Understanding

Transference
The Core Conflictual Relationship
Theme method

LESTER LUBORSKY
PAUL CRITS-
CHRISTOPH
Copyright © 1998 Lester Luborsky & Paul Crits-
Christoph
e-Book 2019 International Psychotherapy Institute
All Rights Reserved

This e-book contains material protected under

International and Federal Copyright Laws and
Treaties. This e-book is intended for personal use
only. Any unauthorized reprint or use of this
material is prohibited. No part of this book may be
used in any commercial manner without express
permission of the author. Scholarly use of
quotations must have proper attribution to the
published work. This work may not be
deconstructed, reverse engineered or reproduced in
any other format.

Created in the United States of America

Table of Contents

Foreword
Martin E. P. Seligman

Foreword
Robert S. Wallerstein

Preface

Acknowledgments

Project Participants

Part I THE BASICS OF THE CCRT METHOD AND

ITS SCORING
Chapter 1. The Early Life Of The Idea For The Core
Conflictual Relationship Theme Method
Lester Luborsky
Chapter 2. A Guide To The CCRT Method
Lester Luborsky
Chapter 3. A Guide To The CCRT Standard
Categories And Their Classification
Jacques P. Barber, Paul Crits-Christoph, and
Lester Luborsky
Chapter 4. Positive Versus Negative CCRT Patterns
Brin F. S. Grenyer and Lester Luborsky
Chapter 5. Illustrations Of The CCRT Scoring Guide
Lester Luborsky and Scott Friedman
Chapter 6. The Reliability Of The CCRT Measure:
Results From Eight Samples
Lester Luborsky and Louis Diguer
 Chapter 7. The Relationship Anecdotes Paradigm
(RAP) Interview As A Versatile Source Of
Narratives
Lester Luborsky
Chapter 8. Why Each Ccrt Procedure Was Chosen
Lester Luborsky

Part II DISCOVERIES FROM THE CCRT METHOD

Chapter 9. The Narratives Told During
Psychotherapy And The Types Of CCRTS Within
Them
Lester Luborsky, Jacques P. Barber, Pamela
Schaffler, and John Cacciola
Chapter 10. Changes In CCRT Pervasiveness During
Psychotherapy
Paul Crits-Christoph and Lester Luborsky
Chapter 11. The Parallel Of The CCRT For The
Therapist With The CCRT For Other People
Deborah Fried, Paul Crits-Christoph, and Lester
Luborsky
Chapter 12. The Parallel Of The CCRT From Waking
Narratives With The CCRT From Dreams
Study 1: The Parallel Of The CCRT From
Waking Narratives With The CCRT From
Dreams
Carol Popp, Lester Luborsky, and Paul Crits-
Christoph
Study 2: The Parallel Of The CCRT From
Waking Narratives With The CCRT From
Dreams: A Further Validation
Carol Popp, Louis Diguer, Lester Luborsky,
Jeffrey Faude, Suzanne Johnson, Margaret
Morris, Norman Schaffer, Pamela Schaffler,
and Kelly Schmidt
 Chapter 13. The Measurement Of Accuracy Of
Interpretations
Paul Crits-Christoph, Andrew Cooper, and Lester
Luborsky
Chapter 14. Self-Understanding Of The CCRT
Paul Crits-Christoph and Lester Luborsky
Chapter 15. The Perspective Of Patients Versus That
Of Clinicians In The Assessment Of Central
Relationship Themes
Paul Crits-Christoph and Lester Luborsky
Chapter 16. Stability Of The CCRT From Age 3 To 5
Lester Luborsky, Ellen Luborsky, Louis Diguer,
Kelly Schmidt, Dorothee Dengler, Jeffrey Faude,
Margaret Morris, Pamela Schaffler, Helen
Buchsbaum, and Robert Emde
Chapter 17. Stability Of The CCRT From Before
Psychotherapy Starts To The Early Sessions
Jacques P. Barber, Lester Luborsky, Paul Crits-
Christoph, and Louis Diguer
Chapter 18. The Measurement Of Mastery Of
Relationship Conflicts
Brin F. S. Grenyer and Lester Luborsky

PART III CLINICAL USES OF THE CCRT

Chapter 19. The Everyday Clinical Uses Of The
CCRT
Lester Luborsky
Chapter 20. Alternative Measures Of The Central
Relationship Pattern
Lester Luborsky

PART IV WHAT’S NOW AND WHAT’S NEXT

Chapter 21. The Convergence Of Freud’s
Observations About Transference With The
 CCRT Evidence
Lester Luborsky
Chapter 22. Where We Are In Understanding The
CCRT
Lester Luborsky

References

About the Editors

 FOREWORD

Are you a psychotherapist of any stripe,

practicing or in training? Are you a clinical

researcher? Are you a patient in psychotherapy? If so,

this book will illuminate and guide.

Sigmund Freud believed that psychotherapy was

not effective merely because of the rapport between

doctor and patient, although he acknowledged that

rapport was vital. He asserted that cure took place
through the analysis of transference, a mysterious

process in which the ancient conflicts of the patient

were played out on the current stage of
psychotherapy. The leading role was, of course,
played by the patient, but the other leads— mother,
father, siblings, lovers, teachers—were played by the

therapist. The reenactment of the early relationships

in the current relationship with the therapist, and the

www.theipi.org 8
www.freepsychotherapybooks.org
 insight gained into them, constituted cure through
analysis of transference. A wealth of clinical

evidence has accumulated to support this idea.

Cognitive and behavioral therapists believe

something very different about the relationship

between patient and therapist. The relationship is
important, and a good one may be necessary for cure
to take place, but it is secondary. Cure is brought
about primarily by these techniques: systematic
desensitization, challenging automatic thoughts,

undoing depresso-genic assumptions, assertiveness

training, and the like. Furthermore, these techniques

can be scientifically studied and refined. An equally
impressive body of clinical and experimental
evidence has accumulated to support this opposing
view.

Understanding Transference is a bridge between

these two seemingly irreconcilable perspectives.

www.theipi.org 9
www.freepsychotherapybooks.org
 Lester Luborsky, Paul Crits-Christoph, and their
colleagues and students have taken the notion of

transference, objectified it, and shown us how and

why it works. These lessons apply to both worlds of

therapy. Luborsky’s thinking shows the way toward

reunification.

Most scientists take phenomena that are in the

light and shed further light on them. Other scientists
(brave souls) take phenomena that are in the dark and
bring them into the penumbra. I number Freud and

most of his followers among them. Still others,

perhaps those from whom we learn the most, take

what is in the penumbra and bring it into the light.
Lester Luborsky has, through his long and productive
career, always been an example of this last kind of
scientist. Understanding Transference is a sterling

example of his work.

www.theipi.org 10
www.freepsychotherapybooks.org
 The authors of this clear and wise book, which is
now revised and expanded, take the up-until-now

mysterious notion of transference (and the notions of

“insight” and “self-understanding,” as well) and show

us what it really is. They propose and confirm a

reliable measure of transference, the Core Conflictual
Relationship Theme (CCRT), which can be
objectively tested.

After validating the basic concept and its

measure, they show how the CCRT stands at the heart

of psychotherapy. The major contribution of this

book is to demonstrate that this reliable measure, the

CCRT, correlates meaningfully with other
theoretically related phenomena. Among these
attributes of the CCRT are (a) its pervasiveness
across relationships, (b) its similarity for the relation

to the therapist and the relations to other people, (c)

its appearance in different modes of expression—in

fact, it appears in both dreams and waking narratives,

www.theipi.org 11
www.freepsychotherapybooks.org
 (d) its usefulness to the therapist as a guide to
beneficial interpretations, and so on. These kinds of

correlates are part of the set of observations that

Freud made when he put forward the transference

concept. The research in this book implies that the

clinical-level psychoanalytic observations are well
worth following up through empirical studies.

What awaits the reader, then, is a powerful act of

demystification. A crucial idea, transference, that has
dwelt in the penumbra since Freud, now emerges into

the light.

With this book, a science of transference is finally

ongoing.

Martin E. P. Seligman

Professor of Psychology and former Director of

Clinical Training,
University of Pennsylvania, Philadelphia

www.theipi.org 12
www.freepsychotherapybooks.org
 FOREWORD

Few theoretical issues in psychoanalysis have

been as constantly and passionately argued as the

status of our discipline as a science. For Freud the

status was self-evident, and he labored unceasingly

over his lifetime to create and maintain a unitary

theoretical structure for psychoanalysis in a natural

science mold. However, he never encouraged the

empirical research through which science normally

tests and expands its hypotheses and accrues new

knowledge. Instead, he relied on his clinical case

study method to accumulate the observational base

that would then generate the causal explanatory

network of the metapsychology (the general theory)

that he was so painstakingly elaborating.

Nonetheless, it was only in the so-called classical,

or mainstream, ego psychology development in

www.theipi.org 13
www.freepsychotherapybooks.org
 America that this legacy of a unified theoretical
structure within a natural science framework endured.

The whole meta-psychological edifice in the first

post-World War II decades was brought to its position

of almost unquestioned hegemony, at least within

American psychoanalysis, in the ego psychology
associated with the names of Hartmann, Kris,
Loewenstein, Rapaport, and a host of others. In

Europe, Latin America, and the rest of the world, it

was quite otherwise from the start. What we have

come to call our psychoanalytic theoretical diversity,

or pluralism, began even during Freud’s lifetime with

the emergence of Kleinian analysis, ideas developed

out of Melanie Klein’s initial work with children.

These ideas evolved into an alternative

metapsychology and an alternative theory of

technique that soon claimed as many adherents as did
Freud’s structural theory and its ego psychology, if

not more. Then, in an almost bewildering

www.theipi.org 14
www.freepsychotherapybooks.org
 development, came the British object relations
school, coalescing out of the innovative thinking and

creative theorizing of a galaxy of original minds—

Fairbairn, Winnicott, Guntrip, Balint, Bowlby, and

others—followed by the far-reaching Bionian

extensions of Kleinian thinking. From France, and in
a distinctively French voice, came Lacan’s linguistic
conceptualization of the nature of the psychoanalytic

enterprise. This in turn was part of the larger fabric of

a Franco-German hermeneutic accounting of

psychoanalysis, sparkplugged by Habermas and

abetted by Ricoeur. It was an effort, embedded in the

critical theory of the Frankfurt school, to turn

psychoanalysis entirely away from what were

declared to be misguided positivist, natural science

directions, unfortunate carryovers of Freud’s 19th-

century natural science roots in Helmholtz school
physiology.

www.theipi.org 15
www.freepsychotherapybooks.org
 Two main trends have become clear in this steady
progression of ramifying theorizing in

psychoanalysis. One is the increasing theoretical

diversification within the overall psychoanalytic

corpus. The other is the progressive erosion of the

(natural) science commitment as the identifying and
determining hallmark of the discipline. Both these
trends have by now become well established in the

United States. Even early in the period of almost

monolithic supremacy of the ego psychological

metapsychology paradigm, dissident perspectives

arose. They included the Sullivanian interpersonal

school and the Horneyan culturalist movement, but

they were pushed to the margins of organized

psychoanalysis or extruded altogether. Subsequent

decades brought first a small enclave of Kleinian

thinking to the United States, followed in more recent
years by the explosive emergence of Kohut’s self

psychology as a completely alternate metapsychology

www.theipi.org 16
www.freepsychotherapybooks.org
 (and theory of technique). Kohut fashioned a
psychology primarily of deficit and its repair, rather

than central conflict and its resolution. In the same

years, a U.S. object relations stream crystallized,

coalescing out of the work of Jacobson, Mahler, and

Kernberg and building on its British progenitors as
well.

Even within the once almost monolithic domain

of ego psychology in the United States, varieties of
divergent and revisionist theoretical positions have

emerged, albeit with contrapuntal, passionate

defenses by its continuing adherents. This is what I

have called the Great Metapsychology Debate in our
field. Certainly in today’s post-ego psychology world,
the United States has staunch and persuasive
advocates (Gill, Schafer, G. Klein, Spence, and

others) of all the varieties of hermeneutic,

phenomenological, exclusively subjectivistic, or

linguistically based conceptualizations of the field.

www.theipi.org 17
www.freepsychotherapybooks.org
 Together, these advocates challenge our traditional
and customary conception of our discipline as a

reasonably uniform entity—and that entity properly a

science.

Where does the history explored leave our

century-old discipline today? Certainly, for those of
us who have commitments to the empirical enterprise
as the route to the organization and advance of
knowledge in disciplines putatively rooted in science,
it is by now abundantly clear that today

psychoanalysis worldwide consists of multiple and

divergent theories of mental functioning, psychic

development, pathogenesis, treatment, and cure.
Many of these theories claim in varying degrees a
natural science framework, whereas some repudiate
such a heritage altogether. My own firm conviction is

that, in their present stage of conceptual development

with regard to the logic of theory construction, our

psychoanalytic general theories, our

www.theipi.org 18
www.freepsychotherapybooks.org
 metapsychologies—and here I interpolate a reminder

of Freud’s metaphoric commentary, our witch

metapsychology—our widely different theoretical

frameworks that mark our psychoanalytic pluralism,
are to a major extent primarily still metaphoric. They

are merely large-scale explanatory metaphors, or

symbolisms, that we employ to give a needed sense

of coherence and closure to our psychoanalytic

understandings and therefore to our interventions. To

me, they are still only the metaphors we live by, our
pluralistic psychoanalytic articles of faith. In our
current developmental stage, as general theory, or
metapsychology, none of them is formulated in ways

amenable to empirical study and scientific process,

even though they seem to be cast within a natural

science explanatory language (the ego psychology

paradigm) or, oppositely, into an avowedly

hermeneutic—and anti-natural science—system of
language.

www.theipi.org 19
www.freepsychotherapybooks.org
 And yet we do have—and must have—a

psychoanalytic common ground that enables us all to

be recognizably psychoanalysts doing reasonably
comparable clinical work with reasonably
comparable patients around the world, whatever our

theoretical allegiance or our regional, cultural, or

linguistic perspective. This common ground I find not

in our overarching high-level general theories, or

metapsychologies, for all the reasons adduced.

Rather, the unifying element of psychoanalysis lies in

our contrasting low-level, experience-near, and
common, clinical theory, the level of theory of

transference and countertransference, of resistance

and defense, of anxiety and conflict and compromise,
of self and object representation. It is at this level that
the empirical referents—the clinical phenomena of
our consulting rooms—link by means of discernible

and traceable canons of inference to these clinical

theoretical constructs. Furthermore, it is at this level

www.theipi.org 20
www.freepsychotherapybooks.org
 that these clinical constructs can be put to the kind of
empirical study and test that will determine the extent

to which, in psychoanalysis, we are indeed fashioning

a body of science that in turn can guide our clinical

therapeutic work ever more precisely.

It is at this point and at this level that Lester

Luborsky and Paul Crits-Christoph, and their
collaborators and students, have pitched their
investigative activity over many years of hard work.
They have fastened on the centrality of the clinically

derived conception of the idiosyncratically evolved

core neurotic conflict that powers the guiding

unconscious fantasies and their manifestations,
lifelong in character and in symptom, and that
emerges in the therapy of our patients as the major
transference paradigm. They have endeavored to

operationalize this conception to render it amenable

to the usual varieties of scientific study and testing.

What has gradually evolved, and is continually being

www.theipi.org 21
www.freepsychotherapybooks.org
 refined and clarified, is their operationalized

conception, the Core Conflictual Relationship Theme

(CCRT). They have developed a strategy for reliably

recognizing the CCRT through the study of the

interpersonal narratives recounted in analytic therapy,

which they call relationship episodes (REs). This
transition from the unguided use of the transference
concept—that is, in the clinical formulations of the

psychoanalytic therapist—to an operational measure,

the CCRT, based on stepwise, systematic, guided

formulation methods, is key in their work. The

possibilities for achieving reliability in clinical
judgment and then pushing on to explore conceptual
validity, empirical usefulness, scientific hypothesis
testing, and solidly established new knowledge
accrual in psychoanalytic investigation are all

dependent on this concept.

This is the ambitious agenda and it is, to me,

squarely in the proper domain. Scientific advance can

www.theipi.org 22
www.freepsychotherapybooks.org
 and must be made, in our discipline, within the
common ground of our experience-near clinical

theory constructions, if indeed scientific advance is to

be made at all. It is, of course, up to readers to decide

for themselves how well Luborsky and Crits-

Christoph succeed in rendering a meaningful study of
the transference concept, in enlarging the possibilities
for increasing useful knowledge about it, and in

garnering investigative rewards from the continuing

lines of research that they outline—as well as how

much all this advances our field scientifically. What

those of us interested in the ultimate fate of

psychoanalysis as a science can certainly fully agree

on is that this is precisely the kind of research that is

necessary to be able to answer these questions

properly and fairly. For this, our discipline owes these

authors and their devoted labor of love, so lucidly
chronicled in this newly revised edition of the book, a

collective vote of great gratitude.

www.theipi.org 23
www.freepsychotherapybooks.org
 Robert S. Wallerstein

Professor of Psychiatry,
University of California at San Francisco

www.theipi.org 24
www.freepsychotherapybooks.org
 PREFACE

This second edition of Understanding

Transference has much in it that is new, along with

the essentially unchanged CCRT method. All of the

chapters have been revised and updated, and six are

entirely new to the book. Some of the chapters had

their origins in coauthored journal articles, but they

too have been reedited. Taken all together, this new

book gives the most complete guidance to users of

the CCRT method, adds to the readers’ knowledge of

discoveries about the patients’ central relationship

patterns, and points to the clinical applications of the

CCRT method.

The revised book’s contents can be best summed

up by listing each chapter’s special contribution. Part

I has eight chapters surveying the idea for the CCRT,

its history, its source in the narratives spontaneously

www.theipi.org 25
www.freepsychotherapybooks.org
 told in psychotherapy sessions (chapter 1), its precise
scoring rules (chapter 2), and its standard categories

(chapter 3). The scoring of the positive and negative

qualities of CCRT patterns is defined and examined

(chapter 4); detailed illustrations are given of the

CCRT scoring of narratives that can speed up the
learning of the scoring procedures (chapter 5); and
the essence of the largest collection of results of

independent scoring of 9 samples of patients is

presented (chapter 6). We also offer another method

for obtaining narratives: a versatile, easily used

alternative source of narratives told on request, the

Relationship Anecdotes Paradigms interview (chapter

7). This major section ends with explanations for why

each of the CCRT procedures was chosen (chapter 8).

The book then takes the reader further into the

facets of validity of the CCRT: In Part II it tells of our

discoveries during 10 explorations to find the

meaning of the CCRT measure. These chapters report

www.theipi.org 26
www.freepsychotherapybooks.org
 the most typical types of narratives (chapter 9), the
pervasiveness and generality of the CCRT within

sessions and across the treatment (chapter 10), the

degree of parallel of the CCRT for the therapist

versus the CCRT for other people (chapter 11), and a

newly expanded study of the parallel of the CCRT
within narratives as compared with the CCRT within
dreams (chapter 12). The chapters then deal with

three aspects of the patient’s and the therapist’s

“accuracy” in using the CCRT: (a) the accuracy of the

therapist’s interpretations (chapter 13), (b) the

accuracy of self-understanding (chapter 14), and (c)

another kind of accuracy, the degree of parallel

between the patient’s and the clinician’s perspective

on the CCRT (chapter 15). Finally, two new

explorations with the CCRT reveal its considerable

consistency across time, (a) from age 3 to age 5
(chapter 16) and (b) from before psychotherapy to

during psychotherapy (chapter 17). This section ends

www.theipi.org 27
www.freepsychotherapybooks.org
 with an extension beyond the CCRT: a reliable
method for moving beyond the assessment of the

conflicts in the usual CCRT to evaluating the degree

of mastery of the conflicts shown in the narratives

(chapter 18).

The book turns in Part III to the even more

clinical applications of the CCRT method, with a
chapter on the everyday uses of the CCRT in practice
(chapter 19), and a new exploration and the most
complete set of descriptions of the continually

lengthening stream of alternative central relationship

pattern measures (chapter 20).

The book ends with an integrative wind up, the

two broad summary chapters of Part IV. The first is

on the convergences of the many CCRT findings with

Freud’s 23 observations about the transference

pattern (chapter 21), and the last chapter is an even

more reflective summary: a status report on the

www.theipi.org 28
www.freepsychotherapybooks.org
 book’s contributions, a takeoff into the stratosphere of
super-clinical theories that account for the existence

and maintenance of a CCRT in everybody, along with

a set of prescriptions for the continued healthy

growth of the CCRT measure (chapter 22).

www.theipi.org 29
www.freepsychotherapybooks.org
 ACKNOWLEDGMENTS

With thanks to these sources of financial support:

The first seed money to nourish the germination

of the CCRT idea that sprouted in 1974 providently
came from the Luborsky Biopsychosocial

Foundation. The earliest substantial sustenance for

the CCRT method was a 2-year grant from the Fund

for Psychoanalytic Research of the American

Psychoanalytic Association (1 January 1984 to 31
December 1985). The long-term continuance of the

research on the CCRT method was sustained by a 3-

year National Institute for Mental Health (NIMH)
grant (R01MH39673) to me, which was then renewed
for another 3 years (1 December 1987 to 30
November 1990), and by another grant

(RO1MH40472) to Paul Crits-Christoph. Over the

expanse of time from 1968 to the present, my

www.theipi.org 30
www.freepsychotherapybooks.org
 program of research has been supported in part by
USPHS NIMH Research Scientist Award MH407010

and National Institute on Drug Abuse Research

Scientist Award DA00168. Paul Crits-Christoph’s

research, which has provided much help to our

collaboration, has been supported in part by USPHS
NIMH Career Development Award MH00756 and
Coordinating Center Grant U18-DA07090.

With thanks to these supportive people and

places:

I thank the Menninger Foundation for providing

the fertile locale for my 13 years of satisfying work
(1947-1959), with the last 6 as part of the Menninger
Foundation Psychotherapy Research Project with

Robert Wallerstein as director. The years at

“Menninger’s” helped me develop sophistication

about central relationship patterns, especially through

the central relationships with such stars in the field as

www.theipi.org 31
www.freepsychotherapybooks.org
 David Rapaport, Karl Menninger, George Klein,
Robert Wallerstein, Robert Holt, Philip Holzman,

Herbert Schlesinger, Donald Spence, Otto Kernberg,

Margaret Brenman-Gibson, Howard Shevrin, Merton

Gill, Gardner Murphy, Lois B. Murphy, and Hartvig

Dahl. After the Menninger period, Robert Rosenthal
of Harvard University has been a crucially generative
guide.

The earliest collaborating group in Philadelphia

(1978-1979) that rallied around to try the CCRT

scoring method on psychotherapy transcripts included

Frederic J. Levine, Richard Kluft, Thomas Wolman,

and myself.

For several years, stimulation of ideas and

support for myself and Paul Crits-Christoph came

from participation in the Open Laboratory on

Conscious and Unconscious Processes (sponsored by

the John D. and Catherine T. MacArthur Foundation),

www.theipi.org 32
www.freepsychotherapybooks.org
 whose director, Mardi Horowitz, shares basic
interests with us in relationship schemas.

The Ulm (Germany) University Department of

Psychotherapy and Psychosomatics, with Horst

Kächele (current chairman) and Helmut Thoma (past

chairman), has continued for almost 2 decades to
keep up an exchange of ideas through seminars that
led to the first published guide to the CCRT
(Luborsky & Kächele, 1988). From the Ulm group
came Dorothee Dengler to help apply the CCRT to

narratives told by 3-year-olds, with data generously

lent by the University of Denver’s Helen Buchsbaum

and Robert Emde. From the Ulm group Robert Eckert
also worked with us for two periods of several
months each and was the mainstay of the CCRT study
of depression. The Breuninger Foundation of West

Germany helped to finance the early Ulm exchanges,

for which I am much indebted. Horst Kächele and his

Ulm group have continued to be outstandingly

www.theipi.org 33
www.freepsychotherapybooks.org
 productive with CCRT research. The culmination of
their collaboration came in 1995, with the 10th

anniversary of their Ulm CCRT group: They staged a

beautiful and illuminating conference and

celebration, with CCRT articles presented by

researchers from the far corners of the globe,
including Australia, Japan, and Russia.

Continual backing has come from colleagues

within our own Department of Psychiatry at the
University of Pennsylvania, led by its chairman, Peter

Why brow, who has effectively encouraged and

supported us in our research. Special among the

steadfast friends and colleagues in the department
who have helped on a variety of projects over many
years have been A. Thomas McLellan, George
Woody, and Charles O’Brien; the many joint papers

with them reflect the productivity of our collaboration

and include various “firsts” in method development.

Our colleague Jacques Barber in our University of

www.theipi.org 34
www.freepsychotherapybooks.org
 Pennsylvania Center for Psychotherapy Research did
a detailed editorial and evaluative review of the first

edition and served as an able participant in many of

its research studies.

A 2-decade-long collaboration with Martin E. P.

Seligman of the Department of Psychology has
generated advances in our powers for explaining
narratives. Hartvig Dahl of Downstate Medical
Center has been generous with research exchanges
and with case materials, in particular for the dream

study described in chapter 12, this volume. Herbert

Schlesinger, head of clinical psychology at the New

School for Social Research until recently, has helped
with clinical and research ideas and assisted with
making transcripts of sessions containing dreams.

Among the people who took part in applying the

CCRT in recent studies were these significant long-

term players in the advancement of CCRT research:

www.theipi.org 35
www.freepsychotherapybooks.org
 Jim Mellon, Scott Friedman, Paul van Ravenswaay,

Anita V. Hole, Anna Rose Childress, Amy Demorest,

Karen Stewart, Jeffrey Faude, Laura Dahl, and David

Mark. Jim Mellon of the East-West University in
Hawaii worked full time for about 4 years as a highly

reliable CCRT judge and data overseer. For at least 5

years, Carol Popp of Emory University directed and

organized studies of dreams for chapter 12. Robert

Waldinger of the Massachusetts Mental Health Center

and the Judge Baker Foundation in Boston has

developed precise CCRT scoring systems and
recently provided a thoughtful, helpful critique of the
entire manuscript based on his deep understanding of

the CCRT. In the management of the computer’s

prodigious output of chapters, Joyce Bell has been

marvelously speedy, reliable, and knowledgeable.

The constant efficient research assistance of Suzanne

Johnson lightened the load of the arduous tasks of
many different aspects of writing the revision of the

www.theipi.org 36
www.freepsychotherapybooks.org
 book and doing the research. John Cacciola did an

expert job as evaluator and diagnostician of the

patients in some samples. For the past year, Jill
Levine has ably overseen many of the revisions;
recently she has been succeeded by the equally

capable Elizabeth Krause. For the past 2 years, David

Seligman has been the organizational overseer of the

research and its presentations; he was helped by

Amanda Horn and, more recently, by Joanna

Liebman, Jessica Kline, Alicia Starkman, Avi Benus,

Julie Kilman, Abraham Cotto, and Niharika Desai
who have lightened the load of the CCRT-related

studies. In 1995-1996 Monica Bishop served as an

editorial coordinator participating in the completion
of the editorial process. Gregory Halpern was
resident guru on permissions and oversaw the last
checking of the final corrections of all chapters.

Ellen Luborsky has been an ingenious general

advisor about style and order of presentation and

www.theipi.org 37
www.freepsychotherapybooks.org
 consultant on developmental theory. Lise Luborsky
has provided expert legal opinion. Miranda Outman

has been helpful with editorial improvements. Peter

Luborsky and Catherine Goubault Luborsky have

been there to consult with when translations were

needed. Paul Gerin of Lyon, France, and James Bond
and Howard Shevrin of the University of Michigan
have a special place in the field of CCRT research

because they were early outside-of-Philadelphia

contributors.

My wife, Ruth Samson Luborsky, as always, has

been a supremely supportive, expressive, and

versatile facilitator of the prospering of this book,
both in its exposition and in its scope.

The American Psychological Association has

been a superb publisher. My special thanks to my old

friend Gary VandenBos, its executive director, and to

his helpful director of APA Books, Julia Frank-

www.theipi.org 38
www.freepsychotherapybooks.org
 McNeil, as well as to his gifted advisor and

acquisitions editor, Peggy Schlegel. Andrea Phillippi,

as development editor, generated a sophisticated,

detailed (20 plus pages) evaluation of the entire

manuscript; it gave a big boost to the quality of the

exposition. The contribution that concluded the
book’s production was carried out by Ed
Meidenbauer, the technical/production editor.

Because some of the multiauthored chapters had

been previously published or presented, their authors

were again included, in order to recognize their past
contribution. Yet in the present edition, I was

primarily responsible for changes from and additions

to the original versions that rounded out and updated

each chapter and made each chapter consistent with

the rest of the book in style and content.

Lester Luborsky

www.theipi.org 39
www.freepsychotherapybooks.org
 PROJECT PARTICIPANTS

Jacques P. Barber, PhD, University of Pennsylvania

Helen Buchsbaum, PhD, University of Colorado, Denver

John Cacciola, PhD, Philadelphia VA Hospital and

University of Pennsylvania

Andrew Cooper, PhD, Philadelphia, PA

Amy Demorest, PhD, Amherst College

Dorothee Dengler, MD, University of Ulm, Germany

Louis Diguer, PhD, Laval University, Canada

Robert Emde, MD, University of Colorado, Denver

Robert Eckert, MD, Numberg, Germany

Jeffrey Faude, PhD, University of Pennsylvania

Deborah Fried, MD, Yale University

Scott Friedman, PhD, Georgia Institute of Technology

Brin F. S. Grenyer, PhD, University of Wollongong,

Australia

Anita V. Hole, PhD, Philadelphia VA Hospital

www.theipi.org 40
www.freepsychotherapybooks.org
 Suzanne Johnson, BA, Temple University

Horst Kächele, MD, University of Ulm, Germany

Jill Levine, BA, Boston University

Ellen Luborsky, PhD, Riverdale Mental Health Center,

New York

Jim Mellon, AB, University of Hawaii at Hilo

Margaret Morris, BA, University of New Mexico

Carol Popp, MD, PhD, Emory University

Norman Schaffer, PhD, Interpsych Associates, King of

Prussia, PA

Pamela Schaffler, AB, Harvard School of Public Health

Kelly Schmidt, BA, George Washington University

David Seligman, BA, Boston University

Paul van Ravenswaay, MD, Philadelphia, PA

Robert Waldinger, MD, Massachusetts Mental Health

Center and Judge Baker Children’s Center, Boston

www.theipi.org 41
www.freepsychotherapybooks.org
 I

THE BASICS OF THE

CCRT METHOD AND ITS
SCORING

www.theipi.org 42
www.freepsychotherapybooks.org
 1

THE EARLY LIFE OF THE IDEA FOR

THE CORE CONFLICTUAL
RELATIONSHIP THEME METHOD

LESTER LUBORSKY

A new measure of personality, the Core

Conflictual Relationship Theme (CCRT) method, is

what this book is about. The CCRT is the central

relationship pattern, script, or schema that each

person follows in conducting relationships. It is
derived from the consistencies across the narratives

people tell about their relationships.

The measure came into being quietly, for at the

time it was conceived there was not much sense of

what it would become. At first it was just an offshoot

of another measure, the therapeutic alliance

www.theipi.org 43
www.freepsychotherapybooks.org
 (Luborsky, 1976). But once the relationship pattern in
the alliance measure had been shaped, its use led to

the natural next question: How does the relationship

pattern in the alliance fit into the broader central

pattern of relationships? If, for example, a patient is

ready to feel helped by the therapist and shows other
signs of a positive alliance, such an alliance might be
part of a general pattern: the patient’s readiness to

feel helped by other people as well as by the

therapist. This broadened perspective led to the first

glimpse of the scope of the idea that would become

the CCRT method.

A closer look at the new measure followed my

playing around with systems for inferring a general
pattern of relationships from the transcripts of a set of
psychotherapy sessions. In this exercise I tried to

trace the bases for my own inferences about a general

relationship pattern as they emerged while reading

transcripts of sessions. My first self-observation was

www.theipi.org 44
www.freepsychotherapybooks.org
 that I was making most inferences while I was most
closely attending to the patient’s narratives of

interactions with the therapist and with other people.

Then, as I continued to read the narratives, I became

aware that I was particularly attentive to their most

recurrent interactions. After that I came to realize
that I was paying most attention to three facets of
these interactions: what the patient wanted from the

other people, how the other people reacted, and how

the patient reacted to their reactions. After trying

these and other facets, I came back to these three as

the most routinely evident and serviceable for

inferring the general relationship pattern.

After a couple of years of practice in identifying

these categories and then of my colleagues’ trying to
do the same, it became clear that the Core Conflictual

Relationship Theme method was ready to be born and

to begin an independent life. I first showed it off at

the Downstate Medical Center meeting on

www.theipi.org 45
www.freepsychotherapybooks.org
 Communicative Structures and Psychic Structures on
Saturday, January 17, 1976, at 2 in the afternoon

(Luborsky, 1977b); it showed clear signs of fitting its

name and already contained all of the essential

qualities of the measure described in chapter 2, this

volume. It looked like this:

The CCRT looked much like it belonged in the

family of Freud’s (1912/1958a) concept of the

transference template. Yet it had the special gift of
being endowed with a reliable system for guiding
inferences about each person’s recurrent central
relationship pattern. It behaved much as many

experienced psychodynamic clinicians do in making

their usual inferences in formulating transference

www.theipi.org 46
www.freepsychotherapybooks.org
 patterns, but it relied on formalized explicit principles
of inference making. Like the inferences of clinical

judges, the inferences that are aided by the CCRT

rely on the three facets of narratives about

interactions with other people: the types of wishes,

needs, and intentions concerning the other person;
responses from the other person; and responses of the
self. The final step in coming to the CCRT measure is

the combination of the most pervasive of each type of

these components found across the sample of

narratives.

The fashioning of the CCRT is a resounding

success story; it has transformed a useful clinical
concept into an even more useful clinical-quantitative
measure (Luborsky, in press). Over the past 50 years,
I was doing what clinicians typically do in the course

of each session: shaping a formulation of the patient’s

central relationship patterns so that I could derive

interpretations that fit the formulation. Now, with the

www.theipi.org 47
www.freepsychotherapybooks.org
 use of the CCRT method, I had the support of a
reliable guided system to help with this routinely

necessary task.

THE LINEAGE OF THE CENTRAL

RELATIONSHIP PATTERN CONCEPT

Naturally, after having looked after its inception,

gestation, and growth, all the way along up to its

young maturity, I began to be more and more curious

about the concept’s lineage, and so I dug up more of

its background and placed what I unearthed of its

relatives into the following five generic categories.

From the Psychoanalysts

I reread Freud’s ur-accounts of transference,

especially his 1912 “Dynamics of Transference.” I

expected some congruence between the observations

that led Freud to his concept of a transference

relationship template and the CCRT results, but the

high degree of congruence was striking. The parallels

www.theipi.org 48
www.freepsychotherapybooks.org
 are marshaled in detail in chapter 21, this volume,

and Paul Crits-Christoph and I have tested some of

them, with results given in other chapters. Freud’s

transference template appeared, with each study, to be
a good candidate for a cognate concept of the central

relationship pattern.

Several analysts since Freud have rediscovered

concepts like the central relationship pattern or the

transference template. A similar concept was posited

by Bios (1941), who used the term residual trauma.

French and Wheeler (1963) suggested the related idea

of a single “nuclear conflict” in each patient.

Arlow (1961, 1969a, 1969b), writing in 1961

about the recurrence of a single, overwhelmingly
pervasive theme, said that “fantasies are grouped
around certain basic instinctual wishes and such a
group is composed of different versions or different

editions of attempts to resolve the intrapsychic

www.theipi.org 49
www.freepsychotherapybooks.org
 conflict over these wishes” (p. 377). In one of his

articles (1969a), he developed the idea further:

The organization of these fantasies takes shape

early in life and persists in this form with only
minor variations throughout life. To borrow an
analogy from literature, one could say the plot
line of the fantasy remains the same although the
characters and the situations may vary. (p. 47)

His last sentence states vividly the observation that

has emerged from research with the Core Conflictual

Relationship Theme method. Arlow, in his discussion

of my Downstate Medical Center findings (Luborsky,

1977b), viewed the Core Conflictual Relationship

Theme as an offshoot of a more basic substrate

composed of unconscious fantasies. Actually, both

may be the product of highly ingrained patterns or

scripts or schemas of relationships. The heart of both

the unconscious fantasies and the Core Conflictual

Relationship Theme of the narratives may be

fruitfully conceived of as related expressions of the
pervasive central relationship patterns that are

www.theipi.org 50
www.freepsychotherapybooks.org
 expressed when wishes are activated toward other
people and even to the self.

Several other analysts, on the basis of follow-up

sessions, have examined recurrent conflictual

relationship patterns that persist long after treatment

has been terminated, even after very successful
treatments. For example, Pfeffer (1963) wrote, “In
analysis repetition is not eliminated but the content
and substance of what is repeatable is changed” (p.
241). Similarly, Schlessinger and Robbins (1975)

described a patient’s follow-up that illustrates the

preservation of conflictual themes after the analysis

but with the difference that “the significant outcome
of the analysis is the development of a self-analytic
function” (p. 781). Time trends in transference within
analyses have been reported by Graff and Luborsky

(1977), who compared two relatively more successful

analyses with two relatively less successful analyses

by means of daily postsession therapist ratings of

www.theipi.org 51
www.freepsychotherapybooks.org
 amount of transference and resistance. Transference

was even more evident in the later than in the earlier

stages of the more successful analyses, suggesting

that the earlier concept of a reduction in the amount

of transference in a successful analysis (Ekstein,

1956) should be reformulated.

From the Personality Researchers

Henry Murray (1938), the preeminent

personologist, proposed a Thematic Apperception
Test (TAT) scoring principle that reflects a

rediscovery of a version of Freud’s (1912/1958a)

concept of the transference template. It is in a little-

known footnote, at the end of a chapter in Murray’s

Explorations in Personality (the passage was pointed

out to me by Robert R. Holt, personal

communication, 1978). Murray referred to the

“principle of unification which raises certain interests
to the apex of the hierarchy of aims” (p. 396). But
Murray never fully presented this idea of a central

www.theipi.org 52
www.freepsychotherapybooks.org
 relationship pattern; in his footnote he deferred to the
need to conserve space. His concept would have

advanced research in the same domain as the

transference pattern and the CCRT if it had been

followed up. Later in the same book, in the

introduction to one of the case histories, Murray
(1938) further explained his “unity-thema” and the
central relationship pattern concept evident in it:

Experience was to teach us that ... it was

possible to find in most individuals an
underlying reaction system, termed by us unity-
thema, which was the key to his unique nature ..
. because if one assumed the activity of this
unity-thema many superficially unintelligible
actions and expressions became, as it were,
psychologically inevitable. A unity-thema is a
compound of interrelated—collaborating or
conflicting—dominant needs that are linked to
[the] press[es] to which the individual was
exposed on one or more particular occasions,
gratifying or traumatic, in early childhood. The
thema may stand for a primary infantile
experience or a subsequent reaction formation to
that experience. But, whatever its nature and

www.theipi.org 53
www.freepsychotherapybooks.org
 genesis, it repeats itself in many forms during
later life.

As soon as we realized the force of the unity-

thema, its importance in the interpretation of
each session began to dawn upon us. For if every
response is the objectification of an aspect of a
particular personality and the most fundamental
and characteristic determinant of a personality is
its unity-thema, then many responses cannot be
fully understood except in terms of their relation
to the unity-thema. (pp. 604-605)

Another method that has had a long career in

academic psychology, Kelly’s Role Construct

Repertory (1955), may well tap a similar
phenomenon, although I do not include it as a central
relationship pattern measure because it is not based

on the patient’s expressions in interviews. Rather, it

uses a concept formation method: The participants

sort people into different categories, for example, a

teacher they liked, a teacher they disliked, their

spouse, their father, and so on. They then sort these
people according to ways they are alike and different.

www.theipi.org 54
www.freepsychotherapybooks.org
 The method identifies the main constructs each

person uses for significant other people. Kelly

believed that his measure, as applied for each person,

was highly stable over time.

Yet another concept, the theoretically based

“nuclear script” (Tomkins, 1979), is similar to the

clinically based central relationship pattern concepts

(and as such is summarized in chapter 20, this

volume). The concept is related to Tomkins’ “nuclear

scene,” which reappears in memories with variations

over a person’s life, as exemplified by Carlson

(1981). A script is composed of the person’s rules for

understanding and dealing with a set of scenes. A

nuclear script involves the interpretation of present

situations in terms of their similarity to childhood

nuclear scenes. A scene is a basic element in the

theory; it is an organized unit that includes persons,

places, actions, and feelings.

www.theipi.org 55
www.freepsychotherapybooks.org
From the Developmental and the Object Relations
Researchers

The CCRT research has much in common with

the work of the object relations theorists, particularly

those who have relied on attachment theory, such as

Bowlby (1973). A series of studies by Sroufe and
Waters (1977) and Sroufe (1983) on children’s
relationship patterns has advanced the study of

attachment patterns. An adult form of the attachment

interview (George, Kaplan, & Main, 1985) has

facilitated this type of research, particularly with

Main and Goldwyn’s (1985) classification system.
Their interview consists of questions about
relationships with parents and requires recounting of

memories of childhood. The relationships revealed

were rated on such qualities as “rejection by parent”

and “loved versus unloved in childhood.”

A systematic clinical assessment method for

relationship patterns, Mayman’s Early Memories Test

www.theipi.org 56
www.freepsychotherapybooks.org
 (1968), has some basic similarities to the CCRT

approach in that the patient is asked to give actual

memories, not stories or fantasies as in the Thematic

Apperception Test. Mayman’s method differs from

the CCRT method in that only early memories are

asked for and the method of scoring is clinical but not
quantitative (Mayman & Faris, 1960).

Some of the basic concepts of the CCRT method

are like those of commonsense social psychology as

presented by Heider (1958). According to Heider, an

intrapsychic analysis helps the analyst understand
interpersonal relations. It is congruent with Heider’s

view that people have a need to form concepts of

their relationship environment. The narratives that are

used as the basis for the CCRT contain the person’s

view of the expected or actual responses of other

people, under the category of “response from other.”
People are seen as having wishes and as trying to
cause changes in their relationships with others.

www.theipi.org 57
www.freepsychotherapybooks.org
 The extended family of central relationship
pattern concepts also includes the “core organizing

principle” of Meichenbaum and Gilmore (1984) and

the “problematic reaction in the description of

events” of Rice and Greenberg (1984). Many other

such concepts are comprehensively reviewed by
Singer (1985) and Singer and Salovey (1991).

In general, it seems fair to conclude that—with

only a few exceptions, such as Kelly’s (1955) Role
Construct Repertory method, Tomkins’s (1979) script

theory, and attachment theory and research (George et

al., 1985)—academic psychology has neglected to

develop operational measures of the central
relationship pattern. Academic psychologists have
generally missed even seeing the expressions of the
concept because they lack a method for measuring

and an appreciation of the value of the concept,

although recently their level of appreciation may be

on the rise (Thome, 1989). Dynamic psychotherapists

www.theipi.org 58
www.freepsychotherapybooks.org
 are able to recognize and to use such a concept in the

form of the transference pattern, because they are

willing to rely on the results of their clinical method
of observation and they believe the concept is

essential for their therapeutic techniques.

From the Early Attempts to Develop a Transference

Measure

The needed transition from the traditional

unguided reliance on a clinical method of inferring

the transference concept to an operational measure of

it has been my recurrent preoccupation. Luborsky and

Schimek (1964), for example, considered making a

measure of “transference resolution” but thought it to

be virtually impossible because of the difficulty of

measuring the transference: “No one has yet gone far

into the measurement ... of this concept” (p. 96).

Now, looking back, I can make better sense of a

series of abortive attempts to achieve a reliable

central relationship pattern measure. These attempts

www.theipi.org 59
www.freepsychotherapybooks.org
 are worth a review here because they are not just a
matter for the record; they attest to the difficulty of

the task yet offer some leads worth following through

on. They are listed here in the order of the timing of

their attempt.

Systematic Clinical Formulation of the Transference

Paradigms

The first attempt goes back to my 1953-1959

participation in the Menninger Foundation

Psychotherapy Research Project (Kernberg et al.,

1972; Wallerstein, 1986). In that first experience, as

the head of the team assigned to the task of

evaluating the patients at the termination of

psychotherapy, I contributed to the construction and

first use of a form that required an assessment of the
transference pattern (Wallerstein, Robbins, Sargent,
& Luborsky, 1956). The form was filled out initially,

at termination, and at follow-up. The form called for

a statement of the essence of the “transference

www.theipi.org 60
www.freepsychotherapybooks.org
 paradigms” and within that of “the kinds of

interpersonal expectations of the patient that will be

recreated in the therapeutic situation. …What are the

earlier models on which these are based?” (p. 249).

“What are the … impulses that are being defended

against and simultaneously seeking gratification?” (p.
244). The research team was given the task of
assembling information about the patient’s

relationships and extracting from them a transference

formulation following its own estimate of the

conventional definition of transference. Today, it still

would be a useful comparison to see how these
relatively unguided conventional transference pattern
formulations compare with CCRT formulations.

Agreement on the Conventional Unguided

Transference Formulations

Before the advent of the CCRT measure, only a

few researchers had tried to judge the reliability of

measures of this concept on the basis of the

www.theipi.org 61
www.freepsychotherapybooks.org
 psychotherapy sessions themselves. One of the

earliest of such research attempts was made by Seitz

(1966) and his research group at the Chicago

Psychoanalytic Institute, who spent several years

trying to decide whether a concept of this sort could

be reliably inferred by clinicians. His conclusion was
that it could not. But his conclusion was based on the
use of the usual unguided judgment system: His

judges were free to respond on any inference level

and with any language, so that it was hard to evaluate

agreement or disagreement. Because he did not have

a method that would permit a decision either for or
against its reliability, his verdict that the concept
could not be judged reliably remained on shaky
ground.

Rating the Amount of Transference

Another episode in the off-and-on search for an

objective relationship pattern measure was played out
by the Analytic Research Group of the Institute of the

www.theipi.org 62
www.freepsychotherapybooks.org
 Pennsylvania Hospital. The group’s aim was to
develop a measure of transference. However, because

this aim presented us with a formidable task, we

settled on a limited version of the goal: to judge the

amount, rather than the content, of transference and

related variables. In our project we studied thirty 5-
minute segments from one psychoanalytic patient
(Luborsky, Graff, Pulver, & Curtis, 1973). The

agreement between judges was only modest (r = .26)

when the amount of overall transference in a segment

was rated. But the agreement was higher when the

assessment was for “transference likely” (r = .46, p <

.01), when the judgment was based on the amount of

transference expressed in relation to each person

referred to in the segment. A similarly aimed study

(Strupp, Chassan, & Ewing, 1966) also found only

slight agreement among independent judges in their
ratings of amount of transference in entire sessions.

www.theipi.org 63
www.freepsychotherapybooks.org
 Formulation of the Patient’s Main Communication in
a Session as a Focus for Interpretations

In an earlier attempt that allowed some progress

with the measurement of the content of the
transference (Auerbach & Luborsky, 1968), I

constructed a two-phase measure. In the first phase,

the judge formulated the patient’s main

communication across the session; in the second

phase, the judge estimated the degree to which the

therapist responded adequately in each interpretation

to this main communication. The study was done on
samples of sessions in which the therapist was judged
to have responded adequately versus inadequately to

the patient’s communications. Three judges achieved

moderate agreement (r = .68) in their global ratings

of the adequacy of the therapist’s main responses to

the patient. Because the concept of the patient’s main

communication requires of the clinical judge a

formulation that has a kinship to the main

relationship pattern, the experience with this measure

www.theipi.org 64
www.freepsychotherapybooks.org
 is included here as part of the history of attempts to

develop an objective measure of the central

relationship pattern.

Formulation and Rating of the Symptom-Context

Theme

One other early attempt appeared in retrospect to

have been formative in moving toward an objective

measure. The idea of a core conflictual relationship

theme was evident in my System-Context Theme

research (Luborsky, 1967; Luborsky & Auerbach,

1969). In that research a particular theme was found
for each patient within the psychotherapy session and
also just before recurrent symptoms appeared. This

theme is recurrent within each patient but different

from patient to patient. In the 1967 paper I noted the

similarity of this symptom-onset theme to the themes

evident in the patient’s dreams and to the other

themes in the same session. The parallel of the CCRT
based on therapy narratives with the CCRT based on

www.theipi.org 65
www.freepsychotherapybooks.org
 dreams has been carried further in the research
reported in chapter 12, this volume. The study of the

commonality between the Symptom-Context Theme

and the CCRT was carried forward in a paper in

which we examined the phobic symptoms of a patient

by the symptom-context approach and found

considerable parallel with the CCRT (Luborsky,

Mellon, & Crits-Christoph, 1985b). A similar

commonality with the CCRT was observed with

another symptom, momentary forgetting (Luborsky,

1988b, 1988c). A core content within each patient’s
context for forgetting was found to be similar to a

component of the patient’s CCRT for 3 of the 4

patients examined, and somewhat similar for the 4th
as well. These parallels were further established in
my book on the symptom-context method (Luborsky,
1996).

In summary, all of these early starts toward

fashioning a transference measure formed part of the

www.theipi.org 66
www.freepsychotherapybooks.org
 preparatory stages that led to the successful

construction of the CCRT. These starts toward

creating the new CCRT measure came from a goal-
directed tinkering process much like the one followed

by the discoverer of the Epstein-Barr (E-B) virus,

Anthony Epstein, who explained his research style
when he was interviewed about how his new
knowledge came about (Wolpert & Richard, 1988):

Interviewer: What do you think your skill is as a

scientist? You’re not a theoretician?

Epstein: No, not at all, I don’t understand any of

that. I think just sort of messing about is the
answer. You’ve got to keep messing about at the
bench. You see how to change this just a little
bit, you see how to change that a bit, and you
want to tinker with something and find a slightly
different and new way of doing it. (p. 165)

CONCLUSION

• From the experiences in constructing the CCRT, I

have come to a new stage in understanding
central relationship pattern measures. Now I can

www.theipi.org 67
www.freepsychotherapybooks.org
 more fully state the inclusion criteria for this
class of measures:
1. The measure must be based on extraction of a
pattern from a sample of self-other
narratives about relationship interactions.
Each one is part of either (a) a narrative
about such interactions or (b) a direct
observation of an enactment within
transcripts of audio or video recordings.
2. The pattern extracted should be of a central
relationship pattern, with central defined as
the most pervasive across the self-other
interactions.
3. The process of extraction of the pattern must
be based in part on clinical-quantitative
judgment, not only on responses to a
questionnaire filled out by the patient or on
unguided clinical judgment.
4. The measure must be at least partly capable
of reliable application.

These decisions about criteria are more

specifically described in chapter 8, this volume,
as they apply to the CCRT measure and in
chapter 20 for the many alternative measures of
the central relationship pattern that have been
appearing since the launching of the CCRT.
There may be other criteria that could apply, but

www.theipi.org 68
www.freepsychotherapybooks.org
 they are too abstract or too inferential, such as
that the measure also serve an integrating
function. There are other measures that come
close to fitting our criteria, such as Gottschalk
and Gleser’s (1969) content codes for free
association samples, but the other measures are
not included here because they describe specific
aspects of the content of the patient’s utterances
rather than of a broad central relationship pattern.
Additional criteria need to be examined
empirically, as I have begun to do in chapter 21,
this volume.

• Freud’s transference concept, as judged by

clinicians each in their own way, has been relied
on routinely since the turn of the century as a
guide to making interpretations by clinicians who
do dynamic psychotherapies as well as by many
other therapists. The concept has not, however,
been as well represented in personality theory
and research as it deserves to be because of the
earlier lack of reliable methods of measurement.
After nearly a century of clinical use of Freud’s
transference concept of a central relationship
pattern, the field now has a defined and measured
version of this pattern: the CCRT. It is a method

www.theipi.org 69
www.freepsychotherapybooks.org
 that should advance the field, for it can be used
with confidence about its measurement
capacities.

• The lineage of the CCRT measure has been traced

back to (a) the concept of the transference and
(b) the early precursors of the eventual CCRT,
such as Kelly’s role construct method (1955) and
the early memories method of Mayman and Faris
(1960). The CCRT captures a pattern that is
much like the transference template. As I
recounted at the beginning of this chapter, at the
outset of examining the concept I had merely
looked for a reliable measure of the central
relationship pattern. But the discoveries that
gradually accumulated suggested that what is
measured by the CCRT has much in common
with what is covered by the transference concept.
Many of the many users of the transference
concept have been realizing that the idea of the
CCRT as a measure of the central relationship
pattern is not a violation of the transference
concept but, in fact, fosters its understanding and
its use in practice.

www.theipi.org 70
www.freepsychotherapybooks.org
 2

A GUIDE TO THE CCRT METHOD

LESTER LUBORSKY

The CCRT is derived from narratives about

relationship episodes that patients typically tell or
sometimes even enact during their psychotherapy

sessions. Two major evaluation phases for these

narratives are required by the CCRT method: Phase A

is for locating and identifying the relationship

episodes, and Phase B is for extracting the CCRT
from the set of narratives. In addition to explaining
how to proceed through Phase A and Phase B, this

chapter also lists six optional steps for CCRT scoring,

each serving special scoring needs. The chapter ends

with explanations for crucial supplementary issues:

reliability of tailor-made categories, the judge’s need

www.theipi.org 71
www.freepsychotherapybooks.org
 for minimal information about the patient, the
necessary sample size of relationship episodes for

deriving the CCRT, the use in the CCRT of dreams

and fantasies, and the best procedures for training

CCRT judges. All of these scoring issues are essential

for the research use of the CCRT method that are
described in this chapter. Note that the clinical uses
of the CCRT method require special guidelines, and

these are explained in chapter 19.

PHASE A: LOCATION RELATIONSHIOP

EPISODES

This section helps to locate narratives about

relationship episodes in the transcripts of
psychotherapy sessions. It includes the definition of a
relationship episode, a classification of its varieties,

and an explanation of how to select passably

complete ones.

Definition of a Relationship Episode

www.theipi.org 72
www.freepsychotherapybooks.org
 A relationship episode (RE) is a part of a session
that is a relatively discrete episode of explicit

narration about relationships with others or with the

self. Although the entire session has some

characteristics of a narrative (Schafer, 1983), the

focus is limited here to the most explicit narratives
about relationships.

In each relationship episode, a main other person

with whom the patient is interacting is identified.
Usually this main other person is easy to identify.

Sometimes the patient talks about other people as

well, but this is not a problem as long as one person is

identifiable as the main other person. If another of the
other persons in an episode is talked about
sufficiently, a separate, additional relationship
episode may be scorable (see “Completeness of

Relationship Episodes,” following).

www.theipi.org 73
www.freepsychotherapybooks.org
 The demarcation of the length of the relationship
episode is facilitated by the fortunate fact that as a

narrative it tends to have a beginning, a middle, and

an end. The intent to begin a story is often signaled

by conventional stereotypical markers, such as the

beginning of a narrative about another person, a
relatively long pause, signs of a transition to a new
topic, or even a direct introductory statement. Often

the narrative is told to explain a self-observation or an

observation about another person with whom the

patient is interacting. Such introductory observations,

like the following two, simplify the task of the judge

in recognizing the beginning point:

Patient: (pause) Anyway, I remember another

incident…

or

Patient: I want to tell you something that

happened ...

www.theipi.org 74
www.freepsychotherapybooks.org
 A narrative is sometimes told as an example of a
characteristic of the patient or of the kind of event

that happens to the patient. Therefore, words such as

like or for example are used as part of the preface to

the narrative, which suggests that the patient is

providing an illustration or analogic representation of
types or relationships. These explanatory
introductions should be included with the narrative,

as in this preliminary to a relationship episode:

Patient: I’ve been bothered … with people

telling me what to do or trying to give me
directions, like, just, well, for example, I’ve been
registering for school all week and …

The length of each relationship episode in a

session transcript is marked off by a continuous line

along the left margin, extending from the beginning

(together with the prefatory comments) to the end of

the relationship episode. The number of the

relationship episode and the name of the main person
with whom the patient is interacting are noted at the

www.theipi.org 75
www.freepsychotherapybooks.org
 start of the line. The relationship episodes remain in
their actual serial order within the session’s transcript.

In that way the judge can read the entire session and
know the context in which the patient told the

relationship episode.

Transcripts are faster to evaluate than tape

recordings. Transcripts also have the advantage of
easy access to rereading that helps the judge to
remember the details of the relationship episodes. On
the other hand, recordings convey additional

information through the voice. Our net conclusion is

that transcripts are adequate and preferable for

purposes of extracting the CCRT, but a combination
of transcript and tape would be optimal.

Types of Other Persons in Relationship Episodes

The relationship episodes used for the CCRT

method are about relationships with people, including

the therapist, and relationships with the self. Most

www.theipi.org 76
www.freepsychotherapybooks.org
 often these relationships with people are with the

parents, spouse, friends, and bosses. Relationships

with inanimate objects in narratives are excluded only

because they are rare. The following paragraphs
identify the main types of persons in the narratives.

RE: Specific Other People

Relationship episodes that involve specific other

people are by far the most frequent type. The main
specific other person is usually readily identifiable,

but at times a patient may narrate an incident that

involves several people or a group of people (for
example, the patient’s family, classmates, or friends)

without indicating a specific person. In these

infrequent cases it is acceptable to designate the

“other person” as a group of people, such as “family”

or “friends.”

RE: Therapist

www.theipi.org 77
www.freepsychotherapybooks.org
 The therapist is one kind of other person who

needs to be considered as a separate category.

Although the entire session can be considered an
interaction with the therapist, some parts of a session
are especially identifiable as either of these two kinds

of relationship episodes manifestly about the

therapist:

RE: Therapist (narrative). The patient recounts an

episode about past or current interactions with the

therapist. Because this subtype of relationship

episode is a narrative, it is like the usual RE.

RE: Therapist (enactment). The patient engages

during the session in a delimited behavioral episode

of conflictual interaction with the therapist. These

relatively infrequent episodes form a special class of

relationship episodes because they are actual

enactments of interactions (as further described by

McLaughlin, 1987), rather than the more usual

www.theipi.org 78
www.freepsychotherapybooks.org
 narratives about interactions. In these enactments the
patient may initiate the episode by asking a

challenging question to which the therapist may

respond non-therapeutically, as happened in this

episode from Mr. T. Dodge’s1 session:

Patient: Would you please mail the bill to my

father?

Therapist: No, it is meant for you.

Patient: I would only have to mail it on again to

my father.

Therapist: No, it is not possible. Your father is

not the patient.

Patient: It is unfair to make me do that.

In this enactment both the patient and the

therapist play a part: The patient makes his wishes

known to the therapist; the therapist does not go

along with the patient’s wishes; and the patient feels
unfairly treated and the therapist does not deal with

the meaning of the exchange.

www.theipi.org 79
www.freepsychotherapybooks.org
 RE: Self

The few relationship episodes that qualify as “RE:

Self” are narratives about the patient’s interactions

with the patient’s own self. In contrast, most of the

patient’s references to self tend to be self-descriptions

and therefore do not qualify as developed relationship

episodes about the self.

Characteristic of “RE: Self” narratives is a

patient’s recollection of a specific interaction with the

self that included feeling or thoughts about the self

that involved confronting herself. For example, Ms.

Sheila Garrett provided a relationship episode about

herself that qualified:

Patient: Even the other day I heard a song. I

started crying. Then I thought to myself, my
God, Katie, you were so happy at first when you
broke up with Dave. You felt like it was a
rebirth. Why now are you crying or why is he
popping back into your dreams again? I wanted
to be over him.

www.theipi.org 80
www.freepsychotherapybooks.org
 In another example, Ms. Carol Kraft provided a
relationship episode about herself that also qualified:

Patient: I’m just very, very independent, very

much a loner. Like I don’t like people to hem me
in at all and back me into a corner, and once I
remember I backed myself into a corner and it
happened in this way…

Current Versus Past Relationship Episodes

Relationship episodes can cover any time span

from the very earliest memories to the present, as

exemplified systematically by Thome (1995a). The

judge should estimate the approximate age of the
patient at the time of the event in the narrative and, if
possible, should estimate the date of the event.
“Current” is defined as within the session or in the
last few days (REc); “recent” is in the last 3 years

(REr); all else is past (REp). (The time of events in

the episode may make a difference in the CCRT:

Perhaps a higher proportion of early events produces

a more pervasive CCRT across the narratives.)

www.theipi.org 81
www.freepsychotherapybooks.org
Completeness of Relationship Episodes

A patient in psychotherapy is usually told nothing

specific about what to say but to “say whatever comes

to mind” or “whatever you want to speak about.” As

a result, what is available are the narratives the
patient has chosen to tell, in the way the patient has

chosen to tell them. Thus, it is to be expected that

narratives about relationship episodes will vary

widely in completeness.

The RE judges have the task of deciding which

relationship episodes are complete enough and which

are too sketchy and incomplete to use for judgments
about the CCRT. It is useful for RE judges to rate

each relationship episode on its degree of

completeness on a scale of 1 to 5, from least to most

detailed. The usual cutoff for inclusion is a mean of

judges’ ratings of 2.5 or more. The following five

principles concerning completeness of narratives can
aid the evaluation:

www.theipi.org 82
www.freepsychotherapybooks.org
 1. An important aspect of the narrative’s
completeness is the specific detailing of the
patient’s interaction with the other person. A
relatively complete relationship episode would
be likely to contain an account of an interaction
in which the narrator includes the exchange of
the patient and therapist, the events that
occurred, the wishes, the responses from the
other person and of the self, and the outcome of
the event. Accounts of specific events are
probably more informative than general accounts
combining several incidents, although the latter
may be acceptable as relationship episodes.

2. The very incomplete relationship episodes—less

than 2.5 on the point scale—should be excluded
because their CCRT components are difficult to
identify. Such exclusion is justified as long as the
decision about whether to use a relationship
episode is based primarily on its degree of detail.
In addition, there is no indication that exclusion
of incomplete relationship episodes distorts the
eventual CCRT, and the supply of detailed
relationship episodes is usually adequate anyway.
The percentage of REs excluded tends to be

www.theipi.org 83
www.freepsychotherapybooks.org
 small—no more than 20-30% and usually much
less.

3. The demarcation of the beginning or the end of an

episode may be unclear. Because this is not a
serious deficiency, such episodes can be used.

4. At times a continuation of an episode occurs later

in the session. If the later addition is clearly part
of the same episode, it should be used and
labeled “continued from RE No. ___ on page
___.” The relationship episode with its
continuations is to be scored as one unit.

5. Some episodes can be understood as subepisodes

of larger narratives, but unless they are discrete,
they are not to be considered as separate
episodes.

Table 1, showing graded examples, will assist the

RE judge in the rating of completeness.

Two RE judges rated this example of a marginally

complete relationship episode from Mr. Ben Nevin on

the 5-point completeness scale, with 1 being the low

end; one judge rated it 1.5, the other 2.0:

www.theipi.org 84
www.freepsychotherapybooks.org
 TABLE 1
Completeness Ratings of Hypothetical Relationship Episodes

Rating Essence of the Relationship Episode

1.0 I met Joe and we talked. (No CCRT components)
1.5 I met Joe and we talked and he said little. (A fairly
vague response from other)
2.0 I met Joe, we talked and he said little. He's an old
friend from school who I like. (More vague
components, a hint of a wish and a response of self)
2.5 I met Joe, we talked, he said little. He’s an old friend
from school who I like. I was disappointed he said so
little about the event we went through together.
(Enough information to score a wish, response from
other, and response of self)
3.0 (Beyond the 2.5 level, the completeness ratings are
based on how much the patient elaborates on the
story and how detailed the information for each of the
components is.)
3.5
4.0 I met Joe, we talked, he said little. He’s an old friend
from school who I like. I was disappointed he said so
little about the event we went through together.
I was kind of trying to relive those days and get back
the feeling of that event we shared, but Joe seemed
distracted. I suggested we meet for lunch next week
and he agreed. (All three components are more
detailed and more explicit)
4.5 (Like the 4.0 description above, but with even more
detail)
5.0 (Like the 4.5 description above, but with even more
detail)

www.theipi.org 85
www.freepsychotherapybooks.org
 Patient: …and uh you know, the same thing with
the drawing course I’d like to get into. Fool
around with something like that. It’s uh, the
course, I guess—the teacher, I hear is pretty
good. Like he doesn’t really care about your uh,
technical ability to render, so much as he cares
about getting you to see, y’know, to be aware of,
uh, space. Architectural space and perspective,
and things like that. And I’m uh, kind of excited
over, over getting y’know, getting into
something like that.…

To include even such a skimpy relationship

episode as this one would do no harm, but its

inclusion is hardly worthwhile because the

description of the interaction with the teacher is so

incomplete. But even that brief description suggests a

possible basis for the patient’s attraction to the

teacher: The patient might anticipate that a

relationship with this encouraging teacher would

minimize the chances that his central relationship
problem—being incapable of asserting himself
against domination—would be activated.

www.theipi.org 86
www.freepsychotherapybooks.org
 The directions for locating relationship episodes
and judging their completeness were examined in a

sample of sessions for which two RE judges

independently selected the relationship episodes on

the basis of the ratings of completeness: Agreement

was good for the selection of the relationship
episodes as well as for their demarcation (detailed in
chapter 7).

Assignment of Scoring Judges

It is best to have two independent sets of judges,

one for locating the relationship episodes and the

other for scoring the CCRT, although this separation

is not essential. The relationship episodes should be

demarcated and rated for completeness before the

CCRT judges begin their job. Having two sets of

judges eases the time-consuming job of the CCRT

judge: Although the task of the RE judge takes only a

little longer than the time needed to read the
transcript, the task for the CCRT judge varies from 1

www.theipi.org 87
www.freepsychotherapybooks.org
 hour to 2½ hours per session depending on the

number of relationship episodes to be scored.

PHASE B: SCORING THE CCRT

To guard against confusion about the details of

scoring, I begin with a simplified diagram of the

essence of the method for extracting the CCRT from
the narratives about relationship episodes (see Figure
1). The diagram shows that a judge usually inspects

in succession 10 narratives about relationship

episodes and scores all the scorable thought units in

each for wishes, responses from others, and responses

of self. The frequency of each is totaled, and the

highest frequencies constitute the CCRT; that is the

essence of the CCRT method in brief.

Two repeated pairs of essential steps for

discerning the CCRT are summarized in this section:
Step 1 calls for the judge to identify types of
components to be scored in each relationship episode,

www.theipi.org 88
www.freepsychotherapybooks.org
 Figure 1.

Diagram of the essentials of the CCRT method

applied to 10 Res from psychotherapy sessions. W1 =
Wish, Standard Category #___ with highest
frequency, W2 = Standard Category #___ with second
highest frequency, etc. Frequency is expressed as a
proportion of the 10 narratives that contain the
component.

www.theipi.org 89
www.freepsychotherapybooks.org
 and Step 2 requires the judge to summarize them.
Then, in Step 1’ the judge re-identifies types of

components and in Step 2’ resummarizes them. More

explanation is given in later sections, and examples

are given in chapter 5, this volume.

Step 1: Identifying the Types of CCRT Components

Locating and Underlining Parts of the Relationship

Episode to Be Scored

After the length of each relationship episode has

been marked off by a vertical line on the left margin

of the transcript, the CCRT judge reads, rereads, and

scores the relationship episodes on the transcript of

the session. While reading, the judge underlines (or

marks off with slash marks) the parts of the text of

each relationship episode that will be the basis for
inferences about CCRT components. Each underlined
part is a single thought unit as defined by Benjamin

(1986b) and as illustrated in the case examples in

chapter 5, this volume. Usually each thought unit is

www.theipi.org 90
www.freepsychotherapybooks.org
 given a score for a single component; occasionally

two components are scored from a single thought

unit.

Identifying Types of Components

Three components are to be identified in each

relationship episode: (a) the wishes, needs, or
intentions: W; (b) the responses from others: RO; and
(c) the responses of the self: RS. The categories
scored for each type of component are to be written
on the left margin of the transcript alongside each

underlined thought unit, as illustrated in chapter 5.

This annotation must be done legibly and in sufficient

detail because it will be relied on by researchers to

make their tallies of scores. To make the linkage

clear, the judge should draw an arrow linking the

underlined thought unit in the text with the scores

from the wishes, responses from other, and responses
of self components derived from them, as shown in
Figure 2.

www.theipi.org 91
www.freepsychotherapybooks.org
 Figure 2.

Scores: Thought Unit in the Text:

NRO: Critical of me → "He was criticizing me and
NRS: Anger → I sort of felt annoyed by it"

Examples of linking of the score with the thought unit

in the text.

www.theipi.org 92
www.freepsychotherapybooks.org
 Preparing the Transcript for Scoring

A further variant simplifies the job of scoring:

The first CCRT judge (a) marks the thought units and
(b) adds a notation of the type of component to be

scored at the start of each thought unit; the

subsequent CCRT judges use the same scorable

thought units and components. This type of

preparation of the transcript eases the job of the

CCRT judges and, even more important, simplifies

the computation of reliability, because all CCRT

judges score the same thought units in terms of the

same components.

A more time-consuming method is to arrange for

a prescoring judge to prepare the transcript for the

CCRT judges by (a) identifying the initial thought

units and (b) adding a notation above the beginning

of each scorable thought unit of the type of
component to be scored.

www.theipi.org 93
www.freepsychotherapybooks.org
 Keeping Within the Range of Levels of Inference
From Literal to Moderately Abstract

The judge infers the wishes or responses within a

range between two levels of inference: the level of

virtually direct expression by the patient, in which the

judge stays close to the literal wording used by the
patient in the transcript, and the level of moderate
inference from what the patient says. Guidance for

staying within this range is given later in this chapter,

although some of the criteria remain inherently

unspecifiable. The responses from other and

responses of self are also divided into positive and

negative categories (as defined later in this chapter

and in chapter 4).

Step 2: Counting the Types of Components and

Formulating the CCRT

The judge counts the scores noted in the left

margin of the transcript for the occurrences of

particular types of components to see which have the

highest frequency across the relationship episodes.

www.theipi.org 94
www.freepsychotherapybooks.org
 Now, to make a preliminary CCRT formulation, the

judge reviews the scores, one type of component at a

time across relationship episodes, to find the theme or
themes that apply to the most relationship episodes.

The most frequent of each type of component

constitutes the preliminary CCRT formulation. This
step usually requires more labor for formulation of
the wish than for the generally more concrete

response from other and response of self.

Only occasionally does the most common theme

across the relationship episodes obligingly leap into

view. Usually, the process requires time and patience

for review and re-review of the wishes across

relationship episodes until a general formulation at

the most fitting level of inference is recognized. The

key to finding thematic consistencies across the

episodes is the rereading of the episodes and,
especially, the re-inspecting of the types of
components across episodes. Earlier episodes become

www.theipi.org 95
www.freepsychotherapybooks.org
 more understandable to the judge after the later ones

are studied, and newly recognized redundant themes

within and across episodes rearrange themselves from

time to time in saltatory accretions of eurekas.

It does not matter if some episodes remain opaque

or do not fit with the others, because the main aim of

the CCRT method is to locate the themes that repeat

themselves the most. The most recurrent components

of the episodes point to where the main conflicts lie;

the inclusion of the word conflictual in the CCRT

label rests on this assumption.

