THENEWLIBRARYOFPSYCHOANALYSIS

G e n e r a l E d i t o r : D a n a B i r k s t e d - B r e e n Projective Identification The Fate of a

Concept

Edited by

Elizabeth Spillius and

Edna O’Shaughnessy

PUBLISHED IN ASSOCIATION WITH THE INSTITUTE OF PSYCHOANALYSIS,

LONDON

Projective Identifi cation

In this book Elizabeth Spillius and Edna O’Shaughnessy explore the development of the concept
of projective identifi cation, which had important antecedents in the work of Freud and others,
but was given a specifi c name and defi nition by Melanie Klein. They describe Klein’s
published and unpublished views on the topic, and then consider the way the concept has been
variously described, evolved, accepted, rejected and modifi ed by analysts of different schools of
thought and in various locations – Britain, Western Europe, North America and Latin America.
The authors believe that this unusually widespread interest in a particular concept and its varied
‘fate’ has occurred not only because of beliefs about its clinical usefulness in the psychoanalytic
setting but also because projective identifi cation is a universal aspect of human interaction and
communication.

Projective Identifi cation: The Fate of a Concept will appeal to any psychoanalyst or
psychotherapist who uses the ideas of transference and counter-transference, as well as to
academics wanting further insight into the evolution of this concept as it moves between
different cultures and countries.

Elizabeth Spillius studied general psychology at the University of Toronto (1945), social
anthropology at the

University of Chicago, The London School of Economics and The Tavistock Institute of Human
Relations (1945–

1957) and psychoanalysis at the Institute of Psychoanalysis in London (1956 to the present). Her
main writings have been Family and Social Network (1957, writing as Elizabeth Bott), Tongan
Society at the Time of Captain Cook’s Visits (1982), Melanie Klein Today (1988) and Encounters
with Melanie Klein (2007). She is the Hon Archivist of the Melanie Klein Trust.

Edna O’Shaughnessy came to psychoanalysis from philosophy she trained fi rst as a Child
Psychotherapist at the Tavistock Clinic in the 1950s, and then in the 1960s, she trained at the
British Psychoanalytical Society, of which she is a training and supervising analyst and also a
child analyst. Her many published papers are written from both a clinical and a conceptual
perspective.

THE NEW LIBR ARY OF PSYCHOANALYSIS

General Editor: Alessandra Lemma

The New Library of Psychoanalysis was launched in 1987 in association with the Institute of
Psychoanalysis, London. It took over from the International Psychoanalytical Library which
published many of the early translations of the works of Freud and the writings of most of the
leading British and Continental psychoanalysts.

The purpose of the New Library of Psychoanalysis is to facilitate a greater and more widespread
appreciation of psychoanalysis and to provide a forum for increasing mutual understanding
between psychoanalysts and those working in other disciplines such as the social sciences,
medicine, philosophy, history, linguistics, literature and the arts. It aims to represent different
trends both in British psychoanalysis and in psychoanalysis generally. The New Library of
Psychoanalysis is well placed to make available to the English-speaking world psychoanalytic
writings from other European countries and to increase the interchange of ideas between British
and American psychoanalysts. Through the Teaching Series , the New Library of Psychoanalysis
now also publishes books that provide comprehensive, yet accessible, overviews of selected
subject areas aimed at those studying psychoanalysis and related fi elds such as the social
sciences, philosophy, literature and the arts.
The Institute, together with the British Psychoanalytical Society, runs a low-fee psychoanalytic
clinic, organizes lectures and scientifi c events concerned with psychoanalysis and publishes the
International Journal of Psychoanalysis . It runs a training course in psychoanalysis which leads
to membership of the International Psychoanalytical Association – the body which preserves
internationally agreed standards of training, of professional entry, and of professional ethics and
practice for psychoanalysis as initiated and developed by Sigmund Freud. Distinguished
members of the Institute have included Michael Balint, Wilfred Bion, Ronald Fairbairn, Anna
Freud, Ernest Jones, Melanie Klein, John Rickman and Donald Winnicott.

Previous general editors have included David Tuckett, who played a very active role in the
establishment of the New Library. He was followed as general editor by Elizabeth Bott Spillius,
who was in turn followed by Susan Budd and then by Dana Birksted-Breen.

Current Members of the Advisory Board include Liz Allison, Giovanna di Ceglie, Rosemary
Davies and Richard Rusbridger.

Previous Members of the Advisory Board include Christopher Bollas, Ronald Britton, Catalina
Bronstein, Donald Campbell, Sara Flanders, Stephen Grosz, John Keene, Eglé Laufer,
Alessandra Lemma, Juliet Mitchell, Michael Parsons, Rosine Jozef Perelberg, Mary Target and
David Taylor.

The current General Editor of the New Library of Psychoanalysis is Alessandra Lemma, but this
book was initiated and edited by Dana Birksted-Breen, former General Editor.

ALSO IN THIS SERIES

Impasse and Interpretation Herbert Rosenfeld

Psychoanalysis and Discourse Patrick Mahony

The Suppressed Madness of Sane Men Marion Milner The Riddle of Freud Estelle Roith
Thinking, Feeling, and Being Ignacio Matte-Blanco The Theatre of the Dream Salomon Resnik
Melanie Klein Today: Volume 1, Mainly Theory Edited by Elizabeth Bott Spillius Melanie Klein
Today: Volume 2, Mainly Practice Edited by Elizabeth Bott Spillius Psychic Equilibrium and
Psychic Change: Selected Papers of Betty Joseph Edited by Michael Feldman and Elizabeth Bott
Spillius

About Children and Children-No-Longer: Collected Papers 1942–80 Paula Heimann. Edited by
Margret Tonnesmann

The Freud–Klein Controversies 1941–45 Edited by Pearl King and Riccardo Steiner Dream,
Phantasy and Art Hanna Segal

Psychic Experience and Problems of Technique Harold Stewart Clinical Lectures on Klein and
Bion Edited by Robin Anderson From Fetus to Child Alessandra Piontelli

A Psychoanalytic Theory of Infantile Experience: Conceptual and Clinical Refl ections E.

Gaddini. Edited by Adam Limentani
The Dream Discourse Today Edited and introduced by Sara Flanders The Gender Conundrum:
Contemporary Psychoanalytic Perspectives on Femininity and Masculinity Edited and introduced
by Dana Birksted-Breen Psychic Retreats John Steiner

The Taming of Solitude: Separation Anxiety in Psychoanalysis Jean-Michel Quinodoz

Unconscious Logic: An Introduction to Matte-Blanco’s Bi-logic and its Uses Eric Rayner
Understanding Mental Objects Meir Perlow Life, Sex and Death: Selected Writings of William
Gillespie Edited and introduced by Michael Sinason What Do Psychoanalysts Want?: The
Problem of Aims in Psychoanalytic Therapy Joseph Sandler and Anna Ursula Dreher Michael
Balint: Object Relations, Pure and Applied Harold Stewart Hope: A Shield in the Economy of
Borderline States Anna Potamianou Psychoanalysis, Literature and War: Papers 1972–1995
Hanna Segal Emotional Vertigo: Between Anxiety and Pleasure Danielle Quinodoz Early Freud
and Late Freud Ilse Grubrich-Simitis A History of Child Psychoanalysis Claudine and Pierre
Geissmann Belief and Imagination: Explorations in Psychoanalysis Ronald Britton

A Mind of One’s Own: A Kleinian View of Self and Object Robert A. Caper Psychoanalytic
Understanding of Violence and Suicide Edited by Rosine Jozef Perelberg

On Bearing Unbearable States of Mind Ruth Riesenberg-Malcolm. Edited by Priscilla Roth

Psychoanalysis on the Move: The Work of Joseph Sandler Edited by Peter Fonagy, Arnold M.
Cooper and Robert S.

Wallerstein

The Dead Mother: The Work of André Green Edited by Gregorio Kohon The Fabric of Affect in
the Psychoanalytic Discourse André Green The Bi-Personal Field: Experiences of Child
Analysis Antonino Ferro The Dove that Returns, the Dove that Vanishes: Paradox and Creativity
in Psychoanalysis Michael Parsons Ordinary People, Extra-ordinary Protections: A Post-
Kleinian Approach to the Treatment of Primitive Mental States Judith Mitrani The Violence of
Interpretation: From Pictogram to Statement Piera Aulagnier The Importance of Fathers: A
Psychoanalytic Re-Evaluation Judith Trowell and Alicia Etchegoyen Dreams That Turn Over a
Page: Paradoxical Dreams in Psychoanalysis Jean-Michel Quinodoz The Couch and the Silver
Screen: Psychoanalytic Refl ections on European Cinema Edited and introduced by Andrea
Sabbadini In Pursuit of Psychic Change: The Betty Joseph Workshop Edited by Edith
Hargreaves and Arturo Varchevker The Quiet Revolution in American Psychoanalysis: Selected
Papers of Arnold M.

Cooper Arnold M. Cooper. Edited and introduced by Elizabeth L. Auchincloss Seeds of Illness
and Seeds of Recovery: The Genesis of Suffering and the Role of Psychoanalysis Antonino Ferro
The Work of Psychic Figurability: Mental States Without Representation César Botella and Sára
Botella Key Ideas for a Contemporary

Psychoanalysis: Misrecognition and Recognition of the Unconscious André Green The

Telescoping of Generations: Listening to the Narcissistic Links Between Generations Haydée
Faimberg Glacial Times: A Journey Through the World of Madness Salomon Resnik This Art of
Psychoanalysis: Dreaming Undreamt Dreams and Interrupted Cries Thomas H. Ogden

Psychoanalysis as Therapy and Storytelling Antonino Ferro Psychoanalysis and Religion in the
21st Century: Competitors or Collaborators?

Edited by David M. Black

Recovery of the Lost Good Object Eric Brenman. Edited and introduced by Gigliola Fornari
Spoto The Many Voices of Psychoanalysis Roger Kennedy Feeling the Words:
Neuropsychoanalytic Understanding of Memory and the Unconscious Mauro Mancia

Projected Shadows: Psychoanalytic Refl ections on the Representation of Loss in European

Cinema Edited by Andrea Sabbadini

Encounters with Melanie Klein: Selected Papers of Elizabeth Spillius Elizabeth Spillius. Edited
by Priscilla Roth and Richard Rusbridger Constructions and the Analytic Field: History, Scenes
and Destiny Domenico Chianese Yesterday, Today and Tomorrow Hanna Segal. Edited by
Nicola Abel-Hirsch Psychoanalysis Comparable and Incomparable: The Evolution of a Method
to Describe and Compare Psychoanalytic Approaches David Tuckett, Roberto Basile, Dana
Birksted-Breen, Tomas Böhm, Paul Denis, Antonino Ferro, Helmut Hinz, Arne Jemstedt, Paola
Mariotti and Johan Schubert

Time, Space and Phantasy Rosine Jozef Perelberg Rediscovering Psychoanalysis: Thinking and
Dreaming, Learning and Forgetting Thomas H. Ogden

Mind Works: Technique and Creativity in Psychoanalysis Antonino Ferro Doubt, Conviction and
the Analytic Process: Selected Papers of Michael Feldman Michael Feldman. Edited by Betty
Joseph Melanie Klein in Berlin: Her First Psychoanalysis of Children Claudia Frank. Edited by
Elizabeth Spillius The Psychotic Wavelength: A Psychoanalytic Perspective for Psychiatry
Richard Lucas Betweenity: A Discussion of the Concept of Borderline Judy Gammelgaard The
Intimate Room: Theory and Technique of the Analytic Field Giuseppe Civitarese Bion Today
Edited by Chris Mawson

Secret Passages: The Theory and Technique of Interpsychic Relations Stefano Bolognini
Intersubjective Processes and the Unconscious: An Integration of Freudian, Kleinian and
Bionian Perspectives Lawrence J. Brown

Seeing and Being Seen: Emerging from a Psychic Retreat John Steiner Avoiding Emotions,
Living Emotions Antonino Ferro TITLES IN THE NEW LIBRARY OF PSYCHOANALYSIS

TEACHING SERIES

Reading Freud: A Chronological Exploration of Freud’s Writings Jean-Michel Quinodoz

Listening to Hanna Segal: Her Contribution to Psychoanalysis Jean-Michel Quinodoz Reading
French Psychoanalysis Edited by Dana Birksted-Breen, Sara Flanders and Alain Gibeault
Reading Winnicott Edited by Lesley Caldwell and Angela Joyce THE NEW LIBRARY OF
PSYCHOANALYSIS

General Editor Alessandra Lemma

Projective Identifi cation

The Fate of a Concept

Edited by

Elizabeth Spillius and

Edna O’Shaughnessy

First published 2012

by Routledge

27 Church Road, Hove, East Sussex, BN3 2FA

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 Copyright © 2012
Selection and editorial matter, Elizabeth Spillius and Edna O’Shaughnessy; individual chapters,
the contributors The right of the editors to be identifi ed as the authors of this work has been
asserted by them 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 identifi cation 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 Projective identifi cation: the fate of a concept / Edited by
Elizabeth Spillius and Edna O’Shaughnessy.

p.;

cm.

Includes bibliographical references and index.

ISBN 978–0–415–60528–1 (hbk) — ISBN 978–0–415–60529–8 (pbk) 1. Projective identifi

cation. 2. Klein, Melanie, 1882–1960.
3. Psychoanalysis—History. I. Spillius, Elizabeth Bott, 1924– editor. II. O’Shaughnessy, Edna,
editor.

[DNLM: 1. Klein, Melanie, 1882–1960. 2. Projection. 3. History, 20th Century. 4.

Identifi cation (Psychology) 5. Psychoanalytic Theory—history.

WM 193.5.P7]

RC455.4.P76P76

2011

616.89’17—dc22

2010050600

ISBN: 978–0–415–60528–1 (hbk)

ISBN: 978–0–415–60529–8 (pbk)

Typeset in Bembo

by Refi neCatch Limited, Bungay, Suffolk

Paperback cover design by Sandra Heath

Printed and bound in Great Britain by TJ International Ltd, Padstow, Cornwall Contents

Notes on contributors xiii

Acknowledgements xvii

Foreword xix

elizabethspilliusandednao’shaughnessy

Part One

Melanie Klein’s work

1 The emergence of Klein’s idea of projective

identifi cation in her published and unpublished work

elizabethspillius
2 Notes on some schizoid mechanisms

19

melanie klein

Part Two

Some British Kleinian developments

47

3 Developments by British Kleinian analysts

49

elizabethspillius

4 Attacks on linking

61

w.r.bion

ix

Contents

5 Contribution to the psychopathology of psychotic

states: the importance of projective identifi cation

in the ego structure and the object relations of

the psychotic patient

76

herbertrosenfeld

6 Projective identifi cation: some clinical aspects

98

bettyjoseph

7 Projective identifi cation: the analyst’s involvement

112
michaelfeldman

8 Who’s who? Notes on pathological identifi cations

132

ignessodré

Part Three

The plural psychoanalytic scene

147

Introduction 149

elizabethspilliusandednao’shaughnessy

1 The British Psychoanalytic Society

151

9 The views of Contemporary Freudians and

Independents about the concept of projective

identifi cation

153

ednao’shaughnessy

10 The concept of projective identifi cation

167

josephsandler

2 Continental Europe

183

Introduction 185

elizabethspillius

11 Projective identifi cation: the fate of the concept

in Germany
186

helmuthinz

Contents

12 Projective identifi cation: the fate of the concept in

Italy and Spain

204

jorgecanestri

13 Projective identifi cation in contemporary

French-language psychoanalysis

218

jean-michelquinodoz

3 The United States

237

Introduction 239

elizabethspillius

14 Projective identifi cation in the USA: an overview

240

royschafer

15 A brief review of projective identifi cation in

American psychoanalytic literature

245

elizabethspillius

16 Projective identifi cation in the therapeutic process

264
arthurmalinandjamess.grotstein

17 On projective identifi cation

275

thomash.ogden

18 Vicissitudes of projective identifi cation

301

albertmason

4 Latin America

321

Introduction 323

LuizMeyer

19 Projective identifi cation: the concept in Argentina

325

gustavojarast

20 Projective identifi cation: Brazilian variations of the concept

339

marinamassi

xi

Contents

21 Projective identifi cation and the weight of

intersubjectivity 354

juanfranciscojordan-moore

Afterword 365

elizabethspilliusandednao’shaughnessy

References 367
Index 393

xii

Contributors

Melanie Klein was born in Vienna in 1882 and had brief analyses with Sandor Ferenczi in
Budapest and with Karl Abraham in Berlin.

In 1925 she was invited by Ernest Jones to give lectures in England and in 1926 to settle in this
country, where she gradually gained adherents and developed new psychoanalytic ideas,
although she strongly asserted her loyalty to the work of Freud and Abraham. She died in
England in 1960. She is usually credited with introducing the concept of projective identifi
cation, although other analysts mentioned it in passing before her.

Wilfred Bion further developed Klein’s and his own seminal ideas, especially, in the case of
projective identifi cation, in distinguishing between psychotic and normal modes of projective
identifi cation.

Herbert Rosenfeld was born in Germany but came to Britain in the 1930s where he was
analysed by Melanie Klein and became well known particularly for his innovative work in
understanding psychotic patients. In the paper reprinted in this book, he helpfully describes
several different types of projective identifi cation in psychotic patients.

Betty Joseph has been awarded the Mary Sigourney award for outstanding contributions to
psychoanalysis. She is renowned for her interest in psychoanalytic technique, and in particular
for her understanding of the expression of projective identifi cation in the analyst–

patient relationship.

xiii

Contributors

Michael Feldman was born in South Africa but studied psychoanalysis in Britain and is
especially interested in the technique of analysing analyst–patient relationships, including the
expression of forms of projective identifi cation.

Ignes Sodré was born in Brazil but came to England and began her study of psychoanalysis
when quite young. She has been particularly interested in psychoanalytic technique, including
the interpretation of projective identifi cation, and she is also especially interested in the relation
between psychoanalysis and literature.

Joseph Sandler was originally from South Africa and came to Britain when quite young where
he became a member of the Contemporary Freudian ‘group’ of British analysts. He was also
interested in Kleinian ideas, especially projective identifi cation. Many Kleinians have adopted
Sandler’s ideas about the ‘actualisation’ of unconscious phantasies and beliefs, an important
aspect of projective identifi cation.
Helmut Hinz together with Elizabeth Spillius, Jorge Canestri, and Jean-Michel Quinodoz gave
papers on projective identifi cation at the 2002 Congress of the European Federation of
Psychoanalysis in Prague. Dr Hinz is a member of the German Psychoanalytical Association
(DPV) and practices in Tübingen. The paper he gave at the Prague Congress on projective
identifi cation is reprinted in this book.

Jorge Canestri is a member of the Italian Psychoanalytic Association and is also an overseas
member of the Argentine Psychoanalytic Association. The paper he gave at the Prague Congress
in 2002 is reprinted in this book.

Jean-Michel Quinodoz is a member of the Swiss Psychoanalytical Society and practices

psychoanalysis mainly in Cologny, Geneva. His paper in this book is a reworked version of the
paper he gave at the Prague Congress.

Roy Schafer is a member of the American Psychoanalytic Association. He practises mainly in

New York, although he lectures and teaches more widely. He has contributed an original paper
on American views on projective identifi cation to this book.

xiv

Contributors

Arthur Malin is a member of the American Psychoanalytical Association and of the Los
Angeles Institute and Society for Psychoanalytic Studies. In 1966 he and James Grotstein were
the fi rst American analysts to write a paper specifi cally on projective identifi -

cation; that paper is reprinted in this book.

James Grotstein is a member of the American Psychoanalytic Association and of the

Psychoanalytic Centre of California. He and Arthur Malin wrote the fi rst main paper on
projective identifi cation in the United States.

Thomas Ogden is a member of the American Psychoanalytic Association, the Psychoanalytic

Institute of Northern California and the Psychoanalytic Centre of California. He wrote one of the
early (1979) and infl uential American papers on projective identifi cation, which is reproduced
in this book.

Albert Mason originally studied psychoanalysis in England and emigrated to Los Angeles in
1967. He is now a member of the Psychoanalytic Centre of California, the American
Psychoanalytic Association and the British Psycho-Analytical Society. His original contribution
to this book gives numerous clinical descriptions of patients’

expressions of projective identifi cation.

Luiz Meyer practices psychoanalysis in São Paulo, Brazil and introduces the contributions to the
description of the use of the concept of projective identifi cation in Latin America.

Gustavo Jarast of the Argentinian Psychoanalytic Association describes the development of the
concept of projective identifi cation in Argentina.

Marina Massi of the Brazilian Society of Psychoanalysis of São Paulo describes the historical
development of psychoanalysis in Brazil, its many complications because of the multiplicity of
psychoanalytic centres, and the many views on the concept of projective identifi cation.

Juan Francisco Jordan-Moore is a member of the Chilean Psychoanalytic Association and

practices in Santiago.

He describes xv

Contributors

the critical reception of the concept of projective identifi cation in Chile, with particular attention
to attempts by Chilean authors to develop an intersubjective or bi-personal fi eld theory that
would enable a fuller account of the complexity of the phenomena involved.

xvi

Acknowledgements

Elizabeth Spillius and Edna O’shaughnessy wish to express their thanks, fi rst, to all the authors
whose contributions make up this volume. Second, we are grateful to the Melanie Klein Trust
both for their general encouragement and also for their fi nancial help which enabled us to
employ Sophie Boswell and Jessica Brighty, our fi rst two Editorial Assistants, and later to seek
the help of Elizabeth Allison who saw the book through to its conclusion with meticulous editing
and energy. Our thanks are also due to John Churcher for valuable assistance, to Richard
Rusbridger for helpful advice throughout, to Dana Birksted-Breen as Editor of the New Library,
and to the three anonymous readers of our earlier manuscript.

Permissions acknowledgements

Chapter 2 is based on: M. Klein, ‘Notes on some schizoid mecha-

nisms’. International Journal of Psychoanalysis , 27: 99–110, 1946. ©

Institute of Psychoanalysis, London, UK. Material reprinted with permission.

Chapter 4 is based on: W.R. Bion, ‘Attacks on linking’. International Journal of Psychoanalysis ,
40: 308–315, 1959. © Institute of Psychoanalysis, London, UK. Material reprinted with
permission.

Chapter 5 is based on: H.A. Rosenfeld, ‘Contribution to the psychopathology of psychotic states:
The importance of projective identifi cation in the ego structure and object relations of the
psychotic patient. In P. Doucet and C. Laurin (Eds.), Problems of Psychosis, Vol. I. , pp. 115–
128.

The Hague: Excerpta Medica. Material reprinted with permission.

xvii

Acknowledgements

Chapter 7 is based on: M. Feldman, ‘Projective identifi cation: the analyst’s involvement’.
International Journal of Psychoanalysis , 78: 227–241, 1997. © Institute of Psychoanalysis,
London, UK. Material reprinted with permission.

Chapter 11 is based on: H. Hinz, ‘Projective identifi cation: The fate of the concept in Germany’.
European Psychoanalytical Federation Bulletin , 56, 2002.

Chapter 12 is based on: J. Canestri, ‘Projective identifi cation: The fate of the concept in Italy
and Spain’.

European Psychoanalytical

Federation Bulletin , 56, 2002.

Chapter 16 is based on: A. Malin and J.S. Grotstein, ‘Projective identifi cation in the therapeutic
process’.

International Journal of

Psychoanalysis , 47: 26–31, 1966. © Institute of Psychoanalysis, London, UK. Material reprinted
with permission.

Chapter 17 is based on: T. Ogden, ‘On projective identifi cation’.

International Journal of Psychoanalysis , 60: 357–373, 1979. © Institute of Psychoanalysis,

London, UK. Material reprinted with permission.

All reasonable efforts have been made to contact copyright holders, but in some cases this was
not possible. Any omissions brought to the attention of Routledge will be remedied in future
editions.

xviii

Foreword

Elizabeth Spillius and Edna O’Shaughnessy

When the term ‘projective identifi cation’ was introduced into the language of psychoanalysis in
1946, there already existed a rich though unsystematised harvest of knowledge about identifi
cation.

Freud himself had described the primary identifi cations of the ego with its objects, and also how
some identifi cations can divide or impoverish the ego, e.g., hysterical, multiple and group
identifi cations, while others, like the narcissistic identifi cations with the object in melancholia,
alter the ego in complex morbid ways, and furthermore, how the identifi cations that form the
super-ego and the ego ideal structure the mind and cause the ego anxiety through their judgments
of guilt and demands, even as they enrich the personality with ideals. In The Language of
Psychoanalysis Laplanche and Pontalis wrote that in ‘Freud’s work the concept of identifi cation
comes little by little to have the central importance which makes it, not simply one psychical
mechanism among others, but the operation itself whereby the human subject is constituted’
(Laplanche and Pontalis, 1973, p. 206).

Abraham and Ferenczi also contributed to the notion of identifi cation, and Anna Freud noted a
defence mechanism of identifi cation with the aggressor. But it was Melanie Klein who extended
the concept in a far-reaching way. She conceived of identifi cation as creating a whole inner
world from the beginning of life, through introjection and

projection, so that the ego has, as she expressed it,

‘an orientation outwards and inwards’ with ‘a constant fl uctuation between internal and external
objects and situations’ (Klein, 1945, xix

Foreword

p. 409). And then in 1946 in ‘Notes on some schizoid mechanisms’

she described another form of identifi cation, projective identifi cation (in contradistinction to
introjective identifi cation), a defence mechanism used by the early ego in states of anxiety when,
in an omnipotent phantasy, it splits off and projects parts of itself into the object with consequent
identifi cations for ego and object that structure their ensuing relations (Klein, 1946).

Klein’s concept of projective identifi cation has aroused an unusual degree of interest among
psychoanalysts, both because of its hypothesised role in early normal and abnormal
development, and more especially for its illumination of the ways in which communication
happens between patient and analyst. Among Kleinian analysts it became a major area of
development. Among analysts of other schools it has been adopted for essential or occasional
use, and also redefi ned, disputed and rejected from a variety of perspectives.

A concept is linked many times over – to the phenomena that give it meaning, to other concepts
in the language in which it has its place, and to its users who over time develop, enlarge, restrict
and even change its meaning. All these have added to the complex, sometimes confusing or even
contradictory uses of the term ‘projective identifi cation’.

In this book we have several aims: to return to the origins of the concept in the published and
unpublished archival writings of Melanie Klein; to try to understand the complexities of the
concept itself; to chart its further Kleinian development and also its evolution in other places and
other psychoanalytic orientations; to note the grounds for its rejection; and overall, to explore the
conceptual and clinical problems that accompany the further evolution and migration of a
concept from one psychoanalytic theory to another. All of these endeavours have been much
assisted by contributions from colleagues in Britain, Europe, and North and South America.

Introducing
The Language of Psychoanalysis by Laplanche and Pontalis, Daniel Lagache wrote ‘ words, like
ideas (and together with ideas,) are not merely created – they have a fate’ (Laplanche and
Pontalis, 1973, p. viii, our italics). What may be seen in this volume about the creation and the
fate of the concept of projective identifi cation, might, we think, in some measure, apply to other
signifi cant psychoanalytic concepts.

xx

PARTONE

Melanie Klein’s work

The emergence of Klein’s idea of projective

identification in her published and

unpublished work

Elizabeth Spillius

Introduction

I will fi rst describe something of the history of the term ‘projective identifi cation’ followed by
discussion of

Klein’s own ideas about it in the two published versions of her paper ‘Notes on some schizoid
mechanisms’ (Klein, 1946, 1952a) and in her 1955 paper ‘On identifi cation’. Then I will discuss
two sets of unpublished entries on the concept of projective identifi cation in the Melanie Klein
Archive.

History of the concept

Although Melanie Klein was the originator of the defi nition and usage of the concept of
projective identifi cation as we know it today, she was not the fi rst person to use the actual term.
It was fi rst used by Edoardo Weiss, in 1925 in a paper called ‘Über eine noch unbeschriebene
Phase der Entwicklung zur heterosexuellen Liebe’ in the context of explaining sexual object
choice (Weiss, 1925). Klein refers to Weiss’s paper in The Psycho-Analysis of Children (Klein,

1932b, p. 250, n. 2) where she explains Weiss’s understanding of sexual object choice, using the
term ‘projection’ in her explanation but not discussing his use of the term ‘projective identifi
cation’ (see also Massidda, 1999 and Steiner, 1999).

Elizabeth Spillius

There was another precursor to Klein’s use of the term projective identifi cation. In 1945, a year
before Klein published her fi rst version of ‘Notes on some schizoid mechanisms’, Marjorie
Brierley mentions projective identifi cation in a paper called ‘Further notes on the implications of
psycho-analysis: Metapsychology and personology’, in which she says

projective identifi cation of ego-ideal with outer object, human or abstract, would appear to be a
feature of the economy of all fanatics [. . .] the pedestrian everyday charity that begins at home,
as distinct from fanatical devotion of ultra-personal interests, may depend upon projective
identifi cation with a fairly well-libidinized operative self.

(Brierley, 1945, p. 96)

Brierley mentions projective identifi cation again in a second paper in 1947. Brierley and Klein
do not refer to each other’s work, but it looks as if considerable thinking and discussion about
introjection, projection and identifi cation was going on among British analysts in the 1940s and
that projective identifi cation was not so much a special focus as part of this general discussion.

Klein’s published views on projective identifi cation Klein fi rst mentions the idea of
projective identifi cation in her paper ‘Notes on some schizoid mechanisms’ in 1946 but at fi rst
only in passing. In the original version of

‘Notes on some schizoid mechanisms’, which was published in the

International Journal of

Psycho-Analysis in 1946, Klein describes the process of projective identifi cation as follows:
Together with these harmful excrements, expelled in hatred, split off parts of the ego are also
projected on to the mother or, as I would rather call it, into the mother. These excrements and
bad parts of the self are meant not only to injure the object but also to control it and take
possession of it. Insofar as the mother comes to contain the bad parts of the self, she is not felt to
be a separate individual but is felt to be the bad self.

Much of the hatred against parts of the self is now directed towards the mother. This leads to a
particular kind of identifi ca-4

Emergence of Klein’s projective identifi cation tion which establishes the prototype of an
aggressive object relation.

(Klein, 1946, p. 102)

In the next paragraph Klein adds:

It is, however, not only the bad parts of the self which are expelled and projected, but also good
parts of the self.

Excrements then have the signifi cance of gifts; and parts of the ego which, together with
excrements, are expelled and projected into the other person represent the good, i.e. the loving
parts of the self.
(Klein, 1946, p. 102)

In essence these paragraphs are a defi nition of projective identifi cation, but the concept is not
mentioned by name.

The actual name

‘projective identifi cation’ is mentioned not as part of this defi nition but only in a passing
comment two pages later where Klein says, ‘I have referred to the weakening and
impoverishment of the ego resulting from excessive splitting and projective identifi cation’
(Klein,

1946, p. 104); 1 this is the only mention of the concept of projective identifi cation by name in
the 1946 version of the paper. It was not until the 1952 version of ‘Notes on some schizoid
mechanisms’ that Klein added the crucial sentence ‘I suggest for these processes the term
“projective identifi cation” ’ to the defi ning paragraphs quoted above.

Somewhat confusingly, it has become customary for this 1952

version of the paper to be cited as ‘1946’ in the Kleinian literature.

It is the 1952 version that is reprinted in Chapter 2 in the present

book.

The 1952 version of the paper differs from the 1946 version not only in including the term
‘projective identifi cation’ in the defi nition but also in other respects. There are two new
paragraphs, one of which is devoted to projective identifi cation, and there is some rearranging of
the order of other paragraphs. There are also thirteen new footnotes, notably one to Paula
Heimann thanking her for her

1 I am grateful to Riccardo Steiner for pointing out that Klein uses the term ‘projective identifi
cation’ in the 1946

version of her paper even though it is not part of her defi nition of the concept.

Elizabeth Spillius

‘stimulating suggestions’ about Klein’s paper, one to Ferenczi (1930) on fragmentation, two to
Rosenfeld, especially to his papers on depersonalisation (Rosenfeld, 1947) and on male
homosexuality and paranoia (Rosenfeld, 1949), and one to an unpublished paper on paranoid
attitudes by Joan Riviere which I have been unable to track down (Riviere, unpublished).

According to Hanna Segal (personal communication) the term

‘projective identifi cation’ was suggested to Klein by her colleague (and patient) Roger Money-
Kyrle. Segal also said that Klein did not like the term ‘projective identifi cation’ and she added
that Klein thought of the term in the context of comparing projection with introjection. In the
case of introjection, once an object has been taken in several things may happen to it: it may
exist inside the subject as an internal object, good or bad; the subject may unconsciously identify
with this internal object or with an aspect of it; or both processes may occur, that is, the internal
object may be recognised as separate from oneself but it may also be identifi ed with. Klein,
according to Segal, thought of projective identifi cation as a parallel process to introjective
identifi cation, meaning that projective identifi cation was only one of several possible outcomes
of projection, although Klein does not describe what these other outcomes might be.

After 1952 Klein makes comparatively few mentions of projective identifi cation except of
course for her paper ‘On identifi cation’ in 1955 in which identifi cation is the central theme
(Klein, 1955). But this paper is an analysis not of a patient but of a character in a novel who
projects his whole self into various other people in order to acquire their identity. The idea of this
sort of projective identifi cation is still used, but nowadays the concept is more frequently used to
describe the projection of aspects of the individual’s self into other objects.

Apart from this 1955 paper, the most frequent mentions of projective identifi cation come in
1957 in Envy and Gratitude where Klein notes the projective character of envy and the
contribution envy makes to diffi culties in making the basic and primal split between good and
bad experiences and feelings, this split being essential for the development of integration of the
ego. Envy leads to attacks on the good object which take the form of projection of bad parts of
the self into the good object, resulting in confusion between the good and the bad self, between
good and bad aspects of the object, and between self and object.

Emergence of Klein’s projective identifi cation Klein’s unpublished notes on projective

identifi cation in the Melanie Klein Archive 2

If one were to know about Klein’s views on projective identifi cation only from her published
work, one would conclude that she did not think the concept was very important, for she
published comparatively little about it. Once one explores the Klein Archive, however, one fi nds
much more material on the topic, both in the form of theoretical thoughts and in clinical
illustration. One fi le of particular importance, B98, dates from 1946/1947 when Klein was
working on her important paper ‘Notes on some schizoid mechanisms’ (1946), which she always
refers to in the Archive as ‘my splitting paper’, never as ‘my projective identifi cation paper’.
The fi le B98 consists of excerpts of clinical material, sporadic notes and some longer theoretical
thoughts especially on splitting and projective identifi cation.

Other relevant material occurs in parts of fi le D17, which probably dates from the late 1950s and
includes both theory and clinical illustration. The comments on projective identifi cation in fi le
D17 are scattered here and there throughout the fi le.

James Gammill (1989), whose work with a three-year-old child was supervised by Klein in the
late 1950s, has told me that she talked to him in considerable detail about the way his patient was
using projective identifi cation. Thus, although Klein published so little about projective identifi
cation, it was important in the thoughts she was formulating and using in her supervisions in the
late 1950s.

2 The Melanie Klein Archive was given to the Wellcome Library for the History of Medicine by
the Melanie Klein Trust in 1984 and has been catalogued by their excellent archivists. The part
of the Archive with which I am mainly concerned has been divided into B, C, and D fi les. B fi
les consist of Klein’s clinical notes on patients. C fi les consist of her papers, published and
unpublished. D fi les consist of her notes on theory and various other matters. In quoting from
the Archive I have made minor corrections of spelling but I indicate other changes or comments
by putting them within square brackets [ ]. When Klein herself puts brackets in her text, I use
rounded brackets ( ). All fi les are prefaced by PP/KLE followed by the letter and number of the
fi le. In most fi les the pages are not numbered consecutively so that I have not been able to cite
page numbers.

Elizabeth Spillius

The notes of fi le PP/KLE, B98, ‘Patients’ material: theoretical thoughts’

This fi le of 106 pages dates from 1946 and 1947, the year before and just after the time when
Klein’s 1946 version of ‘Notes on some schizoid mechanisms’ was fi rst published. Some pages
give theoretical formulations, some give clinical material; some are generalisations about several
patients; some give the material of a single patient. The pages are not at all systematically
arranged and they seem to have been hastily written or perhaps dictated; almost all are typed.
One gets the impression that for this fi le Klein put down or dictated examples and thoughts on
splitting, identifi cation, projection and introjection spontaneously as they occurred to her.
Because fi le B98 fi le is so long it

will not be reproduced in the present book, but I will quote from it in some detail.

Although this material of B98 is interesting in itself, it is not of very much help in showing how
Klein arrived at the novel formulations of her 1946 ‘splitting’ paper. It seems to me that there is
a huge conceptual leap from the notes of B98 to the paper, which describes her formulation of
the paranoid-schizoid position for the fi rst time.

In addition, only one of the clinical examples given in the published paper is to be found in B98;
nor could I fi nd the clinical examples of the paper in any other part of the Archive. I think it
likely that the clinical material and theoretical thoughts of B98 formed a general background for
Klein’s thinking rather than a specifi c source of inspiration for the new ideas of ‘Notes on some
schizoid mechanisms’.

There are certain themes that Klein repeatedly refers to in the notes of B98.

The fi rst theme is that good as well as bad aspects of the self are involved in projective
identifi cation. Klein also states this view in her paper ‘Notes on some schizoid mechanisms’,
although she puts less stress on it than she does in the Archive notes, and most of the literature
since Klein’s work has stressed the projection of bad aspects rather than good. Klein gives
several examples of the projection of good aspects of the subject’s self, as for example in the
case of Patient H, who projected good aspects of himself into one person and bad aspects into
another, his analyst:

June 25th, 1946

Patient H felt particular gratifi cation in a situation which could have stirred his jealousy and
envy through comparison with 8

Emergence of Klein’s projective identifi cation somebody but in fact was only felt to be
gratifying. The enviable object – Mrs. X – not only represented an ally in what was felt to be a
good cause which she had particularly well dealt with (that is where envy would have come in);
but as a good aspect of H

himself. It appeared that he felt that he had put everything good he possessed into her. That is
how she came to represent himself.

During this hour he had a strange feeling of being quite estranged from himself and could not
account for that. He had been very satisfi ed with progress in analysis and insight gained, and
was struggling to maintain an exclusively friendly and grateful relation to me. I could show him
that the satisfaction led to an increased greedy wish to get more, and that he was trying to
prevent his greed because he felt that he would enter my mind violently to rob me.

The interpretation was that he was putting into myself all his valueless products, representing
faeces and urine, and in this process taking such full possession of me when his products had
also come to represent himself. At this moment he said, ‘I would wish to get out’ with quite a
strong physical feeling of breathless-ness and oppression. The interpretation being that he would
like to withdraw himself out of me which was followed by physical relief. Now we could
connect the feeling of strangeness to himself, depersonalisation, with the feeling that he had been
inside me, and getting himself out of me had felt that he had left so much of his personality
inside me that he was estranged to what he took back.

On the other hand, his identifi cation with Mrs. X was based on the feeling that he had put his
good things into her and therefore could enjoy without rivalry or envy her accomplishments.

(Klein Archive, PP/KLE, B98)

Klein’s second main theme in the notes of B98 is that projection and introjection go
together . She says, for example: To Schizoid paper

Chapter Note (to p.18)

Projective identifi cation is the basis for many anxiety situations of a paranoid nature. Since
projection and introjection operate simultaneously, paranoid anxieties focus on persecution
within the self and within the object

which the self has forcefully entered. The attempts to control an external object by entering it
give rise to the fear of being controlled and persecuted by this object. The subject 9
Elizabeth Spillius

may be unable to withdraw from this object; it is kept imprisoned and subjugated by the object.
Once part of the ego, or the whole ego, might be felt to have got lost for ever etc. (I have
described formerly, in The Psycho-Analysis of Children , such fears as being not only at the
bottom of paranoid anxiety but also as a cause for disturbances in the male’s sexuality –
impotence – and as a basis for claustrophobia.) In addition, the re-introjection of this object,
which now represents a combination of a persecutory object and the bad self, reinforces inner
persecution. The accumulations of anxiety situations of this nature – particularly the fact that the
ego is, as it were, caught between external and internal anxieties – is one of the basic features in
paranoia. (Cf. chapter on ‘Notes on some schizoid mechanisms’ also H. Rosenfeld’s paper
‘Analysis of a Schizophrenic state with Depersonalization’, IJPA Vol.

XXVIII, 1947, and Joan Riviere’s paper. [Title? Published?])

(Klein Archive, PP/KLE, B98)

She puts it more concisely in another entry:

To Theoretical Refl ections

Vampire like sucking. In the fi rst sucking of sadistic character you do not only suck out but you
put yourself in.

Projective identifi cation already there as a complementary process to earliest greedy introjection
of the breast.

I believe that persecutory fear of a greedily introjected object

– and later the guilt regarding this – contributes to projection and projective identifi cation.

(Klein Archive, PP/KLE, B98)

Klein also gives a specifi c example, once again from Patient ‘H’.

September 20, 1946

To Splitting paper

Extension into the body, and loneliness derived from the fear of destruction of the object from
which one is parting –

this has been explained as the fear of loss of the object because of one’s own sadistic desires
against it. Now in one case, ‘H’ [this phrase and the letter ‘H’ are crossed out] I have found that
panic connected with the fear of parting or being left alone derived from the feeling that one part
of himself remained in me and that he could not 10

Emergence of Klein’s projective identifi cation withdraw that before going. The interpretation
was that the great grievance and hatred to which parting and being left alone gave rise made the
part of himself left inside me particularly vicious. I was therefore not only [left, but left] in great
danger. But the feeling of being weakened by such an important part of himself being in me
increased the feeling of dependence and anxiety.

Correspondingly we found in the same hour again the feeling of bits inside. The interpretation
was that to this destruction wrought on me by one part of himself acting so ferociously inside me
corresponded a similar state of me as an internal object inside him. This internal ‘me’ was also a
source of great persecution and danger to him, as for that matter I was as an external object when
I was treated by his being in me in that way. Both relations – me inside him and him inside me –
were connected with a feeling of narcissistic withdrawal, loss of feelings, incapacity of re-
establishing the relation to me and to others. Also between these situations on the one hand of
being overwhelmed by an internal me or containing a destroyed internal me, and the situation
described before (the external one) – there was hardly anything of his personality left. The
remaining of feeling and an increased condition of the ego as a

consequence of such interpretations [was]

strengthened.

(Klein Archive, PP/KLE, B98)

A third major theme in the material of the B98 fi le is that a person’s sense of his own
identity is built up around the internalised good object . If this internalisation is not secure,
the person resorts to intensive but

‘unselective’ introjection and to equally intensive and piecemeal projections of aspects of the
self.

This is how Klein puts it:

One way of formulating the focus which the good internal object provides: the ego builds it up
[this probably means

‘builds itself up’ – EBS] from the identifi cations. Round the primary good objects cluster the
parts of the self which are then as it were held together by this strong identifi cation. If it is not
strong enough, the great need of more and more identifi cations, anybody is taken in, and the
connection between this and the unselective throwing out again and dispersing again.

(Klein Archive, PP/KLE, B98)

11

Elizabeth Spillius

Klein describes the process in an example from a patient whom she calls ‘J’.

April 13, 1946

Patient J.: An hour which is characterised by greed and frustration coming up in connection with
the holidays.

During that hour he refers to a decision to be made, on which he would wish to have my advice
because I have common sense and sound judgement. At one point he mentions a friend, X, who
suspects his relative who keeps house for him of taking the cream off the milk. He relates that to
his relation with me, wishing to be fi rst in order to get the cream of the milk, but then the people
following would be deprived, about which he both triumphs and is sorry.

Rivals. Then he speaks of X as being delusional on that point.

K interprets that at each of these points he was identifi ed, or rather represented by the relative
who steals the cream, the rival who is being deprived, the neglected child, the friend who is
delusional, which is in contrast to the other fact where the cream is in fact in his mind taken
away from the milk. He then adds: ‘There must be another part there who watches all that and
tries to be common sense.’

K interprets four split off aspects which seem incompatible with one another. At this moment –
though very impressed at the beginning with the interpretation, there is great diffi culty in
following and understanding what I just said. K interprets that this diffi culty arises from the
thought processes being as split as these different aspects are. Also that his common sense aspect
is in identifi cation with me who was common sense at the beginning of the hour; that round this
identifi cation, which is also part of himself, he is trying to bring together the split off aspects.

(Klein Archive, PP/KLE, B98)

Klein also touches on this matter of the origin of a weak and divided sense of identity in the
description of two sessions with Patient ‘H’ in October, 1946. In the fi rst session (which I will
not report in full) Klein had interpreted that H was suffering from a lack of emotion and that this
was a defence. In the next session H had recovered his capacity to feel. In fact he felt very
anxious and feared a kind of explosion in which he would blow up into bits, or

be 12

Emergence of Klein’s projective identifi cation blown up. Then he had a vision of a monkey, an
unemotional monkey, sitting and eating banana after banana. This was almost immediately
followed by another vision of a very wild monkey tearing and breaking everything up.

Klein interpreted that the senseless force he was anxious about was his own impulse, one part of
himself destroying and disintegrating him.

Following this hour, H had a very striking experience. Somebody rang him and he seemed
interested in what he was told but found it diffi cult to follow. In the conversation he tried to
bring together two points of this special topic and could not fi nd them. He felt himself talking
round about. He was left with an increasing state of anxiety. This conversation did not in fact
contain anything which on the surface one would say needed to stir anxiety in him. Later on he
suddenly felt that the very fact that he could not bring things together during this conversation,
and that his own thoughts had become confused, had roused anxiety in him. He suddenly saw the
monkey in front of him, and now he had torn all his thoughts and words to bits.

In the analysis we found that in this conversation he had suddenly become identifi ed with the
person who had been mentioned in the telephone talk, with the person who telephoned and with
himself in a specially unpleasant situation because the person referred to was an ill person, and
anxiety of illness had been stirred strongly through the recent material. All these quick projective
identifi cations related to the analytic work of the last few sessions and increased the feeling of
being in bits and being unable to integrate himself. He had been persons A, B, C, D, during this
conversation and now he was exposed again to this terrible force represented through the tearing
monkey who, as he suddenly felt, tore his thoughts and words to bits.

All these mechanisms were found in somebody who seems

quite normal, just someone who tells things in a rather slow, deliberate way, a type one often
meets.

(Klein Archive, PP/KLE, B98)

Finally, Klein describes a different sort of patient, a man who is not uncertain about his identity,
but is worried about the way he controls women by projective identifi cation.

13

Elizabeth Spillius

Patient ‘M’

October 12, 1946

The infl uence the projective identifi cations have on sexual intercourse are seen quite clearly in
somebody whose analysis has not been carried to any length yet. His fear of infl uencing and
moulding the women he is interested in such a way that they are greatly changed and become
really like himself. He saw with dismay that a girl he likes and who likes him had changed her
style of dressing in the way in which he sometimes likes women to be dressed and he called this
‘the thin edge of the wedge’. In the same hour he told me that he does not like this infl uence at
all and is strongly trying to prevent it in present relations. He speaks with great concern about an
earlier relation in which this seemed to be one of the factors which made the girl too fond, too
dependent on him and fi nished unsatisfactorily because he cannot bear too great dependence in
the woman. Somebody said that he is apt to choose people (in working conditions) who are so
receptive to his ideas that they will make a perfect staff. In referring to this infl uence he said:
‘They become really too much like myself and [I] don’t want to see so much of myself about.’
The sexual relations too were obviously impeded and infl uenced by these fears. All
relationships, but particularly with women, were affected. He does not seem to feel having such
powers over men.

Conclusion: The penis being used as a controlling object, as an object to be split off, and then the
mechanism of splitting is very active. Not only faeces are split off, but parts of the body which
are entering the body and
controlling it. Now the penis is then felt to remain inside in a controlling, guiding, etc. way. That
too must have a bearing on diffi culties in potency, because if it is too much a sent out part of
oneself it impedes the capacity.

(Klein Archive, PP/KLE, B98)

Klein also notes that the projection of goodness may deplete the self. She says: NOTE

The importance of projection and putting out things from the self also in the service of goodness
has its effect on the process of depersonalisation. The goodness put out of the self into the other
person, and then trying to get it back – a hardly recognisable 14

Emergence of Klein’s projective identifi cation part of the self. One of the important factors in
overdoing this process – the intensity of these processes has already to do with fear and guilt.
Here also enters the sacrifi cial attitude – to take over the badness and suffering from the other
object in order that it should remain perfect and unharmed.

(Klein Archive, PP/KLE, B98)

I have quoted these excerpts from the B98 part of the Klein Archive to give an idea of the
complexity of the way Klein used the idea of projective identifi cation and of the multiplicity of
motives for it. Her brief account of projective identifi cation in ‘Notes on some schizoid
mechanisms’ does not do justice to the variability of patients’

projective identifi cations and the complexity of the relations between projective and introjective
identifi cations.

Klein’s examples give an idea of the richness and fl exibility of her use of her concept, although
she does not try to describe this in a systematic way either in her notes or in her publications.

Summary of the main themes of B98 in 1946 and 1947

First, good as well as bad aspects of the self may be projected and identifi ed with.

Second, projection and introjection usually go together.

Third, if there is no strong and stable internalisation of a good internal object around which
identifi cations can cluster, the ego cannot be satisfactorily built up and introjective and
projective identifi cations will be not only complex but also fragmented, unselective and
unstable.

The fourth theme is closely related to the third: the ego is likely to be weakened by loss of good
and bad parts of the self through excessive projection.

Klein’s unpublished notes on projective identifi cation in 1958: Klein Archive, PP/KLE,
D17, entitled ‘Klein: Technique. New Notes on Technique’

The fi le D17 covers many topics. It consists of 99 ‘frames’ (meaning pages) of which 11 are
devoted to projective identifi cation. One 15

Elizabeth Spillius

page of these notes is dated 1958 and it seems likely that the other pages may date from the same
time. (The 11

pages of D17 that are devoted to projective identifi cation have been printed in Spillius, 2007,
pp. 121–126.) On the fi rst frame (799) of these 1958 notes on projective identifi cation, Klein
states in handwriting: ‘To be used in a paper on projective identifi cation’. I did not fi nd any
such paper in the Archive, but it looks as if by this time, 1958, the term ‘projective identifi
cation’ was worth writing about in its own right. It had apparently acquired an identity of its own
and was no longer thought of mainly in the context of general discussions of introjection,
projection and identifi cation.

In these unpublished notes from D17, Klein states several views on projective identifi cation.

First, she distinguishes between projection and projective identifi cation as two steps in the
same process. In the fi rst step, which she calls ‘projection’, something of oneself that is very
unpleasant or something that one feels one does not deserve to have is attributed to somebody
else. In the second step, which is ‘projective identifi cation’, this something, good or bad, is split
off from the self and put into the object. The two steps, she says, ‘need not be simultaneously
experienced, though they very often are’ (Klein Archive, PP/KLE, D17, frames 802, 840).

I have not found this distinction useful – in fact I fi nd it diffi cult to see how the second step is
really different from the fi rst – and so far as I can see, none of Klein’s colleagues has adopted
her distinction.

Second, Klein thinks it is essential for the analyst to project himself into the patient in order
to understand him, and that it is also essential that the analyst should introject the patient.

She says:

Only is [if ] the analyst can project himself into the analysand will he be able to understand him
deeply enough [. . .]

A conclusion from what I am saying would be that an optimum in identifi cation with the patient,
both by introjection and projection, is essential for a deeper understanding with the patient,
together with a capacity to regain one’s own self and ego suffi ciently not to be misled by the
identifi cation.

(Klein Archive, PP/KLE, D17, frame 804)

16

Emergence of Klein’s projective identifi cation Third, as in her published paper (1946, 1952a)
Klein stresses that good as well as bad parts of the self may be projected into the object (D17,
frame 801). There may be several motives for projecting bad aspects of oneself or for making
‘bad’ penetrations into the object: to get rid of something bad in oneself; greedily to control or
rob the object; or to satisfy one’s aggressive curiosity (D17, frame 805). The motive she cites for
projecting good aspects of oneself is that one feels one does not deserve to have such aspects.

Fourth, Klein says nothing about countertransference in these notes of 1958, but we know from
notes in other parts of the Archive that Klein did not agree with the idea that the analyst’s
countertransference could be a useful source of information about the patient (Klein Archive,
C72, frames 695–724). This is partly because, like Freud (1910), she defi nes
countertransference as a sign of pathology in the analyst. She does imply, however, in the notes
on projective identifi cation in the D17 part of the Archive that the analyst is bound to have a
distinct emotional reaction to the patient, and she says explicitly that the analyst may be
somewhat anxious both about projecting himself into the patient and about the patient’s
projecting himself into the analyst (D17, frames 805–806). ‘But again,’ she adds,

‘if the analyst possesses the strength of ego and the other qualities to which I referred earlier, the
anxiety of the patient projecting himself into him will not disturb him, and he can then analyse
the projection of the patient’ (D17, frame 806). She seems to assume that the analyst’s emotional
reaction needs to be overcome because it will interfere with his capacity to think analytically.

Fifth, I believe that Klein assumes throughout these notes, although she does not explicitly say
so, that projective identifi cation takes the form of an unconscious phantasy , and I assume
that she thought this was also true of introjective identifi cation.

Finally, Klein briefl y describes a clinical session in which she shows how she analyses a
particular instance of projective identifi cation (Klein Archive, D17, frames 802–803). Because
she gives so little published clinical illustration of her ideas about projective identifi cation, I will
quote her example here.

In this connection, I wish again to stress the necessity to go step by step according to the
emotions, anxieties, etc.

activated in the patient, and not to run ahead because the analyst knows already what is behind
that. I have an example of this point. The analysand 17

Elizabeth Spillius

speaks in an early session of his analysis of some experience in . . .

during the War. They had been warned that there were man-eating tigers about, but they had not
met any. Previous to that, his suspicions of me, very much stimulated by remarks made [by
others] about me, had come up and had led to his distrust of his mother. His mother was
supposed to have said, as an aunt reported to him, that, being in very bad circumstances, if she
were starving, she would eat her son. Though Mr [X] actually knew quite well that that was not
what she meant, he had never forgotten this remark, and it had come up in his suspicions of me,
together with his suspicions of my possessiveness, dangerousness, etc. I fi rst analysed these
suspicions and linked them with his suspicions of his mother, who had actually not been able to
give him enough food at a certain period in his life, and linked these with the man-eating tiger he
had mentioned, his fear that she would eat him and that she was starving him for bad and
dangerous reasons. [He] went on about his stay in . . . and said that they had never actually met a
man-eating tiger, but had met a bear, and then added laughingly that the bear did not eat them
they ate the bear. My interpretations were then fully supported, that it was his wish to eat his
mother’s breast and that led to his suspicion that his mother was devouring him. At that moment
[he] felt that the plants on my desk actually belonged to him, corrected this in the next moment,
but found that this was a confi rmation of what I said, because he had appropriated something
that belonged to me.

(Klein Archive, D17, frames 802–3; this example is also cited in Spillius, 2007, pp. 110–111) It
is worth noting here that Klein does not use her distinction between projection and projective
identifi cation in this brief account.

It seems clear that she assumed that projection and projective identifi cation were occurring
simultaneously in this instance.

As I discuss in Part Two of this book, some of these points have

been worked on and developed by several of Klein’s British Kleinian colleagues. Some are no
longer adhered to, and some are used by some colleagues but not by others.

This introduction to Klein’s published and unpublished work on the topic of projective identifi
cation is here followed by reprinting of the 1952 version of Klein’s ‘Notes on some schizoid
mechanisms’.

18

Notes on some schizoid mechanisms 3

Melanie Klein

Introduction

The present paper is concerned with the importance of early paranoid and schizoid anxieties and
mechanisms. I have given much thought to this subject for a number of years, even before
clarifying my views on the depressive processes in infancy. In the course of working out my
concept of the infantile depressive position, however, the problems of the phase preceding it
again forced themselves on my attention. I now wish to formulate some hypotheses at which I

have arrived regarding the earlier anxieties and mechanisms. 4

The hypotheses I shall put forward, which relate to very early stages of development, are derived
by inference from material gained in the analyses of adults and children, and some of these
hypotheses seem to tally with observations familiar in psychiatric work. To substantiate my
contentions would require an accumulation of detailed case material for which there is no room
in the framework of this paper, and I hope in further contributions to fi ll this gap.
3 [Original footnote to 1952 version: This paper was read before the British PsychoAnalytical
Society on December

4, 1946, and has been left unchanged and then published, apart from a few slight alterations (in
particular the addition of one paragraph and some footnotes).] Editors’ note: this chapter
reproduces the text of Klein, M. (1952a).

Notes on some schizoid mechanisms. In Envy and Gratitude and Other Works 1947–1963 .
London: Hogarth Press (1975), pp. 1–24.

4 Before completing this paper I discussed its main aspects with Paula Heimann and am much
indebted to her for stimulating suggestions in working out and formulating a number of the
concepts presented here.

19

Melanie Klein

At the outset it will be useful to summarize briefl y the conclusions regarding the earliest phases
of development which I have already put forward. 5

In early infancy anxieties characteristic of psychosis arise which drive the ego to develop specifi
c defence-mechanisms. In this period the fi xation-points for all psychotic disorders are to be
found. This hypothesis led some people to believe that I regarded all infants as psychotic; but I
have already dealt suffi ciently with this misunderstanding on other occasions. The psychotic
anxieties, mechanisms and ego-defences of infancy have a profound infl uence on development
in all its aspects, including the development of the ego, superego and object-relations.

I have often expressed my view that object-relations exist from the beginning of life, the fi rst
object being the mother’s breast which to the child becomes split into a good (gratifying) and bad
(frustrating) breast; this splitting results in a severance of love and hate. I have further suggested
that the relation to the fi rst object implies its introjection and projection, and thus from the
beginning object-relations are moulded by an interaction between introjection and projection,
between internal and external objects and situations. These processes participate in the building
up of the ego and superego and prepare the ground for the onset of the Oedipus complex in the
second half of the fi rst year.

From the beginning the destructive impulse is turned against the object and is fi rst expressed in
phantasied oral-sadistic attacks on the mother’s breast, which soon develop into onslaughts on
her body by all sadistic means. The persecutory fears arising from the infant’s oral-sadistic
impulses to rob the mother’s body of its good contents, and from the anal-sadistic impulses to
put his excrements into her (including the desire to enter her body in order to control her from
within) are of great importance for the development of paranoia and schizophrenia.

I enumerated various typical defences of the early ego, such as the mechanisms of splitting the
object and the impulses, idealization, denial of inner and outer reality and the stifl ing of
emotions. I also mentioned various anxiety-contents, including the fear of being poisoned and
devoured. Most of these phenomena – prevalent in the
5 Cf. my Psycho-Analysis of Children (1932), and ‘A Contribution to the Psychogenesis of
Manic-Depressive States’ (1935).

20

Notes on some schizoid mechanisms

fi rst few months of life – are found in the later symptomatic picture of schizophrenia.

This early period (fi rst described as the ‘persecutory phase’) I later

termed ‘paranoid position’, 6 and held that it precedes the depressive position. If persecutory
fears are very strong, and for this reason (among others) the infant cannot work through the
paranoid-schizoid position, the working through of the depressive position is in turn impeded.
This failure may lead to a regressive reinforcing of persecutory fears and strengthen the fi xation-
points for severe psychoses (that is to say, the group of schizophrenias). Another outcome of
serious diffi culties arising during the period of the depressive position may be manic-depressive
disorders in later life. I also concluded that in less severe disturbances of development the same

factors strongly infl uence the choice of neurosis.

While I assumed that the outcome of the depressive position depends on the working through of
the preceding phase, I nevertheless attributed to the depressive position a central role in the
child’s early development. For with the introjection of the object as a whole the infant’s object-
relation alters fundamentally. The synthesis between the loved and hated aspects of the complete
object gives rise to feelings of mourning and guilt which imply vital advances in the infant’s
emotional and intellectual life. This is also a crucial juncture for the choice of neurosis or
psychosis. To all these conclusions I still adhere.

Some notes on Fairbairn’s recent papers

In a number of recent papers, 7 W. R. D. Fairbairn has given much

attention to the subject-matter with which I am now dealing. I therefore fi nd it helpful to clarify
some essential points of agreement and disagreement between us. It will be seen that some of the
conclusions

6 When this paper was fi rst published in 1946, I was using my term ‘paranoid position’

synonymously with W. R. D. Fairbairn’s ‘schizoid position’. On further deliberation I decided to

combine Fairbairn’s term with mine and throughout the present book

[ Developments in Psycho-Analysis (Klein et al., 1952), in which this paper was fi rst published]

I am using the expression ‘paranoid-schizoid position’.

7 Cf. ‘A Revised Psychopathology of the Psychoses and Neuroses’, ‘Endopsychic Structure

Considered in Terms of Object-Relationships’ and ‘Object-Relationships and Dynamic
Structure’.

21

Melanie Klein

which I shall present in this paper are in line with Fairbairn’s conclusions, while others differ
fundamentally.

Fairbairn’s approach was largely from the angle of ego-development in relation to objects, while
mine was predominantly from the angle of anxieties and their vicissitudes. He called the earliest
phase the ‘schizoid position’: he stated that it forms part of normal development and is the basis
for adult schizoid and schizophrenic illness. I agree with this contention and consider his
description of developmental schizoid phenomena as signifi cant and revealing, and of great
value for our understanding of schizoid behaviour and of schizophrenia. I also think that
Fairbairn’s view that the group of schizoid or schizophrenic disorders is much wider than has
been acknowledged is correct and important; and the particular emphasis he laid on the inherent
relation between hysteria and schizophrenia deserves full attention. His term ‘schizoid position’
would be appropriate if it is understood to cover both persecutory fear and schizoid mechanisms.

I disagree – to mention fi rst the most basic issues – with his revi-sion of the theory of mental
structure and instincts.

I also disagree with his view that to begin with only the bad object is internalized

– a view which seems to me to contribute to the important differences between us regarding the
development of object-relations as well as of ego-development. For I hold that the introjected
good breast forms a vital part of the ego, exerts from the beginning a fundamental infl uence on
the process of ego-development and affects both ego-structure and object-relations. I also differ
from Fairbairn’s view that ‘the great problem of the schizoid individual is how to love without
destroying by love, whereas the great problem of the depressive individual is how to love
without destroying by hate’. 8 This conclusion is in line not only with his rejecting Freud’s
concept of primary instincts but also with his underrating the role which aggression and hatred
play from the beginning of life. As a result of this approach, he does not give enough weight to
the importance of early anxiety and confl ict and their dynamic effects on development.

Certain problems of the early ego

In the following discussion I shall single out one aspect of ego-development and I shall
deliberately not attempt to link it with the

8 Cf. ‘A Revised Psychopathology’ (1941).

22

Notes on some schizoid mechanisms

problems of ego-development as a whole. Nor can I here touch on the relation of the ego to the id
and super-ego.

So far, we know little about the structure of the early ego. Some of the recent suggestions on this
point have not convinced me: I have particularly in mind Glover’s concept of ego nuclei and
Fairbairn’s theory of a central ego and two subsidiary egos. More helpful in my view is
Winnicott’s emphasis on the unintegration of the early ego. 9

I would also say that the early ego largely lacks cohesion, and a tendency towards integration
alternates with a tendency towards

disintegration, a falling into bits. 10 I believe that these fl uctuations are characteristic of the fi rst
few months of life.

We are, I think, justifi ed in assuming that some of the functions which we know from the later
ego are there at the beginning.

Prominent amongst these functions is that of dealing with anxiety. I hold that anxiety arises from
the operation of the death instinct within the organism, is felt as fear of annihilation (death) and
takes the form of fear of persecution.

The fear of the destructive impulse seems to attach itself at once to an object – or rather it is
experienced as the fear of an uncontrollable overpowering object. Other important sources of
primary anxiety are the trauma of birth (separation anxiety) and frustration of bodily needs; and
these experiences too are from the beginning felt as being caused by objects. Even if these
objects are felt to be external, they become through introjection internal persecutors and thus
reinforce the fear of the destructive impulse within.

The vital need to deal with anxiety forces the early ego to develop fundamental mechanisms and
defences. The destructive impulse is partly projected outwards (defl ection of the death instinct)
and, I think, attaches itself to the fi rst external object, the mother’s breast.

As Freud has pointed out, the remaining portion of the destructive impulse is to some extent
bound by the libido within the organism.

9 Cf. D. W. Winnicott, ‘Primitive Emotional Development’ (1945). In this paper Winnicott also
described the pathological outcome of states of unintegration, for instance the case of a woman
patient who could not distinguish between her twin sister and herself.

10 The greater or lesser cohesiveness of the ego at the beginning of postnatal life should be
considered in connection with the greater or lesser capacity of the ego to tolerate anxiety which,
as I have previously contended (

PsychoAnalysis of Children , particularly p. 49), is a constitutional factor.

23

Melanie Klein
However, neither of these processes entirely fulfi ls its purpose, and therefore the anxiety of
being destroyed from within remains active. It seems to me in keeping with the lack of
cohesiveness that

under the pressure of this threat the ego tends to fall to pieces. 11

This falling to pieces appears to underlie states of disintegration in schizophrenics.

The question arises whether some active splitting processes within the ego may not occur even at
a very early stage.

As we assume, the early ego splits the object and the relation to it in an active way, and this may
imply some active splitting of the ego itself. In any case, the result of splitting is a dispersal of
the destructive impulse which is felt as the source of danger. I suggest that the primary anxiety of
being annihilated by a destructive force within, with the ego’s specifi c response of falling to
pieces or splitting itself, may be extremely important in all schizophrenic processes.

Splitting processes in relation to the object

The destructive impulse projected outwards is fi rst experienced as oral aggression. I believe that
oral-sadistic impulses towards the mother’s breast are active from the beginning of life, though
with the onset of teething the cannibalistic impulses increase in strength – a factor stressed by
Abraham.

In states of frustration and anxiety the oral-sadistic and cannibalistic desires are reinforced, and
then the infant feels that he has taken in the nipple and the breast in bits. Therefore in addition to
the divorce between a good and a bad breast in the young infant’s phantasy, the frustrating breast
– attacked in oral-sadistic phantasies – is felt to be in fragments; the gratifying breast, taken in
under the dominance of the sucking libido, is felt to be complete. This fi rst internal good object
acts as a focal point in the ego. It counteracts the processes of splitting and dispersal, makes for
cohesiveness and integration, and

11 Ferenczi in ‘Notes and Fragments’ (1930) suggests that most likely every living organism
reacts to unpleasant stimuli by fragmentation, which might be an expression of the death instinct.
Possibly, complicated mechanisms (living organisms) are only kept as an entity through the
impact of external conditions. When these conditions become unfavourable the organism falls to
pieces.

24

Notes on some schizoid mechanisms

is instrumental in building up the ego. 12 The infant’s feeling of having inside a good and
complete breast may, however, be shaken by frustration and anxiety. As a result, the divorce
between the good and bad breast may be diffi cult to maintain, and the infant may feel that the
good breast too is in pieces.

I believe that the ego is incapable of splitting the object – internal and external – without a
corresponding splitting taking place within the ego. Therefore the phantasies and feelings about
the state of the internal object vitally infl uence the structure of the ego. The more sadism
prevails in the process of incorporating the object, and the more the object is felt to be in pieces,
the more the ego is in danger of being split in relation to the internalized object fragments.

The processes I have described are, of course, bound up with the infant’s phantasy-life; and the
anxieties which stimulate the mechanism of splitting are also of a phantastic nature. It is in
phantasy that the infant splits the object and the self, but the effect of this phantasy is a very real
one, because it leads to feelings and relations (and later

on, thought-processes) being in fact cut off from one another. 13

Splitting in connection with projection and introjection So far, I have dealt particularly with
the mechanism of splitting as one of the earliest ego-mechanisms and defences against anxiety.

Introjection and projection are from the beginning of life also used in the service of this primary
aim of the ego.

Projection, as Freud described, originates from the defl ection of the death instinct outwards and
in my view it helps the ego to overcome anxiety by ridding it of danger and badness. Introjection
of the good object is also used by the ego as a defence against anxiety.

12 D. W. Winnicott (loc. cit.) referred to the same process from another angle: he described how
integration and

adaptation to reality depend essentially on the infant’s experience of the mother’s love and care.

13 In the discussion following the reading of this paper, Dr W. C. M. Scott referred to another
aspect of splitting. He stressed the importance of the breaks in continuity of experiences, which
imply a splitting in time rather than in space. He referred as an instance to the alternation
between states of being asleep and states of being awake. I fully agree with his point of view.

25

Melanie Klein

Closely connected with projection and introjection are some other mechanisms. Here I am
particularly concerned with the connection between splitting, idealization and denial. As regards
splitting of the object, we have to remember that in states of gratifi cation love-feelings turn
towards the gratifying breast, while in states of frustration hatred and persecutory anxiety attach
themselves to the frustrating breast.

Idealization is bound up with the splitting of the object, for the good aspects of the breast are
exaggerated as a safeguard against the fear of the persecuting breast. While idealization is thus
the corollary of persecutory fear, it also springs from the power of the instinctual desires which
aim at unlimited gratifi cation and therefore create the picture of an inexhaustible and always
bountiful breast – an ideal breast.
We fi nd an instance of such a cleavage in infantile hallucinatory gratifi cation. The main
processes which come into play in idealization are also operative in hallucinatory gratifi cation,
namely, splitting of the object and denial both of frustration and of persecution. The frustrating
and persecuting object is kept widely apart from the idealized object. However, the bad object is
not only kept apart from the good one but its very existence is denied, as is the whole situation of
frustration and the bad feelings (pain) to which frustration gives rise. This is bound up with
denial of psychic reality. The denial of psychic reality becomes possible only through strong
feelings of omnipotence – an essential characteristic of early mentality. Omnipotent denial of the
existence of the bad object and of the painful situation is in the unconscious equal to annihilation
by the destructive impulse. It is, however, not only a situation and an object that are denied and
annihilated – it is an object-relation which suffers this fate; and therefore a part of the ego, from
which the feelings towards the object emanate, is denied and annihilated as well.

In hallucinatory gratifi cation, therefore, two interrelated processes take place: the omnipotent
conjuring up of the ideal object and situation, and the equally omnipotent annihilation of the bad
persecutory object and the painful situation. These processes are based on splitting both the
object and the ego.

In passing I would mention that in this early phase splitting, denial and omnipotence play a role
similar to that of repression at a later stage of ego-development. In considering the importance of
the processes of denial and omnipotence at a stage which is characterized by persecutory fear and
schizoid mechanisms, we may remember the delusions of both grandeur and of persecution in
schizophrenia.

26

Notes on some schizoid mechanisms

So far, in dealing with persecutory fear, I have singled out the oral element. However, while the
oral libido still has the lead, libidinal and aggressive impulses and phantasies from other sources
come to the fore and lead to a confl uence of oral, urethral and anal desires, both libidinal and
aggressive. Also the attacks on the mother’s breast develop into attacks of a similar nature on her
body, which comes to be felt as it were as an extension of the breast, even before the mother is
conceived of as a complete person. The phantasied onslaughts on the mother follow two main
lines: one is the predominantly oral impulse to suck dry, bite up, scoop out and rob the mother’s
body of its good contents. (I shall discuss the bearing of these impulses on the development of
object-relations in connection with introjection.) The other line of attack derives from the anal
and urethral impulses and implies expelling dangerous substances (excrements) out of the self
and into the mother. Together with these harmful excrements, expelled in hatred, split-off parts
of the ego are also projected on to the mother or, as I would rather call it,

into the mother. 14 These excrements and bad parts of the self are meant not only to injure but
also to control and to

take possession of the object. In so far as the mother comes to contain the bad parts of the self,
she is not felt to be a separate individual but is felt to be the bad self.

Much of the hatred against parts of the self is now directed towards the mother. This leads to a
particular form of identifi cation which establishes the prototype of an aggressive object-relation.
I suggest for these processes the term

‘projective identifi cation’. When projection is mainly derived from the infant’s impulse to harm
or to control

the mother, 15 he feels her to be a persecutor. In psychotic disorders

14 The description of such primitive processes suffers from a great handicap, for these
phantasies arise at a time when the infant has not yet begun to think in words. In this context, for
instance, I am using the expression ‘to project into another person’ because this seems to me the
only way of conveying the unconscious process I am trying to describe.

15 M. G. Evans, in a short unpublished communication (read to the British PsychoAnalytical

Society, January, 1946), gave some instances of patients in whom the following phenomena were
marked: lack of sense of reality, a feeling of being divided and parts of the personality having
entered the mother’s body in order to rob and control her; as a consequence the mother and other
people similarly attacked came to represent the patient. M. G. Evans related these processes to a
very primitive stage of development.

27

Melanie Klein

this identifi cation of an object with the hated parts of the self contributes to the intensity of the
hatred directed against other people. As far as the ego is concerned the excessive splitting off and
expelling into the outer world of parts of itself considerably weaken it. For the aggressive
component of feelings and of the personality is intimately bound up in the mind with power,
potency, strength, knowledge and many other desired qualities.

It is, however, not only the bad parts of the self which are expelled and projected, but also good
parts of the self.

Excrements then have the signifi cance of gifts; and parts of the ego which, together with
excrements, are expelled and projected into the other person represent the good, i.e. the loving
parts of the self. The identifi cation based on this type of projection again vitally infl uences
object-relations. The projection of good feelings and good parts of the self into the mother is
essential for the infant’s ability to develop good object-relations and to integrate his ego.

However, if this projective process is carried out excessively, good parts of the personality are
felt to be lost, and in this way the mother becomes the ego-ideal; this process too results in
weakening and impoverishing the ego. Very

soon such processes extend to other people, 16 and the result may be an over-strong dependence
on these external representatives of one’s own good parts. Another consequence is a fear that the
capacity to love has been lost because the loved object is felt to be loved predominantly as a
representative of the self.
The processes of splitting off parts of the self and projecting them into objects are thus of vital
importance for normal development as well as for abnormal object-relations.

The effect of introjection on object-relations is equally important.

The introjection of the good object, fi rst of all the mother’s breast, is a precondition for normal
development. I have already described that it comes to form a focal point in the ego and makes
for cohesiveness of the ego. One characteristic feature of the earliest relation to the good object –
internal and external – is the tendency to idealize it.

In

16 W. C. M. Scott in an unpublished paper, read to the British Psycho-Analytical Society a few

years ago, described three interconnected features which he came upon in a schizophrenic
patient: a strong disturbance of her sense of

reality, her feeling that the world round her was a cemetery, and the mechanism of putting all
good parts of herself into another person – Greta Garbo – who came to stand for the patient.

28

Notes on some schizoid mechanisms

states of frustration or increased anxiety, the infant is driven to take fl ight to his internal
idealized object as a means of escaping from persecutors. From this mechanism various serious
disturbances may result: when persecutory fear is too strong, the fl ight to the idealized object
becomes excessive, and this severely hampers ego-development and disturbs object-relations. As
a result the ego may be felt to be entirely subservient to and dependent on the internal object –
only a shell for it. With an unassimilated idealized object there goes a feeling that the ego has no
life and no value of

its own. 17

I would suggest that the condition of fl ight to the unassimilated idealized object necessitates
further splitting processes within the ego.

For parts of the ego attempt to unite with the ideal object, while other parts strive to deal with the
internal persecutors.

The various ways of splitting the ego and internal objects result in the feeling that the ego is in
bits. This feeling amounts to a state of disintegration. In normal development, the states of
disintegration which the infant experiences are transitory. Among other factors, gratifi cation by
the external good object 18 again and again helps to break through these schizoid states. The
infant’s capacity to overcome temporary schizoid states is in keeping with the strong elasticity
and resilience of the infantile mind. If states of splitting and therefore of disintegration, which the
ego is unable to overcome, occur too frequently and go on for too long, then in my view they
must be regarded as a sign of schizophrenic illness in the infant, and some indications of such
illness may already be seen in the fi rst few 17 Cf. ‘A Contribution to the Problem of
Sublimation and its Relation to the Processes of Internalization’ (1942) where Paula Heimann
described a condition in which the internal objects act as foreign bodies embedded in the self.

Whilst this is more obvious with regard to the bad objects, it is true even for the good ones, if the
ego is compulsively subordinated to their preservation. When the ego serves its good internal
objects excessively, they are felt as a source of danger to the self and come close to exerting a
persecuting infl uence. Paula Heimann introduced the concept of the assimilation of the internal
objects and applied it specifi cally to sublimation. As regards ego-development, she pointed out
that such assimilation is essential for the successful exercise of ego-functions and for the
achievement of independence.

18 Looked at in this light, the mother’s love and understanding of the infant can be seen as the
infant’s greatest stand-by in overcoming states of disintegration and anxieties of a psychotic
nature.

29

Melanie Klein

months of life. In adult patients, states of depersonalization and of schizophrenic dissociation

seem to be a regression to these infantile states of disintegration. 19

In my experience, excessive persecutory fears and schizoid mechanisms in early infancy may
have a detrimental effect on intellectual development in its initial stages. Certain forms of mental
defi ciency would therefore have to be regarded as belonging to the group of schizophrenias.
Accordingly, in considering mental defi ciency in children at any age one should keep in mind
the possibility of schizophrenic illness in early infancy.

I have so far described some effects of excessive introjection and projection on object-relations. I
am not attempting to investigate here in any detail the various factors which in some cases make
for a predominance of introjective and in other cases for a predominance of projective processes.
As regards normal personality, it may be said that the course of ego-development and object-
relations depends on the degree to which an optimal balance between

introjection and projection in the early stages of development can be achieved. This in turn has a
bearing on the integration of the ego and the assimilation of internal objects. Even if the balance
is disturbed and one or the other of these processes is excessive, there is some interaction
between introjection and projection. For instance the projection of a predominantly hostile inner
world which is ruled by persecutory fears leads to the introjection – a taking-back – of a hostile
external world; and vice versa, the introjection of a distorted and hostile external world
reinforces the projection of a hostile inner world.

Another aspect of projective processes, as we have seen, concerns the forceful entry into the
object and control of the object by parts of the self. As a consequence, introjection may then be
felt as a forceful entry from the outside into the inside, in retribution for violent

19 Herbert Rosenfeld, in ‘Analysis of a Schizophrenic State with Depersonalization’ (1947), has

presented case-material to illustrate how the splitting mechanisms which are bound up with
projective identifi cation were responsible both for a schizophrenic state and depersonalization.
In his paper ‘A Note on the Psychopathology of Confusional States in Chronic Schizophrenias’
(1950) he also pointed out that a confusional state comes about if the subject loses the capacity to
differentiate between good and bad objects, between aggressive and libidinal impulses, and so
on. He suggested that in such states of confusion splitting mechanisms are frequently reinforced
for defensive purposes.

30

Notes on some schizoid mechanisms

projection. This may lead to the fear that not only the body but also the mind is controlled by
other people in a hostile way. As a result there may be a severe disturbance in introjecting good
objects – a disturbance which would impede all ego-functions as well as sexual development and
might lead to an excessive withdrawal to the inner world. This withdrawal is, however, caused
not only by the fear of introjecting a dangerous external world but also by the fear of internal
persecutors and an ensuing fl ight to the idealized internal object.

I have referred to the weakening and impoverishment of the ego resulting from excessive
splitting and projective identifi cation. This weakened ego, however, becomes also incapable of
assimilating its internal objects, and this leads to the feeling that it is ruled by them.

Again, such a weakened ego feels incapable of taking back into itself the parts which it projected
into the external world. These various disturbances in the interplay between projection and
introjection, which imply excessive splitting of the ego, have a detrimental effect on the relation
to the inner and outer world and seem to be at the root of some forms of schizophrenia.

Projective identifi cation is the basis of many anxiety-situations, of which I shall mention a few.
The phantasy of forcefully entering the object gives rise to anxieties relating to the dangers
threatening the subject from within the object. For instance, the impulses to control an object
from within it stir up the fear of being controlled and persecuted inside it. By introjecting and re-
introjecting the forcefully entered object, the subject’s feelings of inner persecution are strongly
reinforced; all the more since the re-introjected object is felt to contain the dangerous aspects of
the self. The accumulation of anxieties of this nature, in which the ego is, as it were, caught
between a variety of external and internal persecution-situations, is a basic element in paranoia.
20

20 Herbert Rosenfeld, in ‘Analysis of a Schizophrenic State with Depersonalization’ and

‘Remarks on the Relation of Male Homosexuality to Paranoia’ (1949), discussed the clinical
importance of those paranoid anxieties which are connected with projective identifi cation in
psychotic patients. In the two schizophrenic cases he described, it became evident that the
patients were dominated by the fear that the analyst was trying to force himself into the patient.
When these fears were analysed in the transference-situation, improvement could take place.

Rosenfeld has further connected projective identifi cation (and the corresponding persecutory
fears) with female sexual frigidity on the one hand and on the other with the frequent
combination of homosexuality and paranoia in men.
31

Melanie Klein

I have previously described 21 the infant’s phantasies of attacking and sadistically entering the
mother’s body as giving rise to various anxiety-situations (particularly the fear of being
imprisoned and persecuted within her) which are at the bottom of paranoia. I also showed that
the fear of being imprisoned (and especially of the penis being attacked) inside the mother is an
important factor in later disturbances of male potency (impotence) and also underlines

claustrophobia. 22

Schizoid object-relations

To summarize now some of the disturbed object-relations which are found in schizoid
personalities: the violent splitting of the self and excessive projection have the effect that the
person towards whom this process is directed is felt as a persecutor. Since the destructive and
hated part of the self which is split off and projected is felt as a danger to the loved object and
therefore gives rise to guilt, this process of projection in some ways also implies a defl ection of
guilt from the self on to the other person. Guilt has, however, not been done away with, and the
defl ected guilt is felt as an unconscious responsibility for the people who have become
representatives of the aggressive part of the self.

Another typical feature of schizoid object-relations is their narcissistic nature which derives from
the infantile introjective and projective processes. For, as I suggested earlier, when the ego-ideal
is projected

21 P sycho-Analysis of Children , Chapter 8 , particularly p. 131, and Chapter 12 , particularly

p. 242.

22 Joan Riviere, in an unpublished paper ‘Paranoid Attitudes seen in Everyday Life and in
Analysis’ (read before the British Psycho-Analytical Society in 1948), reported a great deal of
clinical material in which projective identifi cation became apparent. Unconscious phantasies of
forcing the whole self into the inside of the object (to obtain control and possession) led, through
the fear of retaliation, to a variety of persecutory anxieties such as claustrophobia, or to such
common phobias as of burglars, spiders, invasion in wartime. These fears are connected with the
unconscious ‘catastrophic’ phantasies of being dismembered, disembowelled, torn to pieces and
of total internal disruption of the body and personality and loss of identity – fears which are an
elaboration of the fear of annihilation (death) and have the effect of reinforcing the mechanisms
of splitting and the process of ego-disintegration as found in psychotics.

32

Notes on some schizoid mechanisms

into another person, this person becomes predominantly loved and admired because he contains
the good parts of the self. Similarly, the relation to another person on the basis of projecting bad
parts of the self into him is of a narcissistic nature, because in this case as well the object
strongly represents one part of the self. Both these types of a narcissistic relation to an object
often show strong obsessional features. The impulse to control other people is, as we know, an
essential element in obsessional neurosis. The need to control others can to some extent be
explained by a defl ected drive to control parts of the self. When these parts have been projected
excessively into another person, they can only be controlled by controlling the other person. One
root of obsessional mechanisms may thus be found in the particular identifi cation which results
from infantile projective processes. This connection may also throw some light on the
obsessional element which so often enters into the tendency for reparation. For it is not only an
object about whom guilt is experienced but also parts of the self which the subject is driven to
repair or restore.

All these factors may lead to a compulsive tie to certain objects or

– another outcome – to a shrinking from people in order to prevent both a destructive intrusion
into them and the

danger of retaliation by them. The fear of such dangers may show itself in various negative
attitudes in object-relations. For instance, one of my patients told me that he dislikes people who
are too much infl uenced by him, for they seem to become too much like himself and therefore
he gets tired of them.

Another characteristic of schizoid object-relations is a marked artifi ciality and lack of

spontaneity. Side by side with this goes a severe disturbance of the feeling of the self or, as I
would put it, of the relation to the self. This relation, too, appears to be artifi cial. In other words,
psychic reality and the relation to external reality are equally disturbed.

The projection of split-off parts of the self into another person essentially infl uences object-
relations, emotional life and the personality as a whole. To illustrate this contention I will select
as an instance two universal phenomena which are interlinked: the feeling of loneliness and fear
of parting. We know that one source of the depressive feelings accompanying parting from
people can be found in the fear of the destruction of the object by the aggressive impulses
directed against it. But it is more specifi cally the splitting and projective processes which
underlie this fear. If aggressive elements in relation 33

Melanie Klein

to the object are predominant and strongly aroused by the frustration of parting, the individual
feels that the split-off components of his self, projected into the object, control this object in an
aggressive and destructive way. At the same time the internal object is felt to be in the same
danger of destruction as the external one in whom one part of the self is felt to be left. The result
is an excessive weakening of the ego, a feeling that there is nothing to sustain it, and a
corresponding feeling of loneliness. While this description applies to neurotic individuals, I think
that in some degree it is a general phenomenon.

One need hardly elaborate the fact that some other features of schizoid object-relations, which I
described earlier, can also be found in minor degrees and in a less striking form in normal people
– for instance shyness, lack of spontaneity or, on the other hand, a particularly intense interest in
people.

In similar ways normal disturbances in thought-processes link up with the developmental

paranoid-schizoid position. For all of us are liable at times to a momentary impairment of logical
thinking which amounts to thoughts and associations being cut off from one another and
situations being split off from one another; in fact, the ego is temporarily split.

The depressive position in relation to the

paranoid-schizoid position

I now wish to consider further steps in the infant’s development. So far I have described the
anxieties, mechanisms and defences which are characteristic of the fi rst few months of life. With
the introjection of the complete object in about the second quarter of the fi rst year marked steps
in integration are made. This implies important changes in the relation to objects. The loved and
hated aspects of the mother are no longer felt to be so widely separated, and the result is an
increased fear of loss, states akin to mourning and a strong feeling of guilt, because the
aggressive impulses are felt to be directed against the loved object. The depressive position has
come to the fore. The very experience of depressive feelings in turn has the effect of further
integrating the ego, because it makes for an increased understanding of psychic reality and better
perception of the external world, as well as for a greater synthesis between inner and external
situations.

34

Notes on some schizoid mechanisms

The drive to make reparation, which comes to the fore at this stage, can be regarded as a
consequence of greater insight into psychic reality and of growing synthesis, for it shows a more
realistic response to the feelings of grief, guilt and fear of loss resulting from the aggression
against the loved object. Since the drive to repair or protect the injured object paves the way for
more satisfactory object-relations and sublimations, it in turn increases synthesis and contributes
to the integration of the ego.

During the second half of the fi rst year the infant makes some fundamental steps towards
working through the depressive position.

However, schizoid mechanisms still remain in force, though in a modifi ed form and to a lesser
degree, and early anxiety-situations are again and again experienced in the process of modifi
cation. The working through of the persecutory and depressive positions extends over the fi rst
few years of childhood and plays an essential part in the infantile neurosis. In the course of this
process, anxieties lose in strength; objects become both less idealized and less terrifying, and the
ego becomes more unifi ed. All this is interconnected with the growing perception of reality and
adaptation to it.

If development during the paranoid-schizoid position has not proceeded normally and the infant
cannot – for internal or external reasons – cope with the impact of depressive anxieties a vicious
circle arises. For if persecutory fear, and correspondingly schizoid mechanisms, are too strong,
the ego is not capable of working through the depressive position. This forces the ego to regress
to the paranoid-schizoid position and reinforces the earlier persecutory fears and schizoid
phenomena. Thus the basis is established for various forms of schizophrenia in later life; for
when such a regression occurs, not only are the fi xation-points in the schizoid position
reinforced, but there is a danger of greater states of disintegration setting in. Another outcome
may be the strengthening of depressive features.

External experiences are, of course, of great importance in these developments. For instance, in
the case of a patient who showed depressive and schizoid features, the analysis brought up with
great vividness his early experiences in babyhood, to such an extent that in some sessions
physical sensations in the throat or digestive organs occurred. The patient had been weaned
suddenly at four months of age because his mother fell ill. In addition, he did not see his mother
for four weeks. When she returned, she found the child greatly changed. He had been a lively
baby, interested in his surroundings, 35

Melanie Klein

and he seemed to have lost this interest. He had become apathetic.

He had accepted the substitute food fairly easily and in fact never refused food. But he did not
thrive on it any more, lost weight and had a good deal of digestive trouble. It was only at the end
of the fi rst year, when other food was introduced, that he again made good physical progress.

Much light was thrown in the analysis on the infl uence these experiences had on his whole
development. His outlook and attitudes in adult life were based on the patterns established in this
early stage. For instance, we found again and again a tendency to be infl uenced by other people
in an unselective way – in fact to take in greedily whatever was offered – together with great
distrust during the process of introjection. This process was constantly disturbed by anxieties
from various sources, which also contributed to an increase of greed.

Taking the material of this analysis as a whole, I came to the conclusion that at the time when the
sudden loss of the breast and of the mother occurred, the patient had already to some extent
established a relation to a complete good object. He had no doubt already entered the depressive
position but could not work through it successfully and the paranoid-schizoid position became
regressively reinforced. This expressed itself in the ‘apathy’ which followed a period when the
child had already shown a lively interest in his surroundings. The fact that he had reached the
depressive position and had introjected a complete object showed in many ways in his
personality. He had actually a strong capacity for love and a great longing for a good and
complete object. A characteristic feature of his personality was the desire to love people and trust
them, unconsciously to regain and build up again the good and complete breast which he had
once possessed and lost.

Connection between schizoid and

manic-depressive phenomena

Some fl uctuations between the paranoid-schizoid and the depressive positions always occur and
are part of normal development. No clear division between the two stages of development can
therefore be drawn; moreover, modifi cation is a gradual process and the phenomena of the two
positions remain for some time to some 36

Notes on some schizoid mechanisms

extent intermingled and interacting. In abnormal development this interaction infl uences, I think,
the clinical picture both of some forms of schizophrenia and of manic-depressive disorders.

To illustrate this connection I shall briefl y refer to some case-material. I do not intend to present
a case-history here and am therefore only selecting some parts of material relevant to my topic.
The patient I have in mind was a pronounced manic-depressive case (diagnosed as such by more
than one psychiatrist) with all the characteristics of that disorder: there was the alternation
between depressive and manic states, strong suicidal tendencies leading repeatedly to suicidal
attempts, and various other characteristic manic and depressive features. In the course of her
analysis a stage was reached in which a real and great improvement was achieved. Not only did
the cycle stop but there were fundamental changes in her personality and her object-relations.
Productivity on various lines developed, as well as actual feelings of happiness (not of a manic
type). Then, partly owing to external circumstances, another phase set in. During this last phase,
which continued for several months, the patient co-operated in the analysis in a particular way.
She came regularly to the analytic sessions, associated fairly freely, reported dreams and
provided material for the analysis. There was, however, no emotional response to my
interpretations and a good deal of contempt of them.

There was very seldom any conscious confi rmation of what I suggested. Yet the material by
which she responded to the interpretations refl ected their unconscious effect. The powerful
resistance shown at this stage seemed to come from one part of the personality only, while at the
same time another part responded to the analytic work. It was not only that parts of her
personality did not co-operate with me; they did not seem to co-operate with each other, and at
the time the analysis was unable to help the patient to achieve synthesis.

During this stage she decided to bring the analysis to an end. External circumstances contributed
strongly to this decision and she fi xed a date for the last session.

On that particular date she reported the following dream: there was a blind man who was very
worried about being blind; but he seemed to comfort himself by touching the patient’s dress and
fi nding out how it was fastened. The dress in the dream reminded her of one of her frocks which
was buttoned high up to the throat. The patient gave two further associations to this dream. She
said, with some resistance, that the blind man was herself; and when referring to the dress
fastened 37

Melanie Klein

up to the throat, she remarked that she had again gone into her ‘hide’.

I suggested to the patient that she unconsciously expressed in the dream that she was blind to her
own diffi culties, and that her decisions with regard to the analysis as well as to various
circumstances in her life were not in accordance with her unconscious knowledge. This was also
shown by her admitting that she had gone into her
‘hide’, meaning by it that she was shutting herself off, an attitude well known to her from
previous stages in her analysis. Thus the unconscious insight, and even some co-operation on the
conscious level (recognition that she was the blind man and that she had gone into her ‘hide’),
derived from isolated parts of her personality only. Actually, the interpretation of this dream did
not produce any effect and did not alter the patient’s decision to bring the analysis to an end in
that particular hour. 23

The nature of certain diffi culties encountered in this analysis as well as in others had revealed
itself more clearly in the last few months before the patient broke off the treatment. It was the
mixture of schizoid and manic-depressive features which determined the nature of her illness.

For at times throughout her analysis – even in the early stage when depressive and manic states
were at their height

– depressive and schizoid mechanisms sometimes appeared simultaneously. There were, for
instance, hours when the patient was obviously deeply depressed, full of self-reproaches and
feelings of unworthiness; tears were running down her cheeks and her gestures expressed
despair; and yet she said, when I interpreted these emotions, that she did not feel them at all.
Whereupon she reproached herself for having no feelings at all, for being completely empty.

In such sessions there was also a fl ight of ideas, the thoughts seemed to be broken up, and their
expression was disjointed.

Following the interpretation of the unconscious reasons underlying such states, there were
sometimes sessions in which the emotions and depressive anxieties came out fully, and at such
times thoughts and speech were much more coherent.

This close connection between depressive and schizoid phenomena appeared, though in different
forms, throughout her analysis but became very pronounced during the last stage preceding the
break just described.

I have already referred to the developmental connection between the paranoid-schizoid and
depressive positions.

The question now

23 I may mention that the analysis was resumed after a break.

38

Notes on some schizoid mechanisms

arises whether this developmental connection is the basis for the mixture of these features in
manic-depressive disorders and, as I would suggest, in schizophrenic disorders as well. If this
tentative hypothesis could be proved, the conclusion would be that the groups of schizophrenic
and manic-depressive disorders are more closely connected developmentally with one another
than has been assumed. This would also account for the cases in which, I believe, the differential
diagnosis between melancholia and schizophrenia is exceedingly diffi cult. I should be grateful if
further light could be thrown on my hypothesis by colleagues who have had ample material for
psychiatric observation.

Some schizoid defences

It is generally agreed that schizoid patients are more diffi cult to analyse than manic-depressive
types. Their withdrawn, unemotional attitude, the narcissistic elements in their object-relations
(to which I referred earlier), a kind of detached hostility which pervades the whole relation to the
analyst create a very diffi cult type of resistance.

I believe that it is largely the splitting processes which account for the patient’s failure in contact
with the analyst and for his lack of response to the analyst’s interpretations. The patient himself
feels estranged and far away, and this feeling corresponds to the analyst’s impression that
considerable parts of the patient’s personality and of his emotions are not available. Patients with
schizoid features may say: ‘I hear what you are saying. You may be right, but it has no meaning
for me.’ Or again they say they feel they are not there. The expression ‘no meaning’ in such
cases does not imply an active rejection of the interpretation but suggests that parts of the
personality and of the emotions are split off. These patients can, therefore, not deal with the
interpretation; they can neither accept it nor reject it.

I shall illustrate the processes underlying such states by a piece of material taken from the
analysis of a man patient.

The session I have in mind started with the patient’s telling me that he felt anxiety and did not
know why. He then made comparisons with people more successful and fortunate than himself.
These remarks also had a reference to me. Very strong feelings of frustration, envy and grievance
came to the fore. When I interpreted – to give here again only the gist of my interpretations – that
these feelings were directed against the analyst and that he wanted to destroy me, his mood
changed 39

Melanie Klein

abruptly. The tone of his voice became fl at, he spoke in a slow, expressionless way, and he said
that he felt detached from the whole situation. He added that my interpretation seemed correct,
but that it did not matter. In fact, he no longer had any wishes, and nothing was worth bothering
about.

My next interpretations centred on the causes for this change of mood. I suggested that at the
moment of my interpretation the danger of destroying me had become very real to him and the
immediate consequence was the fear of losing me. Instead of feeling guilt and depression, which
at certain stages of his analysis followed such interpretations, he now attempted to deal with
these dangers by a particular method of splitting. As we know, under the pressure of
ambivalence, confl ict and guilt, the patient often splits the fi gure of the analyst; then the analyst
may at certain moments be loved, at other moments hated.

Or the relations to the analyst may be split in such a way that he remains the good (or bad) fi
gure while somebody else becomes the opposite fi gure. But this was not the kind of splitting
which occurred in this particular instance.
The patient split off those parts of himself, i.e. of his ego which he felt to be dangerous and
hostile towards the analyst. He turned his destructive impulses from his object towards his ego,
with the result that parts of his ego temporarily went out of existence. In unconscious phantasy
this amounted to annihilation of part of his personality.

The particular mechanism of turning the destructive impulse against one part of his personality,
and the ensuing dispersal of emotions, kept his anxiety in a latent state.

My interpretation of these processes had the effect of again altering the patient’s mood. He
became emotional, said he felt like crying, was depressed, but felt more integrated; then he also
expressed a feeling of hunger. 24

24 The feeling of hunger indicated that the process of introjection had been set going again under
the dominance of the libido. While to my fi rst interpretation of his fear of destroying me by his
aggression he had responded at once with the violent splitting off and annihilation of parts of his
personality, he now experienced more fully the emotions of grief, guilt and fear of loss, as well
as some relief of these depressive anxieties. The relief of anxiety resulted in the analyst again
coming to stand for a good object which he could trust. Therefore the desire to introject me as a
good object could come to the fore. If he could build up again the good breast inside himself, he
would strengthen and integrate his ego, would be less afraid of his destructive impulses; in fact
he could then preserve himself and the analyst.

40

Notes on some schizoid mechanisms

The violent splitting off and destroying of one part of the personality under the pressure of
anxiety and guilt is in my experience an important schizoid mechanism. To refer briefl y to
another instance: a woman patient had dreamed that she had to deal with a wicked girl child who
was determined to murder somebody. The patient tried to infl uence or control the child and to
extort a confession from her which would have been to the child’s benefi t; but she was
unsuccessful. I also entered into the dream and the patient felt that I might help her in dealing
with the child.

Then the patient strung up the child on a tree in order to frighten her and also prevent her from
doing harm. When the patient was about to pull the rope and kill the child, she woke. During this
part of the dream the analyst was also present but again remained inactive.

I shall give here only the essence of the conclusions I arrived at from the analysis of this dream.
In the dream the patient’s personality was split into two parts: the wicked and uncontrollable
child on the one hand, and on the other hand the person who tried to infl uence and control her.
The child, of course, stood also for various fi gures in the past, but in this context she mainly
represented one part of the patient’s self. Another conclusion was that the analyst was the person
whom the child was going to murder; and my role in the dream was partly to prevent this murder
from taking place. Killing the child – to which the patient had to resort – represented the
annihilation of one part of her personality.

The question arises how the schizoid mechanism of annihilating part of the self connects with
repression which, as we know, is directed against dangerous impulses. This, however, is a
problem with which I cannot deal here.

Changes of mood, of course, do not always appear as dramatically within a session as in the fi rst
instance I have

given in this section.

