of moral judgment Piaget (1932) introduced a fundamental distinction between different types of social relationship, or more specifically he attributed different types of psychosocial processes to different forms of social relationship. Where there is constraint because one participant holds more power than the other the relationship is asymmetrical, and, importantly, the knowledge which can be acquired by the dominated participant takes on a fixed and inflexible form. Piaget refers to this process as one of social transmission, and he refers to the way in which the elders of a tribe initiate younger members into the patterns of beliefs and practices of the group. Similarly where adults exercise a dominating influence over the growing child, it is through social transmission that children can acquire knowledge. By contrast, in cooperative relations, power is more evenly distributed between participants so that a more symmetrical relationship emerges. Under these conditions authentic forms of intellectual exchange become possible because the partners have the freedom to project their own individual thoughts, consider the positions of others, and defend their own independent points of view. In such circumstances, where children’s thinking is not limited by a dominant influence, the conditions exist for the emergence of constructive solutions to problems, or what Piaget refers to as the reconstruction of knowledge rather than social transmission. Here the knowledge which emerges is open, flexible, and regulated by the logic of argument rather than being determined by an external authority. In short, cooperative relations provide the arena for the emergence of operations, which for Piaget requires the absence of any constraining influence, and is most often illustrated by the relations which form between peers. Piaget’s argument is in part motivated by his critical response to Durkheim’s sociology. While he recognizes much that is of value in Durkheim’s work, he also considers its limitations, insofar as it focuses only on the constraining influence of one generation over the succeeding one. This is one particular way in which Durkheim’s theory is oriented toward understanding what it is that holds societies together, and as Moscovici and others have suggested, such a view neglects to consider the equally important process of how societies change (cf. Duveen,

134


2000a). Piaget’s response to Durkheim is framed as an ontogenetic argument, but he too is concerned with the prospects for change, in this case of cognitive structure. In basing his argument on a distinction between different types of social relationship, Piaget takes the adult–child and the child–child relation as the ideal types of what he identifies as constraint and cooperation. Of course he is also aware that while ideal types may have analytical utility, they do not always describe empirical realities very clearly. Here, for instance, it would be absurd to ignore the multiple ways in which child–child relations might take asymmetric forms as one child dominates the other, or the way in which it may be possible for adults to interact in more symmetrical ways with children. It is the distribution of power in the relationship which is central for Piaget, and he insists that it is only under conditions of symmetry that children experience the autonomy necessary for the construction of operational structures. Indeed, so closely did the relationship between cooperation and cognitive development appear to Piaget that in a later work he remarks that “they constitute two indissociable aspects of a single reality that is at once social and individual” (Piaget, 1977/1995, p. 145). And in his comments on Vygotsky, Piaget also notes that “there is an identity between intra individual operations and the inter individual operations which constitute co-operation, in the proper quasi-etymological sense of the word” (Piaget, 1962, pp. 13–14). And he goes on to observe that the reason for this identity is that “actions, whether individual or interpersonal, are in essence co-ordinated and organized by the operational structures which are spontaneously constructed in the course of mental development” (Piaget, 1962, p. 14). Vygotsky too considers social relationships as fundamental to the development of psychological processes. But in Vygotsky’s (1978) account we do not find any distinction between types of social relationship comparable to that found in Piaget. When he formulates his general genetic law of cultural development he frames it simply as internalization as an intrapsychological category of a function initially established between people as an interpsychological category. And when he gives this general principle a specific expression as the zone of proximal development he refers explicitly to asymmetric situations, in which one partner (adult or more capable peer) uses his or her expertise to guide the activity of the other partner (see Duveen, 1997, for a discussion of this point). This formulation emphasizes that the asymmetry in the relationship is one of expertise (or perhaps one should say more precisely that it is an asymmetry in the availability of forms of semiotic mediation), but as always it is difficult to disentangle knowledge and power. Thus while both Piaget and Vygotsky emphasize the importance of social relationships for psychological development they do so in rather different ways. Vygotsky’s emphasis on asymmetrical relations seems to elide any reference to the kind of constructive activity which Piaget suggests is the core of development. Yet, at the same time, Piaget’s abstract theoretical scheme is not without its problems and difficulties either. Piaget himself never undertook any systematic research into the relations between social relations and psychological development, so it is perhaps not surprising that we do not find any clear account of how children (or any other social actors) come to establish cooperative relations rather than relations of

The constructive role of asymmetry in social interaction 135 constraint. In Piaget’s account these two different social relations appear rather as two distinct states of affairs, each with its own properties. But the world in which social actors encounter one another is above all a world constituted around differences, which may be differences in prestige, in authority, in knowledge, or along a multiplicity of other dimensions. Each dimension of difference is itself a potential source of asymmetry in social relations, so that it becomes important to ask how it is possible for cooperative relations to emerge from the communicative exchanges of interaction.

The hypothesis of sociocognitive conflict A central contribution to our understanding of the ways in which social interaction may be related to cognitive development came from the experimental program of research in Geneva initiated by Willem Doise and his colleagues, Gabriel Mugny and Anne-Nelly Perret-Clermont (cf. Doise & Mugny, 1984; Doise, Mugny, & Perret-Clermont, 1975; Perret-Clermont, 1980). By re-presenting some of Piaget’s classic experimental investigations as problems for pairs (and sometimes triads) of children, this research was able to focus on the consequences of interaction for children’s cognitive development. In their designs, children’s performance on a pretest enabled the researchers to control the composition of pairs of children who would be invited to work on the same problem together in an interaction phase, and the outcomes for individual children could be established through subsequent posttests and by comparing the performance of children who had participated in such an interaction with a control group who had not. Using this design, these “socialGenevan” researchers were able to establish that children did indeed benefit from working with another child, especially when their partner was more advanced in his or her understanding of the task. In their work these authors also reported some other interesting observations. First, in a number of cases they reported that children produced novel arguments in the posttest; that is, they produced arguments which they had not heard from their partner during the interaction phase. This was one important element sustaining their claim that in these circumstances children’s development was the product of a structural reorganization of their cognitive functions, rather than simply imitation of their partner. Second, on a more complex task of spatial perspectives they also observed that children working together could produce solutions which were in advance of what either child had achieved in their pretest. As well as providing further evidence that in these situations children were not simply imitating their partners, such observations can also be seen as questioning the Vygotskian assumption about the necessity for a difference in expertise as a facilitator of development. While there have been some studies which have claimed not to have been able to replicate these effects (Russell, 1982; Russell, Mills, & Reiff-Musgrave, 1990), the weight of evidence clearly supports the productive role of social interaction. Not only did Doise and Mugny (1984) find similar effects across a range of tasks, but a number of other studies have also reported congruent findings (Ames & Murray, 1982; Bearison, Magzamen, & Filardo, 1986; Mugny, De Paolis, & Carugati, 1984; Perret-Clermont, 1980).

136


To explain their experimental results, Doise and his colleagues introduced the hypothesis of sociocognitive conflict (Doise & Mugny, 1984; Doise, Mugny, & Pérez, 1998; Mugny, De Paolis, & Carugati, 1984; Perret-Clermont, 1980), by which they meant that in these interactions different perspectives on the task were embedded in the social relations between the children tackling it, and that by facilitating the expression of different perspectives these interactions were also contributing to the resolution of the conflict between these perspectives. Thus, where Piaget himself had suggested that cognitive development emerged through the child’s confrontation with conflicting perspectives, this hypothesis framed this process as a function of social relations. While this hypothesis of sociocognitive conflict appears to be theoretically coherent, some problems are clearly evident. First, in the Genevan research it was never possible to identify specific features or characteristics of the interaction which could be linked to outcomes for individual children, so that sociocognitive conflict remained a general inference about what had happened rather than an observable feature of the interaction. And second, not every child who participated in an interaction with a peer made progress on a posttest. To explain this differential effect of interaction, Doise and his colleagues drew on theories of social influence to suggest that while some children were able to work toward a constructive resolution of the conflict (envisaged as similar to the process of innovation in minority influence), others achieved only a relational resolution, a kind of compliance to the power and authority of the more developed child, which did not lead to the emergence of any more developed understanding. But again, this distinction, while theoretically interesting, is really a post hoc interpretation rather than a direct observation of the interaction (and one could add in parentheses here that while the Vygotskyian proposition that development consists in the internalization of interpsychic relations as intrapsychic relations, his work offers no explanation of why some interactions may not lead to internalization). Notwithstanding these problems, the work of these “social Genevans” nevertheless also introduces a clearer theoretical frame through its focus on the triadic relationships among child-child-task as the fundamental unit for analysis. In other words, theoretical explanations for the relations between social interaction and cognitive Task

Child

Figure 9.1 The triadic structure of relationships

Child

The constructive role of asymmetry in social interaction 137 development need to consider the relations between the children as well as between the children and the task (Chapman, 1991, also introduces a similar idea). This emphasis on triadic relationships is important for a number of reasons. First, it serves to establish a distinction with classical Piagetian accounts which have tended to emphasize the dyadic relationship between subject and object, child and task. Actually, whether this is best characterized as a distinction or a clarification is perhaps a moot point. While there has always been a clear orientation toward the significance of social interaction in Piaget’s work (e.g., Piaget, 1977/1995; and the contributions to Carpendale & Müller, 2004), it is not clear whether this extends to encompassing a triadic model of development. In his methodological comments on the clinical interview Piaget shows himself to be a very sensitive analyst of the intricacies of the relationship between interviewer and child (Duveen, 2000b), even to the point of recognizing the potential structuring influence of the interviewer. Yet when he turns to theoretical explanations of development, social influences are given less weight than the autoregulative process of equilibration, with the former being considered one of the necessary but not sufficient conditions for intellectual development, while only the latter acquires the status of being both necessary and sufficient. Even in his most sociological writings, Piaget (1932, 1977/1995) hesitates over whether social interaction can ever be seen as having a constitutive role in the genesis of structures. At best he usually restricts himself to analyzing the homologies he finds between structures of operations on the one hand and structures of interaction on the other, with both being driven by the rationality implicit in the logic of the coordination of actions. In his broader considerations of the factors in mental development, Piaget always insisted on the necessary role of social interactions, while maintaining that they were not a sufficient condition for development (e.g., Piaget & Inhelder, 1969). Thus, while social relations might be considered a constituting element in the development of operations, for Piaget they did not constitute the whole. By contrast, the hypothesis of sociocognitive conflict suggests precisely a constitutive role for social interaction in the ontogenesis of intelligence. Second, the emphasis on a triadic structure of relationships also opens a theoretical space for considering intellectual development as a social-psychological process, in particular to the genetic social psychology of Serge Moscovici which has also emphasized the triadic structure of self-other-object as the central unit of analysis (Moscovici, 1972, 1976, 1980; and also Marková, 2003). Indeed, Moscovici’s work on social influence should be seen as an important element in the formation of the hypothesis of sociocognitive conflict. For Moscovici, it is precisely the forms of communication within interaction which structure different modalities of social influence, and in particular he draws a distinction between compliance (often described as majority influence) in which patterns of communication lead to the public acceptance of a position, and conversion (or innovation, and often described as a minority influence) in which patterns of communication induce cognitive reorganization. Indeed, in his research he has demonstrated that minorities can be influential to the extent that they are able to structure communications within interactions in such a way as to induce precisely a form of reflection which generates a

138


process of reorganization. There is clearly a parallel between Moscovici’s description of these different forms of social influence and Piaget’s distinction between constraint and cooperation as types of social relation (cf. Butera & Mugny, 2001; Doise, Mugny, & Pérez, 1988; Duveen, 2001; Psaltis, 2005a). The hypothesis of sociocognitive conflict suggests that development through interaction is envisaged as a kind of conversion process or innovation generating cognitive reorganization, and hence the significance of Doise and Mugny’s (1984) observations of novelty in posttests following interactions since they provide an index of such reorganization. In Moscovici’s theory, the capacity for minorities to generate change by influencing majorities is viewed as a function of the behavioral styles which they adopt. This has led researchers (Doise, Mugny, & Pérez, 1998; Leman & Duveen, 1999) to suggest that interactions between children be considered as communicative exchanges framed as social influence processes. Emler and Glachan (1985), for instance, linked the argumentative style of conservers with what Moscovici described as the behavioral style of consistency, since their operational grasp of the task was more likely to provide constant support for their argument throughout the discussion. More generally, Moscovici’s discussion of behavioral styles can be seen as related to Piaget’s (1932, 1977/1995) distinction between social relations of cooperation, which foster the emergence and resolution of conflicts of centration and thus provide the context in which the child is able to reconstruct knowledge, and social relations of constraint in which knowledge is not reconstructed but simply accepted because of the prestige of the source. Piaget defines the relation of cooperation as “reciprocity between autonomous individuals” (1977/1995, p. 232). Piaget also emphasizes the importance of reflection for cognitive development which he sees as linked with relations of cooperation, and notes “reflection is not spontaneous to individuals” (p. 234) but rather arises through social interaction, since the conscious realization of the methods and procedures of one’s actions are less fruitful when an individual’s action is not reflected in that of others; that is, when cooperation does not complete pure individuality. Third, the triadic view also enables a clearer perspective to emerge on the character of the communication between the children themselves as they work together on a task. In this respect, research on sociocognitive conflict may also be connected with research on peer interaction from the Vygotskyian perspective, where the emphasis has been more on the process rather than the outcome, which has also led to a greater reliance on more ethnographic forms of research. In looking at interactions between more and less capable peers, Tudge (1992), for instance, suggested that verbalization of reasoning by both partners was a condition under which cognitive growth is more likely to occur during peer collaboration, and specifically that the more competent partner had to verbalize his or her reasoning and the less advanced child had to accept it. A focus on the communicative demands of the task situation also gave rise to a second generation of studies derived from the initial research on the theme of sociocognitive conflict. These studies (Perret-Clermont, Perret, & Bell, 1991; Perret-Clermont & Schubauer-Leoni, 1981; Schubauer-Leoni & Grossen, 1993; Schubauer-Leoni & Perret-Clermont, 1997; see also Perret-Clermont, 1994) have

The constructive role of asymmetry in social interaction 139 also been more focused on questions of process, and in particular on the pragmatics of the communication between adult and child in the testing situation, emphasizing the form of communicative contract established between them.

Social interaction, social identity, and social representations As children work together on a task the dynamics of the interactions between them may be shaped by a number of different influences. There may be differences between them in their initial levels of understanding of the problem, but any conversation also embraces other sources of difference and conflict, each of which has the potential to influence the form and shape of the conversation which emerges between the children. Thus it becomes important to consider what it is that children bring into these conversations, the background knowledge which, while it may often be implicit, nevertheless exercises an influence. Children may be attentive to the academic or social reputation of their partner, for example, or the character of their relationship with their interlocutor within a broader social network. While it would be interesting to consider the influence of asymmetries derived from such sources (cf. Psaltis, 2005b), gender, or, rather, social representations of gender, is a persistent influence, introducing a complex dynamic into conversation. Children not only enter a conversation with their own sense of their gender identity, but also bring to it expectations about their partner and the appropriate forms of interaction which derive from their shared representations of gender (Lloyd & Duveen, 1992). What a child, whether a boy or a girl, expects of a partner in terms of authority or power in this kind of interaction differs according to the gender of the partner. Such expectations may, of course, be challenged or modified through the course of the interaction—talk itself can also always be productive in this sense—but gender remains a pervasive influence in children’s conversations. A first exploration of this idea came in the work of Leman and Duveen (1999) which modified the design used by Doise and his colleagues by controlling the gender composition of the pairs of children participating in the interaction. In this study children were engaged with one of Piaget’s (1932) moral judgment dilemmas (which contrasts judgments about the consequences of an action with judgments about intentions), and the pretest established whether the child was oriented toward a heteronomous or autonomous solution. Four different dyads could then be formed by controlling both gender and initial level of development (see Table 9.1; all of the children in this study were aged 9–10 years). Analysis of the interactions within each of these pair types showed that the Fm pairing was distinctive in several ways—these conversations took longer to reach an agreed solution than any other pair type, and the girl had to deploy a wider and more sophisticated set of arguments to persuade her partner. Of course this is also the pair type where the initial level of understanding is most strongly contrasted with expectations derived from representations of gender—generally the boys do not expect their female partners to take a leading or assertive role in the conversation, and their initial resistance to the girls’ argument leads to both longer conversations and the broader set of arguments deployed by the girl.

140

Development as decentration Table 9.1 Pairings achieved by manipulating gender and developmental level in Leman and Duveen (1999) Autonomous

Heteronomous

M M F F

m f f m

Note: M=Boy, F=Girl. Upper case letters refer to the autonomous child, lower case letters refer to the heteronomous child.

It is instructive to review the form of these different conversational patterns. In the first example (an Mf pair), it is the boy, James, who opens the conversation, but he does so through an assertive statement which sets out his perspective while leaving little space for the girl, Nadia, to challenge, or even participate in the agreed solution to the task. Her contribution to the solution is the simple repeated monosyllable “Yeah”. Example 9.1: Nadia & James 1. 2. 3. 4. 5. 6. 7. 8.

J He’s not as bad ‘cos he only opened the door to get tea but he’s naughtier ‘cos he’s not meant to be having sweets. N Yeah. J So he’s naughtier? N Yeah J Go get him back in then. Go say we’re finished. Experimenter enters. N That one. J That one.

In the second example (an Fm pair), it is again the autonomous child who opens the conversation, but this time Jade takes a noticeably different approach. Instead of asserting what the answer is, she opens a space for dialogue by asking Leon why he thought John was the naughtier of the two boys in the vignettes. By doing so she opens the problem space for discussion, and as the conversation proceeds she allows Leon the opportunity to articulate his own perspective, while also bringing forward arguments of her own which highlight aspects of the situations described in the vignettes which Leon has not taken into consideration. Example 9.2: Jade & Leon 1. 2.

J That’s David and that’s John. Why did you think John was naughtier? L Because he’s broke more cups and he’s also knocked over a table.

The constructive role of asymmetry in social interaction 141 3. J Yeah but his mum said that when his mum called him for dinner she didn’t tell him that there was a table behind the door and . . . to be careful did she? But David. . . . 4. L David . . . 5. J But David, he’s naughtier ‘cos he got up on a chair and he was stealing sweets. 6. L Yeah, but his mum was out. 7. J Yeah, but his mum still called him for dinner and she didn’t tell him that there was a table behind the door, did she? What one d’you think? 8. L Well, he’s only broken one cup so how’s that being naughty, he’s only broken one cup? And he’s broken about six . . . 9. J But his mum, she called him down for dinner. 10. L Yeah. 11. J And his mum never told him there was a table behind the door and he pushed it a little bit. ‘Cos he just pushed it open and she never told him that there was a table behind the door when she called him down to dinner did she? So I’ve got a point, ain’t I? But he, he knows where the cups are. He knew that them cups were in there but he just went up. 12. L Yeah, but I’ve got a point because he’s broken more cups. 13. J Don’t make no difference, his mum never told him. 14. L Right then, so it was him. 15. Experimenter enters. 16. L David. 17. J David. Where Nadia is a passive presence in her conversation with James, Leon is an active participant in his conversation with Jade. James, it seems, is oriented toward solving the problem, and Nadia is accorded only an instrumental recognition in his strategy. Since they have been asked to reach an agreed solution, he needs to secure her agreement, and shows little interest in her understanding. Jade, by contrast, from the opening of her conversation recognizes Leon as an active agent in the dialogue, seeking to secure agreement through persuasion. Although these conversations are very different, they can both be seen as proceeding from the same underlying representation of gender, one in which the dominant and authoritative position in tackling a problem and structuring a conversation is marked as masculine. James exploits the power this gives him, while Jade finds a discursive solution to escaping such gendered framing of the conversation. These two examples illustrate clearly the features of what Doise and his colleagues described through their hypothesis of sociocognitive conflict. Where Example 9.2 indicates a constructive resolution of the task, Example 9.1 corresponds to what they described as a relational resolution. Naturally one wants to know whether this difference in the form of the conversation has any consequences for the children involved. What do Nadia and Leon carry away from this interaction? Are there any traces of these different conversational experiences visible in their later moral thinking? Unfortunately the design of the study by

142


Leman and Duveen did not include any posttests, so that it is not possible to answer these questions about these children. However, Psaltis and Duveen (2006; 2007) in a study using the familiar conversation of the amount of liquid task used the same manipulation of gender and developmental level embedded in a design with both pretests, posttests, and control groups (see Table 9.2). In this study1 70 pairs of children (aged 6.5 to 7.5 years), distributed more or less evenly across the four pair-types, participated in the interaction phase, with a further 44 children serving as the control groups (28 boys, Cm, and 16 girls, Cf ). In the interaction phase of the study most of the conservers successfully persuaded their nonconserving partners to agree that the amounts of the water were the same (59 out of 70 pairs reached an agreement of equality, and the pairs reaching an agreement of inequality were randomly distributed across the pair-types), and as Table 9.3 shows, performance on the posttest showed clearly the benefits of the interaction for the nonconservers: 34 out of 70 nonconservers (49%) made progress on the posttest compared to only 7 out of 44 (16%) in the control groups. However, comparisons between each pair-type and the corresponding control group showed a clear effect only for the Fm group, with a marginal effect for the Mm group and no effects for either the Ff or Mf groups. As in the social-Genevan studies, the analysis of the posttests also investigated the use of novel arguments (that is arguments they had not heard during the interaction phase) as a stricter criterion of progress and a clearer index of cognitive reorganization. Again while the overall comparison is significant, there is considerable variation among the pair types. Nonconservers interacting with a female Table 9.2 Design of the Psaltis and Duveen (2006, 2007) study Pretest

Interaction

Delayed posttest

Four Pair types (Mm, Mf, Ff, Fm) Controls (Cm, Cf)

Controls (Cm, Cf)

Note: M=Boy, F=Girl; C=Control group. Capital letters refer to the autonomous child, lower case letters to the heteronomous child.

Table 9.3 Progress made from pretest to posttest by gender composition of the group Pair type Group

Mm

Mf

Ff

Fm

Cm

Cf

1 2

No progress Progress without novelty

10 6

11 7

8 4

7 6

24 n/a

13 n/a

3 4

Progress with novelty Global (total) progress

2 8

0 7

3 7

6 12

n/a 4

n/a 3

Note: M=Boy, F=Girl. Upper case letters refer to the autonomous child, lower case letters refer to the heteronomous child.

The constructive role of asymmetry in social interaction 143 conserver (i.e., groups Ff and Fm) made more use of novel arguments than nonconservers interacting with a male conserver (i.e., groups Mf and Mm).

Conversation type and developmental progress While these analyses indicate some relationship between the gender composition of the pairs and outcomes for the nonconservers, the relationship is clearly a complex one. In all pair-types there are nonconservers who make progress on the posttest as well as some who do not, so that no unequivocal relationship between gender composition and outcome can be established. Similarly, closer analyses of the details of the argumentative strategies employed also showed complex relationships between gender composition and outcomes. However, stronger and clearer relationships emerged when the data were analyzed at a molar level by taking the whole conversation as the unit of analysis. By avoiding the fragmented picture given by focusing on specific conversational features, this approach enabled both the sequential nature of the conversations as well as the modalities of resolving the conflict to be taken into account. A review of all the transcripts of the interactions revealed the existence of four distinct types of conversation. As well as considering the joint response given by each pair, the strategy for identifying conversational types also focused on the contributions of the nonconserving child, since the development of the interaction depended largely on their response to the conserving child’s assertion of equality. While the majority of interactions resulted in a conserving response as the joint answer, in a small minority of cases the pairs also came to agree on a nonconserving answer. In such instances the nonconserver persuaded the conserving partner, and such conversations were classified as Nonconserving type. The first type of conversation which resulted in an agreement on a conservation response could be described as having the form of direct conformity, since no resistance was present during the discussion, and such conversations were classified as No Resistance. Conversations where the nonconserver offered an argument in support of his or her position at least once during the interaction seemed qualitatively different from such direct compliance, so that conversations of this type were described as Resistance. In some discussions there were also indications from the nonconserver that agreement with the conserving position was the product of genuine progress. Such indications of understanding could be expressions of the a-ha moment such as “Oh, now I understand!” “I see, you are right!”; or moments where original nonconservers appropriated and used conservation arguments themselves, either elaborated arguments or unelaborated general assertions. A characteristic of these discussions was also that after the turn where the indication of understanding was given the original nonconserver offered no further resistance to the conserving argument. These types of conversations were described as Explicit Recognition. As Table 9.4 indicates, the relationship between conversation type and outcome for the nonconserver is both clearer and stronger than was the case for the gender

144


Table 9.4 Progress on the posttest by conversation type Conversation types

1 2

No progress Progress without novelty

3 4

Progress with novelty Global (total) Progress

Nonconserving

No Resistance

Resistance

Explicit recognition

11 0

7 6

16 9

2 8

0 0

1 7

2 11

8 16

combination of the pairs. Specifically, two of the conversation types have a clearly identifiable relationship with progress on the posttest. None of the participants in Nonconserving conversations made progress, while nearly all of the children who participated in Explicit Recognition did show progress on the post test. Further, almost half of the children from Explicit Recognition conversations also made use of novel arguments in their posttest (which accounted for nearly all of the novelty observed in the posttest), suggesting that this conversation type is particularly linked to cognitive reorganization. The other two conversation types, No Resistance and Resistance, have more mixed results, with only about half of these children showing progress on the posttest, and for those who do progress there is very little evidence of novelty in the posttest. While the relationship between conversation type and outcome is stronger than that between pair type and outcome, analysis also showed that conversation type was not independent of pair type. Indeed, instances of all four conversation types were observed in each of the gender combinations. The most distinctive pair types were Mf and Fm, with many more instances of No Resistance in Mf, and more Resistance and Explicit Recognition in the Fm. When the data were broken down by all three variables, conversation type, pair type, and outcome, the cell sizes were too small for statistical comparisons. However, some patterns were clearly visible. Where the original nonconserver made Progress with novelty on the posttest following an Explicit Recognition conversation, this was overwhelmingly in pairs in which there was a female conserver rather than a male conserver. And while progress is not exclusively restricted to Explicit Recognition, the children who made Progress without novelty following a No Resistance conversation were nearly all females who had participated in an Mf pair. The complexities of the relations between the gender composition of the pairs, the form of communication established between them, and outcomes illustrates the openness and unpredictability of interaction. While the composition of a particular pair type may influence the formation of a particular conversation type, it does not determine it. Further, while the establishment of a particular conversation type may influence the outcome, it also does not determine it. True, all of our nonconservers will eventually achieve conservation, but the process of reaching it

The constructive role of asymmetry in social interaction 145 could be very important in terms of the timing or the modality of knowledge obtained. The evidence here suggests differences in the stability of knowledge depending on the ways that it is legitimized through different conversation types. To understand how these conversations are structured in different ways so as to generate different opportunities for development it is necessary to examine in more detail the different ways in which children participate in each conversation type. In doing so an important focus for attention is on the ways in which children acquire access to the problem space and the different forms of social recognition which this implies. Nonconserving Characteristically in these conversations it is the nonconserver who initiates the discussion, and in doing so sets the frame for the conversation as a whole. Most of these discussions were quite short, and in many instances the conserving child did not support his or her original position even once. However, most of the conservers regained their conserving level in the posttest, suggesting that such discussions largely involved the establishment of a relation of social constraint between the two partners introduced by factors related to other sources of asymmetry such as gender or academic reputation (in fact, the majority of the nonconservers who won the argument were considered academically stronger students in comparison to their partners by their teachers, and girl conservers were twice as likely as boys to submit to the nonconserver’s view). A revealing symptom of the constraint prevailing in these conversations is that in some of them the nonconserver moved to force an ending to the conversation by underlining the need for agreement. In the following example of Nonconserving discussions that follow, this urgency for premature closure can be seen in turn 8. Example 9.3: Mf dyad Nonconserving 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

C This is bigger . . . my glass is bigger. N Yes but mine is smaller than yours. C The water is equal! N No. . . . This is more. C The water here is equal with yours because if we pour it in there it will be the same. N It isn’t. . . . This is small. C Yes but the glass is bigger . . . N That means that the water is more . . . Do you agree? . . . Do we agree? . . . (long pause) Experimenter What did you agree? N That his glass is bigger and he has more . . . Experimenter Do you both agree? C Yes.