It is important to keep the general formulation

only as abstract as is necessary to fit the most

relationship episodes. This step, with its necessity for
keeping within a moderate level of inference,
especially requires the wet, gray software, the cortex
of a human judge—a tool not likely to be supplanted

www.theipi.org 96
www.freepsychotherapybooks.org
 by the dry, any-color hardware or any-style software
of the computer.

Step 1': Re-Identifying the Types of CCRT

Components to Make a Final Formulation

The Step 1 phase can be improved by a review to

be sure all of the components that make up the
general formulation have been considered and scored

in the relationship episodes. In light of the

preliminary formulation, the judge may now see a

particular component with new insight.

It is of research value to preserve the notations of

Step 1 scoring separately from the notations of Step

1' scoring. Therefore, the judge writes the additions

and alterations resulting from Step 1' in capital letters

in the left margin of the transcript. Again, judges

should draw arrows between the scored components

on the left margin and the text from which they are
derived.

www.theipi.org 97
www.freepsychotherapybooks.org
Step 2': Recounting and Reformulating the CCRT on
the Basis of Step 1'

Step 2' is a repeat of Step 2. The judge lists each

type of component on the CCRT summary sheet in

order of frequency across episodes. This is the CCRT:

the most frequent wish, followed by the most
frequent response from other, followed by the most
frequent response of self. The judge should group

similar types of components (for example, the

response “hostile” with “angry” or “afraid” with

“anxious”) as one type of component and add the

frequency of each to yield a single sum. If the same

type of component occurs more than once within a

relationship episode, the frequency of that type of

component is still limited to only 1 for that

relationship episode. As is further explained in

chapter 8, this volume, the rationale for this is that the

measure of the CCRT rests on pervasiveness of each

type of component across narratives, not within
narratives.

www.theipi.org 98
www.freepsychotherapybooks.org
Step 3: Choosing and Rating the Standard Categories
for the CCRT

What is described in Steps 1, 2 and 1', 2' is the

tailor-made system first presented by Luborsky
(1977b). A major asset of tailor-made categories is

their supreme ability to capture the individuality of

each patient. The tailor-made system works well but
has two major limitations for research: (a) the need

for equivalent categories across subjects for large

groups of subjects and (b) the need for the calculation

of standard reliability and validity coefficients. For

example, the standard categories are a way to cope

with a situation where one CCRT judge may have

said the patient “wishes to be close," and another

judge may have said the patient “wishes not to be cut

off from contact.” Are these two wish statements the

same, similar, or different?

Options Among Standard Category Lists

www.theipi.org 99
www.freepsychotherapybooks.org
 The procedure for standard CCRT categories
avoids such ambiguities by requiring that all judges
apply the same categories to the narratives so that
comparison between judges is simplified. The
recommended practice for scoring combines the

assets of tailor-made and the assets of standard

categories: First, one presents the tailor-made

inference, and then one translates it into standard

categories. I have relied on the three lists of standard

categories that are introduced here but are described

in greater detail in chapter 3, this volume.

Edition I: Standard categories. This set of

standard categories (Luborsky, 1986b) was based on

a list of the most frequently used categories within a
normative sample of 16 cases. This list was used in a
number of studies, including those by Luborsky,

Mellon, and Crits-Christoph (1985a); Luborsky,

Crits-Christoph, and Alexander (1990); and Crits-

Christoph, Cooper, and Luborsky (1988), and it

www.theipi.org 100
www.freepsychotherapybooks.org
 continues to be used by some researchers, although
less often. Some of these standard categories are

much like those in the Thematic Apperception Test

(TAT), with scoring categories from Murray (1938)

and Aron (1949). The similar categories were not

deliberately selected; some of the similarity derives
from the fact that the these categories are evident in
both the narratives used for the CCRT and the stories

used for the TAT.

Edition 2: Expanded standard categories. This

list of standard categories (Crits-Christoph &

Demorest, 1988) represents a large expansion of the

earlier categories reported by Luborsky (1986b). The
added categories drew on major category sets, such as
Murray’s (1938) “need” and “press” categories. The

set in the current Edition 2 has 35 wishes, 30

responses from other, and 30 responses of self.

Edition 2 has been used in several studies, including

those by Crits-Christoph and Demorest (1991);

www.theipi.org 101
www.freepsychotherapybooks.org
 Eckert, Luborsky, Barber, and Crits-Christoph

(1990); Luborsky, Luborsky, et al. (1995).

Edition 3: Reduction of Edition 2 to eight

clusters. Naturally, the large number of categories in

Edition 2 has much redundancy; this was

demonstrated by a cluster method that reduced the
number to eight clusters for each component (see
chapter 3).

Related Lists of Standard Categories

Several of these lists are included in the

alternative central relationship pattern measures

described in chapter 20. One of these frequently used

lists is known as the Structural Analysis of Social

Behavior (Benjamin, 1974); an adaptation of it was

developed for the CCRT by Crits-Christoph,
Demorest, Muenz, and Baranackie (1994). Future
improved standard category lists may emerge from an

www.theipi.org 102
www.freepsychotherapybooks.org
 expanded assessment of the personality domain and
from increased normative data.

Applications of Standard Categories

The standard categories can be applied to a

session in either of two ways: (a) by using tailor-

made categories followed by a translation into

standard categories or (b) by using standard

categories directly from the text without the tailor-

made categories. It is the first method that is most
highly recommended.

After the tailor-made categories are inferred from

the text, each one is translated into the standard
categories, as was typically done in most studies.
This system benefits from the virtues of a

combination of the tailor-made and standard

categories, that is, from categories that specifically fit

the case as well as from categories that are standard

across cases. The translation from the tailor-made to

www.theipi.org 103
www.freepsychotherapybooks.org
 the standard can be done by the original scorer or it
can be done by a different judge. One of the

following two systems differing in completeness

must be chosen:

System 1

The one best-fitting category (or a best-fitting

category followed by a next best) is chosen from the

approximately 30 in each of the three lists of types of

components. In making choices from Edition 2,
review the list of standard categories from time to

time to be sure that all of them have been considered.

This procedure of choosing just one (or two) of the

categories has the virtue of a rapid ranking system,

but it misses considerable information. Some of this

loss is inadvertent; it comes from the inherent

difficulty of attending to and then choosing one (or

two) of the categories in a long list.

System 2

www.theipi.org 104
www.freepsychotherapybooks.org
 All 30 or so categories of each component in

Edition 2 are rated for each of the scorable thought

units in a relationship episode: Because of the limits

of System 1, System 2, which involves rating all
categories on the following scale (see Figure 3), can

be used with the three convenient forms in Appendix

B, this chapter. On each form, at the top of each

column is a space to write the thought unit that is to

be rated on all categories.

For the sake of completeness, one also notes on

the listed ratings the best-fitting (by a circle around

the rating) and the next-best-fitting (by a square

around the rating) categories. This combination of the

rating system with the ranking system is desirable

because it provides considerable information. This

system becomes even more convenient when the

number of standard categories is reduced by cluster

analysis (as illustrated in chapter 3, this volume).

www.theipi.org 105
www.freepsychotherapybooks.org
 Figure 3.

A graphic scale for rating each category in the list of

components.

www.theipi.org 106
www.freepsychotherapybooks.org
 On the summary sheet for each type of

component, the judge should list in parentheses the

relationship episodes that contained that component.

For example, for Mr. Ben Nevin’s second most
frequent wish, the summary score sheet would read:

“Wish: to not go along with wishes of others (RE 3,

4, 7).”

Detailed Guides for Scoring the Components: Wishes,

Responses From Other, and Responses of Self

The CCRT judge usually starts scoring after

identifying the relationship episodes and the scorable
thought units within them. The CCRT judge (or a
prescoring judge) identifies by underlining or slash

marks every instance on the transcript in which a

patient reiterates a wish, response from other, or

response of self, either by repeating a phrase verbatim

or by attempting to describe it in different terms, but

each different type of response counts only once for
that relationship episode on the summary sheet. In

www.theipi.org 107
www.freepsychotherapybooks.org
 the following excerpt, from a relationship episode of
Mr. Uri Irion, the patient describes in several ways

his feelings of being relieved and unburdened. The

judge would count the type of response of self as

“relief of pressure” only once on the summary sheet.

Patient: I remember it felt like someone had

been on my shoulders, like I was carryin’ them
around in a swimming pool or somethin’ for 10
minutes … then they got off my shoulders. You
feel as though you just float. I felt as though
somethin’ had been taken off my brain.

The response from other and response of self

categories tend to be easier to score than the wish

category because they are more directly expressed.
They tend to be consequences of a wish, although not
every response is recognizably linked up with a wish.

Even though many of the linkages of the components

with each other are evident, it is sufficient for the

judge to note all wishes and responses without having

to connect them as a sequence.

www.theipi.org 108
www.freepsychotherapybooks.org
 Wishes

The usual two levels of inference need to be

distinguished for the wishes, in particular, but also for

responses from other and responses of self: (a) those

that are explicit or almost explicit: W, and (b) those
that are not explicit but moderately inferable: (W).

Explicit level of inference: W. An explicit wish is

often directly stated as a wish by words such as “I
wish,” “I want,” “I hunger for,” “I need,” and so on.

Moderate level inference: (W). Parentheses

around the component indicate a moderate level of

inference. These often are recognizable when the
judge reviews and makes inferences across different

relationship episodes from the same patient. Instances

of (W) are essential to the CCRT method because

they are more likely than instances of W to be evident
within many of the relationship episodes. In contrast,
the Ws tend to be more specific and limited to each

www.theipi.org 109
www.freepsychotherapybooks.org
 relationship episode. (This explication for using the
moderate level of inference emerged from a personal
communication with James Bond, 1986.) The judge
should list as many Ws and (W)s as are applicable; it
is useful for the judge to consider both an explicit

level score and a moderate level score for each of the

to-be-scored thought units. There is no reason for the

judge to worry about getting too many scores; in the

CCRT method only the most recurrent scores get into

the CCRT anyway. Also, as with any component, if

no wish is moderately clearly inferable in a

relationship episode, no wish score is given.

The wish formulation in the tailor-made system

should be no more abstract than necessary. An
example of an applicable but sometimes overly
abstract wish statement was “I wish to be

independent.” A review of the relationship episodes

showed that a more fitting word than independent
was assertive.

www.theipi.org 110
www.freepsychotherapybooks.org
 Table 2, derived from sessions of Mr. Ben Nevin,
illustrates the levels of inference in scoring W versus
(W) on the basis of the text examples in the right
column.

Responses From Other

A response from other should be scored only with

respect to the main other person (as listed in the left

margin at the beginning of the relationship episode).

Responses of Self

Responses of self should also include the patient’s

symptoms when these are evident in the relationship

episode. The consistent inclusion of the symptom as
response of self each time it appears in a narrative
permits an understanding of the symptom’s CCRT
context (as discussed by Luborsky, 1996).

Positive and Negative Responses

www.theipi.org 111
www.freepsychotherapybooks.org
 TABLE 2
Level of Inference in Scoring Wishes

Scoring of W and (W) Text Example of Wish

RE W: to assert self …in the end of the fantasies,
1 against being put down the position was reversed and I
by putting the other guy was stepping on him like that…
down
RE W: not to get sexually “I really don’t want to get
4 involved with the involved with her sexually…
woman
(W): to assert myself by “I really don’t want to get
not going along with involved …
the womana
(W): to get sexually “I really don’t want to get
involvedb involved …
W: to have people “I’m sort of hungry to have
around people around…

a Note that this formulation of the inference gains specificity by its

inclusion of the response from other.
b This one is even more inferential because it involves the denial of
the wish.

www.theipi.org 112
www.freepsychotherapybooks.org
 Each type of response from other and response of

self is also scored as mainly positive (P) or mainly

negative (N); usually a further subdivision is made at
the extremes as PP or NN, as explained in chapter 4.

A negative response is defined as one in which, to

the patient, interference with satisfaction of the

wishes has occurred or is expected to occur. A
positive response is one in which there is
noninterference or expectation of noninterference
with the satisfaction of wishes or a sense of mastery

in being able to deal with the wishes.

Patient: When I went to see my advisor

yesterday … and told her that, um, that I uh
decided not to go back full-time and she asked
me why … and I explained to her ... I really was
waiting for her to say to me … "Well, why? I
really thought you ought to go on” and “Why
don’t you get finished” and stuff like that that …
and she didn’t say it at all. She said, “Well, it’s
your decision and you’re probably better off
doing that.” I guess I was a little disappointed.

www.theipi.org 113
www.freepsychotherapybooks.org
 In this episode, Ms. Nan Iolanta wishes and

expects her advisor to try to talk her out of her

decision not to go back to school full-time, but the

advisor supports her capacity to make the decision,

leaving the patient feeling disappointed. Although the

judge may perceive the advisor’s response to be a
positive one, the patient perceives the response as one
that interferes with the satisfaction of her wish. Thus,

the response from other would be scored “negative.”

The response of self of disappointment would also be

scored “negative.”

Because all responses in a relationship episode

should be scored, even if they are not clearly

associated with a wish, at times it is hard to determine

whether a response should be labeled positive or

negative. In these instances the judge should choose a

score and add a question mark.

RO and RO-Expected

www.theipi.org 114
www.freepsychotherapybooks.org
 Only when the other person in a relationship

episode actually performs an action or responds in

some way does the judge score an RO. For example,
“He did hang up on me,” is scored “RO: Rejection.”
But if the patient has only an expectation or fantasy

that the other will respond in a particular way, the

incident is scored as in this example: “I didn’t call

him because I expected he’d hang up on me. RO-

expected: Rejection.”

OPTIONAL STEPS

Step 4: Distinguishing Between Expressed and Not

Expressed Responses of Self

This step and all that follow are optional to the

scoring procedure. The distinction of expressed (expr)

versus not expressed (not expr) may be a valuable

one, especially in studying the CCRT as a measure of
change. For example, “that irritated me” is not an
expressed response if the patient makes it clear that
he did not express his irritation to the other person.

www.theipi.org 115
www.freepsychotherapybooks.org
 Expression of irritation in later sessions might be

considered a positive change for this patient.

Step 5: Judging the Intensity of Theme Components

Intensity refers to the degree to which the speaker

expresses and experiences affect. A scale from 1 to 5

can be used for rating each type of component, in
which 1 is little or none and 5 is very much (included

when all standard categories are to be rated, the rating

forms in Appendix A are used when all standard

categories are rated for each thought unit). However,

further research is necessary to specify what is gained

by the addition of intensity ratings, for it may emerge,

that intensity of a type of component is largely
redundant with frequency of appearance, which is the
basis for the CCRT scoring. Table 3 lists three

intensity ratings for Mr. Ben Nevin.

Step 6: Recording the Sequence of the Appearance of

Each Component

www.theipi.org 116
www.freepsychotherapybooks.org
 TABLE 3
Intensity Ratings

Intensity Ratings Text Example

RS (4) “That irritated me.”
RO (3) “He tries to dominate the conversation.”
W (3) “I wish to assert myself.”

www.theipi.org 117
www.freepsychotherapybooks.org
 A record of the sequence of the wish, response
from other, and response of self within each

relationship episode can provide further insight into

each patient’s typical relationship interaction
sequences (as suggested by Ellen Berman, personal

communication, 1979). The sequences may even be

longer and more complex than just W → RO → RS.

A frequent longer sequence is W → RO → RS → W.

While reading the relationship episodes, the judge

can note the sequence of the appearance of each

component by numbering them consecutively on the
transcript and then indicating the sequence numbers

along with the CCRT scores. My experience with this

step indicates that interactional sequences are highly
stereotyped (see chapter 9, this volume).

A newly discovered facet of the sequence of

components was constructed by Mitchell (1995)

called the Coherence of the Relationship Theme. It

measures the degree to which the patient describes

www.theipi.org 118
www.freepsychotherapybooks.org
 himself and others as in interaction. The main

measure is the “link percentage,” which is the

percentage of CCRT components that are linked

together. A higher percentage of links was found to
be associated with the patients’ degree of relatedness

to others and based on the severity of the diagnosis.

For example, “I didn’t want to be close to her

(mother)” is a linked response of self. It is a link of

the wish with the response of self. “She was rejecting

me” is a linked response from other.

Step 7: Estimating the Patient’s Moment-to-Moment

Experience of the Components of the CCRT in the
Relationship With the Therapist

The types of components of the CCRT vary in the

degree to which they are experienced in the

relationship with the therapist; these variations can be

useful for the therapist to note. One system is called

the Patient’s Experience of the Relationship With the
Therapist (Gill & Hoffman, 1982b), described in

www.theipi.org 119
www.freepsychotherapybooks.org
 chapter 20, this volume. Unlike other central
relationship pattern measures, this measure aims to
identify when in the course of the session the
experience of and awareness of the experience of the
relationship with the therapist are clearest, so that the

therapist can consider interpreting them. According to

the psychoanalytic theory of psychotherapeutic
change (Luborsky, 1984), patients who improve in

the course of psychoanalytic therapy should develop

improved access to awareness of their experiences in

relationships, especially the one with the therapist.

This thesis can be examined by having judges score

the degree to which the patient experiences each type

of relationship component and when in the session

the experience is most evident. (For a sample, see

notations in the Appendix A, step 7. Although the

notations and definitions are mine, their basis is the
system by Gill & Hoffman, 1982b.) For example, for

Mr. Edward Howard (chapter 9, this volume) the

www.theipi.org 120
www.freepsychotherapybooks.org
 wish “to be close” is the most frequent wish. The

judge scores the category “Jet” which means the

judges infer that the patient experiences the wish

because the patient appears to be aware of the wish in

relation to the therapist, although he does not directly

express it. Scoring categories are then applied to the
response from other and response of self components.

Step 8: Randomizing Relationship Episodes Before

Scoring

The relationship episodes are usually scored in

the order and in the context in which they are given

because they are more meaningful that way. But for
some research purposes it may be of value to
randomize them to see what is conveyed by each

relationship episode by itself (see Crits-Christoph &

Demorest, 1991; Crits-Christoph, Demorest, Muenz,

& Baranackie, 1994). Eventually, a research-based

comparison of the naturalistic clinical versus

randomized conditions will be made.

www.theipi.org 121
www.freepsychotherapybooks.org
Step 9: Additional Quantitative Specification of the
CCRT

The specification includes two elements: (a) the

exact pervasiveness score across narratives and (b)
the range of scores. An example from chapter 5,

Table 3, conveys this point exactly. The usual CCRT

presentation includes only the most frequent types of
components. This proposed additional specification

would include the following from the CCRT:

Wish (five types of wishes scored); Category 13,

to be helped, 3/5; Category 18, to oppose others, 2/5.

Negative responses from other (types were

scored); Category 4, were rejecting, 3/5; Category 14,

are helpful, 3/5.

Negative response of self (six types scored);

Category 21, feel angry, 2/ 5; Category 22, feel

depressed, 2/5.

SUPPLEMENTARY ISSUES

www.theipi.org 122
www.freepsychotherapybooks.org
Estimating Reliability of Tailor-Made Categories

Agreement Judges

The job of agreement judges is to rate the degree

of agreement of pairs of CCRT judges with each

other. This is the most common reliability method for

tailor-made categories, but it has inherent problems of

subjectivity. Each agreement judge reviews the

scoring of the CCRT judges and merely classifies on
a clinical basis the degree to which CCRT judges’

scores on each of the categories are in clear

agreement, questionable agreement, questionable

disagreement, or clear disagreement. The agreement

judges generally report that this job can be done with

a sense of confidence.

Paired Comparisons

A more controlled reliability system for tailor-

made categories is based on the paired-comparisons

method: Each scored formulation by each judge is

www.theipi.org 123
www.freepsychotherapybooks.org
 paired with the scoring of the other judges on the
same case as well as on other cases; the pairs are then

judged for similarity. This procedure is described as

the method of mismatched cases (Levine &

Luborsky, 1981; Luborsky, Mellon, van Ravenswaay,

et al., 1985). The method gives information on the
level of agreement in the similarity of the tailor-made
pairs of categories for the same-case pairs versus the

similarity for the mismatched pairs. The method as

diagrammed in chapter 11, this volume, is most

suitable for dealing with tailor-made categories
because the tailor-made system is not designed for

conventional reliability methods. We have

demonstrated by this method that judges agree with

each other significantly more when the formulation

by each judge is compared with the formulation of

the other judges on the same case than when the pair

includes formulations from other (mismatched) cases.

www.theipi.org 124
www.freepsychotherapybooks.org
The Need of the CCRT Judge for Uniform Background
Information About the Patient

The judges should have some uniform minimal

information about the patient beyond what may be

inferable from the transcript of the session. At least

the age and sex of the patient should be given; the
adequate reliabilities reported in chapter 6, this
volume, were achieved by judges who had only these

two extra-session items of information. In addition,

the judge should be given an explanation of the

relationship to the patient of certain named people;
for example, “John” is the boyfriend, “Sarah” is the
sister, and so on. It may also be desirable for the

relationship episodes to be judged within the full

transcript so that the CCRT judge can understand the
context in which they are told.

It would be of interest to compare the therapist’s

scoring of the therapist’s own cases with the scoring

of these cases by other judges. The other judges

www.theipi.org 125
www.freepsychotherapybooks.org
 probably would be at a disadvantage because, as
Spence (1983b) would say, they are not “privileged”;

the therapist is “privileged” through knowing much

more about the patient.

The Best Sample Size of Relationship Episodes for

Deriving the CCRT

It is important to have an adequate sample of

relationship episodes to obtain a CCRT that is

representative of the treatment or of the treatment

phases. It is usually sufficient to sample at least two

early and two late sessions to locate 10, or close to

10, relationship episodes in the early sessions and 10,

or close to 10, relationship episodes in the late

sessions. In the inaugural CCRT study (Luborsky,

1977a), the relationship episodes were selected from
four 20-minute segments from each session (two
early and two late). Within each of these segments of
each session, 4 to 6 relationship episodes were
usually found, but occasionally there were as many as

www.theipi.org 126
www.freepsychotherapybooks.org
 10. The fact that there was considerable consistency

in content from the early CCRT to the late CCRT

implies that even fewer than 10 relationship episodes
can be enough to discern the CCRT. More research is
needed to determine how many relationship episodes

in how many sessions need to be sampled as an

adequate basis for the CCRT.

When a single session’s relationship episodes are

scored, as in some examples given in chapter 5, this

volume, the CCRT should be referred to as a “session

CCRT” rather than a “treatment CCRT.” A treatment

CCRT is based on assembling the relationship

episode scores from a sample of more than one

session. A session CCRT is a special version of the

CCRT and may differ some or very much from the

treatment CCRT.

The Use of Dreams and Fantasies

www.theipi.org 127
www.freepsychotherapybooks.org
 Whether to include dreams and fantasies as

relationship episodes, in addition to the usual

narratives, has not yet been fully decided. However,

if they are included, there will be relatively few of
them in each treatment and they will be lost among

the larger number of relationship episodes told as

accounts of patient’s actual interactions with other

people. The best resolution for now is to score

dreams and fantasies but not include them in the

CCRT, then note how they compare with the usual

relationship-episode-based CCRT. A comparative
study along these lines is reported in chapter 12, this
volume.

Training Procedures for Learning to Score the CCRT

For high agreement of judges to be achieved, it is

necessary that judges be well trained in the use of the

method. This training sequence is recommended:

1. Judges should have had some clinical training and

should be interested in learning the CCRT.

www.theipi.org 128
www.freepsychotherapybooks.org
 Despite the preference to use as judges
experienced clinicians who have a
psychodynamic orientation, a few graduate
students, including nonclinical students, have
also performed well as judges; clearly the task
does not require that the judge be trained in or
committed to a particular school of
psychotherapy.

2. It is recommended that each judge study the

CCRT instructions in this chapter and the scored
examples in chapter 5. A good first step in
training is to score the examples in chapter 5, this
volume, after first covering the score on the left
side of each page.

3. The experience of scoring several practice cases is

helpful for improving skill. The practice cases
are scored one by one, with feedback on the
trainee’s performance after each one by the
person in charge of training. The feedback is
based on the trainee’s agreement with other
judges who have scored the same cases as well
as the adherence of the judge to the procedures
outlined in this chapter. The sequence of cases
provided to the trainee is graded to begin with
easier cases. This kind of training followed by

www.theipi.org 129
www.freepsychotherapybooks.org
 feedback on three cases usually is required to
bring the judge up to an adequate level of
agreement with trained judges. (The present
series of practice cases includes Mr. G. Heyman,
Mr. Ben Nevin, Mr. O. Disims, and Ms. G.
Diane.) In fact, Bond, Hansell, and Shevrin
(1987) found evidence for increased agreement
among judges after greater experience in the
task. More about the special procedures for
clinical use of the CCRT is given in chapter 18.

CONCLUSION

• The scoring instructions presented so far in this

chapter provide a foundation for becoming
proficient in the CCRT method. But to become
even more proficient requires building one’s skill
by reviewing the basic steps for scoring outlined
in Appendix A, this chapter, by applying the lists
of standard categories given in Appendix B, also
this chapter, and by practicing with the examples
presented in chapter 5. After scoring three or
four cases, the average judge becomes
reasonably competent; however, even the
competent judge may have to consult this basic
chapter again and again as scoring problems
appear.
Note
www.theipi.org
[1] All names of patients are fictitious. 130
www.freepsychotherapybooks.org
 Appendix A:
Summary of Basic Steps for Scoring the
CCRT

PHASE A: LOCATING RELATIONSHIP

EPISODES

Locate the relationship episodes (REs) in the

session (if these have not already been located for
you by an independent RE judge).

Notations for the Relationship Episodes

RE = relationship episode

REC = current RE (within the session or the last few

days)

REC3 = within past 3 years

REP = past RE (Note patient’s age at the time of the

event and the date of the event. Approximate this
information when it is not available and note
when it is estimated age by parentheses.)

PHASE B: EXTRACTING THE CCRT

www.theipi.org 131
www.freepsychotherapybooks.org
Step 1

1. Underline (or use slash marks) each scorable

thought unit in the relationship episodes (if this
has not already been done for you by an
independent judge).

2. Tailor-made scoring. Write the scores you give for

each underlined scorable thought unit in the left
margin of the page. Connect by an arrow the
underlined unit with your inference in the
margin. Remember to consider scoring each of
the underlined thought units at two levels of
inference: (a) the less inferential level, that is,
close to the manifest level, and (b) the more
inferential, that is, at a moderate level of
inference; these scores are enclosed in
parentheses.

Step 2

Copy all scores from the transcript onto the score

sheets, count them, and formulate the CCRT.

Step 1'

Review the scores to see whether anything was

missed or needs revision. Additions and
revisions are to be written in capital letters.

www.theipi.org 132
www.freepsychotherapybooks.org
Step 2’

Recount the main types of components on the score

sheet summary and reformulate the CCRT. (Be
sure to note alongside each type of component
the relationship episode number in which each is
present.)

Step 3

Standard category scoring:

Top two choices: For each underlined thought
unit, choose the best-fitting standard
category and then the next-best-fitting one.
Write their numbers in parentheses after the
tailor-made category.
Rate all: An even more complete method to use
for the standard categories in Edition 2 and
3 is first to rate all of them for each
underlining and then choose the two best-
fitting ones. (Remember to include only the
ratings for each different type of component
in each relationship episode; Steps 1' and 2'
need not be done for scoring by standard
categories.)

Step 4

www.theipi.org 133
www.freepsychotherapybooks.org
 Differentiate between expressed and not expressed
response of self.

Step 5 (optional)

Consider the intensity: Give a rating of 1 to 5 (1 =

least, 5 = most) of the intensity of each type of
theme component (W, RO, RS).

Step 6 (optional)

Note the sequence of components: (1), (2), (3), etc. =

the position of each W, RO, and RS in each
relationship episode.

Step 7 (optional)

Consider awareness of one’s experience (see Gill &

Hoffman, 1982b) with a rating of 1 to 5 (1 = little
or none, 5 = very much).

et = directly expressed awareness by the patient of

an experience in the relationship with the
therapist in the session

Jet = inference by the judge of awareness of an

experience in the relationship with the therapist

eto = directly expressed awareness of the parallel

between the relationship with the therapist and

www.theipi.org 134
www.freepsychotherapybooks.org
 other relationships

Step 8 (optional)

Randomization of REs before scoring

Step 9 (optional)

Quantitative specification of the CCRT

NOTATIONS FOR THE CCRT COMPONENTS

Steps 1—2 and 1'—2'

Wishes

W = wish, need, or intention, as directly or almost

directly stated by the patient

(W) = wish, need, or intention as inferred by the

judge when moderate inference is used

Responses From Other

RO = actual response from other

RO-expected = response expected from other

N = negative (e.g., NRO = negative response from

other, from the patient’s perspective; use NN for
very negative)

www.theipi.org 135
www.freepsychotherapybooks.org
 P = positive (e.g., PRO = positive response from
other, from the patient’s perspective; use PP for
very positive)

Responses of the Self

RS = response of the self

N = negative (e.g., NRS = negative response of self,

NN = very negative)

P = positive (e.g., PRS = positive response of self,

PP = very positive)

expr = an expressed response of self (e.g., NRSexpr

= negative response of self that is expressed to
other person)

not expr = not expressed response of self (e.g., NRS

not expr = negative response of self that is not
expressed to other person)

www.theipi.org 136
www.freepsychotherapybooks.org
 APPENDIX B

WISHES

Rate intensity of all standard categories for each

thought unit

1 2 3 4 5
slight somewhat moderate much very much

Date:

Rated by:

(Write in each to-be-scored thought unit and its RE#

in the columns(continue on extra pages)

Edition 3 Edition 2
(clusters)
1. to assert self 21. to have self-control
& be 28. to be my own
independent person
34. to assert myself
23. to be independent

www.theipi.org 137
www.freepsychotherapybooks.org
2. to oppose, 18. to oppose others
hurt & control 16. to hurt others
others
19. to have control of
others
3. to be 15. to be hurt
controlled, hurt, 20. to be controlled by
& not others
responsible
29. to be not
responsible/obligated
13. to be helped
27. to be like others
4. to be distant 17. to avoid conflicts
& avoid 14. to not be hurt
conflicts
10. to be distant from
others
5. to be close & 4. to accept others
accepting 5. to respect others
9. to be open
6. to have trust
8. to be opened up to
11. to be close to
others
6. to be loved & 33. to be loved
understood 3. to be respected
1. to be understood
2. to be accepted
7. to be liked
7. to feel good 30. to have stability
& comfortable 31. to feel comfortable
32. to feel happy

www.theipi.org 138
www.freepsychotherapybooks.org
 24. to feel good about
self
8. to achieve & 22. to achieve
help others 25. to better myself
12. to help others
26. to be good
35. to compete with
someone for another’s
affection

RESPONSES FROM OTHER

Rate intensity of all standard categories for each

thought unit

1 2 3 4 5
slight somewhat moderate much very much

Date:

Rated by:

(Write in each to-be-scored thought unit and its RE#

in the columns(continue on extra pages)

Edition 3 Edition 2

www.theipi.org 139
www.freepsychotherapybooks.org
 (clusters)
1. strong 24. strong
23. independent
29. happy
2. controlling 26. strict
20. controlling
3. upset 16. hurt
22. dependent
28. anxious
27. angry
19. out of control
4. bad 8. not trustworthy
25. bad
5. rejecting & 7. don't trust me
opposing 6. don't respect me
2. are not
understanding
4. rejecting
10. dislike me
12. distant
14. unhelpful
17. oppose me
15. hurt me
6. helpful 13. are helpful
18. cooperative
7. likes me 30. loves me
5. respects me
9. likes me
21. gives me

www.theipi.org 140
www.freepsychotherapybooks.org
 independence
8. understanding 11. open
1. understanding
3. accepting

RESPONSES OF SELF
Rate intensity of all standard categories for each
thought unit

1 2 3 4 5
slight somewhat moderate much very much

Date:

Rated by:

(Write in each to-be-scored thought unit and its RE#

in the columns(continue on extra pages)

Edition 3 (clusters) Edition 2

1. helpful 7. am open
1. understand
9. am helpful
2. unreceptive 2. don't

www.theipi.org 141
www.freepsychotherapybooks.org
 understand
8. am not open
6. dislike others
3. respected & 28. feel
accepted comfortable
29. feel happy
30. feel loved
4. feel
respected
3. feel accepted
5. like others
4. oppose & hurt 11. oppose
others others
10. hurt others
5. self-controlled & 14. self-
self-confident controlled
15. independent
18. self-
confident
12. controlling
6. helpless 13. out of
control
17. helpless
19. uncertain
16. dependent
7. disappointed & 21. angry
depressed 20. disappointed
22. depressed
23. unloved
24. jealous
8. anxious & ashamed 27. anxious

www.theipi.org 142
www.freepsychotherapybooks.org
 26. ashamed
25. guilty
31. somatic
symptoms

www.theipi.org 143
www.freepsychotherapybooks.org
 3

A GUIDE TO THE CCRT STANDARD

CATEGORIES AND THEIR
CLASSIFICATION

JACQUES P. BARBER, PAUL CRITS-

CHRISTOPH, AND LESTER LUBORSKY

The CCRT scoring system uses well-constructed

sets of standard categories. These categories describe

the types of components of the CCRT for use in both

psychotherapeutic and research settings. The tailor-

made system tends to be more appropriate for clinical

work or for case studies because it allows the

therapist to derive a psychodynamic formulation that

closely fits each patient. The standard category

systems are usually more appropriate for research
because they allow the researcher to compare reliably
and easily different judges’ selections of categories.

www.theipi.org 144
www.freepsychotherapybooks.org
 Several procedures are recommended in chapter 2,
this volume, to guide users of these lists of standard

categories, but the most versatile procedure is for the

judge to record the tailor-made categories and then to

translate each of them into Edition 2 of the standard

categories, which is described in this chapter. Such a
combination of methods gives the richest information
for both research and clinical purposes.

To facilitate construction of standard categories,

we assembled lists of standard categories: Editions 1

and 2 (see Exhibit 1 and Table 1, this chapter, and the

three rating forms that make up Appendix B of

chapter 2). In this chapter we present an account of
the development of the Edition 2 classification of
CCRT standard categories and the subsequent cluster
analysis of these categories to derive Edition 3.

DEVELOPMENT OF THE STANDARD

CATEGORIES

www.theipi.org 145
www.freepsychotherapybooks.org
 EXHIBIT 1
A Standard List of Scoring Categories: Edition 1 Based on the
CCRT in a Normative Group (n = 16)

Wishes, Needs, Intentions (“I wish, need, or

intend in relation to the other person ...”)

A. To assert myself
A1. To assert my independence and autonomy
A2. To dominate; to impose my will or control on
others
A3. To overcome other’s domination; to be free of
obligations imposed by others; to not be put down
A4. To win in competition with another; to be better
than the other person
A5. To win the affection or attention of another over
someone else (triangle or oedipal situation)
B. To submit
B1. To submit; to give in; to be passive
C. To make contact with others; to be close and intimate with
others
C1. To make contact with others; to be close; to be
friends
C2. To receive affection; to not be deprived of
continued affection
C3. To be receptive (to open up) to others
C4. To please the other person; to avoid hurting the
other person
D. To get sexual gratification
D1. To get sexual gratification
E. To receive acceptance
E1. To receive acceptance; to be respected,

www.theipi.org 146
www.freepsychotherapybooks.org
 recognized, approved, vindicated, reassured; to
maintain one’s self-esteem (to avoid disapproval, and
so on)
E2. To be fairly treated (to not be unfairly treated)
F. To get help and care from others
F1. To get help, care, protection, and guidance from
others
G. To achieve and be competent
G1. To achieve, be competent, be successful
H. To hurt the other person
H1. To hurt the other person; to get back at the other
person; to express anger, hostility, or resentment to
the other person
I. To exert control over myself
I1. To exert control over myself

Responses From Others (“The other person

becomes ...”)

Negative Positive
1. Dominating, controlling, interfering, 1. Supportive,
intimidating, intruding reassuring
2. Unfair, exploiting, taking advantage 2. Treats fairly,
respectfully
3. Resentful, angry, irritated, hostile,
violently aggressive
4. Critical, disapproving 4. Accepting,
approving
5. Insensitive, inconsiderate (does not 5. Concerned,
consider my feelings) interested
6. Unhelpful, uncooperative, 6. Helpful,
noncompliant (does not gratify my cooperative,
wishes) compliant

www.theipi.org 147
www.freepsychotherapybooks.org
7. Distant, withdrawn 7. Close,
expressive
8. Unaffectionate 8. Affectionate
9. Does not understand 9. Understanding
10. Dependent
11. Hurt
12. Honest, open

Responses of Self (“I become ...”)

Negative Positive
1. Passive, submissive, dominated, 1. Assertive,
compliant, deferential, controlled (“I express self
give in to the power of other”) assertively,
gain control
2. Dependent
3. Helpless, less confident, ineffectual (“I 3. Gain self-
do not know how to do things") esteem, feel
affirmed
4. Hopeless (“I give up; I feel nothing
can be done”)
5. Obligated
6. Rejected, unaccepted, disappointed
7. Uninvolved with people, lonely, 7. Feel close to
detached, distant, quiet, others
nonresponsive, untalkative (“I pull
away from people”)
8a. Angry, resentful, hating 8a. Not angry
8b. Violently angry
9. Frustrated
10. Depressed 10. Happy
11. Anxious, tense, upset
12. Frightened, afraid

www.theipi.org 148
www.freepsychotherapybooks.org
13. Guilty, self-blaming
14. Embarrassed
15. Jealous
16. Confused, indecisive, ambivalent

Note. Only the numbered categories, not the headings, are scored.
Parentheses after a category contain a statement summarizing the
essence of the category.

www.theipi.org 149
www.freepsychotherapybooks.org
 TABLE 1
Standard CCRT Categories: Edition 2

Standard Category
Category
Components
WISHES, NEEDS, INTENTIONS
1. TO BE UNDERSTOOD To be comprehended; to be
empathized with; to be seen
accurately
2. TO BE ACCEPTED To be approved of; to not be
judged; to be affirmed
3. TO BE RESPECTED To be valued; to be treated
fairly; to be important to
others
4. TO ACCEPT OTHERS To be receptive to others
5. TO RESPECT OTHERS To value others
6. TO HAVE TRUST Others to be honest; others
to be genuine
7. TO BE LIKED Others to be interested in me
8. TO BE OPENED UP TO To be responded to; to be
talked to
9. TO BE OPEN To express myself; to
communicate
10. TO BE DISTANT FROM To not express myself/my
OTHERS feelings; to be left alone
11. TO BE CLOSE TO To be included; not to be left
OTHERS alone; to be friends
12. TO HELP OTHERS To nurture others; to give to
others
13. TO BE HELPED To be nurtured; to be given
support; to be given
something valuable; to be
protected

www.theipi.org 150
www.freepsychotherapybooks.org
14. TO NOT BE HURT To avoid pain and
aggravation; to avoid
rejection; to protect/defend
myself
15. TO BE HURT To be punished; to be
treated badly; to be injured
16. TO HURT OTHERS To get revenge; to reject
others; to express anger at
others
17. TO AVOID CONFLICT To compromise; not to anger
others; to get along; to be
flexible
18. TO OPPOSE OTHERS To resist domination; to
compete against others
19. TO HAVE CONTROL To dominate; to have power;
OVER OTHERS to have things my own way
20. TO BE CONTROLLED To be submissive; to be
BY OTHERS dependent; to be passive; to
be given direction
21. TO HAVE SELF- To be consistent; to be
CONTROL rational
22. TO ACHIEVE To be competent; to achieve;
to win
23. TO BE INDEPENDENT To be self-sufficient; to be
self-reliant; to be
autonomous
24. TO FEEL GOOD ABOUT To be self-confident; to
MYSELF accept myself; to have a
sense of well-being
25. TO BETTER MYSELF To improve; to get well
26. TO BE GOOD To do the right thing; to be
perfect; to be correct
27. TO BE LIKE OTHER To identify with other; to be
similar to other; to model
after other
28. TO BE MY OWN Not to conform; to be unique

www.theipi.org 151
www.freepsychotherapybooks.org
 PERSON
29. TO NOT BE To be free; to not be
RESPONSIBLE OR constrained
OBLIGATED
30. TO BE STABLE To be secure; to have
structure
31. TO FEEL To relax; to not feel bad
COMFORTABLE
32. TO FEEL HAPPY To have fun; to enjoy; to feel
good
33. TO BE LOVED To be romantically involved
34. TO ASSERT MYSELF To compel recognition of
one’s rights
35. TO COMPETE WITH
SOMEONE FOR
ANOTHER PERSON’S
AFFECTION

RESPONSES FROM OTHERS

1. ARE Are empathic; are sympathetic; see
UNDERSTANDING me accurately
2. ARE NOT Are not empathic; are
UNDERSTANDING unsympathetic; are inconsiderate
3. ARE ACCEPTING Are not rejecting; approve of me;
include me
4. ARE REJECTING Are disapproving; are critical
5. RESPECT ME Treat me fairly; value me; admire
me
6. DON’T RESPECT Don’t treat me fairly; don’t value
ME me; don’t admire me
7. DON’T TRUST ME Don’t believe me; are suspicious of
me
8. ARE NOT Betray me; are deceitful; are
TRUSTWORTHY dishonest

www.theipi.org 152
www.freepsychotherapybooks.org
9. LIKE ME Are interested in me
10. DISLIKE ME Are not interested in me
11. ARE OPEN Are expressive; are disclosing; are
available
12. ARE DISTANT Are unresponsive; are unavailable
13. ARE HELPFUL Are supportive; give to me; explain
14. ARE UNHELPFUL Are not comforting; are not
reassuring; are not supportive
15. HURT ME Are violent; treat me badly; are
punishing
16. ARE HURT Are pained; are injured; are
wounded
17. OPPOSE ME Are competitive; deny/block my
wishes; go against me
18. ARE Are agreeable
COOPERATIVE
19. ARE OUT OF Are unreliable; are not dependable;
CONTROL are irresponsible
20. ARE Are dominating; are intimidating;
CONTROLLING are aggressive; take charge
21. GIVE ME Give me autonomy; encourage self-
INDEPENDENCE direction
22. ARE DEPENDENT Are influenced by me; are
submissive
23. ARE Are self-directed; are not
INDEPENDENT conforming; are autonomous
24. ARE STRONG Are superior; are responsible; are
important
25. ARE BAD Are wrong; are guilty; are at fault
26. ARE STRICT Are rigid; are stem; are severe
27. ARE ANGRY Are irritable; are resentful; are
frustrated
28. ARE ANXIOUS Are scared; are worried; are
nervous

www.theipi.org 153
www.freepsychotherapybooks.org
29. ARE HAPPY Are fun; are glad; enjoy
30. LOVES ME Is romantically interested in me

RESPONSES OF SELF
1. UNDERSTAND Comprehend; realize; see accurately
2. DON’T Am confused; am surprised; have
UNDERSTAND poor self-understanding
3. FEEL Feel approved of
ACCEPTED
4. FEEL Feel valued; feel admired
RESPECTED
5. LIKE OTHERS Am friendly
6. DISLIKE Hate others
OTHERS
7. AM OPEN Express myself
8. AM NOT OPEN Am inhibited; am not expressive; am
distant
9. AM HELPFUL Am supportive; try to please others;
am giving
10. HURT OTHERS Am violent; act hostile
11. OPPOSE Am competitive; refuse/deny others;
OTHERS conflict with others
12. AM Am dominating; am influential;
CONTROLLING manipulate others; am assertive; am
aggressive
13. AM OUT OF Am irresponsible; am impulsive; am
CONTROL unreliable
14. AM SELF- Am responsible
CONTROLLED
15. AM Make my own decisions; am self-
INDEPENDENT directed; am autonomous
16. AM Am submissive; am passive
DEPENDENT

www.theipi.org 154
www.freepsychotherapybooks.org
17. AM HELPLESS Am incompetent; am inadequate
18. FEEL SELF- Am or feel successful; feel proud; feel
CONFIDENT self-assured
19. AM UNCERTAIN Feel torn; am ambivalent; feel
conflicted
20. FEEL Am not satisfied; feel displeased; feel
DISAPPOINTED unfulfilled
21. FEEL ANGRY Feel resentful; feel irritated; feel
frustrated
22. FEEL Feel hopeless; feel sad; feel bad
DEPRESSED
23. FEEL Feel alone; feel rejected
UNLOVED
24. FEEL JEALOUS Feel envious
25. FEEL GUILTY Blame myself; feel wrong; feel at fault
26. FEEL Am embarrassed; feel abashed
ASHAMED
27. FEEL ANXIOUS Feel scared; feel worried; feel nervous
28. FEEL Feel safe; am or feel satisfied; feel
COMFORTABLE secure
29. FEEL HAPPY Feel excited; feel good; feel joy; feel
elated
30. FEEL LOVED
31. SOMATIC Headache; rash; pain
SYMPTOMS

Note. The category label is in capital letters; the lower-case labels are
variants of the capitalized label.

www.theipi.org 155
www.freepsychotherapybooks.org
 Our effort to create standard categories originated
in the typical difficulties we experienced in using

tailor-made assessments and trying to compare them.

These standard categories provide the judges with a

common language that has enabled us to compare

patients’ CCRTs quantitatively as well as to compare
judges’ performance in deriving the CCRT. To
measure the agreement between two raters on a

specific CCRT formulation, we had to make sure that

they used a common language, because one cannot

assess traditional agreement (for example, kappa or

intraclass correlation) from idiographic narrative

descriptions. In the past, we resolved this problem by

presenting matched and mismatched pairs of CCRTs

to judges and asking them to rate the similarity of

these CCRTs (e.g., Levine & Luborsky, 1981). Now

such a uniform language is achieved through the
creation of standard categories.

www.theipi.org 156
www.freepsychotherapybooks.org
 Edition 1 of the standard categories (Exhibit 1)
was abstracted from the judges’ CCRT ratings of the

16 participants available at that time. These patients

were mainly diagnosed with depressive or anxiety

disorders and were considered representative of

patients who were referred for psychotherapy. The
categories chosen were the most frequent in this
sample of narratives. Edition 1 was used by

Luborsky, Mellon, et al. (1985a); Luborsky, Crits-

Christoph, and Alexander (1990); and Crits-

Christoph, Cooper, and Luborsky (1988).

Six main organizational principles were followed

in constructing Edition 1: (a) The categories are the
ones that most frequently fit the 16 cases scored by
the CCRT method; (b) the categories are readily
discriminable from each other, that is,

nonoverlapping; (c) the subsidiary adjectives used

within each category are fairly synonymous with each

other and with the category label; (d) the categories

www.theipi.org 157
www.freepsychotherapybooks.org
 are organized logically, psychologically, and
conveniently from the point of view of ease of

application by the judge; (e) the order of the

categories is similar within each of the three types of

components (for wishes, responses from other, and

responses of self); this organizational principle is
aimed at easing the judges’ task in finding a particular
category; and (f) the words selected for each category

are the same in each of the three components lists,

whenever reasonable, for example, W: to reject

other’s domination; RO: dominates; RS: feels

dominated.

Later, Edition 2 was created to provide judges

with a more representative collection of categories
than was available in Edition 1. To prepare this new
edition, Crits-Christoph and Demorest (1988)

reviewed the relevant literature, such as Murray’s

(1938) list of needs. From these categories, they

created a list of 34 wishes (to which a 35th has been

www.theipi.org 158
www.freepsychotherapybooks.org
 added), 30 responses from others, and 30 responses
of self. They then included up to five, but usually

three, exemplifying subcategories for each one of

these standard categories (see Table 1).

Through using standard categories, we could

begin comparing patients in a more reliable manner,
and we could assess changes in the CCRT during
treatment without needing another special judge to
assess the degree of similarity between the
pretreatment and posttreatment CCRTs. It was clear

to us, however, that the list of standard categories had

some overlapping categories. For example, one judge

would categorize a patient as having the wish “to be
accepted,” whereas another would specify the wish as
“to be respected,” and still another would identify the
wish as “to be loved.” To deal with such overlap, we

went on to examine the underlying structure of our

list of standard categories by a cluster analysis of

similarity judgments of the standard categories.

www.theipi.org 159
www.freepsychotherapybooks.org
PROCEDURE FOR A CLUSTER-BASED
CLASSIFICATION

Another goal of the present study was to reduce

the list of the standard categories of Edition 2 to a
more practical size. Thirty-five wishes, 30 responses

from others, and 30 responses of self were hard for

our judges to use conveniently and reliably. To create

a fairly comprehensive but still manageable list of

categories to guide the judges, each of the standard
categories within each CCRT component was paired
with all other standard categories. All of these pairs

to be compared were written in questionnaire form.

Somatic symptoms (such as response of self No. 31)

were not included in this questionnaire. A 35th wish

was added to the list to represent an “oedipal wish,”

defined as the wish to compete with somebody to get
another person’s affection. There were 595 pairs of
comparisons for the wishes and 435 each for the

responses from other and responses of self.

www.theipi.org 160
www.freepsychotherapybooks.org
 Nine judges were asked to rate the degree of

similarity of each pair of standard categories on a

scale from 1 (not similar at all) to 7 (extremely
similar). To avoid tedium, the judges could spread the
work over more than one sitting.

RESULTS

For each CCRT component, an intraclass

correlation was computed to assess the judges’

reliability. Pooled-judge intraclass coefficients for the

nine judges were .80 for the wishes, .90 for the

responses from other, and .86 for the responses of

self. Because the reliability of these judgments was
acceptable, we averaged the judges’ scores.

For each CCRT component, the average similarity

scores were submitted to the SPSS/PC + (1986)

cluster analysis procedure using the Ward method.
This method is designed to optimize the minimum
variance within each cluster.

www.theipi.org 161
www.freepsychotherapybooks.org
 Inspection of the tree of interrelated groupings
(“dendograms”) obtained from the cluster analyses

led us to retain eight clusters for the wishes, eight for

the responses from other, and eight for the responses

of self. The clusters for each CCRT component—

clustered standard categories—are listed in Table 2.

DISCUSSION

This study has achieved an encompassing level of

description along with a convenient brevity by the

cluster analyses for each of the three components of

the standard categories. We recommend these clusters
primarily for research use, especially in studies
involving groups, whereas we recommend the
standard categories both for clinical and research use.

Wishes

The clusters of wishes consist of the wish to be

independent and have individuality (Cluster 1); the
wish to hurt and control others (Cluster 2); its

www.theipi.org 162
www.freepsychotherapybooks.org
 TABLE 2
CCRT Clustered Standard Categories; Edition 3

Clusters Standard Category Components

WISHES, NEEDS, INTENTIONS
1. TO ASSERT SELF To have self-control; to be my own
AND BE person; to assert myself; to be
INDEPENDENT independent
2. TO OPPOSE, To oppose others; hurt others; have
HURT, AND control over others
CONTROL
OTHERS
3. TO BE To be hurt, controlled by others; not
CONTROLLED, to be responsible or obligated; to be
HURT,AND NOT helped; to be like others
RESPONSIBLE
4. TO BE DISTANT To avoid conflicts; to not be hurt; to
AND AVOID be distant from others
CONFLICTS
5. TO BE CLOSE To accept others, respect others; to
AND ACCEPTING be open; to have trust; to be opened
up to; to be close to others
6. TO BE LOVED To be loved, respected, understood,
AND accepted, liked
UNDERSTOOD
7. TO FEEL GOOD To have stability; to feel comfortable;
AND to feel happy; to feel good about
COMFORTABLE myself
8. TO ACHIEVE AND To achieve; to better myself; to help
HELP OTHERS others; to be good

RESPONSES FROM OTHER

1. STRONG Strong, independent, happy
2. CONTROLLING Strict, controlling

www.theipi.org 163
www.freepsychotherapybooks.org
3. UPSET Hurt, dependent, anxious, angry, out
of control
4. BAD Not trustworthy, bad
5. REJECTING AND Don’t trust me; don’t respect me; are
OPPOSING not understanding; rejecting; dislike
me; are distant; unhelpful; oppose
me; hurt me
6. HELPFUL Are helpful, cooperative
7. LIKES ME Loves me; respects me; likes me;
gives me independence
8. UNDERSTANDING Open, understanding, accepting

RESPONSES OF SELF
1. HELPFUL Am open; understand; am helpful
2. UNRECEPTIVE Don’t understand; am not open;
dislike others
3. RESPECTED AND Feel comfortable, happy, loved,
ACCEPTED respected, accepted; like others
4. OPPOSE AND Oppose others; hurt others
HURT OTHERS
5. SELF- Self-controlled, independent; self-
CONTROLLED confident, controlling
AND SELF-
CONFIDENT
6. HELPLESS Out of control, helpless, uncertain,
dependent
7. DISAPPOINTED Angry, disappointed, depressed,
AND DEPRESSED unloved, jealous
8. ANXIOUS AND Anxious, ashamed, guilty
ASHAMED

www.theipi.org 164
www.freepsychotherapybooks.org
 counterpart, the wish to be controlled and hurt

(Cluster 3); the wish to withdraw (Cluster 4); the

wish to be close (Cluster 5); the wish to be loved and

understood (Cluster 6); the wish to feel good (Cluster

7); and the wish to achieve (Cluster 8).

The clusters are not entirely nonoverlapping,

however, as we wished they would be. Clusters 1 and
2 are close in their meaning because both represent
different aspects of expressing and maintaining

independence. Clusters 5 and 6 also have a similar

meaning to each other; both entail a wish for
closeness. Thus, we are still left with two pairs of

clusters that might lead to ambiguities with regard to

the classification of a person’s specific wishes.

Our eight clusters of wishes seem to represent a

fairly complete list of human motivations as they
appear in accounts of interpersonal encounters. Yet

certain wishes and needs are not in these clusters,

www.theipi.org 165
www.freepsychotherapybooks.org
 probably because for the majority in our society
needs for survival, such as for food and shelter, are

not a frequent source of interpersonal problems or of

problems leading patients into psychotherapy. Higher

level motivations such as self-fulfillment (e.g.,

Maslow, 1970) to not make up a moderate level of
inference category and therefore do not appear in our
list. The wish for mastery is probably included under

Cluster 8, the wish to achieve; our judges did not

usually use the word mastery, although they used to

achieve to explain some narratives. It is likely that the

wish for mastery represents a higher level of

inference and abstraction than is generally used by

judges. An even higher level of abstraction is not

evident in the list and not recommended; reliability is

low for such categories, as exemplified by self-

realization, Eros, libido, masochism, and so on.

In their study of the scaling of interpersonal

problems, Horowitz and Vitkus (1986) asked

www.theipi.org 166
www.freepsychotherapybooks.org
 participants to sort a long list of interpersonal

problems into categories of problems that seemed to

be semantically similar. Following a hierarchical
clustering procedure, they found clusters describing

issues about “intimacy, assertiveness, aggression,

compliance, dependency, independence and
socializing” (p. 448). Although Horowitz and Vitkus
used a list of interpersonal problems as their database,

the results of their cluster analysis are similar to the

ones we obtained by using a database of wishes and

desires. A moderate-to-high level of similarity is also

found between the subscales of the Inventory of
Interpersonal Problems (Horowitz, Rosenberg, Baer,
Ureno, & Villasensor, 1988) and our clusters. Clearly,
there is an intricate relationship between what people
view as problems and what they want for themselves.

For an event to be described as a problem, it must

hinder the fulfillment of some wish, need, or

intention. The similarities between our clusters and

www.theipi.org 167
www.freepsychotherapybooks.org
 the subscales of the Inventory of Interpersonal

Problems increase our confidence in the validity of

the results of our cluster analysis, as well as in the

comprehensiveness of the final list of clustered
categories found in this study.

Responses From Others

The clusters of responses from others also seem

to describe a fairly exhaustive range of interpersonal
responses. People are viewed as or expected to be

strong and independent (Cluster 1), controlling

(Cluster 2), upset (Cluster 3), bad (Cluster 4),
rejecting (Cluster 5), helpful (Cluster 6), liking others

(Cluster 7), and understanding (Cluster 8).

Responses of Self

The clusters of the responses of self also describe

a fairly wide range of personal reactions. Participants

feel open and helpful (Cluster 1), un-receptive to
others (Cluster 2), respected (Cluster 3), in opposition

www.theipi.org 168
www.freepsychotherapybooks.org
 to others (Cluster 4), self-controlled and confident

(Cluster 5), helpless (Cluster 6), disappointed

(Cluster 7), and anxious or ashamed (Cluster 8).

CONCLUSIONS

• This latest set of standard categories is therefore

recommended. Compared with tailor-made
categories, Luborsky, Barber, and Schaffler
(1989) showed that the use of standard categories
and clustered standard categories yielded a
greater agreement among three CCRT judges
when scoring relationship episodes obtained
from a specimen case at intake, termination, and
follow-up evaluation.

• The process of developing a classification of

categories is one of progressive refinement. The
present cluster study is another step in the
direction of creating a comprehensive,
nonredundant list of standard categories. It
represents what we have already achieved, but
we are continually refining the classification. We
certainly share with the readers the wish that a
more finished set were available, although it is to
be expected that we cannot have a final and

www.theipi.org 169
www.freepsychotherapybooks.org
 definitive list describing the various aspects of
interpersonal behavior encompassed in the
CCRT. Nevertheless, we continue to work
toward improved lists, and we have several
studies in progress dealing with this important
issue.

• A note of caution about cluster analysis is in order,

however. One major problem, besides the choice
of which clustering method to use for a specific
data set, is that the researcher must make a
decision about the optimal number of clusters to
describe a set of categories (for a review of these
problems and others, see Aldenderfer &
Blashfield, 1984). Naturally, such decision
making may be biased by the theories of the
investigator.

• For the classification of standard categories, there

are methods other than cluster analysis. One of
them is the reliance on theories of personality.
Perry’s (1993) classification of wishes and fears
is one of these; he followed the eight stages of
development proposed by Erikson (1959). In
Perry’s classification scheme, wishes such as “to
communicate one’s needs, to be protected, or to
survive” are included under Erikson’s first stage,

www.theipi.org 170
www.freepsychotherapybooks.org
 “basic trust versus mistrust.” Fears associated
with this first stage are, for example, fears of
physical harm, being alone, being dependent on
others, and so on. Others have relied on a
different theoretical direction based on
Benjamin’s (1974) Structural Analysis of Social
Behavior (SASB) as an organizational principle.
Two of these groups are Demorest and Crits-
Christoph (1989) and Schacht, Binder, and
Strupp (1984). These theory-based methods have
an important contribution to make alongside the
cluster methods.

www.theipi.org 171
www.freepsychotherapybooks.org
 4

POSITIVE VERSUS NEGATIVE CCRT

PATTERNS

BRIN F. S. GRENYER AND LESTER

LUBORSKY2

A traditional distinction in transference patterns is

between a positive transference and a negative

transference. Freud (1901-1905/1953a; 1912/1958a)

made that distinction, or one like it, almost from the

beginning of his use of the transference concept. The
routine use by clinicians of positive versus negative

transference over so many years suggests that it is a

useful distinction and even that it probably can be

made reliably.

The general aims of this chapter are to define the

positive and negative patterns of the CCRT and their

www.theipi.org 172
www.freepsychotherapybooks.org
 reliability and validity. The specific aims are to look
into the following: (a) the reliability of scoring the

positive and negative dimensions; (b) the results of

other studies of scoring positive and negative

narratives in patients, in nonpatients, and in children;

(c) the changes in positive and negative patterns over
the course of psychotherapy; and (d) the relation of
positive and negative patterns to the outcomes of

psychotherapy.

From the beginning of the CCRT method

Luborsky (1977b) also distinguished between

positive and negative CCRTs (as discussed in chapter

2, this volume). At the start, all CCRTs were labeled
as mainly positive ( + ) or mainly negative ( – ).
Positive means that the patient’s narrative describes
noninterference or an expectation of noninterference

with the satisfaction of wishes or a sense of mastery

about being able to deal with the wishes. Negative

means that the patient describes interference with the

www.theipi.org 173
www.freepsychotherapybooks.org
 satisfaction of wishes that has occurred or is expected
to occur. In CCRT terms, the positive pattern is based

on relationship narratives containing a wish, such as

“I wish to be loved,” followed by a response from

other, such as “the other is affectionate,” followed by

a response of self, such as “I feel close.” This would
be a positive CCRT because the wish to be loved is
satisfied. An example of a negative pattern is the

following: wish: I wish to be loved; response from

other: but the other person rejects me; response of

self: I become frustrated. This example is a negative

pattern because the wish is unsatisfied and the person

feels frustrated about being rejected.

RELIABILITY OF SCORING THE POSITIVE

AND NEGATIVE DIMENSIONS OF THE CCRT

The Two-Category System

The original two-category distinction in scoring

consisted of the scores positive (P) and negative (N)

(Luborsky, 1977b). The agreement between two

www.theipi.org 174
www.freepsychotherapybooks.org
 judges using this distinction is high: 95% for both the
responses from others and the responses of self (see

chapter 9, this volume). In terms of correlations, in

chapter 12 we report that the reliability in scoring the

negative dimension of the response from other was

.67 (p < .0001) and in scoring the negative dimension
of the response of self was .79 (p < .0001).

A neutral or middle category between positive

and negative was not used for both theoretical and
empirical reasons. A theoretical position stated by

Freud (1912/1958a) is that transference responses

must be either positive or negative; for wishes, only

gratification or frustration is possible. Research
results that are probably related appear to be
consistent with this position: Brief exposures of
pictures to participants were experienced as

emotionally tinged with either positive or negative

quality and not neutral quality, and these perceptions

www.theipi.org 175
www.freepsychotherapybooks.org
 appeared seemingly automatic and instantaneous
(Bargh, Chaiken, Govender, & Felicia, 1992).

A New Four-Category Positive and Negative Scale

This section on reliability is based on a scale

expanded from the two-category positive versus

negative system to a four-category scale in which

each response from other (RO) and response of self

(RS) is rated using one of the following four

categories: NN = strong negative, N = moderate
negative, P = moderate positive, and PP = strong

positive. For research purposes, these categories

correspond to scores of –2, –1 and +1, +2,

respectively. Although positive and negative scores

are applied only to the RO and RS components, it is

important in scoring for these to be evaluated in

relation to the expressed or implied wishes. This is

because the RO or RS is only positive or negative

depending on the success in relation to gratifying a
wish.

www.theipi.org 176
www.freepsychotherapybooks.org
 We investigated the reliability of this new four-
category scoring method by having two independent

judges rate the degree of positivity or negativity for

the RS and RO components across a sample of 20

patients from the Penn Psychotherapy Project. Five of

the cases came from the original 10 that were
classified as most improved; 5 cases came from the
original 10 that were classified as least improved; and

10 cases came from the original 21 in the middle

group (see Luborsky, Crits-Christoph, Mintz, &

Auerbach, 1988, for a fuller description of the

sample). In this way we ensured that we had a sample

that covered the full range of positive and negative

CCRT scoring possibilities. Transcripts from both

early in therapy (Sessions 3 and 5) and late in therapy

(at 90% of treatment completion) formed the

database. From these sessions 386 relationship
episodes (generally 10 early in therapy and 10 late)

were extracted, and the CCRT was scored. Prior to

www.theipi.org 177
www.freepsychotherapybooks.org
 the rating task, an independent judge (BG) located all

scorable CCRT components to ensure that judges

each scored the same thought units. The two judges

were both doctoral-level psychoanalytic researchers.

Of the 20 patients, 1 was married, 2 were

divorced, and the rest (85%) were single. The average

age was 23 (range 18-35); there were 10 females and

10 males; and none had children. Educationally they

were diverse; 3 had only finished high school, 7 were

in college, 3 had completed college, 5 were engaged

in getting higher degrees, and another 2 had

completed a postgraduate degree. All were engaged

in psychodynamic psychotherapy, with a mean

duration of 48 weeks (range 27-102 weeks). There

were 19 therapists, with a mean age of 35 (range 26-

47). All therapists were married, 12 had children, all

were psychiatrists, and 13 were psychiatric residents

at the time.

www.theipi.org 178
www.freepsychotherapybooks.org
 We looked at the reliability of the judges’ scoring

of the four categories of positive and negative in two

ways. First, we examined the reliability, taking as the
unit of analysis the relationship episode (RE). Over
386 REs, the interrater reliability using Pearson’s

correlation coefficient was moderate to High. For the

RO dimension, r = .64, p < .0001, and for the RS

dimension, r = .72, p < .0001. Second, we looked at

instances of perfect agreement of scoring. For the

RO, 278 of the 386 scored REs were scored

identically between judges (72%), and for the RS,
274 were scored identically (71%). We then looked at

the interjudge reliability taking subjects (N = 20) as

the unit of analysis rather than REs (N = 386). We
calculated average positive and negative scores for
each subject for each judge. The interrater reliability
for the RO component was r = .77, p < .001, and for

the RS, r = .93, p < .001.

www.theipi.org 179
www.freepsychotherapybooks.org
PROPORTION OF POSITIVE AND NEGATIVE
NARRATIVES IN CHILDREN, ADULT
PATIENTS, AND NORMAL CONTROLS

A few studies have compared the distinction of

positive versus negative narratives across different

samples. These findings have been discussed in detail

in chapter 16, this volume, but we provide a brief
overview here. Narratives were obtained from 18
children who were interviewed at 3 and 5 years of
age using a guided and prompted story-completion
task about a doll family (Buchsbaum & Emde, 1990).

The children’s narratives were scored with the CCRT,

and the percentages of positive and negative

components were computed. At age 3, 69% of the
ROs were positive, and at age 5, 71% were positive.
For the RS component, 63% were positive at age 3,
and 77% were positive at age 5. Thus most of the doll

family narratives told by the young children

contained positive responses from other people and
positive responses from the self.

www.theipi.org 180
www.freepsychotherapybooks.org
 These findings with children are in strong contrast

with those typically found in adult patients in

psychotherapy. For example, in the Penn Depression

Study (N = 30), only 21% of combined RO and RS

components were positive. Similarly, among patients

with major depression seen on an outpatient basis at
Penn (N = 20) only 19% of combined RO and RS
components were positive (Eckert, Luborsky, Barber,

& Crits-Christoph, 1990). These results are highly

similar to those found in patient reports of dreams

versus waking narratives told during psychotherapy

(chapter 12, this volume). It was found that in both
dreams and waking narratives told late in therapy, the
percentages of negative components ranged from
67% to 76% for RO components and from 65% to
80% for RS. It is interesting that the response of self

in waking narratives was significantly more negative

than in dreams, which might suggest the influence of

striving toward wish fulfillment in dreams.

www.theipi.org 181
www.freepsychotherapybooks.org
 For further contrast, two studies provided data on
normal nonpatient adult controls. In Ulm, Germany,

36 normal young women (mainly medical students),

mean age 23.8 years (range 20-30), were asked to

provide narratives about their relationships, and the

resulting narratives were CCRT scored (Dahlbender
et al., 1992). Thirty-seven percent of RO and 43% of
RS components were positive. In another German

study of normal adults (N = 30), 35% of RS

components were positive (Cierpka et al., 1992).

In summary, children’s narratives appear to be far

more positive in both RO and RS components than

those of normal adults, and narratives of
psychotherapy patients are even more negative. The
different conditions under which the sets of narratives
were collected may partly explain the results. With

the children, actual interactions were not prompted;

rather, the children completed a fictitious doll family

story. Some of their answers, therefore, may be

www.theipi.org 182
www.freepsychotherapybooks.org
 heavily infused with wish fulfillment fantasies. Also,

the narratives collected from the patients were told in

the context of psychotherapy, where presumably
patients focus on their most troublesome problems

through the telling of negative and problematic

relationship narratives. Contextual factors within
psychotherapy also may have an influence on the
magnitude of the positive versus negative dimensions

in psychotherapy. For example, as shown in chapter

9, this volume, relationship episodes told about the

therapist by the patient were found to be less negative

than relationship episodes told about others: For the
RO and RS components, 55.8% and 61.5%,
respectively, were negative for the therapist
relationship episodes, compared with 81.5% and
88.5%, respectively, for other relationship episodes.

This suggests either that patients had difficulty telling

negative relationship episodes about the therapist in

the presence of the therapist or that their relationship

www.theipi.org 183
www.freepsychotherapybooks.org
 with the therapist was in fact less negative. In spite of

these possible confounds, it remains plausible that

there are genuine developmental differences in the

emotional quality of relationship narratives from
childhood through adulthood, with adults being

“sadder yet wiser” in appraisal of both others’

motives and their own.

CHANGES IN POSITIVE AND NEGATIVE

DIMENSIONS OVER THE COURSE OF
PSYCHOTHERAPY

We were interested in the distribution of positive

and negative scores for the RO and RS dimensions
using the expanded scoring method. Given the high

interjudge agreement, we pooled the data from the

two judges. Overall, for the RO there were 10% NN
(very negative), 64% N (negative), 24% P (positive),
and 2% PP (very positive) scores. For the RS, there
were 12% NN, 58% N, 24% P, and 6% PP scores.

These results are similar to the distributions of

www.theipi.org 184
www.freepsychotherapybooks.org
 positive and negative components found in other

psychotherapy samples, as reviewed previously. In

general, it seems that the bulk of CCRT narratives

told by patients in psychotherapy are negative; that is,

wishes and needs are often unsatisfied.

To investigate changes in the proportion of

positive and negative components over the course of
psychotherapy, we calculated percentages of each
component early and late in therapy. We further

differentiated the data into the five most improved

and the five least improved patients (based on the
criteria in the Penn Psychotherapy Study of

Luborsky, Crits-Christoph, et al., 1988). This was

done to illustrate graphically some of the patterns of

change in the percentages of the positive and negative

dimensions. The results for the RS dimension appear

in Figure 1 and for the RO dimension in Figure 2.

www.theipi.org 185
www.freepsychotherapybooks.org
 Figure 1.

Percentages of positive versus negative scores for the

response of self (RS) component of the CCRT for five
highly improved patients (best) and five least
improved patients (worst) measured early in
psychotherapy (early) and late in psychotherapy
(late). NN = strong negative; N = negative; P =
positive; PP = strong positive.

www.theipi.org 186
www.freepsychotherapybooks.org
 Figure 2.

Percentages of positive versus negative scores for the

response of other (RO) component of the CCRT for
five highly improved patients (best) and five least
improved patients (worst) measured early in
psychotherapy (early) and late in psychotherapy
(late). NN = strong negative; N = negative; P =
positive; PP = strong positive.

www.theipi.org 187
www.freepsychotherapybooks.org
 For the RS dimension it can be seen that for the

five most improved patients (best, in Figure 1) there

are large reductions in the number of negative (N and

NN) components (about a 35% reduction for N) and a

parallel increase in the number of positive (P and PP)

components from early in therapy to late in therapy (a
35% increase for P). This is indicative of an increase
in the satisfaction of wishes and needs. For the five

least improved patients (worst, in Figure 1), the

pattern is in the reverse direction: There are increases

in the N component late in therapy and parallel

reductions in the P component. For the least
improved patients, there appear to be increases in the
perception of others as blocking the gratification of
wishes and overall their perception of others is very
negative throughout therapy. As the CCRT judges

observed, these patients seemed to be getting worse;

their relationship narratives were more conflictual

www.theipi.org 188
www.freepsychotherapybooks.org
 and negative toward the end of therapy compared
with their initial level.

The pattern for the RO dimension (Figure 2) is

similar to that of the RS dimension, but the changes

are less striking. The most improved patients again

reduced their negativity and increased their positivity
within relationship conflicts, and for the least
improved patients, the reverse trend occurred,
although only slightly. Overall, the response of other
dimension did not change as much over the course of

psychotherapy for either the most improved or the

least improved patients.