But I have repeatedly found that advances in synthesis are brought about by interpretations of the
specifi c causes for splitting. Such interpretations must deal in detail with the transference-
situation at that moment, including of course the connection with the past, and must contain a
reference to the details of the anxiety-situations which drive the ego to regress to schizoid
mechanisms. The synthesis resulting from interpretations on these lines goes along with
depression and anxieties of various kinds. Gradually such waves of depression – followed by
greater integration –

lead to a lessening of schizoid phenomena and also to fundamental changes in object-relations.

41

Melanie Klein

Latent anxiety in schizoid patients

I have already referred to the lack of emotion which makes schizoid patients unresponsive. This
is accompanied by an absence of anxiety.

An important support for the analytic work is therefore lacking. For with other types of patients
who have strong manifest and latent anxiety, the relief of anxiety derived from analytic
interpretation becomes an experience which furthers their capacity to co-operate in the analysis.

This lack of anxiety in schizoid patients is only apparent. For the schizoid mechanisms imply a
dispersal of emotions including anxiety, but these dispersed elements still exist in the patient.
Such patients have a certain form of latent anxiety; it is kept latent by the particular method of
dispersal. The feeling of being disintegrated, of being unable to experience emotions, of losing
one’s objects, is in fact the equivalent of anxiety. This becomes clearer when advances in
synthesis have been made. The great relief which a patient then experiences derives from a
feeling that his inner and outer worlds have not only come more together but back to life again.
At such moments it appears in retrospect that when emotions were lacking, relations were vague
and uncertain and parts of the personality were felt to be lost, everything seemed to be dead. All
this is the equivalent of anxiety of a very serious nature. This anxiety, kept latent by dispersal, is
to some extent experienced all along, but its form differs from the latent anxiety which we can
recognize in other types of cases.

Interpretations which tend towards synthesizing the split in the self, including the dispersal of
emotions, make it possible for the anxiety gradually to be experienced as such, though for long
stretches we may in fact only be able to bring the ideational contents together but not to elicit the
emotions of anxiety.
I have also found that interpretations of schizoid states make particular demands on our capacity
to put the interpretations in an intellectually clear form in which the links between the conscious,
pre-conscious and unconscious are established. This is, of course, always one of our aims, but it
is of special importance at times when the patient’s emotions are not available and we seem to
address ourselves only to his intellect, however much broken up.

It is possible that the few hints I have given may to some extent apply as well to the technique of
analysing schizophrenic patients.

42

Notes on some schizoid mechanisms

Summary of conclusions
I will now summarize some of the conclusions presented in this paper. One of my main points
was the suggestion that in the fi rst few months of life anxiety is predominantly experienced as
fear of persecution and that this contributes to certain mechanisms and defences which are signifi
cant for the paranoid-schizoid position.

Outstanding among these defences are the mechanisms of splitting internal and external objects,
emotions and the ego. These mechanisms and defences are part of normal development and at
the same time form the basis for later schizophrenic illness. I described the processes underlying
identifi cation by projection as a combination of splitting off parts of the self and projecting them
on to another person, and some of the effects this identifi cation has on normal and schizoid
object-relations. The onset of the depressive position is the juncture at which by regression
schizoid mechanisms may be reinforced. I also suggested a close connection between the manic-
depressive and schizoid disorders, based on the interaction between the infantile paranoid-
schizoid and depressive positions.

Appendix

Freud’s analysis of the Schreber case 25 contains a wealth of material which is very relevant to
my topic but from

which I shall here draw only a few conclusions.

Schreber described vividly the splitting of the soul of his physician Flechsig (his loved and
persecuting fi gure). The

‘Flechsig soul’ at one time introduced the system of ‘soul divisions’, splitting into as many as
forty to sixty sub-divisions. These souls having multiplied till they became a ‘nuisance’, God
made a raid on them and as a result the Flechsig soul survived in ‘only one or two shapes’.
Another point which Schreber mentions is that the fragments of the Flechsig soul slowly lost
both their intelligence and their power.

One of the conclusions Freud arrived at in his analysis of this case was that the persecutor was
split into God and Flechsig, and

25 ‘Psycho-Analytic Notes upon an Autobiographical Account of a Case of Paranoia (Dementia

Paranoides)’ (S.E.

12).

43

Melanie Klein

also that God and Flechsig represented the patient’s father and brother. In discussing the various
forms of Schreber’s delusion of the destruction of the world, Freud states: ‘In any case the end of
the world was the consequence of the confl ict which had broken out between him, Schreber, and
Flechsig, or, according to the aetiology adopted in the second phase of his delusion, of the
indissoluble bond which had been formed between him and God. . . .’ (Loc. cit., p. 69).

I would suggest, in keeping with the hypotheses outlined in the present chapter, that the division
of the Flechsig soul into many souls was not only a splitting of the object but also a projection of
Schreber’s feeling that his ego was split. I shall here only mention the connection of such
splitting processes with processess of introjection. The conclusion suggests itself that God and
Flechsig also represented parts of Schreber’s self. The confl ict between Schreber and Flechsig,
to which Freud attributed a vital role in the world-destruction delusion, found expression in the
raid by God on the Flechsig souls. In my view this raid represents the annihilation by one part of
the self of the other parts – which, as I contend, is a schizoid mechanism. The anxieties and
phantasies about inner destruction and ego-disintegration bound up with this mechanism are
projected on to the external world and underlie the delusions of its destruction.

Regarding the processes which are at the bottom of the paranoic

‘world catastrophe’, Freud arrived at the following conclusions: ‘The patient has withdrawn from
the people in his environment and from the external world generally the libidinal cathexis which
he has hitherto directed on to them.

Thus everything has become indifferent and irrelevant to him, and has to be explained by means
of a secondary rationalization as being “miracled up, cursorily improvised”. The end of the world
is the projection of this internal catastrophe; for his subjective world has come to an end since he
has withdrawn his love from it.’ (Loc. cit., p. 70.) This explanation specifi cally concerns the
disturbance in object-libido and the ensuing breakdown in relation to people and to the external
world. But a little further on Freud considered another aspect of these disturbances. He said: ‘We
can no more dismiss the possibility that disturbances of the libido may react upon the egoistic
cathexes than we can overlook the converse possibility – namely, that a secondary or induced
disturbance of the libidinal processes may result from abnormal changes in the ego. Indeed it is
probable that processes of this kind constitute the distinctive characteristic of psychoses’

44

Notes on some schizoid mechanisms

(my italics) . It is particularly the possibility expressed in the last two sentences which provides
the link between Freud’s explanation of the ‘world catastrophe’ and my hypothesis. ‘Abnormal
changes in the

ego’ derive, as I have suggested in this chapter, from excessive split-

ting processes in the early ego. These processes are inextricably linked with instinctual
development, and with the anxieties to which instinctual desires give rise. In the light of Freud’s
later theory of the life and death instincts, which replaced the concept of the egoistic and sexual
instincts, disturbances in the distribution of the libido presuppose a defusion between the
destructive impulse and the libido. The mechanism of one part of the ego annihilating other parts
which, I suggest, underlies ‘world catastrophe’ phantasy (the raid by God on the Flechsig souls)
implies a preponderance of the destructive impulse over the libido. Any disturbance in the
distribution of the narcissistic libido is in turn bound up with the relation to introjected objects
which (according to my work) from the beginning come to form part of the ego. The interaction
between narcissistic libido and object-libido corresponds thus to the interaction between the
relation to introjected and external objects. If the ego and the internalized objects are felt to be in
bits, an internal catastrophe is experienced by the infant which both extends to the external world
and is projected on to it. Such anxiety-states relating to an internal catastrophe arise, according to
the hypothesis discussed in the present chapter, during the period of the infantile paranoid-
schizoid position and form the basis for later schizophrenia. In Freud’s view the dispositional fi
xation to dementia praecox is found in a very early stage of development. Referring to dementia
praecox, which Freud distinguished from paranoia, he said: ‘The dispositional point of fi xation
must therefore be situated further back than in paranoia, and must lie somewhere at the
beginning of the course of development from auto-erotism to object-love.’ (Loc. cit., p. 77.) I
wish to draw one more conclusion from Freud’s analysis of the Schreber case. I suggest that the
raid, which ended in the Flechsig souls being reduced to one or two, was part of the attempt
towards recovery. For the raid was to undo, or, one may say, heal the split in the ego by
annihilating the split-off parts of the ego. As a result only one or two of the souls were left
which, as we may assume, were meant to regain their intelligence and their power. This attempt
towards recovery, however, was effected by very destructive means used by the ego against itself
and its projected objects.

45

Melanie Klein

Freud’s approach to the problems of schizophrenia and paranoia has proved of fundamental
importance. His Schreber paper (and here we also have to remember Abraham’s paper

26 quoted by Freud)

opened up the possibility of understanding psychosis and the processes underlying it.

26 ‘The Psycho-Sexual Differences between Hysteria and Dementia Praecox’ (1908).

46

PARTTWO

Some British Kleinian developments

Developments by British Kleinian analysts 27

Elizabeth Spillius

I believe that nowadays most Kleinian analysts would agree with Klein herself in thinking that
although projective identifi cation is important, it is not the most central and distinctively
Kleinian concept.
Pride of place is reserved for the comprehensive ideas of the paranoid-schizoid and depressive
positions. Indeed, the term ‘projective identifi cation’ was fi rst defi ned by Klein in her paper
‘Notes on some schizoid mechanisms’

(1946), which is her defi nitive statement of the paranoid-schizoid position.

Nevertheless many papers by contemporary Kleinian British analysts use the idea of projective
identifi cation

clinically and a somewhat smaller number of papers have been specifi cally devoted to
conceptual discussion of the concept (the main discussions occur in Bell, 2001; Bion, 1959;
Britton, 1998b; Feldman, 1992, 1994, 1997; Hinshelwood, 1991; Joseph, 1987; Rosenfeld,
1964b, 1971, 1983; Segal, 1964; Sodré,

2004; see particularly Edna O’Shaughnessy’s Chapter 9 in the present

book; Spillius, 1992). In addition there have been unpublished contributions by Ruth Riesenberg
Malcolm, Hanna Segal, Edith Hargreaves, Priscilla Roth, Robin Anderson and Edna
O’Shaughnessy. There have also been two conferences on the topic of projective identifi cation,
one

27 It is not possible to give a precise defi nition of a ‘British Kleinian analyst’. In general usage
it means an analyst whose training analyst has been a Kleinian. However, a few analysts in all
three British psychoanalytic groups, the Kleinians, the Independents and the Contemporary
Freudians, have changed their affi liation after qualifi cation in one group to that of one of the
other groups. A few British analysts prefer to identify themselves as ‘non-aligned’.

49

Elizabeth Spillius

organised by Joseph Sandler in Jerusalem in 1984 (Sandler, 1987b) and one on ‘Understanding
Projective Identifi cation: Clinical Advances’

held at University College London in October 1995.

Klein’s original delineation of the concept of projective identifi cation has of course been central
to later developments in Britain, and she probably conveyed some of the ideas she describes in
the Archive to her patients, colleagues and students. But in spite of Klein’s infl uence the ideas
of contemporary Kleinian analysts differ from hers in several respects.

The fi rst change from Klein’s view is that all Kleinian analysts now assume that the analyst’s
countertransference, using that term in its broadest sense, is at least in part a response to the
patient’s projective identifi cation and can be a useful source of information about the patient.

The second trend is that contemporary Kleinian analysts now follow Bion (1959) in making a
distinction between normal and pathological projective identifi cation, although until recently
most papers have focused primarily on the pathological aspects.
A third trend is that contemporary Kleinian analysts do not make a point of distinguishing
between projection and projective identifi cation as Klein tried to do in her unpublished 1958
notes in the D17

part of the Klein Archive, as I have described in Part One of this book.

A fourth trend consists of distinctions between various sub-types of projective identifi cation
within the general category (Britton, 1998b, pp. 5–6; Spillius, 1988a, pp. 81–86).

A fi fth trend is a recent but growing recognition that there has been a tendency to think of
projective identifi cation as ‘bad’ and introjective identifi cation as ‘good’, whereas it is likely to
be more useful to recognise both ‘concrete’

and ‘symbolic’ forms of phantasy and thought in both types of identifi cation (Sodré, 2004). This
recognition has been accompanied by a tendency to look at the movements between processes of
projective and introjective identifi cation.

The relation between projective identifi cation

and countertransference

Even before Klein’s death several authors, among them Heimann (1950), Racker (1953; 1957;
1958b) and Money-Kyrle (1956), were writing about the use of countertransference as a source
of informa-50

Developments by British Kleinian analysts tion about the patient. This view has become the most
accepted and ubiquitous of the trends I have described above. It is one of the few ways in which
current contemporary Kleinian thinking and technique differ from that of Klein herself.

In her seminal paper on countertransference

Paula Heimann

(1950) clearly thinks of countertransference as a response to what the patient communicates to

the analyst. She says:

‘From the point of view I am stressing, the analyst’s countertransference is not only part and
parcel of the analytic relationship, but it is the patient’s creation , it is a part of the patient’s
personality’ (Heimann, 1950, p. 83, italics in original). The ‘patient’s creation ’ sounds very
much as if Heimann is thinking here of what Klein and later Kleinian analysts would think of as
the patient’s projective identifi cation, but Heimann does not use this term, and indeed she thinks
of projective identifi cation differently, saying, for example: many instances of so-called
‘projective identifi cation’ should be defi ned as the reactivation in the patient of his infantile
experiences with his rejecting and intruding mother [. . .] ‘Projective identifi cation’ occurs as a
countertransference phenomenon, when the analyst fails in his perceptive functions, so that,
instead of recognizing in good time the character of the transference, he on his part
unconsciously introjects his patient who at this point acts from an identifi cation with his
rejecting and intruding mother, re-enacting his own experiences in a reversal of roles.
(Heimann, 1966, pp. 224 and 230, footnote 1)

Thus Heimann’s use of the term ‘projective identifi cation’ is noticeably different from the usage
of Klein and that of Klein’s later Kleinian colleagues.

Among Klein’s later colleagues the relation between the patient’s projective identifi cation and
the analyst’s countertransference is partly a matter of defi nition. When Wilfred Bion , for
example, uses the actual word

‘countertransference’ he often means it in the sense used by Freud (1910) and Klein as an
expression, largely unconscious, of the analyst’s pathology (see, for example, Bion, 1962a, p. 24;
1963, p. 8). At the same time it is clear from his papers, especially ‘Attacks on linking’ (Bion,
1959, that he uses his own emotional response as a source of information about the patient. In the
wider defi nition of countertransference used by Heimann and Racker this 51

Elizabeth Spillius

sort of response would be called countertransference. Bion describes this sort of response in
1955 in his paper

‘Group dynamics: a review’

(Bion, 1955a) and here, somewhat uncharacteristically, he actually uses the word
‘countertransference’ to describe the analyst’s response to the patient’s projective identifi cation:
Now the experience of counter-transference appears to me to have quite a distinct quality which
should enable the analyst to differentiate the occasion when he is the object of a projective
identifi cation from the occasion when he is not. The analyst feels he is being manipulated so as
to be playing a part, no matter how diffi cult to recognize, in somebody’s [ sic ] else’s phantasy –
or he would do if it were not for what in recollection I can only call a temporary loss of insight, a
sense of experiencing strong feelings and at the same time a belief that their existence is
adequately justifi ed by the objective situation without recourse to recondite explanation of their
causation.

(Bion, 1955a, p. 446)

I take this to mean that the emotions the analyst experiences are aroused at least in part by
projections from the patient (or the group) although the analyst is likely to feel that the emotions
are largely of his own making.

In ‘Language and the schizophrenic’, Bion (1955b) describes a similar process with a psychotic
patient, saying that his interpretations depended on his use of Klein’s theory of projective identifi
cation fi rst to illuminate his countertransference, then to frame his interpretation.

He describes how in a session with a psychotic patient Bion at fi rst felt a growing fear that the
patient would attack him. He interpreted that the patient was pushing into his (Bion’s) insides the
patient’s fear that he would murder Bion. After this interpretation the tension lessened, but the
patient clenched his fi sts. Bion then interpreted that the patient had taken the fear back into
himself and was now afraid that he would make a murderous attack. He goes on to say:
This mode of procedure is open to grave theoretical objections and I think they should be faced [.
. .] I think there are signs that as experience accumulates it may be possible to detect and present
facts which exist, but at present elude clinical acumen; they become observable, at second hand,
through the pressure they exert to 52

Developments by British Kleinian analysts produce what I am aware of as counter-transference. I

would not have it thought that I advocate this use of counter-transference as a fi nal solution;
rather it is an expedient to which we must resort until something better presents itself.

(Bion, 1955b, pp. 224–225)

Roger Money-Kyrle describes a similar process of using the analyst’s countertransference as a

way of understanding the patient.

How exactly a patient does succeed in imposing a phantasy and its corresponding affect upon his
analyst in order to deny it in himself is a most interesting problem [. . .] In the analytic situation,
a peculiarity of communications of this kind is that, at fi rst sight, they do not seem as if they had
been made by the patient at all.

The analyst experiences the affect as being his own response to something. The effort involved is
in differentiating the patient’s contribution from his own.

(Money-Kyrle, 1956, p. 342, footnote 10)

In An Introduction to the Work of Melanie Klein in 1964, Hanna Segal seems to do what Bion
describes as

‘resorting to countertransference until something better presents itself’ (Segal, 1964). She
describes projective identifi cation as an ongoing part of clinical work, includes a description of
the defensive uses of projective identifi cation and refers to the withdrawal of projective identifi
cations as an important aspect of reparation (Segal, 1964, see especially pp. 14–17 and 63–67).

All Kleinian analysts now agree that countertransference in the form of the analyst’s feelings in
response to a patient’s projective identifi cation may be a useful source of information about the
patient, provided, of course, that the analyst is well-trained and reasonably sensitive. The work of
Bion, Segal and Joseph has been infl uential in producing this change, but it has also involved a
general change in the defi nition of countertransference. Most British analysts have adopted the
wider defi nition of countertransference advocated by Paula Heimann (1950) and others rather
than the narrower defi nition of Freud and Klein. This change in the defi nition of
countertransference has meant that projective identifi cation is now sometimes considered to be
interpersonal as well as intrapersonal, depending on whether some sort of pressure is put on the
object to conform to the subject’s phantasy.

53

Elizabeth Spillius

Normal and pathological forms of

projective identifi cation

In his infl uential paper ‘Attacks on linking’ (Bion, 1959, reprinted in

Chapter 4 of this book), Bion makes an important distinction between normal and pathological
forms of projective identifi cation. In his view projective identifi cation is an early and very
important form of non-verbal communication. He says: ‘I shall suppose that there is a normal
degree of projective identifi cation, without defi

ning the limits within which normality lies, and that associated with introjective identifi cation
this is the foundation on which normal development rests’ (Bion, 1967, p. 103).

In ‘Attacks on linking’ Bion then describes several fascinating but obscure clinical episodes of
communication between patient and analyst in which something pathological had apparently
happened that had damaged the patient’s capacity for verbal communication.

Bion goes on to say that an infant is dependent on the receptiveness of the object, typically his
mother, to accept projections and to act in such a way that the messages of the projections are
understood and appropriately responded to. If the object fails to receive the projections the infant
will very probably increase his projective efforts, an escalation which is likely to be repeated
later in the individual’s analysis. Bion describes the process as follows: In the analysis a complex
situation may be observed. The patient feels he is being allowed an opportunity of which he had
hitherto been cheated; the poignancy of his deprivation is thereby rendered the more acute and so
are the feelings of resentment at the deprivation. Gratitude for the opportunity co-exists with
hostility to the analyst as the person who will not understand and refuses the patient the use of
the only method of communication by which he feels he can make himself understood. Thus the
link between patient and analyst, or infant and breast, is the mechanism of projective identifi
cation.

(Bion, 1967, pp. 104–105)

But the object’s imperviousness to projection is not the only form of attack on the link between
infant and object.

The infant or later the patient may be so envious of the object that he too attacks the link by
assuming that the object’s capacity to receive is a greedy 54

Developments by British Kleinian analysts attempt to take in the patient’s projections in order to
destroy them.

Between the subject’s envy and the object’s unreceptiveness the links are damaged,
communication is disrupted, curiosity is dulled and a destructive superego is likely to be
installed.

In later papers Bion goes on to develop further the idea of the subject’s projective identifi cation
and the object’s response, particularly in his model of the process of containment (Bion, 1962a,
1963, 1965, 1970).
Herbert Rosenfeld, Betty Joseph and others have reiterated Bion’s distinction between normal
and pathological projective identifi cation. Their descriptions of projective identifi cation,
however, have tended to focus on its pathological aspects.

Herbert Rosenfeld has made several important contributions to the understanding of projective
identifi cation. In 1947 he describes how his depersonalised patient used projective identifi
cation to protect herself and to control others (Rosenfeld, 1947). In 1952 he describes the
intrusiveness of a patient’s use of projective identifi cation to control his analyst (Rosenfeld,
1952a). In a paper on narcissism in 1964 he draws attention to the fact that identifi cations by
projection and by introjection usually occur at the same time (Rosenfeld, 1964b). In omnipotent
identifi cation, which, he says, may occur by both introjection and by projection, the boundary
between self and object is denied.

Identifi cation is an important factor in narcissistic object relations.

It may take place by introjection or by projection. When the object is omnipotently incorporated,
the self becomes so identifi ed with the incorporated object that all separate identity or any
boundary between self and object is denied.

In projective identifi -

cation parts of the self omnipotently enter an object, for example the mother, to take over certain
qualities which would be experienced as desirable, and therefore claim to be the object or part-
object. Identifi cation by introjection and by projection usually occur simultaneously.

In narcissistic object relations defences against any recognition of separateness between self and
object play a

predominant part.

(Rosenfeld, 1964b, pp. 170–171)

In a paper of 1971 Rosenfeld makes an important addition to the understanding of projective

identifi cation, particularly in the case of 55

Elizabeth Spillius

psychotic patients (Rosenfeld, 1971, reprinted in Chapter 5 of this book). He emphasises the
splitting of the patient’s mind that is involved in the projective identifi cation of parts of the mind
into an object. He develops a differentiation of several general motives for projective identifi
cation: communication, which is based on the normal non-verbal communication between infant
and mother; ridding the self of unwanted aspects, which leads to a denial of psychic reality; to
control the mind and body of the other, the other being the analyst in the analytic situation; to get
rid of awareness of separateness and envy; as a special form, particularly in the case of psychotic
patients, of parasitism which involves the patient having an unconscious phantasy of virtually
living inside the mind of the object.

Rosenfeld further elaborates his ideas about these motives and their clinical expression in a later
paper called

‘Primitive object relations and mechanisms’ (Rosenfeld, 1983) and in his book Impasse and
Interpretation (Rosenfeld, 1987a).

Betty Joseph gives a very clear clinical demonstration of her use of the concept of projective
identifi cation in

‘Projective identifi cation: some clinical aspects’ ( Joseph, 1987, reprinted in Chapter 6 of

this book), a paper given at the conference on projective identifi cation at the University of
Jerusalem in 1985.

Three features are distinctive of this paper. First, like Rosenfeld and unlike Klein, Joseph focuses
in this paper almost exclusively on the aggressive and negative aspects of projective identifi
cation: getting rid of unwanted parts of the self; dominating and controlling the object; avoiding
feelings of separateness; taking over the capacities of the object; invading the object in order to
damage or destroy it. She mentions the communicative functions of projective identifi cation
described by Bion and Rosenfeld but adds that projective identifi cation is by its nature a form of
communication even when communication is not its aim. Second, she emphasises the importance
of the patient’s defensive efforts to use projective identifi cation to maintain his psychic
equilibrium, illustrating the effect of such defensive efforts in the case of a psychotic child, an
adult patient ‘more or less stuck in the paranoid-schizoid position’, and fi nally in a patient
beginning to move towards the depressive position. The third feature of Joseph’s paper is her
perceptiveness in understanding the effect of her patient’s projections on her own emotional
state, making it possible for her to understand not only her patients’ emotional states but also
their use of projective identifi cation to maintain their equilibrium. Unlike Klein, 56

Developments by British Kleinian analysts who thought the analyst should overcome the
emotional effect of the patient’s projection, Joseph uses that emotional effect as a central basis of
her understanding.

Ruth Riesenberg Malcolm (1970) describes a perverse phantasy by means of which a patient
attempted to get rid of disturbing parts of herself by projective identifi cation, leading to a
vicious circle of projection and forceful re-entry.

Robert Hinshelwood (1991) gives a comprehensive account of the defi nitions and uses of the
concept of projective identifi cation not only by Klein and her contemporary Kleinian colleagues
but also by American analysts, most of whom, he says, have taken the concept out of its
framework in Kleinian theory. He concludes his extensive discussion by saying, ‘There appears
to be no consensus on the value of the term “projective identifi cation” outside the Kleinian
conceptual framework’ (Hinshelwood, 1991, p. 204).

Michael Feldman in his paper ‘Projective identifi cation: the

analyst’s involvement’ (Feldman, 1997, reprinted in Chapter 7 of this

book) shows in detail how patients may use projective identifi cation in a way that impinges on
the analyst’s thinking, feeling and actions in such a fashion that he is drawn into some form of
mutual enactment with the patient, an enactment which is essential for the fulfi lment of the
patient’s unconscious aims.

Projection and projective identifi cation

Although, as I described in Part One, Klein made a distinction – in my

view a rather unclear distinction – between projection and projective identifi cation in her
unpublished notes of 1958, contemporary Kleinian analysts have not followed her usage in this
respect. The usual though tacit attitude of present Kleinian analysts is that there is no really clear
distinction, especially no clear distinction in clinical work, between projection and projective
identifi cation.

Segal (personal communication) says that Klein viewed ‘projection’ as the mental mechanism
and ‘projective identifi cation’ as the particular unconscious phantasy expressing it. Bell (2001)
also mentions this distinction.

Neither Segal nor Bell nor any other British Kleinian analyst distinguishes between projection
and projective identifi cation in their own clinical work, although such a distinction

has become important to many American analysts (see Chapter 15 , ‘A 57

Elizabeth Spillius

brief review of projective identifi cation in American psychoanalytic literature’).

Varieties of projective identifi cation

I have suggested (Spillius, 1988a, pp. 81–86) that it might be best to use the idea of projective
identifi cation as a general ‘umbrella’ concept covering a wide variety of specifi c types of
projective identifi cation, although it may also be useful to use particular adjectives to distinguish
specifi c sub-types within the general category. I suggested the term ‘evocatory’ to describe the
sort of projective identifi cation in which a patient’s unconscious phantasy is accompanied by
behaviours unconsciously designed to evoke a specifi c sort of response from the object.

Ronald Britton has made a useful distinction between ‘attributive’ projective identifi cation, in
which some aspect of the subject is attributed to the object, and ‘acquisitive’ projective identifi
cation in which the projective phantasy concerns an idea of entering the object to acquire some
attribute that the object is thought to possess (Britton, 1998b, pp. 5–6). Britton’s idea of
acquisitive projective identifi cation describes a process similar to Bollas’s idea of

‘extractive introjection’

(Bollas, 1987, pp. 157–169), indicating the close relation between projective and introjective
processes. Most phantasies involving projective identifi cation include elements of both
attribution and of acquisition, just as most phantasies of psychic interaction involve both
phantasies of projection and of introjection.
Concrete and symbolic thinking in both projective

and introjective identifi cation

In her paper ‘Who’s who? Notes on pathological identifi cation’, Ignes Sodré makes another
valuable addition to the understanding of processes of identifi cation (Sodré, 2004, reprinted in
Chapter 8 of

this book). She reminds the reader of Rosenfeld’s 1964 paper described above in which he says
that omnipotent identifi cation, whether by projection or by introjection, obliterates the boundary
between self and object: 58

Developments by British Kleinian analysts Even though ‘projective identifi cation’ is used to
describe normal as

well as pathological processes, I think that we tend to think of projective processes as more
pathological than introjective ones.

(Sodré, 2004, p. 57)

Sodré argues that introjective processes and introjective identifi cation can be just as pathological
as projective identifi cation. The pathological element in identifi cation is not, she says, whether
it is projective or introjective; it is whether the identifi cation is concrete or symbolic. Sodré
makes clear, I think, why we have tended to describe certain identifi catory processes as
‘projective’ when it might have been more useful to have described them as projective identifi -

cation followed by introjective identifi cation. If the identifi cation is thought to involve ‘bad’
attributes, we tend to think of it as ‘projective’; if it involves ‘good’ attributes, we tend to think
of it as ‘introjective’. Sodré illustrates her thesis with her own clinical material but I think it can
also be applied to other instances.

For example, in Klein’s example of Fabian, the hero of Julian Green’s novel If I Were You
(Klein, 1955), Fabian’s behaviour could be regarded as an instance of acquisitive projective
identifi cation followed by introjection. Fabian projects himself into other people, according to
Klein’s analysis, but could also be seen as taking the resulting amalgam of Fabian plus other
person back into himself to the point of not knowing who was who.

In describing what he terms the ‘identifi cate’ Leslie Sohn (1985) describes the way certain
narcissistic patients project themselves into the object to take possession of its desirable qualities,
an instance of what Britton calls

‘acquisitive’ projective identifi cation. Sohn describes this entirely as a projective process,
although I think one might view it as projection followed by introjection of the desired self/other
identifi cate which is then claimed to be oneself.

With the work of Bion, Rosenfeld, Joseph, Britton, Feldman, Sohn, Sodré and many other
Kleinian analysts, I believe that the ideas of projection and introjection have become more
clearly defi ned but also more interconnected with one another to create a coherent approach to
the analysis of episodes of intrapsychic and interpersonal interaction. Normality and pathology
have come to be regarded not as resting on whether identifi cation is projective or introjective,
but rather on the motive and content of the identifi cations.

59

Elizabeth Spillius

Conclusion

The work of British contemporary Kleinian analysts has added several dimensions to Klein’s
original discussion of projective identifi cation.

The analyst’s countertransference, using that term in its widest sense, can be a useful source of
information about the patient. This change has contributed to the process of regarding projective
identifi cation as an interpersonal as well as an intrapersonal process.

Bion’s distinction between normal and pathological aspects of projective identifi cation (Bion,
1959, reprinted in

Chapter 4 of this book) is

now generally accepted, as are the motives described by Rosenfeld (1971) as characteristic of
projective identifi cation: communication; control of the other; avoidance of envy; parasitism.

I have described the possible usefulness of describing sub-types of projective identifi cation such
as my suggested

‘evocative’ and

‘non-evocative’ and Britton’s ‘attributive’ and ‘acquisitive’ projective identifi cation. Although
this chapter has been

chiefl y concerned with the usage and developments in the use of the concept of projective
identifi cation by

Kleinian analysts, it should be noted that some Kleinian analysts have been infl uenced by
Sandler’s concept of

‘actualization’ as expressed in two of his papers in 1976 (Sandler, 1976a; 1976b).

Finally, there has been a general tendency to focus on the pathological aspects of projective
identifi cation, but this has been corrected, especially clearly shown in the work of Ignes Sodré
(2004, reprinted

in Chapter 8 of this book), by emphasis on the presence of both

constructive and pathological aspects in both projective and introjective identifi cations. There
has also been a gradual recognition that projection and introjection ‘operate simultaneously’, as
Klein put it, a process further stressed by Rosenfeld (1964b). Further, I think that these
simultaneous processes occur in both (or perhaps all) members of an interaction, a process which
gives psychic and social interchanges immense complexity and variability.

60

Attacks on linking 28

W. R. Bion

In previous papers (Bion, 1957a) I have had occasion, in talking of the psychotic part of the
personality, to speak of the destructive attacks which the patient makes on anything which is felt
to have the function of linking one object with another. It is my intention in this paper to show
the signifi cance of this form of destructive attack in the production of some symptoms met with
in borderline psychosis.

The prototype for all the links of which I wish to speak is the primitive breast or penis. The paper
presupposes familiarity with Melanie Klein’s descriptions of the infant’s fantasies of sadistic
attacks upon the breast (Klein, 1935), of the infant’s splitting of its objects, of projective identifi
cation, which is the name she gives to the mechanism by which parts of the personality are split
off and projected into external objects, and fi nally her views on early stages of the Oedipus
complex (Klein, 1928). I shall discuss phantasied attacks on the breast as the prototype of all
attacks on objects that serve as a link and projective identifi cation as the mechanism employed
by the psyche to dispose of the ego fragments produced by its destructiveness.

I shall fi rst describe clinical manifestations in an order dictated not by the chronology of their
appearance in the consulting room, but by the need for making the exposition of my thesis as
clear as I can. I shall follow this by material selected to demonstrate the order which these
mechanisms assume when their relationship to each other is

28 This paper was read before the British Psycho-Analytical Society on 20 October 1957.

First published in 1959 in the International Journal of Psycho-Analysis, 40 , 308–315.

61

W. R. Bion

determined by the dynamics of the analytic situation. I shall conclude with theoretical
observations on the material presented. The examples are drawn from the analysis of two
patients and are taken from an advanced stage of their analyses. To preserve anonymity I shall
not distinguish between the patients and shall introduce distortions of fact which I hope do not
impair the accuracy of the analytic description.

Observation of the patient’s disposition to attack the link between two objects is simplifi ed
because the analyst has
to establish a link with the patient and does this by verbal communication and his equipment of
psycho-analytical experience. Upon this the creative relationship depends and therefore we
should be able to see attacks being made upon it.

I am not concerned with typical resistance to interpretations, but with expanding references
which I made in my paper on ‘The Differentiation of the Psychotic from the Non-psychotic Part
of the Personality’ (Bion, 1957a) to the destructive attacks on verbal thought itself.

Clinical examples

I shall now describe occasions which afforded me an opportunity to give the patient an
interpretation, which at that point he could understand, of conduct designed to destroy whatever
it was that linked two objects together.

These are the examples:

i. I had reason to give the patient an interpretation making explicit his feelings of affection and
his expression of them to his mother for her ability to cope with a refractory child. The patient
attempted to express his agreement with me, but although he needed to say only a few words his
expression of them was interrupted by a very pronounced stammer which had the effect of
spreading out his remark over a period of as much as a minute and a half. The actual sounds
emitted bore resemblance to gasping for breath; gaspings were interspersed with gurgling sounds
as if he were immersed in water. I drew his attention to these sounds and he agreed that they
were peculiar and himself suggested the descriptions I have just given.

ii. The patient complained that he could not sleep. Showing signs of fear, he said, ‘It can’t go on
like this’.

Disjointed remarks gave the impression that he felt superfi cially that some catastrophe would 62

Attacks on linking

occur, perhaps akin to insanity, if he could not get more sleep.

Referring to material in the previous session I suggested that he feared he would dream if he
were to sleep. He denied this and said he could not think because he was wet. I reminded him of
his use of the term ‘wet’ as an expression of contempt for somebody he regarded as feeble and
sentimental. He disagreed and indicated that the state to which he referred was the exact
opposite. From what I knew of this patient I felt that his correction at this point was valid and
that somehow the wetness referred to an expression of hatred and envy such as he associated
with urinary attacks on an object. I therefore said that in addition to the superfi cial fear which he
had expressed he was afraid of sleep because for him it was the same thing as the oozing away of
his mind itself. Further associations showed that he felt that good interpretations from me were
so consistently and minutely split up by him that they became mental urine which then seeped
uncontrollably away. Sleep was therefore inseparable from unconsciousness, which was itself
identical with a state of mindlessness which could not be repaired. He said, ‘I am dry now’. I
replied that he felt he was awake and capable of thought, but that this good state was only
precariously maintained.
iii. In this session the patient had produced material stimulated by the preceding week-end break.
His awareness of such external stimuli had become demonstrable at a comparatively recent stage
of the analysis. Previously it was a matter for conjecture how much he was capable of
appreciating reality. I knew that he had contact with reality because he came for analysis by
himself, but that fact could hardly be deduced from his behaviour in the sessions.

When I interpreted some associations as evidence that he felt he had been and still was
witnessing an intercourse between two people, he reacted as if he had received a violent blow. I
was not then able to say just where he had experienced the assault and even in retrospect I have
no clear impression.

It would seem logical to suppose that the shock had been administered by my interpretation and
that therefore the blow came from without, but my impression is that he felt it as delivered from
within; the patient often experienced what he described as a stabbing attack from inside. He sat
up and stared intently into space. I said that he seemed to be seeing something. He replied that he
could not see what he saw. I was able from previous experience to interpret

that he felt he was ‘seeing’ an invisible object and subsequent experience convinced me that in
the two patients on whose analysis I am 63

W. R. Bion

depending for material for this paper, events occurred in which the patient experienced invisible-
visual hallucinations. I shall give my reasons later for supposing that in this and the previous
example similar mechanisms were at work.

iv. In the fi rst twenty minutes of the session the patient made three isolated remarks which had
no signifi cance for me. He then said that it seemed that a girl he had met was understanding.
This was followed at once by a violent, convulsive movement which he affected to ignore. It
appeared to be identical with the kind of stabbing attack I mentioned in the last example. I tried
to draw his attention to the movement, but he ignored my intervention as he ignored the attack.

He then said that the room was fi lled with a blue haze. A little later he remarked that the haze
had gone, but said he was depressed. I interpreted that he felt understood by me. This was an
agreeable experience, but the pleasant feeling of being understood had been instantly destroyed
and ejected. I reminded him that we had recently witnessed his use of the word ‘blue’ as a
compact description of vitu-perative sexual conversation. If my interpretation was correct, and
subsequent events suggested that it was, it meant that the experience of being understood had
been split up, converted into particles of sexual abuse and ejected. Up to this point I felt that the
interpretation approximated closely to his experience. Later interpretations, that the
disappearance of the haze was due to reintrojection and conversion into depression, seemed to
have less reality for the patient, although later events were compatible with its being correct.

v. The session, like the one in my last example, began with three or four statements of fact such
as that it was hot, that his train was crowded, and that it was Wednesday; this occupied thirty
minutes.

An impression that he was trying to retain contact with reality was confi rmed when he followed
up by saying that he feared a breakdown. A little later he said I would not understand him. I
interpreted that he felt I was bad and would not take in what he wanted to put into me. I
interpreted in these terms deliberately because he had shown in the previous session that he felt
that my interpretations were an attempt to eject feelings that he wished to deposit in me. His
response to my interpretation was to say that he felt there were two probability clouds in the
room. I interpreted that he was trying to get rid of the feeling that my badness was a fact. I said it
meant that he needed to know whether I was really bad or whether I was some bad thing which
had come from inside him. Although the point was not 64

Attacks on linking

at the moment of central signifi cance I though the patient was attempting to decide whether he
was hallucinated or not. This recurrent anxiety in his analysis was associated with his fear that
envy and hatred of a capacity for understanding was leading him to take in a good, understanding
object to destroy and eject it – a procedure which had often led to persecution by the destroyed
and ejected object. Whether my refusal to understand was a reality or hallucination was
important only because it determined what painful experiences were to be expected next.

vi. Half the session passed in silence; the patient then announced that a piece of iron had fallen
on the fl oor.

Thereafter he made a series of convulsive movements in silence as if he felt he was being

physically assaulted from within. I said he could not establish contact with me because of his
fear of what was going on inside him. He confi rmed this by saying that he felt he was being
murdered. He did not know what he would do without the analysis as it made him better. I said
that he felt so envious of himself and of me for being able to work together to make him feel
better that he took the pair of us into him as a dead piece of iron and a dead fl oor that came
together not to give him life but to murder him. He became very anxious and said he could not
go on. I said that he felt he could not go on because he was either dead, or alive and so envious
that he had to stop good analysis. There was a marked decrease of anxiety, but the remainder of
the session was taken up by isolated statements of fact which again seemed to be an attempt to
preserve contact with external reality as a method of denial of his phantasies.

Features common to the above illustrations

These episodes have been chosen by me because the dominant theme in each was the destructive
attack on a link. In the fi rst the attack was expressed in a stammer which was designed to
prevent the patient from using language as a bond between him and me. In the second sleep was
felt by him to be identical with projective identifi cation that proceeded unaffected by any
possible attempt at control by him.

Sleep for him meant that his mind, minutely fragmented, fl owed out in an attacking stream of
particles.

The examples I give here throw light on schizophrenic dreaming.

The psychotic patient appears to have no dreams, or at least not to 65

W. R. Bion

report any, until comparatively late in the analysis. My impression now is that this apparently
dreamless period is a phenomenon analogous to the invisible-visual hallucination. That is to say,
that the dreams consist of material so minutely fragmented that they are devoid of any visual
component. When dreams are experienced which the patient can report because visual objects
have been experienced by him in the course of the dream, he seems to regard these objects as
bearing much the same relationship to the invisible objects of the previous phase as faeces seem
to him to bear to urine. The objects appearing in experiences which we call dreams are regarded
by the patient as solid and are, as such, contrasted with the contents of the dreams which were a
continuum of minute, invisible fragments.

At the time of the session the main theme was not an attack on the link but the consequences of
such an attack, previously made, in leaving him bereft of a state of mind necessary for the
establishment of a satisfying relationship between him and his bed. Though it did not appear in
the session I report, uncontrollable projective identifi -

cation, which was what sleep meant to him, was thought to be a destructive attack on the state of
mind of the coupling parents. There was therefore a double anxiety; one arising from his fear that
he was being rendered mindless, the other from his fear that he was unable to control his hostile
attacks, his mind providing the ammunition, on the state of mind that was the link between the
parental pair. Sleep and sleeplessness were alike inacceptable.

In the third example in which I described visual hallucinations of invisible objects, we witness
one form in which the actual attack on the sexual pair is delivered. My interpretation, as far as I
could judge, was felt by him as if it were his own visual sense of a parental intercourse; this
visual impression is minutely fragmented and ejected at once in particles so minute that they are
the invisible components of a continuum. The total procedure has served the purpose of fore-
stalling an experience of feelings of envy for the parental state of mind by the instantaneous
expression of envy in a destructive act. I shall have more to say of this implicit hatred of emotion
and the need to avoid awareness of it.

In my fourth example, the report of the understanding girl and the haze, my understanding and
his agreeable state of mind have been felt as a link between us which could give rise to a creative
act. The link had been regarded with hate and transformed into a hostile and destructive sexuality
rendering the patient-analyst couple sterile.

66

Attacks on linking

In my fi fth example, of the two probability clouds, a capacity for understanding is the link
which is being attacked, but the interest lies in the fact that the object making the destructive
attacks is alien to the patient. Furthermore, the destroyer is making an attack on projective
identifi cation which is felt by the patient to be a method of communication. In so far as my
supposed attack on his methods of communication is felt as possibly secondary to his envious
attacks on me, he does not dissociate himself from feelings of guilt and responsibility. A further
point is the appearance of judgement, which Freud regards as an essential feature of the
dominance of the reality principle, among the ejected parts of the patient’s personality. The fact
that there were two probability clouds remained unexplained at the time, but in subsequent
sessions I had material which led me to suppose that what had originally

been an attempt to separate good from bad survived in the existence of two objects, but they
were now similar in that each was a mixture of good and bad. Taking into consideration material
from later sessions, I can draw conclusions which were not possible at the time; his capacity for
judgment, which had been split up and destroyed with the rest of his ego and then ejected, was
felt by him to be similar to other bizarre objects of the kind which I have described in my paper
on ‘The Differentiation of the Psychotic from the Non-Psychotic parts of the Personality’.

These ejected particles were feared because of the treatment he had accorded them. He felt that
the alienated judgment – the probability clouds – indicated that I was probably bad. His
suspicion that the probability clouds were persecutory and hostile led him to doubt the value of
the guidance they afforded him. They might supply him with a correct assessment or a
deliberately false one, such as that a fact was an hallucination or vice versa; or would give rise to
what, from a psychiatric point of view, we would call delusions. The probability clouds
themselves had some qualities of a primitive breast and were felt to be enigmatic and
intimidating.

In my sixth illustration, the report that a piece of iron had fallen on the fl oor, I had no occasion
for interpreting an aspect of the material with which the patient had by this time become familiar
(I should perhaps say that experience had taught me that there were times when I assumed the
patient’s familiarity with some aspect of a situation with which we were dealing, only to discover
that, in spite of the work that had been done upon it, he had forgotten it).

The familiar point that I did not interpret, but which is signifi cant for the 67

W. R. Bion

understanding of this episode, is that the patient’s envy of the parental couple had been evaded
by his substitution of himself and myself for the parents. The evasion failed, for the envy and
hatred were now directed against him and me. The couple engaged in a creative act are felt to be
sharing an enviable, emotional experience; he, being identifi ed also with the excluded party, has
a painful, emotional experience as well. On many occasions the patient, partly through
experiences of the kind which I describe in this episode, and partly for reasons on which I shall
enlarge later, had a hatred of emotion, and therefore, by a short extension, of life itself. This
hatred contributes to the murderous attack on that which links the pair, on the pair itself and on
the object generated by the pair. In the episode I am describing, the patient is suffering the
consequences of his early attacks on the state of mind that forms the link between the creative
pair and his identifi cation with both the hateful and creative states of mind.

In this and the preceding illustration there are elements that suggest the formation of a hostile
persecutory object, or agglomeration of objects, which expresses its hostility in a manner which
is of great importance in producing the predominance of psychotic mechanisms in a patient; the
characteristics with which I have already invested the agglomeration of persecutory objects have
the quality of a primitive, and even murderous, superego.
Curiosity, arrogance and stupidity

In the paper I presented at the International Congress of 1957 (Bion, 1957b) I suggested that
Freud’s analogy of an archaeological investigation with a psycho-analysis was helpful if it were
considered that we were exposing evidence not so much of a primitive civilization as of a
primitive disaster. The value of the analogy is lessened because in the analysis we are confronted
not so much with a static situation that permits leisurely study, but with a catastrophe that
remains at one and the same moment actively vital and yet incapable of resolution into
quiescence.

This lack of progress in any direction must be attributed in part to the destruction of a capacity
for curiosity and the consequent inability to learn, but before I go into this I must say something
about a matter that plays hardly any part in the illustrations I have given.

68

Attacks on linking

Attacks on the link originate in what Melanie Klein calls the paranoid-schizoid phase. This
period is dominated by part-object relationships (Klein, 1948). If it is borne in mind that the
patient has a part-object relationship with himself as well as with objects not himself, it
contributes to the understanding of phrases such as ‘it seems’

which are commonly employed by the deeply disturbed patient on occasions when a less
disturbed patient might say

‘I think’ or ‘I believe’.

When he says ‘it seems’ he is often referring to a feeling – an ‘it seems’

feeling – which is a part of his psyche and yet is not observed as part of a whole object. The
conception of the part-object as analogous to an anatomical structure, encouraged by the
patient’s employment of concrete images as units of thought, is misleading because the part-
object relationship is not with the anatomical structures only but with function, not with anatomy
but with physiology, not with the breast but with feeding, poisoning, loving, hating. This
contributes to the impression of a disaster that is dynamic and not static. The problem that has to
be solved on this early, yet superfi cial, level must be stated in adult terms by the question, ‘What
is something?’ and not the question

‘Why is something?’ because ‘why’ has, through guilt, been split off. Problems, the solution of
which depends upon an awareness of causation, cannot therefore be stated, let alone solved. This
produces a situation in which the patient appears to have no problems except those posed by the
existence of analyst and patient. His preoccupation is with what is this or that function, of which
he is aware though unable to grasp the totality of which the function is a part.

It follows that there is never any question why the patient or the analyst is there, or why
something is said or done or felt, nor can there be any question of attempting to alter the causes
of some state of mind. . . . Since
‘what?’ can never be answered without ‘how?’ or ‘why?’ further diffi -

culties arise. I shall leave this on one side to consider the mechanisms employed by the infant to
solve the problem

‘what?’ when it is felt in relation to a part-object relationship with a function.

Denial of normal degrees of projective identifi cation I employ the term ‘link’ because I wish
to discuss the patient’s relationship with a function rather than with the object that subserves a
function; my concern is not only with the breast, or penis, or verbal thought, but with their
function of providing the link between two objects.

69

W. R. Bion

In her ‘Notes on Some Schizoid Mechanisms’ (Klein, 1946) Melanie Klein speaks of the
importance of an excessive employment of splitting and projective identifi cation in the
production of a very disturbed personality. She also speaks of ‘the introjection of the good
object, fi rst of all the mother’s breast’ as a ‘precondition for normal development’. I shall
suppose that there is a normal degree of projective identifi cation, without defi ning the limits
within which normality lies, and that associated with introjective identifi cation this is the
foundation on which normal development rests.

This impression derives partly from a feature in a patient’s analysis which was diffi cult to
interpret because it did not appear to be suffi -

ciently obtrusive at any moment for an interpretation to be supported by convincing evidence.

Throughout the analysis the patient resorted to projective identifi cation with a persistence
suggesting it was a mechanism of which he had never been able suffi ciently to avail himself; the
analysis afforded him an opportunity for the exercise of a mechanism of which he had been
cheated. I did not have to rely on this impression alone. There were sessions which led me to
suppose that the patient felt there was some object that denied him the use of projective identifi
cation. In the illustrations I have given, particularly in the fi rst, the stammer, and the fourth, the
understanding girl and the blue haze, there are elements which indicate that the patient felt that
parts of his personality that he wished to repose in me were refused entry by me, but there had
been associations prior to this which led me to this view.

When the patient strove to rid himself of fears of death which were felt to be too powerful for his
personality to contain he split off his fears and put them into me, the idea apparently being that if
they were allowed to repose there long enough they would undergo modifi cation by my psyche
and could then be safely reintrojected.

On the occasion I have in mind the patient had felt, probably for reasons similar to those I give in
my fi fth illustration, the probability clouds, that I evacuated them so quickly that the feelings
were not modifi ed, but had become more painful.

Associations from a period in the analysis earlier than that from which these illustrations have
been drawn showed an increasing intensity of emotions in the patient. This originated in what he
felt was my refusal to accept parts of his personality. Consequently he strove to force them into
me with increased desperation and violence.

His behaviour, isolated from the context of the analysis, might have appeared to be an expression
of primary aggression. The more violent 70

Attacks on linking

his phantasies of projective identifi cation, the more frightened he became of me. There were
sessions in which such behaviour expressed unprovoked aggression, but I quote this series
because it shows the patient in a different light, his violence a reaction to what he felt was my
hostile defensiveness. The analytic situation built up in my mind a sense of witnessing an
extremely early scene. I felt that the patient had experienced in infancy a mother who dutifully
responded to the infant’s emotional displays. The dutiful response had in it an element of
impatient ‘I don’t know what’s the matter with the child.’ My deduction was that in order to
understand what the child wanted the mother should have treated the infant’s cry as more than a
demand for her presence. From the infant’s point of view she should have taken into her, and
thus experienced, the fear that the child was dying. It was this fear that the child could not
contain. He strove to split it off together with the part of the personality in which it lay and
project it into the mother. An understanding mother is able to experience the feeling of dread,
that this baby was striving to deal with by projective identifi cation, and yet retain a balanced
outlook. This patient had had to deal with a mother who could not tolerate experiencing such
feelings and reacted either by denying them ingress, or alternatively by becoming a prey to the
anxiety which resulted from introjection of the infant’s feelings. The latter reaction must, I think,
have been rare: denial was dominant.

To some this reconstruction will appear to be unduly fanciful; to me it does not seem forced and
is the reply to any who may object that too much stress is placed on the transference to the
exclusion of a proper elucidation of early memories.

In the analysis a complex situation may be observed. The patient feels he is being allowed an
opportunity of which he had hitherto been cheated; the poignancy of his deprivation is thereby
rendered the more acute and so are the feelings of resentment at the deprivation. Gratitude for the
opportunity coexists with hostility to the analyst as the person who will not understand and
refuses the patient the use of the only method of communication by which he feels he can make
himself understood. Thus the link between patient and analyst, or infant and breast, is the
mechanism of projective identifi -

cation. The destructive attacks upon this link originate in a source external to the patient or
infant, namely the analyst or breast. The result is excessive projective identifi cation by the
patient and a deterioration of his developmental processes.

71

W. R. Bion

I do not put forward this experience as the cause of the patient’s disturbance; that fi nds its main
source in the inborn disposition of the infant as I described it in my paper on ‘The Differentiation
of the Psychotic from the Non-psychotic Part of the Personality’ (Bion, 1957a). I regard it as a
central feature of the environmental factor in the production of the psychotic personality.

Before I discuss this consequence for the patient’s development, I must refer to the inborn
characteristics and the part that they play in producing attacks by the infant on all that links him
to the breast, namely, primary aggression and envy. The seriousness of these attacks is enhanced
if the mother displays the kind of unreceptiveness which I have described, and is diminished, but
not abolished, if the mother can introject the infant’s feelings and remain

balanced (Klein, 1957); the seriousness remains because the psychotic infant is overwhelmed
with hatred and envy of the mother’s ability to retain a comfortable state of mind although
experiencing the infant’s feelings. This was clearly brought out by a patient who insisted that I
must go through it with him, but was fi lled with hate when he felt I was able to do so without a
breakdown. Here we have another aspect of destructive attacks upon the link, the link being the
capacity of the analyst to introject the patient’s projective identifi cations. Attacks on the link,
therefore, are synonymous with attacks on the analyst’s, and originally the mother’s, peace of
mind. The capacity to introject is transformed by the patient’s envy and hate into greed
devouring the patient’s psyche; similarly, peace of mind becomes hostile indifference. At this
point analytic problems arise through the patient’s employment (to destroy the peace of mind
that is so much envied) of acting out, delinquent acts and threats of suicide.

Consequences

To review the main features so far: the origin of the disturbance is twofold. On the one hand
there is the patient’s inborn disposition to excessive destructiveness, hatred, and envy: on the
other the environment which, at its worst, denies to the patient the use of the mechanisms of
splitting and projective identifi cation. On some occasions the destructive attacks on the link
between patient and environment, or between different aspects of the patient’s personality, have
their origin in the patient; on others, in the mother, although in the latter instance and in
psychotic patients, it can never be in the mother 72

Attacks on linking

alone. The disturbances commence with life itself. The problem that confronts the patient is:
What are the objects of which he is aware?

These objects, whether internal or external, are in fact part-objects and predominantly, though
not exclusively, what we should call functions and not morphological structures. This is
obscured because the patient’s thinking is conducted by means of concrete objects and therefore
tends to produce, in the sophisticated mind of the analyst, an impression that the patient’s
concern is with the nature of the concrete object. The nature of the functions which excite the
patient’s curiosity he explores by projective identifi cation. His own feelings, too powerful to be
contained within his personality, are amongst these functions. Projective identifi cation makes it
possible for him to investigate his own feelings in a personality powerful enough to contain
them. Denial of the use of this mechanism, either by the refusal of the mother to serve as a
repository for the infant’s feelings, or by the hatred and envy of the patient who cannot allow the
mother to exercise this function, leads to a destruction of the link between infant and breast and
consequently to a severe disorder of the impulse to be curious on which all learning depends.
The way is therefore prepared for a severe arrest of development. Furthermore, thanks to a denial
of the main method open to the infant for dealing with his too powerful emotions, the conduct of
emotional life, in any case a severe problem, becomes intolerable. Feelings of hatred are
thereupon directed against all emotions including hate itself, and against external reality which
stimulates them. It is a short step from hatred of the emotions to hatred of life itself. As I said in
my paper on ‘The Differentiation of the Psychotic from the Non-psychotic Part of the
Personality’ (Bion, 1957a), this hatred results in a resort to projective identifi cation of all the
perceptual apparatus including the embryonic thought which forms a link between sense
impressions and consciousness. The tendency to excessive projective identifi cation when death
instincts predominate is thus reinforced.

Superego

The early development of the superego is effected by this kind of mental functioning in a way I
must now describe.

As I have said, the link between infant and breast depends upon projective identifi cation and a
capacity to introject projective identifi cations. Failure to introject 73

W. R. Bion

makes the external object appear intrinsically hostile to curiosity and to the method, namely
projective identifi cation, by which the infant seeks to satisfy it. Should the breast be felt as
fundamentally understanding, it has been transformed by the infant’s envy and hate into an
object whose devouring greed has as its aim the introjection of the infant’s projective identifi
cations in order to destroy them. This can show in the patient’s belief that the analyst

strives, by understanding the patient, to drive him insane. The result is an object which, when
installed in the patient, exercises the function of a severe and ego-destructive superego. This
description is not accurate applied to any object in the paranoid-schizoid position because it
supposes a whole-object. The threat that such a whole-object impends contributes to the
inability, described by Melanie Klein and others (Segal, 1950), of the psychotic patient to face
the depressive position and the developments attendant on it. In the paranoid-schizoid phase the
bizarre objects composed partially of elements of a persecutory superego which I described in
my paper on

‘The Differentiation of the Psychotic from the Non-psychotic Part of the Personality’ are
predominant.

Arrested development

The disturbance of the impulse of curiosity on which all learning depends, and the denial of the
mechanism by which it seeks expression, makes normal development impossible. Another
feature obtrudes if the course of the analysis is favourable; problems which in sophisticated
language are posed by the question ‘Why?’ cannot be formulated. The patient appears to have no
appreciation of causation and will complain of painful states of mind while persisting in courses
of action calculated to produce them. Therefore when the appropriate material presents itself the
patient must be shown that he has no interest in why he feels as he does. Elucidation of the
limited scope of his curiosity issues in the development of a wider range and an incipient
preoccupation with causes. This leads to some modifi cation of conduct which otherwise
prolongs his distress.

Conclusions

The main conclusions of this paper relate to that state of mind in which the patient’s psyche
contains an internal object which is 74

Attacks on linking

opposed to, and destructive of, all links whatsoever from the most primitive (which I have
suggested is a normal degree of projective identifi cation) to the most sophisticated forms of
verbal communication and the arts.

In this state of mind emotion is hated; it is felt to be too powerful to be contained by the
immature psyche, it is felt to link objects and it gives reality to objects which are not self and
therefore inimical to primary narcissism.

The internal object which in its origin was an external breast that refused to introject, harbour,
and so modify the baneful force of emotion, is felt, paradoxically, to intensify, relative to the
strength of the ego, the emotions against which it initiates the attacks.

These attacks on the linking function of emotion lead to an over-prominence in the psychotic
part of the personality of links which appear to be logical, almost mathematical, but never
emotionally reasonable. Consequently the links surviving are perverse, cruel, and sterile.

The external object which is internalized, its nature, and the effect when so established on the
methods of communication within the psyche and with the environment, are left for further
elaboration later.

75

Contribution to the psychopathology of

psychotic states

The importance of projective identifi cation in the ego

structure and the object relations of the psychotic patient 29

Herbert Rosenfeld

Following the suggestion of the organizers of the Symposium that I should discuss the
importance of projective identifi cation and ego splitting in the psychopathology of the psychotic
patient, I shall attempt to give you a survey of the processes described under the term: ‘projective
identifi cation’.

I shall fi rst defi ne the meaning of the term ‘projective identifi cation’ and quote from the work
of Melanie Klein, as it was she who developed the concept. Then I shall go on to discuss very
briefl y the work of two other writers whose use appeared to be related to, but not identical with,
Melanie Klein’s use of the term.

‘Projective identifi cation’ relates fi rst of all to a splitting process of the early ego, where either
good or bad parts of the self are split off from the ego and are as a further step projected in love
or hatred into external objects which leads to fusion and identifi cation of the projected parts of
the self with the external objects. There are important paranoid anxieties related to these
processes as the objects fi lled with aggressive parts of the self become persecuting and are
experienced by

29 This article was fi rst published in P. Doucet and C. Laurin (Eds.) (1971). Problems of
Psychosis, Volume 1 . The Hague: Excerpta Medica, pp. 115–128.

76

The psychopathology of psychotic states the patient as threatening to retaliate by forcing

themselves and the bad parts of the self which they contain back again into the ego.

In her paper on schizoid mechanisms Melanie Klein (1946) considers fi rst of all the importance
of the processes of splitting and denial and omnipotence which during the early phase of
development play a role similar to that of repression at a later stage of ego development. She
then discusses the early infantile instinctual impulses and suggests that while the ‘oral libido still
has the lead, libidinal and aggressive impulses and phantasies from other sources come to the
fore and lead to a confl uence of oral, urethral and anal desires, both libidinal and aggressive’.

After discussing the oral libidinal and aggressive impulses directed against the breast and the
mother’s body, she suggests that:

the other line of attack derives from the anal and urethral impulses and implies expelling
dangerous substances (excrements) out of the self and into the mother. Together with these
harmful excrements, expelled in hatred, split off parts of the ego are also projected into the
mother. These excrements and bad parts of the self are meant not only to injure but also to
control and to take possession of the object. In so far as the mother comes to contain the bad
parts of the self, she is not felt to be a separate individual but is felt to be the bad self. Much of
the hatred against parts of the self is now directed towards the mother. This leads to a particular
form of identifi cation which establishes the prototype of an aggressive object relation. I suggest
for these processes the term projective identifi cation .

Later on in the same paper Melanie Klein describes that not only bad, but also good parts of the
ego are expelled and projected into external objects who become identifi ed with the projected
good parts of the self. She regards this identifi cation as vital because it is essential for the
infant’s ability to develop good object relations. If this process is, however, excessive, good parts
of the personality are felt to be lost to the self which results in weakening and impoverishment of
the ego. Melanie Klein also emphasizes the aspect of the projective processes which relates to
the forceful entry into the object and the persecutory anxieties related to this process which I
mentioned before. She also describes how paranoid anxieties related to projective identifi cation
disturb introjective processes. ‘Introjection is interfered 77

Herbert Rosenfeld

with, as it may be felt as a forceful entry from the outside into the inside in retribution for violent
projections’. It will be clear that Melanie Klein gives the name ‘projective identifi cation’ both to
the processes of ego splitting and the

‘narcissistic’ object relations created by the projection of parts of the self into objects.

I shall now discuss some aspects of the work of Dr Edith Jacobson who describes psychotic
identifi cations in schizophrenic patients identical with the ones I observed and described as
‘projective identifi cation’. She also

frequently uses the term ‘projective identifi cation’

in her book Psychotic Confl ict and Reality ( Jacobson, 1967).

In 1954 ( Jacobson, 1954) Edith Jacobson discussed the identifi cations of the delusional
schizophrenic patient who may eventually consciously believe himself to be another person. She
relates this to early infantile identifi cation mechanisms of a magic nature which lead to ‘partial
or total blending of the magic self and object images, founded on phantasies or even the
temporary belief of being one with or of becoming the object, regardless of reality’. In 1967 she
describes these processes in more detail. She discusses ‘the psychotic’s regression to a
narcissistic level, where the weakness of the boundaries between self and object images gives
rise to phantasies, or experiences of fusion between these images. These primitive introjective or
projective identifi cations are based on infantile phantasies of incorporation, devouring, invading
(forcing oneself into), or being devoured by the object’. She also says ‘We can assume that such
phantasies, which pre-suppose at least the beginning distinction between self and object, are
characteristic of early narcissistic stages of development and that the child’s relation to the
mother normally begins with the introjective and projective processes’; and that the

‘introjective and projective identifi cations (of the adult patient) depend on the patient’s fi xation
to early narcissistic stages and upon the depth of the narcissistic regression’. In discussing
clinical material of the Patient A she described this fear that any affectionate physical contact
might bring about experiences of merging, which in turn might lead to a manifest psychotic state.
Her views that the introjective and projective identifi cations observed in the adult patient depend
on the fi xation to early narcissistic phases where these identifi cations originate, seem identical
with my own views and there is nothing in her clinical and theoretical observations which I have
quoted above with which I would disagree. She stresses, however, that she differs from Melanie
Klein and my own opinion in so far as 78

The psychopathology of psychotic states she does not believe that the projective identifi cations
of the adult patient observable in the transference or acted out by the patient with objects in his
environment are in fact a repetition of the early infantile projective and introjective processes,
but are to be understood as a later defensive process, as in her view early processes cannot be
observed in the transference. She also disagrees with my analytic technique of verbally
interpreting the processes of projective identifi cation when they appear in the transference,
which I regard as of central importance in working through psychotic processes in the

transference situation. 30

Margaret Mahler in 1952 described symbiotic infantile psychoses and suggested that the
mechanisms employed are introjective and projective ones and their psychotic elaboration
(Mahler, 1952). Her ideas seem to be closely related, but nevertheless quite distinct from what I
have described as projective identifi cation. She describes the early mother/infant relationship as
a phase of object relationship in which the infant behaves and functions as though he and his
mother were an omnipotent system (a dual unity with one common boundary, a symbiotic
membrane as it were). In 1967 she says, ‘the essential feature of symbiosis is hallucinatory or
delusional, somato-psychic, omnipotent fusion with the representation of the mother and, in
particular, delusion of common boundary of the two actually and physically separate individuals’
(Mahler, 1967). She suggests that

‘this is the mechanism to which the ego regresses in cases of psychotic disorganization’. In
describing the symbiotic infantile psychosis she says that the early mother-infant symbiotic
relationship is intense.

The mental representation of the mother remains or is regressively fused with that of the self. She
describes the panic reactions caused by separations ‘which are followed by restitutive
productions which serve to maintain or restore the symbiotic parasitic delusion of oneness with
the mother or father’. It is clear that Mahler has introjective or projective processes in mind as
the mechanisms which

30 When Edith Jacobson describes the defensive nature of the projective identifi cation in her
adult psychotic patients she stresses the projection of bad parts of the self into external objects in
order to avoid psychotic confusions, in other words she sees the projective identifi cation of the
adult psychotic as the attempt to split off and project into a suitable external object those parts of
the self which are unacceptable to the adult ego: the external object would then represent the
patient’s ‘bad self ’.

79

Herbert Rosenfeld

produce the symbiotic psychosis. I have, however, found no clear description of these
mechanisms in her papers.

She seems to see the symbiotic psychosis as a defence against separation anxiety which links up
closely with my description of the narcissistic object relation serving a defensive function. The
symbiotic processes described by Mahler have some resemblance to the parasitical object
relations I shall describe later. Projective identifi cation which includes ego splitting and
projecting of good and bad parts of the self into external objects is not identitical with symbiosis.
For projective identifi cation to take place some temporary differentiation of ‘me’ and ‘not me’ is
essential. Symbiosis, however, is used by Mahler to describe this state of undifferentiation, of
fusion with the mother, in which the ‘I’ is not yet differentiated from the ‘not I’.

In my own work with psychotic patients I have encountered a variety of types of object relations
and mental mechanisms which are associated with Melanie Klein’s description of projective
identifi cation. First of all, it is important to distinguish between two types of projective identifi
cation, namely, projective identifi cation used for communication with other objects and
projective identifi cation used for ridding the self of unwanted parts.

I shall fi rst discuss projective identifi cation used as a method of communication. Many
psychotic patients use projective processes for communication with other people. These
projective mechanisms of the psychotic seem to be a distortion or intensifi cation of the normal
infantile relationship, which is based on non-verbal communication between infant and mother,
in which impulses, parts of the self and anxieties too diffi cult for the infant to bear are projected
into the mother and where the mother is able instinctively to respond by containing the infant’s
anxiety and alleviating it by her behaviour.

This relationship has been stressed particularly by Bion. The psychotic patient who uses this
process in the transference may do so consciously but more often unconsciously. He then
projects impulses and parts of himself into the analyst in order that the analyst will feel and
understand his experiences and will be able to contain them so that they lose their frightening or
unbearable quality and become meaningful by the analyst being able to put them into words
through interpretations. This situation seems to be of fundamental importance for the
development of introjective processes and the development of the ego: it makes it possible for
the patient to learn to tolerate his own impulses and the analyst’s interpretations make his
infantile responses 80

The psychopathology of psychotic states and feelings accessible to the more sane self, which can
begin to think about the experiences which were previously meaningless and frightening to him.
The psychotic patient who projects predominantly for communication is obviously receptive to
the analyst’s understanding of him, so it is essential that this type of communication should be
recognized and interpreted accordingly.

As a second point I want to discuss projective identifi cation used for denial of psychic reality. In
this situation the patient splits off parts of his self in addition to impulses and anxieties and
projects them into the analyst for the purpose of evacuating and emptying out the disturbing
mental content which leads to a denial of psychic reality.

As this type of patient primarily wants the analyst to condone the evacuation processes and the
denial of his problems, he often reacts to interpretations with violent resentment, as they are
experienced as critical and frightening since the patient believes that unwanted, unbearable and
meaningless mental content is pushed back into him by the analyst.

Both the processes of communication and evacuation may exist simultaneously or alternatively
in our psychotic patients and it is essential to differentiate them clearly in order to keep contact
with the patient and make analysis possible.

As a third point I want to discuss a very common transference relationship of the psychotic
patient which is aimed at controlling the analyst’s body and mind, which seems to be based on a
very early infantile type of object relationship.

In analysis, one observes that the patient believes that he has forced himself omnipotently into
the analyst, which

leads to fusion or confusion with the analyst and anxieties relating to the loss of the self.

In this form of projective identifi cation the projection of the mad parts of the self into the analyst
often predominates. The analyst is then perceived as having become mad, which arouses extreme
anxiety as the patient is afraid that the analyst will retaliate and force the madness back into the
patient, depriving him entirely of his sanity.

At such times the patient is in danger of disintegration, but detailed interpretations of the
relationship between patient and analyst may break through this omnipotent delusional situation
and prevent a breakdown.

There is, however, a danger that the verbal communication between patient and analyst may
break down at such times as the analyst’s interpretations are misunderstood and misinterpreted
81

Herbert Rosenfeld

by the patient and the patient’s communications increasingly assume a concrete quality,
suggesting that abstract thinking has almost completely broken down. In investigating such
situations, I found that omnipotent projective identifi cation interferes with the capacity of verbal
and abstract thinking and produces a concreteness of the mental processes which leads to
confusion between reality and phantasy. It is also clinically essential for the analyst to realize
that the patient who uses excessive projective identifi cation is dominated by concrete thought
processes which cause misunderstanding of verbal interpretations, since words and their content
are experienced by the patient as concrete, non-symbolic objects. Segal, in her paper ‘Some
aspects of the analysis of a schizophrenic’ (Segal, 1950), points out that the schizophrenic patient
loses the capacity to use symbols when the symbol becomes again the equivalent of the original
object, which means it is hardly different from it. In her paper ‘Notes on symbol formation’
(Segal, 1957) she suggests the term ‘symbolic equation’ for this process: she writes: The
symbolic equation between the original object and the symbol in the internal and external world
is, I think, the basis of the schizophrenic’s concrete thinking. This non-differentiation between
the thing symbolized and the symbol is part of a disturbance in the relation between the ego and
the object. Parts of the ego and internal objects are projected onto an object and identifi ed with
it. The differentiation between the self and the object is obscured then; since a part of the ego is
confused with the object, the symbol which is a creation and a function of the ego becomes in
turn confused with the object which is symbolized.

I believe that the differentiation of the self and object representation is necessary to maintain
normal symbol formation which is based on

the introjection of objects experienced as separate from the self. 31 It

31 Dr Segal (1957) also stresses greater awareness and differentiation of the separateness
between the ego and object in normal symbol formation. She thinks that symbolization is closely
related to the development of the ego and the objects which occur in the depressive position. She
emphasizes ‘that symbols are in addition to other factors created in the internal world as a means
of restoring, recreating, recapturing and owning again the original object.

But in keeping with the increased reality sense, they are now felt as created by the ego and
therefore never completely equated with the original object.’

82

The psychopathology of psychotic states is the excessive projective identifi cation in the
psychotic process which obliterates differentiation of self and objects, which causes confusion
between reality and phantasy and a regression to concrete thinking due to the loss of the capacity
for symbolization and symbolic thinking. 32

It is, of course, extremely diffi cult to use verbal interpretations with the psychotic patient when
interpretations are misunderstood and misinterpreted. The patient may become extremely
frightened, may cover his ears and try to rush out of the consulting room and the analysis is in
danger of breaking down. At such times it is necessary to uncover the projective processes used
for the purpose of communication between patient and analyst, which will establish some
possibility of simple verbal interpretations to explain to the patient and help him to understand
the terrifying situation due to the concrete experience. It is essential for the analyst to remember
that all three types of projective identifi cation which I have described so far exist simultaneously
in the psychotic patient, and one-sided

concentration on one process may block the analysis and meaningful communication between
patient and analyst.

There is one further aspect of the psychopathology of psychotic patients that is linked with
projective identifi cation

– that is the importance of primitive aggression, particularly envy, and the use of projective
identifi cation to deal with it.

When the psychotic patient living in a state of fusion (projective identifi cation) with the analyst
begins to experience himself as a separate person, violent destructive impulses make their
appearance.

32 The loss of the capacity for abstract and symbolic thinking of the schizophrenic patient, which
leads on to very concrete modes of thinking, has been described by many writers such as
Vigotsky, Goldstein and others. Harold Searles (1962) in his paper

‘The differentiation between concrete and metaphorical thinking in the recovering schizophrenic
patient’ suggests that the concrete thought disorders depend on the fl uidity of the ego boundaries
when self and object are not clearly differentiated. In one of his cases he describes ‘abundant
evidence of massive projection, not only on to human beings around him but also on to trees,
animals, buildings and all sorts of inanimate objects’. Only when ego boundaries gradually
become fi rmly established through treatment can fi gurative or symbolic thinking develop.

Searles’ observations have a close relationship to my own observation that excessive projective
identifi cation, leading to fusion between self and object, always causes loss of the capacity for
symbolic and verbal thinking.

83

Herbert Rosenfeld

His aggressive impulses are sometimes an expression of anger related to separation anxiety, but
generally they have a distinctly envious character. As long as the patient regards the analyst’s
mind and body and his help and understanding as part of his own self he is able to attribute
everything that is experienced as valuable in the analysis as being part of his own self, in other
words he lives in a state of omnipotent narcissism. As soon as a patient begins to feel separate
from the analyst the aggressive reaction appears and particularly clearly so after a valuable
interpretation, which shows the analyst’s understanding. The patient reacts with feelings of
humiliation, complains that he is made to feel small; why should the analyst be able to remind
him of something which he needs but which he cannot provide for himself. In his envious anger
the patient tries to destroy and spoil the analyst’s interpretations by ridiculing or making them
meaningless. The analyst may have the distinct experience in his counter transference that he is
meant to feel that he is no good and has nothing of value to give to the patient.

There are often physical symptoms connected with this state because the patient may feel sick
and may actually vomit. This concrete rejection of the analyst’s help can often be clearly
understood as a rejection

of the mother’s food 33 and her care for the infant repeated in the analytic transference situation.
When the patient had previously made good progress in the treatment this ‘negative therapeutic
reaction’ is often quite violent, as if he wants to spoil and devalue everything he had previously
received, disregarding the often suicidal danger of such a reaction. Many patients experience this
violent envy directed against the good qualities of the analyst as quite insane and illogical and as
the inner saner part of the patient experiences these envious reactions as unbearable and
unacceptable, many defences against this primitive envy are created.

One of these defences relates to the splitting off and projection of the envious part of the self into
an external object, which then becomes the envious part of the patient. This kind of defensive
projective identifi cation follows the model of Melanie Klein’s

33 It is of course important to differentiate between a patient’s rejection of the analyst’s bad

handling or misunderstanding, which would repeat a bad feeding situation from the envious
aggression of the child which occurs in a good setting. The latter is not only diffi cult for the
primitive ego of the child to tolerate but creates a particularly diffi cult problem for any loving
and caring mother.

84

The psychopathology of psychotic states description of the splitting off and projection of bad
parts of the self, which I quoted in the beginning of this paper.

Another defence against envy relates to omnipotent phantasies of the patient of entering the
admired and envied object and in this way insisting that he is the object by taking over its role.
When total projective identifi cation has taken place with an envied object envy is entirely
denied, but immediately reappears when the self and object become separate again. In my paper
‘On the psychopathology of narcissism’ (Rosenfeld, 1964b) I stressed that: projective identifi
cation was part of an early narcissistic relationship to the mother, where recognition of
separateness between self and object is denied. Awareness of separation would lead to feelings
of dependence on an object and therefore to anxiety (see Mahler, 1967). In addition, dependence
stimulates envy when the goodness of the object is recognized. The omnipotent narcissistic
object relations, particularly omnipotent projective identifi cation, obviate both the aggressive
feelings caused by frustration and any awareness of envy.

I believe that in the psychotic patient projective identifi cation is more often a defence against
excessive envy, which is closely bound up with the patient’s narcissism, rather than a defence
against separation anxiety. In my paper

‘Object relations of an acute schizophrenic patient in the transference situation’ (Rosenfeld,

1964a) I tried to trace the origin of the envious projective identifi cation in schizophrenia. I
suggested: If too much resentment and envy dominates the infant’s relation to the mother,
normal projective identifi cation becomes more and more controlling and can take on omnipotent
delusional tones. For example, the infant who in phantasy enters the mother’s body driven by
envy and omnipotence, takes over the role of the mother, or breast, and deludes himself that he is
the mother or breast. This mechanism plays an important role in mania and hypomania, but in
schizophrenia it occurs in a very exaggerated form.

Finally, I want to draw attention to two similar types of object relations: a parasitical and a
delusional one. In the parasitical object relation the psychotic patient in analysis maintains a
belief that he 85

Herbert Rosenfeld

is living entirely inside an object – the analyst – and behaves like a parasite living on the
capacities of the analyst, who is expected to function as his ego. Severe parasitism may be
regarded as a state of total projective identifi cation. It is, however, not just a defensive state to
deny envy or separation but is also an expression of aggression, particularly envy. It is the
combination of defence and acting out of the aggression which makes the parasitic state a
particularly diffi cult therapeutic problem.

The parasitic patient relies entirely on the analyst, often making him responsible for his entire
life. He generally behaves in an extremely passive, silent and sluggish manner, demanding
everything and giving nothing in return.

This state can be extremely chronic and the analytic work with such patients is often minimal.
One of my depressed patients described himself as a baby, which was like a stone heavily
pressing into my couch and into me. He felt he was making it impossible for me either to carry
him or to look after him and he feared that the only thing that I could possibly do was to expel
him, if I could not stand him any longer. However, he was terrifi ed that he could not survive
being left. He not only felt that he had a very paralysing effect on the analysis but that he was
paralysed and inert himself. Only very occasionally was it possible to get in touch with the
intense feelings either of hostility or overwhelming pain and depression bound up with this
process. There was no joy when the analyst was felt to be helpful and alive, as it only increased
the patient’s awareness of the contrast between himself and the analyst and at times produced a
desire to frustrate him, and with this he returned to the status quo of inertia, which was felt to be
unpleasant but preferred to any of the intense feelings of pain, anger, envy or jealousy which
might fl eetingly be experienced. As I suggested before, extreme parasitism is partly a defence
against separation anxiety, envy or jealousy, but it often seems to be a defence against any
emotion which might be experienced as painful. I often have the impression that patients, like the
one I described, who experience themselves as dead and are often experienced by the analyst as
so inactive that they might as well be dead, use their analyst’s aliveness as a means of survival.

However, the latent hostility prevents the patient from getting more than minimal help or
satisfaction from the

analysis. In the more active forms of parasitism the insidious hostility dominates the picture and
is much more apparent.

Dr Bion in his book Transformations (Bion, 1965) describes a more active case of parasitism. He
emphasizes that such patients are partic-86

The psychopathology of psychotic states ularly unrewarding. The essential feature is

simultaneous stimulation and frustration of hope and work that is fruitless, except for
discrediting analyst and patient. The destructive activity is balanced by enough success to deny
the patient fulfi lment of his destructiveness.

‘The helpful summary of such a case is described as ‘chronic murder of patient and analyst’ or
‘an instance of parasitism’: the patient draws on the love, or benevolence of the host to extract
knowledge and power which enables him to poison the association and destroy the indulgence on
which he depends for his existence.

It is important to differentiate the very chronic forms of parasitism from the massive intrusion
and projective identifi cation into the analyst which resembles parasitism but is of shorter
duration and responds more easily to interpretations. It occurs at times when separation threatens
or when jealousy or envy is violently stimulated in the transference or in outside life. Meltzer
(1967) describes a primitive form of possessive jealousy which plays an important role in
perpetuating massive projective identifi cation of a peculiar withdrawn, sleepy sort.

The other form of living entirely inside an object occurs in severely deluded schizophrenic
patients who seem to experience themselves as living in an unreal world, which is highly
delusional but nevertheless has qualities of a structure which suggests that this hallucinatory
world represents the inside of an object, probably the mother. The patient may be withdrawn,
preoccupied with hallucinations, in the analysis occasionally projecting the hallucinatory
experience on to the analyst, which leads to mis-identifying him and others with his delusional
experience.
Sometimes the patient may describe himself as living in a world, or object, which separates him
entirely from the outside world and the analyst is experienced as a contraption, an actor or a
machine and the world becomes extremely unreal. The living inside the delusional object seems
to be defi nitely in opposition to relating to the outside world, which would imply depending on a
real object. This delusional world or object seems to be dominated by an omnipotent and
sometimes omniscient part of the self, which creates the notion that within the delusional object
there is complete painlessness and freedom to indulge in any whim. It also appears that the self
within the delusional object exerts a powerful suggestive and seductive infl uence on saner parts
of the personality in order to persuade or force them to withdraw from reality and to join the
delusional omnipotent world. Clinically, the patient may hear a voice making propaganda for
living inside the mad world by 87

Herbert Rosenfeld

idealizing it and praising its virtue by offering a complete satisfaction and instant cure to the
patient. This persuasion or propaganda to get inside the delusional world implies clinically the
constant stimulus to all parts of the self to use omnipotent projective identifi cation (forcing the
self inside the object) as the only possible method to solve all problems. This situation leads to
constant acting out with external objects which are used for projective identifi cation. When,
however, projective identifi cation becomes directed towards the delusional object, the saner
parts of the self may become trapped or imprisoned within this object and physical and mental
paralysis amounting to catatonia may result.

The psychoanalytic treatment of the processes related to projective identifi cation in the
psychotic patient As this paper deals primarily with the psychopathology of psychotic states, I
can only briefl y discuss my psychoanalytic technique in dealing with psychotic patients to
emphasize my contention that the investigation of the psychopathology of the psychotic and the
therapeutic approach are closely interlinked.

In treating psychotic states it is absolutely essential to differentiate those parts of the self which
exist almost exclusively in a state of projective identifi cation with external objects, or internal
ones such as the delusional object I described above and the saner parts of the patient which are
less dominated by projective identifi cation and have formed some separate existence from
objects. These saner parts may be remnants of the adult personality, but often they represent
more normal non-omnipotent infantile parts of the self, which during analysis are attempting to
form

a dependent relationship to the analyst representing the feeding mother. As the saner parts of the
self are in danger of submitting to the persuasion of the delusional self to withdraw into the more
psychotic parts of the personality, and to get entangled in it, the former need very careful
attention in analysis to help them to differentiate the analyst as an external object from the
seductive voice of the omnipotent parts of the self related to the internal delusional object, which
can assume any identity for the purpose of keeping up the domination of the whole self. As there
is always a confl ict, amounting sometimes to a violent struggle, between the psychotic and saner
parts of the personality, the nature of this 88

The psychopathology of psychotic states confl ict has also to be clearly understood in order to
make it possible to work through the psychotic state by means of analysis. For example, the
structure and the intentions of the psychotic parts of the patient, which are highly narcissistically
organized, have to be brought fully into the open by means of interpretations, as they are
opposed to any part of the self which wants to form a relationship to reality and to the analyst
who attempts to help the ego to move towards growth and development. The interpretations have
also to expose the extent and the method used by the psychotic narcissistic parts of the
personality in attempting to dominate, entangle and to paralyse the saner parts of the self. It is
important to remember that it is only the sane dependent parts of the self separate from the
analyst that can use introjective processes uncontaminated by the concreteness caused by the
omnipotent projective identifi cations; the capacity for memory and growth of the ego depends
on these normal introjective processes. When the dependent non-psychotic parts of the
personality become stronger, as the result of analysis, violent negative therapeutic reactions
usually occur as the psychotic narcissistic parts of the patient oppose any progress and change of
the

status quo , a

problem which I recently discussed in detail in a paper on ‘The Negative Therapeutic Reaction’
(Rosenfeld, 1969).

Case presentation

I shall now bring some case material of a schizophrenic patient in order to illustrate some aspects
of projective identifi cation and ego splitting.

Patient A

Had been diagnosed several years ago as schizophrenic, when he had an acute psychotic
breakdown which was characterized by overwhelming panic, confusion and fears of complete
disintegration. He did not hallucinate during the acute phase, nor are the delusional aspects of the
psychosis dominant at the present time, but he is unable to work or to maintain a close
relationship with men or women in the outside world. He had been treated by another analyst for
several years before starting analysis with me more than a year ago. The previous 89

Herbert Rosenfeld

analyst in his report to me emphasized the patient’s tendency to slip into a state of projective
identifi cation with the analyst at the beginning of each session leading to the patient’s becoming
confused and unable to speak in an audible and understandable way. The analyst interpreted to
the patient that he expected him, the analyst, to understand him even if he could not talk or think,
since he believed himself to be inside the analyst; as a result of such interpretations he generally
started to speak more distinctly. During the analysis with me there were further improvements
and he felt at times more separate, so that the saner parts of his self were able to form to some
extent a dependent relationship to me. However, from time to time, particularly after he had
made some progress, or when there were long separations, he fell back to a parasitical
relationship of living inside me (projective identifi cation), which led to states of confusion,
inability to think and talk, claustrophobia and paranoid anxieties of being trapped by me. When
envy was aroused through experiences in the real world, for example when he met a man who
was successful in his relationship with women or in his work, after a short conscious experience
of envy A would frequently become identifi ed with him. This was followed by severe anxieties
of losing his identity and feelings of being trapped, rather than leading to the delusion that he
was the envied man or that he was able to function in the outside world similarly to the man with
whom projective identifi cation and confusion had taken place.

Last year, in the autumn, I had to interrupt the patient’s analysis for a fortnight which disturbed
him considerably.

Consciously, he seemed unconcerned about my going away which I had of course discussed with
him several months before. However, two weeks before the interruption he became acutely
anxious and confused and for a day he feared that he would have another breakdown and have to
go into hospital. The disturbance started with the patient’s complaint that he could not drag
himself away from the television screen where he was watching the Olympic games. He felt
forced, almost against his will to look at it until late at night. He complained that he was drawn
into the hot climate of Mexico which made him feel that being there would make him well. He
was also compelled to look at the athletes, or wrestlers and weightlifters and felt he was, or ought
to be, one of them. He asked me questions: Why have I to be an athlete? Why can’t I be myself?
He felt that this looking at television was like an addiction which he could not stop and which
exhausted and drained him. At times he felt so strongly ‘pulled inside 90

The psychopathology of psychotic states the television’ that he felt claustrophobic and had diffi
culty in breathing.

Afterwards during the night he felt compelled to get up and see whether the taps of the
washbasin in his fl at were closed and whether the stoppers in the basin were blocking up the
drainage. He was terrifi ed that both his bath and the basin might overfl ow and eventually he
confessed that he was afraid of being drowned and suffocated. I interpreted to him that after he
felt that he was making progress and feeling separate from me he was suddenly overcome with
impatience and envy of me and other men who were able to move about and were active. I
suggested that it was the envious part which drove him into the identifi cation with other men
and myself in order to take over their strength and potency, and in this way the omnipotent part
of himself could make him believe that he could be mature and healthy instantly. He agreed with
the interpretation without any diffi culty and started to speak very fast: he said he knew all this
and was quite aware of it, but he also knew that this belief was quite false and that it was a
delusion and he was angry at having to listen to a voice in him which was very persuasive and
stimulated him to take over the mind and body of other people. I also interpreted to him that I
thought that the threatening separation was stimulating his wish to be suddenly grown up and
independent in order not to have to cope with the anxieties of being separate from me. He then
told me that he was falling every night into a very deep sleep from which he could not easily
awake in the morning and so he had arrived late for his session. He compared the feeling of
being pulled into the television screen, which seemed to have become identifi ed with the
delusional object, to being pulled into this deep sleep. He now spoke fairly fl uently and more
distinctly and conveyed that he felt now more separate from me. He said he felt disgusted with
himself for being a parasite and he also complained that the television experience and his bed
were draining his life out of him, so that he had a strong impulse to smash both; he was glad that
he had been able to control this in reality. I acknowledged his own observation that his looking at
television and being pulled into a deep sleep were experienced by him as parasitical experiences
where he felt he was getting into other objects. I pointed out that he felt angry with that part of
himself which stimulated him to get inside external objects, the athletes representing me as a
successful man who was travelling abroad during the break, and also into internal objects which
were represented by his bed. I stressed that at fi rst he felt he probably could 91

Herbert Rosenfeld

control and possess these objects entirely when he got inside them, but very soon he felt enclosed
and trapped and persecuted, which roused his wish to destroy the bed and the television screen
which had turned into persecuting objects. I thought that his fear of being trapped and his anger
related also to the analysis and the analyst. The patient’s obsessions about the stoppers of the
basin were also related to his fear of being trapped and drowned. It seemed that he had constantly
to fi nd out whether after his intrusion into objects he was trapped and was in danger of drowning
and suffocating inside, or whether there was a hole through which he could escape.

Simultaneously with the projective identifi cation related to the delusional television experience,
the patient was violently pulled into relations to prostitutes. He explained to me that there was a
part of him which persuaded him whenever he felt lonely or anxious that he needed to have a
lovely big prostitute for nourishment and this would make him well. During the session he
assured me that he realized the falsity of the voice, but in fact he very rarely could resist. He felt
he wanted to get inside the prostitutes in an excited way in order to devour them, but after
intercourse he felt sick and disgusted and convinced that he had now acquired syphilis of the
stomach. The patient, during this session, many times asserted that he knew quite well the
difference between reality and the delusional persuasion and he also knew what was wrong. But
it was clear to me that in spite of this knowledge he was again and again put temporarily into a
deluded state by a psychotic omnipotent and omniscient part of him which succeeded in seducing
and overpowering the saner part of his personality and induced him to deal with all his diffi

culties and problems, including his envy, by projective identifi cation. During the session, the
saner part of the patient seemed to receive help and support from the analyst’s interpretations,
but he felt humiliated and angry that he could not resist the domination and persuasion of the
psychotic part when he was left on his own. In attempting to examine the reason for listening so
readily to the internal voice, I found that he was promised cure, freedom from anxiety and from
dependence on myself. I was then able to interpret that the separation made him more aware of
feeling small and dependent on me, which was humiliating and painful and increased his envy of
me. By omnipotently intruding into me, he could delude himself that from one moment to the
next he became grown up and completely all right and could manage without me.

92

The psychopathology of psychotic states I shall now briefl y describe the relationship between
ego splitting, projective identifi cation and the persecutory anxieties related to these processes in
this patient. On the following session he reported that he felt much better, but in the middle of
the session he became very silent and then admitted with shame that he had been intensely anti-
semitic some time ago for a period of over six months. He had regarded the Jews as degraded
people who were only out to exploit others in order to extract money from them in a ruthless
way. He hated exploiters and wanted to attack and smash them for it. I interpreted that while he
was aware that this happened in the past, he now felt awful towards me because after yesterday’s
session he had got rid of the greedy parasitical exploiting part of his self but had pushed it into
me. He felt now that I had become his greedy exploiting self and this made him feel intensely
suspicious about me.

He replied that he feared that I must now hate and despise him, and that the only thing which he
could do was to destroy himself or this hated part of himself. I interpreted his fear of my
retaliation because when he saw me as a greedy, exploiting Jew he attacked and despised me,
and feared that I would hate him because he believed I could not bear that he had pushed his own
greedy self into me, not only as an attack but because he could not bear it himself and wanted to
get rid of it. I suggested that it was when he felt that I could not accept his bad and hated self that
he attacked himself so violently. In fact, the greatest anxiety during this session was related to
violent attacks that were directed against his bad self which built up to a crescendo, so that he
feared he would tear himself to pieces. He calmed down considerably after the interpretations.

The next session showed progress in relation to the splitting processes, followed in subsequent
sessions by some experience of depression. In the beginning of the session the patient reported
that he had some diffi culty in getting up, but he was glad that he remembered a dream. In this
dream he was observing a group of Olympic runners in a race on the television screen. Suddenly
he saw a number of people crowding in on to the track and interfering with the race.

He got violently angry with them and wanted to kill them for interfering and deliberately getting
in the way of the runners. He reported that he had been looking at the television screen for only a
short time the night before and had been thinking about the last session in which he had been
afraid of damaging himself when he tried to cut off and destroy bad parts of himself. He now
was determined to face 93

Herbert Rosenfeld

up to whatever was going on in him. He had no associations to the dream, apart from the fact that
the interfering people looked quite ordinary. I pointed out that in this dream he showed in a very
concrete way what he felt he was doing when he was looking at television. The interfering
people seemed to be the parts of himself which he experienced as worming their way into the
track in Mexico when he was greedily and enviously looking at television. In this dream it was
quite clear that people representing him were not competing by running, but were simply trying
to interfere with the progress of the race. I was then able to show him another aspect of the
extremely concrete form of projection which did not only relate to the Olympic runners but to the
analyst. I interpreted that he felt when the analysis was making good progress he experienced my
interpretations and thoughts as something which he was watching with admiration and envy, like
the athletes on television. He felt that the envious parts of himself actually could worm their way
into my brain and interfere with the quickness of my thinking. In the dream he was attempting to
face up to the recognition that these parts of himself actually existed and he wanted to control
and stop them. I also related this process to the patient’s complaints that his own thought
processes were often interfered with and I related this to an identifi cation with the analyst’s
mind which he often enviously attacked.
Actually, the patient’s co-operation during the last week had been very positive, which had led to
considerable unblocking of his mind, so that a great number of his projective identifi cations and
splitting processes had shown themselves clearly in the analysis and could be related to the
transference situation. In the dream he had actually succeeded in what he announced he tried to
do, namely, to face up to the processes by bringing them into the transference rather than
attempting to destroy and get rid of them by splitting and projection. This also enabled him to
face up to his acute fear of damaging both his objects and his self through his projective identifi
cations.

My interpretations seemed to diminish his anxiety about having completely destroyed me and
my brain so that I could be experienced as helpful and undamaged, and for certain periods I was
introjected as good and undamaged, a process leading gradually to a strengthening of the ego.
One of the diffi culties of working through such situations in the analysis is the tendency to
endless repetition, in spite of the patient’s understanding that very useful analytic work is being
done. It is important in dealing with patients and processes of 94

The psychopathology of psychotic states this kind to accept that much of the repetition is
inevitable. The acceptance by the analyst of the patient’s processes being re-enacted in the
transference helps the patient to feel that the self, which is constantly split off and projected into
the analyst, is acceptable and not so damaging as feared.

I want now to describe briefl y a short depressive spell in the patient’s illness which throws some
light on his internal anxieties related to damage to objects and his self. A few days after the
session I reported before, the patient became increasingly concerned about injuries he believed
he had done to other people, but most of all he was horrifi ed about what was going on inside
himself. For half an hour he experienced intense anxiety and reported that he was too frightened
to look inside himself. Suddenly he saw his brain in a terrible state as if many worms had eaten
their way into it. He feared that the damage was irreparable and his brain might fall to pieces.

Despairingly he said how could he allow his brain to get into such an awful state! After a pause
he suggested that his constant relations to prostitutes had something to do with the state of
affairs. I interpreted that he felt that he had forced himself during the last weeks into people such
as the prostitutes and the athletes and that he was afraid to see that damage outside. The damage
to his brain seemed identical to the damage he feared he had done to external objects. He then
began to talk about his brain as a particularly valuable and delicate part of his body which he had
neglected and left unprotected. His voice sounded now much warmer and more concerned than
ever previously, so I felt it necessary to interpret that his brain was also identifi ed with a
particularly valuable important object relationship, namely, the analysis and the analyst which
represented the feeding situation to him. This he had usually displaced on to the prostitutes to
whom he always went for nourishment. I gave him now detailed interpretations of the intensity
of his hunger for me, his inability to wait and I described his impulses and the self which he had
experienced as boring himself omnipotently into my brain, which contained for him all the
valuable knowledge which he longed to possess.

Throughout the hour the patient felt great anxiety and almost unbearable pain because he feared
he could not repair the damage. However, he was clearly relieved through the transference
interpretations which helped him to differentiate and disentangle the confusion between inside
and outside, phantasy and reality. I think it was particularly the interpretations about my brain,
which showed him that I could still 95

Herbert Rosenfeld

think and function, which both helped him to understand this very concrete phantasy in relation
to his own thought processes and to relieve his anxiety about the damage he feared he had done
to me.

In this case material I have tried to illustrate some of the processes of projective identifi cation
and ego splitting and the part they play in the psychopathology of psychotic patients.

Summary

‘Projective identifi cation’ relates fi rst of all to a splitting process of the early ego, where either
good or bad parts of the self are split off from the ego and are as a further step projected in love
or hatred into external objects, which leads to fusion and identifi cation of the projected parts of
the self with the external objects. There are important paranoid anxieties related to these
processes as the objects fi lled with aggressive parts of the self become persecuting and are
experienced by the patient as threatening to retaliate by forcing themselves and the bad parts of

the self which they contain back again into the ego.

In this paper I have discussed a number of processes related to projective identifi cation which
play an important part in psychotic patients. First of all, I am distinguishing between two types of
pro -

jective identifi cation: the projective identifi cation used by psychotic patients for communication
with other objects, which seems to be a distortion or intensifi cation of the normal infantile
relationship which is based on non-verbal communication between infant and mother; and
secondly, the projective identifi cation used for ridding the self of unwanted parts, which leads to
a denial of psychic reality. As a third point I am discussing projective identifi cation representing
a very common transference relationship of the psychotic patient which is aimed at controlling
the analyst’s body and mind, which seems to be based on a very early infantile type of object
relationship.

My fourth point is projective identifi cation used by the psychotic patient predominantly for
defensive purposes to deal with aggressive impulses, particularly envy. The fi fth point I am
drawing attention to are those object relations of the psychotic patient in analysis where he
maintains the belief that he is living entirely inside an object – the analyst – and behaves like a
parasite using the capacities of the analyst, who is expected to function as his ego.

Severe parasitism may be 96

The psychopathology of psychotic states regarded as a state of total projective identifi cation. I
am also discussing the parasitical state which is related to living entirely in a delusional world.
Sixthly, I am discussing the psychoanalytic treatment of the
processes related to projective identifi cation in the psychotic patient.

Finally, I present case material of a schizophrenic patient in order to illustrate some aspects of
projective identifi cation and ego splitting.

97

Projective identification

Some clinical aspects 34

Betty Joseph

The concept of projective identifi cation was introduced into analytic thinking by Melanie Klein
in 1946. Since then it has been welcomed, argued about, the name disputed, the links with
projection pointed out, and so on; but one aspect seems to stand out above the fi ring line, and
that is its considerable clinical value. It is this aspect that I shall mainly concentrate on today, and
mainly in relation to the more neurotic patient.

Melanie Klein became aware of projective identifi cation when exploring what she called the
paranoid-schizoid position, that is, a constellation of a particular type of object relations,
anxieties, and defences against them, typical for the earliest period of the individual’s life and, in
certain disturbed people, continuing throughout life.

This particular position she saw as dominated by the infant’s need to ward off anxieties and
impulses by splitting both the object, originally the mother, and the self and projecting these
split-off parts into an object, which will then be felt to be like, or identifi ed with, these split-off
parts, so colouring the infant’s perception of the object and its subsequent introjection.

She discussed the manifold aims of different types of projective identifi cation, for example,
splitting off and getting rid of unwanted

34 This article was published in J. Sandler (Ed.) (1987). Projection, Identifi cation, Projective
Identifi cation .

London: Karnac Books, pp. 65–76. Published with the permission of International Universities
Press Inc., Madison,

CT, USA.