146


No resistance Constraint is also characteristic of the No Resistance conversations, though this time it is exercised by the conserver. As can be seen in Example 9.3 they do not leave much space for the nonconserver to be given any kind of recognition, except a kind of instrumental recognition as the conserving child seeks to find agreement in order to meet the requirements of the task set by the experimenter. In this type of conversation it is nearly always the conservers who initiate the discussion and then in a condensed and consistent manner support their view. Notably in turn 3 in Example 9.4 the conserver frames the conversation in a highly structured way in which the nonconserver’s contribution is anticipated, as though they were simply being asked to fill in the conversational slots generated by the conserver. This type of conversation occurred most frequently between male conservers and female nonconservers (cf. Psaltis & Duveen, 2006) suggesting that gender asymmetry contributes to the establishment of this form of constraint. Example 9.4: Mf dyad No Resistance 1. 2. 3. 4. 5. 6. 7. 8.

C Wasn’t this (water in glass C) equal a while ago, when he poured it? N Yes. C This is a tall glass and that’s why they are equal should we call him? N Yes. C (opens the door and calls the experimenter back) Experimenter comes back to the room. What did you agree? N Equal. C Equal.

Interestingly, as Table 9.4 shows, half of the nonconservers who participated in No Resistance conversations nevertheless showed progress on the posttest. Overwhelmingly, however, these were nonconserving girls interacting with a conserving boy, suggesting that these conversations were dominated by expectations derived from social representations of gender. Yet, as the progress evidenced by these girls suggests, compliance is not always the same as passivity, though in these cases it was largely progress without novelty. Resistance Resistance is a less constrained conversation type than the previous two since by definition the nonconservers support their original nonconserving position. Again it is generally the conserver who sets the frame by initiating the discussion, but as the nonconserver now begins to defend their position the conversations not only become longer but also introduce a different form of participation for the children. What kind of conflict emerges in these conversations and how is it resolved? Generally in these conversations it is indeed a sociocognitive conflict (Doise & Mugny, 1984) that emerges in which different points of view are clearly expressed. In these conversations, the conserver also provides some clearer invitation for the

The constructive role of asymmetry in social interaction 147 nonconserver to reflect on the problem, indicating there is also a clearer recognition of the nonconserver as a thinking agent. Both of these features can be seen in Example 9.4, where the conserver can also be observed introducing more elaborated conserving arguments in turn 11 in response to the nonconserver’s repeated assertion of inequality, until the original nonconserver agrees on conservation. Example 9.5: Fm dyad Resistance 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

C They are equal. They are equal! I say they are equal, equal. N No they are not! C They are equal! N But yours is here. (points to the level of water) C Equal. N And mine [is C [equal N Here! (points to the level of the water in the transformation glass C) C But they are equal! Equal! We had both glasses and they were equal. N Mine is up to here and yours is up to here. (points to the level of water in glass C and glass A) C But they are the same! It’s just that the glass is bigger and it can take more, that’s why. N OK should we tell him? C Yes. N Come! Experimenter. What did you agree? N Equal. C Equal.

Explicit recognition This type of conversation has some of the same features which were evident in Resistance conversations, though as they develop the child who was originally a nonconserver comes to a clearer recognition of conservation. Again it is generally the conserver who establishes the frame, though interestingly in a small minority of cases the nonconserving child actually began the conversation by adopting a conserving position, presumably due to spontaneous progress between the pretest and interaction phases. In some conversations Explicit Recognition came after the nonconserving child had not only been exposed to a conserving argument, but had also resisted it at least once by supporting their original position, a type illustrated in Example 9.6. Example 9.6: Mm dyad Explicit Recognition 1. C What should we do? Should we say the same? That we have the same? 2. N No, I have more, you have less.

148 3. 4. 5. 6. 7. 8. 9. 10. 11.

Development as decentration C But, this is taller and that is wider . . . it’s the same. . . . N It’s the same thing. . . . C No . . . this is short, it’s small. . . . N That is small, but is wider. . . . If we put it in here it will be the same! That means it’s the same! C Should we say that it’s the same? N OK. Experimenter. Did you agree? C Yes, it’s the same water. N Yes.

In this example, questioning by the conserver invites the nonconserver to take responsibility for the task and provides an opportunity for reflection. The initial resistance of the nonconserver fades as the conserver restates their argument. By turn 6 the nonconserver has taken advantage of the opportunity for reflection by providing a clear argument for conservation, leading to agreement with their partner. There were also conversations where explicit recognition came after the nonconserver was exposed to a conservation argument by their partner without any explicit resistance by the nonconserver. In these instances the conserver generally adopts a more didactic strategy, as in Example 9.7, where the turn by turn construction of the joint solution is mostly controlled by the conserver. The conservation argument is skilfully built. The nonconserver contributes to the building of the argument by answering the questions of the conserver who asks questions that draw the nonconserver into reflection. Here there is clearly recognition of N as a thinking subject by C, who is looking for recognition of their own arguments. In turn 8 N explicitly recognizes conservation (it might be argued that this is beginning to happen from turn 4). Example 9.7: Fm dyad Explicit recognition 1. C This glass (glass C) had the same water with that one (glass B) right? 2. N Yes. 3. C This glass (glass B) had equal water with that one (glass A) right? As we have said here, it’s the same when we poured it in here (glass C) should we have more? 4. N No . . . so the glass is bigger. 5. C Yes, only the glass is bigger. 6. N Yes. 7. C Therefore is it more in here or are they equal? 8. N Equal, it looks more because the glass is bigger. 9. C Yes! 10. N OK 11. C Come! (calls the experimenter) 12. Experimenter. What did you agree?

The constructive role of asymmetry in social interaction 149 13. N Equal. 14. C Equal. 15. N It’s just that her glass (glass C) is bigger. As these illustrations indicate, the conversation types vary in the forms of participation established between the children. As always in conversation, the partner who initiates the discussion exercises some measure of control by setting the frame for the exchange. But there are clearly different ways of establishing control. In the Nonconserving and No Resistance conversations this control is exercised in a way that limits the access of the other child to the problem space, so that in doing so the child who takes the dominant position extends only an instrumental recognition of the partner. In both of these conversation types the dominant child frames the conversation by aligning the initial asymmetries of knowledge between the children with other asymmetrical dimensions of the situation stemming from representations of gender or of academic ability so that the conversations as a whole remain bounded by the expectations derived from these representations. While this strategy is sufficient to induce compliance from their partner, these conversations are not associated with any persistent change in the understanding of the dominated partner. Where there is Resistance, the conserving child has to find a strategy for persuading his or her partner, which usually involves some form of sustained argument for conservation. Here the problem space of the task becomes more accessible to the nonconserving child, who is able to take on a measure of responsibility for its solution and thus establish a clearer recognition for themselves as an autonomous agent or thinking subject. Lastly, in Explicit Recognition there is a clearer sense of the problem space being accessible to both children, and the conserver is more clearly oriented to a form of persuasion in which the nonconserver is stimulated to reflect on their understanding of the task. In both of these conversation types, the initial asymmetries of knowledge take on a more positive or constructive role in the conversation, which consequently can sometimes move outside the limits associated with representations of gender or academic ability. As it does so, a different kind of social relation becomes established between the children in which both are recognized as thinking subjects contributing to the solution of the task. This is clearest in the Explicit Recognition conversations, where the nonconserver’s reflection on their understanding is highlighted, resulting not only in greater evidence of developmental progress but also in clearer evidence of cognitive reorganization in the extent of novelty in their posttest discussions.

Conclusion: from microgenesis to ontogenesis In returning to the vexed question of the relations between social interaction and cognitive development we have argued for the benefits of extending the analysis of the details of how conversations take shape and form for understanding the ways in which they may also structure the development of children. In this sense,

150


development can be seen to proceed from the microgenesis of the emergence of shared representations of the task to the ontogenesis of cognition. At the center of this process through which social interaction influences and structures development are the different forms of participation through which children engage in the conversations. Where the child becomes an active participant in addressing and solving the task, then he or she also enters into (joint) ownership of the problem space of the task. A balanced triadic structure of child-child-task is established in which not only can different viewpoints be expressed, but they can also be coordinated in a way in which each child is recognized as an active contributor to achieving the solution. In short, in these circumstances, children engage with their peer and the task as active agents, responsible for their own perspective and responsive to the perspective of the other. Participation in this sense is the cooperative moment at the heart of children’s interactions, and if it is not always necessary for participation to occur for children to demonstrate progress on the posttest (cf. the Mf girls who progress), then it does seem to be the case that only participation in this sense can lead to the interiorization which enables the child to produce novelty in the posttest. Participation, though, is not the inevitable consequence of children interacting, but rather something which emerges from the particular type of conversation which is established between them as they address the task. It is in shaping the type of conversation that asymmetries between the children can be seen to have a significant role. Where the asymmetry is given a negative expression, the conversation remains bound by the different expectations associated with the asymmetry itself. Where the asymmetry becomes a more positive force it is because the contrasts, conflicts, and tensions it establishes lead to the conversation escaping from the limits set by prior expectations, thus opening the prospect of both children participating in the solution to the task. Consequently we also need to revise Piaget’s distinction between constraint and cooperation. Rather than being construed as independent states of affairs, they need to be seen as social relations established through the processes of interaction. Indeed, the achievement of a cooperative relation may be just a moment within the flux of interaction, but it is a moment in which asymmetric relations between children can take on a constructive role leading to the emergence of more symmetric relations. As with the “social Genevans,” our approach to the study of social interaction and cognitive development is grounded in a social constructionist perspective, with its assumption that psychological activity is always the product of communication between different voices (cf. Marková, 2003). This perspective differs from Piaget’s own constructionism precisely in relation to this assumption. In his brief comments on the first studies by Doise and his colleagues, Piaget (1976) himself raised this question by asking whether this research concerned the source of operations or their structure. Concretely, the question becomes one of asking whether social interaction is not only the context for disequilibration but also provides the resources for the emergence of new forms of structure (that is for reequilibration). This, however, is not a straightforward or simple empirical question, in the sense that what can be observed might be uniquely interpretable from

The constructive role of asymmetry in social interaction 151 one theoretical perspective or the other. It is only through the theoretical work of interpretation that the constitutive or causal factors in development appear intelligible in relation to observable phenomena. In Piaget’s own work, his analyses of equilibration are likewise interpretations of observable phenomena. Our emphasis on the ways in which different types of conversation generate different outcomes for the participants is a step toward a broader effort to establish a social constructionist perspective on intellectual development. As we have suggested earlier, our analyses of the relations between social interaction and cognitive development are derived from a triadic epistemology framed as the interdependence between the child, the other, and the object. Others (for example, Chapman, 1991; Schubauer-Leoni & Perret-Clermont, 1997) have also emphasized the need for such a triadic epistemology, and also remarked on the way this moves beyond the classical dyadic epistemological frame of Piagetian theory. Working through the consequences of adapting such an epistemology without simply abandoning Piaget’s genetic epistemology is a challenge which still needs to be met.

Notes 1 Details of the statistical analyses of these data can be found in Psaltis and Duveen (2006). 2 Interrater agreement for the four conversation types was excellent (Cohen’s Kappa=0.82, p < 0.001). 3 Details of this qualitative analysis are in Psaltis and Duveen (2007). In the transcripts, C indicates the conserving child, N the nonconserving child. This study was undertaken in Cyprus, and the transcripts are translated from the original Greek-Cypriot dialect.

References Ames, G. J., & Murray, F. B. (1982). When two wrongs make a right: Promoting cognitive change by social conflict. Developmental Psychology, 18, 894–897. Bearison, D., Magzamen, S., & Filardo, E. K. (1986). Socio-cognitive conflict and cognitive growth in young children. Merrill-Palmer Quarterly, 32, 51–72. Butera, F., & Mugny, G. (Eds.) (2001). Social influence in social reality: Promoting individual and social change. Ashland, OH: Hogrefe and Huber. Carpendale, J., & Müller, U. (Eds.) (2004). Social interaction and the development of knowledge. Mahwah, NJ: Lawrence Erlbaum. Chapman, M. (1991). The epistemic triangle. Operative and communicative components of cognitive development. In M. Chandler & M. Chapman (Eds.), Criteria for competence: Controversies in the conceptualisation and assessment of children’s abilities (pp. 209–228). Hillsdale, NJ: Lawrence Erlbaum. Doise, W., & Mugny, G. (1984). The social development of the intellect. Oxford: Pergamon. Doise, W., Mugny, G., & Pérez, J. A. (1998). The social construction of knowledge, social marking and sociocognitive conflict. In U. Flick (Ed.), The psychology of the social (pp. 77–90). Cambridge: Cambridge University Press. Doise, W., Mugny, G., & Perret-Clermont, A. N. (1975). Social interaction and the development of cognitive operations. European Journal of Social Psychology, 5, 367–383.

152


Duveen, G. (1997). Psychological development as a social process. In L. Smith, P. Tomlinson, & J. Dockerell (Eds.), Piaget, Vygotsky and beyond (pp. 67–90). London: Routledge. Duveen, G. (2000a). The power of ideas. Introduction to S. Moscovici, G. Duveen (Ed.), Social representations: Explorations in social psychology (pp. 1–17). Cambridge: Polity Press. Duveen, G. (2000b). Piaget ethnographer. Social Science Information, 39, 79–97. Duveen, G. (2001). Representations, identity, resistance. In K. Deaux & G. Philogeˋ ne (Eds.), Representations of the social (pp. 257–270). Oxford: Blackwell. Emler, N., & Glachan, M. (1985). Apprentissage et developpement cognitif. In G. Mugny (Ed.), Psychologie sociale du developpement cognitif (pp. 71–92). Bern: Peter Lang. Leman, P. J., & Duveen, G. (1999). Representations of authority and children’s moral reasoning. European Journal of Social Psychology, 29, 557–575. Lloyd, B., & Duveen, G. (1992). Gender identities and education. London: Harvester Wheatsheaf. Marková, I. (2003). Dialogicality and social representations: The dynamics of mind. Cambridge: Cambridge University Press. Moscovici, S. (1972). Society and theory in social psychology. In J. Israel & H. Tajfel (Eds.), The context of social psychology (pp. 17–68). London: Academic Press. Moscovici, S. (1976). Social influence and social change. London: Academic Press. Moscovici, S. (1980). Towards a theory of conversion behaviour. In L. Berkowitz (Ed.), Advances in experimental psychology (Vol. 13, pp. 209–239). London: Academic Press. Mugny, G., De Paolis, P., & Carugati, F. (1984). Social regulations in cognitive development. In W. Doise & A. Palmonari (Eds.), Social interaction in individual development (pp. 127–146). Cambridge: Cambridge University Press & Maison des Sciences de l’Homme. Perret-Clermont, A.-N. (1980). Social interaction and cognitive development in children. London: Academic Press. Perret-Clermont, A.-N. (1994). Articuler l’individuel et le collectif. New Review of Social Psychology, 3, 94–102. Perret-Clermont, A.-N., Perret, J. A., & Bell, N. (1991). The social construction of meaning and cognitive activity in elementary school children. In L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.), Perspectives on socially shared cognition (pp. 41–62). Washington, D.C.: American Psychological Association. Perret-Clermont, A.-N., & Schubauer-Leoni, M.-L. (1981). Conflict and cooperation as opportunities for learning. In P. Robinson (Ed.), Communication in development. London: Academic Press. Piaget, J. (1932). The moral judgement of the child. London: Routledge & Kegan Paul. Piaget, J. (1962). Comments on Vygotsky’s critical remarks concerning The Language and Thought of the Child and Judgement and Reasoning in the Child. (Anne Parsons, Trans.). Cambridge, MA: MIT Press. Piaget, J. (1976). Postface. Archives de psychologie, 44, 223–228. Piaget J. (1995). Sociological studies. London: Routledge. (Original work published 1977) Piaget, J., & Inhelder, B. (1969). The psychology of the child. London: Routledge & Kegan Paul. Psaltis, C. (2005a). Communication and the construction of knowledge or transmission of belief: The role of conversation type, behavioural style and social recognition. Studies in Communication Science, 5, 209–228.

The constructive role of asymmetry in social interaction 153 Psaltis, C. (2005b). Social relations and cognitive development: The influence of conversation types and representations of gender. Unpublished doctoral dissertation, University of Cambridge. Psaltis, C., & Duveen, G. (2006). Social relations and cognitive development: The influence of conversation type and representations of gender. European Journal of Social Psychology, 36, 407–430. Psaltis, C., & Duveen, G. (2007). Conservation and conversation types: Forms of recognition and cognitive development. British Journal of Developmental Psychology. Russell, J. (1982). Cognitive conflict, transmission, and justification: Conservation attainment through dyadic interaction. Journal of Genetic Psychology, 140, 283–297. Russell, J., Mills, I., & Reiff-Musgrove, P. (1990). The role of symmetrical and asymmetrical social conflict in cognitive change. Journal of Experimental Child Psychology, 49, 58–78. Schubauer-Leoni, M.-L., & Grossen, M. (1993). Negotiating the meaning of questions in didactic and experimental contracts. European Journal of Psychology of Education, 8, 451–471. Schubauer-Leoni, M.-L., & Perret-Clermont, A.-N. (1997). Social interaction and mathematics learning. In P. Bryant & T. Nunes (Eds.), Learning and teaching mathematics. An international perspective (pp. 265–283). Hove: Psychology Press. Tudge, J. (1992). Processes and consequences of peer collaboration: A Vygotskian analysis. Child Development, 63, 1364–1379. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.

Part III


10 The significance of social identities Gerard Duveen and Barbara Lloyd

Common-sense meanings of the ‘individual’ and the ‘social’ are examined emphasizing the fallacy of viewing them as denoting separate entities given in nature. The concept of social representations is used to explain the construction of these categories in terms of the particular systems of values, ideas and practices of different societies. The concept of social identity marks the individual–social interface interpreted as the construction of individuals in relation to the social representations of significant groups in their society. The earliest constructions of a social identity are traced in the domain of gender because this is a ubiquitous feature of social life, involving the differentiation of only two groups: they are obligatory and use physical differences to provide the signifiers in a semiotic system of social representations. Empirical evidence is reviewed showing that the development of a social gender identity is a complex process and involves many aspects of children’s activity slowly coming to be regulated by the particular social representation of gender dominant in their society.

1 The ‘individual’ and ‘social’ as social representations The term ‘individual–society interface’ readily conjures up the image of two objects in some kind of relationship, with the theoretical problem then being to specify the nature of that relationship. In this image ‘individual’ and ‘society’ are taken as given, as objects existing in nature. We want to argue that this image reflects a particular representation of the issue, and moreover, one which introduces a category error in so far as it identifies as facts of nature what are, in reality, human constructions. As socio-psychological categories ‘individual’ and ‘society’ are not immutable givens. The representation of them as lying on each side of a fixed boundary is only one of many, for the spheres of individual and social existence are marked differently in different cultures (cf. Dumont, 1970) and within our own culture the representations of ‘individual’ and ‘social’ have undergone change over time (cf. Foucault, 1970; Morris, 1972). The argument presented in this paper rests on the premise that individuals are so inextricably interwoven in a fabric of social relations within which their lives are lived that a representation of the ‘individual’ divorced from the ‘social’ is theoretically inadequate. There is no pure ‘individuality’ which can be apprehended independently of social relations. The complex interrelations of the individual and

158


the social mean that, in effect, an individual is inconceivable as a viable entity without a sustaining network of social relations. It is as useless to attempt to separate these two terms as it is to attempt to isolate environmental and hereditary influences on the expression of intelligence (Furth, 1973). Although the terms ‘individual’ and ‘social’ have distinct common-sense meanings this does not vouch for their reality. These categories are important for sociopsychological analysis nonetheless because they constitute a central frame of meaning in the construction of representations of the world. Theoretical confusion results when the constructed and constructive nature of the categories ‘individual’ and ‘social’ is ignored. They may then unwittingly be viewed as though they were given in nature. A recognition that they are constructed categories reduces the risk of invoking common-sense meanings when the terms are employed in more rigorous or scientific discourse. Our aim is to question some of the explicit and implicit meanings of the terms. Problems of meaning would be more simply resolved if we could claim that social psychologists, as objective scientists, were able to reach beyond society’s representations of categories. But it is precisely because we are unable to take up any position of absolute objectivity that we need to examine our use of the terms ‘individual’ and ‘society’ with care. How, then, have common-sense representations influenced the scientific construction of the categories ‘individual’ and ‘social’? The influence of common sense is most apparent when meanings are assumed to be already sufficiently clear that they can be borrowed from common-sense discourse without any reflexive effort at clarification. In contemporary Western societies the common-sense understanding of ‘individual’ implies uniqueness and stands in opposition to the ‘social’. Quantitative psychology describes the ‘individual’ by relating persons to social groups; the individual is measured in relation to a comparison group. The individualities which common sense and quantitative psychology address are very different. Uniqueness is not addressed through conventional methodological procedures. Barratt (1984) has recently presented a strong case for the argument that the individual only becomes visible through a psychoanalytic discourse, and that individual psychic reality remains inaccessible to the procedures of quantitative psychological research. In these terms the ‘individual’ is represented as a collection of private and unique meanings. The issue here turns on the problem of rendering human actions intelligible. Psychoanalysis developed to deal with the apparent opacity of such activities as neurotic symptoms, everyday slips and dreams. As an investigation it operates through the negotiation of private meanings so that the unintelligible is rendered comprehensible. Quantitative psychology, on the other hand, is limited insofar as the meaning systems through which it interprets the actions of individuals are themselves public systems of social representations. In these terms an action is intelligible when comprehended within a social rule system. To make human actions intelligible in this way provides a legitimate level of conceptual analysis but one which is also limited. These limits are rarely acknowledged and the function of social representations in quantitative psychology generally remains implicit (Herzlich, 1972).

The significance of social identities 159 We do not wish to follow Barratt (1984) when he goes on to argue that psychoanalytic investigation of individual psychic realities is the only valid form of psychological research. On the contrary we cite his argument to clarify what is accessible and what must remain inaccessible to quantitative psychological research. The aspect of the individual which he has identified accords with the common-sense notion of individual as unique. To impute uniqueness to the individual in social psychological discourse is to commit what Gilbert Ryle has called a category mistake (1949). The mistake arises through a failure to recognize the influence of social representations in constructing the social psychological category of the individual. The distinction between the individual as unique and the individual construed within a social framework alerts us to the importance of social representations in analyses of social psychological aspects of individuals. The individual defined in relation to the group can also be described as specified in terms of the individual– social interface. It is not easy to maintain a definition of the individual in terms of the group; and it is all too easy to slip back into a view of the individual as a private construction. We employ a particular concept of social identity to help us maintain an awareness of the individual as constructed in terms of the group. According to our usage social identities reflect individuals’ efforts to situate themselves in their societies in relation to the social representations of their societies. Before exploring further any particular social identities we need to examine more closely the concept of social representations in which they are grounded.

2 Social representations and social identities The first point to consider is that social representations are always the representation of something (Moscovici, 1976a). The evaluative dimension of social representations can only be understood through a consideration of the content which is organized in them. This constraint precludes an empty and formal analysis of social representations, though it may be necessary to develop a classification which takes account of differences in content. The relationship between the ‘individual’ and ‘society’ is constantly mediated by the particular context in which it is situated. Thus we can say that the ‘individual–society interface’ does not actually exist. Individuals are related to society through their participation in social groups defined by gender, age, social class, etc., and within the social representations of these systems particular ‘individual–society interfaces’ are defined. Therefore, it is necessary to investigate the relationship in particular contexts without presupposing that it will have the same structure and boundaries in every instance. Our second point is that social representations are features of social groups. They are to be conceptualized at the sui generis level of the social described by Durkheim. Social representations form organized systems of ‘values, ideas and practices’ (Moscovici, 1973, p. xiii) which pre-exist the birth of any individual, and while they form the context in which an individual life is lived they also

160


persist after an individual’s death. This is not to deny the historical, and hence changeable, character of social representations; it is, rather, to insist on their relative resistence to individual modification. Children are born into a particular culture and to become competent, functioning members of that culture must re-construct for themselves the categories of that culture.1 Construed in these terms social representations may usefully be invoked in explaining the development of children’s understanding of the adult community into which they are born. Social representations regulate the child’s construction of reality, just as in Piaget’s analysis of cognitive development it is the closed structures of logical systems which regulate the child’s (re-)construction of logicomathematical structures. Though this view of social representations clearly raises important developmental issues these have only recently been addressed (cf. Moscovici & Hewstone, 1983; Duveen, 1984). In the developmental re-construction of the categories of a culture not all categories are equally significant for the child. The particular significance of certain categories is embedded in the social interactions of everyday life. Categories associated with dimensions such as age, gender and social class exert a great influence on the processes of interpersonal interaction, and this endows them with a high degree of salience. The effects of the social marking of these category systems is to offer considerable ‘scaffolding’ to the young child’s efforts at re-construction. We expect that the more salient and socially marked category systems will be among the earliest to be re-constructed by young children. The social representations which form particularly significant elements in children’s cultural environments are those which confer on them certain positions in the social system. In our culture one cannot be just a human being, one must be male or female, young or old. Membership in particular social categories provides individuals with both a social location and a value relative to other socially categorized individuals. These are among the basic prerequisites for participation in social life, and can be described as social identities.