RELATIONSHIP TO PSYCHOTHERAPY
OUTCOMES

The final aim of this chapter is to investigate the

relationship of the positive-negative dimension to two

psychotherapy outcome measures, the Health-

Sickness Rating Scale (HSRS) and the Mastery Scale.

The HSRS is a commonly used outcome measure in

www.theipi.org 189
www.freepsychotherapybooks.org
 which a 100-point scale of degrees of mental health-

sickness is rated by a trained observer on the basis of

data collected from a clinical interview. The Mastery
Scale is a 6-point content analysis scale measuring

degrees of self-control and self-understanding scored

from relationship narratives told by patients in
psychotherapy (Grenyer, 1994). It has high reliability
and validity, and changes in self-understanding and

self-control have been associated with good

psychotherapy outcome as assessed by the patient,

the therapist, and independent judges (see chapter 18,

this volume). Table 1 presents the results of the
analyses.

The RO component showed little relationship to

the outcome variables. The RO early in therapy was

related to mastery early in therapy, and changes in the

RO over therapy were related to the termination score
on the HSRS. In contrast, the RS component appears
to be related in an important way to clinical changes.

www.theipi.org 190
www.freepsychotherapybooks.org
 TABLE 1
Pearson Correlations Between Positive and Negative CCRT
Response of Other (RO) and Response of Self (RS) Scores with
the Health-Sickness Rating Scale (HSRS) and Mastery Scale

Scale RO.early RO.late RS.early RS.late RO.change RS.change

Mastery.early .67* -.15 .68* .15 -.47* -.00
Mastery.late .26 .35 .24 .82* .30 .57*
HSRS.intake .18 .45* -.12 .38 .29 .39
HSRS.termination .03 .47* -.19 .64* .47* .46*
Mastery.change .09 .31 .12 .70* .38 .51*
HSRS.change -.20 .19 .02 .49* .32 .16

Note. Scores are for early in therapy (early), late in therapy (late), and
residual change over therapy (change) (N = 20).

*p < .05.

www.theipi.org 191
www.freepsychotherapybooks.org
 We found significant concordance between the RS
components and the Mastery Scale scores both early

and late in therapy. The RS late-in-therapy measure

appeared to be a particularly good index of

psychological health. There was a strong relationship

between RS late and residual change in Mastery
Scale scores (r = .70, p < .05) and residual change in
the HSRS score (r = .49, p < .05). There was also a

strong concordance between the RS late scores and

the late-in-therapy Mastery Scale scores (r = .82, p <

.05) and the HSRS termination scores (r = .64, p <
05). In addition, residual change in the RS component

was significantly related to late Mastery Scale scores,

termination HSRS scores, and residual change in

Mastery Scale scores (r = .51, p < .05).

CONCLUSION

This study was the first to investigate the

reliability and validity of the expanded scale method
of scoring the positive and negative dimension of the

www.theipi.org 192
www.freepsychotherapybooks.org
 CCRT. Some impressive results appear but they

require replication with a larger sample. These are the

clearest of the findings:

• Judges are able to agree very well on the scoring of

the positive and negative dimensions. This seems
to be due in part to the simplicity and clarity of
the basic CCRT method of scoring these
dimensions.

• Adult patients in therapy have very negative

relationship narratives, particularly when
compared with adults and with children who
were not in therapy.

• The change in positive and negative dimensions

from early to late in psychotherapy typically is
small, yet the change that does occur appears to
be clinically meaningful.

• These findings about changes during therapy,

however, relate primarily to the changes in the
response of self dimension of the CCRT, which
shifts in clinically meaningful ways (see Table
1). It appears from the data we have so far that
the response of other dimension is only indirectly
related to clinical changes. In fact we found far

www.theipi.org 193
www.freepsychotherapybooks.org
 fewer changes overall from early to late in
therapy in the percentage of positive and
negative components in the response of other as
compared with the response of self dimension
(see Figures 1 and 2). We explain this in the
following way: Psychotherapy does not eliminate
all intrapsychic conflicts, so that most of the
same CCRT patterns found early in therapy
remain late in therapy. In other terms, the
transference template is pervasive and resistant
to change. This is consistent with the conclusion
that what changes in psychotherapy is a person’s
mastery of his or her problems in the form of
greater self-understanding and self-control (see
chapter 22, this volume).

Note

[2] The research in this chapter was supported in part by grants

from the Australian Research Council to Brin Grenyer,
with further support from other sources to Lester Luborsky
by grants listed in the acknowledgments section of this
book. An earlier version was given at the Society for
Psychotherapy Research Annual Meeting, Amelia Island,
Florida, June 1996, and to the International CCRT
Workshop in Ulm, Germany, in May 1995. For assistance
with data collection, we thank Mary Carse, Richard
Rushton, Nadia Solowij, and Kelly Schmidt.

www.theipi.org 194
www.freepsychotherapybooks.org
 5

ILLUSTRATIONS OF THE CCRT

SCORING GUIDE

LESTER LUBORSKY AND SCOTT

FRIEDMAN

This chapter offers scored narratives to serve as

guides to CCRT scoring procedures. These

illustrations, which we draw on throughout the book,

can be used by readers as practice cases for learning

the CCRT procedures. The scoring is given for three
patients’ CCRTs, each engaged in a different type of

dynamic therapy: Ms. Sandy Smyth in a short-term,

time-limited psychodynamic psychotherapy, Mr.
Edward Howard in a moderate-length psychodynamic
psychotherapy, and Ms. Cathy Cunningham in a long-
term psychoanalysis.

www.theipi.org 195
www.freepsychotherapybooks.org
 For each of the three patients, the chapter consists
of (a) a clinician’s brief sketch of the patient’s state

before and after the treatment, (b) a sample of the

relationship episodes in the sessions early and late in

the therapy, and (c) a sample of the CCRT scoring.

The sample of scoring includes for each case, first for
early and then for late sessions, the tailor-made score
sheet, the tailor-made summary, and the standard

category summary. Of all this, the most useful

materials for learning the CCRT method are the

scored text of the relationship episodes and the

summary of the scoring of the standard categories.

These procedures are especially suited for research

uses; the clinical uses are discussed in chapter 19, this

volume.

MS. SMYTH: SHORT-TERM PSYCHODYNAMIC

PSYCHOTHERAPY

The example from Ms. Smyth’s sessions

illustrates the application of the CCRT to a 16-session

www.theipi.org 196
www.freepsychotherapybooks.org
 time-limited psychotherapy. The mode of treatment
was supportive-expressive psychotherapy as guided

by the recommendations of a manual for the

treatment of major depression (Luborsky, Mark,

Hole, Popp, Goldsmith, & Cacciola, 1995), which is a

specific version of the general manual for supportive-
expressive therapy (Luborsky, 1984).

Clinical Evaluation

Initial

At the time of the evaluation, Ms. Smyth was a

32-year-old single woman. The pretreatment

diagnostic evaluation arrived at a DSM-IV diagnosis

of alcohol dependence (303.90), sustained full

remission; major depressive disorder, recurrent,
moderate severity (296.32); and dysthymic disorder
(300.4). She did not meet Axis II personality disorder

criteria. The patient was a recovering alcoholic who

had been abstinent for 3 years. She came for

www.theipi.org 197
www.freepsychotherapybooks.org
 treatment for depression (with a high initial Beck
Depression Inventory score of 25) after having

flunked out of a job training program. The therapy

began inauspiciously when she showed up half an

hour late and said she was unable to schedule a next

appointment. The therapist’s reaction was one of
anger, which the therapist did not express; however,
she used her perception of her own experience to

recognize the feeling that the patient was setting up in

her. When the patient said she was afraid of

“sabotaging” herself, the therapist did say she thought

the patient was correct to be concerned.

Termination

Ms. Smyth continued to have difficulty in keeping

appointments. Nevertheless, she benefited remarkably

well and surprised the therapist by how well she did:

At termination her Beck Depression Inventory score

was 6. The therapist concluded in her termination
evaluation, “I would not have thought someone with

www.theipi.org 198
www.freepsychotherapybooks.org
 such severe depression and who already was making
full use of self-help therapeutic groups [such as

Alcoholics Anonymous] could have resolved her

depression without the use of

psychopharmacotherapy.”

In the termination interview, Ms. Smyth stated

that she was generally feeling “good” and that
“everything’s a lot better.” She had been seeing a man
for 5 months (since shortly after she began therapy)
with whom she was pleased. She had also set up a

stable living arrangement with a female roommate

and was working regularly in a clerical job she was

not pleased with. She still complained of
premenstrual symptoms—tension and a headachy
feeling. Recently her period had been late; she was
concerned about being pregnant and believed she

might have had a miscarriage. She generally seemed

in far less turmoil and was less pessimistic and much

more confident and hopeful. She gave the impression

www.theipi.org 199
www.freepsychotherapybooks.org
 that she could take care of herself, in comparison
with the time of the initial evaluation when she had a

desperate, disorganized quality.

Six-Month Follow-up

Ms. Smyth continued not to feel depressed to any

significant extent. Her Beck Depression Inventory

score was 9. She has continued working full-time,

although still at the same kind of work. Ms. Smyth

has found out she is pregnant by the man she is
involved with. She plans to be married, but the man is

waffling on commitment. The patient is angry,

anxious, and worried about the situation but feels she

can handle whatever happens and will have the baby.

At first news of her pregnancy, she developed a

probable generalized anxiety disorder and missed

some work. She and her boyfriend entered weekly

couples therapy at that time and they continue in it.

She has also maintained involvement with AA. She
continued to live with the roommate and maintained

www.theipi.org 200
www.freepsychotherapybooks.org
 contact with her family and a few close friends.

Although this has been a difficult time because of her

pregnancy and the ambivalent boyfriend, she

expressed a resolve that she will get by, whatever it
takes. Even with these stressors, although she was

initially frantic for a short while, she is now basically

okay and is not taking any medications for depression

or anxiety.

CCRT Data and Results

In scoring each session the judge first reads the

transcript, then scores the relationship episodes by
placing notations in the left margin of the transcript

(as an option, the scorer may also transfer the scores

for each relationship episode to the CCRT score

sheet; see Table 1), and then summarizes these on the

CCRT summary sheet (see Tables 2 and 3). This

juxtaposition of an unguided clinical sketch, next to a
systematically guided clinical-quantitative description
achieved by the CCRT, makes clearer what is

www.theipi.org 201
www.freepsychotherapybooks.org
 contributed by the CCRT to the formulation of the

central relationship pattern (Exhibit 1).

To illustrate the CCRT method, a set of

relationship episodes from Sessions 3 and 5 are

reproduced in a subsequent section, together with

their scoring in the left margin. Only five scored

relationship episodes are presented, to keep the

example brief. For further simplification, the scoring

written in the left margin is only by one experienced

CCRT judge.

To specify precisely what is scored, the text is

divided into single thought units (marked off by
slashes); the marking of thought units can be done

rapidly, as described by Benjamin (1986b). Those

parts of relationship episodes that are to be scored are

connected with the scores by an arrow.

Ms. Smyth’s CCRT data and results include (a)

the scored relationship episodes, (b) the score sheet

www.theipi.org 202
www.freepsychotherapybooks.org
 EXHIBIT 1
SYMBOLS FOR CCRT SCORING TO BE APPLIED TO THE
RELATIONSHIP EPISODES

// Slashes mark the beginning and end of a

thought unit.
W, RO, or RS Wish, response from other, or response of
P or N self, positive or negative (P or N) when
appropriate, are to be scored in the left
margin at the beginning of the thought unit.
--- Interrupted or broken-off speech.
xxx Inaudible word.
xxx– Inaudible text longer than one word.
… Section of text omitted.
(20, 17) Numbers in parentheses refer to the
standard category numbers (see chapter 3,
this volume). The first number indicates the
best-fitting category; the second is for the
next-best-fitting category. A line under a
number means the episode is an exact fit to
a category on the list; a question mark
means a questionable fit. (Note that some
thought units are not scored because they do
not fit any scorable category or they have
already been given a score of the same
type.)

www.theipi.org 203
www.freepsychotherapybooks.org
 for each relationship episode by tailor-made
categories (Exhibit 1), (c) the summary for all

relationship episodes by tailor-made categories (Table

2), and (d) the summary for all relationship episodes

by the standard category system (Table 3).

The main CCRT for the five relationship episodes

that can be read from the tailor-made CCRT score
sheet summary in Table 1 follows. The number in
parentheses after each component is the number of
relationship episodes out of the five scored in which

the component is found:

Wish: I want to end nonsupportive relationships

(3) and to have support and caring (3).

Response from other: But the other person—

others include her ex-employers, her brother, her
boyfriend, and her father—is rejecting (4) and
unsupportive (2).

Response of self: I then feel angry (3) and feel

bad (depressed) (2).

www.theipi.org 204
www.freepsychotherapybooks.org
 The CCRT from the standard category system

using only the best-fitting standard category (Table 3)

is similar to the one from the tailor-made system:

Wish: I want to oppose others (2) (to end

nonsupportive relationships) and to be helped (3)
(nurtured, supported, and given things I need).

Response from other: But the person is rejecting

(3), unhelpful (3), and not trustworthy (2).

Response of self: I then feel angry (2) and

depressed (2) and ashamed (2).

It is noteworthy that by both the tailor-made and

the standard category systems there are also a few

positive responses from the other person, such as

“others are helpful” (2). These may imply that there

is for this patient, apparent in the early phase of
treatment, a potential for positive relationships.

Ms. Smyth: CCRT Scoring of Relationship Episodes

Session 3

RE 2: Ex-employers, Completeness Rating 4.0

www.theipi.org 205
www.freepsychotherapybooks.org
 And then the other job was, ah, they didn’t even
give me a chance. I was supposed to work on
this computer and the computer wasn’t hooked
up and they said, well you don’t have to work on
it. We’re replacing you with somebody else. So I
was replaced with somebody else. It really
pissed me off [pause] ‘cause I gave up another
job to get this job and I ended up with nothing at
all, no unemployment, no nothing. It’s horrible. I
call ‘em everyday, but they always say we don’t
have anything. It’s just terrible. Because I don’t
know what I’m going to do. [pause] It’s really
discouraging. It’s so hard to get out and—get the
door slammed in my face constantly. [pause]

✔ * NRO: //RO And then the other job, was, ah,

Rejecting (4, 14) they didn’t even give me a chance.
NRO: Rejecting (4, //RO I was supposed to work on this
14) computer and the computer wasn’t
hooked up and they said, well you don’t
have to work on it.
✔ NRO: Give no //RO We’re replacing you with somebody
help (14, 19?) else.
NRO: Rejecting (4, //RO So I was replaced with somebody
14) else.
✔ NRS: Anger (21, //RS It really pissed me off
20?)
✔ NRS: Have no //RS [pause] ’cause I gave up another
job, nothing (20?, job to get this job and I ended up with
17) nothing at all, no unemployment, no
nothing.

www.theipi.org 206
www.freepsychotherapybooks.org
NRS: Horrible //RS It’s horrible.
state (20?, 17?)
✔ W: I want job //W I call ’em every day,
(help) (13?, 3?)
NRO: Rejecting (4, //RO but they always say we don’t have
14) anything.// It’s just terrible.
✔ NRS: Helpless //RS Because I don’t know what I’m
(17, 19) going to do.
✔ NRS: //RS [pause] It’s really discouraging.
Discouraged (22,
23?)
NRO: Rejecting (4, //RO It’s so hard to get out and—get the
15) door slammed in my face constantly,
[pause]//

*Checks mean that the items were used in the tallies in the summaries
of the standard category.

RE 3: Brother and His Wife, Completeness Rating 4.5

Anyway, I want to move out of Bob and Jane’s

(brother and sister-in-law) house as soon as
possible. Treated like a second-rate citizen there.
It’s not very good for my self-esteem. Like
they’re both addicts and they have the
personality of addicts. [pause] I guess … I, I
much rather be around sober people. … Yeah.
The old tapes start running and it’s just real bad.
I mean I start thinking negatively as soon as I’m
around them, ‘cause they’re both negative.
They’re dishonest. They’re acting like they’re
doing me a big favor, but I’m paying half the
rent there, for their apartment, and I have this

www.theipi.org 207
www.freepsychotherapybooks.org
 tiny little room, no closet, and their junk’s in the
room, and un I have to work around their lives.
[pause] So, I ah just can’t stand them.

✔ W: To get out of //W Anyway, I want to move out of Bob

bad rel. (13, 23) and Jane’s (brother and sister-in-law)
house as soon as possible.
✔ NRO: Rejecting //RO Treated like a second-rate citizen
(4, 6?) there.
✔ NRS: Feel bad //RS It’s not very good for my self-
about self (26?, esteem.// Like they’re both addicts and
17?) they have the personality of addicts.
✔ W: To be in //W [pause] I guess … I, I much rather
good rel. (3, 2) be around sober people.// … Yeah. The
old tapes start running// and it’s just real
bad.
✔ NRS: Feel bad //RS I mean I start thinking negatively as
(22?, 20?) soon as I’m around them, ’cause they're
both negative.
✔ NRO: Dishonest //RO They’re dishonest.
(8, 15)
NRO: Putting her //RO They’re acting like they’re doing me
down (8, 4) a big favor, but I’m paying half the rent
there, for their apartment,// and I have
this tiny little room, no closet, and their
junk’s in the room, and uh I have to
work around their lives.
✔ NRS: Anger (21, //RS [pause] So, I ah just can’t stand
6) them.

RE 4: Boyfriend, Completeness Rating 5.0

Yeah, I’ve, and I’ve stopped speaking to that

married guy. ‘cause he got to be a real asshole. I

www.theipi.org 208
www.freepsychotherapybooks.org
 mean I’m not taking shit from anybody this year
—for the rest of my life and, uh, he just sort of
stopped talking to me and, uh, he didn’t contact
me, he didn’t even—where I was going to move
to this week? he didn’t contact me so screw him.
I’m not going to contact him, not at all, either…
it just makes me mad. I really don’t want
anything at all to do with him. Never again will I
—Christmas Eve I spent alone in church crying
my eyes out ‘cause it was an intensely lonely
feeling and I said no way am I ever gonna feel
that bad again. no way. I’m isolated from my
friends and family because of this guy I wanted
—this married guy. It’s just a conflict between
honesty and dishonesty. … I just, ah, he pissed
me off. All my other friends gave me all kinds of
moral support, even some financial support for
this horrible dilemma I’m in right now. He didn’t
do shit.… He didn’t buy my a Christmas present
or a card or a birthday.…

✔ (W): To stop bad //W Yeah, I’ve, and I’ve stopped

rel. (18, 23) speaking to that married guy
✔ NRO: Rejecting //RO ’cause he got to be a real ass hole.
(4, 14)
✔ PRS: Assertive //RS I mean I’m not taking any shit from
about stopping anybody this year—for the rest of my life
rejecting rel. (12,
11 ?)
✔ NRO: Stopped //RO and, uh, he just sort of stopped

www.theipi.org 209
www.freepsychotherapybooks.org
talking to me (12?, talking to me and, uh,
14)
NRO: Didn't //RO he didn’t contact me, //he didn't
contact me (12, 4) even—where was I going to move to
this week?// He didn’t contact me
✔ NRS: Anger (21, //RS so screw him,
6)
✔ PRS: I stop //RS I’m not going to contact him, not at
contact with him all, either …
(15, 23)
NRS: Anger (21, 6) //RS it just makes me mad.
✔ PRS: Reject //RS I really don’t want anything at all to
other (6, 21) do with him.
✔ (W): Not to be //W Never again will I—
lonely again (11,
14)
✔ NRS: Lonely, //RSChristmas Eve I spent alone in
crying (23?, 22?) church crying my eyes out ’cause it was
an intensely lonely feeling.
W: Not to feel //W and I said no way am I ever gonna
isolated (11, 14?) feel that bad again. No way.
NRS: Isolated (23, //RS I’m isolated from my friends and
20) family because of this guy I wanted—
this married guy.// It’s just a conflict
between honesty and dishonesty.
NRS: Anger (21, 6) //RS... I just, ah, he pissed me off.
✔ PRO: Other //RO All my other friends gave me all
friends gave kinds of moral support, even some
support (13,3) financial support for this horrible
dilemma I’m in right now.
✔ NRO: Gave no //RO He didn’t do shit….
support (14, 4)
NRO: Gave no //RO He didn’t buy me a Christmas
support (4, 14) present or a card or a birthday… .//

www.theipi.org 210
www.freepsychotherapybooks.org
 Session 5

RE 5: Father, Completeness Rating 3.0

I mean he’s acting just like an asshole. I mean, to

him I think my grandmother was always a pain
in the ass. that’s how he treated his children too.
That’s shy he’s just a total asshole. A couple of
years ago she was sick and she had a sister that
lived out in California. he wanted to ship her out
there so he wouldn’t have to deal with her and
all. it was horrible. And that’s what—and soon
after that when I found out how he treated his
mother, I realized how he [sniff] he felt about
me, and the rest of my sisters. Like he didn’t
give a shit. And I xxx my alcoholic father after
that. It was just a matter of a couple of weeks.
[pause] You know, I saw what he’s really like
went to xxx to see him at Christmas time, New
Year’s and I went hoping to get money,
[Laughing quietly] xxx I didn’t get any. …
[laughing] And I needed it at the time. I felt sort
of like a whore but I needed money. … [pause] I
just want him out of my life. He just xxx. At one
time xxx and denied that he was a bastard and an
asshole but then I saw right through his face
what he really is. I didn’t want to know that I
had a father that was a big asshole. I found out

www.theipi.org 211
www.freepsychotherapybooks.org
 that he was. Just devastated me. [pause] You
know—definitely a dysfunctional family.

✔ NRO: Asshole //RO I mean he’s acting just like an

(25?, 4?) asshole.//I mean, to him I think my
grandmother was always a pain in the
ass.
✔ NRO: Rejecting, //RO That’s how he treated his children
nonsupportive too.// That’s why he’s just a total
(14,4) asshole. A couple of years ago she was
sick and she had a sister that lived out
in California.
NRO: Rejecting, //RO He wanted to ship her out there so
nonsupportive, he wouldn’t have to deal with her and
nonloving (14, 4) all./ It was horrible.// And that’s what—
and soon after that when I found out
how he treated his mother, I realized
how he [sniff] he felt about me, and the
rest of my sisters.
NRO: Noncaring //RO Like he didn’t give a shit.// And I
(14, 4) xxx my alcoholic father after that.
✔ PRS: //RS It was just a matter of a couple of
Awareness of his weeks, [pause] You know, I saw what
nature (1?, 6?) he’s really like// went to xxx to see him
at Christmas time, New Year’s
✔ W: To be given //W and I went hoping to get money.
to (13, 8?)
NRO: //RO [Laughing quietly] xxx I didn’t get
Nonsupportive any.… [laughing] And I needed it at the
(14?, 4?) time.
✔ NRS: Shame //RS I felt sort of like a whore but I
about her asking needed money.
for money (26, 25)
✔ W: To end //W …[pause] I just want him out of my
nonsupportive life. He’s just xxx. At one time xxx and

www.theipi.org 212
www.freepsychotherapybooks.org
relationship denied that he was a bastard and an
(23,18) asshole
✔ PRS: See his //RS but then I saw right through his face
true nature (1?, what he really is.// I didn’t want to know
6?) that I had a father that was a big
asshole.
✔ NRO: He is an //RO I found out that he was.// Just
asshole, although devastated me. [pause]// You know—
he denies it (8?, definitely a dysfunctional family.//
2?)

RE 6: Boss, Completeness Rating 2.5

P: But these people I work for are very, very nice

xxx [pause] Yeah, I feel lucky to have a boss like
my boss. xxx- Yeah I mean I went…

T:…

P: and I was down in town with her xxx her

husband and me. Went to see a trade show this
week and he drove and we were out together. It
was nice. xxx…There’s a lot of them.

T: They included you with them?

P: Yeah.

T: And you just started working there?

P: Yeah. She followed me last week. My first

week on the job when the car broke down, to
make sure my car started. She followed me

www.theipi.org 213
www.freepsychotherapybooks.org
 down there xxx to my house which is like an
hour out of her way.

T: Ha!

P: And she was concerned about, like, you know,

about me, how I was feeling about my
grandmother dying. she let me leave early. Very
nice lady. I feel blessed [pause] I don’t know
[pause] see I’ve had some bad experiences
which ah [whispering] xxx … [long, inaudible].
Yeah, that will be fun. I’ve never been to
Memphis.

✔ PRO: Nice P: //RO But these people I work for are

people (13?, 9?) very, very nice// xxx [pause]
✔ PRS: Feel lucky //RS Yeah, I feel lucky to have a boss
(29?, 28?) like my boss.// xxx- Yeah I mean I
went…
T: …
P: //and I was down in town with her xxx
her husband and me.// Went to see a
trade show this week
✔ PRO: Spends //RO and he drove and we were out
time with me (3?, together.
9?)
PRS: It was nice //RS It was nice.//xxx … There’s a lot of
(29, 28?) them.
T: They included you with them?
P: Yeah.
T: And you just started working there?
P: Yeah.
PRO: Supports her //RO She followed me last week. My first
(13, 9?)

www.theipi.org 214
www.freepsychotherapybooks.org
 TABLE 1
CCRT Score Sheet for Each Relationship Episode, Tailor-Made
System

Patient: Ms. Sandy Smyth Date:

1/26/89
Sessions 3 and 5 Judge: LL
Number of REs: 5
RE No. Wish, Need, Response Response
Person: Intention From Others of Self
RE 2 To have job Rejecting
Ex-employers (help)
Not helpful Angry
Have
nothing, no
job
Feel
horrible;
FEEL BAD*
Feel
helpless
Discouraged
RE 3 To get out of a Rejecting I feel bad
Brother and nonsupportive about
wife relationship and myself
be in a good
one
Put me down Angry
Dishonest
RE 4 To end Rejecting
Boyfriend nonsupportive
relationship
To not be lonely Broke contact Angry

www.theipi.org 215
www.freepsychotherapybooks.org
 or hurt with me
Not helpful/ I break
supportive contact with
him
PRO Lonely,
Others are isolated
supportive
Sad
PRS
Assertive
about
ending
relationship,
rejects other
Won’t be
lonely
RE 5 To get money Rejecting Aware of his
Father nature
To end Nonsupportive Ashamed of
nonsupportive asking for
relationship money
Nonloving
Noncaring
Dishonest
Asshole
Denies that
he’s an
asshole
RE 6 Boss To have PROs: PRSs:
concern and Nice Feel lucky
caring
Spends time It was nice
with me
Supportive Feel
blessed

www.theipi.org 216
www.freepsychotherapybooks.org
 Helpful, even
at a sacrifice
Concerned
Giving

*Additions from Step 1 are put in capital letters, as noted in chapter 2.

Capital letters indicate the name of the standard category.

www.theipi.org 217
www.freepsychotherapybooks.org
 TABLE 2
CCRT Score Sheet, Tailor-Made System: Summary Across All
Relationship Episodes

Patient: Ms. Smyth Date: 1/26/89

Sessions 3 and 5 Judge: LL
Number of REs: 5
Response
Wish, Need, Response of Self (RE
From Others
Intention (RE No.) No.)
(RE No.)
Negative Negative
W1: To end Rejecting (2, 3, Angry (2, 3, 4)
nonsupportive 4, 5)
relationship (3, 4, 5)
W1: To have Are not helpful Feel bad (2, 3)
support and caring or supportive
(2, 5, 6) (2, 4, 5)
W2: To not be Dishonest (3, Feel helpless (2)
lonely (4) 5)
W3: To have job (2) Puts me down Feel discouraged (2)
(3)
W3: To get money Won’t contact Have nothing, no job
(5) me (4) (2)
Nonloving (5) Lonely, isolated (4)
Noncaring (5) I break contact (4)
Asshole (5) Sad (4)
Denies he’s an Aware of other’s
asshole (5) nature (5)
Ashamed (5)
Positive Positive
Others are Assertive (about
supportive (4, ending nonsupportive

www.theipi.org 218
www.freepsychotherapybooks.org
 6) relations) (4)
Nice (6) Feel lucky (6)
Spends time It was nice (6)
with me (6)
Helpful, even at Feel blessed (6)
a sacrifice (6)
Concerned (6)
Giving (6)

www.theipi.org 219
www.freepsychotherapybooks.org
 TABLE 3
CCRT Summary: Standard Categories

Patient: Ms. Smyth

Sessions 3 and 5 Judge: LL
Number of REs: 5 (In RE number 2, 3, 4, 5, 6)
Frequency
Edition 2 Category Across REs
RE No.
Number (first choices
only)
Wishes
13: TO BE HELPED 2, 5, 6 3
18: TO OPPOSE 3, 4 2
OTHERS
3: TO BE RESPECTED 3 1
11: TO BE CLOSE TO 4 1
OTHERS
23: TO BE 5 1
INDEPENDENT
Negative responses from others
4: ARE REJECTING 2, 3, 4 3
14: ARE UNHELPFUL 2, 4, 5 3
8: ARE NOT 3, 5 2
TRUSTWORTHY
12: ARE DISTANT 4 1
25: ARE BAD 5 1
Positive responses from others
13: ARE HELPFUL 4, 6 2
3: ARE ACCEPTING 6 1
11: ARE OPEN 6 1
Negative responses of self

www.theipi.org 220
www.freepsychotherapybooks.org
21: FEEL ANGRY 2, 3, 4 3
22: FEEL DEPRESSED 2, 3 2
26: FEEL ASHAMED 3, 5 2
17: AM HELPLESS 2 1
20: FEEL 2 1
DISAPPOINTED
23: FEEL UNLOVED 4 1
Positive responses of self
1: UNDERSTAND 5 1
6: DISLIKE OTHERS 4 1
12: AM CONTROLLING 4 1
15: AM INDEPENDENT 4 1
29: FEEL HAPPY 6 1

www.theipi.org 221
www.freepsychotherapybooks.org
 week on the job when the car broke
down, to make sure my car started.//
She followed me down there xxx to my
house
PRO: Helps even //RO which is like an hour out of her
at a sacrifice (13, way.//
11?) T: Ha!
PRO: Concern (13, P: //RO And she was concerned about,
3?) like, you know, about me,
✔ (W): Wish for //(W) how I was feeling about my
concern and caring grandmother dying.// She let me leave
(13, 11?) early.//[pause]
✔ PRO: Nice lady //RO Very nice lady.
(11?, 1 ?)
PRS: Feel blessed //RS I feel blessed [pause] I don’t know
(29?, 28?) [pause] see I’ve had some bad
experiences which ah [whispering] xxx
… [long, inaudible]. Yeah, that will be
fun. I’ve never been to Memphis.

MR. EDWARD HOWARD: PSYCHOANALYTIC

PSYCHOTHERAPY

Clinical Evaluation

Initial

Mr. Howard was a 20-year-old man, born and

raised in the same area in which he now was

attending college. He had completed part of his third

year of college when he came for psychotherapy to a

www.theipi.org 222
www.freepsychotherapybooks.org
 practitioner in his hometown with complaints about
these problems: anxiety, guilt, sporadic pain in the

penis, difficulty in dealing with a new girlfriend, and

resentment of his parents. In growing up he never felt

close to his father but was very close to his mother;

he felt the need to comfort and take care of her. In his
relationships he recurrently felt he could not achieve
the closeness and responsiveness from others that he

needed.

He was expected to have a good outcome with the

help of psychotherapy. He seemed to relate well to

doctors; he was warm and open. He appeared to be

well motivated and able to learn. Also, his conflicts
were seen as perhaps no more than an exacerbation of
normal adolescent conflicts, chiefly intense guilt over
sex.

On the other hand, his thoughts about wanting to

be like an exalted spiritual leader were somewhat

www.theipi.org 223
www.freepsychotherapybooks.org
 confused: “I’d like to have what he had, without the

preliminary steps.” By “what he had,” he may have

meant the leader’s spiritual power. This statement
may simply be an expression of the patient’s desire to

be a great person, or it may reflect significant

disturbance.

iw-1.1 It was not clear whether there was

prognostic significance in the fact that his guilt over
sex took the form of a conversion symptom— pain in
his penis—rather than simply the experience of guilt.

The initial evaluation suggested only that later failure

to improve would indicate either more significant

disturbance, or that his underlying guilt was too

strong, particularly his guilt-inducing attachment to

his mother, or that both of these were true.

Termination (After About 90 Sessions)

The initial 2 months of treatment revealed that

Mr. Howard had difficulties in knowing how to be

www.theipi.org 224
www.freepsychotherapybooks.org
 assertive, be less passive, establish an identity,

become separate from his family, and relate better to

his peers, especially women. Although there was no

evidence of an active thought disorder, before therapy
the patient described “panicky feelings” about not

being able to keep himself “in control” and the sense

that he was so weak that he would need to be a

spiritual leader in order to obtain any of his desires.

This was part of the concern about the possibility of

psychotic deterioration.

After his return from a holiday visit with his

parents, the patient appeared to be remarkably more

stable. He had not used the medication that was

offered because he felt that it was not necessary.

During the spring he examined his fear of closeness

in relationships; he also continued a relationship with

a freshman and experienced his first sexual

intercourse with her. Through his relationship with

www.theipi.org 225
www.freepsychotherapybooks.org
 his new girlfriend and with his therapist, he began to
reexperience many of his oedipal conflicts.

During the summer he began more actively to

examine his ambivalence in relationships. He

experienced increased anxiety and described events

that suggested ideas of reference. However, when he

visited his parents during the summer, he found

himself able to respond in a much more satisfactory

and assertive fashion than previously and felt

encouraged by that.

In the fall the patient returned to school; he and

his girlfriend made arrangements to live together. As
the anticipated problems developed, the patient began

to be able to evaluate the transference relationship in

the triangle created between him, his girlfriend, and

his therapist. He seemed to understand more clearly

the mechanism of his “need to be better” and found

himself better able to maintain relationships.

www.theipi.org 226
www.freepsychotherapybooks.org
 He responded to the therapist’s departure and the

impending separation with a reawakening of his

earlier feelings of needing to be all-powerful in order
to survive and to satisfy his desires. But by the time

of the final appointment, the patient had become

better able to understand much of his current
difficulty in terms of his transference and in the light
of the way he had learned to respond to his situation

at home during his earlier years. He entertained

(ambivalently) the idea of marriage to his current

girlfriend and made plans for his education and

training.

CCRT Data and Results

The presentation of illustrations of the data and

results for Mr. Howard include in this sequence:

Early: The relationship episodes from Session 3,

the score sheets for each relationship episode by
tailor-made categories (see Table 4), the summary

www.theipi.org 227
www.freepsychotherapybooks.org
 score sheet for all relationship episodes by tailor-

made categories (see Table 5), and the summary score

sheet for all relationship episodes by standard
categories (see Table 6).

Late: The same types of data are presented for the

late period of treatment, including four relationship
episodes (plus three dreams) selected from Sessions
82 and 83 and the score sheets (see Tables 7, 8, and
9).

Early CCRT: The CCRT for the six relationship

episodes early in treatment in the illustration here is

from Table 5 (session 3) based on the tailor-made
system (The number that follows each component is

the frequency of the component among the six

relationship episodes): I want to be close and receive

affection (3) (and not experience the loss of
relationships [2]). But the other person rejects my
wish (3). I respond by feeling resentment (2) and self-

www.theipi.org 228
www.freepsychotherapybooks.org
 blame (2). A similar pattern appears in Table 6 using
more exact standard categories.

Late: The CCRT for the four relationship

episodes late in treatment (Tables 7, 8, 9) was the

following: I want to be close (3) and to share with

them (2). But the other person is stronger (4) and

shows behavior that justifies distrust and does not
give what I want. I am self-critical (2) and distrustful
of others (2). Note that most of the pattern is
recognizably similar late in treatment to what it was

early in treatment. However, the patient shows

greater signs of recognition of his own pattern, for

example, in RE 3: “I realize I distrust everybody.”

The scores from the dreams are not counted with the
RE scores, but we see that some of their CCRT
components overlap with the most frequent

components from the relationship episodes. (The

overlap of CCRTs for dreams and narratives is shown

more systematically in chapter 12, this volume.)

www.theipi.org 229
www.freepsychotherapybooks.org
 Mr. Howard: CCRT Scoring of Relationship
Episodes (Condensed)
Session 3

RE 1: Mother

This might have been a dream. Mother says it

didn’t happen. Up until we moved, when I had
questions about sex, mother would explain.
when we moved to _____, one day I asked and
she said, sorry we can’t talk bout that. You’re
getting to that age. Bothered me ‘cause my
young sister, age 9 or 10, laughed.

//This might have been a dream.

✔ NRO: Disagrees //RO Mother says it didn’t happen.
with his view (8,
14)
✔ W: To get sexual //W(W) Up until we moved, when I had
information (8?, 11 questions about sex,
?)
✔ (W): To get
close to M. (11?, 8)
✔ PRO: (past) //RO mother would explain.// When we
Explains (11, 13) moved to _____, one day I asked and
✔ NRO: Rejects //RO she said, sorry we can't talk about
(4, 12) that. You’re getting to that age.
✔ NRS: //RS Bothered me ’cause my young
Frustration (21, sister, age 9 or 10, laughed.//
20?)
NRS: Shame (26,
25)

RE 2: Mother

www.theipi.org 230
www.freepsychotherapybooks.org
 Another incident mother said never happened.
We, brother and I before sister was born … when
it was really cold, would sleep with parents.
Parents took my brother in bed with them and
they wouldn’t take me.

✔ NRO: Disagrees //RO Another incident mother said never

with his view (8, happened.
14)
✔ PRO: (past) //RO(W) We, brother and I before sister
Closeness (11?, was born … when it was really cold,
13?) would sleep with parents.
✔ (W) To be
physically close
(11,8?)
✔ NRO: Rejection, //RO Parents took my brother in bed with
choose someone them and they wouldn’t take me.
else (4, 12)

RE 3: Therapist

T: What’s happening now?

P: I feel generally unresponsive. I’m getting a

headache, tense, thinking all week about relating
all this stuff to what I was 10 years ago [sigh]
and not getting any—I mean, nothing comes out
… like groups of guys who have embarrassing
moments of silence. It proves no perfect rapport
exists. I feel blank.

T: What’s happening now?

www.theipi.org 231
www.freepsychotherapybooks.org
✔ NRS: P: //RS I feel generally unresponsive.
Unresponsive,
distant (8,16).
✔ NRS: //RS I'm getting a headache,
Headache(31?)
✔ NRS: Tense (27, //RS tense,
19?)
NRS: Lack of //RS thinking all week about relating all
response on his this stuff to what I was 10 years ago
part (8, 16?) [sigh] and not getting any—I mean,
nothing comes out … like groups of
guys who have embarrassing moments
of silence.
✔ NRO: No //RO It proves no perfect rapport exists.
rapport (12, 14?).
NRS: Lack of //RS I feel blank.
response on his
part.(8, 16?)

RE 4: Mother

Before I went to school, I always used to kiss my

mother. I’m not sure it was a big thing, but it was
a big thing when it stopped. She made a big
thing about how I didn’t want to kiss her
anymore. I was suddenly out in the cold again.

✔ (W): To be //(W)RS Before I went to school, I always

close, have used to kiss my mother.// I’m not sure it
affection (11,7) was a big thing, but it was a big thing
✔ PRS: (past) when it stopped.
Closeness,
affection (30, 5)
✔ NRO: Blames //RO She made a big thing about how I
(4, 27) didn’t want to kiss her anymore.

www.theipi.org 232
www.freepsychotherapybooks.org
✔ NRS: Felt alone //RS I was suddenly out in the cold
(23, 20?) again.

RE 5: Girlfriend

I’m beginning to feel a lot of resentment to E

[girlfriend]. I went with her for a couple of
years. It’s just been severed. I’m fearful of
seeing her and feeling something for her. She
just doesn’t give a damn. Bothers me I used to
be so screwed up about her.

✔ NRS: //RS I’m beginning to feel a lot of

Resentment (21, resentment to E [girlfriend].
20?)
✔ (W): To not //(W)I went with her for a couple of
suffer loss of rel. years.
(11?, 33?)
✔ NRO: Cuts off //RO It's just been severed.
rel. (4, 14?)
✔ NRS: Fear of //RS I'm fearful of seeing her and feeling
wish for something for her.
attachment (19?,
27?)
NRO: Rejects (4?, //RO She just doesn’t give a damn.
15)
✔ NRS: Blames //RS Bothers me I used to be so
self (25, 17?) screwed up about her.//

RE 6: Mother

One thing that started my resentment against my

parents. I told her about E [girlfriend] that

www.theipi.org 233
www.freepsychotherapybooks.org
 everything was cut off. I said E’s not writing and
it upset me. She said, “Well, I’m sure about you
and you aren’t sure about her.” That really cut
me up because of the rel … relationship she …
she … a … assumes between us is like between
E and me.

✔ NRS: //RS One thing that started my

Resentment (21, resentment against my parents.
11?)
✔ (W): To not be //(W) I told her about E [girlfriend] that
cut off from everything was cut off.// I said E’s not
girlfriend (11, 14?) writing and it upset me.
✔ NRO: Hurts P. //RO She said, “Well, I’m sure about you
(15, 4) and you aren’t sure about her.”// That
really cut me up
✔ NRO: Assumes //RO because of the rel … relationship
P. loves girl like his she … she … a … assumes between us
mother—hurtful is like between E and me.//
idea (2, 15)

Mr. Howard: CCRT Scoring of Relationship

Episodes (Condensed)
Sessions 82 and 83

RE 1: Therapist (Session 82)

Things going well but I feel I have to give up my

girlfriend. I resent you because I have to give up
things to get close to you.

✔ (W): To be close //(W) Things going well but I feel I have

www.theipi.org 234
www.freepsychotherapybooks.org
(11, 17?) to give up my girlfriend.
✔ NRS: //RS I resent you
Resentment (21,
?)
✔ NRO: Forces //RO because I have to give up things to
me to give up get close to you.//
girlfriend
STRONGER,
MORE
POWERFUL
(24,17?)

Dream A: Trainers3

War. I enlisted as soldier on our side. The

trainers of the soldiers had superior strength.
You [therapist] will smile. I only got into the
entertainment troops.

✔ (W): To be close //(W) War. I enlisted as soldier on our

to trainer side.
(therapist?) (11?,
27?)
✔ PRO: Stronger //RO The trainers of the soldiers had
(24, ?) superior strength.
✔ NRO: Rejects //RO You [therapist] will smile.
(4, 17?)
✔ NRS: Not good //RS I only got into the entertainment
enough; SELF- troops.//
CRITICAL (17, 20)

RE 2: Boyfriend

www.theipi.org 235
www.freepsychotherapybooks.org
 In preschool, I wanted to trust someone. I
confided in [boyfriend], but he could screw me.

✔ W: To trust, to //W In preschool, I wanted to trust

share (6, 8) someone.
✔ PRS: Am open, //RS I confided in [boyfriend]
trusting (7, 5?)
✔ NRO [expected]: //RO but he could screw me.
Could screw me.
(8,15)

Dream B: Store (Session 83)

I was going into the candy store with other kids.

The place is floating in ice cream. I got sick and
repelled by it.

✔ (W): To be fed //(W) I was going into the candy store

(13, 11) with other kids.// The place is floating in
ice cream.
✔ NRS: Sick (31, //RS I got sick and repelled by it.
26?)

Dream C: Store Owner

I was going into a store. I was naked. Two young

ladies said, “Tsk, tsk.” I had to escape from the
proprietor.

✔ (W): To buy //(W) I was going into a store.

something (13?, ?)
✔ (W): To expose //(W) I was naked.

www.theipi.org 236
www.freepsychotherapybooks.org
self (33?, ?)
✔ NRO: Shamed //RO Two young ladies said, “Tsk, tsk.”
by ladies (4, 17)
✔ NRO: Pursued //RO I had to escape from the
by man (15, 20?) proprietor.//

RE 3: Therapist

I felt bad because of distrusting our relationship.

I saw an article that makes me distrust you. I
realize that I distrust everybody.

✔ (W): To have //(W)RS I felt bad because of distrusting

trusting our relationship.
relationship (6,3)
✔ NRS: Self-
blame (25, 26)
✔ NRO: //RO I saw an article that makes me
Untrustworthy (8, distrust you.
17?)
✔ NRS: Distrust //RS I realize that I distrust everybody.//
(6?, 19?)

RE 4: Father

I wrote to Father for money. It is not all right

with him to send me it. I lose trust.

✔ W: To get //W I wrote to Father for money.

money (13?, 8?)
✔ NRO: Does not //RO It is not all right with him to send
give me (14, 4) me it.
✔ NRS: Distrust //RS I lose trust.//
(20?, ?)

www.theipi.org 237
www.freepsychotherapybooks.org
 TABLE 4
CCRT Score Sheet for Each Relationship Episode, Tailor-Made
System

Patient: Mr. Howard Date: 1/26/89

Session 3 Judge: LL
RE No. Wish, Need, Response Response of
Person: Intention From Others Self
RE 1 Mother To get sexual Disagrees with Frustration
information my view
To get close PRO (past) Shame/SELF-
to Mother Explains BLAME*
Rejects
RE 2 Mother To be Disagrees with
physically my view
close
PRO (past)
Closeness
Rejection
Chooses other
instead of me
RE 3 No rapport Unresponsive,
Therapist distant
To be close, Headache
to have
rapport
Tense
Lack of
response
RE 4 Mother To be close, Blames PRS (past)
have Closeness,
affection affection

www.theipi.org 238
www.freepsychotherapybooks.org
 Felt alone
RE 5 To not suffer Cuts off Resentment
Girlfriend loss of relationship
relationship
Rejects Fear of wish
for attachment
Self
blame/SHAME
RE 6 Mother To not be cut Hurts me Resentment
off from
girlfriend
Assumes I
love my
mother like a
girl

*Additions from Step 1' are in capital letters.

www.theipi.org 239
www.freepsychotherapybooks.org
 TABLE 5
CCRT Score Sheet, Tailor-Made System: Summary Across All
Relationship Episodes

Patient: Mr. Howard Date: 1/26/89

Session 3 Judge: LL
Number of REs: 6 (In RE number 1, 2, 3, 4, 5, 6)
Wish, Need, Response from Response of
Intention (RE No.) Others (RE No.) Self (RE No.)
Negative Negative
To be close, have Rejects (1, 2, 5) Resentment (5,
affection (1, 2, 4) 6)
To not lose Disagrees with my Self-
relationship, be cut view (1, 2) blame/shame (1,
off (5, 6) 5)
To get sexual Chooses other Frustration (1)
information (1) instead of me (2)
No rapport (3) Unresponsive,
distant (3)
Blames (4) Headache, tense
(3)
Cuts off Lack of support,
relationship (5) help (3)
Hurts me (6) Feel alone (4)
Assumes I love Fear of wish for
mother like a girl attachment (5)
(6)
Positive Positive
(past) (past)
Explains things to Closeness,
me (1) affection (4)
(past)
Closeness (2)

www.theipi.org 240
www.freepsychotherapybooks.org
 TABLE 6
CCRT Summary: Standard Categories

Patient: Mr. Howard

Session 3 Judge: LL
Number of REs: 6 (In RE number 1, 2, 3, 4, 5, 6)
Edition 2 Category Frequency Across REs (first
RE No.
Number: choices only)
Wishes
11: TO BE CLOSE 1, 2, 4, 5
TO OTHERS 5, 6
8: TO BE OPENED 1 1
UP TO
Negative responses
from others
4: ARE REJECTING 1, 2, 4, 4
5
8: ARE NOT 1, 2 2
TRUSTWORTHY
2: ARE NOT 6 1
UNDERSTANDING
12: ARE DISTANT 3 1
15: HURT ME 6 1
Positive responses
from others
11: ARE OPEN 1, 2 2
Negative responses of self
21: FEEL ANGRY 1, 5, 6 3
8: AM NOT OPEN 3 1
19: AM UNCERTAIN 5 1
23: FEEL UNLOVED 4 1

www.theipi.org 241
www.freepsychotherapybooks.org
25: FEEL GUILTY 5 1
26: FEEL ASHAMED 1 1
27: FEEL ANXIOUS 3 1
31: SOMATIC 3 1
SYMPTOMS
Positive responses of self
30: FEEL LOVED 4 1

www.theipi.org 242
www.freepsychotherapybooks.org
 TABLE 7
CCRT Score Sheet for Each Relationship Episode, Tailor-Made
System

Patient: Mr. Howard Date:

1/26/89
Sessions 82 and 83 Judge: LL
RE No. Wish, Need, Response Response
Person: Intention From Others of Self
RE 1 To be close T forces me to I must give
Therapist give up up girlfriend
girlfriend
Resentment
Dream A To be with you, on Stronger Not good
Trainers your side; TO BE enough,
CLOSE rejected
Rejects
RE 2 To trust, to share Could screw me Am open,
Friend To be close (expected) trusting
Dream B To be fed Too much ice Overfed
Store cream
Sick
Repelled
Dream C To buy something; Shamed by
Store TO SPEND ladies
Owner MONEY
To expose self Pursued by
sexually to woman man

RE 3 To have a trusting Shows behavior Distrust

Therapist relationship; TO that justifies
SHARE distrust
Self-blame

www.theipi.org 243
www.freepsychotherapybooks.org
RE 4 To get money Does not give Distrust
Father money

www.theipi.org 244
www.freepsychotherapybooks.org
 TABLE 8
CCRT Score Sheet for Each Relationship Episode, Tailor-Made
System: Summary Across All Relationship Episodes

Patient: Mr. Howard Date: 1/26/89

Sessions 82 and 83 Judge: LL
Number of REs: 4
Number of Dreams: 3 (In RE number 1, 2, 3, 4)
Wish, Need, Response from
Response of Self
Intention Others
(RE No.)
(RE No.) (RE No.)
Negative Negative
To be close (1, 2, Stronger, more Distrust (3, 4)
A) powerful (1, A*, 2, C)
To get things (B, Shows behavior that Self-critical and
C, 4) justifies distrust (3) self-blame (3, A)
To trust, to share Does not give me I must give up
(2, 3) what I want (4) what I want (1)
To expose self Too much/excessive Overfed (B)(C)
sexually to (B)
woman
Shames me (C) Sick (B)
Rejects me (A) Repelled (B)
Resentment (1)
Positive
Am open, trusting
(2)

*Capital letters refer to the dreams.

www.theipi.org 245
www.freepsychotherapybooks.org
 TABLE 9
CCRT Summary: Standard Categories

Patient: Mr. Howard

Sessions 82 and 83 Judge: LL
Number of REs: 7 (In RE number 1, A, 2, B, C, 3, 4)
Frequency
Edition 2 Category Across REs
RE No.
Number: (first choices
only)
Wishes
13: TO BE HELPED B*, C, 4 3
6: TO HAVE TRUST 2, 3 2
11: TO BE CLOSE TO 1, A 2
OTHERS
33: TO BE LOVED C 1
Negative responses from others
4: ARE REJECTING A, C 2
8: ARE NOT 2, 3 2
TRUSTWORTHY
14: ARE UNHELPFUL 4 1
15: HURT ME C 1
24: ARE INDEPENDENT 1 1
Positive responses from others
24: ARE INDEPENDENT A 1
Negative responses of self
6: DISLIKE OTHERS 3 1
17: AM HELPLESS A 1
20: FEEL 4 1
DISAPPOINTED
21: FEEL ANGRY 1 1

www.theipi.org 246
www.freepsychotherapybooks.org
25: FEEL GUILTY 3 1
31: SOMATIC B 1
SYMPTOMS
Positive responses of self
7: AM OPEN 2 1

*Capital letters refer to the dreams.

www.theipi.org 247
www.freepsychotherapybooks.org
MS. CATHY CUNNINGHAM: PSYCHOANALYSIS

The clinical sketches for Ms. Cunningham before

and after psychoanalysis draw largely on the material

by Weiss, Sampson, and the Mount Zion

Psychotherapy Research Group (1986); the CCRT
analyses draw on our own examination of the
sessions.

Clinical Evaluation

Initial

Ms. Cunningham was an attractive 30-year-old

social worker in a Catholic agency. She had been

married for 4 years to a successful businessman. Her

main symptom was sexual inhibition. She did not

enjoy sex; she did not have orgasms and held back

from having intercourse. Her inability to relax and

enjoy herself affected other aspects of her life as well.
The second symptom was her self-criticalness. The
third was her passivity; she was afraid of “simply

www.theipi.org 248
www.freepsychotherapybooks.org
 being a nonentity” and wished to be able to be
assertive.

Her father was a successful businessman, and her

mother was a housewife. The patient was the second

of four children. She had an older sister, a younger

sister, and a younger brother. The father tyrannized

the mother with his occasional fits of temper, and the
mother took his abuse passively. The parents
displayed little affection toward each other and
appeared to be joyless and puritanical.

In the following episode, which occurred when

the patient was 6, the parents showed their usual style

of relating to her. The patient’s sister had hit her in

the stomach. She had gone weeping and complaining
to her mother. When she could not get her mother’s
attention, she became frustrated and hit her mother in

the stomach. Her mother made no attempt to defend

herself. She doubled over in pain, wept, and went to

www.theipi.org 249
www.freepsychotherapybooks.org
 her room. When the father heard what the daughter

had done, he became enraged; he beat her and threw

her into a closet. The girl was horrified at the father’s
loss of control, and she believed for a short time her
father had wanted to kill her. Not long after, she

became enraged at her younger brother and wanted to

kill him.

It is of interest to note the formulations about

central relationship conflicts in the early sessions that
were done independently and with different emphases

(from each other and from ours) by two research

groups (Weiss et al., 1986):

By a clinical group. Ms. Cunningham’s

difficulties came after the birth of her brother when

she was 6. She felt her parents, especially her father,

valued her brother more than her. She assumed this

was because the younger brother had a penis and she

did not. Therefore, her primary unconscious wish was

www.theipi.org 250
www.freepsychotherapybooks.org
 shown in envy of men and by aggressively

withholding admiration and sexual response or

criticizing and attacking them.

By a Mt. Zion research group. Ms. Cunningham’s

problems came not from unconscious envy but from

unconscious guilt. She saw her parents as fragile and
vulnerable. She believed the parents would be
damaged if she held ideas or values different from
theirs or disagreed with them or led an independent

life that was less burdened and less joyless than

theirs. She unconsciously felt superior to them. She
protected herself from hurting them by making

herself weak and helpless. (Evidence from the Mt.

Zion research group’s central relation pattern

measure, the plan diagnosis sketched in chapter 17,

this volume, favored this formulation.)

Termination (After About 1,300 Sessions)

www.theipi.org 251
www.freepsychotherapybooks.org
 The initial symptoms were markedly improved by

the end of treatment. In fact, they had gradually

improved throughout the period of treatment.

CCRT Data and Results

We present here a sample of the relationship

episodes (condensed) for Ms. Cunningham for an
early session (Session 5) and for a late session

(Session 1,028). Session results are provided to

illustrate the degree of agreement between

independent judges that can be achieved for the

relationship episodes for Session 5 by tailor-made
scores (see Tables 10 & 11); Table 12 does the same

by standard categories. The two judges show

moderate agreement, especially for the most frequent

wishes, responses from others, and responses of self.

As we explained in chapter 2, this volume, the tailor-
made system is harder to cope with for showing
reliability, even though the tailor-made categories
more specifically fit the particular patient.

www.theipi.org 252
www.freepsychotherapybooks.org
 We used the scored case illustration with five

relationship episodes from Session 5; one can see in

the tailor-made CCRT summary score sheet (Table
11) that the CCRT is as follows: I wish to be in
control (2). I also wish for reassurance (2). But the

other person doesn’t give what I want (2). I respond

to becoming angry (3) and I show poor control of

myself (undercontrol) (2). Table 12 by the more exact

standard categories is similar. (This formulation is not

the same as either of the two groups’ formulations

given by Weiss et al., 1986; however, it is more
congruent with the Mt. Zion research group’s because

of its emphasis on conflicts about her wish to become

more independent.)

For the late session (1,028) with six relationship

episodes, the CCRT is similar but somewhat changed

(see Table 13 & 14): I wish to dominate the other

person and get my way (3) and to control myself (2).

I also wish to be close and feel loving. The other

www.theipi.org 253
www.freepsychotherapybooks.org
 person becomes angered. I am inclined to withdraw.
The more exact standard category summary (Table

15) is especially similar in the wish to control

(dominate) others.

In the late session there is a somewhat greater

inclination to be close and feel loving (as is also

pointed out by Weiss et al., 1986). Otherwise the

wishes show considerable consistency from the early

to the late session, even in this very long therapy (a

finding that is confirmed by the studies of

pervasiveness discussed in chapter 9, this volume).

However, Ms. Cunningham is more able to achieve a

positive response from others; her anger is less
pervasive; and her undercontrol is less evident.

Ms. Cunningham: CCRT Scoring of Relationship

Episodes (Condensed)
Session 5

RE 1: Assistant

I am annoyed by her. I don’t want to share her

because she might prefer the other supervisor to

www.theipi.org 254
www.freepsychotherapybooks.org
 me.

✔ NRS: Annoyed //RS I am annoyed by her.

(21, 6)
✔ W: Want her for //W I don’t want to share her
myself (35, 19)
✔ NRO: Might //RO because she might prefer the other
prefer other (6?, supervisor to me.//
4?)

RE 3: Husband

I got home in a funny mood about what I’d done

with the boys (in RE 4). I wanted his approval or
direction, he didn’t, so I go furious.

✔ NRS: Funny //RS I got home in a funny mood about

mood (17?, 19?) what I’d done with the boys (in RE 4).
✔ W: Wanted //W I wanted his approval or direction.
approval (2, 20)
✔ NRO: No //RO he didn’t
approval (2, 6?)
✔ NRS: Anger (21, //RS so I got furious.
20?)

RE 4: Boys and parents

I was upset. Two boys in the group stick together

rather than mix. I’m trying to separate them. I’m
annoyed. I spoke to one parent—I hadn’t
intended to—but then I couldn’t stop. I tried to
control myself.

www.theipi.org 255
www.freepsychotherapybooks.org
✔ NRS: Upset (19, //RS I was upset.
13)
✔ NRO: Not pay //RO Two boys in the group stick
attention to her together rather than mix.
(25?, 17)
✔ (W): To get //(W) I’m trying to separate them.
control of them
(19, 22?)
✔ NRS: Annoyed //RS I’m annoyed.5
(21, 20?)
✔ NRS: Under //RS I spoke to one parent—I hadn’t
control (13, 17) intended to—but then I couldn’t stop.// I
tried to control myself.//

RE 5: Therapist

T: You didn’t express wanting reassurance here

yesterday.

P: I won’t get reassurance here. I want to be

reassured you’re listening.

T: You didn’t express wanting

reassurance here yesterday.
✔ W: To get P: //W RO I won’t get reassurance here.
reassurance (2,
13)
✔ NRO: Won't be
given (14, 6)
W: Wish for //W I want to be reassured you’re
reassurance (2, listening.//
13)

RE 7: Professor

www.theipi.org 256
www.freepsychotherapybooks.org
 I noticed his tie was nubby, coarse-woven. I
reached out and held it and said, “this is a
wonderful texture.” I was horrified that I had
done this.

✔ (W): To admire //I noticed his tie was nubby, coarse-

what man has woven.//(W) I reached out and held it
(27?, 22?) and said “this is a wonderful texture.”
(W): Wish to have
what the man has
(27?, 19?)
✔ (W): TO HAVE //(W)RS I was horrified that I had done
SELF-CONTROL this.
(19,34)
✔ NRS:
Embarrassed (26,
25)
✔ NRS:
Undercontrol (13,
17)

Ms. Cunningham: CCRT Scoring of Relationship

Episodes (Condensed)
Session 1,208

RE 1: Therapist

I was thinking I should mention it first thing. I

don’t recall getting a bill last month.

✔ W: To rectify not //W(W) I was thinking I should mention it

getting bill (23?, first thing. I don’t recall getting a bill last
26?) month.//
✔ (W):
INCREASED

www.theipi.org 257
www.freepsychotherapybooks.org
CONTROL (19?,
21?)

RE 2: Self

I must have put weight on near the end of last

week. I really felt fat and ugly today.

✔ (W): To lose //(W) I must have put weight on near the

weight (21, 25?) end of last week.
✔ NRS: Feel fat //RS I really felt fat and ugly today.//
and ugly (26, 13?)

RE 3: Assistant

She was depressed. It can’t work out with the

guy she’s dating and she can’t close it off. I was
saying things to make her feel more confident
and she said, “Oh you’ve helped me so much”
and “I love you.” I could feel myself withdraw.

✔ NRO: //RO She was depressed.// It can’t work

Depressed (28?, out with the guy she’s dating// and she
?) can’t close it off.
✔ W: To make her //W(W) I was saying things to make her
more confident feel more confident// and she said “Oh,
(12, 7?) you’ve helped me so much”
(W): To be
therapeutic like T
(12, 22?)
✔ PRO: Loves me //RO and “I love you.”
(30?, 9)
✔ NRS: Withdraw //RS I could feel myself withdraw.

www.theipi.org 258
www.freepsychotherapybooks.org
(8, 19)

RE 4: Husband

It’s like when husband wanted to make love. I

would first have to get him angry at me, then I
would feel loving.

✔ PRO: Wish to //RO It’s like when husband wanted to

make love (30. make love.
11?)
✔ W: Make him //W I would first have to get him angry at
angry (16, 18) me
✔ (W): Dominate,
have control over
other (19,28?)
✔ PRS: Feel //RS then I would feel loving.//
loving (5?, 29?)

RE 5: Self

I was putting recipes in order. This time my

response was different. Before I went through
whatever order the pile of recipes was in. Now
it’s ones I’ve done, favorites. I feel tense. Will I
be able to throw away what I won’t use?

✔ PRS: Feel better //I was putting recipes in order.//RS This

(29?, 15?) time my response was different.// Before
I went through whatever order the pile of
recipes was in.
✔ W: To cook what //W Now it’s ones I’ve done, favorites.
I like (25?, 29?)

www.theipi.org 259
www.freepsychotherapybooks.org
 TABLE 10
CCRT Score Sheet for Each Relationship Episode, Tailor-Made
System

Patient: Ms. Cunningham Date:

1/26/89
Session 5 Judge: LL
RE No. Wish, Need, Response Response of
Person Intention From Others Self
RE 1 To have the Prefers the Annoyed
Asst. assistant to myself; other person
TAKE WHAT I
WANT*
RE 3 To get approval and Doesn’t give Angry
Husband direction; WANT what I want
REASSURANCE
Funny mood
RE 4 To get students to Stay Undercontrol
Boys obey; to dominate; together, go
be in control against my
wish
DOESN’T Annoyed
GIVE WHAT
I WANT
Upset
RE 5 Want reassurance Not
Therapist reassuring
RE 7 To assert my Undercontrol
Professor impulse to admire
the man
TO HAVE SELF- Embarrassed
CONTROL
To have what man
has

*Additions from Step 1' are in capital letters. 260

www.theipi.org
www.freepsychotherapybooks.org
 TABLE 11
CCRT Score Sheet, Tailor-Made System: Summary Across All
Relationship Episodes

Patient: Ms. Cunningham Date: 1/26/89

Session 5 Judge: LL
Number of REs: 5 (In RE number 1, 3, 4, 5, 7)
Wish, Need, Intention Response from Response of
(RE No.) Other (RE No.) Self (RE No.)
Negative Negative
To be in control of self Doesn’t give what Angry,
and others (4, 7) I want (3, 4) annoyed (1, 3,
4)
To take what I want (1) Prefers the other Undercontrol
person (1) (4, 7)
Want reassurance (3, 5) Not reassuring (5)
To assert my impulse to Embarrassed
admire the man (7) (7)
To have what man has Funny mood
(7) (3)
Upset (4)

www.theipi.org 261
www.freepsychotherapybooks.org
 TABLE 12
CCRT Summary: Standard Categories

Patient: Ms. Cunningham

Session 5 Judge: LL
Number of REs: 5 (In RE number 1, 3, 4, 5, 7)
Frequency
Across
REs
Edition 2 Category Number: RE No.
(first
choices
only)
Wishes
2: TO BE ACCEPTED 3, 5 2
19: TO HAVE CONTROL OVER 4, 7 2
OTHERS
27: TO BE LIKE OTHERS 7 1
35: TO COMPETE WITH 1 1
SOMEONE FOR ANOTHER
PERSON’S AFFECTION
Negative responses from others
2: ARE NOT UNDERSTANDING 3 1
6: DON'T RESPECT ME 1 1
14: ARE UNHELPFUL 5 1
25: ARE BAD 4 1
Negative responses of self
21: FEEL ANGRY 1, 3. 4 3
13: AM OUT OF CONTROL 4, 7 2
17: AM HELPLESS 3 1
19: AM UNCERTAIN 4 1
26: FEEL ASHAMED 7 1

www.theipi.org 262
www.freepsychotherapybooks.org
 TABLE 13
CCRT Score Sheet for Each Relationship Episode, Tailor-Made
System

Patient: Ms. Cunningham Date:

12/26/89
Session 1,028 Judge: LL
RE No. Response Response
Wish, Need, Intention
Person From Others of Self
RE 1 To rectify not getting bill
INCREASE CONTROL
RE 2 To lose weight Feel fat
Self and ugly
To make myself be in
control of weight
RE 3 To make the other more Loving
Asst. confident feeling (PRO
Depressed
(NRO)
To be assertively Drawing
therapeutic like T. back
To be close
RE 4 To make the other angry Wants to
make love
To dominate the other Angry Feel
loving
(PRS)
RE 5 To arrange things so I can Feel
Self cook what I want most better
(PRS)
Feel
tense
To establish control over

www.theipi.org 263
www.freepsychotherapybooks.org
 self
RE 6 To insist on getting my way
and have a baby by T.
To have father
To have my way

www.theipi.org 264
www.freepsychotherapybooks.org
 TABLE 14
CCRT Score Sheet, Tailor-Made System: Summary Across All
Relationship Episodes

Patient: Ms. Cunningham Date:

1/26/89
Session 1,028 Judge: LL
Number of REs: 6
Response Response
Wish, Need, Intention (RE No.) From Others of Self
(RE No.) (RE No.)
Negative Negative
To dominate the other; get my Angry (4) Withdrawal
way (1, 4, 6) Depressed (3)
(1)
To control myself and others (2.
5)
To be close/feel loving (3) Feel tense
(5)
To rectify not getting bill (1) Feel fat and
ugly (2)
To lose weight (2)
To make other more confident
(3)
To be assertively therapeutic
like T. (3)
To make other angry (4)
To arrange things so I can cook
what I want most (5)
To have father (6)
To have baby by T. (6)
Positive Positive

www.theipi.org 265
www.freepsychotherapybooks.org
 Loving feeling
(3)
Wants to Feel loving
make love (4) (4)
Feel better
(5)

www.theipi.org 266
www.freepsychotherapybooks.org
 TABLE 15
CCRT Summary: Standard Categories

Patient: Ms. Cunningham

Session 1028 Judge: LL
Number of REs: 6 (In RE number 1, 2, 3, 4, 5, 6)
Frequency Across
Edition 2 Category
RE No. REs (first choices
Number
only)*
Wishes
19: TO HAVE 1, 4, 6 3
CONTROL OVER
OTHERS
21: TO HAVE SELF- 2, 5 2
CONTROL
12: TO HELP 3 1
OTHERS
16: TO HURT 4 1
OTHERS
23: TO BE 1 1
INDEPENDENT
25: TO BETTER 5 1
MYSELF
Negative responses from others
28: ARE ANXIOUS 3 1
Positive responses from others
30: LOVES ME 3, 4 2
Negative responses of self
8: AM NOT OPE•N 3 1
26: FEEL 2 1
ASHAMED
27: FEEL ANXIOUS 5 1

www.theipi.org 267
www.freepsychotherapybooks.org
Positive responses of self
5: LIKE OTHERS 4 1
29: FEEL HAPPY 5 1

*In this tabulation, only the first choice of a standard category is listed.

www.theipi.org 268
www.freepsychotherapybooks.org
 ✔ NRS: Feel tense //RS I feel tense.
(27, 19?)
✔ (W): To // (W) Will I be able to throw away what I
establish control won’t use?//
over self (21, 22)

RE 6: Therapist

T: You want to pregnant and you want your

father’s child or my child and nothing else will
do.

P: I just discussed something like that with the

assistant about “one element that’s missing” with
husband. I’m holding on to my wish with father.

T: You want to be pregnant and you

want your father’s child or my child and
nothing else will do.
✔ (W): To have my P: (W) I just discussed something like
way (19, 34?), that with the assistant about “one
element that’s missing” with husband.
W: To have my //W (W) I’m holding on to my wish with
wish from father father.//
(19?).
(W): To have my
wish from therapist
(19?, ?)

CONCLUSIONS

• This chapter provided illustrations to accompany

the guide to scoring in chapter 2; these should

www.theipi.org 269
www.freepsychotherapybooks.org
 help in learning the CCRT method.

• Early and late CCRTs were scored for three cases

in three kinds of psychotherapies: brief time-
limited psychotherapy, middle-range open-ended
psychoanalytic psychotherapy, and long-term
psychoanalysis. The CCRT method was suitable
and equally applicable to each of these different
therapies.

• Within each therapy, the early versus late

comparisons showed consistency of the CCRT,
even over the long time period of the
psychoanalysis, as well as some changes related
to the patient’s improvement. These trends are
discussed in more detail in later chapters.

Note

[3] Dreams are included here to illustrate their consistency with

relationship episodes (see chapter 12, this volume).

www.theipi.org 270
www.freepsychotherapybooks.org
 6

THE RELIABILITY OF THE CCRT

MEASURE: RESULTS FROM EIGHT
SAMPLES

LESTER LUBORSKY AND LOUIS DIGUER4

It is time for an updated summary of reliability

studies of the Core Conflictual Relationship Theme

method. It is essential that the many CCRT

researchers and educators know the degree to which

the judges can agree. Since the first edition of this
book (Luborsky & Crits-Christoph, 1990), 150

ongoing studies and 55 published articles have been

included in the partial list in the CCRT Newsletter

(Luborsky, Kächele, & Dahlbender, 1997). The eight
samples for which usable and available reliability
studies are reported in terms of weighted kappas are

reviewed here in terms of the level of agreement of

www.theipi.org 271
www.freepsychotherapybooks.org
 judges on the selection of the narratives and the
agreement of independent judges on the CCRT

scoring.

AGREEMENT ON THE SELECTION OF

RELATIONSHIP EPISODES FROM SESSIONS

The RE judge’s first task before scoring the

CCRT is to locate the relationship episodes that

patients commonly tell during psychotherapy sessions

and to identify the main other person within each

episode. These judges also rate each relationship

episode for completeness on a 1-to-5 scale. Only

relationship episodes with a mean rating by two

judges of at least 2.5 are used for CCRT scoring. Ten
relationship episodes for each patient are the usual
basis for scoring the CCRT, and these 10 are usually

all the relationship episodes in the two early sessions:

Sessions 3 and 5. If 10 relationship episodes are not

found in the two sessions, a third session is used;

usually it is Session 4 or 6.

www.theipi.org 272
www.freepsychotherapybooks.org
 Reliabilities are assessed on three basic types of
agreement for each narration about each relationship

episode: (a) the completeness of each episode on a 1-

to-5 scale, (b) the choice of the main other person in

each episode, and (c) the location of the episode in

the session by noting its beginning and end point. The
results for these three aspects are reported here briefly
but in more detail by Crits-Christoph, Luborsky,

Dahl, Popp, Mellon, and Mark (1988) and Crits-

Christoph, Luborsky, Popp, Mellon, and Mark (1990)

on the basis of a sample of 111 relationship episodes

scored by two judges.