98

Projective identifi cation: Some clinical aspects parts of the self that cause anxiety or pain;
projecting the self or parts of the self into an object to dominate and control it and thus avoid any
feelings of being separate; getting into an object to take over its capacities and make them its
own; invading in order to damage or destroy the object. Thus the infant, or adult who goes on
using such mechanisms powerfully, can avoid any awareness of separateness, dependence,
admiration, or its concomitant sense of loss, anger, envy, etc. But it sets up anxieties of a
persecutory type, claustrophobic, panics and the like.

We could say that, from the point of view of the individual who uses such mechanisms strongly,
projective identifi cation is a phantasy and yet it can have a powerful effect on the recipient. It
does not always do so and when it does we cannot always tell how the effect is brought about,
but we cannot doubt its importance. We can see, however, that the concept of projective identifi
cation, used in this way, is more object-related, more concrete and covers more aspects than the
term projection would ordinarily imply, and it has opened up a whole area of analytic
understanding. These various aspects I am going to discuss later, as we see them operating in our
clinical work; here I want only to stress two points: fi rst, the omnipotent power of these
mechanisms and phantasies; second, how, in so far as they originate in a particular constellation,
deeply interlocked, we cannot in our thinking isolate projective identifi cation from the
omnipotence, the splitting and the resultant anxieties that go along with it. Indeed, we shall see
that they are all part of a balance, rigidly or precariously maintained by the individual, in his own
individual way.

As the individual develops, either in normal development or through analytic treatment, these
projections lessen, he becomes more able to tolerate his ambivalence, his love and hate and
dependence on objects, in other words, he moves towards what Melanie Klein described as the
depressive position. This process can be helped in infancy if the child has a supportive
environment, if the mother is able to tolerate and contain the child’s projections, intuitively to
understand and stand its feelings. Bion elaborated and extended this aspect of Melanie Klein’s
work, suggesting the importance of the mother being able to be used as a container by the infant,
and linking this with the process of communication in childhood and with the positive use of the
counter-transference in analysis. Once the child is better integrated and able to recognize its
impulses and feelings as its own, there will be a lessening in the pressure to project, accompanied
99

Betty Joseph

by an increased concern for the object. In its earliest forms projective identifi cation has no
concern for the object, indeed it is often anti-concern, aimed at dominating, irrespective of the
cost to the object.

As the child moves towards the depressive position, this necessarily alters and, although
projective identifi cation is probably never entirely given up, it will no longer involve the
complete splitting off and disowning of parts of the self, but will be less absolute, more
temporary and more able to be drawn back into the individual’s personality

– and thus be the basis of empathy.

In this paper I want, fi rst, to consider some further implications of the use of projective identifi
cation, and then to discuss and illustrate different aspects of projective identifi cation, fi rst in
two patients more or less stuck in the paranoid-schizoid position, and then in a patient beginning
to move towards the depressive position.
To begin with: some of the implications, clinical and technical, of the massive use of projective
identifi cation as we see it in our work.

Sometimes it is used so massively that we get the impression that the patient is; in phantasy,
projecting his whole self into his object and may feel trapped or claustrophobic. It is, in any case,
a very powerful and effective way of ridding the individual of contact with his own mind; at
times the mind can be so weakened or so fragmented by splitting processes or so evacuated by
projective identifi cation that the individual appears empty or quasi-psychotic.

This I shall show with C, the case of a child. It also has important technical implications; for
example, bearing in

mind that projective identifi cation is only one aspect of an omnipotent balance established by
each individual in his own way, an interpretative attempt on the part of the analyst to locate and
give back to the patient missing parts of the self must of necessity be resisted by the total
personality, since it is felt to threaten the whole balance and lead to more disturbance. I shall
discuss this in case T. Projective identifi cation cannot be seen in isolation.

A further clinical implication that I should like to touch on is about communication. Bion
demonstrated how projective identifi -

cation can be used as a method of communication by the individual putting, as it were,

undigested parts of his experience and inner world into the object, originally the mother, now the
analyst, as a way of getting them understood and returned in a more manageable form.

But we might add to this that projective identifi cation is, by its very nature, a kind of
communication, even in cases where this is not its aim or its intention. By defi nition projective
identifi cation means the 100

Projective identifi cation: Some clinical aspects putting of parts of the self into an object. If the
analyst on the receiving end is really open to what is going on and able to be aware of what he is
experiencing, this can be a powerful method of gaining understanding. Indeed, much of our
current appreciation of the richness of the notion of counter-transference stems from it. I shall
later try to indicate some of the problems this raises, in terms of acting-in, in my discussion of
the third case, N.

I want now to give a brief example of a case to illustrate the concreteness of projective identifi
cation in the analytic situation, its effectiveness as a method of ridding the child of a whole area
of experience and thus keeping some kind of balance, and the effect of such massive projective
mechanisms on her state of mind. This is a little girl aged 4, in analytic treatment with Mrs
Elizabeth Da Rocha Barros, who was discussing the case with me. The child had only very
recently begun treatment, a deeply disturbed and neglected child, whom I shall call C.

A few minutes before the end of a Friday session C said that she was going to make a candle; the
analyst explained her wish to take a warm Mrs Barros with her that day at the end of the session
and her fear that there would not be enough time, as there were only three minutes left. C started
to scream, saying that she would have some spare candles; she then started to stare through the
window with a vacant, lost expression. The analyst interpreted that the child needed to make the
analyst realize how awful it was to end the session, as well as expressing a wish to take home
some warmth from the analyst’s words for the weekend. The child screamed: ‘Bastard! Take off
your clothes and jump outside.’ Again the analyst tried to interpret C’s feelings about being
dropped and sent into the cold. C

replied: ‘Stop your talking, take off your clothes. You are cold. I’m not cold.’ The feeling in the
session was extremely moving. Here the words carry the concrete meaning, to the child, of the
separation of the weekend

– the awful coldness. This she tries to force into the analyst and it is felt to have been concretely
achieved. ‘You are cold, I am not cold.’

The moments when C looked completely lost and vacant, as in this fragment, were very frequent
and were, I think, indicative not only of her serious loss of contact with reality, but of the
emptiness, vacant-ness of her mind and personality when projective identifi cation was operating
so powerfully. I think that much of her screaming is also in the nature of her emptying out. The
effectiveness of such emptying is striking, as the whole experience of loss and its concomitant
emotions 101

Betty Joseph

is cut out. One can again see here how the term ‘projective identifi -

cation’ describes more vividly and fully the processes involved than the more general and
frequently used terms, such as ‘reversal’ or, as I said, ‘projection’.

In this example, then, the child’s balance is primarily maintained by the projecting out of parts of
the self. I want now to give an example of a familiar kind of case to discuss various kinds of
projective identifi cation working together to hold a particular narcissistic omnipotent balance.
This kind of balance is very fi rmly structured, extremely diffi cult to infl uence analytically and
leads to striking persecutory anxieties. It also raises some points

about different identifi catory processes and problems about the term ‘projective identifi cation’
itself.

A young teacher, whom I shall call T, came into analysis with diffi culties in relationships, but
actually with the hope of changing careers and becoming an analyst. His daily material consisted
very largely of descriptions of work he had done in helping his pupils, how his colleagues had
praised his work, asked him to discuss their work with him, and so on. Little else came into the
sessions. He frequently described how one or other of his colleagues felt threatened by him,
threatened in the sense of feeling minimized or put in an inferior position by his greater insight
and understanding. He was, therefore, uneasy that they felt unfriendly to him at any given
moment. (Any idea that his personality might actually put people off did not enter his mind.) It
was not diffi cult to show him certain ideas about myself – for example, that when I did not seem
to be encouraging him to give up his career and apply for training as an analyst, he felt that I,
being old, felt threatened by this intelligent young person coming forward, and, therefore, would
not want him in my professional area.
Clearly, simply to suggest, or interpret, that T was projecting his envy into his objects and then
feeling them as identifi ed with this part of himself might be theoretically accurate, but clinically
inept and useless; indeed it would just be absorbed into his psychoanalytic armoury. We can see
that the projective identifi cation of the envious parts of the self was, as it were, only the end
result of one aspect of a highly complex balance which he was keeping. To clarify something of
the nature of this balance, it is important to see how T was relating to me in the transference.

Usually he spoke of me as a very fi ne analyst and I was fl attered in such ways. Actually he
could not take in interpretations meaningfully, he appeared not to listen properly; he 102

Projective identifi cation: Some clinical aspects would, for example, hear the words partially and
then re-interpret them unconsciously, according to some previous theoretical and
psychoanalytical knowledge, then give them to himself with this slightly altered and generalized
meaning. Frequently, when I interpreted more fi rmly, he would respond very quickly and
argumenta-tively, as if there were a minor explosion which seemed destined, not only to expel
from his mind what I might be going to say, but enter my mind and break up my thinking at that
moment.

In this example we have projective identifi cation operating with various different motives and
leading to different identifi catory processes – but all aimed at maintaining his narcissistic
omnipotent balance. First we see the splitting of his objects – I am fl attered and kept in his mind
as idealized; at such moments the bad or unhelpful aspect of myself is quite split off, even
though I don’t seem to be achieving much with him; but this latter has to be denied. He projects
part of himself into my mind and takes over; he ‘knows’ what I am going to say and says it
himself. At this point, a part of the self is identifi ed with an idealized aspect of myself, which is
talking to, interpreting to, an idealized patient part of himself; idealized because it listens to the
analyst part of him. We can see what this movement achieves in terms of his balance. It cuts out
any real relationship between the patient and myself, between analyst and patient, as mother and
child, as a feeding couple. It obviates any separate existence, any relating to me as myself; any
relationship in which he takes in directly from me. T was, in fact, earlier in his life slightly
anorexic.

If I manage for a moment to get through this T explodes, so that his mental digestive system is
fragmented, and by this verbal explosion, as I said, T unconsciously tries to enter my mind and
break up my thinking, my capacity to feed him. It is important here, as always with projective
identifi cation, to distinguish this kind of unconscious entering, invading and breaking up from a
conscious aggressive attack. What I am discussing here is how these patients, using projective
identifi cation so omnipotently, actually avoid any such feelings as dependence, envy, jealousy,
etc.

Once T has in phantasy entered my mind and taken over my interpretations, and my role at that
moment, I notice that he has

‘added to’, ‘improved on’, ‘enriched’ my interpretations, and I become the onlooker, who should
realize that my interpretations of a few moments ago were not as rich as his are now – and surely
I should feel threatened by this young man in my room! Thus the two 103
Betty Joseph

types of projective identifi cation are working in harmony, the invading of my mind and taking
over its contents and the projecting of the potentially dependent, threatened and envious part of
the self into me. This is, of course,

mirrored in what we hear is going on in his outside world – the fellow students who ask for help
and feel threatened by his brilliance – but then he feels persecuted by their potential
unfriendliness. So long as the balance holds so effectively, we cannot see what more subtle,
sensitive, and important aspects of the personality are being kept split off, or why – we can see
that any relationship to a truly separate object is obviated – with all that this may imply.

A great diffi culty is, of course, that all insight tends to get drawn into this process; to give a
minute example: one Monday, T really seemed to become aware of exactly how he was subtly
taking the meaning out of what I was saying and not letting real understanding develop. For a
moment he felt relief and then a brief, deep feeling of hatred to me emerged into consciousness.
A second later he added quietly that he was thinking how the way that he had been feeling just
then towards me, that is, the hatred, must have been how his fellow students had felt towards him
on the previous day when he had been talking and explaining things to them! So, immediately
that T has a real experience of hating me because I have said something useful, he uses the
momentary awareness to speak about the students, and distances himself from the emerging envy
and hostility, and the direct receptive contact between the two of us is again lost. What looks like
insight is no longer insight but has become a complex projective manoeuvre.

At a period when these problems were very much in the forefront of the analysis, T brought a
dream, right at the end of a session. The dream was simply this: T was with the analyst or with a
woman, J, or it might have been both, he was excitedly pushing his hand up her knickers into her
vagina, thinking that if he could get right in there would be no stopping him. Here, I think under
the pressure of the analytic work going on, T’s great need and great excitement were to get
totally inside the object, with all its implications, including, of course, the annihilation of the
analytic situation.

To return to the concept of projective identifi cation; with this patient I have indicated three or
four different aspects: attacking the analyst’s mind; a kind of total invading, as in the dream
fragment I have just quoted; a more partial invading and taking over aspects or 104

Projective identifi cation: Some clinical aspects capacities of the analyst; and fi nally putting
parts of the self, particularly inferior parts, into the analyst. The latter two are mutually
dependent, but lead to different types of identifi cation. In the one, the patient, in taking over,
becomes identifi ed with the analyst’s idealized capacities; in the other, it is the analyst who
becomes identifi ed with the lost, projected, here inferior or envious parts of the patient. I think it
is partly because the term is broad and covers many aspects that there has been some unease
about the name itself.

I have so far discussed projective identifi cation in two cases caught up in the paranoid-schizoid
position, a borderline child and a man in a rigid omnipotent narcissistic state. Now I want to
discuss aspects of projective identifi cation as one sees it in a patient moving towards the
depressive position. I shall illustrate some points from the case of a man as he was becoming less
rigid, more integrated, better able to tolerate what was previously projected, but constantly also
pulling back, returning to the use of the earlier projective mechanisms; then I want to show the
effect of this on subsequent identifi cations and the light that it throws on previous identifi
cations. I also want to attempt to forge a link between the nature of the patient’s residual use of
projective identifi cation and its early infantile counterpart and the relation of this to phobia
formation. I bring this material also to discuss briefl y the communicative nature of projective
identifi cation.

To start with this latter point, as I said earlier, since projective identifi cation, by its very nature
means the putting of parts of the self into the object, in the transference we are of necessity on
the receiving end of the projections and, therefore, providing we can tune into them, we have an
opportunity par excellence to understand them and what is going on. In this sense, it acts as a
communication, whatever its motivation, and is the basis for the positive use of
countertransference. As I want to describe with this patient, N, it is frequently diffi cult to clarify
whether, at any given moment, projective identifi cation is primarily aimed at communicating a
state of mind that cannot be verbalized by the patient or whether it is aimed more at entering and
controlling or attacking the analyst, or whether all these elements are active and need
consideration.

A patient, N, who had been in analysis many years, had recently married and, after a few weeks,
was becoming anxious about his sexual interest and his potency, particularly in view of the fact
that his wife was considerably younger. He came on a Monday, saying that he felt that ‘the
thing’ was never really going to get right, ‘the sexual thing’, 105

Betty Joseph

yes, they did have sex on Sunday, but somehow he had to force himself and he knew it wasn’t
quite all right, and his wife noticed this and commented. It was an all-right kind of weekend, just
about. He spoke about this a bit more and explained that they went to a place outside London, to
a party; they had meant to stay the night in an hotel nearby, but couldn’t fi nd anywhere nice
enough and came home and so were late. What was being conveyed to me was a quiet, sad
discomfort, leading to despair, and I pointed out to N how he was conveying an awful long-term
hopelessness and despair, with no hope for the future.

He replied to the effect that he supposed that he was feeling left out, and linked this with what
had been a rather helpful and vivid session on the Friday, but now, as he made the remark, it was
quite dead and fl at. When I tried to look at this with him, he agreed, commenting that he
supposed he was starting to attack the analysis, etc.

The feeling in the session now was awful; N was making a kind of sense and saying analytic
things himself, which could have been right, for example about the Friday, and which one could
have picked up, but, since they seemed fl at and quite unhelpful to him, what he seemed to me to
be doing was putting despair into me, not only about the reality of his marriage and potency, but
also about his analysis, as was indicated, for example, by the useless, and by now somewhat
irrelevant, comment about being left out. N denied my interpretation about his despair about the
progress of the analysis, but in such a way, it seemed to me, as to be encouraging me to make
false interpretations and to pick up his pseudo-interpretations as if I believed in them, while
knowing that they and we were getting nowhere. He vaguely talked about this, went quiet and
said: ‘I was listening to your voice, the timbre changes in different voices. W (his wife), being
younger, makes more sounds per second, older voices are deeper because they make less sounds
per second, etc.’ I showed N his great fear that I showed with my voice, rather than through my
actual words, that I could not stand the extent of his hopelessness and his doubts about myself,
about what we could achieve in the analysis and, therefore, in his life, and that I would cheat and
in some way try to encourage. I queried whether he had perhaps felt that, in that session, my
voice had changed in order to sound more encouraging and encouraged, rather than contain the
despair he was expressing.

By this part of the session my patient had got into contact and said with some relief that, if I did
do this kind of encouraging, the whole bottom would fall out of the analysis.

106

Projective identifi cation: Some clinical aspects First, the nature of the communication, which I
could understand primarily through my counter-transference, through the way in which I was
being pushed and pulled to feel and to react. We see here the concrete quality of projective
identifi cation structuring the counter-transference. It seems that the way N was speaking was not
asking me to try to understand the sexual diffi culties or unhappiness, but to invade me with
despair, while at the same time unconsciously trying to force me to reassure myself that it was all
right, that interpretations, now empty of meaning and hollow, were meaningful, and that the
analysis at that moment was going ahead satisfactorily.

Thus it was not only the despair that N was projecting into me, but his defences against it, a false
reassurance and denial, which it was intended I should act out with him. I think that this also
suggests a projective identifi cation of an internal fi gure, probably primarily mother, who was
felt to be weak, kind, but unable to stand up to emotion. In the transference (to over-simplify the
picture) this fi gure is projected into me, and I fi nd myself pushed to live it out.

We have here the important issue of teasing out the motivation for this projective identifi cation:
was it aimed primarily at communicating something to me; was there a depth of despair that we
had not previously suffi ciently understood; or was the forcing of despair into me motivated by
something different? At this stage, at the end of the session, I did not know and left it open.

I have so much condensed the material here that I cannot convey adequately the atmosphere and
to and fro of the session. But towards the end, as I have tried to show, my patient experienced
and expressed relief and appreciation of what had been going on. There was a shift in mood and
behaviour as my patient started to accept understanding and face the nature of his forcing into
me, and he could then experience me as an object that could stand up to his

acting in, not get caught into it, but contain it. He could then identify temporarily with a stronger
object, and he himself became fi rmer. I also sensed some feeling of concern about what he had
been doing to me and my work – it was not openly acknowledged and expressed but there is
some movement towards the depressive position with its real concern and guilt.

To clarify the motivation as well as the effect of this kind of projective identifi cation on
subsequent introjective identifi cation, we need to go briefl y into the beginning of the next
session, when N brought a dream, in which he was on a boat like a ferry boat, on a grey-green
sea surrounded by mist; he did not know where they were going.

107

Betty Joseph

Then nearby there was another boat which was clearly going down under the water and
drowning. He stepped on to this boat as it went down. He did not feel wet or afraid, which was
puzzling. Amongst his associations we heard of his wife being very gentle and affectionate, but
he added that he himself was concerned; was she behind this really making more demands on
him? She, knowing his fondness for steak and kidney pudding, had made him one the night
before. It was excellent, but the taste was too strong, which he told her!

Now the interesting thing, I think, was that, on the previous day I had felt rather at sea, as I said,
not knowing exactly where we were going, but I was clear that the understanding about the
hopelessness and the defences against it was right, and, though I had not thought it out in this
way, my belief would have been that the mists would clear as we went on. But what does my
patient do with this? He gratuitously steps off this boat (this understanding) on to one that is
going down, and he is not afraid! In other words, he prefers to drown in despair rather than
clarify it, prefers to see affection as demands, and my decent, well-cooked steak and kidney
interpretations as too tasty. At this point, as we worked on it, N could see that the notion of
drowning here was actually exciting to him.

Now we can see more about the motivation. It becomes clear that N was not just trying to
communicate and get understood something about his despair, important as this element is, but
that he was also attacking me and our work, by trying to drag me down by the despair, when
there was actually progress. After a session in which he expressed appreciation about my work
and capacity to stand up to him, he dreamt of willingly stepping on to a sinking boat, so that
either, internally, I collude and go down with him or am forced to watch him go under and my
hope is destroyed and I am kept impotent to help. This activity also leads to an introjective
identifi cation with an analyst-parent who is felt to be down, joyless and impotent, and this
identifi cation contributes considerably to his lack of sexual confi dence and potency. Following
this period of the analysis, there was real improvement in the symptom.

Naturally these considerations lead one to think about the nature of the patient’s internal objects,
for example, the weak mother, that I described as being projected into me in the transference.
How much is this fi gure based on N’s real experience with his mother, how much did he exploit
her weaknesses and thus contribute to building in his inner world a mother, weak, inadequate and
on the 108

Projective identifi cation: Some clinical aspects defensive, as we saw in the transference? In
other words, when we talk of an object projected on to the analyst in the transference, we are
discussing an internal object that has been structured in part from the child’s earlier projective
identifi cations, and the whole process can be seen being revived in the transference.

I want now to digress and look at this material from a slightly different angle, related to the
patient’s very early history and anxieties.

I have shown how N pulls back and goes into an object, in the dream, into the sinking boat, as in
the fi rst session he goes into despair, which is then projected into me, rather than his thinking
about it. This going into an object, acted out in the session, is, I believe, linked with a more total
type of projective identifi cation that I indicated in the sexual dream of T, and referred to briefl y
at the beginning of this paper, as being connected with phobia formation. At the very primitive
end of projective identifi cation is the attempt to get back into an object, to become, as it were,
undifferentiated and mindless and thus avoid all pain. Most human beings develop beyond this in
early infancy;

some of our patients attempt to use projective identifi cation in this way over many years. N,
when he came into analysis, came because he had a fetish, a tremendous pull towards getting
inside a rubber object which would totally cover, absorb, and excite him. In his early childhood
he had nightmares of falling out of a globe into endless space.

In the early period of analysis he would have severe panic states when alone in the house, and
would be seriously disturbed or lose contact if he had to be away from London on business. At
the same time there are minor indications of anxieties about being trapped in a claustrophobic
way, for example, at night he would have to keep blankets on the bed loose or throw them off
altogether; in intercourse phantasies emerged of his penis being cut off and lost inside the
woman’s body. As the analysis went on, the fetishistic activities disappeared and real
relationships improved, and the projecting of the self into the object could clearly be seen in the
transference. He would get absorbed in his own words or ideas or in the sound of my words and
my speaking, and the meaning would be unimportant compared with the concrete nature of the
experience. This type of absorption into words and sounds, with the analyst, as a person, quite
disregarded, is not unlike the kind of process that one sometimes sees in child patients, who
come into the playroom, on to the couch, and fall so deeply asleep that they are unable to be
woken by interpretations. It is, therefore, interesting to see in N how he has always concretely
attempted to get into an object, 109

Betty Joseph

apparently largely in order to escape from being outside, to become absorbed and free from
relating and from thought and mental pain.

And yet we know that this is only half the story, since the object he mainly got into was a fetish
and highly sexualized. And still in the modern dream of getting into the drowning boat there was
masochistic excitement that he tried to pull me into and in this sense it needs to be compared
with T. I described how, as his constant invading and taking over was being analysed, we could
see in T’s sexual dream an attempt totally to get inside me with great excitement. I suspect there
is much yet to be teased out about the relation between certain types of massive projective
identifi cation of the self and erotization.

Now I want to return to the material that I quoted and to the question of projective identifi cation
in patients who are becoming more integrated and nearer to the depressive position. We can see
in the case of N, unlike T who is still imprisoned in his own omnipotent, narcissistic structure,
that there is now a movement, in the transference, towards more genuine whole object relations.
At times he can really appreciate the strong containing qualities of his object; true he will then
try to draw me in and drag me down again, but there is now potential confl ict about this. The
object can be valued and loved, at times he can consciously experience hostility about this, and
ambivalence is present.

As his loving is freed, he is able to introject and identify with a whole valued and potent object,
and the effect on his character and potency is striking. This is a very different quality of identifi
cation from that based on forcing despairing parts of the self into an object, which then in his
phantasy becomes like a despairing part of himself. It is very different from the type of identifi
cation we saw in T, where the patient invaded my mind and took over the split and idealized
aspects, leaving the object, myself, denuded and inferior. With N, in the example I have just
given, he could experience and value me as a whole, different and properly separate person with
my own qualities, and these he could introject and thereby feel strengthened. But we still have a
task ahead, to enable N to be truly outside and able to give up the analysis, aware of its meaning
to him and yet secure.

Summary

I have tried in this paper to discuss projective identifi cation as we see it operating in our clinical
work. I have described various types of 110

Projective identifi cation: Some clinical aspects projective identifi cation, from the more
primitive and massive type to the more empathic and mature. I have discussed how we see
alterations in its manifestation as progress is made in treatment and the patient moves towards
the depressive position, is better integrated and able to use his objects less omnipotently, relate to
them as separate objects and introject them and their qualities more fully and realistically, and
thus also to separate from them.

111

Projective identification

The analyst’s involvement 35

Michael Feldman

In Klein’s original formulation of the mechanism of projective identifi cation she referred to an
unconscious phantasy in which the patient expelled what were usually disturbing contents into
another object. This object is partially transformed in the patient’s mind as a consequence of the
projection, being now possessed of qualities the patient has expelled. In addition to its use as a
method of evacuation, Klein suggested that projective identifi cation may fulfi l a variety of other
unconscious functions for the patient, such as leading to him believing that he possesses the
object, or controls it from within.

These projective processes usually alternate with introjective ones.

Thus the phantasy of forceful entry into the object by parts of the self in order to possess or
control the object creates problems with normal introjection, which the patient may fi nd diffi
cult to distinguish from forceful entry from the outside, in retribution for his own violent
projections (Klein, 1946, p. 11).

The exploration of these unconscious phantasies has increased our understanding of the functions
and defensive needs these primitive mental mechanisms serve for the patient. While the
elucidation of these processes has, in the past, often seemed to emphasise the analyst’s role as a
dispassionate observer, the impingement of the patient’s

35 This chapter reproduces the text of Feldman, M. (1997). Projective identifi cation: the
analyst’s involvement.

International Journal of Psycho-Analysis, 78 , 227–241.

112

Projective identifi cation: Analyst’s involvement phantasies and actions on the analyst has in fact
been recognised from the earliest days of psychoanalysis. Following the early work of Heimann
(1950) and Racker (1958a) there has been increasing interest in the systematic investigation of
the way in which the patient’s phantasies, expressed in gross or subtle, verbal or non-verbal
means, may come to infl uence the analyst’s state of mind and behaviour.

Fairbairn wrote: ‘in a sense, psychoanalytical treatment resolves itself into a struggle on the part
of the patient to press-gang his relationship with the analyst into the closed system of the inner
world through the agency of transference’ (Fairbairn, 1958, p. 385).

We now recognise that while this conscious or unconscious pressure on the analyst may interfere
with his functioning, it can also serve as an invaluable source of information concerning the
patient’s unconscious mental life

– his internal object relations in particular.

More recently, a number of authors have been concerned to elaborate the concept of
countertransference into what is described as an

‘interactive’ model of psychoanalysis, where the emphasis is on the signifi cance of the analyst’s
own subjective experiences in his understanding of and his method of responding to his patient.
Tuckett (1997) has provided an excellent commentary on some of the interesting work in this
area. Building upon the notions of Racker (1958a), Sandler (1976a) and Joseph (1989a), he
elaborates a model of the analytic situation in which both the patient and the analyst engage in
unconscious enactment, placing more or less subtle pressure on the other to relate to them in
terms of a present unconscious phantasy.

He makes the point that ‘Enactment makes it possible to know in representable and
communicable ways about deep

unconscious identifi cations and primitive levels of functioning which could otherwise only be
guessed at or discussed at the intellectual level.’

In this paper I want to focus particularly on the nature of the involvement by the analyst that the
patient seems to require as an essential component of the defensive use of projective identifi
cation.

I will suggest that the projection of elements of a phantasised object relationship represents an
attempt by the patient to reduce the discrepancy between an archaic object relationship and an
alternative object relationship that might be confronting the patient and threatening him. There
are times when the analyst is used primarily as the recipient of projections by which he is
transformed in the patient’s phantasy alone. More commonly, as described above, it seems
necessary for the patient that the analyst should become involved in the 113

Michael Feldman

living out of some aspects of phantasies that refl ect his internal object relations.

I hope to illustrate some of the ways in which the patient’s use of projective identifi cation exerts
subtle and powerful pressure on the analyst to fulfi l the patient’s unconscious expectations that
are embodied in these phantasies. Thus the impingement upon the analyst’s thinking, feelings
and actions is not an incidental side-effect of the patient’s projections, nor necessarily a
manifestation of the analyst’s own confl icts and anxieties, but seems often to be an essential
component in the effective use of projective identifi cation by the patient. Later in the paper, I
will consider some of the defensive functions these processes serve. Confronted with such
pressure, the analyst may apparently be able to remain comfortable and secure in his role and
function, involved in empathic observation and understanding, recognising the forces he is being
subjected to, and with some ideas about their origins and purpose. He may, on the other hand, be
disturbed by the impingement and transformation in his mental and physical state, becoming
sleepy, confused, anxious or elated. Finally, it may become apparent to the analyst that he has
unconsciously been drawn into a subtle and complex enactment that did not necessarily disturb
him at fi rst, but which can subsequently be recognised as the living out of important elements of
the patient’s internal object relationships.

We are concerned with a system in which both patient and analyst are dealing with the anxieties
and needs aroused in each of them by the phantasies of particular object relationships. The
disturbance in either the patient, or the analyst, or both, arises from the discrepancy between the
pre-existing phantasies that partly reassure or gratify, and those with which each is confronted in
the analytical situation, which are potentially threatening. I am suggesting that this unwelcome
discrepancy drives each to deploy either projective mechanisms or some variety of enactment in
an attempt to create a greater correspondence between the pre-existing unconscious phantasies
and what they experience in the analytic encounter. As I hope to illustrate, part of the analyst’s
struggle involves the recognition of some of these pressures, and the capacity to tolerate the gap
between the gratifying or reassuring phantasies and what he is confronted with in the analytical
situation, which includes the unconscious anxieties evoked by the patient’s projections.

Rosenfeld (1971) describes a psychotic patient who, when confronted with interpretations he
admired, was fi lled with envy and driven to 114
Projective identifi cation: Analyst’s involvement attack his analyst’s functions. In his phantasy,
he wormed his way into the analyst’s brain, like a parasite, interfering with the quickness of his
thinking. This use of projective identifi cation was often accompanied by the patient becoming
confused, unable to think or talk properly, with claustrophobic and paranoid anxieties about
being trapped in the analyst. Rosenfeld describes the need for the analyst empathically to follow
the patient’s description of both real and fantasised events, which are often re-enacted by being
projected into him. The analyst has to bring together the diffuse, confused or split-up aspects of
the patient’s pre-thought processes in his own mind so that they gradually make sense and have
meaning (Rosenfeld, 1987c, p. 160).

When Rosenfeld was able to interpret the dynamics of the patient’s state to him in a clear and
detailed way, his anxiety about having completely destroyed the analyst’s brain diminished, and
the patient was able, with relief, to experience him as helpful and undamaged.

When it became possible for the patient to introject this object in a good state, he could, for a
while, recover his own capacities for clearer thought and speech.

Bion (1958) gives a complex description of the beginning of a session with a psychotic patient,
who gave the analyst a quick glance, paused, stared at the fl oor near the corner of the room, and
then gave a slight shudder. He lay down on the couch, keeping his eye on the same corner of the
fl oor. When he spoke, he said he felt quite empty, and wouldn’t be able to make further use of
the session. Bion spells out the steps in the process by which the patient fi rst used his eyes for
introjection, and then for expulsion, creating a hallucinatory fi gure that had a threatening
quality, accompanied by a sense of internal emptiness. When he made an interpretation along
these lines, the patient became calmer and said, ‘I have painted a picture’.

Bion writes, ‘His subsequent silence meant that the material for the analyst’s next interpretation
was already in my possession’ (p. 71).

Bion suggested that his task was to consider all the events of the session up to that point, try to
bring them together and discern a new pattern in his mind which should be the basis for his next
interpretation.

A young man, Mr. A, encountering me for the fi rst time after a holiday break, was initially
disconcerted by fi nding someone new with him in the waiting room, and then came to the view
that I might have made a mistake, which would cause me discomfort and embarrassment which I
would not be able to face, and he imagined I 115

Michael Feldman

would send a colleague to deal with the problem. Once he had arrived at this construction, the
patient became the calm and confi -

dent observer of his muddled analyst. The patient later told me that during my absence he had
found himself in a mess, he had lost his watch, and felt he hadn’t known what was going on.

I suggested that the patient’s experience of confusion and his diffi -

culties over time had become projected, in phantasy, into me. After fi nding himself briefl y
discomforted in the waiting room, he ‘cured’

himself of his disturbing experience, so he became the calm analytic observer, while, in his
phantasy, I had to summon help to rescue me from the mistake I had made over my timetable.

These examples illustrate patients’ unconscious belief in the effectiveness of a concrete process
by which (usually) undesirable and threatening parts of the personality can be split off and
projected.

The motives for this projection vary, but the involvement of the object as a recipient of this
projection is a defi ning characteristic of projective identifi cation, as is the belief in the
transformation of the object by the projection. This transformation may take place in relation to a
delusional or hallucinatory object, an absent object or a dream object, but central to our work is
the investigation of the process in relation to the analyst in the room with the patient. In the
examples quoted, the patients seemed to have no doubt about the effectiveness of the
transformation of themselves that accompanied the transformation of the object. I think there was
a general assumption, based on previous experience, of the sympathy, understanding and
receptivity of the analyst, but it is a feature of the projective processes manifested in these
examples that they did not depend on concurrent evidence of the analyst’s capacity or
willingness to receive the projections.

Indeed, the noteworthy feature of these examples is the contrast between the picture we have of
the analyst’s actual mental state, and the way in which this is represented in the patient’s
phantasy. As Bion has pointed out, patients vary in the extent to which they are able to take
‘realistic steps’ to affect their object by projective identifi cation, and vary in their capacity to
recognise and respect the actual properties of the object. Thus with some patients, the omnipotent
phantasy is likely to have little counterpart in reality. While Rosenfeld and Bion have made
important contributions to our understanding of the impact of the patient’s projections on the
analyst, in the situations I have quoted, they both convey thoughtful, calm, benign 116

Projective identifi cation: Analyst’s involvement attention, in marked contrast to the phantasy
either of a persecutory object, or an analyst whose mind has been invaded and damaged.

When Rosenfeld talked to his patient, in a clear, insightful and empathic way, taking the
phantasy into account, but clearly demonstrating a state of affairs diametrically opposed to that
which obtained in the patient’s phantasy, the patient was relieved, and was able to recover some
of his lost ego functions.

With my own patient, Mr A, I found myself interested in and concerned about the patient’s
experience and the properties with which I had temporarily been invested in the patient’s mind. I
did not actually feel uncertain or confused, and I was confi dent I was seeing the right patient at
the right time. What my patient said did not, on this occasion, discomfort me. The other feature
of this brief example is that when I did talk to the patient in a way that conveyed that I was
neither confused nor particularly anxious, and gave him the impression that something was being
understood, he was able to recall and integrate more of his own experiences. Later in the session
he told me that during the holidays he had moved out of his offi ce to a larger, more spacious offi
ce on a higher fl oor. The two people with whom he had shared the old offi ce had been away,
and when they returned they complained bitterly about the terrible mess he had left.

Mr A said, indignantly, that there might have been a bit of untidi-ness: he had intended to clear it
up, but he had been busy with other things. He went on to suggest that his colleagues were being
unreasonable and neurotic, and he gave other examples of their childish behaviour. He began to
sound like the confi dent and superior person in the larger offi ce whom I had encountered at the
start of the session.

What I think I had failed to question initially was why I should have felt so comfortable and
secure, presented with the material at the start of the fi rst session after a break. I suspect that I
was, in part, enacting the object relationship that the patient subsequently made clearer to me. I
was the confi dent, sane and sensible fi gure in a superior position, dealing with someone into
whom almost all the disturbance and confusion had been projected. This projection and the slight
enactment it gave rise to failed to disturb me, or even to alert me at the time, since my role as the
unruffl ed observing analyst in the offi ce above was congruent with a version of myself with
which I was reasonably comfortable, at least for a while.

Refl ecting on this material, what I also failed to recognise initially was the patient’s unconscious
communication of a bitter complaint 117

Michael Feldman

about my responsibility for having left him with such a mess during the holiday, defensively
claiming that I had intended to do something about it, but largely denying my responsibility for
the disorder.

As I will illustrate later, we have learnt not only to take notice of our feelings of discomfort as
possible refl ections of the patient’s projective identifi cation, but also to consider situations in
which we fi nd ourselves perhaps feeling a little too secure and comfortable, confi -

dent about where the pathology lies, and who is responsible for the mess. I think this example
illustrates that there is in fact a complex relation between the projection into an object in
phantasy (even in the absence of the actual object), and what happens as soon as the patient and
analyst encounter one another, when quite subtle, non-omnipotent interactions begin to take
place, usually based on unconscious projections into the analyst.

Of course, it is not diffi cult to see the advantages of projection into a hallucinatory, delusional or
absent object.

Since it is an omnipotent process, there is no doubt about the object’s receptivity, and the
consequent transformation (there also seem to be no problems about the corresponding
introjection of the object’s valuable properties).

The patient is not confronted with the contrast between phantasy and reality, which is disturbing,
nor with the differences between himself and his object.

What were the factors that allowed the more benign, integrative process, which Rosenfeld
describes, to take place, albeit temporarily?

How can a patient sometimes tolerate, and indeed feel greatly relieved by, being confronted with
an analyst in a state quite discordant with their psychic reality at that moment? Why, on the other
hand, do some patients feel driven to use other methods, more subtle or more violent, to involve
the analyst through projective identifi cation?

While Bion’s patient had split off and projected a dangerous persecutory version of the analyst
into the hallucinatory object in the corner, he did at least have some conception of benign
symbolic communication, which is implied in the belief that it was possible to paint a picture in
the mind of a suitably receptive analyst. Other patients either seem to have no belief in this
possibility, or cannot tolerate such a confi guration. Bion (1959) has vividly described how the
infant, confronted with what seems like an impenetrable object, is driven to attempt to project
into such an object with more and more force. The early experience of such diffi culties with the
object’s receptivity may drive the patient to involve the analyst in such a way 118

Projective identifi cation: Analyst’s involvement that his mind is actually disturbed, or actually to
force him to become compliant or persecutory. It is as if the patient has such doubts about the
possibility either of symbolic communication or the object’s receptivity to any form of projection
that he cannot relent until he has evidence of the impact on the analyst’s mind and body. If this
consistently fails, confi rming an early experience of an unavailable, hateful object, he may give
up in despair.

We tend to assume that once the patient has felt understood, in the sense of some important part
of him being accepted, he would be relieved by the contrast between the more sane and benign
imago of the analyst and the archaic one projected into him (to use Strachey’s

[1934] terms). We sometimes assume that it is only the operation of the patient’s envy that
militates against this.

However, it often seems that there is a different drive in operation, namely the pressure towards
identity, which seems paradoxical and diffi cult to reconcile with the longing for a better, more
constructive experience. It is as if the patient requires the analyst’s experience or behaviour to
correspond in some measure to his unconscious phantasy, and is unable to tolerate or make use
of any discrepancy, however reassuring we might assume that to be.

On the contrary, as Sandler and Sandler (Sandler, 1990; Sandler and Sandler, 1978) have pointed
out, the patient’s attempts to ‘actualise’ such phantasies can be regarded as a form of wish-fulfi
lment, serving a reassuring and gratifying function.

Joseph (1987) describes a session in which an analyst interpreted a deprived child’s reaction to
the imminent end of a Friday session.

The analyst interpreted the child’s urgent wish to make a candle as an expression of her desire to
take a warm object away with her. The child screamed, ‘Bastard! Take off your clothes and jump
outside’.

The analyst tried to interpret the child’s feelings about being dropped and sent into the cold, but
the child replied,

‘Stop your talking, take off your clothes! You are cold. I’m not cold’. While the projection into
the representation of the analyst leads to the child saying, ‘You are cold. I’m not cold’, this will
not suffi ce for the child. Her non-delusional perception of the analyst as being relatively warm
and comfortable drives her to try to force the analyst actually to take off her clothes, so that she
would indeed be cold, and there would not be the immensely painful and disturbing discrepancy
between the internal representation and the fi gure she encounters in the external world. This
dramatic scenario is reproduced in more subtle ways with many of our patients.

119

Michael Feldman

I am suggesting this goes beyond and seems to confl ict with the need to feel understood, or
reassured about the capacity of the object to take in and to ‘contain’ the projections. The lack of
this identity between the internal and external reality may not only stir up envy, or doubts about
the object’s receptivity, but create an alarming space in which thought and new knowledge and
understanding might take place, but which many patients fi nd intolerable.

Incidentally, I am assuming some familiarity with the way in which Rosenfeld and Bion have
expanded and

deepened our understanding of the use of projective identifi cation as a means of communication
and recognised the forceful or even violent use of projective identifi cation in an attempt to get
through to an impenetrable, rejecting object.

Clinically, of course, the patient’s use of more forceful projection may be driven by his
experience of the analyst as a non-understanding, non-receptive fi gure, which the analyst may
not perceive.

There have been important developments in our recognition and understanding not just of the
ways in which the patient might need to project a feeling of confusion, inadequacy or excitement
into the analyst, but the more complex and subtle ways in which the analyst is induced into states
of mind, sometimes accompanied by various forms of enactment, which are relevant to the
patient’s early history, and his current anxieties, defences and desires.

I want to consider what functions these interactions serve for the patient, and how he might
succeed in involving the analyst. Sometimes the analyst will recognise that there is something
slightly alien, disturbing, discordant with a view of himself that he can comfortably tolerate, and
we have learnt to consider this state as a result of the patient’s projective identifi cation. This
recognition can lead us to a better understanding of our own diffi culties, as well as the important
confi gurations in the patient’s object relationships which are being lived out in the analytic
situation.

What writers such as Joseph and O’Shaughnessy (1992) have described are the diffi culties in
easily or quickly recognising the analyst’s involvement resulting from the projective identifi
cation. On the contrary, the analyst may have the sort of comfortable, benign, dispassionate
involvement I described at the beginning of the paper.

What sometimes emerges is that this state represents the unconscious convergence of the
patient’s and the analyst’s defensive needs and may militate against real progress.

Money-Kyrle (1956) has described the process taking place in the analyst as follows: ‘As the
patient speaks, the analyst will, as it were, 120

Projective identifi cation: Analyst’s involvement become introjectively identifi ed with him, and
having understood him inside, will re-project him and interpret’ (p. 361). When there are
particular diffi culties in understanding or helping the patient, two factors may contribute to this.
Firstly, there is the patient’s projection and disowning of unwanted aspects of himself. Secondly,
when these projections correspond to aspects of the analyst himself that are unresolved and not
understood, he may have diffi culty in appropriately re-projecting the patient. If he then ‘cannot
tolerate the sense of being burdened with the patient as an irreparable or persecuting fi gure
inside him, he is likely to resort to a defensive kind of re-projection that shuts out the patient and
creates a further bar to understanding’.

He makes the point that for some analysts – for example, those who most crave the reassurance
of continuous success – the strain of not being able to understand or help the patient is felt more
acutely than others. Money-Kyrle suggests that the extent to which an analyst is emotionally
disturbed by periods of non-understanding will probably depend, in the fi rst instance, on another
factor: the severity of his own superego. If our superego is predominantly friendly and helpful,
we can tolerate our own limitations without undue distress, and, being undisturbed, will be the
more likely to regain contact quickly with the patient. But if it is severe, we may become
conscious of a sense of failure as the expression of an unconscious persecutory or depressive
guilt. Or, as a defence against such feelings, we may blame the patient.

While I fi nd Money-Kyrle’s descriptions familiar and convincing, what we have become more
aware of is that when the analyst is confronted with the anxieties and strain he describes, he may
be unconsciously drawn to diminish them by enacting a complex object relationship with the
patient that initially serves to reassure both. I believe this is achieved by the analyst striving to
create a closer correspondence between a relatively comfortable or gratifying internal
representation of himself and the way in which he experiences and interprets the external
situation. Indeed, while I think Money-Kyrle is describing the process by which the analyst
disentangles himself from the patient’s projection in order to understand and communicate, the
re-projection he describes may actually be a form of enactment by which the analyst deals with
an uncomfortable version of his relationship with the patient.

To return for a moment to Rosenfeld’s paper describing his work with the psychotic patient,
which I quoted at the beginning:

121

Michael Feldman

One of the diffi culties of working through such situations in the analysis is the tendency to
endless repetition, in spite of [the patient’s] understanding that very useful analytic work was
being done. It is important in dealing with patients and processes of this kind to accept that much
of the repetition is inevitable. The acceptance by the analyst of the patient’s processes being re-
enacted in the transference helps the patient to feel that the self, which is constantly split off and
projected into the analyst, is acceptable and not so damaging as feared.

(Rosenfeld, 1987c, p. 180)

Why does Rosenfeld address his colleagues in this way? I think the point he is making is that
unless the analyst recognises the fact of and perhaps even the necessity for the repetition and re-
enactment, he may become disheartened, confused or resentful. In other words, far from being
able to feel reasonably confi dent in the representation of himself as a helpful, effective, patient
analyst, he might be burdened by an intolerable version of himself that he may then try to deal
with very concretely. This could be enacted by the analyst blaming or accusing the patient in a
hostile and critical way, entering into a defensive collusive arrangement, or by terminating the
treatment in despair.

What I am thus suggesting is that what is projected is not primarily a part of the patient, but a
phantasy of an object relationship. It is this that impinges upon the analyst, and may allow him to
remain reasonably comfortable, or may disturb him and incline him to enact. This enactment is
sometimes congruent with the phantasy that has been projected, so that the analyst becomes a
little too compliant or too harsh. On the other hand, the enactment might represent the analyst’s
attempt at restoring a less disturbing phantasy to the fore (for example, having to distance
himself consciously or unconsciously from an impotent or sadistic archaic fi gure). Finally, we
must also be aware that the impulse towards enactment may refl ect unresolved aspects of the
analyst’s own pathological internal object relations.

I believe some of these issues are addressed by O’Shaughnessy (1992) with great clarity and
insight. She describes how a patient initially drew her into making denuded, un-disturbing
interpretations, and offering what seemed like reasonable links with the patient’s history. Thus, it
seems, the analyst initially felt reasonably comfortable with her role and functions. After a period
of time, 122

Projective identifi cation: Analyst’s involvement however, she became uneasy and dissatisfi ed
with such interpretations, which felt inauthentic, and which did not seem to promote any change.
The insight, and work involved in the recognition of something in the patient’s limited and over-
close relationship with her, and her own denuded functioning with the patient, which needed
exploration and thought, led, I believe, to a crucial transformation in the analyst’s representation
of herself, and consequently in her ability to function. There is a convergence between the
internal representation of herself as a thoughtful, reparative fi gure and the person who has now
been able to recognise the degree of acting out that inevitably occurs, and this can be used to
further understanding.

This shift in internal perspective promotes the change from the situation in which the analyst is
unwittingly involved in the enactment of the patient’s problems, to the emergence of the
potential for containment and transformation by the analyst, refl ected in a shift in the style and
content of the interpretations.
What O’Shaughnessy was then able to recognise was the function this over-close, secluded and
denuded relationship served for the patient. The fact that the patient made a refuge of symmetry
and over-closeness suggested that she was afraid of differences and distance between herself and
her objects. The placation between analyst and patient was necessary because the patient feared
either too intense erotic involvement or violence between them. I assume she had unconsciously
evoked corresponding versions of these disturbing phantasies in the analyst’s mind, which
resulted in her functioning in the way she initially described. O’Shaughnessy describes how, in
sessions when acute anxiety threatened, the patient worked to rebuild her refuge, subtly and
powerfully controlling the analyst to be over-close and to operate within its limits.

Thus, at the beginning of the analysis, the patient transferred her highly restricted object relations
into the analytic situation. She must have communicated with words and non-verbal projections
her intense anxieties about a fuller and freer object relationship, with the terrifying erotic and
violent phantasies associated with this.

I believe the analyst’s anxieties about being experienced both by the patient and herself, in these
disturbing and destructive roles, led her to function in the way the patient apparently required.
While this may have served as a necessary temporary refuge at the start of the analysis, the
analyst subsequently felt uneasy and dissatisfi ed with her role, and was then able to think about
it in a different way.

123

Michael Feldman

I think the patient always fi nds this shift very threatening – it creates an asymmetry, and may
arouse envy and hatred, with powerful attempts to restore the status quo ante. This may be
successful if the analyst cannot tolerate the uncertainty, anxiety and guilt associated with the
emergent phantasies of the relationship as a frightening, disappointing and destructive one, and
we sometimes need the internal or external support of colleagues to sustain our belief in what we
are attempting to do.

Meltzer describes a somewhat similar dynamic in relation to a group of disturbed patients who
use extensive projective identifi cation, which results in a compliant, pseudomature personality:
the pressure on the analyst to join in the idealization of the pseudomaturity [is] . . . great, and the
underlying threats of psychosis and suicide so covertly communicated . . . the
countertransference position is extremely diffi cult and in every way repeats the dilemma of the
parents, who found themselves with a ‘model’ child, so long as they abstained from being
distinctly parental, either in the form of authority, teaching, or opposition to the relatively modest
claims for privileges beyond those to which the child’s age and accomplishments could
reasonably entitle it.

(Meltzer, 1966, pp. 339–340)

The parental fi gure is thus faced either with the phantasy of being helplessly controlled, or the
phantasy of driving the child into madness or suicide.

In the fi nal part of this paper, I should like to illustrate in more detail fi rst the way in which I
believe a patient was able to use projection into the internal representation of the analyst (in his
absence), to free herself from anxiety, whereas in the subsequent analytic sessions she needed to
involve the analyst in different ways. I believe she achieved this through her projection of
phantasies of disturbing object relations that were not only refl ected in her verbal
communications, but also partially enacted by her in the sessions. I suspect that if the analyst is
receptive to the patient’s projections, the impact of the patient’s disturbing unconscious
phantasies that concern the nature of his relationship with the patient inevitably touch on the
analyst’s own anxieties. This may evoke forms of projection and enactment by the analyst, in an
attempt at restoring an internal equilibrium, of which the analyst may initially be unaware. The
diffi cult 124

Projective identifi cation: Analyst’s involvement and often painful task for the analyst is to
recognise the subtle and complex enactments he is inevitably drawn into with his patient, and to
work to fi nd a domain for understanding and thought outside the narrow and repetitive confi nes
unconsciously demanded by the patient, and sometimes by his own anxieties and needs. While
the achievement of real psychic change is dependent on this process, it is threatening for the
patient and liable to mobilise further defensive procedures.

The patient I want to describe is a single woman, who has been in analysis for several years. She
arrived on a Monday morning and after a silence told me she was very involved in something
that had occurred on Saturday, and which she hadn’t thought about since –

not until she was actually here. A friend, who works as a psychotherapist, told her about a young
male supervisee who confessed to her that he had seduced one of his patients. My patient’s
friend told her not to tell anyone, and as soon as she said that my patient immediately thought of
me. My patient proceeded to give some details of the complicated connections between
therapists, supervisors and the patient involved. She seemed very concerned about

who discussed what with whom, and commented on how incestuous it all seemed.

She added that there was something almost sinister about all these people knowing about it.
Then, after a silence she said, ‘thinking about it here, I was wondering why it should come to my
mind here.

I feel reasonably calm about it, it doesn’t make me want to curl up in horror. I feel suffi ciently
removed from it, otherwise it would be horrifi c’.

There was a tense and expectant silence, and I felt aware of a pressure to respond quickly to what
she had brought.

When I did not do so, she commented that the silence seemed rather ominous.

When, on the Saturday, my patient was confronted with the disturbing image of a therapist’s
incestuous involvement with his patient, and told not to tell anyone, I was conjured up in her
mind, and I believe she projected the knowledge, the anxiety and disturbance into me. It was
then not something she had in mind to tell me about – on the contrary, it had become unavailable
to her until she actually encountered me on Monday. I suggest we are thus dealing not with
ordinary thinking or communication but rather with the omnipotent projection in phantasy not
only of mental contents but also of the capacity to think about them. Since the process is an
omnipotent one, the patient does not need to use symbolic means of 125

Michael Feldman

communication. In this case the phantasy involves an object immediately receptive to the
patient’s projections, and apparently neither disturbed by them, nor changed into something
threatening. Involving the object in this way seems to have succeeded in completely freeing the
patient of anxiety and discomfort.

When she encountered me at the beginning of the session on Monday, and became aware that in
reality I did not have possession of what she had got rid of, she recovered that part of her mind,
and its contents, which had in phantasy been projected. She was then driven to use verbal and
non-verbal communication in a non-omnipotent way, apparently in order to achieve the same
outcome. While telling me about all the incestuous connections between therapists, supervisors
and patients, it was striking that my patient wondered why all of this should come into her mind
while she was with me, apparently failing to make the link between the story she reported and
the phantasies connected with her own relationship with her analyst. I believe that by the
combination of conscious and unconscious actions involved in this procedure, the patient was
able both to communicate with and to ‘nudge’ the analyst into thinking about and taking
responsibility for the thoughts, phantasies and impulses towards action that threatened her.

The point I wish to emphasise is that the projective mechanisms served several functions. Firstly,
they evidently allowed the patient to disavow the disturbing or potentially disturbing responses
to what her friend had elicited.

Secondly, they involved the analyst in the sense that it was now his function to make the
connections, and think about the signifi cance of what she had communicated. Thirdly, I hope to
illustrate the way in which they served to draw the analyst into the partial enactment of some of
the underlying phantasies that had been elicited, which had to be dealt with by the patient, in
spite of the analyst’s conscious attempts to avoid such an enactment, and to fi nd a working
position with which he could feel reasonably comfortable.

In the session, I was made aware of the obvious role I was expected to play by the palpable
pressure to respond quickly to what she had brought, and make some half-expected comment or
interpretation.

My long experience with this patient suggested that if I had complied, and directly addressed the
material she had brought, offering some rather obvious answers to why it should come to her
mind in the room with her analyst, there were a limited number of repetitive, and unproductive
scenarios.

126

Projective identifi cation: Analyst’s involvement The fi rst, and most common one, involved the
patient relaxing and withdrawing, re-enacting with me the procedure that had taken place on
Saturday when her friend had spoken to her,
making it clear that the diffi cult and potentially disturbing material was no longer in her
possession, but in mine.

The second involved a less complete projection, in which the patient retained some contact with
what had been projected, but resisted the dangerous prospect of thinking for herself about these
issues, insisting that it was my function to do so. The third scenario was one in which my
interpretations were themselves concretely experienced as threatening and demanding intrusions.
In the session I have described, I was not aware of being disturbed by the contents of the
patient’s material, but I was troubled and disheartened by the prospect of enacting one of these
repetitive and unproductive roles with her. However, when I remained silent for a while,
attempting to fi nd a way of understanding and approaching the patient, my silence nevertheless
evoked the patient’s phantasy of a disturbing archaic object relationship, in which she was
involved with a threatening, ‘ominous’ fi gure, fi lled with unspoken, alarming things, potentially
intrusive and demanding.

I believe she had partially re-created an important archaic object relationship through the
interaction of two powerful factors. Firstly, the phantasised projection into the analyst of some of
these archaic qualities and functions.

Secondly, by communicating and behaving in the way she had, she was indeed faced with an
analyst whose mind was fi lled with thoughts about what she had told him, who did indeed want
something from her, and might make diffi cult and

‘intrusive’ demands on her. When these expectations and experiences were coloured by the
qualities projected into them, the patient was indeed living out an archaic, familiar object
relationship.

In this session, and those that followed, I felt the need to try and fi nd a way of working that I
hoped would partially avoid the repetitive interactions I have described. I remained silent at
times, trying to understand what was taking place, or made comments on what I thought the
patient was doing with me, or expecting of me. I also attempted to get the patient to explore what
was making her so uncomfortable, and some of the links between her material, her family
history, and the analytical situation that I thought were available to her. I was made aware of the
threat my efforts posed to the patient’s equilibrium, and her extreme reluctance to allow either of
us to escape from familiar interactions that appeared, paradoxically, 127

Michael Feldman

to be necessary and reassuring for her. I felt subjected to powerful pressure either to allow
myself to be used in such a way that I had to take responsibility for the disturbing material that
the patient projected, or to enact some elements of the phantasy of a forceful seductive or
intrusive relationship. I was thus confronted with painful and unwelcome representations of my
role in relation to my patient, and continued to struggle to fi nd an approach that I felt might be
more constructive, and with which I could be more comfortable.

There is always the idea that by remaining more silent, or speaking more, understanding the
situation in a different way, taking a different tack, one can free oneself from such repetitive and
unproductive interactions. Sometimes this is manifested in the thought (held by the analyst, or
the patient, or both) that if the analyst changed, or were a different kind of analyst, these
problems would not arise. Of course, these considerations have to be taken seriously, and will
often have some element of truth. However, for much of the time in dealing with my patient, I
came to believe that whatever I said or did was liable to be experienced in accordance with the
limited, archaic phantasies I have briefl y indicated, and that the repetitive living-out of these
phantasies in the sessions served important and reassuring functions for the patient. There were
brief periods of thoughtful refl ection that were a relief to me, as I felt I could regain a sense of
my proper function.

However, it was evidently painful and diffi cult for the patient to be anywhere outside the
familiar and reassuring enactments, and she would quickly withdraw again, or re-evoke the
excited provocative relationship in which, paradoxically, she seemed to feel safer.

For example, after a period of diffi cult work the patient said, thoughtfully, ‘I can see . . . both
sides . . . in what has been going on.

I can appreciate you want me to . . . look rather more closely at the things that have come up.
After all, just putting them out in an extremely cautious way as “ideas” doesn’t get me any
further’. Her voice then became fi rmer and

more excited: ‘At the same time it seems remarkable to me that I’m even prepared to mention
these things. In fact I’m amazed. I must feel very confi dent that I am not going to be pushed into
anything more’. Her excitement escalated, and she repeated how extraordinary it was that she
had said as much as she had, what a risk she had taken that I would seize on the opportunity. She
said that normally her main concern was to avoid saying things if she could foresee some sort of
opening she might give me, so she has to make sure this doesn’t occur.

128

Projective identifi cation: Analyst’s involvement Thus, having briefl y and uncomfortably
acknowledged the existence of an analyst who was actually trying to help her, and the
recognition of the defensive processes she was so persistently caught up in, she moved in to a
state of erotised excitement that gripped her for much of the rest of the session. The patient thus
seemed compulsively driven to involve me in interactions in which she either experienced a
tantalising, ominous withholding or exciting demanding sexual intrusion. These were, of course,
aspects of the powerful oedipal confi guration that had been evoked in her mind by the episode
her friend had originally reported to her, and which had important links with her early history.

While it is familiar to us, I fi nd that the recurrent pressure on the analyst to join the patient in the
partial enactment of archaic, often disturbed and disturbing object relationships is one of the
most interesting and puzzling phenomena we encounter. With my patient, what functions did it
serve to involve me not as a helpful benign fi gure, but a version of a disturbing archaic one? I
suspect there are many answers to this. This interaction frees the patient from knowledge of and
responsibility for her own impulses and phantasies: she is predominantly a helpless victim. It
was very evident in the sessions that it provided her with a degree of gratifi cation and
excitement. It may have served as a means of making me recognise and understand aspects of
her history, or her inner life, which I had thus far failed to address, although I am uncertain about
suggesting this as her motive.

What I want to add is the way in which it seems to serve a reassuring function if what is enacted
in the external world corresponds in some measure with an object relationship that is
unconsciously present.

The alternative, when she is confronted with the discrepancy between the two, is painful and
threatening.

From the analyst’s point of view, I suspect that if he is receptive to the patient’s projections, the
phantasies of archaic object relationships must inevitably resonate with the analyst’s own
unconscious needs and anxieties. If these relate too closely to areas of confl ict that remain
largely unresolved, there are dangers that the analyst will be driven into forms of enactment that
either gratify some mutual needs or defend him against such gratifi cation.

Hoffman points out: Because the analyst is human, he is likely to have in his repertoire a
blueprint for approximately the emotional response that the patient’s transference dictates and
that response is likely to be elicited, whether 129

Michael Feldman

consciously or unconsciously . . . Ideally this response serves as a key – perhaps the best key the
analyst has – to the nature of the interpersonal scene that the patient is driven by transference to
create.

(Hoffman, 1983, p. 413)

As Joseph (1987, 1988), O’Shaughnessy (1992) and Carpy (1989) have suggested, we may have
to recognise that a degree of enactment is almost inevitable; part of a continuing process that the
analyst can come to recognise, temporarily extricate himself from, and use to further his
understanding. Indeed, in the clinical situation I have just described, it seemed important to
recognise the pressure towards enactment within the patient, and the corresponding pressures felt
by the analyst. The recognition of the compulsive and repetitive nature of these interactions may
have important consequences. As Rosenfeld and O’Shaughnessy have indicated, it may allow the
analyst to recover some sense of his own proper function. This diminishes the discrepancy
between his own phantasies of his role and what is manifested in the analytical situation. If the
analyst is also more able to tolerate whatever discrepancies exist, he will be less driven to use
projective mechanisms and the forms of enactment I have been describing. In the space thus
created, he may be able to think differently about his patient.

In this chapter I have tried to emphasise that what is projected into

the analyst is a phantasy of an object relationship that evokes not only thoughts and feelings, but
also propensities towards action. From the patient’s point of view, the projections represent an
attempt to reduce the discrepancy between the phantasy of some archaic object relationship and
what the patient experiences in the analytical situation.

For the analyst too, there are impulses to function in ways that lead to a greater correspondence
with some needed or desired phantasies.

The interaction between the patient’s and the analyst’s needs may lead to the repetitive
enactment of the painful and disturbing kind that I have described. It may be very diffi cult for
the analyst to extricate himself (or his patient) from this unproductive situation and recover his
capacity for refl ective thought, at least for a while.

As I have indicated, the diffi culty is compounded when the projection into the analyst leads to
subtle or overt enactments that do not initially disturb the analyst, but on the contrary constitute a
comfortable collusive arrangement, in which the analyst feels his role is 130

Projective identifi cation: Analyst’s involvement congruent with some internal phantasy. It may
be diffi cult to recognise the defensive function this interaction serves both for the patient and the
analyst and the more disturbing unconscious phantasies it defends against.

The analyst’s temporary and partial recovery of his capacity for refl ective thought rather than
action is crucial for the survival of his analytical role. The analyst may not only feel temporarily
freed from the tyranny of repetitive enactments and modes of thought himself, but he may
believe in the possibility of freeing his patient, in time.

However, such moves are likely to provoke pain and disturbance in the patient, who fi nds the
unfamiliar space in which thought can take place frightening and hateful.

131

Who’s who?

Notes on pathological identifi cations 36

Ignes Sodré

Freud’s (1917) discovery, in ‘Mourning and melancholia’, of the process through which the ego
unconsciously identifi es with the introjected bad object (the rejecting loved object) thus
becoming a victim of its own superego, was one of the most important breakthroughs in
psychoanalysis: perhaps as important as the discovery of the meaning of dreams and of the
Oedipus complex. The idea that when individuals feel ‘I am the worst person in the world’, they
may in fact be unconsciously accusing somebody else whose victim they feel they are, but who,
through a pathological process of introjection and identifi cation, they have ‘become’, was
indeed a revolutionary one, and one which is still of tremendous clinical importance for us today.

Freud (1917) describes the establishing of what he calls a narcissistic identifi cation with the
abandoned object in two ways: as a passive taking in of the object – ‘thus the shadow of the
object fell upon the ego’ (p. 249) – and as an active process in which ‘the ego wants to
incorporate this object into itself, and, in accordance with the oral or cannibalistic phase of
libidinal development in which it is, it wants to do so by devouring it’ (p. 249). He also describes
the ego as being overwhelmed by the object. It would seem then that there
36 This chapter reproduces the text of Sodré, I. (2004). Who’s who? Notes on pathological
identifi cations. In E.

Hargreaves and A. Varchevker (Eds.), In Pursuit of Psychic Change: The Betty Joseph Workshop
. London: Brunner-Routledge, pp. 53–68.

132

Who’s who? Notes on pathological identifi cations is no differentiation between self and object
at that point – the introjected object occupies the entire ego – except of course that this is not
entirely true, since the ego has undergone a ‘cleavage’ and some of it has now become the
‘special agency’ that judges it (the ego who has become the object) so harshly. As we know, the
superego was subsequently also seen by Freud as the product of introjections. In psychoanalytic
theory, introjections leading to identifi cations with primary objects very soon became linked
with normal development; but the kind of identifi cation described in ‘Mourning and
melancholia’ is a massive, pathological one, characterised by an extraordinary clinical event: the
subject seems to have ‘become’ the object.

In his paper ‘On the psychopathology of narcissism’ Rosenfeld (1964b) stated: Identifi cation is
an important factor in narcissistic object relations.

It may take place by introjection or by projection. When the object is omnipotently incorporated,
the self becomes so identifi ed with the incorporated object that all separate identity or any
boundary between self and object is denied.

In projective identifi -

cation parts of the self omnipotently enter an object, for example, the mother, to take over certain
qualities which would be experienced as desirable, and therefore claim to be the object or part-
object. Identifi cation by introjection and by projection usually occur simultaneously.

(Rosenfeld, 1964b, p. 170)

This is an extremely clear differentiation between two modes of identifi cation; the fi rst
description corresponds to Freud’s mechanism in melancholia, the second follows Klein’s (1946)
discovery of the mechanism of projective identifi cation. But I think it is worth noticing that
Freud’s description of the more active (cannibalistic) form of incorporation is in fact similar to
the description of projective identifi cation. Rosenfeld (1964b) stressed the

omnipotent quality of this type of projective identifi cation; Freud had, as we know, pointed out
the hidden mania implied in melancholia. It seems to me that some states of massive projective
identifi cation are like a manic version of what Freud described as the melancholic’s narcissistic
identifi cation with the (now externally annihilated) object.

In this chapter I shall focus mainly on the interaction of projections,

introjections and manic mechanisms in the creation and perpetuation 133

Ignes Sodré

of those states of pathological identifi cation which are usually described as ‘the subject is in
massive projective identifi cation with the object’, as opposed to states where the subject ‘gets
rid of something’ or ‘does something to the object’ by the use of projective identifi cation. You
will have gathered from my title that I am concerned with exploring extreme shifts in a person’s
sense of identity. I will bring clinical examples to illustrate both the question of the loss of a
sense of identity and the shift into a different identity through the use of excessive introjective
and projective identifi cation.

The sense of identity stems simultaneously from the differentiation of the self from its objects
and from various identifi cations with different aspects of the objects. All object relations depend
on the capacity to remain oneself while being able to shift temporarily into the other’s point of
view. Any meaningful interchange between two people involves of necessity an intricate process
of projections, introjections and identifi cations. ‘Projective identifi cation’

is an umbrella term which includes many different processes involving the operation of both
projection and introjection; it is used to describe normal modes of communication as well as
extremely pathological manoeuvres and even permanent pathological states which are at the root
of some character traits.

One way of differentiating the complex processes involved in the various aspects of what is
called ‘projective identifi cation’ from ‘classical’ projection is that projective identifi cation takes
place in an object relationship and, therefore, necessarily affects both subject and object (in
phantasy, but often in external reality too), whereas it should be at least possible, in theory, to
conceive of projection as not necessarily related to a specifi c object into which something is
being projected. But having said that, I must confess that I fi nd it diffi cult to imagine a
projection into outer space, or into something inanimate or abstract, without imagining too that
whatever has been projected into has become personifi ed in some way.

Projective identifi cation as a defence mechanism has as its primary aim the wish to get rid of a
particular experience; I do not think it is true to say that what characterises projective identifi
cation is that the subject (the

‘projector’, as it were) maintains links with the part of the self that is now felt to be inside the
‘object, the ‘receptor’

(see for instance Ogden’s [1979] discussion). This may occur, but the hall-mark of projective
identifi cation – and especially of pathological projective identifi cation – is the wish to sever
contact with some-134

Who’s who? Notes on pathological identifi cations thing that provokes pain, fear, discomfort; the
word ‘identifi cation’

should, in this particular instance, refer to the object’s identifi cation (in the subject’s mind) with
the projected experience, and not to the subject’s identifi cation with the object: as Sandler
(1987b) clearly pointed out, the self wants to dis-identify with that which is projected.

Projective identifi cation, by defi nition, affects the sense of self, since it involves getting rid of
aspects of the personality through splitting them off and locating them in the object, so that in the
subject’s phantasy the identities of both subject and object are affected. It also may involve
acquiring aspects of the object, in which case the identities are further modifi ed. In her seminal
papers where she fi rst discovered and conceptualised projective identifi cation, Klein (1946,
1955) described both archaic processes furthering communication and development (a concept
developed and expanded by Bion [1962a] in his theory of the container) and a pathological
process leading to loss of contact with the self and aiming at omnipotent control of the object.
Massive projective identifi cation with the object implies a phantasy of ‘becoming’ the object or
a particular aspect or version of the object (and here the

word identifi cation refers also to the subject’s identifi cation with aspects of the object) whereas
the object

‘becomes’

the self, or personifi es an unbearable aspect of the self (a process fi rst described by Anna Freud
[1936] as ‘identifi cation with the aggressor’).

I will suggest that such states of pathological identifi cation imply the excessive use not only of
violent projections but also of concrete, pathological introjections and that this mode of
functioning also relies for its ‘success’ on the massive use of manic defences.

Excessive use of projective identifi cation can lead, on the one hand, to confusion and loss of a fi
rm sense of self and, on the other, to an extreme rigidity in character, where artifi cial new
boundaries are created between subject and object, but are then tenaciously adhered to. In this
case, the new boundaries between what is ‘me’

and what is ‘you’ have to be maintained as a fortress against the threat of the return of the split-
off projected parts of the self, which results not in confusion but in its extreme opposite, in an
absolute certainty which has to be maintained at all costs, to the impoverishment of the
personality and serious disturbance in the capacity for object relations. Arrogance as a character
trait is, I think, a good example of this state of affairs; it is essentially a state of permanent
projective identifi cation with an idealised bad object (I will explain what I mean by this later).

135

Ignes Sodré

Looking rather schematically into what happens in projective identifi cation, one could say that
from the point of view of the ‘projector’, a part of the self becomes in phantasy a part of the
object through a complicated manoeuvre which, for the sake of simplicity, we could temporarily
call ‘projective dis-identifi cation’; the projector is not consciously aware of that aspect of the
self, since he believes that it belongs to the object. This process, which happens in unconscious
phantasy, can of course have an effect on the object – the ‘receptor’
– in external reality (Sandler describes this as the ‘actualisation’ of the projective identifi cation,
whereas Spillius uses the term ‘evocative’).

If this is the case then, from the point of view of the ‘receptor’, there is an intrusion of something
foreign into the self which causes a partial

– or total – ‘forced introjective identifi cation’.

What the outcome of this situation will be will depend on the degree of intrusiveness and
violence of the projection matched up with the

‘receptor’s’ capacity (or lack thereof ) to introject and partially identify with what has been
introjected without losing the boundaries of the self. In other words a helpful ‘receptor’ should
be able to function as a container (Bion) who can simultaneously experience what it is like to feel
what the other person feels (for instance, a mother who can empathise with her baby) through
introjecting what is being projected as the experience of an object. This experience is, in the
inner world, incorporated into the picture of the internal object and not into that of the self. (It is
obvious that if a mother felt totally identifi ed with her distressed baby she would not be able to
help the baby. For example, she has to take in her baby’s fear of not surviving and to be able
partially, and temporarily, also to fear for its survival. But if she becomes so persecuted and
overwhelmed by the baby’s terrifi ed crying to the point of feeling, ‘I will not survive’, then she
will ‘be’ the baby and the baby will ‘be’ a persecutor; more like a bad mother to her. This then
might lead to her emotionally abandoning or even attacking the baby.) The central characteristic
of the use of ‘projective identifi cation’ is the creation in the subject of a state of mind in which
the boundaries between self and object have shifted. This state can be more or less fl exible,
temporary or permanent. The motives for such unconscious manoeuvres are manifold, from the
need to maintain psychic equilibrium and avoid pain, to the more intrusive ones of robbing and
depleting the object. The object’s perception of and method of dealing with what is being
projected will also affect the development of the object relationship that is taking place at that
moment.

136

Who’s who? Notes on pathological identifi cations Even though ‘projective identifi cation’ is
used to describe normal as well as pathological processes, I think that we tend to think of
projective processes as more pathological than introjective ones.

When we think of somebody being identifi ed with somebody else, we tend to think rather
loosely of introjective identifi cation as healthier than projective identifi cation. We visualise two
very different object relationships: one in which the self receives something from the object, and
the other in which there is massive intrusion into the object.

And of course emotional development does depend essentially on taking in from our objects and
identifying with them. But we can excessively polarise these different modes, seeing one as a
peaceful welcoming of the object into the inner world, and the other as the warlike invasion of
the object. In fact, as we know, there is pathological introjection as much as pathological
projection. Furthermore, projection and introjection are psychic mechanisms based on phantasies
which are felt to have the power of concrete actions, and phantasies are totally coloured by affect
and motive. If identifi cation is based on the wish to become the object (and therefore to rob the
object of its identity), as opposed to the wish to be like the object, therefore allowing the object
to continue existing with its identity preserved –

then this is pathological and destructive. And although it is important in analysis to investigate
the unconscious phantasy manoeuvres used to achieve this taking over of the object – to
differentiate what happens through concrete introjection from what happens through intrusive
massive projection – the fundamental point is that the integrity of the object has been damaged or
destroyed in this process. We are talking here of an ‘imperialist’ attitude towards the object and
in this universe the different phantasies and mechanisms employed are simply tactical
manoeuvres to defeat the enemy.

Pathological introjective identifi cation implies a phantasy of concretely taking something in,
whereas a normal identifi cation with an internal object presupposes a capacity to introject
symbolically while allowing the object to retain its separate identity. The same is true of normal
projection, of course: when the ego is functioning in a depressive position mode, symbolic
projection into the other’s mind – being able to put oneself imaginatively in the other’s place –

helps us to understand who the other person is.

In his paper ‘Remarks on the relation of male homosexuality to paranoia, paranoid anxiety and
narcissism’

Rosenfeld (1949) uses a very interesting dream from his patient to illustrate the origin of 137

Ignes Sodré

projective identifi cation. I will quote it here because it is such a clear example of two points I
want to stress: fi rst, the fact that not only affects and parts of the personality are projected, but
also modes of functioning; and second, the role of wholesale, concrete introjection of the object
in states of pathological, massive identifi cation.

Rosenfeld describes this patient as consciously afraid that the analyst will become too interested
in him; he is therefore often silent when he has thoughts that he thinks are of special interest to
the analyst.

Dream: He saw a famous surgeon operating on a patient, who observed with great admiration the
skill displayed by the surgeon, who seemed intensely concentrated on his work. Suddenly the
surgeon lost his balance and fell right inside the patient, with whom he got so entangled that he
could scarcely manage to free himself. He nearly choked, and only by administering an oxygen
apparatus could he manage to revive himself.

Rosenfeld comments that the patient had paranoid fears of being controlled by the analyst from
inside and that later on in the analysis he became more aware of his fear of falling inside the
analyst and becoming entangled inside him.

This dream is a very striking example of how the whole process of projective identifi cation is
itself projected. The surgeon/analyst in the dream relates to the patient via intrusive projective
identifi cation: such is his curiosity that he gets concretely inside his patient. What is projected is
not only curiosity but also a mode of functioning. You could

say that this happens because this is the only mode of relating that the patient knows. This is a
patient who thinks very concretely; but the fact that the surgeon ‘administers to himself an
oxygen apparatus’ seems to me to indicate that the patient thinks that the analyst can save his
own life – his separate identity – by recovering his capacity to function as an analyst. I think the
fact that the word ‘apparatus’

appears in the dream text, rather than simply ‘oxygen’, reinforces this idea. I suspect that
‘administering an oxygen apparatus’ stands for a capacity of the analyst’s that the patient has, in
phantasy, robbed him of through the projection of his all encompassing infantile curiosity.

In the dream this capacity is now available for reintegration into the patient’s picture of the
analyst. (In the patient’s inner world, the analyst ‘cures’ himself by re-establishing himself as the
analyst, with a 138

Who’s who? Notes on pathological identifi cations separate identity and capacities.) This
suggests that this patient is therefore capable of conceiving of such a function. I imagine also that
this patient has already begun to fi nd out, in his analysis, that massive projective identifi cation
is not a great method through which to live one’s life! I think this dream is a rather beautiful
metaphor for moments when the analyst feels entirely at the mercy of violent projections and
then recovers his capacity to function.

I also wanted to use this dream to illustrate something else. What we have here is the patient
ending up with the analyst in his belly, as opposed to ending up inside the analyst. He has power
over the analyst because the analyst is inside him, not him inside the analyst. In other words, not
only has he projected a whole way of functioning into the analyst, but also he has swallowed the
analyst: a pathological massive introjection. There is an expression in Portuguese to describe
somebody who feels he is superior to everybody else: ‘He thinks he’s got the king in his belly’.
(So, through swallowing the king, he is superior even to the king.) An extremely complex
interplay between projections and introjections takes place continuously to perpetuate this
peculiar state of affairs, but I think it may be useful when describing states of massive projective
identifi cation – ‘becoming’ the object – to picture not only the patient inside the analyst
(following Klein’s description of the infantile impulses to invade the mother) but also the patient
with the analyst inside (related to phantasies of primitive oral incorporation of the mother).
Triumph in this case comes from having swallowed the whole object, thus totally controlling and
owning its power and strength.

Manic mechanisms are involved in this process by which the self becomes so much bigger than
the object and so much more powerful.

I hope to illustrate with the following clinical example the interplay of projective and introjective
mechanisms in massive projective identifi cation, as well as the manic fl avour of such
operations.

Mr A: ‘becoming’ the idealised bad object

A narcissistic young man comes into the session and looks at me more closely than usual, staring
intensely into my eyes in a way that feels uncomfortable, intrusive. He lies on the couch and,
with a rather superior tone of voice, tells me that he can clearly see that I must be quite
shortsighted, I have that kind of unfocused look in my 139

Ignes Sodré

eyes. It is ridiculous that I do not wear glasses but I am obviously too vain to do so. I say rather
hesitantly that perhaps there is a reason why he feels today that I cannot see him properly and I
get an absolutely furious, indignant and self-righteous response: I want everything to be his
problem, I don’t want to admit to my own failures, and I clearly suffer from an inferiority
complex about my eyesight. He adds that he has had his eyes tested and has 100

per cent vision.

I think a very complex process of projections, introjections and identifi cations has occurred to
produce this state of affairs and I will now look in detail into what I think may have happened.

Something has obviously taken place that is connected to vision, specifi cally to do with seeing
into the other person.

He may have felt misunderstood in the previous session but I am more inclined to think that he
felt understood in a way that was threatening to him.

My capacity to see inside him made him too anxious, lest insight would threaten his
pathological, but desperately needed, psychic equilibrium (Joseph, 1989b) or, perhaps, because
he feels envious when he thinks I have better

‘eyes’ than he – probably both. I do not know what has happened, but I ask you to accept this as
a working hypothesis so that this can be used as an example of the kind of process that can take
place.

What gives him the absolute certainty that I am shortsighted and pathologically vain (preferring
not to see than to wear glasses) is, I suspect, a projection of his fear of insight and of his
narcissism. This is one aspect of his use of projective identifi cation whereby I, his object, am
now identifi ed with unwanted aspects of himself. From his point of view, though, this could also
be described as projective dis-identifi cation, since through this process he loses part of his
identity. He has lost contact with his narcissistic hurt, his fear of being inferior and despised, etc.

Another aspect of projective identifi cation, the phantasy of intrusively being able to get inside
the object – is illustrated by his omnisciently ‘knowing’ what is in my mind: he ‘knows’ that I
cannot see properly and he also

‘knows’ that this makes me feel inferior to him.

There is something else going on though, which I think has to do with pathological introjection
rather than projection. How has he acquired these omniscient (100 per cent vision) malevolent
eyes and whose eyes are they? I suggest that these were originally my perceptive and therefore
threatening analytic eyes, infl ated by idealisation. I am shortsighted not only because I now
contain the projection of his 140

Who’s who? Notes on pathological identifi cations lack of insight – and his narcissistic inability
to see as far as the other person – but also that my eyes that could see into him have been
concretely incorporated by him. In this defective, concrete introjection, if he has the ‘analyst’s
eyes’, then I obviously do not have them any more. In other words, if we assume that I made
some interpretation yesterday that revealed something to him that he had not been able to see
before, and by interpreting made him aware that I could

‘see’ (had good ‘eyesight’ and was interested in him), he perhaps did not take this in a healthy
introjective way which would make it possible for both of us to see, but instead took over my
function concretely. He acquired my capacity to describe some aspect of himself or some
situation in his internal world, rather than taking in my description of what I thought I was
observing in him, so that the interpretation couldn’t be used to further his capacity to think about
himself. Instead, he became the Me who can see into somebody else’s mind.

(In normal identifi cation, I introject your perceptive eyes, they are now felt to be symbolically in
my mind and, through identifi cation, I may then be able to see like you see, but you remain the
owner of your eyes. And since this is a benign interchange, the relationship remains one of
mutual co-operation. In pathological identifi cation, not only do I become the sole possessor of
the eyes because of a failure in symbolisation, but also the relationship is now dominated by a
struggle for power in which omniscient knowledge acts as a barrier against insight.) The person
who arrives in the consulting room is now an ‘analyst’

(or rather, a caricature of one) whose primary concern is to look inside the other’s mind and
reveal what can be seen there to a disturbed ‘patient’ me, but in a cruel, humiliating way.

This is then what is described as ‘being in massive projective identifi cation with’ the object.
This cruel, self-righteous, omniscient person lying on my couch is my patient in a state of total
projective identifi cation with . . .

me! A rather distorted (I hope!) version of his analyst. And this is what it feels like to be seen
through ‘my’ eyes, which are now my patient’s property: it is to be seen as inferior, vain, blind.
If this is what has happened, then the analyst in my patient’s mind is defi nitely a bad object and
a very powerful one. I suggest that this particular brand of badness – cruel omniscience – is the
product of an idealisation of a hated but also envied capacity of the object.

The feared perceptive eyes are certainly a very desirable attribute, 141

Ignes Sodré

which is why they get stolen. No shame or guilt are apparent in this process, only manic triumph.

In other sessions, the process may happen slightly differently. An object with helpful eyes may
be temporarily introjected – sometimes he can feel helped by an interpretation and feel some
relief at being understood – but then the separation at the end of the session may cause an
upsurge of hostility due to the pain of jealousy or envy or to an increase in persecutory anxiety.
In his phantasy, if he takes in what I give him he will lose his defences, will become dangerously
dependent, etc. (In this case there would be hostile projections into the internal object and the
perceptive eyes, now transformed into cruel eyes, have to be stolen as if they were a weapon to
be stolen from an enemy.)

Miss B: the loss of parts of the self

I will now bring an example from a patient who can get into massive states of projective identifi
cation with a bad object, but who does so much more temporarily. She is a very fragile,
borderline young woman, whose identifi cations shift rather quickly, producing a sense of
fragmentation and of loss of a sense of identity (I am very grateful to Richard Rusbridger for
allowing me to use his clinical material).

The previous weekend had been extremely distressing for Miss B.

Her boyfriend, C, a pop star musician, had been on tour around the country for several weeks and
was coming back to visit her. She had been waiting for his arrival in an eager but also rather
desperate mood. He arrived in a manic state, very much the star, made absolutely no emotional
contact with her, found it intolerable when she started clinging to him, and fi nally left with his
friends for an excited, drunken night out, leaving her behind in a distraught condition.

Throughout the following week Miss B was in a very bad state, on the brink of completely
falling apart. What follows is a summary of the following Monday session.

She starts talking, hesitating, and in a very croaky voice, ‘Last night I just cancelled everything,
and went out and got drunk, and had quite a nice time, and at one point felt much better about
everything in a drunken way, and kind of went round the place’. She then said,

‘It was really strange, because I thought I would do some work, but

. . .’ and went on to describe meeting several people for drinks in 142

Who’s who? Notes on pathological identifi cations what seemed to be a very excited, possibly
dangerous, way which seemed to the analyst to be exactly how she had described her boyfriend’s
activities in the previous weekend.

The narrative was punctuated with comments like, ‘I’ve erased everything from my mind’. At
some point in the session she exclaimed, ‘I am not afraid of C [the boyfriend] any more!’

It seemed clear to her analyst that her way of ‘cancelling everything’ was via getting into a state
of massive projective identifi cation with the manic boyfriend (in the transference a cruel
weekend analyst). The patient in the room seemed to have come out of that state, now felt to be
in the past, so that she seemed able to listen to his interpretations. But there was an interesting
misunderstanding at one point. The narrative about the night’s events had started with her
explaining that she ‘had driven John, a boy I know, home to X’

(a place quite far away); it ended with her driving around very late at night, until she was ‘fl
agged down by two chaps’ and she had driven one of them ‘home’. The analyst asked her if she
meant that she had driven the man to his home and she answered, as if it was obvious, that she
meant her own home.

The analyst was rather alarmed at this, feeling that his patient had been putting herself through a
really dangerous experience, and took this description of the end of her manic night out to mean
her acting out an identifi cation with an unfaithful, promiscuous boyfriend/analyst.

Miss B made it clear then that this was not the case at all, that she had recognised a friend, Paul,
in the road and that it had been helpful to have him at home, had made it possible for her to
sleep. She then explained that ‘cancelling everything’ had begun as trying to stop a terrible pain
in her back, and then it had become exciting to feel so very strong; but at the end of the night she
had felt terrible about

‘total disengagement’. It became quite clear, then, that for this patient, the state of massive
projective identifi cation with a manic bad object starts with ‘driving away’ to some far away
place some part of herself and that she can only come ‘home’ (to her house, to her own identity,
and also to her session) if she takes back inside parts of herself that have been fragmented and
spread ‘around town’, as it were. So whatever actually happened in the previous night in external
reality, what we have in the session is a narrative that gives a particular shape and meaning to
psychic events.

This patient is on some level able to communicate to her analyst the temporary loss of contact
with essential parts of herself that have 143

Ignes Sodré

to be recovered so as to enable her to go ‘home’, that is to say, to recover some sense of who she
is. Her state of projective identifi cation with the manic object is only temporary and threatens
her with a loss of her sense of identity. Ultimately she knows that this powerful manic person in
the night is not her real self. This is a temporary state, a defence that becomes threatening. I am
not suggesting, of course, that this patient is suddenly ‘cured’ of her need to relate to her object
through pathological projective identifi cation. For instance, I suspect that even though she could
come to the session and take in her analyst’s interpretations, which involves a relationship with a
less malevolent object and some awareness of a need to be helped to be more in contact with
herself, her dependency still resides partly in this other, more receptive object. She mentioned at
some point how much time she spends looking after others and there was a distinct sense in the
material that the analyst in the here-and-now of the session is ‘fl agging her down’ with his
attention and his comments and that her listening to him is probably coloured by her

‘helpfulness’

to him. But it is clear that the objects involved in this interchange are in fact kinder and saner,
and that she is consciously aware of her need to be ‘reconnected’ again.

In my patient, Mr A, who was also identifi ed with a manic object, these states are much more
infl exible. He is much more solidly identifi ed with the idealised bad object and there is great
commitment to keeping things this way. The equilibrium of the personality depends on
maintaining this identifi cation, and splitting mechanisms are constantly used to prevent any
awareness of weaker, dependent parts. Mr A’s pathological identifi cation is more or less
permanent and when this equilibrium is threatened his reaction is paranoid. Miss B’s projective
identifi cation with the object, although extensive, is only temporary.

She is much more fragile, her defensive solutions do not last and the state of ‘becoming’ the
object very quickly becomes a threat in itself. It is as if Mr A is in possession of the object, has
taken it over; whereas Miss B seems possessed by the manic object. She never entirely loses her
awareness that she has been invaded by something alien to her. Or perhaps one could say that she
does not idealise the bad object in the same way Mr A does, and her identifi cations shift. In the
presence of her sensitive analyst she is also projectively identifi ed with a helpful parent who
picks people up, drives them home, etc.

I will now bring another example from Mr A to illustrate the technical diffi culties the analyst
may feel confronted with when an object 144

Who’s who? Notes on pathological identifi cations relation that seems very fi xed and
unchangeable is dominating the transference. I fi nd it useful to refer here to Joseph’s careful
exploration and development of Klein’s concept of transference as a ‘total situation’. I think this
concept can help us to keep in mind the fact that a whole mode of functioning between two
people is being repeated in the situation in the session. (I am of course taking it for granted that,
as the analyst, one must always try to differentiate between what is being projected and the effect
this has on oneself, which is due at least partly to one’s own psychological make-up.) Mr A, who
is by profession a journalist, wrote a novel and sent it to a well-known publisher. He received a
letter of rejection from one of the directors of the publishing house (a writer himself). His
reaction in the session was one of moral indignation and contempt.

It was absolutely clear that this director had been motivated in his action by envy of Mr A’s
superiority as a novelist.

As soon as I took up what I thought was Mr A’s terrible hurt and disappointment at this rejection
he became enraged with me, clearly feeling that I was trying to project into him feelings that
were absolutely not his. I seemed to have become the publisher who was rejecting (refusing to
publish, as it were) his point of view. Soon an impasse was created in the session. I felt I had
either to accept his version of the situation or I would become entirely identifi ed with the
publisher in his mind and not only suffer a barrage of hatred and contempt but also, as the

‘enemy’, be entirely unable to help my patient. I began to feel more and more trapped in a
situation in which I had either to remain silent or agree with what seemed to me a rather mad
version of events; that the only conceivable reason for being rejected is that one is far superior to
the rejecting person. (As with the ‘eyesight’ situation in the fi rst example, what seems so
disturbing in these states is his certainty about the state of mind of the object.) It would seem that
Mr A has projected his envy of a creative parent who can produce a viable baby into the

‘publisher’, an aspect of me as a parent whom he sees as wanting to thwart his creativity:

possibly of course an internal object created originally by introjecting a disturbed parent. But is
this what happens in the session? I did not feel envious of my patient, I felt isolated as if I had
lost any hope of ever getting through to him. It was impossible for me to ‘publish’ my thoughts
(for instance, about the pain of being rejected, the defensive nature of his superiority, and his
hatred of me as a cruel publisher trying to put him down).

145

Ignes Sodré

Thinking about the session afterwards, though, I became aware that all the interpretations I could
think of were really aiming at changing the situation by reversing it: either his version is
published, or my version is. No wonder we didn’t get anywhere!

These are really diffi cult situations to get out of and often only through thinking carefully about
it afterwards can one begin to visualise what actually took place, without having to be either
victim or aggressor, which is what one undoubtedly is (and not only in the patient’s phantasy)
when trying to deal with projections by (unconsciously) re-projecting them. What I am talking
about is a necessary shift in the analyst to a position from which it would be possible to observe
what is happening in the interaction between these two people in the session. From this position
it becomes more possible to see who is who and what is the object relationship which is being
enacted in the transference. In this case, this could be seen to be one between somebody who is
trying to get something through, something that absolutely must be seen to be of value, and
somebody else who is impenetrable, unreachable, who says ‘No!’ to any attempt at
communication. (This experience links closely to what I have learned about Mr A’s fi rst two
years of life, when his mother was severely depressed and withdrawn.) Mr A’s change of identity
gives me fi rst-hand experience, as it were, of contact with his internal object. By trying to
visualise the total situation, I have some hope of understanding his underlying despair and of fi
nding a way out of the impasse in which we could be trapped into only repeating and not
working through.

146

PARTTHREE

The plural psychoanalytic scene

Introduction

Elizabeth Spillius and Edna O’Shaughnessy

In Part Three we describe the views held by members of several

psychoanalytic schools of thought about the concept of projective identifi cation. Because it is
somewhat unusual for a concept to be so widely known as the term ‘projective identifi cation’
has come to be, we have asked ourselves several questions about how this has come about. What
is it about the concept that has aroused such interest? Are there any particular factors that
contribute to a receptive attitude towards the adoption of the concept, and what factors mitigate
against such receptiveness? Is it possible to ‘lift’ a concept from one psychoanalytic milieu and
use it in another without altering the concept or the receiving psychoanalytic school of thought in
the process? We have
looked at these questions with the help of colleagues from four psychoanalytic cultures: fi rst, the
Contemporary Freudian and Independent members of the British Psychoanalytical Society;
second, the views about projective identifi cation held by three schools of thought in Europe –
German psychoanalysts, Italian and Spanish psychoanalysts, and French-speaking
psychoanalysts; third, we look at the attitude of American psychoanalysts towards the concept;
and fi nally we examine the views of Latin American psychoanalysts.

149

SECTION1

The British Psychoanalytic Society

The views of Contemporary Freudians and

Independents about the concept of

projective identification

Edna O’Shaughnessy

Some of the reasons and some of the passions for using, or not using, the concept of projective
identifi cation are to be found in the history of the British Psychoanalytic Society. After the
Controversial Discussions (fi nely chronicled by King and Steiner, 1991) held during the Second
World War, the British Society divided itself into Classical Freudians, later known as the
Contemporary Freudians; a Middle group, later called the Independents, and a Klein group, the
Kleinians.

This division into three groups had its beginnings in a dispute over child analysis during the
1920s between Anna Freud in Vienna and Melanie Klein in Berlin. The dispute involved issues
of transference, unconscious phantasy, the super-ego and the Oedipus Complex.

Milton, Polmear and Fabricius (2004) describe how

Klein was glad to accept Ernest Jones’s invitations, fi rst in 1925 to lecture in the British
Psychoanalytic Society, and then in 1926 to live in London. Klein was grateful for the warm and
generally open-minded reception of the British somewhat distanced as they were from the
analytic fray of mainland Europe.

(Milton et al., 2004, p. 64)

They record the terrible circumstances that in 1938 forced Sigmund and Anna Freud to come to
London as refugees from the Nazis –

153

Edna O’Shaughnessy
books burned, Freud himself suffering from advanced cancer of the jaw. About the Classical
Freudian group they write:

From 1933 many European refugee analysts were helped to settle in London . . . the emigres
closest to Freud and his daughter . . .

felt they had to protect the legacy of their mortally ill friend and teacher. For them Klein’s ideas
were not psychoanalytic at all; she was heretical in an analogous way to Jung or Adler.

(Milton et al., 2004, pp. 64–65)

But ‘Klein and close supporters . . . felt deeply that their developments were true to Freud’s
vision, and were

dismayed at the prospect that they might even be expelled from the British Society’ (Milton et
al., 2004, p. 65).

Expulsion would have been a grim reversal of fortune for Melanie Klein, whose ideas had
interested British colleagues such as Ernest Jones, Edward Glover (for a time), Ella Sharpe,
Marjorie Brierley, and Sylvia Payne. Joan Riviere and Susan Isaacs eventually became Klein’s
close colleagues, and so for a time did the young Donald Winnicott. All of these analysts were
making fresh discoveries.

Thus, on one side there was a battle for the scientifi c right of analysts to make clinical advances
and develop new concepts, and, on the other, the Classical Freudians fought for the preservation
of psychoanalysis as a heritage that had been cruelly lost in Europe.

Eric Rayner explains in his book The Independent Mind in British Psychoanalysis (Rayner, 1990)
that ‘The majority of the British Society, particularly those who had been its original members,
wished to take sides with neither group’, and he stresses the Independents’ pride in origins that
go back to the founding members of the British Society (Rayer, 1990, pp. 2 and 45). In The
British School of Psychoanalysis: The Independent Tradition (Kohon, 1986) Gregorio Kohon
describes how ‘a Middle Group was created: the Society remained one, but divided into three
separate groups [. . .] It was another characteristic achievement of the British Psychoanalysts,
accomplished through their remarkable capacity for compromise’ (Kohon, 1986, p. 45).

Once the three groups had formed themselves during the 1940s and 1950s, each tended to see
itself in ways that were discordant with how it was seen by the other two, and psychoanalytic
questions, sometimes mixed with personal animosities and training and institutional problems,
were oppositional and militarised. While both Independents and Kleinians paid special attention
to object relations, Independents 154

Views of Contemporary Freudians and Independents focused on pathology caused by external

objects. From a Kleinian perspective, this was a neglect of inner sources of disturbance.

Independents, in their turn, saw Kleinians as over-focused on the inner world, especially on
innate destructiveness, and as failing to take into account real life experiences. Both Kohon and
Rayner see Independents as ready to add to their own tradition with ideas from Anna Freud and
Melanie Klein, and from elsewhere in the world too; Independents prized their variety of
orientations as against the more unifi ed approach of the other two groups, particularly the
Kleinians. But the other two groups tended to see Independents as relying on a set of too loosely
related ideas and seeking an identity through what they were not.

Yet, in this anxious and divided institution, with its mix of facts and shibboleths about its groups
and its members, there remained a will to stay together, and, moreover, there came a remarkable
fl owering of psychoanalytic ideas.

Kleinian papers were specially written for the series of Scientifi c Discussions on Controversial
Issues by Susan Isaacs, Paula Heimann and Melanie Klein.

37 Among the

Classical Freudians, Anna Freud, Dorothy Burlingham, Willi Hoffer and Edward Glover were all
writing papers, as were Ernest Jones, Ella Sharpe, Marjorie Brierley and Sylvia Payne, soon to
be followed by Michael Balint, William Gillespie, Donald Winnicott, John Bowlby and John
Klauber among the Independents. And Melanie Klein was formulating a new theory of psychic
development (see

Spillius, Chapter 1 in this book), in which she stated the hypothesis

of a depressive position fi rst (Klein, 1935, 1940), following it with an account in ‘Notes on
some schizoid mechanisms’ (Klein, 1946) of an earlier paranoid-schizoid position in the fi rst
three months of life. It

was in this latter paper, reprinted in Chapter 2 of this book, that Klein introduced the term
‘projective identifi cation’

as the name of

an early defence mechanism of the ego in its struggles with anxiety. 38

37 These discussions, now generally known as the Controversial Discussions, took place in the
British Society from 1941 to 1945, and have been impressively discussed in King and Steiner
(1991), where the papers are collected.

38 Though neither refers to the other, it is of particular interest that Majorie Brierley (1945) had
used the same term,

‘projective identifi cation’, in a sense not so different from Klein’s the year before; Brierley
repeated it in her paper of 1947, though thereafter she did not pursue

this line of thought. See Spillius, Chapter 1 in this book. The very fi rst use of the term

‘projective identifi cation’ would seem to be in 1925 by Weiss, who also took it no further
(Weiss 1925). See an interesting exchange of letters between G. B. Massida and R. Steiner
on the ‘origins of concepts’ (Massida, 1999). See also Spillius, Chapter 1 in this book.

155

Edna O’Shaughnessy

Klein’s colleagues, Joan Riviere (Unpublished), Herbert Rosenfeld (1947, 1949, 1971) and
Hanna Segal (1950), quickly saw how the phenomenon of projective identifi cation might
advance the understanding of patients, and Paula Heimann (1950), then still a member of Klein’s
group, read a much-cited paper ‘On countertransference’

at the Amsterdam Congress, in which, without using the term

‘projective identifi cation’, she stated: ‘From the point of view I am stressing, the analyst’s
countertransference is not only part and parcel of the analytic relationship, but it is the patient’s
creation , it is part of the patient’s personality’, and further, ‘The analyst’s countertransference is
an instrument of research into the patient’s unconscious’

(Heimann, 1950, pp. 83 and 84, italics in original). However, Melanie Klein herself had
misgivings about the idea of countertransference, fearing that the analyst might confuse what he
felt for reasons of his own with what his patient felt. As we shall see, the clinical use of
countertransference remains a controversial issue.