3 Social gender identity In order to elaborate this schematic outline further it is necessary to locate the argument within some particular system of social representations. In the remainder of this paper these issues are considered in the context of the gender system (cf. Archer & Lloyd, 1985; Connell, 1985). There are a number of reasons why gender is a particularly felicitous context within which to pursue a discussion of social identity. Although ‘male’ and ‘female’ are terms applied to individuals, to be male or to be female also establishes a social identity for the individual as a member of a gender group. The social categories of ‘male’ and ‘female’ constitute a gender system which exhibits a number of characteristics relevant to a discussion of social identity: 1

Gender is a ubiquitous dimension of social organization, its influence is apparent in every social encounter. Indeed gender is such an essential dimension that it

The significance of social identities 161

2 3

4

provides a framework for the organization of many other aspects of social life, which appear as ‘marked’ for gender. The division of people into two gender groups establishes an exhaustive binary social categorization in most societies and as such provides the simplest possible form of category differentiation. Membership in a gender group is obligatory and initially assigned by the genital characteristics present at birth. (Such assignations are held to be of sufficient importance that in cases of ambiguous genital characteristics a decision is taken to assign the child to a gender group and to ‘complete’ the work of nature through surgical intervention.) Universal physical sex differences are given particular gender meanings by different societies. The elaboration of systematic gender categorization is a product of social life employing biological sex differences as signifiers in a semiotic system in which social representations of gender are the signified.

For all these reasons social representations of gender are apposite to any discussion of social identity. We interpret the development of gender knowledge in children as the elaboration of a social gender identity by the child; that is, children must situate themselves, take cognizance of themselves, in relation to social representations of gender. The development of social representations of gender affords the child access to a centrally important and significant explanatory frame of reference. This system provides children with the means of interpreting actions and events according to the cultural rule system in which they live, as well as allowing them to participate in their culture by acting, thinking and feeling in ways that are comprehensible to themselves and others. In short, social representations of gender provide an important shared frame of reference which makes possible the exchange of differentiated signifiers between people; social representations of gender correspond to a semiotic code which allows children to participate in social life by providing an explanatory framework for the interpretation of their own and others’ actions, thoughts and feelings. In speaking of social gender identity we want to emphasize that this refers to the sense in which people recognize their membership in one of the two gender groups. It does not refer to all the ways in which persons construct their maleness or femaleness, that is the way ‘being male’ or ‘being female’ figures in a person’s lived experience. While social representations of gender and social gender identity are amenable to quantitative socio-psychological investigation, there are aspects of a person’s lived experience of gender which lie outside this perspective, and may properly fall within the domain of psychoanalytic inquiry. The systematic marking of social relations for gender, and the salience which this distinction has in social life, ensures that it becomes highly ‘visible’ or significant for the child. One may speak of social representations of gender as being instantiated in these systematic markings. Mugny et al. (1984) characterize social marking by saying that it ‘connects relations of a cognitive order with those of a social order’ (p. 137). This is indeed what happens very early in the child’s life.

162


Gender is one of the earliest conceptual systems to develop in children’s understanding of the world. Social representations are complex systems of values, ideas and practices which both establish a particular conception of the world and make possible communication about it. If research into children’s re-constructions of gender representation is to be adequate it must be based upon an explicit codification of the adult system. It also needs to take account of the changing capacities of children as they develop. In designing empirical research it becomes important to undertake developmentally appropriate assessments in evaluating the significance of adult gender representations for children’s re-constructive activities. The development of social gender identity cannot be isolated as a single event marked by a particular milestone, nor can it be indexed by a single observation or experimental procedure. From the arguments already presented it follows that any type of human activity for which gender is a relevant consideration may serve as a medium for the investigation of social representations of gender. For the development of social gender identity this leads to the general hypothesis that in the course of children’s development their activity will progressively come to be regulated by the social representations of gender dominant in their cultural environment. In the following section we report on our own investigations which take up issues concerning the development of social gender identity. These include studies of gender marking in infancy, symbolic gender understanding in preschool children and their representations of economic life.

4 The development of social gender identity 1. Gender-marked infancy. Social gender identity is derived from the social symbol system in which gender is expressed as a set of social representations. In a strict sense the empirical assessment of social gender identity is dependent upon the child’s capacity for representational activity. Developmentally, the origin of social gender identity is to be sought not only in the adult symbol system but also in the emergence of a distinction between self and other in the infant. The emergence of this distinction in the first year of life, together with adults’ marking of gender through action must be regarded as the precursors of a symbolic social gender identity. It is apposite to examine the development of the self–other distinction and to focus on infants’ earliest recognition of a representational self and other (Lewis & Brooks, 1975; Lewis & Rosenblum, 1975). Studies using mirror techniques have shown that by the second half of the second year infants recognize their own faces and have the beginnings of a representational concept of the self. Earlier in the second year they are able to apply names correctly to pictures of familiar others. Stern (1983) and Emde (1983) have argued that the origins of the self/other distinction can be found in developments in the first year. The origins of a social gender identity can also be traced to experiences of socially defined gender in the first year.

The significance of social identities 163 Studies of mothers of first-born six-month-olds (e.g. Smith & Lloyd, 1978) have demonstrated that gender influences mothers’ handling of, and speech to, six-month-olds. When asked to play for six minutes with an infant presented as ‘John’ or ‘Jane’, these mothers chose the initial toy to match the baby’s gender. When babies made gross motor movements mothers responded with further gross motor behaviour if the baby had been presented as a boy. However, if the baby had been presented as a girl mothers attempted to calm the baby. The mother’s accompanying speech either emphasized physical skills, or, in the case of girls, cleverness and attractiveness. At 13 months the play behaviour of infants offers evidence of gender marking (Goldberg & Lewis, 1969; Smith, l982). Goldberg & Lewis have shown that 13-month-old boys preferred toys with wheels while girls chose toys with faces. Although Smith was unable to replicate gender differences in toy choice among 13-month-olds playing on their own, she showed that while playing with their mothers boys made more gross motor responses, pushing, pulling and banging the toys with which they were playing. Thus, by the beginning of the second year children’s play is marked according to adult representations of gender although their activity is of a sensorimotor nature. 2. Toy choice. Our studies of children aged 19 to 42 months provide clearer evidence of the influence of the adult gender system on toy choice (Lloyd & Smith, 1985; Lloyd, 1986). Using toys which were rated for gender appropriateness by parents it was shown that boys spend more time playing with masculine toys and that girls spend more time in action play with feminine toys. Boys’ use of masculine toys increased with age, but girls’ use of masculine toys showed no systematic variation with age. There were no significant effects of age for either boys’ or girls’ use of feminine toys. We interpret this systematic patterning of children’s activity with toys as evidence that their toy use is regulated by a practical appreciation of adults’ gender marking of these objects. These results derive from observations of children playing with familiar peers, either in same-gender or mixed-gender pairs. The differences in durations of play with masculine and feminine toys are modulated by the gender of the child’s partner. Durations of action play with own-gender toys were lower in boy–girl pairs. Girls in boy–girl pairs played less with feminine toys and more with masculine toys than girls in girl–girl pairs while boys in boy–girl pairs played less with masculine toys and more with feminine toys. In other words the regulation of young children’s practical activity by social representations of gender is susceptible to social influences deriving from the gender of their partner. Two children of the same gender are able collectively to construct a consistent social gender identity as shown in their use of gender-marked toys. The inconsistent use of gender-marked toys in boy–girl pairs suggests that these young children find difficulty in maintaining their social gender identity without the support of a same-gender peer.2 In same-gender pairs the social gender identity of self and other are congruent and children in these pairs are not required to shift their point of view when

164


selecting a toy for use. In boy–girl pairs the perspectives of the self and the other conflict and difficulties in shifting viewpoints may be observed in compromises in toy use which result in less clear differentiation. Young children have difficulty in simultaneously coordinating two socially marked dimensions. These considerations suggest that in boy–girl pairs of older children whose capacities for simultaneous coordination are more fully developed there may be clearer differentiation of toy use. Although the settings are not directly comparable Serbin et al. (1979) have reported results suggestive of such a change; they found that there is less cross-gender toy choice in the presence of a cross-gender peer in children of three and a half to five years. 3. Social behaviour. The social behaviour of the 19- to 42-month-old children we observed also showed the function of the adult gender system in regulating social action (Lloyd & Smith, 1986). Studies of gender stereotypes in adults have consistently reported that men are judged to be more aggressive, ambitious, assertive, dominant and forceful than are women while women are described as gentle and understanding (cf. Archer & Lloyd, 1985, pp. 38–47). In our oldest age group boys displayed more assertive behaviour than girls and only in this age group did girls show more withdrawal than boys. Examination of sequences of social behaviour revealed that boys’ attempts to be assertive were significantly more successful than those of girls. These results show that young children’s interpersonal behaviour is progressively regulated by social representations of gender. Assertion on the part of boys comes to be recognized as a legitimate expression of a masculine social gender identity while girls’ assertive bids, though more frequent than those of boys, are less successful. The same behaviour, assertion, is given a different interpretation according to whether it is the action of boys or of girls. These interpretations are systematic, and applied by children of both genders. We believe that it is their access to the semiotic system of gender representations which allows children to impose this interpretive scheme on the events they observe. 4. Categorization. The symbolic capacities of children are also changing in the period between 19 and 42 months. We tested the children whom we observed at play on four sorting tasks (Lloyd, 1986). They were asked to place in separate piles six red and blue geometric shapes, photographs of six cats and dogs, of six boys and girls, and six men and women as well as photographs of the masculine and feminine toys they would use in the play sessions. Predictably performance improved with age reflecting the changing capacities of children in their second and third years. Children of all ages found it easier to sort for colour and species (cats and dogs) than to sort people or toys according to gender. Of the 30 children aged 19 to 24 months only one was totally successful in sorting colour and one in sorting species but no one managed to sort people by gender. Forty per cent of children aged 25 to 30 months sorted colour successfully, 37 per cent species and 17 per cent people by gender. Among children approaching three years of age colour was successfully sorted by 77 per cent,

The significance of social identities 165 species by 80 per cent but gender by only 47 per cent. Sixty-three per cent of children aged 37 to 42 months sorted people correctly by gender: 83 per cent were successful with colour and 90 per cent with species. Children’s ability to sort toys according to gender improved with age but girls were more successful than boys. Boys were better when sorting masculine toys but girls sorted with equal facility masculine and feminine toys; though more successful than boys, girls in the oldest group scored barely 60 per cent correct. In other words, these young children demonstrate that although they are capable of sorting a collection of things into binary categories, the nature of the task, whether it has a perceptual possibly physiological basis, as in the sorting of colours, or requires discrimination at a basic object level, as in separating cats and dogs, affects the ease with which the categorization is made. Sorting people for gender requires the conceptual representation of gender, while sorting toys according to gender involves applying this conceptual distinction to objects which provide no perceptual scaffolding. Both gender tasks require capacities not yet well developed in two- and three-year-olds. 5. Linguistic development. The linguistic development of young children is of particular interest, since not only does performance on linguistic tasks depend on the emergence of symbolic capacities, but gender marking is itself also a significant linguistic feature. The same group of 19- to 42-month-old children were also tested for their ability to recognize and produce gender-marked nouns and pronouns. Recognition was tested first by presenting two photographs and asking the child a question in the form ‘Show me the MAN’ or ‘Which one is HE?’ Even if the child failed to choose the photograph of a man, the next question focused on the alternative picture, that of a woman, and the child was asked ‘Who is it?’ and expected to produce the reply LADY or SHE. The gender-marked nouns used in these tasks were MUMMY, DADDY, LADY, MAN, GIRL and BOY. The pronouns were SHE, HE, HER and HIM. Performance on these tasks showed two effects familiar from studies of language development. As expected there were improvements on all tasks with age, and also, production tasks were more difficult at every age than recognition tasks. Gender-marked noun recognition was near perfect by 30 months. Even the youngest children were correct 60 per cent of the time. Pronoun recognition was more difficult with only 15 per cent correct in the children approaching two years and was correct only 91 per cent of the time in the oldest children. Performance in self-recognition, i.e. applying the correct label, GIRL or BOY to their own photograph, was similar to that for pronoun recognition. The difficulty which younger children experienced in applying the gendermarked nouns BOY and GIRL to their own photographs may be related to a lack of familiarity with these terms. Recognition scores for younger children reached 49 and 73 per cent respectively for the terms BOY and GIRL but scores for the adult terms, MAN and LADY, were significantly higher. The terms BOY and GIRL may be considered appropriate in labelling the other (the observer) while the child’s own proper name may be considered the correct label for the self

166


(the actor). Here again, the need to function both as actor and observer may inhibit the child’s performance. In each age group scores for production of gender-marked nouns and pronouns and of common nouns were lower than scores on the corresponding recognition task. Having passed the recognition test by pointing to the picture of a woman when asked to ‘Show me MUMMY’ the child might produce the term DADDY, to the photograph of the man, but many other terms are possible, among them MAN. One source of the discrepancy in performance on recognition and production tasks may be the application of very stringent criteria of correctness to children’s verbal responses (cf. Bloom, 1974). Young children had great difficulty producing complements for the gender-marked nouns which they were generally able to recognize. The majority of errors resulted from a failure to meet the specific demands of the test situation, i.e. MAN was produced rather than DADDY. Gender confusion accounted for only 2.4 per cent of errors in the production of gender-marked nouns, and age confusion 3.5 per cent. Errors can be informative and in a further small study children of two to three years were given recognition tasks in which errors were inevitable. They were shown pairs of photographs, each pair including one male and one female picture. Mismatches were likely since for the terms BOY and GIRL the photographs presented were of adults while for MAN and LADY they were of children. Children of two to two and a half years responded with little hesitation, selecting the appropriate gender while apparently ignoring the age discrepancy. The older children often failed to make a choice and replied saying ‘There isn’t one!’. Two of the older children who responded willingly appeared deliberately to suspend judgement about age. Gender-marked nouns may simultaneously encode other socially marked distinctions in adult use of language. Children may use the same terms but limitations in their capacity to coordinate simultaneous variations in two dimensions may produce characteristic error patterns. In this task the younger children fail to appreciate the contradiction while the further development of symbolic functions in older children allows them to recognize the contradiction. Nevertheless, the youngest children show an awareness of gender marking even though they may not be able to deal with the other demand characteristics of this task. As with the original study, it appears that whatever difficulties young children encounter in linguistic tasks these are of a cognitive–developmental order. Representations of gender are sufficiently well established to enable young children to deal with the encoding of gender in language. 6. Representations of economic life. Gender representations may function for the child as operative schemes in the manner identified by Piaget. They may serve as instruments of knowing which allow the child to assimilate areas of social life which lie beyond their immediate field of action. Besides suggesting that young children’s representations of social life are marked for gender, our studies have also noted the differential effects of the social gender identities established by boys and girls. Where social representations of gender assign different

The significance of social identities 167 evaluations to male and female activities, differences also emerge in the developing representations of young children when they re-construct these social representations. The differentiation process can be observed in the influence of gender on young children’s representations of economic life. Studies of the development of children’s understanding of economic life in middle childhood and adolescence have emphasized the systematic quality of their representations (e.g. Jahoda, 1979, 1981; Furth, 1980; Berti & Bombi, 1981). Goldstein & Oldham (1979) reported that children followed adult representations in associating occupations with gender. In our own society there is still a coordination of the division of labour with gender; a contrast appears between the public and private spheres of social life, with men being the dominant public figures and women the dominant private figures (e.g. Rosenblatt & Cunningham, 1976). Such a contrast also suggests a greater value attached to men rather than women in occupational roles, and to women rather than men in domestic roles. The distinction between work and domestic settings is recognized by very young children. Duveen & Shields (1984) found that a majority of 3-year-olds could successfully sort a collection of photographs into two groups, those showing people at work and those showing people at home. In another task in the same study 3-, 4- and 5-year-old children were shown photographs of men at work and asked if women could perform these roles, and also photographs of women engaged in domestic activities and asked if men could perform such roles. Initially children construed the contrasting associations of public with male and private with female as constituting exclusive categories so that the gender distinction between public and private constrained their judgements about the accessibility of work and domestic roles to men and women. Three-year-olds mapped the distinction between work and domestic roles onto that between male and female in a one-to-one correspondence, so that work and domestic roles were perceived as exclusively male and female respectively. By the age of 4 this mapping had become less rigorous and domestic roles were seen as being accessible to men. The distinction between work and domestic settings had been separated to some degree from the distinction between male and female, although the two frames of reference were still not construed as being orthogonal. Even at 5 years women were frequently denied access to work roles. No differences between boys and girls were observed in these results. Specific effects of differing social gender identities were observed in a more complex task in the same study. Children were shown photographs of men and women engaged in various work roles and asked whether these figures could also occupy domestic and parental roles. Girls recognized a multiplicity of roles for the female figures at younger ages than did the boys. At 5 years only half the boys recognized that these female figures could take on domestic roles, while 85 per cent of the girls made such judgements. For the male figures, on the other hand, neither boys nor girls at any of the ages studied (3, 4 and 5 years) recognized that they could simultaneously occupy both work and domestic roles. Why, then, should girls recognize a multiplicity of roles for women at younger ages than do boys when neither girls nor boys recognize any such multiplicity for

168


men? The male figures presented to the children showed them engaged in public, ‘male’ occupational settings. In this case there is an alignment between the gender of the stimulus figure and the gender marking of the setting, an alignment which is also congruent with the social gender identity of both boys and girls. In spite of the lack of alignment between the gender of the female stimulus figures and the ‘male’ marking of the settings in which they were presented, girls’ social gender identity allowed them to accept this discrepancy. Although young girls recognized the gender markings of occupational and domestic settings they did not construe these markings as establishing exclusive categories. This flexibility may derive from the salience which the multiplicity of roles undertaken by women has for girls’ gender identity. In the absence of data from other studies this interpretation must be regarded as provisional. Nevertheless, it does illustrate how the concepts of social gender identity and social representation may be used to account for gender differences in the development of children’s representations of economic life. In our review of the development of social gender identity we have emphasized both the need for an explicit codification of adult representations and an awareness of the developing capacities available for their reconstruction. We have reviewed our own research on infancy and the transition to symbolic functioning from this perspective. Our focus on early development in the reconstruction of social representations allows us to deal with publicly marked aspects of social action, and is also particularly appropriate to the periods of sensorimotor and preoperational thought. Social gender identity undergoes further developments as children become capable of concrete and formal operational thinking.

5 Conclusions Adults’ social representations of gender structure the environment in which children develop. In their re-construction of reality children need to situate themselves in relation to (or to adapt themselves to) this significant dimension of social organization. They need to construct what we have described as a social gender identity. In this re-construction the crucial developmental parameter is that of decentration. We are employing here a classic Piagetian term to identify the child’s increasing ability to coordinate simultaneous variations in many dimensions. Frequently, and importantly for the present discussion, the dimensions to be coordinated are the perspectives of self and other (actor and observer) and of male and female. In constructing a social gender identity children become able to coordinate perspectives in such a way that they can function effectively as both actor and observer (which, as we have seen, is a source of difficulty for very young children). In addition, they must also be able to locate other persons in relation to these perspectives (which, as the data again illustrate, is another source of difficulty). In this paper we have construed the ‘individual–society interface’ in terms of the concept of social identity. We have argued that social identities correspond to human subjects as they are construed in quantitative social psychology, that is

The significance of social identities 169 social identities describe individuals as they are engaged in the field of social action. The limit of this construction is that it does not describe the psychic identity of individuals, it does not grasp the uniqueness of human individuals. Piaget distinguished between the epistemic subject and the psychological subject in order to emphasize the sense in which his developmental theory considered the communalities between individuals and not their differences (Piaget, 1966; Duveen, 1984; Apostel, 1985). Social identities may be viewed as occupying the space between communalities and differences. Interpreting Piaget’s logical structures as a particular set of social representations we can adapt his binary distinction to include the epistemic subject, the social psychological subject and the psychic subject. Our view of social identities as the product of individuals’ reconstruction of their social world may be compared with that of symbolic interactionists. While there are many points of similarity between these perspectives they are not isomorphic. Berger’s (1966) discussion of the social psychology of identity in the context of the sociology of knowledge allows us to clarify the differences between them. Central to Berger’s analysis is the notion that ‘psychological reality refers . . . to the manner in which the individual apprehends himself, his processes of consciousness and his relations with others’ (Berger, 1966, p. 106). This view tends to merge the levels of social identity and psychic identity which we have sought to separate. The developmental emphasis apparent in our analysis is an integral element in the social psychological perspective we have described. As Piaget and Vygotsky have argued, to understand something it is necessary to understand how it has been constructed. To understand social gender identity it is necessary to understand how it arises in the course of children’s reconstruction of their social world. This re-construction is itself, of course, set in the context of a whole series of socio-psychological processes, and for this reason we might identify the theoretical perspective outlined in this paper with what has recently been termed genetic social psychology (cf. Moscovici, 1976b; Doise, 1978; Perret-Clermont, 1980). We have argued that social representations constitute the primitive terms from which conceptions of individuality and sociality are constructed. The theoretical intention behind our use of the concept of social gender identity is to provide a means for grasping and analysing the dialectic of this construction.

Acknowledgements Parts of the research reported here were supported by Grant No. HR 5871 from the Social Science Research Council, and Grant No. C00232113 from the Economic and Social Research Council. We would also like to thank Liza Catan, Eric Clarke, Marie Jahoda, Peter Lloyd. Maureen Shields and Carpline Smith for their comments on earlier drafts of this paper. We have also benefitted from the editorial advice of Gün Semin, the comments of two unnamed referees and from discussions with colleagues in seminars at Bristol University and the University of Florida.

170


Notes 1 2

In this sense one can note that social representations belong to that order of things for which Sartre (1976) coined the term ‘practico-inert’ to designate the manner in which the products of the history of human praxis confront present human activities. These results do not accord with some of the theoretical propositions current in the theory of intergroup relations, which suggest that the clearest differentiation between social identities would occur in situations which offer the greatest opportunity for social comparison (Tajfel, 1978: Turner, 1982). They emphasize the importance of the criticisms which both Breakwell (1979) and Deschamps (1982, 1984) have directed against the formulations of intergroup relations. General propositions derived from laboratory studies employing arbitrary categorizations are insufficient to account for the effects of categorical differentiations grounded in social representations. The evaluative aspects of social representations influence cognitive processing in a non-arbitrary way.

References Apostel, L. (1985). The future of genetic epistemology (commentary on Gruber). In G. Butterworth, J. Rutkowska & M. Scaife (eds). Evoluton and Developmental Psychology. Brighton: Harvester. Archer, J. & Lloyd, B. (1985). Sex and Gender. New York: Cambridge University Press. Barratt, B. (1984). Psychoanalytic Knowing and Psychic Reality. Hillsdale, NJ: Analytic Press. Berger, P. (1966). Identity as a problem in the sociology of knowledge. Archives of European Sociology, VII, 105–115. Berti, A. E. & Bombi, A. S. (1981). Il mondo economico nel bambino. Florence: La Nuova Italia (English translation forthcoming: Cambridge University Press). Bloom, L. (1974). Talking, understanding and thinking. In R. L. Schiefelbusch & L. L. Lloyd (eds), Language Perspectives: Acquisition, retardation and intervention. London: Macmillan. Breakwell, G. (1979). Women: Group and identity. Women’s Studies International Quarterly, 2, 9–17. Connell, R. W. (1985). Theorising gender. Sociology, 19, 260–272. Deschamps, J.-C. (1982). Social identity and relations of power between groups. In H. Tajfel (ed.). Social Identity and Intergroup Relations. Cambridge: Cambridge University Press. Deschamps, J.-C. (1984). Identité sociale et différenciations catégorielles. Cahier de Psychologie Cognitive, 4, 449–474. Doise, W. (1978). Groups and Individuals. Cambridge: Cambridge University Press. Dumont, L. (1970). Homo Hierarchicus. London: Weidenfeld & Nicholson. Duveen, G. M. (1984). From social cognition to the cognition of social life: An essay in decentration. Unpublished DPhil thesis, University of Sussex. Duveen, G. & Shields, M. (1984). The influence of gender on the development of young children’s representations of work roles. Paper presented to the First European Conference on Developmental Psychology, Groningen, Netherlands. Emde, R. N. (1983). The prerepresentational self and its affective core. The Psychoanalytic Study of the Child, 38, 165–192. Foucault, M. (1970). The Order of Things: An Archaeology of the Human Sciences. London: Tavistock.

The significance of social identities 171 Furth, H. (1973). Piaget, IQ and the nature–nurture controversy. Human Development, 16, 61–73. Furth, H. (1980). The World of Grown-Ups. New York: Elsevier. Goldberg, S. & Lewis, M. (1969). Play behavior in the year-old infant: Early sex differences. Child Development, 37, 21–31. Goldstein, B. & Oldham, V. (1979). Children and Work. New Brunswick, NJ: Transaction Books. Herzlich, C. (1972). La représentation sociale. In S. Moscovici (ed.), Introduction à la Psychologie Sociale. Paris: Librairie Larousse. Jahoda, G. (1979). The construction of economic reality by some Glaswegian children. European Journal of Social Psychology, 9, 115–127. Jahoda, G. (1981). The development of thinking about economic institutions: The bank. Cahiers de Psychologie Cognitive, 1, 55–73. Lewis, M. & Brooks, J. (1975). Infants’ reactions to people. In M. Lewis & L. Rosenblum (eds), The Origins of Fear. New York: Wiley. Lewis, M. & Rosenblum, L. (eds) (1975). The Origins of Fear. New York: Wiley. Lloyd, B. (1986). The social representation of gender. In J. Bruner & H. Weinreich-Haste (eds), Making Sense: The Child’s Construction of the World. London: Methuen. Lloyd, B. & Smith, C. (1985). The social representation of gender and young children’s play. British Journal of Developmental Psychology, 3, 65–73. Lloyd, B. & Smith, C. (1986). The effects of age and gender on social behaviour. British Journal of Social Psychology, 25, 33–41. Morris, C. (1972). The Discovery of the Individual 1050–1200. New York: Harper & Row. Moscovici, S. (1973). Foreword. In C. Herzlich, Health and Illness. London: Academic Press. Moscovici, S. (1976a). La Psychanalyse, Son Image et Son Public. Paris: Presse Universitaires de France. Moscovici, S. (1976b). Social Influence and Social Change. London: Academic Press. Moscovici, S. & Hewstone, M. (1983). Social representation and social explanation: from the ‘naive’ to the ‘amateur’ scientist. In M. Hewstone (ed.), Attribution Theory. Oxford: Blackwell. Mugny, G., De Paolis, P. & Carugati, F. (1984). Social regulation in cognitive development. In W. Doise & A. Palmonari (eds), Social Interaction in Individual Development. Cambridge: Cambridge University Press. Perret-Clermont, A.-N. (1980). Social Interaction and Cognitive Development. London: Academic Press. Piaget, J. (1966). Mathematical epistemology and psychology. In E. W. Beth & J. Piaget (eds), Mathematical Epistemology and Psychology. Dordrecht, Holland: Reidel. Rosenblatt, P. & Cunningham, M. R. (1976). Sex differences in cross-cultural perspective. In B. Lloyd & J. Archer (eds), Exploring Sex Differences. London: Academic Press. Ryle, G. (1949). The Concept of Mind. London: Hutchinson. Sartre, J. P. (1976). Critique of Dialectical Reason. London: New Left Books. Serbin, L. A., Connor, J. M., Burchardt, C. J. & Citron, C. C. (1979). Effects of peer presence on sex-typing of children’s play behaviour. Journal of Experimental Child Psychology, 27, 303–309. Smith, C. (1982). Mothers’ attitudes and behaviour with babies and the development of sex-typed play. Unpublished DPhil thesis, University of Sussex. Smith, C. & Lloyd, B. (1978). Maternal behaviour and perceived sex of infant. Child Development, 49, 1263–1265.