Completeness of Each Relationship Episode

The agreement of the judges on completeness was

satisfactory: The pooled-judge agreement in

intraclass correlation was .68 (p < .001); the per-

judge intraclass correlation was .51. However, when

faced with episodes of very low completeness, such
as a brief reference to another person, the judges

www.theipi.org 273
www.freepsychotherapybooks.org
 tended not to give a score. Thus, there was a
significant restriction of range for these completeness

ratings.

Selection of the Main Other Person in Each RE

The choice of the main other person with whom

the patient was interacting in the RE was made with

considerable agreement. This study involved

relationship episodes of acceptable completeness with

a mean judge’s rating of 2.5 or better on the 5-point
scale. Of 80 episodes, 97% had the same other person

identified by both judges, although for 8% of these

cases the judges used different labels but were

referring to the same other person. In only 3% of the

cases was a different other person identified by the

two judges, usually when several other persons were

equally alluded to in the relationship episode.

Location of the Relationship Episode

www.theipi.org 274
www.freepsychotherapybooks.org
 The third type of agreement on the location of the
relationship episode in the transcript was also

moderately good. The complete relationship episode

averaged 47.3 lines of text. The two RE-selection

judges differed by an average of only 4.8 lines at the

beginning of an episode but 7.9 lines at the end of an
episode; it was apparently harder to agree about the
end point of an episode than about the beginning of

an episode. For the beginnings of episodes, the judges

were within 7 lines of each other 85% of the time; for

the end of relationship episodes, the judges were

within 7 lines 70% of the time. Bond et al. (1987)

reported similar findings using a somewhat different

method.

AGREEMENT ON THE CCRT

To begin at the beginning of the stream of

agreement studies, we first note three preliminary

studies; then, in the body of this chapter, we describe

the eight usable samples of patients in which we

www.theipi.org 275
www.freepsychotherapybooks.org
 could compute kappas for the agreement of the CCRT
judges. Most of these studies relied on training

procedures for the CCRT judges that are described at

the end of chapter 2, this volume.

The Earliest Study of CCRT Reliability

The inaugural study of the CCRT (Luborsky,

1976, 1977b) showed moderate interjudge agreement,

but because it was done with only a few patients, the

results are not given here.

Agreement on the CCRT for Matched Versus

Mismatched Cases

A subsequent agreement study (Levine &

Luborsky, 1981) was also based on a small sample:
16 graduate psychology student judges who

individually scored the CCRT for one patient. The

scoring of each of these judges was compared with

the composite scoring of four research judges;

agreement was good, with an average correlation of

www.theipi.org 276
www.freepsychotherapybooks.org
 .88. In addition, agreement was demonstrated by the

ingenious method of “mismatched cases,” that is, a

comparison between cases in which the components

to be compared were drawn from the same case
(showing high agreement) compared with a

comparison in which the components were purposely

drawn from a mismatched case (showing low

agreement).

Agreement on the CCRT for One Patient

Another carefully done study of reliability

(Guitar-Amsterdamer, Stahli, Schneider, & Berger,
1988) used only one patient with a sample of

relationship episodes from four sessions. The method

is a variation on Levine and Luborsky’s (1981) study.

The main result of interest for a review of reliability

is that the weighted kappas for agreement judges in

this study were not as good as they were in the study
by Levine and Luborsky (1981) even though the

www.theipi.org 277
www.freepsychotherapybooks.org
 procedures have much in common in terms of

method.

From this point onward, we review the main

studies constituting nine samples, eight of which

contained kappas, that are summarized in Table 1 as

Samples B through I.

Sample A: Agreement on the CCRT by Three

Independent Judges for 8 Patients

This study (Luborsky, Crits-Christoph, & Mellon,

1986), although also based on a small sample (8
patients), used standard reliability method. Two well-

trained judges compared the similarity of CCRT

formulations of three CCRT judges in terms of

whether their formulations were basically similar or

basically different. The agreement between these two

judges was high (96%): The three CCRT judges

arrived at similar formulations across the 8 cases.

Specifically, on the wish component, the three CCRT
judges had similar formulations 75% of the time (6

www.theipi.org 278
www.freepsychotherapybooks.org
 TABLE 1
Summary of Eight Samples for Reliabilities of CCRT
Components

Type of Reliability
Type of
Sample Relationship Kappa,
Sample %
Episodes Weighted
A. Luborsky 8 adult Therapy W: 100%a
et al. patients sessions RO: 88%
(1986) RS: 88%
B. Crits- 35 adult Therapy W: kw =
Christoph, patients sessions .61
Luborsky, RO: kw =
et al. .70
(1988) RS: kw =
.61
C. Popp et 13 adult Therapy W: kw =
al. (1996) patients sessions .67
RO: kw =
.74
RS: kw =
.75
D. Popp et Dreams W: kw = .58
al. (1996) RO: kw =
.70
RS: kw = .83
E. Barber et 19 adult Therapy W: kw = W: 94%
al. (1995) patients sessions .81 RO: 100%
RO: kw = RS: 88%
.64
RS: kw =
.73
F. Barber et RAPs W: kw = W: 84%
al. (1995) .68 RO: 100%
RO: kw = RS: 89%
.60

www.theipi.org 279
www.freepsychotherapybooks.org
 RS: kw =
.65
G. Luborsky, 18 Storiesb W: kw = W: 70%
Luborsky, normal .33 RO: 100%
et al. children RO: kw = RS: 90%
(1995) at age 3 .69
and 5 RS: kw =
.60
H. Waldinger 40 Peer W: kw =
(1997a) young relationship .37d
adults c
narratives RO: kw =
(20 .69
normals RS: kw =
and 20 .76
patients)
I. Lefebvre 50 adult RAPs W: kw =
et al. patients .71e
(1996) RO: kw =
.71
RS: kw =
.71
Averaged W: kw = %
results: .60 agreement
RO: kw = W: 87%
.68 RO: 97%
RS: kw = RS: 89%
.71

Note. W = wish; RO = response from other; RS = response of self. % =

percentage of agreement; kw = weighted Cohen’s kappa; RAP =
Relationship Anecdote Paradigm.

a Agreement of two judges.

b Narratives that appear to combine accounts of imaginary events
with real events.
c Examples of relationship events, taken from peer relationship
interviews, not from formal RAP interviews.
d Averaged weighted kappas for four judges.
e Averaged weighted kappas for three judges.

www.theipi.org 280
www.freepsychotherapybooks.org
 out of the 8 cases); two of the three judges had
similar formulations 100% of the time. For the

negative responses from other, in 63% of the cases

the three judges arrived at similar formulations,

whereas two of the three judges arrived at similar

formulations 88% of the time. For the negative
responses of self, the three judges came to similar
formulations 38% of the time, whereas two of the

three reached similar formulations 88% of the time.

Sample B: The First Large-Scale Study of Interjudge

Agreement on the CCRT

For examining the reliability of the application of

CCRT scoring to 35 cases, two psychodynamically

oriented clinicians served as CCRT judges (Crits-

Christoph, Luborsky, et al., 1988). Only those
narratives (relationship episodes) were included that
reached or surpassed the minimum acceptable
completeness rating of 2.5 by the mean rating of the
RE-selection judges. We consider this preselection of

www.theipi.org 281
www.freepsychotherapybooks.org
 the relationship episodes to be a reasonable hurdle;

we were interested in applying the CCRT only to

narratives that were at least moderately complete

because the themes might be more easily scored in

more complete narratives.

The agreement procedure on the standard CCRT

categories relied on the independent CCRT judges
and on Cohen’s weighted kappa (Cohen, 1968). The
kappa is defined simply as the proportion of

agreement after chance agreement is removed.

According to Cohen (1968), it is often necessary to
weight the agreement to make it more precise. The

rationale is that certain disagreements are less

important than others and should be given a value

somewhere between 1 (perfect agreement) and 0

(perfect disagreement). In scoring matches between

judges, the highest weight (1.0) was given for full
agreement, that is, when both CCRT judges listed the
same wish, response from other, or response of self as

www.theipi.org 282
www.freepsychotherapybooks.org
 the most frequent across the 10 relationship episodes;
a lower weight of .66 was given when the highest

frequency of a component of one CCRT judge

matched the second highest frequency of the same

component of the other CCRT judge; an even lower

weight of .33 was given when the match was with the
second highest frequency of each CCRT judge.

Landis and Koch (1970) provided useful

standards on ranges for evaluating the degree of
agreement using kappa: 0 to .39 = poor; .4 to .74 =

fair to good; .75 to 1.00 = excellent. Our results show

a range of agreement from fair to good in terms of

weighted kappa: The wish and negative response of
self components were .61, and the negative response
from other component was .70 (n = 35, p < .001).

Samples C and D: A Comparison of the CCRT in

Dreams Versus Narratives

We reported on a sample including 13 adult

patients in psychoanalysis (Popp, Luborsky, & Crits-

www.theipi.org 283
www.freepsychotherapybooks.org
 Christoph, 1990; Popp et al., 1996). For each patient

we studied an average of 14.5 narratives from early

and late phases of therapy and an average of 13.0
dreams from early and late phases.

To compare the two pairs of judges’ agreements

on the CCRT standard category clusters, we used the
following system: We noted for each pair of judges
(the pair for dreams and the pair for narratives) the
most frequent first choice and the most frequent
second choice wish, response from other, and

response of self. To determine the closeness of a

match between judges, we gave the following

weights: a weight of 1.0 when the match was based

on agreement between the first-choice CCRT

component of Judge 1 and the first-choice CCRT

component of Judge 2, a weight of .66 when the first-

choice CCRT component of Judge 1 matched the
second-choice CCRT component of Judge 2, and a
weight of .33 when the match was based on

www.theipi.org 284
www.freepsychotherapybooks.org
 agreement between the second-choice CCRT

components of the two judges.

The results, seen in Table 1, showed weighted

kappas for dreams ranging from .58 to .83. These

weighted kappas showed fair-to-excellent agreement,

according to the convention established by Landis

and Koch 0970) noted earlier. Table 1 also shows that

the kappas for the wishes in both narratives and

dreams were somewhat less than the kappas for

response from other and response of self. These

kappas also suggested a new observation: that the

level of agreement of two judges in scoring the CCRT

is about as good for the dreams as it is for the

narratives (dreams average kappa = .70; narratives

average kappa = .72).

Samples E and F: A Comparison of the CCRT for

Session Narratives Versus Relationship Anecdote
Paradigm Narratives

www.theipi.org 285
www.freepsychotherapybooks.org
 The comparison in another study (Barber,

Luborsky, Crits-Christoph, & Diguer, 1995) was

between two forms of narratives: those that appear

spontaneously in psychotherapy sessions and those

that appear in the Relationship Anecdote Paradigm

(RAP) interview (Luborsky, 1990b) after a request to
tell narratives about relationships with others. We
therefore carried out two separate reliability studies,

one for the session narratives and one for the RAP
narratives.

Reliability of the CCRT From Psychotherapy

Sessions (Sample E)

The CCRTs were rated by two independent judges

for 19 patients. The percentage of agreement between

the two judges on the clustered standard categories

was 94% for wishes, 100% for responses from other

and 88% for responses of self. The corresponding

weighted kappas were .81, .64, and .73. It is notable

and may be meaningful that the percentages of

www.theipi.org 286
www.freepsychotherapybooks.org
 agreement and weighted kappas for the therapy

sessions were higher than for the RAPs.

Reliability of the CCRT from RAP Narratives

(Sample F)

The narratives in the 19 patients’ RAP interviews

were rated by two other independent judges. The
percentage of agreement between the two judges on
the clustered standard categories was 84% for wishes,

100% for responses from others (ROs), and 89% for

responses of self (RSs). The corresponding weighted

kappas were .68, .60, and .65. Weighted kappas were

used to allow for different degrees of agreement: A
weight of 1.0 was used if exact agreement was found
on the most frequent clustered standard categories, a

weight of .66 was used if the components were

second highest for one set versus highest frequency

for the other set, and a weight of .33 was given for

agreement between two sets that were both next to

the highest in frequency. The kappas were in the fair-

www.theipi.org 287
www.freepsychotherapybooks.org
 to-good range according to Landis and Koch (1970).
The percentage of agreement is also given (Table 1)

to demonstrate that it would be difficult to find much

higher kappas, especially in the ROs. Part of the basis

for the high percentage of agreement, but only fair

kappas, is the narrow range of the CCRT components
for these patients.

Sample G: A Comparison of the CCRTs of Children

From Age 3 to Age 5

Each of the 18 children in another study

(Luborsky, Luborsky, et al., 1995) told 10 videotaped

stories at age 3 and again at age 5. The experimenter

started each story by offering, as an initial stimulus, a

story-stem about a puppet family in which an

upsetting event occurred, such as the loss of the key
to the family car. The experimenter then conducted an
inquiry to stimulate the telling of a story about what
would happen next after the event in the story-stem.

Percentage of Agreement Between Two Judges

www.theipi.org 288
www.freepsychotherapybooks.org
 We used the narratives from 10 of the 3-year-olds

that had been rated by the two judges. The judges

completely agreed with each other on the wishes of 7
of the 10 children, on the responses from other of 10

of the 10 children, and on the responses of self of 9 of

the 10 children. Agreement was defined as a match
between the two judges in identifying the same
cluster of each child with the highest average sum of

standard categories.

Weighted Kappa Agreement

The weighted kappa was based on the assignment

of 1.0 to instances in which the first choice was a
match; partial matches were assigned a .5. The
weighted kappas for each CCRT component were
wishes, .33; responses from other, .69; responses of

self .60. The lower kappa for the wishes may have

reflected the very low variability of wish clusters.

Sample H: A Comparison of Adolescents at Age 14

and at Age 18 on the Basis of RAP Narratives

www.theipi.org 289
www.freepsychotherapybooks.org
 In a recent study (Waldinger, 1997a), the CCRT
was applied to the Early Adult Close Peer

Relationship Interview given to 40 persons at age 24

(20 male, 20 female). Twenty participants had had

psychiatric care, and 20 had never had this

experience. This semistructured interview, developed
by Shultz, Hauser, and Allen (1990), asks participants
to describe in depth their two closest relationships,

one a platonic and one a romantic relationship.

Transcripts of these interviews were coded for

relationship episodes, and the CCRT was extracted
from them by four independent judges. The averaged

weighted kappas across the four judges showed

evidence for the good reliability of the CCRT. W: kw

(weighted Cohen’s kappa) = .37; RO: kw = .69; RS:

kw = .76.

Sample I: The Core Conflictual Relationship Themes

of Borderline Personality Disorder

www.theipi.org 290
www.freepsychotherapybooks.org
 The aims of another study (Lefebvre, Diguer,

Morrissette, Rousseau, & Normandin, 1996) were to

examine (a) whether the CCRT can be reliably

applied to transcripts of patients with borderline

personality disorder and (b) whether the CCRTs from

this sample differ from those of patients with other
disorders.

Twenty patients (6 female, 14 male) were

included in the study. All presented with either an

adjustment disorder or a mood or anxiety disorder

using DSM-III-R criteria. Six of them met five DSM-
III-R borderline personality disorder criteria and four

met four criteria; the other 10 patients had no Axis II

diagnosis. All patients were evaluated by three

psychologists and three graduate students using the

SCID for Axis I and the SCID II for Axis II

(American Psychiatric Association, 1990). The
patients were asked to tell 10 narratives involving
incidents or events about themselves in relation to

www.theipi.org 291
www.freepsychotherapybooks.org
 another person, following the RAP interview method

(Luborsky, 1990b). Weighted Cohen’s kappas showed

high reliability among the three judges, who scored

the transcripts independently: W: kw = .72; RO: kw =

.61; RS: kw = .76. The results show that CCRT can

indeed be reliably applied to narratives of patients
with borderline personality disorder.

An additional study (Zander et al., 1995) is noted

just because it is a reliability study of the CCRT.

However, it is not included among the samples

reviewed here because it has some variations in
procedure that may explain why its level of reliability

in applying the CCRT scoring categories is lower

than in the other studies. The study is also difficult to

summarize because only ranges of kappa were

presented but not means or medians. It reported two

blocks of data. In Study 1 the video-presentation
kappas ranged from .44 to .58, whereas in the
transcript presentation the kappas ranged from .38 to

www.theipi.org 292
www.freepsychotherapybooks.org
 .47. In the second study the video presentations
ranged from .14 to .52 and the transcript presentation

from .35 to .48. One factor that may have contributed

to the lower reliabilities is the high number of RAP

narratives that the authors found were “not accepted

for coding.” For this to happen in a RAP interview
implies that the RAP interviewer did not instruct the
participants sufficiently on how to present reasonably

complete narratives; if this had been done, the

participants would have complied and told complete

narratives most of the time, as is typically the case

with RAP narratives. It also seems that the authors

did not compute the usual weighted kappas. The net

contribution of Zander et al. (1995) is the

demonstration that (a) their video scoring and

transcript scoring gave somewhat similar reliabilities

and (b) not all CCRT reliability studies demonstrate
the usual level of reliability in terms of kappas that is

www.theipi.org 293
www.freepsychotherapybooks.org
 found in the more typically conducted studies

reported in this chapter.

CONCLUSIONS

• Completeness of relationship episodes, the basic

unit of analysis, can be reliably judged, as
previously shown (Crits-Christoph, Luborsky, et
al., 1988). Likewise, the main other person and
the location of the relationship episode are both
reliably judged.

• The findings on reliability of scoring the CCRT by

the percentage agreement method show high
agreement, although this method is not as precise
as the weighted kappa method. For four samples,
the means were W: 87%; RO: 97%; RS: 89%.

• The main findings on reliability of scoring the

CCRT by weighted kappas, as summarized in
Table 1, based on eight samples in six main
studies, showed the mean level of reliability
across the eight samples is “good,” according to
the categories proposed by Landis and Koch
(1970), and not markedly different from sample
to sample or even from component to
component: W: kw = .60; RO: kw = .68; RS: kw

www.theipi.org 294
www.freepsychotherapybooks.org
 = .71. Because of the consistency of the level of
reliability across the eight samples with weighted
kappas shown in Table 1, we expect future
studies to show the same high level of reliability.

• An explanatory factor that may be significant,

although it has not been tested, is that the wish
component has a slightly lower mean weighted
kappa than the response from other or response
of self. This is especially evident for the wish
component in the children’s stories, which turns
out to be atypically unreliable with a weighted
kappa of .33 (all other samples have weighted
kappas for wishes of .58 or higher). It may have
been hard for judges to infer the wishes in
children’s stories.

• This analysis of reliability gives results only in

terms of the agreement for each of the three
types of components; we are working on studies
that also give the level of agreement for each of
the 94 standard categories in Edition 2 (34
wishes, 30 responses from others, 30 responses
of self).

• We hope also that in future reliability studies some

of the other factors influencing the CCRT’s

www.theipi.org 295
www.freepsychotherapybooks.org
 reliability will become better known. Among
these may be the scoring skill of the judges.

Note

[4] The research reported in this chapter was partially supported

by Research Scientist Award (NIDA) DA 00168-23A and
N1DA Grant 5418 DA 07085 (to Lester Luborsky), by
NIMH Clinical Research Center Grant P50 MH45178 and
Coordinating Center Grant U18-DA 07090 (to Paul Crits-
Christoph), and by Quebec FCAR Research Grant 95-NC-
1277 (to Louis Diguer). We thank Robert DeRubeis and
Carol Popp for helpful reviews of this chapter.

www.theipi.org 296
www.freepsychotherapybooks.org
 7

THE RELATIONSHIP ANECDOTES

PARADIGM (RAP) INTERVIEW AS A
VERSATILE SOURCE OF NARRATIVES

LESTER LUBORSKY

The idea for an interview-style “RAP session” for

eliciting narratives grew out of my restiveness with

the initial format of the CCRT method, which was

based only on the narratives spontaneously told in

psychotherapy. What if the person was not in
psychotherapy, or what if the psychotherapy sessions

were not available? These “what-ifs” led me to

construct a Relationship Anecdotes Paradigms (RAP)
interview that could be applied to almost any sample
of people and could serve to elicit narratives to use as
data for the same variety of purposes as the narratives
drawn from psychotherapy (Luborsky, 1990b). In

www.theipi.org 297
www.freepsychotherapybooks.org
 fact, versions of the original unpublished
Relationship Anecdotes Paradigms interview guide

(Luborsky, 1978b) have been widely circulated

among researchers for over 19 years.

The primary purpose of this chapter is to give a

more complete and time-tested guide to the
administration of the RAP interview. Other purposes
are to examine what is known about characteristics of
the narratives elicited by the RAP interview and to
review research applications of the RAP, especially as

a data source for the CCRT measure.

ADMINISTRATION OF THE RAP INTERVIEW

In the RAP interview the narrator tells about

actual events in relationships with other people. Each

narrative is an account of a specific interaction with a

specific other person. The narrator is free to tell a

narrative about any relationship episode, present or

past, and is encouraged to describe the episode

www.theipi.org 298
www.freepsychotherapybooks.org
 concretely and to include a sample of conversation
with the other person: what the narrator said, what the

other person said, and what happened at the end of

the interaction.

Instructions

The interviewer gives the following instructions:

Please tell me some incidents or events, each

involving yourself in relation to another person.
Each one should be a specific incident. Some
should be current and some old incidents. For
each one tell (1) when it occurred, (2) who was
the other person it was with, (3) some of what
the other person said or did and what you said or
did, (4) what happened at the end, and (5) when
the event in the narrative happened. The other
person might be anyone—your father, mother,
brothers and sisters, or other relatives, friends or
people you work with. It just has to be about a
specific event that was personally important or a
problem to you in some way. Tell at least ten of
these incidents. Spend about three but no more
than five minutes in telling each one. I will let
you know when you come near to the end of five
minutes. This is a way to tell about your

www.theipi.org 299
www.freepsychotherapybooks.org
 relationships. Make yourself comfortable and
engage in this RAP session as you would with
someone who you want to get to know you.

The total expected time for each patient’s telling

of 10 episodes, is between 30 and 50 minutes. The

interview is tape-recorded, beginning with the

patient’s name (or initials). The interviewer adds the

date, name of the therapist (or initials) if the patient

has a therapist, and the interviewer’s name.

In the usual form of the RAP procedure, the

patient is free to tell any incidents about any people.
It is desirable that there be some variety among the

people chosen. It is also useful if all the narratives are

not from one time period; some should be from the

present and some from the past. Each narrative

should be about a specific incident, not a generalized

amalgam of several incidents.

For special purposes, the form of the RAP

procedure can be varied. The interviewee may be

www.theipi.org 300
www.freepsychotherapybooks.org
 requested to provide narratives about specific types of
other people, such as narratives about the spouse for

patients in marital therapy. Each member of a couple

may be asked to tell narratives about the other (with

scoring suggested by concepts such as those from

Bernal & Baker, 1979). In an “object relations” RAP,
the instructions include a request for four narratives
about each of the main people in the participant’s

early life. In Mayman’s (1968) version, only early

memories are asked for. Another format for scoring

concentrates on aspects of the structure of the

narrative (Wilson, Passik, Morral, Turner, & Kuras,

1994).

Finally, the set of narratives can be drawn from

those told as part of interviews done for a variety of
other purposes, for example, interviews done as part

of anthropological fieldwork or as part of a

psychiatric evaluation. After such narratives have

been located in the text of the interview, they can be

www.theipi.org 301
www.freepsychotherapybooks.org
 scored as if the set were elicited by a RAP interview
(as illustrated in the study of adolescents by

Waldinger, 1997a).

Improving Rapport and Dealing With Special

Contingencies

The patient should be sufficiently comfortable to

tell a set of narratives about incidents that, from the
patient’s point of view, are reasonably accurate

accounts of actual events that the patient has

experienced. Once patients are comfortable with the

RAP interviewer, they usually can perform the task

without much difficulty. The interviewer may find it

useful to increase rapport by explaining the

confidentiality of the interview and its value in

understanding the patient’s relationships. These are
some ways to deal with special contingencies in a
RAP interview.

• To help patients who have difficulty being detailed

or whose episodes are too brief: “Could you tell

www.theipi.org 302
www.freepsychotherapybooks.org
 more about that incident?”

• To help patients who have difficulty finding any

relationship anecdotes to tell: “Just tell about any
incident, event, or interaction with anyone, either
an old or a recent incident. Tell any that you
happen to remember now.”

• To help patients who find the words given in the

instructions, to “describe events that were
important to you,” a hindrance to finding any
events, the interviewer should deemphasize these
words by saying, “Just give any incidents or
events as you think of them.”

• To help patients who find it hard to remember any

conversation in relationship episodes: “It is not
necessary to put in exact conversation; just say
the general idea of it” or “just put in what you
remember.”

After all instructions have been given to the

narrator, the interviewer should see that the narrator

provides reasonably complete narratives. The best

time to provide further guidance is at the end of the
first narrative, saying, “That first one is fine; go on in

www.theipi.org 303
www.freepsychotherapybooks.org
 the same way with the other narratives” or if further

direction is needed, “go ahead with the other

narratives; but please remember to _____” (fill in
with whatever needs fine-tuning).

CHARACTERISTICS OF RAP NARRATIVES

When participants follow the instructions for the

RAP, what kinds of narratives do they tell? The

characteristics of narratives presented here are the

first systematically recorded. They are based on 24

patients in psychotherapy at the outpatient psychiatric

clinic (OPD) of the Department of Psychiatry of the

University of Pennsylvania. This OPD sample is like
other samples drawn from the OPD as described by

Luborsky, Crits-Christoph, et al. (1988). The most

frequent DSM-IV diagnoses in the present sample, in

order of frequency from most to least, were typical
nonpsychotic diagnoses: depression, anxiety,
adjustment disorder, and personality disorders.

www.theipi.org 304
www.freepsychotherapybooks.org
 The areas covered in this description of
characteristics of RAP narratives (see Table 1)

include the time taken per narrative, the main type of

other person the participant interacts with in the

narrative, the number of others in the narrative, the

stage of life from which the narrative is drawn, and
the dominant emotion in the narrative.

The mean time taken to tell the narratives was 2½

minutes. For the telling of 10 narratives, the mean
time per interview was 26 minutes, with a mean time

per narrative of about 2½ minutes. These data show

that, although the instructions allow a maximum of 5

minutes, for this sample the interviews consumed
only about half of that per tape-recorded narrative.

The main type of other persons in the narrative

was greatest for “other people.” Each type of other

person was classified into one of four categories,

which occurred with the following frequencies:

www.theipi.org 305
www.freepsychotherapybooks.org
 TABLE 1
Descriptive Characteristics of RAP Narratives in an Outpatient
Psychiatric Clinic Sample (N = 24)

Time Mean SD
Average time per narrative (min) 2.4 0.7
Total time per interview (min) 26.3 7.9
N of
Content Percentage
narratives
Other person
Parents 21 15
Love relationships 18 14
Siblings 9 8
Other people 52 21
Number of different people 81 18
represented
When episode occurred
Adulthood, recent (past week) 31 26
Adulthood, past 53 23
Adolescence (10-18 years) 8 11
Childhood (less than 10 years) 8 14
Dominant emotion in narratives
Positive 19 21
Negative 73 22
Mixed or neutral 8 7

www.theipi.org 306
www.freepsychotherapybooks.org
 parents (21%), love relationship (18%), siblings

(9%), other people (52%). Clearly, the largest

category was “other people.”

“Different other people” were most represented

in the narrative. For this sample, on the average 81%

of the relationship episodes of each narrator were
about different other people. For example, if eight
different other people were featured in 10 episodes of
a narrator, the percentage would be 80.

The “current" time in the teller’s life was when

most of the action in the narrative took place. Four

time periods were analyzed: recent adulthood (past

week, 31%), past adulthood (53%), adolescence (10-

18 years of age, 8%), and childhood (less than 10

years of age, 8%). Therefore, most often the time of
the action of the narrative is either within the era of
the participant’s life that is current (past week) or in

the recent past (past adulthood).

www.theipi.org 307
www.freepsychotherapybooks.org
 The negative dominant emotion of the
relationship episode was very high. The scoring of

the dominant emotion of each relationship episode

was based on a rating of these categories: positive,

negative, and mixed or neutral. Most of the

relationship episodes (73%) had a negative dominant
emotion. Similar percentages for negative emotions
have been found in other patient samples, described

in chapters 4 and 17.

USES OF RAP NARRATIVES

RAP Narratives as Data for the CCRT Measure

RAP interviews provide a supply of narratives

that are more accessible but similar to those from the

psychotherapy sessions. The scoring system applied
to each type of narrative is the same as given in
chapter 2, this volume. Our experience with each type

of narrative leads us to expect that the CCRTs from

the sessions versus from the RAP interview are likely

www.theipi.org 308
www.freepsychotherapybooks.org
 to be much the same. Only one study so far gives

quantitative support to this impression (Barber et al.,

1995); in a sample of depressed patients given the
RAP before therapy started, the CCRTs scored from

the early sessions were not significantly different

from the RAP results. The two independent judges
were in agreement concerning the CCRT components
77% of the time for the wishes and responses of self

and 100% of the time for the responses of other.

A brief example is provided in this chapter’s

appendix, taken from the Penn Psychotherapy Study,

of a CCRT from Mr. Edward Howard based on a

RAP interview at a follow-up session 8 years after the

end of psychotherapy. This example not only

illustrates the CCRT scoring of RAP narratives (see

Table 2) but also allows comparison with the RE-

based narratives from Session 3 of the psychotherapy
(see chapter 4, this volume). The appendix includes
only the first three relationship episodes as scored by

www.theipi.org 309
www.freepsychotherapybooks.org
 only one of three judges who scored the 10

relationship episodes elicited in the RAP session.

Another independent judge summarized the three

judges’ scores in the summary table (see Table 2).
Several findings can be read from the summary table.

Wish A shows considerable agreement among the

three independent judges. Eight or 9 out of the 10

relationship episodes were scored by the three judges

as containing the general version of Wish A. Wish A1

is one subcategory of Wish A that was found less

frequently. For the response from others, the
frequency of the negative responses (“rejects,
criticizes me”) was from 2 to 4 out of the 10 for the

three judges.

For the responses of self, the highest frequency

negative type was “frustrated, angry.” The entire

CCRT, on the basis of the usual principle of the

highest frequency type of components, is “I wish to
be close with, liked by, and cared about by the other

www.theipi.org 310
www.freepsychotherapybooks.org
 TABLE 2
CCRT Score Sheet Summary (Number of Relationship
Episodes Containing Each Component)

Patient: Mr. Edward Howard, No. 44

Session: RAP
Number of REs: 10
Wish, Need, Intention
Judge
V C M
A. To be close with, liked by, and cared about by 8 9 8
others
A1. To be accepted, to receive approval 3 2 4
A2. To be close, connected 1 5 1
A3. To be cared for, emotionally supported 2 1 2
A4. To have more satisfying relationships with me 1 1 1
Negative Response From Other
Judge
V C M
1. Rejects, criticized me 4 2 4
2. Angry 1 1 1
3. Indecisive, changes mind 2 0 2
4. Feels silly, embarrassed 0 0 1
5. Needs support 0 0 1
Negative Response of Self
Judge
V C M
1. Frustrated, angry 3 2 4
2. Confused 1 1 2
3. Afraid of rejection 1 1 2

www.theipi.org 311
www.freepsychotherapybooks.org
4. Inhibits desire to be close 1 1 1
5. Doesn’t express anger 1 1 1

www.theipi.org 312
www.freepsychotherapybooks.org
 person, but the other person rejects or criticizes me

and I become frustrated and angry.”

The reader may have a sense of having read this

example before. In fact, the reader is partly right: The

CCRT in this example from the RAP is much like the

CCRT from Session 3 and Sessions 82 and 83 of Mr.
Howard. What tends to change most in improved
patients, such as this one, is that some of the
responses from others and responses of self shift from

negative to positive. Thus, it appears not only that the

CCRT is consistent across time, but also that, even
after an 8-year interval, there remains a congruence

between the CCRT from the RAP narratives and the

CCRT from psychotherapy sessions.

The CCRT has shown adequate reliability in

terms of agreement between judges when scored from
the narratives told during psychotherapy (see chapter

7, this volume); reliability studies using RAP

www.theipi.org 313
www.freepsychotherapybooks.org
 narratives, such as those of Mr. Howard, tend to show

similar themes and similar levels of reliability to

those from psychotherapy (Barber et al., 1995; van

Ravenswaay, Luborsky, & Childress, 1983).

Other Uses of the RAP Interview Procedure

The four uses reviewed here constitute only a

small sample of a widening field of applications.

As a Database for Studies of Explanatory Style

The RAP interview contains descriptions of

events and often also gives explanations of the causes

of the events. Such explanations are scorable in terms

of a concept called “explanatory style” (Seligman et

al., 1984). The explanatory style that is measured in

this way can be thought of as part of an inference

pattern about causes of the events. The explanations
for good and bad events are scored on three
dimensions: internal-external (“it is me” versus “it is
not me”), stable-unstable (“it will always be me”

www.theipi.org 314
www.freepsychotherapybooks.org
 versus “it will not always be me”), global-specific (“it
will affect all aspects of my life” versus “it will affect

just this aspect”). These explanatory styles have been

shown to be associated with the development of

depressive symptoms (Peterson & Seligman, 1984).

The explanatory style measure that has been

applied to the RAP interview data is a procedure
called the Content Analysis of Verbatim Explanations
(CAVE) technique (Peterson & Seligman, 1984).
Reliabilities of coding for each of the three

dimensions are high: .93, .89, and .90 (Cronbach’s

alpha; four judges pooled for the internal, stable, and

global ratings, respectively; Peterson, Bettes, &
Seligman, 1985). In a study of a patient in
psychotherapy who had precipitous mood swings, it
was found that spontaneously given causal

explanations about spontaneously reported negative

events predicted the appearance of a mood swing

(Luborsky, 1996, chap. 5; Peterson, Luborsky, &

www.theipi.org 315
www.freepsychotherapybooks.org
 Seligman, 1983). For this patient the combined

dimensions of internal, stable, and global

explanations for bad events preceded shifts toward
increased depression, whereas the combined

dimensions of external, unstable, and specific

explanations preceded shifts toward decreased
depression.

As a Source of Data for Developmental Studies of

Central Relationship Patterns

A special kind of interview developed by

Buchsbaum and Emde (1990) has been applied to a

sample of young children at age 3 and again at age 5

(Luborsky, Luborsky, et al., 1995). In their stories a

combination of fantasy constructions and recounting

of real events appears. To make the telling of stories

about events easier for these very young children,

family figure dolls and initial story-stems about

conflictual situations are used.

www.theipi.org 316
www.freepsychotherapybooks.org
 In another developmental study, a RAP-type set
of memories was also obtained from 23-year-olds

retested as part of the Berkeley Longitudinal Study

(Jack Block, personal communication, 1989). They

were used to assess the developmental correlates of

the central relationship pattern of the stories
stemming from the participant’s different age periods
(Thorne, 1989, 1995a, 1995b; Thorne & Michaelien,

1996).

As a Basis for Studies of Self-Understanding

The RAP narratives are a convenient source of

data for studies of the ability of a person to

understand his or her own central relationship pattern,
as described in chapter 15, this volume, and in a
procedure for self-interpretation of the RAP
narratives (Luborsky, 1978a).

As a Basis for the Comparison Among Diagnostic

Groups

www.theipi.org 317
www.freepsychotherapybooks.org
 We are launched on a set of studies of RAPs in

which the comparison is of the central relationship

pattern of three groups: patients with major

depression (Luborsky, Diguer, et al., 1996), those

with schizophrenia, and normal persons (defined as

without a psychiatric diagnosis on the Schedule for
Affective Disorders and Schizophrenia–Research
Diagnostic Criteria (SADS-RDC; Demorest, Crits-

Christoph, Hatch, & Luborsky, 1997).

As a Basis for Intergenerational Comparisons

RAP narratives are being compared for the patient

versus the patient’s parents. The narratives are about

events selected by the patient; a subset of these are

narratives about the same event as told both by the

patient and by the patient’s parents (Waldinger,

1997a).

CONCLUSIONS

www.theipi.org 318
www.freepsychotherapybooks.org
 • The RAP interview produces narratives that have
much in common with those produced in
psychotherapy sessions (Barber et al., 1995). The
format is well accepted by both patients and
nonpatients.

• The RAP procedure appears to have some

similarities to the Thematic Apperception Test
(TAT; Murray, 1938), and its uses are often
similar. Both are interviews in which narratives
are elicited, but there is one major difference.
The RAP narratives are told by the participant as
accounts of actual experiences with actual
people; the TAT stories are told as fictional
narratives that are stimulated by the people and
settings depicted in the TAT cards. As an
assessment method, the advantage may go to the
RAP interview: Conclusions based on the RAP
may be less inferential than those based on the
TAT because RAP narratives are based on
accounts of events that more directly reveal the
patient’s relationship patterns.

In fact, the heavy emphasis in the past 40

years on projective techniques in diagnostic
psychological testing (Holt, 1978; Rapaport, Gill,
& Schafer, 1968), although it has had significant

www.theipi.org 319
www.freepsychotherapybooks.org
 benefits, may have led to a slighting of the
potential gains for accurate assessment that can
be derived from narratives about actual events in
the patient’s present and past relationships. The
RAP procedure is congruent with the trend
toward giving less attention to projective tests
and even to psychodiagnostic testing generally
(Holt, 1967; Piotrowski & Keller, 1984).

• The RAP interview relies on a relatively natural

format; it requires only doing what people like to
do—tell narratives about events in relationships
that have occurred between themselves and other
people. For this reason the RAP interview can be
used for assessing a broad range of
developmental, intellectual, and cultural
qualities, as is illustrated in the research studies
already completed or in progress.

www.theipi.org 320
www.freepsychotherapybooks.org
APPENDIX

The RAP Interview for Mr. Edward Howard: An 8-

Year-Follow-up

The following are 3 of 10 relationship episodes

from a RAP interview. In the left margin are the

tailor-made CCRT scoring annotations by one of the

judges; the parts of the text from which the inferences
were derived are underlined.

RE 1: Therapist

And uh, I was thinking the therapy was the first

time in my relationships that I felt like uh that I
was worth something, and, and that I was
important in myself. And even though it was a
professional kind of relationship, it struck me uh
that that was kind of strange, you know that. I
hadn’t felt that at all in my family. Uh, but that’s
really how it was. That was the first time. And I
guess in terms of my uh a lot of my individual
growth, that I was somebody uh separate and an
individual in my own right and that somebody
cared about me just for that and not what I could
do for them or not because of who I knew or
something like that…

www.theipi.org 321
www.freepsychotherapybooks.org
 And uh, I was thinking
PRS: Feelings of //RS the therapy was the first time in my
self-worth relationships that I felt like uh that I was
worth something// and and that
PRS: Felt //RS I was important in myself.// And
important even though it was a professional kind
of relationship, it struck me uh that that
was kind of strange, you know that
NRS: Low self- //RS I hadn’t felt that at all in my
worth family.// Uh, but that’s really how it was.
That was the first time. And I guess in
terms of my uh a lot of my individual
growth, that
W: To be cared //W I was somebody uh separate and
about for himself an individual in my own right and that
somebody cared about me just for
that// and not what I could do for them
or not because of who I knew or
something like that…

RE 2: Work Supervisor

My first-year placement was in Law Center, and

my supervisor was a man who grew up in
Hungary and as a supervisor he really showed a
lot of interest in me as a professional and a
person. We spent a couple of hours doing an
evaluation of…of my work, because of his
concern he brought out a lot of things that had
been going on with me and he says, “Frankly,
I’ve been a little concerned.” I said, “What
about?” He said, “I think you’re a loaf.” I said,
“A loaf!” I’m not answering, “a loaf.” I said, “A

www.theipi.org 322
www.freepsychotherapybooks.org
 loaf, huh.” ‘Cause actually, I’d been…I’d been
waiting for him to say something like this
because my, my worst fear was that I would be
seen as a, as an ineffectual do-nothing, you
know, and since I really worked hard to
counteract this image that I had of myself and
saw myself…. And so he talked a little bit more
about that then I figured out what he was saying.
He wasn’t saying “a loaf, “ he was saying
“aloof.” And it was his accent, and it was so
weird because that was like my worst fear. And
he said that uh, yeah, that I…I seem to always be
walking around like absentminded, and I would
never have time to talk to the staff people. He
thought that a lot of the staff lime much more
than I, I liked them, and uh, and that was kind of
nice to hear but uh also it… I began to think
more about how I turned people off because of
my feeling that they wouldn‘t be interested in
me, like my assumption, and working so hard to
get their approval uh, I ended up turning them
off instead of being direct with them.

My first-year placement was in Law

Center, and my supervisor was a man
who grew up in Hungary and
PRO: Showed //RO as a supervisor he really showed a
interest in me lot of interest in me as a professional
and a person.// We spent a couple of
hours doing an evaluation of…of my

www.theipi.org 323
www.freepsychotherapybooks.org
 work, because of his concern he
brought out a lot of things that had been
going on with me and he says,
NRO: Express //RO “Frankly, I’ve been a little
concern concerned.”// I said, “What about?”
NRO: Criticizes //RO He said, “I think you’re a loaf.” I
him said, “A loaf!” I’m not answering, “a
loaf.” I said, “A loaf, huh.” ’Cause
actually I’d been…
NRS: Fear of //RS I’d been waiting for him to say
criticism something like this// because
W: To be seen as a //W my, my worst fear was that I would
competent, be seen as a, as an ineffectual do-
effective worker nothing,// you know, and since
PRS: Tried to be //RS I really worked hard to counteract
competent and this image that I had of myself and saw
effective myself…
PRS: //RS And so he talked a little bit more
Misunderstood about that then I figured out what he
was saying.// // He wasn’t saying “a
loaf,” he was saying “aloof.” And it was
his accent, and it was so weird because
NRS: Fear of //RS that was like my worst fear.//And he
criticism said that uh, yeah, that I … I seem to
always be walking around like
absentminded, and I would never have
time to talk to the staff people.
PRO: Reassures, //RO He thought that a lot of the staff like
encourages me much more than I, I liked them,// and
uh, and that was kind of nice to hear but
uh also it…
NRS: Reject //RS I began to think more about how I
others turned people off//
NRS: Fear of //RS because of my feeling that they
rejection wouldn’t be interested in me, like my
assumption, and

www.theipi.org 324
www.freepsychotherapybooks.org
W: To receive //W working so hard to get their
approval from approval//
others
NRS: Reject uh,//RS I ended up turning them off
others instead of being direct with them.

RE 3: Sister

Uh (pause) I don’t’ have too many angry

interactions, which is a problem. I have to work
on that. Uh, I remember a sad interaction that I
had when I was about 16 or 17. This…like the
first time I could remember having any kind of
nonnegative feelings for my sister who was
younger than I was. I always like either felt
really cold and distant and uh just aloof, like I
hated her guts, but uh this time she came in in
the morning off her delivery route and uh and
she was crying ‘cause her, her hands and her feet
were really cold. And uh and she was like a
really courageous kid, I mean she doesn’t cry,
you know, and I just felt so bad that her fingers
and toes were a little bit frostbitten, you know.
And uh, it made me feel really sad and it made
me feel angry, too some, because I felt like she
shouldn’t have had to do that and and I knew
that that the reason…I guess it made me angry at
my parents some.

Uh (pause) I don’t have too many angry

www.theipi.org 325
www.freepsychotherapybooks.org
 interactions, which is a problem. I have
to work on that. Uh, I remember a sad
interaction that I had when I was about
16 or 17. This … like
(W): To be close //W the first time I could remember ever
with his sister having any kind of nonnegative feelings
for my sister// who was younger than I
was.
NRS: Felt cold, //RS I always like either felt really cold
distant, aloof and distant and uh just aloof, like I hated
her guts,// but uh this time she came in
in the morning off her delivery route and
uh and
NRO: Cried; in //RO she was crying ’cause her, her
distress hands and her feet were really cold.//
PRO: Courageous //RO And uh and she was like a really
courageous kid,// I mean, she doesn’t
cry, you know, and
PRS: Empathized //RS I just felt so bad that her fingers and
toes were a little bit frostbitten,// you
know.
NRS: Sad //RS And uh, it made me feel really sad//
NRS: Angry //RS and it made me feel angry, too
some,// because I felt like she shouldn't
have had to do that and and I knew that
that the reason ... I guess
NRS: Angry //RS it made me angry at my parents
some.//

www.theipi.org 326
www.freepsychotherapybooks.org
 8

WHY EACH CCRT PROCEDURE WAS

CHOSEN

LESTER LUBORSKY

To measure such a complex concept as a central

relationship pattern reliably and validly requires both

plain luck and great care in decision making about

methods of assessment. The decisions I made had

profound effects on the structure of the method and

on the observations derived from it. In this chapter I

explain the virtues and vices of each of the many
decisions about construction of the CCRT measure.
The decisions are grouped into three main bunches:
the database, the scoring system, and the inference

level.

DATABASE DECISIONS

www.theipi.org 327
www.freepsychotherapybooks.org
To Rely on Psychotherapy Sessions

There is a special virtue in using psychotherapy

sessions as the database of the new measure; after all,

sessions were the original data from which Freud

(1912/1958a) generated the concept of a central

relationship pattern and, from it, his transference

template. We therefore hoped that the phenomena of

transference that were supposed to be present in

sessions could be captured within them. For locating

such a bounty, it would be worth putting up with the

time and expense of transcribing and of making

judgments from sessions. My approach was to use

psychotherapy sessions but to try to rise to the

challenge of transforming such data into objectively

scorable form by means of the procedures listed in
chapter 2, this volume.

My decision to use data from psychotherapy

sessions to find central relationship patterns did not

follow the most popular route. Workers in the field

www.theipi.org 328
www.freepsychotherapybooks.org
 had concentrated for several decades on
questionnaires about the patient’s relationship with

the therapist and with others, perhaps because of the

difficulty of transforming data from sessions into

objectively scorable measures. True, the

questionnaire method could save time, but so far the
gains from relying on it solely appear to have been
penny-wise, because the questionnaire approach, as

reviewed in chapter 20, has not convincingly

captured what we were after.

To Restrict the Database to Narratives About

Relationship Episodes

The decision to restrict the scoring to the

narratives about relationship episodes within the

sessions came early in the development of the CCRT
method. As I tried to formulate the relationship
patterns from whole sessions, it became obvious that
the judges’ inferences about transference were mainly
derived from the patients’ narratives about

www.theipi.org 329
www.freepsychotherapybooks.org
 relationship interactions with people. Narratives
provide rich data for the method because they are

concrete examples of the patient’s interactions with

others and self. Thus, it seemed simpler and without

significant loss of information to focus the judges’

attention on relationship episodes. The narrowing of
the focus to relationship episodes not only reduced
the data to be inspected and scored but, more

importantly, highlighted the relationship ideas and

behaviors that were of most interest.

The decision to score only the relationship

episodes was reinforced by the conclusion from a

study of ratings of transference (Luborsky, Graff, et
al., 1973). In that study, ratings of “transference as
expressed to specific objects” were found to yield
higher interjudge agreement than ratings of

“transference as expressed in the entire segment”;

that is, ratings of the entire segment without attention

www.theipi.org 330
www.freepsychotherapybooks.org
 to rating the specific people were less reliable than
ratings focused on specific people.

In light of further experience, the decision to

restrict scoring to the relationship episodes appears to

have been a good one, even though in principle any

restriction of focus entails some loss of information.
In fact, the judge can and should read the transcript of
the whole session, which is easy to do because the
relationship episodes are typically presented to the
judge within the transcript of the whole session. The

whole session can be thought of as a context for

further understanding of each relationship episode.

The session can be considered to be the associative
context for the relationship episode in much the same
sense as the dream in dream analysis, for which much
of the session can be viewed as associations to the

dream. As a whole, the loss entailed in the restriction

to the relationship episodes is offset by a gain in

simplicity of the procedure and is further offset by a

www.theipi.org 331
www.freepsychotherapybooks.org
 crucial gain in focus because the relationship
episodes provide a good basis for inferring typical

relationship patterns.

Finally, I decided not to use the thought unit as an

alternative to the relationship episode—at least not

now. It can, however, be used as a subunit within the
relationship episodes (see chapters 3 and 5, this
volume). The thought unit is an operationally defined
single thought (Benjamin, 1974). For the purpose of
producing a database that can measure interactional

patterns, however, it seemed too small and too

noninteractional. The thought unit is a much smaller

unit than the relationship episode; it is often about the
length of an average sentence. Another unit that could
have been used, patient utterances, was relied on by
Schacht and Binder (1982) in their dynamic focus

method. It is defined as a single uninterrupted turn at

talk. It, too, appeared to be too short in comparison

with the relationship episode.

www.theipi.org 332
www.freepsychotherapybooks.org
To Include Behavioral Enactments of the Relationship
With the Therapist

The database for the CCRT consists both of

narratives told by the patient and of enactments of
these. The enactments are the behavioral expressions

of interactional sequences during the session in

relation to the therapist. In dramatic terms, they are

the central scenes within the play. Although the

whole session can be considered an interaction

between patient and therapist, the enactments are

limited, discrete behavioral episodes within the
session. I assume, so far, that in terms of CCRT
content the narratives and the enactments have much

similarity.

The use of enactments offers some advantages for

the CCRT method. First, they are actual behavioral
interactions between patient and therapist, not just
narrative accounts of episodes with the therapist.
Therefore, they offer an opportunity to examine the

www.theipi.org 333
www.freepsychotherapybooks.org
 validity of the CCRTs from the narratives by a

comparison with the enactments. Second, the use of

enactments increases the number of relationship
episodes with the therapist, a category that is
typically sparse. One disadvantage, however, is that

more work needs to be done to improve the reliability

of recognition of enactments.

SCORING SYSTEM DECISIONS

To Use Guided Judgments

In the everyday relatively unguided approach to

inferring the transference, the clinician is free to rely

on any principle and any level of abstraction in

making judgments that his or her training and
intuition suggest is appropriate. But the use of
unguided judgments of transference —for example,

those examined by Seitz (1966)—has yielded poor

and ambiguous agreement among clinicians. In

contrast, measures of central relationship patterns,

www.theipi.org 334
www.freepsychotherapybooks.org
 such as the CCRT, the Plan Diagnosis, or the
Dynamic Focus, use guided clinical judgment, which

gives high levels of interclinician agreement, as

summarized in chapters 6 and 20, this volume. These

guides specify in advance the judgment principles

and levels of abstraction that should be relied on for
making inferences. Holt (1978) reviewed the
extensive research on unguided versus guided clinical

judgments and concluded that guided approaches

yield benefits for reliability and validity; measures

that provide some degree of guidance to the judge

produce results that are psychometrically more

promising.

To Identify the Main Other Person Within Each

Relationship Episode With Whom the Speaker Is
Interacting

The judges who select the relationship episodes

from the psychotherapy sessions also identify in each

relationship episode one main other person with

www.theipi.org 335
www.freepsychotherapybooks.org
 whom the narrator is interacting. (The agreement on

this task is given in chapter 6.) A CCRT built on this

basis provides an opportunity to learn the degree to

which the relationship pattern in the CCRT is

pervasive across many types of relationships.

An alternative decision might have been to

analyze the relationship episodes grouped for each
type of other person. This method would yield a
separate CCRT about father (and father figures),

about mother (and mother figures), about the

therapist, and so on. Developing separate CCRTs in
this way would have caused a practical problem: For

many participants we would not have had a sufficient

sample of relationship episodes about each type of

other person. This was certainly the most difficult

restriction faced by Fried et al. (see chapter 11, this

volume) in her study comparing the CCRT for the
therapist with the CCRT for other people. (A study on
this topic by Crits-Christoph & Demorest, 1991, had

www.theipi.org 336
www.freepsychotherapybooks.org
 a large number of relationship episodes per different
other person but relied on one treatment only with

many sessions transcribed and scored.) Fried et al.

looked only at the degree of parallel between the

CCRT for the therapist and the CCRT for other

people; she has not yet explored the kinds of
differences. When this task is done, she will probably
find both a specific prototype for each type of other

person, as well as what we have already found, a

basic prototype that encompasses and pervades the

narratives about most other persons. Such a general

as well as specific pattern would be no surprise to

clinicians; in fact, Freud (1912/1958a) expected both

kinds of results on the basis of his observation about

the characteristics of the pattern (reviewed in chapter

21).

To Rely on Three Components of Narratives: Wishes,

Responses From Others, and Responses of Self

www.theipi.org 337
www.freepsychotherapybooks.org
 The CCRT scoring system relies on inferences

based on signs in the session of central classes of

components that are often reflected in clinicians’
descriptions of the transference pattern. These

components are often in conflict with each other. The

first is the wish class: wishes, needs, and intentions;
psychoanalytic theorists call these drive derivatives.
The second is in the response class: responses from

others and responses of self; psychoanalytic theorists

generally consider these as containing control,

executive, or ego functions (Rapaport & Gill, 1967).

The wish class and the response of self class are
recognizable as representatives of two of the main
entities posited in psychoanalytic constructions of the
“mental apparatus.” The responses from other contain
that part of the perceived relationship environment

that the person must deal with. Clinicians tend to

believe that through greater insight the patient’s

responses from others and of the self can become

www.theipi.org 338
www.freepsychotherapybooks.org
 more conflict free in the course of psychotherapy

because the patient’s insights offer more information

about the internal and external conditions affecting

the possibilities of satisfaction of the wish, need, or
intention. Thus, the built-in similarity of the content

of the CCRT to the transference concept increases the

meaningfulness of the CCRT method in clinical

practice.

To Use a Theme Format That Highlights Conflicts

The format of the CCRT is set up to point to the

locus of conflicts. The usual sequence of the
components is wishes, followed by responses from

others, followed by responses of self. Among these

components, the two most prominent types of

conflicts are (a) among the wishes (“I want this, but it

conflicts with something else I want”) and (b)

between the wishes and the responses from other and
of self. The second type of conflict is much more
frequent.

www.theipi.org 339
www.freepsychotherapybooks.org
 Example: The most obvious conflict for Mr.
Edward Howard (see chapter 5, this volume)
concerns his wish to be close, which conflicts
with his expected response from others of being
cut off from closeness. Although he is somewhat
aware of this conflict, he is less aware of some
aspects of the conflict, such as that his wish to be
close conflicts with his wish to be distant. And
perhaps his wish to be distant arises because of
his expectation that he will be cut off from
closeness. He is probably even less aware of
other aspects of the conflict and that the wish to
be close, because it is so intimately associated
with sexual wishes, heightens his expectation of
a negative response from others.

To Judge All Responses as Either Positive or Negative

The inclusion of the qualities of the narratives for

the CCRT as positive or negative followed the work
of Freud (1912/1958a), who designated transference

patterns as positive or negative. I have been using this

aspect with considerable satisfaction since the

launching of the CCRT idea (Luborsky, 1977b). The

judgment of positive or negative was confined to

www.theipi.org 340
www.freepsychotherapybooks.org
 responses from other or responses of self; wishes are

not typically in themselves positive or negative.

Much has been learned since then about positivity

and negativity; for example, Grenyer and I (chapter

4) have shown that positivity and negativity can be

reliably judged and (chapter 9) that there is a general
proclivity of people toward negative narratives and
therefore negative CCRTs.

To Use Both Tailor-Made and Standard Categories

In the original form of the CCRT (Luborsky,

1977b), as well as in part of its present form in this

book, categories that were fashioned to suit each

patient were selected by each judge in a scoring

system appropriately called tailor-made. In the

language of psychometrics, the system is idiographic

because it outfits each patient with case-specific

descriptive categories. This fine asset, however, is
offset by a practical liability: The variability from
patient to patient and from judge to judge in the

www.theipi.org 341
www.freepsychotherapybooks.org
 selection and wording of tailor-made categories
makes comparisons among cases problematic.

The contrasting current system based on standard

categories, also discussed in this book, is nomothetic.

It escapes the problem of variability in the wording of

case formulations by asking judges to fit their
formulations into the standard categories described in
chapter 3. Other standard category systems for central
relationship pattern measures differ in their choice of
categories; for example, the Cyclical Maladaptive

Pattern (Schacht & Binder, 1982; also see chapter 20,

this volume) and the Quantitative Analysis of

Interpersonal Themes (Crits-Christoph, Demorest, &
Connolly, 1990; see chapter 20, this volume) require
judges to code formulations into the categories
provided by the Structural Analysis of Social

Behavior (Benjamin, 1974).

www.theipi.org 342
www.freepsychotherapybooks.org
 In conclusion, although an idiographic approach

is desirable because it is closer to the clinical process

of making transference formulations, the use of
standard categories provides definite benefits to the

researcher, including a greater ease of establishing

interjudge agreement. Yet an evaluation remains to be
done of the gains from the use of standard categories
and the degree to which these make up for the loss of

the uniqueness of each case that is provided by the

tailor-made categories. I believe, therefore, that it is

advantageous to report results by both tailor-made

and standard categories.

To Rely on Redundancy Across the Narratives

(Pervasiveness) as the Indicator of the CCRT

From the inception of the CCRT method

(Luborsky, 1977b), CCRT scores were based on their

pervasiveness across narratives about REs, that is, on

the proportions of the REs containing each type of

component. This definition of pervasiveness was first

www.theipi.org 343
www.freepsychotherapybooks.org
 applied to 20-minute segments of sessions from the
Penn Psychotherapy Study.

The reliance on the frequency of each category to

define the CCRT makes good sense as an indicator of

the relationship schema reflected in the CCRT, but it

also might risk missing what is salient but infrequent.

This risk was noted by Howard Shevrin in his

comments about my paper on the nature of the CCRT

at the MacArthur Conference in 1985. This concern

was also expressed in Hartvig Dahl’s critique of the

CCRT at the psychotherapy conference in Sweden’s

Skokloster Castle in April 1980; he was concerned

that the formulation could be limited because of
reliance on frequency and suggested it would be
better to call it the main theme, not the core theme.

Several findings, however, helped in coping with this

concern. The main one is that the frequency with

which a theme is expressed tends to be a good

indicator of its importance; according to Murray

www.theipi.org 344
www.freepsychotherapybooks.org
 (1938), the most frequently expressed theme is the

most central one in the sense that it tend to be the

locus of the most pervasive relationship problem and

a concomitant of the greatest intrapsychic conflict.

Example: The heart of Ms. Sally Simpson’s

CCRT illustrates the issue. Her CCRT was “I
want to be given reassurance that I’m okay and
even special (mentally, professionally, and
sexually), but the other person will not give it
and I feel rejected and defective.” Her main wish
to be okay or special and the responses to
frustration of that wish, as reflected in her
CCRT, caused her the greatest recurrent
suffering. It was responsible for her need to start
treatment when she did. At that time she was
suffering acutely because she realized that she
was not going to be special for the married man
she was having an affair with and he would go
back to his wife. She tried throughout the
treatment to be special for the therapist and soon
began to feel that she was. This theme, therefore,
not only was expressed in a salient instance but
was frequent.

Example: The wish in the CCRT of another

patient, Ms. Rachael Apfel (see chapter 12, this

www.theipi.org 345
www.freepsychotherapybooks.org
 volume), to get a positive response, especially
from men, and to be able to cope with negative
ones, was most frequent and most central.
Recurrently, throughout most of the treatment,
she maintained the painful expectation that she
was going to be rejected. She expressed
instances of this expectation in both dreams and
relationship episodes.

Another reason that the theme with the highest

frequency is worth designating the core theme is that

it provides a framework within which to understand

the network of intertwined themes. The other themes,

many of them clearly subsidiary themes, can be

represented as related to the core theme. Rather than

considering all the other themes as subthemes, it is

often more fitting to refer to them as alternative
expressions of the CCRT.

Example: For Ms. Sally Simpson, her oedipal

theme is a version of her CCRT. Two examples
from her two early and two late sessions support
this inference: (a) her wish to be special
compared with the wife of the married man she
was having an affair with and (b) a triangular

www.theipi.org 346
www.freepsychotherapybooks.org
 situation evident in her relationship episodes in
which a woman telephone operator was
continually seen as preventing her from
contacting the man.

Example: For Ms. Cathy Cunningham, her

oedipal theme also appeared to be one version of
her CCRT. The general version of her main
wishes was “I need and want to assert myself
and to get the support and attention of a man.” In
some sessions, especially the later ones (e.g.,
Session 1,028), the wish appeared as “I want and
felt I must have a baby from Father" and “I want
and must have a baby from you [the therapist].”

Finally, the reliance on frequency as a criterion

appears not to have interfered with and even is likely
to have fostered the high association of the CCRT-

based results with clinically based observations about

the transference as summarized in chapter 21, this

volume. This type of association needs to be explored
more systematically, for it offers a basis for
examining the value of reliance on frequency as a

criterion for assessing the CCRT.

www.theipi.org 347
www.freepsychotherapybooks.org
To Allow Only One Score per Relationship Episode for
Each Different Category

The score for each category is the frequency of

that category on the basis of a count of the number of
relationship episodes out of the total number of

relationship episodes in which the category appeared

(that is, its pervasiveness). Each relationship episode

is allowed only one score for each different type of

scoring category. The desirable effect of this decision

is to emphasize the degree of redundancy of the

category across relationship episodes.

An alternative CCRT scoring system would

permit the frequency to reflect the total number of

times the category appeared, regardless of how often

it appeared in a relationship episode. A high score,

however, that is based on a count done in this way
might come from a few relationship episodes in
which the category was frequent. We therefore opted
for our current system, which clearly captured

www.theipi.org 348
www.freepsychotherapybooks.org
 pervasiveness of categories across relationship

episodes. But a scoring in terms of the now-rejected

alternative score is worth examining in future

research.

To Use Our Usual System for Counting All Scorable

Components Regardless of Sequence

The decision was made to include all scorable

components, not just those with an explicit, complete

sequence of the CCRT components. The decision

means that all types of CCRT components are

counted even when there are missing components in
the sequence; for example, a type of wish might be
expressed with no expected response from the other

to that wish. The decision to use all scorable

components was based on the need for simplicity of

scoring, the concern that sequences often might not

be explicitly stated, and the expectation that the

decision would provide meaningful data about the
central relationship pattern.

www.theipi.org 349
www.freepsychotherapybooks.org
 I have covered all bases, however, by adding as
an optional scoring system the designation of explicit

sequences (chapter 2, this volume). Research needs to

be done in which the present score-all-components

system is compared with the score-only-explicit-

sequences system to determine the degree of

difference between the two (as my colleagues and I

have briefly done in chapter 9). Since this section was

initially written in 1989, further development of a

sequence method has appeared (Dahlbender, Albani,

Pokorny, & Kächele, in press).

INFERENCE LEVEL AND FOCUS DECISIONS

To Stay Within the Range of Moderate Inference

The decision to stay within a moderate level of

inference was based on the impression that when the
level of inference is high, it is hard to get agreement

among judges. Yet the restriction of inference to a

moderate range also means that the present guidelines

www.theipi.org 350
www.freepsychotherapybooks.org
 for scoring appear to be adequate in terms of judges’

agreement with each other (see chapter 6). Although

the decision to stay within the limits of moderate
inference means that some unconscious-level

inferences may be excluded for some patients,

fortunately the restriction of the range of inference
does not mean that unconscious processes are
excluded; in fact, according to one study (see chapter

15), important inferences within the usual CCRT

appear to be at a restricted-awareness, or

unconscious, level.

Several ideas for future research are implied by

these decisions. A study should grade the degree of

inference used for each score; it could then be

determined whether different levels of inference are

associated with different levels of reliability. Another

study might investigate the types of CCRTs when no
limit on inference level is imposed on the judge. This
study would check systematically on the reliability of

www.theipi.org 351
www.freepsychotherapybooks.org
 deep-level inferences and show what kinds of

inferences they are. My colleagues and I are working

on research to develop an “unconscious conflict”
measure that extends the CCRT to include deeper
levels of inference (see the brief description of this

study in chapter 21).

To Add a Re-Review of the Relationship Episodes by

Steps 1' and 2'

The purpose of the re-review of the relationship

episodes is to achieve a more complete scoring and

formulation of the CCRT by the tailor-made system.

Is the expected gain worth the effort? Why not just do

a collation of the scores from Steps 1 and 2 and stop

there? It would be simpler, more straightforward, and
more like the coding of the original TAT stories by
many quantitative scoring systems.

But the rationale for the re-review and

reformulation required by Steps 1' and 2' comes from
an enduring clinical observation, for example, by

www.theipi.org 352
www.freepsychotherapybooks.org
 Freud (1958d, p. 112): “It must not be forgotten that
the things one hears are for the most part things
whose meaning is only recognized later on.” A basic
theme may not be discerned until a later occasion on
which the same theme is re-presented; the delay

followed by a further review sometimes sparks the

judge’s recognition of a common theme. In terms of

the CCRT scoring, one may not discern a theme until

the step of reviewing across relationship episodes,

because much of the realization of a theme’s

centrality is based on the well-established clinical
inference principle of attending to redundancy. The
following example from Mr. Howard’s treatment

makes this virtue more apparent:

Example: In Session 3, in RE 3 about the

therapist, Mr. Edward Howard suddenly feels he
has nothing to say to the therapist. A judge who
reads all the REs in context reads this
relationship episode immediately after reading
the preceding and before reading the subsequent
relationship episodes and so can readily infer

www.theipi.org 353
www.freepsychotherapybooks.org
 that the experience of not having anything to say
to the therapist could be the patient’s response to
this realization that the therapist did not have
enough to say to him. In fact, the therapist states
exactly that inference in his interpretations to the
patient. The theme is so recurrent that it might
well have been recognized simply on the basis of
Steps 1 and 2, but the review in Steps 1' and 2'
would clinch its recognition.

The possible criticism needs to be considered that

Steps 1' and 2' open the door to the inclusion of

inferences that are not inherent in the relationship

episode by itself. Such a criticism is tenable, but in

fact it does not correspond to the judges’ experience.
Steps 1' and 2’ generally suggest or bolster inferences
that have the same degree of cogency and range of
inference as inferences derived during Steps 1 and 2.
Furthermore, as the example shows, it is not only

Steps 1' and 2' that are vulnerable to this criticism but

also any knowledge derived from another relationship

episode, even in Steps 1 and 2. To be safe from this

criticism, each relationship episode would have to be

www.theipi.org 354
www.freepsychotherapybooks.org
 scored by judges who knew only the single

relationship episode being scored (as in the QUAINT

system discussed in chapter 20, this volume). But

such a procedure would limit the CCRT, because the

inference system for the CCRT was set up to parallel

the experience of the therapist, who is attentive to
redundancy across relationship episodes as the usual
basis for forming inferences during psychotherapy.

Nevertheless, it would be of interest to know (a)

how often inferences are added by Steps 1' and 2' and
(b) how often this additional review of inferences
makes a difference. “Making a difference" might

mean that the correlations with other measures based

on Steps 1 and 2 alone are different from the

correlations based on the addition of Steps 1' and 2'.

To Focus on the Patient’s Perspective

The clinician’s focus should be on inferring the

patient’s perspective in the patient’s narratives, not

www.theipi.org 355
www.freepsychotherapybooks.org
 anyone else s. The carrying out of this basic
injunction fits the definition of a psychoanalytic

method, as Klein (1970) described it: The intention of

the clinician who is following this method is to

achieve understanding of “intentionally” from the

point of view of the patient.

CONCLUSIONS

• This chapter examines the rationales for and

justifies each of the procedures chosen for the
CCRT method. Three basic types of decisions
had to be made: decisions for the database, for
the scoring system, and for the inference level. I
decided to focus on redundancy across narratives
from psychotherapy sessions as the database for
inferring the CCRT. The scoring system was
based on a three-part clinical judgment of the
types of wishes, needs, and intentions; responses
from other; and responses of self, together with
the positivity versus negativity of each response.

• The inference level was to be limited to moderate,

which mainly serves to avoid very high-level
abstractions.

www.theipi.org 356
www.freepsychotherapybooks.org
 • Now, with hindsight, I can see that the rationale for
these decisions was reasonable and useful in
terms of achieving adequate reliability, reported
in chapter 6, and several types of validity,
reported in the next major section on
“Discoveries with the CCRT.” Further research
remains to be done to test the competing options
considered for each of the decisions.

www.theipi.org 357
www.freepsychotherapybooks.org
 II

DISCOVERIES FROM THE

CCRT METHOD

www.theipi.org 358
www.freepsychotherapybooks.org
 9

THE NARRATIVES TOLD DURING

PSYCHOTHERAPY AND THE TYPES OF
CCRTs WITHIN THEM

LESTER LUBORSKY, JACQUES P. BARBER,

PAMELA SCHAFFLER, AND JOHN
CACCIOLA

Each time we have told people about our book’s

topic, their first question has been, “And what are the
types of narratives people tell and what are the types

of CCRTs?” In this chapter, we begin with a

description of the narratives told during

psychotherapy and go on to explore the types of

CCRTs extracted from them.

All psychotherapists know that narratives are

often told during psychotherapy sessions and that

they are clinically very informative. Although the

www.theipi.org 359
www.freepsychotherapybooks.org
 special clinical values of narratives are evident, the
narrative as a unit in psychotherapy had never been

systematically investigated before; until the launching

of the CCRT (Luborsky, Barber, & Diguer, 1992), its

exact formal characteristics (such as frequency,

length, and variety of people in them) remained
unexplored territory. Because narratives are the
database for the CCRT, our focus on the CCRT

requires us to scrutinize the narratives themselves

closely. This chapter sketches a few features of

narratives that are important to the studies in this

book, such as the number and completeness of

narratives, the length of the narratives, and the main

other people in the narratives. This chapter also

examines the frequency of the types of CCRT

components, such as the types of standard category

CCRTs within narratives, the sequential versus the
regular CCRT, the positive versus negative responses

www.theipi.org 360
www.freepsychotherapybooks.org
 within CCRTs, and the possible significance of the
diagnosis of dysthymia for CCRTs.

NUMBER AND COMPLETENESS OF

NARRATIVES

Narratives are common in psychotherapy

sessions: The average session in psychotherapy in the

Penn Psychotherapy Study had 4.1 passably complete

narratives (with a range approximately from 1 to 7

narratives per session). This average is for Sessions 3

and 5, the usual sessions in this book chosen as a

basis for extracting the CCRT early in therapy, and

the estimation of narratives’ completeness is based on

the rules outlined in chapter 2.

It is worth taking a moment to consider why

patients tell so many narratives about relationships

with other people within psychotherapy sessions.

Although the patients themselves have not been asked

this question directly, we do have explanatory leads

that are based on the context in which the narratives

www.theipi.org 361
www.freepsychotherapybooks.org
 are told: (a) Narratives are clearly useful to the
patient as a means of illustrating for the therapist

examples of the patient’s problems. A patient may

say, for instance, “I have a problem with dependency.

Let me tell you this event. ...” (b) The central

relationship pattern and the conflicts within it, as this
book shows, are present within the narratives and
make them memorable to the patient, (c) The

therapist occasionally asks for specific examples. The

first and second of these leads imply that patients find

they can communicate to their therapist the nature of

their problems better through the narrative mode than

through more direct modes of communication;

Bruner (1987) made a similar point about this

property of narratives.

LENGTH OF NARRATIVES TOLD DURING

THERAPY SESSIONS

The length of the usual narrative is now known. A

convenient measure of the length of a narrative is the

www.theipi.org 362
www.freepsychotherapybooks.org
 number of typed lines it takes up in a session
transcript. The average number of lines per narrative

within the early sessions is 51.1, which is about two

double-spaced pages. The range is large: 7-207 lines.