The decade after the formation of the three groups that preserved the unity of the British Society
was full of troubles.

Pearl King observes that at that time ‘it was painful for those Members and students who became
involved in the inter-group tensions’ (King and Steiner, 1991, p. 908). It must have been painful
for everyone. In 1956 Paula Heimann left Klein’s group to join the Independents. Regarding
projective identifi cation, she eventually adopted a usage at odds with standard Kleinian usage.
‘Projective identifi cation’ she wrote ‘occurs as a countertransference phenomenon when the
analyst fails in his perceptive functions’ (Heimann, 1969). Margret Tonnesmann discusses
Heimann’s changing views in her Editor’s Introduction to Heimann’s collected papers
(Tonnesmann, 1989).

Meanwhile, British Kleinians continued to develop the concept.

Rosenfeld explored its intrusive aspects, especially in the psychotic and borderline transference,
Segal linked it to her work on symbolic equations and concrete thinking, Money-Kyrle examined
its varying roles in countertransference, Bion made it central to his new formulations about the
development and failure of communication and thinking; Joseph explored it clinically in
transference and countertransference. Projective identifi cation became an integral part of
Kleinian theory and technique.

On the world scene in the 1960s and 1970s psychoanalysis was changing. In the USA the school
of ego-psychology was beginning to lose its exclusive predominance. Non-medical analysts
gained 156
Views of Contemporary Freudians and Independents admission to the American Psychoanalytic
Association in 1988. The work of Loewald, Kernberg, Schafer and Ogden, infl uenced by contact
with the two British object relations schools of Klein and Winnicott, contributed to a new
complex plural scene (see Roy

Schafer, Chapter 14 in the present volume). Relational and interper-

sonal groups were legitimised. Kohut founded the American school of self psychology. In the
British Society the groups became less sectarian. Anna Freud (1967), for instance, in a paper
‘About losing and being lost’, described these phenomena in classical Freudian terms but also
mentioned a link to Klein’s idea of projective identifi -

cation. Sandler and Dreher (1996) offer an overview of these events.

By the 1980s Wallerstein could ask the question ‘One psychoanalysis or many?’, and give the
answer ‘Many’ – a

‘Many’ with relations to the ‘One’ of clinical practice (Wallerstein, 1988). And with the change
to pluralism in the 1970s and 1980s came a growing interest in Kleinian ideas. Martin Miller
(Independent) charted the occurrence in journals of the concept of projective identifi cation and
found that while it is less frequently referred to in American journals, the trend is the same: from
the 1970s on, the term ‘projective identifi cation’ occurs with increasing frequency in UK and
US journals (Martin Miller, personal communication).

In 1984 Joseph Sandler, who at that time held the Chair of Psychoanalysis in Jersualem,
organised a conference

‘because of the current interest in projective identifi cation’. Speakers were interna-

tional: Betty Joseph, a London Kleinian (see Chapter 6 in this book),

W. W. Meissner and Otto Kernberg from the USA and Rafael Moses from Israel. The conference
proceedings were published under the title Projection, Identifi cation, Projective Identifi cation
(Sandler, 1987b) with Sandler himself adding an important paper on ‘The concept of projective
identifi cation’ (reprinted here in Chapter 10 ).

Sandler’s stated aim was to make projective identifi cation ‘comprehensible to those lacking a
Kleinian psychoanalytic background’

(Sandler, 1987b, p. xi). It is instructive to see how he does this. In a scholarly manner he
discusses the wide, and still widening, range of phenomena that the concept is used to refer to,
and also its affi nities with some Contemporary Freudian thinking: how Anna Freud (1936) in
The Ego and the Mechanisms of Defence had a similar idea about

‘identifi cation with the aggressor’ and ‘living through another person’ as ‘a form of altruism’
(A. Freud, 1936, pp.

117 and 132). He also links projective identifi cation to his own clinical fi ndings about the
wishes 157

Edna O’Shaughnessy

of patients to actualise phantasies with their analysts (Sandler, 1976a).

He has a struggle with what he calls ‘concrete terms’ in Mrs Klein’s formulations in a way a
Kleinian might not have, because for a Kleinian such ‘concrete terms’ are readily clarifi ed by
other Kleinian concepts, like the paranoid-schizoid position, Segal’s (1957) differentiation of
symbolic equations from symbols, and Rosenfeld’s (1952a) notion of intrusions. Sandler writes
in ‘The concept of projective identifi cation’: Let me stress, however, that together with the
Kleinians, I believe that the processes of projective identifi cation (in the sense in which I can
make use of them) play a highly signifi cant part in development and in the clinical
psychoanalytic situation.

Projective identifi cation has given an added dimension to what we understand by transference,
in that transference need not now be regarded simply as a repetition of the past. It can also be a
refl ection of fantasies about the relation to the analyst created in the present by projective
identifi cation and allied mechanisms.

(Sandler, 1987a, p. 20)

I think Sandler’s account of the actualisation of role-relationships contributed both to the further
understanding of projective identifi -

cation itself, and also to the success of his aim of making projective identifi cation
‘comprehensible to those lacking a Kleinian psychoanalytic background’ (Sandler, 1987a, p. xi).

Sandler also addresses another problem, the problem of taking a concept into a different frame of
reference. He writes, ‘I believe it is mistaken to assume that the idea of projective identifi cation
as a mechanism is a part of a package which includes a theory of development and which has to
be accepted in its entirety’ (Sandler, 1987a, p. 19, italics in original). Unless in one gulp Sandler
were to become a Kleinian and so lose his identity as a Contemporary Freudian, some conceptual
unlinking must be done: the mechanism of projective identifi cation, in Sandler’s phrase, needs
to be ‘fruitfully separated’

from its Kleinian theory. How otherwise can an analyst of one persuasion take ideas from
another?

Indeed, a willingness to do just this has always been part of the Independent tradition that has
allowed them to experiment with concepts. In his book Freedom to Relate , Roger Kennedy, an
Independent analyst, discusses this issue in a chapter called ‘Are psychoanalytic concepts too
rigid?’ (Kennedy, 1993). He shows how Freud’s central 158

Views of Contemporary Freudians and Independents concepts had a fl exibility of use that could
accommodate new discoveries and also how Freud’s new uses remained alongside his old.

Kennedy writes:
There is, in fact, a problem about making any psychoanalytic defi nition: there are so many
different analytic assumptions that it is diffi cult to be certain of making any commonly accepted
statement, and opinions on theory and practice differ widely. This could be interpreted as
evidence of psychoanalytic chaos, but it may also indicate a rich variety of approaches, or of
different ‘voices in the psychoanalytic conversation’.

(Kennedy, 1993, p. 34)

And, in fact, Contemporary Freudian and Independent voices have participated in the
‘conversation’ about projective identifi cation, to different degrees and in different ways. Pearl
King (1974), Denis Carpy (1989) and Eric Rayner (1992), all Independents, make reference to
projective identifi cation in connection with transference and countertransference. In more
extensive contributions, Michael Sinason (1993, 1999) examines its role in psychosis, as does
Paul Williams in a book written with Murray Jackson,

Unimaginable

Storms ( Jackson and Williams, 1994), and also Brian Martindale, for example, in his paper on
‘New discoveries concerning psychoses and their organizational fate’ (Martindale, 2001).
Michael Brearley, in an unpublished paper,

‘The psychoanalyst’s neutrality’, uses the notion of projective identifi cation for a fresh
exploration of the concept of analytic neutrality. In her writings Dana Birksted-Breen has linked
the concept to her themes, especially in ‘Time and the après-coup’

(Birksted-Breen, 2003).

I also made some informal enquiries by email of Independent and Contemporary Freudian
colleagues who tend to participate in the exchange of ideas in the British Society. Some did not
reply. Those who did, whether by email or in conversation, offered a range of views. A few said
they never used the concept, but worked in a different way with their preferred notion of
projection. Several others said they had trained in the British tradition and so had been infl
uenced by groups other than their own, and sometimes used projective identifi cation in work
with more disturbed patients, or to understand a primitive defence, or in supervision to help a
supervisee recognise a countertransference experience. By far the majority of my respondents
159

Edna O’Shaughnessy

said they found projective identifi cation an essential concept for clinical work and supervision,
and for the

understanding of groups and strange psychic phenomena. ‘Base my work on it’, ‘couldn’t work
without it’,

‘embedded in my thinking’ were frequent phrases. Even so, with only one exception, every
respondent, from the most negative to the most enthusiastic, was critical of the Kleinians’
clinical use of projective identifi cation –
criticisms I come to later.

These replies from colleagues were lively. Many refl ected on the concept itself, with
illustrations from their clinical experience. They questioned whether ‘projective identifi cation’
was too wide a term and whether the phenomena it denotes were intrinsically yoked or disparate.
They asked: What is the relation of projective identifi cation to its components of splitting,
projection and identifi cation? Can there be projection without an object, or projection without
identifi cation?

Is projective identifi cation both intra-psychic and inter-psychic?

Several suggestions were made to limit or to redefi ne the concept in one way or another, or to
introduce some additional new term.

These discussions, many at some length, by Contemporary Freudians and Independents show, as
do their books and papers, that there are indeed gains to be had from the ‘fruitful separation’ of a
concept from its home theory. There are also pitfalls. I think we can live with terms having
different uses, which, as Roger Kennedy points out, is not untypical of psychoanalysis in either
classical or recent more plural times. What is more serious is that misunderstandings of the
original concept can occur through the undoing of signifi cant connections with its home theory.
One frequent misconception in the answers to my enquiry was that projective identifi cation, as
introduced by Klein in ‘Notes on some schizoid mechanisms’ (Klein, 1946), was essentially an
aggressive mechanism. In her paper Klein fi rst describes the ego’s projection of hated parts of
itself and how after this event ‘the hatred of parts of the self is now directed towards the mother.’
She continues, ‘This leads to a particular form of identifi cation which establishes the prototype
of an aggressive object-relation. I suggest for these processes the term “projective identifi
cation”.’ But she goes on to say

It is, however, not only the bad parts of the self which are expelled and projected, but also good
parts of the self.

Excrements then have the signifi cance of gifts, and parts of the ego which, together with
excrements, are expelled and projected into the 160

Views of Contemporary Freudians and Independents other person represent the good, i.e. the
loving parts of the self

[. . .] The projection of good feelings and good parts of the self into the mother is essential for
the infant’s ability to develop good object-relations and to integrate the ego.

(Klein, 1946, pp. 8–9)

That is to say, classically, ‘projective identifi cation’ was introduced by Klein to name processes
both hostile and loving.

The nature of Klein’s overall opus may lend itself to misunderstanding. In her writings of the
1920s and early 1930s, one of her emphases had been on aggressive infantile phantasies of
intrusion into the mother, and these form part of the later concept of projective identifi cation.
However, by the time she introduced projective identifi cation in 1946

Klein had formulated a new theory of psychic development, at the core of which was the ego’s
anxiety for the survival of both itself and its objects, and its anxious vicissitudes with love and
hate, depression as well as persecution, problems of guilt and reparation as much as aggression
(Klein, 1935, 1940). Indeed, immediately preceding her introduction of the concept of projective
identifi cation Klein writes of the ego’s ‘vital need to deal with anxiety’ (Klein, 1946, p. 4). As I
see it, a common misunderstanding of Klein’s concept comes from the omission of the part of
her hypothesis which states that one of the driving forces of projective identifi cation is the ego’s
need to survive . And as Elizabeth Spillius describes in Chapter 1 , Klein’s stress

on the projection of good feelings was even more marked in her unpublished notes in the
Melanie Klein Archive

than in her published paper of 1946.

Respect for the integrity of a theory may contribute to a reluctance to make Sandler’s ‘fruitful
separation’. I think that the coherence of a theory in which terms have interlocking meanings is
sometimes a reason in the British Psychoanalytic Society for not using a concept from a group
different from one’s own. I have long thought that Winnicott’s concept of transitional space
(Winnicott, 1953) identifi es important phenomena. Yet I have never used the idea, because to
separate ‘transitional space’ from Winnicott’s theory of development and use it, or its several
cognate terms, in a Kleinian context that in some relevant respects is a theory opposite to
Winnicott’s, would so change the concept as to become a misuse of it. Winnicott’s trajectory
goes from illusion through transitional space towards object relations, while Klein’s
developmental path starts with 161

Edna O’Shaughnessy

object relations – intermixed with other states of mind. Again, the Contemporary Freudian term
‘individuation’

captures what happens as the depressive position sets in when projections are withdrawn and
objects start to be experienced as whole and separate. But I feel unable to borrow the term and
say of the onset of the depressive position that ‘self and object become individuated’, because the
Contemporary Freudian theory of original merging in which individuation has its home is so
different from Klein’s theory. Canestri

(see Chapter 12 in this book) examines some of these conceptual

problems in the plural state of psychoanalysis.

Longstanding unresolved scientifi c differences between the groups are yet other reasons why
Contemporary Freudians and Independents might not use the concept of projective identifi cation
or use it somewhat differently.

Contemporary Freudians and Independents tend to see the early ego as less active than Kleinians
do and conceptualise the connections between mother and infant with their own, different ideas.
A patient’s internalised psychic history could not then be seen by them as repeatable in the same
way via communicative projective identifi cations to an analyst whose countertransference could
reveal something of the nature of these fundamental links –

the early containments, or failures and projections from objects into infant, or the attacks by
object or infant on the communicative links. Balint (1950) and Winnicott (1971) for instance
offer other perspectives for clinical practice that are still alive today in the British Society. For a
comparison of Contemporary Freudian, Independent and Kleinian views on countertransference,
see Hinshelwood (1999). Rather than the notion of projective identifi cation, many
Contemporary Freudians and Independents prefer for clinical practice Pearl King’s (1978) idea
of ‘the analyst’s total affective response’. The related problem of the analyst disentangling his
feelings from those coming to him from the patient is also approached differently. In the
Kleinian tradition there are papers by Money-Kyrle (1956), Joseph (1978), Brenman-Pick (1985)
and Feldman (1997), while Fonagy (1991) and Parsons (2000) put forward other, distinctive
views.

I come now to Contemporary Freudian and Independent criticisms of the Kleinians and
projective identifi cation that were made in response to my email enquiry. First, there is the
charge of over-use, that Kleinians ‘explain’ as projective identifi cation all the patient’s
communications and all the analyst’s feelings, both of which, it was often insisted, are made up
of many different things that cannot be 162

Views of Contemporary Freudians and Independents reduced to one. Second, many respondents
contended (sometimes contradicting earlier appreciative remarks) that the concept of projective
identifi cation was an aggressive mechanism through which Kleinian analysts over-focused on an
assumed innate destructiveness of their patients. And third, a few Contemporary Freudians and
Independents thought the concept tempted Kleinians into making interpretations with
unwarranted haste in the ‘here and now’

and furthermore – a view elaborated in print by Pearl King (2004)

– that the concept of projective identifi cation is to blame for a thin style of ‘me-you’ analysing
that has become current in the British Society.

The fi rst thing to be said is that a Kleinian analyst, like any other, needs to question himself, his
relations to his patient, and his way of conceptualising. And we have also to remember that any
concept and any technique can be poorly used. Every orchard has its bad apples. However, while
I agree that not all a patient’s communications are made by projective identifi cation and that the
analyst’s countertransference too is formed by many things, I would point out that because of the
major and multiple roles of projective identifi cation in Kleinian theory, projective identifi cation
is bound to have much Kleinian use. I contest the charge that Kleinians use the concept to over-
focus on destructiveness. Nor is it true that projective identifi cation in itself implies a denuded
‘me-you’, ‘you-me’

technique of analysis. Klein wrote in 1952: ‘It is my experience that in unravelling the details of
the transference it is essential to think in terms of total situations transferred from the past into
the present, as well as of emotions, defences, and object relations’ (Klein, 1952b, p. 55, italics in
original). She added on the next page: It is only by linking again and again [. . .] later
experiences with earlier ones and vice versa , it is only by consistently exploring their interplay
that present and past can come together in the patient’s mind. This is one aspect of the process of
integration.

(Klein, 1952b, p. 56)

Hanna Segal, discussing the curative factors in psychoanalysis, reminds us of how very many
there are: a full interpretation will involve interpreting the patient’s feelings, anxieties and
defences, taking into account the stimulus in the 163

Edna O’Shaughnessy

present and the reliving of the past. It will include the role played by his internal objects and the
interplay of phantasy and reality.

(Segal, 1962, p. 212)

She adds: ‘though we cannot always make a full interpretation, we aim eventually at completing
it’ (p. 212). No analyst can address everything at once; each analyst – approaches differ – makes
his choice of what is urgent for now and what will have to wait for later.

Pearl King (2004) raises a further issue. ‘What has happened to psychoanalysis in the British
Society?’ she asks As I understand it the way the phrase ‘here and now’ is presently used by
many people in the Society from all three groups, i.e. many Kleinians, some Contemporary
Freudians and some Independents, they use it to refer to the actual relationship in the session and
what the analyst thinks the patient is doing to him.

(King, 2004, p. 127)

It seems to me that it is the doing involved in projective identifi cation

that arouses most unease (see Hinz’s excellent discussion in Chapter 11

of this book). That we – whether the ‘we’ is patient and analyst, individual colleague and
individual colleague, or the three groups of the British Society – that we psychically do things to
each other , as, of course, we do to people in our inner and external lives and also to society at
large, brings enormous anxiety and guilt. It is not merely that we have thoughts and phantasies
about, or exchange words with, our fellow human beings, but, to use Sandler’s notion,
actualisations occur: projections from one psyche intrude into another and affect it – leaving the
recipient feeling disturbed, inferior, manipulated into action, wrong, guilty, or anxious.
Conversely, projections can also bring relief, or a sense of affi rmation or gratitude or pleasure.
Bion’s theory of thinking and his writings on container-contained formulate the general
hypotheses of this area (Bion, 1962b). Over the years Betty Joseph (1989b) has illustrated the
anxieties and pressures that lead to ‘acting in’, the ‘doing’

in analysis that occurs between patient and analyst, while Michael Feldman (1997) describes the
many factors, including unresolved problems in the analyst, that lead to the analyst becoming
actually involved in ‘the complex

intrapsychic and interpersonal process which constitutes projective identifi cation’ (see Chapter 7
in this book).

Indeed, in

1915 Freud commented, ‘It is a very remarkable thing that the Ucs of 164

Views of Contemporary Freudians and Independents one human being can react upon that of
another, without passing through the Cs ’ (Freud, 1915b, p. 194). Projective identifi cation gives
a name to the way one psyche does things to, and for, another.

Well – where does this leave us after I, a Kleinian, have tried to answer some of the criticisms of
Contemporary Freudians and Independents? I suspect it leaves us where we were – in
disagreement. Disagreements, especially those between analytic schools, are complex, involving
plural theories and concepts, unsettled controversies about development and the causes of
pathology, disagreements over analytic technique, plus all the problems of verifying and
falsifying competing clinical claims about non-replicable analytic encounters that can be
approached in more than one way even from within a single school of analytic thought, though
this does not, as David Tuckett has put it, mean ‘that anything goes’ (Tuckett, 2005).

Over many years in the British Society, while some early differences have been settled others
have not, and I think this is due to yet another complication: our old group passions.

By tradition, analysts in the British Society belong to the same group as their training analyst. All
candidates go through the same training and can choose supervisors from any group. In addition,
all students now go to a third consultant, who must have a different theoretical orientation from
their fi rst supervisor, for ten consulta-tions on their work with their fi rst patient. We all end
with a combination of conscious, and also unconsciously infl uential, identifi cations, immediate
analytic and supervisory experience, plus further knowledge from case presentations of the work
of the other two groups. In this internal admixture are loyalties, hostilities, anxieties and guilts
connected with the groups — one’s own and the other two. Among things said or emailed to me
in response to my enquiries were profes-sions of exasperation with the notion of projective
identifi cation, even when the analyst used it himself/herself, along with the charge of Kleinian
over-use that I discussed above. One Contemporary Freudian explained: ‘Projective identifi
cation is the Kleinian fl ag.

That’s how non-Kleinians see it.’ When projective identifi cation is felt to hoist a group fl ag, it
means our three groups are not yet out of their old battlefi eld, what Riccardo Steiner (1985) has
called ‘the political militarization of ideas’ where there are winners, losers, triumphs and useless
guilts for having analytic ideas.

What can we learn from this account of a local institution, a psychoanalytic village whose
neighbours in the 1950s battled over concepts 165

Edna O’Shaughnessy
and partly defi ned their psychoanalytic identity by concepts? The concept of projective identifi
cation was born in those times and concepts still matter intensely – and not only for their scientifi
c use.

Projective identifi cation is a Kleinian concept, and sometimes it has an emblematic meaning that
wakes up memories of antipathies and clashes and opposed group loyalties and identities. Yet
times have also changed

– perhaps more than we all realise. There is freer thinking about psychoanalytic ideas and
clinical experience, and many in the British Society wish to discard habits anchored in
procedural memories about groups and historic personages, and younger analysts and students
have yet to make their voices – the voices of the future – known.

When a concept like projective identifi cation, because of its intrinsic value, gains a primacy over
others we have to worry about how we experience this. We harm ourselves if acceptance of a
Kleinian idea is seen as and/or becomes a Kleinian triumph rather than a contribution to a shared
Freudian enquiry. Even so, inequality of conceptual achievement, along with a scientifi c
evaluation, brings concerns about losing signifi cant Contemporary Freudian and Independent
ideas, and anxieties about the sources of psychoanalytic identity.

We all have to continue to work under the tensions intrinsic to pursuing one approach while
colleagues pursue another within the same institution, trying to remember also that analysts from
one orientation may gain from and contribute to another. For those with a liking for so doing,
there is the chance to engage with the plural aspects of the

contemporary psychoanalytic scene, local and world-wide, a formidable enterprise about

concepts and the psychic realities named by concepts. As we gain in understanding it is likely
there will be more resolution to our rival conceptualisations, though who knows when – or
whether – we shall eventually arrive at a universal language for psychoanalysis.

Meanwhile, in the British Society we are still trying to fi nd ourselves; we are again full of
troubles as we struggle with our reasons and our passions in a new epoch that has come upon us
so untidily.

166

10

The concept of projective identification 39

Joseph Sandler

The introduction of the concept of projective identifi cation by Melanie Klein in 1946 was set
against a rather confused and confusing background of literature on various forms of
internalization and externalization – imitation, identifi cation, phantasies of incorporation and
many varieties of projection. Projective identifi cation is a broad concept, as the following
description by Hanna Segal indicates.
In projective identifi cation parts of the self and internal objects are split off and projected into
the external object, which then becomes possessed by, controlled and identifi ed with the
projected parts.

Projective identifi cation has manifold aims: it may be directed towards the ideal object to avoid
separation, or it may be directed towards the bad object to gain control of the source of danger.

Various parts of the self may be projected in order to get rid of them as well as to attack and
destroy the object, good parts may be projected to avoid separation or to keep them safe from
bad things inside or to improve the external object through a kind of primitive projective
reparation. Projective identifi cation starts when the paranoid-schizoid position is fi rst
established in relation to the breast, but it persists and very often becomes intensifi ed when the
mother is perceived as a whole object and the whole of her body is entered by projective identifi
cation.

(Segal, 1973)

39 This chapter reproduces the text of Sandler, J. (1987). The concept of projective identifi
cation. Bulletin of the Anna Freud Centre, 10 , 33–49.

167

Joseph Sandler

In this context Segal explicitly treats projective identifi cation as a mechanism of defence, but
elsewhere she describes it as ‘the earliest form of empathy’, and as providing ‘the basis for the
earliest form of symbol-formation’

(Segal, 1973). Melanie Klein saw it as a vehicle for distinguishing ‘me’ from ‘not-me’ (Klein,
1946). Rosenfeld (1965) has described processes entering into psychotic states and comments
that ‘Melanie Klein described these primitive object relations and the ego disturbances related to
them under the collective name “projective identifi cation”.’ In the last 35 years, projective
identifi cation has become increasingly seen by Kleinian analysts as a central mechanism in
countertransference, and in this context Bion’s addition of the container-contained model has
played a specially important part (Grinberg, Sor, and de Bianchedi, 1977). Ogden (1979, 1982)
has emphasized the role of projective identifi cation as the pathway from the intrapsychic to the
interpersonal.

It is possible to cite many other ways in which the concept of projective identifi cation has been
employed, and it is clear that its

‘collective’ and necessarily ‘elastic’ nature must render any precise defi nition implausible. But it
must be equally

clear that the term projective identifi cation has carried with it an idea which has proved to be of
substantial value to a signifi cant group of analysts. It has become a central Kleinian clinical
concept, although it is also used by analysts of other orientations. However, it is a notion which
is diffi -
cult to discuss from a non-Kleinian perspective. This may be due in part to the fact that those
who use the concept tend to speak of it as a single mechanism, while in fact it is one which (like
so many others in psychoanalysis) shifts its meaning according to the context in which it is being
used. It has as a result acquired a certain mystique, with the unfortunate consequence that it is
sometimes either dismissed entirely or thought to be understandable only with special ‘inside
knowledge’.

This paper refers to some of my own attempts to get to grips with the concept, and I have no
doubt that I will be seen by some to have done violence to it. Because my own frame of
reference is in signifi -

cant respects different from that of the Kleinians, it has been necessary to break the concept
down in my own mind in order to digest and absorb it. For the sake of convenience I shall
present the material which follows under a number of different headings. Like all concepts in
psychoanalysis, that of projective identifi cation has undergone a progressive development since
its introduction, and it is convenient 168

The concept of projective identifi cation to separate three stages in this development. These
stages are, I believe, conceptually clear-cut, although they overlap considerably, and the ideas of
the fi rst and second stages have persisted alongside those of the third.

Some remarks on the background

In the period between the two world wars, and particularly after the mid-twenties, processes of
internalization and externalization became increasingly important in psychoanalytic thinking.
These processes were particularly evident in psychotic patients, and this led Melanie Klein to
construct and elaborate a theory of development in which object relationships were seen as being
built up on the basis of such internalizing and externalizing processes. She distinguished the two

‘positions’ in normal development on the basis of the two major psychotic states in which such
processes could be seen most clearly.

Although one might disagree with Mrs Klein’s theory of development and with the concrete
nature of her formulations, in retrospect we can see that she was the analyst who gave the earliest
and greatest recognition to projective and identifi catory processes in the development of object
relationships, and to their operation in the here-and-now of the transference. I believe that it was
clinical pressure which prompted this development, and in this context it is interesting to note
that Anna Freud has remarked, in her work on the mechanisms of defence (Freud, 1936), that she
was prompted to develop her ideas by the need to understand resistances in analysis more fully.
In relation to Anna Freud’s introduction of a number of new defences in 1936, I have
commented elsewhere that What Anna Freud did at this time was to introduce a whole class of
what might be called object-related defences, which involved reversal of roles or some
combination of identifi cation and projection. These are defences in which there is an active
interchange between aspects of the self and of the object, the unacceptable aspects of one’s own
self being dealt with by producing (or attempting to produce) their appearance in the external
object.

Often, simultaneously, frightening or admired aspects of the object may be taken into the self.
(Sandler, 1983)

169

Joseph Sandler

First stage projective identifi cation

When Melanie Klein introduced the term projective identifi cation in 1946, she said: Much of the
hatred against parts of the self is now directed towards the mother. This leads to a particular form
of identifi cation which establishes the prototype of an aggressive object-relation. I suggest for
these processes the term

‘projection identifi cation’. When projection is mainly derived from the infant’s impulse to harm
or to control the mother, he feels her to be a persecutor. In psychotic disorders this identifi cation
of an object with the hated parts of the self contributes to the intensity of the hatred directed
against other people.

(Klein, 1946)

Although this formulation was put by Mrs Klein in very concrete terms, it can be understood as
referring to processes which occur in phantasy processes of change in the mental representation
of self and object occurring at various levels of unconscious phantasy. The concreteness of the
formulations can be taken to refer to processes imagined as concrete, i.e. involving images of
literal incorporation or

‘forcing’ something into an object. The processes described served the function of being
defensive or adaptive in the immediate present, although when they occur in extreme form in
infancy they may have harmful effects on later development. Melanie Klein can be taken to be
referring here to shifts and displacements within the child’s representational world (Sandler and
Rosenblatt, 1962). Identifi cation with parts of an object can be regarded as a

‘taking into’ the self-representation aspects of an object-representation. Projection is then a

displacement in the opposite direction, i.e. aspects of the self-representation are shifted to (and
made part of ) an object-representation.

In my view it does not matter very much, in this context, whether we speak, for example, of

‘the bad parts of the self ’ or of ‘the hated (or despised) unwanted aspects of the self-
representation’, as long as we agree that the processes as described by Mrs Klein occur within
phantasy life. Moreover, in my view, for conceptual purposes we can separate the idea of
projective identifi cation as a mechanism operating in the here-and-now from processes
described in the Kleinian theory of early development.

I shall touch on this point again later, but I want to emphasize that what is of the greatest
importance for me in Mrs Klein’s formulation is 170
The concept of projective identifi cation her description of a mental mechanism or set of
mechanisms. For Melanie Klein projective identifi cation involved splitting, which I would
understand in this context as a splitting-off of parts of the self-representation or of the object-
representation. Projective identifi cation involves projection in that it is an identifying of the
object with split-off parts of the self. When the process moves in the opposite direction, i.e.

identifying oneself with aspects of the object, Mrs Klein speaks of introjective identifi cation.

I have referred here to fi rst stage projective identifi cation in order to emphasize the point that
for Mrs Klein projective identifi cation was a process which occurred in phantasy. Let me put it
another way. The real object employed in the process of projective identifi cation is not regarded
as being affected – the parts of the self put into the object are put into the phantasy object, the
‘internal’ object, not the external object. This is borne out by the way in which transference and
countertransference are treated in Mrs Klein’s writings. Transference refl ects infantile object
relationships (Klein, 1952b), and is a phantasy about the analyst which needs to be analysed. It is
a phantasy which creates a distortion of the patient’s perception of the analyst, a distortion based
on, among other things, the projective identifi cation which has affected the patient’s phantasy
about the analyst.

Countertransference, in its turn, is scarcely mentioned, and when it is (e.g. in ‘Envy and
gratitude’: Klein, 1957), it is regarded as a hindrance to the analyst’s technique.

Second stage projective identifi cation

It was probably inevitable that the concept of projective identifi cation came to be widened soon
after its introduction, and I want to touch here on a particular extension of the concept to object
relationships in general and to the transference-countertransference relation between patient and
analyst in particular. In 1950, Paula Heimann (then very much a Kleinian) drew attention to the
positive value of countertransference thoughts and feelings in analysis. She stated, ‘The analyst’s
countertransference is an instrument of research into the patient’s unconscious’, and remarked
‘From the point of view I am stressing, the analyst’s countertransference is not only part and
parcel of the analytic relationship but it is the patient’s creation, it is part of the patient’s
personality’ (Heimann, 1950).

171

Joseph Sandler

It is of interest that a number of writers made a very similar point at about this time, some linking
countertransference specifi cally with projective identifi cation. Thus Racker, in a series of
papers beginning in 1948

(Racker, 1968), connected the analyst’s countertransference response to projective identifi cation
on the part of the patient.

In regard to a case which he was discussing, he said:

[. . .] the ‘projective identifi cation’ . . . frequently really obtains its ends – in our case to make
the analyst feel guilty, and not only implies (as has been said at times) that ‘the patient expects
the analyst to feel guilty’, or that ‘the analyst is meant to be sad and depressed’. The analyst’s
identifi cation with the object with which the patient identifi es him, is, I repeat, the normal
countertransference process.

Racker makes a valuable distinction in this context between concordant and complementary
identifi cation on the part of the analyst.

To put it simply, countertransference based on a concordant identifi cation occurs when the
analyst identifi es with the patient’s own phantasy self-representation of the moment.
Countertransference based on a complementary identifi cation occurs when the analyst identifi es
with the object-representation in the patient’s transference phantasy.

Heimann, Racker and others, such as L. Grinberg, 1957 and 1985

(referred to in Grinberg, 1962), made a signifi cant extension of projective identifi cation by
bringing it into conjunction with the analyst’s identifi cation with the self- or object-
representation in the patient’s unconscious phantasies, and with the effect of this on the
countertransference. The countertransference reaction could then be a possible source of
information for the analyst about what was occurring in the patient.

I have called the formulations referred to above second stage projective identifi cation because
they represent an extension of Melanie Klein’s original propositions, whereas projective identifi
cation occurring within the person’s phantasy life (refl ected in a phantasy distortion of the
analyst), can be called fi rst stage projective identifi cation. If either the self or object represented
in such unconscious phantasies is identifi ed with by the analyst to a degree suffi

cient to contribute to the analyst’s countertransference, we have a second stage.

172

The concept of projective identifi cation Third stage projective identifi cation In this stage of
the development of the concept it is no longer one or other aspect of the unconscious phantasies
that is identifi ed with by the analyst.

Projective identifi cation is now described as if the externalization of parts of the self or of the
internal object occurs directly into the external object. This extension was given its main impetus
by work of Bion in the late 1950s and found explicit expression in his concept of the ‘container’
(Bion, 1962a, 1963).

Bion describes the function of the container by presenting what is essentially Melanie Klein’s
description of projective identifi cation, to which he adds the following (Grinberg et al., 1977):
One of the consequences of this process is that, by projecting the bad parts (including phantasies
and bad feelings) into a good breast (an understanding object), the infant will be able – insofar as
his development allows – to reintroject the same parts in a more tolerable form, once they have
been modifi ed by the thought (reverie) of the object . . . . (my italics).
By no stretch of the imagination can this be understood as occurring in phantasy only, nor is this
what Bion intended to imply. What he describes here is a concrete ‘putting into the object’. He
says: An evocation of the bad breast

takes place through a realistic projective identifi cation. The mother, with her capacity for
reverie, transforms the unpleasant sensations linked to the ‘bad breast’ and provides relief for the
infant who then reintrojects the unmitigated and modifi ed emotional experience, i.e. reintrojects

. . . a non-sensual aspect of the mother’s love.

Bion’s formulations can be related to Winnicott’s ‘holding’ function of the good-enough mother
(Winnicott, 1958), and are echoed too in aspects of Ogden’s conception (Ogden, 1982). What
Bion has proposed in the container-contained metaphor is clearly of great relevance for analytic
technique, and I shall return to this projective identifi cation later.

It must be clear that all the theoretical propositions put forward in connection with transference
and countertransference apply equally to object relationships outside the analytic situation.

173

Joseph Sandler

Comments

The comments which follow relate to various aspects of the projective identifi cation, and are
somewhat disjointed.

However, they will give an indication of my own view of projective identifi cation.

1. An integral part of Kleinian theory is that which relates to very early infantile development,
and to early object-relations. Later adult functioning is seen as rooted in this particular
developmental view. As a consequence, Kleinian formulations regarding adult mental processes
tend to use the same concepts as those which are used to describe infantile ones. However, I
believe that it is a mistake to assume that the idea of projective identifi cation as a mechanism is
part of a

‘package’ which includes a theory of development and which has to be accepted in its entirety.
Nor need we be put off by the concreteness of the metaphors used, for we all use metaphors in
our theory (and in our interpretations), and what is important is to be aware that we are using
metaphors. Consequently I have no diffi culty in absorbing Mrs Klein’s description of projective
identifi cation (fi rst stage projective identifi cation) into my own frame of reference, in which it
can be regarded as a mechanism involving shifts and displacements in mental representation or
in phantasy. I would put emphasis on its role as a mechanism for regulating unconscious feeling
states and emphasize, too, that this mechanism can be divorced if necessary from the specifi c
phantasy content associated with projective identifi cation by Mrs Klein and her followers. Let
me stress, however, that together with the Kleinians, I believe that processes of projective identifi
cation (in the sense in which I can make use of them) play a highly signifi cant part in
development and in the clinical psychoanalytic situation. Projective identifi cation has given an
added dimension to what we understand by transference, in that transference need not now be
regarded simply as a repetition of the past. It can also be a refl ection of phantasies about the
relation to the analyst created in the present by projective identifi cation and allied mechanisms.

2. The element of control of the objects (into which parts of the self have been projected) has
been consistently stressed by Kleinian writers on projective identifi cation, and it is an element
which is, I believe, central to the concept. What one wants to get rid of in oneself can be
disposed of by projective identifi cation, and through controlling the object one can then gain the
unconscious illusion that one is controlling the unwanted and projected aspect of the self. The
174

The concept of projective identifi cation urge to control the object is evident in the process of
‘living through another person’ – Anna Freud’s ‘altruistic surrender’ (Freud, 1936) can be taken
to be a good example of this process. But there is a further aspect of the wish to control the
object which is worth stressing. It is a common clinical observation that patients who feel guilty
(‘attacked by an internal persecutor’) may deal with the guilt and gain powerful narcissistic
supplies by projecting the guilty (‘bad’) part of the self on to another, while at the same time
identifying (by

‘introjective identifi cation’) with the persecutor. This provides a double gain, i.e. the gain of
identifying with the

(usually idealized) part of the superego introject as well as that of getting rid of the ‘bad’

unwanted part of the self. This gives a powerful motive for control of the object into which the
projective identifi cation has taken place (Sandler, 1960).

3. I have made use of the phrase ‘into the object’ rather than ‘on to’. For many years, of course,
projection ‘into’ was Kleinian usage, and non-Kleinians took care to speak of ‘projection on to’.
However, if we accept that projection is a process that involves self- and object-representations,
then we need have no diffi culty in accepting the phrase

‘into the object’. But this does not mean, in my view, that we have necessarily to accept the idea
that projective identifi cation is accompanied by phantasies of entering and invading. However,
the idea of ‘forcing’ an aspect of oneself into the object raises no diffi culty because projective
identifi cation as a mechanism of defence aims at reducing anxiety, and this is a strong motive
for applying force to keep the projected aspect on the other side of the self-object boundary.

This force is refl ected in the resistance shown by patients to accepting projected aspects of the
self back into the self-representation.

4. The role of the self-object boundary is clearly important in projective identifi cation and
deserves a few comments. For projective identifi cation to function as a mechanism of defence
the existence of a boundary between self and object is essential so that a person can feel
dissociated from the split-off parts of the self. In psychotic states it is diffi cult to maintain
representational boundaries so projective identifi -

cation may be intensifi ed as an attempt to establish the boundaries.

If they cannot be satisfactorily established then a state of panic may follow. It is worth noting
that although we, as observers, may see the massive use made of projective identifi cation in
certain psychotic states, for the psychotic the existence of a persecutor in his phantasies means
that he has (perhaps only temporarily) established a self-boundary 175

Joseph Sandler

of sorts. Here I fi nd myself in disagreement with those who tend to see projective identifi cation
as a psychotic mechanism only. While it may be massive in psychosis, it is nevertheless
ubiquitous; it might also be preferable, in line with this, to speak of pathogenic rather than
pathological projective identifi cation.

5. It has been stated (Klein, 1946) that projective identifi cation is a mechanism whereby the
boundary between self and object is established in infancy. This notion of projective identifi
cation is diffi cult to reconcile with the requirement of a self-object boundary for successful
projective identifi cation. I would suggest that the concept need not be applied to these early
differentiating processes, but, if it is, it could be understood as referring to the differentiation of
mental representations of self and object in a way which is different from projective identifi
cation as a mechanism of defence. What I mean is that we can conceive of attempts at projection
and identifi cation (or, perhaps better, identifi cation and dis-identifi cation) occurring as the
infant struggles to organize what has been regarded as a state of primary confusion between
experiences of self and object, and thereby to gain control over his feeling states. These attempts
can be taken to contribute, in their turn, to the gradual establishment of self and object
boundaries.

6. If we turn now to second stage projective identifi cation we must immediately be concerned
with the relation between a phantasy of the object (the object representation) as modifi ed by
projective identifi cation and its effect on a real external person. This is, as has been pointed out
by many, most evident in the dimension transference-countertransference. From the point of
view of my own frame of reference it is insuffi cient to say that the internal phantasy object is
‘put into’ the analyst. Rather, I understand the process in the following way: (a) A phantasy is
created, involving the analyst. Projective identifi cation enters into the creation of the phantasy,
which is a wishful one, i.e. it has behind it a pressure towards gratifi cation or fulfi lment.

(b) The patient attempts to actualize (Sandler, 1976a) the unconscious wishful transference
phantasies, to make them real, to experience them (either overtly or in disguised form) as part of
reality. He will do this in numerous ways in

the analytic situation, some of which will evoke a countertransference response which can be
meaningful to the analyst in the ways described by Heimann, Racker, Grinberg and many others.
In a previous paper (Sandler, 1976b), relating to 176

The concept of projective identifi cation actualization and object relationships, I commented on
the wishful phantasy as follows:

[. . .] it involves a self-representation and object representation, and an interaction between the

two. There is a role for both self and object. Thus the child who has a wish to cling has, as part of
this wish, a mental representation of clinging to someone else; but he also has, in his wish, a
representation of the object responding to his clinging in a particular way. Those role
relationships which appear in the transference are representative of the important wishful aspect
of the unconscious phantasy life. This is rather different from the idea of a wish consisting of a
wishful aim being directed towards the object. A notion of an aim which seeks gratifi cation
needs to be amplifi ed by an idea of a wished-for role interaction, with the wished-for or
imagined response of the object being as much a part of the wishful phantasy as the activity of
the subject in that wish or phantasy . . . it can be said that the patient in analysis attempts to
actualise the role relationship inherent in his current dominant unconscious wish or phantasy, and
that he will try to do this (usually in a disguised and symbolic way) within the framework of the
analytic situation . . . it is not a great step to say that the striving towards actualisation is part of
the wish-fulfi lling aspect of all object relationships. I do not use the term actualisation here in
the same sense as it is used by a number of other psychoanalytic authors, but quite simply in the
dictionary sense, that is, as a making actual; a realisation in action or fact.

This quotation relates to the evocation of a response in the analyst which refl ects in some way
the role of the object in the current wishful phantasy of the patient. This would correspond to
Racker’s (1968) notion of the analyst’s complementary identifi cation with the phantasied object
of the active internal object relationship. The concordant identifi cation, on the other hand, can
be said to be the analyst’s identifi cation with the self-representation involved in the patient’s
wishful phantasy. However, I want to suggest that countertransference response due to second
stage projective identifi -

cation is always based on identifi cation with a phantasy object, and that when it appears to be
identifi cation with aspects of the self-representation, a further intrapsychic step has occurred in
the patient’s phantasying process, i.e. a further projective identifi cation has taken 177

Joseph Sandler

place into a new phantasy analyst-object which then contains projected aspects of the self.

7. The central contribution to third stage projective identifi cation is that of Wilfred Bion, and the
container model is of substantial clinical and theoretical interest. What I understand by it, in my
own frame of reference, is the capacity of the caretaking mother to be attentive to and tolerant of
the needs, distress, anger as well as the love of the infant, and to convey, increasingly, a
reassurance that she can ‘contain’ these feelings and, at an appropriate time, respond in a
considered and relevant way. Through this the infant learns that his distress is not disastrous, and
by internalizing the ‘containing’ function of the mother (through identifi cation or introjection)
gains an internal source of strength and well-being. The ‘reverie’ of the mother is to be
distinguished from an immediate ‘refl ex’ response to the child. The latter process does not
require the identifi cation with the child’s distress in the same way as does the

‘reverie’ of the

‘containing’ mother.

As far as the analytic situation is concerned, there is a parallel with the description given for the
mother and infant.
The analyst as

‘container’ is, as I see it, the analyst who can tolerate the patient’s distress, hostility, love –
indeed, all his phantasies and feelings, and who as a consequence of his ‘reverie’ can return them
to the patient in the form of interpretations which will allow the patient to accept as aspects of
himself those parts which he had previously considered to be

dangerous and threatening, and which had been dealt with defensively, with ensuing cost. What I
fi nd unacceptable is the notion that this process is one of projective identifi cation, unless the
concept is stretched to extreme limits.

We would have to say, for example, that the child’s cry of distress is ‘put into’ the mother by
projective identifi cation, and it seems to me that this represents a caricature of the original
concept. The ‘container’ model can, I believe, be fruitfully separated from the developmental
theory to which it is attached, as well as from the concept of projective identifi cation (although
what the analyst will ‘contain’ will encompass the patient’s transference projections as well as
his distress), and has value in its own right.

In this regard the following comments (Sandler and Sandler, 1984) are relevant:

. . . to achieve what we regard as the aim of the analytic work we need to bring the patient to the
point where he can tolerate, in a 178

The concept of projective identifi cation safer and more friendly fashion, the previously
unacceptable aspects of himself. In order to do this he will need to gain insight, in an emotionally
convincing manner, not only into the content of his unconscious phantasies, but also into the
nature of what, for present purposes, we shall refer to as his

‘inner world’, i.e. his unconscious relation to his introjects, with whom he had a continual
unconscious internal dialogue . . ., his unconscious anxieties and confl icts as well as his methods
of resolving such confl icts.

This includes, of course, an understanding of his own usual defensive mechanisms and
manoeuvres, with particular reference to the projections and externalizations that occur in his
unconscious phantasy life – the spectrum of mechanisms that have come to be called projective
identifi cation. To the extent that we can achieve our analytic aims we will . . . be able to bring
about the reduction of confl ict and associated painful affects, and to permit a defl ection of what
was not previously tolerated near consciousness into conscious or preconscious thought and
phantasy

. . . it is our task to work with the patient in such a way that as much as possible of the content of
what has come close to the surface layers . . . can be made readily available to consciousness. In
our work we strive to bring about the liberation of such material through appropriate
interpretations, particularly the interpretation of confl ict. But because what we have available is
the adult form of the relevant unconscious content, in order to anchor the progressive mapping of
the patient’s inner world and the central and recurring themes in his present unconscious, we
have also to reconstruct the patient’s past in a relevant way, just as much as we have to make
construc-tions about his current inner world (which is, of course, a direct descendant of the inner
world he formed in childhood); and we link the two together.

8. Projective identifi cation has been regarded as the basis for empathy, but a simple statement of
this sort does not have a great deal of explanatory power without amplifi cation. I want to
suggest that the state of primary confusion between self and object which I referred to earlier in
this paper (usually called primary identifi cation) is one which persists in modifi ed form
throughout life, and which can provide the basis for the capacity for empathy. Some time ago W.
G. Joffe and I put forward the following formulation in a paper on the tendency to persistence
(Sandler and Joffe, 1967), and I have 179

Joseph Sandler

taken the liberty of quoting from it at length because of its relevance to projective identifi cation:
Identifi cation (we refer here to secondary identifi cation), as a number of authors now see it,
involves a change in the self-representation on the model of an object-representation; and
projection is the attribution to an object-representation of some aspect of the self-representation.
These processes can occur after the boundaries between self- and object-representations have
been created; before that, we have the state referred to by Freud as primary identifi cation,

‘adualism’ by Piaget and primary identity by others. A better term to designate this early state
might be ‘primary confusion’ [. . .] If we apply the idea of persistence to processes of identifi
cation and projection in the older child or adult we can postulate that there will always be a
momentary persistence of the primary state of confusion, however fl eeting, whenever an object
is perceived or its representation recalled. What happens then is that the boundaries between self
and object become imposed by a defi nite act of inhibiting and of boundary-setting.

It is as if the ego says ‘This is I and that is he.’ This is a very different idea from that of a static
ego boundary or self-

boundary which remains once it has been created. What develops [. . .] is the ego function of
disidentifying, a mental act of distinguishing between self and object which has to be repeated
over and over again; and the function of disidentifying makes use of structures which we can call
boundaries.

The persistence of this genetically earlier primary confusion in normal experience is evident
when we think of the way in which we move and tense our bodies when we watch ice skaters, or
see a Western. We must all surely have had the experience of righting ourselves when we see
someone slip or stumble. In these everyday experiences there is a persistence of the primary
confusion between self and object; and this may more readily occur in states of relax-ation or of
intense concentration in which the bringing into play of boundary-setting may temporarily be
suspended or delayed.

[. . .] The persistence of this genetically early state [. . .] must surely provide the basis for
feelings of empathy, for aesthetic appreciation, for forms of transference and countertransference
in analysis

[. . .] and in connection with what we call secondary identifi cation and projection, we would
suggest that the bridge to these processes 180
The concept of projective identifi cation is the persisting momentary state of primary confusion
or primary identifi cation which occurs before the process of ‘sorting out’ or

‘dis-identifying’ occurs. One result of this ‘sorting out’ may be that aspects of the object-
representation are incorporated into the self-representation and vice-versa.

The existence of fl eeting primary identifi cations after infancy can give us a tool for improving
our understanding of processes of projective identifi cation, for the notion of persistence allows
us to account for the fact that we must, in some way, be aware that what we have projected is our
own in order to feel the relief of being rid of it. I would suggest that in all forms of defensive
projection there is a constant to-and-fro, an alternation between the momentary state of

‘oneness’, of primary identifi cation or primary confusion, and the

‘sorting out’ referred to earlier. This would allow one to feel that what is projected is fl eetingly
‘mine’, but then reassuringly ‘not mine’.

I want to end these comments by expressing some concerns which I am sure are shared by
others. First, because projective identifi cation is more of a descriptive than an explanatory
concept, and because its range of meanings is wide, its use without further elaboration provides a
ready pseudo-explanation. Such pseudo-explanations are tempting, and we should be on our
guard against them. If projective identifi cation is used as an explanation, its specifi c meaning in
the relevant context should, I think, always be given. Second, because of the close link between
the concept of projective identifi cation and our extended understanding of countertransference,
it is tempting to see all feelings, phantasies and reactions of the analyst to his patient as being an
outcome of what the patient has ‘put into’ the analyst by means of projective identifi cation.

Unfortunately, the differentiation of what belongs to the patient and what to the analyst is likely
to remain with us for some time as a diffi cult technical problem.

181

SECTION2

Continental Europe

Introduction

Elizabeth Spillius

The three papers in this section were given at a meeting of the

European Psychoanalytical Federation at Prague in 2002. They were part of an attempt to

achieve greater clarity about the differences and agreements among European analysts on topics
of mutual interest.

Eike Wolff, the editor of the journal Psychoanalysis in Europe , introduced the topic of
projective identifi cation by saying: In the last fi fteen years, the term ‘projective identifi cation’
has expanded at an infl ationary rate and in some places has thrown concepts such as
‘transference’ and ‘transference neurosis’ into the background. Refl ection on the various uses of
the term in different European countries and analytical schools can strengthen the awareness of
the advantages and the dangers involved in extensive habitual use of the concept.

At the conference Elizabeth Spillius briefl y introduced the topic, followed by Helmut Hinz, who
described the reception of the concept in Germany, Jorge Canestri of the Italian Association, who
talked about the fate of the concept in Italy and Spain, and Jean-Michel Quinodoz of the Swiss
Society, who described the response of French-speaking psychoanalysts to the concept. All four
contributions were published in 2002 in Psychoanalysis in Europe: Bulletin 56

(Canestri, 2002; Hinz, 2002; J.-M. Quinodoz, 2002a; Spillius, 2002).

The papers of Hinz and Canestri are reproduced here, followed by Quinodoz’s 2003 version of
his paper, somewhat revised for inclusion in the present book.

185

11

Projective identification

The fate of the concept in Germany

Helmut Hinz

Reception, positive adoption, critical and aversive voices: German contributions In order to
outline the reception of the concept of projective identifi cation, I began by looking in the index
of key words to the psychoanalytic journal Psyche to see when this term fi rst started to appear in
articles there. It is fi rst mentioned in 1957 with the translation of Melanie Klein’s

Envy and Gratitude , followed in 1960 by Klein’s

‘Some theoretical conclusions regarding the life of the infant’ (Klein, 1952c). These in turn are
followed by Betty Joseph’s ‘Some characteristics of the psychopathic personality’ (Joseph,
1961), as well as A. Bonnard’s ‘Pre-body ego types of (pathological) mental functioning’
(Bonnard, 1961), and J. O. Wisdom’s ‘A methodological approach to the problem of hysteria’
(Wisdom, 1962). The fi rst German contribution to make use of the key term projective identifi
cation was published in 1962. It was by Wolfgang Loch and bore the title ‘Anmerkungen zur
Pathogenese und Metapsychologie einer schizophrenen Psychose’ (Loch, 1962). It was followed
by H. A. Thorner (1963). In 1963, Bion’s Theory of Thinking was made available in translation,
followed by another contribution by Wolfgang Loch on the subject of identifi cation-introjection.

During the period up to 1992, the term projective identifi cation appeared in a total of 40 articles
in the journal Psyche . In 1985,

R. Zwiebel wrote a report about a conference in Jerusalem organized 186

Projective identifi cation in Germany by Joseph Sandler on the theme ‘Projection, Identifi cation
and Projective Identifi cation’ (Zwiebel, 1985) and, in 1988, a fi rst original work by the same
author was published using the term projective identifi cation in the title (Zwiebel, 1988). This
paper offers a sound theoretical overview of the concept of projective identifi cation. It was
followed in 1989 by a clinical presentation by Hinz with the title: ‘Projektive Identifi zierung
und psychoanalytischer Dialog’ (Hinz, 1989).

As early as 1965, Wolfgang Loch was giving projective identifi cation a central place in his
lectures because its

interpretation could contribute to a ‘restructuring of the ego’ (Loch, 1965, p. 55) insofar as

‘strongly defended material’ and ‘split-off parts’ of the ego are brought into words, and this
‘strongly stimulates a new balance’ and a ‘better foundation of the ego’. In a 1975 lecture, in his
apt formulation, the analyst gets projected parts of the patient ‘put inside himself ’, and in such a
way that he is forced to take up these parts, and then, for example, feels the urge to do
something. This suggests that projective identifi cation holds an intrusive potential for action.

(Loch, 2001, p. 127)

The oral transmission of the term, for example through lectures, guest speakers, supervisions, is
not easy to trace. In 1995, R. Klüwer assessed the overall situation as follows: ‘It is safe to say
that Melanie Klein did not have much infl uence on German psychoanalysis until the end of the
eighties and was quite insignifi cant in the seventies and eighties before that’ (Klüwer, 1995, p.
47).

One important reason for the delayed reception of Melanie Klein’s research lies in German
history. This became the subject of increased refl ection in Germany from 1977 after the IPA
Members’ meeting turned down the German invitation to hold the 1981 IPA Conference in
Berlin. Commenting on this episode in her preface to the German edition of the Complete Works
of Melanie Klein , Ruth Cycon (1995) had the following to say: The cruel and destructive
psychotic phantasies discovered by Melanie Klein, of chopping up, tearing up, robbing and
defecating, of burning (through urine), of poisoning (through excrement), of gassing (through
intestinal gases) and the total annihilation of the object that has become absolutely bad, through
excessive projection, provoked horror, rejection and hostile defence when 187

Helmut Hinz

described in detail, because they had become a reality in German history. . . . It was unbearable
to look this time of destruction and irreparable guilt in the eye, something that only seems
possible for us human beings over a longer span of time.

(Cycon, 1995, p. xii)

Although this is a German problem, it is more generally rooted in a refusal to recognize human
destructiveness. Refl ecting on this diffi -

culty, Freud commented: ‘I presume that a strong affective factor is coming into effect in this
rejection. Why have we ourselves needed such a long time before we decided to recognize an
aggressive instinct?’ (Freud, 1933, p. 103).

Seventy years later, a strong affective factor is still at work, resulting in continuing widespread
rejection of the theory of the death instinct, of destructive narcissism and certain aspects of
projective identifi cation (see Frank, 2002).

In Germany, from about 1980, a small, but growing number of analysts began to experience and
become aware of a pervasive lack of clinical understanding. Their intense personal wish for
better understanding and practice in their day-to-day work with patients led them to London for
supervision. They also formed supervision groups in Germany under the direction of
psychoanalysts from the Kleinian group. Herbert Rosenfeld headed one of the fi rst of these
groups, in Heidelberg, from 1981 to his death at the end of 1986. Other groups followed and
individual analysts and smaller groups worked regularly with Kleinians in London. This
continuous joint clinical work resulted in a more precise knowledge and refi nement of
interpretative strategies for dealing with severely disturbed patients and with perverse, addictive
and psychotic mechanisms in neurotic patients. In this way, practical experience was gained in
working with early anxieties and defence formations, including projective identifi cation.

Ruth Cycon, a participant in the supervision group with Rosenfeld and later Feldman, sums it up
this way: ‘It was above all Herbert Rosenfeld who imparted a new understanding of the
transference and countertransference processes and a Kleinian technique that seeks to
comprehend everything the patient says, step by step, as expressing the current transference
situation’ (Cycon, 1995, p. xv). Such experiences had an impact on training analyses and
supervisions of candidates. My own psychoanalytic development was shaped by them. In my
estimation, the greatest

differences between analysts in Germany today lie not so much in theoretical differences, such as
their rejection 188

Projective identifi cation in Germany or acceptance of the concept of projective identifi cation,
but in their practical fi rmness in examining countertransference and transference in the analytic
situation precisely every step of the way.

However, my experience has shown that this microanalysis in hic et nunc et mecum is refi ned
and becomes more fl exible through the concept of projective identifi cation.

Then, in the 1990s, the discussion came to a head. A number of interesting papers on projective
identifi cation began to be published by German authors (Cycon, Beland, Gutwinski-Jeggle,
among others). One Frankfurt author’s comment on this development seems revealing to me:
‘The concept is having a heyday at the moment, and I would even say it is running amok in the
psychoanalytic literature.

Nearly every author has his or her own idea of this process’ (Dornes, 1993, p. 1145). This way of
putting it, as almost ‘running amok’, shows what characterizes the term’s reception in Germany:
on the one hand, a broad positive reception of the concept, and on the other, the suspicion that it
might be a source of dangerous infl uences.
On the positive reception of the concept

In 1995, Gutwinski-Jeggle summarized the reception of the concept in this way: ‘The analytic
world could not get around acknowledging Paula Heimann’s concept of the countertransference
nor Melanie Klein’s observations on projective identifi cation’ (Gutwinski-Jeggle, 1995, p. 71).

In my opinion, this has to do with the intellectual strength of this Kleinian concept, which
condensed a complex clinical experience to one single concept, or even tapped into it for the fi
rst time: namely, the experience that intrapsychic and interpersonal processes run in parallel and
intersect and infl uence one another. This understanding of the analytic process is at the level of
modern epistemology, which gave us the idea of the observer who is dependent on the
observation. These interactions are easier to examine in the analytic situation with the help of an
elastic term such as projective identifi cation. The term ‘projective identifi cation’ proves to be
especially useful because it provides a common denominator for the multifarious, often
simultaneous functions of this process: the desire to penetrate the object, to control and
manipulate it, the perception or illusion of actually being able to do this, taking on the object’s
characteristics, getting rid 189

Helmut Hinz

of a bad quality by putting it into the object, protecting a good one in the same way (Spillius,
1988b, pp. 83–84).

The concept of projective identifi cation also proved to be useful in enabling analysts to explain
intense countertransference feelings and impulses, which compelled one to act, in consequence
of a non-verbal, action-based form of communication. Accordingly, in the subsequent period,
many distinctions and further developments of the concept were also eagerly seized on and
accepted: the important differentiation between normal and pathological projective identifi
cation, for example, and/or between a purely communicative and an evacuative function of
projective identifi cation. Spillius now speaks in terms of evocatory and non-evocatory, and
Britton of attributive and acquisitive forms.

Technical refi nements in treatment resulting from the concept of projective identifi cation were
also adopted. One example was Segal’s suggestion that if an interpretation of the projectively
identifi ed object-relations aspects is given too early and is not suffi ciently worked through in
the countertransference, it may be experienced by the patient as a ‘persecutory pushing back’ of
what had been projected (Segal, 1964, p. 121). Another was Steiner’s differentiation between
analyst-centred and patient-centred interpretation ( J. Steiner, 1993, p. 144). Contributions on
technique were also forthcoming in Germany, such as the distinction between ‘interpretation of
the projective identifi cation’ versus ‘interpretation from within the projective identifi cation’
(Hinz, 1989, p. 616), and the distinction between communicating as conveying an experience
and communicating as sharing an experience (Hinz, 1989, p. 611).

My own clinical experience in dealing with the concept of projective identifi cation is shaped by
the analysis of a patient in her late twenties with a borderline syndrome. She had a severe object-
relations disorder, behaved

promiscuously, quickly broke off every relationship she began and had therefore also already had
very many initial analytic contacts that failed or never got beyond the initial contact, without
being able to begin an analysis. In this analysis, what initially played a key role was that she
talked a lot and very fast, but more with the unconscious intention of avoiding understanding.

Understanding would have meant the beginning of a relationship and that had to be avoided,
because this was connected to evoking the early trauma she had suffered. The analyst was
projectively identifi ed with not-understanding, which could thus be brought into a meaningful
interpretative context in this way.

190

Projective identifi cation in Germany When this hurdle was overcome, a great inner unrest came
to the fore, which also projectively affected the analyst. Over a long phase of the analysis, the
interpretative work consisted in taking the tormenting unrest that the patient created in the
analyst as a key message, which the patient was sending without being able to put it into words.
Simply stating what the analyst felt and understood as projective identifi cation, initially without
making any further causal connection beyond that, led with impressive clarity to fi rst moments
of calm in this analysis. All I said, for example, was: ‘You’re full of agitation, you feel like a
bundle of nerves.’ I think this is an example of interpretation from within the projective identifi
cation.

This example is characterized by an understanding of a long-lasting countertransference

experience in terms of projective identifi cation.

With a second clinical illustration, I would like to show how countertransference perceptions of
short duration can be used to leave aside the content level of communication in favour of an
underlying level of unconscious communication that has greater relevance. A patient spoke so
quietly that several words and then the context of what was being said remained
incomprehensible. In situations like this it can be extremely helpful to think in terms of a
projective identifi cation and to resist the impulse to ask for clarifi cation. Leaving aside the
substantive context of the content and taking my hunches about this patient and the current
situation into account, I interpreted that what she probably wanted to tell me more than anything
else at the moment was how mysterious her relationship was to her, and she was wondering how
she would be able to understand what was going on here. She immediately reacted very
defensively and full of cynical self-deprecation: ‘I’m an idiot, mentally deranged.’ This self-
destructiveness is otherwise usually hidden behind superfi cial friendliness. It is activated when it
becomes possible for her to articulate this level of her search for a relationship and her
experience of not-knowing. At this level it is entirely unclear to her whether she is dealing with a
helpful or a destructive object.

Critical voices and aversions to the concept

Critics of the concept of projective identifi cation take their cue from Melanie Klein’s own
ambivalence about her concept. She apparently had doubts about its usefulness and was afraid
that it would be misused 191

Helmut Hinz
as an alibi for the analyst’s own insuffi ciencies, because he might downplay the complexity of
the analytic relationship and blame his own emotional reactions on the patient without working
on them, integrating and converting them further refl exively. Hanna Segal has often described
how Melanie Klein once told an analyst: ‘You need a little self-analysis’ (Gutwinski-Jeggle,
1995, p. 72) because he felt confused and took this to be an expression of confusion in his
patient.

This concern about potential misuse of the concept is often coupled with the accusation that
Melanie Klein had a mechanistic understanding, that she believed parts of the self are put into
the object directly without taking the object’s capacity to work on them into account. This is
identical with the prejudice or preconceived notion, that Kleinians deny or underestimate the infl
uence of the analyst, the object or external reality, both in the present and genetically.

Accordingly, in Germany there is a widespread opinion that the Kleinian technique is ‘highly
theory-driven’ (Thomä and Kächele, 1985, p. 150), intrusive and that Kleinian technique leads to
superego intropression and causes the

clinical phenomena that are found.

These critical opinions, which to some extent are due to weaknesses of early Kleinian work, and
to the adoption of an inferior imitation of her theory and technique, easily become solidifi ed into
the form of prejudices if they spring from theory that is too quickly received and intellectualized.

In many cases, I think, there is also a fear of the clinical experience that could be gained if these
concepts were applied skilfully. Another factor involved in the rejection of the concept is the
suspicion that the fuzziness in the concept’s defi nition leads to confusion, directly to
countertransference mysticism, to a lack of scientifi c rigour and to self-isolation in the scientifi c
community. However, it has already been pointed out that the elasticity of a concept has
advantages in clinical practice for its clinically fl exible usefulness. We should not follow the
example of those philosophers who are always cleaning their spectacles in order to see better but
never actually looking through them in order to see anything.

Instead, let us recall Max Planck, who illustrated the usefulness and importance of basic concepts
that cannot be defi ned with perfect clarity by demonstrating to his friends while washing dishes
how a wine-glass can be polished clean with a dirty tea-cloth. The diffi culty of defi ning
projective identifi cation more clearly lies in the high degree of complexity of the processes that
it seeks to grasp. It is an 192

Projective identifi cation in Germany attempt to make explicit that an archaic form of
communication leads to an interpenetration of the boundaries between two psychic systems
(subject and object) for the purpose of building up primarily one of these systems.

A quote from a textbook on psychoanalytic therapy will serve as an illustration of these critical-
aversive voices: If most of the most important part of the exchange between patient and analyst
could be explained according to the model of projective and introjective identifi cations, then
psychoanalysis would have its own and original theory of communication. This would largely be
beyond critical examination by other disciplines because in cases of doubt it could always be
said by recourse that unconscious processes are involved here.
(Thomä and Kächele, 1985, p. 151)

This seems to suggest that analysts are not allowed to discover anything that others cannot
scrutinize with their methods, and that a fruitful concept should not be used because it could be
misused.

It is repeatedly contended that the concept of projective identifi -

cation has been disproved by the results of infant research. This is not borne out on closer
examination: infant researchers come to the interesting conclusion that

psychoanalytic theory so far underestimates the infants, yet on the one hand it tends to
overestimate the infant’s abilities by deeming it capable of complicated mental operations – such
as hallucinatory wish-fulfi lment, omnipotent phantasies and projective identifi cation – which it
does not even have yet.

(Dornes, 1993, p. 1116)

Many results of infant research provide impressive proof of the infant’s abilities for coherent
self- and object perception (Dornes, 1993, p. 1129). As I see it, this result confi rms an important
prerequisite for Kleinian developmental theory: at the beginning is the experience of
separateness, not non-separateness or symbiosis. These are already a form of defence if
separateness and infantile need for help are felt to be unbearable. Projective identifi cation can
then be understood to be an early active mechanism of protection and defence. Winnicott already
assumed that there was an alternation of 193

Helmut Hinz

coherence and incoherence and Bion described the ongoing oscillation between the paranoid-
schizoid and the depressive position (or between PS–D). Wolfgang Loch assumed that the (more
coherent) depressive position might precede the (incoherent) paranoid-schizoid position.

In my view, the controversy regarding projective identifi cation has to do with divergent
concepts of representation and symbol.

There are those who would not speak in terms of representation and symbol formation until
language development starts and elaborated conditions prevail; while others assume that early,
archaic coherence-experiences close to the body exist, i.e. precursors of representation and
symbol formation. These are stored in the memory as action sequences and interaction patterns
between the baby and the person taking care of it (in connection with physical sensations and the
care-taking of great bodily needs). Gaddini speaks here in terms of body-phantasy and phantasy
in the body. At the same time, image impressions are stored as iconic representations. Under
these conditions experiences are inscribed in the memory and rewritten – in other words pre-
represented, while not yet represented in the word-memory. It then is possible to imagine that
defence processes become operative from the very beginning in moments of excessive strain on
the coherence of the child’s psyche, such as Gaddini’s imitation and projective identifi cation.
Infant research appropriately assumes that
‘Thought by the pre-symbolic child [is] tied to action and perception, senso-motoric and not
symbolic’ and that ‘Its

“object presentation” [is] an object sensation and identical with the sensory perceptions that the
object triggers in him’ (Dornes, 1993, pp. 1144 and 1143).

At this point, one problem of infant researchers often sets in, namely an epistemological and
psychoanalytic naivety.

The infant is imagined as being the passive receptacle of sense perceptions. It is forgotten that
the infant too can only tap into his world and make perceptions by means of unconscious
phantasies that act like pieces of theory.

To conclude my discussion of the concept’s rejection by infant researchers, I will once again
quote a passage from Dornes that provides a good illustration of the clinical relevance,
practicability and modernity of the term projective identifi cation, which is fi fty years old: A
depressive infant does not project any depressive phantasies that it does not have, but its physical
posture has become sunken, its 194

Projective identifi cation in Germany motor skills and breathing has slowed down, its face is
lifeless, its interaction lacking in vitality: (. . .) even adults who are not depressive are infi ltrated
by this affect and style of interaction.

Their interactions slow down, their facial expression becomes less lively, and after a short time
they feel exhausted.

An unpleasant sensation of the infant has settled on them. It was not ‘projected’, but
communicated via interactional forms of behaviour. The adult cannot evade the communicative
affect entirely as in the case of projective identifi cation; it now interacts depressively itself and
gives the affect back to the infant.

(Dornes, 1993, p. 1146)

Finally, I would like to mention one more reason for the aversion to the term projective identifi
cation which seems important to me.

Bion had taken over the concept of projective identifi cation and expanded it to describe a
fundamental form of emotional mental exchange and the emergence of the psyche. Projective
identifi cation is the decisive communicative bridge between inside and outside, infant and
caring person. Projective identifi cation can therefore aptly be referred to as a ‘social umbilical
cord’ (Frank, 2001). Initially, however, the concept was much more narrowly defi ned. Melanie
Klein emphasized the mode of aggressive penetration into the object, taking possession of it,
manipulating and controlling it. She spoke of the ‘prototype of an aggressive object relation’. By
expanding the concept into a fundamental mechanism in the emergence of the psychic apparatus,
of thought and thinking, this aspect receded into the background. And yet precisely this violent
and perhaps ugly side has great theoretical and clinical importance. This is the case not only
when we have to deal with serious pathologies and unbearable mental states that can only be
projected. The packaging of the concept in the concept of ‘containment’ and its popularization
contributed further towards making it pleasing. Thus, the ugly but realistic aspects were extracted
from the concept. Yet, these are necessary in order to be able to perceive and modify psychic
reality in its cruel and deadly aspects. When terms like containment or, fi fteen years earlier,
holding function became fashionable, these watered down and lost the complexity of the concept
of projective identifi cation. In the popularized form of

containment there is only successful containing, but no longer any form of failure, which Bion
clearly described.

The following playful formulation nevertheless captures something true 195

Helmut Hinz

about the story of the reception of projective identifi cation in Germany. There are good and bad
Kleinians: Melanie Klein is a bad Kleinian and Bion is a good Kleinian. My personal, only
slightly ironic statistic reads as follows: in the Federal Republic of Germany, about 20 per cent
are bad Kleinians, and 80 per cent are good Kleinians. Only very few want to be counted among
the bad ones.

The action potential in unconscious phantasy,

in the transference, and the intrusive action potential of projective identifi cation Ever since
‘Remembering, repeating and working through’ (Freud, 1914), analysts have known that the
analytic aim of conscious remembering can usually only be achieved via the detour of repeating.
The technical knowledge that transference is a form of remembering has also been established.
Ideas, experiences and actions emanating from an unconscious object relations experience and
fantasy are repeated.

Positive and negative feelings, libidinal and aggressive aspirations, fantasies and behaviour
patterns – in other words, the totality of an object relationship – are repeated and actualized,
which in turn implies an element of action.

The concept of countertransference as an emotional-mental and initially unconscious reaction by

the analyst to the transference, has also become generally indispensable. Anxiety had to be
overcome just to recognize the transference, for the treatment ideal of rapid dissolution of a
disorder failed when it came up against the phenomenon of repetition and the transference as
resistance to the continuing process of investigation. It took courage to give serious consideration
to the phenomenon of countertransference, given the fact that it shattered the scientifi c ideal of
the distanced, contemplative analysis of a phenomenon and the independence of the observation
from the observer. Years of psychoanalytic work were required before the fact that transference
is not just a hindrance to the analysis, but can also be its ally could be established and accepted.
The same is true with regard to the long and hard struggle to gain recognition that the
countertransference is an important emotional indicator for and instrument of psychoanalytic
investigation and not a failure on the analyst’s part, as it were. As we know, Paula Heimann and
Heinrich Racker were revolutionary pioneers in this endeavour, while Melanie Klein and the
Kleinians 196

Projective identifi cation in Germany were still dragging their feet. And today we know that the
countertransference is ‘the best of servants but the worst of masters’ (Segal, 1981, p. 86).

This discovery of the interaction and reciprocal dynamics of transference and

countertransference was extremely uncomfortable for psychoanalysis, but also extremely
interesting. As a consequence, the category of causality was pushed into the background and the
category of reciprocity moved to the fore as a complementary paradigm.

This development was further radicalized by the discovery of projective identifi cation, as the
concept of projective identifi cation is the description of an interpersonal process which draws
the analyst, actively and with pressure, into affi rming the patient’s unconscious system of
perception, thinking and fantasy. The patient’s intensive yearning for identity seeks to bring
external reality into harmony with their unconscious fantasy (Feldman, 1997; Hinz, 2002b)
because if they do not harmonize, this can trigger intense anxiety.

On the internal affi nity of transference and

projective identifi cation

This unconscious but effective pressure to bring the object relationship in line with the
unconscious object-relations fantasy comes from an action component and/or action potential
that is already inherent in the simple transference.

Sandler, Dare, and Holder (1973) had defi ned the following with regard to the transference: that
transference need not be restricted to the illusory appercep-tion of another person . . . but can be
taken to include the unconscious (and often subtle) attempts to manipulate or to provoke
situations with others which are a concealed repetition of earlier

experiences and relationships (. . .). It is likely that such acceptance or rejection of a transference
role is not based on a conscious awareness of what is happening, but rather on unconscious cues.

(Sandler et al., 1973, p. 48, quoted by Klüwer, 1995, p. 52) The intrinsic affi nity between
transference and projective identifi -

cation becomes clear from Sandler’s defi nition of the concept of projective identifi cation in
1984 at the conference in Jerusalem: 197

Helmut Hinz

In our opinion, it is useful to regard projective identifi cation as a mechanism in which undesired
aspects of the self (or desired, but unachievable states of the self ) are perceived and called forth
in another person. This is accompanied by the attempt to control this other person, in order thus
to gain the unconscious illusion of control over the externalised aspects of the self. The projected
behavior is evoked and induced by means of subtle unconscious pressure and cues . . . and can be
seen best in the transference-countertransference situation in the therapy.

(quoted in Zwiebel, 1985, p. 458)

The common features of transference and projective identifi cation seem important to me. They
have in common the inherent endeavour to push the other person unconsciously towards a
repetition, towards a behaviour, towards a pattern of action or to an action, to manipulate the
person subtly, in order to control him. This common basis implies that the concept of projective
identifi cation does not describe a special case or an exceptional situation, but rather deepens and
refi nes the concepts of transference and of projection. For those dimensions of the personality in
which paranoid-schizoid mechanisms are predominant, in other words, where the symbolization,
or representation, of the absent object is damaged, splitting and omnipotence hold sway as
defences. Making a sharp distinction between remembering and repeating, or as the case may be,
between thinking and action, may be misleading for the analyst’s understanding, and this is not
only the case in severely disturbed patients. The patterns of action and behaviour in the analytic
situation are the concretized forms of thinking and experiencing respectively available in a given
case,

‘because we basically have to assume that patients bring their customary attitudes and modes of
behaviour as a whole into the relationship with the analyst, not just their way of thinking’ (
Joseph, 2001, p. 135) An important German contribution in this area was made by Wolfgang
Loch, who offered the following formulation in 1965: Transference and countertransference, are,
as we know, ‘motivated’

forms of behaviour in the sense that they represent object relationships that are driven by or
brought about by instinctual drive needs.

But this means that the ego, as the executive organ, is under pressure; for in such cases, it works,
energetically speaking with motivational 198

Projective identifi cation in Germany cathexes which press for a rapid discharge. (. . .) 40 In the
analyst’s introspection, the thus instinctually determined interpretations are heralded by their
urgent character. One feels virtually forced to give a certain association as an interpretation. If
one can analyze this pressing association quickly, it is possible to allow the countertransference
motivating it to become conscious and thus put oneself in a position to verbalize a ‘constructive’
interpretation.

(Loch, 1965, pp. 41–42)

Freud established a system of coordinates for the analyst with his statement: When I instruct a
patient to abandon refl ection of any kind and to tell me whatever comes into his head, I will rely
fi rmly on the presumption that he will not be able to abandon the purposive ideas inherent in the
treatment and I feel justifi ed in inferring that what seem to be the most innocent and arbitrary
things which he tells me are in fact related to his illness. There is another purposive idea of
which the patient has no suspicion – one relating to myself.

(Freud, 1900, p. 532)

Between this X-axis and Y-axis lies the fi eld of research of analysis.

To the extent that the patient is always talking about his relationship with the doctor, everything
that he tells him is in the sense of the transference at the same time a stimulus for the doctor, in
which – unconscious – expectation is brought forth that this partner will react in just the same
way as the decisive others (those persons who had a formative infl uence on him) have done in
the past. The fact is, the patient must want this; consequently, he keeps holding on to the illness,
indeed he is constantly driven to nip in the bud every attempt to overcome the illness, insofar as
he is holding on to the illness, as indeed he must, (. . .) want the

40 The counterpart to such ‘motivational cathexes’ are so-called ‘attention cathexes’. Only if the
ego works with attention cathexes, that do not have any aims such as specifi c objects or
selectivity . . . is there a guarantee that not inner needs, but real features decisively determine
perception and experience.

(Loch, 1965, pp. 41–42)

199

Helmut Hinz

same manner and attitude [as those] of the persons in which the illness thrived (. . .) In that sense,
transference means evoking those conditions which perpetuate the pathological solution.

(Loch, 1965, pp. 39–40)

As already indicated above, the common basis of transference, projection and projective identifi
cation has an important source in memory. We can assume that the fi rst traces of memory come
from the perception of bodily processes and physical interactions with the caregiver. For
analytical investigation and treatment, it is probably important to be able to work not just at the
level of the lexical memory and iconic memory, in other words, at the level of memory in words
and images, respectively, but also at the level of action-patterns, in other words, to search for the
action-memory, which is manifested in the behaviour patterns in the doctor-patient relationship.

It can be assumed that the analyst will not be able to make headway with his patient into areas
that are relevant for the genesis of the illness and for its healing process until he recognizes and
interprets processual fi gures and patterns of action, ‘process identifi cations’

(Danckwardt, 2001) below the level of the spoken contents. Closely connected to this are
recommendations to attend not only to the content of the verbal communication, but also its
form, not only to what is being communicated, but also how and what function that which is
being communicated presumably has in the current relationship, taking note not only of the
verbal communications, but also the gesticular, phonetic, the scenic and the action dialogue
(Bion, Loch, Argelander, Klüwer, Joseph, Feldman). Form is sedimented content,
psychoanalytically expressed: the form of the message holds a hidden meaning which
corresponds to a deeper transference layer, which initially cannot be communicated verbally.

Scenic function of the unconscious and action dialogue: event and interpretation,
interaction of unconscious phantasy and behaviour

In 1946, Melanie Klein described the processes to which, according to Elizabeth Spillius, she
initially rather casually, then in 1952 defi nitively, gave the name of ‘projective identifi cation’.
At the same time,
‘split-off parts of the ego are also projected on to the mother or, as I 200

Projective identifi cation in Germany would rather call it, into the mother’ (Klein, 1952a, p. 8) In
a footnote to this statement she adds, almost apologetically: The description of such primitive
processes suffers from a great handicap, for these phantasies arise at a time when the infant has
not yet begun to think in words. In this context, for instance, I am using the expression ‘to
project into another person’ because this seems to me the only way of conveying the unconscious
process I am trying to describe.

As far as I can tell, this situation remains unchanged, even if developmental psychologists use
the new term

‘interactional communication’ instead of projective identifi cation and unconscious phantasy. If

the infant researchers then also forget that the infant is not a passive receptacle but creates its
environment at the same time in accordance with its inner reality, this would be a setback behind
the insights of psychoanalysis.

Therefore, one can agree with Dornes, for example, who makes a case for a theory of the
interaction between phantasy and behaviour.

Although these processes still cannot be described more precisely or more simply even today, the
auxiliary construction that, based on the principle of human similarity, ‘analogous and/or
complementary affects’ can be evoked in the analyst seems helpful to me (Dantlgraber, 1982).
Understanding grows from classifying the evoked affects in

‘emotionally accented ideational images, in other words, phantasies’, i.e. through the integration
of a previously

‘split-off affect into a meaning context’ (Dantlgraber, 1982). What seems important about this is
that emotion results from situational interpretation and situational interpretation leads to emotion
(Beland, 1992, pp. 63 and 66).

Together they form an elementary unit. Early processes of exchange, in other words introjection,
projection and projective identifi cation, thus mean that psycho-physical sensitivities, phantasies
and affects of the parents and the infant are communicated via interactional processes and mutual
assignments of meaning and taken up, understood or misunderstood by the infant and the
parents, that is to say, are interpreted or misinterpreted by them.

My impression is that the current focus of the psychoanalytic discussion has more or less shifted
away from projective identifi cation and towards examining its results. Numerous works on
enactment, involvement and on the microanalysis of the patient’s reactions 201

Helmut Hinz

to an interpretation and the effort to listen to the listening prove this.

In this connection, it should also be pointed out that in Germany Alfred Lorenzer had a great deal
of infl uence with the concept of Tiefenhermeneutik (hermeneutics of depth) and the
Hermeneutik des Leibes (hermeneutics of the body), as did Hermann Argelander, with his works
at the end of the 1960s on the scenic function of the unconscious.

And Rolf Klüwer, with his studies on the communicative function of acting out in the analytic
situation, apparently independently of the Kleinian research and using different terminology,
assigned great importance to the observation of enactments and modes and patterns of behaviour
in the doctor-patient relationship.

In particular, the works of Argelander had a strong infl uence in shaping the analytic
understanding in Germany, as did the works of Racker on concordant and/or complementary
countertransference, which are closely related to the theme of projective identifi cation.

In spite of these developments in the direction of microanalytic research on the situational and
actional aspects, on enactment and role adoption, it is important to note with some modesty that
even the analyst who works conscientiously and in a modern fashion in this way and observes his
countertransference feelings in order to make use of them as a treatment instrument is not
immune to the risk of fi nding only what he already knew was there anyway, in the manner of a
subsuming logic. However, the same is also true for all other psychoanalytic means of access.
This is why the methodological and utopian guiding principle of evenly suspended attention
(disciplined pushing back of theory, memory and wish) is of the utmost importance.

How much development takes place in an analysis also depends essentially on the analyst’s
person; into which areas he has advanced in his training analysis, what theoretical concepts he
favours, what experiences he can take in (cf.

Money-Kyrle, 1991): the degree to which the analyst can, for example, think and feel in terms of
sharing the responsibility for ‘mistakes, misunderstandings and failures’

(Steffens, 1999, p. 79). Can he uphold or restore a ‘thinking and feeling space’ for himself
(Gutwinski-Jeggle, 1995,

p. 76) under the infl uence of an intrusive action potential? The concept of projective identifi
cation can help in any case to feel complex clinical situations more fl exibly and think them
through. To give an example: anyone who knows the difference between a projective identifi
cation for the purpose of communicating an unspeakable (or inexpressible) 202

Projective identifi cation in Germany experience and a projective identifi cation that has the
purpose of getting rid of an unconscious experience, knowing that it depends on his own
personal make-up, his own reaction, whether or not projective identifi cation can be turned into a
communication (from ‘Notes on some schizoid mechanisms’) (Spillius, 1990, p. 106) can in a
corresponding situation wind up deciding that the patient can, at the moment, under the greatest
pressure, do nothing more than get rid of everything and throw it out of himself. Even in this
situation, this could still prove to have a value as an acceptable interpretation and secure its status
as a verbal communication.

203

12
Projective identification

The fate of the concept in Italy and Spain

Jorge Canestri

In order to trace an outline, even though incomplete, of the fate and use of the concept of
projective identifi cation in Italian and in Spanish psychoanalysis, I have examined its literature
and have benefi ted from the help of colleagues from these two countries. I am grateful to them,
although the responsibility for anything that is missing or inaccurate, and for the general layout
of this presentation, obviously falls on me. 41

Some general premises

Some preliminary statements will be useful to help identify the problem we are dealing with.
First, even when we have in mind the study of the evolution of a specifi c concept, it is necessary
fi rst to outline the history of the general evolution of the psychoanalytical theory in each
country, and to describe the modalities for training in the psychoanalytical institutes and the
general cultural orientation in the society. The acceptance of a particular concept by the
psychoanalytical community is the result of many factors: some of them are exquisitely
individual, for example training undergone in another

41 I especially thank for their assistance my Italian colleagues Jacqueline Amati Mehler, Stefano
Bolognini and Federico Flegenheimer, and my Spanish colleague Luis Martin Cabré.

204

Projective identifi cation in Italy and Spain country; others are more general, for example the
theoretical orientation and the traditions of the society. This outline exceeds our present capacity
and, in the case of the history of psychoanalysis in Italy, we can rely on some excellent treatises.

Second, regarding the concept itself, there appears to be a certain amount of consensus within the
literature around the fact that Melanie Klein, in 1946, proposes a defi nition of projective identifi
-

cation adhering to that of a typical defence mechanism in the classical sense: it seems to be
linked to instinctual problems and reveals the need for the ego to defend itself. In 1955 she
extends the process to normality, both in order to account for empathy as well as to consider its
functions in normal development. These two meanings would be explored in various Italian and
Spanish works on the matter, but the use of the second meaning is predominant.

This orientation does not appear to follow the British orientation: an initial study of the psychotic
pathologies with a particular focus on the psychopathological role played in them by projective
identifi cation; a subsequent analysis of the communication functions in the mother-child and in
the analyst-patient relationships.

Third, the initial clarity of Klein’s defi nition, describing a specifi c and unitary mechanism,
becomes progressively obscured. A concept that has become infl ated and confused in its clinical
and theoretical use (Amati Mehler, 2003) is referred to, and there is a tendency to consider it as
‘a general term indicating a certain number of distinct processes that are however correlated,
connected to splitting and projection’, according to the defi nition of O’Shaughnessy (1975).

Some epistemological premises

First, from an epistemological point of view, it would be advisable to make some preliminary
specifi cations. Two interpretative positions can be identifi ed from today’s theoretical pluralism.
One of these states that psychoanalysis has a central indispensable nucleus composed of a small
number of fundamental theoretical propositions, to which

‘puzzle’ solutions are linked in an attempt to solve partial problems.

The other position says that we are dealing with diverging and accomplished theories concerning
the psychic apparatus. My opinion is that each of the psychoanalytical positions presents a
different theoretical picture, both in the sense of a global theory as well as 205

Jorge Canestri

concerning the details of the functioning of the psychic apparatus.

Therefore, whoever produces interpretative hypotheses of a phenomenon, does so within a

general theoretical framework that can be more or less well characterized and more or less
explicit.

Second, the unity of analysis, from an epistemological point of view, is in the theory. The
empirical data on which we work are the data of the methodological empirical basis, in other
words data that presuppose the use of material or conceptual instruments that in their turn
respond to a theory (this is the case with projective identifi cation). A different theory of the
instrument (or the use of a different instrument) has an inevitable consequence on the
methodological empirical basis, on the method itself and consequently on the theory (this also
applies to projective identifi cation).

Third, if we agree with what has been said, we must ask ourselves whether we can use a concept
taken from one theory in the context of another theory, without altering it or modifying it into
something else, and without the concept in question entering into obvious contradiction with the
theory into which it has been imported. I must say that, looking again at the literature on this
theme, it is diffi cult not to conclude that in certain uses the concept of projective identifi -

cation is rendered unrecognizable and incompatible with its guest theory, as well as making it
incoherent. The argument invoked for the use of this concept within theoretical positions that
would apparently discourage it, is that the phenomenon described by projective identifi cation
exists ‘in nature’ and can easily be observed in clinical practice. This objection is naive and does
not take into consideration what was said above, in other words it confuses the epistemological
empirical basis with the methodological one. What we can testify to is a phenomenon that
certainly exists: how it is observed, conceived and theorized belongs to a different order of
things, to a level of abstrac-tion that is not equivalent to the phenomenal.
Fourth, any hypothesis that tries to describe a mechanism, whether it be defensive or normal,
cannot be formulated, discussed and put into practice outside a more general hypothesis about
the development of the psychic apparatus.

Consciously or not, every hypothesis of this type is embedded in a theory of development and
cannot be intrinsically in contradiction with it. The concept of projective identifi cation is
incompatible with some psychoanalytical theories that have elaborated hypotheses that differ
greatly from the Kleinian ones on psychic development.

206

Projective identifi cation in Italy and Spain The area of research In Italy there are two
psychoanalytical societies, the Italian Psychoanalytical Society (SPI) and the Italian
Psychoanalytical Association (AIPsi.), and both have a pluralistic tradition, in other words
different theoretical orientations cohabit within the societies and within their

training. There are also two societies in Spain, the Spanish Psychoanalytical Society and the
Madrid Psychoanalytical Association. As far as I know, the Madrid Association is also
pluralistic, while the Spanish Society has a mainly Kleinian tradition. In all that follows it must
be borne in mind that, as I live and work in Italy, my knowledge of Italian psychoanalysis is
more complete than that of Spanish psychoanalysis, which is accessible to me through personal
relationships, presentations of works, visits and supervisions, but mainly due to my knowledge of
the Spanish, or in some cases Catalan, literature.

Italy: the fi rst works

In Italy, from the 1960s to the 1980s, psychoanalysis appears to undergo a process of de-
provincialization and liberation from previous cultural and political tendencies (Corti, 1983).
This is also the period during which the British infl uence is felt, through its literature as well as
through supervisions, seminars and the periodic visits of some of the more well-known British
analysts of the Kleinian school: Rosenfeld, Bion, Meltzer, and so on. Concepts from the Kleinian
theory appear in the literature and among them we fi nd the concept of projective identifi cation.

However, as I mentioned earlier, the importance of this concept is mainly emphasized for its
signifi -

cance and its value in the analyst-patient relationship. This fact explains why – as within the
British Society – the concept of projective identifi cation remained intimately linked to that of
countertransference, which ended up by fi nding its rationale in projective identifi cation. This is
very clear in some of the pioneering work of two Italian analysts who are also quite well known
abroad, S. Manfredi Turilazzi (1974, 1984a, 1984b) and L. Nissim Momigliano (1974, 1984,
1991). The main concern of these two authors seems to be the fact that the continuous use of
projective identifi cation on the part of the patient may make the analyst blind. The analyst does
not 207

Jorge Canestri

fully understand the countertransference and he suffers manipulations and intrusions without
being able to diagnose them or use them therapeutically. An adequate knowledge of projective
identifi cation would, on the other hand, enable him to understand the countertransference more
clearly and would favour its containing function.

The introjection of the analyst’s containing function on the part of the patient would allow for
mutative changes.

Another analyst, G. Di Chiara, describes a similar position in a paper published in 1983 and
much quoted in Italy (Di Chiara, 1983).

The author’s refl ections are mainly focused on protecting the analyst and the patient from
intrusive and parasitic operations. Some important ideas can be found in this work, written ten
years after the initial papers of Manfredi and Nissim. First of all, the fact that the author speaks
of the protection of the analyst and the patient implies acceptance of L. Grinberg’s hypothesis of
counterprojective identifi cation (Grinberg, 1962). The analyst who makes a counterprojective
identifi cation as a result of the patient’s projective identifi cation produces a counter-acting. This
phenomenon is not confi ned to particularly serious situations, but appears to be ubiquitous.

Another difference lies in the fact that the author already speaks of a group of phenomena
gathered together under the common denominator of projective identifi cation, and no longer of a
single phenomenon. He says: ‘It is diffi cult to fi nd a suffi ciently clear defi nition of projective
identifi cation, that is, a defi nition that includes all the phenomena that are grouped under this
denomination’. Di Chiara states that these different phenomena have in common a mental
operation that originates in a part of the psychic apparatus and terminates either in a part of
another psychic apparatus or in another part of the same apparatus in which it originated,
producing alterations in the emitting part as well as in the receiving part.

(Di Chiara, 1983, p. 467)

This defi nition already emphasizes some of the problems that will subsequently be raised; for
example, can an operation that terminates in another part of the same apparatus be considered to
be projective identifi cation? This question was re-examined only ten years later in a work by J.
Amati Mehler. Moreover, Di Chiara (contrary to his predeces-sors) is sensitive to the possible
incompatibility of the simultaneous or 208

Projective identifi cation in Italy and Spain alternative use – in terms of projective identifi cation
– of the Freudian

structural model and of the internal objects model. However, he concludes that the problem is
manifest not in clinical practice or in the observed phenomenology, but in the incapacity of the
theory to represent the facts, therefore running into the objections that I mentioned earlier: our
data are not empirical, but rather empirical methodological; in other words interpreted with a
conceptual instrument and through a theory. From this perspective no clinical phenomena exist
that are exempt from theory.

A year later, the same author wrote with F. Flegenheimer an accurate historical-critical note
about this concept in which they review the main stages in the diffusion of projective identifi
cation in the international and Italian psychoanalytical community (Di Chiara and Flegenheimer,
1985). Some of the problems mentioned – a weighing down of the concept that has become too
broad, loose and inclusive, its incompatibility with certain theoretical patterns, the divergence of
opinions between those who think the concept of projective identifi cation should also include
projection, and those who think the contrary, the objections to Grinberg’s concept of
counterprojective identifi cation, and so on – are already put forward in this note, as they are in
the notes that R. Speziale-Bagliacca adds to the Italian publication of Grotstein’s book on this
subject (Speziale-Bagliacca, 1983). There is a notable omission in the historical-critical note by
Di Chiara and Flegenheimer (1985): they do not consider E. Gaddini’s contribution on imitation,
but this will be taken into account in a work by M. Mancia (1996).

In 1984 Manfredi Turillazzi published a work of which I quote the summary: This contribution
endeavours to fi ll the gap between our daily clinical work and psychoanalytical language and to
rediscover the lost meaning of words which we continue to use. I particularly examine the
meaning of the term

‘projective identifi cation’ and the widening of the gap between its conceptual defi nition and its
clinical use. I think that what we want to describe with this term can be understood only in the
context of a psychoanalytical object relations theory and I have attempted to provide a basis for
the defi nition of ‘internal object’. In my opinion projective identifi cation, within this
framework, is nothing but the externalization through acting of an internal object relation.

(Manfredi Turilazzi, 1984b)

209

Jorge Canestri

If we compare this summary with her work of 1974 we can see that she no longer concentrates
on theory: the phenomenon of projective identifi cation must be understood within an object
relations theory, and it is nothing but an externalization through acting. But she also reveals that
there is a gap between clinical work and theory, that psychoanalytical language has become
babelized, and that the words we use to describe and conceptualize the phenomena are worn out
and function as umbrellas sheltering more differences than similarities. This last level of
awareness is not dissimilar to that of many other analysts of the international community. On the
other hand, there may be doubts about the theoretical decision proposed by the author: is
Kleinian theory part of object relations theory? Some analysts would be prepared to uphold the
contrary. Are projective identifi cation and externalization synonymous? Here also it is
reasonable to have doubts.

Italy, twenty years later

In the following years, during the 1990s, I believe that the concept of projective identifi cation
was incorporated into current use and charged with all the ambiguities already mentioned.
Analysts use it in the discussion of their clinical cases without – or so it seems to me

– making much effort to suggest intersections between apparently incompatible models. At times
one has the feeling that the problem of incompatibility is not even noticed or taken into
consideration. A specifi c analysis of the concept is not re-proposed until 1993 by Amati Mehler.
I will deal with this work later.

Among the developments during this period I will only mention a paper by A. Ferro (1987), in
which we can observe theoretical intersections that will become increasingly frequent in the
literature. Ferro wrote a work on the

‘Inversion of the fl ow of projective identifi cations’, a situation in which the projective identifi
cation originates in

the analyst, ‘in three different situations: that of an analyst weighed down by his own anxiety;
that of an analyst undergoing intrusion by a particular disturbing patient, and fi nally that of an
analyst performing his normal work’

(Ferro, 1987, p. 70), but it is not conceived in terms of counterprojective identifi cation. It is an
interesting work, with abundant clinical material illustrating the author’s theses that refl ect
Ogden’s observations about the technical errors to be attributed to the analyst’s 210

Projective identifi cation in Italy and Spain projective identifi cation (Ogden, 1979). Ferro
moves within a Kleinian theoretical framework, in a Bionian-Meltzerian version (inversion of
alpha function), but he takes into consideration some of Winnicott’s contributions, as he will
continue to do later. I may add that during the 1990s, in Italy Kleinian theory was mainly defi
ned in Bionian terms.

It goes without saying that many other authors (F. Fornari, L. Generali Clemens and others) and
articles could be mentioned, but I have tried to quote only the bare minimum of works, specifi
cally regarding projective identifi cation, that will allow readers who are not familiar with Italian
psychoanalytical literature to fi nd their way along the historical pathway that this concept has
followed in Italy.

The work by J. Amati Mehler, mentioned above, is in the form of a comment and simultaneously
a discussion of Joseph Sandler’s contributions on projective identifi cation (Sandler, 1987b). The
title of this paper, ‘Internalization and externalization processes: projective identifi cation’,
already suggests a framework for the concept in question: it is interpreted as being part of
complex and controversial processes of internalization and externalization that have different
characteriza-tions in the different theoretical models. If these psychic events to which projective
identifi cation belongs are considered in terms of processes, then they must be inserted into an
overall theory of the relationships of outside/inside, subjective/objective, self/other. I will supply
an example given by the author: from the Kleinian point of view projective identifi cation is a
mechanism whereby self-object boundaries are established in infancy; from Sandler’s point of
view, for projective identifi cation to function as a mechanism of defence it is necessary to
postulate the existence of a boundary between self and object. Therefore, if we want to make a
coherent examination of the concept, we are obliged to consider (paraphrasing the author): fi rst,
the degree of psychic maturation (and, I would add, the kind of theory we have on it), second,
whether the projection that we are describing and conceptualizing is the projection of instinctual
drives or other contents, such as representations of the self, representations of whole or part
objects, representations of fused parts of the self and of the objects, other aspects of their
relation, and so on, third, whether we consider projective identifi cation as a mechanism of
defence, or as a physiological phenomenon that contributes towards differentiation between the
self and the external world, and represents a normal way of communication both between mother
and child and between adults.

211

Jorge Canestri

Amati Mehler’s work is therefore in the form of an examination articulated around each of the
concepts that are in one way or another involved in the concept of projective identifi cation, and
also around the theoretical frameworks inside which each of them could be inserted. This fact, as
well as her own personal theoretical beliefs, leads her to take into consideration E. Gaddini’s
concept of imitation, which had not found a place in the notes by Di Chiara and Flegenheimer.

The concept is internationally known and needs no comment; it is suffi cient to say that its
insertion in this context is justifi ed to the extent that it describes an operation that differs from
that of introjection, that contributes to the identifi cation and appears to be based on omnipotent
fantasies connected with primitive perceptions of the modifi cation of the baby’s body. I quote
Gaddini’s concept because it will be included in the last work that I shall mention.