172


Stern, D. (1983). The early development of schemas of self, of other and of various experiences of ‘self with other’. In J. D. Lichtenberg & S. Kaplan (eds), Reflections on Self Psychology. New York: International Universities Press. Tajfel, H. (1978). Social categorization, social identity and social comparison. In H. Tajfel (ed.), Differentiation Between Social Groups. London: Academic Press. Turner, J. (1982). Towards a cognitive redefinition of the social group. In H. Tajfel (ed.), Social Identity and lntergroup Relations. Cambridge: Cambridge University Press.

11 Social representations as a genetic theory Gerard Duveen and Barbara Lloyd

Social representations as a perspective in social psychology The concept of social representations introduced into social psychology by Moscovici and his collaborators has had a chequered reception in the Englishspeaking world. It is in La Psychanalyse, son image et son public that Moscovici (1976a) elaborated the concept of social representation most fully, both theoretically and empirically. In the absence of a translation, even Moscovici’s own English presentations of the concept have an abstract, general or programmatic character, since they introduce a theoretical perspective without the benefit of a clear demonstration of its value for empirical research (see Moscovici, 1973, 1981, 1983, 1984, 1988; Moscovici and Hewstone, 1983). In calling the first chapter of his book ‘Social representation: a lost concept’ Moscovici implies that social psychology has become disengaged from a concern with the situation of psychological processes in social life. The concept of social representation is intended to restore to social psychology an awareness of the social by providing the means for comprehending social life from a psychological perspective. Such a perspective is a necessary prerequisite for understanding the influence of social relations on psychological processes. Moscovici defines social representations as system(s) of values, ideas and practices with a twofold function; first, to establish an order which will enable individuals to orient themselves in their material and social world and to master it; and secondly to enable communication to take place among the members of a community by providing them with a code for social exchange and a code for naming and classifying unambiguously the various aspects of their world and their individual and group history. (Moscovici, 1973, p. xiii) This definition establishes social representations as particular kinds of structures which function to provide collectivities with intersubjectively shared means for understanding and communicating. As well as referring to social representations as structures in this way, Moscovici also uses the term to designate the process through which such structures are constructed and transformed. (If the English

174


language were more flexible one could refer to this process as social representing, but to avoid such solecisms we shall use the singular form without an article to refer to social representation as process, and the singular with an article or the plural to refer to a social representation or social representations as structures.) As process, social representation is not bound by the canons of logical discourse, nor is it regulated by procedures of empirical verification or falsification. Rather, social representation is construed as being composed of two complementary functions, anchoring (whereby the unfamiliar or remote is absorbed into the familiar categories of everyday cognition) and objectification (whereby representations are projected into the world, so that what was abstract is transformed into something concrete). These two functions are interdependent, in the sense that a representation can become securely anchored to the extent that it is also objectified, and vice versa, that objectification would be impossible unless a representation were anchored. Nevertheless, they can be analytically distinguished as two moments in the process of social representation. Moscovici’s conceptualisation of the process of social representation is related to his distinction between the consensual universe of social representations and the reified universe of scientific discourse which respects the laws of logic and whose products are open to empirical investigation (1981). His purpose in making this distinction is not to propose a particular philosophy of science, but to point to a central phenomenon of our own society whereby the category of scientific understanding is distinguished from the category of everyday or common-sense understanding. What is proposed, then, is that these two universes, the reified and the consensual, correspond to a particular social representation in which the realm of the scientific is distinguished from that of common sense. The distinction is, nevertheless, a powerful one, as Moscovici notes. Science ‘attempts to construct a map of the forces, objects and events unaffected by our desires and consciousness. [Social representation] stimulates and shapes our collective consciousness, explaining things and events so as to be accessible to each of us and relevant to our immediate concerns’ (Moscovici, 1981, p. 187). Social psychology, in this view, is concerned with the analysis of the consensual universe, for which the theory of social representations provides the conceptual apparatus. Social representations thus provide the central, integrative concept for a distinct perspective on social psychology, a perspective which is not an entirely new departure but one which recovers and enriches traditions which had become marginalised in the discipline. It shares with both Piagetian theory and other constructivist trends in psychology and the social sciences an epistemological basis in treating the subject and object of knowledge as correlative and co-constitutive and rejecting the view that these terms designate independent entities. The ontological corollary to this position is that social representations are constitutive of the realities represented, a constitution (or construction) effected through anchoring and objectification. Thus the content of what is constructed is accorded the same significance as the process of construction, and hence Moscovici’s dictum that social representations are always the representation of something (Moscovici, 1976a, 1984).

Social representations as a genetic theory 175 In this respect the theory of social representations is not a psychology of cognitions about social life, but rather a theory in which psychological activities are located in social life. Indeed, social representations can be contrasted with social psychological theories based on narrower definitions of psychological activity focussed on notions of attitudes or attribution. In such theories social cognition is viewed as cognitive processes in relation to social stimuli, but these ‘social stimuli’ are taken as given, since social life itself remains untheorised. The effect of this theoretical lacuna is to present a view of social cognition as the activity of individual minds confronting the social world. For social representations, on the other hand, attitudes and attributions arise as consequences of participation in social life; they form, as it were, the visible tip of an iceberg whose submerged portion comprises the very structures which enable the subject to construct meaningful attitudes and attributions. As Moscovici notes, the concept of social representations may be difficult to grasp because it occupies a ‘mixed position, at the crossroads of a series of sociological concepts and a series of psychological concepts’ (1976a, p. 39). The focus for this perspective is the systems of social representations through which groups construct an understanding, or theorise, social life. Thus as well as being always the representation of something, social representations are also always representations of someone or some collective (for example, Moscovici, 1976a, 1984). The interdependence between social representations and the collectives for which they function means that social life is always considered as a construction, rather than being taken as a given. The duality of social representations in constructing both the realities of social life and an understanding of it recalls a similar duality in Piaget’s conceptualisation of operational structures. Piaget’s task was facilitated by the availability of a knowledge of physics, mathematics and logic. These sciences describe a reified universe which provided him with a perspective from which it was possible to understand and interpret the behaviours of subjects at different levels of development. Without access to the logic of class inclusion it would have been difficult for him to have understood the attempts of children to answer the question whether there were more flowers or more roses in a given collection. In the consensual universe of social life there is no privileged vantage point which offers an objective perspective from which to orient any investigation. In some circumstances it is possible for investigations of social representations to locate a point of reference comparable to the ‘objectified’ perspective available to Piaget. Moscovici’s (1976a) study of psychoanalysis, for instance, takes the body of psychoanalytic theory originating in the work of Freud as an objectified point of reference from which to compare and contrast the social representations of psychoanalysis constructed by different social groups. He was able to observe the transformation of this body of knowledge as it was reconstituted in the network of representations held by the different groups. Again, without access to the corpus of psychoanalytic theory it would be difficult to understand and interpret the responses of members of different social groups to questions about psychoanalysis.

176


The study of social representations of psychoanalysis is an example of the way in which the reified universe of science is represented in the consensual world of everyday understanding. But not every social representation originates as a body of knowledge in the reified universe of scientific discourse. Indeed, most of the studies presented in this book [Social Representations and the Development of Knowledge] concern social representations of themes drawn from within the realm of the consensual universe. In dealing with themes such as gender or the emotions a primary task for the investigators is to locate a perspective from which to organise the investigation. In some of the chapters the investigators have been able to use a knowledge of the social representations among adults as a point of reference for analysing the development of these representations through childhood. This is the case, for example, in the chapters by Lloyd and Duveen on gender, Corsaro on rules of social interaction and Semin and Papadopoulou on reflexive emotions. But in other cases the perspective from which to organise an investigation is more difficult to locate. In the studies of social representations of childhood by D’Alessio and by Emiliani and Molinari, for instance, no specific social group provides a fixed point of orientation. These studies resolve the issue by focussing on the object, childhood, and comparing the ways in which different groups represent this object. Similarly, both Carugati’s study of the social construction of intelligence and De Paolis’ analysis of social representations of psychology illustrate this perspective focussed on the object. The issue outlined here is methodological in the sense that it concerns the relation of an epistemological position to empirical investigation. It is a strategic problem for research on social representations rather than a question of specific techniques. In each case the researcher has to proceed by identifying what Lucien Goldmann describes as a significant structure (Goldmann, 1976, 1980), by which he means a structure which has a functional necessity for a particular group. Social representations as significant structures identify both the group which constructs a representation and the content which is represented. The notion of social representations as significant structures also helps to distinguish this theory from other recent attempts to construct theories concerning the social psychological analysis of social life in terms of ordinary explanations (Antaki, 1981), linguistic repertoires (Potter and Wetherell, 1987) or rhetoric (Billig, 1987). What all of these approaches share is that they identify particular processes independently of any specific content, so that, again, the particular characteristics of specific aspects of social life remain outside the theory. While certain forms of ordinary explanation or particular linguistic repertoires or rhetorical devices all describe identifiable features of social discourse, these are also all features of the discourses of particular groups about specific aspects of social life, and thus draw on underlying social representations. Although they may provide useful analytical tools for investigating social representations, without explicating their implicit references to social representations, the analysis of these features of discourse cannot describe the social-psychological representation of social life. These features describe formal structures which it is difficult to locate in concrete social-psychological terms bound to some particular content. In this sense it can be said that these features do not constitute significant structures.

Social representations as a genetic theory 177

Social representations as a genetic theory A genetic perspective is implied in the conception of social representations, in the sense that the structure of any particular social representation is a construction and thus the outcome of some developmental process. The works of Piaget and Goldmann again offer a comparable point of view. Both of these authors insisted on describing their approach as a genetic structuralism in which a structure is always viewed as a particular moment in development. A structure is the relatively enduring organisation of a function, while the realisation of a function implies its organisation in a structure. For similar reasons the theoretical perspective of social representations can be described as a genetic social psychology. Even if social representations as structures do not meet the strict formal criteria proposed by Piaget (1971), they nonetheless constitute organised wholes with the specific function of making communication and understanding possible. Conceived in this way the concept of social representation appears to have a general application as a means of comprehending the way in which socio-epistemic structures exercise a psychological influence. But to grasp the complexities subsumed in this concept it is useful to distinguish three types of transformations associated with social representation. There are processes of sociogenesis, which concerns the construction and transformation of the social representations of social groups about specific objects, ontogenesis, which concerns the development of individuals in relation to social representations, and microgenesis, which concerns the evocation of social representations in social interaction. Sociogenesis Sociogenesis is the process through which social representations are generated. Moscovici’s (1976a) study of psychoanalysis is an example of the diffusion of scientific knowledge through the community as it is reconstructed by different social groups. But, as we noted above, it is not only knowledge originating in scientific discourses which gives rise to social representations: other themes also circulate in society through the medium of social representations. In recent years social representations of gender, for instance, have clearly been undergoing transformations, providing another example of a sociogenetic process. One chapter which clearly touches on issues of sociogenesis is De Paolis’ comparisons of the social representations of psychology held both by psychologists and other professional groups working in the field of mental health. Sociogenesis takes place in time, so that even when social representations are investigated at a particular moment in time, the resulting description needs to be viewed in a diachronic perspective. Moscovici’s study, for instance, was originally published in 1961, and describes the structure of social representations of psychoanalysis at that time. Clearly the sociogenesis of these representations had taken place over the years since Freud’s work began to appear. In the intervening years since Moscovici’s study the theory of psychoanalysis has itself evolved and the characteristics of many social groups have also changed. A comparable study

178


undertaken today might reveal transformations in the social representations of psychoanalysis. Sociogenesis thus also points to the historical dimension of social representations. Ontogenesis Human infants are born into a social world constructed in terms of the social representations of their parents, siblings, teachers, etc., representations which also structure the interactions of these others with the child. If, as Moscovici asserts, the society into which children are born is a ‘thinking society’, it is social representations which constitute the ‘thinking environment’ for the child. Developing the competence to participate as actors in this thinking society implies that children can acquire access to the social representations of their community. It is this process which we refer to as the ontogenesis of social representations, although ontogenesis as a process is not restricted to childhood, but occurs whenever individuals, children or adults, engage with novel social representations in order to participate in the life of a group. An adequate account of ontogenesis needs to describe how social representations become psychologically active for individuals. Elsewhere (Duveen and Lloyd, 1986) we have suggested that ontogenesis is a process through which individuals re-construct social representations, and that in doing so they elaborate particular social identities. It is as social identities that social representations become psychologically active for individuals. Thus we can say that in expressing or asserting a social identity individuals draw on the resources made available through social representations. As this formulation implies, there is a distinction between social identities and social representations. Our own research on gender has shown that the same social representation can support distinct social identities. As we note in our chapter, in many respects boys and girls develop similar representations of gender, but they do not behave in similar ways. The influence exercised by social representations on individuals can take different forms. Some social representations impose an imperative obligation on individuals to adopt a particular social identity. This is the case, for example, with representations of gender or ethnicity where individuals are constrained to construct the corresponding social identity. In such cases there is an external obligation which derives from the ways in which others identify an individual in terms of these social categories. In other instances the influence of social representations is exercised through a contractual obligation rather than an imperative one. In these cases an individual joining a social group contracts to adopt a particular social identity. Social representations of psychoanalysis provide an example of such a contractual obligation. As a body of knowledge psychoanalysis exercises no external obligation on individuals to interiorise the categories of analytic thinking as psychologically active constructions. But entry to some social groups (principally that of psychoanalysts themselves, but also other social groups for whom an analytic perspective forms part of their world-view) is dependent upon individuals contracting to construe the world in terms of psychoanalytic categories.1

Social representations as a genetic theory 179 Microgenesis A third genetic aspect of social representations is in social interaction, where individuals meet, talk, discuss, resolve conflicts – in short, communicate with one another. Social representations are evoked in all social interactions through the social identities asserted in the activity of individuals. Social identities, however, are not fixed attributes which individuals carry into each interaction and which remain invariant through the course of interaction. Rather, particular social identities are constructed through the course of social interaction, or through the successive encounters which make up the history of a particular interpersonal relation. There is a genetic process in all social interaction in which particular social identities and the social representations on which they are based are elaborated and negotiated. It is this process which we refer to as the microgenesis of social representations. The evocation of social representations in social interaction occurs first of all in the ways in which individuals construct an understanding of the situation and locate themselves and their interlocutors as social subjects. In many circumstances, of course, there will be a mutuality in the understandings constructed by different participants which will obviate the need for any explicit specifications or negotiation of social identities, though one can still describe the course of such social interactions as the negotiation of social identities in the same sense as one speaks of a ship negotiating a channel. But where the mutuality of understanding cannot be taken for granted, or where an assumed mutuality breaks down, the negotiation of social identities becomes an explicit and identifiable feature of social interaction. In these circumstances the negotiation of social identities may involve the coordination of different points of view and the resolution of conflicts. In every social interaction there is a microgenetic process in which social identities are negotiated and shared frames of reference established, processes for which social representations provide the resources. Corsaro’s chapter gives a vivid illustration of the microgenetic process and presents a view of ontogenetic and sociogenetic dimensions. Language is, of course, a central medium through which social interactions are conducted, and recent studies in sociolinguistics have emphasised the construction of social identities in discourse (Gumperz, 1982) as well as the role of social representations (Rommetveit, 1974, 1984). Through the course of social interaction participants may come to adopt positions distinct from those with which they entered the interaction, and in this sense microgenesis is always a process of change. In many instances the changes which can be observed through the course of social interaction are transitory rather than structural as individuals adopt particular social identities in order to pursue specific goals or accomplish specific tasks. Yet social interaction is also the field in which social influence processes are most directly engaged (see Moscovici, 1976b), and in some instances the influences at work in social interaction may also lead to structural change in the representations of participants. These changes may be ontogenetic transformations in the development of social representations in individual subjects

180


(see Doise and Mugny, 1984), but they may also be sociogenetic transformations resulting in the restructuration of social representations.2 Some examples may help to illustrate the possible relationships between these three types of genetic transformation of social representations. Consider first of all the scientist who proposes a new theory, and let us assume that we are dealing with an Einstein or a Freud proposing a radical new interpretation of the human situation or human experience. Through various forms of social interaction (publications or lectures) the scientist tries to communicate this theory to colleagues. The communication will have been successful to the extent that other scientists will have understood the concepts being proposed and also accepted that these concepts are well founded and not in error. The outcome will be ontogenetic transformations in the representations held by these scientists as individuals, as well as a sociogenetic transformation in the representation held by the scientific community as a social group. By contrast, consider the developing child as they grasp some social representation of their community, gender or nationality, for example. For this development to occur the child needs to receive some communication, whether through interaction with other children or adults or from the public representations presented in the media. These microgenetic processes will have led to ontogenetic transformations in the child’s representation of the world, but the social representations of their community are unlikely to be influenced by these particular microgenetic processes. In this case there is ontogenesis without sociogenesis, a state of affairs which is a characteristic feature of childhood given the negligible influence which children are able to exert on the representations held by their community. In both of these examples ontogenesis and sociogenesis are the consequence of microgenetic processes. Indeed, microgenesis constitutes a motor, as it were, for the genetic transformations of social representations.

Notes 1 2

We are grateful to Serge Moscovici for suggesting these particular terms, imperative and contractual obligations, to characterise this distinction. Lucien Goldmann makes a similar point when he notes that there may be information which can be communicated to either groups or individuals only on condition of a transformation in their socio-psychological structures (Goldmann, 1976).

References Antaki, C. (ed. 1981). The Psychology of Ordinary Explanations of Behaviour. London: Academic Press. Billig, M. (1987). Arguing and Thinking. Cambridge: Cambridge University Press. Doise, W. and Mugny, G. (1984). The Social Development of the Intellect. Oxford: Pergamon. Duveen, G. and Lloyd, B. (1986). The significance of social identities. British Journal of Social Psychology, 25, 219–30. Goldmann, L. (1976). Cultural Creation in Modern Society. Saint Louis: Telos Press. —— (1980). Essays on Method in the Sociology of Literature. Saint Louis: Telos Press.

Social representations as a genetic theory 181 Gumperz, J. (1982). Language and Social Identity. Cambridge: Cambridge University Press. Moscovici, S. (1973). Foreword to C. Herzlich, Health and Illness. London: Academic Press. —— (1976a). La Psychanalyse, son image et son public. 2nd edn. Paris: Presses Universitaires de France. —— (1976b). Social Influence and Social Change. London: Academic Press. —— (1981). On social representations. In J. Forgas (ed.), Social Cognition. London: Academic Press. —— (1983). The coming era of social representations. In J. P. Codol and J. P. Leyens (eds.), Cognitive Analysis of Social Behaviour. The Hague: Martinus Nijhoff. —— (1984). The phenomenon of social representations. In R. Farr and S. Moscovici (eds.), Social Representations. Cambridge: Cambridge University Press. —— (1988). Notes towards a description of social representations. European Journal of Social Psychology, 18, 211–50. Moscovici, S. and Hewstone, M. (1983). Social representations and social explanations: from the ‘naive’ to the ‘amateur’ scientist. In M. Hewstone (ed.), Attribution Theory. Oxford: Blackwell. Piaget, J. (1971). Biology and Knowledge. Edinburgh: Edinburgh University Press. Potter, J. and Wetherell, M. (1987). Discourse and Social Psychology. London: Sage. Rommetveit, R. (1974). On Message Structure. London: Wiley. —— (1984). The role of language in the creation and transmission of social representations. In R. Farr and S. Moscovici (eds.), Social Representations. Cambridge: Cambridge University Press.

12 Representations, identities, resistance Gerard Duveen

From the perspective of social representations, social identity appears as a function of representations themselves. In elaborating this view, this chapter draws primarily on research on the development of social gender identities to argue the following points: (a) identity is as much concerned with the process of being identified as with making identifications; (b) identities can be construed as points or positions within the symbolic field of a culture; in other words, identities are constructed externally and not simply elaborated internally; (c) representations always imply a process of identity formation in which identities are internalized and which results in the emergence of social actors or agents; (d) identities provide ways of organizing meanings so as to sustain a sense of stability; (e) an identity is essentially an asymmetry in a relationship which constrains what can be communicated through it; and (f) we need to consider the possible varieties of forms of social identity.

Introduction For almost as long as there has been social psychology the question of identity has been construed as the answer to the question “Who am I?” And the answer has generally been that “I am who the Other says I am.” We can look back to William James as the locus classicus of this sense of identity in his famous dictum that “A man has as many social selves as there are individuals who recognize him” (1981, p. 281, first published 1890). James, as is well known, construed the Self as a duality of the I – the Self as a knowing subject – and the Me – the Self as it is known. Following James, this perspective was further elaborated in Cooley’s notion of the “looking-glass self,” and particularly in G. H. Mead’s conception of the Self as a dialogue between the I and the Me, between the I which reacts to the attitudes of others and the Me which is the sum of the attitudes of others which have been internalized. (An important exception to the tradition of James, Cooley and Mead is Jean-Paul Sartre (1974), for whom the answer to this question is not that I am who the other says I am, but rather I am what I make of what the other says I am.) While these classical authors wrote almost exclusively about the self, more recent writers have tended to focus on the concept of social identity conceived, in

Representations, identities, resistance 183 Tajfel’s terms, as “that part of an individual’s self-concept which derives from his knowledge of his membership of a social group (or groups) together with the value and emotional significance attached to that membership” (1981, p. 255). In this conception, Tajfel is careful to leave a space within the self which is not accounted for by any social identification, although he has little to say about what this space might be. And although this conception has continued to animate work in contemporary Social Identity Theory, there have also been some dissident voices. Jean-Claude Deschamps (1982), for instance, sketches a perspective in which social identity is conceived as a function of relations of power between groups where access to a sense of self as a bounded and autonomous individual is seen as the expression of a specific form of symbolic capital. On this view self becomes a function of the constructions of identity, rather than identity being a subset of the self. However, it is not my intention in this paper to try and unravel any consistent pattern in the relations between self and identity, in part because the issues being addressed in this area of social psychology continue to be considered by different authors under the rubric of “self or “identity.” When Oyserman and Markus (1998), for example, discuss the self as social representation, they are recognizably referring to the same phenomena which Barbara Lloyd and I discussed as social identity (Duveen & Lloyd, 1986). There may be good – or bad – reasons for preferring one term to the other, either for some substantive purpose or in relation to the way these terms have figured in the discourse of social psychology. But we should not lose sight of the phenomenon under discussion, namely the ways in which individuals or persons or agents come to have a sense of who they are through a recognition of their position within the symbolic space of their culture. Although in more recent years the question of identity has been more usually framed as a discussion of social identity rather than directly in terms of the self, nevertheless, the question “Who am I?” has remained central to this discussion. At the same time it has also been noticeable that there has been relatively less attention to the nature of the answers given to this question in terms of the content of the categories through which identities are defined. Henri Tajfel was a notable exception, who, while interested in formulating general propositions about the sources and consequences of social identity, was also deeply concerned with the nature of these categorizations. What it meant to be defined as a Jew in Europe in the 1930s, or as Black in Britain in the 1960s and 1970s were enduring questions for him. And although there have been some other examples of work which has tried to address similar questions – Breakwell (1979) on women for example – this strand has been a minority interest in work on social identity, and a minority which has struggled to exert any strong influence on a dominant and wellorganized majority for whom social identity theory has been framed as a discussion of the processes through which identity is sustained or manipulated rather than on the content of those identities. A lack of concern with the content of identities in contemporary social psychology is also accompanied by a one-sided view of the process of identity itself. Consider, for example, the newborn child, newly arrived into a world which

184


is already structured by the social representations of the communities into which they arrive. The circulation of representations ensures that the child is born into a world in which meanings already exist, and which have, in some way or another, already prepared the way for the child. Even before they are born, at times even before they are conceived, children are already the objects of the hopes, fears, aspirations, and anxieties of the people who will become its parents. The child may be wanted or unwanted, parents may wish for a boy or a girl, and so on. Once the child is born, the parents’ representations which underlie their construction of the child as a social object also influence and regulate the ways in which they interact with the child. Representations of gender, and of the process of development itself, structure the relationships which emerge through the interaction between parents and children. As the child develops, they also begin to internalize the identities which have been extended to them, and in their turn they develop into social actors capable of taking their place as competent and functioning members of their community. Before they are capable of independent activity in the field of gender (or any other social field) children are the objects of the representations of others. A child is always a construction before it is a reality, but a construction of others, and through this construction its parents and others extend to it a social identity, they locate it in a social space. The situation of the child expresses something fundamental about the nature of identity, namely that identity is as much about the process of being identified as it is about the process of identification. My first concern in this chapter is to illustrate how approaching identity from the perspective of social representations can take us back to a concern with the content of identities. This concern has grown out of an interest in the development of social representations, particularly in relation to gender (cf. Lloyd & Duveen, 1990,1992). And it is no accident that questions of the relations between representations and identities are posed so clearly from a developmental perspective focused on the process of children’s emergence as social actors, that is, as active participants in a symbolic community.