These figures are based on a representative sample of

18 patients from the Penn Psychotherapy sample
(Luborsky, Crits-Christoph, et al., 1988).

MAIN OTHER PEOPLE IN NARRATIVES

The main other types of persons with whom the

teller of the narrative interacts most often, in a sample

of 33 patients, are given in Table 1. The 33 patients
are also a representative sample of the 73 patients in
the Penn Psychotherapy Project (Luborsky, Crits-
Christoph, et al., 1988).

The therapist is often a main other person in the

patients’ narratives. Of the 33 patients studied, 25

(76%) told about the therapist as the main other

person in their narratives, and these 25 patients told a

www.theipi.org 363
www.freepsychotherapybooks.org
 TABLE 1
Frequency of Different Other Persons in Narratives (N = 33)

REs
Patients
(N =
Person (N = 33)
323)
%
%
Therapist 76 16
Family (father, mother, siblings, or 85 27
relatives)
Intimate relation (e.g., spouse) 73 29
For males (5) 56a 13b
For females (19) 79c 35d
Friends (same sex) 30 8
Friends (opposite sex) 15 2
Friends in general 15 2
Authority figures 45 7
Coworkers 15 2
People in general 21 3

Note. RE = relationship episode.

a 9 patients. b 89 REs. c 24 patients. d 234 REs.

www.theipi.org 364
www.freepsychotherapybooks.org
 total of 52 narratives. Of the 10 narratives told by
each patient, a mean of two of these were about the

therapist.

Family members (father, mother, siblings, or

other relatives) made up another frequent type of

other person in the narratives. Twenty-eight (85%) of
the 33 patients told narratives about family members
as the main other person. The total number of
narratives they told was 87, meaning almost 3 out of
10 of their narratives were about family members.

Among the large number of narratives about family

members, when they were broken down into

subcategories, was a relatively small percentage
about siblings, which is surprising in view of the
importance of conflict among siblings as
overwhelmingly demonstrated by Sulloway (1996).

Intimate relationships with other people are an

especially frequent category for female participants.

www.theipi.org 365
www.freepsychotherapybooks.org
 Of the 24 women in this sample, 19 (79%) told

narratives about intimate relationships with others;

there were 82 of these narratives, meaning that

narratives about intimate relationships constituted 4
out of 10 relationship episodes. Narratives about

nonintimate relationships were told less frequently.

We conclude that the distribution of narratives

follows the principle that the more intimate the

relationship with a type of other person, the more

narratives are told about that type of other person.

That principle about intimacy is so obvious it

evokes memories of Groucho Marx’s routine of

challenging his audiences with the question, “Who is

buried in Grant’s tomb?” (Grant, who else?) The

rediscovery of an obvious principle catches our

interest because of its reassuring fit with what was

expected. The principle is consistent with the

meaningfulness of narratives as our database—

www.theipi.org 366
www.freepsychotherapybooks.org
 intimate relationships are important for people; thus,

what is buried in narratives reflects meaningful data.

CCRTs OF PATIENTS

What do patients in psychotherapy want from

other people, how do they expect them to respond,

and how do they react? These familiar questions can

be answered from the narratives patients tell about

their interactions with other people in the course of

psychotherapy sessions and from the CCRTs derived

from these narratives. The purpose of this part of the

chapter is to examine the frequency of different types

of CCRT components and their patterns. Such data
have never been systematically reported before for a

sizable sample of patients in dynamic psychotherapy.

We also searched through the clinical writings of

dynamic therapists to classify the types of
descriptions of central relationship patterns or
transference patterns. Although we found such a

www.theipi.org 367
www.freepsychotherapybooks.org
 formulation in each case study, there is no summary
anywhere of the frequency of different types of

transference patterns.

Nor is there much about central relationship

patterns in Murray’s (1938) book on Thematic

Apperception Test (TAT) stories. His main variables
were needs and presses, with the exception of his
brief suggestion of a central relationship pattern in
terms of his “unity theme” (see chapter 1). The same
applies to other guides to the TAT, such as Tomkins’

(1947), which also presents sets of needs and presses.

We have already observed in the case examples

(in chapter 5, this volume) that the CCRTs derived
from the three specimen patients’ narratives were

very different from each other. Their diversity is

apparent even when one looks only at the main

wishes: Ms. Smyth’s main wish was to be given

support and care; Mr. Howard’s was to be close and

www.theipi.org 368
www.freepsychotherapybooks.org
 not cut off from affection; Ms. Cunningham’s was to

be independent and assertive. This chapter shows

how common these wishes and the responses to them

were among the narratives told by a sample of

patients about their interactions with other people.

We did CCRT analyses on the basis of Sessions 3

and 5 in the sample of 33 patients selected from the
Penn Psychotherapy Project’s sample of 73 patients
(Luborsky, Crits-Christoph, et al., 1988). From these

sessions, approximately 10 narratives were selected

(sometimes a third session was needed to reach a
sample of 10 narratives). These sessions had been

independently scored for their CCRT by two judges

(CP and DM) using the tailor-made system (chapter

2).

We report on the rescoring of the tailor-made

categories by two other independent judges (JC and

PS) using the standard categories. This chapter also

www.theipi.org 369
www.freepsychotherapybooks.org
 presents the frequency of each of the three usual
CCRT components among the 33 patients.

TYPES OF TAILOR-MADE CCRTs WITHIN

NARRATIVES

One of the authors (PS) summarized the tailor-

made CCRT frequencies for the 33 patients as scored

by two judges (CP and DM; see Table 2, first

column). The table includes the frequencies, both the

most frequent and next most frequent, for each

patient for each CCRT component in the narratives.

The summarizing judge tried to remain faithful to the

tailor-made categories offered by each of the original

scoring judges. Even though the tailor-made method
has inherent limits that make comparisons difficult,
the experience of the summarizing judge was that the

categorization usually could be done in a fairly

straightforward way.

We now offer an answer to the age-old question,

“What do people want most from people they interact

www.theipi.org 370
www.freepsychotherapybooks.org
 TABLE 2
Tailor-Made CCRT Categories for the Penn Psychotherapy Project Sample
(N = 33) With Subsamples of Dysthymic (n = 12) and Nondysthymic (n =
21) Groups

Percentage of Patients
Total Sample
Frequent Dysthymic Nondysthymic

Frequent

Frequent

Frequent

Frequent

Frequent
Most

Most

Most

Most

Most

Most
Next

Next

Next
Cluster

Wish
To be close 39 21 50 17 33 24
to other
To assert 30 45 33 75 29 29
myself, be
independent
To get 18 0 0 0 29 0
attention
from other
To be 9 12 8 17 10 10
helped,
taken care
of
To be 9 3 25 8 0 0
accepted
Negative responses from other
Rejects or 33 12 25 8 38 14
criticizes me
Dominates 27 15 42 17 19 14
or controls
me
Distant 18 6 17 0 19 10
Unhelpful or 12 0 0 0 19 0
unreliable
Positive responses from other
Close to me 21 0 25 0 19 0
Likes me 6 0 17 0 0 0
Negative responses of self

www.theipi.org 371
www.freepsychotherapybooks.org
 Angry 36 24 25 25 43 24
Withdrawn, 24 24 33 17 19 29
distant
Feel 21 21 33 33 14 14
inadequate,
helpless
Self- 18 15 25 17 14 14
blaming
Positive responses of self
Close to 15 0 17 0 14 0
other
Assertive 9 0 8 0 10 0
Like other 9 0 8 0 10 0

www.theipi.org 372
www.freepsychotherapybooks.org
 with?” The most frequent wishes are listed in Table 2

and are summarized here; in parentheses is the

number of patients having each wish: to be close
(13), to assert myself (10), and to get attention and

interest from other (6). The next most frequent wishes

expressed by patients were also to be close (5) and to
assert myself (6). As would be expected, the next
most frequent categories tended to be similar to the

most frequent.

The most frequent responses from other were

negative: rejects or criticizes me (11) and dominates

or controls me (9). The positive responses from other

were moderately frequent. The two most frequent

negative responses of self were angry (12) and

withdrawn, distant (8).

TYPES OF STANDARD CATEGORY CCRTs

WITHIN NARRATIVES

To ease the task of summarizing the tailor-made

scoring done by different judges, we developed the

www.theipi.org 373
www.freepsychotherapybooks.org
 lists of standard categories presented in chapter 3, this

volume. The one used here is the result of a cluster

analysis of the standard categories of Edition 2,

described in chapter 3.

The agreement in the translation into standard

categories by the two judges is moderately good.

More specifically, the kappas—that is, the chance-

corrected agreement between the two judges—were

.59 for wishes, .60 for responses from others, and .59

for responses of self. Table 3 reports the frequencies

in standard categories for all 33 patients combined

and for the two diagnostic groups within the sample.

The same age-old question can now be answered

in terms of standard categories: The most frequent
wishes were “to be close and accepting” (13), “to be
loved and understood” (12), and “to assert self and to
be independent” (11). The first and second most

www.theipi.org 374
www.freepsychotherapybooks.org
 TABLE 3
Clustered Standard Categories Within Narratives for Penn
Psychotherapy Project Sample (N = 33) With Subsamples of
Dysthymic (n = 12) and Nondysthymic (n = 21) Groups

Percentage of Patients
Cluster Total Dysthymic Nondysthymic
Wishes
1. To assert self and be 33 42 29
independent
2. To oppose, hurt, or 18 25 14
control others
3. To be controlled, hurt, or 24 25 24
not responsible
4. To be distant and avoid 27 33 24
conflicts
5. To be close and 39 42 38
accepting
6. To be loved and 36 50 29
understood
7. To feel comfortable and 15 17 14
good
8. To achieve and help 18 17 19
others
Responses from other
1. Strong 3 8 0
2. Controlling 36 42 33
3. Upset 27 17 33
4. Bad 6 0 17
5. Rejecting and opposing 73 58 81
6. Helpful 6 0 17
7. Likes me 6 17 0

www.theipi.org 375
www.freepsychotherapybooks.org
8. Understanding and 9 8 17
accepting
Responses of self
1. Helpful 9 17 8
2. Unreceptive 42 25 52
3. Respected and 6 17 0
accepted
4. Oppose and hurt others 3 0 8
5. Self-controlled and self- 3 8 0
confident
6. Helpless 36 33 38
7. Disappointed, 45 33 52
depressed, angry
8. Anxious and ashamed 21 0 33

www.theipi.org 376
www.freepsychotherapybooks.org
 frequent wishes, “to be close” and “to be loved,”

undoubtedly have much in common.

By far the most frequent responses from others

were “rejecting and opposing” (14) and “controlling”

(12). By far the most frequent responses of self were

“disappointed and depressed” (15), “unreceptive”
(14), and “helpless” (12). Taken together, the results
of these analyses show that much of what appeared
by the tailor-made method appears again by the

standard categories method (see Table 4).

CCRT SEQUENCES OF COMPONENTS FOR

THE WISH TO BE CLOSE VERSUS THE WISH
TO BE INDEPENDENT

We report here on one of the earliest systematic

attempts to examine the sequence of the CCRT

components as they appear in the patients’ sessions;
these results build on the earlier work of Luborsky
(1984, p. 202) and Barber (1989). The usual CCRT
method looks only at the frequency of CCRT

www.theipi.org 377
www.freepsychotherapybooks.org
 TABLE 4
Most Frequent Tailor-Made and Standard Clustered Categories
for the Penn Psychotherapy Project Sample (N = 33)

Standard
Tailor-Made Percentage Percentage
Cluster
Wishes
To be close to 39 To be close and 39
other accepting
To assert myself, 30 To assert self 33
be independent and be
independent
To get attention 18 To be loved and 36
from other understood
Responses from other
Rejects or 33 Rejecting and 73
criticizes me opposing
Dominates or 27 Controlling 36
controls me
Close to me 21 Upset 27
Responses of self
Angry 36 Disappointed, 45
depressed,
angry
Withdrawn, 24 Unreceptive 42
distant
Feel inadequate, 21 Helpless 36
helpless

www.theipi.org 378
www.freepsychotherapybooks.org
 components without taking into account their

sequence. Because we could not look at the

sequences of CCRT components that follow all

wishes, we decided to focus only on the wish to be

close and the wish to be independent. Not only are

these two wishes the most common, but also they
carry much theoretical weight. To make this task
more realistic, we considered ratings belonging to the

clusters “to be loved’ and “to be close” as

interchangeable because their meaning is similar (see

the clustered standard categories called Edition 3 in

chapter 3).

A judge (PS) did a sequence analysis for these

wishes in all relationship episodes, noting the

responses from others and responses of self that came

just after these wishes. In many cases, we could only

find either responses from others or responses of self;
in a few cases we found both or neither. Because
there were many relationship episodes in which

www.theipi.org 379
www.freepsychotherapybooks.org
 several wishes were present or the same wish was
present on numerous occasions, we kept the most

complete sequence; when there was more than one

complete sequence or there were two incomplete

sequences, we kept the first identified sequence. This

selection process was performed in order to keep only
one main sequence for each relationship episode, that
is, the most frequent combinations of wishes and

responses from others or responses of self. The

results of this analysis are presented separately for

each of the two types of wishes.

The Wish to Be Close or Be Loved

This wish was the main wish of 13 patients. It

was followed by the response from others of

“rejecting and opposing” in 24-4% of the relationship

episodes in which the wish itself was expressed; the

sequence was in 67% of the relationship episodes in

which the wish was followed by any responses from
others.

www.theipi.org 380
www.freepsychotherapybooks.org
 The wish to be close or loved was followed by the

response of self “helpless” or “disappointed or

depressed” in 44.2% of the relationship episodes in
which the wish was expressed; the sequence was in

50% of the relationship episodes in which the wish

was followed by any responses of self. The following
breakdown of the RE components makes even clearer
the major sequences and subdivisions in these results:

The wish to be close or loved was followed by the

response of self “helpless,” “disappointed or

depressed,” or “ashamed or anxious” in 62% of the
relationship episodes in which the wish was
expressed; the sequence was in 76% of the

relationship episodes in which the wish was followed

by any responses of self.

www.theipi.org 381
www.freepsychotherapybooks.org
 Regarding the complete sequence of wish,
response from others, and response of self, the wish

to be close or loved was followed by the response

from others “rejecting and opposing” and responses

of self “helpless,” “disappointed or depressed,” or

“ashamed or anxious” in 11.6% of the relationship
episodes in which the wish was expressed; the
sequence was in 55.4% of the relationship episodes in

which the wish was followed by any responses from

others and responses of self. The findings regarding

the complete sequence are even more impressive if
one considers the fact that only 8 of the 13 patients

had any relationship episode with such a complete

sequence.

The Wish to Be Independent

This wish was the main wish of 7 patients. It was

followed by the response from others “rejecting” in

25.4% of the relationship episodes in which the wish

was expressed; the sequence was in 47.6% of the

www.theipi.org 382
www.freepsychotherapybooks.org
 relationship episodes in which the wish was followed

by any responses from others.

The wish to assert oneself and to be independent

was followed by the response of self, “helpless” or

“disappointed or depressed” in 42.7% of the

relationship episodes in which the wish was
expressed and in 47.3% of the relationship episodes
in which the wish was followed by any responses of
self.

The wish to assert oneself and to be independent

was followed by the response of self, “helpless,”

“disappointed or depressed,” or “ashamed or

anxious” in 57% of the relationship episodes in which

the wish was expressed and in 61.6% of the

relationship episodes in which the wish was followed
by any responses of self. The wish to assert oneself
and to be independent was followed by the response

of self “unreceptive” in 17.9% of the episodes in

www.theipi.org 383
www.freepsychotherapybooks.org
 which the wish was expressed and in 19.4% of the

episodes in which the wish was followed by any

responses of self.

The wish to assert oneself and to be independent

was followed by the responses from others

“rejecting,” “upset,” or “bad” in 34.6% of the
relationship episodes in which the wish was
expressed; the sequence was in 64.3% of the
relationship episodes in which the wish was followed

by any responses from others. If we added the

response from others “controlling,” these responses
from others followed the wishes in 41.5% of the

relationship episodes in which the wishes were

expressed and in 78.5% of the relationship episodes

in which the wishes were followed by any responses

from others.

However, only 3 patients had the complete

sequence of the wish, followed by responses from

www.theipi.org 384
www.freepsychotherapybooks.org
 others, and followed by responses of self.

COMPARISON OF THE SEQUENTIAL CCRT

AND THE REGULAR CCRT

According to the frequency analysis of the CCRT,

as noted earlier, the most frequent wish was “to be

loved”; the more frequent responses from others were

“rejecting” and “controlling”; and the most frequent

of the responses of self were “disappointed and

depressed,” “unreceptive,” and “helpless.” In the

sequential analysis, we found that the same wish was

actually followed by a “rejecting and controlling”

response from others and by a “helpless,”

“disappointed and depressed,” or “ashamed and
anxious” response of self in 8 of 13 patients who
provided relationship episodes including the three

CCRT components. The same analysis does not seem

worth doing in this sample for the wish to be

independent because only 3 patients had a complete

www.theipi.org 385
www.freepsychotherapybooks.org
 sequence of CCRT components involving this wish in

their relationship episodes.

In conclusion, our results so far suggest that the

usual procedure of a mere compilation of CCRT

components generally yields similar results to those

obtained with a sequential analysis of the CCRT
components. Recently, a more formal sequence-of-
components CCRT method has been developed to
enable more of such analyses (Dahlbender, Albani, et
al., in press; Dahlbender, Kurth, Stübner, Kalmykova,

& Pokorny, in press).

POSITIVE VERSUS NEGATIVE RESPONSES

WITHIN CCRTs

We also scored each type of component for its

positive or negative quality. The idea for scoring the
positive or negative quality within the CCRT came
from Freud’s (1912/1958a) designation of positive or

negative in his transference formulations. In our

operational measure of positive and negative, positive

www.theipi.org 386
www.freepsychotherapybooks.org
 is defined from the patient’s point of view as

noninterference or expectation of noninterference

with the satisfaction of wishes; negative is defined

from the patient’s point of view as an interference or
expectation of interference with satisfaction of the

wishes. We scored each response from other and

response of self as either mainly positive or mainly

negative. But because wishes are not easily assigned

a positive or negative rating, we did not do such

scoring for wishes. A wish to be close, for example,

is not in itself either positive or negative. It is its
association with the responses that tends to give it a
positive or negative quality.

We found that independent judges can reliably

assign either a positive or negative score. The

agreement of the two judges was 95% for both the

responses from others and the responses of self. (The

two judges used here were the two who translated the
tailor-made into the standard categories.) The study

www.theipi.org 387
www.freepsychotherapybooks.org
 by Grenyer and Luborsky reported in chapter 4, this
volume, found a similar high level of agreement.

The results shown in Table 5 in terms of

percentages of positive and negative CCRT responses

constitute a remarkable set of findings. There is an

overwhelming trend for people to tell narratives about

others reflecting negative rather than positive patterns

in their responses from others and responses of self:

Negative responses from others and self were found

in 81.5% and 88.5% of the responses, respectively;

positive responses from others and self were found in

only 14.4% and 10.7%, respectively. Does this

phenomenon occur because these are patients in
psychotherapy who might be expected to tell negative
narratives? Probably not, but more data are needed,

including related data from other groups, before this

question is answered. Another interpretation of these

results is that negative interactions are more

memorable because they deal with relationship

www.theipi.org 388
www.freepsychotherapybooks.org
 TABLE 5
Positive Versus Negative CCRT Components Related to the
Therapist or Others

Therapist (52 REs) Others (271 REs)

Percentage

Percentage
Frequency

Frequency
Type of
Responses

Negative 29 55.8 220 81.2

responses
from others
Negative 32 61.5 239 88.2
responses
of self
Positive 6 11.5 39 14.4
responses
from others
Positive 7 13.4 25 9.2
responses
of self

Note. RE = relationship episode.

www.theipi.org 389
www.freepsychotherapybooks.org
 interactions that are harder to master. Remembering

them and telling them are in the service of efforts at

mastery (Loevinger, 1976; White, 1952). Such
possible meanings of these results are discussed

further in chapter 22.

We also found an interesting property of the

positivity and negativity of narratives about the
therapist (Table 5). Although for the narratives about
the therapist (n = 52) the percentages of positive
responses from other and positive responses of self

are somewhat similar to each other, they are higher

than in the narratives told about other people. Perhaps

the relationship with the therapist is less negative;

alternatively, it may be harder to tell something

negative about the person one is speaking to than to

tell something negative about someone not present.

Freud (1912/1958a) had the latter alternative in mind
when he said that it is hard to talk about aspects of

www.theipi.org 390
www.freepsychotherapybooks.org
 the negative transference when it involves someone

—the therapist—who is directly present.

DIFFERENCES IN CCRTs BY DIAGNOSIS

(DYSTHYMIA AND NONDYSTHYMIA)

The main results we have reported so far are for

the entire sample of 33 participants, but this group

included 12 patients with dysthymia who may have

differed from the remaining 21. Because of the small

number of dysthymics, we discuss here only the

results from the clustered standard categories (see
Table 3) in which percentage differences between
patients with and without dysthymia are very large

(two times in size or close to it). Among the

differences for the wishes, the most impressive is the

larger percentage of dysthymics (25% vs. 14%) with

the wish to oppose, hurt, or control others. Among

the responses from others, the other is seen less often
as “upset” by those with dysthymia (17% vs. 33%).
Among the responses of self, “unreceptive” is less

www.theipi.org 391
www.freepsychotherapybooks.org
 evident for the dysthymics (25% vs. 52%). In
essence, patients with dysthymia more often see
themselves as wishing to oppose, hurt, or control the
other; they see the other as less upset; and they see
themselves as more responsive (more receptive) than

nondysthymic patients.

RESULTS, DISCUSSION, AND CONCLUSIONS

• A novel piece of information about relationship

episodes reveals just how common they are in
every psychotherapy session: We found 4-1 at
least fairly complete narratives per session, on
the basis of the sample from Sessions 3 and 5.
Each narrative took up a mean of about 51 lines,
which is approximately two double-spaced
pages. The main other people with whom the
narrator of the relationship episode interacted
were, in the following order, family members,
the therapist, and other intimate relationships.

• Certain wishes and responses were moderately

frequent in the narratives told by patients during
psychotherapy (in our sample of 33 patients), on
the basis of either the tailor-made or the standard

www.theipi.org 392
www.freepsychotherapybooks.org
 category methods. Because a standard category
method is more likely to produce reliable results
than a tailor-made method, we concentrated in
this chapter on the standard category clustered
results using Edition 3.

• These results were summarized in terms of the

CCRT components. The two most frequent
wishes expressed by our patients were “to be
close” and “to assert myself and to be
independent.” The most frequent of the responses
from other were “rejecting” and “controlling”;
the most frequent of the responses of self were
“disappointed and depressed,” “unreceptive,”
and “helpless.”

• We reported on a pioneer analysis of the CCRT

components that followed the two most common
wishes: “to be close” versus “to be independent.”

• It is interesting to notice the degree to which the

most common wishes are socially acceptable (see
both the tailor-made and standard category
results in Tables 2 and 3). Almost never is there
an example of a strongly socially unacceptable
wish such as to steal or to murder someone. Our
impression is that this tendency to tell socially

www.theipi.org 393
www.freepsychotherapybooks.org
 acceptable wishes has most to do with the fact
that these narratives are told by the narrator
about the narrator’s own relationships with
others; from their own point of view, their wishes
are socially acceptable. A few socially
unacceptable themes tend to appear in the
response of self. A paraphrase of the narrative
from this point of view is that the person is
saying “I want socially acceptable things and
others impede me or hurt me, so I sometimes end
up with socially unacceptable responses.”

• It is impressive to find the high frequency of

negative responses in contrast to positive
responses (as we discuss further in chapters 4
and 22, this volume); it would be useful to have
even more comparison groups to know how to
evaluate the frequency of these responses.

• We now know from the data presented in this

chapter how common the wishes and responses
were for the three patients whose CCRTs serve as
illustrations throughout this book. On the basis of
the frequencies of the clustered standard
categories (Table 3), Ms. Smyth’s main wish, “to
be helped,” fits with only part of Cluster 3 and is
therefore relatively infrequent. Mr. Howard’s

www.theipi.org 394
www.freepsychotherapybooks.org
 wish fits with the most common wish, Cluster 5,
to be close and accepting. Ms. Cunningham’s
main wish fits with the next most frequent wish,
Cluster 1, to assert self and to be independent.

• As we said at the start, there are no surveys of

frequencies of types of CCRTs or types of
transference patterns in clinical sources, although
we have general impressions from a review of
Freud’s case histories (e.g., 1893-1895, 1901-
1905) and many other clinical case studies.

In terms of source of the transference patterns,

we know that clinicians tend to refer to negative
transference as modeled on the patient’s early
relationship with the father or mother (as in
Freud’s Observation 10; see chapter 21, this
volume). Examples of such interpretations are
not difficult to find in the cases used for
illustration in chapter 5: (a) In Mr. Howard’s
Session 3, for the relationship episode about the
therapist, the patient describes feeling “generally
unresponsive, I’m getting a headache, tense. ...”
The therapist interprets the negative relationship
with the therapist by conveying the idea that the
patient is expecting to be disappointed by the
therapist’s response (with the implication that

www.theipi.org 395
www.freepsychotherapybooks.org
 this was the disappointment he had also felt with
his mother), (b) A similar interpretation of the
negative experience in the relationship with the
therapist is present in the case of Ms.
Cunningham. After the patient speaks about
needing reassurance from her husband, the
therapist says, “You didn’t express wanting
reassurance here yesterday.” Such interpretations
of the negative experience of the relationship
with the therapist are not uncommon. They fit
Freud’s (1913/1958c) recommendations to
therapists to deal with negative transference
manifestations.

In terms of content of the most frequent

patterns, many clinicians believe that the most
common transference pattern is one involving the
oedipal triangle. Yet the oedipal theme is not
immediately obvious in our three main examples
of early sessions (chapter 5), and it is not obvious
in the other 30 cases of the Penn Psychotherapy
Project sample discussed in this chapter. Instead,
the oedipal theme is often indirectly expressed in
the early sessions, and it becomes more directly
exposed in the later sessions. This sequence of
indirect expression becoming more direct later in

www.theipi.org 396
www.freepsychotherapybooks.org
 the therapy fits the data from Mr. Howard. His
CCRT derived from the early sessions reflected a
strong wish for closeness and affection from his
mother; the same wish appeared later in the
treatment in ways that more directly reflected an
oedipal theme. In addition, a version of the
oedipal theme often appears in the CCRT, when
it is either directly or indirectly expressed, as we
saw for Ms. Cunningham in the late sessions.

• We conclude that the two most frequent types of

wishes, to be close and to be independent, are
probably associated with each other in the sense
that they often conflict, probably even in people
in our society who do not have psychiatric
diagnoses. The likelihood of spawning symptoms
from this conflict is related to (a) the conflict’s
intensity and (b) the fears aroused by thoughts or
behaviors that coincide with the expression of
one or the other side of the conflict. A good
example of the second source came when Ms.
Cunningham (in Session 5) reached out and
stroked the professor’s tie to express admiration
for it. This impulsive break from her usual
control increased her anxiety and her symptoms

www.theipi.org 397
www.freepsychotherapybooks.org
 of inhibition and constriction, the symptoms that
had served to forestall such breaks.

This view about the likelihood of conflict

between these wishes is consistent with a review
by Bonanno and Singer (1990). They see the
wishes “to be close” and “to be independent” as
an expression of an inherent conflict between
two classes of wishes: a basic desire to blend
with another person versus a desire to have
individuality, power, and competence. Their
review implies that maladaptive behavior may
appear as one moves closer to one extreme at the
expense of the other. Further evidence of the
meaningfulness of the conflict between these two
types of wishes comes from a study (Luborsky,
Crits-Christoph, & Alexander, 1990) relating
CCRT wishes and repressive style measures.
Participants who scored low on the Weinberger
measure of repression (1990) tended to score
high on the wish to receive affection (which is
related to the wish to be close) and low on
wishes for dominance and competitiveness
(which are related to the wish for independence).

• Our findings must reflect our sample of

nonpsychotic outpatients in dynamic

www.theipi.org 398
www.freepsychotherapybooks.org
 psychotherapy (Luborsky, Crits-Christoph, et al.,
1988). We still do not have data from other
diagnostic groups to show the distinctiveness of
the CCRT components. Although the CCRT
Newsletter (Luborsky et al., 1997) lists 26
studies of diagnostic groups, none has yet been
published. In time we will have results from at
least three of such groups: (a) normal persons
without any psychiatric diagnosis (Demorest et
al., 1997), (b) patients with borderline
personality disorder who are in dynamic
psychotherapy (Lefebvre, Diguer, Morissette,
Rousseau, & Normandin, 1996), and (c) patients
with major depression who are in dynamic
psychotherapy (Luborsky, Diguer, et al., 1996).

www.theipi.org 399
www.freepsychotherapybooks.org
 10

CHANGES IN CCRT PERVASIVENESS

DURING PSYCHOTHERAPY

PAUL CRITS-CHRISTOPH AND LESTER

LUBORSKY5

In case descriptions by psychodynamic writers, it is

the relationship conflicts that are pointed out as the

fomenters of symptom outbreaks. When the symptoms

decrease, these relationship conflicts are seen as having

become less pervasive. To become less pervasive means

that the relationship conflicts appear in fewer narratives

about interaction with other people. This sequence of

reduction in generality of relationship conflicts across

relationship episodes, followed by reduction in

symptoms, is illustrated in the examples we gave of

improvement in patients assessed before versus after

psychotherapy (chapter 5).

www.theipi.org 400
www.freepsychotherapybooks.org
 Example: The following CCRT emerged from a
review of Mr. Howard’s narratives of interactions
with other people: (a) I wish not to be cut off from
closeness and affection; (b) I expect the other person
to cut me off from closeness and affection; (c) I
respond by feeling rejected and becoming angry,
self-blaming, and highly anxious, that is, by
developing a symptom. In this example the main
relationship conflict is between the wish and the
expected response from the other person. The
patient’s relationship conflicts were considerably
better controlled by the end of psychotherapy,
although when they reappeared from time to time the
patient’s anxiety symptom reemerged, as in the
episode reported 8 years later (described at the end
of chapter 7, this volume), when the patient’s
supervisor at first appeared to him to be cutting him
off from approval. On that occasion the fomenters of
the patient’s relationship conflict were reactivated
and anxiety symptoms started to develop, but he
showed greater mastery, for he had acquired a greater
capacity to recognize the reappearance of a familiar
relationship conflict and then even to see the humor
in his mistaken perception that almost set off its
recurrence.

The decreased pervasiveness of the conflictual

relationship patterns appears to operate as a curative

factor through fostering reduction of the symptoms, and

www.theipi.org 401
www.freepsychotherapybooks.org
 the amount of change in itself is a theoretically relevant

measure of the outcome of dynamic psychotherapy. One

of the most significant needs of research on

psychotherapy, particularly dynamic psychotherapy, is for

measures of the outcome of therapy that are relevant to

the theory of psychodynamic change. Both the

behaviorists and supporters of the Diagnostic and

Statistical Manual (DSM) approach have pulled the field

of psychotherapy research toward the use of bare

measures of overt behavior and symptoms as outcome

criteria. For psychodynamic psychotherapies, the lack of

a reliable and valid measure of psychodynamic change

has forced researchers to rely on other types of

assessment, such as general symptom inventories or

global ratings of improvement. Although such measures
have the virtue of simplicity and applicability to many

different kinds of treatment, they are not derived from

sound theory. For psychoanalytic psychotherapy, relevant

measures include change in the main conflictual

www.theipi.org 402
www.freepsychotherapybooks.org
 relationship pattern and changes in the patient’s

awareness of this pattern.

The work by Malan and his associates (Malan, 1963;

Malan, Bacal, Heath, & Balfour, 1968) is an exception to

the trend because it deals with theory-related outcome

measures. Malan has argued against the reliance on

behavioral manifestations alone and has developed an

individualized method of assessment that is guided by

psychodynamic hypotheses. In brief, his method involves

(a) an initial detailed account of the patient’s presenting

problems, (b) a consensus formulation of a dynamic

hypothesis by a group of clinicians, and (c) the

specification of emotional and behavioral changes that

would indicate favorable outcome for each case.

Posttreatment assessment is based on a clinical interview.

An account of the interview is given to the group of

clinicians so that they can rate improvement on a global

9-point scale. However, problems with Malan’s method

have been detailed by Mintz (1981); they include the lack

of reliability and validity information about the

www.theipi.org 403
www.freepsychotherapybooks.org
 psychodynamic hypothesis and the reliance on one

interviewer’s account after treatment rather than

independent assessment of outcome by different

clinicians. Using Malan’s published case reports, Mintz

(1981) has also demonstrated that a large component of

the Malan outcome rating is simple symptomatic

improvement that can be readily assessed by nonclinician

judges.

Other individualized methods of outcome assessment

have been developed, such as the Target Complaints

method (Battle et al., 1966) and Goal Attainment Scaling

(Kiresuk, 1973), but these measures are not

psychodynamically based and have been criticized on

psychometric grounds (Mintz & Kiesler, 1982). Still

other measures go beyond assessment of symptoms (e.g.,

Weiss, DeWitt, Kaltreider, & Horowitz, 1985), yet these

methods do not assess the types of individual themes and

conflicts that are important in psychoanalytic

psychotherapy.

www.theipi.org 404
www.freepsychotherapybooks.org
 It is clear that the major stumbling block to the

development of a measure suited to the evaluation of

outcomes of psychoanalytic psychotherapy has been the

lack of a reliable and valid measure of the nature of the

patient’s particular dynamic conflicts and themes.

Without a measure of the relevant conflicts for each

patient, researchers cannot determine whether

improvement that is consistent with the theory and

techniques of the therapy has occurred.

Several studies (Seitz, 1966; DeWitt, Kaltreider,

Weiss, & Horowitz, 1983) comparing independent

clinicians’ formulations of patients’ dynamic themes

have reported a lack of consensus among clinicians in

unguided judgments of such themes. But more recently,

guided clinical case formulation methods that are applied

to psychotherapy session material have arrived on the

scene and appear to be more promising. There are 17 of

these alternative methods listed in chapter 20, this

volume, that have appeared since information about the

www.theipi.org 405
www.freepsychotherapybooks.org
 Core Conflictual Relationship Theme (CCRT) method

was first published in 1976.

The concept of pervasiveness as the recurrence of

components across narratives is at the heart of the CCRT

method, as first described by Luborsky (1976, 1977b).

Because the focus of dynamic psychotherapy tends to be

on themes that are maladaptive, repetitive, and

inappropriately applied, we propose that one index of

change in dynamic therapy is the extent to which the

maladaptive theme becomes less pervasive in the

relationships of a patient by the end of treatment. For the

CCRT method, this concept translates into a decrease

from the beginning to the end of treatment in the

percentage of relationship episodes in which the

maladaptive theme is present.

The purposes of the study presented here were (a) to

assess the agreement of independent judges on our

measure of change based on the CCRT, (b) to compare

results from applying the measure to sessions early and

late in treatment, and (c) to examine the relationships

www.theipi.org 406
www.freepsychotherapybooks.org
 between the psychodynamic measure of change in the

CCRT and the more conventional measures of change in

self-reported symptoms (Derogatis, Lipman, Covi,

Rickels, & Uhlenhuth, 1970) and change on a clinician-

rated psychological health-sickness scale (Luborsky,

1962, 1975; Luborsky & Bachrach, 1974; Luborsky,

Diguer, et al., 1993).

PROCEDURE

Combination of Judges’ CCRT Formulations

For each case the final CCRT selected for inclusion in

the study was a composite of two judges’ independent

CCRT formulations. It included five components: wish,

negative response from other, negative response of self,

positive response from other, and positive response of

self. Each of the two judges’ tailor-made formulations for

each type of component was coded into standard

categories (Edition 1) to permit direct comparisons

between their formulations. This task of coding the tailor-

made into the standard categories was done with greater

than 95% agreement between judges.

www.theipi.org 407
www.freepsychotherapybooks.org
 For each case, each judge’s CCRT formulation, as

coded into standard categories, was examined. A

composite CCRT was derived by selecting wishes and

responses that were in common among the two judges’

listings of the most frequent CCRT components.

Frequency scores were derived for each type of

component by taking the average of the different judges’

frequency scores.

Finally, these average frequency scores were divided

by the number of relationship episodes used for each case

to derive a percentage score, which we have termed the

CCRT pervasiveness score or the pervasiveness of

conflicts across the relationship episodes. Because this

score is a central measure in the current study, we restate

it here for the sake of clarity (RE = relationship episode):

Number of REs that include the CCRT component

CCRT = _________________________________________
pervasiveness
Total REs in the session or sessions

www.theipi.org 408
www.freepsychotherapybooks.org
 This measure is built on a theory-derived expectation

of the gains from successful psychoanalytic

psychotherapy: The main conflictual relationship patterns

should become less pervasive across the relationship

episodes because the patterns should become less

stereotyped as more relationship options are opened up to

thought and expression.

Although multiple wishes and responses are generally

evident for each patient, the categories with the highest

frequency for each of the five types of CCRT components

were chosen for study. Because we were examining

change on these components, it was necessary to use the

percentage score for the same thematic category both

early and late in treatment, although the same category

was not necessarily the one with the highest frequency at

both times (for example, if “anxious” was the most

common negative response of self early in treatment, it

was not always the most common response of self late in

treatment). For the wish, negative response from other,

and negative response of self, the early-in-treatment

www.theipi.org 409
www.freepsychotherapybooks.org
 category with the highest frequency was selected as the

focus to examine change. For the positive response from

other and positive response of self, the late-in-treatment

category with the highest frequency was chosen first, and

then the frequency of this same component early in

treatment was noted.

A modified version of the Hopkins Symptom

Checklist, the Symptom Checklist (SCL; Derogatis et al.,

1970), containing 85 items, was used as a general

measure of level of self-reported symptoms. The measure

was obtained from patients before the start of therapy and

at termination. The internal consistency (Cronbach’s

alpha coefficient) of this measure was .96.

Patients and Therapists

The sample in this study was 33 patients chosen to

approximate the range of improvement in the larger

sample of 73 patients in the Penn Psychotherapy Project.

The sample consisted of 8 men and 25 women, with a

median age of 24 years. The DSM-III diagnoses included

www.theipi.org 410
www.freepsychotherapybooks.org
 dysthymic disorder (13), generalized anxiety disorder (7),

schizoid personality disorder (7), and histrionic

personality disorder (4); the rest of the diagnoses are

represented by only 1 to 3 patients each. All patients were

nonpsychotic.

A total of 25 therapists (all psychiatrists) treated the

33 patients, with each therapist generally working with

only one or two patients. The therapists ranged in age

from 26 to 55 years, with a median age of 34. Thirteen of

the therapists were 4th-year psychiatric residents who

were supervised by experienced clinicians. Eight

therapists had fewer than 10 years of postresidency

experience, and four therapists had more than 10 years of

postresidency experience.

Treatment, Sessions, and Judges

The patients were in psychodynamic psychotherapy,

attending once or twice per week; two patients attended

four sessions per week of psychoanalysis. Treatment

www.theipi.org 411
www.freepsychotherapybooks.org
 length varied from 21 to 149 weeks, with a median length

of 43 weeks.

Sessions drawn from the early and later parts of

treatment were used to score the CCRT. The number of

early and late sessions was a function of the number of

sessions needed to obtain the minimum of 10 relationship

episodes. This was generally two sessions, but for a few

patients it was three or four.

The early sessions used were typically Sessions 3 and

5. For the late sessions we chose to avoid sessions close

to termination so that issues related to termination (such

as resurgence of symptoms) would not affect our data. On

the average, the late sessions represented the point of

treatment at which 90% of sessions had been completed.

Trained judges working independently were used for

each task. The judges were clinical psychologists,

psychiatrists, and research assistants highly familiar with

the methods.

RESULTS

www.theipi.org 412
www.freepsychotherapybooks.org
CCRT Pervasiveness Scores Were Highly Reliable

The agreement between judges in the pervasiveness

of the main CCRT components was examined by

calculating intraclass correlation coefficients for early-

session data in which the same two judges had scored the

cases. The results indicated relatively high agreement

between the two judges; pooled judges’ intraclass

correlations were as follows: wish, .82; negative response

from other, .90; negative response of self, .80; positive

response from other, .84; positive response of self, .85.

Judges’ pervasiveness scores were therefore averaged for

all subsequent analyses.

Measures of CCRT Change Were Low to Moderately

Intercorrelated

To find the degree of intercorrelation among change

scores on each of the CCRT pervasiveness measures,

Pearson correlations were computed (see Table 1). Of the

10 intercorrelations, 3 were statistically significant: (a)

residual gain on the wish measure was significantly

correlated with residual gain on the negative response of

www.theipi.org 413
www.freepsychotherapybooks.org
 TABLE 1
Intercorrelations of CCRT Pervasiveness Change Measures
(Residual Gain Scores)

Negative Negative Positive Positive

Dimension
RO RS RO RS
Wish .25 .45** -.24 -.24
Negative response .28 -.52** -.28
from other
Negative response of -.22 -.16
self
Positive response .41*
from other

Note. Residual gain is corrected for initial level of the variable. RO =

response from other; RS = response of self.

*p < .05. **p < .01.

www.theipi.org 414
www.freepsychotherapybooks.org
 self scores, r = .45, p < .01 (the residual gain is the gain

corrected for initial level); (b) change in pervasiveness on

the positive response from other dimension was related to

change on the negative response from other measures, r =

–.52, p < .01; and (c) change in positive response from

other was related to change in positive response of self, r

= .41, p < .05.

CCRT Pervasiveness Was Greatest for the Wishes

The early and late pervasiveness scores for the CCRT

were moderately high (see Table 2). When we looked at

each CCRT component separately, we found the wishes

to be much more pervasive than the responses. This was

true for the wishes at both the early (66.3%) and late

(61.9%) points in the psychotherapy; the negative

responses averaged only a little more than half the

pervasiveness of the wishes. The positive responses were

much less pervasive than the negative ones.

CCRT Pervasiveness Decreased From Early to Late in

Therapy

www.theipi.org 415
www.freepsychotherapybooks.org
 To examine differential change across the measures,

the early treatment CCRT pervasiveness and late

treatment CCRT pervasiveness on each of the five CCRT

measures were subjected to a two-factor repeated-

measure analysis of variance (ANOVA). One factor,

Measure, had five levels corresponding to the five types

of pervasiveness measures (wish, negative response from

other, negative response of self, positive response from

other, and positive response of self), and a second factor,

Time, consisted of the early—late dimension. The

interaction term of primary interest, Measure by Time,

addressed the question of differential change across the

five CCRT measures.

The ANOVA produced statistically significant main

effects for both Measure, F(4, 128) = 93, p < .001, and
Time F(1, 32) = 7.4, p < .01. In addition, the Measure by

Time interaction was highly significant, F(4, 128) = 12.6,

p < .001, indicating that the early—late changes were not

uniform across the five measures.

www.theipi.org 416
www.freepsychotherapybooks.org
 Table 2 presents mean early and late pervasiveness

scores on each of the five CCRT measures. For each

measure, the statistical significance of the early-to-late

changes was tested by a paired t test. Small but reliable

changes occurred on four of the five CCRT measures.

The pervasiveness of the CCRT main wish decreased

nonsignificantly from 66.3% to 61.9% over the course of

treatment. Changes on the negative response from other

(12.2% decrease), negative response of self (18.9%

decrease), and positive response from other (10.1%

increase) dimensions were all highly significant (p <

.001, two-tailed). Change on the positive response of self

score (5.7% increase) was also significant (p = .055, two-

tailed).

Some Early-in-Therapy Pervasiveness Scores Were

Related to Initial Symptoms and to Initial Health–Sickness

Pearson correlation coefficients were calculated

between the five CCRT pervasiveness measures and (a)

initial symptoms as measured by the total score on the

Hopkins Symptom Checklist and (b) initial scores on the

www.theipi.org 417
www.freepsychotherapybooks.org
 TABLE 2
Mean Percentages Early and Late in Treatment for CCRT
Pervasiveness Scores (N = 33)

Dimension Early SD Late SD

Wish 66.3 15 61.9 25
Negative response from other 40.7 14 28.5 18
Negative response of self 41.7 14 22.8 18
Positive response from other 8.6 10 18.7 12
Positive response of self 13.4 12 19.1 16

Note. SD = standard deviation.

www.theipi.org 418
www.freepsychotherapybooks.org
 composite clinician-rated Health-Sickness Rating Scale

(HSRS). Four of the five measures of pervasiveness

yielded correlations near zero with the scores on the

Symptom Checklist. But the positive response of self

pervasiveness measure correlated significantly (r = –.48,

p < .005) with the Symptom Checklist scores, indicating

that higher symptom levels were associated with fewer

positive responses of self. It is consistent with this that a

significant correlation (r = .41, p < .05) was also found

between the HSRS and the positive response of self

pervasiveness measure. Again consistently, scores on

HSRS correlated significantly (r = —.34, p < .05) with

the negative response from other measure.

Change in CCRT Pervasiveness Was Moderately

Correlated With Change in Symptoms and Change in
Health-Sickness

The relationships between change in CCRT

pervasiveness and both change in symptom levels and

change in the HSRS were assessed through partial

correlation analyses. The purpose of using this method

was to allow for an assessment of change corrected for

www.theipi.org 419
www.freepsychotherapybooks.org
 initial level on each type of measure. Posttreatment

Symptom Checklist scores were correlated with late-in-

treatment pervasiveness, with the effects of pretreatment

symptom scores and early treatment pervasiveness

partialed out. The same analysis was done using the

HSRS.

Table 3 presents the partial correlations between

change on each of the five CCRT measures and change

on the Symptom Checklist and the HSRS. Of the five

CCRT measures, three showed statistically significant

partial correlations with the change in the Symptom

Checklist, and the other two evidenced near-significant

partial correlations, all in the expected direction. For

change in the HSRS, only change in the pervasiveness of

the negative response of self was significantly correlated

(r = –.53, p < .01), again in the expected direction:

Improvement in HSRS was associated with less change

in pervasiveness of the negative response of self.

DISCUSSION

www.theipi.org 420
www.freepsychotherapybooks.org
 TABLE 3
Partial Correlations of Change in CCRT Pervasiveness
Measures With Change in Symptoms and Change in Health-
Sickness

Hopkins Symptom Health-Sickness

Dimension
Checklist Change Rating Scale Change
Wish .41*** -.14
Negative .34* -.27
response from
other
Negative .40*** -.53****
response of
self
Positive -.32** .11
response from
other
Positive -.40*** .14
response of
self

*p = .06. **p = .08. ***p < .05. ****p < .01.

www.theipi.org 421
www.freepsychotherapybooks.org
 This study provided a theory-relevant measure of

psychodynamic change based on the CCRT method.

Good interjudge reliability was obtained with the

measure of pervasiveness of the CCRT across

relationship episodes. The theory-derived expectation in

relation to successful psychotherapy did emerge: The

main conflictual relationship pattern did become less

pervasive across the relationship episodes; the change

was relatively small but consistent.

CCRT changes from early to late in treatment were

found. Changes in pervasiveness were significantly

correlated with changes in other outcome measures,

namely changes in symptoms and changes in clinician

rated health-sickness. Each of these findings is discussed

in turn.

Reliability of Pervasiveness Score

The relatively high reliability of the pervasiveness

score can be attributed to several factors. First, the system

calls for a two-step task: categorizing the main

www.theipi.org 422
www.freepsychotherapybooks.org
 relationship pattern and then scoring its pervasiveness.

Second, the CCRT method on which the pervasiveness

measure is based is a guided clinical scoring system with

demonstrated interjudge reliability (see chapter 6, this

volume). A dynamic formulation method of questionable

or unknown reliability clearly would be problematic as a

basis for assessing change in dynamic conflicts. The

concept is clear: It is the percentage of interactions with

other people across narratives that contain the main

relationship theme.

CCRT Pervasiveness Across the Narratives

What we have labeled Freud’s (1912/1958a)

Observation 9 (see chapter 21, this volume), states that

there is a single central relationship pattern. One way to

examine this observation is to measure the degree of

pervasiveness of the CCRT components across the

narratives in psychotherapy sessions. A high

pervasiveness of components would be consistent with

Freud’s observation. The fist bit of precise evidence of

this kind was presented in Table 2. It revealed that the

www.theipi.org 423
www.freepsychotherapybooks.org
 pervasiveness percentage was especially high for the

wish (greater than 60%) within the early sessions; it was

only slightly lower within the late sessions, but at both

early and late points the pervasiveness of the wishes was

greater than that of the responses.

CCRT Pervasiveness Changes in Therapy

In terms of the pattern of changes from early to late in

treatment in the pervasiveness scores, it was found that

the wishes changed less than the responses. Apparently,

one’s wishes, needs, and intentions in relationships are

relatively intractable, and yet the expectations about

others gratifying or blocking one’s wishes and one’s

emotional responses to the others’ actions or expectations

have more flexibility, or malleability. Through successful

therapy, patients learn to recognize and cope with their

wish-response patterns in ways that lead to fewer

negative and more positive responses. In essence, the

reduction in pervasiveness of the conflictual relationships

can be seen as a theoretically crucial curative factor.

www.theipi.org 424
www.freepsychotherapybooks.org
 The magnitude of the changes in the pervasiveness

scores during psychotherapy also raises important

questions. Despite the fact that on standard outcome

measures the majority of patients in the Penn

Psychotherapy Project improved—65% were moderately

or much improved, which is the usual percentage in

psychotherapy outcome studies (Luborsky, Crits-

Christoph, et al., 1988)—the dynamic changes were

small on the average, and even patients who improved

considerably retained some of their basic relationship

components. These results have implications for theories

of the curative process in psychoanalytic psychotherapy.

Clinical discussions of the theory of change offer two

views of what happens in the course of psychoanalysis or

psychodynamic psychotherapy: One view holds that the
transference patterns and the conflicts within them are

resolved (see, for example, Ekstein, 1956; Davanloo,

1980); the other view (e.g., Pfeffer, 1963; Schlessinger &

Robbins, 1975) holds that transference patterns and the
conflicts within them remain evident even in the most

www.theipi.org 425
www.freepsychotherapybooks.org
 successful psychotherapy, although some components of

the pattern are altered. Our results clearly favor the

second clinical view of change, which emphasizes the

stability of the transference pattern. An early CCRT study

(Baguet, Gerin, Sali, & Marie-Cardine, 1984) of five

patients also showed basic stability during group

psychotherapy.

Another implication has special interest for the still

very hot debate in academic psychology about traits and

states. We have developed what is mostly a trait measure,

the CCRT, with a wish component that is especially

consistent across situations concerning parents,

therapists, and others while remaining consistent across

time, despite dedicated therapeutic efforts to change the

patient. Our finding would be opposite to the expectations

of such writers as Mischel (1968), who emphasized

inconsistencies over time in personality measures, and

more consonant with the writings of Block (1971, 1977),

who emphasized the findings of consistencies over time.

www.theipi.org 426
www.freepsychotherapybooks.org
CCRT Change Tends to Correlate With Symptom
Reduction

The changes in pervasiveness found in our data,

although small, apparently were meaningful, as

evidenced by the significant correlations between change

in CCRT pervasiveness and symptom reduction. Whether

the change drives the symptom reduction or is merely a

correlate cannot be rigorously answered with the current

data. Assessing both dynamic and symptomatic change at

multiple time points would test hypotheses about whether

dynamic change precedes and predicts symptom change.

The correlations between CCRT change and symptom

change were not so high, however, as to suggest that

change in the CCRT is redundant with change on

symptom inventories. Our data indicate that change in the

CCRT provides reliable extra information that is not

captured by symptom inventories or clinician ratings.

Limitations of the Measure of CCRT Pervasiveness

Despite the promising results obtained, the methods

used here have certain limitations. The CCRT

www.theipi.org 427
www.freepsychotherapybooks.org
 pervasiveness measure is based on all of the relatively

complete relationship episodes that we found in the two

or three sessions used at each time point. Variations in the

size of this database may affect the final measure of

dynamic change. For example, it may be that a larger

number of episodes than the 10 early and 10 late ones

used here might be necessary to obtain a more

representative index of the frequency of the main

relationship theme in the person’s life.

Another problem was a product of the freedom given

to the patient to select the main other person in the

narratives told about relationship episodes. For some

patients, the same type of other person may be present in

several of the episodes early in treatment but not included

in the episodes sampled late in treatment, a situation that

potentially could bias the data through comparison of

episodes from early and late points that are based on

narratives about different types of other people. Of

course, the simple fact that the patient chooses which

episode to tell means that the group of episodes cannot be

www.theipi.org 428
www.freepsychotherapybooks.org
 considered a representative sampling of the interpersonal
interactions in the patient’s life. On the other hand,

similar relationship patterns tend to emerge in narratives

about different other persons, and it is our impression that

the main relationships in most patients’ lives are covered

by the sampling used here. In addition, the reliance on

material as it unfolds to the clinician allows our measure

to have a closer tie to the clinical theories and methods.

But more research is necessary to examine the role of the

patient’s choice of significant other people in affecting

the measures of early versus late changes in the

frequencies within the CCRT.

Another limitation on our results came into view only

after we applied a different method of data analysis: a

rating of each relationship episode by itself after

randomizing them (see the QUAINT method described in

chapter 20, this volume; Crits-Christoph et al., 1994;

Connolly, Crits-Christoph, Demorest, Azarian, Muenz, &

Chittams, 1996). The new results indicated slightly less

pervasiveness, high levels of chance pervasiveness, and

www.theipi.org 429
www.freepsychotherapybooks.org
 the presence of multiple themes. However, a limitation

may also emerge from the rating of each relationship

episode by itself, because the clinician makes use of the

context of other relationship episodes to help in

understanding each relationship episode.

Finally, our usual index of CCRT pervasiveness is

only one operational conceptual translation that might be

used as an outcome measure for psychoanalytic

psychotherapy or psychoanalysis. Other aspects of the

psychoanalytic process, such as changes in defenses and

changes in awareness of relationship patterns, need more

study as criteria for improvement in psychoanalysis or

psychoanalytic psychotherapy.

We also need to know how our measure of

pervasiveness compares with similar measures used in

the field. We have defined pervasiveness as the frequency

of the types of relationship theme components across

different narratives. Yet this definition is only a variation

of a common one for pervasiveness, that is, the frequency

of all instances of the types of components, regardless of

www.theipi.org 430
www.freepsychotherapybooks.org
 whether they are in different narratives. Many TAT

scoring systems use such a frequency-of-all-occurrences

measure. Murray (1938) discussed the merits of the use

of frequency of occurrence, regardless of how often a

component appears in a particular narrative. Frequency,

he argued, has much to offer as a measure; it even

corresponds generally to the salience of the component.

CONCLUSIONS

The purpose of the study presented in this chapter

was to examine a theory-relevant measure of change in

moderate-length psychoanalytic psychotherapy in 33

patients’ psychotherapy transcripts from the Penn

Psychotherapy Project. The measure chosen was change

in pervasiveness of the CCRT from early to late in

therapy; the aim was to learn whether it was related to

more usual measures of change: the self-reported change

in symptoms and the clinician-rated health-sickness

rating. The main results were as follows.

• CCRT pervasiveness scores showed high agreement

among judges.

www.theipi.org 431
www.freepsychotherapybooks.org
 • There were small but meaningful changes in
pervasiveness from early to late in treatment,
although the changes were not uniform across the
five pervasiveness measures. Essentially, wishes did
not decrease significantly in pervasiveness, but
responses from others and responses of self did
decrease significantly. The largest changes were a
decrease in negative responses of self, a decrease in
negative responses from other, and an increase in
positive responses from other.

• The early-in-treatment pervasiveness of the positive

responses of self was significantly correlated with
initial level of symptoms and clinician-rated mental
health-sickness. Health-sickness ratings were
significantly correlated with pervasiveness of
negative responses from other. These correlations
were in the expected direction.

• Change in CCRT pervasiveness from early to late in

therapy was significantly correlated with change in
symptoms for three of the five CCRT measures;
change in health-sickness was correlated
significantly with change in the negative responses
of self measure. These correlations were in the
expected direction.

• As a whole, these results demonstrate another aspect of

validity for the CCRT method of measuring dynamic

www.theipi.org 432
www.freepsychotherapybooks.org
 change. Beyond that, the data have implications for
psychoanalytic and other theories of change and in
particular lend support to clinical theories
maintaining that aspects of the core conflictual
relationship pattern are still apparent even after
successful treatment.

Note

[5] An earlier version of this chapter appeared in L. Luborsky, P.

Crits-Christoph, J. Mintz, and A. Auerbach, 1988, Who Will
Benefit from Psychotherapy? (pp. 250-262), New York: Harper-
Collins. It has been adapted, revised, and printed with
permission of the publisher.

www.theipi.org 433
www.freepsychotherapybooks.org
 11

THE PARALLEL OF THE CCRT FOR THE

THERAPIST WITH THE CCRT FOR
OTHER PEOPLE

DEBORAH FRIED, PAUL CRITS-

CHRISTOPH, AND LESTER LUBORSKY

During much of his clinical career, Freud

(1895/1955b, 1901-1905/1953a, 1912/1958a)

observed a basic parallel in relationship patterns:

Soon after psychotherapy starts, the relationship

pattern with the therapist is experienced as similar to

the patient’s relationship pattern with other people.

This observation gave rise to Freud’s concept of a

relationship template and to the term transference, a

word that implies that there is a transfer of attitudes
and behavior from earlier relationships with

www.theipi.org 434
www.freepsychotherapybooks.org
 personally important people to the later relationship
with the therapist as well as others.

Freud’s first use of the term transference (1955b)

focused on this parallel between the relationship with

the therapist and a much earlier relationship, as the

following quotation makes clear:

In one of my patients the origin of a particular

hysterical symptom lay in a wish, which she had
had many years earlier and had at once relegated
to the unconscious, that the man she was talking
to at the time might boldly take the initiative and
give her a kiss. On one occasion, at the end of
the session, a similar wish came up in her about
me. She was horrified at it…. What had
happened therefore was this. The content of the
wish had appeared first of all in the patient’s
consciousness without any memories of the
surrounding circumstances which would have
assigned it to a past time. Since I have
discovered this, I have been able, whenever I
have been similarly involved personally, to
presume a transference and a false connection
had once more taken place, (pp. 302-303)

www.theipi.org 435
www.freepsychotherapybooks.org
 In his postscript to the Dora case Freud (1901-
1903/1953a) again wrote about this parallel:

They [transferences] are new editions or

facsimiles of the impulses and fantasies which
are aroused and made conscious during the
progress of the analysis; but they have this
peculiarity … . that they replace some earlier
person by the person of the physician. To put it
another way; a whole series of psychological
experiences are revived, not as belonging to the
past, but as applying to the person of the
physician of the present moment. Some of these
transferences … are merely new impressions or
reprints. Others … may even become conscious,
by cleverly taking advantage of some real
peculiarity in the physician’s person or
circumstances, (p. 116)

Freud (1901-1905/1953a) showed in a further

description of the treatment of Dora how literally he

followed his definition of transference in terms of this

parallel:

At the beginning it was clear that I was replacing

her father in imagination … she was even
constantly comparing me with him consciously,

www.theipi.org 436
www.freepsychotherapybooks.org
 and kept anxiously trying to make sure whether I
was being quite straightforward with her, for her
father “always preferred secrecy and roundabout
ways.” But when the first dream came, in which
she gave herself a warning that she had better
leave my treatment just as she had formerly left
Herr K’s house, I ought to have listened to the
warning myself. “Now,” I ought to have said to
her, “it is from Herr K that you have made a
transference on to me. Have you noticed
anything that leads you to suspect me of evil
intentions similar ... to Herr K’s?” (p. 118)

Until recently all of Freud’s observations about

transference—and 23 of them are listed in chapter 21

of this volume—have remained unexamined by

clinical-quantitative measures applied to the sessions

themselves. This is a gaping gap in research in

dynamic psychotherapy because the concept of

transference has been in everyday clinical use for the
last 100 years. The main exception, for many years,
has been studies of the concept by a questionnaire
approach (see chapter 20), but these were not
generally taken seriously, even though they were

www.theipi.org 437
www.freepsychotherapybooks.org
 confirmatory of a parallel, perhaps because of
uncertainty about the validity of the questionnaire

approach.

AIMS AND PROCEDURES

Freud’s initial observation about transference—

that it involves a parallel between the current
relationship with the therapist and past relationships
—is reexamined here by clinical-quantitative means.

Thirty-five cases were drawn from the Penn

Psychotherapy Project (Luborsky et al., 1988). All

were outpatients who were diagnosed according to
DSM-III as having nonpsychotic disorders, mostly
personality disorders and anxiety disorders. The
diagnoses were revised into DSM-III terms by

research psychologists on the basis of case reviews.

The patients were treated in psychodynamically

oriented psychotherapy with a mean of 45 sessions

during a mean of 41 weeks. The therapists were either

psychiatric residents in supervision at the University

www.theipi.org 438
www.freepsychotherapybooks.org
 of Pennsylvania or therapists who had completed
their training and had had several years of experience.

The data for the project were transcripts of

psychotherapy sessions, two sessions from early in

therapy and two from late in therapy. The two early

ones were generally Sessions 3 and 5, and the late
ones were at approximately the 90% point in the
treatment.

This study required a sufficient number of

relationship episodes from the sessions to extract a

relationship pattern both toward the therapist and

toward other people. As is usually true, relationship
episodes about other people (other people-REs) were
plentiful, whereas relationship episodes about the

therapist (therapist-REs) were sparse: Only a small

percentage of all relationship episodes were about the

relationship with the therapist—16% in the early

sessions and 22% in the late sessions. With so few

www.theipi.org 439
www.freepsychotherapybooks.org
 therapist-REs it is difficult to extract a therapist-

CCRT. We therefore decided to provide the judge

with the maximum information available: the

therapist-REs themselves rather than the therapist-
CCRTs. The basic comparison, therefore, was

between the other people-CCRTs and the therapist-

REs.

Not all transcribed sessions included therapist-

REs. When there were none in the late sessions, only

those in early sessions were used. When there were

none in the early sessions, only those in late sessions

were used. When both early and late sessions

included therapist-REs, both sources were used for

the relationship episodes and for formulating the

other person-CCRTs.

The study was carried out in two phases, as

explained in the following paragraphs:

Phase 1: Formulating Other Person-CCRTs

www.theipi.org 440
www.freepsychotherapybooks.org
 Because our interest was in the CCRT, which is

based on patients’ experiences with people in their

lives other than the therapist, the data from the
therapist-REs were deleted from the originally scored
CCRTs. Other person-CCRTs were formulated by the

CCRT method using the relationship episodes that did

not refer to the therapist: The number of relationship

episodes that contained a specific wish, response

from other, or response of self were counted; the most

frequently found wishes, responses from other, and

responses of self were used for the final CCRT
formulation.

Phase 2: Comparing Other Person-CCRTs With

Therapist-REs

Three judges familiar with the CCRT method and

blind to our hypothesis were employed for this study.
Each judge first read through the therapist-REs

collated from the early and late sessions. They were

instructed to use what they read to form a “gestalt”

www.theipi.org 441
www.freepsychotherapybooks.org
 Figure 1.

Matched case pair: Case A Therapist-RE

Case A Other person-CCRT
Mismatched case pair: Case A Therapist-RE
Case B Other person-CCRT

Matched versus mismatched pairs used for similarity

judgments.

www.theipi.org 442
www.freepsychotherapybooks.org
 view of the patient. The judges then compared the
group of therapist-REs with the other person-CCRTs

for the patient and with the other person-CCRT

formulations for 7 other patients. Pairs were judged

for similarity by the method of mismatched pairs

(Levine & Luborsky, 1981): A “matched” pair was
composed of the therapist-REs from “Case A” and
the other person-CCRT from Case A; a “mismatched”

pair was composed of the therapist-REs from Case A

and the other person-CCRT from a randomly chosen,

different case, such as Case B, as diagrammed in

Figure 1. This use of mismatched cases served as a

control for chance levels of similarity. We decided to

use this method and only 7 mismatched cases because

a previous pilot study involving 30 cases showed that

similarity ratings for 29 mismatched pairs correlated

highly (r = .9) with similarity ratings for 7
mismatched pairs. Judges were blind to the match-

mismatch design. The judges compared the CCRT

www.theipi.org 443
www.freepsychotherapybooks.org
 components separately; that is, they examined each

wish, response from other, and response of self in

turn and made the comparison to the gestalt derived

from reading the relationship episodes. These judges

were asked to rate how well each CCRT component

identified a theme of the relationship episodes, that is,
to note how much similarity they saw between the
CCRT component and the relationship episodes,

using a 1-to-7 scale (1 = no similarity, 7 = high

similarity). Data were analyzed by paired t tests for

differences between means with two-tailed

probability values.

The reader can readily get a sense of the judge’s

task by reading the case examples in chapter 5, this

volume, starting with the relationship episodes.

Briefer impressions of the judges’ task are conveyed

in the following two examples:

Example: Mr. Howard. Consider the similarity

of Mr. Howard’s therapist-RE 3 to the CCRT

www.theipi.org 444
www.freepsychotherapybooks.org
 formulation. The relationship episode is about
people being unresponsive to each other, and the
patient’s response is anxiety and tension. The
CCRT is about the wish to be close and to get a
response, with the expectation that the other
person will be unresponsive and the response of
self will be anxiety and tension. There is
moderate similarity between this therapist-RE
and this CCRT.

Example: Ms. Cunningham. The therapist-RE 5

for Ms. Cunningham is about the patient’s
conviction that she will not get reassurance from
the therapist. The CCRT contains the wish for
reassurance and the expected response from
other that she will not get the reassurance.
Again, there is moderate similarity between this
theme of the therapist-RE and this CCRT.

The kind of comparison made in these two

examples is exactly the type the judges had to make
for the matched samples. For the mismatched
material, the same therapist-REs were presented to

the judges along with the CCRT from another case.

For Mr. Howard’s data, the clinical judge would

almost certainly see considerable similarity for the

www.theipi.org 445
www.freepsychotherapybooks.org
 matched pair and dissimilarity for the mismatched

pair.

RESULTS

Interjudge Reliability of the CCRT Was Adequate

The intraclass correlation coefficient (pooled

judges) ranged from .55 to .75 for the different CCRT

components, with a median of .69. These figures
demonstrate adequate interjudge reliability.

Therapist-REs and Other Person-CCRTs Showed

Similarity

The main finding of this study was that patients

demonstrate a pervasive relationship pattern that can
be discerned when they interact with the therapist as

well as with other people. We found that correctly

matched pairs of therapist-REs and other person-

CCRTs were more similar than were mismatched

pairs. This was found for all three CCRT components

(wishes, responses from other, responses of self) and

www.theipi.org 446
www.freepsychotherapybooks.org
 reached statistical significance for the wish and

response of self components (see Figure 2).

Averaging the ratings for wishes, responses from

other, and responses of self showed that the mean
similarity between other person-CCRTs and therapist-

REs was 3.5 for the correctly matched pairs and 3.0
for mismatched pairs. The difference, although

apparently small, was highly significant, t(34) = –

3.51, p = .001.

Similarity Was Greater for Cases With Three or More

Therapist-REs

When only cases with three or more therapist-REs

were examined, the findings were in the same

direction but of greater magnitude: The similarity

between correctly matched cases was 4.1 for the

wish, 3.1 for the response from other, and 4.3 for the

response of self component (mismatched cases had

lower average similarity ratings, as shown,
respectively, in Figures 3, 4, and 5).

www.theipi.org 447
www.freepsychotherapybooks.org
 Figure 2.

* p < .01.

Mean similarity between therapist-REs and other

person-CCRTs for each CCRT component (1–7
scale).

www.theipi.org 448
www.freepsychotherapybooks.org
 This trend was maintained; that is, the more

therapist-REs per case, the greater the similarity

between therapist-REs and other person-CCRTs, with

significantly greater similarity for the wish, response

from other, and response of self components for the

correctly matched pairs (as shown, respectively, in
Figures 3, 4, and 5).

THE DIRECTIONS THAT LEAD FROM HERE

Increasing the Number of Relationship Episodes

Because the late sessions offered a few more

therapist-REs, we used a combined sample of early

and late sessions. Future work with larger subsamples
should examine the possibility that the parallel

between the relationship with the therapist and

relationship patterns with others might well fluctuate

in the course of the treatment.

We have emphasized the vital role of the number

of relationship episodes in the measurement of the

www.theipi.org 449
www.freepsychotherapybooks.org
 Figure 3.

Note: 3+ = 3 or more therapist-REs, 4+ = 4 or more therapist-REs, 5+ =

5 or more therapist-REs.

*All differences significant, p ≤ .04.

Mean similarity between therapist-REs and other

person-CCRTs by number of therapist-REs for wishes
only.

www.theipi.org 450
www.freepsychotherapybooks.org
 Figure 4.

Note: 3+ = 3 or more therapist-REs, 4+ = 4 or more therapist-REs, 5+ =

5 or more therapist-REs.

All differences nonsignificant.

Mean similarity between therapist-REs and other

person-CCRTs by number of therapist-REs for
responses from other only.

www.theipi.org 451
www.freepsychotherapybooks.org
 Figure 5.

Note: 3+ = 3 or more therapist-REs, 4+ = 4 or more therapist-REs, 5+ =

5 or more therapist-REs.

*p ≤ .002.

Mean similarity between therapist-REs and other

person-CCRTs by number of therapist-REs for
responses of self only.

www.theipi.org 452
www.freepsychotherapybooks.org
 relationship pattern: With each increase in the

number of therapist-REs, all similarity ratings rose,

and more so for correctly matched pairs. It is likely,

that a replication of the present study with a larger

sample of therapist-REs would show the same

parallel, but even more strongly.

We were struck by the concomitant increase in

similarity between incorrectly matched pairs of
therapist-REs and other person-CCRTs. This finding

indicates that with more therapist-RE data available,

a judge can discern more overlap with any other
person-CCRT, particularly for that of the same patient

but also for randomly chosen patients. This finding

can be understood as illuminating two points: (a)

There is a degree of personal specificity to a patient’s

relationship pattern as expressed by the CCRT

formulation (Luborsky, Mellon, van Ravenswaay, et
al., 1985; see also Freud’s Observation 11, chapter
21, this volume), and (b) some aspects of the CCRT

www.theipi.org 453
www.freepsychotherapybooks.org
 may be commonly found among different patients.
The specificity may lie in patients’ particular

combinations of the three CCRT components, so that,

taken as a whole, the CCRT tends to be specific to

each patient; but looking only at the CCRT

components separately may highlight the commonly
expressed aspects of relationship patterns across
different patients.

Examining Concomitants of Different Other Persons

Comparing the three CCRT components

separately gives researchers the ability to clarify

which aspects of the relationship pattern are

consistent when the “other person” changes. We

found that the wish and response of self components

were the most consistent, whereas the response from

other component did not show this pattern. The

former can be considered patient-specific

components, determined mostly, by the patient’s
personality. The response from other, although seen

www.theipi.org 454
www.freepsychotherapybooks.org
 through the patient’s eyes, is presumably influenced

to some degree by what was actually said or done by

the other person; the patient’s perception of this
behavior may be relatively insusceptible to influence

by the patient’s own pattern. Our sample consisting of

patients with anxiety disorders and depression was a
fairly high functioning one. Perhaps patients with
psychotic disorders or borderline personality

structures would have more distorted perceptions of

the responses from others.