At this point we should recall the statement made by Sandler in the title of a work he wrote: ‘On
communication from patient to analyst: not everything is projective identifi cation’ (Sandler,
1993).

This serves as an introduction to the work of M. Mancia (1996). The author says that with the
concept of projective identifi cation, Klein (1946) introduced a new scientifi c paradigm to
psychoanalysis, that of the unconscious preverbal modalities that operate, also defensively, in the
analytical relationship. In this defi nition he appears to deal

with projective identifi cation both as a defence mechanism and as a communicative modality,
even though his research seems more oriented towards this latter aspect of the concept. The
author proposes a pathway through very different disciplines and developmental theories, ‘in the
attempt to elaborate an integrated hypothesis on the role of the very early imitations, the
representations and the various forms of identifi cation’ (Mancia, 1996, p. 226) (among which is
projective identifi cation). It is within this framework that Gaddini’s imitation – referred to
earlier – rightly fi nds its place, a concept

‘neglected and underestimated by psychoanalysis’. Mancia is aware of the implicit risks in the
construction of an integrated hypothesis: Although I am aware of the diffi culty of integrating
knowledge deriving from different methods of study, I think it is legitimate to try to elaborate
complex hypotheses that take into consideration the relevant phenomena also beyond the
different theoretical contexts inside which they are constructed.

(Mancia, 1996, p. 226)

212

Projective identifi cation in Italy and Spain For example, this leads him to confront Piaget’s
theory of imitation, Meltzoff’s genetic imitative competences, Brazelton’s primary
intersubjectivity, Trevarthen’s proto-conversations, Condon and Sander’s auto- and hetero-
synchrony, and so on. I think it is easy to see how the problem that I referred to earlier becomes
complicated: is it possible and valid to interpret a concept by isolating it from the overall theory
from which it originates? Mancia’s attempt is in any case part of a group of works that seek to
integrate data deriving from different disciplines, and it refl ects a defi nite interest in acquiring a
better knowledge of those modalities of communication that previously were only ‘obscure’.

Spain: general considerations

I will now try briefl y to examine projective identifi cation in Spanish psychoanalysis, although
there have been some obstacles for me to overcome. Some of them have already been mentioned;
others derive from the scarcity of specifi c publications, although I have been able to undertake
only limited bibliographical research. I mentioned earlier that in order to understand and follow
the adoption and development of a concept deriving from a different cultural context to the guest
one, it is necessary to investigate the local psychoanalytical history and culture. I think this is
also true for Spain, a country that has two different psychoanalytical traditions and that uses two
different languages, Spanish and Catalan. As I have already said, the theoretical tradition of the
Madrid Association is more pluralistic; the infl uence of French psychoanalysis has been and is
important, also for reasons connected with training (some members of the association trained in
Paris). On the other hand, the tradition of the Spanish Society, based in Barcelona, is much closer
to Kleinian theory.

In the psychoanalytical history of the Madrid Association, one signifi cant fact from the point of
view of our discussion was the arrival of a migratory current from Latin America, specifi cally
from Argentina Uruguay and Peru, that was assimilated by the Association.

It is particularly necessary to remember the presence of L. Grinberg, an author mentioned

previously – especially in terms of his concept of counterprojective identifi cation – who was of
a clearly Kleinian stamp. The presence of Grinberg as a training analyst certainly played a role in
encouraging the use of Kleinian concepts in clinical work 213

Jorge Canestri

and in theory. It is obvious that Grinberg’s interest in the concepts of projective identifi cation
and counterprojective identifi cation has meant that they are now used in the Association’s
scientifi c discussions. I have several doubts about the effects that this may have had on the
scientifi c developments in the long term. My impression, confi rmed by conversations with
colleagues and friends of the Association, but in any case open to discussion, is that with
Grinberg’s departure from Madrid, the place of Kleinian theory in the theoretical thinking of the
Association was progressively weakened. I will not examine the works of this author here as they
are very well known and, moreover, it is not entirely legitimate to consider him as a
representative of Spanish analysis.

Spain: some relevant works

Research into the Revista de Psicoanálisis de Madrid in order to fi nd works specifi cally
dedicated to projective identifi cation has not yielded much information. One paper called ‘Sobre
identifi cación proyectiva’, published in the Revista , is by a Catalan analyst, E. Jiménez, from
the Spanish Psychoanalytical Society (Jiménez, 1995).

Jiménez describes the origins of the concept of projective identifi cation in Klein’s work as a
prototype of a sadistic object relationship as well as the basis of empathy. The communicative or
evacuative aspect of the operation is connected not only to the contents of the unconscious
phantasy, but above all to the level of splitting and omnipotence. There follows an examination
of the ideas of the post-Kleinian analysts (British) who broadened and enriched the concept. The
novel idea that this author introduces is that of considering Esther Bick’s ‘adhesive identifi
cation’

(Bick, 1968) as being a ‘second mechanism of narcissistic identifi cation’. Jiménez adheres to
Meltzer’s dimensional model of the mind, and he particularly refers to the bi-dimensional state,
in which anxiety is manifested through the phantasy of breaking through the surfaces.

The author emphasizes that in this mental state projective identifi cation cannot exist inasmuch
as it needs a three-dimensional space (Bion), but adhesive identifi cation can exist for the
interpretation of psychopathological problems such as autism and ‘as if ’ personalities.

Jiménez is fully aware that when a concept such as this is introduced into Kleinian theory,
consequences are unavoidable. He says that there are problems connected with the need to redefi
ne certain 214

Projective identifi cation in Italy and Spain Kleinian hypotheses related to the functioning of the
very early ego at the dawn of mental life and certain relational concepts, for instance that of
primary narcissism. The author’s proposal to confront the problem from a Bionian genetic point
of view – the passage from bi-dimensionality to tri-dimensionality and from adhesive to
projective identifi cation – is accompanied by all the diffi culties inherent in the joining together
of two patterns that are basically incompatible unless some of the fundamental concepts of one
of them, in this case the Kleinian one, are relinquished. From this point of view, the work well
exemplifi es the confl ict that is created if a concept corresponding to one particular schema of
development is introduced into another.

Contrary to other authors, although he proposes an integrative solution that is diffi cult to
accomplish, Jiménez is aware of the problems that need to be resolved and states that it is
necessary to promote greater theoretical rigorousness.

In 1991, in the IPA monograph On Freud’s ‘Analysis Terminable and Interminable’ , T.

Eskelinen de Folch published an article entitled

‘The obstacles to the analytic cure’, in which she studies the role of projective identifi cation in
provoking forms of alienation of the personality in external reality (Eskelinen de Folch, 1987).
These forms constitute a diffi culty in the analytic process, a particular form of repetition
compulsion linked to concealed nuclei of the personality. The author believes that the concept of
projective identifi cation and a detailed knowledge of the various forms that it can take would
allow us to recuperate the object relationship which exists behind the more or less severe
fragmentation of the self and objects through explorations of the transference. Like others by the
same author, this article is generally of a Kleinian stamp, without any concessions to hazardous
theoretical intersections.

In the same fi eld of Catalan analysis, I would like to mention a book by A. Pérez-Sánchez,
Análisis terminable (Pérez-Sánchez, 1997), published in Spanish. Although he does not deal
specifi cally with the subject of projective identifi cation, when speaking about the termination of
the analytic process and in the analysis of the case presented, he uses the concept with pertinence
and in adherence to Kleinian theory.

I think that this short review, although incomplete, provides evidence of a considerable
familiarity with and use of the concept of projective identifi cation in Catalan psychoanalysis
coherent with the Kleinian theoretical affi liation of the Spanish Society. It is also 215

Jorge Canestri

possible to observe a greater preoccupation with theoretical coherence, whenever use is to be

made of concepts in some way connected to that of projective identifi cation, but deriving from
different theoretical patterns.

I will now conclude this very brief incursion into Italian and Spanish psychoanalysis regarding
the destiny of the concept of projective identifi cation. Can we make a diagnosis of the present
situation and a prognosis for the future? Can we propose some suggestions?

Some provisional conclusions

As far as the present situation is concerned, at least four different ways of conceptualizing and
using projective identifi cation can be identifi ed.

First, analysts who are theoretically and clinically Kleinian and who use the concept both in
theory and in practice in a way that is not dissimilar to that described by Spillius (1992), both for
‘pathological’ and for ‘ordinary’ defensive processes. The description of the sub-types of
projective identifi cation varies according to preferences for one Kleinian author or another:
evocatory or non-evocatory (Spillius, 1994), attributive or acquisitive (Britton, 1998a), and so
on.

My impression is that this category is prevalent in the Spanish Society, and more in the minority
in the Italian and Madrid Associations.

Second, analysts who integrate the concept in a general theoretical grouping that includes
segments of various theories: Freudian-Kleinian when there is a tradition binding them to the
French analysis of the past or the present; Kleinian-Bionian-Winnicottian when they are close to
British psychoanalysis. I think that this is the majority group and it is in these cases that there is
need for refl ection about the coherence of the resulting theory and the possible denaturalization
of the original concept. Generally speaking, the authors in question do not seem to be too
preoccupied with taking these risks into consideration or with averting them.

Third, analysts who, besides favouring concepts stemming from various theories, add to them
data and hypotheses from other disciplines that can be used to support or as ‘proof’ of the
validity of the psychoanalytic concepts, or as
‘updates’ of the theory in general.

Fourth, analysts who are aware of the diffi culties deriving from these operations, as well as the
diffi culties perhaps present in Kleinian 216

Projective identifi cation in Italy and Spain theory itself, and who seek to investigate the analysis
of the concept and to relate it to an integral theory of the development of the psychic apparatus.
This group of analysts – to my mind very much in the minority – postulates the need for greater
theoretical rigorousness.

Making a prediction for the future is not easy and would in any case be presumptuous. It is,
however, reasonable to hypothesize that the concept of projective identifi cation will continue to
be central to clinical practice. If it has been integrated into theoretical models that at fi rst sight
seem to be incompatible with those from which it stems, this is also a sure indication of its
validity and usefulness both as a theory and in the clinical fi eld. It is likely that these theoretical
patch-works will continue, with the addition of interdisciplinary concepts.

This might lead to an increase in the babelization of psychoanalytic language and could put its
theoretical coherence at risk. The call of certain authors for theoretical rigorousness would
therefore seem legitimate.

Bion emphasized the need to identify points of congruence between the different psychoanalytic
models, and this is undoubtedly possible and desirable. However, the points of congruence can
reveal not only the convergence, but also the divergence, of certain positions within one or more
theories. I think that a careful analysis of the differences and the compatibilities can only be of
benefi t to our discipline.

217

13

Projective identification in contemporary

French-language psychoanalysis 42

Jean-Michel Quinodoz

The gradual advance of Kleinian concepts in

psychoanalysis as practised in the

French-speaking world

A concept considered to be typically Kleinian

Projective identifi cation was described by Klein in 1946. From then on, it has been considered
by many analysts to be one of the characteristic concepts of the Kleinian school, a source of
admiration for some and an excuse for criticism for others. It has been said that the concept of
projective identifi cation is the pons asinorum of Kleinian theory, in other words that it is the
prerequisite for understanding Kleinian thinking, just as Pythagoras’s theorem is a key concept in
descriptive geometry. René Diatkine would poke gentle fun at those psychoanalysts, Kleinians or
not, who used the concept: It is easy to measure the degree of Kleinianism of a keynote speaker:
just calculate the time that elapses between the beginning of the talk and the fi rst use of the term
‘projective identifi cation’. The shorter the time, the more Kleinian the speaker!

(an anecdote told me by Paul Denis)

42 Translated by David Alcorn.

218

PI in French-language psychoanalysis This joke is not without its ironic side: it hints at the
resistance that Kleinian thinking has encountered in French-speaking countries,

and still does, though to a lesser degree than before. In this chapter,

I shall explore the importance of the concept of projective identifi cation for contemporary
French-language psychoanalysts – focusing on those who belong to the International
Psychoanalytical Association

– before concluding this overview with a few hypotheses concerning the reasons for such
resistance.

A phenomenon often described in clinical terms

before being conceptualized

Though it is often labelled as typically Kleinian, the phenomenon that, in 1946, Melanie Klein
called projective identifi cation does not in fact belong to Kleinian thinking as such – far from it,
indeed, because the idea was already fl oating around long before Klein conceptualized it. The
idea of projection is implicit in Ferenczi’s article

‘Introjection and transference’ (Ferenczi, 1909), and exists in many of Freud’s own writings too.
Similar phenomena are also described in various psychoanalytic texts dating from the 1920s and
1930s, although the authors do not make any clear distinction between these and projection as
such. French-speaking psychoanalysts too have made similar statements in the past. For
example, with reference to female homosexuality, R. de Saussure highlighted the partial
character of what is projected onto the homosexual’s partner, as well as the narcissistic nature of
this kind of identifi cation:

To my mind, homosexual fi xation has much more to do with a division of narcissism than with
an object-related fi xation.

[. . .] Such women project outside of themselves either their masculinity or their femininity, and
they are attracted to women who are the opposite of themselves. Indeed, their aim is much more
to identify with their partner than to love her in an object-related way.
(de Saussure, 1929, p. 70)

In this complex movement, the idea of projective identifi cation highlights the details that come
into play in projective and identifi catory phenomena.

219

Jean-Michel Quinodoz

France: a two-stage development

Klein’s ideas made only very slow headway in the French-language countries, and in France in
particular, where it was a two-stage process. Though her theories on child psychoanalysis were
acknowledged in the 1950s, her ideas then fell into relative disuse before resurfacing once again
in the 1970s and thereafter continuing to develop right up to the present day.

Back in the 1930s, Jacques Lacan was undoubtedly one of the fi rst psychoanalysts in France to
recognize the importance of Klein’s thinking. What interested him in particular was her idea of
the early stages of the Oedipus complex; this led him to ‘refocus the Oedipal question in terms of
triangulation, while taking into account the various contributions of the Kleinian school’
(Roudinesco and Plon, 1997, p. 746). Later, from the 1950s on, Klein’s ideas aroused increasing
interest among child psychoanalysts such as René Diatkine and Serge Lebovici. At the same
time, Lacan was beginning to criticize Klein’s theses, even though he himself had made use of
them in constructing some of his own conceptions; Lacan’s ambivalent attitude towards Klein
was to have a signifi cant

impact on many psychoanalysts in French-speaking countries. One particular anecdote is a

signifi cant indication of Lacan’s attitude towards Klein. In the early 1950s, René Diatkine
translated Klein’s The Psycho-Analysis of Children (Klein,
1932b) into French for the fi rst time. At that time, he was in analysis with Lacan. One day,
Lacan asked the enthusiastic young translator to lend him a copy of the translation. Lacan never
gave it back to Diatkine, claiming that he had mislaid it. Unfortunately, it was the only copy, and
Diatkine’s translation was lost for ever. It was only many years later that the book was fi nally
translated, but by someone else.

In France, after a silence that lasted 20 years, renewed interest in Klein’s work began with the
arrival in Paris of James Gammill in 1969, followed by that of Jean Bégoin and Florence Bégoin-
Guignard in 1970. Gammill, an American-born psychoanalyst, was trained in London; his
analyst was Paula Heimann, and Melanie Klein was one of his supervisors. Jean Bégoin came
from Annecy and was analysed by Marcelle Spira in Geneva, while Florence Bégoin-Guignard
had trained in Geneva. These psychoanalysts joined forces with Donald Meltzer, whose ideas
were rapidly spreading throughout French psychoanalytic circles, thanks in particular to the
publication of his The Psycho-Analytical Process (Meltzer, 1967), translated into French by Jean
Bégoin and 220

PI in French-language psychoanalysis Florence Bégoin-Guignard. In the years that followed,

this initially small group gradually expanded and came to include Geneviève Haag and Didier
Houzel (both of whom had supervision in London), as well as Cléopâtre Athanassiou, a pupil of
Esther Bick. In addition, Pierre and Claudine Geissmann in Bordeaux were trying to discover
how to overcome the diffi culties they were having in treating severely psychotic children; they
travelled to London on a regular basis between 1982 and 1987 in order to have supervision with
Hanna Segal and to attend meetings of the British Psychoanalytical Society. Thus, in France,
Kleinian thinking began initially to spread among child and adolescent psychoanalysts, as well as
those who treated psychotic patients.

Gradual growth in Switzerland, Belgium and

French-speaking Canada

In other French-language countries, Kleinian ideas began to make headway in the 1950s and
1960s. In the French-speaking part of Switzerland, it was Marcelle Spira who initiated the
movement; she settled in Geneva in 1956 after leaving Argentina, which was also where she
trained ( J.-M. Quinodoz, 2002b). During several decades, Spira trained many Swiss and other
European psychoanalysts, and had as guest speakers at her seminars Melanie Klein, Betty Joseph
and Herbert Rosenfeld, who all contributed to the spread of Kleinian ideas, especially in Geneva,
in spite of the strong opposition which that theory encountered. In the 1980s, Hanna Segal
conducted a seminar in Geneva which lasted for all of ten years. In Belgium, Kleinian ideas
began to make some progress in the 1960s, again among child psychoanalysts in particular,
thanks to contacts with neighbouring countries such as the Netherlands. In Canada, French-
speaking psychoanalysts learned of Klein’s ideas thanks to J-B. Boulanger, who translated
Klein’s book, The Psycho-Analysis of Children (1932), to C. Scott in 1954, and then to Henri
Rey who made regular visits to the country in the 1970s.

The main contributions of the French-speaking world

Making projective identifi cation more familiar to the French world

Starting in the 1970s, several French-speaking psychoanalysts who use the concept of projective
identifi cation in a Kleinian or 221

Jean-Michel Quinodoz

post-Kleinian context have published papers, only a few of which have been translated into
English.

When we talk of projective identifi cation, Florence Guignard is, for most French-language
psychoanalysts, the name that most readily springs to mind. Her many publications (successively
as F. Bégoin, F. Bégoin-Guignard and F. Guignard) testify to her commitment to the idea; one
such is the interesting exchange of correspondence she had with M. Fain (1984) on the topic of
projective identifi cation and hysterical identifi cation. In the course of that discussion, F.
Guignard emphasized the stumbling blocks that prevented French-speaking psychoanalysts from

having a proper understanding of the notion of projective identifi cation. Most of them, she
argued, remained attached to a ‘classic Freudian’ approach to splitting – i.e., one that related only
to splitting of the ego and not to that of the object at the same time. She also pointed out that they
had some diffi culty in understanding the concept of part-object introjection (Guignard, 1984, p.
521). G. Bayle’s controversial point of view on the concept of splitting –

according to him, the word ‘splitting’ designates simply the result of a defensive process, never
the defence mechanism itself; in other words, it is not an action that takes place in the mind
(Bayle, 1996, pp. 1334–1335) – led Guignard (1996) to put the notion back into its proper
conceptual context. It must be noted, all the same, that ideas have gradually evolved since the
1970s and that the term is no longer taboo; anyone who talks nowadays of splitting would not
automatically be labelled ‘Kleinian’, as once was the case.

It may be useful to remind French-language psychoanalysts, who sometimes have their own
particular idea of what splitting is all about, that a distinction should be drawn between what
Freud called splitting and Melanie Klein’s defi nition of the term (Canestri, 1989). For Freud,
splitting of the ego is the result of a denial of some perception of reality, such that the ego fi nds
itself divided – in a passive sort of way, one could say. On the other hand, for Klein and the post-
Kleinians, splitting is an active defence mechanism that has many different modalities. In the
preface to the French translation of his Dictionary of Kleinian Thought , Hinshelwood (2000, p.
3) points out that the English language can make subtle distinctions between different kinds of
splitting (splitting up, splitting off, separated off, separated apart, etc.), a possibility that does not
exist in French. ‘I wonder if the linguistic constraints of the French language, in relation to this
fundamental concept (splitting), make Melanie Klein’s 222

PI in French-language psychoanalysis thinking seem less subtle to the French-speaking reader,

leading to less interest in her ideas.’

As for more recent publications, there is the chapter that Guignard devoted to projective identifi
cation in her Épître à l’objet (Guignard, 1997), in which she reminds the reader that the concept
includes a wide spectrum of associated phenomena. Several child and adolescent psychoanalysts
refer to the idea of projective identifi cation in their work –
A. Anzieu, C. Athanassiou, G. Haag, and D. Houzel.

When D. Ribas (1992) emphasizes the absence of projection in infantile autism, he is in fact
referring exclusively to adhesive identifi cation, which, as he says, is a more primitive
mechanism than projective identifi cation.

Projective identifi cation and projection – in what

way are they different?

Alain Gibeault asked this question in his paper ‘De la projection et de l’identifi cation projective’
(‘On projection and projective identifi cation’) (Gibeault, 2000a). In that text, he considers
whether the concept, as defi ned by Klein in 1946, simply overlaps that of projection as
described by Freud or whether it adds something to Freud’s notion.

After discussing the various meanings of introjection and projection as processes, Gibeault
highlights the diffi culty that we encounter when trying to specify the differences between
projection and projective identifi cation, given the contradictory positions to which the question
gives rise. In his conclusion, Gibeault goes back to Freud’s defi nition of projection: ‘an aspect
of oneself which, denied or repressed, is then attributed to an external object; in so doing, the
subject does not want to rediscover in the other person what has been projected into that person’
(Gibeault, 2000a, p. 742).

Contributions from Switzerland, Belgium and Canada

In the French-speaking part of Switzerland, Danielle Quinodoz has published two remarkable
papers that deal with projective identifi cation stricto sensu . In ‘Interpretations in projection’ (D.
Quinodoz, 1989), she suggests an innovative technical approach:

223

Jean-Michel Quinodoz

In a nutshell, this form of interpretation consists in the analyst’s lending his voice directly to the
part of the patient projected into him. The analyst speaks in the fi rst person, saying aloud what
he thinks the projected part of the patient would have told him if it had been able to speak.

(D. Quinodoz, 2002, pp. 112–113)

In Words that Touch , D. Quinodoz makes a detailed study of projective identifi cation in the
light of clinical examples, emphasizing the notion of projective counter-identifi cation that León
Grinberg (1962) introduced.

According to D. Quinodoz, the diffi culty that French-language psychoanalysts encounter in

imagining the idea of bodily fantasy leads to a problem with mentally representing the notion of
projecting a fantasied part of the ego that involves bodily fantasies; this obstacle, she feels, is
linked inter alia to the continuing impact of Lacan’s ideas – for him, the body could refer only to
concrete reality, thereby excluding bodily fantasies from the psychoanalytic fi eld.
As far as my own work is concerned, I have in the main highlighted the phenomena that involve
integration, when projective identifi cation weakens and leaves more room for introjection; this
occurs both in the experience of

‘buoyancy’ when separation anxiety is worked through in the psychoanalytic process ( J.-M.
Quinodoz, 1992), and in the kind of dream that ‘turns the page’ – dreams that appear to be
regressive but in fact have to do with integration ( J.-M.

Quinodoz, 2001). In French-speaking Switzerland, projective identifi cation is a notion that is

used also by child analysts, including those who work with psychotic children (Manzano and
Palacio Espasa, 1986), as well as in mother-infant therapies where it is particularly useful for
explaining normal and pathological projection (Manzano, Palacio Espasa, and Zilkha, 1999).
Contributions from various other analysts also take the idea of projective identifi cation into
account, for example the creative and innovative Kleinian perspective suggested by M. Spira
(1985).

Projective identifi cation would appear to be a concept that is both well accepted by Belgian
psychoanalysts and used by them in their clinical work. The reports drawn up by M. Haber and J.
Godfrind-Haber (2002) bear witness to this, as do the various discussions that followed on from
their work.

As regards French-speaking Canadian analysts, Kleinian and post-Kleinian conceptions have

been further developed, in particular 224

PI in French-language psychoanalysis by P. Drapeau and G. da Silva. Recently, L. Brunet and

D. Casoni (2001) have tried to ‘sort things out in the conceptual Tower of Babel’ that is
projective identifi cation; they argue that the connection between projective identifi cation and
the containing function is particularly helpful as a model for understanding how some analysands
use the object in the transference–countertransference relationship. However, projective identifi
cation may be used defensively, with the risk that some analysts could attribute to the notion all
sorts of meanings, ‘some [of which] have resulted in dubious countertransferential “alibis”
(Aulagnier 1984), and hence have contributed to the hesitation many psychoanalysts have in
referring to the concept’

(Brunet and Casoni, 2001, pp. 137–138).

The present-day situation: informal soundings

What do contemporary French-language psychoanalysts think of the concept of projective

identifi cation? To what extent do they use it in their clinical practice? To answer these
questions, I sought the opinion of some of my analyst colleagues, members of the International
Psychoanalytical Association, either in informal discussions with them or through examining
what they have written on the subject ( J.-M. Quinodoz, 2002a). These soundings enabled me to
divide them into four main groups.

Psychoanalysts who refer directly to Kleinian concepts

This fi rst group comprises not only those psychoanalysts who were trained by Kleinians but also
those who are receptive to Kleinian thinking without necessarily having had any formal training
in that approach. For analysts who belong to this group, the concept is a fundamental
contribution to psychoanalytic theory and they make use of it in their clinical work – just as they
refer to Kleinian thinking in general as an integral part of their clinical practice. In

France, those who belong to this group are, for the most part, psychoanalysts who work with
children and adolescents, and with psychotic patients; in other French-language countries, there
are some adult psychoanalysts who belong to this group.

225

Jean-Michel Quinodoz

Psychoanalysts who consider projective identifi cation

to be one of many projective phenomena

It becomes obvious when discussing the question with them that, since the early 1980s, many
French-language psychoanalysts have shown increasing interest in projective mechanisms as
such.

Particularly attentive to what is projected into the analyst’s mind, they raise issues that have
mainly to do with what the analyst does with such projections. For the most part, they tend to
speak in terms of the countertransference rather than of projective identifi cation as such.

There is also some considerable variation in the way that these analysts describe the projective
phenomenon on which they are focusing; this is more easily detected in their writings than in
oral discussions with them.

For example, in his ‘The dead mother’, André Green (1980) describes clinical phenomena that
closely resemble projective identifi cation, yet in that paper he writes only of ‘projection’, never
of ‘identifi cation’. In a later paper, Green (1990) introduces the idea of ‘foreclosed projection’ (
projection forclose ), which he locates somewhere between ‘ex-corporation’ and ‘projection’, in
order no doubt to underline the distinction he draws between that notion and projective identifi
cation. For Racamier, post-Kleinian psychoanalysts use the term ‘projective identifi cation’ in
much too wide a sense; he preferred the term ‘projective injection’ ( injection projective ) to the
one Segal suggested – ‘identifi catory projection’ – because, in his opinion, ‘projective injection
is carried out at one go and in the original tongue, as it were, directly and proximately’
(Racamier, 1992, p. 99). Michel de M’Uzan (1994) calls

‘paradoxical thinking’

( pensée paradoxale ) a clinical phenomenon that closely resembles what the analyst may
experience in the countertransference when the patient resorts to projective identifi cation. He
does not, however, make any reference to splitting, projection or identifi cation; in his view,
‘paradoxical thinking’ belongs to an inter-subjective space that he calls the ‘chimera’, one which
is neither wholly the analyst’s nor wholly the patient’s. The psychosomatic school of thought in
Paris refers to ‘projective reduplication’ (
reduplication

projective ), a term introduced by C. David and M. de M’Uzan.

R. Roussillon describes the type of communication found in the 226

PI in French-language psychoanalysis

‘narcissistic transference’ in terms that would seem to imply, inter alia , the idea of splitting: A
kind of transference

‘by reversal’ replaces or is added to the

‘displacement’ transference typical of the transference neurosis; here, the patient, split-off from
any possibility of integration and therefore ‘in parallel’, as it were, makes the analyst experience
what proved impossible to experience properly in the patient’s own life.

(Roussillon, 1999, p. 14)

A. Eiguer highlights a particularly corrosive attempt to control the object by means of a kind of
projection that he

calls ‘narcissistic induction’ (Eiguer, 1989). J-Cl. Rolland (quoted in Smith, 2002) described a
session in which he felt overwhelmed by terror as if he were ‘identifi ed with a nightmare’, so
powerful was the impact of his patient’s transference. In his view, an ‘over-restrictive theory’

prevented him from discovering that some clinical experiences force the analyst to participate in
the ‘action’ of the transference, in such a way as to ‘embody’ the patient’s states of anxiety.

It is obvious from what I have just said that practically every one of the psychoanalysts I have
mentioned has his own way of putting things. This makes it diffi cult to decide what
differentiates any one concept from the others –

and all the more so, indeed, because the analysts concerned rarely refer to similar concepts that
have been defi ned earlier. The resulting impression is one of dispersal and a lack of conceptual
unity. In suggesting some new notion, they hardly ever make any reference to the idea of
projective identifi cation, though the term is a well-defi ned one; and when it is referred to, it is
often treated as banal, or classed as just one modality of projection among many.

Psychoanalysts who accept only those psychoanalytic

concepts established by Freud

This is the third category – the concept of projective identifi cation is not referred to at all by
these analysts, because they adhere strictly to Freud’s writings. Since projective identifi cation is
not one of the 227

Jean-Michel Quinodoz
concepts specifi cally described and developed by Freud, it hardly ever rates a mention in their
work.

Psychoanalysts who are more or less ambivalent

towards Kleinian ideas

Finally, there are some psychoanalysts who are more or less ambivalent towards Kleinian
thinking – and, therefore, towards projective identifi cation. Some are quite scathing in their
criticism, which, moreover, is often based on prejudice: they unwittingly reveal their lack of true
knowledge concerning Kleinian theory (though they think they understand it well). Some
analysts, for example, are still convinced that Klein spoke only of the mother’s role, never of the
father’s, in the parent–infant relationship – it is as though they had never read Klein’s writings.
For example, in his article on projective identifi cation, F. Pasche (1982) is convinced that his
criticism of projective identifi cation is perfectly justifi ed when he claims that a Kleinian
psychoanalyst simply returns the patient’s projections without fi rst transforming them and that,
as a result, the countertransference is always ‘put to one side’ (Pasche, 1982, p. 410). When confi
rmed analysts speak with apparent authority on such topics, their albeit unreliable statements
often go unchallenged among their readers or their audience, who do not check their sources.

On the other hand, the ambivalence of some analysts is tinged with respect. This is the case, for
example, of J.

Laplanche, who has never disguised his interest in Klein’s contributions even though he remains
wary of

‘Kleinianism’, of its ‘proselytism’, its ‘failure to address basic questions’ and its ‘hegemony’
(Laplanche, 1992, p.

215). In expressing his fear of Kleinian attempts at hegemony, Laplanche, in my view, has
identifi ed one of the sources of the ambivalence that some analysts feel – fascinated by (and
sometimes drawing their inspiration from) Kleinian concepts, they do not want to admit as much.

Projective identifi cation in its own specifi c context After this overview that has enabled us to
see just how diverse the points of view on the question are, I fi nd it necessary to refocus my 228

PI in French-language psychoanalysis ideas and concentrate on the main theme of this essay, i.e.
on projective identifi cation and its own specifi c theoretical and technical context. This brief
reminder will then enable me to highlight the convergences and divergences between the various
points of view. I should point out, however, that, in

making comparisons, I do not intend to give the impression that one point of view is superior to
another; my aim is to further our understanding of how the various analysts work and of the
genuine differences between them.

Careful attention to the transference–

countertransference situation
In order to pinpoint and to analyse the phenomena linked to projective identifi cation, technical
considerations are important, because we have to adopt an approach that closely monitors the
development of the transference–

countertransference situation. The constant and meticulous attention given to the details of what
is happening in the transference and countertransference enables us to distinguish between what
comes from the patient and what our own contribution is to the continuous to and fro movement
of projections and introjections typical of the analytical relationship. This is particularly useful
when the emotionality involved is tinged with persecutory aspects, which encourage projection,
or depressive elements, which favour insight; these differences in tone must be closely followed.
For Klein and Kleinian psychoanalysts, the most appropriate way to obtain this kind of clinical
material is to establish a clear and well-defi ned psychoanalytic setting, in order to analyse the
fantasies and affects in a ‘global-situation’ context and to limit as far as possible any acting-out.
From this point of view, the Kleinian approach is a demanding one. That said, I do not believe
that following so precisely the hic et nunc of what is being experienced in the setting necessarily
implies that the psychoanalyst who adopts such a perspective loses sight of the process as a
whole.

Bion’s contribution is inseparable from that of Klein

W. R. Bion’s contribution (Bion, 1962a, 1967) to the concept of projective identifi cation is
fundamental in terms of the development of Klein’s ideas; if it is isolated from the sources on
which it is based, 229

Jean-Michel Quinodoz

some degree of misunderstanding is only to be expected. As J. Chasseguet-Smirgel (personal

communication, 2002) has said, to omit Klein and refer only to Bion is ‘an indelicate way of
nullifying the line of descent that runs from Klein to Bion’.

Bion’s starting point was the pathology of the paranoid-schizoid position. His fundamentally
innovative development of Klein’s work was the distinction he drew between pathological
projective identifi cation and the normal, neurotic form of projective identifi cation that
underpins empathy. The difference between paranoid anxiety and depressive position anxiety is
also the basis on which Bion developed his theory of the container and the transformation of β-
elements into α-elements; the latter are required for symbol formation and for growth, while β-

elements can only be evacuated – having no developmental potential, they lead only to concrete
thinking (Segal, 1957).

The idea of ‘capacity for reverie’, the distinction between normal and pathological projective
identifi cation and the transformation of β-elements into α-elements throw light on the
psychoanalyst’s work and help us to understand it better: the patient’s projections are taken in,
transformed, then returned – they are not simply pushed back into the patient without having
been worked through, as some unfounded criticism of projective identifi cation would have us
believe. If that were the case, the psychoanalyst would resemble the dysfunctional mother in
Bion’s description, so incapable of tolerating her infant’s projections that she in fact reinforces
her child’s projective identifi cation and renders it meaningless. Also, the idea of the mother’s
‘capacity for reverie’ is often misunderstood as implying a dyadic relationship that excludes the
father; that, however, was not Bion’s idea. For him, maternal reverie means that the baby and the
father are both present in the mother’s mind. From that point of view, Winnicott, not Bion, was
the one who explicitly declared that the containing mother is in a dyadic fusional relationship
with her infant.

Convergence and divergence

Increasing communication between different

currents of thought in psychoanalysis

As I have pointed out, projective identifi cation is a relatively well-defi ned concept that covers a
wide range of projective phenomena 230

PI in French-language psychoanalysis extending from neurosis to psychosis. Moreover, it

participates in the differentiation between the more elaborate defence mechanisms linked to
repression and the more primitive ones; as such, it is a particularly suitable instrument for the
analysis of narcissistic, borderline or psychotic patients.

Over a long period of time, projective identifi cation was felt to belong exclusively to the
Kleinian school of thought, but this is no longer entirely the case. Many contemporary
psychoanalysts accept the concept and integrate it into their clinical work, just as they do its
underlying notions such as denial of psychic reality, splitting, projection and introjection of parts
of the self, and the object’s identifi cation with the projections received (Sandler, 1987b).

This is particularly true of French-language analysts in Switzerland, Belgium and Canada, who
are probably much more in contact with various international currents of thought in
psychoanalysis than their counterparts in France.

That said, the situation has changed somewhat in France too over the past few decades, as the
various articles and papers I mentioned earlier go to show. Many French analysts have told me
that they pay very close attention to the transference–countertransference experience and that, in
their view, projective identifi cation is a fairly well-recognized concept even though they
themselves may prefer to use others. As a result, it is no longer possible to contrast Kleinian and
Freudian technique in so radical a manner as de M’Uzan (1994) – he considered the Kleinian
approach to be too ‘tactical’

because it focused to an excessive degree on the details of the transference–countertransference

experience and contrasted it with the

‘strategic’ approach that he claimed was exclusive to the Freudian model.

Nowadays, when psychoanalysts belonging to the French school of thought show interest in
patients who are diffi cult to treat, their perspective is mainly that of ‘borderline’ cases, and when
they refer to the British school, they readily quote Winnicott and Bion, as though to avoid
mentioning Klein. Generally speaking, there now seem to be fewer differences between the
technique employed by these analysts and that of practitioners – in Italy, Germany or the
Scandinavian countries as far as Europe is concerned – who have integrated at least some
Kleinian conceptions. I believe that these developments are the result of an increasing exchange
of ideas on the international level –

especially through symposia such as the one in 231

Jean-Michel Quinodoz

Brighton which brought together French and British psychoanalysts; the latter, for example,
showed considerable interest in notions that have done much for the originality of the French
current of thought in contemporary psychoanalysis.

Signifi cant differences in technique

Narrowing the gap between different schools of thought is a highly positive development, but we
must not play down the differences that remain, some of which are particularly relevant to the
topic of this paper. It seems to me that it is in the technical approach to the psychoanalytic
treatment of neurotic, narcissistic, borderline or psychotic patients that one of these differences is
particularly evident.

For some psychoanalysts, the classic analytical technique – with its couch-and-armchair setting –
is suited above all to neurotic patients, those who, from the preliminary interviews onwards,
show that they can make use of their capacity for symbolic representation in dealing with their
intrapsychic confl icts. When there are defi ciencies in that capacity, these analysts feel that
classic psychoanalytic treatment is not indicated and tend to offer the patient psychoanalytic
treatment with each sitting opposite and in full view of the other. According to Gibeault (2000b),
this

‘face-to-face’ approach is a way of avoiding pathological regression and the consequent risk of
psychic disorganization or even disintegration.

Analytic work on the couch is necessarily ‘traumatic’ because of its suspension of sight and
action, for it involves a return to an anxiety about non-representation that has to be negotiated
through capacities for topographical and formal regression. Face-to-face analysis restores what
was suspended, in this case visual support for the fragility of the ego during a resumption of the
psychic functioning that is also a capacity for forming a potential space for play (Winnicott
1971).

(Gibeault, 2000b, p. 383)

This approach is very different from the Kleinian one, which applies equally to neurotic patients
and to those who are narcissistic, borderline or psychotic – and perhaps even more so to the
latter, if we refer to the clinical examples that are reported in various 232

PI in French-language psychoanalysis publications. It is true, of course, that Kleinian

psychoanalysts do resort to the facing-each-other technique whenever the patient fi nds lying on
the couch to be unbearable – but when they do so, they maintain the well-defi ned
psychoanalytic setting because they feel that that setting creates the best possible conditions for
identifying and analysing the transference–countertransference experience.

These different technical approaches are based on differences in theoretical conceptualisations

that I cannot develop here, because

these issues go somewhat beyond the scope of this chapter. The

broad overview that I have given should, all the same, leave enough space for different shades of
opinion – some of which are quite contrasted – to be expressed.

Infl uences that go far back

How is it that French-language psychoanalysts – and especially those in France itself – have only
recently shown any interest in the projective phenomena that occur in the transference–
countertransference relationship, as well as in the concept of projective identifi cation and the
theoretical and technical context that surrounds these notions?

There is of course a tradition that goes back to the introduction of Freud’s ideas into France in
the 1920s; at that point, they dealt mainly with the neuroses, since he was still working on their
application to depression and the psychoses. However, I think that in spite of disagreement on
certain fundamental issues, the ideas that Lacan began to put forward from the 1950s on have
infl uenced and continue to infl uence psychoanalytic thinking in France and in other French-
language regions.

I feel that the ‘return to Freud’ that Lacan advocated in 1953

focused attention more or less permanently on the neuroses and verbal communication to the
detriment of the psychoanalytic approach to more severe pathological states and non-verbal
communication. The ‘return to Freud’

did not concern the whole range of his work, but mainly the period between 1900 and 1905,
when his emphasis was on the part played by language, symbolism and condensation in dreams
and in jokes. As a result, a considerable amount of Freud’s contributions from 1915 on were
pushed into the background – and this was when Freud was exploring the possibility of using
psychoanalysis to treat depression and the psychoses, thereby 233

Jean-Michel Quinodoz

opening the way to post-Freudian ideas ( J.-M. Quinodoz, 2000).

Moreover, Lacan was fi rmly against making use of the countertransference in the course of an
analysis, as F.

Duparc (2001) has pointed out; that is why it took several decades for the concept to be accepted
into mainstream French psychoanalytic thinking even though it had been introduced in the early
1950s. Among the reasons for rejecting it, Duparc suggests that Lacan considered the
countertransference to carry the risk of setting up a mirror relationship that would be dyadic in
nature and exclude the presence of any tiers or third party. The point of view of those
psychoanalysts who belong to the Paris Psychoanalytical Society or the French Psychoanalytical
Association

has changed over the years as regards the countertransference; the idea has gradually come to be
generally accepted, thanks in particular to the work of S. Vidermann, M. Neyraut, P. Aulagnier
and J. McDougall. On the other hand, the point of view of those psychoanalysts who belong to
the Lacanian school of thought is the same today as it was in Lacan’s day: rejecting the
countertransference is one of the principles on which their work is based (Widlöcher, 2003). In
addition, Lacan felt that, in cases of psychosis, the transference could not be analysed. For
Lacan, foreclosure ( forclusion ) is characteristic of the psychoses, but he never indicated, how,
in his view, foreclosure could be reversed; as a result, his point of view concerning the treatment
of psychosis remained a purely speculative one (Diatkine, 1997).

Lacan’s infl uence made itself felt not only through his theoretical and technical conceptions but
also via the fascination his personality evoked. I cannot help thinking that the ambivalent attitude
of some analysts towards Kleinian ideas is very similar to that of Lacan himself: there is perhaps
a hint here of unconscious transgenerational identifi cation with his personality.

In a recently published book, the psychoanalyst Maria Pierrakos (2003) discusses the
transgenerational indentifi cations that she has identifi ed in post-Lacanian analysts – with
whom, indeed, she is very familiar, having for twelve years served as stenotypist at Lacan’s
Seminars . She points out that many of Lacan’s followers identify not only with the ‘Master’s’
way of putting things – plays on words, puns and spoonerisms – but also with many aspects of
his personality such as his love of paradox and enigmatic statements or his ‘disdainful mocking’
of anything that had to do with feelings. ‘Have they all caught the Witz -bug?’ she asks. ‘What is
it that is both hidden by and 234

PI in French-language psychoanalysis yet revealed through this desire to astound, to captivate,

to hypnotize?’

(Pierrakos, 2003, p. 35). She goes on to say that those who were analysed by Lacan and attended
his Seminars had the impression that Lacan’s

public face, with its cynicism and its coldness, was completely at odds with what they knew of
the man himself.

That transference, forever unresolved, has kept them irrevocably in slavery, because each of
them thought that the entire Janus-like fi gure was for him or her alone.

(Pierrakos, 2003, p. 38)

A plea for genuine controversy

It is time for me to conclude. To my mind, the concept of projective identifi cation is

fundamental to our work as analysts; I was delighted to accept the invitation that Elizabeth
Spillius and Edna O’Shaughnessy addressed to me, all the more so since, based in Geneva, I feel
close both to the British (and, more specifi cally, Kleinian) tradition and to the French school of
thought. When I focus on projective identifi cation and its use in the psychoanalytic approach to
narcissistic, borderline or psychotic patients, it does not mean that I am neglecting the original
contributions of the French tradition as regards neurotic patients. On the contrary – I believe that
the variety of points of view in contemporary psychoanalytic thinking is a great advantage; but if
we are to avoid frittering away the inheritance that Freud left us, we need a genuine debate more
than ever. For this to be productive, we have to create favourable conditions for true dialogue, as
R. Bernardi (2002, p. 851) noted: ‘When this occurs, controversies promote the discipline’s
development ( i.e. that of psychoanalysis ), even when they fail to reach any consensus’.

235

SECTION3

The United States

Introduction

Elizabeth Spillius

The idea of projective identifi cation has aroused considerable interest among American
psychoanalysts and psychotherapists, a process that is here described in two papers. The fi rst, by
Roy Schafer, is based on his very

extensive knowledge of the various schools of thought in American psychoanalysis and their
response to the idea of projective identifi cation. The second, by Elizabeth Spillius, briefl y
reviews some of the considerable body of American psychoanalytic literature on the topic of
projective identifi cation. These two expositions are followed by three noteworthy papers. The fi
rst, by Arthur Malin and James Grotstein, is one of the earliest American papers on the topic of
projective identifi cation. The second, by Thomas Ogden, expounds his views on the topic about
a decade later. The third paper, written for the present volume, presents the views of Albert
Mason, a British analyst who settled in Los Angeles in the late 1950s, and whose views on the
concept express both his British origin and his subsequent clinical experiences of projective
identifi cation with American patients.

239

14

Projective identification in the USA

An overview

Roy Schafer

I do not fi nd a center to the uses of projective identifi cation (PI) in the USA. The concept is
being used as though it fi ts into whatever happens to be one’s established theory and mode of
practice.

Variations extend from Grotstein’s close adherence to Klein, Bion, and Meltzer to informal or
improvisational applications in specifi c clinical instances. Some are not clearly different from
Freud’s projection or from attribution, imitation, manipulation and persuasion.

I will organize this scattered material around a listing of what I consider the conceptual issues
encountered or created by recent authors. Some of these issues are new versions of age-old
epistemological controversies.

1 The clinical scene in the USA has featured the rising infl uence of such object relational
thinkers as Klein, Winnicott, Bion, Fairbairn, Ferenczi and Loewald and such interpersonal-
relational thinkers as Harry Stack Sullivan, Stephen Mitchell, and Jay Greenberg. This change
has been evidenced by widespread attention to the interplay of transference and
countertransference. Sometimes, projective identifi cation is implied rather than named in
clinical interpretations and discussions.

2 In general, use of the idea tends to be stripped of its specifi cally Kleinian origin (the same may
be said of those counterproduc-tively overshadowed concepts introjective identifi cation and
introjection ). My impression is that affi liation anxiety might be playing a 240

Projective identifi cation in the USA: an overview part in trendy and selective use of the
terminology of projection.

Nevertheless, a broad overview of clinical practice suggests that the awareness of projective
identifi cation has led to a gain in clinical effectiveness at the expense of conceptual rigor,
technical consistency, and professional candor.

3 Judging by citations in texts and reference lists of recent years, Thomas Ogden seems to have
been the most infl uential USA writer on projective identifi cation, though Grotstein remains the
most prolifi c in this regard. Kernberg has also made notable use of the concept. In his early
publications Ogden stayed close to Klein and Bion. Like Bion, he emphasized projective identifi
cation as both defence and communication, in this way opening himself to somehow having to
include in one approach both intrapsychic and interpersonal orientations. Grotstein’s recent
projective transidentifi cation reaches toward the same goal. However, Ogden’s interest in
defensive uses of projective identifi cation has declined as he has shifted his focus to
intersubjectivity – what he now discusses under

‘the third.’ Ogden designates projective identifi cation as one aspect of the intersubjective third.
The third is a realm of subjective experience and discourse that comes into being uniquely
between each analyst-analysand pair; in addition to the two individual subjectivities, the third is
a source of words, feelings, desires, mental states, and fantasies, and it must be taken into
account as a or the crucial sign of a genuine analytic process. I do not think Ogden has so much
integrated the intrapsychic and interpersonal orientations – they may be irreconcilable – as

linked them verbally and applied them eclectically in his clinical work.

His prominence in this area suggests that many students and readers fi nd his approach quite
helpful. However, it can be argued that in this way he is reinforcing widespread use of eclectic
tendencies while perpetuating theoretical disarray.
4 Intersubjectivity is being used, especially by those in the relational school, as a suffi cient basis
for disestablishing the analyst’s expertise in interpreting the interplay of transference and
countertransferences. Now, it is to be as though two transferences are interacting on ‘a level
playing fi eld’ (in the words of Owen Renik). On this understanding, it remains for the two
participants to ‘negotiate’

the nature of the confl icts or desires being expressed in the clinical interaction, if necessary with
some personal disclosure provided by the analyst.

241

Roy Schafer

In the published illustrations of this approach, I do not fi nd much, if any, convincing evidence of
any traditional analytic sort that this leveling move has had the desired egalitarian result. The
evidence provided usually cites behavioral change assessed at face value while setting aside
patient exploration of unconscious fantasies about would-be egalitarianism. The end result is a
mix of interpersonal and relational presuppositions and practices and modifi ed, often ‘wild,’
premature or superfi cial attempts at interpretation. The context is defi ned in terms of
enactments within simultaneous interpersonal and intersubjective realities.

5 The interest in projective identifi cation as communication has led some analysts to make
prominent use of Bion’s idea of containment. Together, these two concepts are being employed
in several new lines of study in an old research area: infant and child development. Outstanding
in this development are neuropsychological imaging studies and studies of attunement and
attachment. Some of these studies use rapid photographic sequences of mother-child interactions.
However, projective identifi cation is not often featured in these contexts. It is, of course, linked
to Kleinian propositions based on Freud’s speculatively introduced ideas about Life and Death
Instincts and to the assumed infl uence of unconscious phantasies so early in development as to
seem inborn. Neither of these propositions fi

t readily into current neuro-cognitive-developmental

contexts that deny the possibility of self-object differentiation suffi -

cient to accommodate Klein’s formulations. Consequently, the concept tends to be condemned,

avoided or disavowed on scientifi c grounds. I believe contemporary review, clarifi cation and
perhaps reformulation of Klein’s propositions and their origins and place in clinical practice
might protect projective identifi cation from being explicitly barred from general psychoanalytic
discourse.

The developments just summarized entail a major shift of emphasis in thinking about projective
identifi cation: the prime assumption in this regard being that human relatedness itself could not
take place in its absence. Thus, Ogden emphasizes that the projective identifi cation, containment
and reverie in the third represent a ‘self-curative’

‘reaching out.’ I take him to be positing a primary need to be understood and contained by
others, to be in their thoughts and their modes of response; also, in the case of psychic damage, a
need for others to ‘be these’ in a new, safer and better way. To my mind, Ogden is developing
another way of 242

Projective identifi cation in the USA: an overview thinking about Loewald’s ‘new object’ and
‘ego core held in trust’

as well as contemporary Kleinian formulations concerning latent phantasies of, and hopes for,
the good object relationships that might emerge from entering the depressive position. All of
which brings us back to Freud’s propositions concerning expressions of a Life Instinct aimed
toward unity and survival.

The new focus on adaptation can also be found in clinical psychological research, for instance in
the work on

adaptive projective identifi cation by Blatt and associates.

6 Three major sources of diverse usage and confusion regarding projective identifi cation remain
to be mentioned.

(a) In the USA, the reference to identifi cation is often understood to mean a two-step process,
the fi rst being projection and the second, identifi cation with the changed object. A model for
this view can be found in Freud’s theory of superego formation: the boy identifi es with the
projectively exaggerated, threatening and vengeful father.

As I understand Klein’s concept, it involves only one step: unconscious identifi cation is implied
whenever parts of the self are intruded into the object.

(b) There exists a common disinclination to think of projection into the object. In effect, this
disinclination implies balking at the interpretation of unconscious, concrete phantasies of
substantial, bodily substantial parts of the self and the object. I believe the preference for
projection on to the object representation is also based on a perhaps unrecognized vestige of
Freud’s theories of distributions of cathexes on the model of electrical charges.

(c) Some clinicians are troubled by the relatively large number of features and uses of projective
identifi cation: defensive; protective of good parts of the self; a way to control others or to merge
with them; an aspect of idealization; a prerequisite of human communication; crucial in
empathizing; and so on.

The concept’s defi nitions have varied accordingly, and led to the objection that projective
identifi cation covers too much territory. It loses meaning by meaning too much. This objection
is, I would say, based on analysts’ impatience with their having to practice contextual thinking
and engage in close listening to phenomenological accounts of subjective experience. In
principle, there is no reason why a series of aims cannot be 243

Roy Schafer

served by one mechanism or process. Meanings vary; today’s analysts are beyond routine
interpretations; analytic work is more challenging and interesting than ever.
There is one fi nal thought I consider it necessary to introduce before concluding this sketch of
recent developments surrounding projective identifi cation in the USA: Whatever the benefi ts of
the increasing focus on communication, intersubjectivity, and interpersonal relatedness and
adaptation, they are offset by the shift of emphasis away from the intrapsychic, endogenous,
destructive, and from confl ict in the internal world.

These emphases have, however, established the value of conceptualizing projective identifi
cation in psychic functioning. Future efforts to strike a balance in this regard will be a source of
unending controversy. The history of ideas shows this to be so.

244

15

A brief review of projective identification in

American psychoanalytic literature

Elizabeth Spillius

In this chapter I attempt to illustrate some of the general themes

described by Roy Schafer (in Chapter 14 ) in his helpful overview

of the work of American psychoanalysts on the concept of projective identifi cation. The chapter
is based partly on general reading and partly on information provided by Psychoanalytic
Electronic

Publishing (‘PEP’ CD-Rom).

43 My brief survey suggests that although the interest of American psychoanalysts in the idea of
projective identifi cation began slowly, by the 1970s it had begun to quicken and by the 1990s
nearly four times as many American as British authors were writing about the topic.

This surge of interest in projective identifi cation coincided with increasing American interest in
British object-relations theory, especially in the work of Winnicott, Fairbairn, Klein and Bion,
and with the growth in American analysis of various sorts of relational, interpersonal and
intersubjective approaches to the relationship between analyst and patient. At fi rst, interest in
projective identifi cation was focused on the term itself – its meaning and possible usefulness or
lack of usefulness. As time has gone on – it is now more than forty-fi ve years since Malin and
Grotstein wrote what I believe was the fi rst American paper on projective identifi cation – it
seems to me

43 I am grateful to Professor David Tuckett for assistance in using the PEP CD-Rom in locating
relevant papers concerning projective identifi cation in 2003 and again in 2008.

245
Elizabeth Spillius

that the analysts and therapists who have been using the idea have been taking it more for
granted. They tend to mention the term in the course of discussing other clinical or conceptual
problems; sometimes they do not mention the term explicitly, although they seem to be using the
idea. Perhaps this is a sign that the idea of projective identifi cation has gradually come to be
more accepted.

The European contributors to the present book have suggested that when a new psychoanalytic
concept is adopted into a different psychoanalytic tradition it is likely not to fi t very well,
leading sometimes to changes in the concept, sometimes to changes in the recipient tradition.
The process of evaluation of a new and unfamiliar concept by members of another
psychoanalytic tradition tends to be accompanied by much attention to defi nition, especially at fi
rst. This has certainly been the case in the United States, where there has been a great deal of
discussion about what projective identifi cation really is and how it should be defi ned, which has
only recently begun to decline.

Most though by no means all of the American authors who write about the concept of projective
identifi cation describe the differences between Klein’s and Bion’s usages, usually by saying that
Klein’s usage is intrapsychic whereas Bion’s is interpersonal, or intrapsychic

and interpersonal (Bion, 1959; Klein, 1946, 1955). Most authors do not mention Bion’s
distinction between normal and pathological projective identifi cation or his discussion of the
processes that tend to lead to pathological projective identifi cation (Bion, 1959). Only the
American analyst Judith Mitrani, and to some extent Grotstein and Ogden, stress Bion’s
emphasis on projective identifi cation as a normal mode of communication between infant and
mother, with the accompanying assumption that it is a normal mode of communication between
adults as well (Mitrani, 2001).

Very few American authors mention the work of Rosenfeld on projective identifi cation
(Rosenfeld, 1971, 1987b) and the few who include Rosenfeld are usually those who have worked
with very disturbed patients. But Rosenfeld’s most important paper on projective identifi cation
(Rosenfeld, 1971), reprinted in Chapter 5 in the

present book, is not cited in the American literature, perhaps because it was fi rst published in an
obscure book. This

paper by Rosenfeld includes a comprehensive statement about the various motives for projective
identifi cation (see

Chapter 5 ) a topic which, surprisingly,

is not usually systematically described or discussed in the American literature.

246

Review of PI in American psychoanalytic literature A feature that has been much discussed in
the American literature is whether or not there is a difference between projection and projective
identifi cation. Grotstein is

apparently the only American analyst who explicitly follows the current but usually tacit British
usage in stating that it is not useful to make such a distinction (Malin and Grotstein, 1966).
Virtually all other American authors say or imply that in ‘projection’ the projector loses contact
with what he has projected into the other person, whereas in

‘projective identifi cation’

the link is (unconsciously) maintained. Some American authors defi ne projection as

‘intrapersonal’ and projective identifi cation as

‘interpersonal’ (e.g. Gilhooley, 1998). Often it is also pointed out that in the case of projective
identifi cation the projector uses some sort of evocative behaviour designed to get the recipient
of the projection to experience feelings appropriate to the content of the projection. This
distinction is very similar to, perhaps the same as the distinction I have made between
‘evocative’ and ‘non-evocative’

projective identifi cation (Spillius, 1988b, pp. 81–86). The difference, however, is that I describe
both the evocative and non-evocative varieties as projective identifi cation, whereas in American
usage the tendency is to describe the non-evocative type as ‘projection’ and the evocative type as
‘projective identifi cation’ (see, for example, Garfi nkle, 2005). In any case I think that these two
types are much more diffi cult to distinguish in practice than in theory.

Another difference between American and British usage is that most British Kleinian analysts
tend to use or at least to be aware of the distinction that Ronald Britton has made between what
he calls ‘attributive’ and ‘acquisitive’

projective identifi cation (Britton, 1998a), whereas these terms and the distinction they describe
are seldom used in the American literature, although Stanley Rosenman’s idea of ‘assaultive’
projective identifi cation describes a very much intensifi ed form of what Britton describes as
‘acquisitive’ projective identifi cation (Rosenman, 2003). Both Britton’s and Rosenman’s terms
are similar to Donald Meltzer’s idea of ‘intrusive’ projective identifi cation (Meltzer, 1967, p.
xi).

Another theme in connection with projective identifi cation is its link to the concepts of
transference and countertransference. The analyst’s response to the patient’s projective identifi
cation is a central preoccupation for many American authors; indeed, James Grotstein says:
‘American interest in projective identifi cation was largely due, in my considered opinion and
experience, to the rapid growth in 247

Elizabeth Spillius

interest in countertransference phenomena, and this was due in turn to the increased interest in
the treatment of borderlines’ ( James Grotstein, personal communication, 2005). Most American
authors who write about countertransference as a response to projective identifi cation now
appear to be using the term ‘countertransference’
in the widened sense suggested by Paula Heimann (1950) and others.

As Glen Gabbard points out, the idea of projective identifi cation in its interpersonal dimension,
together with the idea of countertransference enactment, has been an important aspect of what he
calls ‘the emerging common ground’ in which countertransference is gradually coming to be
conceived as a joint creation by analyst and patient (Gabbard, 1995).

In discussing the connection between projective identifi cation and countertransference, several
American authors mention the work of León Grinberg (1962), especially his idea of ‘projective
counteridentifi cation’. But, with the exception of Otto Kernberg (1989, p. 80), the authors who
cite Grinberg usually alter his defi nition to mean the way the analyst responds to the patient’s
projective identifi cation, a countertransference response, using that term in its broadest sense.
That is not, however, what Grinberg meant. It is perhaps what he should have meant, but not
what he actually meant. In his view projective identifi cation is not a countertransference
response, for he defi nes countertransference as involving the analyst’s psychopathology. He
thinks that projective counter-identifi cation is the patient’s fault, so to speak. The patient has
projected something into the analyst with such force that any analyst would be compelled to
identify with the projection. I think that this defi nition limits the term’s usefulness, for there is
no way of being sure that all analysts would react to a particular case of projective identifi cation
in the same way, a point also made by J.S. Finell (1986). It is perhaps for this reason that
Grinberg’s term has rarely been used in

Britain and that in American usage Grinberg’s defi nition has usually been altered to mean what
in Britain would be called the analyst’s responses to the patient’s projection.

American views on projective identifi cation have been profuse, complex, and varied. I have
attempted to simplify their description by dividing them into three sets. In the fi rst set are those
whose main concern is to use the concept in clinical work: I call this set the adopters . In the
second set are those who defi ne the term and relate it to other concepts but do not make very
much use of it clinically: I 248

Review of PI in American psychoanalytic literature call this set partial adopters . In the third set
are those who focus on the defi nition of the term, usually in order to disagree with it or dismiss
it: I call this set the defi ners and doubters . There is also, of course, a very large fourth set:
those who have not written about the concept and so are presumably indifferent or negative
about it. I assume that many orthodox ego-psychologists would consider themselves to belong to
this fourth set.

The adopters

The best known and most prolifi c analysts of this group are James Grotstein, Thomas Ogden
and Otto Kernberg, although by now (2010) there are many others.

James Grotstein (Grotstein, 1981, 1982, 1994a, 1994b, 1995, 1997, 2001, 2002, 2005; Malin
and Grotstein, 1966) James Grotstein is one of the fi rst American analysts to write about
projective identifi cation and it has continued to be a cornerstone of his psychoanalytic thinking.
He has written a great many papers in which the concept is central, more than ten, and he has
been concerned to develop a theory of mind and of the analytic relationship in which projective
identifi cation plays an important part.

Grotstein’s fi rst paper on projective identifi cation was written in 1966 with Arthur Malin
(Malin and Grotstein, 1966, reprinted in

Chapter 16 of this book). They give Segal’s (1964) version of Klein’s

defi nition of the concept and go on to say that ‘the external object now receives the projected
parts, and then this alloy – external object plus newly arrived projected part – is reintrojected to
complete the cycle’ (Malin and Grotstein, 1966, p. 26). They do not quite make clear the fact that
in Klein’s view this procedure is an unconscious phantasy, not an interpersonal process. They
cite Robert Knight’s 1940 paper ‘Introjection, projection and identifi cation’ as an anticipation of
Klein’s and their own view (Knight, 1940). They also describe Harold Searles’ (1963) paper as
an exemplifi cation of projective identifi cation, although Searles himself makes it clear in his
paper that he thought of what he was describing as ‘symbiotic relatedness’

rather than projective identifi cation.

Malin and Grotstein refer to Rosenfeld’s and Bion’s papers, which use the idea of projective
identifi cation in analysing psychotic 249

Elizabeth Spillius

patients. They do not quite in my view recognise the extent to which Bion’s approach led to an
idea of projective identifi cation as a basic preverbal process of communication or the extent to
which he emphasises the way in which the recipient is likely to be affected by the projection, in
contrast to Klein’s more exclusive emphasis on projective identifi cation as the individual’s
phantasy.

Malin and Grotstein stress that projective identifi cation is a normal process and that it is ‘the
way in which the human organism grows psychically, nurtured by his environment’ (Malin and
Grotstein, 1966, p. 28). ‘Transference phenomena’, they say, ‘are obviously very closely related
to projective identifi cation’ and they stress that there is no difference between projection and
projective identifi cation. ‘A projection’, they say, ‘of itself seems meaningless unless the
individual can retain some contact with what is projected’ (Malin and Grotstein, 1966, pp. 28 and
27).

Finally, they think that projective identifi cation is a normal process existing from birth, though it
can also become

pathological and defensive.

These attributes of projective identifi cation have continued to be essential in Grotstein’s later
work. In his book Splitting and Projective Identifi cation (Grotstein, 1981) written fi fteen years
after his paper with Malin, Grotstein emphasises the varied motives for projective identifi cation
much in the fashion of Rosenfeld (1971): fusion, control of the object, evacuation and disavowal
of aspects of the self, communication to other aspects of the self. He does not, however,
emphasise the projection of good aspects of the self, which was important in Klein’s thinking.
Once again he asserts that there is no difference between projection and projective identifi cation.
He now emphasises much more than before the concept of splitting, which is an essential aspect
of Klein’s view of projective identifi cation. He emphasises the difference between intrapsychic
and interpersonal projective identifi cation more than before, but he does not discuss the
importance of evocative behaviour. Indeed, he thinks that we do not actually project into external
objects, but into ‘our images of them’. In other words, projective identifi cation is intrapsychic,
although, as Dorpat (1983) points out, Grotstein’s usage is inconsistent for at some points he
regards projective identifi cation as interactional. Dorpat thinks, and I agree, that ‘the
unconscious phantasy is actualised through verbal and non-verbal communications
unconsciously designed to provoke in another person various emotions and attitudes’ (Dorpat,
1983, p. 119).

250

Review of PI in American psychoanalytic literature In his many later papers that use the concept
of projective identifi cation, Grotstein relates it to other concepts of object-relations theory and to
ideas of his own to form a somewhat new and complex conceptual system in which projective
identifi cation is central (Grotstein, 1981, 1982, 1994a, 1994b, 1995, 1997, 2001, 2002). He also
stresses that the concept of projective identifi cation was absolutely central in Klein’s thinking.
In my view this is an exaggeration.

To Klein and to contemporary Kleinians, projective identifi cation is important clinically, but its
importance as a theoretical concept derives from its place in the rest of Klein’s conceptual
system, most especially the ideas of the paranoid-schizoid and depressive positions.

Thomas Ogden (Ogden, 1978a, 1978b, 1979, 1982, 1986, 1994a, 1994b, 1994c, 2004)

As Roy Schafer describes in Chapter 14 , Ogden’s work, including

his work on projective identifi cation, is well known and frequently cited in the literature. In
understanding projective identifi cation and then in developing his own conceptualisation of the
mind and of the analyst–patient relation, Ogden builds on the ideas of Klein and Bion on the one
hand and of Winnicott on the other. He has also been infl uenced by Rosenfeld, Balint, Searles,
Langs, and by his friends James Grotstein and Bryce Boyer, and, inevitably, by the intensity of
his work with severely disturbed patients, the sort of patient with whom an understanding of
projective identifi cation is particularly important. He also spent some time at the Tavistock
Clinic in the 1970s, which I surmise was an important introduction to the thinking of British
object-relations theorists.

His fi rst paper discussing projective identifi cation was ‘A developmental view of identifi cation
resulting from maternal impingements’

(Ogden, 1978a). This paper was followed in 1979 by ‘On projective identifi cation’ (reproduced
in Chapter 17 ), and

then by the book

Projective Identifi cation and Psychoanalytic Technique (Ogden, 1982). In

‘On projective identifi cation’ Ogden (1979) describes projective identifi cation as a process in
which the fi rst step consists of a phantasy in which the patient wishes to rid himself of unwanted
aspects of the self into another person.

The second part of the process consists of the projector unconsciously putting pressure on the
recipient to think and behave in a manner congruent with the projection. The third part of the
process consists of the recipient processing the 251

Elizabeth Spillius

projected feelings so that they can be re-internalised by the projector.

He does not make clear that in Klein’s view the whole process of projective identifi cation takes
place in the projector’s phantasy and that it was Bion who added the interpersonal
communicative element and the containing function described in the third part of Ogden’s view
of the process. One gets the impression that Ogden is not particularly concerned about academic
niceties concerning attribution. His aim is to develop a workable interpretation of the processes
involved.

In his 1979 paper ‘On projective identifi cation’ Ogden, unlike Grotstein, distinguishes
projection from projective identifi cation. He thinks that in projection an aspect of the self is in
phantasy expelled, disavowed and attributed to the recipient whom the projector often
experiences as foreign, strange and frightening. In projective identifi cation the link with what
has been projected and attributed to the recipient is to some extent retained. In practice I have
found this distinction diffi cult to maintain for it has been on the occasions when a patient has
found me foreign, strange, frightening, hateful even, that the patient has had the most intense
although denied relationship with what he has located and evoked in me. The diffi culty, as
always, is to metabolise the experience suffi ciently to be able to use it constructively and to
express one’s understanding of it in attitudes and words that the patient can at least partially
begin to take in.

In his book Projective Identifi cation and Psychotherapeutic Technique Ogden (1982) explains
how he uses the related concepts of externalisation, introjection and introjective identifi cation.
He also explains his own technique in more detail, contrasting it with that of other
psychoanalytic traditions, and he describes his work with hospital in-patients and seriously
disturbed schizophrenic patients.

Ogden’s later work builds on these early formulations and experiences, leading to an integration
of it with the work of Klein, Winnicott, Esther Bick, Donald Meltzer, Frances Tustin and others:

see Chapter 6 of T he Matrix of the Mind (Ogden, 1986) and Chapters

6 and 10 of Subj ects of Analysis (Ogden, 1994c). Projective identifi cation continues to play a
part in his later formulation of the ‘dialectical’

relationship between the ‘autistic-contiguous’, the paranoid-schizoid and the depressive

positions, and also in the dialectical relationship of the two subjectivities of the analyst and the
patient and their intersubjectivity, which Ogden calls the ‘analytic third’ (Ogden, 1994a, 1994c
[especially Chapter 5 ], 2004).

252

Review of PI in American psychoanalytic literature It is clear that projective identifi cation has
been a central organising concept for Ogden, as for Grotstein and others, because of its
interpersonal dimension. He has used it extensively in the development of his conceptual system,
which takes some ideas from Klein herself and from other Kleinians and object-relations
theorists, but he weaves these ideas together into a new and rather different synthesis.

Both Grotstein and Ogden have had considerable personal contact with Kleinian and
Independent analysts, which supports Helmut Hinz’s thesis that such contact encourages
adoption of new concepts (see

Hinz’s Chapter 11 in the present book). But both Ogden and Grotstein have not just adopted
some aspects of the Kleinian and object-relations traditions, they have used parts of these
traditions, including the idea of projective identifi cation, in developing new syntheses.

Otto Kernberg (Kernberg, 1975, p. 56; 1980, pp. 27, 45; 1984,

pp. 16–17, 113–115; 1989, pp. 6, 24; 1992, especially Chapter 10 )

Like Grotstein, Ogden, and others, Otto Kernberg is particularly interested in severely ill and
borderline patients and it is mainly (though not exclusively) in the context of their treatment that
he uses the concept of projective identifi

cation (Kernberg, 1986, 1987).

Kernberg defi nes projective identifi cation as a more primitive defence than projection
(Kernberg, 1986; 1987; 1989, pp. 6, 24). He thinks it consists of three processes: fi rst, projection
of badness into the object (he does not say anything about the projection of goodness); second,
maintaining empathy with what has been projected; and third, inducing the object to experience
what has been projected.

He regards projection as a more mature defence in which unacceptable experience is repressed

and then projected into the object.

Empathy is not maintained with what has been projected (Kernberg, 1986, 1987). This is not a
defi nition that I think other American or British authors have proposed.

In psychosis, Kernberg says, there is a loss of ego boundaries accompanied by regression to an

abnormal symbiotic phase bringing about an obliteration of the self and destruction of the object
world

‘under the infl uence of projective identifi cation’. He does not advocate interpreting projective
identifi cation to psychotic patients because he thinks that such interpretations would exacerbate
the loss of ego boundaries, a fi nding which has not, to my knowledge, been emphasised by other
analysts who use the concept of projective identifi cation in the analysis of psychotic patients.
253

Elizabeth Spillius

He thinks, on the contrary, that it is useful to interpret projective identifi cation to borderline and
narcissistic patients, and that in the case of neurotic patients projective identifi cation is less
important and one would fi nd oneself interpreting projection.

In ‘Projection and projective identifi cation: developmental and clinical aspects’, Kernberg
(1987) illustrates his formulations by three clinical cases: fi rst, an hysterical woman who
Kernberg thought did not use either projective identifi cation or projection; second, a narcissistic
woman who used projective identifi cation; and third, a paranoid borderline man who also used
projective identifi cation.

One gets the impression that although Kernberg is familiar with British Kleinian theory, he
focuses especially on clear defi nition of its concepts more than its own practitioners do, and in a
form that they would probably not use, particularly in the distinction he makes between
projection and projective identifi cation and in his classifi cation of patients for whom
interpretation of projective identifi cation is or is not appropriate. Perhaps some of this focus on
precise defi nition and application comes about because of the importance of Kernberg’s role in
the United States in communicating psychoanalytic ideas to psychiatrists and psychologists as
well as to psychoanalysts.

Although Ogden, Grotstein and Kernberg have made the main contributions in this fi eld, there
have been contributions by many others: Bryce Boyer (1978, 1986, 1989, 1990a, 1990b); Lucy
La Farge (1989); Harold Boris (1993, 1994a, 1994b); Albert Mason (a

current paper is included as Chapter 18 in the present book); Robert Caper (1988, 1999); Dan
Gilhooley (1998); Louis Brunet and Diane Casoni (Brunet and Casoni, 1996, 2001); Judith
Mitrani (2001); Roger Karlsson (2004); Jeffrey Eaton (2005); Stephen Purcell (2006); Henry
Smith (2000; 2006).

Bryce Boyer (1978, 1986, 1989, 1990a, 1990b)

Bryce Boyer, like Ogden and Grotstein, but in his own unique way, integrates the theory of
British object relations with that of ego-psychology and, like Grotstein and Ogden, his chief area
of clinical work has been with psychotic and seriously disturbed borderline patients (Boyer,
1978, 1986, 1989, 1990a, 1990b). He uses the concept of projective identifi cation but he is less
concerned than Grotstein and Ogden with formulating it conceptually and using it to develop a
conceptual system. He also discusses countertransference, especially the contributions of Racker,
Heimann, Rosenfeld and 254

Review of PI in American psychoanalytic literature Balint. But it is as a talented clinician that his
writings are most memorable.

Lucy La Farge (1989)

Lucy La Farge’s contribution to projective identifi cation is briefer than those discussed so far,
but is of the same general type. She uses the idea, along with other concepts, to analyse the way
severely ill borderline patients create empty states to ward off regression to states of
fragmentation or pathological fusion.

Harold Boris (1993, 1994a, 1994b)

Harold Boris is in a special category for he was not an analyst but he became a skilled analytic
clinician at a time when non-medical people were not allowed to be formally trained analytically
in the United States. He is one of the few American practitioners who took in the whole ethos of
Klein’s and Bion’s thinking as if he had known it emotionally in himself all along. His
conceptualisation of it, however, is all his own. Unlike most other American authors he does not
defi ne projective identifi cation or try formally to relate it to other concepts. His concern is to
use it clinically especially in trying to understand envy, hope as a defence, anorexia and bulimia
(Boris, 1993, 1994a, 1994b). His most intensive discussion of projective identifi cation is to be
found towards the end of his paper

‘Torment of the object: a contribution to the study of bulimia’ (Boris, 1988).

In a later paper (Boris, 1994a) he gives an engaging description of projective identifi cation and
envy in a four-year-old girl. Boris is not particularly concerned to make conceptual defi nitions
of projective identifi cation. He explains the concept by example rather than by formal defi
nition, and he uses the idea as a tool in understanding both normal and ‘diffi cult’ patients.

Albert Mason (his current paper is included as Chapter 18 in the

present book)

Albert Mason is an English Kleinian analyst who went to the United States (Los Angeles) in
1957. He has written

several unpublished papers about projective identifi cation, one of which, ‘Vicissitudes of
projective identifi cation’,

is published in Chapter 18 in the present

book for the fi rst time. In it Mason describes Klein’s use of the term and he gives many
examples of his own use of it in clinical practice.

255

Elizabeth Spillius

Robert Caper (1988, 1999)

Caper has not written specifi c papers on projective identifi cation although he discusses the
concept in his two books, Immaterial Facts (1988) and A Mind of One’s Own (1999). He is
familiar with the work of Klein, Bion, Mason, Grotstein and contemporary Kleinian authors
generally. Caper’s discussions of projective identifi cation are distinguished by particularly
detailed and evocative clinical examples. He is especially interested in the role of projective
identifi cation in developing relations with internal objects. ‘The most prominent of the internal
objects with which the child must learn to live’, he says, ‘is the superego, an unconscious
melding of his or her external objects and impulses toward them brought about by projective
identifi cation’ (Caper, 1988, p. 236). He also describes the role of projective identifi cation in
developing a differentiation between identifi cation in the paranoid-schizoid position
(‘narcissistic identifi cation’) and in the depressive position (‘depressive identifi cation’) (Caper,
1999, pp. 5–6, 96–104).

Dan Gilhooley (1998)

Dan Gilhooley describes several episodes in the life of a three-year-old boy which he then
analyses in terms of his views on projection, which he defi nes as intrapsychic, and projective
identifi cation, which he defi nes as interpersonal.

Louis Brunet and Diane Casoni (Brunet and Casoni, 1996, 2001) Brunet and Casoni describe
the symbolisation, projective identifi cation and use of the analyst by a schizophrenic adolescent
boy. In this analytic experience of projective identifi cation the analyst was able to act as a
container for the patient’s actions and thoughts in such a way that he could symbolise the
patient’s archaic and overwhelming thoughts and feelings in a way that the patient could take in.

Judith Mitrani (2001)

Judith Mitrani describes several episodes of clinical material illustrating the way the analyst
needs to be fully open to feeling the patient’s emotional experience. Only such meaningful
containment of emotionally charged projective identifi cations can hope to develop emotional
understanding and a potential for psychic change. Intellectual understanding is not enough, and
projective identifi cation is an important aspect of emotional communication.

256

Review of PI in American psychoanalytic literature Roger Karlsson (2004) Roger Karlsson

thinks that in some cases the analyst needs to be able to experience prolonged tolerance of the
patient’s projecting intolerable aspects of himself into the analyst before the patient is able to
progress from an experience of ‘oneness’ to an experience of

‘twoness’. Experience of ‘twoness’, of separateness, was feared by the particular patient he

describes because it involved recognition of the analyst’s and the mother’s imperfections, which
threatened the patient’s defence of idealisation.

Stephen Purcell (2006) and Henry Smith (2000, 2006) Both Purcell and Smith describe in
clinical papers how they used understanding of their patients’ particular projective identifi
cations.

In all three papers it was evident that the authors were taking it for granted that the reader would
understand that they were using the concept of projective identifi cation implicitly even when
they were not formally defi ning it or, in Smith’s case, hardly using it by name.
Projective identifi cation was not the theme of these papers – the focus of the papers was the
particular clinical problem of each patient-analyst pair.

The partial adopters

The analysts of this group have made use of the concept of projective identifi cation, but usually
in comparative isolation from the other concepts and the general ethos of the Kleinian tradition,
and their clinical use of the concept is less intensive than that of the fi rst set of analysts
described above.

Robert Langs (1978a, 1978b, 1978c, 1979)

In the Preface to his book Technique in Transition (Langs, 1978c) Langs says that he
encountered the idea of projective identifi cation in the 1960s when ‘looking for a language that
could help me to conceptualize my observations in the interactional sphere’. He thought that the
Kleinian literature was interactional although not in quite the way he had in mind. It was not
until he read a paper by Madeline and Willi Baranger (1966) with their concept of the bi-
personal fi eld and the role of projective identifi cation in it that Langs fully realised the immense
signifi cance that the two concepts 257

Elizabeth Spillius

could have in his work. For Langs his developing interest in this theme was the essence of his
shift from a ‘classical’

to a ‘neo-classical’

approach.

Like many American analysts, Langs describes projection as an intrapersonal concept and
projective identifi cation as interactional.

Projection he defi nes as

an intrapsychic mechanism through which the patient unconsciously attributes his own inner
contents and impulses to another individual without actual interactional efforts to place those
contents into the other person and to have the other person experience and deal with them.
Projective identifi cation is specifi cally interactional in that it is unconsciously designed to
create an intrapsychic effect within the recipient.

(Langs, 1978c, p. 317)

Langs describes his understanding and use of projective identifi cation by describing how,
through observations of patients and analysts, combined with much reading of Bion, Khan and
Winnicott, he delineated three sub-types of bi-personal fi eld. Langs illustrates the three sub-
types of bi-personal fi eld in the work of his supervisees, and in the description of the three fi
elds he fi nds the concept of projective identifi cation useful, though I do not think his main
purpose is to develop or expand upon the concept of projective identifi cation in itself.
Walter Burke and Michael Tansey (Burke and Tansey, 1985; Tansey and Burke, 1985, 1989)
Burke and Tansey explore the relationship of projective identifi cation to empathy and
countertransference, all three concepts being important for them in linking intrapsychic and
interpersonal experiences.

They maintain, rightly I think, that Heinrich Racker is mistaken in thinking that empathy is likely
to accompany what Racker calls

‘concordant’ countertransference and that lack of empathy and projective identifi cation are
likely to accompany what he calls ‘complementary’ countertransference (see Racker, 1957).
Burke and Tansey think that projective identifi cation and complementary countertransference
may lead to empathy; concordant countertransference may lead to discord. Projective identifi
cation, the authors say, can be defensive, adaptive or communicative – and perhaps all three at
the same time, although they do not go quite so far as to say that.

258

Review of PI in American psychoanalytic literature Burke and Tansey seem to make a tacit
assumption that where empathy is, understanding will follow. Judging from their case material, I
think it is often the other way around.

Once one understands what is going on between analyst and patient, one is more likely to feel
empathy.

Tansey and Burke have developed a very complex set of categories for analysing progress or
lack of progress in analytic empathy and understanding. The fi rst phase of such understanding
they categorise as ‘reception’, the second as ‘internal processing’ and the third as

‘communication’. Within each of these phases, at least in their fi rst exposition of them, there are
three sub-phases.

At fi rst I thought this scheme unnecessarily elaborate and ponderous – how could one use it in a
session?

Apparently I am not alone, for in a review of Tansey and Burke’s (1989) book Understanding
Countertransference: From

Projective Identifi cation to Empathy , James Frosch says, ‘The clinical examples lack the
richness, depth and individuality that we are accus-tomed to in analytic case reports’ (Frosch,
1990). But on reading their papers and the book again, I thought Frosch’s judgement was a little
harsh. Although their system is too cumbersome to be used in the heat of a session, it does give a
means for examining what has gone wrong after it is over, and once one knows what has gone
wrong, one is half way towards putting it right.

Jill Savege Scharff (1992)

In her book Projective and Introjective Identifi cation and the Use of the Therapist’s Self
(Scharff, 1992), Jill Savege Scharff reviews a great many defi nitions and uses of the concepts of
projection, projective identifi cation, introjection and introjective identifi cation in both Britain
and the United States, and illustrates her own usages extensively in work with individuals,
couples, and families. She considers too the usefulness of these ideas in understanding art and
culture. In spite of its comprehensiveness, her work is somewhat piecemeal; she does not attempt
to choose among the many defi nitions or to develop an integrated approach in which the various
defi nitions of identifi cation would be related to one another, and in her clinical applications she
gives the impression of using whatever particular defi nition or aspect of a defi nition seemed to
make sense to her on that occasion.

Morris N. Eagle (2000)

Eagle disapproves of facile assertions by analysts, perhaps especially facile assertions about
projective identifi cation in transference and 259

Elizabeth Spillius

countertransference as a response to it. His view is that in the process of correcting the idea of
the analyst as a ‘blank screen’, the misleading idea has developed that everything that the analyst
thinks and feels about his patient, all his countertransference reactions, can be seen as a virtually
unerring guide to the patient’s mental contents. He is equally critical of Racker’s belief that when
the analyst feels resentful about a patient’s critical attacks on him this means that the analyst is in
a state of ‘complementary identifi cation’ with one of the patient’s internal objects (Racker,
1957). Nor does Eagle agree with those analysts who explain their thoughts about their patients
by asserting that the thoughts have been ‘put into’ them by the patient. He expects more rigorous
attention and thinking from the analyst:

‘explicit and specifi c knowledge of the other, theoretical knowledge, explicitly searching for
patterns and cues, examination of evidence, and clinical reasoning and inference based on the
patient’s productions’. His paper is reminiscent of Joseph Sandler’s paper ‘On communication
from patient to analyst: not everything is projective identifi cation’ (Sandler, 1993).

Ely Garfi nkle (2005)

Garfi nkle gives a careful review of American and British uses of the concept of projective
identifi cation and concludes that it has been confusing to extend the defi nition of the concept to
include interpersonal as well as intrapersonal components. He rejects Ogden’s defi nition
because it includes interpersonal components, and he rejects the defi nitions of the British
analysts Priscilla Roth (1999, pp. 4–5) and Ronald Britton (1998a, pp. 4–5) because they require
familiarity with Kleinian concepts. He suggests that projective identifi cation should be defi ned
as follows:

Projective identifi cation is to be defi ned strictly as an unconscious phantasy in which split off
parts of the self are disowned, projected, and attributed to someone else. In addition, a
conclusion by the analyst based on clinical evidence, that an unconscious motive in the
unconscious phantasy of the analysand is to control and/or to infl uence the thinking, feeling
and/or action of the object, would be a suffi cient (though not necessarily exclusive) criterion to
defi ne such an unconscious phantasy as a projective identifi cation.
(Garfi nkle, 2005, pp. 202–203)

260

Review of PI in American psychoanalytic literature He also defi nes the term ‘projection’, as
follows: In order to distinguish projective identifi cation from simple projection, I would also
propose that we defi ne simple projection in the following manner: that the term projection refl
ect a situation in which the subject perceives the object in accordance with the subject’s internal
reality without an unconscious intent to affect the mind of the object.

(Garfi nkle, 2005, pp. 203–204)

I do not know whether other analysts have adopted Garfi nkle’s defi -

nitions, but I suspect that the inclusion of interpersonal components in American usage,
especially by Ogden, has become too well accepted for it to be abandoned. Further, deductions
about a patient’s thoughts and feelings are necessarily made through the perceptions of the
analyst, fallible though they may be, so that some sort of interpersonal component is inevitably
part of the perceiving and defi ning process.

The researchers

Within the set I have called ‘partial adopters’ is a sub-set who have used the concept of
projective identifi cation in psychological and psychosocial research in ways that required
conceptual understanding and application of the concept, although without very extensive
clinical use of it. There have been three such projects, reported by John Zinner and Roger
Shapiro (Zinner and Shapiro, 1972), Roy Muir (1982, 1990) and Paolo Migone (1995).

In summary, the analysts of the set I have called ‘partial adopters’

tend to defi ne and to some extent to use the concept of projective identifi cation, but they stress
its formal aspects rather than its clinical usefulness.

The defi ners and doubters

Almost all the large group of authors of this group have written just one paper on projective
identifi cation. Most complain in one way or another about a lack of clarity in the concept which
many try to remedy by giving their own defi nition. Some authors are neutral in their views
(Carveth, 1992; Crisp, 1986; Goldman, 1988; Hamilton, 261

Elizabeth Spillius

1986, 1990; Issacharoff and Hunt, 1994; Jaffe, 1968). Two authors (Adler, 1989; Feinsilver,
1983) have tried to translate projective identifi cation into a Winnicottian framework. The
remaining authors variously fi nd the concept

‘mysterious’, ‘confusing’, ‘jargon’, ‘impre-cise’, ‘obfuscating’, ‘psychobabble’, or part of

‘Klein’s fantastic metapsychology’ (Blechner, 1994; Finell, 1986; Meissner, 1980, 1987;
Ornston, 1978; Pantone, 1994; Porder, 1987; Whipple, 1986). Not surprisingly, these authors do
not use the concept of projective identifi cation clinically, and, with the exception of an
occasional expression of interest in the work of Winnicott, they do not express any interest in
other aspects of the Kleinian or object-relations traditions.

Considering their generally negative attitude, it is surprising that the members of this group have
written even one paper on the topic of projective identifi cation. It seems possible that they have
done so because they thought at fi rst that the concept might be relevant to an interpersonal view
of the analyst–patient relationship and other relationships more generally. The fact that other
well-known analysts were writing about projective identifi cation may also have stimulated their
interest.

Conclusion

This examination of American responses to the concept of projective identifi cation provides an
interesting and perhaps unusual example of the spread of a concept from one analytic school of
thought to another. I have described three sorts of response: adoption, partial adoption, defi ning
but doubting. There is of course the fourth response, that of simply ignoring the ‘new’ concept
entirely. The fact that so much notice has been taken of the concept, even though the noticing is
negative as well as positive, suggests that there is something about the idea that intrigues
analysts. I think the ‘something’ is its potential for helping to understand the relationship of
analyst and patient from a slightly new and unfamiliar angle.

There has also been a gradual change over time in the types of paper written by the ‘adopters’.
There is less attention to defi nitions of projective identifi cation, and more tacit use of the idea in
analytic work on particular clinical problems; sometimes the actual term is used in such papers,
sometimes it is hardly mentioned. Perhaps, in addition to projective identifi cation coming to be
taken more for 262

Review of PI in American psychoanalytic literature granted, there is some reluctance to use the
term by name in case readers might assume that the writer was too ‘Kleinian’, for Kleinian ideas
have for many years been regarded in the United States as quite ridiculous, and have only
recently acquired a certain respectability in some quarters.

263

16

Projective identification in the

therapeutic process 44

Arthur Malin and James S. Grotstein

Recent articles by Loewald (1960) and Searles (1963) having to do with certain aspects of the
therapeutic process have stimulated us to investigate what we believe may be the basis of the
therapeutic effect in psycho-analysis. In our view the concept of projective identifi cation can be
fruitfully applied to an understanding of the therapeutic process. We shall attempt to describe the
concept of projective identifi cation and then discuss the relevance of this idea to normal and
pathological development with a view toward clarifying the therapeutic process in light of it.

The term projective identifi cation was fi rst used by Melanie Klein (1946) and was meant to
indicate a process in which parts of the self are split off and projected into an external object or
part object.

Hanna Segal (1964) states:

Projective identifi cation is the result of the projection of parts of the self into an object. It may
result in the object’s being perceived as having acquired the characteristics of the projected part
of the self, but it can also result in the self becoming identifi ed with the object of its projection.

44 This chapter reproduces the text of Malin, A. and Grotstein, J. S. (1966). Projective identifi
cation in the therapeutic process.

International Journal of Psychoanalysis ,

47 ,

26–31.

264

Projective identifi cation in the therapeutic process This idea was developed from Klein’s
(1932b; 1935) earlier concept of object relations existing from the start of extrauterine life. Klein
had indicated that the relation to the fi rst object, the breast, is through introjection. She also
demonstrated that object relations from the beginning depend for their development on projective
and introjective mechanisms. Klein (1946) suggested that these mechanisms are seen in the
earliest period of normal development, which she described as the paranoid-schizoid position.
She stated further that these mechanisms are also a type of defence found particularly in
schizophrenic patients.

We wish to emphasize at this point that projective identifi cation to us has come to mean many
different things and embraces many concepts. Our paper is an attempt both to clarify and to
expand on it, and to place it in its proper perspective in psycho-analytic theory and practice.

First, we should like to say why we use the term projective identifi cation and not projection.
Projection alone is a mechanism for dealing with instinctual drives, akin to incorporation. It is an
instinctual mode. We feel, as does Fairbairn (1952), that all intra-physic and inter-personal
relations are transacted on the basis of object relationships, rather than on the basis of instinctual
drives alone. The object is the irreducible vehicle in human interaction.

Once we make this assumption, we then conceive of the psychic apparatus as a dynamic
structure composed of internalized objects (and part-objects) with drive charges inseparably
attached to them.

We feel that these charged parts of self (or identifi cations) are projected outward and that the
status of the identifi cation changes by virtue of the projection, thus enabling the ego to
discharge, for instance, unwanted or disclaimed parts of the self (purifi ed pleasure ego of Freud,
1915a). The external object now receives the projected parts, and

then this alloy – external object plus newly arrived projected part – is re-introjected to complete
the cycle.

In the preceding paragraph, we have dealt with the defensive nature of projective identifi cation.
We wish to emphasize that it is also, at the same time, a way of relating to objects. As Freud
(1921) has stated, the infant relates by identifi cation prior to making anaclitic object choices. We
agree with this and go two steps further; fi rst, we believe that all identifi cation includes
projection, as we hope to show; and second, that projective identifi cation is also a normal, as
well as abnormal, way of relating which persists into mature adulthood.

265

Arthur Malin and James S. Grotstein We hope to develop the reasons why these burdensome
emenda-tions of theory are necessary, especially since the advent of object-relations theory has
imposed this task upon us.

An article by Knight (1940) appears to anticipate the concept of projective identifi cation
although it is not described directly by that name. In this short article Knight attempts to describe
the different ways in which identifi cation may be used and defi ned. Knight states,

‘Identifi cation is never an irreducible process or state of affairs, but is always based on a subtle
interaction of both introjective and projective mechanisms.’ Knight makes a point that Bibring’s
term, ‘altruistic surrender’, involves a projection of one’s own desires for pleasure and gratifi
cation into another person with whom one then identifi es.

Knight goes further and states,

The awareness of how we would feel under similar circumstances enables us to project our own
needs and wishes on to the object and then to experience his feelings as if they were ours through
the resultant temporary identifi cation with him. Even though this vicarious experience would
appear to be an instantaneous process, it seems to me valuable to reduce it to its constituent
mechanisms of projection and possibly also introjection.

It is obvious that Knight is referring to identifi cation with whole objects rather than part objects
as emphasized by Klein, but Knight’s ideas are certainly compatible with the concept of
projective identifi cation.

In line with Knight’s thinking, we want to emphasize what seems obvious in the concept of
identifi cation, namely, that all identifi cation includes projection, and all projection includes
identifi cation. 45

Before we are ready to internalize (take in psychically, incorporate), we must be in some state of
readiness for this process. That is, we must tentatively project out a part of our inner psychic
contents in order to be receptive to the object for introjection and subsequently
45 We defi ne introjection as a psychic phenomenon in which the object is taken into the psychic
apparatus but is kept separate from the self; in other words, it is within the ego but unassimilated,
much like a foreign body.

Following introjection, identifi cation may take place by the object’s becoming assimilated into
the ego or self. See Greenson (1954).

266

Projective identifi cation in the therapeutic process to form an identifi cation with it. When we
start with the projection it is necessary that there be some process of identifi cation or
internalization in general, or else we can never be aware of the projection.

That is, what is projected would be lost like a satellite rocketed out of the gravitational pull of the
earth. Eventually all contact with the satellite will be lost. Although the satellite has left Earth, it
must remain under the infl uence of Earth’s gravitational pull to remain in orbit in order for it to
maintain some contact with Earth. A projection, of itself, seems meaningless unless the
individual can retain some contact with what is projected. That contact is a type of
internalization, or, loosely, an identifi cation. We want to show that Klein’s concept of projective
identifi cation can be broadened greatly in order to understand many phenomena in psychic life
both normal and pathological, and to enhance our knowledge of identifi cation itself.

Rosenfeld (1952a, 1952b, 1954) and Bion (1955b, 1956) have applied the concept of projective
identifi cation to the understanding and treatment of the psychotic patient. They state that when a
patient splits off a part of himself and projects it into the object, such as the analyst, he has a
feeling of relatedness to the analyst but with some corresponding feelings of inner
impoverishment. Very often the patient feels that the split-off part, now in the external object, is
a persecutor. They emphasize the importance of projective identifi cation in understanding
delusional transference material.

Searles (1963) describes very similar phenomena. He relates much of his material to the Kleinian
concept of projective identifi cation, but he does emphasize some important differences between
his ideas and Klein’s. In a more broadly defi ned manner, however, we would view Searles’s
ideas on transference psychosis as being another aspect of projective identifi cation. Searles
makes an important point, for instance, of the schizophrenic patient’s need to project a part of
himself into the therapist. The therapist must provide, according to Searles, a suitable and
receptive object in himself to receive this projection from the patient. Searles suggests,
Moreover, it is my experience that he [the chronic schizophrenic patient] actively needs a degree
of symbiotic relatedness in the transference, which would be interfered with were the analyst to
try, recurrently, to establish with him the validity of verbalized transference interpretations.

267

Arthur Malin and James S. Grotstein Searles suggests here that the projective identifi cation
from the patient to the analyst must fi rst be accepted by the analyst before verbal interpretations
will be of any help.

Loewald (1960) writes of therapeutic change as involving structural change in the ego. In
speaking of the patient’s reaction to the analyst, Loewald states, ‘A higher stage of organization,
of both himself and his environment, is thus reached, by way of the organizing understanding
which the analyst provides.’

Loewald emphasizes throughout his article the importance of higher levels of ego integration
which the patient can achieve through the analytic treatment. We suggest that projective identifi
cation helps explain the development of these higher levels of ego integration.

Transference phenomena are obviously very closely related to projective identifi cation.
Transference implies the projection of inner attitudes which came from earlier object
relationships into the fi gure of the analyst during the analysis. A much broader concept of
transference would state that all subsequent relationships are modifi ed on the basis of the earliest
object relationship of the individual which is now established in the inner psychic life. This view
very closely approximates the concept of primary objects which was advanced by Balint (1937).
If we accept a broad view of transference to include all object relations, internal and external,
after the primary relationship with the breast-mother which is now internalized, then we are
stating that all object relations and all transference phenomena are examples, at least in part, of
projective identifi cation. This implies that there must be a projection from within the psychic
apparatus into the external object. We emphasize that this includes parts of self as well as
internal object representations. To go back to Klein’s ideas for a moment, some of her lack of
emphasis on the environment in human development can be understood in terms of projective
identifi cation.

It can be understood in the sense that the early instinctual representations, including the death
instinct, are projected into the breast-mother, and then the bad breast-mother is introjected on the
basis of the earlier projection and not so much on the basis of the actual environmental situation
of that breast-mother. We should like to modify this idea, however, with the suggestion that it is
just the fact that the inner psychic contents related to earliest object relations are projected into
the external objects that makes for the tremendous infl uence of the environment. It seems to us
that it is only upon perceiving how the external object receives our projection and deals 268

Projective identifi cation in the therapeutic process with our projection that we now introject
back into the psychic apparatus the original projection, but now modifi ed and on a newer level.
Hopefully, the mother has helped the infant by allowing this projection to be met with a response
of understanding, care, and love. It is the mother who cannot do this, and who sees the child’s
projections as destructive and frightening, who will confi rm the infant’s fears of his own bad
destructive self. 46 We suggest, moreover, that this method of projecting one’s inner psychic
contents into external objects and then perceiving the response of these external objects and
introjecting this response on a new level of integration is the way in which the human organism
grows psychically, nurtured by his

environment. The environment must meet the needs of these projections and be able to
reinterpret for the developing individual the inner workings of their psychic apparatus and to
demonstrate that these are not destructive, ‘bad’

parts. The external object must confi rm those constructive and ‘good’ aspects of the developing
individual and thus facilitate higher ego integration which will mitigate the effect of the
destructive components of the self.

We propose that these concepts are of crucial importance in understanding the earliest
experiences of the infant, the further growth and development of children and adults, and to a
great extent the therapeutic effect of psychoanalysis.

We have all observed how patients must project into the analyst their inner psychic contents.

These consist of objects and part objects with associated feelings and attitudes. It is mainly
through his perception of the manner in which the analyst handles these projections that the
patient can fi nd a new level of integration. As Searles (1963) emphasizes, what is important is a
receptiveness without an encouragement of these projections, and an attempt at understanding
their meaning without the fear that these projections will destroy the analyst.

The essence of the therapeutic process is through modifi cation of internal object relationships
within the ego, and this is largely brought about by projective identifi cation. Correct
interpretations can be seen as an important way in which the patient can observe how his
projections have been received and acknowledged by the analyst. If this does not take place the
patient is left with futility, despair, and doubt in regard to his inner self worth.

46 Erikson (1959) has shown that the mother also projects her needs and feelings into the infant
and responds to the child’s perception of these needs.

269

Arthur Malin and James S. Grotstein One of the most common defences of the schizophrenic
borderline patient, as well as of many neurotics, is the need to preserve the analyst as a good
object by maintaining a distance which paradoxically is not very helpful to developing
understanding. Much of this is related to what seems to be a negative therapeutic reaction. It
would appear that these patients are trying to preserve the analyst by avoiding closeness to him,
i.e. not projecting any of their bad parts into the analyst which they feel will destroy the analyst
and therefore their only hope for survival. For example, a borderline patient could rarely speak of
any positive feelings toward the analyst, but would occasionally, with great disappointment,
point out what he felt was an error on the part of the analyst. It was learned in the analysis that in
this way the patient would demonstrate his great reliance and positive attitude toward the analyst,
but only through this method of expressing disappointment. To speak directly of his concern and
closeness to the analyst would be forbidden because the patient felt that any closeness and trust
would mean that the analyst would have to handle the patient’s destructiveness and would
therefore be destroyed.