The development of social gender identities Becoming an actor in the field of gender Like all good developmental narratives, this one begins with the newborn baby, who, with its instinctual reactions and emerging schemes, remains without any sense of self as either agent or identity. But for others around the child – principally parents, siblings, and so on – this newborn infant is not a neutral object, but one which is invested with the characteristics of a social identity. For the other, the baby is already a little boy or a little girl, and by extending a social identity to the child the other incorporates the child into a representational system which both gives the child a social location as well as regulates the action of the other towards the child. In this sense the child is first of all an object in the representational world of others, who anchor this new and unfamiliar being in a particular classification

Representations, identities, resistance 185 and give them a particular name, and who objectify their representations through the ways in which they interact with the child. For the newborn infant one of the key constitutive acts of the other is the assignment of the child to a sex group. In a famous study Rubin, Provenzano, and Luria (1974) asked the parents of newborn children to describe their babies. Even when there were no differences in birth weight, birth length, or apgar scores, these descriptions varied systematically as a function of the biological characteristics of children’s external genitalia. At this point in their lives these children had literally done nothing, and yet their very physical form served as signifiers for others to project a gender identity on to them. As we have shown elsewhere (Lloyd & Duveen 1990, 1992; Duveen 1994), this extended social identity is progressively internalized by children who, by the age of four, are already capable of establishing themselves as independent actors in the field of gender. It may be helpful here to trace out some of the key moments in this development. In the first 18 months of children’s lives their gender identity is regulated by the actions of others. Observations of women who were themselves the mothers of young children playing with six-month-old infants illustrate some of the ways in which this occurs. In these studies (Smith & Lloyd, 1978; Smith, 1982) the women had not previously met the babies they were asked to play with, and the children were presented to them dressed and named stereotypically as either girls or boys. In fact half the time these children were cross-dressed and cross-named. Regardless of the biological sex of the child, the first toy offered to them for play was usually one which carried a gender marking consistent with the child’s apparent gender. Thus a child presented as a girl was usually offered a doll, while a child presented as a boy was more likely to be offered a hammer. Even more striking were these women’s reactions to these infants’ gross motor activity. Such activity is not itself an indicator of gender, being a feature of the behavioral repertoire of both girls and boys. However, when this activity was produced by a child dressed and named as a boy the women offered both verbal and motor encouragement, stimulating further activity. The same behavior produced by a child dressed and named as a girl was viewed by the women as a sign of distress, to which they responded by soothing and calming the child. From these studies it is apparent that for young infants toys and other objects of material culture do not carry any clear gender signification. It is the actions of others1 which recognize the marking of objects and connect these significations with the signifying aspects of the child’s presence in a coherent representational praxis. With the emergence of the semiotic function at the end of the sensorimotor period, children themselves begin to take on a more active role in controlling and regulating their expression of gender. The emergence of children as independent actors in the field of gender can be seen in observational studies of children playing in pairs with familiar peers of the same age (these studies are reviewed in Lloyd & Duveen, 1989, 1990). In one of these studies boys and girls aged three to four years were each observed twice, once playing with a girl and once playing with a boy. All of these observations took place in a room furnished with a variety of toys with known gender markings

186


as either feminine (e.g., dolls, saucepans, shopping bags) or masculine (e.g., guns, trucks, briefcases).2 Each time a child picked up a toy, therefore, their choice could be identified as being either congruent with their own gender, or incongrucnt with their own gender. Analyses of these toy choices showed a strong and asymmetrical pattern. Boys made far more congruent than incongruent toy choices, whereas girls tended to choose more evenly between both congruent and incongruent toys. This asymmetry could also be clearly seen in the length of time children played with different gender marked toys. The asymmetry between boys and girls evident in their play with gender marked toys contrasted sharply with these children’s performance on a range of cognitive and linguistic tasks, in which they were asked to sort pictures of people and toys according to gender, or to identify appropriate gender marked words. Although children’s knowledge of these different codes for gender marking increased as they developed, few differences were found between girls and boys. The differences observed between girls and boys in their use of toys, therefore, could not be attributed to any difference in their knowledge of how these objects were marked for gender. Rather, it seems that even at this young age children are using their knowledge to construct different identities. For the boys in these studies, the material culture of toys seems not just to carry particular significations of gender, but also to provide an arena in which they need to express a clearly differentiated identity. The girls, by contrast, while not oblivious to the gender marking of toys, do not use this material culture as a resource for expressing a differentiated identity. In this early acquisition of gender identities one can see clearly how their thoughts and actions become structured in terms of the representations of the communities in which they are growing up. Their identities are not simply internal elaborations of meanings, but the reconstruction of externally constructed patterns of meanings. Further, one can also see in these studies something of the complexity of the process of internalization. Children, as we have argued (Duveen & Lloyd, 1986; Lloyd & Duveen, 1990, 1992), are not simply internalizing social representations, but, as they do so they are also constructing particular identity positions in relation to these representations. Representations in pretend play The emergence of the semiotic function is marked by the appearance of a wide range of representational activities, and in a recent contribution Hans Furth (1996) has argued that the emergence of these activities also marks the emergence of the child as an active participant in the collective world. His claim, as he puts it, is that “these mental constructions have an inherently societal character, such that the mental object itself is a societal object” (1996, p. 26, original emphasis). On this basis he also suggests an endogenous origin for human society, for which the clearest evidence is to be seen in the pretend play of young children. Pretend play begins to emerge in young children with the emergence of the semiotic function itself, and by the fourth and fifth years it is established as a social and collective

Representations, identities, resistance 187 form of activity in which children not only evoke and recreate aspects of the world they inhabit, but also a forum in which their understanding of this world can be given structure and organization. Furth himself analyzes some examples of the construction of what he calls the “societal frame” in young children’s play, as have other authors such as in Corsaro’s (1990) account of Italian kindergarten children’s recreation of a banking system. The following example, however, returns to the theme of gender, and is taken from our own research with children in the first year of formal schooling (in England this means with five-year-olds). This example, transcribed from video records of children during free play in the classroom of an ordinary infants’ school (Lloyd & Duveen, 1992), illustrates children’s belief that physical/sexual contact between sex group members needs to be validated through marriage. Oscar is chased for some time by the girls but once he is kissed by Christine, perhaps somewhat to his surprise, he proclaims (turn 16) that he is going to marry her. Children create a simple world in which physical contact between sex group members is construed as sexual and involves marriage. In this world actions have direct and predictable consequences. 1. Edith: 2. Christine: 3. Edith: 4. Lulu: 5. Edith: 6. Lulu: 7. Edith: 8. Edith: 9. Joan: 10. Edith: 11. Lulu: 12. Edith: 13. Christine: 14. Oscar: 15. Edith: 16. Oscar: 17. Edith: 18. Oscar: 19. Sally: 20. Oscar: 21. Sally: 22.

. . . and Lulu kiss, uhm, Oscar. Go on. I’m not playing now. Go away, then No, you kiss Oscar and I kiss Darren. I know. Look. You (Joan) kiss him Darren. And I’ll kiss Oscar. Joan kiss Oscar. Joan kiss Darren, and Oscar kiss . . . you! (Starts for Darren, who runs) Hey! Come here. (Grabs Lulu and moves her towards Oscar, not unwillingly) No, kiss! Kiss her on the lips. Kiss her on the lips. Come on! No way! Go on. Kiss her. Kiss her. (Makes a dash for Oscar) I kissed him. I kissed HER! Oooh! (Points at Christine) I’m going to marry her. (With Lulu, no longer struggling, very close) Kiss her. I’m going to marry her. (Also closing in on Oscar) All right (But which one should he kiss?) Kiss me. (They kiss) (All laugh. Oscar throws himself back on sofa)

Central to the social representations of gender in this extract is a reproductive metaphor that offers an image of gender in terms of the bipolar opposition of the masculine and feminine. This is an image which children appear to have acquired

188


very early in their lives and which persists into adulthood. (De Rosa, 1987, also notes that the iconic aspects of social representations of madness are acquired early in life.) In discussing the process of objectification Moscovici refers to the figurative nucleus of a social representation, “an image structure that reproduces a conceptual structure in a visible manner” (Moscovici, 1981). The most graphic examples of iconic aspects of social representations of gender in our work concern children’s evocation of sexuality in their play, where sexuality is evoked precisely as the union of bipolar opposites, and once established is celebrated through the rituals of marriage and domestic life. Indeed for these children there is a syncretic fusion of sexual relations, the institution of marriage, and the complementarity of gender roles in domestic life. The structure of a bipolar opposition is the connecting thread between these different elements, each of which implicates the others, so that when one element is evoked in play it can lead to the evocation of the others. The figurative nucleus of bipolar opposites also supports a conceptualization of social life in terms of two complementary but exclusive categories. This conceptual structure influences how children interpret the world around them, while their participation in collective life provides a scaffolding which confers further legitimacy on this conceptual structure. In sexuality, or more precisely heterosexuality, difference is both asserted because it depends on the presence of bipolar opposites, and also overcome at the same time through the union of these opposites. Sexuality, therefore, can take on a privileged status for young children because it offers the clearest resolution to the problem of difference. As I argued earlier, the image of bipolar opposition connects sexuality with marriage and domestic life, and in their play children’s engagement with this theme expresses and celebrates a certain understanding of the world. In this understanding sex and gender are reduced to a single dimension, and it is the difference between the categories of masculine and feminine which is emphasized, while differences within each of these categories are obscured. Variations in social identities A primary function of social representations is the construction of social objects that provide a stable pattern of meanings for social actors. As in Piaget’s analysis, this conception emphasizes the correlative relationship between the construction of the object and the construction of a sense of self. In other words, the construction of an object is one aspect of a developmental process that also extends to the formation of identities. Thus in addition to the construction of the object as a stabilizing function in the process of social representation, we also need to consider the identity function of representations. Piaget, of course, conducted his analysis at the level of epistemological relations between subject and object, a level of abstraction which assumes a homogeneity in the forms of subjects and objects which are constructed (cf. Duveen, 1994, 1997). But in the social world, such homogeneity cannot be taken for granted; rather, the social world is more frequently marked by heterogeneity, difference, and hierarchy. Different social

Representations, identities, resistance 189 groups construct different understandings as they establish a stable view of the world and their place within it. As this suggests, there is a close relationship between stability and identity as functional aspects of social representations. An identity is first of all a way of making sense of the world, a way of organizing meanings that provides a sense of stability. In the preceding section I reviewed some aspects of the early development of gender identities, emphasizing the transition from extended identities as children are incorporated into the social world through the actions of others, to internalized identities as children become independent actors in the field of gender. In this research identities are considered as characteristics of gender groups, but further research revealed a need to consider variations in identities within gender groups and as a consequence of differentiated patterns of social organization. In this study (Lloyd & Duveen, 1992) we followed children through their first year of compulsory schooling (which in England means the year in which children become five), using firstly ethnographic techniques to describe the structure of gender organization in the classroom, and then more structured observational techniques. The pattern of gender organization varied from classroom to classroom. Both the school as an institution and the teacher as key actor influenced the formation of a local gender culture in each classroom, so that at times the same items of material culture could take on different gender meanings in different classrooms. When this study was undertaken there was much public discussion about the influence of gender in education, and various initiatives set in motion by Local Education Authorities which aimed to provide equal opportunities for girls and boys in school. Within the schools this sometimes produced conflict and discussion among teachers and between teachers and headteachers, a division of opinion which could loosely be characterized as contrasting views of “sex differences” as natural, with views of “gender differences” as open to social influences. Such conflicts were an important source for the variations we observed in local gender cultures in different classrooms. The second theme that this study highlighted was the importance of investigating variations among girls and among boys. While it was possible to see some general patterns in the ways girls and boys behaved – for example in their choice of whom to play with, or in the way they used the material culture of the classroom, including both toys and other objects as well as different spaces – considerable variations within sex groups were also observed. Thus, while in general patterns of association among children showed a tendency toward gender segregation, with boys choosing to play with other boys and girls with girls, we also observed significant variations in these patterns, especially among the girls. For example, some girls formed themselves into very tightly knit little groups of two or three who rarely interacted with any other children in the classroom, girls or boys. Other girls showed a general tendency to associate primarily with girls, while a third group also included some of the boys among their play partners. These different styles of associating with their peers can be seen as the expression of different forms of feminine identity, and indeed, when we used information about children’s pattern of peer association as an index of gender identity we

190


found some consistent effects linking different forms of identity to children’s pattern of using both objects and space in the classroom. By the end of their first year of schooling, children were able to express differentiated and varied gender identities in their classroom practice. However, the same sophistication was not nearly so clearly expressed in children’s reflective knowledge of the codes for marking gender in the classroom. Again, as in the earlier studies, we devised a series of judgment tasks to explore children’s knowledge of gender marking, using as stimuli the kinds of materials familiar from the classroom. In only some isolated examples did children’s performance on these tasks relate to the peer-association index of gender identity. More generally, children’s responses could be divided according to the demands of the task. When children were asked only about their knowledge of gender marking, girls and boys displayed a similar level of competence. However, when children were also able to express a preference their responses were divided on the basis of their membership in a sex group. No doubt these results reflect something of the developmental process involved as children move from practical activity to reflexive awareness. But some of the difficulty is also related to the content of the representations themselves, to the meanings that gender holds for these children. Our ethnographic observations suggested that children are often the most conservative elements in the gender culture of the classroom. As we saw in the analysis of children’s pretend play in the classroom, their representations of gender are structured around a figurative nucleus of a bipolar opposition, a structure that suggests some reasons for children’s conservatism about gender in the classroom. As an image, a bipolar opposition offers a degree of clarity and simplicity which is also consistent with children’s limited capacity for any cognitive elaborations which might require greater sophistication. Children’s resistance to any influence of an egalitarian voice in representations of gender is also a resistance to losing this clear and sharp image of the world. Thus the image of bipolar opposition crystallizes for the child a state of understanding which also fuses the form of knowledge (its categorical structure) with the content of knowledge (the separation between things masculine and things feminine). All things masculine tend to cohere together and to separate from things feminine. As we noted in our ethnography, separation along these lines can come to characterize the pattern of interaction in the classroom, and once established in this way the dynamic interplay of activity and understanding is capable of sustaining such moments over extended periods of time. Yet as well as representing difference the image of gender as a bipolar opposition also represents hierarchy, for the difference between the genders is also a relation of power. Notwithstanding the extent to which this image is saturated with notions of hierarchy and power, so long as the difference between the bipolar opposites can be resolved, the hierarchy can be obscured. The union of bipolar opposites in sexuality, or the complementarity of marriage and domestic roles, presents an image in which the aspect of hierarchy is masked. Yet the masculine and the feminine are not equals, and the shadow which this inequality casts can be observed in the disputes which break out over access to resources, and in the

Representations, identities, resistance 191 psychological patterns of overvaluing same gender group and devaluing the opposite gender. For the girls, of course, the reproduction of the hierarchy of gender also brings with it the devaluation of their own gender. It is perhaps not surprising that it is among girls that we have seen evidence of a break with the hegemony of a strict bipolar opposition. From this point of view the girls’ refusal to attribute as feminine any socially undesirable behaviors in one of our interview measures can be seen as one such indication. In our ethnography we also recorded a number of episodes of girls challenging this image, usually by competing with boys for masculine marked resources. In comparison to the dominance of an image of gender as a bipolar opposition, such examples are marginal to the general run of life in the reception class. Yet they serve to illustrate that even among the children gender is not an entirely uncontested terrain, even here relations of power can generate resistance. Indeed, in so far as any identity is as much a system of exclusion as of inclusion, identity formation always implies the prospect of points of resistance.

Identities and resistance Thus as well as enabling individuals to sustain a stable sense of themselves and the world they inhabit, identities also project individuals into a social world marked by a complex set of relationships between social groups. The idea of identifications as a form of positioning of the self in relation to representations also needs to take account of the complex dynamics which may be involved. Wherever representations are internalized they are linked to a process of identity formation, which can sometimes take surprising forms. For example, we do not usually think of children’s internalization of representations of mathematics as being linked to specific social identities, but this can indeed be the case. When the form of mathematics which children internalize is linked to their identity as the member of a marginalized social group, this can lead to a disruptive relationship in their schooling, and it is only when we see the consequences of difficulties and failures in school that the sense in which representations of mathematics also express a social identity becomes apparent (de Abreu, 1993, 1995). If the relationship between representation and identity is usually opaque in the field of mathematics, it can nevertheless become clear in some contexts. The pervasiveness of variations and differences associated with gender ensure that the relationship between representations and identities is clear across a very wide range of contexts. That this should be so is due to the significance of gender as a dimension of power in the social world. In recent years identity in social psychology has become more of a formula than a concept – an operating variable in the calculus of experimental designs. This is not the fault – or the inheritance – of Henri Tajfel (1981) who originally proposed what has become known as social identity theory. Tajfel himself always strove to make identity a concept that could bear the weight of the presence of the individual in the social world, a complex, problematic and sometimes contradictory presence. In all Tajfel’s work we can see the struggle to connect forms of

192


social psychological thought and explanation to the complexities of social life and social experience, and without the weight and concrete struggle for this presence his work would have been quite different. In too much of the work around the theme of social identity since Tajfel’s death there is little sign of this struggle continuing. The forms of the theory initiated by Tajfel have persisted and even been extended, but too often they have been empty forms so that identity has become a vacuous notion, a transparent glass case enclosing nothing. Yet Tajfel’s work itself cannot be said to have been successful in its struggle. By focusing on processes with the hope that content would take care of itself in the concrete particularity of specific cases, his work opened up the possibilities which John Turner (1987) and others have developed. Identity is a process – of course it is, and no doubt the consequences of categorization can be seen in the processes of identity. But to initiate the discussion of identity in the motivational drive to acquire a positive self-concept is to start from the wrong place. Before it becomes thematized as a struggle for the individual, an identity is first a social location, a space made available within the representational structures of the social world. It is this which gives categorizations their power, not categorizations which determine identities. The fundamental problem with social identity theory is that it offers a theory of the consequences of categorization, but is mute on the question of why individuals should categorize themselves in particular ways. Why is it that young children come to categorize themselves in terms of gender? The answer might seem to be extraordinarily obvious – gender is one of the central dimensions of power in our societies, and as such is articulated in a representational field which not merely surrounds the child from the moment of their birth, but can even predate their conception in the hopes, wishes, fears, and anxieties of their parents. But if we follow social identity theory we find there is no means for grasping this priority of representations over identities. To paraphrase a famous dictum of Sartre’s we can say that representations precede identities. And just as in Sartre’s analysis, essence is a projective work of experience, so we can say that identities take shape through the engagement of the individual in the world of representations. In fact even the term “individual” is problematic in this discussion. The person who takes on an identity is himself or herself a product of a representational system. As Mauss (1985), Geertz (1993, first published 1973), Shweder (1991), and others have shown, the category of the person is a cultural construction, a point which actually reinforces the argument that representations precede identities. To be sure there is a dialectic of representations and identities in the life of individuals and groups. As Lucien Goldmann (1976) saw so penetratingly, the identities which emerge in the course of development constrain the representations which individuals or groups might accept. In his terms the limiting case was one where the conditions for the acceptance of a new representation entailed the dissolution of an existing identity – which means change for the individual, or disbandonment, schism, or reorganization for the group. Identity, then, is not some thing, like a particular attitude or belief, it is the force or power which attaches a person or a group to an attitude or a belief, in a word, to a representation.

Representations, identities, resistance 193 Goldmann’s argument is important since it links identity to communication, to what it is possible to communicate in a relationship, and what is incommunicable, or where communication itself can lead to change and reorganization. The stability of particular forms of identity is therefore also linked to the stability of the network of social influences which sustain a particular representation. As the balance of influence processes change, so too does the predominant representation, and consequently the patterns of identity which are a function of that representation (cf. Duveen, 1998). We can then consider identity as an asymmetry in a relationship which constrains what can be communicated through it – both in the sense of what it becomes possible to communicate and in the sense of what becomes incommunicable (and potentially a point of resistance), or communicable only on condition of a reworking of that identity. Resistance, then, is the point where an identity refuses to accept what is proposed by a communicative act, that is, it refuses to accept an attempt at influence. Points, or moments of resistance, can remain limited within the immediate contexts in which they occur. Or they can also develop into a broader social response, and escape from the limiting horizons of particular exchanges and become linked to a coordinated and constructive attempt at influencing the pattern of social thought. Resistance which occurs first in the microgenetic evocation of social representations (cf. Duveen & Lloyd, 1990), can lead both to ontogenetic transformations (where identities themselves are restructured) and to sociogenetic change (where resistance becomes first a resistance to a change in identity, and then linked to an effort to influence the wider social world to recognize that identity).

Conclusion – varieties of social identities I began this discussion by referring to William James’s notion of the multiplicity of social selves, and I want to conclude by returning to this theme from the perspective I have outlined in which identity is considered a function of social representations. But, as Barbara Lloyd and I (Duveen & Lloyd, 1990) have suggested, we can distinguish between different types of social identity, or more precisely between different types of relation between representations and identities. On the one hand, there are social representations which impose an imperative obligation on individuals to adopt a particular identity, that is circumstances where there is an “external obligation which derives from the ways in which others identify an individual in terms of particular social categories” (Duveen & Lloyd, 1990, p. 7). This is the case with the gender identities I have been discussing. On the other hand, however, the influence of social representations is exercised through a contractual obligation rather than an imperative one, so that an individual voluntarily joining a social group contracts to take on a particular social identity. This distinction between imperative and contractual obligations as forms of identity may not exhaust all the possibilities, but it serves to introduce a discussion of differences and variety in forms of identity. Indeed, as Deaux, amongst others, has recently emphasized, a focus on the contents of identities leads to a consideration of

194


variations in the form of identities and the different consequences which may follow for thought, feeling and action (Reid & Deaux, 1996; Deaux, Reid, Mizrahi, & Ethier, 1995). The implication here is that we need to extend the analysis of the relationship of representations and identities to provide a frame for understanding varieties of social identities, a theme which would also enlarge the scope of the discussion of social identity in contemporary social psychology.

Notes This paper was originally presented at the conference in New York in October 1998, and I am grateful to the British Academy for their support for my participation in this conference through the award of an Overseas Conference Grant. I should also like to thank Kay Deaux and Gina Philogène for their helpful editorial comments. Soon after the completion of this chapter I learned of the death of Hans Furth in November 1999. He was my first doctoral supervisor and I benefited not only from his unparalleled grasp of Piaget’s genetic epistemology but also from the example of his intellectual rigour and integrity. I would like to dedicate this chapter to his memory. 1 In these studies the others were always women who were themselves mothers, a choice dictated by the contingencies of organizing laboratory studies rather than any theoretical presupposition. There is no reason to suppose that such effects are peculiar to women rather than men, nor to adults rather than older children. 2 The gender marking of these toys had been established by asking a sample of adults to select toys which were appropriate for either a girl or a boy.

References Breakwell, G. (1979) Women: Group and identity. Women’s Studies International Quarterly, 2, 9–17. Corsaro, W. (1990) The underlife of the nursery school: Young children’s social representations of adult rules. In G. Duveen and B. Lloyd (Eds.) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Class. de Abreu, G. (1993) The Relationship Between Home and School Mathematics in a Farming Community in Rural Brazil. PhD thesis, University of Cambridge de Abreu, G. (1995) Understanding how children experience the relationship between home and school mathematics. Mind, Culture and Activity, 2, 119–142. De Rosa, A. (1987) The social representation of mental illness in children and adults. In W. Doise and S. Moscovici (Eds.) Current Issues in European Social Psychology, Volume 2. Cambridge: Cambridge University Press. Deaux, K., Reid, A., Mizrahi, K. and Ethier, K. (1995) Parameters of social identity. Journal of Personality and Social Psychology, 68, 280–291. Deschamps, J.C. (1982) Social identity and relations of power between groups. In H. Tajfel (Ed.) Social Identity and Intergroup Relations. Cambridge: Cambridge University Press. Duveen, G. (1994) Crianças enquanto atores sociais. In S. Jovchelovitch and P. Guareschi (Eds.) Textos em representações sociais. Petrópolis, Brazil: Vozes. Duveen, G. (1997) Psychological development as a social process. In L. Smith, J. Dockrell and P. Tomlinson (Eds.) Piaget, Vygotsky and Beyond (pp. 67–90). London: Routledge. Duveen, G. (1998) The psychosocial production of knowledge: Social representations and psychologic. Culture and Psychology, 4, 455–472.

Representations, identities, resistance 195 Duveen, G. and Lloyd, B. (1986) The significance of social identities. British Journal of Social Psychology, 25, 219–230. Duveen, G. and Lloyd, B. (1990) Introduction. In G. Duveen and B. Lloyd (Eds.) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Furth, H.G. (1996) Desire for Society. London: Plenum Press. Geertz, C. (1993) The Interpretation of Cultures. London: Fontana Books. Goldmann, L. (1976) Cultural Creation in Modern Society. Saint Louis: Telos Press. James, W. (1981). The Principles of Psychology. Cambridge, MA: Harvard University Press. Lloyd, B. and Duveen, G. (1989) The reconstruction of social knowledge in the transition from sensorimotor to conceptual activity: The gender system. In A. Gellaty, D. Rogers and J. Sloboda (Eds.) Cognition and Social Worlds. Oxford: Oxford University Press. Lloyd, B. and Duveen, G. (1990) A semiotic analysis of the development of social representation of gender. In G. Duveen and B. Lloyd (Eds.) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Lloyd, B. and Duveen, G. (1992) Gender Identities and Education. London: Harvester Wheatsheaf. Marx, K. (1970). Capital, Volume I. London: Lawrence and Wishart. Mauss, M. (1985) A category of the human mind: The notion of the person; the notion of self. In M. Carrithers, S. Collins, and S. Lukes (Eds.) The Category of the Person (pp. 1–25). Cambridge: Cambridge University Press. Moscovici, S. (1981). On social representation. In J. Forgas (Ed.) Social Cognition. London: Academic Press. Oyserman, D. and Markus, H. (1998) The self as social representation. In U. Flick (Ed.) The Psychology of the Social. Cambridge: Cambridge University Press. Reid, A. and Deaux, K. (1996). Relationship between social and personal identities: Segregation or integration? Journal of Personality and Social Psychology, 71, 1084–1091. Rubin, J.Z., Provenzano, F.J. and Luria, Z. (1974) The eye of the beholder: Parent’s views on the sex of newborns. American Journal of Orthopsychiatry, 44, 512–519. Sartre, J.P. (1974) Between Existentialism and Marxism. London: New Left Review of Books. Shweder, R.A. (1991) Thinking through Cultures: Expeditions in Cultural Psychology. Cambridge, MA: Harvard University Press. Smith, C. (1982). Mothers’ Attitudes and Behaviour with Babies and the Development of Sex-typed Play. Unpublished doctoral dissertation, University of Sussex. Smith, C. and Lloyd, B. (1978). Maternal behavior and perceived sex of infant: Revisited. Child Development, 49, 1263–1265. Tajfel, H. (1981) Human Groups and Social Categories: Studies in Social Psychology. Cambridge: Cambridge University Press. Turner, J.C., Hogg, M.A., Oakes, P.J., Reicher, S.D., and Wetherell, M. (1987) Rediscovering the Social Group: A Self-Categorisation Theory. Oxford: Blackwell-Publishers.