Improving the Level of Similarity of Therapist-REs

and Other Person-REs

Another finding to consider is that even the

correctly matched pairs involving many therapist-REs

had similarity ratings less than 5 on a 1-to-7 scale (7

= highly similar)—at best they were only moderately

similar. This may be the result of the nature of the

task required of the judges: to compare data as

disparate as therapist-REs with a CCRT formulation.

www.theipi.org 455
www.freepsychotherapybooks.org
 One set of data, the therapist-REs, is composed of
narratives from which the judges were asked to form

a gestalt sense of the patient. The other, the other

person-CCRTs, is a brief summary of complex ideas:

the patient’s wish, the expected response from others,

and the patient’s response of self. It may be that even
when the underlying motives and responses were
quite similar, their expression in such different

formats made their comparison difficult, thereby

lowering the similarity ratings.

CONCLUSIONS

The results of the study reported here offer

quantitative support for the initial hypotheses of the

transference concept.

• A patient’s relationship with the therapist has

parallels with the patient’s relationships with
other people (Freud’s Observation 4; chapter 21,
this volume). We demonstrated this for the first
time ever through a comparison of therapist-REs
with other person-CCRTs for 35 patients: Each

www.theipi.org 456
www.freepsychotherapybooks.org
 patient’s three CCRT components (wish,
response from other, response of self) were
compared separately with the patient’s therapist-
REs considered en masse.

In essence, this work has examined the

observation of Freud’s (1912/1958a) that is
absolutely basic to the concept of transference
and therefore vital to all psychodynamic
therapies: that the patient’s experience with the
therapist partially parallels the pattern of
experiences with other people. Through
inspection of the central relationship patterns by
the CCRT method, we have shown that there is a
significant degree of similarity between patients’
wishes and responses toward the therapist and
those toward other people.

• Patients have a consistent relationship pattern,

which was demonstrated by the greater similarity
of correctly matched than of incorrectly matched
pairs of therapist-REs and other person-CCRTs.

www.theipi.org 457
www.freepsychotherapybooks.org
 12

THE PARALLEL OF THE CCRT FROM

WAKING NARRATIVES WITH THE CCRT
FROM DREAMS

Editors’ introduction6: We have shown so far in

this book that the Core Conflictual Relationship

Theme can be reliably recognized within the

narratives that a person tells. What we have not

looked into until the venture in this chapter is whether

the CCRT can also be reliably extracted from another

kind of narrative told in psychotherapy, that is, from

dreams. Although dreams provide a wealth of
conscious and unconscious revelations about
relationships (Freud, 1900/1953b), we do not know
(a) the extent to which the CCRT can be reliably

extracted from dreams and (b) when it is extracted,

how much it is like the CCRT obtained from the

narratives.

www.theipi.org 458
www.freepsychotherapybooks.org
 On the basis of Freud’s observations (for
example, Observation 21; see chapter 21, this

volume), we thought they might reveal parallel

information. But in fact, until the studies reported in

this chapter were done, we could not easily predict

the degree of parallel of the central relationship
patterns in the waking narratives with those in the
dreams.

In this chapter we report on two studies. The first

study is an evaluation of three sample cases presented

in detail; we also examine the degree to which it is

necessary to have the associations to the dream to

understand and score the dream for the CCRT and to
make the comparisons with waking narratives. The
second study was to be done if the results of the first
were encouraging. It was to be a precise and exact

examination of the same two questions, that is, (a)

can the CCRT be applied reliably to dreams, and (b)

www.theipi.org 459
www.freepsychotherapybooks.org
 is the CCRT from dreams like the CCRT from
waking narratives?

STUDY 1: THE PARALLEL OF THE CCRT

FROM WAKING NARRATIVES WITH THE
CCRT FROM DREAMS[7]

Carol Popp, Lester Luborsky, and Paid Crits-

Christoph

The data selected for the first study were drawn

from the psychotherapy of three patients, to be called

here Ms. Apfel, Ms. Bauman, and Mr. Crane. Ms.

Apfel’s and Ms. Bauman’s data were transcriptions of

psychoanalytic sessions conducted by expert

psychoanalysts; Mr. Crane’s data were a series of

dreams published by Gottschalk (1985, p. 66). The

transcribed cases were selected from the Penn

Psychoanalytic Collection (Luborsky, Stuart, et al.,

1997) only on the basis of the availability of

transcribed dreams and narratives. Diagnoses were

made according to DSM-III criteria applied to the

www.theipi.org 460
www.freepsychotherapybooks.org
 initial interviews. The diagnosis for Ms. Apfel was
obsessive-compulsive disorder, and for the diagnosis

for Ms. Bauman was dysthymia and mixed

personality disorder.

Selection of Dreams, Narratives, and Judges

For Ms. Apfel (Case 1), transcripts of Sessions 2

and 3 were used to select 10 relationship episodes for

the CCRT scoring. For Ms. Bauman (Case 2), the

narrative CCRT formulations were based on 39
relationship episodes from five early and five late

sessions. For both cases the usual criteria for

selection of relationship episodes on the basis of

completeness of content were used (see chapter 2,

this volume). The 10 dreams and associations for

Case 1 and the 7 dreams and associations for Case 2

were obtained from sessions in both early and late

phases of treatment. Three of the 7 dreams for Case 2

were obtained from incomplete transcripts, and
therefore associations were available for only 4 of

www.theipi.org 461
www.freepsychotherapybooks.org
 these dreams. To delineate the beginnings and ends of
associations, we followed the usual clinical criteria,

which involve the patient’s designated associations

and the analyst’s designations, usually adjacent to the

dream, of what he or she considered to be

associations to the dream. Essentially all the dreams
available to us in 1987 were used for the study; that
is, no discrimination was made on the basis of

completeness of content or other characteristics.

Psychodynamically oriented clinical psychologists

served as CCRT judges. A psychiatrist and a research

assistant identified the relationship episodes. The

psychiatrist identified the dreams and associations.

All CCRT judges were experienced in the use of the

CCRT method and all worked independently.

CCRT Tailor-Made and Standard Categories

CCRT components were scored both in narratives

of relationship episodes and in dreams by application
of the CCRT method in the usual tailor-made fashion.

www.theipi.org 462
www.freepsychotherapybooks.org
 When scoring the dreams for Ms. Apfel and Ms.
Bauman, the judges used three successive sets of

clinical material: (a) the dreams alone, (b) the dreams

with the addition of associations to assist in

identifying CCRT components, and (c) the dreams

with the addition of associations as part of the entire
sessions that included the relationship episodes. For
Mr. Crane, no relationship episodes were available,

so that only sets A and B were given to the judges.

Each of the individual tailor-made types of CCRT

components were assigned a standard category from

three lists of 14 wishes, 7 responses from other, and

14 responses of self (Crits-Christoph, 1987). Two

research assistants working as independent judges

made the assignments. Agreement between judges

was 91%. Differences were resolved by a third judge.

The results are given for each of the three patients
separately and then summarized. Preceding each

www.theipi.org 463
www.freepsychotherapybooks.org
 patient’s results, a brief clinical sketch is provided for

background.

Results

Ms. Apfel (Case 1)

At the time she came for analysis, Ms. Apfel was

a 31-year-old woman who had had repeated

unsuccessful relationships with men. In the first

session she explained the onset of her symptoms and

her goals for the therapy. “I’m unmarried although I’d

like to be… . The thing that made me seek therapy

was that I was in an unusually emotional state, with

crying and depression ... and some encounter with a

young man touched or culminated the whole thing.”

She described choosing younger men and taking a

maternal role in relation to them. In a previous brief

therapy, “the therapist pointed out I always set up the

relationship with men or I always choose one which I

could feel terrible anger against, and that I would be

www.theipi.org 464
www.freepsychotherapybooks.org
 forbearing and accept the conditions as laid down by

them meekly on the surface but meanwhile would be

feeling very resentful and ill-treated.” Her other
symptoms included a fear that she was homosexual
and a work inhibition involving years of unsuccessful

attempts to finish her graduate work. Of all of her

symptoms, she felt the recurrent failure with men was

the most pressing. She was seen four times a week for
8 months, after which her male analyst became ill and

had to stop the treatment. She resumed treatment with

a female analyst and continued for 3 more years.

Wishes derived from dreams alone. Ten dreams

were analyzed using the CCRT method. An example
of the tailor-made wish formulations is given in Table
1, which lists the number of dreams of the 10 dreams

scored in which the wishes were identified. The

wishes were some of the highest frequency wishes

obtained from the tailor-made application of the

CCRT method to the dreams alone.

www.theipi.org 465
www.freepsychotherapybooks.org
 TABLE 1
Tailor-Made Wishes Scored From Dreams Alone for Ms. Apfel

Judge 1 Number a Judge 2 Number a

1. To be close to, 6 For others to be 3
intimate with interested in
other me, give to me
2. To have sexual 3 To show 3
intimacy, affection and
physical concern for a
affection woman
3. To reject other, 3 To do things my 1
get away from way
other

a
Number of dreams and narratives, out of 10, containing the wish.

www.theipi.org 466
www.freepsychotherapybooks.org
 To facilitate comparisons between judges and
between results that were based on different sets of

clinical material, each tailor-made formation was

assigned a standard category. All of the standard

category wishes obtained from the dream content

alone are shown in Table 2, Column A. We selected
for each judge the wishes that occurred with the
highest frequency. If one wish was most frequent, the

wish or wishes that occurred at the second highest

level of frequency was also included. Consequently,

up to three components per judge were used for

comparisons, as shown in Table 2, Column A. The

highest frequency wishes for Judge 1 were for

closeness and independence and for Judge 2 for

closeness, for independence, and to be helped. Here,

the judges agreed on two of the highest frequency

wishes.

Wishes Derived From Dreams Plus Associations and

From Dreams Plus Whole Sessions

www.theipi.org 467
www.freepsychotherapybooks.org
 TABLE 2
Standard Category Wishes for Ms. Apfel

A B C D
Dreams Dreams Dreams REs
Alone + + Alone
Assoc. Session
Judges Judges Judges Judges
Wish 1 2 1 2 1 2 1 2
1. Closeness 6a 3a 6a 3a 6a 2a 9b 7b
2. Independence 3b 3b 3b 3b 3a 4a 10a 7a
3. To be helped 2 3 2b 4b 2 1 1 5
4. Respect 2 0 2 0 4 1 1 0
5. To help others 2 1 3 0 3 0 0 0
6. To be good 0 0 0 0 3 0 0 0
7. To feel good about 0 0 0 0 0 2 0 0
myself
8. To be understood 0 0 0 0 0 0 1 0
9. To avoid conflict 0 0 0 0 0 0 2 0

Note. Entries are number of dreams and narratives, out of 10,

containing each wish. Assoc. = associations; REs = relationship
episodes.
a Identified as having the highest mean frequencies for the pair of

judges.
b Identified as having the next highest mean frequencies for the
pair of judges.

www.theipi.org 468
www.freepsychotherapybooks.org
 In the clinical situation, dreams are assessed not

in isolation but together with their associations and

together with the entire session. Additional material

was given to the judges in two sets: first, as the

dream’s associations, and second, as the entire

sessions, including all of the relationship episodes.
Results of these additions are shown in Table 2,
Columns B and C. There were some changes when

associations and relationship episodes were used.

Judge 1 scored up to six more wishes when the

additional clinical material was available. Both

judges formulated some additional, entirely new
wishes. There was also some change in the relative
frequency of wishes. Regarding interjudge agreement
on the most frequently scored wishes, there was at
least one match between the most frequently

occurring wishes for both sets of clinical material.

Comparisons between the formulations obtained

from the three sets of clinical material (Table 2,

www.theipi.org 469
www.freepsychotherapybooks.org
 Columns A, B, and C) can be approached by first
selecting from each set the components that were

identified by both judges. These include Wishes 1, 2,

3, 4, and 5. The wishes for closeness, for

independence, and to be helped were obtained by

both judges for all three sets of clinical material. To
compare the relative frequency of occurrence, the
average of the frequencies of occurrence determined

by the two judges was obtained. The wish

formulation for closeness occurred with the highest

average frequency for all three sets of clinical

material. The wish for independence occurred at the

level of the second highest average frequency, in all

three sets.

Wishes derived from narratives. The wish

formulations obtained from relationship episodes by

the standard category CCRT are shown in Table 2,

Column D. The wishes that were identified by both

judges are Wishes 1, 2, and 3, and they are the same

www.theipi.org 470
www.freepsychotherapybooks.org
 as the wishes identified by both judges in all three

sets of clinical material used to formulate the dream

CCRT. The relationship episode wishes that occur
with the highest and second highest average

frequency were Wishes 1 and 2, which is comparable

to the dream formulations.

Responses from other and responses of self for

dreams and for narratives. Results for the responses
from other and responses of self are shown in Table
3. When scoring the dream-based CCRT components,

Judge 1 obtained a few more formulations with the

use of additional clinical material, whereas Judge 2

sometimes decreased the number of components

scored. For these formulations, there was at least one

match between judges for the highest frequency

components for all sets of clinical material. A

comparison between sets of clinical material for the
highest average frequency dream-based CCRT
components shows a considerable degree of

www.theipi.org 471
www.freepsychotherapybooks.org
 TABLE 3
Standard Category Responses From Other or of Self for Ms.
Apfel

A B C D
Dreams Dreams Dreams + REs
Alone + Assoc. Session Alone
Judges Judges Judges Judges
Responses 1 2 1 2 1 2 1 2
Responses from other
1. Rejecting 2a 5a 4b 1b 4b 3b 0 2
2. Not trustworthy 4b 2b 5a 3a 5 0 9a 3a
3. Dislikes me 3 1 3 0 3a 5a 2b 2b
4. Are hurt 3 0 2 0 2 0 1 2
5. Accepting 3 1 3b 2b 3 1 0 0
6. Happy 2 0 2 0 2 0 0 0
7. Are attractive 0 0 0 0 0 0 0 1
Responses from self
1. Angry 2b 1b 2b 3b 2b 1b 5b 4b
2. Uncertain 3a 2a 4a 3a 4 0 8a 4a
3. Guilty 0 3 3 1 5a 4a 1 2
4. Hurt other 4 0 4b 1b 4 0 0 0
5. Am not open 0 1 0 1 0 0 8 0
6. Dependent 0 0 0 0 0 0 3 0
7. Jealous 0 0 0 0 0 0 3 1
8. Accepted 0 0 0 1 0 0 0 0
9. Self-confident 0 1 0 0 0 1 0 0

www.theipi.org 472
www.freepsychotherapybooks.org
 Note. Entries are number of dreams and narratives, out of 10,
containing each response. REs = relationship episodes.
a Identified as having the highest mean frequencies for the pair of

judges.
b Identified as having the next highest mean frequencies for the

pair of judges.

www.theipi.org 473
www.freepsychotherapybooks.org
 consistency. The response from other of rejecting was

common in all three sets, and the category not

trustworthy was frequent in two sets. The responses

of self of angry and uncertain were most frequent in
three and two sets of clinical material, respectively.

The most frequent response from other formulations

obtained from the relationship episodes were not

trustworthy and dislikes me. The category not

trustworthy agreed with the dream formulations that

were based on dreams alone and on dreams with their

associations. The category dislikes me agreed with a
frequent response from other formulation identified in
dreams with the additional clinical material of

relationship episodes. The most frequent response of

self components identified in the relationship

episodes were angry and uncertain; they agreed with

the dream formulations for two or three sets of

clinical material, respectively.

www.theipi.org 474
www.freepsychotherapybooks.org
 Composite CCRT formulations. To complete the

present discussion, we include Ms. Apfel’s composite

CCRT formulation. A composite formulation can be

obtained by collecting the components that occur

with the highest average frequencies. Thus, the CCRT

formulation obtained from the dreams alone is “I
wish for closeness or independence, but the other
person is not trustworthy and rejecting and then I am

angry and uncertain.” One of the dreams used as the

basis for the CCRT is the following:

I dreamt on Thursday night that … I was in

some kind of meeting … And it was a very long
meeting … and I wanted to get out … we were
asked to go back in for another meeting, but uh,
there was something about it being a trick…

// I dreamt on Thursday night that … I

was in some kind of meeting// … And it
was a very long meeting
W: Independence //W … and I wanted to get out
NRO: Not //RO ... we were asked to go back in for
trustworthy another meeting, but uh, there was
something about it being a trick … //

www.theipi.org 475
www.freepsychotherapybooks.org
 The CCRT formulation from the narratives alone

is almost identical to the one from the dreams: I wish

for closeness or independence, but the other person is

not trustworthy and dislikes me and then I am angry

and uncertain. One of the narratives used as the basis

for the CCRT is the following:

And, uh, my friend, … and I was betrayed by

that [the friend] too… not exactly had been used
but … something like that. That she hadn’t my
best interests at heart…I felt betrayed. And I was
very unwilling to have another such relationship
with a girlfriend…

// And, uh, my friend, … and I was

betrayed by that [the friend] too … // not
exactly had been used but … something
like that.
NRO: Not //RO That she hadn’t my best interests at
trustworthy heart … // I felt betrayed.
W: Independence //W And I was very unwilling to have
another such relationship with a
girlfriend … //

Ms. Bauman (Case 2)

www.theipi.org 476
www.freepsychotherapybooks.org
 Ms. Bauman came for treatment at age 35
because she felt burdened by abnormal hair growth

all over her body (idiopathic hirsutism). She felt it to

be a stigma and a reason for social withdrawal. She

avoided social activities because of anxiety about

being seen and then rejected. In addition, she feared
being observed to be blushing. Her mood was
depressed, her self-esteem was low, and she felt

lonely and full of guilt.

Her relationships with her parents remained close.

She spent weekends and most of her vacations with

them, but these relationships were strained because

she felt overly influenced by her mother and
depreciated in relation to her brothers. The same
tension appeared as well in relation to other women:
She felt inadequate in comparison and questioned her

femininity. The tension was reinforced because she

never had had heterosexual relations. She attributed

this to her hirsutism and her Catholic background.

www.theipi.org 477
www.freepsychotherapybooks.org
 She did feel competent in her profession as a teacher,

but she found it hard to tolerate rivalries with the

other teachers and to be assertive with other teachers,
with the head of the school, and with her pupils. She

felt guilty each time she was able to behave

assertively. She felt too dependent for approval on
people in authority. Her analysis continued for 517
sessions, with gradual improvement in most of her

symptoms, especially in the depression and the

feelings of guilt.

Seven dreams were present in the 10 sessions

available. Summaries of the results based on the

standard categories are given in Table 4, which lists

only the components identified by both judges in at

least one set of clinical material.

Only a few changes in the CCRT occurred when

dreams were scored with the assistance of additional
session material. Agreement between judges was

www.theipi.org 478
www.freepsychotherapybooks.org
 TABLE 4
Standard Category Wishes and Responses From Other or of
Self for Ms. Bauman

A B C D
Dreams Dreams Dreams + REs
Alone + Assoc. Session Alone
Judges Judges Judges Judges
Responses 1 2 1 2 1 2 1 2
Wishes
1. Independence 2b 2b 2 1 2 1 24a 18a
2. Closeness 3b 1b 3b 1b 3b 1b 8b 5b
3. To be helped 4a 1a 4a 1a 4a 1a 8b 5b
4. Respect 0 1 0 1 0 1 3 5
Responses from other
1. Rejecting 2a 2a 2a 2a 2a 2a 9a 5a
2. Are hurt 2a 2a 2a 2a 2a 2a 0 0
3. Not trustworthy 2 0 2 0 2 0 8b 2b
4. Dislikes me 0 1 0 1 0 1 3 3
5. Happy 2 1 2 1 4 1 0 0
6. Accepting 0 0 0 0 0 0 1 1
Responses of self
1. Incompetent 4a 1a 5a 1a 5a 1a 8a 8a
2. Angry 2 1 2b 2b 2 2 3 2
3. Uncertain 2 1 2 1 2 1 8a 8a
4. Hurt other 2 1 2 1 2 1 0 6
5. Dependent 2 1 2 1 2 1 7b 8b
6. Am not open 0 1 0 1 0 1 2 3

www.theipi.org 479
www.freepsychotherapybooks.org
7. Guilty 0 1 0 1 4b 1b 8a 8a
8. Accepted 3b 1b 0 1 0 1 0 0
9. Self-confident 0 0 0 0 0 1 5 6

Note. Entries are number of dreams and narratives, out of ten,

containing each wish or response. REs = relationship episodes.
a Identified as having the highest mean frequencies for the pair of
judges.
b Identified as having the next highest mean frequencies for the

pair of judges.

www.theipi.org 480
www.freepsychotherapybooks.org
 difficult to assess because of the low number of
components scored. There was some similarity of

highest average frequency formulations for dreams

on the basis of the three sets of clinical material.

There was also moderate agreement between the

dream CCRT based on that material and the CCRT
based on the relationship episodes. For example, a
dream CCRT formulation obtained from the dreams

alone is “I wish for closeness and to be helped or for

independence, but the other is rejecting and is hurt

and then I feel incompetent.” One of the dreams used
as a basis for Ms. Bauman’s CCRT is the following:

… So it was my wish … that I was waiting to

sleep with him … but it didn’t work out, so I
practically, as it were had offered myself but was
rejected.

W: Closeness … //W So it was my wish … that I was

waiting to sleep with him … // but it
didn’t work out, so I practically, as it
were had offered myself
NRO: Rejecting //RO but was rejected.//

www.theipi.org 481
www.freepsychotherapybooks.org
 The CCRT formulation from narratives about

relationship episodes is “I wish for closeness and to

be helped or for independence, but the other is

rejecting and not trustworthy and then I feel

incompetent, guilty, and uncertain.” One of the

narratives used as a basis for Ms. Bauman’s CCRT is
the following:

… if I were to call him for once because I need

him, sometime when I really want to talk to him
… would he actually listen? Would I think that
was possible? And I have the damnable feeling
that it just isn’t possible.

W: Closeness … //W if I were to call him for once

because I need him, sometime when I
really want to talk to him// … would he
actually listen?// Would I think that was
possible?
NRO: Rejecting //RO And I have the damnable feeling
that it just isn’t possible.//

Mr. Crane (Case 3)

When Mr. Crane came for treatment, he was a 35-

year-old lawyer. His main physical symptoms were

www.theipi.org 482
www.freepsychotherapybooks.org
 associated with a duodenal ulcer; he had a diagnosis

of peptic ulcer by a gastroenterologist. He was an

intelligent, perceptive, and well-adjusted man. He had

the typical conflicts reported by some members of the

Chicago Psychoanalytic Institute for patients with

peptic ulcers: a conflict over dependency urges versus
feeling of shame. After his personal analysis, he
never experienced peptic ulcer symptoms again and

successfully advanced in his career. Gottschalk

(1985) summarized the patient’s conflicts as follows:

He was experiencing separation anxiety and guilt

about his desires for affiliation and care during the
termination phase of his therapy. This formulation
appears to be consistent with the dream CCRT
results.

Table 5 shows only the dream CCRT formulations

(Gottschalk. 1985). Again, a relatively small number
of components were scored on the basis of this small
sample of dreams. A few more wishes were scored

www.theipi.org 483
www.freepsychotherapybooks.org
 TABLE 5
Standard Category Wishes and Responses From Other or of
Self for Mr. Crane

A B
Dreams Dreams
Alone + Assoc.
Judges Judges
Components 1 2 1 2
Wishes
1. Independence 4a 4a 4a 5a
2. To be helped 2b 4b 3b 5b
3. Closeness 0 2 3 2
Responses from other
1. Dislikes me 2 1 2 0
Responses of self
1. Shame 2 2 2 2
2. Guilt 2 2 2 2
3. Controlling 2 2 2 2

Note. Entries are number of dreams, out of six, containing each wish or
response.
a Identified as having the highest mean frequencies for the pair of
judges.
b Identified as having the next highest mean frequencies for the
pair of judges.

www.theipi.org 484
www.freepsychotherapybooks.org
 when the judges could use the associations to the
dreams.

The results showed that for the wish and response

of self formulations, there was agreement between

judges on the highest frequency components. A

CCRT formulation that was based on the preceding
dream content is as follows: I wish for independence
or to be helped, but the other dislikes me and then I
feel shame and guilt and am too controlling.

Summary and Discussion

The main findings in Study 1 were based on a

comparison of dreams and narratives of three cases
presented in detail.

• Dreams can be reliably judged by the use of the

CCRT method. A study of two psychoanalytic
cases and one published set of dreams from a
psychoanalytic psychotherapy case indicates that
independent judges can obtain similar CCRT
formulations from dreams and from waking
narratives.

www.theipi.org 485
www.freepsychotherapybooks.org
 In terms of the aims of this research, the
results suggest that the CCRT method can be
reliably applied to dreams; that is, judges
working independently can obtain similar
formulations. In all comparisons between the
judges in which both judges scored components
with a frequency of 2 or greater, there was at
least one match between the highest frequency
components. Using this match as a criterion for
interjudge agreement, there was agreement
between judges for all three cases on the
relatively more frequency scored CCRT
components. Further assessment of interjudge
agreement is limited at this time because of the
small number of cases and the relatively small
number of components scored per case.

• Dream-based CCRTs were not significantly altered

by additional material from the session. An
examination of the dream-based CCRT
components using graded sets of clinical material
indicated that the judges made some changes
when additional clinical material was provided to
them. However, assessment of the components
with the highest average frequency showed more
consistency than change between the graded sets

www.theipi.org 486
www.freepsychotherapybooks.org
 of clinical material. Considering the importance
accorded to the use of associations to arrive at an
understanding of conflictual issues revealed in
dreams, more differences might have been
expected. However, associations are often used
clinically to help in decoding the manifest
content on the basis of an understanding of
dream work mechanisms, such as condensation,
displacement, and the meaning of dream symbols
(Pulver, 1987). None of these decoding
techniques were employed in deriving the CCRT
themes, which were intentionally obtained at a
level of inference that was no more than
moderately beyond the manifest level.

• Dream-based and narrative-based CCRTs show

agreement. With regard to the second aim of this
research, we found that agreement is usual
between the CCRTs that are extracted from
dreams and those that are extracted from
narratives about relationship episodes. In all
comparisons, there was at least one match
between dreams and narratives on the highest
average frequency components. Using these
matches as a criterion, we found interjudge
agreement for both psychoanalytic cases between

www.theipi.org 487
www.freepsychotherapybooks.org
 the relatively more frequently scored dream
CCRT components and the relatively more
frequently scored narrative CCRT components.
We concluded, therefore, that there is similarity
between the central, recurrent relationship
patterns revealed in dreams and those revealed in
waking narratives. The parallel between dreams
and narratives in terms of the CCRT is consistent
with the concept of a basic relationship schema
that shapes similar versions of itself in each
mode of expression—in dreams and in
narratives.

Some studies have demonstrated the clinical

observation of a parallel between waking
ideation and dreams. Goldhirsh (1961) showed
that the dream themes of convicts were related to
the particular crime that they committed. Miller
(1970) showed that the dreams of patients with
depression were about being defeated, frustrated,
and coerced. Beck (1967, 1971) examined the
themes of waking ideation and fantasies of
depressed patients and showed their similarities
to the themes in dreams; one general theme
common to both dream and waking life was
negative outcome of an activity. The studies by

www.theipi.org 488
www.freepsychotherapybooks.org
 Beck (1967) have much in common with our
own. However, our study is focused specifically
on the parallels of waking narratives and dreams
on the basis of our measure of the central
relationship pattern.

Two other comparable studies were reported

by Greenberg and Pearlman (1975) and
Greenberg (1987). Although no quantitative data
were given, the authors described similarity
between the manifest dream content and
discussions within the analytic hours or
interviews conducted, before and after the nights
of the dreams. This similarity concerned issues
of central import in the patients’ waking lives,
including the transference.

• Refinement of the standard categories may show

even greater agreement among judges. A final
comment deals with the similarity of the standard
category wish formulation obtained in these three
cases. For all cases, the components identified
with the highest average frequency were wishes
for “closeness” for “independence,” and “to be
helped.” These appear to be very common
wishes. However, it is useful to note that this
similarity is based on a comparison of standard

www.theipi.org 489
www.freepsychotherapybooks.org
 category formulations, and their use may obscure
differences between cases. For example, the
tailor-made wish formulations derived from the
dreams alone that were assigned the standard
category of “independence” were these:

Ms. Judge 1: To reject other, get away

Apfel from other
Judge 2: To be free of
responsibilities, to do something
illicit, to do things my way
Ms. Judge 1: For privacy
Bauman Judge 2: To protect my privacy

Mr. Judge 1: To get ahead of, to beat

Crane others, to intrude into others’
property
Judge 2: To compete with others,
to take what I want, not to be
dependent, to be alone

For each case, tailor-made descriptions appear

by inspection to have congruence but also some
differences. Further refinement of the standard
category lists, which is an ongoing process in our
group (see chapter 3, this volume), may enable
more discriminating comparisons to be made
between different cases and between dreams and
waking narratives.

www.theipi.org 490
www.freepsychotherapybooks.org
STUDY 2: THE PARALLEL OF THE CCRT
FROM WAKING NARRATIVES WITH THE
CCRT FROM DREAMS: A FURTHER
VALIDATION

Carol Popp, Louis Diguer, Lester Luborsky,

Jeffrey Faude, Suzanne Johnson, Margaret
Morris, Norman Schaffer, Pamela Schaffler, and
Kelly Schmidt

The learning experiences from Study 1, the pilot

study of the comparison of dreams and waking

narratives, encouraged us to go ahead with Study 2,

which was a more exact comparison of dreams and
waking narratives across a larger sample of patients,

with greater focus on the same theoretically and

clinically important questions as in Study 1: (a) the
reliability of the CCRT in dreams and (b) the
comparison of the CCRT in dreams and in waking
narratives.

Method

Our research relied on psychoanalytic treatment

because it is a type of psychotherapy that often

www.theipi.org 491
www.freepsychotherapybooks.org
 involves many dreams. We were fortunate to have
access to a collection of audiotapes and transcripts

from the Penn Psychoanalytic Collection

administered by Lester Luborsky, Sydney Pulver, and

George Woody (Luborsky, Stuart, et al., 1997). The

main criteria for inclusion in our sample of 13 cases
were that the treatment was a psychoanalysis and that
the case contained a minimum of 20 usable dreams.

(We have included one case that contained only 19

usable dreams.) Although our number of cases is

small, it is quite large for a psychoanalytic collection
because psychoanalytic cases are difficult and

expensive to collect; in total, these 13 cases consist of

approximately 8,500 sessions.

Of the 13 patients, 10 were female and 3 were

male. They all were in formal psychoanalysis and

were treated by well-known and experienced

analysts. The number of sessions in each treatment

www.theipi.org 492
www.freepsychotherapybooks.org
 ranged approximately from 110 to 1,200. Ten cases

each consisted of more than 400 sessions.

The CCRT method was performed as described in

chapter 2, this volume. As the first step, transcripts of

psychotherapy sessions were reviewed for the

presence of waking narratives about interactions
between the participant and others. These narratives
were rated for the degree of detailed information
given about the interaction. The ratings were

performed using a 5-point completeness-of-content

rating scale in which a complete narrative would
contain a description of the events that occurred, the

patient’s wishes, the responses from others, and the

responses of the self. Reasonably complete narratives

—those that were 2.5 or above on the 5-point scale—

were selected for CCRT scoring; they are called

relationship episodes (REs). Because differences have
been seen between CCRT components of early versus
late phases of therapy, we selected relationship

www.theipi.org 493
www.freepsychotherapybooks.org
 episodes separately from early and late sessions
(Crits-Christoph & Luborsky, 1990). For waking

narrative relationship episodes, early sessions were

typically Sessions 2, 3, 4, 5, or 6. Late sessions for 9

cases were located about 75% of the way through the

analysis, and late sessions for the 4 other cases were
located closer to the 90% point of the completion of
treatment. The total number of waking narrative

relationship episodes used in this study was 346.

Although transcripts of full sessions were used to

locate waking narrative relationship episodes,

transcripts of full sessions were not available to

locate dreams for most cases because only the dream
was transcribed. The criteria for rating dreams for
completeness of content were similar to those for
narratives except that dreams were selected as usable

if they had a completeness-of-content rating of 2.0 or

above. We were able to obtain a minimum of 20

usable dreams per case from 12 cases; a 13th case

www.theipi.org 494
www.freepsychotherapybooks.org
 included 19 usable dreams. The dreams of these 13
cases were separated into 26 sets: 13 sets of early

dreams and 13 sets of late dreams. Early dreams were

located, on average, in approximately the first 25% of

the analytic sessions, and late dreams were located in

approximately the last 37% of the sessions. Of the 26
sets, 3 sets contained only 9 dreams, and 1 set
contained only 8 dreams. Dreams and narratives were

almost always selected from different sessions—of

the 309 dreams used in this study, only 7 dreams were

located in sessions that also contained scored waking

narratives.

A count of words in the dreams was made as

described for the dream-related word count by
Stickgold, Pace-Schott, and Hobson (1994, pp. 20-
21). There were 168 early dreams that met

completeness-of-content criteria; 84% of these

consisted of 100 or more words, and 94% consisted

of 50 words or more. Of the 141 late dreams used,

www.theipi.org 495
www.freepsychotherapybooks.org
 87% consisted of 100 words or more, and 96%

consisted of 50 words or more.

We scored the waking narratives and dreams for

the three elements called CCRT components: wishes,

responses from other, and responses of self. Two

methods of formulating components were used:

tailor-made and standard category. Tailor-made

CCRT components present the wish, response from

other, and response of self in words of the CCRT

judge’s choice, often using the participant’s own

words. CCRT scoring using tailor-made components

allows the CCRT judges a great deal of flexibility and

results in CCRT components that can be uniquely

formulated for each participant. However, tailor-made

components are not optimal for many types of

statistical analyses of reliability and validity. Thus,

standardized lists of CCRT wishes and responses

must also be used. These standardized content
categories were translated by CCRT judges from the

www.theipi.org 496
www.freepsychotherapybooks.org
 tailor-made components. The standard categories

used here were selected from Edition 2 lists of the

standard categories (in Barber, Crits-Christoph, &

Luborsky, 1990; see also chapter 3, this volume). The
lists consist of 33 wishes, 33 responses from others,

and 31 responses of self categories.

Within the standard category lists, there is some

overlap among the standard categories. For example,

there can be some overlap in the meaning and

assignment of the standard category wishes “to be

independent” and “to be my own person.” Because of

this overlap, it was found that the standard categories

could be condensed or clustered into smaller sets

(Barber et al., 1990). That is, the 33 standard category

wishes could be condensed into only 8 categories

called CCRT standard category clusters. The 33

responses from other could be condensed into 8

response from other standard category clusters, and
the 31 responses of self could be condensed into 8

www.theipi.org 497
www.freepsychotherapybooks.org
 response of self clusters. (These standard category
clusters of Barber et al., 1990, are listed in Appendix

B, chapter 2, this volume.) In this study, all the

standard category CCRTs assigned by the CCRT

judges were put into the appropriate standard

category clusters. Finally, we determined which

CCRT standard category clusters occurred most

frequently and second most frequently for each set of

waking narrative relationship episodes or dream

narratives. The most frequent and second most

frequent CCRT standard category clusters were
identified because they are the most repetitive

relationship patterns and they have the greatest

similarity to the transference concept (Luborsky,
Crits-Christoph, & Mellon, 1986; Luborsky, 1990b).
To do this, we averaged the frequency scores of the
two judges, ranked the frequencies, and then selected

the clusters with the highest and the second highest

www.theipi.org 498
www.freepsychotherapybooks.org
 average frequency. These are the first rank and

second rank scores, respectively.

To assess reliability of the CCRT scoring, we

used the weighted kappa (Cohen, 1968), which has

become standard for evaluating reliability of the

CCRT (Crits-Christoph, Luborsky, et al., 1988;
chapter 6, this volume). This statistic allows one to
estimate the agreement of judges on nominal scales
with provision for agreement occurring by chance. In
counting “matches” between judges, a weight of 1.0

was given when the match was based on agreement

between the first rank CCRT cluster of Judge 1 and

the first rank cluster of Judge 2. A weight of 0.66 was

given when the match was based on agreement

between the first rank CCRT cluster of Judge 1 and

the second rank cluster of Judge 2. When the match

was based on agreement between the second rank
CCRTs of both judges, a weight of 0.33 was given.
Finally, a weight of 0 was given when there was no

www.theipi.org 499
www.freepsychotherapybooks.org
 agreement between judges for either first or second
rank CCRTs.

To evaluate agreement between waking narrative

and dream CCRTs paired by each case, a similar

method to the preceding was used; the weighted

kappa and the same weights as those used in
reliability estimation were used. A further analysis
was conducted using similarity scores (see chapter 6,
this volume). Similarity scores weight the agreement
between two judges for first and second rank CCRTs

in a more graded fashion than does the preceding

method and account for the occurrence of ties. A tie

occurs when more than one CCRT cluster may rank
equally as first (or second) rank clusters. The weights
are as follows: weight of 1.0 for first rank-first rank
and second rank-second rank match without any ties;

weight of 0.75 for first rank-first rank match with or

without ties; weight of 0.50 for first rank-second
match with or without ties; weight of 0.25 for second

www.theipi.org 500
www.freepsychotherapybooks.org
 rank-second rank match with or without ties; and
weight of 0 for “none of the above” matches. The
highest appropriate similarity score was assigned to
each comparison. Hotelling’s T2 multivariate analysis
was used to examine the similarity scores.

Results

Reliability

The reliability of CCRT scoring for the three

components—wish, response from other, and
response of self—in the dream reports and waking
narratives was assessed by examining agreement

between judges for the most frequent (first rank) and

second most frequent (second rank) CCRT standard

category clusters (see the section on Methods,
preceding). The most frequent and second most
frequent CCRT components were examined because
they describe the most repetitive relationship patterns

and show many similarities to the transference

www.theipi.org 501
www.freepsychotherapybooks.org
 concept. We used Cohen’s weighted kappa statistic to

estimate reliability (Cohen, 1968). Table 6 shows our

results. Kappas of 0.0 to 0.39 suggest poor
agreement, of 0.40 to 0.74 fair-to-good agreement,
and of 0.75 to 1.0 excellent agreement (Landis &

Koch, 1970). Our results fall in the fair-to-good or

excellent ranges of kappa. They indicate good

agreement between judges for standard category

CCRT scoring of dream reports. These values of

kappa for dreams are similar to those found in this

study for waking narratives (see chapter 6, this

volume). Thus, the reliability of the CCRT method

when applied to dreams appears comparable to the

reliability of the method when applied to waking
narratives.

Common Types of CCRT Components in Dreams and

Narratives

We wished to know which CCRT standard

category clusters were most common. In Table 7, we

www.theipi.org 502
www.freepsychotherapybooks.org
 TABLE 6
Agreement by Weighted Kappa of Two Judges in Scoring the
CCRT

Component Dreams Narratives

Wish .58 .67
Response from other .70 .74
Response of self .83 .75

www.theipi.org 503
www.freepsychotherapybooks.org
 TABLE 7
Percentage of Cases in Which a Standard Category CCRT
Cluster Appeared as the First Rank, Most Frequent CCRT
Cluster

Wish
Waking
Dreams
Narratives
Cluster Early Late Early Late
1. To assert self 0 8 0 0
2. To oppose, hurt others 0 0 15 8
3. To be helped, controlled 15 8 23 31
4. To be distant, avoid 31 62 31 54
conflict
5. To be close, accept others 8 0 8 15
6. To be loved, understood 38 38 23 23
7. To feel good, comfortable 0 0 0 0
8. To achieve, help others 23 0 0 0

Response From Other

Waking
Dreams
Narratives
Cluster Early Late Early Late
1. Strong 0 8 8 0
2. Controlling 0 8 0 0
3. Upset 54 23 23 8
4. Bad 0 0 0 0
5. Rejects, opposes me 62 62 69 92
6. Helpful 0 8 0 0
7. Likes me 8 8 15 8
8. Understands 0 0 0 0

www.theipi.org 504
www.freepsychotherapybooks.org
Response of Self
Waking
Dreams
Narratives
Cluster Early Late Early Late
1. Helpful 0 0 0 0
2. Unreceptive 0 0 0 0
3. Happy, respected 15 31 15 0
4. Oppose, hurt others 0 0 0 0
5. Self-confident 0 0 0 0
6. Helpless 0 8 0 31
7. Sad, angry 31 23 23 38
8. Anxious, ashamed 69 46 62 46

Note, n = 13 for dreams, n = 13 for narratives. The total percentage can

be larger than 100% owing to ties of CCRT cluster categories for
first choice rank.

www.theipi.org 505
www.freepsychotherapybooks.org
 show the percentage of cases in which a standard

category CCRT cluster appeared as the first rank, or

most frequent, CCRT cluster. For example, in dreams

located in early sessions, the wish cluster “to be

distant,” “avoid conflict” was the most common wish

in 31% of the 13 cases. The wishes “to be helped,
controlled;” “to be loved, understood;” and “to
achieve, help others” were also found to be the most

common wishes in early dreams in a relatively high

percentage of cases. Similar results were seen for

narratives, except that the wish “to achieve, help

others” did not appear as a first rank wish in
narratives. The most prevalent first rank response
from other clusters in dreams and narratives were
“rejects, opposes; upset;” and “likes me.” The most
prevalent first rank response of self clusters were

“anxious, ashamed;” “sad, angry;” “happy,

respected;” and “helpless.” Thus, it appears that most

CCRT standard clusters that presented as highest

www.theipi.org 506
www.freepsychotherapybooks.org
 ranking CCRTs in dream reports also appeared as

highest ranking CCRTs in waking narratives.

Similarity Between Dream and Narrative CCRT

Components

To look at the degree of similarity between dream

and narrative CCRT formulations, we compared the

dreams to waking narratives paired by each case.

First, we examined how often the two different sets of

CCRT scorings had the same first rank or second rank

standard category clusters. Table 8 shows that, for

each of the three CCRT components, usually from
about half to more than half of the participants had
the same frequency cluster. As examples, 5 of the 13

patients had the same first rank, highest frequency

wish cluster in both early waking narratives and

dreams, and 10 out of the 13 had at least one match

between the first rank or second rank wish clusters in

their early dreams and waking narratives. Similar
comparisons were made between late dreams and late

www.theipi.org 507
www.freepsychotherapybooks.org
 TABLE 8
Number of Participants With the Same Highest Frequency
Clusters in Dreams and Narratives

Early Dreams Versus Late Dreams Versus

Early Waking Narratives Late Waking Narratives
CCRT Weighted CCRT Weighted
I II I II
Match Kappa Match Kappa
W 5 10 .63 W 5 12 .60
RO 11 13 .91 RO 8 13 .68
RS 9 13 .79 RS 9 10 .45

Note. The numbers represent the number of participants, out of 13,

whose highest frequency wish, response from other, or response
of self clusters were the same for the comparison. A match in
Column I means that the CCRT formulations that were compared
had the same first rank, highest frequency CCRT standard
category clusters. A match in Column II means that the two
CCRT formulations that were compared had at least one
agreement between the first rank or second rank CCRT standard
category clusters. W = wish; RO = response from other; RS =
response of self.

www.theipi.org 508
www.freepsychotherapybooks.org
 narratives. Cohen’s weighted kappas were calculated

to account for agreement between CCRTs occurring

by chance. Table 8 shows kappas for comparisons
between dreams and narratives. The kappas ranged
from 0.45 to 0.91 and indicated considerable

agreement for all three components: wish, response

from other, and response of self.

Next, Hotelling’s T2 multivariate analysis was

used to assess further the agreement between dreams

and waking narratives of the most frequent CCRT

components. The analysis was made on the similarity

scores, weighting the agreements between dreams
and waking narratives. The results indicated that at

the early stage of psychotherapy, the similarity

between dreams and waking narratives varied from

one CCRT component to another, Hotelling’s T2 (2,

11) = 9.35, p = 0.04. The F statistics indicated that
the similarity between dreams and waking narratives
was lower for wishes than for responses of self, F(1,

www.theipi.org 509
www.freepsychotherapybooks.org
 12) = 5.47, p = 0.04, and for responses from others,
F(1, 12) = 9.12, p = 0.01; response from other and

response of self were not different, F(1, 12) = 0.0, ns.

At the late stage of psychotherapy, the similarity
between dreams and waking narratives was again

found to differ from one CCRT component to

another, Hotelling’s T2(2, 11) = 10.17, p = 0.03. The

similarity between dreams and waking narratives was

higher for response from other than for response of

self, F(1, 12) = 6.72, p = 0.02; wishes, F(1,12) =

7.80, p = 0.02; response from other and response of
self were not different, F(1, 12) = 0.03, ns.

Negativity of Responses

The CCRT response from other and response of

self can be rated as either negative or positive
(Luborsky, 1990a; chapter 4, this volume). Reliability

of negativity scoring was evaluated with Pearson

correlations: for response from other, r = .76, p =

.0001; for response of self, r = .79, p = .0001.

www.theipi.org 510
www.freepsychotherapybooks.org
 In early dreams, 69% of response from other

CCRTs were negative (Table 9). In early narratives,

76% of responses from other were negative.

Comparable assessments were made for late dreams
and late narratives. The range of average percentage

of negative responses was 67% to 76% for response

from other and 65% to 80% for response of self. The

paired t test was used to examine the negativity

scores of response from other and response of self;

the results indicated that the negativity of response

from other early dreams versus response from other
early waking narratives, of response of self early
dreams versus response of self early narratives, and

of response from other late dreams versus response

from other late narratives was not significantly

different but that the negativity of response of self

late narratives was significantly different, t(12) = 4-

46, p = .0008.

www.theipi.org 511
www.freepsychotherapybooks.org
 TABLE 9
Negativity of CCRT Components

Percentage of Negative Responses

Source of Data Response From Other Response of Self
Early dreams 69 (11) 71 (11)
Early narratives 76 (16) 79 (12)
Late dreams 66 (15) 65 (14)
Late narratives 71 (10) 80 (11)

Note. The negative and positive responses were counted, and the
percentage of negative responses per total number of responses
was calculated for each participant’s group of early or late dreams
or waking narratives. The numbers are the mean percentages of
negative responses obtained by averaging over the 13
participants. Numbers in parentheses are standard deviations.

www.theipi.org 512
www.freepsychotherapybooks.org
 In summary, our comparisons of CCRT
components in dreams and waking narratives show
considerable agreement. The results indicate the
presence of a central relationship pattern that is
commonly expressed in both waking narratives and

dreams. Clinically, the similarities provide a basis for

application of the CCRT method to both dream and
waking narratives to help find the repetitive patterns,

which can be a guiding focus for the therapist.

It should be noted that we almost always selected

dreams from different sessions than were used for

scoring relationship episodes in waking narratives. In

so doing, we chose a more demanding condition for
assessment of similarity; it might be expected to be
easier to find similarities if many of the dreams and
waking narratives came from the same sessions, in

which particular themes would predominate. The fact

that we found similarities between dreams and

waking narratives selected from many different

www.theipi.org 513
www.freepsychotherapybooks.org
 sessions suggests that the contents that were similar

were pervasive across many sessions.

We also found that the participants tended to

express more negative than positive responses and

that they did so in both dreams and waking

narratives. A preponderance of negative emotions in

dreams has also been reported in work by Carlson

(1986), Hall and Van de Castle (1966), Merritt,

Stickgold, Pace-Schott, Williams, and Hobson

(1994), and Rhode, Geller, and Farber (1992). Our

results are similar to those of several other studies

done with the CCRT, reported in chapter 4 and in this

chapter. We considered the possibilities, in chapter 4,

that the patients expressed more negative responses in

their waking narratives because they were in

psychotherapy and hence were presumably
experiencing pronounced difficulties or because
negative experiences were more strongly remembered
or because describing negative relationship episodes

www.theipi.org 514
www.freepsychotherapybooks.org
 might help master them. Regarding dreams, Merritt et

al. (1994) invoked several possible explanations for

the large amount of negative emotion in dreams,
including activation of limbic areas by
pontogeniculo-occipital waves and synthetic activity

of cortical association systems occurring during the

process of dream generation. Although the relevance

of these hypotheses is not known, further

investigations of negativity in either dreams or

waking narratives should give consideration to the

fact that considerable negativity has been found in

both dreams and waking narratives.

Further studies would benefit from use of both a

larger number of participants and control over the

conditions of dream reporting. These circumstances

are unlikely to be met using psychotherapy sessions.

Finding psychotherapy cases, which usually involve
one session per week, that contain 20 usable dreams
would require an enormous database of patients

www.theipi.org 515
www.freepsychotherapybooks.org
 because dreams are relatively uncommon in

psychotherapy treatments. It is for this reason that we

chose to study psychoanalytic cases. However,
psychoanalytic cases are difficult and expensive to
record because they consist of a large number of

sessions; some of our cases were 1,200 sessions long.

An alternative approach would be to study patients in

psychotherapy, conducted once per week, in which

the patients report their dreams at home on a nightly

basis using controlled protocols such as described by

Hobson and Stickgold (Hobson & Stickgold, 1994;

Merritt et al., 1994; Stickgold et al., 1994). In this

way, the number of patients could be increased and

the circumstances of obtaining dream reports could
be controlled.

SUMMARY AND CONCLUSIONS

• Study 2 confirms and expands on Study 1. This

systematic comparative study of relationship
themes in dreams is the first that is based on the
CCRT and confirms and adds to the findings of

www.theipi.org 516
www.freepsychotherapybooks.org
 Study 1. Study 2 was also supported by the
finding in Study 1 that the dream association
material in the sessions was not essential for the
comparison with narratives.

• CCRTs of dreams can be reliably judged. Our

results confirm our hypothesis that the CCRT can
be reliably identified in dreams.

• The dream and narrative CCRTs are similar, even

when the samples of dreams and narratives are
collected from different sessions. The repetitive
elements that form the CCRT in dreams can be
compared with relationship themes in waking
narratives in a relatively straightforward fashion.
We found significant similarities between the
recurrent relationship themes in dreams and
those in waking narratives. The results for
relationship narratives are new and imply that the
narrative also has qualities that also justify its
being considered a “royal road” to deeper levels
of the personality.

• We also found that the similarity between dreams

and narratives is highest for the responses from
others. This high similarity may be of special

www.theipi.org 517
www.freepsychotherapybooks.org
 significance and should be followed up in future
analyses.

• The positive and negative quality of waking

narratives and dreams can be judged reliably.

• Negativity of responses of self in late dreams is less

than that of the responses of self in late
narratives. This finding should also be followed
up in future studies for it is contrary to what
might be expected.

• The finding of similar CCRTs in the waking

narratives and dreams implies similarity of
CCRTs across states of consciousness. These
results support the concept of a central
relationship pattern that shows itself in two
different states or modes of thinking: in dreams
and in waking narratives.

Note

[6] Study 2 of this chapter is a revised version of “Repetitive

Relationship Themes in Waking Narratives and Dreams,”
by Popp, Diguer, Luborsky, Faude, Johnson, Morris,
Schaffer, Schaffler, and Schmidt, 1996, Journal of
Consulting and Clinical Psychology, 64, pp. 1073-1078.
Reprinted by permission.

The research was partially supported by a grant from

the Fund for Psychoanalytic Research (to Carol Popp); by
Research Scientist Award (NIDA) DA-00168-24 and

www.theipi.org 518
www.freepsychotherapybooks.org
 Research Scientist Award (NIMH) MH 40710-22 (to
Lester Luborsky); by NIMH Clinical Research Center
Grant P50NH45170 (to Paul Crits-Christoph); and by
FCAR Research Grant 95-NC-1277 and SSHRCC
Research Grant 410-93-1388 (to Louis Diguer).

[7] Grateful acknowledgment is given to Hartvig Dahl, Merton

Gill, Horst Kächele, Sydney Pulver, and George Woody for
help in assembling the Penn Psychoanalytic Collection of
recorded psychoanalyses administered by Lester Luborsky,
Sydney Pulver, and George Woody. We thank Paul Crits-
Christoph for some of the transcriptions of the waking
narrative relationship episodes. Portions of this study were
presented in 1995 at the international meeting of the
Society for Psychotherapy Research, Vancouver, Canada.

www.theipi.org 519
www.freepsychotherapybooks.org
 13

THE MEASUREMENT OF ACCURACY

OF INTERPRETATIONS

PAUL CRITS-CHRISTOPH, ANDREW

COOPER, AND LESTER LUBORSKY8

For years a steady succession of clinical papers

have shown therapists how to recognize the accuracy

of interpretations. A representative one by Kubie

(1952) offered these criteria: “Evidence for the

accuracy of the interpretations is sought in the
patient’s further free associations and in the

unlocking of doors to his lost memories” (p. 76).

Kubie went on to provide clinical examples of this

kind of match between the interpretation and the
patient’s associations and concluded that the tests of
the accuracy of an interpretation derive from “1) the

patient’s associations to it, which may confirm,

www.theipi.org 520
www.freepsychotherapybooks.org
 correct or reject it; 2) the alterations in symptoms; 3)
finally, and only rarely, the increase in our ability to

predict future behavior. But we need better tests than

these” (p. 89). He was right; we do need better tests

than these. Kubie’s main test, a fit between the

interpretations and the patient’s verbal and repeated-
in-action associations, echoed Freud’s (1912/1958a,
1913/1958c, 1914/1958e) views of how to determine

the accuracy of interpretation.

Although such criteria for accuracy of the match

between the interpretation and the associations are

often clinically recognizable, the field lacks

information about the interjudge agreement on such
criteria. An early method of Luborsky’s (included in
Auerbach & Luborsky, 1968), intended to handle this
research problem, requires ratings of the degree of

convergence between the therapist’s interpretations

and the essence of the patient’s main

communications. The degree of such convergence

www.theipi.org 521
www.freepsychotherapybooks.org
 was found to be moderately reliably judgeable (mean
r of about .6), but it had a major problem: its

impressionistic assessment of the patient’s main

communication.

The current method improved on the previous one

by operationalizing further the concept of accurate
interpretation by means of this new measure:
convergence of the therapist’s interpretations with the
patient’s CCRT (Crits-Christoph, Cooper, &
Luborsky, 1988). The improvement is accomplished

by the safer reliance on the CCRT (see chapter 6, this

volume) as a measure of the patient’s main

communication. The improved assessment rests on
the hypothesis that a cogent measure of accuracy is
the extent to which the therapist’s interpretations in a
session deal with the main independently established

CCRT components of the session.

www.theipi.org 522
www.freepsychotherapybooks.org
 Our initial aims in the current work were (a) to
develop the improved measure of accuracy of

interpretation and (b) to examine the predictive

capacity of the measure for the outcome of

psychotherapy. Our subsequent aims were (c) to learn

whether accuracy of interpretation would predict
outcome only in the context of a positive therapeutic
alliance and (d) to learn the relative predictive

capacity of a broader measure of the therapist’s

responses, the Therapist’s Errors in Technique Scale

(Sachs, 1983).

PROCEDURES

Of the 73 patients in the Penn Psychotherapy

Project (Luborsky et al., 1988), 43 were included in

this study because their transcripts were available and

they roughly reflected the range of outcomes in the

total sample. We applied our CCRT measure to

transcripts of two sessions drawn from early in

treatment (usually Sessions 3 and 5). Sets of

www.theipi.org 523
www.freepsychotherapybooks.org
 independent judges could then rate each of the
interpretations for their accuracy.

With the exception of one judge who coded

interpretations and one judge who marked off

relationship episodes (both trained research

assistants), judges were experienced clinicians
(clinical psychologists and psychiatrists) trained in
each task. All judges were blind to treatment outcome
and worked independently. Separate sets of judges
scored each measure.

Two previously developed measures, the Helping

Alliance Counting Signs method (Luborsky, Crits-
Christoph, Alexander, Margolis, & Cohen, 1983) and
the Errors in Technique Scale (Sachs, 1983), a

subscale of the Vanderbilt Negative Indicators Scale

(O’Malley, Suh, & Strupp, 1983), were also applied

to the sessions. The Errors in Technique Scale was

used to examine the overlap between this scale and

www.theipi.org 524
www.freepsychotherapybooks.org
 our measure of accuracy of interpretations. The
Helping Alliance method was applied to examine

whether alliance and accuracy measures contribute

independently to the predictions of outcome and to

test the hypothesis that interpretations have more

impact in the context of a positive therapeutic
relationship.

Patients

The patient sample consisted of 30 women and 13

men, ranging in age from 18 to 48 years, with a mean

age of 25.2 years. About two-thirds of the patients

were between the ages of 18 and 24. Descriptive

characteristics of the patients are presented in Table

1. Patient diagnoses are summarized in Table 2. The

majority of patients were diagnosed as having

dysthymic disorder, generalized anxiety disorder, or

one of a variety of personality disorders.

Therapist and Treatment Characteristics

www.theipi.org 525
www.freepsychotherapybooks.org
 TABLE 1
Characteristics of Patients (N = 43)

Characteristic Number
Age
15-19 3
20-24 25
25-29 6
30-34 4
35-39 4
40-44 0
45-49 1
Sex
Female 30
Male 13
Race
Black 3
White 40
Marital status
Single 29
Married 8
Divorced, separated, or widowed 6
Education
High school degree 5
Some college 21
College degree 6
Some graduate or professional school 7
Graduate or professional degree 4

www.theipi.org 526
www.freepsychotherapybooks.org
 TABLE 2
Patient Diagnosis (N = 43)

With Axis
Diagnosis Number
II
DSM-III Axis I
Atypical eating disorder 1 —
Dysthymic disorder 16 8
Ego dystonic homosexuality 2 1
Generalized anxiety disorder 11 4
Inhibited sexual excitement 2 —
Obsessive-compulsive disorder 2 —
No Axis I diagnosis 13 13
DSM-III Axis II
Atypical personality disorder 1
Compulsive personality disorder 4
Histrionic personality disorder 4
Narcissistic personality disorder 1
Passive-aggressive personality 4
disorder
Schizoid personality disorder 8
Schizotypal personality disorder 3
Mixed personality disorder 1

Note. Several patients had more than one diagnosis.

www.theipi.org 527
www.freepsychotherapybooks.org
 Twenty-eight therapists participated in the

research project. Each therapist usually treated one or

two patients. The therapists ranged in age from 27 to

55 years, with a mean age of 35.6 years. They had
between 1 and 22 years of prior clinical experience,

with an average of 5.4 years. Twelve of the therapists

were psychiatrists in private practice; the remaining

16 were psychiatric residents. The residents attended

weekly 1-hour individual therapy supervision

sessions led by experienced clinicians.

All patients were seen in individual

psychodynamic psychotherapy. Approximately two-

thirds of the patients were treated at the outpatient

clinic of the Hospital of the University of

Pennsylvania. The rest were seen in private practice

settings. Treatment length ranged from 21 to 149

weeks, with an average of 53.5 weeks.

Measures

www.theipi.org 528
www.freepsychotherapybooks.org
 Identifying Interpretations

Two judges (an experienced clinician and a

research assistant very familiar with the task) coded

therapist statements into interpretations versus all

other types of responses. A response was considered
an interpretation if it met at least one of the following
criteria: (a) The therapist explained possible reasons
for a patient’s thoughts, feelings, or behavior (such
as, “Yes, but one of the benefits of using drugs is that

it keeps you in the role of the child”); and (b) the

therapist alluded to similarities between the patient’s

present circumstances and other life experiences

(such as, “And what’s happening is that you keep
getting yourself into these kinds of situations, like
what happened on Saturday where you put yourself in

a hell of a big rejection experience”).

The judges were kept blind to treatment outcome

and independently read the typed transcripts of two
therapy sessions. Responses that were scored as

www.theipi.org 529
www.freepsychotherapybooks.org
 interpretations by both judges were included in the
study. If a case yielded no interpretations in the two

sessions (as occurred for two patients), judges read

additional sessions until agreement was reached on at

least one interpretation. The number of interpretations

obtained per patient ranged from 1 to 16, with a mean
of 6.1.

Interrater reliability on the basis of judges’ ratings

for all 43 cases, was assessed for distinguishing
interpretations versus other statements. The

interjudge agreement was 95%; Cohen’s (1960)

kappa statistic, a measure of chance-corrected

agreement for nominal scales, was .56 (p < .0001).

Combining Judges’ CCRT Formulations

Two, or occasionally three, experienced clinician

judges scored each of the 43 patients for the CCRT,

according to the method outlined in chapter 2, this

volume. For each case the final CCRT selected for

www.theipi.org 530
www.freepsychotherapybooks.org
 inclusion in the study was a composite of the judges’

independent CCRT formulations. It included the

wish, negative response from other, and negative

response of self (positive responses from other and

positive responses of self were of low frequency in

this sample). The judges’ tailor-made CCRT
formulations were coded into standard categories
(Edition 1; see Luborsky, 1986b) by three other

judges, with greater than 95% agreement.

A composite CCRT was derived by selecting the

most frequent wishes and responses among the CCRT
judges. The final CCRT formulation for each patient

consisted of up to two wishes, three negative

responses from other, and three negative responses of

self.

Accuracy of Interpretations

Accuracy of interpretations as scored here

represents the degree of congruence between the

www.theipi.org 531
www.freepsychotherapybooks.org
 contents of the therapist’s interpretations and the
contents of a patient’s CCRT. Because the CCRT is

composed of three main subtypes, accuracy is

conceptualized as a multidimensional concept.

Consequently, the assessment of accuracy involves

multiple ratings on each interpretation. A 4-point
rating scale was used to assess the degree to which a
clinical judge believed that the therapist’s

interpretation addressed a particular CCRT wish,

response from other, or response of self.

The following CCRT and therapist’s

interpretation, drawn from one of the cases used in

this study, is presented to illustrate the nature of the
accuracy ratings. The patient’s CCRT consisted of
one wish (“to make contact with others, be close”),
one negative response from other (“rejects, distant”),

and three responses of self (“lonely, depressed,

anxious”). One of the therapist’s interpretations

follows:

www.theipi.org 532
www.freepsychotherapybooks.org
 Well, I’m beginning to get a picture of a—a lot
of involvement that you have with this guy still,
even though he’s cut things off, you haven’t.
And you’re not able to begin replacing him yet
and, uh, the emotional investment, emotional tie
you’ve got still to him, and pretty strongly. And
that’s inhibiting you. Now, what’s behind that,
y’know, obviously he was very important to you,
more important than any other guy has been.
And that makes it harder to give him up. And the
fact that he really is the one who decided—made
the choice to break, not you, makes it harder to
give him up too. I—I see some reaction: What’s
going on?

This interpretation was rated as accurate in regard to

the wish and response from other but not for the

responses of self. For the wish, the average of the

accuracy of the judges’ ratings for the congruence of

content was 3.67; for the response from other, it was

4.0; and for the three responses of self, 1.33, 1.33,

and 1.0, respectively.

Examples of interpretations from two other

patients are given in Figures 1 and 2 (from chapter 5,

www.theipi.org 533
www.freepsychotherapybooks.org
 Figure 1.

PATIENT’S CCRT
🙁 Wish: To not be cut off from closeness
Response From Others: Rejects
Response of Self: Anger; self-devaluation; upset
THERAPISTS INTERPRETATION
🙂 Naturally you feel upset now—You see me as
unresponsive to you

A therapist’s interpretation that is highly congruent

with the patient’s CCRT (patient: Mr. Howard,
Session 3).

www.theipi.org 534
www.freepsychotherapybooks.org
 Figure 2.

PATIENT’S CCRT
🙁 Wish 3: I want to be reassured; to get approval; to avoid
disapproval
THERAPISTS INTERPRETATION
🙂 What strikes me … you went home after you left and
talked to Henry [husband] about it… wanting
reassurance, but not here.

A therapist’s interpretation that is moderately

congruent with the patient’s CCRT (patient: Ms.
Cunningham, Session 5).

www.theipi.org 535
www.freepsychotherapybooks.org
 this volume). These interpretations are given here in

simplified form. The one by Mr. Howard’s therapist

was given a high rating for accuracy (3.8); the one by

Ms. Cunningham’s therapist got only a moderate
rating for accuracy (2.4).

For each case, three experienced clinician judges

who were kept blind to treatment outcome were

presented with composite CCRT formulations and

interpretations that were extracted from transcripts

for each case. The judges, working independently,

were directed to familiarize themselves with the

patient’s CCRT formulation and to make ratings of

accuracy on each wish, response from other, and

response of self contained in that patient’s CCRT

formulation.

Ratings for the wishes were averaged to form a

composite wish dimension for each patient. Similarly,

ratings for the responses from other and responses of

www.theipi.org 536
www.freepsychotherapybooks.org
 self were averaged to yield composites for each. For

each patient, these accuracy scores were then

averaged across all interpretations. Interrater

reliability of the accuracy scales was computed using

the intraclass correlation coefficient. On the basis of

the sample of 43 cases, the pooled interjudge
reliabilities were as follows: (a) .84 for accuracy with
respect to the patient’s wishes, (b) .76 for accuracy

with respect to the patient’s responses from other, and

(c) .83 for accuracy with respect to the patient’s

responses of self.

Intercorrelations between the accuracy scales

were computed to examine the overlap between these

dimensions. The correlation between the wish and

response from other scales was .68. Virtually no

correlation was found between the response of self

and wish (r = .07) or between the response of self and
response from other scales (r = .04). Given the
sizable correlation between the wish and response

www.theipi.org 537
www.freepsychotherapybooks.org
 from other components, these two dimensions were

combined into a composite dimension to avoid

problems of multicollinearity of predictors (Cohen &

Cohen, 1975) in subsequent multiple regression

analyses.

Errors in Technique Scale

The Errors in Technique subscale of the

Vanderbilt Negative Indicators Scale (O’Malley et al.,
1983) is a set of 10 items, all of which are
hypothesized to be inversely related to beneficial

treatment outcome (Sachs, 1983). The 10 items are

(a) failure to structure or focus the session, (b) failure

to address maladaptive behaviors or distorted

apperceptions, (c) insufficient examination of

potentially harmful behavior or attitudes, (d) failure

to address signs of resistance, (e) failure to examine

the patient-therapist interaction, (f) superficial
interventions, (g) poorly timed interpretations, (h)
destructive interventions, (i) inappropriate use of

www.theipi.org 538
www.freepsychotherapybooks.org
 silence, and (j) inflexible use of therapeutic
techniques. Possible ratings for each item range from

0 (errors not present or within normal limits) to 5

(strong evidence for errors).

In Sachs’s (1983) study of negative factors in

short-term therapy, interrater reliability was
calculated for the Errors in Technique Scale. Of the
original 10 items, 7 had adequate levels of interrater
reliability (>.60). The average interrater reliability of
the 7 scales was .73. The level of internal consistency

(coefficient alpha) was .46 (p. 559). In addition, the

scale was significantly correlated in the expected

direction with treatment outcome (r = –.56, p < .01)
in a sample of 18 male college students in brief
therapy.

For each case in the present study, two

experienced clinician judges (including the author of

the scale) independently listened to the tape

www.theipi.org 539
www.freepsychotherapybooks.org
 recordings while reading the typed transcripts of the

first 30 minutes of each of two early therapy sessions.

To reduce the complexity of the task, each 30-minute
segment was divided into two 15-minute segments,

which were separately rated on the 10 items of the

Errors in Technique Scale. The ratings of each item
were averaged across the two segments, and then the
scores of the two sessions were combined. Of the

original 10 items on the scale, only 6 had some

variance; these were summed to form a final scale

score. The 6-item version of the Errors in Technique

Scale used in the current study had a mean of 5.5 (out
of a possible 30 points) and a standard deviation of
3.4.

On the basis of the sample of 43 cases, the pooled

interjudge reliability (intraclass correlation) of the 6-

item scale was .61. Additionally, the level of internal
consistency of the scale, as assessed by Cronbach’s
alpha, was .60.

www.theipi.org 540
www.freepsychotherapybooks.org
 Helping Alliance Counting Signs Scale

The Helping Alliance Counting Signs method

(Luborsky, 1976; Luborsky, Crits-Christoph, et al.,

1988) was applied by two experienced clinician

judges to the first 30 minutes of each of the two early

sessions for each patient. The score for positive

helping alliance signs was selected for use in this

analysis because this measure had proved to be the

most successful predictor of outcome in the

comparison of the 10 most improved and 10 least

improved cases from the Penn Psychotherapy Project.

The pooled judge reliability of this measure

(intraclass correlation = .57) was lower than expected

and appeared to be a function of one judge’s scoring

many more indicators of a helping alliance than the
other judge. Nevertheless, we combined the two
judges’ scores for a final helping alliance measure.

Treatment Outcome

www.theipi.org 541
www.freepsychotherapybooks.org
 Test and interview evaluations were conducted for

the 43 patients when they began therapy and again

when they terminated treatment. From these data two

outcome measures were devised: residual gain

(meaning a gain corrected for initial level) and rated

benefits (meaning a composite rating of
improvement; Luborsky, Crits-Christoph, et al.,
1988).

RESULTS

The Average Level of Accuracy of Interpretations Was

Low

Table 3 gives the means and standard deviations

of the accuracy dimensions. It can be seen that the
average level of accuracy was low, yet enough

variability was present to allow for relationships with

other variables to emerge.

The mean level on a 5-point scale of the different

component measures ranged from 1.49 to 1.81. It

www.theipi.org 542
www.freepsychotherapybooks.org
 TABLE 3
Means and Standard Deviations for Accuracy Dimensions (N =
43)

Accuracy Dimension Mean Standard Deviation

Wish (W) 1.81 .56
Response from other (RO) 1.49 .38
Response of self (RS) 1.69 .41
W + RO 1.65 .43

Note. Accuracy dimensions were rated on a 1-to-4 scale, with 1

indicating no congruence between the content of the
interpretation and the patient's CCRT and 4 indicating high
congruence.

www.theipi.org 543
www.freepsychotherapybooks.org
 might be argued on clinical grounds that this is not

really low, because clinicians may give piecemeal,

partial interpretations. (What would be of interest,

therefore, would be to examine the level of accuracy

of the top 20% so that we would know how accurate

the level is for the most accurate interpretations; we
should also do an initial study on what follows the
most and least accurate of interpretations.)

The Predictors of Accuracy Were Unrelated to Each

Other

The relationships among the predictors were

examined as a preliminary to the prediction of
outcome. An intercorrelation matrix for the four
predictors is given in Table 4. None of the

correlations attained statistical significance.

Accuracy of the Wish Plus Response From Other Was

the Best Predictor of Outcome

Multiple regression analyses were performed

using the two accuracy measures (wish plus response

www.theipi.org 544
www.freepsychotherapybooks.org
 TABLE 4
Intercorrelations of Predictors

Predictor 1 2 3 4
1. Accuracy of wish + response from other — .06 .12 -.11
2. Accuracy of response of self — .17 -.21
3. Helping Alliance Scale — -.08
4. Errors in Technique Scale —

www.theipi.org 545
www.freepsychotherapybooks.org
 from other, and response of self), the Errors in
Technique Scale, and the Helping Alliance Scale as

predictors, and rated benefits and residual gain as

outcome criteria. Simple correlations between each

predictor and the two outcome measures are given in

Table 5, as well as partial correlations (each variable
controlling for the others) and a multiple correlation
combining the predictors.