Therefore, to keep some distance from the analyst was to preserve him. Conversely, a patient
may often keep his distance because he has already projected bad objects into the analyst and
therefore sees the analyst as a persecutor.

The following case history will illustrate some of the above ideas: A 23-year-old civil engineer
came into analysis because of increasing anxiety over his loneliness. He found himself very aloof
from his fellow offi ce workers toward whom he felt a mixture of fear and contempt and did not
dare, as a consequence, get close to them. His sexual life, other than masturbation, consisted of a
few contacts with prostitutes and one contact with a girl toward whom he had begun to develop
feelings. Subsequent sexual attempts with her resulted in humiliating impotent failures, however,
so he abruptly terminated the relationship with her. His life otherwise was characterized by a
lonely, stark impoverishment in which he spent most of his spare time in his apartment, drinking,
playing the guitar, or reading.

He is the second eldest of four children, having an older sister and a younger brother and sister
respectively. His father was described as an angry, loud, drunk, martinet of a man who once was
a prize-fi ghter. His mother was a willowy, soft-spoken, subtly patronizing martyr of a woman
who was frequently beaten by the father while the children watched in paralyzed horror. When
the patient was 12, 270

Projective identifi cation in the therapeutic process the mother ‘escaped’ from the father and
encouraged her children to come with her. Only the oldest child obliged, however; the others
remained with the father. Immediately thereafter the father moved them away from New York to
a small town in California where he forced them to use

assumed names so that the mother could not trace them and have them brought back to New
York.

Life with father consisted of hearing his insults and temper fi ts, subjecting oneself to Spartan
discipline (the father enforced regular calisthenics upon them as if they were in training), and
consistently being reminded of what a better parent he was than their mother who, he claimed,
wanted them sent to an orphan’s home. After grad-uating from high school the patient left home
against his father’s will and used his savings to enter college to become an engineer.

His initial behaviour in analysis was cold, formal, and detached.

He would describe a very lonely, impoverished life with an eerie detachment. He did not seem to
be involved with his own life.

Provocative gestures at work, such as frequently arriving late, allowing himself to be seen idle,
and arguments with the supervisor, changed into transference phenomena of professing mild to
enormous contempt and ridicule toward the analyst, whose weaknesses and defi ciencies almost
invariably bore a striking resemblance to the patient’s own shortcomings, in addition to
shortcomings of both parents. Examples of some of the projections are as follows: frequently he
would accuse the analyst of being weak and poorly integrated and possibly suffering from a huge
inner impoverishment. Along with this he would state that he felt the analyst also had a hidden
homosexual problem.

These all were projections of his weak self-concept.

On other occasions he would berate the analyst as being too rigid and demanding, and he would
freely express how he hated pleasing him

– that would be like giving in. This perception of the analyst as rigid, autocratic, and hard to
please, represented a projection of the father identifi cation. On still other occasions he would
perceive the analyst as supercilious, polite, ingratiating, insincere, and martyr-like. All these
qualities belonged to his mother identifi cation.

The projections were accepted by the analyst for their psychic validity, and then interpreted as
his need to put bad parts of himself, including bad objects and part objects, into the analyst in
order to rid his ego of these bad contents.

In addition he was symbolically entering the analyst through these projections, to take control of
him by weakening his self-esteem through consistent criticism and 271

Arthur Malin and James S. Grotstein denigration. Not only was he repeating with the analyst
what he had experienced with his father and mother, but he was also taking possession, in
fantasy, of the analyst from within to guarantee total possession of the object. In his life history
there was no precedent for him to assume he could have any relationship with anyone without
total control or total subjugation. Without this guarantee, as it were, there existed no relationship
for him.

The projection of bad parts of himself (and bad objects and part objects) had still another purpose
which closely dovetailed with the mechanisms described. This patient was so trapped in his
schizoid world that he could not trust his good, positive love feelings to be truly good. He had
the conviction that his very love was bad and would be rejected; thus he related with his overtly
bad self in order to establish a relationship and, paradoxically, protect the external object and
himself from destruction. Moreover, he got a particular delight if he felt the analyst was hurt by
his tirades of abuse. As long as the analyst was hurt (i.e. affected), then he as an individual was
having some effect on another person and was therefore asserting his identity and was at the
same time dealing with his deep envy of the analyst’s immutability.

Consistent interpretations of all of these mechanisms wherever they occurred considerably

lessened the negative transference, and the patient was subsequently able to recognize that he
was warding off his deep feelings of dependency on the analyst. Changes occurred by virtue of
analysing the projections rather than by the analyst’s unconsciously or consciously responding as
if they were objectively valid. In other words, this was a new experience for the patient which
allowed him to integrate the previously projected parts, now reintegrated into the ego, so that a
higher level of functioning could occur.

This is an example of transference, but it is also something more than is ordinarily conveyed by
that term. The patient was not merely displacing from the past; he was projecting from within
himself bad contents into the analyst.

By permitting the patient to project into the analyst, that is, to accept the psychic validity of the
projection, a way of establishing a relationship with the patient was developed which allowed
successful interpretation and resolution of this archaic way of relating. It also anticipated and
precluded a negative therapeutic reaction and aided the patient to heal his ego fragments.

In the light of all the above material we should like to offer some speculative ideas in regard to
the general concept of identifi cation.

272
Projective identifi cation in the therapeutic process We suggest the possibility that there is an
early primary identifi cation with the breast-mother and that in a sense no further real identifi
cation takes place. Instead, there is a constant modifying and integrating of this earliest identifi
cation. This might explain the contradiction that appears in the literature in regard to identifi
cation at one point appearing as a normal process of development and at another point as a
pathological defence mechanism. It would seem that normal identifi cation refers to the primary
identifi cation and that any further identifi cation later on in life would be of a more pathological
defensive nature more likely on the level of introjection, that is, an unassimilated foreign-body
reaction in the psychic apparatus. However, normal development does include identifi cation, but
of a far more transient nature than originally assumed, which really has to do with further
structuring, integrating, and synthesizing of the earliest primary identifi cation. What is
commonly thought of as good identifi cations can be seen to be growth of the self through these
mechanisms. It may be stated that we can never change the facts of what has happened to the
patient in his life. What we hope to do, however, is to help the patient integrate his experiences
in a new way so that he

may have a choice in the way he relates to the world. 47

Fairbairn (1952) has made an interesting contribution to the concept of identifi cation. He feels
that primary identifi cation takes place with the pre-ambivalent object, which is then split into
good and bad objects. All future identifi cations are made solely with the bad objects. The good
objects, he states, do not need to be identifi ed with. There is a different kind of internalizing of
the good objects, but this is transitory and is given up as one matures. In other words, the good
objects are loosely held as a scaffolding, as it were, for ego growth and differentiation. As this
takes place, the scaffolding is removed.

To summarize, we are suggesting that projective identifi cation is a normal process existing from
birth. It is one of the most important mechanisms by which growth and development take place
through object relations. This mechanism can be described as one in which objects and
associated affects are re-experienced on a new integrative level so that further synthesis and
development will take place within the ego.

47 See Lichtenstein’s (1961) concept of ‘identity theme’.

273

Arthur Malin and James S. Grotstein We have taken Klein’s concept of projective identifi cation
and have attempted to show how this idea can be greatly broadened to increase our
understanding of normal and pathological development and the therapeutic process. In our view
projective identifi cation seems to be the way in which human beings are able to test their own
inner psychic life by projecting psychic contents out into the environment and perceiving the
environment’s reaction to these projected parts of oneself. This process gives rise to newer
psychic integrations leading to normal growth and development, and is, moreover, of crucial
importance in the therapeutic process.

274

17
On projective identification 48

Thomas H. Ogden

I. Introduction

It is a continuing task of psychoanalytic thinking to attempt to generate concepts and consistent

language that are helpful in understanding the interplay between phenomena in an intrapsychic
sphere (e.g. thoughts and feelings) and phenomena in the sphere of external reality and
interpersonal relations (e.g. the reality of the other person in an object relationship as opposed to
the psychological representation of that person). Psychoanalytic theory suffers from a paucity of
concepts and language that help to bridge these areas. Since projective identifi cation can be
understood as representing one such bridging formulation, it is to the detriment of psychoanalytic
thinking that this concept remains one of the most loosely defi ned and incompletely understood
of psychoanalytic conceptualizations.

This paper attempts to make some steps towards a wider understanding of projective identifi
cation, as well as towards an increased precision of defi nition in this area. The concept of
projective identifi cation will be located in relation to other related psychoanalytic concepts, such
as projection, introjection, identifi cation, internalization, and externalization. In addition, there is
an effort to arrive at a more precise understanding of the nature and function of fantasy in
projective identifi cation, and the relation of that fantasy component to external reality and to real
object relations, specifi cally how

48 This chapter reproduces the text of Ogden, T. (1979). On projective identifi cation.

International Journal of Psychoanalysis , 60 , 357–373.

275

Thomas H. Ogden

projective fantasies (intrapsychic phenomena) abut with real, external objects. Further, the paper
attempts to specify more clearly the expe-riential referents of projective identifi cation. Once
what is meant by projective identifi cation has been clarifi ed, a brief historical overview of the
concept is offered. Finally, on the basis of the understanding of projective identifi cation arrived
at earlier in the paper, there is a discussion of the resulting implications for psychotherapeutic
technique and for clinical theory. This will include an examination of sources of problems in the
handling of projective identifi cations and a view of the role of interpretation in a therapeutic
interaction characterized by projective identifi cations.

II. Projective identifi cation as fantasy and

object relationship

Projective identifi cation is a term that was introduced by Melanie Klein in 1946. Since then,
there has been considerable lack of clarity about what is meant when the term is used, how it
differs from projection on the one hand and from identifi cation on the other, and its relation to
fantasy. The term has been used to refer to a type of projection wherein the person projecting
feels ‘at one with’ the object of the projection (Schafer, personal communication, 1974).

The term is also commonly used to refer to a class of fantasy wherein a part of the self is felt to
be located in another person (Segal, 1964).

Without going further into the different usages of the term, it will suffi ce at this point to say that
the term

‘projective identifi cation’ has been used to refer to a variety of different, but often
complementary, conceptualizations. The defi nition of projective identifi cation that will be
presented in this paper represents a synthesis of, and extension of, contributions made by a
number of analysts.

Projective identifi cation will be used in this paper to refer to a group of fantasies and
accompanying object relations having to do with the ridding of the self of unwanted aspects of
the self; the depositing of those unwanted ‘parts’

into another person; and fi nally, with the ‘recovery’ of a modifi ed version of what was
extruded.

Projective identifi cation will be discussed as if it were composed of a sequence of three parts,
phases, or steps (Malin and Grotstein, 1966). However, the idea of there being three aspects of a
single psychological event better conveys the sense of simultaneity and 276

On projective identifi cation

interdependence that befi ts the three aspects of projective identifi cation that will be discussed.
In a schematic way,

one can think of projective identifi cation as a process involving the following sequence: fi rst,
there is the fantasy of

projecting 49 a part of oneself into another person and of that part taking over the person from
within; then there is

pressure exerted via the interpersonal interaction such that the ‘recipient’ of the projection
experiences pressure to think, feel, and behave in a manner congruent with the projection; fi
nally, the projected feelings, after being

‘psychologically processed’ by the recipient, are reinternalized by the projector.

The fi rst step of projective identifi cation must be understood in terms of wishes to rid oneself of
a part of the self either because that part threatens to destroy the self from within, or because one
feels that the part is in danger of attack by other aspects of the self and must be safeguarded by
being held inside a protective person. This latter psychological use of projective identifi cation
was prominent in a schizophrenic adolescent who vehemently insisted that he opposed
psychiatric treatment and was only coming to his sessions because his parents and the therapist
were forcing him to do so. In reality, this 18-year-old could have resisted far more energetically
than he did and had it well within his power to sabotage any treatment attempt.

However, it was important for him to maintain the fantasy that all of his wishes for treatment and
for recovery were located in his parents and in the therapist so that these wishes would not be
endangered by the parts of himself that he felt were powerfully destructive and intent on the
annihilation of himself.

The type of projective identifi cation involving the fantasy of getting rid of an unwanted, ‘bad’
part of the self by putting it into another person is exemplifi ed by a psychotic obsessional patient
who frequently talked about wishing to put his ‘sick brain’ into the therapist, who would then
have to add up obsessively the numbers on every licence plate that he saw and be tormented by
fears that every time he touched something that was not his, people would accuse

49 The term ‘projection’ will be used to refer to the fantasy of expelling a part of the self that is
involved in the fi rst phase of projective identifi cation even though it is understood that this is
not the same as a projection that occurs outside of the context of a projective identifi cation. The
nature of the difference between projection as an independent process and projection as a part of
projective identifi cation will be discussed later in this paper.

277

Thomas H. Ogden

him of trying to steal it. This patient made it clear that his fantasy was not one of simply ridding
himself of something; it was equally a fantasy of inhabiting another person and controlling him
from within.

His ‘sick brain’ would in fantasy torment the therapist from within in a way that it was currently
felt to be tormenting the patient. This type of fantasy is based on a primitive idea that feelings
and ideas are concrete objects with lives of their own. These ‘objects’ are felt to be located inside
oneself, but it is also felt that they can sometimes be removed from one’s insides and placed into
another person, thereby relieving the self of the effects of containing such entities. The
obsessional patient just described would often in the course of a therapy hour turn his head
violently to the side in an effort to ‘shake loose’ a given worry.

The fantasy of putting a part of oneself into another person and controlling them from within refl
ects a central

aspect of projective identifi cation: the person involved in such a process is operating at least in
part at a developmental level wherein there is profound blurring of boundaries between self and
object representations. The projector feels that the recipient experiences his feeling, not merely a
feeling like his own, but his own feeling that has been transplanted into the recipient. The person
projecting feels ‘at one with’ (Schafer, personal communication, 1974) the person into whom he
has projected an aspect of himself. This is where projective identifi cation differs from
projection. In projection, the projector feels estranged from, threatened by, bewildered by, or out
of touch with, the object of the projection. The person involved in projection might ask,
‘Why would anyone act in such an angry way when there is nothing to be angry about? There’s
something the matter with him.’ In projection, one feels psychological distance from the object;
in projective identifi cation, one feels profoundly connected with the object. Of course, the
contrasting processes are rarely found in pure form; instead, one regularly fi nds a mixture of the
two, with greater or lesser preponderance of feelings of oneness or of feelings of estrangement.

In the second phase of projective identifi cation (more accurately, a second aspect of a single
unit), there is a pressure exerted by the projector on the recipient of the projection to experience
himself and behave in a way congruent with the projective fantasy. This is not an imaginary
pressure. This is real pressure exerted by means of a multitude of interactions between the
projector and the recipient.

278

On projective identifi cation

Projective identifi cation does not exist where there is no interaction between projector and
object. A 12-year-old in-patient, who as an infant had been violently intruded upon
psychologically and physically, highlights this aspect of projective identifi cation. The patient
said and did almost nothing on the ward, but made her presence powerfully felt by perpetually
jostling and bumping into people, especially her therapist. This was generally experienced as
infuriating by other patients and by the staff. In the therapy hours (often a play therapy), her
therapist said that he felt as if there were no space in the room for him. Everywhere he stood
seemed to be her spot. This form of interaction represents a form of object relationship wherein
the patient puts pressure on the therapist to experience himself as inescapably intruded upon.
This interpersonal interaction constitutes the ‘induction phase’ of this patient’s projective identifi
cation.

The psychotic obsessional patient mentioned earlier consistently generated a type of therapeutic
interaction that illuminated the induction phase of projective identifi cation. This 14-year-old
patient was born with pyloric stenosis and suffered from severe projectile vomiting for the entire
fi rst month of his life, before the condition was diagnosed and surgically corrected. His
psychological experience since then has been continuous in the sense that he has imagined
himself to be inhabited by attacking presences: scolding parents, burning stomach pains,
tormenting worries, and powerful rage over which he feels little or no control. The initial phases
of his therapy consisted almost exclusively of his attempt to torment the therapist by kicking the
therapist’s furniture, repeatedly ringing the waiting room buzzer, and by ruminating without
pause in a high-pitched whine. All of this invited retaliatory anger on the part of the therapist,
and it was to the extent that the therapist experienced feelings of extreme tension and helpless
rage that the patient felt momentarily calmed. The patient was fully conscious of both his
attempts to get the therapist to feel angry, as well as the calming and soothing effect that that had
on him. I would understand this therapeutic interaction as an enactment of the patient’s fantasy
that anger and tension are noxious agents within him that he attempted to get rid of by placing
them in the therapist. However, as with his projectile vomiting, a solution is not simple: the
noxious agents within that he wishes to rid himself of (anger/food/parents) are also essential for
life. Projective identifi cation offers a compromise solution wherein the patient could in fantasy
rid himself of the noxious, but life-giving, objects within 279
Thomas H. Ogden

himself while at the same time keeping them alive inside a partially separate object. This solution
would be merely a fantasy without the accompanying object relationship in which the patient
exerted terrifi c pressure on the therapist to conform to the projective fantasy.

When there was evidence of verifi cation of the projection (i.e. when the therapist showed
evidence of tension and

anger), the patient experienced a sense of relief since that offered confi rmation that the
noxious/life-giving agents had been both extruded and preserved.

I would like to mention very briefl y a third clinical example in which the induction phase of
projective identifi cation will be focused upon. T. A. Tähkä of Finland (Tähkä, 1977) has
reported that a profound lack of concern for a patient on the part of the therapist often
immediately precedes the patient’s suicide. Although Dr Tähkä does not approach this
phenomenon from the point of view of projective identifi cation, his observations can be
understood as refl ecting the patient’s attempt to induce in the therapist his own state of total lack
of caring for himself or for his life. This could be viewed as an attempt on the part of the patient
to: (1) Rid himself of this malignant absence of concern for life. (2) Make himself understood by
the therapist by inducing the feeling in him. The process of this

‘induction’

of feelings constitutes the second stage of projective identifi cation.

Warren Brodey (1965), from a family observational viewpoint, has studied one mode of
interaction that serves to generate pressure to comply to a projective fantasy. He describes very
vividly the way one member of a family may manipulate reality in an effort to coerce another
member into ‘verifying’ a projection. Reality that is not useful in confi rming a projection is
treated as if it did not exist. This manipulation of reality and the resultant undermining of reality
testing is but one technique in the generation of pressure for compliance with a projective
fantasy.

One further point that needs to be made with regard to the induction of a projective identifi
cation is the ‘or else’ that looms behind the pressure to comply with a projective identifi cation. I
have described elsewhere (Ogden, 1976, 1978a) the pressure on an infant to behave in a manner
congruent with the mother’s pathology, and the ever-present threat that if the infant were to fail
to comply, he would become non-existent for the mother. This threat is the

‘muscle’ behind the demand for compliance: ‘If you are not what I need you to be, you don’t
exist for me,’ or in other language, ‘I can only see in you what I put there, and so if I don’t see
that in you, 280

On projective identifi cation

I see nothing.’ In the therapeutic interaction, the therapist is made to feel the force of the fear of
becoming non-existent for the patient if he were to cease to behave in compliance with the
patient’s projective identifi cation. (See Ogden, 1978a for a detailed discussion of a therapy
revolving around this issue.) So far, I have talked about two aspects of projective identifi cation:
the fi rst involves a fantasy of ridding oneself of an aspect of the self and of the entry of that part
into another person in a way that controls the other person from within. The second aspect of
projective identifi cation that has been discussed is the interpersonal interaction that supports the
fantasy of inhabiting and controlling another person.

Through the projector’s interaction with the object, two aspects of the fantasy are verifi ed: (1)
The idea that the object has the characteristics of the projected aspects of the self. (2) That the
object is being controlled by the person projecting. In fact, the ‘infl uence’ is real, but it is not the
imagined absolute control by means of transplanted aspects of the self inhabiting the object;
rather, it is an external pressure exerted by means of interpersonal interaction. This brings us to
the third phase of projective identifi cation, which involves the

‘psychological processing’ of the projection by the recipient, and the re-internalization of the
modifi ed projection by the projector. In this phase of projective identifi cation, the recipient of
the projection experiences himself in part as he is pictured in the projective fantasy.

The reality is that the recipient’s experience is a new set of feelings experienced by a person
different from the projector. The recipient’s feelings may be close to those of the projector, but
those feelings are not transplanted feelings. The recipient is the author of his own feelings albeit
feelings elicited under a very specifi c kind of pressure from the projector. The elicited feelings
are the product of a different personality system with different strengths and weaknesses. This
fact opens the door to the possibility that the projected feelings (more accurately, the congruent
set of feelings elicited in the recipient) will be handled differently from the manner in which the
projector has been able to handle them. A different set of defences and other psychological
processes may be employed by the recipient so

that the feelings are ‘processed’, ‘metabolized’ (Langs, 1976), ‘contained’

(Bion, 1961), or managed differently. The fact that the projector is employing projective identifi
cation indicates that he is dealing with a given aspect of himself by attempting to rid himself of
the unwanted feelings and representations. Alternative psychological processes that 281

Thomas H. Ogden

could potentially be employed by the recipient to handle the same set of feelings would include
attempts at integration with other aspects of the personality, attempts at mastery through
understanding, and sublimation. These methods of dealing with feelings contrast with projective
identifi cation in that they are not basically efforts to avoid, get rid of, deny, or forget feelings
and ideas; rather, they represent different types of attempts to live with, or contain, an aspect of
oneself without disavowal. If the recipient of the projection can deal with the feelings projected

‘into’ him in a way that differs from the projector’s method, a new set of feelings is generated
which can be viewed as a ‘processed’ version of the original projected feelings. The new set of
feelings might involve the sense that the projected feelings, thoughts and representations can be
lived with, without damaging other aspects of the self or of one’s valued external or internal
objects (cf. Little, 1966). The new experience (or amalgam of the projected feelings plus aspects
of the recipient) could even include the sense that the feelings in question can be valued and at
times enjoyed. It must be kept in mind that the idea of ‘successful’ processing is a relative one
and that all processing will be incomplete and contaminated to an extent by the pathology of the
recipient.

This ‘digested’ projection is available through the recipient’s interactions with the projector for
internalization by the projector. The nature of this internalization (actually a re-internalization)
depends upon the maturational level of the projector and would range from primitive types of
introjection to mature types of identifi cation (cf.

Schafer, 1968). Whatever the form of the re-internalization process, the internalization of the
metabolized projection offers the projector the potential for attaining new ways of handling a set
of feelings that he could only wish to get rid of in the past. To the extent that the projection is
successfully processed and re-internalized, genuine psychological growth has occurred. (The
consequences of inadequate reception of, or processing of, projective identifi cations are
discussed later in this paper.)

The following is an example of projective identifi cation involving a recipient more integrated
and mature than the projector. Mr J had been a patient in analysis for about a year and the
treatment seemed to both patient and analyst to be bogging down. The patient repetitively
questioned whether he was ‘getting anything out of it’, ‘maybe it’s a waste of time’ etc. Mr J had
always grudgingly paid his bills, but gradually they were being paid later and later, leaving the
analyst to 282

On projective identifi cation

wonder whether the bill was going to be paid at all. The analyst found himself questioning
whether the patient might drop out of treatment, leaving that month’s and the previous month’s
bills unpaid.

Also, as the sessions dragged on, the analyst thought about colleagues who held fi fty minute
sessions instead of fi fty-fi ve minute ones, and charged the same fee as this analyst. Just before
the beginning of one session, the analyst considered shortening the ‘hour’ by making the patient
wait a couple of minutes before letting him into the offi ce.

All of this occurred without attention being focused on it either by the patient or the analyst.
Gradually, the analyst found himself having diffi culty ending the sessions on time because of an
intensely guilty feeling that he was not giving the patient ‘his money’s worth’. After this diffi
culty with time repeated itself again and again over several months, the analyst was gradually
able to begin to understand his trouble in maintaining the ground rules of the analysis. It began to
be apparent to the analyst that he had been feeling greedy for expecting to be paid for his

‘worthless’ work and was defending himself against such feelings by being so generous with his
time that no one could accuse him of greed. With this understanding of the feelings that were
being engendered in him by the patient, the analyst was able to take a fresh look at the patient’s
material. Mr J’s father had deserted him and his mother when the patient was 15 months old. His
mother, without ever explicitly saying so, had held the patient responsible
for this. The unspoken, shared feeling was that it was the patient’s greediness for the mother’s
time, energy and affection that had resulted in the father’s desertion. The patient developed an
intense need to disown and deny feelings of greed. He could not tell the analyst that he wished to
meet more frequently because he experienced this wish as greediness that would result in
abandonment by the (transference) father and in attack by the (transference) mother that he saw
in the analyst. Instead, the patient insisted that the analysis and the analyst were totally
undesirable and worthless. The interaction with the analyst subtly engendered in the analyst
intense feelings of a type of greed that was felt to be so unacceptable to the analyst that the
analyst at fi rst also made an attempt to deny and disown it. For the analyst, the fi rst step in
integration of the feeling of greediness was the ability to register a perception of himself
experiencing guilt and defending himself against his feelings of greed. He could then mobilize an
aspect of himself that was interested in understanding his greedy and guilty feelings, rather than
trying to deny, disguise, displace or project them.

283

Thomas H. Ogden

Essential for this aspect of psychological work was the analyst’s feeling that he could have
greedy and guilty feelings without being damaged by them. It was not the analyst’s greedy
feelings that were interfering with his therapeutic work; rather, it was his need to disavow such
feelings by denying them and by putting them into defensive activity.

As the analyst became aware of, and was able to live with, this aspect of himself and of his
patient, he became better able to handle the fi nancial and time boundaries of the therapy. He no
longer felt that he had to hide the fact that he was glad to receive money given in payment for his
work. After some time, the patient commented as he handed the analyst a cheque (on time), that
the analyst seemed happy to get ‘a big, fat cheque’ and that that wasn’t very becoming to a
psychiatrist. The analyst chuckled and said that it is nice to receive money. During this
interchange, the analyst’s acceptance of his hungry, greedy, devouring feelings, together with his
ability to integrate those feelings with other feelings of healthy self-interest and self-worth was
made available for internalization by the patient. The analyst at this point chose not to interpret
the patient’s fear of his own greed and his defensive, projective fantasy. Instead, the therapy
consisted of the digesting of the projection and the process of making it available for re-
internalization through the therapeutic interaction.

In the light of the above discussion, it is worth considering whether this kind of understanding of
projective identifi cation may not bear directly on the question of the means by which
psychotherapy and psychoanalysis contribute to psychological growth. It may be that the essence
of what is therapeutic for the patient lies in the process of the therapist or analyst making himself
available to receive the patient’s projections, utilizing facets of his more mature personality
system in the processing of the projection, and then making available the digested projection for
reinternalization through the therapeutic interaction (Langs, 1976; Malin and Grotstein, 1966;
Searles, 1963).

To summarize, projective identifi cation is a set of fantasies and object relations that can be
schematically conceptualized as occurring in three phases: fi rst, the fantasy of ridding oneself of
an unwanted part of oneself and of putting that part into another person in a controlling way;
then the induction of feelings in the recipient that are congruent with the projective fantasy by
means of an interpersonal interaction; and fi nally, the processing of the projection by the
recipient, followed by the re-internalization by the projector of the

‘metabolized projection’.

284

On projective identifi cation

III. The early developmental setting

Projective identifi cation, as described in the previous section, is a

psychological process that is simultaneously a type of defence, a mode of communication, a

primitive form of object relationship, and a pathway for psychological change. As a defence,
projective identifi -

cation serves to create a sense of psychological distance from unwanted (often frightening)
aspects of the self; as a mode of communication, projective identifi cation is a process by which
feelings congruent with one’s own are induced in another person, thereby creating a sense of
being understood by or of being ‘at one with’ the other person. As a type of object relationship,
projective identifi cation constitutes a way of being with and relating to a partially separate
object; and fi nally, as a pathway for psychological change, projective identifi cation is a process
by which feelings like those that one is struggling with, are psychologically processed by another
person and made available for re-internalization in an altered form.

Each of these functions of projective identifi cation evolves in the context of the infant’s early
attempts to perceive, organize, and manage his internal and external experience and to
communicate with his environment. The infant is faced with an extremely complicated,
confusing, and frightening barrage of stimuli. With the help of a ‘good enough’ mother
(Winnicott, 1952), the infant can begin to organize his experience. In this effort towards
organization, the infant discovers the value of keeping dangerous, painful, frightening
experiences separate from comforting, soothing, calming ones (Freud, 1920). This kind of
‘splitting’ becomes established as a basic part of the early psychological modes of organization
and of defence ( Jacobson, 1964; Kernberg, 1976). As an elaboration of, and support for, this
mode of organization, the infant utilizes fantasies of ridding himself of aspects of himself
(projective fantasies) and fantasies of taking into himself aspects of others (introjective
fantasies). These modes of thought help the infant to keep what is valued psychologically
separate from, and in fantasy safe from, what is felt to be dangerous and destructive.

These attempts at psychological organization and stability occur within the context of the
mother-infant dyad. Spitz (1965) describes the earliest ‘quasi-telepathic’ communication
between mother and infant as being of a ‘conesthetic type’ wherein sensing is visceral and
stimuli are ‘received’ as opposed to being ‘perceived’. The mother’s 285

Thomas H. Ogden
affective state is ‘received’ by the infant and is registered in the form of emotions. The mother
also utilizes a conesthetic mode of communication. Winnicott beautifully describes the state of
heightened maternal receptivity that is seen in the mother of a newborn: I do not believe it is
possible to understand the functioning of the mother at the very beginning of the infant’s life
without seeing that she must be able to reach this state of heightened sensitivity, almost an
illness, and then recover from it. . . . Only if a mother is sensitized in the way I am describing can
she feel herself into the infant’s place, and so meet the infant’s needs.

(Winnicott, 1956)

It is in this developmental setting that the infant develops the process of projective identifi cation
as a mode of fantasy with accompanying object relations that serve both defensive and
communicative functions. Projective identifi cation is an adjunct to the infant’s efforts at keeping
what is felt to be good at a safe distance from what is felt to be bad and dangerous. Aspects of
the infant can in fantasy be deposited in another person in such a way that the infant does not feel
that he has lost contact either with that part of himself or with the other person. In terms of
communication, projective identifi cation is a means by which the infant can feel that he is
feeling. The infant cannot describe his feelings in words for the mother; instead, he induces those
feelings in her. In addition to serving as a mode of interpersonal communication, projective
identifi cation constitutes a primitive type of object relationship, a basic way of being with an
object that is only partially separate psychologically. It is a transitional form of object
relationship that lies between the stage of the subjective object and that of true object relatedness.

This brings us to the fourth function of projective identifi cation, that of a pathway for
psychological change. The following hypothetical interaction will be presented in an attempt to
describe the place of this aspect of projective identifi cation in early development.

Let us imagine that a child is frightened by his wish to destroy and annihilate anyone who
frustrates or opposes him.

One way of his handling these feelings would be to project unconsciously his destructive wishes
in fantasy into his mother, and through the real interaction with her, engender feelings in her that
she is a ruthless, selfi sh person who wishes to demolish anything standing in the way of the 286

On projective identifi cation

satisfaction of her aims and wishes. One way a child could engender this feeling in his mother
would be through persistently stubborn behaviour in many areas of daily activity, e.g. by making
a major battle out of his eating, his toileting, his dressing, getting him to sleep at night and up in
the morning, leaving him with another caretaker, etc.

The mother might unrealistically begin to feel that she perpetually storms around the house in a
frenzy of frustrated rage ready to kill those that stand between her and what she desires. A
mother who had not adequately resolved her own confl icts around such destructive wishes and
impulses would fi nd it diffi cult to live with the heightening of these feelings. She might attempt
to deal with such feelings by withdrawing from the child and never touching him.

Or she might become hostile or assaultive toward him or dangerously careless with him. In order
to keep the child from becoming the target, the mother might displace or project her feelings on
to her husband, parents, employer, or friends. Alternatively, the mother may feel so guilty about,
or frightened of, these frustrated, destructive feelings that she might become overprotective of
the child, never allowing him out of her sight and never allowing him to be adven-turous for fear
that he might get hurt. This type of ‘closeness’ may become highly sexualized, e.g. by the mother
constantly caressing the child in an effort to demonstrate to herself that she is not harming him
with her touch. Any of these modes of dealing with the engendered feelings would result in the
confi rmation for the child of his feeling that angry wishes for the demolition of frustrating
objects are dangerous to himself and to his valued objects. What would be internalized from the
mother in this case would be an even stronger conviction than he had held before that he must
get rid of such feelings. In addition, the child could internalize aspects of the mother’s
pathological methods of handling of this type of feeling (e.g. excessive projection, splitting,
denial, or violent enactment as a mode of tension relief or as a mode of expression of feeling).
On the other hand, ‘good enough’

handling of the projected feelings might involve the mother’s ability to integrate the engendered
feelings with other aspects of herself, e.g. her healthy self-interest, her acceptance of her right to
her anger and resentment at her child for standing in the way of what she wants, her confi dence
that she can contain such feelings without acting on them with excessive withdrawal or
retaliatory attack. None of this need be available to the mother’s conscious awareness.

This act of psychological integration constitutes the 287

Thomas H. Ogden

processing phase of projective identifi cation. Through the mother’s interactions with the child,
the processed projection (which involves the sense of the mother’s mastery of her frustrated
feelings and destructive, retaliatory wishes) would be available to the child for re-internalization.

It can be seen from this developmental perspective that the concept of projective identifi cation is
entirely separable from a Kleinian theoretical or developmental framework, and for that matter,
from that of any other school of psychoanalytic thought. In particular, there is no necessary tie
between projective identifi cation and the death instinct, the concept of envy, the concept of
constitutional aggression, or any other facet of specifi cally Kleinian clinical theory or
metapsychology.

Moreover, there is nothing to tie the concept of projective identifi -

cation to any given developmental timetable. The concept of projective identifi cation requires
only that: (1) The projector (infant, child or adult) be capable of projective fantasy (albeit often
very primitive in its mode of symbolization) and specifi c types of object-relatedness that are
involved in the induction and re-internalization phases of projective identifi cation. (2) That the
object of the projection be capable of the type of object-relatedness that is involved in ‘receiving’

a projection in addition to being capable of some form of ‘processing’

of the projection. At some point in development, the infant becomes capable of these
psychological tasks and it is only at that point that the concept of projective identifi cation would
become applicable. It is unfortunate that the discussion of projective identifi cation so often
becomes ensnared in a debate over the Kleinian developmental timetable which is in no way
inherent to the concept of projective identifi cation.

IV. An historical perspective

Before discussing the technical and theoretical implications of the above discussion, it will be
useful to present a brief historical overview of the important contributions to the development
and application of the concept of projective identifi cation. The concept and term

‘projective identifi cation’ were introduced by Melanie Klein in

‘Notes on Some Schizoid Mechanisms’ (Klein, 1946). In this paper, Mrs Klein applies the term
‘projective identifi cation’ to a psychological process arising in the paranoid-schizoid phase of
development, 288

On projective identifi cation

wherein ‘bad’ parts of the self are split off and projected ‘into’ another person in an effort to rid
the self of one’s

‘bad objects’, which threaten to destroy oneself from within. These bad objects (psychological
representations of the death instinct) are projected in an effort to

‘control and take possession of the object’. The only other paper in which Mrs Klein discusses
projective identifi cation at any length is

‘On Identifi cation’ (Klein, 1955). In that paper, Mrs Klein, by means of a discussion of a story
by Julian Green (‘If I Were You’) offers a vivid account of the subjective experience involved in
the process of projective identifi cation.

In Green’s story, the devil grants the hero the power to leave his own body and enter and take
over the body and life of anyone he chooses. Mrs Klein’s description of the hero’s experience in
projecting himself into another person captures the sense of what it is like to inhabit someone
else, control them, and yet not totally lose the sense of who one really is. It is the sense of being a
visitor in the other person, but also of being changed by the experience in a way that will make
one forever different from the way one was before. In addition, this account brings home an
important aspect of Mrs Klein’s view of projective identifi cation: the process of projective
identifi cation is a psychologically depleting one that leaves the projector impoverished until the
projected part is successfully reinternalized. The attempt to control another person and have them
act in congruence with one’s fantasies requires tremendous vigilance and a very great
expenditure of psychological energy that leaves a person psychologically weakened.

Wilfred Bion (1959, 1961) has made important steps in elaborating upon and applying the
concept of projective identifi cation. He views projective identifi cation as the single most
important form of interaction between patient and therapist in individual therapy, as well as in
groups of all types. Bion’s strongly clinical perspective is helpful in emphasizing an aspect of
this process that is very little elucidated by Mrs Klein: ‘The analyst feels that he is being
manipulated so as to be playing a part, no matter how diffi cult to recognize, in somebody else’s
phantasy’ (Bion, 1961). Bion is consistently aware that in addition to projective identifi cation’s
being a fantasy, it is also a manipulation of one person by another, i.e. an interpersonal
interaction.

Bion’s work also manages to capture some of the strangeness and mystery that characterize the
experience of being involved as the container (i.e. the recipient) of a projective identifi cation.
He likens the experience to the idea of ‘a thought without a thinker’ (Bion, 289

Thomas H. Ogden

1977). In a sense, being the recipient of a projective identifi cation is like having a thought that is
not one’s own. A further point that Bion makes is the idea that there is a severely destructive
impact of a parent (or therapist) who cannot allow himself to receive the projective identifi
cations of the child (or patient): ‘The environment . . . at its worst denied to the patient the use of
the mechanisms of splitting and projective identifi cation’ (Bion, 1959). An essential part of
normal development is the child’s experience of his parents as people who can safely and
securely be relied upon to act as containers for his projective identifi cations.

Herbert Rosenfeld contributed several important early papers (Rosenfeld, 1952b, 1954) on the
clinical applications of projective identifi cation theory to the understanding and treatment of
schizophrenia. In particular, he used the concept of projective identifi cation to trace the genetic
origins of depersonalization and confusional states.

The development and application of the concept of projective identifi cation has not been limited
to the work done by Melanie Klein and her followers. Even though the term projective identifi
cation is not always used by members of other schools of analytic thought, the work of non-
Kleinians has been a fundamental part of the development of the concept. For example, Donald
Winnicott rarely used the term projective identifi cation in his writing, but I would view a great
deal of his work as a study of the role of maternal projective identifi cations in early development
and of the implications of that form of object relatedness for both normal and pathological
development, e.g. his concepts of impingement and mirroring (Winnicott, 1952, 1967).

Michael Balint’s accounts (Balint, 1952, 1968) of his handling of therapeutic repression
(especially in the phase of treatment that he calls the ‘new beginning’) focuses very closely on
technical considerations which have direct bearing on the handling of projective identifi cations.
Balint cautions us against having to interpret or in other ways having to act on the feelings the
patient elicits; instead, the therapist must ‘accept’, ‘feel with the patient’, ‘tolerate’,

‘bear with’

the patient and the feelings he is struggling with and asking the therapist to recognize.

The analyst [when successfully handling the patient’s regression] is not so keen on
‘understanding’ everything immediately, and in particular, on ‘organizing’ and changing
everything undesirable by 290

On projective identifi cation

his correct interpretations; in fact, he is more tolerant towards the patient’s sufferings and is
capable of bearing with them – i.e. of admitting his relative impotence – instead of being at pains
to

‘analyse’ them away in order to prove his therapeutic omnipotence.

(Balint, 1968, p. 184)

I would view this in part as an eloquent statement on the analyst’s task of keeping himself open
to receiving the patient’s projective identifi cations without having to act on these feelings.

Harold Searles enriches the language that we have for talking about the way a therapist (or
parent) must attempt to make himself open to receiving the projective identifi cations of the
patient (or child). In his 1963 paper on

‘Transference Psychosis in the Psychotherapy of Chronic Schizophrenia’, Searles discusses the

importance of the therapist’s refraining from rigidly defending himself against the experiencing
of aspects of the patient’s feelings:

‘The patient develops ego-strengths . . . via identifi cation with the therapist who can endure, and
integrate into his own larger self, the kind of subjectively non-human part-object relatedness
which the patient fosters in and needs from him’. And later in the same paper, Searles adds,

The extent to which the therapist feels a genuine sense of deep participation in the patient’s
‘delusional transference’

relatedness to him during the phase of therapeutic symbiosis . . . is diffi cult to convey in words;
it is essential that the therapist come to know that such a degree of feeling-participation is not
evidence of

‘counter-transference psychosis’ but rather is the essence of what the patient needs from him at
this crucial phase of the treatment.

Searles is here presenting a view that therapy, at least in certain phases of regression, can
progress only to the extent that the therapist can allow himself to feel (with diminished intensity)
what the patient is feeling, or in the terminology of projective identifi cation, to allow himself to
be open to receiving the patient’s projections. This

‘feeling-participation’ is not equivalent to becoming as sick as the patient because the therapist,
in addition to receiving the projection, must process it and integrate it into his own ‘larger’
personality, and make this integrated experience available to the patient for reinternalization. In
his recent article, ‘The Patient as Therapist to the Analyst’

(Searles, 1975), Searles describes in detail the opportunity 291

Thomas H. Ogden

for growth in the analyst that is inherent in his struggle to make himself open to his patient’s
projective identifi cations.

There has been a growing body of literature that has attempted to clarify the concept of
projective identifi cation and has made efforts to integrate the concept into a non-Kleinian
psychoanalytic framework.

Malin and Grotstein (1966) present a clinical formulation of projective identifi cation in which
they help make this very bulky concept more manageable by discussing it in terms of three
elements: the projection, the creation of an

‘alloy’ of external object plus projected self, and re-internalization. These authors present the
view that therapy consists of the modifi cation of the patient’s internal objects by the process of
projective identifi cation.

Interpretation is seen as a way in which the patient can be helped to observe ‘how his projections
have been received and acknowledged by the analyst’.

Finally, I would like to mention the work of Robert Langs (1975, 1976) who is currently
involved in the task of developing an adaptational-interactional framework of psychotherapy and
psychoanalysis.

His efforts represent a growing sense of the importance and usefulness of the concept of
projective identifi cation as a means of understanding the therapeutic process (see also Kernberg,
1968, 1976; Nadelson, 1976). Langs contends that it is necessary for analytic theory to shift from
viewing the analyst as a screen to viewing him as a ‘container for the patient’s pathological
contents who is fully participating in the analytic interaction’ (Langs, 1976). By making such a
shift, we clarify the nature of the therapist’s response to the patient’s transference and non-
transference material and are in a better position to do the self-analytic work necessary for the
treatment of the patient, in particular for the correction of errors in technique. For Langs,
projective identifi cation is the basic unit of study within an interactional frame of reference.

V. Implications for technique and for clinical theory I would like now to move to a discussion
of several technical and theoretical implications of the view of projective identifi cation
presented above.

1. A question that immediately arises is, ‘What does a therapist

“do” when he observes that he is experiencing himself in a way that is congruent with his
patient’s projective fantasy, i.e. when he is 292

On projective identifi cation

aware that he is the recipient of his patient’s projective identifi cation?’ One answer is that the
therapist ‘does’

nothing; instead, he attempts to live with the feelings engendered in him without denying his
feelings or in other ways trying to rid himself of the feelings. This is what is meant by ‘making
oneself open to receiving a projection’.
It is the task of the therapist to contain the patient’s feelings. For example, when the patient is
feeling that he is hopelessly unmother-able, unloveable, and untreatable, the therapist must be
able to bear the feeling that the therapist and the therapy are worthless for this hopeless patient,
and yet at the same time not to act on the feelings by terminating the therapy (cf. Nadelson,
1976). The ‘truth’ about himself that the patient is presenting must be treated as a type of
transitional phenomenon (Winnicott, 1953) wherein the question of whether the patient’s ‘truth’
is reality or fantasy is never an issue. As with any transitional phenomenon, it is both reality and
fantasy, subjective and objective at the same time. In this light, the question

‘If the patient can never get better, why should the therapy continue?’

never needs to be acted upon. Instead, the therapist attempts to live with the feeling that he is
involved in a hopeless therapy with a hopeless patient and is, himself, a hopeless therapist. This
of course is a partial truth that the patient

experiences as a total truth. The ‘truth’

of the patient’s feelings must be experienced by the therapist as emotionally true just as the
good-enough mother must be able to share the truth in her child’s feelings about the comforting
and life-giving powers of his piece of satin.

There are several further aspects of the question raised about the handling of projective identifi
cation that need to be considered. The fi rst is that the therapist is not simply an empty receptacle
into which the patient can ‘put’

projective identifi cations. The therapist is a human being with his own past, his own repressed
unconscious, his own confl icts, his own fears, his own psychological diffi culties. The feelings
that patients are struggling with are, by their nature, highly charged, painful, confl ict-laden areas
of human experience for the therapist as well as for the patient. Hopefully, the therapist, through
the benefi t of greater integration in the course of his own developmental experience and in the
course of his analysis, is less frightened of, and less prone to run from, these feelings than is the
patient.

However, we are not dealing with an ‘all or nothing’ phenomenon here, and the handling of the
feelings projected by the patient require considerable effort, skill, and ‘strain’ (Winnicott, 1960)
on the part of 293

Thomas H. Ogden

the therapist. One major tool at the disposal of the therapist in his efforts at containing his
patient’s projective identifi cations is his ability to bring understanding to what he is feeling and
to what is occurring between himself and his patient. The therapist’s theoretical training, his
personal analysis, his experience, his psychological-mindedness, and his psychological language
can all be brought to bear on the experience he is attempting to understand and to contain.

The question now arises, ‘How much of the therapist’s effort at understanding the patient’s
projective identifi cation is put to the patient in the form of interpretations?’ The therapist’s
ability not only to understand but also to formulate clearly and precisely his understanding in
words is basic to his therapeutic effectiveness (Freud, 1914; Glover, 1931). In the case of
working with projective identifi cations, this is so not only because such verbal understandings
may be of value to the patient in the form of well-timed clarifi cations and interpretations, but
equally because these understandings are an essential part of the therapist’s effort to contain the
feelings engendered in him. The therapist’s understanding may constitute a correct interpretation
for the therapist, but may not be at all well-timed for the patient. In this case, the interpretation
should remain ‘a silent one’ (Spotnitz, 1969), i.e. it is formulated in words in the therapist’s
mind, but not verbalized to the patient. Another aspect of the importance of the silent
interpretation is that it can contain a much heavier weight of self-analytic material than one
would include in an interpretation offered to the patient. Continued self-analysis in this way is
invaluable in a therapist’s attempts to struggle with, contain, and grow from the feelings his
patients are eliciting in him.

The other side of this must also be mentioned. There is a danger that the therapist in his handling
of projective identifi cations may be tempted to use the patient’s therapy exclusively as an arena
in which to fi nd help with his own psychological problems. This can result in a repetition for the
patient of an early pathogenic interaction (frequently reported in the childhood of pathologically
narcissistic patients) wherein the needs of the mother were the almost exclusive focus of the
mother-child relationship. (See Ogden, 1974, 1976, 1978a for further discussion of this form of
mother-child interaction.) 2. The subject of the recognition of errors in the handling of projective
identifi cations and the corrective steps that can be taken has been addressed in various places in
the above discussion. Errors in technique very often refl ect a failure on the part of the therapist
to 294

On projective identifi cation

process adequately the patient’s projective identifi cation. Either through an identifi cation with
the patient’s methods of handling the projected feelings, or through reliance on his own
customary defences, the therapist may come to rely excessively on denial, splitting, projection,
projective identifi cation, or enactment in his efforts to defend against the engendered feelings.
This basically defensive stance can result in ‘therapeutic misalliances’

(Langs, 1975) wherein the patient and therapist ‘seek gratifi cation and defensive reinforce-ments
in their

relationship’. In order to support his own defences, the therapist may introduce deviations in
technique, and may even violate the basic ground rules and framework of psychotherapy and
psychoanalysis, e.g. by extending the relationship into social contexts, by giving gifts to the
patient or by encouraging the patient to give him gifts, by breaches of confi dentiality, etc. A
therapist’s failure adequately to process a projective identifi cation is refl ected in one of two
ways: either by his rigidly defending himself against awareness of the feelings engendered, or by
allowing the feeling or the defence against it to be translated into action. The consequences of
either type of failure to contain a projective identifi cation are that the patient re-internalizes his
own projected feelings combined with the therapist’s fears about, and inadequate handling of,
those feelings.

The patient’s fears and pathological defences are reinforced and expanded. In addition, the
patient may despair about the prospect of being helped by a therapist who shares signifi cant
aspects of his pathology.

3. The patient is not the only person in the therapeutic dyad who employs projective identifi
cation. Just as the patient can apply pressure to the therapist to comply with his projective
identifi cations, the therapist similarly can put pressure on the patient to validate his own
projective identifi cations. Therapists have an intricately over-determined wish for their patients
to ‘get better’ and this is often the basis for an omnipotent fantasy that the therapist has turned
the patient into the wished-for patient. Very often, the therapist, through the therapeutic
interaction, can exert pressure on the patient to behave as if he were that wished for, ‘cured’
patient. A relatively healthy patient can often become aware of this pressure and alert the
therapist to it by saying something like, ‘I’m not going to let you turn me into another of your
“successes”.’ This kind of statement, however over-determined, should alert the therapist to the
possibility that he may be engaged in projective identifi cation and that the 295

Thomas H. Ogden

patient has successfully processed his projections. It is far more damaging to the patient and to
the therapy when the patient is unable to process a projective identifi cation in this way and has
either to comply with the pressure (by becoming the ‘ideal’ patient) or rebel against the pressure
(by an upsurge of resistance or by termination of therapy).

Winnicott (1949) also reminds us that therapists’ and parents’

wishes for their patients and children are not exclusively for cure and growth. There are also
hateful wishes to attack, kill or annihilate the patient or child. A stalemated therapy, a perpetually
silent patient, a fl urry of self-destructive or violent activity on the part of the patient, may all be
signs of the patient’s efforts to comply with a therapist’s projective identifi cation that involves
an attack upon or the annihilation of the patient. As Winnicott suggests, it is imperative that a
parent or therapist be able to integrate his or her anger and murderous wishes toward their
children and patients without enacting these feelings or having to get rid of them through denial
and projection.

Persistent and unchanging projective identifi cations on the part of the therapist should, if
recognized, alert the therapist to a need to examine seriously his own psychological state and
possibly to seek further analysis.

4. In the light of the understanding of projective identifi cation outlined in this paper, I would
like to clarify the relationship of projective identifi cation to a group of related psychological
processes: projection, introjection, identifi cation, and externalization. As mentioned earlier,
projection in a broad sense is a mode of thought in which one experiences oneself as having
expelled an aspect of oneself. A distinction has to be drawn between the projective mode of
thought involved in projective identifi cation and projection as an independent process.

In the former, the individual employs a projective mode of thought in his fantasy of ridding
himself of a part of himself and inhabiting another person with that part. The subjective
experience is one of being at one with the other person with regard to the expelled feeling, idea,
self-representation, etc. In contrast, in projection as an independent process, the aspect of oneself
that is expelled is disavowed and is attributed to the object of the projection. The projector does
not feel kinship with the object and, on the contrary, often experiences the object as foreign,
strange, and frightening. In projective identifi cation, the projective mode of thought is but one
aspect of a dynamic interplay between projection and internalization. However, it must be borne
in 296

On projective identifi cation

mind that the distinction between projection and projective identifi cation is not an all-or-nothing
affair. As Knight (1940) pointed out, every projective process involves an interaction with an
introjective one and vice versa.

Projection and projective identifi cation should be viewed as two ends of a gradient in which
there is increasing preponderance of interplay between the projective and introjective processes
as one moves toward the projective identifi cation end of the gradient.

Just as a projective mode of thought, as opposed to projection, can be seen as underlying the fi
rst phase of projective identifi cation, one can understand the third phase as being based on an
introjective mode as opposed to introjection. In the fi nal phase of projective identifi cation, the
individual imagines himself to be repossessing an aspect of himself that has been ‘reposing’
(Bion, 1959) in another person. In conjunction with this fantasy is a process of internalization
wherein the object’s method of handling the projective identifi cation is perceived and there is an
effort to make this aspect of the object a part of oneself. Following the schema outlined by
Schafer (1968), introjection and identifi cation are seen as types of internalization processes.
Depending upon the projector’s maturational level, the type of internalization process he
employs may range from primitive introjection to mature types of identifi cation. In introjection,
the internalized aspect of the object is poorly integrated into the remainder of the personality
system and is experienced as a foreign element (‘a presence’) inside oneself. In identifi cation,
there is a modifi cation of motives, behaviour patterns, and self-representations in such a way
that the individual feels that he has become ‘like’ or ‘the same as’ the object with regard to a
given aspect of that person. So the terms introjection and identifi cation refer to types of
internalization processes that can operate largely in isolation from projective processes or as a
phase of projective identifi cation.

To expand briefl y upon what has been said earlier, the concept of externalization (as discussed
by Brodey, 1965) would be used narrowly to refer to a specifi c type of projective identifi cation
wherein there is a manipulation of reality in the service of pressuring the object to comply with
the projective fantasy. However, in a broader sense, there is ‘externalization’ in every projective
identifi cation in that one’s projective fantasy is moved from the internal arena of psychological
representations, thoughts and feelings, to the external arena of other human beings and one’s
interactions with them.

Rather than simply altering the psychological representations of an 297

Thomas H. Ogden

external object, in projective identifi cation one attempts to, and often succeeds in, effecting
specifi c alterations in the feeling state and self-representations of another person.
5. Finally, I would like to attempt briefl y to locate projective identifi cation in relation to
projective transference, counter-transference and projective counter-identifi cation. Transference
involves the attribution to the therapist of qualities, feelings and ideas that originated in relation
to an earlier object. Transference projection is a type of transference wherein aspects of the self
are attributed to the therapist. When projective identifi cation is an aspect of the transference
relationship, it would be differentiated from transference pro -

jection in that a transference projection is largely an intrapsychic defensive phenomenon. In

contrast, projective identifi cation not only involves an intrapsychic event (a projective fantasy)
but also involves an interpersonal interaction in which the object is pressured to become the way
he or she is represented in the projection. Also, as with other forms of projection, the term
projective transference would imply a greater weight of disavowal of an aspect of the self than is
involved in projective identifi cation, and would entail less of the feeling of being at one with the
object than is encountered in projective identifi cation.

Counter-transference has been defi ned in a number of different ways. It has been viewed by
some as the set of feelings of the therapist elicited by the patient which refl ect the therapist’s
unanalysed pathology. Such feelings interfere with his ability to respond therapeutically to his
patient. Others have viewed counter-transference as the totality of the response of the therapist to
the patient. Still others refer to that portion of the counter-transference that represents the
therapist’s mature, empathic response to the patient’s transference, as the ‘objective
countertransference’ (Winnicott, 1949). This aspect of the therapist’s response to the patient is
viewed as the complement to the aspect of the earlier relationship portrayed by the patient in the
transference. The remainder of

the countertransference would then be seen as a refl ection of the therapist’s pathology. I fi nd
Winnicott’s view to be the most useful in clarifying the role of a therapist’s feelings in the
successful handling of a patient’s projective identifi cations. As an object of the patient’s
projective identifi cations, it is the task of the therapist both to experience and process the
feelings involved in the projection. The therapist allows himself to participate to an extent in an
object relationship that the patient has constructed on the basis of 298

On projective identifi cation

an earlier relationship. In so doing, the therapist has the opportunity to observe the qualities of
the previously internalized object relationship and, over time, process the feelings involved in
such a way that the patient is not merely repeating an old relationship in the therapy.

In Winnicott’s terminology, this aspect of the therapist’s work would represent the observation
of and therapeutic use of the objective counter-transference. A failure on the part of the therapist
in his handling of the patient’s projective identifi cations is often a refl ection of the fact that
instead of his therapeutically making use of the objective counter-transference data, he is
involved in what Grinberg (1962) calls ‘projective counter-identifi cation’.

In this latter form of countertransference, the therapist, without consciously being aware of it,
fully experiences himself as he is portrayed in the patient’s projective identifi cation. He feels
unable to prevent himself from being what the patient unconsciously wants him to be. This
would differ from therapeutically ‘being open to’ a patient’s projective identifi cation, because in
the latter case, the therapist is aware of the process and only partially, and with diminished
intensity, shares in the feelings that the patient is unconsciously asking him to experience. The
successful handling of projective identifi cation is a matter of balance – the therapist must be
suffi ciently open to receive the patient’s projective identifi cation, and yet maintain suffi cient
psychological distance from the process to allow for effective analysis of the therapeutic
interaction.

Summary

This paper presents a clarifi cation of the concept of projective identifi cation through a
delineation of the relation of fantasy to object relations that is entailed in this psychological-
interpersonal process.

Projective identifi cation is viewed as a group of fantasies and accompanying object relations
involving three phases which together make up a single psychological unit. In the initial phase,
the projector fantasies ridding himself of an aspect of himself and putting that aspect into another
person in a controlling way. Secondly, via the interpersonal interaction, the projector exerts
pressure on the recipient of the projection to experience feelings that are congruent with the
projection. Finally, the recipient psychologically processes the projection and makes a modifi ed
version of it available for re-internalization by the projector.

299

Thomas H. Ogden

Projective identifi cation, as formulated here, is a process that serves as: (1) A type of defence by
which one can distance oneself from an unwanted or internally endangered part of the self, while
in fantasy keeping that aspect of oneself ‘alive’ in another. (2) A mode of communication by
which one makes oneself understood by exerting pressure on another person to experience a set
of feelings similar to one’s own. (3) A type of object relatedness in which the projector
experiences the recipient of the projection as separate enough to serve as a receptacle for parts of
the self, but suffi ciently un-differentiated to maintain the illusion that one is literally sharing a
given feeling with another person. (4) A pathway for psychological change by which feelings
similar to those with which one is struggling are processed by another person, following which
the projector may identify with the recipient’s handling of the engendered feelings.

Projection and projective identifi cation are viewed as representing two poles of a continuum of
types of fantasies of expulsion of aspects of the self with the former being seen as predominantly
a one-person phenomenon involving a shift in self- and object-representations; in contrast, the
latter requires that one’s projective fantasies impinge upon real external objects in a sequence of
externalization and internalization.

300

18

Vicissitudes of projective identification

Albert Mason

In 1946 Melanie Klein published one of her most important works entitled ‘Notes on some
schizoid mechanisms.’ In this paper Klein described the psychic processes occurring in the fi rst
three months of life; she delineated the characteristics of the early ego as well as the form and
nature of its object relations and anxieties. Klein described schizoid states including idealization,
ego disintegration, and projective processes connected with splitting for which she introduced
the term ‘projective identifi cation’. For Klein, splitting was the key concept of this stage of
development and state of mind.

Klein (1921) had also examined splitting processes in her fi rst published paper and continued
her investigation of splitting throughout the 1930s, from the pathological to the more normal
forms. She suggested that the ego’s fi rst defence against anxiety was not repression, wherein the
anxiety, although unconscious, remained attached to the self, but expulsion, a violent form of
splitting that relieved the ego of pain, distress, and anxiety. Since Klein believed that the
phantasy of expulsion was also object related, any split-off part of the self would go into the
object.

Just as introjection, like feeding, is always from an object, so projection, like evacuation, is
always onto or into an object. Feeding (or introjection) can be gentle, loving, voracious, greedy,
biting, tearing or violent; and projection can be invasive, fragmenting, erotic, possessive, or a
host of other qualities which depend upon the particular phantasies of the projector at any given
moment. It is precisely the same with projective identifi cation, which is a combination of
expulsion and acquisition.

301

Albert Mason

To continue the analogy between projective identifi cation and the gastrointestinal tract, one
could describe two methods of feeding or acquiring what one needs. One method entails feeding
from an object repeatedly, digesting the food (that is, working internally and thinking), and
assessing what one retains so that one may eventually become like the nourishing object if one
chooses to do so. In this system the anxieties attendant on the awareness of the separateness of
self and object, or mouth and breast, are essentially tolerated. This is in contrast to devouring the
nourisher (breast) whole in order to become the nourishing object in as little time and with as
little work as possible, thereby eschewing the painful awareness of separateness altogether.
Acquiring and becoming the envied object or breast by projective identifi cation and acquiring
and becoming it by greedy devouring are two omnipotent phantasies producing many complex
consequences.

A second main stream of Klein’s ideas took shape in a paper entitled ‘Early stages of the
Oedipus confl ict (Klein, 1928) and was concerned with phantasies of intrusion into the mother’s
body. This intrusion had many motives. The motives could stem from any affective state
including love, hate, sadness, sadism, possessiveness, or the desire for fusion. Klein saw the
infant as having an internal world (itself ) and an external world that was equated with the
mother’s body. The boundaries between these two worlds were frequently missing or porous
because of the continual processes of introjection and projection which often blurred the
distinction between self and object.

These earlier ideas all formed part of the broader concept of projective identifi cation, which is
an overall term for a number of distinct yet related processes connected to splitting and
projection. Klein believed that projective identifi cation was the leading defence against anxiety
in the paranoid schizoid position, and that it constructed the narcissistic object relations
characteristic of this period, in which objects became equated with the split-off and projected
parts of the self. Simultaneous with the projection was an introjection and an identifi cation for
the purpose of acquiring the desired and envied qualities and capacities of the object. Klein also
described how the ego became impoverished due to the excessive use of projective identifi
cation. For example, weak functioning of the ego would occur as a consequence of projecting
away hostile, aggressive parts of oneself, thereby losing the strength that these qualities could
impart to the ego were they retained.

302

Vicissitudes of projective identifi cation Projective identifi cation has been confused with
projection from the outset, and both terms have been used in overlapping ways historically to
describe phenomena that are not completely distinguished. Freud described projection as one
person’s ideas being attributed to someone else, creating a state of paranoia. His concept of
projection seemed to be largely limited to the projection of impulses and feelings into the object.
Abraham seemed much more concerned with projection into the external world of an internal
object. Klein developed these ideas and elaborated them in terms of projecting split-off parts of
the ego. She would postulate, for example, that it was not simply anger that was being projected
into an object, but, say, an angry baby part of the personality that was jealous of a mother’s
attention to a sibling.

An example of projective identifi cation occurred before a holiday break in a patient who
historically had had temper tantrums at the birth of his baby sister and subsequently could not be
separated from his bottle. He had the following dream on his sister’s birthday thirty years later:
his mother and father were searching for his sister who was lost.

When they returned, his mother had a white moustache which the patient thought was due to her
having snorted cocaine.

His father was driving his car out of control, and the patient felt he now had to take care of them
both. When he tried to leave them, his parents pleaded with him to stay and became very angry
when he could not. We can see in this dream that the parents who were exhibiting loss of control,
jealousy, and anger were containing the projections of the baby part of the patient who felt both
angry and jealous when his sister was born. The patient’s early addiction to milk (his inability to
part from his bottle) was now projected into his mother who had a white moustache, and the
patient had become both the caretaking parent as well as the parent whose attention was
elsewhere, producing jealousy in the left-out child. In the transference I later became the left-out
parent containing the patient’s left-out child part of himself when he pre-empted my vacation
with a break of his own.
Klein also pointed out that parts of the ego which are projected are often connected to an internal
object. Since both the ego and its internal objects are constructed from mixtures and integrations
of the self and the external world, the whole process is extremely complex.

What is important to understand is that Klein added depth and complexity to Freud’s concept of
projection by emphasizing that one 303

Albert Mason

cannot project impulses without simultaneously projecting parts of the ego that are associated to
these impulses.

Hence, a split in the ego occurs. The aspect or aspects of the self that are projected go into the
object and consequently distort the appearance of the object. In addition, the perception of the
self is altered due to the impoverishment which occurs as a result of splitting and projecting parts
of the self.

This projection of parts of the self often produces a deep sense of connection and responsibility
on the part of the projector towards the recipient of the projection – whether this be mother and
child or child and mother – and can occur in therapists who may compulsively look after very
disturbed patients, or very poor people who may unconsciously represent hated or disowned
aspects of the therapist him/herself. Such compulsive care of a particular population may also be
connected to the exculpation of unconscious guilt, as though the therapist were somehow
responsible for their plight. A dentist, for example, removed his patients’ teeth at the slightest
provocation, claiming this to be a necessary operation in each case. The dentist, in fact, had had
his own teeth removed at the age of thirty-six. He had frequent dreams of being pursued by packs
of wolves –

animals with prominent teeth. As might be anticipated, the dentist’s own appetites, both oral and
genital, were voracious and felt to be dangerously destructive to his objects. The teeth extraction
was both a punishment for and a prevention of his destructive phantasies.

One can see that the differences between projective identifi cation and the simpler mechanism of
projection are due to the consequences of the immensely varied phantasies connected to
projective identifi -

cation. These may include paranoia, a result of the projection into others of threatening parts of
the self.

Claustrophobia can be the result of identifi cation with the part of the self that is felt to be lodged
inside the object.

This consequence contrasts with identifi cation with the object that contains the projected part of
the self. The reintrojection of a hostile projection meant to damage the object can cause
hypochondriasis when the introject is located somatically, whereas the reintrojection of a
damaged object into the mind can be experienced as a damaged or dead internal object, which is
a frequent cause of depression. Internalized objects, containing projections, become complex
superegos that can be experienced as internal persecutors capable of depriving one of all pleasure
and good experiences.
An idealized object, when introjected and identifi ed with, produces manic states of mind or
obsessional perfectionism, in contrast to the 304

Vicissitudes of projective identifi cation depressive consequence of the introjection of the

devalued or damaged object.

Hostile projections due to envy and jealousy damage and distort the child’s view of the parents’
pleasure or creativity, and psycho-sexual disorders can frequently be due to an identifi cation
with these damaged views of parental genitality. Money-Kyrle (1968) has described how the
patient produces a misconception of parental intercourse by his/her projections, concluding that
‘every conceivable representation of it (parental intercourse) seems to proliferate in the
unconscious except the right one’ (p. 417). Attacks on parental intercourse can be directed at any
form of intercourse including the parents’ speech or their emotional involvement with each other
as well as at their genital activity. These phantasies commonly have the aim of splitting the
parents away from each other, thus diminishing the envy and jealousy that the child feels as a
consequence of recognizing that he/she lacks the capacity to take part in the parents’

intercourse at any level. Damaged or separated parents, when introjected and identifi ed with, can
produce any disorder of articulation including sexual disorders, speech disorders, emotional
disorders, and/or thought disorders.

Thinking itself, or the process of putting thoughts together, may be equated to the parents’
intercourse. This fi nding of Klein was later amplifi ed by Bion (1959) in his paper entitled
‘Attacks on Linking’. The ideas set forth in Bion’s paper have proven invaluable in the
understanding and treatment of thought disorders.

A patient with a thought disorder had a dream of connecting an electric wire with a water-pipe to
form a ‘jagger’.

When asked what a jagger was, the patient responded that he didn’t know. So in phantasy he was
connecting two things that were incompatible to make something which was unknown and
probably dangerous. These symbols stood for the patient’s envious and jealous devaluation of his
parents’ intercourse, and the consequence of identifi cation with the devalued parental introjects
was the destruction of his own capacity to think and create. Flattening and shallowness of affect
is a consequence of the identifi cation with parental intercourse that has been enviously drained
of all pleasure and enjoyment. It is also created when projective identifi cation produces
phantasies of fusion with an object, as affect depends to a large degree upon separateness.

Perception, phantasy, and thinking develop in the gap between self and object and are a result of
separateness.

Consequently, the more 305

Albert Mason

massive the projective identifi cation, the more the individual’s capacity for perception,
phantasy, and thinking will be impaired.
The projection of the aggressive parts of one’s personality linked to hating, biting, excreting,
mocking, or devaluing produce various versions of paranoia so that the paranoid object will be
seen to be either hating, biting, mocking, or devaluing oneself, depending upon which part of the
self is projected. Projections into animals or insects will result in typical animal or insect
phobias. Similarly, feelings of violence fragmented and projected into the world around one can
produce fear of every earthquake tremor, every shadow, every virus, every sneeze – all
depending upon the quality and phantasy of the actual projection.

A patient literally saw his father’s stern face (an internal object) in the middle of a mountain
when he was leaving town. Frightened, he turned his car around and headed for the safety of his
home, which was, in phantasy, his mother’s body. There seems to be no limit to the persecutory
phantasies produced by projecting into the outside world. For example, one can project one’s
demanding nature into work, traffi c, one’s wife, or one’s children. One then experiences a
persecutory pressure coming from these various situations or people which appear to pursue one
with relentless expectations, turning life into a series of burdens rather than satisfying effort.

Part-objects can also become the subject of projective identifi cation, wherein a part of the self (a
part of the mind or the body) is projected and confused with a part of the object. An easily
recognizable version of this is the thumb sucking or tongue thrusting of the infant and the adult
versions of this such as cigarette or pipe smoking, or nail biting. Freud felt that ‘the thumb-
sucking child was involved in auto-erotic stimulation, gratifying or soothing itself through
sensuous pleasure to avoid a painful state of mind’. Kleinians would see the thumb sucking as
part-object projective identifi cation in which the infant in phantasy takes into its thumb the
nipple of the mother or the penis of

the father depending upon the source of the gratifi cation needed. This defence against
dependency, and the anger or anxiety experienced because of the absence of the nipple would be
dealt with in Klein’s view by the child’s phantasised possession of a nipple with which it can
feed itself. Such an example of part-object projective identifi cation was given by Susan Isaacs
(1952) in the case of a young boy who watched his mother breast-feed her new baby.

He pointed to her breast and exclaimed, ‘That’s what you bit me 306

Vicissitudes of projective identifi cation with’ (p. 88). The little boy was clearly projecting his
hostile, greedy, biting mouth into the breast which had now become a biting persecutor, rather
than an envied or jealousy-producing source of nourishment.

A patient during analysis was revealing envy of me as a paternal object. Her material also led me
to interpret that she was particularly envious of what she felt to be the magical quality of the
erect penis

– how it drew attention to itself, how it excited others who observed it (both herself and other
women), and how it was felt to be superior to her own genital, which was, in her eyes, just a little
hole. When I pointed this out to her and linked her feelings of inadequacy to her idealized views
of the penis, she came to her next session wearing a pointed hat, standing erect on high heels,
and carrying a string bag containing two cabbages that she had bought on the way. She wore a
tight-fi tting, shiny yellow slicker which matched the hat. The patient felt this to be an eye-
catching and exciting outfi t, having observed many women on the subway watching her. It
seemed clear that she felt herself to be a penis, stimulating admiration and excitement, originally
in Mommy. The interpretation of her phantasised erect penile state produced a rather rapid
collapse of this patient on to the couch.

In his paper ‘A psychoanalyst looks at a hypnotist: a study of folie à deux ’ (1994), Mason has
described a patient who projected both loving and hating parts of herself into her objects. Her
loving aspects were meant to repair a damaged sister, and her hating aspects were aimed at
stealing and possessing envied qualities of her objects. Eve had persistent fears of being raped,
which were explicated using dream and transference material. She felt and feared she could not
resist rape because she herself wished and, in fact, believed that she could magically invade her
objects –

her sister, mother, and the analyst in the transference. Being raped was the persecutory mirror
image of her own invasive phantasies. Her fear of rape was also experienced as a terror of many
situations including visits to her doctor or dentist. The patient dreamt that she was in a room with
seven other women. A man came in and raped everyone, one after the other, and there was
nothing anyone could do. Her association to the dream was that the rapist had hypnotized
everyone with his baby-blue eyes (which were, by association, the same colour as the patient’s).
This patient had come for treatment because she awakened one morning with her legs paralysed
– a condition which lasted several weeks. The 307

Albert Mason

paralysis had developed the morning after she had had her fi rst sexual experience. She
remembered how wonderful this experience had felt, and at the same time she had had the
thought that her sister would never ever know what it felt like to be made love to. This sister,
who was four years older than Eve, had been institutionalized since the age of four with spastic
diplegic paralysis. Following Eve’s sexual encounter, she had a dream about two fi r trees
standing side by side. One tree was alive and the other was dead. The dead tree sprouted a few
leaves, and the patient said that she now knew that the live tree would die.

Eve had been extremely jealous of her older sister who was often held up to her by her parents as
a saint. She felt that her parents loved this sister more than herself. It seemed that Eve’s
unconscious guilt following her sexual triumph over the sister led to her identifi cation with the
sister’s paralysis. This identifi cation with a damaged object seemed to be a defence against guilt,
as suffering the same damage as the object was felt to exculpate the crime. In addition, the dream
of the fi r tree suggested that the identifi cation was an unconscious attempt to omnipotently
repair the damaged sister by the sacrifi ce of her own life. In her mind and in her body through
her symptom of paralysis, the repair had been achieved by the projection of Eve’s healthy legs
and life into the ‘dead sister tree legs’.

One could also infer that this patient’s fear of rape was a projection into men of her own
omnipotent, intrusive

powers that she would be helpless to resist. The baby blue eyes of the rapist in her dream were
her own eyes which she felt had the power to invade, possess and control whomever she looked
at. This patient had many conscious voyeuristic fantasies of ‘knowing all about people’; the
knowing – like the Biblical knowing – implying taking possession of.

She also displayed ongoing wishes and attempts to get into my private life through her questions
and fantasies.

Eve’s projected omnipotence and consequent feelings of helplessness toward the imagined
power of the rapist were potent factors in the production of her terror. The analysis of her
intrusive omnipotent phantasies gradually led to the disappearance of her phobias and to her
marriage several years later.

A striking example of part-object projective identifi cation was demonstrated by a forty-six-year-

old lawyer who sought treatment for depression and diffi culty in her relationships with men.
Janet had had several relationships which lasted a few years, after which time she would fi nd
something wrong with her partners, lose interest in 308

Vicissitudes of projective identifi cation them sexually, and end the relationship. She had been
engaged twice but never married. Janet had been hospitalized at the age of two for polio and had
been isolated from her parents for four months. During this period she could only see them
through her hospital window and speak to them by telephone.

One day, through the window, her parents showed her their new baby. After that visit, the patient
smeared faeces continuously during her hospitalization.

This patient masturbated frequently, often several times before coming to her session. An
unusual feature of the masturbation was that she came to orgasm by contracting her levator ani
muscles on a hard stool held in her rectum.

Janet had been doing this as long as she could remember; she particularly recalled masturbating
in this way in elementary school before a class bell would ring. She would try to achieve orgasm
before she had to change rooms for a new class.

I believe that the masturbation was a defence against depression and jealousy; the stool with
which she masturbated sometimes stood for Daddy’s penis and at times became Mommy’s
nipple, swallowed and possessed. This part-object projective identifi cation was also meant to
project her jealousy and left-out feelings into her objects while she enacted the phantasy of
sexual intercourse in her anus. The masturbating activities became more frequent prior to breaks
in treatment and, like the masturbation in class, the excitation took her mind away from loss and
depression.

The trapped and idealized stool was sometimes equated with God who was always with Janet,
unlike the breast that at times went away, fed her baby brother, and was unavailable to soothe her
persecutory anxieties.

When Freud said of the Wolf Man, ‘There is a wish to be back in a situation in which one was in
the mother’s genitals; and in this connection the man is identifying himself with his own penis
and is using it to represent himself
’ (Freud, 1918, p. 102), he was anticipating the fi ndings of Klein and her followers in the
understanding of a vast number of somatic and hypochondriacal disorders. A man or woman can
identify not only with his penis or her vagina, but also with his/her stools, urine, fl atus, breath –
in fact with any part of the body or body products and/or the body of the object. Meltzer wrote
about a patient who ‘functioned in a state of projective identifi cation with a father’s penis, alive
but detached as a part-object from the rest of the dead father’ (Meltzer, 1968, p. 70).

The use of projective identifi cation for the purpose of elevating one’s status can also be
achieved by identifying with one’s car 309

Albert Mason

(a Bentley), one’s football team (when it wins), one’s child (my son the doctor), one’s diamond
ring, one’s fur coat (sable, my dear), or even, like Charles De Gaulle, one’s country: ‘La France,
c’est moi!’

These examples demonstrate one’s phantasy of acquiring the power, prestige, or the ‘goodness’
of the selected thing, and that, in a similar fashion, one can also project one’s own badness into
hated objects or things, which is the

principle of scapegoating. In his paper

‘Megalomania’, Money-Kyrle has written about projecting into clothing for the purpose of
acquiring the emperor’s power and prestige (Money-Kyrle, 1965). Mason has described
projection into parts of one’s own body in his paper

‘The suffocating superego: psychotic break and claustrophobia’ (Mason, 1981).

Angela, a woman of twenty-three, was referred for treatment because her parents could not stand
her keeping glasses of her urine in their refrigerator which she would drink daily. The patient
said that drinking urine was healthy –

since this was a tradition practiced in India – and that she could not understand her parents’
concern.

Angela’s biological parents had divorced when she was eleven. After a brief period of living
with her mother, she and her younger sister had moved in with her father and step-mother, who
had had four more daughters in the next eight years. Angela described her step-mother as looking
like a witch and said that she hated the woman.

But she herself felt no need for therapy and came only because her father and step-mother
insisted.

On the surface, she did not seem worried about the fact that she had had only one menstrual
period at age sixteen and none since.

Nor was she upset that she had no breast development and needed continuous enemas and
suppositories for normal defecation. Her abdomen protruded like a six-month pregnancy. The
patient had never masturbated nor had she had sexual intercourse. She thought that sex was
‘dirty’. When asked if her lack of periods concerned her, she responded by recounting a dream
she had had the night before:

‘I lived in Communist Russia and was married to a monster who sewed up a pregnant woman’s
vagina.’ She had no idea what the dream might mean.

I interpreted that living in Communist Russia refl ected a state of mind in which there were no
aristocrats or peasants

– all people were equal. It was as though she wanted babies to be equal to mommies.

This state of mind also hated her grown-up step-mother’s breasts and pregnant belly and sewed
up her vagina to kill any babies and to 310

Vicissitudes of projective identifi cation prevent intercourse with her father. She took this sewn-
up mother into herself and so she became sewn up – with no periods, no sex, and no baby. She
had no doubt lost her own breasts in a similar way, identifying with a step-mother whose breasts
were enviously destroyed. Drinking her own urine was Angela’s way of showing she could
produce food and feed herself just as well as her step-mother could produce milk and feed
babies. Her shit-fi lled abdomen was like the dead baby she felt she had caused in the step-
mother by her monstrous, envious wishes. Upon hearing all this Angela changed her mind and
decided to come for two sessions a week which she kept for two years, lying on the couch the
opposite way round. Her periods began after three months as did her breast development, and her
bowel function eventually became normal. She married a year later and now has two children
and a thriving business. Angela’s rather rapid improvement was unusual and somewhat
inexplicable.

Perhaps it was related to her youth and perhaps to the fact that the major trauma in this patient’s
life – the envy and jealousy of the step-mother and her babies – developed when she was a child
of eleven rather than an infant. But whatever the reasons, Angela has maintained her therapeutic
gains to the present time.

William was a young man of twenty-fi ve who came to treatment complaining of severe and
persistent depression and an inability to work and concentrate. He was lonely and shy, and
reported an incapacity to make friends and a strong feeling of being stared at and laughed at in
the streets. He also suffered from claustrophobia, the analysis of which occupied the fi rst stage
of his treatment. Following the initial period of claustrophobia and its resolution, there was a
period in which material emerged slowly – sometimes with great diffi culty or pain – which I
equated with constipation. The patient revealed that he was indeed literally constipated and had
been severely so all his life.

He remembered and recounted having written an essay at school in which he advocated

constipation as a solution for the world’s starving people. He had thought that if the people were
constipated and therefore had their bellies full, they would not feel hunger.

A long portion of the analysis dealt with William’s confused phantasies concerning the function
of his anus and the nature of his own faeces, which were commonly equated with food. He used
faecal equivalents in place of food to fi ll himself up, turning frequently to the use of alcohol,
barbiturates, cigarettes, snuff, and junk food which 311

Albert Mason

he would eat in enormous quantities. Eating these ‘bad foods’ was his phantasized effort to deny
and replace his need for an external living object. Concurrent with his idealization of faeces were
scathing and contemptuous attacks on breasts – actual breasts as well as their derivatives and
function. Interpretations or food for thought were treated equally badly except for past
interpretations which became the equivalent of faeces. The patient would ruminate about them,
play around with them, and elaborate on them. He would make notes on his sessions, or secretly
tape the sessions, and re-read the notes or replay the tapes continuously, particularly at times
when he felt ill, as he believed they would help him. It was only after much painful working
through that he began to accept that the interpretations of yesterday were only meaningful in a
particular context, and might have no relevance to today’s analysis.

William’s phantasies about stools were accompanied by all sorts of manipulations of his anus.
Suppositories were used, there was frequent and prolonged cleansing with water and ointments,
and also bouts of pruritis and piles. I felt that these activities were thinly disguised masturbation
seen (together with his phantasies about his stools) as a support of a projective identifi cation
process with his mother’s body, particularly her breast. The patient dreamed that he mocked a
female statue and bit off pieces of it, which he swallowed. He then allowed his penis to be
sucked by little boys who were trussed up like chickens (William was swaddled as an infant). By
swallowing the breast, he became the breast and could identify with this envied part-object by
projective phantasies or introjective cannibalistic phantasies.

Another dream at this time illustrated a similar dynamic. In reality there existed in the patient’s
country of origin a tribe who believed that their dead must be buried in a certain district to
achieve paradise.

Often these people had to save for half their lives to acquire enough money for this purpose and
until that time, the bodies were entrusted to a certain group for safe-keeping. In the dream the
people who were keeping the bodies turned out to be gangsters who buried the bodies without
correct preservation. When an effort was made to recover them, it was discovered that the bodies
had disintegrated and been washed out to sea by the rains.

This dream could be interpreted as his own waste of life and energy by idealizing and
worshipping dead and useless stools; that is to say, a material remnant and representation of
something like the breast, once alive and loved. This dream and the extensive working through
312

Vicissitudes of projective identifi cation of its meaning marked a turning point in William’s
analysis as he gradually relinquished his need to record and concretely retain specifi c
interpretations.

Following the analysis of his anal masturbation and idealization of faeces, there emerged a
severe and persistent delusion that stemmed imperceptibly from the preceding material. At fi rst
it made itself felt as a constant consciousness of his anus. This consciousness grew in intensity
until it became what he described as a tormenting mental irritation. Finally, the depression and
persecution produced by his symptoms became so acute that William almost broke down. He had
frequent suicidal thoughts during this episode and was sure that he was going mad. The fi nal
form that the delusion took was William’s conviction that his anus was a vagina. He became
constantly preoccupied with fears, wishes, and phantasies of being penetrated and felt that
penises were jumping at him and into him in the street. Frequently he felt that my penis was
being taken out and was going to jump into him.

Sometimes he even felt that I looked like a penis and that everything I said went into him and
produced excitation or pain. An aspect of the analysis demonstrated that his delusion was a part-
object state of projective identifi cation with his mother’s body to escape the pains of being the
dependent child. He also had many complex phantasies related to his father and his envy of his
mother’s ‘power and control’ over the father. In contrast, most of his

conscious fantasies were heterosexual.

It should be noted that in addition to the phenomenon of identifying with an idealized part-
object, one may also identify with a devalued part-object. An example of this phenomenon
occurs in the anorectic girl who feels her body to be ugly if she gains a pound or two. Her
envious hatred of the beauty and fecundity of her mother’s breasts and body causes damaging
attacks on them which then become a source of self-hatred when these damaged maternal
qualities are introjected and identifi ed with.

It seems clear that projective identifi cation exists along the whole continuum of human
development – that it is both healthy and, indeed, necessary as well as pathological – and can be
used as a defence against all anxieties. ‘I am the object’ defends against loss of the object either
by separation (separation anxiety), to another person ( jealousy), or through envy, as one now
possesses the object’s envied qualities or capacities. It must also follow if one accepts Klein’s
model of development from paranoid-schizoid states (unintegrated) to 313

Albert Mason

depressive states (separate, realistic, integrated), that projective identifi cation will occur with
both integrated whole objects and unintegrated part-objects. The transsexual who in phantasy
‘becomes’ a woman or the boy who

‘becomes’ his father (whole object) is quite different from the baby’s sucked thumb which
becomes, in phantasy, a nipple (part-object). Perhaps identifi cation with a whole object is most
often a depressive defence against loss of the object, while identifi cation with a part-object is
most often a defence against paranoid anxieties. Klein frequently stated that the breast is a source
of protection against persecutors as well as a source of nourishment.

A particular consequence of projective identifi cation occurs when the identifi cation with the
envied object is a mixture of the feelings of love and hate experienced simultaneously. Love
stimulates a faithfully preserved or even an idealized identifi cation, and hate produces a
parodied or debased identifi cation. For example, a fi fty-two-year-old patient who described
himself as a ‘fl aming fag’ seems to have identifi ed with his beautiful baby sister (three years his
junior) whom he believed was his father’s favourite child. His envy and jealousy of her produced
an admiring identifi cation but also a devalued identifi -

cation as the envy stirred up both admiring and hating (mocking) feelings. The resultant identifi
cation was a parodied version of femininity that caused the patient’s ‘swishy’ behaviour to be
laughed at and mocked rather than admired.

One important characteristic of projective identifi cation is its unconscious use as a method of
communication. This property of projective identifi cation is a signifi cant factor in the
production of countertransference responses during analysis. This phenomenon was originally
described by Heimann, Racker, Money-Kyrle, and Bion and has been elaborated by many
analysts since. Projective identifi cation is an important contributor to non-verbal
communication, a property which is particularly important for non-verbal infants. I believe when
we ‘intuit’ some aspect of our patient’s psychic state we are responding to the patient’s
projective identifi cation. In fact, intuition could be called a countertransference response. An
example of this function was told to me by a supervisee, whose patient was a borderline female.
When this analyst opened her waiting room door, the patient was standing with her ear pressed
against the door which separated the waiting room from the consulting room. The patient entered
the consulting room muttering quietly and intensely to herself. The analyst responded by asking,
‘What did you say?’

314

Vicissitudes of projective identifi cation Here we see how the patient successfully projected into
the analyst her frustrated curiosity at detecting a couple in the adjoining room whom she had not
understood or heard clearly. This enactment was probably a replay of the young patient’s
exclusion from her parents’

bedroom and the projection of this frustrated curiosity into the analyst. It was also a repetition of
the many secret, forbidden activities that the patient had indulged in throughout her childhood
and had phantasized would make her parents feel left out and envious.

Following his treatment of some one hundred and fi fty patients by hypnosis and extensive
research into the

outcomes of other hypnosis practitioners, Mason (1994) suggested that: the most dramatic and
powerful effects of projective identifi cation are produced when it exists in its most primitive,
i.e., magical or omnipotent, form in a patient, and when the patient meets a therapist in whom
this phantasy is also powerfully present. Then ‘I wish to possess’ and the corollary and mirror
image of this wish, i.e., ‘I can be possessed,’ has found a practitioner who also wishes and
believes he or she can possess another human being. A

duet such as this is what I believe to be the basis of the hypnotic state.

This highly charged folie à deux takes many forms in medicine, religion, and politics, for
messiahs will always fi nd devoted disciples. I believe that some phantasies of invasive
possessiveness are universal and part of normal development, but when phantasies are extreme
and are colluded with by a parent with similar phantasies, then a folie à deux can result.
This phenomenon can be observed when the entitled ‘prince’ and

‘princess’ wishes of children are met by the parents’ own omnipotent phantasies about these
children. Meltzer wrote, ‘Where a child and an adult form a stable acting out collaboration, the
folie à deux, so refractory to analysis, arises’ (Meltzer, 1967, pp. 5–6).

In his paper ‘Quick Otto and slow leopard: The Freud-Fliess Relationship’ (Mason, 1997),
Mason described a dramatic folie à deux that developed in the years 1895–1905 between Freud
and his colleague Fliess. At this time Freud suffered from severe depressions as well as the
physical symptoms of migraine, nasal diffi culties, cardiac pain, arrhythmia, and dyspnoea. He
obtained relief from these symptoms with the use of cocaine, nicotine (twenty cigars daily), and
nasal cauterization performed by Fliess. These physical ‘cures’

315

Albert Mason

accompanied and were clearly supported by Freud’s manic episodes connected to his love of
Fliess and his belief in Fliess’s delusional theory concerning numbers (number twenty-three and
number twenty-eight stood for the male and female periods) which were seen as determining
human growth stages, the dates of illness, and each individual’s date of death. Freud believed his
own death would occur at the age of fi fty-one which was the sum of twenty-three and twenty-
eight. Fliess thought that his numbers theory applied not only to human beings but also to
animals and all living organisms.

Additionally, Fliess developed a ‘nasal refl ex neurosis’ theory which was meant to explain
almost all physical illness. Not only did Freud have Fliess operate on his own nose and on
several patients’ noses, but he referred to Fliess as ‘The Kepler of Biology’ and ‘The Messiah’.

Freud supported the irrational neurological connections made by Fliess despite his own
neurological training and a classical paper Freud (1898) wrote differentiating hysterical from
organic paralysis.

The view that a folie à deux existed between Freud and Fliess was stated originally by Strachey
(1951) who communicated to Jones about Freud’s letters to Fliess. In a letter to Jones, Strachey
stated, ‘It’s really a complete instance of folie à deux with Freud in the unexpected role of a
hysterical partner to a paranoiac.’ Mason also described a folie à deux between Freud and Fliess
concerning the denial of the forgotten gauze left in his patient’s nose post-operatively by Fliess.
Freud went to great lengths to exculpate Fliess’s serious oversight and later became

‘amnesic’ himself concerning the cause of the patient’s haemorrhages. Freud’s life-long
physician Max Schur observed that even after his break with Fliess ‘. . . Freud still needed to
cling to the fi ction that Fliess’s speculations about periodicity were valid, . . . he did not totally
overcome this conviction for decades, if indeed he ever did’

(Schur, 1972, p. 47).

The formation of the superego was fi rst described by Freud (1923) who called it the heir to the
Oedipus complex.

He theorized that a child between the age of three and fi ve, faced with the impossibility of his
Oedipal wishes –

because of his love for his parents and his fear of punishment – permanently incorporates and
installs these parental fi gures inside his/her mind. These internal parents now become internal
objects controlling the child’s dangerous and destructive impulses.

Klein (1932a) continued Freud’s investigation into the nature of the superego and the processes
of introjection and projection as well 316

Vicissitudes of projective identifi cation as her understanding of the nature of projective identifi
cation, which is, in effect, a combination of both these processes. It was Klein’s idea that many
pathological superegos (harsh, cruel, envious, perfec-tionist) were due not just to external reality
but to the child’s own primitive impulses projected into the external parents and then
reintrojected. The formation of the pathological superego also gave us a tool for the possible
modifi cation of this structure which is the source of so much distress and mental illness. The
tool was, of course, the observation of the mechanisms of projective identifi cation that occur in
the transference and, through this, the possibility of showing the patient his/her transference
distortions. The patient’s understanding of his projective processes through interpretation
hopefully enables him to reclaim what has been projected. This, in turn, gives the patient a truer
view of himself and his objects present and past and is a major aspect of the therapeutic value of
psychoanalysis. It is the understanding of projective identifi cation with the analyst and its
consequences now and with parents in the past that comprise a great deal of the therapeutic value
of psychoanalysis.

The formation of a pathological superego was illustrated by a thirty-six-year-old man who

sought treatment because of relationship diffi culties and obsessive procrastination. He dreamt
that he parked his car outside a market and went in to shop. He returned to fi nd a dent in his
fender and, upon inspecting the dent, noticed that it contained a piece of mirror. He could also
see his own eye refl ected in the mirror. This patient’s associations were that he had seen me
driving and that my car had large side-view mirrors which stuck out a long way (an accurate
observation).

He surmised that I had dented his car.

I could interpret his view of me as violently intruding into him and mirroring him back to
himself. However, my view of him contained a part of himself – his eye that looked back at him.
This patient was, in fact, a voyeur who spied on people to triumph over them and feel superior by
‘knowing their dirty secrets’. This was the me he took into himself; a me looking at him with his
own eye. The introjection of this view of me became a very persecuting internal object which
looked at him critically and harshly – noticing all his fl aws – and was primarily the cause of the
inhibition of all his activities. He feared being judged severely and mocked by cruel eyes.

As previously stated, the confl ict between advocates of the difference between projective
identifi cation and projection no longer 317
Albert Mason

generates much interest among Kleinians. There does, however, seem to be a present-day
disagreement between those who believe projective identifi cation to be a purely intrapsychic
phenomenon and those who regard it as an interpersonal phenomenon. Klein’s original view was
that projective identifi cation was an omnipotent phantasy –

that is, purely in the mind of the projector who is projecting into an internal object and changing
his perception of himself and the object. This can occur in the presence or the absence of the
object in question.

It is now also widely accepted that the phantasies of the projector can at times be conveyed in
some way to the object who is present.

The countertransference response of the analyst would be one way of detecting this. Another
noticeable effect of projection would be some kind of overt response from the recipient that
would be described an enactment. What part of any response is due to unanalysed targets in the
recipient of the projections is certainly food for much investigation and speculation. However, I
believe that a good case can be made for seeing our countertransference responses or

‘intuitions’ as benign or empathic, temporary folies à deux. Unlike the psychotic folie à deux, the
folie à deux described here is benign because the projection on the part of the patient is to
communicate and the projection on the

part of the analyst is to understand. The mechanism involved is mutual rather than one way. It
also appears that this dynamic is the essential basis of the phenomenon we call containment, in
which there is a shared desire for and attempt at understanding. It does seem clear that both
intrapsychic and interpersonal phenomena occur with the complex and fascinating mental
mechanism that Klein brought to our awareness, and that it will take much work and evidence to
fl esh out all the implications of this universal human phenomenon.

Summary

In this paper projective identifi cation has been described and contrasted with Freud’s view of
projection. Various motives for projective identifi cation and some of the consequences of these
motives are discussed. A dream illustrating projective identifi cation is recounted, followed by a
brief description of attacks on linkage by projective identifi cation and the consequences of
identifi cation with 318

Vicissitudes of projective identifi cation the damaged links. An example of the production of a
thought disorder following a patient’s phantasized attack on the parental intercourse is discussed.

Part-object projective identifi cation is demonstrated with examples of breast-mouth confusion

due to jealousy and the phantasy of the body as a part-object due to envious identifi cation.
Projective identifi cation as a method of communication is illustrated with a clinical vignette.
Various examples of projective identifi cation – both whole-object and part-object – are
elucidated, and the phenomenon of folie à deux due to mutual projective identifi cation is
discussed as related to the state of hypnosis and to the relationship of Freud and Fliess from 1895
to 1905. A link between intuition, countertransference, containment, and benign, temporary folie
à deux, is suggested.

Finally, there is a discussion of the role of projective identifi cation in the formation of the
superego and a mention of the present speculations about intersubjective and interpersonal
aspects of projective identifi cation.

319

SECTION4

Latin America

Introduction

Luiz Meyer

The papers that make up the South American contribution to this book are informative about the
way that psychoanalysis has developed and matured in the three countries which are represented:
Argentina, Chile and Brazil.

Argentina is the Latin American country in which psychoanalysis has gained the greatest
preeminence, ‘socialized’

and disseminated itself. It is also the one whose psychoanalysts had, and still have, the greatest
participation at an international level (although still less than they deserve). This is likely to
result from the association of a privileged economic setting, characterized by the presence of a
large middle class, with a high level of education among the population.

This has enabled Gustavo Jarast, author of the paper about Argentina, to choose an authorial
approach to study the concepts of projective identifi cation most used in that country.

His study discusses the contribution of four important authors: Heinrich Racker, León Grinberg,
and Willy and Madeleine Baranger.

It is worth emphasizing not only the cosmopolitan nature of this group but also the type of
training they had. Willy Baranger is of French origin, studied philosophy and it was as a
philosophy professor that he went to Buenos Aires.

Racker was born in Poland but gradu-ated in philosophy in Vienna, having migrated to
Argentina due to the war.

There was certainly a mutual fertilization between the receptive environment they encountered in
Buenos Aires and the European cultural experience/background they were able to offer.

Juan Francesco Jordan-Moore, author of the paper about Chile, centres his presentation on a
conceptual discussion: is projective 323

Luiz Meyer
identifi cation a solipsistic concept? What is the weight of intersubjectivity? He describes how
this question has gained relevance in the Chilean psychoanalytic

milieu , lists the psychoanalysts that have

discussed it, and presents the arguments they have used, supported either by ‘classical’ authors
(such as Rosenfeld and Bion), by the already mentioned Argentinean authors, or by
contemporary authors (such as Ogden), as well as Ignacio Matte-Blanco, the Chilean
psychoanalyst whose work is best known internationally. His presentation can be read as an
article that addresses questions relevant to the understanding of the functioning of the concept of
projective identifi cation.

Marina Massi, in order to write her article about Brazil, used a different methodology from the
previous authors.

Basing herself on publications about the history of psychoanalysis in Brazil and on bibliographic
research, she sent a questionnaire to various analysts in the country. Brazil is a country of
continental proportions. It is the only country in Latin America in which the language is
Portuguese, and not Spanish. This results in a certain amount of isolation when compared to the
intense psychoanalytic interchange that takes place among Spanish-speaking countries.

Furthermore, Brazil was not aided by the migration of intellectuals who already had undergone/

undertaken psychoanalytic training or that would be interested in it upon arrival in the country.
Massi’s article implicitly makes the point that, if in the case of Argentina we can refer to a
‘socialization’ of psychoanalysis, in the case of Brazil we should talk about its popularization:
today the country has twelve societies and eleven nuclei of psychoanalysis spread through its
entire territory. This implies a certain pulverization of theoretical concepts, refl ected in the
article.

Massi presents the work of various authors who have approached the subject of projective
identifi cation and tried to offer a personal interpretation. However, none achieved an authorial
dimension, as was the case in Argentina, neither did a burning question arise, as was the case in
Chile, capable of stimulating a fruitful debate regarding the subject of projective identifi cation.
There is, however, a rich and continuous contact with the most varied international schools of
psychoanalytic thinking. Numerous foreign analysts visit Brazil, establishing contact with its
members and their production. It is certainly a way to enrich, to make known, and to discuss the
different concepts of projective identifi cation.

324

19

Projective identification

The concept in Argentina

Gustavo Jarast
The Kleinian movement in Latin America was at its peak between the 1950s and the 1970s.
During this time Melanie Klein’s own contributions, as well as the contributions of the Kleinians
and post-Kleinians, were fruitfully taken up by the various psychoanalytical groups that were
gradually forming. Argentina’s Psychoanalytical Association (APA) was founded in 1942, and
articles by Klein, Susan Isaacs and Joan Riviere were published in the fi rst issues of its Revista
de Psicoanálisis.

In 1952 Enrique Pichon-Rivière published an important contribution on working with projective

identifi cations in the context of the psychoanalytic process in the Revue Française de
Psychanalyse .

Authors such as Heinrich Racker, León Grinberg, Willy and Madeleine Baranger, David
Liberman and José Bleger, among others, were able to see the richness of these new theories, and
how they shed light on other areas of clinical research. Of all the above, it was probably
Grinberg who developed the most explicitly Kleinian train of thought.

First Racker and then the Barangers also contributed to the theory of Kleinian psychoanalytical
technique, although in Racker’s case the links are not made explicit. The Barangers made a
creative contribution with their concept of the ‘psychoanalytic fi eld’. This concept actually
includes the mechanism of projective identifi cation in its confi guration.