13 Culture and social representations Gerard Duveen

The main aim of the theory of social representations is clear. By focusing on everyday communication and thinking, it hopes to determine the link between human psychology and modern social and cultural trends. (Moscovici, 1988, p. 225)

Culture as a field of representations Since the inception of work on social representations nearly half a century ago, there has been a persistent interest in the articulation between these representations and the broader field of culture, as Anne Parsons’ (1969) pioneering studies indicate. In part, her investigations explored the forms of systematic misunderstanding that arise when expressions of mental states grounded in one specific culture are projected into a different cultural context. In one of her studies, she examines the reasons for the failure of her attempted treatment of a South Italian immigrant to the United States through a psychoanalytically based psychotherapy. In her reflections she focuses on her failures in grasping Mr Calabrese’s communicative actions, both in the sense of understanding what he was saying in relation to his own cultural context (including the specific context of being an immigrant to the United States), and in the sense of understanding how his cultural resources were being employed to make sense of what she said and did. Such forms of misunderstanding are now familiar, but they can be no less instructive for that. Parsons herself explores the ways in which different meanings are constructed or connoted in the course of such exchanges, and as she demonstrates, the meanings of utterances always refer back to a very specific context. While she herself does not elaborate a semiotic analysis of the context, we could extend her analysis by noting that utterances are always particular instances of forms of semiotic mediation, and that the context which is so important for their comprehension is always the broader network of mediational forms which enable the signifiers of particular signs to be related to their signifieds. As we have argued before (Lloyd & Duveen, 1990), one of the functions of representations is precisely to provide the framework through which signifiers and signifieds can be associated in a meaningful way, to reduce what Saussure called the arbitrariness of the sign. Such a perspective also allows us

Culture and social representations 197 to comprehend how misunderstanding arises when signifiers are associated with signifieds through a different representational context. Anne Parsons, for instance, records the conflict that arose around Mr Calabrese’s hostile feelings towards his wife. From her permissive and professional perspective, such feelings should have been a focus of therapeutic concern, while from his perspective the norms were different, “even if a marriage is difficult you keep quiet about it in order to preserve the institution of marriage as such” (Parsons, 1969, pp. 327–8). But what has just been said about representations could just as easily be said of culture, indeed one might define culture as the totality of forms of semiotic mediations and their associated practices available within a community. At the beginning of her recent essay on Social Representations in the Field of Culture Denise Jodelet (2002) remarks on the significance of Durkheim as a common source for both the notions of social or collective representations and of culture as a concept in the social sciences. As she notes, Durkheim’s sources for his work on representations were ethnographic materials related to traditional societies, and his work had, as she puts it, “posed the question of the relation between the individual and the collective in the functioning of thought” (Jodelet, 2002, p. 112, my translation). Within the Durkheimian tradition, which has been one of the central sources not only for the theory of social representations but also for much thinking about the notion of culture itself, society can be understood as consisting of representations. But is this common ancestry sufficient to justify an elision between these two terms and consider “culture” and “social representations” as synonymous? Even to pose the question in this way suggests the absurdity of such an idea, with its clashing category error. And yet without wishing to sustain any such equivalence between these terms, it is nevertheless interesting to observe two complementary movements within recent thinking about both these concepts which underlines not only their common ancestry but also increasingly common fields of interest, one might even say frames of analysis. On the one hand, recent research on the theory of social representations has emphasized the need both to distinguish between levels and types of representations, as well as the significance of their thematic origins (cf Moscovici, 1988, 2000; Marková, 2003). On the other hand, recent work in cultural psychology has drawn from the Vygotskyian perspective a concern with the cultural framing of everyday interactions between people as the focus for analyzing the forms of semiotic mediation which are seen as the expression of cultural forms in psychological life. On the one hand, then, the theory of social representations has been reaching back to connect with some of the basic and fundamental structures (which Moscovici, following Holton, has described as themata) which could be said to characterize culture in its broadest and most general sense, while on the other, recent cultural psychology has been concerned with analyzing the close-at-hand as the arena in which cultural forms are most accessible. Given these complementary movements in recent research, it is hardly surprising that the relations between culture and social representations have been a focus for a number of important contributions (cf Jodelet, 2002; Valsiner, 2003; Valsiner and Van der Veer, 2000). If these emerging common points of interest

198


and concern hold the promise of a productive engagement between cultural psychology and the theory of social representations, we should also note that while the terms “culture” and “social representations” appear to refer to different levels of analysis, nevertheless whatever it is that we take to be connoted by the term culture only becomes accessible through the observation and analysis of specific representations. But again, this methodological note is not sufficient to elide the difference between these two terms. The crucial difference remains one of scale and scope. As an analytical concept, the term “social representation” carries a dual meaning. On the one hand, it refers to a system of values, ideas, and practices related to a specific object, while also referring to the process through which such representations are formed, a sense more closely conveyed in English by social representing (cf Duveen & Lloyd, 1990). This latter meaning, with its focus on process, already conveys a more dynamic sense than is often found in the use of the term culture. But representations (both in the substantive sense as well as process) always occur within a broader context of other representations which may also contribute towards the ways in which meanings become organized within specific representations. For instance, in her study of representations of Zen Buddhism in both Japanese and British contexts, Saito (1996) notes that Japanese respondents (even those not active practitioners) produced a coherent, well-structured account of this aspect of their local culture, while the British followers of Zen produced more fragmented images which were relatively isolated from other social practices within their local culture. What formed a part of the everyday experience and context for the Japanese appears differently in the British context where it has a more exotic quality precisely because it has been differentiated from more familiar everyday experience. Culture, then, can be taken as referring to a broader network of representations held together as an organized whole by a community. Social representations, in this sense, can be seen as particular cultural forms, and the analysis of social representations will always refer back in some way to the cultural context in which they take shape. I have already illustrated this idea in relation to Anne Parsons’s work, but one might also consider the way in which Moscovici (1976) traces the emergence of representations of psychoanalysis within different social groups in France to the particular contexts in which they are generated. Indeed we could go further and suggest that it is through changing social representations that cultures themselves undergo change and transformation. Not only is it the case that all representations – even those which seem most deeply embedded in our culture – can at some point become the active focus of representational work leading to their transformation, but also what we might call cultural representations can themselves change through influences operating within the communicative practices of a community, that is, sociogenesis (cf Duveen and Lloyd, 1990) may lead to cultural change.

Types of social representations For Moscovici social representations are considered as the form of collective ideation which has appeared in the context of the modern world. Whereas

Culture and social representations 199 pre-modern civilizations are generally characterized by unitary structures of power, authority and legitimation, the modern world is, rather, characterized by a diversity of forms of belief, understanding, and practice in which different social groups construct their own understanding of social processes and social life, in short, their own representations which may not only distinguish one group from another, but can also be the source of conflicts between them. Thus in relating his theory to the work of Durkheim he is not so much concerned with the terminological question of whether these representations are more accurately described as “social” or “collective,” as in distinguishing the modes of construction and functioning of representations in the modern world. Representations are the products of patterns of communication within social groups and across society as a whole, and thus, importantly, are also susceptible to change and transformation. While acknowledging the significance of the Durkheimian concept of collective representations, he also marks his distance from the French sociologist by describing his concept as too static (Moscovici, 1984, p.17; cf also the discussion in Duveen, 2000), referring to a stable and settled order within a society. By introducing the idea of social representations he aims to capture the dynamic processes of change and transformation in the representations which circulate in the modern world. From this point of view, stability is only ever provisional, reflecting a particular moment in a more general process of transformation in which the social influences embedded in patterns of communication achieve a certain balance and closure. Expressing a similar idea in a slightly different context Piaget wrote that “sooner or later reality comes to be seen as consisting of a system of transformations beneath the appearance of things” (Piaget & Inhelder, 1971, p. xiii). Or to paraphrase Karl Marx, we might say that all that is solid can melt into air, and then re-crystallize in a different form. Moscovici defines a social representation as: a system of values, ideas and practices with a twofold function; first to establish an order which will enable individuals to orient themselves in their material and social world and to master it; and secondly to enable communication to take place among the members of a community by providing them with a code for social exchange and a code for naming and classifying unambiguously the various aspects of their world and their individual and group history. (Moscovici, 1973, p. xiii) We can, then, think of social representations as structures of semiotic processes. However, while the socio-cultural tradition stemming from the work of Vygotsky has emphasized the importance of semiotic mediation as the process through which sign usage organizes psychological activities, Moscovici’s attention is not so much focused on the way in which individual signs operate as with the question of how ensembles of signs are held together in a structured and organized way so as to constitute a particular image of an aspect of social reality in which the arbitrariness of signs can be reduced and meaning and reference secured for a particular community. Social representations, then, are collective structures which are

200


both established through communication as well as enabling communication to take place among members of a social group through the exchange of signs with common or shared meanings. A corollary to this definition is that the realities in which we live are constituted by social representations. Defined simply as structures in this way, however, we would not be able to distinguish between different types of social representation, between, for example, a common idea emerging from the informal talk among a group of friends at a table in a café and the more pervasive and resilient representations of madness which exclude and isolate the mad. If there is a need to distinguish between different types of social representations, there is, as yet, no clear and settled means of identifying which aspects or dimensions of social representations might enable such distinctions to be made. This remains a rather open question within the theory, and indeed, Moscovici himself has offered more than one suggestion for how this might be achieved. Perhaps the most familiar of Moscovici’s suggestions is his proposal to distinguish between the consensual and the reified universes, which he sets out in the following terms: The division into the consensual and the reified categories is a distinctive feature of our culture. In the former, society recognizes itself as a visible, continuous creation which is imbued with meaning and aims; it speaks with a human voice, is part and parcel of our lives and acts and reacts like a human being. In short, man is the measure of all things. In the latter, which comprise solid, fundamental, immutable entities and where particularities and individual identities are disregarded society fails to recognize itself and its works, which appear to it under the guise of isolated objects. In as much as the scientific disciplines are linked to these objects, scientific authority is able to impose this way of thinking and experiencing on each of us, prescribing in each case what is and what is not true. Under such circumstances, things thus become the measure of man. The contrast between these two universes is psychologically powerful. The border between them splits collective reality, even physical reality, in two. Obviously science is the mode of knowledge corresponding to the reified universes and social representations the one corresponding to the consensual universes. The former attempts to construct a map of the forces, objects and events unaffected by our desires and consciousness. The latter stimulates and shapes our collective consciousness, explaining things and events so as to be accessible to each of us and relevant to our immediate concerns. (Moscovici, 1981, pp. 186–7) While this distinction remains interesting for the way it articulates how modern societies construct a specific way of representing different types of knowledge (cf. Duveen & Lloyd, 1990), nevertheless as a general proposal for distinguishing between types of social representation this proposal remains too limited. Partly because it derives from Moscovici’s original interest in the processes through

Culture and social representations 201 which scientific knowledge becomes absorbed into the world of everyday life (which was the inspiration for his pioneering study of social representations of psychoanalysis, Moscovici, 1976), and partly because it does not address questions about how different types of social representation might differ in their structure and functioning. A second, more recent proposal is presented in the context of another broader discussion, this time of the relations between the so-called primitive mentality and the mentality of modern societies. Here Moscovici proposes that: A great deal of ink has been poured over this difference between a “primitive mentality” and a “civilized” or “scientific” mentality. In fact it seems to me to refer to the difference between belief and knowledge, so important but so little understood, as can be established by reading Wittgenstein’s (1953) late reflections on belief. In my opinion a great many misunderstandings would be dispelled if the following suggestion were to be accepted: The difference with which we are concerned takes on a new meaning when we pay attention to the distinction between: (a) common representations whose kernel consists of beliefs which are generally more homogenous, affective, impermeable to experience or contradiction, and leave little scope for individual variations; and (b) common representations founded on knowledge which are more fluid, pragmatic, amenable to the proof of success or failure, and leave a certain latitude to language, experience, and even to the critical faculties of individuals. (Moscovici, 2000, p. 136) This distinction already has a greater utility insofar as it offers a clearer characterization of the qualities which distinguish at least these two types of representations. Although in this chapter Moscovici does not immediately extend this argument, it would be in keeping with his general approach to add that this distinction between representations based on belief and those based on knowledge is not limited to a distinction between the mentalities of different epochs or of different levels of society, but, rather, can be seen as a distinction between types of representation which circulate today within our own cultures (cf Moscovici’s comments on Bartlett’s discussion of Lévy-Bruhl, Moscovici, 1990). Yet this distinction, too, is also limited, primarily because it does not yet include any clear discussion of the functional aspects of these representations, of the modalities through which they circulate or are communicated, or the ways in which they serve to structure different types of social groups, or may be structured by different types of social relations. The proposal in which these functional aspects of representations become most clearly visible is to be found in Moscovici’s (1988) response to Jahoda’s (1988) critical commentary on the theory of social representations itself. One point which Jahoda emphasizes is what he sees as a lack of clarity in Moscovici’s discussion of the relations between social representations and social groups. In responding to this argument, Moscovici suggests a tripartite distinction between:

202

Thinking through social representations (i)

Hegemonic representations can become shared by all members of a highly structured group – a party, a city, a nation – without their having been produced by the group. (ii) Emancipated representations are the outgrowth of the circulation of knowledge and ideas belonging to subgroups that are in more or less close contact. (iii) Polemical representations are generated in the course of social conflict, social controversy and society as a whole does not share them. They are determined by the antagonistic relations between its members and intended to be mutually exclusive. (Mosocvici, 1988) What Mosocvici describes as hegemonic representations correspond to the representational forms closest to those described by Durkheim as collective, and could also be said to be largely representations based on beliefs, since they consist of patterns of values and ideas deeply embedded in the practices of everyday life. As such, these representations are not only extraordinarily stable and resistant to change, but also only rarely become the focus of any sustained reflection, remaining more or less closed systems of meaning. They constitute, as it were, the unreflexive assumptions of a particular form of life, the fraglos gegeben as Schütz (1972) described them. On the other hand, both emancipated and polemical representations imply a degree of reflection, of discussion, of argument. In both cases, these representations embrace ideational and evaluative positions which need to be defended and legitimized, while at the same time securing a position within the symbolic world of a society which can provide the locus for the social identity of the group affiliated around these ideas and values. The need to secure legitimation for these positions necessarily means that these representations must be more open to argument or debate. The forms of discourse through which legitimation can be secured may be very varied, from the rigorous procedures of the natural sciences to the peculiar inverted logic of a more or less hermetically sealed religious sect. No doubt these different patterns of legitimation correspond to representations with very different types of presence within the social world, but insofar as they share a certain openness they also tend towards being constituted as representations based on knowledge (even if what is taken as constituting knowledge may also be as varied as the different patterns of legitimation through which it can be secured). What characterizes these representations, then, is a certain activity of reflection as a social group seeks to establish a sense of closure by legitimizing the view of the world embodied by a particular representation.

Reflection in the genesis of representations What I have described here as an activity of reflection is, in fact, a key characteristic of social representations. When he considers the origin of representations, Moscovici notes that the “purpose of all representations is to make the unfamiliar, or unfamiliarity itself, familiar” (2000, p. 37). It is within this context of a dynamic

Culture and social representations 203 of familiarization that he introduces the concepts of anchoring and objectification as the central processes in the genesis of social representations. It is an elegantly simple idea. Something (an idea, a person, a group, a phenomenon) appears which is initially opaque to the existing structures of meaning, and for this very reason is also troubling. In seeking to give the unfamiliar a place within the known world, some kind of representational work, or activity of reflection, is required, and the traces of this activity can be observed in the communications which circulate around this unfamiliar object. At times, of course, unfamiliarity can be very dramatic and even engender considerable change within existing structures of meaning as it becomes familiarized. If we think, for instance, of the fall of the Berlin Wall we can see both the sense of the dramatic emergence of the unfamiliar as well as the profound transformations in structures of meaning which had seemed, at least since the close of World War II, extraordinarily stable and resistant to change. Or again, if we consider the emergence of HIV/AIDS we can recognize successive transformations in its representation from being a “gay plague” to a disease controllable by medication as it has become anchored within different structures of meaning (cf Preda, 2005). Both of these examples illustrate the centrality of representational work or the activity of reflection in the genesis of representations as the unfamiliar is familiarized. Indeed, we could say that what characterizes emancipated or polemical representations as distinct from hegemonic representations is precisely that they are representations which are active centres of reflection in this sense. So familiar were the geo-political divisions of the Cold War, at least in the popular mind, that for decades they seemed part of the given structure of the world. The collapse of the Wall, and the processes which followed, had the effect of bringing this stable and settled representation into question so that they became centres of activity as new representations have emerged. The representational work engendered by the unfamiliar is of course focused on the establishment of a more familiar and stable framing of the object. But this is only one part of the process, for this new understanding still has to be legitimized and, perhaps, defended against alternative interpretations. The activity of reflection evident in the communications around emancipated and polemical representations embraces both aspects of this representational work.

Culture as representation and the representation of culture Culture is sometimes considered as the discriminating principle through which all members of a community are alike in sharing some set of beliefs, values, and practices, and different from other communities which have their own sets of beliefs, values, and practices. While such a view may be helpful to some degree, it can also be a hindrance. First, it presents culture as a categorical phenomenon, which, like all such phenomena, tends to emphasize differences between cultures while minimizing variations within cultures. Each culture is viewed as though it were a homogenous entity, free of internal division. But as the discussion of social representations has already indicated, such a view is unsustainable. Societies,

204


certainly modern societies but perhaps also other societies, are always marked by internal divisions, to the point where we might find some justification for claiming that society is a machine for producing difference. Nor can this issue be avoided by suggesting that the term culture refers only to hegemonic representations, since it would be absurd to consider emancipated or polemical representations as somehow not also being cultural forms. Second, this categorical view of culture can also be a hindrance if it is taken as suggesting that the point of demarcation between one culture and another is a fixed or rigid point. As Bartlett (1923, 1932) pointed out many years ago, even what used to be called primitive cultures were communities which were also often engaged in a variety of forms of contact with other communities rather than being isolated from them. And from that contact new representational elements enter into the life of cultures, which may then be incorporated into their own representational structures. What Bartlett describes as a process of conventionalization is also a process of cultural change. A degree of fluidity between cultures is more the norm than the exception. In the modern world this fluidity can produce seemingly paradoxical situations for a researcher, since what are often presented as comparisons across cultures turn out on closer inspection to be rather studies of how a single more globalized culture is spreading across the world. If people engaged in the market economies of the West turn out to have similar ideas about economic processes to the people engaged in similar activities in the East, is this really a comparison across cultures, or simply an indication that in some important sectors of life both West and East find themselves participating in a common culture? But this is not to say that there are not real boundaries or borders between cultures – regions where horizons emerge and where some form of representational work or negotiation is required. While there may be a fluidity to cultures, this does not mean that any cultural element can be substituted for any other cultural element. There are points where communication only becomes possible if it engenders a change of some kind. Or to put this another way – a cultural identity is revealed at the point where something can no longer be communicated. As Lucien Goldmann (1976) saw so penetratingly, the identities which emerge in the course of development constrain the representations which individuals or groups might accept. In his terms, the limiting case was one where the conditions for the acceptance of a new representation entailed the dissolution of an existing identity – which means change for the individual, or disbandonment, schism or re-organization for the group. Goldmann’s argument is important since it links identity to communication, to what it is possible to communicate in a relationship, and what is incommunicable, or where communication itself can lead to change and re-organization. The stability of particular forms of identity is therefore also linked to the stability of the network of social influences which sustain a particular representation – as the balance of influence processes changes so too does the predominant representation, and consequently the patterns of identity which are a function of that representation (cf Duveen, 1998). We can then consider identity as an asymmetry in a

Culture and social representations 205 relationship which constrains what can be communicated through it – both in the sense of what it becomes possible to communicate and in the sense of what becomes incommunicable (and potentially a point of resistance), or communicable only on condition of a reworking of that identity (cf Duveen, 2001). Cultural identities, then, become evident at the point where communication becomes problematic; it is when the implicit sense of culture can no longer be taken for granted that the explicit search for culture begins. This is the point where boundaries or horizons between cultures emerge. For the social sciences, the corollary to this view is that culture cannot be assumed to be a given property of a community which marks out a difference from another community. Rather, difference or similarity is something which needs to be established and articulated through the analysis of specific and particular situations.

Culture and time Culture in the sense in which it is being used here must not only be seen as having structure, but as a structure which has a functional role in relation to the community which is sustained by it, and which also sustains the culture. For the community, its culture is a stabilizing element, and as a structure it acquires stability because it is sustained within a context of some kind – a context of other cultures or the history of the culture itself. The community which sustains a culture, like every other social group, defines itself in part through the oppositions it generates in relation to other communities. “We are like this, whereas they are like that.” But as well as establishing stability in the present, for a culture to sustain a community it also needs to account for the past through which the present has emerged, and project a future to which the present is oriented. While there has been considerable interest in the ways in which social groups and communities represent their past in their collective memories, the future as an orienting perspective for a group has rarely figured in the social psychological literature. Indeed, for the most part, social psychological conceptualizations of the group remain largely static, or, one might even say, timeless, that is, they are seen to exist almost as though they were outside the dimension of time. There is something quite astonishing in the neglect of the dimension of time in the social psychology of the group or of inter-group relations, and it is surely a theme which deserves more attention than it can be given here. For the moment, one can only note its absence in almost every social psychological account of the group. A recent, and rare, exception is in the work of Gina Philogène (2001) who introduces the notion of anticipatory social representations in her account of the changing nomination from black to African American, in which representations formed in the present may anticipate expected (or desired) futures. But there is also an earlier contribution to this theme in the work of Bartlett, which is more immediately germane to the present discussion of the relations between culture and social representations. Bartlett’s early work in psychology was much concerned with questions of culture, or more particularly with analyzing the psychological aspects of cultural processes. In his first book, Psychology and Primitive Culture (1923), Bartlett sets

206


out to analyze the specifically social psychological aspects of what he (along with most other writers of this time) reluctantly accepts to describe as “primitive” culture – reluctantly because, of course, the word primitive not only inevitably carries with it pejorative overtones, especially when used by an academic representative of a dominant society about a dominated society, but also carries implications of simplicity and lack of sophistication which simply do not do justice to the richness, complexity, and subtlety of cultural forms in what we would now describe as non-literate societies. Central to Bartlett’s account is his treatment of primitive cultures not as the product of a more or less isolated social group which sustains a hegemonic and unchanging culture, but rather he considers such societies as being always in touch with – and hence influenced by their contact with – other social groups, and the culture they sustain as being also something which changes, whether externally through these contacts with other social groups or internally through what he calls the constructiveness of social responses, by which he means the “tendency which produces new modes of social organisation” (Bartlett, 1923, p. 29). As he considers the question of cultural change effected through the contact between peoples, Bartlett also comes to focus more specifically on what he describes as the “group difference tendencies,” which differentiate one group from another, and which “cluster about a group’s established institutions and act directly as determining factors of individual social behaviour” (Bartlett, 1923, p. 29). Such group difference tendencies give a direction to the ways in which a group may take up and elaborate novel cultural elements introduced through their contacts with other people, indeed he goes on to argue that “how, as well as what, material is received through contact depends to a considerable degree upon the operation of these differentiating tendencies of social groups” (Bartlett, 1923, p. 155). In his later, more familiar book on Remembering (1932), Bartlett returns to this theme in his discussion of conventionalization, by which he means the process through which “cultural material . . . introduced into a group from the outside . . . suffers change until it eventually either disappears or reaches a new stable form” (Bartlett, 1932, p. 268). In this later, slightly more elaborated version of his argument, what Bartlett described as group difference tendencies in his earlier book have been subsumed into his account of social constructiveness, considered as the most interesting form of conventionalization (the others being assimilation, simplification and retention of details). It is in his discussion of social constructiveness that Bartlett offers one of his most radical suggestions for social psychology. “Every well-established social group,” he writes, “possesses not only a structure which has been built up in its past, but also a function, or a group of functions, within the community of which it is a part. These functions have to be expressed in co-ordinated human activity, and all such activity has not only a history, but also a prospect [my emphasis]” (1932, p. 275). So clear is this prospect, that one can say of a group that “it inevitably tends to develop in certain more or less specific directions; and if we know enough, we can state in some detail the paths along which it is tending” (1932, p. 275). And he continues by

Culture and social representations 207 suggesting that “when any cultural features come from outside, they may be transformed, not only by assimilation, by simplification and elaboration, and by the retention of apparently unimportant elements, but positively in the direction along which the group happens to be developing at the time these features are introduced,” so that “the imported elements change, both in the direction of existing culture and along the general line of development of the receptive group” (1932, p. 275). For Bartlett, the idea of social constructiveness expresses his conviction that social groups do not simply accept or receive cultural elements from outside, but actively transform them to produce new cultural elements. But this activity of transformation is undertaken in relation to the existing structure of the group, so that elements are drawn into existing projects and reshaped or reworked to make them of service to these projects. And as the quotations from his text indicate, for Bartlett to understand this process it is necessary that we construe social groups having, as he puts it, a prospect, a direction of development which guides and shapes the absorption and transformation of elements encountered through the groups’ exchanges or contacts with other groups. In this sense, Bartlett’s notion of the social group is not that of a static organization of social relations, values, and practices, but rather he sees the social group as having a dynamic organization in which culture not only provides a stable image of the present and its relation to the past but also projects the group into a future. We might envisage this notion of the group in terms of a cultural vector which by coordinating past, present and future gives shape and meaning to the life of a group, that is, furnishes the group with a project which determines not only aspects of the behavior and practice of members of the group but also the ways in which environmental influences are transformed as they become part of the group’s repertoire of resources. In short, Bartlett’s work can be seen as an argument for the necessity of incorporating the dimension of time into our understanding of the social group, and hence also into our considerations of the relations between culture and representations.