Most striking is the fact that the accuracy on the

wish plus response from other scale is the best

predictor of outcome, yielding statistically significant

results in all cases (both outcome measures and

simple and partial correlations). The Errors in
Technique subscale and the accuracy on the response
of self scale were not significantly related to outcome.
The Helping Alliance measure correlated

significantly with both outcome measures, as we

expected on the basis of a study by Luborsky, Crits-

Christoph, et al. (1983), which included a sample of

www.theipi.org 546
www.freepsychotherapybooks.org
 TABLE 5
Prediction of Outcome by Accuracy, Helping Alliance, and
Errors in Technique Measures (N = 43)

Simple Partial
Correlations Correlations
Rated Residual Rated Residual
Predictor
Benefitsa Gainb Benefitsa Gainb
Accuracy of wish + .38* .44** .36* .43**
response from other
Accuracy of .16 .07 .07 -.02
response of self
Helping Alliance .31* .36* .26 .35*
Counting Signs
Errors in Technique -.21 -.10 -16 -.04
Scale
Multiple R .49* .54**

a
Ratings of improvement. b Gain corrected for initial level.
*p < .05. **p < .01.

www.theipi.org 547
www.freepsychotherapybooks.org
 20 patients that overlapped with the sample of 43

used here. In addition, the Helping Alliance Scale

demonstrated a significant partial correlation with
residual gain and a near-significant correlation with

rated benefits. Thus, the predictive effects of accuracy

of interpretations and Helping Alliance Scales appear
to be independent.

The Impact of Accurate Interpretations Was Not

Limited to When the Alliance Was Positive

To test the reasonable hypothesis that accuracy

interacts with helping alliance (that is, that accurate

interpretations have an impact only when the

therapeutic alliance is positive), cross-product terms

between accuracy on the wish plus response from

other and helping alliance were entered after main

effects in the multiple regressions. These interactions

were nonsignificant.

Because one item (“failure to address maladaptive

behaviors or distorted apperceptions”) of the Errors in

www.theipi.org 548
www.freepsychotherapybooks.org
 Technique Scale overlapped conceptually with the

concept of accuracy of interpretation, it was of

interest to examine the correlations of this item with

the accuracy scales. For both accuracy scales, the
correlations were nonsignificant (r = –.11 for wish

plus response from other; r = –.19 for response of

self).

DISCUSSION

Congruence of Interpretations With the CCRT

We developed a reliable measure of the accuracy

of therapists’ interpretations that is based on their

congruence with the Core Conflictual Relationship

Theme, and we examined its relationship to the

outcome of dynamic psychotherapy, Accuracy of

therapists’ interpretations was assessed from two

early-in-treatment sessions of 43 patients receiving

moderate-length therapy. The results indicated that

accuracy with respect to the main wishes and

www.theipi.org 549
www.freepsychotherapybooks.org
 responses from others that were contained in the
relationship themes was significantly related to

outcome, even when we controlled for the effects of

general errors in therapist technique and the quality

of the helping alliance. The hypothesis that accurate

interpretations have their greatest impact in the

context of a positive helping alliance was not

confirmed. The main findings have provided new

information about the validity of the CCRT.

It is important to note that the interrater reliability

of the accuracy of interpretation scales was

reasonably high compared with the general levels of

reliability usually found in psychotherapy process
research. The specific nature of the rating task (the
scales were tailored to each patient’s CCRT) and the

use of experienced clinical research judges probably

contributed to the reliability level. By combining the

ratings made on all interpretations identified in each

of two complete therapy sessions, as well as

www.theipi.org 550
www.freepsychotherapybooks.org
 averaging the ratings over three judges, a robust

measure was constructed.

The major hypothesis of this research received

strong support: A statistically significant and

moderately strong relationship was found between

accuracy of interpretations (for the wish plus
response from other dimension) and treatment
outcome. In a later study of compatibility of
interpretations using the Plan Diagnosis (Norville,
Sampson, & Weiss, 1996) with seven patients in

therapy, a high correlation was found with outcome at

the end of therapy and 6-months after therapy ended.

In the study reported in this chapter, larger and more

diverse groups of patients and therapists were

examined, allowing for the first systematic

investigation of the relationship between accuracy of

interpretations and treatment outcome.

www.theipi.org 551
www.freepsychotherapybooks.org
 Our results suggest that the therapist’s technical

performance in dynamic psychotherapy has an impact

on outcome. The overall pattern of results also
suggests that a specific technique factor, not a more
general one, accounts for the finding. The predictive

strength of accuracy of interpretation on the wish plus

response from other dimension was not accounted for

by other variables such as errors in technique or

quality of the therapeutic alliance.

The approach used in the current study has

possible implications for research concerning the

effects of other treatment techniques on

psychotherapy outcome (Frank, 1979; Orlinsky &

Howard, 1978, 1986). Assessing the “quality” or

“skillfulness” of the treatment techniques under

investigation, as we did, may be necessary before

documented relationships between treatment
techniques and outcome are observed. This type of
research strategy has recently been advocated by a

www.theipi.org 552
www.freepsychotherapybooks.org
 few psychotherapy researchers (e.g., Schaffer, 1982,
1983; Silberschatz, Fretter, & Curtis, 1986) and has
now been confirmed by the significant association of
competence and outcome reported by Barber, Crits-
Christoph, and Luborsky (1996).

It is of interest that accuracy with respect to the

wish plus response from other, rather than accuracy
with respect to the response of self, predicted
treatment outcome. It appears that correctly
addressing the patient’s stereotypical patterns of

needs and wishes, the responses of others, is an

effective strategy. In contrast, limiting the focus of

interpretations to the patient’s usual responses of self

(typically, feeling states and symptoms) in
interpersonal situations may offer more limited
benefits. It may be that the responses of self are closer

to awareness than the wishes and expected responses

from others. It also may be that the wishes and

expected responses from others capture the main

www.theipi.org 553
www.freepsychotherapybooks.org
 facets of relationship conflicts that are antecedent to
and lead to symptoms; they often are, in the language

of cognitive-behavioral therapy, dysfunctional

attitudes.

Noninteraction With the Helping Alliance

The helping alliance predicts outcome

significantly, but it is independent of and
noninteractive with accuracy. The finding suggests
that the greater the accuracy of an interpretation, the
more it is beneficial, regardless of the state of the

alliance. The lack of a significant interaction between

accuracy of interpretation and the quality of the

therapeutic alliance was surprising, given the clinical
lore that a strong alliance is necessary for patients to
tolerate and make use of interpretations. Perhaps this
association would emerge with more severely

disturbed patients than the ones used here,

particularly if there were a higher frequency of poor

alliances present. In our study, only three therapist-

www.theipi.org 554
www.freepsychotherapybooks.org
 patient dyads showed no signs at all of a positive

alliance. It is of interest, however, that Crits-

Christoph, Barber, & Kurcias (1993) showed that the

accuracy on the wish plus response from other

strongly predicted changes in the helping alliance.

Nonsignificant Relationship With the Errors in

Technique Scale

The results for the Errors in Technique Scale are

discrepant with the findings in Sach’s (1983) study,

which showed a significant inverse relationship

between errors in technique and outcome. There are a

few possible reasons for the nonsignificant finding for
errors in technique in the current research. For one,
the relatively limited reliability of the Errors in

Technique Scale may partly explain the results. The

limited reliability may have been a function of the

generally low level of errors in this sample (four

items did not occur and several others occurred

infrequently). In addition, items on this scale may be

www.theipi.org 555
www.freepsychotherapybooks.org
 more appropriate for time-limited psychotherapy. The
treatment in Sachs’s (1983) study was specified as

brief therapy (a maximum of 6 months), whereas

open-ended therapy was used in the current research

(mean length of about a year).

Limits of Correlational Findings

Interpretation of the main findings in the study is

subject to the inherent limitations of all correlational

research. First, the direction of the relationship
between interpretations of the wish plus response

from other dimension and treatment outcome is not

clear. For example, it is possible that patients who are

making good progress in treatment are more likely to

elicit accurate interpretations from their therapists,

particularly if they are becoming aware of their own

relationship patterns and can articulate these issues

during the sessions. However, the fact that the finding

was observed very early in treatment (usually by the

www.theipi.org 556
www.freepsychotherapybooks.org
 fifth session) provides some support for the opposite
position—that accuracy leads to favorable outcome.

It is also possible that an alternative hypothesis,

or “third variable,” accounts for the relationship

between accuracy and treatment outcome. For

example, the complexity of a patient’s CCRT may
influence both accuracy and outcome. Perhaps
therapists are more likely to make accurate
interpretations with patients who have less
complicated, and therefore easier to discern,

relationship patterns. These patients may improve in

treatment, not as a result of the impact of accuracy

but simply because patients with less complicated
relationship patterns may make greater treatment
gains. An informal inspection of the CCRTs in the
sample, however, did not reveal any major

differences in the complexity of the relationship

patterns.

www.theipi.org 557
www.freepsychotherapybooks.org
 Another alternative hypothesis is that the therapist

uses accurate interpretation more often on patients

who are healthier psychologically (that is, exhibiting

lesser psychiatric severity). We know that there is a
trend for patients with better psychological health to

show greater improvement in psychotherapy

(Luborsky, Crits-Christoph, et al., 1988).

It should be mentioned that focusing

interpretations on issues that are not captured by the

patient’s CCRT may also be important. For example,

it would be interesting to learn whether focusing on

defenses is related to patient improvement. (The

CCRT does not require the judge to infer types of

defenses.) Inspection of the interpretations in the

sample did not reveal any consistent differences

among therapists in focusing on defense mechanisms.

In fact, with only a few exceptions, the content of the

interpretations appeared to fit the structure of a
typical CCRT formulation: wishes, responses from

www.theipi.org 558
www.freepsychotherapybooks.org
 others, and responses of self. Formal research in this

area is clearly needed.

A further caveat is in order here. Our use of the

word accuracy is only shorthand for convergence of

the interpretation content with the CCRT content. Our

measure of accuracy is not based on an external

criterion of validity—if such a criterion is even

possible—and it does not deal with the larger concept

of the adequacy of the interpretation, that is, accuracy

plus timing and tact.

CONCLUSIONS

The net gain from our new measure of accuracy

was these three findings:

• Our operational measure of accuracy of

interpretations was based on their congruence
with the Core Conflictual Relationship Theme.
The measure turned out to be reliable and it
showed validity in terms of a significant
correlation with outcomes of psychotherapy.

www.theipi.org 559
www.freepsychotherapybooks.org
 • Accuracy with respect to the congruence of the
main wishes and responses from others with the
interpretations was significantly related to
outcome even after controlling for the effects of
errors in therapist technique and the quality of
the helping alliance.

• The hypothesis that accurate interpretations would

have their greatest impact in the context of the
positive helping alliance was not confirmed.

Where do these findings lead in terms of clinical

and research directions? We believe that the present

operational definition of accuracy has advantages

over our early measure and that both of the accuracy

measures have advantages for research and clinical
practice over the clinical definition from which they

sprang. In clinical terms, our findings imply that an

effective interpretation tends to be a therapist
response that presents to the patient a part of the
patient’s relationship pattern, both what is wished for
and the expectations from others. In addition, the
selection of the part of the patient’s relationship

www.theipi.org 560
www.freepsychotherapybooks.org
 pattern that is presented in interpretations should not

be too far removed from the patient’s awareness

(Bibring, 1954; Freud, 1912/1958d).

It is fitting here to reassure clinical readers that,

on the basis of applying the measure to the sample of

sessions from the Penn Psychotherapy Project, we
found that therapists already tended to make
interpretations that were consistent with our measure.
The content of the usual interpretations appeared to
fit the structure of the typical CCRT formulation:

wishes, responses from others, and responses of self.

Yet the therapists in the sample were doing what they

understood to be dynamic psychotherapy as it existed

before the era of psychotherapy manuals.

Several other operational recastings of the clinical

concept of interpretation have been suggested, as we
mentioned earlier. One of the best known is that
proposed by Weiss et al. (1986), in which accuracy of

www.theipi.org 561
www.freepsychotherapybooks.org
 interpretation is evaluated by the congruence of the
interpretation with the patient’s “unconscious plan”
as measured by the Plan Diagnosis method
(Rosenberg, Silberschatz, Curtis, Sampson, & Weiss,
1986). We expect many more operational measures to

be generated in the next few years for the concept of

accurate interpretation, and clinicians and researchers

will be able to choose from these the best for their

purpose.

In future studies of the accuracy of

interpretations, we intend to apply our method to a

larger number of sessions than the two early ones

used in the present study. It is likely that the
predictability of accuracy will be heightened by the
increase in number of sessions. We will also learn
then how consistent the therapist’s behavior is and

whether it is equally predictive at different points in

the treatment.

www.theipi.org 562
www.freepsychotherapybooks.org
 Our accuracy finding, if replicated, could have
several implications for the teaching of the practice of
psychodynamic psychotherapy. For example,
therapists could be trained in formulating their
patients’ CCRTs and in correctly addressing these

issues in their interpretations. Assuming that the

therapist in this study are representative of the

general population of psychodynamic clinicians, the

relatively low mean scores for accuracy in this

study’s sample of therapists suggests that there is

much room for improvement in the quality of
therapists’ interpretations. These findings further
imply that when the improvement occurs, it will be

associated with improved patient outcomes.

Note

[8] From “The Accuracy of Therapists Interpretations and the

Outcome of Dynamic Psychotherapy,” by P. Crits-
Christoph, A. Cooper, and L. Luborsky, 1988, Journal of
Consulting and Clinical Psychology, 56, pp. 490-495.
Reprinted by permission.

www.theipi.org 563
www.freepsychotherapybooks.org
 14

SELF-UNDERSTANDING OF THE CCRT

PAUL CRITS-CHRISTOPH AND LESTER

LUBORSKY9

In psychotherapy most patients need to increase

their self understanding as a way to help achieve their

treatment goals. This proposition is clinically

accepted as central among the curative factors in

dynamic psychotherapies (Luborsky, 1984). Both the

clinical views and the quantitative studies of the

relation of insight to treatment outcomes, as reviewed

by Crits-Christoph, Barber, Miller, and Beebe (1993),
stress this factor. Yet few quantitative studies exist of
the association of self-understanding with the

outcome of psychotherapy. The studies reviewed by

Luborsky et al. (1988) have provided the following

www.theipi.org 564
www.freepsychotherapybooks.org
 results: Of three studies that measured insight
pretreatment (Raskin, 1949; Stein & Beall, 1971;

Zolik & Hollon, 1960), the last two showed

significant prediction of outcomes; but of two studies

in which the level of insight was measured during

psychotherapy (Rosenbaum, Friedlander, & Kaplan,
1956; Morgan, Luborsky, Crits-Christoph, Curtis, &
Solomon, 1982), neither showed a significant

prediction of outcome. Therefore, it is a fair verdict

that self-understanding measures have tended to vary

in their significance as predictors. The studies are

weak in other ways as well. Several investigators

relied on single-item ratings of insight and did not

present reliability data; only two of the studies were

based on psychodynamic psychotherapy; and all of

the measures of self-understanding were unguided

clinical ratings—each judge defined insight as he or
she saw fit.

www.theipi.org 565
www.freepsychotherapybooks.org
 In contrast, the measure of self-understanding that
we have begun to develop is a guided clinical rating.

The judge assesses the extent to which the patient has

developed self-understanding relative to a specific,

independent criterion: the Core Conflictual

Relationship Theme. Because guided clinical ratings
have generally been found to have better predictive
validity than unguided ratings (Holt, 1978), our use

of a guided clinical rating represents a potential

methodological advance.

The purpose of the present study was to develop a

method that would allow us to assess the extent to

which patients in the Penn Psychotherapy Project
acquired self-understanding of their central
relationship patterns. If we could develop such an
operational measure, we could then examine a main

theory of psychoanalytic psychotherapy: that

improvement is related to gains in this form of self-

understanding.

www.theipi.org 566
www.freepsychotherapybooks.org
PROCEDURE

Patients and Sessions

The sample used for this study consisted of the

same 43 patients used in the study of accuracy of

interpretation (see chapter 13, this volume). These 43
were a subset of the total sample of 73 patients in the
Penn Psychotherapy Project (Luborsky et al., 1988).

Two sessions drawn from the early part of

treatment were used to score our measures of self-
understanding and also to score the CCRT. The

CCRT method was generally scored on Sessions 3

and 5, but occasionally a third session was needed to

obtain the minimum number of 10 relationship

episodes needed for the method. For the ratings of
self-understanding, only the two sessions that
contained the most relationship episodes were used.

Trained judges working independently were used

for each task. The judges were clinical psychologists,

www.theipi.org 567
www.freepsychotherapybooks.org
 psychiatrists, and research assistants highly familiar
with the methods.

Measures

The CCRT Measure

The CCRTs represented a composite of individual

judges’ CCRTs; they were composed of up to two

wishes, three responses from other, and three
responses of self.

Self-Understanding of the CCRT

The items that constituted the self-understanding

rating scale were derived mostly from the principles

provided in Luborsky’s (1984) manual for
supportive-expressive psychoanalytic psychotherapy.
The items tap the patient’s self-understanding in
different areas. Ratings, each on a 5-point scale from

none to very much, were made of self-understanding

of (a) the CCRT in general, (b) the CCRT in

relationship to the therapist, (c) the CCRT in

www.theipi.org 568
www.freepsychotherapybooks.org
 relationship to parents, and (d) the CCRT in
relationship to each of two main other people

discussed by the patient within each session; these

people were the most frequent topics of relationship

episodes.

For each of these areas, separate ratings were

made for each wish, response from other, and
response of self present in each patient’s CCRT. The
ratings of all wishes, responses from others, and
responses of self were averaged for each patient to

create four final scores corresponding to the four

areas listed previously.

The judges who rated self-understanding were

given the following information for each patient: (a)

transcripts of two early sessions, (b) a list of the two

main other people for each session, and (c) the CCRT

formulation. Judges read each transcript and then

rated the session as a whole on the self-understanding

www.theipi.org 569
www.freepsychotherapybooks.org
 scale items. This operational measure of self-

understanding reflects a clinical judge’s rating of the

degree of convergence of what the patient is aware of

understanding in a session with what independent
clinical observers have judged to be the CCRT.

Health-Sickness Rating Scale

The Health-Sickness Rating Scale (HSRS)

(Luborsky, 1975; Luborsky, Diguer, et al., 1993) was
included so that we could examine gains in self-

understanding controlling for the patient’s general

level of psychological health. This control was
considered necessary because the level of self-

understanding might be a function of psychological

health.

Outcome Measures

Two composite outcome measures were

employed: measures of change (corrected for initial

www.theipi.org 570
www.freepsychotherapybooks.org
 level) and ratings of benefits (see Luborsky et al.,

1988).

RESULTS

Reliability of Judging Self-Understanding Was Good

Interjudge reliability was assessed for the four

scores from the self-understanding scale (global,

therapist, parents, other people) and a total combining

the four subscales, using the intraclass correlation
coefficient. The results are given in Table 1. It can be
seen that interjudge reliability was quite good. Scores
for the two judges were combined for subsequent

analyses.

Level of Self-Understanding Was Low

The l-to-5 rating scale defined an average level as

2.5; according to this definition, the mean levels

shown in Table 2 were generally low. The mean level

of self-understanding in Session 3 was compared with

that in Session 5 to assess whether there was any

www.theipi.org 571
www.freepsychotherapybooks.org
 TABLE 1
Interjudge Reliability of the Self-Understanding Scale

Self-Understanding Pooled Judges’ Intraclass

Scale Item Correlation
Global .77
Therapist .87
Parents .89
Other People .87
Total (sum of subscales) .85

www.theipi.org 572
www.freepsychotherapybooks.org
 noticeable increase in self-understanding over these
early sessions. Means, standard deviations, and the

results of paired t tests on each of the four subscales

are shown in Table 2. In general, the level of self-

understanding of the CCRT was low in these early

sessions; it remained at about the same level from
Session 3 to Session 5 except for a small but
significant decrease in self-understanding toward the

therapist. No other significant changes were found.

Level of Self-Understanding Predicted Outcome;

Change in Self-Understanding Did Not

Prediction of outcome of psychotherapy was

performed in two ways. First, level of self-

understanding (averaging Sessions 3 ad 5) was

correlated with the two outcome measures. Because it
might be expected that the healthier patients would
display more insight, these correlations were done
controlling for patients’ pretreatment level of health-
sickness using the HSRS. Second, change in self-

www.theipi.org 573
www.freepsychotherapybooks.org
 TABLE 2
Mean Self-Understanding for Session 3 and Session 5 (N = 43)

Self-Understanding Session Session Paired

SD SD
Scale Item 3 5 t Test
Global 2.21 (.7) 2.31 (-9) .7
Therapist 1.66 (.9) 1.33 (.5) 2.4*
Parents 2.31 (1.1) 2.13 (1.1) .9
Other people 2.50 (.7) 2.74 (.8) 1.6
Total score 2.16 (.5) 2.12 (.5) .4

Note. Self-understanding was rated on a 5-point scale.

*p < .05.

www.theipi.org 574
www.freepsychotherapybooks.org
 understanding from Session 3 to Session 5, correcting

for initial (Session 3) level of self-understanding by

regression analysis, was correlated with outcome,

again partialing for pretreatment scores on the HSRS.

Although this is a short period of time to assess gain

in insight, and there were no significant increases in
insight over this period, previously reported results
(see chapter 13, this volume) had indicated that the

therapist’s interpretations in these sessions had an

impact on outcome. It might be possible, therefore, to

detect trends toward increasing self-understanding in

patients who improve the most in treatment.

The results of the predictive analyses are given in

Table 3. Change in self-understanding from Session 3

to Session 5 yielded no significant partial correlations

with the two outcome measures. Level of self-

understanding of CCRT, however, produced two
significant relationships with outcome: (a) Self-
understanding of the CCRT in relation to the therapist

www.theipi.org 575
www.freepsychotherapybooks.org
 TABLE 3
Prediction of Outcome From Level of and Change in Self-
Understanding (N = 43)

Outcome Measures
Rated Residual
Self-Understanding Scale Item
Benefits Gain
Level (Session 3 + Session 5)
Global .03 .16
Therapist .31* .29
Parents -.28 -.30
Other people .22 .34*
Total .04 .12
Change from Session 3 to Session 5
Global .06 .11
Therapist .23 .25
Parents -.23 -.22
Other people -.07 .11
Total -.07 .04

Note. Coefficients are partial correlations, controlling for Health-

Sickness Rating Scale score and, in the case of change in self-
understanding, for Session 3 level.

*p < .05.

www.theipi.org 576
www.freepsychotherapybooks.org
 correlated with rated benefits (.31, p < .05), and (b)
self-understanding of the CCRT in relation to other

people demonstrated a significant partial correlation

with residual gain (.34, p < .05).

SUMMARY, DISCUSSION, AND CONCLUSIONS

The Findings and Their Meaning

• The results provide evidence that our measure of

self-understanding of specific relationship
themes can be rated reliably.

• We found evidence that there are associations

between the level of self-understanding and
outcome. These associations between our
measure of self-understanding and outcome are
small, however, and we are not able to decide at
this time on the relative merits of two possible
interpretations of the results: (a) that self-
understanding is hard to operationalize in a
meaningful way or (b) that any measure of it will
achieve low associations with outcome because
clinical theory about the necessity of self-
understanding is not well founded.

www.theipi.org 577
www.freepsychotherapybooks.org
 • Several methodological issues lead us to be
cautious in interpreting our findings. The specific
hypothesis that gain in self-understanding is
related to more favorable outcome was not
confirmed, although we believe that this
hypothesis was not given an adequate test. It
probably is necessary to evaluate change in self-
understanding over a longer period of treatment
than two early sessions. To evaluate change over
a longer period of time on a specific criterion
(that is, the CCRT), it would be necessary for
that criterion to be equally relevant to both the
early and later points in therapy. In the case of
the treatments studied here, this condition may
not have applied: The relatively large number of
sessions on the average (mean number of
sessions was 53.5 for the 43 patients) meant that
it was likely that some change in the content of
the clinical material and even some change in the
CCRT would occur over therapy. In a brief focal
dynamic therapy, it might be a simpler task to
track level of self-understanding of the same
content over the course of the whole treatment.

Although the significant results for level of

self-understanding are promising, the

www.theipi.org 578
www.freepsychotherapybooks.org
 correlational nature of the study limits inferences
about the causal role of self-understanding.
Statistical control of one major variable,
psychological health-sickness, allows us to rule
out one potential third variable as an explanation
of the results, yet other third variables may well
exist.

• To the extent that we can speculate from these data,

it is of interest that ratings of self-understanding
of the CCRT in relation to the therapist and to
other people were predictive but self-
understanding of the CCRT in a global sense or
in relation to parents was not. The major
importance of the relationship pattern with the
therapist is of course evident in many clinical
theories of psychodynamic psychotherapy
(Luborsky, 1984; Strupp & Binder, 1984). The
data may be more consistent with a focus on the
“here and now” relationships of the patient and
less on the past parental relationships, although
we did not specifically code each relationship for
past versus present.

The Search for Better Operational Measures

www.theipi.org 579
www.freepsychotherapybooks.org
 The only measure that has been partially

examined in this chapter is the degree of convergence

of the patient’s self statements with the independently

established CCRT. Other reasonable types of

operational measures can be derived from the clinical

observations reported by Luborsky (1984, pp. 124-
25) about the kinds of improvements in self-
understanding that develop in the course of

psychotherapy. The gains are observed to appear in

five stages, (a) Early in psychotherapy, the patient

reports involvement in relationship interactions with

the therapist and other people, but there is relatively
little ability to recognize the patient’s own usual
relationship pattern, (b) Later in therapy, the
experience of having one’s own increasingly familiar
pattern becomes clearer to the patient. The effect of

this recognition of one’s own pattern is that the

patient is better able to distinguish between what he

or she brings to relationship problems and what the

www.theipi.org 580
www.freepsychotherapybooks.org
 external circumstances bring, (c) The patient

develops a healthy respect for how deeply embedded

in his or her personality the central relationship

problem is. It leads, as one patient said, to an

appreciation of its “slippery power to reappear.” (d)

The recognition of having one’s own central
relationship pattern leads to other developments,
including a greater understanding of how the pattern

might have originated, usually in relation to

interactions with the parents, (e) It is also crucially

beneficial that the recognition lead to a range of new

ways of dealing with and mastering the relationship
problems in the patterns.

An example of the benefits provided by a greater

understanding of the pattern is exemplified by the

following words of Mr. Howard in Session 83, with

which he describes his feelings about the therapist: “I
felt bad because of distrusting our relationship. I saw
an article that makes me distrust you. I realize that I

www.theipi.org 581
www.freepsychotherapybooks.org
 distrust everybody.” It is clear in this abbreviated
example that the patient became more distrustful on

seeing the article. But it is also clear that the patient

was able to distinguish the two sources of distrust: the

article and his central relationship pattern, which

included a readiness to distrust.

It might be useful to compare a set of measures

that is based on these clinical observations with an
expanded operational measure of self-understanding
of the kind reported in this chapter. The set would be

composed of rating scales that reflect the degree to

which the patient experiences each of the five stages

listed previously. Finally, in future studies we plan to
try a self-report approach (described in chapter 15,
this volume) but using as part of a postsession
questionnaire items that include the five stages of

self-understanding as well as the patient’s reports of

their own capacity to see the degree of self-

applicability of the four specific relationship themes

www.theipi.org 582
www.freepsychotherapybooks.org
 of the CCRT: global, therapist, parents, and other

people.

• The data presented in this chapter have taken us by

a new route one step toward testing aspects of
the clinical concepts of dynamic psychotherapy.
This study examined our experiences in
investigating an operational measure of the role
of self-understanding in psychotherapy and
offers suggestions about where to look in future
studies for the therapeutic “action” in the
patient’s level of and change in self-
understanding during psychotherapy.

Note

[9] An earlier version of this chapter appeared in L. Luborsky, P.

Crits-Christoph, J. Mintz, and A. Auerback, 1988, Who
Will Benefit From Psychotherapy? New York:
HarperCollins. It has been adapted, revised, and printed
with permission of the publisher.

www.theipi.org 583
www.freepsychotherapybooks.org
 15

THE PERSPECTIVE OF PATIENTS

VERSUS THAT OF CLINICIANS IN THE
ASSESSMENT OF CENTRAL
RELATIONSHIP THEMES

PAUL CRITS-CHRISTOPH AND LESTER

LUBORSKY

In carrying out the CCRT method, clinical judges

extract relationship patterns from narratives about

interactions with other people. Although the clinical

judges do the task with reliable and valid results,

some of the procedures have drawbacks. A major one

is that making transcripts of psychotherapy sessions

to be scored by clinical judges is extremely time

consuming. Then the clinical judges themselves
spend, on the average, about 1½ to 3 hours to score
one session. Thus, large-scale research with the

www.theipi.org 584
www.freepsychotherapybooks.org
 CCRT method is difficult; and the other transference-
related measures generally require even more time

(see chapter 20).

AIM

Our purpose in this study is to examine the

possibility of obtaining CCRT information directly
using a self-report method. Although there will
always be a role for clinician-derived measures in

data sets when self-report questionnaires are not

available (for example, with psychotherapy tapes) or

need to be amplified, the practical advantages of a
self-report measure for research could be great. In
addition, a comparison of self-reported relationship
patterns with clinical judge-scored patterns may tell

more about the nature of a core conflictual theme in

terms of which aspects are typically conscious and

which may not be fully in awareness. According to

psychodynamic theories, some part of a conflictual

pattern is generally outside of a patient’s awareness.

www.theipi.org 585
www.freepsychotherapybooks.org
 It is not known, however, how much the CCRT or the
other judge-scored measures discussed in chapter 20

routinely assess major aspects of a conflictual

relationship pattern that are unconscious.

We proceeded with the development of a self-

report CCRT measure well aware of the potential
pitfalls of self-report methods. One main problem is
that people may be inaccurate in their reporting of
their own behavior (Nisbett & Wilson, 1977) for a
variety of reasons. For example, when the frequency

of occurrence of a behavior is very low, people are

less likely to be able to describe retrospectively what

went on in the situation in which it was expressed.
Social desirability effects can also distort self-report
measures.

On the other hand, it seems likely that important

parts of some major relationship patterns are readily

available to conscious self-report. Such relationship

www.theipi.org 586
www.freepsychotherapybooks.org
 themes can be part of highly redundant patterns, so
that interpersonal interactions with friends, parents,

colleagues, and others that might contain such themes

occur with high frequency for most people. The

ability to observe and describe such patterns may be

an important individual difference variable, however,
perhaps highly correlated with psychological
mindedness or interpersonal awareness. Self-

knowledge about interpersonal patterns may also be

acquired through feedback from other people

concerning such patterns.

PROCEDURE

The data collection centered on a comparison of

the following three main measures:

Self-Report CCRT Questionnaire (SR-CCRT);

(Crits-Christoph, 1986). In this questionnaire
participants simply rate on a l-to-5 scale each of
a number of wishes and responses on how
typical those wishes and responses are of their
relationships. In addition, respondents are asked

www.theipi.org 587
www.freepsychotherapybooks.org
 to describe in a sentence or two what their main
conflicts are.

Self-Interpretation of RAP Narratives (SI-RAP).

This instrument is the self-interpretation (SI-
CCRT) of the RAP interview (see chapter 7).
During the RAP interview, participants are asked
to relate detailed episodes of actual interactions
with other people, including what happened and
what was said during the interaction. The
instructions ask for reports of interactions that
were emotionally involving and included a
significant other person (such as parents, friends,
or lovers). The self-interpretation procedure
comes after all of the narratives have been told.
After telling each of the narratives, subjects are
asked to rate each of a number of wishes and
responses on the degree to which the wishes and
responses apply to their behavior in the reported
interaction. They are also asked to describe what
conflict, if any, was present for them in each
interaction. We elicited 10 such narratives in the
current study.

Standard Clinician-Judged CCRT of RAP

Narratives (CCRT). This measure is the standard
CCRT measure scored by a clinical judge from
the RAP interview.

www.theipi.org 588
www.freepsychotherapybooks.org
 Although each of the three measures assessed all
three CCRT components (wishes, responses from

other, and responses of self), this presentation is

reduced to a single focus only on the results for the

wishes. Two groups of subjects were used for these

studies. The first sample consisted of 70 University of
Pennsylvania undergraduates who completed the SR-
CCRT Questionnaire. One week later, 30 of these

students were readministered the questionnaire to

assess test-retest reliability.

The main comparison of the SR-CCRT

Questionnaire, SI-RAP, and clinical judge-scored

CCRT was performed on the second sample,
consisting of 16 patients. These patients were all
receiving psychotherapy at the outpatient psychiatry
clinic at the University of Pennsylvania.

RESULTS

Self-Report CCRT Questionnaire (SR-CCRT)

www.theipi.org 589
www.freepsychotherapybooks.org
 EXHIBIT 1
Wishes Rated in the CCRT Self-Report Questionnaire

1. To assert my independence and autonomy

2. To dominate; to impose my will or control on others
3. To overcome other’s domination; to be free of
obligations imposed by others; to not be put down
4. To win in competition with another; to be better than the
other person
5. To win someone’s affection or attention, through
competition with another person
6. To submit, to give in, to be passive
7. To make contact with others, to be close, to be friends
8. To receive affection, to not be deprived of continued
affection
9. To be receptive (to open up) to others
10. To please the other person
11. To avoid hurting the other person
12. To get sexual gratification
13. To receive acceptance; to be respected, recognized,
approved, vindicated, reassured; to maintain self-
esteem
14. To be fairly treated
15. To get help, care, protection, and guidance from others
16. To achieve, be competent, be successful
17. To hurt the other person, to get back at the other; to
express anger, hostility, or resentment to the other
person
18. To exert control over myself

Note. Wishes are from Edition 1 Standard Categories (see chapter 3,

this volume).

www.theipi.org 590
www.freepsychotherapybooks.org
 The item pool for the SR-CCRT measure was

drawn from the standard categories list, Edition 1, of

wishes developed as an aid in reliability studies (see

chapter 3, this volume). These 18 wishes (see Exhibit
1) were distilled from the set of wishes obtained

empirically by clinical judges who applied the CCRT

method to a different sample of 16 patients.

Retest Reliability Was Moderate; Internal

Consistency Was Marginal

Test–retest correlations were computed for each

of the 18 wish items. The results (see Table 1)
indicated that item reliability ranged from .24 to .90,
with a median reliability of .61. Although the test-

retest reliability results were adequate, the internal

consistency (Cronbach’s alpha) coefficients were

marginal. Subsequent analyses were performed at

both the item and factor levels.

Three Factors Were Found

www.theipi.org 591
www.freepsychotherapybooks.org
 TABLE 1
Reliability of Self-Report CCRT Questionnaire on Wishes

Wish Item Test-Retest Correlation

1. To assert myself .61
2. To dominate .72
3. To overcome domination .73
4. To be better than others .66
5. To win attention .53
6. To submit .62
7. To be close .60
8. To get affection .24
9. To be receptive .70
10. To please .72
11. To avoid hurting .90
12. To have sex .64
13. To get acceptance .61
14. To be treated fairly .34
15. To get care .40
16. To achieve .50
17. To hurt .57
18. To exert control .47

www.theipi.org 592
www.freepsychotherapybooks.org
 The existence of higher order dimensions within

the set of 18 items was examined by factor analysis

(principal components analysis, varimax rotation)
using the responses of the 70 students. Inspection of
the successive solutions indicated that a three-factor

solution was most appropriate. Table 2 gives the

factor loadings and reliability statistics for the three

factors.

Interpretation of the factors was straightforward.

The first factor contained wishes related to giving and

getting affection and intimacy in relationships and

was labeled Wish for Closeness. The second factor
consisted of the wishes (a) to dominate, (b) to be
better than the other person, and (c) to get back at the
other person. This factor was described as a

Competition dimension. The third factor,

Independence, was characterized by high loadings on

wishes related to independence, achievement,

overcoming others’ domination, and desire to be

www.theipi.org 593
www.freepsychotherapybooks.org
 TABLE 2
Factor Analysis of Wishes

Independence
Competition
Closeness
Factor Loadings
7. To be close .75
8. To get affection .68
9. To be receptive .68
10. To please .66
11. To avoid .59
hurting
15. To get care .50
2. To dominate .66
4. To be better .70
than others
17. To hurt .70
1. To assert .62
myself
3. To overcome .65
domination
14. To be treated .60
fairly
16. To achieve .72
Reliability
Internal .67 .52 .60
consistency
Test-retest .71 .74 .82
Note. Loadings are listed only for the variables that defined the
subscales. Five variables did not load on any of the three main
factors.

www.theipi.org 594
www.freepsychotherapybooks.org
 treated fairly. The factors were scored by summing
the salient items as listed in Table 2.

Self-Interpretation of the RAP Interview

The Self-Interpretation of the RAP (SI-RAP)

procedure (Luborsky, 1978b) yielded patients’ ratings

of the 18 wishes for each of the 10 relationship

episodes. Of initial interest was the extent to which

patients were consistent in their self-ratings of themes

across 10 items. For example, if a patient described a
relationship episode as containing a wish for

closeness, was the patient more likely to rate this

wish as present in other relationship episodes as well?

Put another way, can we construct a scale of a

patient’s characteristic level of experienced wishes by

aggregating over the multiple episodes?

Consistency Over Narratives Was High

This question was answered by computing

Cronbach’s alpha coefficient for each of the 18

www.theipi.org 595
www.freepsychotherapybooks.org
 wishes, considering the 10 replications to be

analogous to items on a test. Table 3 presents the

results. It can be seen that the internal consistencies

of the aggregate scores were all reasonably high, with

the exception of the wish for sex item. The low (.40)
value for wish for sex is not surprising, given that this
item focuses more on a particular behavior than on a
general psychological need or intention (such as a

wish to be close to others) that might be manifested

behaviorally in many different ways. In addition, the

RAP test requested narratives about a variety of

people (for example, parents) for whom a conscious
wish for sexual gratification and a narrative about it
would be unusual. The high internal consistencies for
the other wishes indicate that these patients perceived
the same themes to be present in many relationships.

Correlation With the Self-Report CCRT Was Poor

Table 4 presents the correlations between the

aggregate scores from the RAP (averaging over the

www.theipi.org 596
www.freepsychotherapybooks.org
 TABLE 3
Self-Interpretation of RAP: Consistency of Wishes Across 10
Relationship Episodes

Wish Item Internal Consistency

1. To assert myself .80
2. To dominate .86
3. To overcome domination .66
4. To be better than others .84
5. To win attention .88
6. To submit .91
7. To be close .85
8. To get affection .88
9. To be receptive .83
10. To please .88
11. To avoid hurting .85
12. To have sex .40
13. To get acceptance .81
14. To be treated fairly .88
15. To get care .81
16. To achieve .86
17. To hurt .74
18. To exert control .81

Note. Internal consistency was computed using Cronbach's alpha.

www.theipi.org 597
www.freepsychotherapybooks.org
 TABLE 4
Correlations Between Self-Report and Self-Interpretation

Wish Item Correlation

1. To assert myself -.03
2. To dominate .41
3. To overcome domination .54*
4. To be better than others .58*
5. To win attention .23
6. To submit -.14
7. To be close -.15
8. To get affection -.18
9. To be receptive .18
10. To please .16
11. To avoid hurting -.21
12. To have sex .19
13. To get acceptance .05
14. To be treated fairly -.11
15. To get care -.25
16. To achieve -.06
17. To hurt .43
18. To exert control -.14
Factor
Closeness -.14
Competition .79*
Independence .20

*p < .05.

www.theipi.org 598
www.freepsychotherapybooks.org
 10 episodes for each wish) and the corresponding
wish rating from the SR-CCRT Questionnaire. One

would expect that if a patient rated a specific wish

highly, for example, the item Wish “to assert myself,”

in many different relationship episodes on the RAP,

the patient also would rate this wish as typical of
relationships on the SR-CCRT Questionnaire. The
results in Table 4 indicate that the two measures were

generally not correlated, however. An exception was

the Competition factor, for which there was a .79

correlation between the Self-Report CCRT and the

Self-Interpretation of the RAP scores. The items

constituting this factor (Items 2, 4, and 17) as well as

Item 3 (wish to overcome domination) showed

modest correlations.

There may be several reasons for the general lack

of correlation between the levels of the same wish in

the two tests, (a) The process of telling RAP

narratives and then interpreting them probably leads

www.theipi.org 599
www.freepsychotherapybooks.org
 to increased self-awareness about one’s relationship

patterns. The Self-Report CCRT questionnaire was

completed, therefore, before the self-interpretation
process was initiated. Completion of the Self-Report

questionnaire after the Self-Interpretation RAP may

yield more agreement, (b) The narratives about
relationship episodes chosen by the patient for the
RAP interview may not be representative of the

patient’s typical interactions with others. For

example, a bias toward remembering problematic

interactions rather than typical interactions would

lead to a lack of concordance between the two
measures, (c) Finally, social desirability effects are
likely to be more pronounced for questions asking
about “your typical needs and desires” than for
questions about specifically exemplified interactions,

thereby leading to disagreements.

Clinical Judge’s Scoring: Comparisons With Self-

Report and Self-Interpretation

www.theipi.org 600
www.freepsychotherapybooks.org
 To facilitate comparison of the three methods, we

first selected for each patient the first and second

most frequent wishes as scored by the clinical judge.

These wishes were then translated into the wordings
used in the list of 18 wishes. These wishes had

average frequencies of 7.6 and 4.4, respectively, out

of the 10 RAP episodes. We then examined the

patient’s ratings on the same wishes from the Self-

Report CCRT Questionnaire. For example, if the

clinical judge identified the wish to dominate others

as the most frequent wish from the RAP, we
examined the patient’s ratings on this wish, that is,
how much this wish was typical of the patient’s

relationships in general (Self-Report CCRT

Questionnaire) and, on the average, how much this

wish applied to the RAP episodes (Self-Interpretation

of RAP). In deriving the patient’s self-interpretation

mean ratings on the wishes the clinical judge scored,
we chose to include only episodes in which the

www.theipi.org 601
www.freepsychotherapybooks.org
 clinical judge identified the wish as being present.

Thus, we asked the question: When a clinical judge

infers a main wish to be present in a relationship

episode, to what degree does the patient rate this wish

as applicable to the same relationship episode? We

were also concerned with the possibility that patients
have a general sense of the nature of their main
wishes, which may be similar to that of the clinician,

but that they do not choose the identical wish that the
clinician identified. For this reason, we turned to the

subscales that emerged from the factor analysis of the

Self-Report CCRT Questionnaire completed by the
student sample. We speculated that average score of
the patients on the particular subscale containing the
wish item identified by the clinician would provide an
index of whether the patient was in the ballpark in

designating which wishes applied to his or her

relationships.

www.theipi.org 602
www.freepsychotherapybooks.org
 Patients Tended to Give High Ratings to Wishes
Identified by the Clinicians

Table 5 displays the results comparing the SR-

CCRT and the SI-RAP with the clinical judge’s

CCRT scoring. It can be seen that patients generally

rated highly the wishes that were identified by the
clinicians. For example, the main wish identified by
the clinician was rated, on the average, 4 0 on the 5-

point scale (5 = very typical of me) of the Self-Report

CCRT Questionnaire, but it is conceivable that

patients rate all wishes highly and therefore were not

making the same discrimination that was made by the
clinician. The column Mean Rating for Other Wishes
in Table 5 allows a comparison of the patients’

ratings of the wish identified by the clinician with the

patients’ ratings of other wishes not identified by the

clinical judge. The statistical significance of these

comparisons was tested with a paired t test.

www.theipi.org 603
www.freepsychotherapybooks.org
 TABLE 5
Comparison of Patient’s Report With Clinician-Identified
Wishes

Mean Rating for Mean Rating

Significance
Measure Wish Identified by for Other
(p Value)
Clinician Wishes
Main wish
Self-Report
Item 4.0 3.5 .04
Subscale 3.5 3.5 ns
Self-Interpretation
Item 3.7 2.7 .001
Subscale 3.5 2.8 .001
Secondary wish
Self-Report
Item 3.7 3.6 ns
Subscale 4.0 3.7 ns
Self-Interpretation
Item 3.6 2.7 .004
Subscale 3.3 2.8 .06

Note. Ratings were made on a 1-to-5 scale, in which 5 = very typical of

me.

www.theipi.org 604
www.freepsychotherapybooks.org
 Although patients tended to rate many other
wishes as applicable, discrimination among the

wishes was apparent. This was particularly true for

the Self-Interpretation RAP, on which, for example,

patients rated the clinical judge’s main wish 1 point

higher on the rating scale than they rated other wishes
(3.7 vs. 2.7, p < .001).

However, the Self-Report CCRT Questionnaire

provided poorer discrimination, especially for the
secondary wish identified by the clinical judges.

Examination of subscale scores did not improve

discrimination for either the Self-Report CCRT

Questionnaire or the Self-Interpretation of the RAP.

Because mean comparisons can obscure

important individual differences, we inspected each

patient’s ratings in comparison to the clinician’s

CCRT. Fourteen patients showed ratings consistent

with the mean differences. Two patients, however,

www.theipi.org 605
www.freepsychotherapybooks.org
 had somewhat lower ratings for the main wish that

the clinician identified than for other wishes. Whether

this reflects a selective lack of awareness of an
important wish or is simply a chance finding with less

than perfectly reliable measures is not known.

Finally, we compared the patients’ description of

their main conflicts (written out as part of the Self-
Report CCRT Questionnaire and Self-Interpretation
of the RAP tests) with the clinician’s description of
the main conflict in the CCRT. In only 25% of the

cases did the patients’ descriptions from either

measure reasonably match the content of the

clinician’s description of a conflict. In most of the

mismatches, the patients attributed the “conflict” to

problematic people in their lives, whereas the

clinician saw the conflict as intrapsychic (for

example, as a conflict between wishes).

SUMMARY AND DISCUSSION

www.theipi.org 606
www.freepsychotherapybooks.org
 For this initial study of patients’ versus clinicians’
evaluations, only the wish component of the CCRT

was examined. Three procedures for assessing a

patient’s main wishes within the CCRT were

compared: (a) the Self-Report CCRT Questionnaire

composed of 18 standard category wishes, (b) the
Self-Interpretation of the RAP test, and (c) the
clinician’s interpretation of the RAP test. The SR-

CCRT Questionnaire is a newly developed procedure

that has been shown to have adequate test-retest

reliability and to be composed of three main factors.
The SI-RAP test (Luborsky, 1978a) has been further

developed by the inclusion of ratings of the standard

categories for the CCRT.

• The results suggest that, typically, patients are able

consciously to report on the Self-Report CCRT
Questionnaire the same wishes that are identified
by a clinical judge.

• Although this is an important finding that increases

our understanding of the CCRT method, for

www.theipi.org 607
www.freepsychotherapybooks.org
 several reasons it would not be accurate to
conclude that self-report measures are
interchangeable with clinician-scored measures.
First, despite statistically significant differences
between patients’ ratings of the clinician-
identified wish versus other wishes, patients did
not make as large a discrimination between
important or relevant wishes and unimportant or
less relevant wishes. The use of the mean of “all
other wishes” as a comparison for patients’
ratings of the main clinician-identified wish does
not indicate that the patient’s rated the clinician’s
wish the highest. In fact, other wishes were often
rated by the patient equally highly (of course,
many wishes were rated lower by the patient,
thus bringing the mean of “all other wishes”
down).

• It seems likely that patients vary in the extent to

which they are aware of their main wishes, even
if most tend to be aware of them. Two patients in
this study showed particularly low awareness of
the main wish. Without knowing a priori which
patients give accurate self-reports (that is, are
aware of the CCRT), a self-report measure might
be misleading. In future research, patients with

www.theipi.org 608
www.freepsychotherapybooks.org
 low self-awareness could be evaluated by other
tests (such as repressive style measures) before
their assessment with self-report or clinician
methods.

• Although patients often are aware of the wish

component of the CCRT, other aspects of the
CCRT may be less available to conscious self-
report. In fact, our results indicated that patients
do not usually describe the main conflict in the
same way the clinician does. It might be
important to provide patients with specific
questions about their conflicts, including an item
defining a conflict in the way the clinician
formulated it for each patient, rather than relying
only on open-ended questions as we did here. In
addition, although it is expected that relatively
high awareness of the other components of the
CCRT (responses from other and responses of
self) will be obtained, the sequence of particular
wish &rarr; response from other &rarr; response
of self that defines the thematic nature of the
CCRT may not be evident to many patients. Our
data suggest that the method of inquiry affects
the types of self-report information given. The
Self-Interpretation measure which consists of

www.theipi.org 609
www.freepsychotherapybooks.org
 questions anchored to specific accounts of
interpersonal interactions, yielded better
discrimination than did the general Self-Report
CCRT Questionnaire. The use of the Self-
Interpretation measure requires that a RAP
interview be performed (usually lasting about
one-half hour), followed by a self-interpretation
phase (lasting 1 hour). The hoped-for time
advantage of a 5-minute questionnaire does not
seem to be achievable at this stage. Further
research experimenting with different
instructions in a self-report questionnaire may
allow better congruence with clinician-based
measures. In fact, further research has yielded an
expanded and improved central relationship
questionnaire (CRQ) (Barber, 1993).

CONCLUSION

Our overall conclusion is that although patients

are conscious of many of the individual components

of the CCRT, a better discrimination of more versus
less important wishes is achieved by the clinician. In

addition, clinician formulations of conflict do not

www.theipi.org 610
www.freepsychotherapybooks.org
 agree with patients’ open-ended descriptions of their

conflicts.

www.theipi.org 611
www.freepsychotherapybooks.org
 16

STABILITY OF THE CCRT FROM AGE 3

TO 5

LESTER LUBORSKY, ELLEN LUBORSKY,

LOUIS DIGUER, KELLY SCHMIDT,
DOROTHEE DENGLER, JEFFREY FAUDE,
MARGARET MORRIS, PAMELA
SCHAFFLER, HELEN BUCHSBAUM, AND
ROBERT EMDE10

Starting in the late 1970s, the advent of a reliable

operational measure of central relationship patterns in
psychotherapy gave backing to the view of a stable,

distinctive, central relationship pattern for each

person (L. Luborsky, 1976, 1977b; L. Luborsky &
Crits-Christoph, 1990; L. Luborsky, Mellon, van
Ravenswaay, et al., 1985; L. Luborsky et al., 1986).

This new measure of the Core Conflictual

Relationship Theme showed some stability over a

www.theipi.org 612
www.freepsychotherapybooks.org
 period of approximately 1 year of psychotherapy
(Crits-Christoph & Luborsky, 1990). But early-in-life

consistency on this new measure was not examined

until the present study.

If the CCRT method could be applied to very

young children’s narratives, a study might reveal
whether a central relationship pattern appeared in
their narratives about close relationships with their
parents. We planned to score the CCRT in children’s
narratives at age 3 and repeat the same scoring of the

CCRT with the same children at age 5 to see the

consistency of the relationship patterns over time.

Beyond the study at ages 3 and 5, we also planned to
compare these relationship patterns with those of
adult groups whose narratives were scored by the
same standard CCRT categories.

BACKGROUND OF RELATED RESEARCH

www.theipi.org 613
www.freepsychotherapybooks.org
 Three distinct scores have been tapped so far for
evaluating children’s relationship patterns in their

very early years: (a) clinical retrospection on the

basis of adults’ narratives about their early childhood,

(b) infancy research with direct observations of early

relationships, and (c) narratives told by young
children.

Clinical Retrospection

The retrospection method has had a long history

in clinical practice. The method relies on early

memories for reconstructing scenes, usually traumatic

ones, that may have prefigured the current

relationship pattern. There is impressive clinical

evidence that some early traumatic scenes act like a

template so that later episodes contain replications of

components of the earlier scene (as suggested by

Reiser, 1984). Tomkins’s (1987) script theory

similarly places emphasis on the replication of
scenes, as illustrated in the case presentation by

www.theipi.org 614
www.freepsychotherapybooks.org
 Carlson (1986). Such a concept of a long-lasting
consistency of relationship patterns is also suggested

by the famous 30-year longitudinal study of Monica,

a young child with a gastric fistula (Engel &

Reichsman, 1956).

Infancy Research

Research on relationship patterns in infancy has

mushroomed in the last 2 decades. The two methods

that have been increasingly used are the study of the
mother-infant exchange and the study of attachment

patterns. The microanalysis of the mother-infant

exchange has revealed much about the structure of

the interactions (Tronick, 1982); the enduring

meaning of this exchange for personality

development has been examined by Stern (1985,

1989) and by Beebe and Lachmann (1988).

A systematic comparison of developmental

changes in coping styles (E. Luborsky, 1987), based

www.theipi.org 615
www.freepsychotherapybooks.org
 on videos of 30 mother-child pairs of children at age
1 and 2, gave specific evidence for the expected

greater differentiation of coping styles at age 2.

Dahl and Teller (1993) cited a dissertation

(Davies, 1989) describing twelve 3-year-olds

involved in 10 interactions with each of their mothers
and 10 interactions with each of two other children.
Similar “frames” were found in each child’s
interactions with his or her mother compared with
those of the other children; the frames were also

different for each child.

Research based on attachment theory (Bowlby,

1969, 1973) has provided a means of examining
different patterns of attachment (Ainsworth, Blehar,

Waters, & Wall, 1978) at 1 year and more recently

beyond 1 year. The linking of Bowlby’s (1969, 1973)

concept of internal working models of relationships

with attachment patterns has extended knowledge of

www.theipi.org 616
www.freepsychotherapybooks.org
 the age range of attachment behavior and located the
bonds between internal representations and behavior

(Bretherton, 1995; Bretherton, Ridgeway, & Cassidy,

1990; Main, Kaplan, & Cassidy, 1985).

Important work on delineating relationship

patterns is represented by Sroufe (1983) and by
Sroufe and Fleeson (1986), who pointed out that
continuity and coherence in attachment patterns
remain evident in the early years and beyond.

Narratives Told by Young Children

Very young children who are just becoming

verbal can tell narratives, but they find it difficult to
do so consistently unless they are given considerable
structure and assistance. Providing these children

with a set of interesting stimulus pictures is a

technique for stimulating narratives, as in Bellak’s

(1954) Children’s Apperception Test (CAT) derived

from Murray (1938). Yet even with the CAT it is

www.theipi.org 617
www.freepsychotherapybooks.org
 difficult to elicit organized narratives regularly before

age 3 or 4. Buchsbaum and Emde (1990), partly on

the basis of work by Bretherton, Prentiss, and

Ridgeway (1990), initiate the narratives using a story-
stem and a doll family. Such devices have the effect

of extending to earlier ages the ability of children to

provide coherent narratives consistently.

PROCEDURE

Subjects

In our study, twenty-six 3-year-old children were

evaluated. These children from Denver or its vicinity

had been in an earlier study of normal development

(Buchsbaum & Emde, 1990). They were firstborn and

normal at birth, and 16 were female. The parents

were white, middle-class, and married; 25 of the 26

couples were still married at the time of the study.

Eighteen of the mothers had full-time jobs, and their

children were in day care. Fourteen of the children

www.theipi.org 618
www.freepsychotherapybooks.org
 had a younger sibling and the mothers of two of them

were expecting a second child.

After their third birthday, these children were

interviewed briefly in their home, where they told

four narratives to provide them with preliminary

practice. Seven to 10 days later, they were

interviewed and videotaped in the laboratory by a

different interviewer. For the analyses involving a

comparison of age 3 with age 5, we restricted our

sample to the 18 children who took part at both times.

A Doll Family Story Method For Collecting Narratives

The basic data from each of 25 children from

Buchsbaum and Emde’s (1990) sample of 3-year-olds

were based on the use of a doll family story method.

Each child told 10 videotaped narratives in the

laboratory. The duration of the session was from 25

to 30 minutes. For each narrative, the experimenter

presented a stimulus story-stem peopled with a doll

www.theipi.org 619
www.freepsychotherapybooks.org
 family— a father, mother, and two children of the
same gender as the child—in which an upsetting

event that had just happened was related.11 The

experimenter then inquired about what would happen

next, after the event in the story-stem. For example,

one of these stimulus story-stems is called “The Lost
Car Keys”; the experimenter starts the narrative by
saying that the keys to the family car were lost. The

doll mother accuses the doll father of losing the keys;

the doll father denies this. The child is then asked,

“What happens next?” Like the lost key story, most

of the other stimulus stories are conflictual in content.

An example of data from the lost key story is given in

Figure 1, and the scoring of the wishes in that story is

shown in Figure 2.

Basically, the doll family story procedure is a

guided and prompted method of eliciting narratives.

A scene is set by the experimenter’s stimulus story,

and at each stage of the child’s narrative the child is

www.theipi.org 620
www.freepsychotherapybooks.org
 Figure 1.
"Lost Car Keys" Story

Top
Tailor-made Standard Top Clusters
Categories
E: Mom and PRS: #7: am 1: helpful
Dad look for facilitates open
the car keys? dialogue #9: am
C: Un-huh. between helpful
//And... and... parents
and then Jane
comes into the
room.////And...
and Jane says,
"How about
you talk about
it." (Brings
dolls
together.)//
E: And then W: to help #12: to help 8: to achieve
what happens parents solve others and help
after Jane says problem #17: to others
that? avoid 4: to be
C: Well, //then conflict distant and
Susan comes avoid conflict
in (reaches for
Susan, brings
to other dolls)//
//and they all
sit down and sit
there and talk
about it.//
E: So, they all PRS: sisters #7: am 1: helpful
sit down and discuss open 5: self-
talk about it. sharing #15: am controlled
PRO: parents independent 8:

www.theipi.org 621
www.freepsychotherapybooks.org
What are they talk about lost #11: are understanding
talking about? key open 6: helpful
C: Well, #18: are
//they're two cooperative
are having a
little talk
(moves Jane
and Susan
together)// and
//they’re two
having a little
talk (faces
Mom and
Dad).//
E: And they’re W: to open #9: to be 5: to be close
talking about... communication open and accepting
what are they #11: to be
talking about? close to
C: Well, others
//they’re talking
about the car
keys (touches
Mom and
Dad)//, and
//they’re talking
about playing
(touches Jane
and Susan)//.
E: And they’re
talking about
playing?
C: Like... like...
like sharing
and stuff.

The story told by Constance at age 5, to the “Lost Car

Keys” stem. The ratings of the wishes are given in
Figure 2. The double slashes (//) in the story mark off
each thought unit to be scored. PRS = positive
response of self; PRO = positive response from other;
W = wish.
www.theipi.org 622
www.freepsychotherapybooks.org
 Figure 2.

Ratings of Thought Units

To help
parents To open
Standard Category: Wishes
solve communication
problems
1. TO BE UNDERSTOOD 1 1
2. TO BE ACCEPTED 1 1
3. TO BE RESPECTED 1 1
4: TO ACCEPT OTHERS 3 2
5. TO RESPECT OTHERS 3 3
6. TO HAVE TRUST 3 2
7. TO BE LIKED 1 1
8. TO BE OPENED UP TO 4 4
9. TO BE OPEN 1 1
10. TO BE DISTANT FROM 1 1
OTHERS
11. TO BE OPEN TO OTHERS 4 4
12. TO HELP OTHERS 5a 4b
13. TO BE HELPED 1 1
14. TO NOT BE HURT 1 1
15. TO BE HURT 1 1
16. TO HURT OTHER 1 1
17. TO AVOID CONFLICT 5b 5a
18. TO OPPOSE OTHER 1 1
19. TO HAVE CONTROL OVER 1 1
OTHER
20. TO BE CONTROLLED BY 1 1
OTHER

www.theipi.org 623
www.freepsychotherapybooks.org
21. TO HAVE SELF-CONTROL 1 1
22. TO ACHIEVE 1 1
23. TO BE INDEPENDENT 1 1
24. TO FEEL GOOD ABOUT 1 2
MYSELF
25. TO BETTER MYSELF 1 1
26. TO BE GOOD 1 2
27. TO BE LIKE OTHER 1 1
28. TO BE MY OWN PERSON 1 1
29. TO NOT BE OBLIGATED 1 1
30. TO HAVE STABILITY 5 4
31. TO FEEL COMFORTABLE 4 3
32. TO FEEL HAPPY 4 3
33. TO BE PROTECTED 3 3

a Top choice.

b
Second choice.

A sample of the ratings by judge MM on the standard

categories of wishes for Constance for the Lost Car
Keys story (see Figure 1).

www.theipi.org 624
www.freepsychotherapybooks.org
 prompted to explain what happened next. The

procedure is much like that of the Thematic

Apperception Test (TAT; Murray, 1938), but in the

TAT a pictorial scene is presented, and the narrator

creates a story about it. In contrast, the relationship

episodes told for the Relationship Anecdotes
Paradigm (RAP) interview (L. Luborsky, 1990b) or in
psychotherapy (L. Luborsky & Crits-Christoph,

1990) are intended by the narrator to be about actual

events. In Buchsbaum and Emde’s (1990) method of

guided narratives, the child’s presentations of what

the dolls do or say appear to contain a variable
mixture of depictions of relationship events that have
occurred along with fantasies about these
relationships.

Data Analyses

Transcripts of the narratives were made from

videos of the interview and were CCRT scored by the
method described in chapter 2, this volume. First, a

www.theipi.org 625
www.freepsychotherapybooks.org
 specially trained text-preparation judge marked the
transcript with the CCRT-scorable thought units and

the type of component to be scored for each thought

unit. Second, trained CCRT judges (EL, PS, KS, and

MM) inferred the tailor-made categories of each of

the thought units and rated them for all of the Edition
2 standard categories (Appendix A, chapter 2) listed
(on the form provided in Appendix B, chapter 2, with

34 wishes, 30 responses from others, and 31

responses of self).

For the present analysis of results, only a single

judge’s scores for each child were used, which

allowed the transcripts to be apportioned for scoring
among the three judges. The use of only one judge
per case appeared justified by the moderately high
level of agreement in scoring among judges, as is

discussed in the results section.

An Example of a Scored Narrative

www.theipi.org 626
www.freepsychotherapybooks.org
 To help explain the data analysis, we include the

scored story told by 5-year-old Constance after

hearing the lost car keys story-stem, along with the
nondirective promptings by the experimenter (see

Figure 1). The first ruled column on the left has the
tailor-made scores, that is, the judge’s own inferences
about each marked-off thought unit. The middle
column contains the top (most frequent) standard

categories, and the column on the right contains these

top categories expressed in terms of clusters.

As shown in Figure 2, each judge starts with the

thought unit’s tailor-made inferences and rates each

one on all of the standard categories. The rating

reflects the degree to which the judge believes the

standard category is expressed in the thought unit. In

this example, two of the thought units and their tailor-

made inferences are named in the heading of the
columns, and the ratings from 1 to 5 (with 5 being the

www.theipi.org 627
www.freepsychotherapybooks.org
 highest) for each of 33 standard categories given in

each column.

One judge’s ratings of the wishes in the story by

Constance are given in Figure 2. Constance is clearly

a child of our psychotherapy era for she believes in

conflict resolution by means of the people talking

things over. That ethos can be seen as reflected in the

thought unit’s tailor-made inference: “to help parents

solve problems.” This judge gave a score of 5 to the

standard category “to avoid conflict.” Judges

identified their top choice by putting a circle around a

rating and their second choice by putting a square

around it. The standard category ratings in each

column were summed across the narratives, which

allowed us to locate the most pervasive across-

narratives standard categories expressed in the form

of clusters.

Compliance of 3- and 5-Year-Olds With the Narrative-

Telling Task

www.theipi.org 628
www.freepsychotherapybooks.org
 As the first step in the analysis of results, the

narratives were scored for the degree to which the

child complied with the request to tell a narrative.

This information was useful to us in two ways: (a)

We could limit our sample to children who provided

narratives that were complete enough to score, and
(b) we would have an estimate of the degree to which
the children included in our sample were willing or

able to comply with the request to tell narratives. We

rated compliance on a 5-point well-defined scale; the

scale points were defined as follows: Point 1: The

child does not respond to any of the stimulus stories
despite additional prompting. Point 2: The child may
begin to engage in responding to the stimulus story in
a superficial way, but there is no attempt to develop
it. Point 3: The child makes an initial attempt to deal

actively with the stimulus story but breaks this off

entirely or is distracted from creating an ending or

resolution. Point 4: The child directly engages the

www.theipi.org 629
www.freepsychotherapybooks.org
 doll in the content of the child’s story; the child needs

prompting but tends to be responsive to this

prompting. Point 5: The child becomes actively

involved with the stimulus story, needs little

prompting, and works effortfully and agreeably

within the stimulus story format to find a satisfactory
resolution or outcome.

At 3 years of age, 21 of the 25 children were able

to comply sufficiently to be included in the sample.

At age 5, all of the children complied sufficiently to

be included. For the 18 children who were included at
both age 3 and age 5, the average compliance rating

for age 3 was 3.49 and for age 5 was 3.96, showing
that, as a whole, the compliance at age 3 was fairly

good and that it improved only moderately by age 5.

We also found that at age 3 the ratings of

compliance differed across the 10 stories. We used a

repeated-measures analysis of variance for the mean

www.theipi.org 630
www.freepsychotherapybooks.org
 ratings of the 10 stories and found a significant
difference across the 10 stories, F(9, 180) = 2.14, p <

.05, two-tailed.

Two of the stories at age 3 were of special interest

because the children were markedly less compliant in

completing them; they were the exclusion-departure
story and the reunion story. The average rating of
compliance for these two stories was compared with
the average of the other eight stories using a paired t-
test procedure. The means were significantly

different, t (20) = 2.73, p < .05, two-tailed, and the

difference was in the predicted direction, that is,

lower compliance was evident for these two stories.
These story themes of departure and reunion
prompted more disruption in compliance with
storytelling than did the other themes, although some

of them were also conflictual. This inadvertent

finding dovetails with the use of Ainsworth’s Strange

Situation in attachment research (Ainsworth et al.,

www.theipi.org 631
www.freepsychotherapybooks.org
 1978). The departure-reunion sequence in both cases

seems to activate stress reactions.

Although children of both age groups could tell a

story and adequately follow directions, an interesting

shift was noted concerning how they handled a

conflict element within a story. When a story-stem
was not how a 3-year-old child wanted it, he or she
sometimes changed the original story line and looked
for a solution there. The 5-year-olds did not change
the story-stem. Instead they were likely to have the

main character do something else about the problem,

including having the main character pretend or hide

their character’s agenda.

RESULTS

Do the Judges Agree in Scoring the CCRT?

As noted earlier, the level of agreement among

judges was generally satisfactory and supported our
decision to rely on the scoring of only a single judge

www.theipi.org 632
www.freepsychotherapybooks.org
 in later analyses. Three types of agreement were
examined for the CCRT scoring of the children’s

narratives. First, agreement was based on scoring by

two judges (JF and DD) with a subsample of 12 of

the 25 children at age 3 (Dengler, 1990). The two

judges exactly agreed in their composite CCRT
scores for 75% of the items. Second, from the sample
of the 18 children in the present study for whom we

had data at both age 3 and age 5, we used the 3-year-

olds (n = 10) who were rated by two judges (EL and

PS). These judges agreed on the wishes (W) of 7 of
the 10 children, on the responses from other (RO) of

10 of the 10 children, and on the responses of self

(RS) of 9 of the 10 children. (Agreement is defined as

a match between the two judges in identifying the

same cluster of each child with the highest average

sum of standard categories.) Finally, a weighted

kappa measure of agreement was also computed by

assigning 1.0 to instances in which both judges gave

www.theipi.org 633
www.freepsychotherapybooks.org
 the same first choice to each component; other partial

matches were assigned weights of .66 or .33. These

weighted kappas for each CCRT component were: W

= .33, RO = .69, and RS = .60 (for comparisons with

other studies see chapter 6, this volume). The lower

kappas, especially for the wishes, appear to be largely
attributable to the very low variability. The kappas
are therefore not as representative of the reliability as

the agreement percentages.

How Pervasive Are the Clusters of CCRT Standard

Categories Within Narratives at Age 3 and Age 5?

Our interest was in which clusters of standard

categories were most pervasive, that is, which
reappeared most often across each child’s 10

narratives. Pervasiveness was scored for each of the

three components that made up the central

relationship pattern that is measured by the CCRT:

wishes, responses from others, and responses of self.

www.theipi.org 634
www.freepsychotherapybooks.org
 Our judges rated the narratives for each scorable

thought unit for every standard category using the 1-5

scale, with a high rating indicating that the standard

category was strongly reflected in that thought unit.

In our analysis, we summed the ratings given to each

thought unit for each of the approximately 30
standard categories. The lists of standard categories
were then simplified by cluster analysis in Edition 3

to only eight clusters (see chapter 3, this volume).

Our tables of results reflect the frequency of highest

and next highest cluster scores for each child; a

cluster score is the mean of the sums for the standard
categories within each cluster.

For each of the three components (W, RO, RS),

just one or two cluster scores among the eight clusters

had a high frequency. For example (see Table 1), the

wish “to be loved, understood” was presented in the
narratives among the 18 children as highest or next
highest 10 times at age 3 (28%) and 15 times at age 5

www.theipi.org 635
www.freepsychotherapybooks.org
 TABLE 1
Comparison of Children at Ages 3 and 5 for the Number of
Times Highest or Next Highest for Each Wish Cluster (N = 19)

Wish Cluster
To To be To
To To be To To be To Feel
Oppose, Close, Achieve,
Age Assert Controlled Avoid Loved, Good,
Hurt Accept Help Total*
(years) Self or Hurt Conflict Understood Comfortable
Others Other Others
1 3 4 6 7
2 5 8
27.7% 41.6%
3 4 0 0 5 1 10 15 1 36
41.6% 33.3%
5 5 0 0 2 2 15 12 0 36

*The totals are because each of the 18 children has a highest and next
highest cluster.

www.theipi.org 636
www.freepsychotherapybooks.org
 (42%). Similarly, the wish “to feel good and to be
comfortable” was presented 15 times for the 18

children at age 3 (42%) and 12 times at age 5 (33%).

Among the responses from other component

scores (see Table 2), the most frequent in the

narratives was “helpful” (both at age 3 with 15 [42%]
and at age 5 with 16 [44%]) and “understands” (with
16 [45%] at both age 3 and age 5). The responses of
self (see Table 3) that were most frequent were in the
clusters “self-confident” and “helpful.” (At age 3, the

frequency of “helpful” was 7 [19%] and at age 5 it

was 16 [44%]. The increase at age 5 of “helpful” was

one of the largest increases from age 3 to age 5.)

Is There a Core Relationship Theme for Each Child?

For examining this question of a single versus

several CCRTs for each child, the most telling data

were the rankings of the frequency profiles of the

eight clusters of the CCRT components for each

www.theipi.org 637
www.freepsychotherapybooks.org
 TABLE 2
Comparison of Children at Ages 3 and 5 for the Number of
Times Highest or Next Highest for Each Response From Other
Cluster (N = 18)

Responses From Other Cluster

Rejects, Likes
Age Strong Controlling Upset Bad Helpful Understands
Opposes Me Total*
(years) 1 2 3 4 6 8
5 7
41.6% 44.4%
3 0 2 0 1 1 15 1 16 36
44.6% 44.4%
5 0 2 0 0 0 16 2 16 36

*The totals are 36 because each of the 18 children has a highest and
next highest cluster.

www.theipi.org 638
www.freepsychotherapybooks.org
 Table 3
Comparison of Children at Ages 3 and 5 for the Number of
Times Highest or Next Highest for Each Response of Self
Cluster (N = 18)

Responses of Self Cluster

Oppose,
Self- Anxious,
Age Helpful Unreceptive Respected Hurt Helpless Sad
Confident Ashamed Total*
(years) 1 2 3 Others 6 7
5 8
4
19.4% 38.8%
3 7 1 3 5 14 4 0 2 36
44.4% 27.7%
5 16 1 6 3 10 0 0 0 36

*The totals are 36 because each of the 18 children has a highest and
next highest cluster.

www.theipi.org 639
www.freepsychotherapybooks.org
 child. Our reasoning was that the more each child’s

profile concentrated on just a few clusters, the more

we can conclude that the concept of a core theme for
each child is a cogent one. The data from the profiles,

both at age 3 and at age 5, tended to be consistent

with this concept of a core theme. A telling
illustration of this concentration can be seen in the
rankings of each child’s pervasiveness on each

cluster. Most children had a high pervasiveness

within their top two clusters, with the remaining six

clusters having considerably less pervasiveness. The

drop in mean pervasiveness from the top two to the
remaining six was about one third. This drop was
about the same at age 3 as at age 5 and about the
same for wishes and responses of self, but the drop
from responses from others was very small.