In the years following the founding of the APA, Kleinian thought circulated in a familiar way
through the regular publication of papers 325

Gustavo Jarast

in which these ideas became common language. These papers laid the foundations for clinical
practice and the theoretical development of authors such as Arminda Aberastury (child
psychoanalysis), and of Enrique Pichon-Rivière (the treatment of psychotic patients).

These authors were considered pioneers in the fi eld. Even in groups who were not wedded to
Kleinian theory, theoretical and technical debate took place mainly through a central reference to
Kleinian thought.

In 1956 Liberman published ‘Identifi cación proyectiva y confl icto matrimonial’, in the Revista
de Psicoanálisis (Liberman, 1956).

In 1966, Emilio Rodrigué wrote ‘Relación entre descubrimiento e identifi cación proyectiva’, in
his book written with his wife Genevieve, El contexto del proceso analítico (Rodrigué, 1966).

In 1967 José Bleger published his very important book Simbiosis y ambigüedad , but he did not
draw attention to his use of the concept of projective identifi cation. He made a more explicit
contribution in

‘Psicoanálisis del encuadre psicoanalítico’, published in 1967 in the Revista de Psicoanálisis

(Bleger, 1967a, 1967b).

After the 1970s, debate grew more complex, taking on board recent French thinking, including
Lacanian theory. The contributions of ‘post-Kleinian’ authors such as Bion, Rosenfeld, and
Meltzer were also regarded as invaluable, and helped strengthen Kleinian thought. There is no
doubt that Klein’s concept of projective identifi cation has been one of the contributions which
laid the ground-work for the theoretical, clinical and technical development of psychoanalytical
practice. ‘Notes on some schizoid mechanisms’ was published in Spanish in 1947 in the Revista
de Psicoanálisis , and from that time onwards there was regular theoretical and clinical debate on
this topic.

Contributions, developments, problems

Heinrich Racker

Melanie Klein was never keen on extending the concept of projective identifi cation to apply to
the countertransference. Despite her disagreement, the idea of countertransference as a response
to the patient’s projective identifi cation was widely used from Paula Heimann onwards – in
Argentina, initially by Heinrich Racker. Racker 326

Projective identifi cation in Argentina has made a clear distinction between two sorts of
countertransference based on the two types of identifi cations suggested by Helene Deutsch.

He speaks of ‘concordant countertransference’, where the analyst identifi es parts of himself with
the corresponding psychological part in his patient (ego, id, superego), in an empathic tendency
to understand his own emotional

experience, together with that of the patient. The patient’s transference encourages the analyst’s
emotional psychological processes, in which the latter responds with a sublimated positive
countertransference. Racker wrote that

Concordant identifi cation is based on introjection and projection, or, in other words, on the
resonance of the exterior in the interior, on recognition of what belongs to another as one’s own
(‘this part of you is I’) and of the equation of what is one’s own with what belongs to another
(‘this part of me is you’).

(Racker, 1957, p. 312)

When the analyst identifi es himself with the patient’s internal objects, feels treated like them,
and experiences them as his own, he is experiencing a ‘complementary countertransference’, in
which the patient now represents an internal object belonging to the analyst.

This countertransference reaction implies that the analyst’s neurotic remnants are activated by
the projection of the patient’s objects.

Racker (1957, p. 312) said, ‘The complementary identifi cations are produced by the fact that the
patient treats the analyst as an internal (projected) object, and in consequence the analyst feels
treated as such; that is, he identifi es himself with this object’. In very few cases does Racker
explicitly use the phrase projective identifi cation to refer to this identifi catory process:

For instance, the analyst perceives the analysand’s intense rejection of his own libidinous
feelings towards the analyst (which may be caused by feelings of guilt, paranoid angst, rivalry,
masochism, an internal object’s boycott, etc.); he perceives the insistent annulment of his
interpretations, which might have enabled this rejection to be overcome, and he reacts with
anxiety, which communicates to his conscience as tension. But the perception of exterior danger
– of the analysand’s resistance – is just one of two factors, and countertransferential anxiety is
the result. Another factor is the analyst’s unconscious perception of interior danger, for example,
danger of 327

Gustavo Jarast

being frustrated by an internal object of his own, of being a victim of his own masochism or
counter-resistances.

Whatever the proportions between the subjective and the objective factors (that is to say,
between ‘the danger’ from the analyst or the analysand’s interior or between the death instinct of
either analyst or analysand), these factors cause the ‘tension’. If the analyst is aware of this
tension, it may serve as a sign enabling him to discover that part of the analysand’s internal
object, which is opposed to the libidinous relationship to the analyst. Violent irruptions of
countertransferential anxiety occur at times – as was already mentioned – as a consequence of
the analyst’s identifi cation with abruptly threatened internal objects of his own, or as a
consequence of his identifi cation with parts of the analysand’s ego, which have been intensely
dissociated by the analysand and projected on to the analyst. Often it is the analysand’s diffi
culty in tolerating excessive feelings of guilt which underlies these intense projections on to the
analyst – in this case, projection of a part of the ego experienced as guilty.

It has been observed on a number of occasions that the analyst feels compelled to return the
dissociated part to the analysand as soon as possible. This is because the analyst has trouble
tolerating the guilt deposited in him. This fact can serve to remind us of how much harder it is
likely to be for the analysand (whose ego is usually weaker than the analyst’s) to accept this
dissociated part as belonging to his ego. The anxiety the analyst has experienced once again
indicates what is going on within the analysand and what he is fi ghting against; the intensity of
this countertransferential anxiety can help the analyst to gauge the appropriate dose of
interpretation of this confl ict. I would like to add that the analysand’s defence mechanism
mentioned above (the ‘projective identifi cation’) usually achieves its goal. In our example: the
analyst feels guilty, and this not only implies (as it has sometimes been said) that ‘the analysand
expects the analyst to feel guilty’ or that the analysand ‘supposes that the analyst is sad and
depressed’.

The analyst’s identifi cation with the object with which the analysand identifi es him is – I repeat
– the normal countertransferential process.

(Racker, 1968, pp. 65–66)

For some authors this omission (of the phrase ‘projective identifi -

cation’) is due to a desire to use Freudian terminology, and to the fact 328

Projective identifi cation in Argentina that Racker’s main therapeutic ideal was ‘to remember’,
rather than a belief in

‘mechanisms’, implying essentially the idea of ‘reliving’, and that this explains his reluctance to
use the expression

‘projective identifi

cation’ explicitly (Lichtman, 1979). But the fact

is that he uses the concepts in a clearly ‘Kleinian’ way: a part of the patient’s ego that refuses a
‘good’ or a

‘persecutory’ content, and projects it on the analyst, so that the analyst becomes identifi ed with
it. The analyst must be unaware of having received the projected part, e.g. of the patient’s ego,
and for that reason he cannot interpret it.

For some authors, it was Klein’s seminal 1946 paper that allowed further studies of the
countertransference to fl ourish. In 1948 both Paula Heimann and Heinrich Racker, without
knowing each other, began their studies of countertransference, using the concept of projective
identifi cation as an instrument for understanding it, as well as for interpreting it. Racker
presented his fi rst work on the subject in APA in that year: ‘La neurosis de contratransferencia’,
published in the International Journal of Psychoanalysis in 1953 as ‘A contribution to the
problem of countertransference’. These two presentations of countertransference phenomena are
closely interdependent, and enable the analyst to be more aware of the complex effects of the
persistent, strong forces of the patient’s projections. This awareness is not only useful to the
analyst in understanding more of his own unsolved childhood neurosis, but also helps him to
avoid the danger of a defensive fi xation of a countertransferential position, and enables him to
explore the patient’s objects with which he originally is partially identifi ed. The
conceptualization of projective identifi cation and its effects on the object allows the analyst to
re-evaluate his experiences and opens up new areas of discovery. A deeper awareness of the
strong presence of projective identifi cation, and the splitting processes of massive
characterological pathologies in very disturbed patients, have enabled new work with the
countertransference technique to be done using the idea of the analyst’s countertransference as a
source of information about the patient, although Klein never made her peace with the term
being used in this way, as described above.

Racker’s work had a deep impact on psychoanalytic technique in Latin America; the most well-
known analysts of the time went on to refl ect upon these ideas in their work, and to develop
them.

329

Gustavo Jarast

León Grinberg

As noted above, Grinberg was one of the analysts who used Kleinian theory most directly in his
theoretical writing.
He coined the term

‘projective counteridentifi cation’ in 1956 to refer to a kind of countertransferential reaction

brought about when a patient makes particularly intense use of the mechanism of projective
identifi cation.

He presented this idea in a paper published later as ‘Sobre Aspectos mágicos en la transferencia
y en la contratransferencia. Sus implican-cias técnicas. Identifi cación y “contraidentifi cación”
proyectivas’, in the Argentinean Revista de Psicoanálisis . He explained that this mechanism
eventually leads the analyst passively to enact what the patient unconsciously wants him to. He
was thinking of a particular response of the analyst in reaction to the patient’s projections and
introjections, which would be independent of the analyst’s own emotions.

Grinberg emphasizes that this reaction should be attributed to the extreme violence of the
patient’s projective identifi cation, which is related to traumatic childhood experiences, and is a
result of violent projective identifi cations that the patient himself has received from others. The
counteridentifi cation reaction shows the analyst’s inability to tolerate the projection or to be
aware of what is going on: he only suffers the impact of the material projected by the analysand,
and reacts to it in a ‘real and concrete way’, acquiring and assimilating it. The analyst may feel
that he can no longer manage his own feelings, and that he is bewildered. He may rationalize this
state of affairs, but he cannot become aware that he has become an object of the analysand, or a
part of him (e.g., the id, the ego).

Grinberg gives detailed description and analysis in order to differentiate his concept from
Racker’s complementary countertransference. As described above, Racker’s description refers to
a kind of object relationship in which the patient becomes an internal object of the analyst
because of the analyst’s own infantile neurotic remnants, reactivated by the anxieties and
projections resulting from the patient’s confl icts. The patient’s internal objects may be
experienced as the analyst’s own. Thus, different analysts might react in different ways to the
same situation posed by a patient, depending on the type and nature of their own confl icts. In
projective counteridentifi cation Grinberg’s hypothetical patient’s projective identifi cation
would provoke the same countertransferential response in different analysts.

This is because the quality, the intensity and the force of the 330

Projective identifi cation in Argentina projection will no longer permit a critical threshold in an
analyst’s perception, and the extra-verbal pressure on him will produce a particular reaction.

Grinberg uses the following clinical vignette as an example: A male patient began his session in
the following way:

‘I feel very nervous today. I don’t know how to describe it, but it is absolutely necessary that I
do. I would like to tell you what I have discovered or what has been revealed to me.’ [With great
emotion]: ‘It was so surprising the other day when the diarrhoea stopped as a result of what you
said to me . . . Besides I remember that something else you said gave me a physical stitch.
Diarrhoea is a physical process

. . . Since you spoke, these words seem to produce a physico-chemical reaction in some or other
of my nerve cells; but before that, when you think, there is also a transformation in other cells to
the point where the voice comes from the lungs, lips, tongue, etc., and a string of words which
are now sound, vibrations, comes out. At this moment the receptive process begins in my ear,
through various means until it becomes conscious listening. I ask myself if all those words,
instead of being spoken by you, came from someone else, would they have the same therapeutic
meaning? I think not. It is extremely important for me that those words came from Dr. Grinberg
and no one else.’

All this material was said with force and with a resounding voice which surprised me. It was not
common for him to express himself like that and therefore I felt particularly attracted, as much
by what was said as by the way it was said. Using my impressions as a guide, I interpreted that
he was trying to produce in me the same effect that he said I had had on him, and to show me
that it was his voice and no one else’s which produced this special effect on me.

That is to say that my interpretation was made showing his positive transference. I did not yet
realize that it was only a defence against his deepest paranoid situation.

The patient went on to say: ‘Now that we are talking about sound and listening, I would like to
talk about music: it is divided into three basic parts, rhythm, melody and harmony which are
indissociably joined together. I play jazz; in that we see rhythm and the harmony of the song we
are playing. The melody is improvised. In modern jazz, the rhythm and the harmony are also
improvised. I can improvise for hours in melodies with rhythm but I fi nd 331

Gustavo Jarast

it diffi cult to carry on a specifi c harmony. A melody in 8/4 time in the chord of A for four beats
and the other four in A sharp is impossible for me. The same happens with written music; I
cannot give the timing correctly to each note. On the other hand, when my music teacher played
one of the pieces I was studying, I could play it afterwards exactly by ear. In the session, for
example, I fi nd it diffi cult to adapt to the reality of time. I don’t even know what time it is. It is
as if I made my own time, which is different from your time. I can compare it to my inhibition in
music; this specifi c harmony which we improvise is the kind which allows people who don’t
know each other to

improvise a jam session.’

While he was telling me all this, I did not fully understand what was happening. I felt quite
uneasy. I felt sorry that I did not understand suffi ciently the theory and the technique of music
which I have always loved. I admired and envied his knowledge and the apparent precision with
which he described and explained it . . .

with its technical jargon, the relationship between rhythm, melody, and harmony. I felt the need
to interpret it in his own words; it was a way of showing him that I could also play in the same fi
eld as him and which he knew so well.

My interpretation was that I represented the chord of A major and he the chord of A minor, but
that between our beats there was no harmony and we needed to fi nd a rhythm and a timing
between the two of us which would harmonize so that we could improvise (free association)
together in a common melody. The interpretation was now spoiled; it only demonstrated a partial
aspect and deviated from what was essential.

The important thing would have been to show him his envy and that he was really interested in
discovering which was my timing and which was my rhythm.

The meaning of his deepest phantasy began to dawn on me and I paid attention to the following
material. The patient went on: ‘I don’t know why I thought that one could do all kinds of tests on
a patient; encephalograms, BMR’s, a tape recording, a thermometer to take temperature with, an
oscilloscope to record sound waves; anyway the use of all those appliances so that you would
have a better knowledge of the patient, both inside and out.’ While I was listening to him I
surprised myself with a parallel and simultaneous phantasy of having a metronome to regulate,
control, and direct the time on him, that is to say, to have something which I already knew was
lacking in him. I realized exactly at that moment the full 332

Projective identifi cation in Argentina play of his unconscious phantasy, contained in his intense
projective identifi cation, and also how I ‘counteridentifi ed’ myself projectively with a partial
aspect of him, full of envy and anxiety.

One of the major effects of my ‘projective counteridentifi cation’

was a blind spot towards the paranoid content of his attitude and my having stressed instead the
positive aspect of the transference.

The patient used this in a defensive way to pacify the persecutor, which I represented. But that
was only his defence because of his anxiety and panic due to the power he attributed to me.

My words not only cured him of his diarrhoea but also gave him a physical stitch. I was in
possession of a secret which he envied and feared because I could do what I wanted with him.
He wanted to take this over so that he could limit its danger and so that he could dominate me at
the same time. For this he needed both to know me and to control me. It was for this that he
‘took’

me into his own fi eld, acoustics and music. He made me feel, projectively, what he had felt,
with me. My feeling of dislike corresponded to his feeling of anxiety. My admiration and envy
refl ected similar feelings which he had felt, and my need to use his terminology and concepts
was the equivalent of his desire to take onto himself my special terminology and concepts. My
fantasy of the metronome formed the response to his desire to use all kinds of medical apparatus
so as to get to know me completely, that is to say, to control me. As a last resort, and as a
transactional solution, he offered me his beat and timing in exchange for knowing mine.

(Grinberg, 1979, pp. 180–183, author’s translation)

To summarize, Grinberg thought that the analyst’s reaction to the patient’s projections is
essentially independent of the analyst’s confl icts. What he refers to as a pathological projective
identifi cation, he saw as a further development of Melanie Klein’s projective identifi cation, as
an omnipotent phantasy. He thought of it as a kind of projection that produced real effects on the
analyst as a receptor, so that he became dominated by it. Hence his view that all analysts,
regardless of their character and abilities, would have the same

‘projective counteridentifi cation’. Grinberg found support for his ideas in Bion’s theory of
thinking processes: the

concept of the beta screen in the psychotic patient, through which the patient’s projections of
beta elements go beyond the analyst’s countertransference, 333

Gustavo Jarast

leading to the possibility of understanding, and transforming them into thinkable or ‘alpha’
elements.

Horacio Etchegoyen questioned Grinberg’s affi rmation, suggesting that when these processes
have this kind of impact, it is only because the analyst is under the infl uence of his
‘countertransferential neurosis’

(Etchegoyen, 1986). In 1982 Grinberg modifi ed his original position, asserting that projective
counteridentifi cation enables the analyst to feel a kind of emotion that, when understood and
sublimated, would convert into a very useful tool for keeping in touch with the most profound
levels of a patient’s confl icts (Grinberg, 1982). For this to happen, the analyst must be prepared
to receive and contain the patient’s projections.

Madeleine and Willy Baranger

At the beginning of the 1960s Madeleine and Willy Baranger introduced one of Latin America’s
most fruitful contributions to psychoanalysis with their concept of the ‘psychoanalytic fi eld’.
They published a paper in the Revista Uruguaya de Psicoanálisis entitled ‘La situación analítica
como campo dinámico’ (Baranger and Baranger, 1961–1962). In this paper they presented an
original idea of the psychoanalytic fi eld as a dynamic and bipersonal one, made up of the
unconscious phantasies of both members of the therapeutic couple during the analytic process,
thus developing the concept of unconscious phantasy as described by Melanie Klein and Susan
Isaacs. The idea of a dynamic fi eld was founded on the concepts of Gestalt theory and on
Merleau-Ponty’s phenomenology. It was Etchegoyen who said that the Barangers, and Latin
American analysts in general, place the psychoanalytic process not in the patient, but ’in
between’ the analyst and the patient (Etchegoyen, 1986). The Barangers thought that in the
regressive situation of analysis, a bipersonal phantasy is formed, different from that of either the
patient or the analyst, considered individually. This phantasy underlies the dynamic of the
analytic fi eld. It can be seen not as a sum of the unconscious phantasies of each member of the
couple, but as something created between them, enabled by the unity created during the session.

Central to this phantasy are the mechanisms of projective and introjective identifi cation, and
projective counteridentifi cations.

The Barangers thought that the patient’s and the analyst’s bodily 334

Projective identifi cation in Argentina experiences and phantasies were involved in this dynamic.
The use of projective identifi cation in the present reactivates patterns of past experience, which
have not up until that moment been crystallized.

In the process of countertransference exploration, the analyst may realize that he is identifi ed
with split-off aspects of the patient’s internal world, and he can then interpret this phenomenon.
In such situations projective identifi cation is limited, so that the analyst’s regression is partial.
But in other situations these processes are much more active, so that the transference–
countertransference neurosis may paralyse the analytic process, through the constitution of a

‘bastion’, a structure which immobilizes the process.

In the 1970s, French thinking gained increasing infl uence among Latin American
psychoanalysts, and this had some impact on the Barangers’ conception of the psychoanalytic fi
eld. They continued thinking of their original structure as before, but they included the
conception of ‘intersubjectivity’ in response to Lacanian criticism of the idea of the analytic fi
eld as a ‘specular’ (mirror-like) relationship.

Lacan accented the idea of a ‘third’, which established a difference in the ‘specular couple’.

However, the most fertile element of the Barangers’ theory of the psychoanalytic dynamic comes
from their exploration of countertransference phenomena, for instance through their defi nition
and analysis of the bastion. As

mentioned earlier, the bastion is a pathological structure which immobilizes the dynamic analytic
process, as a consequence of the patient’s resistance and the analyst’s counter-resistance. It may
become chronic, leading to an analytic impasse, or a negative therapeutic reaction. In their
prepublished presentation in the Thirty-third International Congress of Madrid in 1983, the
Barangers offered several clinical vignettes which illustrated what they conceived of as the
bastion: ‘a neo-formation set up around a shared phantasy assembly which involves important
areas of the personal history of both participants and attributes a stereotyped imaginary role to
each’

(Baranger, Baranger, and Mom, 1983, p. 2).

Here are some of the brief examples which they presented to illustrate the concept of a bastion.

A. A manifestly perverse patient. He behaves like a ‘good patient’, complies with the formal
aspects of the pact, manifests no resistances, does not progress. The sessions, over a certain
period, seem to be a condensed version of the whole of ‘Psychopathia Sexualis’

335

Gustavo Jarast

by Krafft-Ebing (1886). The analyst ‘has never seen anyone with so many perversions’. The
bastion here is set up between an exhibitionist analysand and a fascinated-horrifi ed analyst, the
forced

‘voyeur’, complacent with regard to the perverse display.

B. An analysand, veteran of a number of analytic treatments.

Apparently, each session bears the fruit of some ‘discovery’; in reality, nothing is happening.
The analyst is delighted by the subtlety of the analysand’s descriptions of his internal states,
enjoying his own Talmudism. Until he realizes that, while they are toying with their
disquisitions, the analysand is monthly placing the analyst’s fees at interest, speculating with his
delay in paying.

The analysis of this bastion reveals a shared fantasy set-up: the analysand’s old, surreptitious
vengeance on his stingy father and the analyst’s guilt-ridden compulsion to set himself up as the
cheated father.

C. Example of a bastion which has invaded the fi eld. A seriously psychopathic patient. The
analyst is terrifi ed, fearing the analysand’s physical, homicidal aggression without being able
either to suspend or to carry the treatment forward. The nodular fantasy of this bastion is the
patient’s as torturer in a concentration camp, and the analyst’s as tortured, powerless victim.
With the conscious formulation of this manoeuvre, the analyst’s terror disappears.

The two individual histories converge in the creation of this pathological fi eld.

(Baranger et al., 1983, p. 2)

In the fi rst vignette the fascinated analyst cannot separate himself from his exhibitionistic
patient. In the second, the guilty identifi cation of the analyst combines with the vengeance of the
analysand.

In the third, the combination is between the sadomasochistic aspects of patient and analyst,
expressed through the phantasy of victim and torturer. In each case a shared unconscious
phantasy is the result of mutual projective identifi cations, coming from the infantile histories of
each participant. This situation converges in the formation of a bastion structure, which requires
a special ‘second look’ at the immobilized situation. The Barangers concluded: Each of us
possesses, explicitly or not, a kind of countertransferential dictionary (bodily experiences,
movement phantasies, appearance of certain images, etc.) which indicates the moments in which
336

Projective identifi cation in Argentina one abandons one’s attitude of ‘suspended attention’ and
proceeds to the second look, questioning oneself as to what is happening in the analytic situation
. . . This structure (the bastion) never appears directly in the consciousness of either participant,
showing up only through indirect effects: it arises, in unconsciousness and in silence, out of a
complicity between the two protagonists to protect an attachment which

must not be uncovered. This leads to a partial crystallization of the fi eld, to a neoformation set
up around a shared phantasy assembly, which implicates important areas in the personal history
of both participants and attributes a stereotyped imaginary role to each. Sometimes the bastion
remains as a static foreign object while the process apparently goes forward. In other situations,
it completely invades the fi eld and removes all functional capacity from the process,
transforming the entire fi eld into a pathological fi eld.
(Baranger et al., 1983, p. 2)

When ‘things go well’, the analyst’s second look may create timely interpretations which will
help to undo the bastion and enable the patient to gain insight into the projective identifi cations
which have been active. According to the Barangers, ‘Projective identifi cations of the analyst
toward the analysand and his reactions to the projective identifi cations of the latter [. . .] provoke
pathological structuring of the fi eld, require a second look toward it, also demand priority in
interpretive management’ (Baranger et al., 1983, p. 5). The breaking of a bastion implies
returning to the analysand the aspects which were placed in the analyst by projective identifi
cation. Extreme forms of bastion crystallizations in a psychoanalytic fi eld may become stag-
nations, in which the analyst may feel ‘as though he were “inhabited” by the analysand, a
prisoner of worry which goes beyond the sessions’ (through fear of a self-destructive or criminal
act of the analysand, of the imminence of a psychotic ‘outbreak’, or of other, less dramatic
situations). ‘These parasitic situations (equivalent to micro-psychoses in the analytic fi eld) tend
to lead either to a violent rupture of the analytic situation or to its re-channelling by reducing
splitting and by returning the projective identifi cations to the analysand’

(Baranger et al., 1983, p. 9).

The Barangers believed that the pioneering papers of Heimann and Racker showed that
countertransference was not only a universal phenomenon, but an instrument of analytic
technique as well. They 337

Gustavo Jarast

also thought that Melanie Klein’s discovery of projective identifi cation demanded profound
modifi cations in transference theory. For them the concept of projective identifi cation was
overextended, to the point that fi nally transference became quite synonymous with ‘a
continually active projective identifi cation’, and that ‘this led her to defi ne the movement of the
analytic session as a succession of projective and introjective identifi cations, resulting from the
analyst’s interpretive activity’. Later they circumscribed the transference by projective identifi
cation as follows: ‘This type of transference is distinguishable from the others thanks to the very
well-defi ned countertransferential expressions accompanying it, and intervenes determinatively
in the constitution of the pathology of the fi eld’

(Baranger et al., 1983, pp. 3 and 4).

Authors such as Beatriz de Leon de Bernardi use the concept in line with the Barangers, at the
Uruguayan Psychoanalytic Association.

Conclusion

The authors discussed in this chapter are the Latin Americans who

have most richly used and extended the concept of projective identifi cation. They pioneered new
possibilities in therapeutic practice, allowing analysts to get in touch with the more regressive
aspects and defended parts of their patients. In that sense these authors can be clearly located in
the tradition of post-Kleinian thinking which has done so much to benefi t the most disturbed
patients, from Bion to Rosenfeld. Moreover, they are dignifi ed heirs to the intuitions and
conquest of Kleinian research.

In this short enumeration and narration, I have discussed what I consider to be the most
important and fruitful contemporary contributions of these authors, and only mentioned some
others whose contributions have not infl uenced today’s debates on technique to the same degree
as, for example, the Barangers’ contributions.

Researchers like Joseph Aguayo are working hard at the heart of the concept in order to uncover
the controversial

roots of Klein’s rich concept.

338

20

Projective identification

Brazilian variations of the concept 50

Marina Massi

The aim of this paper is to present the use, or different uses, of the concept of projective identifi
cation within the scope of analysts belonging to the IPA in Brazil. A brief summary of the
historical facts can help us to understand and contextualize some of the characteristics of this
psychoanalytic production.

The psychoanalytic movement in Brazil has its origins in the mid 1920s through the pioneering fi
gure of Durval Bellegarde Marcondes, through whose efforts the IPA-qualifi ed training analyst
Dr Adelheid Koch arrived in Brazil, in 1936, giving rise to the education of a group of
candidates. After years of intense educational and institutional organizational efforts, in 1951, the
fi rst Society of Psychoanalysis was founded, gaining IPA recognition, in Brazil.

English analyst Mark Burke, member of the British Psychoanalytical Society, arrived in Rio de
Janeiro in 1948, and began the analysis of candidates. In 1949, German psychoanalyst Werner
Kemper

50 This is a condensed version of a research study conducted by the author. I would like to thank
Luiz Meyer for the valuable discussions during the writing of this paper; my colleagues who
undertook interviews for the research, including Aloísio de Abreu, Alírio Dantas Jr., Gley P.
Costa, José Carlos Zanin, Maria Elena Salles, Maria Inês E.

Carneiro, Marilza Taffarel, Nelson José de Nazaré Rocha, Paulo Marchon, Sergio Leukowics
and Sonia Azambuja; Aurea Rampazzo and Greice Klem for help with the text; and Henrik
Carbonnier, who translated it into English.

339
Marina Massi

(considered a member of the Nazi party through documentation unearthed by Riccardo Steiner)
also began the analysis of other candidates as part of their education. Due to a misunderstanding
between these two groups, a third group of candidates was formed, seeking education in
Argentina, and thus becoming known as the

‘Argentine group’. In the 1950s, some analysts such as Décio Soares de Souza and Henrique
Mendes, who became a member of the British Society, returned to Brazil (Sagawa, 1980, 1992).

In Rio Grande do Sul, due to its proximity to Argentina, psychoanalytic activities were infl
uenced by the APA (Argentine Psychoanalytic Association). Thus, in 1963, the Porto Alegre
Psychoanalytical Society was recognized (Sagawa, 1980, 1992).

The infl uence of the Kleinian school of thought in Brazil was introduced during the 1950s, when
the fi rst generation of Brazilian psychoanalysts went to London to come into contact with Klein
and her followers. The point of entry of the concept of projective identifi cation into Brazil was
through the direct infl uence of seminars and clinical supervisions experienced by Brazilian
analysts with members of the Kleinian group in London, or through visits by foreign
psychoanalysts.

However, as opposed to other countries, it seems that Brazil was

not widely regarded as a destination by foreign psychoanalysts 51 with

a solid theoretical background. Nor has Brazil had pioneering analysts with an organizing power
of infl uence – in the theoretical or clinical fi elds – that nurtured an original psychoanalytical
production.

The Revista Brasileira de Psicanálise (Brazilian Psychoanalysis Journal) featured Melanie Klein
on the cover and a reproduction of the letter she sent to Durval Marcondes, on 1 October 1956, in
which she writes: ‘It makes me happy knowing that my work has been both stimulating and
useful to the Society of São Paulo.’ This was, in fact, a truthful assessment.

The Kleinian school of thought prevailed during the 1960s, concurrently with growing interest in
the thinking of Wilfred Bion.

Virgínia Bicudo and Frank Philips played important roles in the introduction of Bion’s ideas to
São Paulo. The fi rst translation of Bion’s work – Os Elementos da Psicanálise (Elements of
Psychoanalysis)

– was published in 1966.

51 Heinrich Racker and Marie Langer.

340

Projective identifi cation in Brazil Frank Philips, a controversial psychoanalyst, went to London
in 1940 and underwent analysis with Melanie Klein and then with Bion, during which period he
maintained contact with Brazil through supervisions and a few lectures. He returned to Brazil in
1968, establishing a new frontier of psychoanalysis at the SBPSP (Brazilian Society of
Psychoanalysis of São Paulo).

According to Sagawa’s historical studies, Philips had a decisive infl uence on the implantation of
Bion at the SBPSP, ‘which based itself on the presentation of a new focus of Psychoanalysis, to
the point that, in the 1970s and 1980s, it became common to oppose a classic psychoanalysis
with a new one and it was even designated the

“real” Psychoanalysis.’ 52 Philips was the analyst for the re-analysis of many of the SBPSP’s
training analysts (Oliveira, 2005, p. 266).

The 1990s saw a different scenario. The participation of new generations of training analysts
within the Societies introduced new ideas and opened the door to restlessness regarding what
constitutes Brazilian psychoanalysis. The return of analysts seeking education in London
contributed to the debate on Brazil’s theoretical and clinical psychoanalytical productions
(Perestrello, 1992, p. 176).

The various theories and schools of thought seem to be better represented within Brazil’s
Societies, even though the English School still predominates. Brazilian psychoanalysts now have
a greater interest in dialogue between schools and this could well represent one of the important
characteristics of contemporary psychoanalysis in Brazil.

Contributions to the use of the concept

When discussing the contributions of psychoanalysts to the use of the concept of projective
identifi cation, in some ways, we are seeking the origin and structure of the universality of
certain psychoanalytical concepts.

The creation of a concept or the formulation of a change to a concept involves specifi c

contributions that come from daily clinical thought, refi ned by the parameters of the theory and
the technique.

52 Sagawa, R. Y. (n.d.).

Um recorte da História da Psicanálise no Brasil [An Extract

of the History of Psychoanalysis in Brazil], p. 9, www.cocsite.coc.fi ocruz.br/psi/pdf/

artigos1.pdf .

341

Marina Massi

That which is original is born out of that which is known and transformed.
Luiz Meyer, in a piercing refl ection on the universality of psychoanalytical production, said: it is
not built as an absorption or dilution resulting in a placid and aseptic whole. On the contrary, it is
formed by specifi c contributions of a remarkable character – a bumpy terrain – each of which
have an organic quality that confers its coherence.

What marks a school, what gives it consequence is not merely originality.

Originality sustains itself in the weave of the thinking, in the fruit of the transformation effected
on what is already known.

(Meyer, 1989, p. 364)

Although the problem is not one of originality, what needs to be faced in this paper is the fact
that Brazil lacks a psychoanalyst with international recognition as an expressive collaborator to
some aspect or use of this concept.

I am not proposing to ignore Brazil’s contributions, but attempting to identify a specifi c reality
that, once recognized, needs a strategy to be approached.

Brazil currently has twelve societies and eleven nuclei of psychoanalysis over a vast national
territory, which implies a wide variety of social contexts and a production that is spread out and
not always known and discussed by the country’s analysts. Faced with this context, I sought to
create a tool – a questionnaire – to enable a mapping of the main theoretical infl uences used by
Brazilian analysts.

What can be seen is that Klein and her followers have had the greatest infl uence regarding the
concept of projective identifi cation among Brazilian psychoanalysts, with special mention of
Herbert Rosenfeld, Betty Joseph, Heinrich Racker, Paula Heimann, Meltzer, Joseph Sandler and
Grinberg (through his concept of projective counteridentifi cation). Recently, Thomas Ogden,
Antonino Ferro and James S. Grotstein have contributed with the concept of transidentifi cation .

However, Bion is considered the follower who most expanded the use of the concept when he
described it as a means of communication, re-evaluating the importance of projective identifi
cation in the analytic relationship, as well as human relationships.

Based on the history of psychoanalysis in Brazil, bibliographic research and the questionnaire
replies from psychoanalysts belonging 342

Projective identifi cation in Brazil to various societies in the country, it was possible to obtain
some information about the authors who have developed a different use of the concept of
projective identifi cation, and I have tried to convey this below.

Mario Pacheco de Almeida Prado (1917–1991)

Mario Pacheco de Almeida Prado was the founding member of the Brazilian Society of
Psychoanalysis of Rio de Janeiro (SBPRJ). He was the Director of the Institute and President of
the Society. He underwent analysis with the English psychoanalyst Mark Burke, responsible for
his Kleinian infl uence.
In his book Identifi cação Projetiva no Processo Analítico [Projective Identifi cation in the
Analytical Process], the author writes that the use of projective identifi cation happens ‘as a
vehicle of perception; that is, conceiving perception as an identifi cation resulting from a
projective identifi cation of something that is separated and placed outside the ego or self ’
(Almeida Prado, 1979, p. 32). Perception and identifi cation are possible only as a reintrojection
of projective identifi cation.

In a paper about the work of Almeida Prado, José Carlos Zanin (2004) summarizes his ideas, affi
rming that: we verify that, due to the use of projective identifi cation before perception, what is
perceived is not the totality, the authenticity of the lost object, but a mixture of that which
belongs to the object and that which the observer inserts from within into the perceived object.

(Zanin, 2004, p. 318)

That which is mainly inserted can be love (life instinct) or hate (death instinct), according to the
conception developed by the author.

Through his study on projective identifi cation, Almeida Prado (1979) formulated, in an original
manner, the concepts of states of ingrainment and the ingrained object .

In analysis, the state of ingrainment can be understood as a phenomenon unconsciously lived by

the patient, of feeling undifferentiated or intermingled with the analyst, which the patient
constructs through projective identifi cation and the reintrojection of the fi gure of the analyst
into the patient.

343

Marina Massi

Therefore, the interference of the destructive impulses in the appearance and maintenance of the
self leads to the consequent state of ingrainment of the subject within its objects.

Isaías Melsohn (1921– )

Isaías Melsohn, a psychoanalyst of the SBPSP, critically situates the concept of projective
identifi cation in the scope of the phenomenology of Max Scheler, the empirical research of
Klages, Kurt Goldenstein, Köhler and Kofka and, fi nally, in the concept of expressive
perception of Ernest Cassirer.

Based on these authors, Melsohn locates the concept of the schizoid position as a state of
consciousness in which there is an absolute split.

There are no connections. It relates to a purely expressive moment of the consciousness, which
may reappear later in, for example, psychosis. These expressive experiences are not only
persecutory.

The child may be captivated by fascination, tranquillity, pleasure, grimness, fright, etc. This is
where Melsohn disagrees with Klein regarding the denomination of the paranoid-schizoid
position (Sister and Taffarel, 1996).

In his book, A Psicanálise em nova Chave [Psychoanalysis in a New Key], Melsohn (2000)
relates his concept of expressive perception to projective identifi cation, the paranoid-schizoid
position and the perception of ‘you’ , as described by Max Scheler.

The author believes that the roots of the perceptive process in children are not syntheses of
sensorial ‘elements’, but understandings of totalities. They are ‘original, primitive and immediate
expressive characters’ (Melsohn, 2002, p.

250). For Melsohn, the infant’s perceptive experiences are expressive experiences, as described
by the psychology of form, and revisited by Max Scheler and Cassirer.

Melsohn insists on clarifying two points: there is no understanding of isolated sensory content
that is then synthesized; and there is no

‘objective’ perception that is deformed by emotions.

According to the author, expressive perception, as a unit and fusion of the external-internal, is
the notion that corresponds to Klein’s projective identifi cation. ‘It is not just the subjective
projected externally, nor a copy of the exterior; both aspects determine each other as objectifying
is only possible in contact with suitable signifi -

cant forms.’ (Melsohn, 2002, p. 252).

344

Projective identifi cation in Brazil

Walter Trinca (1938– )

Walter Trinca’s work on

intrapsychic projective identifi cation departs from Klein’s view of projective identifi cation and
moves on to Bion’s so-called normal projective identifi cation. For Trinca, projective identifi
cation is processed in object relationships and the same process can be considered to take place
inside the actual self. In intrapsychic projective identifi cation, the subject unconsciously splits
the self, projecting one of the resulting parts inside another, equally split part.

The part that contains the aspects placed within tends to be taken as the identifi cation of the
whole self, becoming dominant, while the part whose aspects were relocated becomes empty and
poor.

Trinca’s fi rst studies date back to 1991 (Trinca, 1991), 53 when he took the self as the
representation of the whole person. It is precisely this self that, through intrapsychic projective
identifi cation, divides, constituting a nucleus that Trinca calls the self ’s sensory system. This
system occupies the self to varying degrees, occasionally taking it over entirely. Instead of
occurring only in object relationships, the projective identifi cation takes place between the parts
of the self.

Trinca considers it necessary to delimit a sensory system within the self in order to account for
the nature of sensory experiences.

The part of the self that contains the sensory system tends to be taken as the whole of the self.
When this happens, the non-sensory and immaterial part of the self becomes impoverished,
leading the person to become mentally and spiritually compromised.

Trinca wrote a paper for Free Associations , published in London, where he separated the
interior being from the self (Trinca, 1992).

For Trinca, the interior being and the self form distinct entities, albeit ones that communicate
between themselves.

His book,

O ser interior na psicanálise [The Interior Being in Psychoanalysis] (Trinca, 2007), shows us
that while the self is an entity of confl icts and turbulence, the interior being is in itself

53 Trinca, W. (1991). A etérea leveza da experiência [ The Ethereal Lightness of Experience ].

São Paulo: Siciliano, p. 100. In 2006, Vetor Editora published the second revised and updated
edition, under the title Psicanálise e transfi guração: a etérea leveza da experiência

[Psychoanalysis and Transfi guration: The Ethereal Lightness of Experience]. See also Trinca,
W. (1991). Notas para um estudo da sensorialidade da mente [Notes for a study of the mind’s
sensory sensitivity]. Revista Insight Psicoterapia , 12 (2), 20–23.

345

Marina Massi

harmonious and non-sensory. A greater or lesser infl uence can result in diversifying psychic
disturbances.

Instead of a psychoanalysis of elements, the author describes a psychoanalysis of factors that are,
basically: 1 a constellation of the internal enemy

2 a distancing of contact with the interior being

3 a fragility of the self

4 the sensory ability

5 an unconscious structuring

6 self dissipation anxiety.

As projective identifi cation plays the role of an active and mobilizing element within the psychic
apparatus, it continues to exercise a function of internal communication, whether between the
entities or between the factors. It only specializes in this function of communication.

Trinca believes that projective identifi cation in the self plays a role in the constitution of the
psychic entities we call the id and superego.

Trinca affi rms that projective identifi cation assists the operating of factors and the moving of
systems (pertaining to a comprehensive psychoanalysis). Its function is made relative and not
absolute, because it constitutes an element instead of a psychic factor.

Trinca believes that its use refl ects more on what he calls the closing state, related to psychotic
personalities (Bion).

Therefore, the communication exercised by projective identifi cation under these conditions
consists not of a bonding function, but instead of a depositing of the products of one or several
parts into another or others, as in anxiety defence, although conserving the original characteristic
proposed by Klein.

Projective identifi cations in phobias and panic are related to the diffi culties in what Bion calls
normal projective identifi cations, responsible for the emptying of the self, the precarious
position of the basic confi dence matrix and the instability of the internal support centre.

As for the uses of Bion’s normal projective identifi cation, Trinca employs them in relation to the
phobic personality. He starts from the principle common to psychoanalysis that the infant,
experiencing mental situations of

‘unimaginable anguish’, requires special 346

Projective identifi cation in Brazil conditions of psychic ‘metabolization’ that should be assumed
by the mother or caregiver. If there are no confi guring resonances from the mother’s responses
to the infant, the infant’s emotions tend to remain turbulent and he will have diffi culties in
representing himself. The failure of the primary container results in the maintenance of the
feelings of threat regarding life.

For Trinca, projective identifi cations in intuitive images are an individual’s direct or symbolic
representations of the meaning of the deep emotional states of another or other individuals.
Beyond the nature and intelligible meaning of the patient’s oral communication, and apparently
without any clear relationship with it, intuitive images appear spontaneously in the mind of the
professional. These images lead to non-verbal communication through the medium of Bion’s
normal projective identifi cation.

Luiz Carlos U. Junqueira Filho (1943– )

For some years, Junqueira has worked on the importance of specular phenomena in emotional
development, trying to elaborate the concept of specular identifi cation as part of a broader
investigation regarding the essence of metapsychology.

The material presented below comes from the author himself, who has generously provided
sections of a chapter soon to be published in book form.

Psychoanalysts are convinced that the human mind is the stage par excellence where the drama
of emotional transformations is played out. Consider, for example, two commonly used
characters, Narcissus and Echo, often seen as each other’s opposites: indeed, Narcissus is
unknowingly deprived of a partner with whom he can identify or into whose interior he can
project his excluded parts; Echo, in turn, represents a clandestine second self to exercise these
functions. [. . .]

The double is, at heart, an illusory resource used by the subject when his tolerance for reality
runs out: in creating a

second character that steals the scene, the subject diverts his attention from the painful focus, but,
at the same time, sets a trap for himself. This is what happened with Oedipus who, under the
illusion at the cross-roads that his fi ght with Laius was no more than an incident of 347

Marina Massi

mere ‘road rage’, could not be conscious that he was fulfi lling the oracle’s prophecy. The
perception of the subject under the illusion thus becomes split in two: a theoretical aspect
(theoréin, meaning

‘that which is seen’) emancipating itself from the practical aspect (i.e. ‘that which is done’).

With this backdrop of the ‘multiplication of personalities’, we can and should now reconsider
projective identifi cation in strictly psychoanalytical terms, trying to map the variation of its
meaning.

It seems clear that the Freudian and Kleinian subject uses it to defend himself from the
vicissitudes of his own personality or from external traumas [. . .]. However, from the point of
view of Lacanian, Winnicottian and Bionian subjects, its power is used to develop the identity of
the self, to introduce the subject to the fi eld of culture and to establish the bases of human
communication and thinking.

Some analysts, such as Grotstein [. . .] and Ogden [. . .] have offered interesting developments of
the theories of Klein, Bion and Meltzer. Grotstein admits that when the image of the mother-as-
object is internalized by the infant, child or adult, this image is altered by the fantasized effect of
subsequent projective identifi cations. To understand the installation of that which he called

‘psychic presences’, and which in my opinion correspond to ‘relational ghosts’, into the self,
Grotstein posited the immanent subject (the one recognized as the ‘I’ in conscious terms) and an
ineffable subject of the unconscious.

In Ogden’s vision, the ineffable subject of the unconscious communicates to the immanent
subject the individual’s anonymous pains through symptoms, dreams, actings and so on; in the
opposite manner, the confl icts of daily life are transmitted to the ineffable subject of the
unconscious, which processes them through unconscious dreaming.
Thus, an internal process of projective identifi cations is created, through which emotional
experience can be thought and so transmitted to oneself and others.

As a consequence of this situation, Grotstein proposes the idea of ‘autoctonia’, the fantasy of
auto-creation through which we

‘personalize’ the world by imagining it as a refl ection of who we are [. . .].

In my experience, to better understand the constitutive dynamic between the ideal I and the I’s
Ideal, I propose

‘mirroring’ as a 348

Projective identifi cation in Brazil term to denominate the process of reception on the part of the
parents of the infant’s projections and communications in an atmosphere of reception and
comprehension: in this case, the parents’

response is always constructive, collaborating in the elaboration of the infant’s anxieties or its
success in communicating. It would be something very akin to the notion of reverie proposed by
Bion.

The term ‘refl ectment’, I reserve to designate to parental insensitivity that returns, unmodifi ed,
to the infant, everything that it authentically produces in its efforts to face the reality of life.

Thus, it would be expected that the healthy infant could count on a positive response from its
parents as regards the

‘consultation’

directed to them when the infant aims to validate its ideal I: when this takes place, the infant feels
that its parents legitimate that which belongs to it, independent of any judgement of value.

Should refl ectment prevail, under the terms described above, there is the risk that, in addition to
not having its anxieties comforted, the infant may receive a package of parental judge-ments and
values in return, which implicitly disqualify it as a subject. From the metapsychological point of
view, in this last case everything would happen as if the infant had sent its ideal I for validation
and in return received it dressed in the I’s Ideal: of course, the doors for the constitution of a
false self would be open under these conditions.

Junqueira’s concept of specular identifi cation deals with the importance of specular phenomena
in the constitutive dynamic between the ideal I and the I’s ideal. The mirroring and refl ectment
are different receptive processes by the parents of the infant’s projective identifi cations.

Elizabeth Lima da Rocha Barros (1948– ) and Elias

Mallet da Rocha Barros (1946– )

Elizabeth and Elias Rocha Barros have presented works on projective identifi cation in which
they discuss how the splits and subsequent projective identifi

cations appear differently in the

paranoid-schizoid and depressive positions, as the introduction of verbalization in the projective

processes impacts on various issues in the area of how communication takes place between
patient and analyst, mother and infant, etc.

349

Marina Massi

The authors affi rm that the process involved is highly complex and cannot be encompassed by a
single term without specifying its evolution. They view projective identifi cation as a diffi cult
concept to be understood in its phenomenology and inherent processes, in addition to being
intrinsically linked to the issue of the production/construction of symbols, countertransference
and its expression in an evocative-expressive plane belonging to affections and emotions.

In other works, these authors have discussed the operation of projective identifi cation as a basis
for a kind of special empathy, a way of getting to know the other in that it allows us to place
ourselves

‘in the other’s place’, albeit with the addition of a new ‘metapho-rizing’ function – the authors
believe that the projection needs to allow for the capturing of meanings and their elaboration.

For the authors, projective identifi cation produces evocations, invitations for the construction of
a mental representation coloured by emotion. Thus, the evocation is a manner of non-discursive
expression, even if permeated by the patient’s verbalized discourse, thus allowing connections
other than those belonging to discursive logic and mediated by words to appear, and in this
manner expanding the forms of representation of emotional relationships.

Infl uenced by Susanne Langer, they came to work on the issue of the nature of symbolic
communication and how it presents itself in the processes of projective identifi cation.

From the clinical point of view, the authors point to the fact that the transmission of
communication through projective identifi cation is infl uenced by the actual structure of the
symbol, which in turn, can be deformed or limited by attacks aimed at its own structure. In other
words, the capacity for symbolic expression present in projective identifi cation can still be
affected during the actual process of symbolic production. These attacks thus limit the fi eld of
meaning present in the symbolic structure.

Roosevelt M.S. Cassorla (1945– )

Roosevelt M. S. Cassorla has developed the theme of crossed projective identifi cations since
1997. This refers to group occurrences (including the group formed by the patient and analyst) in
which each of the members impacts the other with mass projective identifi cations 350

Projective identifi cation in Brazil and is, in turn, also impacted by similar phenomena
originating from the other.

According to the author, crossed projective identifi cations are also

the basis for ‘enactment’. Cassorla studies and extends this concept, 54

describing it as situations in which the analytical dyad produce an obstructive plot thanks to
projective identifi cations taking place in a double sense between the patient and analyst. This
plot, which is not perceived by the members of the pair, is termed a ‘chronic enactment’ or
‘paired-non-dream’ by the author. The projective identifi -

cations are understood, based on a Bionian reference, as ‘non-dreams’; that is, as undigested
facts (beta elements) that neither the patient nor the analyst is able to dream or turn into dreams
through alpha-functions. When this plot is unravelled, a very noticeable event, called an ‘acute
enactment’ by the author, takes place. Cassorla’s

‘acute enactment’ is what is described by other authors as ‘enactment’, but Cassorla considers
such enactments to be the fruit of recapturing the pair’s capacity for thought which unravels the

‘chronic enactment’. The author demonstrates that the ‘acute enactment’ is created in the revival
of archaic traumatic situations that have been implicitly elaborated (implicit alpha-function)
during the

‘chronic enactment’. Cassorla believes that mutual suggestion, a fact that is part of any analytic
process, is the fruit of crossed projective identifi cations between the members of the pair. The
analyst’s art would be to allow himself to be infl uenced in a fi rst instance (to experience at-one-
ment with the patient and what he experiences) and simultaneously (or subsequently) to ‘dream’
this fact, permitting his entry into the symbolic network of thought.

There is no doubt that other authors also describe crossed projective identifi cations, but Cassorla
calls attention, in a decisive manner, to the fact that crossed projective identifi cations are the
basis for enactment. The author promotes the relationship between the use of the concept of
projective identifi cation and enactment in an original way.

54 Cassorla highlights the importance of several authors who dissected the importance of
countertransference as an instrument (Racker, Heimann, Betty Joseph and Money-Kyrle), and
those who followed the Barangers (and their concept of Field) stimulating the intersubjective
view (of whom Cassorla believes Ogden and Ferro to be the most important).

351

Marina Massi

Final considerations

The main objective of this article was to track the different uses of the concept of projective
identifi cation in Brazil.
As previously stated, there is no single psychoanalyst who has made an original contribution in
the consensus of the psychoanalytical community. Hence the need for a bibliographic survey and
the adoption of a questionnaire as a means to fi nd out which authors have been infl uential
within the psychoanalytical community through their original uses of the concept of projective
identifi cation.

What can be noted, then, is that there are at least two branches: the fi rst the more frequent use of
the concept based on the contributions of the English school of thought; while the second relates
to some original contributions by Brazilian analysts who are not particularly widely known,
either in Brazil or abroad.

As the English School’s infl uence over the various psychoanalytic societies in Brazil is signifi
cant, we can state that the most frequent use of the concept is founded on Klein and Bion, as well
as authors such as Rosenfeld, Meltzer, Joseph, Grinberg and, more recently, Ogden, Grotstein
and Antonino Ferro. Both the bibliographic survey and the questionnaire results point to these
authors as having the greatest infl uence on the use of the concept of projective identifi cation in
Brazil.

The second branch includes the seven Brazilian authors presented who have made original
contributions using

different approaches.

Almeida Prado describes the state of ingrainment and the ingrained object as phenomena derived
from projective identifi cation.

Melsohn, in turn, through his studies on expressive perception and the representative function ,
presents a different manner of understanding the relationship between the perception and the
perceived object. It is possible to say that, in this case, the author tries to restrict the defi -

nition of projective identifi cation in order to clarify the concept.

There is a criticism of the concrete manner of thinking that projects itself into the object and
what actually belongs to the object (its content).

Elizabeth and Elias Rocha Barros propose the presence of a ‘meta-phorizing’ and metabolizing
function within the concept.

Trinca ( intrapsychic ) and Junqueira ( specular ) seek to specify different uses of the concept. In
other words, we could say that these authors describe the actual concept of projective identifi
cation in a diverse manner.

352

Projective identifi cation in Brazil Cassorla attributes a special relevance to projective identifi
cation by promoting an original relationship between the concept of crossed projective identifi
cation and enactment.
This research has made it clear that there is a need to further the discussion regarding Brazil’s
production. I believe we still lack the productive experience of what I would term as ‘Brazilian
controversies’, a deep and intense debate on theoretical and clinical experience that does, in fact,
deserve to constitute a common ground for us all.

353

21

Projective identification and the weight

of intersubjectivity

Juan Francisco Jordan-Moore

Introduction

To recount the history of projective identifi cation in Chile I will use the publications that have
addressed the concept in my country. There are not many of them but they give a fair picture of
its development.

After an initially uncritical reception the concept has been challenged on the grounds that it is
based in a monadic conception of the mind that diminishes the contribution of the analyst to the
interactive nature of the psychoanalytic process. However, the acknowledgment of the analyst’s
contribution to projective identifi cation in a dyadic conception of the analytic process may not
suffi ce to give a full account of projective identifi cation as an intersubjective phenomenon.

Development of the concept

Edy Herrera (2000) describes the progressive interpersonalization of the concept and its
development from a pathological defense mechanism to its consideration as a normal process of
psychic life. Projective identifi cation was at fi rst considered to be an intrapsychic pathological
aggressive phantasy and its effect on the recipient was not taken into account. Later, it was
linked to a deepening of the theory of object relations and the connection between transference
and countertransference. Projective identifi cation is thus linked to the 354

Projective identifi cation and intersubjectivity analyst’s countertransference. Hererra highlights

León Grinberg’s (1985) concept of projective counteridentifi cation. Grinberg distinguishes
between the latter and complementary countertransference as defi ned by Racker. The analyst’s
neurotic residues are activated by the patient’s confl icts in complementary countertransference,
whereas in counter-projective identifi cation, the analyst’s response is mainly the consequence of
the intensity and unconscious intention of the patient’s projective identifi cation.

Finally Herrera discusses Bion’s contribution to the theory of projective identifi cation through
the link he makes between projective identifi cation and the model of the container-contained.

Projective identifi cation is considered as a mode of communication of primitive mental states

that affect the container. Depending on the containing capacity of the latter, this process may
result in psychic growth or in intolerable anxiety and psychic depletion. This is the most
interactive defi nition of projective identifi cation, yet, it seems that there is an assumption
concerning the original unconscious intention of PI in the projector. Is the intention
communicative or is it destructive?

Polemical stances regarding projective identifi cation Ayuy et al. (1997), commenting on
clinical material presented at the Trans Andean Meetings, are concerned with the clinical
problem presented when a patient who is feeling persecuted by the analyst due to massive
projective identifi cation violently rejects his interpretation. This refusal is described as ‘an acute
rejection situation’, to stress a complex experience that is not simply a consequence of a negation
of the meaning of the interpretation. The main problem is the fragmentation it suffers, hence
losing almost all signifi cance. They present a thorough investigation of the different strategies
found in the publications of Kleinian authors to confront this problem. John Steiner’s
recommendation to formulate analyst-centred interpretations; Betty Joseph’s emphasis on the
need to focus interpretation on the formal aspects of the patient’s communication; Priscilla
Roth’s interpretations that address what is enacted in the relationship. They emphasize the
analyst’s unavoidable enactments, described by Irma Brenman, Elizabeth Bott Spillius, Edna
O’Shaughnessy and Denis Carpy, as well as the need for close scrutiny of these. This allows the
analyst to 355

Juan Francisco Jordan-Moore

become aware of subtle pressures, coming from the patient, that induce the analyst to act. This
pressure can then be interpreted. Yet all these strategies, designed to contain the patient’s
projection, and fi nally formulate a verbal interpretation, may not be enough. They distinguish
two stances concerning the problem of containment of the patient’s projective identifi cation.
One conceives an initial unconscious intention in the patient, communicative or evacuative.

The latter is related to the rejection of reality by the psychotic part of personality. This intention
can overfl ow the containment capacity of the object. The other conceives projective identifi
cation as a dyadic phenomenon. In this frame, meaning cannot be discerned from the beginning,
it is not a priori. It depends on the quality of the analyst–

analysand interaction. Thus, the intention of projective identifi cation is defi ned a posteriori.
Ayuy et al. (1997) subscribe to the idea that there is an unconscious initial meaning of projective
identifi cation in the patient that can determine a fateful analytic process, independently of the
perturbation in the analyst’s countertransference.

Juan Pablo Jimenez (1992) argues that the analyst’s contribution to projective identifi cation has
not been given due signifi cance. He makes the point that the communicative, or evacuative and
destructive intention of projective identifi cation, is a meaning that depends on the analyst’s
capacity to contain the patient’s projections. If the analyst fails, projection is signifi ed as
destructive, if he succeeds, projection is connoted as communicative. He criticizes León
Grinberg’s (1985) phenomenology of different a priori meanings (communicative, destructive,
evacuative, reparative, controlling) of projective identifi cation. He recognizes that Bion and
Rosenfeld tried to go beyond a monadic conception of projective identifi cation, attributing to the
analyst, as the patient’s object, an active function as a subject who tries to put an end to the
vicious circle of projections and reintrojections. Yet, from the point of view of a dyadic
communication theory, this concept still implies a monadic conception of mind. Projective
identifi cation is described as almost always coming from the patient to the analyst as the
container. The reverse of the contained-container metaphor, the patient as the container of the
analyst, should also be considered. In Jimenez’

view, the theory of projective identifi cation emerges in the analyst as an explanation for
emotionally intense countertransference experiences.

Edy Herrera (2000) notices that this arises frequently with psychotic 356

Projective identifi cation and intersubjectivity and borderline patients. She thinks that the value
of the concept becomes manifest in working with these patients.

Juan Pablo Jimenez (1992) illustrates his thesis with two vignettes.

One of them, Veronica, seems to be a demanding borderline patient.

For three years the analyst had to manage intense acting out in the session, till fi nally, as a result
of an interpretation of a dream, the intense acting out subsided. She was then able to continue her
analysis fruitfully. The patient’s projective processes were thus understood to be communicative
projective identifi cation. During the period of intense acting out the analyst experienced intense
countertransference feelings, such as anger, which he took home with him.

He had to confront the patient repeatedly, sometimes very forcefully.

Frequently he had feelings of triumph over the patient, sensing that he could control and defeat
her. He understood these troubling feelings as a dim awareness that he was ‘projecting’ himself
into the patient with more or less violence, with the hope of being introjected by her with his
analytical function. The analyst recognized that this process took place in the structure of a sado-
masochist relationship, stating that the latter was determined by the patient’s psychopathology.

It is noteworthy that although Jimenez states that the contribution of the analyst is
underestimated in the theory of projective identifi cation, he ends by assigning the determination
of their relationship as sado-masochistic to the patient. There is no acknowledgement of the
analyst’s contribution to the patient’s massive projective identifi cation prior to the interpretation
of the dream, and no attention is paid to the possibility of the analyst’s own masochistic and
sadistic aspects contributing to the sado-masochistic structure of the relationship.

This discussion can be expected in an intersubjective frame that gives a full account of the infl
uence of the psychoanalyst.

Josefi na Figueroa (1997), in her summary of the group discussions at the above mentioned
meeting, reports a consensus that what tran-spires in the analytic dyad is two minds interacting
closely and mutually infl uencing each other. The analyst is emotionally involved with his
patient. ‘Living through’ this experience is the main way to know disavowed aspects of the
patient. There was also consensus that the concept of projective identifi cation had developed to
encompass an interactive notion of the analytic situation. Enactment by the analyst is inevitable
and, perhaps, the only way to make a valid interpretation, through the careful analysis of
countertransference. Priscilla 357

Juan Francisco Jordan-Moore

Roth’s suggestion of interpretations that include the analyst’s involvement in the enactment was
valued as an alternative to classical technique, although the need clearly to distinguish this kind
of intervention from a countertransference confession was enhanced.

Others considered that any change in technique could dilute the analyst’s identity, and should be
considered psychoanalytic psychotherapy. Figueroa concludes that the concept of projective
identifi -

cation has evolved from Melanie Klein’s original understanding to one that takes into account
the ‘ever-growing weight of the intersubjective aspect of the analytic couple’.

Jaime Coloma (2002) understands the psychoanalytic session as determined by a theory that
includes projective identifi cation and a preponderance of oneiric thought in both participants.
The coupling of PI and oneiric thought is considered the main way of dealing with emotions.
Based on Matte-Blanco, he thinks that every emotion connects with a symmetrical mode of
being, and thus, with an existential dimension that is lost if psychoanalysis is defi ned as a
positivist science. Starting from this defi nition of a psychoanalytic session, he posits that the
technique and

setting are not defi ned in terms of expected behavior of the analyst, but, instead, as different
possible interventions that emerge from an analytical understanding of what happens in the
session, taking into account the therapeutic benefi t for the patient. In the same vein as J.P.
Jimenez, he thinks that the meaning of the analyst’s interventions is defi ned a posteriori, i.e. it
becomes clear only retrospectively whether it is an acting-out by the psychoanalyst, or a proper
understanding of the total existential situation of the patient in which his emotional experience is
included.

From this defi nition of the session other kinds of analytic interventions may emerge, not
restricted to verbal interpretations, as in the case of interpretative actions (Casaula et al., 1994).
The issue of the interactive nature of projective identifi cation is not addressed. Yet an interactive
approach seems to be implicit when Coloma posits the analytic session as structured by
projective identifi cation and oneiric thought in both participants. This can be interpreted as
meaning that the patient also functions as the analyst’s container. Coloma does not address the
patient’s participation in defi ning whether the meaning of the analyst’s intervention is or is not
acting out. It seems that the analyst should be able to decide on his own which is the case.

The above summary of these papers enables us to identify two polemical issues regarding the
theory of projective identifi cation in 358

Projective identifi cation and intersubjectivity Chile. One is the question of how the meaning of
projective identifi cation is determined; the other, the question of what modifi cations are
acceptable to classical analytic technique in order to deal with the inevitable enactments linked to
projective identifi cation.
Discussion: the weight of intersubjectivity

Josefi na Figueroa’s fi nal commentary that stresses the ‘ever-growing weight of

intersubjectivity’ in the analytic situation seems an apt place to begin to discuss the fi rst issue. In
what sense is intersubjectivity a weight? Perhaps it presents diffi culties for a theory and practice
that understands mind as generating unconscious and conscious meaning mainly as an
intrapsychic phenomenon. The intrapsychic model seems to have given way under the weight of
accumulated clinical evidence, leading to an understanding of mind as engendering meaning in
an intersubjective fi eld and its associated dialogue.

Recognition of the participation of the analyst in a dyadic model of communication does not
necessarily take us all the way to intersubjectivity, as can be seen in the incongruence between
the theoretical importance assigned to the analyst’s infl uence, and what is reported in clinical
vignettes. Every so often it seems that what is written with one hand is erased by the other. A
dyad may still be two monads communicating their emotions through windows in the monads.
Take for example the following description of projective identifi cation by Thomas Ogden
(1994c, p. 105):

‘The projector and the recipient of a projective identifi cation are unwitting, unconscious allies in
the project of using the resources of their individual subjectivities and their intersubjectivity to
escape the solipsism of their own separate existences’. This is a striking affi rmation about
existence considering that through his conceptualization of PI as manifestation of the subjugating
intersubjective third, Ogden has given one of the fullest account of projective identifi cation in
terms of intersubjectivity, and thus of the primeval being-with-other in our existence.

I have posited ( Jordan, 2005) that the diffi culty for an intersubjective theory of projective
identifi cation lies in positing projective identifi cation as a mechanism that is independent from
the meanings that are negotiated in the relationship. The Barangers (Baranger and Baranger,
1961–1962) fi rst thought of their concept of a bi-personal fi eld as structured by projective
identifi cations and introjections from 359

Juan Francisco Jordan-Moore

patient to analyst and vice versa. 55 In this fi rst version of their fi eld theory, their clinical
vignettes also present the analyst as the rather passive receptor of the patient’s projective identifi
cation, as when they state, for example, that the patient’s desperate projective identifi cation
ultimately creates a tranference-countertransference psychosis in the session. In this
conceptualization of their fi eld theory, constituted by crossed projections and introjections, from
which unconscious phantasy emerges as a bi-personal phenomenon, no attention is paid to the
fact that projective identifi cation is also a phantasy and, as such, should also be considered as a
bi-personal phantasy, i.e. ‘the processes that are postulated as structuring the unconscious bi-
personal phantasies of the fi eld are themselves an unconscious

phantasy’

(Baranger and Baranger, 1961–1962, p. 162).

Later on, Madeleine Baranger (1993, 2005) posits that the concepts of projective and
counterprojective identifi cation may be attempts to elude the need to acknowledge the
participation of the analyst in the events of the fi eld. If the analyst fails he can easily slide into
viewing the patient as ‘guilty’ and as taking the initiative in projective identifi -

cation. Applying Bion’s idea about ‘basic assumption’ group phenomena

to the analytic couple, 56 they defi ned a ‘basic unconscious phantasy’

created by the same fi eld situation. One basic phantasy might be the solipsistic existence of two
human beings desiring to migrate into one another in order to escape from their desperate
isolation.

I proposed ( Jordan, 2005) to understand the fi eld as isomorphic with Matte-Blanco’s (1998)
notion of a basic matrix of projection and introjection. In this region of the ‘deep unconscious,
things –

including those that regarded from outside appear as projection and introjection – do not occur,
they simply are’ (

Jordan, 2005, p. 194).

At this level events do not exist, no inside or outside can be discerned, there is a simple
manifestation of being. This is not an easy matter to understand, we are in the region of ontology,
of being as such.

Perhaps Thomas Ogden’s description of the experience of projective identifi cation may help us
to understand (Ogden, 1994c). He says:

55 When thinking interactively projective identifi cation as going from patient to analyst and
vice versa it is always stressed that it is desirable that the intensity of this kind of exchange is
more intense and frequent from patient to analyst than vice-versa. This is due to the asymmetry
of the analytic situation.

56 Bion (1961) specifi es that the ‘basic assumption’ of the group should be sought in the matrix
of the group and not in the individuals that conform the group.

360

Projective identifi cation and intersubjectivity

‘The recipient is not simply identifying with an other (the projector); he is becoming an other
and experiencing himself through the subjectivity of a newly created other/third/self ’ (Ogden,
1994c, p. 102, my italics). Projective identifi cation is a matter of ‘being-the-other’

and the other ‘being-myself’ because, ‘In part, . . . there is never a recipient who is not
simultaneously a projector in a projective identifi catory experience’ (Ogden, 1994c, p. 102).
This kind of experience can be expressed succinctly as:
I am yourself;

you are

myself; we are together. I think that what has been called communicative projective identifi
cation can be included today in a theory that posits an indivisible system of self-regulation and
mutually regulated affects (Beebe and Lachmann, 2005). When this system fails, affects become
dysregulated, and a phantasy of projective identifi cation may emerge as a desire and a need to fi
nd an other who can help in the regulation of distressing emotions like anxiety, shame, envy,
depressive affect, rage, etc., that threaten to disintegrate the self. Projective identifi cation can
function, in a given interaction, as a self-regulating phantasy in a subject that experiences himself
as emotionally

isolated, expecting to trust someone and, thus, to use the opportunity for successful mutual
regulation, in order to ease the burden of self-regulation at the expense of frail mutual regulation.

Perhaps the diffi culty of developing projective identifi cation into a fully intersubjective concept
lies in the fact that at its inception it was linked with drive theory. In this frame, meaning is
generated endogenously from the drives.

Anxiety, for example, is originally understood as the consequence of the impact of the death
drive on the ego. In this theoretical imaginary fi rst moment, a solipsistic subject without a world
is posited. The world is a creation of the subject. A persecutory object comes into being through
projection to protect the ego from annihilation. A good object is created afterwards by the libido,
to protect the ego from the bad object.

If we substitute affect or emotion for drive,

57 we start with a

different picture. Affects, as currently understood, are, from the very beginning, regulated in an
intersubjective fi eld. They are inherently communicative. There is not fi rst an emotion and then
the intention to communicate it.

Affects open up a meaningful world from which

57 Affect and emotion are used as synonymous. Some authors use affect to indicate the
description from the point of view of an observer and emotion to connote the subjective
experience.

361

Juan Francisco Jordan-Moore

we are not separated. This world is with-the-other by whom we are affected and whom we affect
in many ways. The stuff of the basic matrix of projection and introjection are affects which, ‘on
the deepest level . . . are often transpersonal . . . [they] are evoked interpersonally through dense
resonances between people, without regard for who, specifi cally, is feeling what’ (Mitchell,
2000, p. 61).

Bion’s conclusion in his paper ‘On arrogance’ (Bion, 1957b) states: the denial to the patient of a
normal employment of projective identifi cation precipitates a disaster through the destruction of
an important link. Inherent in this disaster is the establishment of a primitive superego which
denies the use of projective identifi cation.

(Bion, 1957b, p. 92)

A premature superego implies an internalization of an object and consequently a premature

caesura of inside and outside, coupled with a premature knowledge of the other and self as
objects. This implies there is an untimely isolation of the self, a solipsistic subject, deprived of
emotional contact with another subject, a denial of intersubjectivity. This psychic state can
precipitate the need and desire to invade another in search of the intersubjective experience that
has been denied. The phantasy of projective identifi cation can be understood as emerging, a
posteriori, from a failure in the mutuality of affect regulation, i.e. ‘the normal employment of
projective identifi cation’.

If mutual affect regulation is going ahead smoothly there is no need to represent the process in
phantasy, because emotions are inherently communicative and transpersonal. This emotional
experience is embedded in the matrix of projection and introjection, in which what comes to light
is a continuity of being.

The same may apply to the clinical situation. As stated above, Juan Pablo Jimenez suggests that
the theory of projective identifi cation may emerge in the analyst as an explanation for
emotionally intense countertransference experiences or dysregulated emotions. The same can be
said of the patient when going through emotionally intense transference experiences. In my view,
the emergent theory of projective identifi cation in the analyst is linked to a mutually elaborate
unconscious phantasy, of introducing one’s own intolerable emotions into the other, who thus
becomes myself, perhaps searching for an opportunity to restore a lost sense of being-with-other.

362

Projective identifi cation and intersubjectivity The other controversial question concerns the
interventions used to handle events in which projective identifi cation and impasses appear on
the scene. In a previous paper (Casaula et al., 1994) we noted that interpretation in
psychoanalysis has been restricted to the analyst’s verbal utterances.

Following Heidegger (1927) we proposed an extension of the psychoanalytic notion of

interpretation to include the assignation of meaning through actions performed by the analyst.

We put forward the concept of interpretative action. Joan’s case is presented in the paper. The
analyst recalls the confused countertransference with a child in analysis. In a climax of confusion
and desperation on the part of the analyst, with the child in a fragmented manic state in a session
in which she experienced herself as being in many places at the same time, the analyst, starting
to think from her desperation and confusion, decides to refl ect Joan’s image on the back of the
lid of a round tin pencil box, interpreted jointly, in that moment, as a mirror. The child,
anticipating the analyst’s intention, screams:
‘Don’t do it Leito, don’t do it’. The analyst insists forcefully in her act. The child, seeing her refl
ection, dramatically calms down in an experience that the analyst understands as an integration
of the patient and herself.

In the same year, Thomas Ogden (1994b) published a paper with the same concept of
interpretative action. He explores new ways of transmitting meaning in analytic processes that
are experienced as futile. This should not surprise us given that affective regulation is based on
pre-verbal exchanges in adults as well. After all, projective identifi cation is based on actions that
transmit emotions in conso-nance with the origin of the word, i.e. to move.

Interpretative actions convey meaning, in a direct and holistic way. They are what Susanne
Langer described as non-discursive or presentational symbols. This kind of ‘symbol has an
immediate, idio-syncratic sensory reference’

(quoted in Rayner, 1995, p. 16). For example, bang, standing for the noise of a pistol.
Interpretative actions seem to have the capacity to restore to language the proper function it loses
when it is used to conceal meaning instead of disclosing it.

It is for analysts of the future to decide what interventions are properly analytic and how
projective identifi cation may be conceptualized most appropriately. In the meantime it is still a
useful concept in clinical practice. Our controversies may be our best means of bringing the
future of the concept to the fore.

363

Juan Francisco Jordan-Moore

Final commentary: news from the past

Edy Herrera (2000), trying to untangle the mystery of how projective identifi cation manages to
bring on psychic states in the analyst, thinks that it functions by inducing something akin to an
hypnotic state in the analyst. It is surprising to discover a reversal in the history of
psychoanalysis. From a psychoanalyst that hypnotizes his patients we have reached a situation in
which the patient is considered to hypnotize the analyst. This may be a consequence of the need
to negate the analyst’s infl uence on the process. The whole truth may be that the analytic couple
is frequently on the threshold of mutual hypnosis. Freud (1921) called this latter situation a ‘mass
of two’.

364

Afterword

Elizabeth Spillius and Edna O’Shaughnessy The idea of projective identifi cation, of which there
were forerun-ners in the writings of Freud and others, emerged as a distinct concept within the
language of psychoanalysis in Melanie Klein’s paper ‘Notes

on some schizoid mechanisms’ in 1946 (see Chapter 2 ). In this book

we chart the further evolution of the concept in the work of British Kleinians, and also its
acceptance, rejection and alteration by analysts of different psychoanalytic persuasions in
Britain, Europe, and in North and South America.

So widespread an interest in a new concept is unusual. It is due, we think, to the illumination

projective identifi cation can offer of central features of the clinical situation in the formation of
transference and countertransference between patient and analyst. In the future it will be
interesting to see what becomes of the concept as psychoanalysis develops further in Eastern
Europe and Russia, Africa, Southeast Asia and the Far East.

In our view, the concept of projective identifi cation is not particular to the clinical situation but a
universal of human communication, one that Freud was questing for. In 1915 in his paper ‘The
Unconscious’, he writes: ‘It is a very remarkable thing that the Ucs.

of one human being can react upon that of another, without passing through the Cs. This
deserves closer investigation’ (Freud, 1915b, p. 194). And later, in 1933, he again describes this
process.

There is, for example, the phenomenon of thought-transference

[. . .] It claims that mental processes in one person – ideas, emotional states, conative impulses –
can be transferred to another 365

Elizabeth Spillius and Edna O’Shaughnessy person through empty space without employing the
familiar methods of communication by means of words and signs.

(Freud, 1933, p. 39)

We think that the concept of projective identifi cation gives a name to, and a clarifi cation of, the
dynamics of direct communication and the phenomena of transference and countertransference
that are universal among humankind.

366

References

Abraham, K. (1908). Psycho-sexual differences between hysteria and Dementia Praecox. In

Abraham (1927).

Selected Papers on Psycho-Analysis . London: Hogarth Press.

Adler, G. (1989). Transitional phenomena, projective identifi cation, and the essential ambiguity
of the psychoanalytic situation. Psychoanalytic Quarterly , 58 , 81–104.

Almeida Prado, M. P. (1979). Identifi cação projetiva no processo analítico [Projective Identifi
cation in the Analytical Process]. Rio de Janeiro: José Fagundes do Amaral.

Amati Mehler, J. (2003). Identifi cazione proiettiva: fondamenti teorici

[Projective identifi cation: theoretical foundations]. Psicoanalisi , 7 (1), 48–83.

Ayuy, C., Feuerhake, O., Pola, A. and Rowe, J. (1997). Material teórico.

Contención, comunicación y resistencia en la transferencia: el abordaje desde la teoría de la

identifi cación proyectiva [Containment, Communication and Resistance in the Transference:
The Approach from the Theory of Projective Identifi cation]. Revista Chilena de Psicoanalisis ,
14 (2), 57–65.

Balint, M. (1937). Early developmental stages of the ego: primary object love.

In Balint (1965).

Primary Love and Psychoanalytic Technique . London: Tavistock, pp. 90–108.

Balint, M. (1950). Changing therapeutical aims and techniques in psychoanalysis. International

Journal of

Psychoanalysis , 31 , 117–124.

Balint, M. (1952).

Primary Love and Psychoanalytic Technique . New York: Liveright (1965).

Balint, M. (1968). The Basic Fault . London: Tavistock.

Baranger, M. (1993). The mind of the analyst: from listening to interpretation.

International Journal of Psychoanalysis , 74 , 15–24.

Baranger, M. (2005). Field theory. In S. Lewkowicz and S. Flechner (Eds.), Truth, Reality, and
the Psychoanalyst: Latin American Contributions to Psychoanalysis . London: IPA International
Psychoanalysis Library.

367

References

Baranger, M. and Baranger, W. (1961–1962). La situación analítica como campo dinámico [The
analytic situation as a dynamic fi eld]. Revista Uruguaya de Psicoanálisis , 4 (1), 3–54.

Baranger, W. and Baranger, M. (1966). Insight and the analytic situation.

In R. Litman (Ed.), Psychoanalysis in the Americas . New York: International Universities Press,
pp. 56–72.

Baranger, W., Baranger, M. and Mom, J. (1983). Process and non process in analytic work.
International Journal of Psychoanalysis , 64 , 1–15.
Bayle, G. (1996). Les clivages. Revue Française de Psychanalyse , 60 , 1303–1547.

Beebe, B. and Lachmann, F. (2005).

Infant Research and Adult Treatment:

Co-constructing Interactions . Hillsdale, NJ: Analytic Press.

Beland, H. (1992). Die zweifache Wurzel des Gefühls [The twofold root of emotions]. In
Jahrbuch der Psychoanalyse, 29 . Stuttgart-Bad Cannstatt: Frommann-holzboog, pp. 63–91.

Bell, D. (2001). Projective identifi cation. In C. Bronstein (Ed.), Kleinian Theory: A

Contemporary Perspective .

London: Whurr.

Bernardi, R. (2002). The need for true controversies in psychoanalysis.

International Journal of Psychoanalysis , 83 , 851–873.

Bick, E. (1968). The experience of the skin in early object-relations. International Journal of
Psychoanalysis , 49 , 484–486.

Bion, W. R. (1955a). Group dynamics: a review. In M. Klein, P. Heimann and R. Money-Kyrle

(Eds.), New Directions in Psycho-Analysis . London:

Tavistock, pp. 440–477. Also in Bion (1961). Experiences in Groups and Other Papers , London:
Tavistock, pp.

141–191.

Bion, W. R. (1955b). Language and the schizophrenic. In M. Klein, P. Heimann and R. Money-
Kyrle (Eds.), New Directions in Psycho-Analysis .

London: Tavistock, pp. 220–239.

Bion, W. R. (1956). Development of schizophrenic thought.

International

Journal of Psychoanalysis , 37 , 344–346.

Bion, W. R. (1957a). Differentiation of the psychotic from the non-psychotic personalities.

International Journal of Psychoanalysis , 38 (3–4), 266–275.

Bion, W. R. (1957b). On arrogance. In Bion (1967). Second Thoughts . New York: Jason
Aronson, pp. 86–92.

Bion, W. R. (1958). On hallucination. In Bion (1967). Second Thoughts . New York: Jason
Aronson, pp. 65–85.

Bion, W. R. (1959). Attacks on linking. International Journal of Psychoanalysis , 40 , 308–315.

Also in Bion (1967). Second Thoughts . London: Heinemann, pp. 93–109. Also published in E.
B. Spillius (Ed.) (1988). Melanie Klein Today: Developments in Theory and Practice . Volume 1:
Mainly Theory .

London: Routledge, pp. 87–101.

Bion, W. R. (1961). Experiences in Groups . London: Tavistock.

Bion, W. R. (1962a). Learning from Experience . London: Heinemann.

368

References

Bion, W. R. (1962b). A theory of thinking. In Bion (1967). Second Thoughts .

New York: Jason Aronson, pp. 110–119.

Bion, W. R. (1963). Elements of Psychoanalysis . London: Heinemann.

Bion, W. R. (1965). Transformations . London: Heinemann.

Bion, W. R. (1967). Second Thoughts . New York: Jason Aronson.

Bion, W. R. (1970). Attention and Interpretation . London: Tavistock.

Bion, W. R. (1977). Unpublished children’s hospital presentation. San Francisco, CA.

Birksted-Breen, D. (2003). Time and the après-coup. International Journal of Psychoanalysis , 84

, 1501–1515.

Blechner, M. (1994). Projective identifi cation, countertransference, and the

‘Maybe-me’. Contemporary Psychoanalysis , 30 , 619–630.

Bleger, J. (1967a). Psicoanálisis del encuadre psicoanalítico [Psychoanalysis of the

psychoanalytic setting]. Revista de Psicoanálisis, 24 , 241–258.

Bleger, J. (1967b). Simbiosis y ambigüedad [Symbiosis and Ambiguity]. Buenos Aires: Paidós.

Bollas, C. (1987). The Shadow of the Object: Psychoanalysis of the Unthought Known . London:
Free Association Books.

Bonnard, A. (1961). Über pathologische Funktionsweisen aus der Vorstufe des Körper-Ichs [On
pathological functioning of the precursor of the body-ego psychology]. Psyche: Zeitschrift für
Psychoanalyse und ihre
Anwendungen , 15 , 274–297.

Boris, H. (1988). Torment of the object: a contribution to the study of bulimia.

In H. J. Schwartz (Ed.),

Bulimia: Psychoanalytic Treatment and Theory .

Madison, CT: International Universities Press, pp. 89–110. Reprinted in Boris (1994). Sleights of
Mind: One and Multiples of One . Northvale, NJ: Jason Aronson, pp. 187–205.

Boris, H. (1993). Passions of the Mind: Unheard Melodies – A Third Principle of Mental
Functioning . New York: New York University Press.

Boris, H. (1994a). Envy . Northvale, NJ: Jason Aronson.

Boris, H. (1994b). Sleights of Mind: One and Multiples of One . Northvale, NJ: Jason Aronson.

Boyer, L. B. (1978). Countertransference experiences with severely regressed patients.

Contemporary Psychoanalysis , 14 , 48–71.

Boyer, L. B. (1986). Technical aspects of treating the regressed patient.

Contemporary Psychoanalysis , 22 , 25–44.

Boyer, L. B. (1989). Countertransference and technique in working with the regressed patient:
further remarks.

International Journal of Psychoanalysis , 70 , 701–714.

Boyer, L. B. (1990a). Countertransference and technique. In L. B. Boyer and P. Giovacchini

(Eds.), Master Clinicians on Treating the Regressed Patient .

Northvale, NJ: Jason Aronson, pp. 303–324.

369

References

Boyer, L. B. (1990b). Psychoanalytic intervention in treating the regressed patient. In L. B.

Boyer and P.

Giovacchini (Eds.),

Master Clinicians on

Treating the Regressed Patient . Northvale, NJ: Jason Aronson, pp. 1–32.

Brenman-Pick, I. (1985). Working through in the counter-transference.

International Journal of Psychoanalysis , 66 , 157–166.

Brierley, M. (1945). Further notes on the implications of psycho-analysis: metapsychology and

personology.

International Journal of Psychoanalysis , 26 , 89–114.

Brierley, M. (1947). Notes on psycho-analysis and integrative living. International Journal of

Psychoanalysis , 28 , 57–105.

Britton, R. (1998a).

Belief and Imagination: Explorations in Psychoanalysis .

London: Routledge.

Britton, R. (1998b). Introduction. In Britton, Belief and Imagination: Explorations in

Psychoanalysis . London: Routledge, pp. 1–7.

Brodey, W. (1965). On the dynamics of narcissism. I: Externalization and early ego

development. Psychoanalytic Study of the Child , 20 , 165–193.

Brunet, L. and Casoni, D. (1996). A review of the concepts of symbolization and projective
identifi cation in regards

[ sic ] to the patient’s use of the analyst.

Canadian Journal of Psychoanalysis , 4 , 109–127.

Brunet, L. and Casoni, D. (2001). A necessary illusion: projective identifi cation and containing
function. Canadian Journal of Psychoanalysis , 9 (2), 137–163.

Burke, W. and Tansey, M. (1985). Projective identifi cation and countertransference turmoil:
disruptions in the empathic process.

Contemporary

Psychoanalysis , 21 , 372–402.

Canestri, J. (1989). La Babel de l’inconscient: Langue maternelle, mangues étrangères et

psychanalyse . Paris: Presses Universitaires de France (PUF) [translated 1993

as

Babel of the Unconscious: Mother Tongue and Foreign Languages in the Psychoanalytic
Dimension . Madison, CT: International Universities Press].

Canestri, J. (2002). Projective identifi cation: the fate of the concept in Italy and Spain.
Psychoanalysis in Europe, Bulletin 56 , 130–139.
Caper, R. (1988). Immaterial Facts . New York: Jason Aronson.

Caper, R. (1999). A Mind of One’s Own . London: Routledge.

Carloni, G. (1983). Con Melanie Klein e Anna Freud. Rivista Psicoanalisis , 29 , 429–434.

Carpy, D. V. (1989). Tolerating the countertransference: a mutative process.

International Journal of Psychoanalysis , 70 , 287–294.

Carveth, D. (1992). Dead end kids: projective identifi cation and sacrifi ce in Orphans [a play
and fi lm].

International Review of Psychoanalysis , 19 ,

217–227.

Casaula, E., Coloma, J., Colzani, F. and Jordan, J. F. (1994). La bilógica de la interpretación.
Revista Chilena de Psicoanálisis .

Coloma, J. (2002). Bilógica, identifi cación proyectiva y pensamiento onírico

[Biological relevance, projective identifi cation and dream thought]. Gradiva , 3 (1), 9–18.

370

References

Corti, A. (1983). Melanie Klein e Anna Freud: riunione commemorativa nell’ambito del
Convegno a Seminari multipli, Bologna 22.5.83. Rivista di Psicoanalisi , 29 (4).

Crisp, P. (1986). Projective identifi cation: an attempt at clarifi cation. Journal of the Melanie
Klein Society , 4 (1), 47–76.

Cycon, R. (1995).
Vorwort zur Gesamtausgabe der Werke von Melanie Klein

[Foreword to the Complete Works of Melanie Klein]. Stuttgart: Frommann-holzboog.

Danckwardt, J. F. (2001). Die Hilfl osigkeit des Unbewussten und die Prozessidentifi zierungen
als Arbeitsebene bei schweren Konfl ikten. In W. Bohlheber and S. Drews (Eds.), Die
Gegenwart der Psychoanalyse – die Psychoanalyse der Gegenwart [The Presence of
Psychoanalysis – Psychoanalysis of the Present]. Stuttgart: Klett-Cotta, pp. 409–423.

Dantlgraber, J. (1982). Bemerkungen zur subjektiven Indikation für Psychoanalyse [Comments

on the subjective indication for psychoanalysis].

Psyche: Zeitschrift für Psychoanalyse und ihre Anwendungen , 36 , 193–225.

de M’Uzan, M. (1994).

La Bouche de l’inconscient [The Mouth of the

Unconscious]. Paris: Gallimard.

de Saussure, R. (1929). Les fi xations homosexuelles chez les femmes névrosées

[Homosexual fi xations in neurotic women]. Revue Française de Psychanalyse , 3 , 50–91.

Diatkine, G. (1997). Jacques Lacan . Paris: Presses Universitaires de France (PUF).

Di Chiara, G. (1983). La fi aba della mano verde o dell’identifi cazione proiettiva [The tale of the
green hand or projective identifi cation]. Rivista di Psicoanalisi , 29 (4), 459–475.

Di Chiara, G. and Flegenheimer, F. (1985). Identifi cazione proiettiva, Nota Storico-Critica.

Rivista di Psicoanalisi , 31 (2), 233–243.

Dornes, M. (1993). Psychoanalyse und Kleinkindforschung [Psychoanalysis and infant research].

Psyche: Zeitschrift für Psychoanalyse und ihre Anwendungen 12 , 1116–1152.

Dorpat, T. (1983). Review of ‘Splitting and Projective Identifi cation’, by J. Grotstein.

International Journal of Psychoanalysis , 64 , 116–119.

Duparc, F. (2001). The countertransference scene in France.

International

Journal of Psychoanalysis , 81 , 151–169.

Eagle, M. N. (2000). A critical evaluation of current conceptions of transference and

countertransference.

Psychoanalytic Psychology , 17 , 24–37.

Eaton, J. (2005). The obstructive object. Psychoanalytic Review , 92 , 355–372.

Eiguer, A. (1989). Le Pervers narcissique et son complice [The Narcissistic Pervert and his
Accomplice]. Paris:

Dunod.

Erikson, E. H. (1959). Identity and the Life Cycle . New York: International Universities Press.

Eskelinen de Folch, T. (1987). The obstacles to the analytic cure: comments on ‘Analysis
Terminable and Interminable’. In J. Sandler (Ed.), On Freud’s 371

References

‘Analysis Terminable and Interminable’ . New Haven, CT: Yale University Press, pp. 111–126.

Etchegoyen, R. H. (1986).

The Fundamentals of Psychoanalytic Technique .

London: Karnac.

Fain, M. (1984). De l’identifi cation hystérique et de l’identifi cation projective

[On hysterical identifi cation and projective identifi cation]. Revue Française de Psychanalyse ,
48 , 784.

Fairbairn, W. R. D. (1941). A revised psychopathology of the psychoses and neuroses.

International Journal of Psychoanalysis , 22 , 250–279.

Fairbairn, W. R. D. (1944). Endopsychic structure considered in terms of object-relationships.

International Journal of Psychoanalysis , 27 , 70–93.

Fairbairn, W. R. D. (1946). Object-relationships and dynamic structure.

International Journal of Psychoanalysis , 27 , 30–37.

Fairbairn, W. R. D. (1952). Psychoanalytic Studies of the Personality . London: Tavistock Also

published as Fairbairn (1954). An Object-Relations Theory of the Personality . New York: Basic
Books.

Fairbairn, W. R. D. (1958). On the nature and aims of psychoanalytical treatment. International

Journal of Psychoanalysis , 39 , 374–385.

Feinsilver, D. (1983). Reality, transitional relatedness and containment in the borderline.

Contemporary Psychoanalysis , 19 , 537–569.

Feldman, M. (1992). Splitting and projective identifi cation. In R. Anderson (Ed.), Clinical
Lectures on Klein and Bion . London: Routledge, pp. 74–88.
Feldman, M. (1994). Projective identifi cation in phantasy and enactment.

Psychoanalytic Inquiry , 14 , 423–440.

Feldman, M. (1997). Projective identifi cation: the analyst’s involvement.

International Journal of Psycho-Analysis , 78 , 227–241.

Ferenczi, S. (1909). Introjection and transference. In Ferenczi, Contributions to Psychoanalysis .

Boston, MA: R. G.

Badger.

Ferenczi, S. (1930). Notes and fragments, II. In Ferenczi (1994).

Final

Contributions to the Problems and Methods of Psycho-Analysis . London: Karnac, pp. 219–231.

Ferro, A. (1987). Il mondo alla rovescia: l’inversione di fl usso delle identifi ca-zioni proiettive
[The world turned upside down: inversion of the fl ow of projective identifi cation]. Rivista di
Psicoanalisi , 33 (1), 59–77.

Figueroa, J. (1997). Resumen de los grupos de discusión del material teórico.

Revista Chilena de Psicoanalisis , 14 (2), 66–67.

Finell, J. S. (1986). The merits and problems with the concept of projective identifi cation.
Psychoanalytic Review , 73 , 103–128.

Fonagy, P. (1991). Thinking about thinking: some clinical and theoretical considerations in the
treatment of a borderline patient. International Journal of Psycho-Analysis , 72 , 1–18.

Frank, C. (2001). Vorlesungsreihe zur projektiven Identifi zierung, bisher unveröffentlichtes

[Lectures about projective identifi cation]. Unpublished manuscript.

372

References

Frank, C. (2002). Identifi ed with the bombing and dangerous Hitler-father (M. Klein, 1945, p.
377) Zu Melanie Kleins zeitgenössischer Bezugnahme auf Hitler in Ihren Behandlungen.
Unpublished manuscript.

Freud, A. (1936). The Ego and the Mechanisms of Defence . London: Hogarth Press (1987).

Freud, A. (1967). About losing and being lost. Psychoanalytic Study of the Child , 8 , 9–19.

Freud, S. (1898). Die infantile Cerebrallähmung. Jahrbuch Leist. Neurol , 1 , 613

[translated 1968 as Infantile Cerebral Paralysis . Coral Gables, FL: University of Miami Press].

Freud, S. (1900). The Interpretation of Dreams . In J. Strachey (Ed.), The Standard Edition of the
Complete Psychological Works of Sigmund Freud . London: Hogarth Press: S.E. 4–5 .

Freud, S. (1910). The future prospects of psycho-analytic therapy. In S.E. 9 , pp. 144–145.

Freud, S. (1911). Psycho-analytic notes on an autobiographical account of a case of paranoia

(Dementia Paranoides).

S.E. 12: 1–79.

Freud, S. (1914). Remembering, repeating, and working through. In S.E. 12 , pp. 145–156.

Freud, S. (1915a). Instincts and their vicissitudes. In S.E. 14 , pp. 109–140.

Freud, S. (1915b). The unconscious. In S.E. 14 , pp. 159–215.

Freud, S. (1917). Mourning and melancholia. In S.E. 14 , pp. 237–259.

Freud, S. (1918). From the history of an infantile neurosis. In S.E. 17 , pp. 3–122.

Freud, S. (1920). Beyond the pleasure principle. In S.E. 18 , pp. 3–64.

Freud, S. (1921). Group psychology and the analysis of the ego. In S.E. 18 , pp. 67–143.

Freud, S. (1923). The Ego and the Id. In S.E. 19 , pp. 12–66.

Freud, S. (1933).

New Introductory Lectures on Psychoanalysis . In S.E. 22 , pp. 1–182.

Frosch, J. P. (1990). Review of Michael Tansey and Walter Burke,

‘Understanding Countertransference: From Projective Identifi cation to Empathy’, 1989,

Hillsdale, New Jersey: The Analytic Press. International Journal of Psychoanalysis , 71 , 351–
354.

Gabbard, G. (1995). Countertransference: the emerging common ground.

International Journal of Psychoanalysis , 76 , 475–485.

Gammill, J. (1989). Some personal refl ections on Melanie Klein. Melanie Klein and Object
Relations , 7 (2), 1–15.

Garfi nkle, E. (2005). Towards clarity in the concept of projective identifi cation: a review and a
proposal (Part 1).

Defi ning projective identifi cation as an intrapsychic unconscious phantasy.

Canadian Journal of Psychoanalysis ,

13 ,

202–229.

Gibeault, A. (2000a). De la projection et de l’identifi cation projective

[Projection and projective identifi cation]. Revue Française de Psychanalyse , 64 , 723–742.

373

References

Gibeault, A. (2000b). In response to Otto F. Kernberg’s ‘Psychoanalysis, psychoanalytic

psychotherapy and supportive psychotherapy: contemporary controversies’. International
Journal of Psychoanalysis , 81 , 379–383.

Gilhooley, D. (1998). Projection and projective identifi cation in a three-year-old boy.

Psychoanalytic Review , 92 , 355–372.

Glover, E. (1931). The therapeutic effect of inexact interpretation. International Journal of

Psychoanalysis , 12 , 397–411.

Goldman, H. A. (1988). Paradise destroyed, the crime of being born: a psychoanalytic study of
the experience of evil. Contemporary Psychoanalysis , 24 , 420–450.

Green, A. (1980). La mère morte. In Green (1983). Narcissisme de vie, narcis-sisme de mort .
Paris: Minuit

[translated 1986 as The dead mother. In Green, On Private Madness . London: Hogarth,
reprinted 1997 by Karnac, pp. 142–173].

Green, A. (1990). La Folie privée . Paris: Gallimard.

Greenson, R. R. (1954). The struggle against identifi cation. Journal of American Psychoanalytic
Association , 2 , 200–217.

Grinberg, L. (1958). As pectos mágicos en la transferencia y en la contratransferencia. Revista de

Psicoanálisis , 15 , 15–26.

Grinberg, L. (1962). On a specifi c aspect of countertransference due to the patient’s projective

identifi cation.

International Journal of Psychoanalysis , 43 , 436–440.

Grinberg, L. (1979). Countertransference and projective counteridentifi cation.

In L. Epstein and A. H. Feiner (Eds.), Countertransference . New York: Jason Aronson, pp. 169–
191.

Grinberg, L. (1982). Los afectos en la contratransferencia: Más allá de la contraidentifi cación

proyectiva. XIV

Congreso de FEPAL. Actas , 258–259.

Grinberg, L. (1985). Teoría de la Identifi cación [Identifi cation Theory]. Madrid:

Tecnipublicaciones.

Grinberg, L., Sor, D. and de Bianchedi, E. T. (1977). Introduction to the Work of Bion . New
York: Jason Aronson.

Grotstein, J. S. (1981). Splitting and Projective Identifi cation . New York: Jason Aronson.

Grotstein, J. S. (1982). Newer perspectives in object relations theory.

Contemporary Psychoanalysis , 18 , 43–91.

Grotstein, J. S. (1994a). Projective identifi cation and countertransference: a brief commentary

on their relationship.

Contemporary Psychoanalysis ,

30 ,

578–592.

Grotstein, J. S. (1994b). Projective identifi cation re-appraised. Part 1: Projective identifi cation,
introjective identifi cation, the transference/countertransference neurosis/psychosis, and their
consummate expression in the cruci-fi xion, the Pietà, and ‘therapeutic exorcism’. Contemporary
Psychoanalysis , 30 , 708–746.

374

References

Grotstein, J. S. (1995). Projective identifi cation re-appraised. Part II: The countertransference
complex.

Contemporary Psychoanalysis , 31 , 479–511.

Grotstein, J. S. (1997). Integrating one-person and two-person psychologies: autochthony and

alterity in counterpoint. Psychoanalytic Quarterly , 66 , 403–430.

Grotstein, J. S. (2001). A rationale for the psychoanalytically informed psychotherapy of

schizophrenia and other psychoses: towards the concept of ‘reha-bilitative psychoanalysis’. In P.
Williams (Ed.), A Language for Psychosis .
London: Whurr.

Grotstein, J. S. (2002). Projective identifi cation and its relationship to infant development. In S.
Alhanati (Ed.), Primitive Mental States , Volume 2:

Psychobiological and Psychoanalytic Perspectives on Early Transference and Personality

Development . London: Karnac, pp. 67–98.

Grotstein, J. S. (2005). ‘Projective

trans identifi cation’: an extension of the

concept of projective identifi cation. International Journal of Psychoanalysis , 86 , 1051–1069.

Guignard, F. (1984). Identifi cation hystérique et identifi cation projective

[Hysterical identifi cation and projective identifi cation]. Revue Française de Psychanalyse , 48 ,
515–527.

Guignard, F. (1996). Un trouble de la pensée sur la métapsychologie

[Disturbance of thought on metapsychology]. Revue Française de Psychanalyse , 60 , 1551–

1571.

Guignard, F. (1997).

Épître à l’objet [Epistle to the Object]. Paris: Presses Universitaires de France (PUF).

Gutwinski-Jeggle, J. (1995). Zum Verhältnis von Gegenübertragung und projektiver Identifi

zierung [On the relationship between counter-transference and projective identifi cation].
Luzifer-Amor: Zeitschrift zur Geschichte der Psychoanalyse [ Journal of the History of
Psychoanalysis ], 15 , 61–83.

Haber, M. and Godfrind-Haber, J. (2002). L’expérience agie partagée [The shared enacted
experience]. Revue Française de Psychanalyse , 66 , 1417–1460.

Hamilton, N. G. (1986). Positive projective identifi cation. International Journal of

Psychoanalysis , 67 , 489–496.