Time, change, and cognitive polyphasia In Moscovici’s work, the theory of social representations has always been oriented to the analysis of social, or cultural, change. That is, he envisions social representations as dynamic structures in which knowledge is constantly being transformed as social groups construct and re-construct their grasp of the social world and their place within it. Like every genetic psychology, the theory of social representations is oriented around the dimension of time as a fundamental parameter of social processes. As I noted earlier, particular representations are only ever the product of the balance of influence processes at a specific point in time. As this balance changes, so too will the representations sustained by these communicative processes. Yet while these points are theoretically clear, there has been relatively little empirical analysis of how representations change. There have been studies in the ontogenesis of social representations (e.g., Lloyd and Duveen, 1990;

208


Psaltis and Duveen, 2006) which have explored the dynamics of specific processes of change within the psychological development of individuals. In these studies children are investigated as they develop within a social world structured by a stable set of representations within their communities, and something similar can be seen in other developmental research within the field of cultural psychology (e.g., Lave, 1988; Rogoff, 1990), even if in these cases it is a stable set of social practices which is considered to provide the structure within which children’s developing competencies take shape. While the study of ontogenesis is a crucial arena for the investigation of specific types of representational change in which the orientation to time is quite clear, it nevertheless rarely reaches the level of sociogenesis, where representations themselves are formed and transformed. As children learn how to play marbles, or come to understand something about the world of gender, or grasp the idea of conservation, or learn to weave using a traditional loom, they are always engaged in some process of construction in which something novel emerges. But the novelty is for the child, since it is the construction of something new in relation to their existing knowledge and understanding, rather than something novel for the culture in which they are growing up. It is the study of sociogenetic transformation which is central for an understanding of cultural change, a focus that can pose particular methodological challenges. From time to time phenomena emerge in the social world which are recognized as significant and important phenomena, but whose initial unfamiliarity means that the process of familiarization (that is, the sociogenesis of new representations, or, rather, representations of this new object) itself becomes visible and open to investigation. The emergence of HIV/AIDS is one such phenomena of recent times, and research has indeed shown how the representation of this phenomena has undergone considerable transformation in the relatively brief period since the condition was first identified, both within the Western world and within developing societies. In these circumstances successive representations of the same object can be identified either in the public discourse of the mass media, or through the discourse of participants in successive waves of research investigations. Not only does the identification of successive representations of the same object provide data for the analysis of sociogenetic transformation, but, especially where public discourse is available, it also includes much information about the communicative processes influencing these transformations (Alex Preda’s, 2005, book is a good illustration of this approach, which broadly corresponds to what Moscovici (1990) has described as “Piaget’s Way”). But sociogenetic change can also occur in more subtle ways over much more extended periods of time, and here the social psychological analysis of such change encounters more complex methodological issues which concern first of all the visibility of the phenomenon itself. Characteristically a more synchronic perspective has been adopted to the investigation of such transformation through a strategy by examining the ways in which the same object becomes represented by different social groups (which Moscovici describes as “Bartlett’s Way” or “Vygotsky’s Way”). Bartlett’s own work provides a good illustration of this approach, as can be seen in his (1923) accounts of how specific

Culture and social representations 209 cultural elements (and he is mostly concerned with elements of the material culture) move from one primitive group to another. Another example, of course, is Moscovici’s (1976) own study of the transformations of psychoanalysis as it becomes represented by different social groups in French society. In this work Moscovici draws on both the responses of participants to a questionnaire investigation as well as on a systematic content analysis of the French media. And it is this content analysis which enables Moscovici to analyze the different communicative practices (propagation, propaganda, and diffusion) through which different representations of psychoanalysis are constructed and projected. If one misses a diachronic perspective in this study, it nevertheless provides elegant testimony to the rich possibilities which a careful and thorough synchronic analysis can provide for understanding the dynamics of sociogenetic transformation. In the course of his work on psychoanalysis, Moscovici introduces a concept which has come to be seen as both theoretically and methodologically significant for the analysis of sociogenetic transformation (cf Jovchelovitch and Gervais, 1999; Wagner et al., 1999, 2000). In the course of his research on representations of psychoanalysis Moscovici (1976, pp. 279ff) observed the co-existence of different and even contradictory modes of thinking in his research on psychoanalysis. In moving between these different registers or voices, he notes that people in contemporary societies are “speaking” medical, psychological, technical, and political languages. As he puts it, “the same group, and mutatis mutandis, the same individual are capable of employing different logical registers in the domains which they approach with different perspectives, information and values” (1976, p. 286, my translation). By extending this phenomenon to the level of thought he suggests that “the dynamic co-existence – interference or specialization – of the distinct modalities of knowledge, corresponding to definite relations between man and his environment, determines a state of cognitive polyphasia” (p. 286, my translation, emphasis in the original). Indeed, he suggests that far from being an exceptional state of affairs, it is probable that the “coexistence of cognitive systems should be the rule rather than the exception” (1976, p. 285, my translation). From what he describes as the hypothesis of cognitive polyphasia, Moscovici suggests that research should be concerned with “the analysis of transformations – equilibrium and evolution – of these modalities of knowledge, of the relations which are established between them and their adaptation” (1976, p. 287, my translation), and that more generally social psychology should focus on the “movement of forms of reflection and their order, comparing them with events and factors of interaction and culture” (1976, p. 287, my translation, emphasis in the original). Moscovici’s discussion of cognitive polyphasia thus suggests that within the complex of representations characteristic of a social group we can expect to find discontinuities, and even contradictions between different elements, and further, that if we consider the dynamic of the social group we shall see that these different elements are not random collections but that they are organized in systematic ways. Methodologically this proposal is important for the analysis of sociogenetic

210


change, since it alerts us to the probability of finding different elements within the representations of a social group, and to exploring the character of the structure through which they are held together within a system. Hence we need to consider cultures not as finished or complete systems, but rather as systems in constant transformation in which the present can be expected to be not only polyphasic, but also structured around the relations (tensions and contradictions) between the various elements of this polyphasia. One example of polyphasia can be seen in Anne Parsons’s (1969) account of the Southern Italian immigrants to the United States who have converted from the Roman Catholicism of their native culture to a form of Pentacostalism. She identifies many ways in which characteristics of the specific pattern of Southern Italian Catholicism survive or re-emerge within the new forms of Pentacostalism, sometimes even in almost unconscious ways, almost like the Freudian return of the repressed. For instance, within the services of this group she observed the “very frequent repetition by the women of the phrase santa, santa, santa (santa=saint), sometimes appearing in the more complex form of a series of indistinguishable words beginning with santa, but always following a rhythmic pattern and phrase length exactly equivalent to that of Santa Maria, piena di grazia . . . or the Hail Mary which is intimately known to every Catholic” (p. 252). Indeed it is in the services that she finds these combinations of the rejected past and the accepted present most clearly expressed. She notes for instance the way in which the sermons characteristically moved through cycles of affirmation, doubt, regression, and resolution, and that as they did so the pastor also shifted his religious style. While the affirmative phases were characterized by a more or less rational and coherent discourse in an identifiably Protestant tradition, the intermediate phases were notable for their expression of anxiety and skepticism through modulations of voice, of body posture, and of cognitive incoherence, or as she puts it “a process of free association or a kind of dramatic role-playing that served to act out a number of mutually contradictory attitudes” (p. 269). In her analysis of this material, Parsons suggests that the ritual of the service is more than just a bipolar structure which combines elements from two distinctive cultural traditions, but that the specific forms of the ritual need to be referred to the context of the social and cultural changes through which this particular group is passing. While Parsons’ work predates Moscovici’s formulation of the hypothesis of cognitive polyphasia, we can nevertheless see in her account of these Pentacostal immigrants a specific form of polyphasia in which the sociogenesis of cultural change is oriented by the dynamic of the group. While Parsons’ research is a synchronic study, it can nevertheless be considered as a genetic study precisely because the orientation to social change allows the present to be understood within the dimension of time.

Representations of mental health and illness in contemporary India One last example will serve to bring these threads of discussion together. It concerns a study of changing representations of mental health and illness among contemporary

Culture and social representations 211 urban middle-class North Indians (Wagner et al., 1999, 2000).1 Traditional Indian patterns of thought have, or course, a rich set of beliefs about the phenomena of mental illness, which in milder forms can be attributed to an imbalance of the humors and in more serious instances to spirit or ghost possession. These different forms of disturbance are also associated with different forms of treatment, ranging from encouragement and support within the family, through various forms of ritual healing in the temple, up to severe physical treatments of trying to beat the possessing spirit out of the body it has occupied (called in Hindi the jhar-phook). Traditional Hindi society in North India includes arranged marriage, and public knowledge of mental illness in the family is a source of considerable anxiety, since it prejudices the marriage prospects for other members of the family. Over the past few decades Western psychiatric medicine, with its very different forms of etiology and treatment, has also become a stronger presence in the life of these Indian groups, both in the sense of the availability of psychiatric clinics as well as through the media in various forms. Faced with these different and disjunctive representations of mental illness, how do these urban Indians now think and talk about the phenomena of mental illness? To explore these questions the participants were interviewed on the basis of a short vignette describing a person engaged in strange patterns of behavior (strange, but stereotypical for this community of the behavior of the mentally ill). This strategy allowed the interviews to range over a number of issues related not only to participants’ knowledge of both traditional and psychiatric models of mental illness and its treatment, but also to asking how the participants would respond if someone in their own family were to begin to behave in such ways. What emerged from the interviews was that the participants could talk knowledgeably about both traditional and psychiatric notions of mental illness (see Table 13.1 for a summary of the characteristics attributed to each of these representations), though it was also clear that in general they had a richer, more elaborated understanding of traditional patterns of belief and a more superficial grasp of modern psychiatry; a difference which was especially marked in their descriptions of the details of forms of treatment.

Table 13.1 Characteristics of modern and traditional healing

Etiology Treatment Dominant principle

Psychiatrist

Traditional healer

Unfulfilled desires, fear, shock, pressure, depression Talking, finding the reasons, friendliness, removing ideas, medicine Psychological agency of the patient

Spirit possession, disequilibrium of humors, etc. Exorcism, rituals, sacrifice, medicine

Adapted from Wagner et al (2000).

Agency of the causes of the illness, e.g. spirits

212


More interestingly for the present discussion, when they considered the relative merits of different forms of treatment, many of the participants expressed a disbelief in traditional forms and faith in the efficacy of the modern scientific approach of the psychiatrist. For instance, this was one man’s view of the effectiveness of different forms of treatment2: Interviewer: Respondent:

Do you think that there is a TECHNIQUE with traditional healers? NO, I THINK NOT. If a TECHNIQUE would be with them then SUCCESS must be there 100%. But they do not get SUCCESS anywhere, I THINK.

Another expressed the difference between traditional and psychiatric forms of treatment in this way: Interviewer: Respondent: Interviewer: Respondent:

If any member of your family would start to have such a behavior as I told earlier, what would you do? I would take him to a psychiatrist, who does an up-to-date treatment. Would you do anything more? No, nothing . . . Once a boy came to me and got ill. The people told me to go for traditional treatment, but I said I would not do it . . . What would the traditional, the traditional healer do? The jharphook? That is nothing!

At the same time the respondents were often very aware of the power of traditional patterns of thought within the extended family, as this woman explains: Because . . . Look, suppose in my family or maybe in the LONG RUN if it will happen to my child or my daughter-in-law or with me, then all the people who will come will suggest to show him [the ill person] to that maulavi [Urdu word for spirit healer], “in Patna there is a good maulavi. Take him to Biharsharif [a village near Patna] to the shrine.” So all minds are of the same kind. Isn’t it? So I do this even if I don’t want to do it, but I won’t get SUCCESSFUL [the patient won’t be cured there]. But I do it. And when I am not cured, I go to the DOCTOR at LAST and the DOCTOR does the treatment . . . As these extracts indicate, for many of the respondents it was the claim to scientific knowledge and a practice derived from it which was the source of their preference for a psychiatric representation. As one man put it: These people [psychiatrists] are becoming more successful. They are doing research, and they are improving everything. They improve their MEDICINES and their GENERAL SCIENCE. Or as this woman expressed it:

Culture and social representations 213 It is possible that he [traditional healer] is not able to understand the BAD BEHAVIOUR [of the sick person] and it might increase the sickness. It is just possible that he [sick person] can turn mad or become aggressive [because the healer is treating him in a wrong way]. He [healer] might do it unknowingly. But a PSYCHOLOGIST would move in scientifically and would take ACTION continuously. He would always consider whether his ACTION is contributing towards cure or harming him [sick person]. While this preference for the psychiatric over the traditional was common, especially among those of the younger generation among the participants, at times respondents were more conscious of the conflict between these different representations, especially when, as in the case of this woman, there was also experience of successful intervention by a traditional healer: As I told you, it happened with my aunt. But I think these things are nothing. Many people die, I have read that in books. Haven’t you heard that one ojha (spirit healer) killed somebody beating him to death? That man died. His story is finished. So I have some faith and at the same time I don’t believe in them. Her final sentence crystallizes something important about cognitive polyphasia as the coexistence of contradictory representations. More generally, though, the preference for the psychiatric was linked to the claims of science and the idea of progress which this enshrines. The polyphasic quality of contemporary representations of mental illness among the Indian middle-class has to be seen within the context of their general vision of the shifts and changes in their society. They envisage India as a country which is developing away from a traditional culture and society towards a more modern form. Development in this context means becoming more like the advanced societies of the West, in which a scientifically based medicine and treatment for the mentally ill serves as a token of this process, providing a bridge, as it were, between the traditional and the modern. In this example the “traditional” elements of Indian representations of mental illness which might have been considered as deeply embedded within the communal life of these societies are being drawn into a more active form of reflection and change through this process of cultural contact, communication, and exchange. We cannot predict how these representations will develop. Perhaps traditional patterns of thought will disappear altogether and simply be replaced by Western psychiatric notions. However, given that we know that even in the West social representations of mental illness retain strong and powerful aspects of our historical patterns of thought (Jodelet, 1991) this seems unlikely. Much more likely is that in India new forms of polyphasia will emerge in which psychiatric notions will acquire a characteristically Indian patterning, just as Denise Jodelet shows us how the representations of mental illness among the villagers of Ainay-le-chateau have been able to assimilate notions derived from modern psychiatry within more traditional patterns of belief. For Jodelet the dynamic at work in structuring the representations of these villagers is the profound need to sustain a distinction between the mad who

214


are resident in the village and the villagers themselves. If they could not make this differentiation between self and other, then these villagers risk finding themselves collapsing into madness. Striving to sustain a sense of difference is what generates their endless inventiveness in elaborating structures through which to construct their understanding of mental illness. For the Indian middle-classes it is the theme of modernization that establishes the fundamental dynamic for this social group as it engages in these changes, and also serves to orient their appropriation of Western psychiatric notions to produce a specific form of polyphasia.

Conclusion This chapter began with the difficulties inherent in the concept of culture itself, and thus also of establishing a clear relationship between culture and social representations. At best it seems that we can distinguish them by noting that while social representations always bear on specific objects, the term culture refers to a more diffuse body of representations which serve to characterize something distinctive about a community. But this is not a sharp distinction, since, of course, in constructing specific representations social groups may also use elements from this broader repertoire of cultural resources. Further, insofar as culture itself can be considered a set of representations then we can also see that the processes of formation and transformation are similar whether we think of social representations or of cultures. Indeed, the interplay between these two terms is what is central – culture being the sedimentation of sociogenetic changes which appear first as changes in the social representation of a specific object. If we think, for instance, of the position of women in society we can note how great the change has been over the past century, from being excluded even from participation in the political life of society towards a situation structured around complex discourses of equality. As always, change was initiated by a minority seeking to establish a different position, creating as it were novel polemical representations. Over time these have shifted towards becoming emancipated representations, and in important ways now constitute hegemonic representations, so that we can accept that there has been a shift in our culture. No doubt there will be further changes in these representations as the theme of equality is expanded and explored further. But in providing a conceptual frame for considering questions of social and cultural change in this way, the theory of social representations has something significant to contribute to contemporary discussions of cultural psychology. By emphasizing the dynamic aspect of social groups, and thus also orienting the discussion of representations to the dimension of time, and through the concept of cognitive polyphasia, the theory can help to elucidate central aspects of processes of cultural change.

Notes 1 This is not an intuition which is unique to Moscovici. One can find similar ideas in the work of many other psychologists concerned with the genesis of social knowledge, stretching back at least to the work of McDougall and Bartlett, for instance.

Culture and social representations 215 2 This research was undertaken in Patna, North India with a group of 39 participants, ranging in age from 20 to 55+ years and including equal numbers of men and women. All of the participants had at least begun University level courses. The interviews were conducted in Hindi by our native Hindi speaking research assistants, who also transcribed them and translated them into English. Occasionally the respondents used English words in their replies, and these are transcribed in capital letters.

References Bartlett, F. (1923). Psychology and Primitive Culture. Cambridge: Cambridge University Press. Bartlett, F. (1932). Remembering: A Study in Experimental and Social Psychology. Cambridge: Cambridge University Press. Duveen, G. (1998). The psychosocial production of knowledge: Social representations and psychologic. Culture and Psychology, 4, 455–472. Duveen, G. (2000). The Power of Ideas. Introduction to S. Moscovici (Ed. G. Duveen) Social Representations: Explorations in Social Psychology (pp. 1–17). Cambridge: Polity Press. Duveen, G. (2001). Representations, identities, resistance. In K. Deaux and G. Philogène (Eds.), Representations of the Social (pp. 257–270). Oxford: Blackwell. Duveen, G. and Lloyd, B. (1990) Introduction. In G. Duveen and B. Lloyd (Eds.) Social Representations and the Development of Knowledge (pp. 1–10). Cambridge: Cambridge University Press. Goldmann, L. (1976). Cultural creation in modem society. Saint Louis: Telos Press. Jahoda, G. (1988). Critical notes and reflections on “social representations.” European Journal of Social Psychology, 18, 195–209. Jodelet, D. (1991). Madness and Social Representations. London: Harvester. Jodelet, D. (2002). Les representations socials dans le champ de la culture. Social Science Information, 41, 111–133. Jovchelovitch, S. and Gervais, M.-C. (1999). Social representations of health and illness: The case of the Chinese community in England. Journal of Community and Applied Social Psychology, 9, 247–260. Lave, J. (1988). Cognition in practice. Cambridge: Cambridge University Press. Lloyd, B. and Duveen, G. (1990). A Semiotic Analysis of the Development of Social Representations of Gender. In G. Duveen and B. Lloyd (Eds.) Social Representations and the Development of Knowledge (pp. 27–46). Cambridge: Cambridge University Press. Marková, I. (2003). Dialogicality and Social Representations: The Dynamics of Mind. Cambridge: Cambridge University Press. Moscovici, S. (1973). Foreword. In C. Herzlich Health and Illness. London: Academic Press. Moscovici, S. (1976). La Psychanalyse, son image et son public. Paris: Presses Universitaires de France. Moscovici, S. (1981). On Social Representation. In J. Forgas (Ed.), Social Cognition. London: Academic Press. Moscovici, S. (1984). The Phenomenon of Social Representations. In R. Farr and S. Moscovici (Eds.) Social Representations. Cambridge: Cambridge University Press. (Reprinted in Moscovici, 2000) Moscovici, S. (1988). Notes towards a description of social representations. European Journal of Social Psychology, 18, 211–250.

216


Moscovici, S. (1990). Social psychology and developmental psychology: Extending the conversation. In G. Duveen and B. Lloyd (Eds.), Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Moscovici, S. (2000). Social Representations: Explorations in Social Psychology. (Edited by G. Duveen). Cambridge: Polity Press. Parsons, A. (1969). Belief, Magic and Anomie: Essays in Psychological Anthropology. New York: The Free Press. Piaget, J. and Inhelder, B. (1971). The Mental Imagery of the Child. New York: Basic Books. Preda, A. (2005). AIDS, Rhetoric and Medical Knowledge. Cambridge: Cambridge University Press. Psaltis, C. and Duveen, G. (2006). Social relations and cognitive development: The influence of conversation type and representations of gender. European Journal of Social Psychology, 36, 407–430. Rogoff, B. (1990). Apprenticeship in thinking. New York: Oxford University Press. Saito, A. (1996). ‘Bartlett’s Way’ and social representations: The case of Zen transmitted across cultures. The Japanese Journal of Experimental Social Psychology, 35, 263–277. Schütz, A. (1972). The Phenomenology of the Social World. London: Heinemann. Valsiner, J. (2003). Beyond social representations: A theory of enablement. Papers on social Representations (www.psr.juk.at), 12, 7.1–7.16. Valsiner, J. and Van de Veer, R. (2000). The social mind: Construction of the idea. Cambridge: Cambridge University Press. Wagner, W., Duveen, G., Themel, M., and Verma, J. (1999) The modernisation of tradition: Thinking about madness in Patna, India. Culture and Psychology, 5, 413–445. Wagner, W., Duveen, G., Verma, J., and Themel, M. (2000). ‘I have some faith and at the same time I don’t believe’ – Cognitive Polyphasia and Cultural Change in India. Journal of Community and Applied Psychology, 10, 301–314.

14 Social actors and social groups A return to heterogeneity in social psychology Gerard Duveen

Groups and communicative genres What is it that holds collectives together? Is it simply a shared identity of belonging to the same collective? Or do we need to call upon some other resources to understand the forms of collective functioning? Contemporary social psychology has opted fairly unanimously for the former of these alternatives, and chosen to ignore the second. Yet an attentive reader of Serge Moscovici’s (1961/1976/2008) Psychoanalysis: Its Image and Its Public will find a focus on precisely this second alternative, which is elaborated through the analyses of communicative systems which occupies Part II of the book. As Willem Doise (1993) remarked in his comments on the reception of the theory of social representations in the AngloSaxon world, it was notable how little attention had been paid to this part of the book. It is here that Moscovici formulates a set of original hypotheses about the communicative systems of diffusion, propagation and propaganda as distinct communicative genres. Through his content analysis of the French press, one can see Moscovici tracing the outlines of different types of groups structured through distinctive forms of social-psychological organisation. What is challenging here is a recognition that there can be different types of social-psychological organisation, each founded on a particular form of communication. The forms of communication which Moscovici describes appear as types of communicative genres within the mass media of the French Press of the 1950’s. Two points about this focus for his analysis stand in need of clarification. First, the French press of the 1950’s was a more differentiated body of opinion than we are used to finding today in the press as we have come to know it. France in those years was a country recovering from a war, but more importantly, also a country divided by social schisms of considerable force and intensity. A rightwing dominated by the Catholic Church confronted a leftwing dominated by the Communist Party, with a middle ground of liberal professionals. Within the framework established by the relations between these political forces, a topic such as psychoanalysis could not escape scrutiny and re-presentations. In this context, the division and confrontation between these forces meant that even such a seemingly obscure topic as psychoanalysis could become a focus for creative re-imagination. Each of these three powerful forces sought to harness a truth about psychoanalysis and

218


project it into the world through the mass media. For the liberal professionals, the truth was wrapped in a kind of ironic detachment which allowed them to frame the theme as much in a positive as in a negative light. For the rightwing, psychoanalysis was a complex form, which they sought to define by its relationship to the priestly function of the Church, on the one hand, and its social utility on the other. For the leftwing, it was regarded with unremitting hostility as a force, which was a representative of the bourgeois ideology of the United States. There is some caricature in these descriptions, but not much. The three frameworks each took in psychoanalysis, and projected its own version outwards again. This brings me to the second point I want to make. To whom were these images of psychoanalysis projected? Or in other words, what was the relation between the mass media and the individual? The main object of these projections were the people who constituted each of these three social forces. By recognising the image which was being projected at them, these people also contributed to the solidification of these images. As this phrase implies, there is a circularity here, with individuals being constituted and re-constituted through the messages projected by the media. But this is the circularity of the social actor, of the form called into being by the communications that circulate through a divided culture.

Social actors and social groups An immediate question arises as to where did Moscovici uncover the resources for such a distinctive vision? One response is to note that the work was originally published in 1961, that is at a time when such a question would have been thought to be absurd. What characterised thinking about social actors and social groups in the early 1960’s was precisely a recognition of some multiplicity of conceptual forms. In the course of a chapter (not much cited), Moscovici (1963) outlines an overview of the field, in which it is interesting to see how he characterises research in this area. Following some attempts at disposing of uninteresting questions, he comes to focus on dissonance theory, which provides a central frame of reference for the chapter as a whole, for it is here that some of the most productive questions are articulated. The structure of the review is interesting in itself, since each section closes with questions which open to the next section. In this instance, the section on dissonance follows the section on structural issues, which ends with the questions: “What is meant by structure in social psychology, and what are the important social psychological structures?” (Moscovici, 1963, p. 248). It is in responding to these questions that the issues of (cognitive) dissonance appear. Equally modestly, the section on dissonance, which ends with a call from the reviewer for a “more thorough and broader theoretical discussion” (Moscovici, 1963, p. 251), is followed by a section on social representations—then a quite recent topic of research making almost its first appearance in the English language. [And in parenthesis, we can add that this is the penultimate section of the review, which ends with a discussion of the open and closed mind.] Perhaps the most interesting thing about this chapter is the fact that the terms “social actors” and “social groups” do not make any appearance in it. Why should

Social actors and social groups 219 this be so? Perhaps it is no more than that these terms designate fields of study, which have emerged since 1963, but this would not necessarily be the best answer. For the contemporary reader of Psychoanalysis: Its Image and Its Public the analyses of communicative systems in Part II of the book provide a challenging occasion for reconsidering current social psychological thinking about the character of social groups. In Moscovici’s careful delineation of the communicative systems of diffusion, propagation and propaganda through his content analysis of the French press, one can also see the description of different types of group structured through distinctive social psychological organisations. What is challenging here is the recognition that there can be different types of social-psychological organisation, for one of the consequences of the dominance of Social Identity Theory/Self-Categorisation Theory in recent years has been the homogenisation of the social psychological concept of the group itself. While SIT/SCT have certainly made a contribution to our understanding of the group, following John Turner’s proposal for the cognitive redefinition of the social group it has become common practice to consider all groups as sharing the basic form of social psychological structure. Moscovici’s work has the benefit of predating this approach, drawing as it does on the traditions of Kurt Lewin and the study of group dynamics, as well as on earlier theories of communication and opinion. What one sees in Psychoanalysis is not only the way in which distinct social representations both generate and are sustained by different communicative genres, but that these different communicative systems also reveal different forms of affiliation amongst the publics drawn together around each communicative system.

Sympathy, communion and solidarity In the hypothesis he develops in Chapter 6, Moscovici himself suggests that the genres of diffusion, propagation and propaganda are each linked to the construction of a specific type of social-psychological object, opinion, attitude and stereotype respectively. But the interest in this part of the book extends even further, since the discussion of these different communicative genres also invites some reflection on the way in which the concept of group itself is conceptualised. As I have pointed out, in contemporary social psychology the notion of group has taken on a rather homogenous form, so that all groups are considered equal in their pattern of social-psychological functioning, differing only in the set of values, ideas or attitudes they espouse. Yet the identification of these different communicative genres, and the different patterns of social-psychological functioning associated with each of them, suggests that we need to expand our understanding of the character of the group to consider also the ways in which they may also be characterised by distinctive communicative processes. In short, we need to recognise that there is an intimate relation between the values and attitudes of a group and the characteristic patterns of communication which sustain it. To do so, of course, brings with it the implication that groups are more heterogeneous forms of social-psychological organisation than is envisaged in contemporary theorising. But one can also identify different forms of affiliation

220


corresponding to each communicative genre, and consequently also different representations of the people who constitute both the in-group and the out-group in each instance. Diffusion, for instance, is characterised by the voluntary association of independently minded individuals, where the in-group is characterised by the possession of a certain sceptical intelligence, while the out-group is seen as embracing various forms of dogmatism. The affiliative bonds linking the members of this group might be described as a form of sympathy. Propagation is more circumscribed since it is an association founded on belief, which sets limits to the intellectual curiosity of individuals, or to their creativity in responding to new problems. The limits of belief are established by a central authority, though on the margins or the periphery the legitimacy of this authority may also be questioned. The out-group(s) are characterised either by their lack of belief or by their adherence to alternative beliefs or commitment to a political ideology incompatible with the belief. This group can be characterised as being affiliated through some form of communion. Propaganda draws together people who not only share a specific political commitment but also envisage the appropriate form of political organisation as one in which the centre dominates by defining realities. The in-group is thus composed of militants who may be active in various forms of agitation, but who are dependent on the centre for the intellectual content of their representation. The out-group(s) are defined either by their lack of commitment to this ideology, or by their commitment to a different ideology. The group constituted through propaganda can therefore be seen to be affiliated through a particular form of solidarity. Considered in this way, sympathy, communion and solidarity can be seen as identifying distinct types of group structured through distinct forms of socialpsychological organisation.