Hamilton, N. G. (1990). The containing function and the analyst’s projective identifi cation.
International Journal of Psychoanalysis , 71 , 445–453.

Heidegger, M. (1927).

Being and Time , translated by Macquarrie, J. and Robinson, E. Oxford: Blackwell (1962).

Heimann, P. (1942). A contribution to the problem of sublimation and its relation to the
processes of internalization.
International Journal of Psychoanalysis , 23 , 8–17.

Heimann, P. (1950). On countertransference. International Journal of PsychoAnalysis , 31 , 81–

84.

Heimann, P. (1966). Comment on Dr Kernberg’s paper ‘Structural derivatives of object

relationships’. International Journal of Psycho-Analysis , 47 , 254–260.

Also published in M. Tonnesmann (Ed.) (1989).

About Children and

375

References

Children-No-Longer: Collected Papers of Paula Heimann, 1942–80 . London: Routledge, pp.

218–230.

Heimann, P. (1969). Postscript to ‘Dynamics of transference interpretation’

(1955/6). In M. Tonnesmann (Ed.) (1989). About Children and Children-No-Longer: Collected

Papers of Paula Heimann, 1942–80 . London: Routledge, pp. 252–261.

Herrera, E. (2000). Identifi cación proyectiva: el desarrollo de un concepto

[Projective identifi cation: development of a concept]. Revista Chilena de Psicoanalisis , 17 (1),

32–41.

Hinshelwood, R. D. (1991). Entry on ‘Projective identifi cation’. In Hinshelwood, A Dictionary

of Kleinian Thought

, 2nd edn. London: Free Association Books, pp. 179–208.

Hinshelwood, R. D. (1999). Countertransference.

International Journal of

Psychoanalysis , 80 , 797–818.

Hinshelwood, R. D. (2000).

Dictionnaire de la pensée kleinienne

[Dictionary of Kleinian Thought]. Paris: Presses Universitaires de France (PUF).

Hinz, H. (1989). Projektive Identifi zierung und psychoanalytischer Dialog

[Projective identifi cation and psychoanalytical dialogue]. Psyche: Zeitschrift für Psychoanalyse
und ihre Anwendungen , 43 , 609–631.

Hinz, H. (2002a). Projective identifi cation: the fate of the concept in Germany.

Psychoanalysis in Europe, Bulletin 56 , 118–129.

Hinz, H. (2002b). Wer nicht verwickelt wird, spielt keine Rolle. Zu Money-Kyrle: ‘Normale
Gegenübertragung und mögliche Abweichungen [Who is not involved does not matter. To
Money-Kyrle: ‘Normal countertransference and possible deviations’]. In C. Frank, L. Hermanns
and H. Hinz (Eds.), Jahrbuch der Psychoanalyse 44 . Stuttgart-Bad Cannstat: Frommann-
holzboog, pp. 197–223.

Hoffman, I. Z. (1983). The patient as interpreter of the analyst’s experience.

Contemporary Psychoanalysis , 19 , 389–422.

Isaacs, S. (1948). Un caso de ansiedad psicótica aguda en un niño de 4 años [An acute psychotic
anxiety occurring in a boy of four years, originally published in English in 1938]. Revista de
Psicoanálisis , 6 , 433–483.

Isaacs, S. (1949). Naturaleza y función de la fantasía [The nature and function of phantasy,
originally published in English in 1943]. Revista de Psicoanálisis , 7 , 555–609.

Isaacs, S. (1952). The nature and function of phantasy. In M. Klein, P.

Heimann, S. Isaacs and J. Riviere (Eds.), Developments in Psychoanalysis .

London: Hogarth Press, pp. 68–121.

Issacharoff, A. and Hunt, W. (1994). Transference and projective identifi cation. Contemporary
Psychoanalysis , 30

, 593–603.

Jackson, M. and Williams, P. (1994). Unimaginable Storms: A Search for Meaning in Psychosis .
London: Karnac.

376

References

Jacobson, E. (1954). Contribution to the metapsychology of psychotic identifi cations. Journal of

the American Psychoanalytical Association , 2 ,

239–262.

Jacobson, E. (1964). The Self and the Object World . New York: International Universities Press.

Jacobson, E. (1967). Psychotic Confl ict and Reality . New York: International Universities
Press.

Jaffe, D. (1968). The mechanism of projection: its dual role in object relations.

International Journal of Psychoanalysis , 49 , 662–677.

Jiménez, J. P. (1992). La contribución del analista en los procesos de identifi -

cación proyectiva [The contribution of the analyst to processes of projective identifi cation].
Revista Chilena de Psicoanalisis , 9 (2), 54–66.

Jiménez, J. P. (1995). Sobre identifi catión proyectiva. Revista de Psicoanálisis de Madrid , 5 ,

111–123.

Jordan, J. F. (2005). External reality/internal reality: a real dichotomy? In S. Lewkowicz and S.

Flechner (Eds.), Truth, Reality, and the Psychoanalyst: Latin American Contributions to
Psychoanalysis . London: IPA International Psychoanalysis Library.

Joseph, B. (1961). Über einige Persönlichkeitsmerkmale des Psychopathen

[Some characteristics of the psychopathic personality]. Psyche: Zeitschrift für Psychoanalyse und
ihre Anwendungen

, 15 , 132–141.

Joseph, B. (1978). Different types of anxiety and their handling in the analytic situation. In M.
Feldman and E.

Spillius (Eds.) (1989). Psychic Equilibrium and Psychic Change: Selected Papers of Betty Joseph
. London: Routledge, pp. 105–114.

Joseph, B. (1987). Projective identifi cation: some clinical aspects. In J. Sandler (Ed.),
Projection, Identifi cation, Projective Identifi

cation . London: Karnac,

pp. 65–76. Also published in E. Spillius (Ed.) (1988). Melanie Klein Today: Developments in
Theory and Practice .

Volume 1: Mainly Theory . London:

Routledge, pp. 138–150. Also published in M. Feldman and E. Spillius (Eds.) (1989). Psychic
Equilibrium and Psychic Change: Selected Papers of Betty Joseph . London: Routledge, pp. 169–
180.

Joseph, B. (1988). Object relations in clinical practice. Psychoanalytic Quarterly , 57 , 626–642.

Also published in M. Feldman and E. Spillius (Eds.) (1989).

Psychic Equilibrium and Psychic Change: Selected Papers of Betty Joseph . London: Routledge,
pp. 203–214.

Joseph, B. (1989a). Psychic change and the psychoanalytic process. In M. Feldman and E. B.
Spillius (Eds.), Psychic Equilibrium and Psychic Change: Selected Papers of Betty Joseph .
London: Routledge, pp. 192–202.

Joseph, B. (1989b). Psychic Equilibrium and Psychic Change: Selected Papers of Betty Joseph ,
edited by M.

Feldman and E. B. Spillius. London: Routledge.

Joseph, B. (2001). Zur Übertragungsliebe: Aktuelle Überlegungen [About transference-love:

recent refl ections]. In Freud heute: Wendepunkte und

377

References

Streitfragen , Band 3: Über Freuds, Bemerkungen über die Übertragungsliebe [Freud Today:
Turning Points and

Controversies, Volume 3: On Freud’s

‘Observations on Transference-Love’]. Stuttgart-Bad Cannstatt: Frommann-holzboog, pp. 129–

148.

Karlsson, R. (2004). Collusion as interactive resistances and possible stepping-stones out of

impasses.

Psychoanalytic Psychology , 21 , 567–579.

Kennedy, R. (1993). Freedom to Relate . London: Free Association Books.

Kernberg, O. F. (1968). The treatment of patients with borderline personality organization.

International Journal of Psychoanalysis , 49 , 600–619.

Kernberg, O. F. (1975). Borderline Conditions and Pathological Narcissism . New York: Jason
Aronson.

Kernberg, O. F. (1976). Normal and pathological development. In Kernberg, Object-Relations

Theory and Clinical Psychoanalysis . New York: Jason Aronson.

Kernberg, O. F. (1980). Internal World and External Reality . New York: Jason Aronson.

Kernberg, O. F. (1984). Severe Personality Disorders: Psychotherapeutic Strategies .

New Haven, CT: Yale University Press.

Kernberg, O. F. (1986). Identifi cation and its vicissitudes as observed in psychosis.

International Journal of Psycho-Analysis , 67 , 147–158.

Kernberg, O. F. (1987). Projective identifi cation: developmental and clinical aspects. In J.

Sandler (Ed.), Projection, Identifi cation, Projective Identifi cation .

Madison, CT: International Universities Press, pp. 93–115. Also published in 1988, London:
Karnac. Also published in 1987 in Journal of the American Psychoanalytic Association , 35 ,
795–819.

Kernberg, O. F. (1989). Psychodynamic Psychotherapy of Borderline Patients . New York:

Basic Books.

Kernberg, O. F. (1992). Aggression in Personality Disorders and Perversions . New Haven, CT:
Yale University Press.

King, P. (1974). On a patient’s unconscious need to have ‘bad parents’. In King, Time Present
and Time Past: Selected Papers of Pearl King . London: Karnac.

King, P. (1978). Affective response of the analyst to the patient’s communications. International
Journal of Psychoanalysis , 59 , 329–334.

King, P. (2004). What has happened to psychoanalysis in the British Society?

In A. Casement (Ed.), Who Owns Psychoanalysis? London: Karnac.

King, P. and Steiner, R. (1991).

The Freud-Klein Controversies: 1941–45 .

London: Routledge.

Klein, M.

Note. In 1975 Klein’s writings were brought together as The Writings of Melanie Klein , divided
into four volumes: Volume 1, (1975) Love, Guilt and Reparation and Other Works, 1921–1945 ;
Volume 2, (1975) (Originally published in 1932) The Psycho-Analysis of Children ; Volume 3,
(1975) Envy and Gratitude and Other Works,

1946–1963 ; Volume 4, (1980) Narrative of a Child Analysis . All 4 volumes were published by
The Hogarth Press and The Institute 378

References

of Psycho-Analysis, London. The Klein References in the present book are all to be found in
Volumes 1, 2, and 3.

Because the general contents of Vols 1, 2, and 3 are listed in this note, the specifi c content of
each Klein reference will be noted below in the title of the reference, but the general content of
the volume cited can be seen by checking its description in this note.
Klein, M. (1921) The development of a child. In Writings (1975), Volume 1, pp. 1–53.

Klein, M. (1928) Early stages of the Oedipus confl ict. International Journal of Psycho-Analysis,
9 , 167–180. Also in Klein, M. Writings (1975) Vol. 1, pp. 186–198.

Klein, M. (1932) The Psycho-Analysis of Children, Writings (1975) Vol. 2.

Klein, M. (1933) The early development of conscience in the child. In Writings , (1975) Vol. 1,
pp. 248–257.

Klein, M. (1935) A contribution to the psychogenesis of manic-depressive states. In Writings

(1975) Vol. 1, pp.

262–289.

Klein, M. (1940) Mourning and its relation to manic-depressive states. In Writings (1975) Vol. 1,
pp. 344–369.

Klein, M. (1943). Primeros estadios del confl icto de Edipo y de la formación del superyó [Early
stages of the Oedipus confl ict, originally published in English in 1928]. Revista de Psicoanálisis
, 1 (1), 83–110.

Klein, M. (1945) The Oedipus complex in the light of early anxieties. Writings , (1975) Vol. 1,
370–419.

Klein, M. (1946) Notes on some schizoid mechanisms. International Journal of Psycho-Analysis,

27 : 99–110. Also published in 1952a in Klein, M. Writings , (1975) Vol. 3, 1–24.

Klein, M. (1947a). Contribución a la psicogénesis de los estados maníaco-depresivos [A

contribution to the psychogenesis of manic-depressive states, originally published in English in
1935].

Revista de Psicoanálisis ,

4 (3),

508–539.

Klein, M. (1947b). El papel de la escuela en el desarrollo libidinoso del niño

[The role of the school in the libidinal development of the child, originally published in English
in 1923]. Revista de Psicoanálisis , 5 (2), 480–500.

Klein, M. (1948a) A contribution to the theory of anxiety and guilt, International Journal of
Psycho-Analysis, 29 , 114–123. Also in Writings (1975) Vol. 3, as On the theory of anxiety and
guilt, pp. 25–43.

Klein, M. (1948b). Notas sobre algunos mecanismos esquizoides [Notes on some schizoid
mechanisms, originally published in English in 1946]. Revista de Psicoanálisis , 6 (1), 82–113.
Klein, M. (1949). El duelo y su relación con los estados maníaco-depresivos

[Mourning and its relation to manic-depressive states, originally published in English in 1940].
Revista de Psicoanálisis, 7 (3), 415–449.

Klein, M. (1952a) Notes on some schizoid mechanisms. In Klein, M. Writings, (1975) Vol. 3,
pp. 1–24. (Already noted above under Klein, M. 1946.) 379

References

Klein, M. (1952b) The origins of transference. In Klein, M. Writings , (1975) Vol. 3, pp. 48–56.

Klein, M. (1952c) Theoretische Betrachtungen Uber das Gefuhlsleben des Sauglings (Some
theorectical conclusions regarding the life of the infant. In R. Cycon and H. Erb (Eds .)
Gesammelte schriften Melanie Klein 1995–2002) .

Volume III : Das Seelenleben des Kleinkindes . (The Mental Life of the Infant.) Stuttgart: Klett-
Cotta.

Klein, M. (1955) On identifi cation. In Klein, M. Writings , (1975) Vol. 3, pp. 141–175.

Klein, M,. (1957) Envy and Gratitude. In Klein, M. Writings,(1975) Vol. 3, pp. 176–235.

Klein, M., Heimann, P., Isaacs, S. and Riviere, J. (1952) Edited by Joan Riviere. Developments
in Psycho-Analysis , London: The Hogarth Press and The Institute of Psychoanalysis.

Klüwer, R. (1995). Agieren und Mitagieren – zehn Jahre später [Acting and co-acting – ten years
later]. Zeitschrift für Psychoanalytische Theorie und Praxis , 10 , 45–70.

Knight, R. P. (1940). Introjection, projection and identifi cation. Psychoanalytic Quarterly , 9 ,

334–341.

Kohon, G. (1986). The British School of Psychoanalysis: The Independent Tradition .

London: Free Association Books.

Krafft-Ebing, R. von (1886). Psychopathia sexualis: Eine klinisch-forensische Studie

[Aberrations of Sexual Life]. Stuttgart: Ferdinand Enke.

La Farge, L. (1989). Emptiness as defense in severe regressed states. Journal of the American
Psychoanalytic Association , 37 , 965–995.

Langs, R. (1975). Therapeutic misalliances. International Journal of Psychoanalysis , 45 , 77–

105.

Langs, R. (1976). The Therapeutic Interaction . New York: Jason Aronson.

Langs, R. (1978a). The adaptational-interactional dimension of countertransference.

Contemporary Psychoanalysis , 14 , 502–533.

Langs, R. (1978b). Some communicative properties of the bipersonal fi eld. International

Journal of Psychoanalytic Psychotherapy , 7 , 89–161.

Also published in Grotstein, J. (Ed.) (1981). Do I Dare Disturb the Universe: A Memorial to
Wilfred R. Bion .

Beverly Hills, CA: Caesura Press, pp.

441–487.

Langs, R. (1978c). Technique in Transition . New York: Jason Aronson.

Langs, R. (1979). Interventions in the bipersonal fi eld. Contemporary Psychoanalysis , 15 , 1–

54.

Laplanche, J. (1992). Faut-il brûler Melanie Klein? In Laplanche, La Révolution Copernicienne

inachevée [The Unfi nished Copernican Revolution]. Paris: Aubier, pp. 213–226.

Laplanche, J. and Pontalis, J.-B. (1973). The Language of Psychoanalysis . London: Hogarth
Press.

380

References

Liberman, D. (1956). Identifi cación proyectiva y confl icto matrimonial

[Projective identifi cation and matrimonial confl ict]. Revista de Psicoanálisis , 13 , 1–20.

Lichtenstein, H. (1961). Identity and sexuality: a study of their interrelation-ship in man. Journal
of the American Psychoanalytic Association , 9 , 179–260.

Lichtman, A. (1979). Consideraciones acerca de la transferencia y los mecanismos psicóticos en

Freud y Melanie Klein. Revista de Psicoanálisis , 1 , 55–68.

Little, M. (1966). Transference in borderline states.

International Journal of

Psychoanalysis , 47 , 476–485.

Loch, W. (1962). Anmerkungen zur Pathogenese und Metapsychologie einer schizophrenen

Psychose [Remarks on the pathogenesis and metapsychology of a schizophrenic psychosis].
Psyche: Zeitschrift für Psychoanalyse und ihre Anwendungen , 15 , 684–720.

Loch, W. (1965). Voraussetzungen, Mechanismen und Grenzen des psychoanalyt-ischen

Prozesses [Preconditions, Mechanisms and Limits of the Psychoanalytical Process]. Stuttgart:
Klett-Cotta.

Loch, W. (2001).

Mit Freud über Freud hinaus: Ausgewählte Vorlesungen zur Psychoanalyse [With Freud and
Beyond Freud: Selected Lectures on Psychoanalysis], edited by J. Dantlgraber and W. Damson.
Tübingen: Diskord.

Loewald, H. W. (1960). On the therapeutic action of psycho-analysis.

International Journal of Psychoanalysis , 41 , 16–33.

Mahler, M. S. (1952). On child psychosis and schizophrenia: autistic and symbiotic infantile
psychoses.

Psychoanalytic Study of the Child , 7 , 286–305.

Mahler, M. S. (1967). On human symbiosis and the vicissitudes of individuation. Journal of the
American Psychoanalytical Association , 15 (4), 740–763.

Malin, A. and Grotstein, J. S. (1966). Projective identifi cation in the therapeutic process.
International Journal of Psychoanalysis , 47 , 26–31.

Mancia, M. (1996). Imitazione, rappresentazione, identifi cazione: loro ruolo nello sviluppo e nel
transfert

[Imitation, representation, identifi cation: their role in the development and transfer]. Rivista di
Psicoanalisi , 42 (2), 225–247.

Manfredi Turilazzi, S. (1974). Dalle interpretazioni mutative di Strachey alle interpretazioni

delle relazioni tra gli oggetti interni [Strachey’s mutative interpretations from interpretations of
the relationship between internal objects].

Rivista di Psicoanalisi , 20 , 127–143.

Manfredi Turilazzi, S. (1984a). L’Unicorno: saggio sulla fantasia e l’oggetto nel concetto di
identifi cazione proiettiva. Rivista di Psicoanalisi , 31 (4), 462–477.

Manfredi Turilazzi, S. (1984b). Sull’identifi cazione proiettiva: uno studio preliminare. Letto al
Centro Milanese di Psicoanalisi .

Manzano, J. and Palacio Espasa, F. (1986). Etude sur la psychose infantile [Study of Infantile
Psychosis]. Lyon: Simed.

Manzano, J., Palacio Espasa, F. and Zilkha, N. (1999). Les Scénarios narcissiques de la
parentalité [Narcissistic Scenarios of Parenthood]. Paris: Presses Universitaires de France (PUF).

381
References

Martindale, B. (2001). New discoveries concerning psychoses and their organizational fate. In P.
Williams (Ed.), A Language for Psychosis . London:

Whurr, pp. 27–36.

Mason, A. (1981). The suffocating superego: psychotic break and claustrophobia. In J. Grotstein
(Ed.), Do I Dare Disturb the Universe? London: Karnac, pp. 139–166.

Mason, A. (1994). A psychoanalyst looks at a hypnotist: a study of folie à deux.

Psychoanalytic Quarterly , 63 , 641–679.

Mason, A. (1997). Quick Otto and slow leopard: the Freud–Fliess relationship.

Paper presented to the Fortieth International Psychoanalytic Congress, Barcelona, Spain.

Massidda, G. B. (1999). Shall we ever know the whole truth about projective identifi cation?
International Journal of Psychoanalysis , 80 , 365–367.

Matte-Blanco, I. (1998). Thinking, Feeling and Being: Clinical Refl ections of the Fundamental
Antinomy of Human Beings and World . London: Routledge.

Meissner, W. W. (1980). A note on projective identifi cation. Journal of the American

Psychoanalytic Association , 28 , 43–66.

Meissner, W. W. (1987). Projection and projective identifi cation. In J. Sandler (Ed.), Projection,
Identifi cation, Projective Identifi cation . Madison, CT: International Universities Press, pp. 27–
49.

Melsohn, I. (2002). A psicanálise em nova chave [Psychoanalysis in a New Key].

São Paulo: Perspectiva.

Meltzer, D. (1966). The relation of anal masturbation to projective identifi cation. International
Journal of Psychoanalysis , 47 , 335–342.

Meltzer, D. (1967). The Psycho-Analytical Process . London: Heinemann.

Meltzer, D. (1968). The introjective basis of polymorphous tendencies in adult sexuality. In

Sexual States of Mind (pp. 64–73). Strath Tay: Clunie Press.

Meyer, L. (1989). Editorial. Revista Brasileira de Psicanálise , 23 (3), 364.

Migone, P. (1995). Expressed emotion and projective identifi cation: a bridge between
psychiatric and psychoanalytic concepts? Contemporary Psychoanalysis , 31 , 617–640.

Milton, J., Polmear, C. and Fabricius, J. (2004).

A Short Introduction to

Psychoanalysis . London: Sage.

Mitchell, S. (2000). Relationality: From Attachment to Intersubjectivity . Hillsdale, NJ: Analytic

Press.

Mitrani, J. (2001). Taking the transference: some technical implications in three papers by Bion.

International Journal of Psychoanalysis ,

82 ,

1085–1104.

Money-Kyrle, R. (1956). Normal counter-transference and some of its deviations. International

Journal of PsychoAnalysis , 37 , 360–366. Also published in D. Meltzer with E. O’Shaughnessy
(Eds.) (1978). Collected Papers of Roger Money-Kyrle . Strath Tay: Clunie Press, pp. 330–342.

382

References

Money-Kyrle, R. (1965). Megalomania. American Imago , 22 (1–2), 142–154.

Money-Kyrle, R. (1968). Cognitive development.

International Journal of

Psychoanalysis , 49 , 691–698. Also published in D. Meltzer with E. O’Shaughnessy (Eds.)

(1978). Collected Papers of Roger Money-Kyrle . Strath Tay: Clunie Press, pp. 416–433.

Money-Kyrle, R. (1991). Normale Gegenübertragung und mögliche Abweichungen [Normal

countertransference and possible deviations]. In E. B. Spillius (Ed.), Melanie Klein Heute:
Entwicklungen in Theorie und Praxis .

Band 2: Anwendungen [Melanie Klein Today: Developments in Theory and Practice. Volume 2:
Mainly Practice].

Stuttgart: Klett-Cotta, pp. 29–44.

Muir, R. (1982). The family, the group, transpersonal processes and the individual. International
Review of Psychoanalysis , 9 , 317–326.

Muir, R. (1990). Transpersonal processes: a bridge between object relations and attachment
theory in normal and psychopathological development.

British Journal of Medical Psychology , 68 , 243–257.

Nadelson, T. (1976). Victim, victimizer: interaction in the psychotherapy of borderline patients.
International Journal of Psychoanalysis , 50 , 115–129.

Nissim Momigliano, L. (1974). Come si originano le interpretazioni nello psicoanalista [As they
originate in the psychoanalyst’s interpretations]. Rivista di Psicoanalisi , 20 , 144–165.

Nissim Momigliano, L. (1984). ‘. . . Due persone che parlano in una stanza

. . .’ (Una ricerca sul dialogo analitico) [‘. . . Two people talking in a room . . .’ (Research on the
analytic dialogue).

Rivista di Psicoanalisi , 30 (1), 1–17.

Nissim Momigliano, L. (1991). Il tè nel deserto: ulteriori considerazioni a proposito de ‘Lo

psicoanalista allo specchio’ [The sheltering sky: further considerations on ‘The psychoanalyst in
the mirror’]. Rivista di Psicoanalisi , 37 (4), 772–819.

Ogden, T. (1974). A psychoanalytic psychotherapy of a patient with cerebral palsy: the relation
of aggression to self and body representations. International Journal of Psychoanalysis , 53 ,
419–433.

Ogden, T. (1976). Psychological unevenness in the academically successful student.

International Journal of Psychoanalysis , 55 , 437–448.

Ogden, T. (1978a). A developmental view of identifi cations resulting from maternal

impingements. International Journal of Psycho-Analytic Psychotherapy , 7 , 486–587.

Ogden, T. (1978b). A reply to Dr Orston’s discussion of ‘A developmental view of identifi

cations resulting from maternal impingements’. International Journal of Psycho-Analytic
Psychotherapy , 7 , 528–532.

Ogden, T. (1979). On projective identifi cation. International Journal of Psychoanalysis , 60 ,

357–373.

Ogden, T. (1982). Projective Identifi cation and Psychotherapeutic Technique . New York: Jason
Aronson.

383

References

Ogden, T. (1986). The Matrix of the Mind: Object Relations and the Psychoanalytic Dialogue .
Northvale, NJ: Jason Aronson.

Ogden, T. (1994a). The analytic mind: working with intersubjective clinical facts. International
Journal of Psychoanalysis , 75 , 3–20.

Ogden, T. (1994b). The concept of interpretative action.

Psychoanalytic

Quarterly , 63 (2), 219–245.

Ogden, T. (1994c). Subjects of Analysis . Northvale, NJ: Jason Aronson.

Ogden, T. (2004). The analytic third: implications for psychoanalytic theory and technique.
Psychoanalytic Quarterly , 73 , 167–195.

Oliveira, C. L. M. V. (2005).

História da psicanálise – São Paulo

(1920–1969) [History of Psychoanalysis – São Paulo (1920–1969)]. São Paulo: Escuta.

Ornston, D. (1978). Projective identifi cation and maternal impingement.

International Journal of Psychoanalytic Psychotherapy , 7 , 508–528.

O’Shaughnessy, E. (1975). Explanatory notes. In M. Klein (1975). Envy and Gratitude and Other
Works 1947–1963

. London: Hogarth Press, pp. 324–336.

O’Shaughnessy, E. (1992). Enclaves and excursions.

International Journal of

Psychoanalysis , 73 , 603–611.

Pantone, P. J. (1994). Projective identifi cation: affective aspects. Contemporary Psychoanalysis ,

30 , 604–618.

Parsons, M. (2000). The Dove that Returns, the Dove that Vanishes: Paradox and Creativity in
Psychoanalysis .

London: Routledge.

Pasche, F. (1982). A propos de l’identifi cation projective [About projective identifi cation].
Revue Française de Psychanalyse , 46 , 408–411.

Perestrello, M. (1992). A Psicanálise no Brasil. Encontros: psicanálise . Rio de Janeiro: Imago.

Pérez-Sánchez, A. (1997). Análisis terminable: estudio de la terminación del proceso

psicoanalítico [Terminable Analysis: A Study of the Termination of the Psychoanalytic Process].
Valencia, Spain: Promolibro.

Pichon-Rivière, E. (1952). Quelques observations sur le transfert chez des patients psychotiques.
Revue française de psych analyse, 16 , 252–262.
Pierrakos, M. (2003). La ‘tapeuse’ de Lacan: souvenirs d’une sténotypiste fâchée, réfl exions
d’une psychanalyste navrée . Paris: L’Harmattan.

Porder, M. (1987). Projective identifi cation: an alternative hypothesis.

Psychoanalytic Quarterly , 56 , 431–451.

Purcell, S. (2006). The analyst’s excitement in the analysis of perversion.

International Journal of Psychoanalysis , 87 , 105–123.

Quinodoz, D. (1989). Les ‘interprétations dans la projection’. Revue Française de Psychanalyse ,

53 , 103–110

[translated 1994 as ‘Interpretations in projection’. International Journal of Psychoanalysis , 75 ,

755–

762].

Quinodoz, D. (2002). Les Mots qui touchent: une psychanalyste apprend à parler .

Paris: Presses Universitaires de France (PUF) [translated 2003 as Words that Touch:
Psychoanalyst Learns to Speak

. London: Karnac].

384

References

Quinodoz, J.-M. (1992).

La Solitude apprivoisée: l’angoisse de séparation en Psychanalyse . Paris: Presses Universitaires

de France (PUF)

[translated 1993

as

The Taming of Solitude: Separation Anxiety in Psychoanalysis . London: Routledge].

Quinodoz, J.-M. (2000). Mélancolie maniaque: quelle issue? [The outcome of manic
melancholia]. Revue Française de Psychanalyse , 64 , 1825–1835.

Quinodoz, J.-M. (2001). Les Rêves qui tournent une page [The Dreams that Turn a Page]. Paris:
Presses

Universitaires de France (PUF).

Quinodoz, J.-M. (2002a). L’identifi cation projective: qu’en pensent les psychanalystes de
langue française?

Bulletin de la Fondation Européene pour la

Psychanalyse , 56 , 148–156. Also published as Projective identifi cation: what do French-

speaking psychoanalysts think? Psychoanalysis in Europe, Bulletin 56 , 139–147.

Quinodoz, J.-M. (2002b). Suisse romande (histoire de la psychanalyse). In A. de Mijolla (Ed.),

Dictionnaire de la psychanalyse [Dictionary of Psychoanalysis]. Paris: Calmann-Lévy.

Racamier, P. C. (1992). Le Génie des origines: psychanalyse et psychoses [The Genius of

Origins: Psychoanalysis and Psychoses]. Paris: Payot.

Racker, H. (1953). A contribution to the problem of counter-transference.

International Journal of Psycho-Analysis , 34 , 313–324.

Racker, H. (1957). The meaning and uses of counter transference. Psychoanalytic Quarterly , 26 ,
303–357.

Racker, H. (1958a). Classical and present technique in psycho-analysis. In H. Racker (Ed.)

(1968).

Transference and Countertransference . London: Hogarth Press, pp. 23–70.

Racker, H. (1958b). Counterresistance and interpretation. International Journal of Psycho-

Analysis , 6 , 215–221.

Racker, H. (1968). Transference and Countertransference . New York: International Universities

Press.

Rayner, E. (1990). The Independent Mind in British Psychoanalysis . London: Free Association
Books.

Rayner, E. (1992). Matching attunement and the psychoanalytic dialogue.

International Journal of Psychoanalysis , 73 , 39–54.

Rayner, E. (1995). Unconscious Logic: An Introduction to Matte Blanco’s Bi-logic and its Uses .
London: Routledge.

Ribas, D. (1992). Un cri obscur: l’énigme des enfants autistes . Paris: Calmann-Lévy.

Riesenberg Malcolm, R. (1970). El espejo: una fantasia sexual perversa en una mujer, vista como
defensa contra un derrumbe psicótico. Revista Psicoanalisis , 27 , 793–826. Also published as
‘The mirror: a perverse sexual phantasy in a woman seen as a defense against a psychiatric
breakdown’, in E. Spillius (Ed.) (1988).
Melanie Klein Today . Volume 2:

Mainly Practice . London:

Routledge, pp. 1115–1137.

385

References

Rivière, J. (1949). Contribución al análisis de la reacción tarepéutica negativa

[A contribution to the analysis of the negative therapeutic reaction, originally published in

English in 1936]. Revista de Psicoanálisis , 7 , 121–142.

Rivière, J. (Unpublished). Paranoid attitudes seen in everyday life and in analysis. Paper read
before the British Psycho-Analytical Society in 1948.

Rodrigué, E. (1966). Relación entre descubrimiento e identifi cación proyectiva. In Rodrigué, El

contexto del proceso analítico [The Context of the Analytical Process]. Buenos Aires: Paidós.

Rosenfeld, H. A. (1947). Analysis of a schizophrenic state with depersonalization. In Rosenfeld

(1965).

Psychotic States: A Psycho-Analytical Approach .

London: Hogarth Press and Institute of Psycho-Analysis, pp. 13–33.

Rosenfeld, H. A. (1949). Remarks on the relation of male homosexuality to paranoia, paranoid

anxiety and narcissism. In Rosenfeld (1965). Psychotic States: A Psycho-Analytical Approach .
London: Hogarth Press and Institute of Psycho-Analysis, pp. 34–51.

Rosenfeld, H. A. (1950). Note on the psychopathology of confusional states in chronic

schizophrenias. International Journal of Psychoanalysis , 31 , 132–137.

Rosenfeld, H. A. (1952a). Notes on the psychoanalysis of the superego confl ict in an acute
schizophrenic patient.

International Journal of Psychoanalysis , 33 , 111–131.

Rosenfeld, H. A. (1952b). Transference-phenomena and transference-analysis in an acute

catatonic schizophrenic patient. International Journal of Psychoanalysis , 33 , 457–464.

Rosenfeld, H. A. (1954). Considerations regarding the psycho-analytic approach to acute and

chronic schizophrenia.

In Rosenfeld (1965). Psychotic States: A Psycho-Analytical Approach . London: Hogarth Press

and Institute of Psycho-Analysis, pp. 117–127.
Rosenfeld, H. A. (1964a). Object relations of an acute schizophrenic patient in the transference
situation. In P.

Solomon and B. C. Glueck (Eds.), Recent Research on Schizophrenia: Psychiatric Research

Reports of the American Psychiatric Association . Washington, DC: American Psychiatric
Association.

Rosenfeld, H. A. (1964b). On the psychopathology of narcissism: a clinical approach.

International Journal of Psycho-Analysis , 45 , 332–337. Also published in Rosenfeld (1965).
Psychotic States: A Psycho-Analytical Approach . London: Hogarth Press and Institute of
Psycho-Analysis, pp. 1169–1179.

Rosenfeld, H. A. (1965). Psychotic States: A Psycho-Analytical Approach . London: Hogarth

Press and Institute of Psycho-Analysis.

Rosenfeld, H. A. (1969). The negative therapeutic reaction. In P. Giovacchini (Ed.) (1975).

Tactics and Techniques in Psychoanalytic Theory , Volume 2. New York: Jason Aronson.

Rosenfeld, H. A. (1971). Contribution to the psychopathology of psychotic states: the importance

of projective identifi cation in the ego structure and object relations of the psychotic patient. In P.
Doucet and C. Laurin (Eds.), 386

References

Problems of Psychosis , Volume 1. The Hague: Excerpta Medica, pp. 115–128.

Also published in E. Spillius (Ed.) (1988). Melanie Klein Today: Developments in Theory and
Practice . Volume 1: Mainly Theory . London: Routledge,

pp. 117–137.

Rosenfeld, H. A. (1983). Primitive object relations and mechanisms.

International Journal of Psycho-Analysis , 64 , 261–267.

Rosenfeld, H. A. (1987a). Impasse and Interpretation . London: Routledge.

Rosenfeld, H. A. (1987b). The infl uence of projective identifi cation on the analyst’s task. In
Rosenfeld, Impasse and Interpretation . London: Routledge, pp. 157–264.

Rosenfeld, H. A. (1987c). Projective identifi cation in clinical practice. In Rosenfeld, Impasse

and Interpretation .

London: Routledge, pp. 157–189.

Rosenman, S. (2003). Assaultive projective identifi cation and the plundering of the victim’s
identity. Journal of the American Academy of Psychoanalysis , 31 , 521–540.

Roth, P. (1999). General introduction. In R. Riesenberg Malcolm and P. Roth (Eds.), On bearing
unbearable states of mind . London: Routledge, pp. 1–7.

Roudinesco, E. and Plon, M. (1997). Dictionnaire de la psychanalyse . Paris: Fayard.

Roussillon, R. (1999). Agonie, clivages et symbolisation [Agony, Splitting and Symbolization].

Paris: Presses Universitaires de France (PUF).

Sagawa, R. Y. (1980). Durval Marcondes e o início do movimento psicanalítico brasileiro

[Durval Marcondes and the Beginning of the Psychoanalytic Movement in Brazil]. São Paulo:
Cortez, Cadernos Freud/Lacan.

Sagawa, R. Y. (1992). Redescobrir as psicanálises [Rediscovering Psychoanalysis].

São Paulo: Lemos.

Sandler, J. (1960). Identifi cation in children, parents and doctors. In R. MacKeith and J. Sandler
(Eds.), Psychosomatic Aspects of Paediatrics .

London: Pergamon.

Sandler, J. (1976a). Countertransference and role-responsiveness. International Review of

Psychoanalysis , 3 , 43–

48.

Sandler, J. (1976b). Dreams, unconscious fantasies and ‘identity of perception’.

International Review of Psycho-Analysis , 3 , 33–42.

Sandler, J. (1983). Refl ections on some relations between psychoanalytic concepts and
psychoanalytic practice.

International Journal of Psychoanalysis , 64 , 35–45.

Sandler, J. (1987a). The concept of projective identifi cation. In J. Sandler (Ed.), Projection,
Identifi cation, Projective Identifi cation . Madison, CT: International Universities Press, pp. 13–
26. Also published as Sandler (1987). The concept of projective identifi cation. Bulletin of the
Anna Freud Centre , 10 , 33–49.

Sandler, J. (Ed.) (1987b).

Projection, Identifi

cation, Projective Identifi

cation .

Madison, CT: International Universities Press.

Sandler, J. (1990). On internal object relations.

Journal of the American

Psychoanalytic Association , 38 , 859–880.

387

References

Sandler, J. (1993). On communication from patient to analyst: Not everything is projective

identifi cation. Thirty-eighth International Psychoanalytical Congress Presidential Address: From
Listening to Interpretation (1993, Amsterdam, Netherlands). International Journal of Psycho-
Analysis , 74 (6), 1097–1107.

Sandler, J. and Dreher, A. U. (1996). What Do Psychoanalysts Want? London: Routledge.

Sandler, J. and Joffe, W. G. (1967). The tendency to persistence in psychological function and
development, with special reference to fi xation and regression. Bulletin of the Menninger Clinic
, 31 , 257–271.

Sandler, J. and Rosenblatt, B. (1962). The concept of the representational world. Psychoanalytic
Study of the Child , 17 , 128–162.

Sandler, J. and Sandler, A.-M. (1978). On the development of object relationships and affects.
International Journal of Psychoanalysis , 59 , 285–296.

Sandler, J. and Sandler, A.-M. (1984). The past unconscious, the present unconscious and
interpretation of the transference. Psychoanalytic Inquiry , 4 , 367–399.

Sandler, J., Dare, C. and Holder, A. (1973). The Patient and the Analyst: The Basis of the
Psychoanalytic Process .

New York: International Universities Press.

Schafer, R. (1968).

Aspects of Internalization . New York: International Universities Press.

Scharff, J. S. (1992). Projective and Introjective Identifi cation and the Use of the Therapist’s Self
. New York: Jason Aronson.

Schur, M. (1972). Freud: Living and Dying . New York: International University Press.

Searles, H. F. (1962). The differentiation between concrete and metaphorical thinking in the
recovering schizophrenic patient. Journal of the American Psychoanalytic Association , 10 , 22–
49.

Searles, H. F. (1963). Transference psychosis in the psychotherapy of chronic schizophrenia.

International Journal of Psycho-Analysis , 44 , 249–281. Also in Searles (1965). Collected
Papers on Schizophrenia and Related Subjects

. London: Maresfi eld Library, pp. 654–716.

Searles, H. F. (1975). The patient as therapist to the analyst. In P. Giovacchini (Ed.), Tactics and
Techniques in Psychoanalytic Theory , Volume 2. New York Jason Aronson.

Segal, H. (1950). Some aspects of the analysis of a schizophrenic. International Journal of

Psychoanalysis , 31 , 268–278.

Segal, H. (1957). Notes on symbol formation. International Journal of PsychoAnalysis , 38 ,

391–397.

Segal, H. (1962). The curative factors in psycho-analysis. International Journal of Psycho-

Analysis , 43 , 212–227.

Segal, H. (1964). Introduction to the Work of Melanie Klein . London: Heinemann.

Segal, H. (1973). Introduction to the Work of Melanie Klein , new enlarged edn.

London: Hogarth Press and Institute of Psycho-Analysis.

388

References

Segal, H. (1981). Countertransference. In Segal, The Work of Hanna Segal: A Kleinian Approach
to Clinical Practice . New York: Jason Aronson, pp. 81–88.

Sinason, M. (1993). Who is the mad voice inside? Psychoanalytic Psychotherapy , 7 , 207–221.

Sinason, M. (1999). How can you keep your hair on? In P. Williams (Ed.), Psychosis (Madness)
. London: Karnac.

Sister, B. M. and Taffarel, M. (1996). Isaías Melsohn, A psicanálise e a vida: setenta anos de
histórias paulistanas e a formação de um pensamento renovador na psicanálise .

São Paulo: Escuta.

Smith, H. F. (2000). Countertransference, confl ictual listening and the analytic object
relationship.

Journal of the American Psychoanalytic Association , 48 , 95–128.

Smith, H. F. (2002). Creating the psychoanalytical process, incorporating three panel reports:
opening the process, being in the process and closing the process. International Journal of
Psychoanalysis , 83 , 211–227.
Smith, H. F. (2006). Analyzing disavowed action: the fundamental resistance of analysis.
Journal of the American Psychoanalytic Association , 54 , 713–737.

Sodré, I. (2004). Who’s who? Notes on pathological identifi cations. In E. Hargreaves and A.
Varchevker (Eds.), In Pursuit of Psychic Change: The Betty Joseph Workshop . London:
Brunner-Routledge, pp. 53–68.

Sohn, L. (1985). Narcissistic organisation, projective identifi cation and the formation of the
identifi cate.

International Journal of Psychoanalysis , 66 , 201–214.

Speziale-Bagliacca, R. (1983). Introduction to Grotstein J.J., Scissione e identifi cazione

proiettiva . Rome: Armando.

Spillius, E. B. (1988a). Introduction to ‘projective identifi cation’. In Spillius (Ed.), Melanie

Klein Today: Developments in Theory and Practice . Volume 1: Mainly Theory . London:
Routledge, pp. 81–86.

Spillius, E. B. (Ed.) (1988b). Melanie Klein Today: Developments in Theory and Practice .
Volume 1: Mainly Theory . London: Routledge.

Spillius, E. B. (Ed.) (1990). Melanie Klein Heute: Entwicklungen in Theorie und Praxis . Band 1:
Beiträge zur Theorie [Melanie Klein Today: Developments in Theory and Practice. Volume 1:
Mainly Theory]. Stuttgart: Klett-Cotta.

Spillius, E. B. (1992). Clinical experiences of projective identifi cation. In R. Anderson (Ed.),

Clinical Lectures on Klein and Bion . London: Routledge, pp. 59–73.

Spillius, E. B. (1994). Developments in Kleinian thought: overview and personal view.

Psychoanalytic Inquiry , 14 , 324–364.

Spillius, E. B. (2002). A brief introduction to the concept of projective identifi cation.

Psychoanalysis in Europe, Bulletin 56 , 115–117.

Spillius, E. B. (2007). Encounters with Melanie Klein , edited by Priscilla Roth and Richard
Rusbridger. London: Routledge.

Spira, M. (1985). Créativité et liberté psychique [Creativity and Psychic Freedom].

Lyon: Césura.

389

References

Spitz, R. (1965). The First Year of Life . New York: International Universities Press.

Spotnitz, H. (1969). Modern Psychoanalysis of the Schizophrenic Patient . New York: Grune &
Stratton.

Steffens, W. (1999). Bemerkungen über Verfolgung, Verzweifl ung und Wiedergutmachung

[Remarks on persecution, despair and reparation]. In C. Frank (Ed.), Stillstand, Veränderung und
die Angst vor einer Katastrophe.

Perspektiven Kleinianischer Psychoanalyse [Standstill, Change and the Anxiety of a

Catastrophe. Perspectives in Kleinian Psychoanalysis]. Tübingen: Diskord, pp. 61–84.

Steiner, J. (1993). Psychic Retreats: Pathological Organisations in Psychotic, Neurotic and

Borderline Patients .

London: Routledge.

Steiner, R. (1985). Some thoughts about tradition and change arising from an examination of the
British Psychoanalytical Society’s Controversial Discussions (1943–1944). International Review
of Psychoanalysis , 12 , 27–71.

Steiner, R. (1999). Who infl uenced whom? And how? International Journal of Psychoanalysis ,
80 , 349–375.

Strachey, J. (1934). The nature of the therapeutic action of psychoanalysis.

International Journal of Psychoanalysis , 15 , 127–159.

Strachey, J. (1951). Letter from Strachey to Jones, October 24, 1951. Strachey Archives, Institute
of Psychoanalysis, London.

Tähkä, T. A. (1977). Presentation at the Sixth World Congress of Psychiatry, Honolulu, Hawaii.

Tansey, M. and Burke, W. (1985). Projective identifi cation and the empathic process:
interactional communications.

Contemporary Psychoanalysis ,

21 ,

42–69.

Tansey, M. and Burke, W. (1989).

Understanding Countertransference: From

Projective Identifi cation to Empathy . Hillsdale, NJ: Analytic Press.

Thomä, H. and Kächele, H. (1985).

Lehrbuch der Psychoanalytischen

Therapie [Textbook of Psychoanalytical Therapy], Volumes 1 and 2. Berlin: Springer.

Thorner, H. A. (1963). Ursache, Grund und Motiv. Ein psychoanalytischer Beitrag zum
Verständnis psychosomatischer Phänomene [Cause, Reason and Motive: A Psychoanalytical
Contribution to Understanding Psychosomatic Phenomena].

Psyche: Zeitschrift für Psychoanalyse und ihre

Anwendungen , 16 , 670–685.

Tonnesmann, M. (1989). Editor’s Introduction. In Tonnesmann (Ed.), About Children and

Children-No-Longer: Collected Papers of Paula Heimann, 1942–80 .

London: Routledge.

Trinca, W. (1991). A etérea leveza da experiência [The Ethereal Lightness of Experience]. São
Paulo: Siciliano.

Trinca, W. (1992). Dreams, psychic mobility and inner being. Free Associations , 8 (4), 562–
575.

390

References

Trinca, W. (2007). O ser interior na psicanálise: fundamentos, modelos e processos

[The Interior Being in Psychoanalysis]. São Paulo: Vetor.

Tuckett, D. (1997). Mutual enactment in the psychoanalytic situation. In J. L. Ahumada, J.

Olagaray, A. K. Richards and A. D.Richards (Eds.), The Perverse Transference and Other
Matters: Essays in Honor of R. Horacio Etchegoyen

Northvale, NJ: Jason Aronson.

Tuckett, D. (2005). Does anything go? Towards a framework for the more transparent
assessment of psychoanalytic competence. International Journal of Psychoanalysis , 86 , 31–49.

Wallerstein, R. (1988). One psychoanalysis or many? International Journal of Psycho-Analysis ,

69 , 5–21.

Weiss, E. (1925). Über eine noch nicht beschriebene Phase der Entwicklung zur heterosexuellen
Liebe.

Internationale Zeitschrift für Psychoanalyse , 11 , 429–443.

Whipple, D. (1986). Discussion of ‘The merits and problems with the concept of projective
identifi cation’ by Janet Finell. Psychoanalytic Review , 73 , 121–128.

Widlöcher, D. (2003). L’avenir de la psychanalyse: débat entre Daniel Widlöcher et Jacques-

Alain Miller [The future of psychoanalysis: debate between Daniel Widlöcher and Jacques-Alain
Miller]. Psychiatrie, Sciences Humaines, Neurosciences , 1 (1), 10–18.

Winnicott, D. W. (1945). Primitive emotional development.

International

Journal of Psychoanalysis , 26 , 137–143.

Winnicott, D. W. (1949). Hate in the countertransference. International Journal of Psycho-

Analysis , 30 , 69–75.

Winnicott, D. W. (1952). Psychoses and childcare. In Winnicott (1975).

Through Paediatrics to Psycho-Analysis . New York: Basic Books.

Winnicott, D. W. (1953). Transitional objects and transitional phenomena.

International Journal of Psycho-Analysis , 34 , 1–9.

Winnicott, D. W. (1956). Primary maternal preoccupation. In Winnicott (Ed.) (1958).

Collected Papers: Through Paediatrics to Psycho-analysis . London: Tavistock, pp. 300–305.

Winnicott, D. W. (1958). Collected Papers . London: Tavistock.

Winnicott, D. W. (1960). Countertransference. In Winnicott (1965).

The

Maturational Processes and the Facilitating Environment . New York: International Universities
Press.

Winnicott, D. W. (1967). Mirror-role of the mother and family in child development. In P.

Lomas (Ed.), The Predicament of the Family: A Psycho-Analytical Symposium . London: Hogarth
Press, pp. 26–33.

Winnicott, D. W. (1971). Playing and Reality . London: Tavistock.

Wisdom, J. O. (1962). Ein methodologischer Versuch zum Hysterieproblem

[A methodological approach to the problem of hysteria]. Psyche: Zeitschrift für Psychoanalyse

und ihre Anwendungen , 15 , 561–587.

391
References

Zanin, J. C. (2004). Uma época fértil e evolutiva da Psicanálise no Rio de Janeiro. In P. Marchon
(Ed.), A psicanálise no Rio de Janeiro e sua difusão pelo Brasil . São Paulo: CNPQ, Coleção
Memória do Saber.

Zinner, J. and Shapiro, R. (1972). Projective identifi cation as a mode of perception and
behaviour in families of adolescents. International Journal of Psychoanalysis , 53 , 523–530.

Zwiebel, R. (1985). Das Konzept der projektiven Identifi zierung. Bericht über die Tagung
‘Projektion, Identifi kation, projektive Identifi zierung’

( Jerusalem, 27–29 May 1984). [The concept of projective identifi cation.

Report on the conference ‘Screening, Identifi cation, Projective Identifi cation’]. Psyche:
Zeitschrift für Psychoanalyse und ihre Anwendungen , 39 , 456–468.

Zwiebel, R. (1988). Einige klinische Anmerkungen zur Theorie der Projektiven Identifi zierung
[Some clinical remarks about the theory of projective identifi cation]. Zeitschrift für
Psychoanalytische Theorie und Praxis , 3 (2), 165–186.

392

Index

Aberastury, A. 326

Amati Mehler, J. 204, 208, 210, 211,

Abraham, K. 24, 46, 303

212

abstract thinking 82, 83

American Psychoanalytic Association

acquisitive projective identifi cation

157

(Britton) 58–60, 190, 216, 247

analyst ( passim ): as blank screen 260; as

acting out 72, 86, 88, 123, 143, 315,

container for patient’s pathological

357, 358; in analytic situation,

contents 292; countertransference of,

communicative function of 202; see

as sign of pathology in 17;

also enactment

involvement of, in defensive use of

actualization (Sandler) 60, 136, 158,

projective identifi cation 112–31;

164, 177

projection of bad parts into, and

acute enactment 351

control of (clinical example) 270–2;

adaptive projective identifi cation (Blatt)

therapeutic omnipotence of 291

243, 258

analytic couple on threshold of mutual

adhesive identifi cation (Bick) 214, 223

hypnosis, as ‘mass of two’ (Freud)

Adler, G. 154

364

adualism (Piaget) 180

analytic/intersubjective third (Ogden)

affect regulation 362

241–2, 252, 359, 361

aggression: early role of 22; oral 24;

analytic schools, disagreements

primary 70, 72

between 165

aggressive instinct (Freud) 188

analytic situation, projective

aggressive mechanism, projective

identifi cation in, concreteness of

identifi cation as (Klein) 160

(clinical example) 101–3

aggressive object relation (Klein) 5,

Anderson, R. 49

77, 195

annihilation, fear of 23, 32

aggressor/persecutor, identifi cation

anorexia 103, 255

with (A. Freud) 135, 157, 175

anxiety(ies): defences against 25, 301,

Almeida Prado, M. P. 343–4, 352

302; in early infancy 20; latent, in

alpha (α) elements (Bion) 230, 334

schizoid patients 39, 42; paranoid 9,

alpha (α) function(s) (Bion) 211, 351

10, 31, 76, 77, 90, 96, 115, 230, 314

altruistic surrender 175, 266

[early 19]; persecutory 10, 22, 26, 27,

393

Index

29, 32, 35, 77, 93, 102, 142, 309 [in

transformation into alpha (α)

infancy 23, 43]; primary 23, 24;

elements 230

schizoid, early 19; separation 23, 80,

bi-personal fi eld (Barangers) 257, 258,

84, 85, 86, 224, 313

359

Anzieu, A. 223
Bibring, E. 266

archaic object relationship 113, 127,

Bick, E. 214, 221

130

Bion, W. R. ( passim ): alpha (α) elements

Argelander, H. 200, 202

230, 334; alpha (α) function(s) 211,

Argentina, concept of projective

351; basic assumption group

identifi cation in 325–38

phenomena 360; beta (β) elements

Argentine Psychoanalytical Association

333, 351; [transformation into alpha

325, 329, 340

(α) elements 230]; communication

‘as if ’ personalities 214

and thinking, failure of 156;

assaultive projective identifi cation

container [-contained 164, 168, 173,

(Rosenman) 247

355; theory of 135, 136, 178, 230];

Athanassiou, C. 221, 223

containment 123, 195, 242, 256, 318,

attachment 242, 337

319, 356 [model of process of 55];

attention cathexes 199

countertransference [as analyst’s

attributive projective identifi cation

response to the patient’s projective

(Britton) 58, 60, 190, 216, 247

identifi cation 52–4; as expression of

attunement 242

analyst’s pathology 51–2; resorting to

Aulagnier, P. 234

53; use of 51–2]; infant, dependence

autism 214; infantile 223

of on receptiveness of mother to

autistic-contiguous position (Ogden)

accept projections 54–5; infl uence

252

on Brazilian psychoanalysis 341;

autoctonia (Grotstein) 348

linking, attacks on 51, 54, 61–75

Ayuy, C. 355, 356

[clinical examples 62–8]; mother,

capacity of for reverie 173, 178, 230,

bad parts of self/ego, splitting off and

242, 349; on parasitism 86–7;

projection of 5, 17, 28, 77, 84, 85,

paranoid-schizoid position [and

160, 270–2

depressive position, oscillation

Balint, M. 155, 162, 251, 255, 268, 290

between 194; pathology of 230];

Baranger, M. 257, 323, 325, 334–8, 359,

projective identifi cation ( passim )

360

[concept of, fundamental

Baranger, W. 257, 323, 325, 334–8, 359

contribution to 229–30; normal and

basic assumption group phenomena

pathological forms of 50, 54–7, 60;

(Bion) 360

use of, clinical examples 52–3, 115];

bastion (Barangers), clinical examples

reverie 173, 230, 242, 349 [of

335–8

‘containing’ mother 178]; thinking,

Bayle, G. 222

theory of 164 [thought without a

Bégoin, F., see Bégoin-Guignard, F.;

thinker 289]

Guignard, F.

Birksted-Breen, D. 159

Bégoin, J. 220, 222

birth, trauma of 23

Bégoin-Guignard, F. 220–2

Blatt, S. J. 243

Bell, D. 57

Bleger, J. 326

Bernardi, R. 235

body-phantasy and phantasy in the

beta (β) elements (Bion) 333, 351;

body (Gaddini) 194

394

Index

Bollas, C. 58

cannibalistic incorporation 133

Bolognini, S. 204

Caper, R. 256

Bonnard, A. 186

Carpy, D. V. 130, 159

borderline patient(s) 231, 232, 235, 248,

Casoni, D. 225, 256

253–5, 270, 357; clinical example

Cassorla, R. M. S. 350–1, 353

190–1

Chasseguet-Smirgel, J. 230

borderline psychosis 61

childhood neurosis 329

borderline transference 156

Chile, concept of projective

Boris, H. 255

identifi cation in 354–64

Boulanger, J.-B. 221

chronic enactment (Cassorla) 351

Bowlby, J. 155

Classical Freudians 153, 154, 155.

Boyer, L. B. 251, 254

See also Contemporary Freudians

Brazelton, T. B. 213

claustrophobia 10, 32, 90, 91, 99, 100,

Brazil, concept of projective

109, 115, 311

identifi cation in 339–53

collusion, defensive, analyst–patient

Brazilian Psychoanalytic Society: of

122, 130

Rio de Janeiro 343; of São Paulo

Coloma, J. 358

341, 344

communication: conesthetic mode

Brearley, M. 159

of 286; direct, dynamics of,

breast, mother’s 18, 20, 23, 24, 27, 28,

projective identifi cation as 366;

70; attacks on 20, 27; bad/frustrating

failure of (Bion) 156; interactional

24–6, 173, 268; good/gratifying

201; projective identifi cation

24–6, 40, 173; ideal 26; infant’s

as 56, 80, 100, 120, 242, 300

fantasies of sadistic attacks upon

[clinical example 105–9]; symbolic

(Klein) 20, 24, 61, 72; introjected 28,

118, 119, 350; and thinking, failure

70 [good 22; greedy 10]; persecuting

of 156;

26; split into good and bad 20

complementary countertransference

breast-mother 26; primary

(Racker) 202, 258, 327, 330, 355

identifi cation with 273

complementary identifi cation(s) 172,

Brenman-Pick, I. 162

177, 260, 327

Brierley, M. 4, 154, 155

concordant countertransference

British Kleinians 156, 365; use of

(Racker) 258, 327

concept of projective identifi cation

concordant identifi cation 172, 177

by 49–60

concrete thinking 82, 83, 156, 230

British object-relations theory 245, 251

conesthetic mode of communication

British Psychoanalytical Society 28, 32,

286

149–66, 207, 221, 339, 340; training

confi dentiality, breach of 295

in 165

confusional states 290

Britton, R. 58, 59, 60, 190, 247, 260

container 136; – contained model

Brodey, W. 280, 297

(Bion) 164, 168, 173, 178, 355,

Brunet, L. 225, 256

356; theory of (Bion) 135, 230

bulimia 255

containing function, and projective

Burke, W. 258, 259, 339, 343

identifi cation 225

Burlingham, D. 155

containment 55, 123, 195, 242, 256,

318, 319, 356

Cabré, L. M. 204

Contemporary Freudians 49, 149,

Canestri, J. 162, 185, 204–17

153–66

395

Index

control, via projective identifi cation: of

302; ego 20; mechanisms of 169;

analyst 55, 81, 105, 123, 271, 333; of

object-related 169; primitive 159,

mother 20, 27, 170; of object 4, 9,

253; projective identifi cation as

27, 30–4, 56, 77, 99, 112, 135, 139,

53, 79, 96, 134, 168, 175, 176, 211,

174, 175, 189, 195, 227, 250, 260,

241, 265, 285, 300 [analyst’s

289 [greedy 17]; of other(s) 55, 56,

involvement in 112–31; clinical

60, 198, 243, 278, 281, 289, 308 [in

example 115–16, 125–9]; schizoid

obsessional neurosis 33]; of projected

39–41

parts 167, 174, 198, 281, 299; of

defensive collusion, analyst–patient

women, by patient (Klein) 13;

122

Controversial Discussions between

delusional transference 267, 291

Melanie Klein and Anna Freud 153,

dementia praecox 45

155

denial 20, 65, 71, 74, 77, 107, 222, 287,

counterprojective identifi cation

295, 296, 316; of psychic reality 26,

(Grinberg) 208, 209, 213, 214, 330,

56, 81, 96, 231

333, 334, 355, 360;

Denis, P. 218

countertransference ( passim ): analyst’s

depersonalization 6, 9, 14, 30, 290

[as response to the patient’s projective

depressive identifi cation 256

identifi cation (Bion) 52–3; as sign of

depressive individual 22

pathology 17 (Freud) 51]; clinical use

depressive position (Klein) 82, 155;

of 156; complementary (Racker) 202,

anxiety, vs. paranoid anxiety 230;

258, 327, 330, 355; concordant

central role of in early development

(Racker) 258, 327; defi nition of,

21; concept of 19; and identifi cation

change in 53; as instrument 351;

256; and individuation 162;

neurosis 334; objective (Winnicott)

infantile 19; moving towards 99,

298; positive use/value of 105, 171;

100, 105–10, 111, 243; onset of 43;

and projective identifi cation 50–3,

and paranoid–schizoid position 21,

168; psychosis 291; resorting to (Bion)

251, 252; fl uctuations between

53; as response to patient’s projective

35–9, 194; and projective

identifi cation 50; and transference

identifi cation 49, 56, 105–10, 349;

156, 159, 171, 173, 180, 188, 197, 229,

and symbolic projection 137;

240, 247, 354, 365; use of, for

working through of, 34, 35 [failure

understanding patient 53

to 21, 35, 74]

crossed projective identifi cations

depressive processes in infancy 19

(Cassorla) 350, 351

destructive impulse, fear of 23

curiosity 17, 68–9, 138, 315; impulse

destructive narcissism 188

of, disturbance of 74

Deutsch, H. 327

Cycon, R. 187–8

Di Chiara, G. 208, 209, 212

Diatkine, G. 218, 220

Da Rocha Barros, E. 101

dissociation, schizophrenic 30

da Silva, G. 225

Dornes, M. 194, 195, 201

David, C. 226

Dorpat, T. 250

de M’Uzan, M. 226, 231

Drapeau, P. 225

de Saussure, R. 219

dreaming, schizophrenic 65

death instinct/drive 23, 24, 25, 45, 73,

Dreher, A. U. 157

188, 242, 268, 288, 289, 328, 343, 361

Duparc, F. 234

defence(s): against anxiety 25, 301,

396

Index
Eagle, M. N. 259–60

excrements, expulsion of (Klein) 4, 5,

early development, projective

20, 27, 28, 77, 160

identifi cation in 285–8

expressive perception (Melsohn) 344,

ego: boundaries 83, 253; central and

352

two subsidiary egos (Fairbairn) 23;

externalization 167, 169, 173, 209, 210,

defences, early 20; development 20,

211, 252, 275, 296, 297, 300

26, 29, 30, 77, 80, 82 [early 22–4];

extractive introjection 58

disintegration 301; early [splitting

process of 76, 96; structure of 23;

Fain, M. 222

unintegration of, pathological

Fairbairn, W. R. D. 23, 113, 245, 265,

consequences of (Winnicott) 23];

273; views on schizoid mechanisms,

-ideal 4, 28, 32; impoverishment/

vs. Klein’s views 21–2

weakening of 34, 302 [due to

fantasy: function of, in projective

splitting (Klein) 5, 31]; integration 6,

identifi cation 275; of inhabiting and

30, 35, 268, 269; nuclei (Glover) 23;

controlling another (clinical

psychology 156, 249, 254; relation to

example) 278–80; and object

id and superego 23; split-off parts of, relationship, projective identifi cation projecting of 303
[onto mother 27, as 276–84

200]; splitting within 24, 25, 31, 76,

Feldman, M. 57, 59, 112–31, 162, 164,

78, 80, 89, 93, 96, 97, 222; structure,

188

of psychotic patient, role of

Ferenczi, S. 6, 24, 219

projective identifi cation in 76–97;

Ferro, A. 210, 211

unconsciously identifi cation with

Figueroa, J. 357, 358, 359

introjected bad object (Freud) 132;

Flechsig, P. E. 43, 44, 45

vital need of, to deal with anxiety

‘Flechsig soul(s)’ 44, 45

(Klein) 161

Flegenheimer, F. 204, 209, 212

Eiguer, A. 227

Fliess, W. 315, 316, 319

empathy 100, 168, 179, 180, 205, 214,

folie à deux 307, 315–19

230, 253, 258, 259, 350

Fonagy, P. 162

enactment(s) 295, 315, 318; analyst

forced introjective identifi cation 136

114, 117, 295, 355–9 [–patient,

foreclosed projection (Green) 226

mutual 57; observation of 202;

Frank, C. 188

unconscious 113]; acute (Cassorla)

French Psychoanalytical Association

351; in analytic situation (Feldman)

234

120–31; chronic (Cassorla) 351;

French-speaking psychoanalysts

countertransference 248; as response

218–35; main contributions of

to projection 318

221–5

envy: of parental couple 68; primitive

Freud, A. 169: altruistic surrender 175,

84; projective character of 6

266; Controversial Discussions 153,

Erikson, E. H. 269
155; fl ight to London 153;

Eskelinen de Folch, T. 215

identifi cation with aggressor (A.

Etchegoyen, R. H. 334

Freud) 135, 157, 175

Evans, M. G. 27

Freud, S. ( passim ): aggressive instinct

evocative/evocatory (non-evocative/

188; analytic couple on threshold of

non-evocatory) projective

mutual hypnosis, ‘mass of two’ 364;

identifi cation (Spillius) 58, 60, 136,

countertransference as sign of

190, 216, 247

pathology in analyst 17, 51; death

397

Index

instinct, and projection 25;

Grinberg, L. 172, 176, 323, 325, 331,

defl ection of the death instinct

356; counterprojective identifi cation

outwards 25; dementia praecox 45;

208, 209, 213, 214, 330, 333, 334,

ego [concept of splitting 222;

355, 360; projective

unconscious identifi cation of, with

counteridentifi cation 224, 248, 298,

introjected bad object 132]; and

299, 330, 333, 334, 342, 355;

Fliess 315, 316, 319; fl ight to

Grotstein, J. S. 209, 239–41, 245–54,

London 153; free association 199;

256, 264–74, 292; as ‘adopter’ of

judgement, and reality principle 67;

concept of projective identifi cation

life and death instincts 45; narcissistic

249–51; autoctonia 348; projective

identifi cation with abandoned object

transidentifi cation 241; psychic

132–3; paranoic ‘world catastrophe’

presences 348;

44–5; primary identifi cation 180;

Guignard, F. 222, 223

primary instincts, concept of,

Gutwinski-Jeggle, J. 189

rejection of 22; projection 219, 223,

303, 318; psychoanalysis as

Haag, G. 221, 223

archaeological investigation 68;

Haber, M. 224
purifi ed pleasure ego 265; return to

hallucinations 64, 87; visual, of invisible

(Lacan) 233; Schreber case, analysis

objects 66

of 43–6; structural model 209;

Hargreaves, E. 49

superego formation, theory of 243;

hatred, early role of 22

thought-transference 365; Wolf Man

Heidegger, M. 363

309

Heimann, P. 5, 155, 172, 220, 337;

Frosch, J. P. 259

analyst’s countertransference 50, 51,

53, 156, 171, 176, 189, 196, 248, 254,

Gabbard, G. 248

314, 329; patient’s phantasies, effect

Gaddini, E. 194, 209, 212

on analyst 113

Gammill, J. 7, 220

‘here and now’ 163, 164

Garfi nkle, E. 247–8, 260–2

Herrera, E. 354, 355, 356, 364

gastrointestinal tract and projective

Hinshelwood, R. D. 57, 222

identifi cation, analogy between 302

Hinz, H. 185, 186–203, 253

Geissmann, C. 221

Hoffer, W. 155

Geissmann, P. 221

Hoffman, I. Z. 129, 130

genetic imitative competences

homosexuality 6; female 219

(Meltzoff) 213

Houzel, D. 221, 223

Germany, concept of projective

hysteria 22, 186

identifi cation in 186–3

hysterical identifi cation 222

Gibeault, A. 223, 232

hysterical patient(s) 254

Gilhooley, D. 256

Gillespie, W. 155

id 23, 330, 346

Glover, E. 23, 154, 155

idealization 20, 26, 124, 243, 301, 312,

Godfrind-Haber, J. 224

313

good-enough mother (Winnicott) 173,

idealized object: bad, identifi cation

285, 293

with 144; unassimilated 29

Green, A. 59, 226, 289

identifi cation(s) ( passim ): adhesive

Greenberg, J. 240

(Bick) 214, 223; with aggressor (A.

Greenson, R. R. 266

Freud) 135, 157, 175;

398

Index

complementary 172, 177, 260, 327;

330–3]; interpretation of projective

concept of, early discussion on 4;

identifi cation vs. interpretation from

concordant 172, 177; depressive 256;

within projective identifi cation

hysterical 222; introjective 15, 60,

(Hinz) 190

193, 338; mechanisms, early infantile, interpretations, resistance to 62

of a magic nature 78; narcissistic 132,

interpretative action 363

133, 214, 256; omnipotent 55, 58;

intersubjective/analytic third (Ogden)

pathological 132–46; primary

241–2, 252, 359, 361

179–81, 273; process (Danckwardt)

intersubjectivity 241, 244, 252, 324,

200; projective: see projective

335; and projective identifi cation

identifi cation(s); psychotic, in

354–64

schizophrenic patients 78

intrapsychic to interpersonal, projective

identifi catory projection (Segal) 226

identifi cation as the pathway from

identity, sense of 12, 134, 142, 144; and

168

internalized good object 11

intrapsychic projective identifi cation

imitation, concept of: Gaddini 167,

345

192, 194, 209, 212, 240; Piaget 213

introjected object, persecutory fear

impingement (Winnicott) 112, 114,

of 10

290

introjection ( passim ): analyst’s capacity

impotence 10, 32, 291

for 72; of breast 28, 70 [greedy 10;

incorporation 78, 139, 167, 170, 265;

good 22]; concept of, early discussion

cannibalistic 133

on 4; defi nition 266; effect of on

Independents 49, 149, 153–66, 253

object-relations 28; extractive 58; of

individuation 162

object 21, 118, 138 [bad, unconscious

infancy: depressive processes in 19;

identifi cation with (Freud) 132;

early, anxieties in 20; hallucinatory

persecutory fear of 10]; pathological

gratifi cation in 26; psychotic

137, 140; primitive 297; and

anxieties in 20; splitting of objects in

projection [interaction between 16,

61; see also mother

20, 30, 223, 302, 316; simultaneous

infant: dependence of, on mother’s

operation of (Klein) 60]

receptiveness to projections 54–5;

introjective identifi cation(s) 6, 15, 17,

development of, depressive and

54, 60, 70, 107, 108, 171, 175, 193,

paranoid-schizoid positions 34–6;

240, 252, 259, 334, 338; concrete and

fusion with mother 80; oral-sadistic

symbolic thinking in 58–9; forced

and cannibalistic desires of 24; see

136; pathological 137; ‘symbolic’

also mother

forms of phantasy and thought in 50;

infantile autism 223

intrusions 127, 158, 208

infantile psychosis(es), symbiotic 79

intrusive projective identifi cation

ingrainment, state of (Almeida Prado)

(Meltzer) 247

343, 344, 352

Isaacs, S. 154, 155, 306, 325, 334

interactional communication 201

Italian Psychoanalytical Association 207

internal objects 29, 30, 31, 82, 91, 164,

Italian Psychoanalytical Society 207

167, 209, 256, 282, 303, 316, 328

Italy, concept of projective

internalisation of good object, and

identifi cation in 207–13

sense of identity 11; patient’s 108,

260, 292, 327, 330 [experienced as

Jackson, M. 159
analyst’s own (clinical example)

Jacobson, E. 78, 79, 285

399

Index

Jarast, G. 323, 325–38

of parts of self-representation or of

Jiménez, E. 214, 215

object-representation 171;

Jimenez, J. P. 356, 357, 358, 362

transference of total situations 163;

Joffe, W. G. 179

unconscious functions of projective

Jones, E. 153, 154, 155, 316

identifi cation 112; see also depressive Jordan, J. F. 323, 354, 359, 360

position, paranoid–schizoid position

Jordan-Moore, J. F. 354–64

Kleinians/Kleinian concepts ( passim ): in

Joseph, B. 53, 55–7, 59, 113, 120, 130,

Belgium 221; in Brazil 340; British

162, 186, 198, 221, 355; acting in

18, 49–60, 156, 247, 254, 365; in

164; projective identifi cation

Canada 224–5; contemporary, ideas

[aggressive and negative aspects 56;

of, vs. Klein’s 50–1, 57, 60; vs.

clinical aspects 56, 98–111; clinical

Contemporary Freudian and

example 119; normal vs. pathological

Independent thinking 154–66; in

55]; transference 145, 156

France 220–1; in French-speaking

Jung, C. G. 154

world 218–19, 221–2, 225, 231

Junqueira Filho, L. C. U. 347–9, 352

[ambivalence towards/divergence

from 228, 232–4]; in Germany 188,

Karlsson, R. 257

189, 192, 193, 196, 202; growing

Kemper, W. 339

interest in 157; in Italy, 207, 210,

Kennedy, R. 158, 159, 160

211; in Latin America 325–6; in

Kernberg, O. F. 157, 241, 248, 249, 285;

Spain 213–16; in Switzerland 221,

as ‘adopter’ of concept of projective

224; in USA 242, 243, 247, 253,

identifi cation 253–4

[psychoanalytic literature 245–63]

King, P. 153, 156, 159, 162, 163, 164

Klüwer, R. 187, 202

Klauber, J. 155

Knight, R. P. 249, 266

Klein, M. ( passim ): aggressive object

Kohon, G. 154, 155

relation 5, 77, 195; and Bion,

Kohut, H. 157

contributions of inseparable 229–30;

Krafft-Ebing, R. von 336

concept of transference as a ‘total

situation’ 145; Controversial

La Farge, L. 255

Discussions 153, 155; ego’s vital need

Lacan, J. 224, 233–5, 335; ambivalent

to deal with anxiety 161; Fabian 59;

attitude towards Klein 220

infant’s fantasies of sadistic attacks

Langer, S. 363

upon breast 61, 72; ‘Notes on Some

Langs, R. 251, 257–8, 281, 292, 295

Schizoid Mechanisms’ 19–46; Patient Laplanche, J. 228

H 8, 9, 10, 12, 13; projection and

latent anxiety in schizoid patients 39, 42

introjection, simultaneous operation

Lebovici, S. 220

of 60; projection and projective

Liberman, D. 326

identifi cation, distinction between

libido: narcissistic 45; oral 27, 77

57; projective identifi cation [as

life instinct 45, 242, 243

aggressive mechanism 160;

linking, attacks on (Bion) 61–75

conceptualization of ( passim );

Loch, W. 186, 187, 194, 198, 199, 200

occurring in phantasy 171; published

Loewald, H. W. 157, 240, 243, 264, 268

mentions 4–6; unpublished notes in

Lorenzer, A. 202

Klein Archive 7–18]; schizoid

mechanisms, views on 77–9 [vs.

Madrid Psychoanalytical Association

Fairbairn’s views 21–2]; splitting off

207, 213

400

Index

Mahler, M. S. 79, 80, 85

feelings 73; split-off parts of ego

Malin, A. 239, 245, 249, 250, 264–74,

projected onto 27, 200

292
motivational cathexes 199

Mancia, M. 209, 212, 213

Manfredi Turilazzi, S. 207, 208, 209

narcissism 85, 140, 219; destructive 188;

manic-depressive disorders 21, 37, 39

omnipotent 84; primary 75, 215

manic-depressive patient(s) 21, 36, 43;

narcissistic identifi cation 214, 256; with

clinical example 37–8, 39–40

abandoned object 132–3

Martindale, B. 159

narcissistic induction (Eiguer) 227

Mason, A. 239, 255, 256, 301–19

narcissistic libido 45

Massi, M. 324, 339–53

narcissistic object relations 78, 302;

masturbation 270, 309, 312, 313

identifi cation in 55, 133; omnipotent

maternal projective identifi cations, role

85

of (Winnicott) 290

narcissistic patient(s) 231, 232, 235,

Matte-Blanco, I. 324, 358, 360

254

McDougall, J. 234
narcissistic regression 78

melancholia 39, 132; Freud’s

narcissistic transference (Roussillon)

mechanism in 133; hidden mania

227

implied in 133

nasal refl ex neurosis (Fliess) 316

Melanie Klein Archive 3, 7–18, 50, 161

negative therapeutic reaction 84, 270,

Melsohn, I. 344, 352

272, 335

Meltzer, D. 87, 124, 207, 214, 220, 240,

negative transference 272

309, 315; intrusive projective

neurosis(es) 21, 33, 35, 231, 233;

identifi cation 247

childhood 329; countertransference

Meltzoff, A. N. 213

334; nasal refl ex (Fliess) 316;

Merleau-Ponty, M. 334

transference 185, 227 [–

Meyer, L. 323–4, 342

countertransference 335]

Middle Group 153

neurotic patient(s) 188, 232, 235, 254,

Miller, M. 157

270

mirroring (Winnicott) 290, 317, 348,

Neyraut, M. 234

349

Nissim Momigliano, L. 207, 208

Mitchell, S. 240

non-discursive symbols 363

Mitrani, J. 246, 256

non-dream, paired (Cassorla) 351

Money-Kyrle, R. 6, 50, 53, 120, 121,

normal projective identifi cation 85,

156, 162, 305, 310, 314

345, 346, 347

mother ( passim ): as bad object, hatred

of 4; breast- 268 [primary

object: idealized, unassimilated 29;

identifi cation with 273]; breast of 18,

incorporation of 25; infant’s

20, 23, 24, 27, 28, 70; capacity of for

destructive impulse against 20;

reverie (Bion) 178; as container for

introjection of 21, 118, 138; splitting

infant 99; as ego-ideal 28; good-

of 20, 25, 26, 44

enough (Winnicott) 173, 285, 293

object relations: aggressive 5, 77, 195;

[‘holding’ function of 173]; infant’s

archaic 113, 127, 130; narcissistic 78

fusion with 80; -as-object 348;

[omnipotent 85]; of psychotic

phantasies of primitive oral

patient [role of projective

incorporation of 139; refusal of to

identifi cation in 76–97]; schizoid

serve as repository for infant’s

32–4, 43; theory 245, 251, 266

401

Index

objective countertransference 298

in normal development 313]; and

Oedipus complex 20, 61, 132, 220, 316

disturbances in thought-processes 34;

Ogden, T. H. 134, 157, 173, 246, 249,

and expressive perception (Melsohn)

254, 348, 363; analytic/

344; formulation of, 8, 98; and

intersubjective ‘third’ 241–2, 252,

identifi cation 256; mechanisms/

359, 361; projective identifi cation

defences in 43; in normal

241, 275–300, 359–61 [as cause of

development, earliest period of 265;

technical errors by analyst 210–11;

and omnipotence 198; pathology of

concept of, ‘adopter’ of 251–3; role

(Bion) 230; patients stuck in 100–5;

of, between intrapsychic and

and persecutory superego 74; and

interpersonal 168, 241, 260];

projective identifi cation 49, 56,

transidentifi cation 342

100–5, 167, 288–9, 302, 349; and

omnipotence 26, 77, 85, 99, 198, 214,

splitting 198; working through of,

308; therapeutic, analyst’s 291

failure to 21

omnipotent identifi cation 55, 58

parasitism 56, 60, 86, 87, 96

omnipotent narcissism 84

parental intercourse 66, 305, 319

omnipotent narcissistic object relations

Paris Psychoanalytical Society 234

85

Parsons, M. 162
omnipotent phantasy(ies) 85, 116, 193,

part-object projective identifi cation

302, 308, 315, 318, 333

(clinical examples) 306–15

omnipotent projective identifi cation

part-object relationships 69

82, 85, 88

part-objects 73, 265, 306, 314

oral aggression 24

Pasche, F. 228

oral libido 27, 77

pathological fusion 255

oral-sadistic impulses towards mother’s

pathological identifi cation(s) 58,

breast 24

132–46

O’Shaughnessy, E. 49, 120, 122, 123,

pathological introjection 137, 140

130, 149, 153–66, 205, 365–6

pathological introjective identifi cation

137

paired-non-dream (Cassorla) 351

pathological projection 137, 224

paradoxical thinking (de M’Uzan) 226

pathological projective identifi cation

paranoia 6, 10, 20, 31, 32, 45, 46, 303,

(Bion) 50, 55, 134, 144, 176, 190,

304, 306

230, 246, 333

paranoid anxiety(ies) 9, 10, 31, 76, 77,

pathological superego 317

90, 96, 115, 230, 314; early 19

patient’s projective identifi cations 72,

paranoid borderline patient(s) 254

291, 294, 298, 299; analyst’s openness

paranoid position 21

to 292

paranoid-schizoid phase 69, 74, 288

Payne, S. 154, 155

paranoid–schizoid position (Klein) 74,

Pérez-Sánchez, A. 215

155, 158; anxiety in 302 [and

persecutor/aggressor, identifi cation

schizophrenia 21, 35, 43, 45, 265];

with (A. Freud) 135, 157, 175

attacks on linking in 69; as

persecutory anxiety(ies) 22, 23, 26, 27,

combination of paranoid position

29, 32, 35, 77, 93, 99, 102, 142, 309;

and schizoid position 21; and

in infancy 43

depressive position 251, 252

persecutory fear of introjected object

[fl uctuations/oscillation between

10

(Bion) 34–8, 194; movement towards persecutory phase in infancy 21

402

Index

personality: aggressive parts of,

concept of, early discussion on 4;

projection of 306; psychotic part of

foreclosed (Green) 226; Freud’s

61, 75

concept of, vs. projective

phantasy: and behaviour, interaction

identifi cation 301–19; of goodness

between 201; of expulsion 301;

14, 253; and introjection [interaction

omnipotent 85, 116, 193, 302, 308,

between 16, 20, 30, 223, 302, 316;

315, 318, 333; projective

simultaneous operation of (Klein)

identifi cation as occurring in (Klein)

60]; of loving and hating parts, into

171; unconscious ( passim ) [projective objects; metabolized, internalization identifi cation as 17,
112]
of (clinical example) 282; into

Philips, F. 340, 341

object, of good and bad/loving and

phobia formation 105, 109

hating split-off parts of self 5, 17, 28,

Piaget, J. 180, 213

33, 77, 160, 264 [clinical example

Pichon-Rivière, E. 325, 326

307–8]; pathological 137, 224;

Pierrakos, M. 234

patient’s 57, 121, 356 [analyst’s

Planck, M. 192

capacity to contain 334, 356]; of

pleasure ego, purifi ed (Freud) 265

phantasy of object relationship 122,

Porto Alegre Psychoanalytical Society

130; and projective identifi cation,

340

distinction between 16, 18, 50, 57,

post-Kleinian conceptions 214, 222,

200, 201, 223, 247, 250, 254, 297;

224, 226, 325, 326, 338

unconscious, into analyst 118

potential space for play (Winnicott)

projective counteridentifi cation

232

(Grinberg) 224, 248, 298, 299, 330,

premature superego 362

333, 334, 342, 355

presentational symbols 363

projective dis-identifi cation 136, 140

primary aggression 70, 72

projective identifi cation(s): acquisitive

primary anxiety 23, 24

58, 59, 60, 247; adaptive (Blatt) 243,

‘primary confusion’ (Sandler) 180

258; aggressive and negative aspects

primary identifi cation (Freud) 179–81,

of (Joseph) 56; assaultive

273

(Rosenman) 247; attributive

primary instincts, Freud’s concept of 22

(Britton) 60, 247 [vs. acquisitive 58];

primary inter-subjectivity (Brazelton)

clinical value of 98–111; concept of,

213

use of (clinical examples) 357;

primary narcissism 75, 215

conceptualization of ( passim ) [critics

primary objects, concept of (Balint)

of 191–6; fi rst use of, by Weiss 3;

268

history of 3–4; implications of, for

primitive defence(s) 159, 253

technique and for clinical theory

primitive envy 84

292–9; Klein’s, central role of 50];

primitive introjection 297

conferences on 49 [Jerusalem (1984)

primitive projective reparation 167

50, 56, 157, 186, 197; University

process identifi cations (Danckwardt)

College London (1995) 50]; crossed

200

(Cassorla) 350, 351; and ego splitting

projection(s) ( passim ): of aggressive

76, 97 [clinical example 89–96];

parts of personality 306; into analyst,

evocative/evocatory (non-evocative/

of mad parts of self 81; into animals

non-evocatory) (Spillius) 58, 60, 136,

or insects 306; ‘classical’, vs.

190, 216, 247; as fantasy and object

projective identifi cation(s) 134;

relationship 276–84; good and bad

403

Index

(hostile and loving) aspects of self

80, 83, 96, 98–9, 104, 216;

involved in 8, 15, 17, 161; as

unconscious functions (Klein) 112;

interactional 250, 258; as

use of by depersonalised patient 55

interpersonal as well as intrapersonal

projective injection (Racamier) 226

process 60; intrapsychic 345;

projective reparation, primitive 167

intrusive (Meltzer) 247; Klein’s views projective transference 298

of ( passim ) [published 4–6;

projective transidentifi cation

unpublished 7–18]; massive 87, 110,

(Grotstein) 241

133, 134, 355, 357 [interplay of

psychic presences (Grotstein) 348

projective and introjective

psychoanalysis: as archaeological

mechanisms in, clinical examples

investigation (Freud) 68; of factors

139–46]; maternal 290; as

346;
mechanism 158, 168, 170, 174–6,

psychoanalytic fi eld (Barangers) 224,

198, 211, 359 [aggressive (Klein)

325, 334–7

160; Klein’s discovery of 133]; as

psychological change, projective

method of evacuation 112; motives

identifi cation as pathway for 285,

for 56, 246, 250, 318; and narcissistic

286, 300

omnipotent balance (clinical

psychopathology of psychotic states

example) 102–5; neurotic 230;

76–97

normal (Bion) 85, 345–7 [denial of

psychosis(es) 20, 21, 44, 46, 124, 159,

69–72; vs. pathological 50, 54–7, 60,

176, 231–4, 253, 344; borderline 61;

190, 230, 246]; omnipotent 82, 85,

countertransference 291; infantile,

88; origin of 137, 138; pathological

symbiotic 79; symbiotic 80;

(Bion) 50, 55, 134, 144, 176, 190,

transference 267 [transference–

230, 246, 333 [vs. constructive

countertransference 360]

aspects 60]; of patient 50–3, 120,

psychotic anxieties, in infancy 20

208, 247, 248, 279, 281, 293–5, 299,

psychotic disorders 20, 27, 170

314, 326, 330, 355, 360 [analyst’s

psychotic disorganization 79[

capacity to introject 72; containment

psychotic identifi cations in

of 294, 356]; permanent, with

schizophrenic patients 78

idealized bad object 135; as phantasy

psychotic part of personality 61, 75

99, 171; unconscious 17, 112;

psychotic patient(s) ( passim ): ego

processing phase of 288; and

structure and object relations of,

projection, distinction between 16,

projective identifi cation in 76–97;

18, 50, 57–8, 200, 201, 223, 247, 250,

projective identifi cation in,

254, 297; in psychotic patient,

psychoanalytic treatment of 88–96

psychoanalytic treatment of (clinical

psychotic states 168, 169, 175;

example) 89–96; ridding self of

psychopathology of 76–97

unwanted parts 80, 96; role of, in

psychotic transference 156

ego structure and object relations of

Purcell, S. 257

psychotic patient 76–97; as sequence

of steps 277; stages (Sandler) [fi rst

Quinodoz, D. 223, 224

170–1; second 171–2; third 173];

Quinodoz, J.-M. 185, 218–35

‘symbolic’ forms of phantasy and

thought in 50; term, use of,

Racamier, P.-C. 226

variations in 3, 51, 218; total, state of Racker, H. 51, 113, 176, 196, 254, 314, 85, 86, 97, 141;
types of 50, 58, 60, 323, 325, 337; complementary

404

Index

identifi cation, analyst’s with patient’s

Rosenman, S. 247

internal objects 177, 260;

Roth, P. 49, 260, 355, 357

countertransference 50

Roussillon, R. 226, 227

[complementary 172, 202, 258, 327,

Rusbridger, R. 142
330, 355; concordant 172, 202, 258,

327; and projective identifi cation

sadism 25, 302

326–9]

sado-masochistic analytic relationship

Rayner, E. 154, 155, 159

357

re-enactment, necessity for 122

Sandler, A.-M. 119

re-projection as enactment 121

Sandler, J. 113, 161; actualization,

reality principle 67

concept of 60, 119, 136, 158,

refl ectment (Junqueira Filho) 349

164, 177; projective identifi cation

regression 30, 35, 43, 83, 253, 255, 290,

135, 158, 167–81, 211–12

291, 335; narcissistic 78;

[conference, Jerusalem (1984)

topographical and formal 232

50, 56, 157, 186–7, 197; University

relational ghosts 348

College London (1995) 50]; on

Renik, O. 241

transference 197
reparation 33, 35, 53, 161; projective, Schafer, R. 157, 239, 240–4, 245, 251, primitive 167

297

repetition: compulsion 215; endless,

Scharff, J. S. 259

tendency for in analysis 94, 122;

schizoid anxieties, early 19

necessity for 122

schizoid defences 39–41

representative function (Melsohn) 352

schizoid mechanisms (Klein) 19–46, 77,

repression 26, 41, 77, 231, 301;

198; of annihilation of part of self 41

therapeutic 290

schizoid object relations 32–4, 43

resistance to interpretations 62

schizoid personality(ies) 22, 32; latent

reverie (Bion) 242, 349; maternal 178;

anxiety in 39, 42

mother’s capacity for 173, 230;

schizoid phenomena 22, 35, 41; and

Rey, H. 221

manic-depressive phenomena 36–9

Ribas, D. 223

schizoid position (Fairbairn) 21, 22

ridding oneself of part of self (clinical

schizoid states 29, 42, 301, 313

example) 277–8

schizophrenia(s) 37, 42, 46, 87, 97, 252,

Riesenberg Malcolm, R. 49, 57

270, 277; concrete thinking in 82;

Riviere, J. 6, 32, 154, 156, 325

delusions of grandeur and

Rocha Barros, E. L. da 101, 349–50,

persecution in 26; and melancholia,

352

distinction between (Klein) 39;

Rocha Barros, E. M. da 101, 349–50, 352

origins in early childhood [Fairbairn

Rodrigué, E. 326

22; Klein 20–1, 29–30, 35, 43, 45,

Rosenfeld, H. 6, 10, 55–6, 58, 59, 130,

265]; projective identifi cation in 85,

156, 188, 221; identifi cation in

290; splitting at root of 31; states of

narcissistic object relations 133;

disintegration in 24

intrusions 158; projective

schizophrenic dissociation 30

identifi cation 60, 76–97, 133, 168,

schizophrenic dreaming 65
246, 290, 356 [application of, in

schizophrenic individual 24

treatment of psychotic patient 267;

schizophrenic patient(s) 42, 87;

contributions to understanding of

borderline 270; clinical example

55–6; normal vs. pathological 55]

89–96; need of to project into

405

Index

therapist 267; psychotic

and denial 26; Klein’s views 7;

identifi cations in 78, 277

linguistic constraints in French 222;

Schreber (Freud’s patient) 43–6

of object 20, 25, 26, 98 [and self, and Schur, M. 316

projecting split-off parts into object

Scientifi c Discussions on Controversial

98]; of part of personality, violent

Issues 155

(clinical example) 41; and projection

Scott, C. 221

and introjection 25–32; in relation to

Scott, W. C. M. 25, 28

the object 24–5; of self, violent 32

Searles, H. F. 83, 249, 251, 264, 267–9,

splitting off and projection of bad/

291

envious parts of self 84, 85

Segal, H. 6, 49, 53, 57, 82, 156, 158,

Steiner, R. 153, 165, 190, 340, 355

163, 167, 168, 190, 192, 197, 221,

Strachey, J. 119, 316

249, 264; identifi catory projection

structural model (Freud) 209

226

Sullivan, H. S. 240

self, splitting of, violent 32

superego 153, 346; analyst’s 121;

self-object boundary 55, 58, 133, 180,

destructive 55, 74; early development

211; role of 175, 176

of 20, 73–4; formation (Freud) 243,

self psychology 157

316; internalized objects as 68, 74,

separation anxiety 23, 80, 84–6, 224,

132–3, 256, 304, 316 [identifi cation

313

with 175]; intropression 192;

sexual object choice 3

pathological (clinical example) 317;

Sharpe, E. 154, 155

primitive/premature 362

sibling jealousy, projection of (clinical

symbiosis 79, 80, 193; therapeutic 291;

example) 303

symbiotic infantile psychosis(es) 79

Sinason, M. 159

symbiotic membrane, mother–infant

Smith, H. F. 257

79

Sodré, I. 58, 59, 60, 132–46

symbiotic psychosis 80

Sohn, L. 59

symbiotic relatedness (Searles) 249, 267

Spain, concept of projective

symbol formation 82, 194, 230; and

identifi cation in 213–16

projective identifi cation 168;

Spanish Psychoanalytical Society 207,

symbolic communication 118, 119, 350

213–16

symbolic equation(s) 82, 156, 158

Speziale-Bagliacca, R. 209

symbolic production 350

Spillius, E. 3–18, 49–60, 149, 161, 185,

symbolic thinking 58, 83; and concrete

239, 245–63, 365–6; evocative/

thinking, in projective identifi cation

evocatory (non-evocative/non-

58–9;

evocatory) projective identifi cation

symbolization 82, 83, 198, 256, 288;

58, 60, 136, 190, 216, 247

failure in 141;

Spira, M. 220, 221, 224

Spitz, R. 285

Tähkä, T. A. 280

split-off parts of ego/self 27, 33, 45,

Tansey, M. 258, 259

171, 175, 200, 303; identifi cation of

Tavistock Clinic 251

object with 171; projected, return of

therapeutic misalliances (Langs) 295

135; projection onto mother 27, 200

therapeutic omnipotence, analyst’s 291

splitting ( passim ): within ego 24, 25, 31,

therapeutic repression 290

76, 78, 80, 89, 93, 96, 97, 222; of

therapeutic symbiosis 291

fi gure of analyst 40; and idealization

thinking: abstract 82, 83; concrete 82,

406

Index

83, 156, 230; symbolic 58, 83; theory

infant’s 194, 242; of living inside

of 164, 333; verbal 83

object 56; patient’s 58, 124, 131, 172;

Thorner, H. A. 186

in projective dis-identifi cation 136;

thought disorder (clinical example)

projective identifi cation as 17, 57,

305–6

112, 170, 249, 260

thought without a thinker (Bion) 289

Uruguayan Psychoanalytic Association

transference ( passim ): borderline 156;

338

–countertransference ( passim ) 225,

229, 231, 233; neurosis 335;

verbal thinking 83

psychosis 360; situation 229;

Vidermann, S. 234

delusional 267, 291; narcissistic

(Roussillon) 227; negative 272;

Wallerstein, R. 157

neurosis 185, 227; as phantasy about

Weiss, E. 3, 155

analyst (Klein) 171; projective 298;

Williams, P. 159

and projective identifi cation

Winnicott, D. W. 25, 154, 155, 157, 162,

197–200; psychosis 267; psychotic

193, 211, 231, 245, 251; ego, early,

156; relationship, aimed at

unintegration of, 23; good-enough/

controlling analyst’s body and mind

containing mother 230, 285–6, 293,

81, 96; of total situations (Klein) 163

296 [’holding’ function of 173];

transidentifi cation 241, 342

impingement 112, 114, 290;

transitional phenomena (Winnicott)

maternal projective identifi cations,

293

role of (Winnicott) 290; mirroring

transitional space 161

290, 317, 348, 349; objective

trauma of birth 23

countertransference 298; potential

Trinca, W. 345–7, 352

space for play 232; role of maternal

Tuckett, D. 113, 165

projective identifi cations 290;

transitional phenomena 293;

unconscious enactment 113

transitional space 161

unconscious phantasy 137, 153, 200,

Wisdom, J. O. 186

214, 250, 360; action potential of

wish-fulfi lment 119, 193

196–7; in analytic situation 113, 119,

Wolf Man (Freud) 309

177, 179, 333, 334, 336, 362; of

Wolff, E. 185

annihilation of part of ego 40; and

behaviour, interaction of 201–3;

Zwiebel, R. 186, 198

407

Document Outline

Projective Identification The Fate of a Concept

Copyright

Contents

Notes on contributors

Acknowledgements
Foreword

Part One Melanie Klein’s work

1 The emergence of Klein’s idea of projective identification in her published and unpublished
work

2 Notes on some schizoid mechanisms

Part Two Some British Kleinian developments

3 Developments by British Kleinian analysts

4 Attacks on linking

5 Contribution to the psychopathology of psychotic states: the importance of projective

identification in

the ego structure and the object relations of the psychotic patient

6 Projective identification: some clinical aspects

7 Projective identification: the analyst’s involvement

8 Who’s who? Notes on pathological identifications

Part Three The plural psychoanalytic scene

Introduction

SECTION 1 The British Psychoanalytic Society

9 The views of Contemporary Freudians and Independents about the concept of projective
identification

10 The concept of projective identification

SECTION 2 Continental Europe

Introduction

11 Projective identification: the fate of the concept in Germany

12 Projective identification: the fate of the concept in Italy and Spain

13 Projective identification in contemporary French-language psychoanalysis

SECTION 3 The United States

Introduction

14 Projective identification in the USA: an overview

15 A brief review of projective identification in American psychoanalytic literature

16 Projective identification in the therapeutic process

17 On projective identification

18 Vicissitudes of projective identification

SECTION 4 Latin America

Introduction

19 Projective identification: the concept in Argentina

20 Projective identification: Brazilian variations of the concept

21 Projective identification and the weight of intersubjectivity

Afterword

References

Index

Document Outline

Projective Identification The Fate of a Concept

Copyright

Contents

Notes on contributors

Acknowledgements

Foreword

Part One Melanie Klein’s work

1 The emergence of Klein’s idea of projective identification in her published and unpublished
work

2 Notes on some schizoid mechanisms

Part Two Some British Kleinian developments

3 Developments by British Kleinian analysts

4 Attacks on linking

5 Contribution to the psychopathology of psychotic states: the importance of projective

identification in

the ego structure and the object relations of the psychotic patient

6 Projective identification: some clinical aspects

7 Projective identification: the analyst’s involvement

8 Who’s who? Notes on pathological identifications

Part Three The plural psychoanalytic scene

Introduction

SECTION 1 The British Psychoanalytic Society

9 The views of Contemporary Freudians and Independents about the concept of projective
identification

10 The concept of projective identification

SECTION 2 Continental Europe

Introduction

11 Projective identification: the fate of the concept in Germany

12 Projective identification: the fate of the concept in Italy and Spain

13 Projective identification in contemporary French-language psychoanalysis

SECTION 3 The United States

Introduction

14 Projective identification in the USA: an overview

15 A brief review of projective identification in American psychoanalytic literature

16 Projective identification in the therapeutic process

17 On projective identification

18 Vicissitudes of projective identification

SECTION 4 Latin America

Introduction

19 Projective identification: the concept in Argentina

20 Projective identification: Brazilian variations of the concept

21 Projective identification and the weight of intersubjectivity

Afterword

References

Index
Document Outline
Projective Identification The Fate of a Concept
Copyright
Contents
Notes on contributors
Acknowledgements
Foreword
Part One Melanie Klein’s work
1 The emergence of Klein’s idea of projective identification in her published and
unpublished work
2 Notes on some schizoid mechanisms
Part Two Some British Kleinian developments
3 Developments by British Kleinian analysts
4 Attacks on linking
5 Contribution to the psychopathology of psychotic states: the importance of projective
identification in the ego structure and the object relations of the psychotic patient
6 Projective identification: some clinical aspects
7 Projective identification: the analyst’s involvement
8 Who’s who? Notes on pathological identifications
Part Three The plural psychoanalytic scene
Introduction
SECTION 1 The British Psychoanalytic Society
9 The views of Contemporary Freudians and Independents about the concept of
projective identification
10 The concept of projective identification
SECTION 2 Continental Europe
Introduction
11 Projective identification: the fate of the concept in Germany
12 Projective identification: the fate of the concept in Italy and Spain
13 Projective identification in contemporary French-language psychoanalysis
SECTION 3 The United States
Introduction
14 Projective identification in the USA: an overview
15 A brief review of projective identification in American psychoanalytic literature
16 Projective identification in the therapeutic process
17 On projective identification
18 Vicissitudes of projective identification
SECTION 4 Latin America
Introduction
19 Projective identification: the concept in Argentina
20 Projective identification: Brazilian variations of the concept
21 Projective identification and the weight of intersubjectivity
Afterword
References
Index