Social representations and influence The analyses of communicative genres in Part II of the book also bring us back to another feature of Mosocovici’s social-psychological imagination, the relationship between representation and influence. It is tempting to consider diffusion, propagation and propaganda as also constituting different forms of social influence. Certainly, if, as I have suggested before, the shape and form of representations is structured by the balance of influence processes operating in the communicative practices of a group at a particular time (Duveen, 2000), then it would seem appropriate to consider these communicative genres as forms of social influence. Yet a word of caution is also necessary before making the link directly to Moscovici’s (1976) genetic model of social influence processes. The communicative genres analysed in Part II of the book emerge from the content analysis of mass media, while the genetic model of social influence stems from an analysis of face-to-face communication within a controlled experimental paradigm. While we might expect to see some continuity in the form of social influence as we move from the interpersonal to mass-mediated communication, we should also be attentive to significant differences between these

Social actors and social groups 221 contexts. In the experimental studies individuals are engaged precisely in attempts to change the thoughts or behaviour of other individuals, whereas the influence exercised through the mass media is often less immediate or more indirect. In relation to the mass media people find themselves surrounded by a flowing current of influence which sustains their affiliation with a particular group and a particular way of seeing things rather more than finding themselves exposed to a direct challenge to their point of view. We choose which magazines or newspapers we read, or which television programmes we watch, and our choices are already an expression of our affiliation with a specific social group or section of society. In this sense the contexts which frame participation in these different communicative encounters need to be examined more closely if we are to articulate a theory of social influence which can extend from the interpersonal to mass-mediated forms of communication.

Conclusion As well as serving to engage in a critical discussion about the concept of the group, these reflections also suggest two significant questions which need to be addressed if the perspective of Psychoanalysis is to be extended to the contemporary world. First, what repertoire of communicative genres can be identified in current mass communication? Do we still find diffusion, propagation and propaganda? And have any other forms appeared? Secondly, the perspective of Psychoanalysis is derived from the analyses of texts, undertaken at a time when mass communication was largely based on the circulation of texts. In the contemporary world it is not so much that texts have disappeared (though the circulation of newspapers has decreased, the internet or the verbal communication of television and radio still carry much that is textual), as that mass communication has seen an extraordinary increase in the circulation of images. Is the circulation of images linked to distinctive forms of social-psychological organisation?

References Doise, W. 1993. Debating Social Representations. In G.M. Breakwell and D.V. Canter (Eds.), Empirical Approaches to Social Representations (pp. 157–170). Oxford: Clarendon. Duveen, G. 2000. The Power of Ideas. Introduction to S. Moscovici (Ed. G. Duveen) Social Representations: Explorations in Social Psychology (pp. 1–17). Cambridge: Polity Press. Moscovici, S. 1963. Attitudes and Opinions. Annual Review of Psychology, 14, 231–260. Moscovici, S. 1976. Social Influence and Social Change. London: Academic Press. Moscovici, S. 1961/1976/2008. Psychoanalysis: Its Image and Its Public. Cambridge: Polity Press.

Bibliography The published papers of Gerard Duveen

Books and edited volumes 1990

1991 1992 1998 2000

2001 2002 2008

(edited with Barbara Lloyd) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press. Italian translation: Rappresentazioni sociali e sviluppo della conoscenza. Rome: Armando Editore (1998). Editor of English Translation of Madness and Social Representations by Denise Jodelet, London: Harvester Wheatsheaf. (with Barbara Lloyd) Gender Identities and Education. London: Harvester Wheatsheaf. (with M. Arnot, J. Ruddock, J. Gray and M. James) Gender and Performance: A Review of Recent Research. Report commissioned and published by OFSTED. Editor of S. Moscovici Social Representations: Explorations in Social Psychology. Cambridge: Polity Press. (A Portuguese translation of this book was published by Editora Vozes, Petrópolis, Brazil in 2003.) (Edited with Colin Fraser, Brendan Burchell and Dale Hay) Introducing Social Psychology. Cambridge: Polity Press. Représentations et croyances. Special Issue of Psychologie et Societé, edited by T. Apostolidis, G. Duveen and N. Kalampalikis. Editor of English translation of Serge Moscovici Psychoanalysis: Its Image and Its Public. Cambridge: Polity Press.

Papers in refereed journals 1981

1985 1985 1986 1986

(with Kirk Weir) Further development and validation of the Prosocial Behaviour Questionnaire for use by teachers. Journal of Child Psychology and Psychiatry, 22, 357–374. A developmental study of the influence of situation and actor on children’s judgements about friendship. Cahiers de Psychologie Cognitive, 5, 411. (with M. M. Shields) Children’s ideas about work, wages and social rank. Cahiers de Psychologie Cognitive, 5, 411–2. (with Barbara Lloyd) The significance of social identities. British Journal of Social Psychology, 25, 219–230. (with Barbara Lloyd) Authors’ rejoinder to Palmonari and Emler. British Journal of Social Psychology, 25, 235–236.

Bibliography 223 1988

1988 1988 1991 1992 1993 1994 1996 1996 1998 1999

1999 1999 2000

2000 2002 2002 2003 2003 2005 2006

2006

2007

(with Barbara Lloyd and Caroline Smith) Social representations of gender and young children’s play: A replication. British Journal of Developmental Psychology, 6, 83–88. (with Barbara Lloyd) Gender as an influence on the development of scripted pretend play. British Journal of Developmental Psychology, 6, 89–95. (with Barbara Lloyd and Caroline Smith) A note on the effects of age and gender on children’s social behaviour. British Journal of Social Psychology, 27, 275–278. (with Barbara Lloyd) Expressing gender identities in the first year of school. European Journal of Psychology of Education, 6, 437–447. (with Annamaria De Rosa) Social representations and the genesis of social Knowledge. Ongoing Productions on Social Representations, 2, 94–108. The development of social representations of gender. Papers on Social Representations, 2, 171–177. Unanalysed residues: Representations and behaviours. Papers on Social Representations, 3, 207–12. The development of social representations of gender. Japanese Journal of Experimental Social Psychology, 35, 256–262. (with Patrick Leman) Developmental differences in children’s understanding of epistemic authority. European Journal of Social Psychology, 26, 683–702. The psychosocial production of knowledge: Social representations and psychologic. Culture and Psychology, 4, 455–472. (with W. Wagner, R. Farr, S. Jovchelovitch, F. Lorenzi-Cioldi, I. Marková and D. Rose) Theory and method of social representations. Asian Journal of Social Psychology, 2, 59–89. (with P. Leman) Representations of authority and children’s moral reasoning. European Journal of Social Psychology, 29, 557–575. (with W. Wagner, M. Themel and J. Verma) The modernisation of tradition: Thinking about madness in Patna, India. Culture and Psychology, 5, 413–445. (with W. Wagner, J. Verma and M. Themel) ‘I have some faith and at the same time I don’t believe’ – Cognitive Polyphasia and Cultural Change in India. Journal of Community and Applied Psychology, 10, 301–314. Piaget ethnographer. Social Science Information, 39, 79–97. (with T. Apostolidis and N. Kalampalikis) Représentations et croyances. Introduction to the special issue of Psychologie et Societé, 3, 7–11. Construction, belief, doubt. Psychologie et Societé, 3, 139–155. (with T. Zittoun, A. Gillespie, G. Ivinson and C. Psaltis) The use of symbolic resources in developmental transitions. Culture and Psychology, 9, 415–448. (with P. Leman) Gender identity, social influence and children’s arguments. Swiss Journal of Psychology, 62, 149–158. (with G. Ivinson) Classroom structuration and the development of social representations of the curriculum. British Journal of Sociology of Education, 26, 627–642. (with C. Psaltis) Social relations and cognitive development: The influence of conversation type and representations of gender. European Journal of Social Psychology, 36, 407–430. (with E. Ruwanpura, S. W. Mercer and A. Ager) Cultural and spiritual constructions of mental distress and associated coping mechanisms of Tibetans in exile: implications for western interventions. Journal of Refugee Studies, 19, 187–202. (with C. Psaltis) Conversation types and conservation: Forms of recognition and cognitive development. British Journal of Developmental Psychology, 25, 79–102.

224 2007 2008 2009

Bibliography (with B. Wagoner and A. Gillespie) Bartlett in the digital age. The Psychologist, 20(11), 680–681. Social actors and social groups: A return to heterogeneity in social psychology. Journal for the Theory of Social Behavior, 38(4), 369–374. (with C. Psaltis & A.N. Perret-Clermont) The social and the psychological: Structure and Context and Intellectual Development. Human Development, 52, 291–312.

Chapters in edited volumes 1986

1986

1988

1989

1989

1990

1990

1991

1993

1993

1994

1997 1997

(with M. M. Shields) The Young Child’s Representation of Persons. In J. CookGumperz, W. Corsaro and J. Streeck (Eds) Children’s Worlds and Children’s Language. Berlin: Mouton de Gruyer (pp. 173–203). (with Maureen Shields) Dialectics, Dialogue and the Social Transmission of Knowledge. In J. H. Danks, I. Kurcz and G. Shugar (Eds) Knowledge and Language Amsterdam: North-Holland (Elsevier) (pp. 517–539). The Child’s Reconstruction of Economic Life. In K. Ekberg and P. E. Mjaavatn (Eds) Growing into a Modern World, Centre for Child Research, University of Trondheim, Norway (pp. 177–199). (with Barbara Lloyd) The Reconstruction of Social Knowledge in the Transition from Sensorimotor to Conceptual Activity. In A. Gellatly and J. Sloboda (Eds) Cognition and Social Worlds (Keele Cognition Seminar 1987). Oxford University Press (pp. 83–98). (with Barbara Lloyd) Social Representations and the Development of Knowledge. In J. P. Forgas and J. M. Innes (Eds) Recent Advances in Social Psychology: An International Perspective. Elsevier Science Publications (North-Holland). (with Barbara Lloyd) Introduction. In G. Duveen and B. Lloyd (Eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press (pp. 1–10). (with Barbara Lloyd) A Semiotic Analysis of the Development of Social Representations of Gender. In G. Duveen and B. Lloyd (Eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press (pp. 27–46). Asimmetria nello sviluppo della identità di genere (Asymmetry in the Development of Gender Identity). In C. Arcidiacono (Ed.) Identità, Genere, Differenza. Franco Angeli, Milan (pp. 29–42). (with Barbara Lloyd) An Ethnographic Approach to Social Representations. In G. Breakwell and D. Canter (Eds) Empirical Approaches to Social Representations. Oxford: OUP (pp. 90–109). (with Barbara Lloyd) The Development of Social Representations. In C. Pratt and A. Garton (Eds) Systems of Representations in Children: Development and Use. New York: John Wiley & Sons (pp. 167–183). Crianças enquanto atores sociais: As representações sociais em desenvolvimento. (Children as Social Actors: A Developmental Perspective on Social Representations.) In S. Jovchelovitch and P. Guareschi (Eds) Textos em Representações Sociais. Petropolis, Brazil: Vozes (pp. 261–293). Psychological Development as a Social Process. In L. Smith, J. Dockrell and P. Tomlinson (Eds) Piaget, Vygotsky and Beyond. London: Routledge (pp. 67–90). Prosocial Behaviour Questionnaire. In K. Sylva and J. Stevenson (Eds), Child Psychology Portfolio: Social Behaviour and Competence in Childhood. Windsor:

Bibliography 225

1998

1999

2000

2000 2001 2001

2001 2006

2007

2008 2008

NFER-Nelson. This portfolio of research instruments includes substantial material drawn from an earlier paper (Weir and Duveen, 1981). A construção da alteridade (The Construction of Otherness). In A. Aruda (Ed.) Representando a alteridades: Abordagens psicosociais (Representing Otherness: Psychosocial approaches). Petropolis, Brazil: Vozes (pp. 83–107). Le Développement des Représentations Sociales (The Development of Social Representations). Chapter in C. Garnier and M-L. Rouquette (Eds) La Genèse des Représentations Sociales. Montréal, QC: Editions Nouvelle (pp. 114–135). La culture dans les jeux imaginaires de jeunes enfants [Culture in the Pretend Play of Young Children]. In D. Saadi-Mokrane (Ed.) Sociétés et cultures enfantines. Lille: Editions du Conseil Scientifique de l’Université Charles-de-Gaulle – Lille 3 (pp. 111–117). The Power of Ideas. Introduction to S. Moscovici (Edited by G. Duveen) Social Representations: Explorations in Social Psychology. Cambridge: Polity Press (pp. 1–17). Representations, Identities, Resistance. In K. Deaux and G. Philogène (Eds) Representations of the Social. Oxford: Blackwell (pp. 257–270). Genesis and Structure: Piaget and Moscovici. In F. Buschini and N. Kalampalikis (Eds) Penser la vie, le social, la nature: Mélange en l’honneur de Serge Moscovici. Paris: Editions de la Maison des Sciences de l’Homme (pp. 163–173). Social Representations. In C. Fraser, B. Burchell, D. Hay and G. Duveen (Eds) Introducing Social Psychology. Cambridge: Polity Press (pp. 268–287). (with G. Ivinson) Children’s Recontextualisations of Pedagogy. In R. Moore, M. Arnot, J. Beck and H. Daniels (Eds) Knowledge, Power and Educational Reform: applying the sociology of Basil Bernstein. London: Routledge (pp. 109–125). Culture and Social Representations. In J. Valsiner and A. Rosa (Eds) Cambridge Handbook of Socio-Cultural Psychology. Cambridge: Cambridge University Press (pp. 543–559). Introduction to Serge Moscovici (Edited by G. Duveen) Psychoanalysis: Its Image and Its Public. Cambridge: Polity Press (pp. xi–xvii). (with C. Psaltis) The Constructive Role of Asymmetries in Social Interaction. In U. Mueller, J. I. M. Carpendale, N. Budwig and B. Sokol (Eds), Social Life and Social Knowledge: Toward a process account of development. Mahwah, NJ: Lawrence Erlbaum (pp. 183–204).

Book reviews 1994 1995 1996 1997

Review of E. J. S. Sonuga-Barke and P. Webley, Children’s Saving: A Study in the Development of Economic Behaviour. Journal of Economic Psychology, 15, 375–378. Comment on R. Augner ‘On Ethnography: Storytelling or Science’. Current Anthropology, 36, 117–118. Review of J. Piaget Sociological Studies. Education Section Review, 20, 8–9. Review of A. Pollard, The Social World of Children’s Learning, British Journal of Sociology of Education, 18, 287–289.

Translations 1988

Translation from Italian of Il mondo economico nel bambino (La Nuova Italia, 1981) by Anna Emilia Berti and Anna Silvia Bombi. English title: The Child’s Construction of Economics. Cambridge: Cambridge University Press.

226 1989 1989 1990

1998

1998

2000

Bibliography Translation from French of Serge Moscovici “Leon Festinger”, European Journal of Social Psychology, 19, 163–170. Translation from French of Gerard Lemaine “Styles of Explanation in Social Psychology”, European Journal of Social Psychology, 19, 369–370. Translation from French of Serge Moscovici “Social Psychology and Developmental Psychology: Extending the Conversation”. In G. Duveen and B. Lloyd (Eds) Social Representations and the Development of Knowledge. Cambridge: Cambridge University Press (pp. 164–185). Translation from the French of Serge Moscovici “The History and Actuality of Social Representations”. In U. Flick (Ed.) The Psychology of the Social. Cambridge: Cambridge University Press (pp. 209–247). Translation from the French of W. Doise, G. Mugny and J. A. Pérez “The Social Construction of Knowledge: Social Marking and Socio-cognitive Conflict”. In U. Flick (Ed.) The Psychology of the Social. Cambridge: Cambridge University Press (pp. 77–90). Translation from the French of S. Moscovici and G. Vignaux “The Concept of Themata”. In S. Moscovici (edited by G. Duveen) Social Representations: Explorations in Social Psychology. Cambridge: Polity Press (pp. 156–183). (Originally published in C. Guimelli (Ed.) Structures et transformations des représentations sociales. Neuchâtel: Delachaux et Niestlé.)

Index

action: conceptualization 17; coordination 69, 72, 78, 80, 103, 137; interpretation 16-17, 158, 161, 164; schemes 49, 61 adolescence 29-30, 49, 91, 167 affectivity 84-85, 87 animism 47, 59 assessment: methods 125, 131, 162; psychiatric 46, 58, 211-3 assimilation and accommodation 18, 60-1, 69 argument: child’s thought 48, 50-52; communication 117-18, 121, 138, 140, 193, 204; construction of the world 130; development 53, 72, 99-100, 115, 130, 135; gender 109, 117-18, 139, 142-5, 160, 162; representation 201-2, 207; resistance 146-9; social 28, 68, 77, 81-2, 85, 90, 96, 99-100, 102, 133-5, 138, 157, 160, 192-3, 201, 206-7; structure 70, 206; subject and object 18, 79 authority 11,13-14, 17, 21, 32, 68, 96, 133, 135-136, 139, 199-200, 220 autonomy 11, 15, 73, 100, 117, 129, 134 Bartlett, F.C. xviii, xxii, 91-2, 94, 201, 204-8, 214 Bernstein, B. 52 Billig, M. 176 Binet 58, 125 Breakwell, G. 170, 183 Categorization: development 164-5; gender 161; identity 183, 190 Child’s Conception of the World 44, 46, 59 clinical method (methode clinique) 44, 48-9, 125-7 cognition: development xi, xv-xvi, xxi, 96, 150; of social life 68, 82-87; polyphasia 63, 118-9, 207-214; processes xix, 85-7;

social xvi, 22-4, 75-82, 114, 175; structures 94, 108 cooperation and constraint 63, 115-6, 119, 133-5 communication: influence 137-9; knowledge 29, 113-17; mass mediated 62; representation 162, 173, 177, 180, 193, 199-200, 203-5, 217-21 conformity xviii, 61, 143 conversation: free 50; type 143-9 culture x, 68, 79, 81-5, 100-8, 113, 128-9, 157, 160-1, 182-6, 189, 190, 196-214, 218 Danziger, K. 29 decentration xvii-xviii, xxi, 60, 112-21, 126- 51, 168 Deschamps, J.C. 170, 183 dissonance 118, 218 Doise, W. 18, 30, 34, 38-40, 72, 78, 91, 96, 114-15, 135-6, 138, 141, 146, 150, 169, 180, 217 Durkheim, E. 56, 61, 99, 133-4, 159, 197, 199, 202 Einstein, A. 131-2, 180 Emler, N. 30, 39, 95-7, 99-100, 109, 138 empiricism 61, 91 epistemic subject xi, xv, xvii, xxii, 68-76, 85 epistemology: biological 43, 58, 60, 62, 70; constructionist; constructivist 57, 90, 94, 112-13; existence 83; genetic 23, 68-9, 72-6, 79, 151, 194; idealist 80; triadic 151 equilibrium 115, 209 ethnography xxi, 28, 42-3, 105, 190-1 Existence precedes essence 83

228

Index

Farr, R. xviii,131 Freud, S. xvii, 45, 57, 59-60, 71, 74, 83, 117, 124-8, 130-2, 175, 177, 180 friends 3-25, 27, 67, 83, 106, 200 Furth, H xvii, 29, 69, 80, 87, 97, 109, 158, 16, 186-7, 194 genetic: social psychology ix-x, xix-xx, 91, 114, 137, 169, 177; structuralism 61-2, 177 Genevan 18, 23, 44, 54, 57, 59, 72, 75-7, 87, 96, 114, 117, 135-6, 142, 150 Goldmann, L. 57, 61, 69, 71-2, 80, 83, 176-7, 180, 192-3, 204 heteronomy 11, 15 identity xviii, 38-9, 63, 93-4, 96, 105-8, 112, 134, 139, 157, 159-61, 167-170, 178, 182-194, 202-5 intelligence 45, 50, 58, 93-4, 125, 130, 137, 158, 176, 220 internalisation 92-3, 100-2, 104 interviews: clinical 3, 45-9, 52, 70 Jahoda, G. 29-30, 39, 109, 167, 201 James, W. ix, 182, 193 Jodelet, D. xvii, 91, 197, 213 Jovchelovitch, S. xviii, 94, 118, 121, 209 Kohlberg, L. xi, 24, 127-9 Lave, J. 102-3, 208 Levy-Bruhl, L. 47, 60-1, 94, 128, 201, Lewin, K. 219 Lloyd, B. 38, 43, 92, 96-7, 105, 109, 113-4, 139, 157-170 Logic 5-6, 8-10, 43, 45, 47, 49-54, 58, 6063, 69-70, 75, 80, 82, 84, 86, 95-7, 119, 125-7, 133, 137, 160, 169, 174-5, 202, 209 Marková, I. xvii, 56-7, 137, 150, 197 Marx, K. 199 Mead, G.H. xviii, 182 figumental illness 118, 211, 213 mentality: children 44-5, 47, 59-60, 62-3, 126-8; primitive 47, 60, 201 microgenesis xxii, 149-50, 177, 179-80 minorities xviii, 137-8 moral judgment 24-5, 130, 139 Moral Judgment of the Child 115, 130 Moscovici, S. xi-xii, xxii, 56-9, 61-3, 86-7, 90-4, 99, 101-2, 109, 116, 118, 122,

133, 137-8, 159-60, 169, 173-5, 177-80, 188, 196-202, 207-10, 214, 217-20 Mugny, G. 78, 94, 96, 114, 135-6, 138, 146, 161, 180 observation: of children’s thinking 48-9; and interpretation 43-4; participant 45, 47 operational 58, 77, 116, 134, 138, 168, 175 ontogenesis xvii, xxii, 92, 137, 149-50, 177-8, 180, 208 Parsons, A. 196-8, 210 Perret-Clermont, A-N 18, 72, 78, 81, 114, 135-6, 138, 151, 169 Philogène, G. 120, 205 Piaget, J. 3, xi, xvi-xxii,10-11, 17-8, 20-1, 23, 42-54, 56-63, 68-70, 72-82, 84-87, 90-92, 94-6, 99-100, 107-9, 115-17, 119-20, 124-8, 130-9, 150-1, 160, 166, 168-9, 174-5, 177, 188, 194, 199 Play, Dreams and Imitation in Childhood 45, 95 Play xvii, 6, 97, 104-7, 130, 163-4, 185-90, 208 pre-operational 9, 52, 114-16 Psaltis, C. xvi, 120, 138-9, 142, 146, 151, 208 Psychoanalysis xxii, 45, 59, 158, 175-8, 198, 201, 209, 217-19, 221 Psychoanalysis: Its image and its public (La Psychoanalysis) xxii, 62, 217, 219 Potter, J. 124, 176 power xviii, 11, 14, 16, 68, 93-4, 96, 100, 104, 107-8, 115-16, 121, 133-4, 136, 139, 141, 183, 190-2, 199, 212, propaganda 91, 209, 217, 219-21 realism 47, 59-60 recognition 10, 20, 31, 33, 38, 72, 74-5, 108, 143, 158, 162, 165-6, 183, 217-19; explicit 143, 147-9; instrumental 141, 144-6 resistance xviii, xxii, 51, 68, 103, 113, 139, 143-44, 146-9, 190-3 Rogoff, B. 102, 208 rules 31, 47, 67, 72, 84, 97,108, 103, 176 Sartre, J.P. xv, xix, 67, 83-4, 99, 170, 182, 192 schema xviii, 37, 69 school 34, 44-5, 67-8, 93, 97-8, 104-09, 187-91 sexuality 97, 188, 190

Index 229 Shweder, R. 42, 93, 192 Signs 104, 108, 196, 199-200 social actors xv, 43, 92-3, 134-5, 182, 184, 188, 218-9 social change 61, 104, 210 social identity theory 183, 191-2, 219 social influence 28-9, 62-3, 83, 86, 95, 116, 122, 136-8, 163, 179, 189, 193, 204, 220-1 social psychological imagination 61-2, 220 social psychology: genetic xix, 91, 137, 169, 177; systematic 91 social relations: symmetrical and asymmetrical 97, 100, 108, 116, 133-4 social representations: anchoring and objectification 61, 91, 109, 174, 203; consensual and reified universes 174, 200; development of xxvii, 39, 91-4, 100, 108, 161, 179; of economics 29-40, 166-8; epistemology of 112-3; hegemonic, emancipated and polemic 202-4, 214; phenomenon of 56-7, 63; of gender 35, 38-9, 92-4, 104-8, 139, 146, 149, 157, 161-6, 168, 177-8, 187-8,190; and social identities 139-142, 159-160, 178-9; sociogenesis 92

social transmission 29, 33-4, 40, 63, 96, 99, 107-8, 115, 133, socio-cognitive conflict 18, 87 Sociological Studies xvi, 76, 95 Solidarity xx, 219-20 Tajfel, H. 85-6, 170, 183, 191-2 themata 57, 60-2, 197 thick description 43, 52 thinking: adult 44, 119-20; egocentric 47, 60; of Children 5-9, 24-5, 44, 46-8, 52-3, 59-60, 94, 119-20, 133; operational 58, 116, 168; magical 47, 60, 63, 119-20; primitive 60; society 56, 100, 178 Valsiner, J. ix-x, 197 Vygotsky, L.S. ix, xvi, xix, xxii, 42, 61, 72, 90-2, 100-2, 107-8, 134, 169, 199 Wagner, W. 63, 112, 118, 209, 211, Wittgenstein, L. xix, 85, 113, 201 zone of proximal development 72, 101-2, 104, 108, 134

Taylor & Francis

eBooks F 0 R liBRARIES Over 23,000 eBook titles in the Humanities, Social Sciences, STM and Law from some of the world's leading imprints.

Choose from a range of subject packages or create your own! ~ Free MARC records

~ COUNTER-compliant usage statistics ~ Flexible purchase and pricing options ~ Off-site, anytime access via Athens or referring URL

~ Print or copy pages or chapters

~ Full content search ~ Bookmark, highlight and annotate text ~ Access to thousands of pages of quality research

at the click of a button For more information, pricing enquiries or to order a free trial, contact your local online sales team . UK and Rest of World: online.sales®tandf.co.uk US, Canada and Latin America: e-reference®taylorandfrancis.com

www.ebooksubscriptions.com

DEVELOPMENT AS A SIOCIAL PROCESS as a Siocial Process

Recommend Documents