How Constant Does Each Child’s Profile of Clusters

Remain From Age 3 to Age 5?

www.theipi.org 640
www.freepsychotherapybooks.org
 The profile of clusters remains relatively constant

from age 3 to age 5 (even though 4 of the 10 story-

stems were different at the two ages). The Spearman

rank correlations between cluster scores at age 3 and
at age 5 are high: W, .84 (p < .01); RO, .89 (p < .01);

RS, .74 (p < .05). The components of the CCRT that

have the most similarity from age 3 to age 5 are the

wishes and responses from others (the same W,

83.4%; the same RO, 94.5%; the same RS, 72.3%).

Example, John: John’s responses from other and

responses of self remained consistent from age 3
to age 5. At age 3, the story-stem given to him
was about spilled juice; in John’s continuation of
the story he has the mother clean up the juice
(RO), and John then says, “Yum, yum” as he
eats the meal she provides (RS). At age 5, the
story-stem given to him is about falling off his
bike while going to get ice cream; John has the
mother give the child a Band-Aid (RO), and then
the child picks up his bike and rides off again
(RS).

How Constant From Age 3 to Age 5 Were the Highest

Clusters of the CCRT Patterns?

www.theipi.org 641
www.freepsychotherapybooks.org
 We now look at the pervasiveness from age 3 to
age 5 of the combined patterns of the CCRT

components, that is, the wish, response from others,

and response of self together, rather than singly.

Specifically, we examined each child’s combination

of his or her three top frequency clusters for the wish,

response from others, and response of self. In this

measure, the highest number that can be the same is

three: one for the same highest frequency cluster for

the wish, one for the same response from others, and
one for the same response of self.

Example, Robert: In this example, the wish and

the response from others stayed the same from
age 3 to age 5. At age 5, the most pervasive wish
was still “to feel good and comfortable” and the
most pervasive response from the doll parents
was still “to be helpful.” The one component that
changed for Robert was the response of self: At
age 5 his most pervasive response of the doll
child was to be “self-confident” and “assertive.”
Robert’s sameness score was 2.

www.theipi.org 642
www.freepsychotherapybooks.org
 We found across the groups of children that there

was considerable consistency in the highest clusters

from age 3 to age 5: The mean sameness of the
highest clusters for the 18 children at age 3 and again

at age 5 was 2.5. Sixty-one percent of the children

had a score of 3. Twenty-eight percent had a score of
2, and 11% had a score of 1.

The consistency over time is greatest for the wish

(10 children with the same wish), next for the
response of others (8 children), and least for the

response of self (3 children). This preponderance of

consistency over time for the wish is similar to the

finding for adult groups (see chapter 10, this volume),

for whom wishes were found to have more consistent

pervasiveness than the responses over a mean time of

1 year.

How Positive and How Negative Were the

Relationship Patterns at Ages 3 and 5?

www.theipi.org 643
www.freepsychotherapybooks.org
 Central relationship patterns can be classified as

positive or negative, after Freud’s usual practice of

labeling the transference as positive or negative (e.g.,
Freud, [1912/1958a]; see also chapter 4, this volume).
Positive or negative classifications imply satisfaction

of wishes (positive responses from others or of self)

or lack of satisfaction of wishes (negative responses

from others or of self). With this concept in mind, we

classified as positive or negative the CCRT patterns

of each child at age 3 and age 5. Positive and

negative responses from other and of self were

counted for all scored thought units for each child

(Table 4).

Our sample of relatively high-functioning Denver

children showed very low percentages of negative

responses and very high percentages of positive

responses, both at age 3 and at age 5 (see Table 4).
For example, the responses from others at age 3 were
31% negative, and at age 5, 29% negative. The

www.theipi.org 644
www.freepsychotherapybooks.org
 TABLE 4
Positive and Negative CCRT Responses (Percentages)

Component Positive Negative Neutral

Denver Nonclinical Children
(N =18)
RO Age 3 69 31
RO Age 5 71 29
RS Age 3 63 37
RS Age 5 77 23

Penn Depression Patients a

(N = 30)
RO, RS (combined) 21 72 7

Penn OPD Patients a

(N = 20)
RO, RS 19 73

Gottingen Nonclinical b
(N = 30)
RS 35 43 21

Ulm Nonclinical c
(N = 35)
RO 38 57 5
RS 47 48 9

Note. RO = responses from others; RS = responses of self; OPD =

Outpatient Department
a
Luborsky & Crits-Christoph (1990, p. 225) b Cierpka et al.
(1992) c Dahlbender (1992)

www.theipi.org 645
www.freepsychotherapybooks.org
 responses of self at age 3 were 37% negative, and at
age 5 they were 23% negative.

When we describe the level of these negative

responses of the children as very low, we base our

evaluation on our main experience with adult patients

(L. Luborsky & Crits-Christoph, 1990; see chapter 4,

this volume). In contrast, the adults’ level of negative
responses is very high: The Penn Depression Study
sample and the Penn Outpatient Department patient
sample (Luborsky et al., 1988) had levels of 72% and

73%, respectively (see Table 4). These higher

percentages of negative responses may be attributable

to the fact that we have studied groups of adults and

adults tend to have more negative responses.

An additional obvious explanation is that the

adults were patients, and patients tend to have a high

percentage of negative responses. For evidence we

cite two unpublished papers reporting studies of

www.theipi.org 646
www.freepsychotherapybooks.org
 normal German adults. One of these papers reports
work with normal college students (15 men and 15

women) at the University of Göttingen in Germany

(Cierpka et al., 1992). In this sample (Table 4) 43%

of the responses of self were negative for women, and

42% were negative for men. The other sample
consisted of 35 normal women from Ulm, Germany
(Dahlbender, 1992); the negative responses from

others amounted to 57%, and the negative responses

of self, 48%. A likely implication of these studies is

that normal adults are less negative than patients but
much more negative than children.

Are There Gender Differences in Central Relationship

Patterns at Ages 3 and 5?

The commonalities across the frequencies of the

eight clusters for boys versus girls were much more
impressive than the differences. This was true for all

three CCRT components. Because there were only 11

girls and 7 boys, however, high percentages are

www.theipi.org 647
www.freepsychotherapybooks.org
 needed to reveal the few differences. For example,

under the wish “to be loved and understood,” at age 3

there were two top or next-to-top frequency clusters

for the boys (that is, the total number of top or next-

to-top clusters is 14 for the seven boys, and 2 out of

14 would be 14%). By comparison, at age 3 there
were eight top or next-to-top clusters for the girls
(this would be 8 out of 22 = 36%; 22 is the number of

top or next-to-top clusters for 11 girls).

The wish “to be loved and understood” was more

frequent for girls, but it is hard to test whether 36%
for girls is significantly greater than the 14% for

boys. At age 5, the comparable figures are 36% for

boys and 45% for girls, but the figure for girls is

probably not significantly larger.

The differences between boys and girls in

response from others clusters were small; the

www.theipi.org 648
www.freepsychotherapybooks.org
 differences in response of self clusters were also
small.

SUMMARY, DISCUSSION, AND CONCLUSIONS

• Compliance with the task of telling narratives.

Both the 3-year-olds and the 5-year-olds were
able to comply and tell moderately complete
narratives when provided with story-stems.

• Pervasiveness of the CCRT. Originally we asked, is

there a pervasive central relationship pattern in
the narratives about close relationships at age 3,
and if there is, does it continue at age 5? The
answer to both questions is yes, on the basis of
both (a) a high level of pervasiveness of CCRT
components across each child’s set of narratives
at age 3, and (b) the number of these pervasive
CCRT components that reappeared at age 5 (with
a high percentage of similar components at the
two ages).

The two most pervasive clusters were the

wish “to be loved and understood” and the wish
“to feel good and comfortable.” The two most
pervasive responses from others were “helpful”
and “understands”; the two most pervasive

www.theipi.org 649
www.freepsychotherapybooks.org
 responses of self were “self-confident” and
“helpful.” The combination of the wish, response
of other, and response of self also showed high
stability from age 3 to age 5. The most
impressive changes from age 3 to age 5 were that
(a) the wish “to be loved, understood” increased
slightly, and (b) the response of self of being
“helpful” increased markedly, possibly reflecting
greater maturity and responsibility.

These findings are new, and as new findings

tend to do, they lead to questions even harder to
answer: Is the CCRT pattern at age 3 likely to
continue to reappear at even later ages through
adolescence and adulthood? To answer this
question, we need to reevaluate children at later
times. Waldinger et al. (1997) are assessing the
CCRT through interviews with adolescents at age
14 and then again at age 23 on the basis of
longitudinal data collected by Stuart Hauser, Gil
Noam, Sally Powers, Alan Jacobson, and Joseph
Allen.

• Positive versus negative quality of the CCRT

pattern. Positivity and negativity are meaningful
modifiers of the central relationship pattern. We
found that both at age 3 and at age 5 the CCRT’s

www.theipi.org 650
www.freepsychotherapybooks.org
 were overwhelmingly positive. Further research
is needed to learn whether this is a representative
finding to be taken at face value as a
characteristic of the developmental stage of this
age group. Older people and people who are
patients tend to have more negative responses. In
interpreting the results concerning positive and
negative responses in narratives, several factors
must be taken into account:
1. Most of the CCRT research so far has been
with adults who are patients. Both of these
conditions are associated with an increase
in the number of negative responses.
2. The type of narratives the children told may
have contributed to the positivity of their
responses. These narratives are partly
fictional rather than accounts of actual
events, and fictional accounts may be
particularly prone to idealization. In
contrast, the narratives collected from
adults are intended to be descriptions of
actual events (L. Luborsky, 1990a).
3. This sample of Denver children may have
been especially healthy. We need, therefore,
to study other groups of children. From
such studies, we may still emerge with the
conclusion that the preponderance of

www.theipi.org 651
www.freepsychotherapybooks.org
 positive responses is a typical early
developmental characteristic. We are told
by Seligman (1975), by Seligman, Kamen,
and Nolen-Hoeksema (1988), and by
Seligman (1991), for example, that children
are more optimistic than adults in their style
of explaining the causes of negative events.
4. We need to check our assumption about the
relationship patterns we found—that the
normal adult groups are not unduly affected
by differences in cultural backgrounds, such
as between American and German groups.

• Gender differences in CCRT patterns. Gender

similarities clearly are more prominent than
gender differences in the CCRT components. The
only difference that may be gender related is an
increase at age 3 for the girls in the wish “to be
loved and understood,” which may go along with
the slight decrease at ages 3 and 5 in the wish “to
feel good and comfortable.” These gender
differences may involve differences in
relatedness to others, a characteristic that is
thought by some (e.g., Gilligan, 1982), to be
found more commonly in women than men. A
larger and older body of research based on the
work of Witkin (1949) has shown similar

www.theipi.org 652
www.freepsychotherapybooks.org
 differences, for example, that females are more
responsive to the “field” than males and that,
among ways of showing this, females are more
attentive to faces of other people.

WHAT IS NEXT IN THIS LINE OF RESEARCH?

We found that children at the early ages of 3 and

5 already have a pervasive pattern in close

relationships, and that pattern is mostly positive. To

get more perspective on this finding, a study should

be done with data from children that also includes

narratives about actual events, which would allow for

better comparisons with our adult data.

The vulnerabilities of each child in close

relationships are likely to be associated with each

child’s most pervasive central relationship pattern;

when special stresses give rise to symptoms, they

should appear as part of that pattern. We have begun
to examine this premise for the few children in this
sample who experienced traumatic conditions, and

www.theipi.org 653
www.freepsychotherapybooks.org
 Waldinger (1997b) is examining a group of children

who experienced rejecting conditions in their early

environment.

The CCRT patterns we have found at ages 3 and 5

may be similar to patterns described as “attachment

patterns,” and “internal working models” (Bretherton,
1995; Bretherton, Ridgeway, & Cassidy, 1990; Main,
Kaplan, & Cassidy, 1985), or as “transference
patterns.” Its relation to transference gets support
from the work of Fried, Crits-Christoph, and

Luborsky (1990c; also see chapter 11, this volume),

who showed that the relationship episodes about the

therapist provide a CCRT that is much like the CCRT

derived from the relationship episodes about other

people. This parallel is a crucial one, perhaps even

more central than many of the rest of the 23 facets of

Freud’s definition of transference (see chapter 21).
Our next broad agenda is to examine more
systematically the relations among these differently

www.theipi.org 654
www.freepsychotherapybooks.org
 labeled concepts. We will then be able to examine the
overlap among different researchers’ apparently
different but probably similar conceptual models.

Our results on age-related themes may be linkable

to Piagetian and other “transformational

psychologies.” The themes we uncovered may be
related to stages of development (Loevinger, 1976).
Only longitudinal analysis will be able to differentiate
core themes that remain across the life span from
those that change over time (Noam, 1991).

Notes

[10] This chapter is a modified version of “Extending the Core

Conflictual Relationships Into Childhood,” by Luborsky,
Luborsky, et al., 1995, in Development and Vulnerability in
Close Relationships (pp. 287–308), Mahwah, NJ:
Lawrence Erlbaum Associates. Copyright 1996 by
Lawrence Erlbaum Associates, Inc. Reprinted by
permission.

[11] At 36 months, the 10 story-stem events were about spilled

juice, toilet, monster, car keys, argument, ice cream, naps,
restraint of aggression, departure/reunion, couch, and
moral dilemma. At 60 months, the story-stems were about
ice cream, monster, sad, car keys, argument, nap, bicycle,
clean room/new toy, departure/reunion, couch, and Band-
Aid (6 of the 10 story-stems were the same). The fuller

www.theipi.org 655
www.freepsychotherapybooks.org
 account of the story-stem can be found in a chapter by
Buchsbaum and Emde (1990).

www.theipi.org 656
www.freepsychotherapybooks.org
 17

STABILITY OF THE CCRT FROM

BEFORE PSYCHOTHERAPY STARTS TO
THE EARLY SESSIONS

JACQUES P. BARBER, LESTER LUBORSKY,

PAUL CRITS-CHRISTOPH, AND LOUIS
DIGUER12

We come now to an issue about the sources of the

CCRT: Is it primarily a quality that the patient brings
to therapy or a product of the therapist’s responses to
the patient? Some understanding of this issue might
be achieved by examining the stability of the CCRT
between a time before the therapist is met to a time

after the treatment has started.

Psychoanalytic data is mostly inferred from what

patients say and from their behavior. Because these
data require inference to be understood, the

www.theipi.org 657
www.freepsychotherapybooks.org
 therapist’s interpretation of these words and other
behaviors is likely to play a major role in determining

their significance (Eagle, 1983). The therapist’s

theoretical stance is likely to influence how he or she

interprets the patient’s words and other behaviors.

These interpretations, in turn, influence the patients,
who are prone to accept their therapists’
interpretations for a variety of reasons. Thus, argued

Grunbaum (1984), patients’ data obtained from

therapy sessions may be contaminated by the

therapists’ theoretical point of view and

indoctrination and, therefore, cannot be used to

validate the underlying theory of treatment. In other

words, Grunbaum (1984) claimed that clinical data

has little, if any, scientific value because it tends “in

any case to be artifacts of the analysts’ self-fulfilling

expectations, thus losing much of their evidential
value” (Grunbaum, 1986, p. 217). Although

Grunbaum referred to psychoanalysis proper, his

www.theipi.org 658
www.freepsychotherapybooks.org
 criticism applies also to contemporary psychoanalytic
theories (Eagle, 1983), to dynamic psychotherapy,

and to other therapies. This critique has the power to

undercut the use of treatment sessions as a way of

validating the scientific aspect of the theory, because

if the patient’s responses are merely a result of

brainwashing, then Freudian analysis might have
beneficial emotional effects not because it allows
the patient to acquire genuine self-knowledge,
but because of suggestion operating as a placebo
under the guise of non-directive therapy.
(Grunbaum, 1986, p. 221)

The comparison of CCRTs obtained through

narratives from therapy sessions with those obtained

from clinical interviews conducted before the

therapist is even met might begin to address this

criticism (Luborsky, 1986a). To the extent that the

CCRTs obtained from interviews preceding

psychotherapy are similar to the ones extracted from

therapy sessions, we can be confident that the
clinician’s influence on the patient’s central

www.theipi.org 659
www.freepsychotherapybooks.org
 relationship pattern, at least in the early sessions of
therapy, is not as pervasive as Grunbaum has

suggested and that psychodynamic psychotherapy

(and other therapies) may be something more than

suggestion.

METHOD

We compared a measure of the central

relationship patterns before the treatment started with

a measure obtained after it started. A review of such

measures developed in the last 15 years has been

reported (see Barber & Crits-Christoph, 1993; also
see chapter 20, this volume). The Core Conflictual
Relationship Theme method, the oldest of these
measures, is the one we selected. Its interjudge

reliability has been shown to be fair to good

(weighted kappas of .60 to .71) across eight samples

(see chapter 6). Various findings have supported its

validity by showing that the CCRT method assesses a

construct that is consistent with many characteristics

www.theipi.org 660
www.freepsychotherapybooks.org
 of the transference pattern described by Freud
(1958a) and discussed in chapter 21, this volume.

Nineteen patients (15 women, 4 men; mean age,

40, SD = 9.6) participated in a study involving 16

sessions of time-limited supportive-expressive

dynamic psychotherapy for depression (Luborsky,
1984; Luborsky, Mark, et al., 1995). Eight patients
were never married, and 4 were divorced, separated,
or widowed. Patients either were referred from other
clinics within the hospital of a major Northeastern

medical center or had responded to advertising in the

community. Only patients with a Research Diagnostic

Criteria diagnosis of major depression without
psychotic features, brain impairment, or current drug
or alcohol abuse were entered into the study. Patients
needed to have been diagnosed using the Schedule

for Affective Disorders and Schizophrenia (Endicott

& Spitzer, 1978) on two consecutive interviews

spaced 1 week apart before entering treatment. At the

www.theipi.org 661
www.freepsychotherapybooks.org
 second intake interview, the average level of
depressive symptoms as measured by the Beck

Depression Inventory was 28 (SD = 7.5); patients’

average score on the Health-Sickness Rating Scale

was 49.0 (SD = 6.1). Eleven patients had at least one

probable or definite coexisting personality disorder
diagnosis. A more complete description of the larger
sample may be found in articles by Diguer, Barber,

and Luborsky (1993) and by Luborsky, Diguer, et al.

(1996).

The Core Conflictual Relationship Theme Method

The Core Conflictual Relationship Theme method

(see chapter 2, this volume) describes the relationship

pattern that is most pervasive across narratives using

the following steps: (a) Relationship episodes are

delineated in the transcribed material; (b)

independent judges read each relationship episode in

the transcript and identify each of three components
(wishes, responses from others, and responses of

www.theipi.org 662
www.freepsychotherapybooks.org
 self); (c) for each component, the types with the

highest frequency across all relationship episodes are

identified and combined constituting a preliminary

CCRT formulation; (d) on the basis of this
preliminary CCRT formulation, the same judge re-

identifies, when needed, the types of wishes,

responses from others, and responses of self; (e) the

judge can change the original rating on the basis of

the recount of all wishes, responses from others, and

responses of self. In addition, judges were asked to

translate their tailor-made scoring into standard
categories. It has been reported, in a sample of 35
psychotherapy patients, that interjudge agreement as

measured by weighted kappas was .70 for responses

from others and .61 for wishes and responses of self

(Crits-Christoph, Luborsky, et al., 1988).

The RAP Interview Method

The difficulty and expense of extracting

narratives from sessions, as well as their possible

www.theipi.org 663
www.freepsychotherapybooks.org
 contamination by therapists’ suggestions, have led

researchers to use alternative data obtained from

clinical interviews. Luborsky (1990b; see chapter 7,
this volume) developed the Relationship Anecdotes
Paradigm (RAP) interview to collect such

interpersonal narratives from which CCRTs could be

extracted. It has been assumed that a CCRT

formulation based on narratives told during RAP

interviews conducted by an independent researcher or

clinician is similar to one obtained from therapy

sessions. The present study has investigated this
assumption by examining such a comparison.

Instructions for administration of the RAP

interview (see chapter 7) require the participant to tell
at least 10 incidents or events, each about an

interaction between the participant and another

person. Those interviews are recorded and then

transcribed. The CCRTs are extracted from the

interviews in the same manner as they are extracted

www.theipi.org 664
www.freepsychotherapybooks.org
 from therapy sessions. The mean time usually
required to tell 10 episodes is about 30 minutes.

Procedure

The patients in the present study were seen in

supportive-expressive dynamic psychotherapy for 16

sessions by four different experienced therapists

(Diguer et al., 1993). The therapists participated in

the training phase of a treatment development project.

The RAP interviews were given by a research
assistant before therapy began. Sessions 3 and 5 were

transcribed, but for the two patients from whom we

found fewer than 10 complete relationship episodes,

Session 4 was added. The transcribed RAP interviews

and therapy sessions were then rated by two different

teams of two judges. Each judge worked

independently, used the standard categories (Barber et

al., 1990; chapter 3, this volume), and followed the

CCRT scoring manual (see chapter 2). All judges
were experienced psychodynamic clinicians who had

www.theipi.org 665
www.freepsychotherapybooks.org
 been trained in the CCRT method by Lester

Luborsky.

Because there are many standard categories (35

wishes, 30 responses from others, and 31 responses

of self) with some having similar meanings (e.g.,

wish to be understood vs. to be respected vs. to be
accepted), assessing the judges’ agreement on the
most frequent standard categories would have been
too stringent a criterion for calculating reliability; that

is, we did not want to say that if one judge decided

that the main wish was “to be understood” and the
other judge thought it was “to be accepted,” the

interjudge agreement was 0. In addition, there were

many cases in which different standard categories

were high in frequency; that is, more than two or

three standard categories were the most frequently

used by one judge for a specific patient. To resolve
these two problems, we used Barber et al.’s (1990)
grouping of the standard categories into eight

www.theipi.org 666
www.freepsychotherapybooks.org
 clustered standard categories for each CCRT
component. All standard category ratings were

recorded by the research assistant in their appropriate

clusters. For all analyses involving the clustered

standard categories, the two most frequent ratings for

each CCRT component from each judge were chosen.

RESULTS

Reliability of the CCRTs Derived From the RAP

Interviews

All 19 RAP interviews were rated by two

independent judges, the degree of interjudge

agreement on the clustered standard categories is
presented in the top tier of Table 1. To correct for

chance agreement, we followed Crits-Christoph,

Luborsky, et al.’s (1988) use of the weighted kappa
(Cohen, 1968) for assessing interjudge reliability of
the rating for each of the three CCRT components. In
contrast to regular kappa, weighted kappa allows

different weights for different levels of agreement;

www.theipi.org 667
www.freepsychotherapybooks.org
 TABLE 1
Interjudge Agreement and Reliability for the CCRT From the
RAP Interview and Therapy Sessions, and Comparisons of
CCRTs From RAP Interview Versus Therapy Sessions

Agreement Weighted Kappa

Variable Between Clustered Standard
Judges Categories Categories
CCRT from RAP interviews
Ws 84 .68 —
ROs 100 .60 .56
RSs 89 .65 —
CCRT from Sessions 3 and 5
Ws 94 .81 —
ROs 100 .64 .77
RSs 88 .73 —
Comparing CCRT from session to CCRT from RAP
Ws 77 .52 —
ROs 100 1.00 —
RSs 77 .40 —

Note. CCRT = Core Conflictual Relationship Theme; RAP =

Relationship Anecdotes Paradigm; ROs = responses from others;
RSs = responses of self.

www.theipi.org 668
www.freepsychotherapybooks.org
 that is, a higher weight can be given if agreement

between the two judges occurred on the most

frequent clustered standard categories, a lower weight
if the second highest rating from one judge matched

the most frequent rating of the other judge, and the

lowest weight if judges agreed only on the second
most frequent ratings. More specifically, the two most
frequent clustered standard categories of wishes (or

responses from others or responses of self) for each

patient from one judge were compared with the two

most frequent wishes of the other judge. If the most

frequent wish rated by each judge matched, a weight
of 1.0 was given; if the most frequent clustered
standard wish category of one judge matched the next
most frequent of the other judge, a weight of .66 was
given; and if only the two second most frequent

categories matched, a weight of .33 was given. Crits-

Christoph, Luborsky, et al. (1988) used identical

weights. This computation was performed separately

www.theipi.org 669
www.freepsychotherapybooks.org
 for wishes, responses from others, and responses of

self. The results are in the second column of Table 1.

All of these kappas were in the acceptable range.

The high degree of agreement but only fair size of

kappas is likely due to the narrow range of categories

of CCRT components, especially responses from

others, that these patients displayed. Seventeen

patients (89%) were rated as having the response

from others of “rejecting and opposing,” one had the

response from others “understanding or accepting,”

and one patient’s response from others was “upset.”

Because the judges used only three of the eight

clustered standard categories for the responses from

others, we recalculated the degree of agreement and

the weighted kappa for the responses from others

using the 30 standard categories instead of the eight

clusters. Using the standard categories, we observed

that all response from others standard categories were
used at least once by one of the two judges. The

www.theipi.org 670
www.freepsychotherapybooks.org
 weighted kappa obtained using the 30 standard

categories was .56 (shown in the third column of

Table 1).

Reliability of the CCRTs Derived From Therapy

Sessions

Two other independent judges rated the CCRTs

from the sessions for the 17 of the 19 patients who
entered treatment and for whom audiotapes were

available (see the middle tier of Table 1). The

adequate reliability coefficients found in the present

study for the CCRT components derived from therapy

sessions replicate Crits-Christoph, Luborsky, et al.’s
(1988) findings in another moderate-sized sample.
Again, the same problem outlined in the previous

section occurred with the responses from others from

the sessions (16 of 17 patients had the clustered

standard categories response from others of

“rejecting,” whereas the other patient’s response from

others was “like me”). We therefore recalculated the

www.theipi.org 671
www.freepsychotherapybooks.org
 weighted kappa for the responses from others using

the standard categories; as presented in Table 1, the

kappa was adequate. As in the ratings from the RAPs,

these two different judges used all 30 response from

others standard categories at least once.

Correspondence Between CCRTs From RAPs (Before

Therapy) and CCRTs From Sessions

To compare the two sets of ratings, we needed

first to combine the ratings from each independent

team of judges. In the cases in which there was

agreement between the two judges who scored the

RAPs, the categories that were agreed on were used
in the comparison with the CCRT from sessions, and
vice versa. In the cases in which there was no

agreement between the two judges, the clustered

category that was the most frequently rated across

relationship episodes by any of the two judges for a

specific patient was selected for comparison with the

clustered category from the other team of judges. The

www.theipi.org 672
www.freepsychotherapybooks.org
 same process was used for the second most frequent
category. The results from this comparison were

summarized in part in a previous review of the CCRT

(Luborsky et al., 1992).

The comparison of the CCRT ratings from the

RAP interviews and the therapy sessions for the 17
patients indicated a relatively high level of agreement
between the two methods of deriving the CCRT,
suggesting a relatively high level of similarity
between the CCRT obtained from pre-treatment data

and the CCRT obtained from sessions early in

treatment (see the bottom tier of Table 1). Thus, even

when we corrected for chance agreement, we found a
moderate-to-high level of correspondence between
the CCRTs derived from the two different sources of
material. In other words, moderate alternate-form

reliability was found for the wishes and responses

from self across the two methods of deriving

narratives for CCRT formulations. In regard to the

www.theipi.org 673
www.freepsychotherapybooks.org
 responses from others, the two methods yield

excellent alternate-form reliability when one uses the

clustered standard categories. From a psychometric
point of view, one needs to realize that the “alternate

forms” and the “responses” (patients’ narratives) are

very different in the two methods, at least on the
surface. As an anonymous reviewer noted, this lack
of perfect match between the two “forms” may have

reduced the reliability estimates.

CONCLUSIONS

• These results support the conclusion that the

relationship themes that emerge early in
treatment are quite similar to the themes that
emerge during an independent interview, with a
person other than the therapist, that precedes the
therapy. These findings are likely to increase
researchers’ confidence that the RAP interview
can be used to determine patients’
psychodynamic themes independently of
treatment.

www.theipi.org 674
www.freepsychotherapybooks.org
 •In response to Grunbaum’s (1984) critique, we have
presented preliminary empirical evidence that the
CCRT in early sessions of psychodynamic
therapy is not likely to be primarily the result of
therapists’ influence. At the same time, our data
do not indicate that the CCRTs obtained before
treatment and early in treatment are identical.
The present findings are especially meaningful to
the extent that the CCRT indeed measures the
complex and controversial but central
psychoanalytic concept of transference. Indeed,
Fried et al. (1992) showed that the CCRT
expressed in the relationship with the therapist is
similar to the CCRT expressed in other
relationships. Additional studies are needed to
replicate our preliminary findings using material
from before and during psychoanalytic sessions.

• One major limitation regarding the generalizability

of the results of this study is that it is based on a
sample of patients who had received a diagnosis
of major depressive episode in accordance with
the Diagnostic and Statistical Manual of Mental
Disorders (3rd ed.; DSM-III, American
Psychiatric Association, 1980). It may be, for
example, that the restricted range of responses

www.theipi.org 675
www.freepsychotherapybooks.org
 from others obtained in the present sample is
characteristic of depressed patients but not of
other groups of patients; that is, depressed
patients tend to see others as rejecting. Thus,
replication in larger samples as well as in
heterogeneous groups of patients is
recommended.

• Other factors could have affected the results of this

study. The correspondence between pretreatment
and early-in-treatment CCRTs may be due to the
relatively severe state of depression in which the
patients presented at the time. Depression may
have influenced the content of the narratives in a
convergent direction at both times; that is, in the
two kinds of narratives, depressed patients may
tend to perceive others as “rejecting” or include
others who “are rejecting” or cause others to
reject them. The kappa coefficient, however, was
intentionally used to deal with this base-rate
problem.

• It is also possible that the judges’ use of only two

or three clustered standard categories for the
responses from others indicates some problems
with the current version of the clustered standard
categories.

www.theipi.org 676
www.freepsychotherapybooks.org
 • At least 1 month had passed between the RAP
interview and the third treatment session. During
this month, many changes may have occurred
(e.g., slight changes in the CCRT and moderate
relief of depression) that could have lowered the
reliability estimates. Therefore, the two
procedures may be even more similar than the
results suggest.

Note

[12] This chapter is a reedited version of an article by Barber et

al. (1995) from the Journal of Consulting and Clinical
Psychology, 63, 145-148. Reprinted with permission.

www.theipi.org 677
www.freepsychotherapybooks.org
 18

THE MEASUREMENT OF MASTERY OF

RELATIONSHIP CONFLICTS

BRIN F. S. GRENYER AND LESTER

LUBORSKY13

MEASUREMENT OF MASTERY

A special gift to the clinician from the CCRT

method is its capacity to describe the central

relationship conflicts. But what the clinician also

needs is a method that shows the level of mastery of

these central relationship conflicts. That is the agenda

of this chapter.

One of the central propositions in the

psychoanalytic theory of change in psychotherapy is

that symptoms arise after the activation of

relationship conflicts. This proposition has been

www.theipi.org 678
www.freepsychotherapybooks.org
 supported by two strands of research: first, studies
that reveal the structural pattern of these conflicts

using the Core Conflictual Relationship Theme

method (see chapters 2 and 9 and other chapters of

this book); second, studies that link these relationship

conflicts with the emergence of symptoms (as was
shown in Luborsky, 1996). When patients seek
psychotherapy, it is often because they are

overwhelmed by relationship conflicts and

consequent symptoms; the almost universal goal in

psychotherapy is to promote mastery over these

problems (Liberman, 1978). Therefore, the goal of

our study was to assess changes in the mastery of the

core interpersonal conflicts over the course of

psychotherapy and examine their relation to changes

in symptoms.

Mastery is defined as the acquisition of emotional

self-control and intellectual self-understanding in the

context of interpersonal relationships (Grenyer,

www.theipi.org 679
www.freepsychotherapybooks.org
 1994). Gains in mastery come about as part of the
working-through process. We propose that the

conflictual relationship narratives in psychotherapy

are partly told in the service of mastery, just as Freud

wrote that children’s repetitive games were attempts

to master traumatic situations (Freud, 1920/1955a). In
addition, Freud was the first to make the connection
between patterns in the patient’s narratives about

conflictual problems outside of therapy with the kind

of problems experienced within therapy. To our

knowledge, there have been no previous attempts to

measure the process of mastery of the conflicts in

psychotherapy. However, research with the CCRT

presents some findings that need to be considered in

relation to mastery (Crits-Christoph & Luborsky,

1990; also see chapter 10, this volume). We expected

that the repetitive maladaptive relationship conflicts
would become less pervasive over the course of

therapy, that is to say, the CCRT pattern would

www.theipi.org 680
www.freepsychotherapybooks.org
 become more positive and include a wider range of
relationship patterns indicating greater flexibility in

emotional response to conflicts. The results reported

in chapter 10 supported the hypothesis, with the

striking finding that despite the decrease in

pervasiveness, much of the CCRT patterns were still
evident, supporting the view that central relationship
patterns tend to remain recognizable over a

psychotherapy.

Although the changes in the positive and negative

components of the CCRT provide some indication of

changes in the quality of the pattern and in its

pervasiveness (see chapters 4, 8, and 10), the CCRT
is limited in its scope as a measure that reflects
mastery. The Mastery Scale was therefore
constructed, which can be applied to the same

database of narratives of relationship episodes but

also focuses on quantifying degrees of mastery. Table

1 shows a brief summary of the Mastery Scale; for

www.theipi.org 681
www.freepsychotherapybooks.org
 TABLE 1
Mastery Scale, Version I

Level and
Components
Score
Level 1. Lack of impulse control
1A Expressions of being emotionally overwhelmed
1B References to immediacy of impulses
1C References to blocking defenses
1D References to ego-boundary disorders
Level 2. Introjection and projection of negative affects
2E Expressions of suffering from internal negative
states
2F Expressions indicative of negative projection
onto others
2G Expressions indicative of negative projection
from others
2H References to interpersonal withdrawal
2I Expressions of helplessness
Level 3. Difficulties in understanding and control
3J Expressions of cognitive confusion
3K Expressions of cognitive ambivalence
3L References to positive struggle with difficulties
Level 4. Interpersonal awareness
4M References to questioning the reactions of
others
4N References to considering the other’s point of
view
4O References to questioning the reaction of the
self
4P Expressions of interpersonal self-assertion

www.theipi.org 682
www.freepsychotherapybooks.org
Level 5. Self-understanding
5Q Expressions of insight into repeating
personality patterns of self
5R Making dynamic links between past and
present relationships
5S References to interpersonal union
5T Expressions of insight into interpersonal
relations
Level 6. Self-control
6U Expressions of emotional self-control over
conflicts
6V Expressions of new changes in emotional
responding
6W References to self-analysis

Note. For the full details of this scale, consult Grenyer (1994).

www.theipi.org 683
www.freepsychotherapybooks.org
 the full scale including scoring conventions and
practice examples, consult Grenyer (1994).

The content of the scale was developed on the

basis of our definition of mastery from our review of

the literature and after the intensive study of verbatim

transcripts of two successful pilot cases of dynamic
psychotherapy applying a task-analysis approach
(Rice & Greenberg, 1984). Task analysis is a
structured discovery-oriented approach to studying
psychotherapy transcripts to reveal recurrent patterns

of clinical importance. We were interested in dynamic

concepts that were likely to indicate self-control and

self-understanding, such as having insight into
common personality traits, making links between past
and present ways of relating, the development of
tolerance for thoughts and feelings, and the ability to

self-analyze and monitor internal states. These were

identified by Luborsky as among the key curative

factors in dynamic therapy (Luborsky et al., 1988).

www.theipi.org 684
www.freepsychotherapybooks.org
 We saw, just as Gottschalk had some 25 years earlier,
that psychological constructs could be reliably and

validly located, classified, and measured within the

patient’s speech (Gottschalk, Winget, & Gleser,

1969).

The Mastery Scale has three broad levels. Scores

1 and 2 relate to failures of mastery manifested by
problems such as cognitive disturbances. Scores 3
and 4 relate to the struggle to improve, such as the
self-questioning of perceptions of relationship

conflicts. Scores 5 and 6 demonstrate high levels of

mastery, for example, having awareness of one’s

transference patterns and being able to derive
pleasurable experiences from relationships. Self-
control was accorded a higher rating than self-
understanding on the basis of Freud’s well-known

view that intellectual self-understanding by itself

does not guarantee therapeutic change (Freud,

www.theipi.org 685
www.freepsychotherapybooks.org
 1958c). Mastery is gained when one not only

understands a situation but also feels in control.

We evaluated the following hypotheses: (a) that

patients rated as showing greater gains in mastery

will have larger gains on measures of general

functioning and symptoms than patients showing

fewer gains in mastery and (b) that changes in

mastery will parallel changes in the components of

the CCRT. The first hypothesis is important because

dynamic theory holds that improvements in the

mastery of interpersonal conflicts are associated with

higher levels of functioning as judged by independent

assessors.

METHOD

Forty-one patients (29 female, 12 male; mean age,

25; range, 18-48) were chosen as a representative

sample from the group of 72 patients who

participated in the Penn Psychotherapy Project

www.theipi.org 686
www.freepsychotherapybooks.org
 (Luborsky et al., 1988). Twenty-six patients were

single, 7 were married, and 6 were divorced or

separated (with 2 with missing data). Five had
graduated from high school only, 19 had completed
some college education, 6 had completed college, and

11 were undertaking or had completed a graduate

degree. The sample had a mixed diagnostic picture

according to the criteria of the Diagnostic and

Statistical Manual of Mental Disorders (3rd ed.;

DSM-III, American Psychiatric Association, 1980).

Fifteen had primary diagnoses of dysthymia, and 11
had generalized anxiety disorders; the rest of the

primary and secondary diagnoses were mainly

Cluster A (8 schizoid, 3 schizotypal), Cluster B (4
histrionic, 1 narcissistic), and Cluster C (3
compulsive, 3 passive-aggressive) Axis II personality
disorders.

Treatment was based on weekly individual time-

unlimited psychoanalytic psychotherapy with a mean

www.theipi.org 687
www.freepsychotherapybooks.org
 treatment length of 54 weeks (range, 91-149 weeks).
Therapy was conducted by 31 psychiatrists (mean

age, 36; range, 26-56). Of these, 17 were psychiatric

residents, 9 had up to 10 years of postresidency

experience, and 5 had more than 10 years of

experience. The residents saw their patients in an
outpatient clinic, and the postresidents saw their
patients in private practice. Thirty of the therapists

were married, and 23 had children. The orientation of

the group was divided between “psychodynamic

eclectic” (21 adherents) and “Freudian analytic” (10

adherents).

Verbatim transcripts of psychotherapy sessions

collected during the Penn Psychotherapy Project for
each patient formed the database. These were mainly
transcripts from early in therapy (generally Sessions 3

and 5) with two or three transcripts from late in

therapy (when treatment was 90% completed).

Narratives of interactions (relationship episodes,

www.theipi.org 688
www.freepsychotherapybooks.org
 REs) served as the units for the analysis of mastery.

Relationship episodes had been identified from the

transcripts of early and late sessions in an earlier

CCRT study (Luborsky, 1977b, 1990a). There were

usually 10 relationship episodes from early in therapy

and 10 relationship episodes from late in therapy, and
they were randomized among sessions and patients.
The relationship episodes were divided into

grammatical clauses (whether independent or

dependent) by marking off the claused speech units

with a slash according to the conventions adopted by

Gottschalk et al. (1969). The following is an example
of three marked clauses, with Mastery Scale scores in
parentheses: /I’m afraid of myself (2E)/because it’s a
father-lover sort of thing (5Q)I It's also this hangover
from when I was real young (5R)/. To facilitate the

process of scoring, one prescoring judge read all the

relationship episodes and identified all the clauses

that could be scored with the Mastery Scale, a

www.theipi.org 689
www.freepsychotherapybooks.org
 technique also used in CCRT research to control for
possible disagreement about location (as opposed to

scoring disagreement).

All the data were scored twice. Each relationship

episode was independently scored by two of a pool of

four trained judges. Each judge was given a random
portion of the total number of relationship episodes to
score. No individual judge scored the same
relationship episode twice. Judges were not informed
of which patient told the relationship episode, the

time in therapy at which the relationship episode

occurred, treatment outcome status, or other clinical

variables. Judges were trained in the methods of
scoring to an interrater reliability of greater than .90.
One of the 23 Mastery Scale categories from 1A to
6W was assigned to each of the codable clauses by

the judges. Each of the 23 category choices comes

with its own built-in score ranging from 1 to 6 to

represent one of the six levels in the scale. These

www.theipi.org 690
www.freepsychotherapybooks.org
 scores were used in the compilation of statistics. We

calculated Mastery Scale scores for each relationship

episode by summing all the scores and dividing by
the number of scorable clauses to arrive at a mean

score per narrative. These scores were then used to

calculate average levels of mastery for each patient
early and late in therapy.

Outcome measures were collected at the

beginning and at the termination of therapy by an
independent assessor using the Health-Sickness

Rating Scale (HSRS; Luborsky, 1962) and the

Control and Insight ratings of the Prognostic Index

(Luborsky et al., 1988); by the therapist’s composite

rating of patient satisfaction, success, and

improvement; and by patient self-report with the

Hopkins Symptom Checklist (SCL; Derogatis et al.,

1970) and a rating of improvement on the primary
target complaint identified by the patient at the start
of therapy (Battle et al., 1966). The response of self

www.theipi.org 691
www.freepsychotherapybooks.org
 and response from other components from late in

therapy were used. To obtain a score late in therapy

that reflected the overall degree of positivity-

negativity for each of the two CCRT components
(response of self and response from other) for each

patient near termination, we subtracted the sum of the

negative responses from the sum of the positive

responses and divided that by the total number of

responses.

RESULTS

Interjudge agreement was uniformly high, with

correlation coefficients among the four independent

judges as follows: A versus B, r = .75 (n = 187 REs

scored in common); A versus C, r = .77 (n = 161); A
versus D, r = .81 (n = 89); B versus C, r = .79 (n =
149); B versus D, r = .85 (n = 127); C versus D, r =

.89 (n = 81). The judges’ Mastery Scale scores were

therefore averaged in all subsequent analyses.

www.theipi.org 692
www.freepsychotherapybooks.org
 To investigate changes in Mastery Scale scores

across the 41 patients over the course of therapy, we

performed a paired t test between early and late

scores. The change in mastery was highly statistically

significant, t(40) = 4 94, p > .0001. The effect size

was large (1.35). When compared with the published
effect sizes in other psychotherapy studies, the
changes detected by the Mastery Scale can be

considered to be of clinical significance (Lambert &

Bergin, 1994). Thus, the trend in this psychotherapy

sample was for patients to display greater levels of

self-understanding and self-control in their
interpersonal relations late in therapy.

Pretreatment–posttreatment change estimates

were corrected for initial levels by the calculation of

residual gain scores for the Mastery Scale and other

outcome variables in which change estimates were
required. The relationships between Mastery Scale
change scores and outcome variables were calculated

www.theipi.org 693
www.freepsychotherapybooks.org
 TABLE 2
Pearson Correlations Between Mastery Scale Residual Change
Scores and Clinical Outcome Scores

Rating r
Observer ratings of outcome
Health-Sickness Rating Scale residual change .51***
score
Prognostic Index, Control item .30
Prognostic Index, Insight item .01
Therapist ratings of outcome
Therapist rating of patient satisfaction, success, and .47**
improvement
Therapist rating of patient achieving insight .12
Patient ratings of outcome
Rating of change of primary target complaint .59***
Symptom Checklist residual change score -.53***
CCRT outcome ratings
CCRT: Response of self late in the therapy .37*
CCRT: Response of other late in therapy .06

Note. N = 41. CCRT = Core Conflictual Relationship Theme.

*p< .05. **p< .01. ***p< .001.

www.theipi.org 694
www.freepsychotherapybooks.org
 and appear in Table 2. Significant relationships were

found among Mastery Scale change scores and

observer, therapist, patient, and CCRT ratings of

outcome. Figure 1 shows the HSRS residual change

scores plotted against the Mastery Scale residual

change scores.

Figure 2 shows the percentage of change in the

frequency of Mastery Scale categories appearing in
narratives from early to late in psychotherapy for all

41 patients. To illustrate these typical changes in

mastery from early to late in therapy, we briefly
describe one patient. Ms. Simpson, a 24-year-old

divorced graduate student with no children, was seen

in weekly therapy for 41 weeks with the goal to help

change her difficult “personality patterns.” Her

psychodynamic therapist was a 31-year-old married

psychiatric resident. Early in therapy, she expressed
suffering (2E) that was due to conflictual interactions
with others, which led to her avoiding relationships

www.theipi.org 695
www.freepsychotherapybooks.org
 Figure 1.

Distribution of Health-Sickness Rating Scale residual

change scores versus Mastery Scale residual change
scores for all 41 patients.

www.theipi.org 696
www.freepsychotherapybooks.org
 Figure 2.

Change in the frequency of Mastery Scale categories

appearing in narratives late in psychotherapy,
expressed as percentage change from early in therapy.
Data are for all 41 patients. Dimensions indicative of
poor mastery (Categories A-L) show a reduction in
appearance in narratives late in therapy, whereas
interpersonal awareness, self-understanding, and self-
control dimensions (Categories M-W) show a
corresponding increase in appearance late in therapy.
Data at 0% indicate no change in the percentage of
appearance of categories from early to late in therapy.

www.theipi.org 697
www.freepsychotherapybooks.org
 (2H). When in close relationships, she felt worthless
and guilty (2F) and encouraged men to hit her (ID).

Toward the end of therapy, she could see (5Q) that

her global view that “men are evil” was due to

unconscious hostility toward an abusing person from

her childhood (5R). She began to struggle free from
these bonds (4P) and enjoy relationships (5S) in a
new way (6V). These conflictual patterns also

appeared within the early transference relationship

with her therapist. Toward the end of therapy, she

could express with confidence to her therapist that

“you basically seem good to me now” (6V), thus

showing some mastery over her interpersonal

problem.

SUMMARY, DISCUSSION, AND CONCLUSIONS

This chapter is a departure from the theme of the

book in that it is only partly on the subject of the

CCRT method; its focus is on the Mastery Scale,

which can be used to complement the CCRT. The

www.theipi.org 698
www.freepsychotherapybooks.org
 study used the same clinical sample as in the studies

discussed in the rest of this book, that is, a sample

from the Penn Psychotherapy Project. The Mastery
Scale, like the CCRT, uses the narratives within

psychotherapy transcripts to quantify gains in

mastery of core interpersonal conflicts and symptoms.

• The results show that theoretically relevant and

central psychodynamic variables can be reliably
measured directly from the content of verbal
communications that patients give in therapy.
The high interrater reliability obtained for the
Mastery Scale was achieved in part because the
judging task was highly structured, obviating
disagreement that can result from methods based
on the usual unguided, complex inferential
judgments. Judges did not assess levels of
mastery per se; the corresponding scores were
already built into the category choices. This
method of content analysis has proved to be a
powerful way of identifying underlying
constructs (Gottschalk et al., 1969).

• The fundamental conclusion of this study is that

established maladaptive interpersonal patterns

www.theipi.org 699
www.freepsychotherapybooks.org
 for meeting needs and wishes can be better
mastered (understood and controlled) during the
course of more successful psychotherapy. The
primary hypothesis was that patients who were
rated as showing greater gains in mastery would
have greater gains on measures of general
functioning and symptoms than patients showing
lesser gains in mastery. The HSRS is an
important global outcome measure of
functioning (Luborsky et al., 1993); a slight
revision of the HSRS, the Global Assessment
Scale, constitutes Axis V of DSM-III and DSM-
IV. Changes in HSRS were significantly related
to changes in mastery (Figure 1); patients who
are sicker are less able to see the interpersonal
dynamics of their predicament and tend to react
with more helplessness and pain to problems in
getting their needs met. Perhaps of most interest,
the patient’s own judgments of changes in their
main complaint (that is, a target symptom)
paralleled changes in the mastery of
interpersonal conflicts found in their narratives.
These changes were related not only to changes
in reported symptoms but also to the fulfilling of
the patient’s main goal in therapy.

www.theipi.org 700
www.freepsychotherapybooks.org
 • The second hypothesis was that changes in mastery
would parallel changes in components of the
CCRT. We found that mastery was significantly
related to the response of self, which suggests
that our scale is tapping important aspects of the
measure and may be seen as complementary to
the CCRT method. We hypothesized that the
response of other would also be related to
mastery because the ability to elicit positive
responses from others should be a part of gains
in mastery. When we rescored the data from 20
patients using a finer grained measure of
positivity and negativity (see chapter 4, this
volume), the results were essentially the same:
Changes in the CCRT-RS dimension significantly
paralleled changes in mastery, but changes in the
CCRT-RO dimension were not significantly
related to mastery. That we did not find such a
relationship is less surprising in retrospect
because the Mastery Scale specifically limits
scoring to self-statements and self-reflections on
others, whereas in the CCRT statements made by
others are scored. For example, direct quotations
of others in narratives are considered scorable in
the CCRT but not in the Mastery Scale.

www.theipi.org 701
www.freepsychotherapybooks.org
 • It is noteworthy that both the observer Insight and
Control ratings from the Prognostic Index
(Luborsky et al., 1988) and the therapist’s rating
of insight failed to show a relationship with
mastery (Table 2). It may be that the Prognostic
Index and therapist measures suffer in validity
because of the demands made on judges’ scoring
patterns (Luborsky et al., 1988). In addition, our
scale subsumes insight and control into a single
concept, mastery, which is different in important
ways from either variable.

• Further inspection of the change data for the

individual categories of the Mastery Scale
(Figure 2) reveals some interesting findings. In
general, psychotherapy leads to a diminution in
three of the lower levels of mastery: lack of
impulse control, introjection and projection of
negative affects, and difficulties in understanding
and control (from Categories A-L). This
indicates that there is a general reduction in
distress and confusion in interpersonal
relationships over the course of psychotherapy.
As predicted, certain dynamic variables showed
an amplification over therapy, such as
expressions of self-control (6U), which showed a

www.theipi.org 702
www.freepsychotherapybooks.org
 large (16.5%) increase. It is noteworthy that a
few dynamic variables showed no change (e.g.,
making dynamic links between past and present
relationships [5R]). This could be partly
attributed to the infrequent appearance of this
category: Only 6.5% of narratives in therapy
contained scorable clauses for this category. It
may be that this category is dynamically
important in therapy but that our method is not
sensitive enough to reveal its significance. As
indicated by the modest percentage changes in
Figure 2, we have found support for the view that
psychotherapy does not completely eliminate
relationship conflicts but helps people to gain
mastery over them. Our case study of Ms.
Simpson illustrates the clinical relevance of the
scale’s categories.

The major strength of the present study is that

important variables have been systematically scored

from the content of the patient’s verbal

communications and linked to central outcome

variables in therapy. We recognize, however, that

what we have gained in predictive power has been at

www.theipi.org 703
www.freepsychotherapybooks.org
 the expense of specificity. Our concept and

measurement of mastery is broad and reflects many

separate cognitive and affective psychological
constructs such as mood, hope, anxiety, helplessness,

and locus of control. We did not set out to investigate

such specific interrelationships, but we subsumed
aspects of these constructs within our view of
mastery. We also recognize that our measurement of

mastery may not include some factors that are

important to the concept, such as the assessment of

the degree to which an insight statement is salient to

the person’s core problems. But our limitation was
also our strength: We have avoided complex and
possibly unreliable methods of scoring, and in so
doing, we may have forgone some subtle therapeutic
factors; yet the method devised is robust and captures

an important dynamic change variable.

Note

[13] This chapter is a revised version of an article by Grenyer

and Luborsky (1996) from the Journal of Consulting and

www.theipi.org 704
www.freepsychotherapybooks.org
 Clinical Psychology, 64, pp. 411-416. Reprinted with
permission.

This chapter was supported in part by an APRA-

Australian Research Council award, Research Scientist
Award MH 40710-22, National Institute on Drug Abuse
Grants 2 K05 DA00168-23A 24 and RO-I DA0785, and
National Institute of Mental Health Clinical Research
Center Grant MH 45178. Acknowledgment is made to the
Penn Psychotherapy Research Project for providing access
to data and support. Our thanks are also extended to Vera
Auerbach, Mary Carse, Annalisa Dezarnaulds, Louis
Diguer, Suzanne Johnson, Nigel Mackay, Richard Rushton,
Kelly Schmidt, and Nadia Solowij.

www.theipi.org 705
www.freepsychotherapybooks.org
 III

CLINICAL USES OF THE

CCRT

www.theipi.org 706
www.freepsychotherapybooks.org
 19

THE EVERYDAY CLINICAL USES OF

THE CCRT

LESTER LUBORSKY

Ever since the bright day in 1975 when the idea

for the CCRT was conceived, I and then others have

been learning about its capacities for helping

therapists and patients who are in dynamic as well as

other psychotherapies. The CCRT method has shown

clear advantages over the usual, unguided clinical

methods. Its assets come both from its guided method
of formulation and from the evidence of its
consequent reliability (see chapter 6, this volume).
Clinicians whose formulations are guided by the

CCRT method, therefore, are likely to be blessed with

more interclinician concordance than those who are

not so aided.

www.theipi.org 707
www.freepsychotherapybooks.org
 As I explain and illustrate in this chapter, each of
the triad of tasks of therapists in dynamic and related

psychotherapies can be helped by the CCRT method.

These tasks, as outlined in my 1984 manual, include

(a) listening to the patient’s communications; (b)

figuring out formulations, often about the conflicts
within the CCRT; and (c) giving interpretative
responses selected from the CCRT. The therapist then

listens further, aided by the feedback circuit from the

patient’s response to the therapist’s previous

response, and then gives further interpretations. The

chapter begins with detailed explanations of the uses

of the CCRT in dynamic psychotherapy.

Subsequently, I discuss the ways that different

systems of psychotherapy result in different choices of

interpretations. Lastly, I discuss how therapists can

learn to rely on the CCRT to help with their tasks in
psychotherapy.

www.theipi.org 708
www.freepsychotherapybooks.org
THE USES OF THE CCRT IN DYNAMIC
PSYCHOTHERAPY

As a Guide to Formulations and Interpretations

My account of the applications of the CCRT

method to the basic triad of therapeutic tasks will

sound familiar to psychodynamic clinicians, even

those who are not acquainted with the CCRT method.

In fact, the following example of the application of

the principle took place before CCRT was ever

invented:

Example: Mr. Howard (see chapter 5, this

volume) was a college student who came to
treatment because of his extreme proneness to
anxiety and guilt. In Session 3 he told a series of
narrative, in the first of which, at the age of
about 13, he asked his mother a question about
sex, and his mother frustrated him by saying that
he was getting too old for that. This was
followed by a second narrative from about the
same age in which his mother turned down his
wish to get into bed with his parents on a cold
night; she explained that it was okay for his
younger brother but not for him, because he was

www.theipi.org 709
www.freepsychotherapybooks.org
 now too old for that. The third narrative was an
enactment with the therapist of a similar
relationship pattern: The patient had become
anxious, developed a headache, and explained
that he was becoming “unresponsive” to the
therapist. The therapist used the information
from the two antecedent relationship episodes to
formulate a relationship pattern and then to
interpret to the patient that the patient was
expecting the therapist to be unresponsive to the
patient’s wishes.

The therapist in this example had been listening

to the patient’s relationship episodes. He based his

formulation and then his interpretation on one CCRT

component of these episodes: the patient’s expected

response from the other person. The therapist’s

inference about the patient’s expected response

appeared to be based on the three relationship

episodes, especially the third, in which there was an
enactment of the pattern in relation to the therapist.
Another way to describe the therapist’s technique is
that the therapist’s interpretation of the content of the

www.theipi.org 710
www.freepsychotherapybooks.org
 relationship episodes had a high degree of
convergence with the CCRT. Such convergence as a

basis for interpretation is desirable, as demonstrated

in chapter 13, this volume, on accuracy of

interpretation.

The conventional style of using the CCRT for

deciding on interpretations is demonstrated in the
example from the treatment of Mr. Howard (see
chapter 5): The therapist includes within the
interpretation a recurrent facet of the CCRT derived

from the relationship episodes. Both within sessions

and over time, parts of the CCRT that are

incorporated in interpretations help patients to build
up a concept of their recurrent CCRT pattern. This
conventional style might be called a piecemeal
presentation of parts of the therapist’s CCRT

formulation.

www.theipi.org 711
www.freepsychotherapybooks.org
 Another, more focused style is used less
commonly by some therapists: After several sessions

of piecemeal presentation of parts of the CCRT, the

therapist suggests trying to make a joint formulation

of the central relationship pattern that guides the

patient’s conduct of relationships. The therapist
introduces the joint task by a comment such as, “Let’s
try making a sketch of the pattern that we hear in the

events that you tell me about in your relationships. In

the last few episodes, what is it that you wanted from

the other person?” After that part is agreed on, the

patient and therapist concentrate on the question,

“What is it that you tend to expect from the other

person?” After that is agreed on, the therapist asks,

“How is it that you react?” A further procedure might

have benefits for some patients: The therapist may

say, “I will write this out and you can write this out,
so that we can examine together how well they fit.”

The method might be called a joint patient-therapist

www.theipi.org 712
www.freepsychotherapybooks.org
 construction of the CCRT formulation. The method

has something in common with the system used by

adherents of the cognitive-analytic school, as

reflected, for example, in work by Ryle (1990, 1991).
This unusual patient-therapist system of joint

construction of the CCRT seems especially useful for

patients who have a hard time, with the usual

piecemeal presentation of the therapist’s formulation,

becoming aware of the pattern in the conduct of their

relationships.

In fashioning an interpretation, it can be helpful to

choose language that will minimize the patient’s

defensiveness; generally, the greater the

defensiveness the less a patient can use an

interpretation. Language should be used in which

sympathy for and recognition of the patient’s typical

responses are conveyed, along with the CCRT-related

content (consistent with suggestions by Wachtel,
1993). In the case of Mr. Howard described earlier, a

www.theipi.org 713
www.freepsychotherapybooks.org
 therapist’s interpretation might be “When you

experienced rejection of your wish for closeness, it

actually felt like there was nothing you could do
about it, at that time in your life.” This interpretation
clearly contains parts of the CCRT, but the language

chosen for the interpretation conveys sympathy for

and understanding of why the patient felt so helpless

and anxious at that time in his life.

As an Aid to Maintaining a Treatment Focus

The treatment focus is kept on the therapist’s and

patient’s attention to the CCRT for providing

formulations and interpretations. The CCRT for each

patient may change some from session to session, but
typically it does not change drastically, so that it
offers a fairly consistent focus. Therapists find such a

focus especially useful for brief therapy; in fact,

according to Koss and Butcher (1986, p. 650), short-

term therapies are more focused than long-term

therapies.

www.theipi.org 714
www.freepsychotherapybooks.org
 One gain to the therapist from reliance on the
CCRT as the treatment focus is that it offers a

framework from which to select interpretations. The

patient also benefits because a focused treatment

tends to increase the patient’s motivation and

concentration of effort, which may “hothouse” the
growth toward the specific goals of the treatment.
This may be why it has been found that changes that

are related to the specific goals of treatment are

greater than changes in general outcome measures

(Luborsky et al., 1988): The interpretative focus can

serve as a motivator because the target for change has

been made explicit. The patient then puts more

concentrated effort into changes that are within his or

her defined targets of the treatment. It happened that

way in the treatment of Mr. Howard: The therapist

tended to focus on the patient’s wish not to be cut off
from affection and closeness and on his recurrent

expectation that he would be cut off from these

www.theipi.org 715
www.freepsychotherapybooks.org
 qualities. The patient spent a lot of time in the course

of the treatment suffering from the experience of that

theme but also in trying to deal with it. He became

familiar with his own readiness to experience the

wish and his negative expectations of other people’s

responses. In his follow-up (see chapter 7, this
volume) many years later, he confirmed how familiar
he had become with that theme and the extent to

which his gain in familiarity with it mitigated his self-

blame and anxiety when he experienced it.

As a Help in Choosing a Part of the CCRT for Each

Interpretation

Less is sometimes more. There is no special gain

from using the entire CCRT every time one needs to

draw on it for an interpretation. My colleagues and I

have discovered guides to the selection of

components of the CCRT for making interpretations;

these were noted in our presentation of results in

chapter 13 on the predictive value of accuracy of

www.theipi.org 716
www.freepsychotherapybooks.org
 interpretation. We found that accuracy based on the
use in interpretations of the response of self was not

correlated with outcome but that accuracy based on

interpretation of the wish plus response from other

was correlated with better outcomes (note that

because the wish and response from other were
highly correlated, they were combined). These
findings led to the conclusion that, in terms of

potential benefit for the patient, interpretations that

combine the wish and response from other are

effective in the sense that they are associated with the

patient’s greater benefit from treatment.

Further guidelines for fashioning interpretations

come from observations of the interpretative behavior
of effective therapists. A clinical analysis of these
observations suggests two principles: (a) One should

choose the aspect of the CCRT that is most involved

in conflict, or most related to the patient’s main

source of suffering and symptoms; and (b) one should

www.theipi.org 717
www.freepsychotherapybooks.org
 choose the part of the CCRT the patient is most

responsive to and seems able to use best in effecting

change. The second principle is much like the
recommendation to listen to the patient’s responses

after trial interpretations.

Example, Ms. Cunningham: An example of the

complexity of making a choice from within the
CCRT is derived from Ms. Cunningham’s
Session 5 (see chapter 5, this volume). The
therapist interpreted her wish for reassurance,
which was part of her CCRT in that session,
even though in the CCRT it was only the third-
ranked wish in order of frequency in the session;
the more frequent wishes were “to dominate and
control” and “to overcome the other’s
domination.” But the therapist may have decided
to make the interpretation on other clinical
grounds, such as the second of the two principles
listed above, that the patient should be able to
use the interpretation in effecting change.

As a Help in Timing Interpretations

The therapist’s familiarity with the CCRT in a

session eases the tracking of the meaning of what is

www.theipi.org 718
www.freepsychotherapybooks.org
 unfolding in the patient’s communications and in that

way helps to ensure the best timing for an

interpretation. The exact timing of an interpretation

requires a sensitive clinical judgment of favorable
conditions. First, the theme that is to be interpreted

must be experienced by the patient; second, it must

be near to awareness; and third, it must be impeding

the treatment. These judgments by the therapist form

the basis for the classical principles of timing, which

are based on when the transference has become a

resistance and the patient shows that related ideas are
near to awareness (Bibring, 1954).

Example, Mr. Howard: The example given

earlier from the treatment of Mr. Howard also
provides an illustration of a therapist’s
application of the principles of timing. In the
third session the patient told two relationship
episodes dealing with fear of being cut off from
closeness and affection from his mother; then, in
a third episode with the therapist, he stopped
talking freely and referred to himself as being
“unresponsive.” The therapist, therefore, inferred

www.theipi.org 719
www.freepsychotherapybooks.org
 that transference material was involved in this
resistance behavior and decided that this was a
good moment for making an interpretation about
the transference.

As a Clue to the Conflicts Sparking the Formation of

Symptoms

It can be useful to the therapist and therapeutic to

the patient to understand the conditions connected

with the onset of the symptoms (see chapter 15 in

Luborsky, 1996). The format of the CCRT helps to

locate these conditions. Symptoms, in CCRT terms,

emerge from (a) conflicts between wishes and (b)

conflicts between wishes and expected responses

from others. When these conflicts become more
intense, the symptoms are more likely to appear in
the responses of self. Ms. Smyth’s depression (see

chapter 5, this volume) was most likely to worsen

when this conflict was experienced: a greater

intensity of the wish to end nonsupportive

relationships and to get support but with the

www.theipi.org 720
www.freepsychotherapybooks.org
 expectation that the other person would be rejecting

and unsupportive. Mr. Howard’s symptoms were

more likely to worsen when the following conflict

heated up: greater intensity of the wish to be close

and receive affection and not experience a loss of

relationships but with the expectation that the other
person will reject his wish, followed by the responses
of resentment, self-blame, and anxiety. Ms.

Cunningham’s symptom of inhibition increased,

along with her conflicts around experiencing reduced

control and intensification of her wish to be in control

but with her expectation that the other person would
not give her what she wants.

A more exact assessment of the preconditions of

symptoms can be achieved by a related method called

the symptom-context method (Luborsky, 1996). That

method can begin to be used directly after a symptom
materializes in a session. When a series of recurrent
episodes of symptoms has occurred, their context can

www.theipi.org 721
www.freepsychotherapybooks.org
 be further inspected, and it is possible to identify their
typical antecedents. A comparison of the two

methods, the symptom-context and the CCRT, has

revealed that similar conflicts are found through each

method (Luborsky, 1996; Luborsky et al., 1985b).

As a Supplement to DSM Diagnoses

Patients nowadays tend to come to

psychotherapists with already determined diagnoses,

and more and more often these are based on the DSM
family of diagnoses. But these diagnoses tend not to

be very helpful to the therapist for the conduct of the

psychotherapy. The major exceptions are diagnoses

that include psychotic features, for these usually

imply the need for a greater use of supportive

techniques, the reduced use of expressive techniques,

and the possible use of pharmacotherapy (Luborsky,

1984).

www.theipi.org 722
www.freepsychotherapybooks.org
 The DSM diagnoses by themselves are empty of

content about the patient’s psychodynamics, as

pointed out by Karasu and Skodol (1980) and
Auerbach and Childress (1988). Such diagnoses do

not give information about the patient’s typical

relationship patterns and the conflicts within them. In
contrast, I suggest a simple but informative addition:
to supplement the DSM diagnosis with the CCRT.

Such supplementation would reveal more about the

association between the two classes of diagnostic

information, the DSM diagnosis and the central

relationship pattern.

As a Special Aid in the Functioning of Inpatient Units

The CCRT can have a vital place in the proper

functioning of inpatient treatment units, helping

professionals and other workers keep focused on

treatment goals rather than only on custodial or

diagnostic goals. The recommended treatment

procedures are the following (Luborsky, van

www.theipi.org 723
www.freepsychotherapybooks.org
 Ravenswaay, Ball, Steinman, Sprehn, & Bryan,
1993): The initial conference of the treatment team

includes the completion of a CCRT. One of the

members of the team brings to the conference a set of

narratives told by the patient about relationships and

constructs a CCRT from them. The team then
discusses the logical treatment goals for the patient,
derived from the CCRT, to be achieved during the

patient’s stay. Each of the team members agrees on

these goals, so that in their contacts with the patient

they are able to be as helpful as possible toward
achieving them. Because the initial team has agreed

on these goals, the team members’ treatment aims can

have a concerted impact on the patient. Before the

patient leaves the inpatient unit, one of the team

members reviews with the patient his or her

accomplishments in terms of these goals.

THE DIFFERENT CONSEQUENCES OF

DIFFERENT PSYCHOTHERAPEUTIC SYSTEMS
ON THE INTERPRETATIVE FOCUS

www.theipi.org 724
www.freepsychotherapybooks.org
 Although virtually all exponents of brief

psychotherapies advocate consistency of

interpretative focus, the method of finding the focus

differs for each type of psychotherapy. The emphasis

on maintaining a focus is found even in long-term

psychotherapy and in psychoanalysis, but there the
urgency for maintaining it tends to be less strong. The
following discussion of different systems of therapy

shows their somewhat different methods for finding

the focus as well as the consequences of each

method.

Dynamic Therapy

In this therapy the selection of an interpretative

focus necessarily requires that the therapist make a

prior psychodynamic formulation. This formulation

often involves an aspect of the central relationship

pattern. In the example we gave from Ms.
Cunningham’s Session 5 (see chapter 5, this volume),
the therapist must have considered it noteworthy that

www.theipi.org 725
www.freepsychotherapybooks.org
 the patient needed reassurance and, possibly, that she

wanted reassurance from him but that she did not

think she could get it from him or from anyone. As

his main interpretation in that session, the therapist

said, “You expressed a wish to your husband for

reassurance but not here.” The patient responded by
acknowledging that she did not think she would get it
here. In making that main interpretation, the therapist

was following a body of traditional wisdom about

what to focus on in an interpretation that represents

two well-established criteria: (a) that the content

should be close to awareness and (b) that the content
should be related to the patient’s current symptoms in
the sense that the symptoms are impeding the patient
at the moment. In Freud’s words, the therapist should
“wait until the transference … has become a

resistance” (1913/1958c, p. 139).

CCRT-Guided Dynamic Therapy

www.theipi.org 726
www.freepsychotherapybooks.org
 Even with the preceding two criteria, clinical
practice in dynamic therapy is a relatively unguided

system in contrast to the guided systems that use the

CCRT or one of the other relationship pattern

measures (see chapter 20, this volume). Of course,

the guided and unguided systems often coincide in
their implied formulations and consequent
interpretations, as was seen in the example from Mr.

Howard’s therapy presented at the beginning of this

chapter.

A consequence of relying on the CCRT for

finding the focus in a dynamic therapy is that the

focus selected in this way differs from patient to
patient depending on the particular CCRT for each
patient. The differences among patients derive from
the empirical grounding of the CCRT method: In this

format the content of the therapy is not determined in

advance. This patient-specific appropriateness of the

focus is likely to increase the patient’s inclination to

www.theipi.org 727
www.freepsychotherapybooks.org
 recognize that when interpretations are made, he or

she has been listened to and understood.

These and other advantages are offered by the

CCRT-guided short-term dynamic psychotherapy

described by Book (in press); it offers the only fully

recorded and fully published CCRT-guided
psychotherapy manual, together with vividly
illustrated recommended techniques for drawing on
the CCRT.

Davanloo’s Therapy

In contrast to the appropriate diversity of focus

when the CCRT is used, in some forms of brief

psychotherapy the focus of interpretations is fairly

uniform across all patients. In Davanloo’s (1978)

brief psychotherapy the focus of interpretation is
likely to be on understanding the patient’s current
situation in terms of the patient’s passivity as a way
to deal with anger. This was noted in one sample by

www.theipi.org 728
www.freepsychotherapybooks.org
 Gustafson (1986): “All interviews of Davanloo

discover this passivity” (p. 175).

Sifneos’s Therapy

In Sifneos’s (1979) brief psychotherapy the focus

of interpretation across patients typically is on

understanding the patient’s current situation in terms
of the parallels between the early and the current
oedipal triangle themes.

Mann’s Therapy

In Mann’s (1973) 12-session psychotherapy a

uniform focus is recommended for all patients, that

is, to improve the patient’s self-image. According to

Mann, however, the focus is different from the

“central complaint,” and Mann’s formulations differ

from patient to patient because the relationship

problems that determine the poor self-image differ

from patient to patient.

Cognitive Therapy

www.theipi.org 729
www.freepsychotherapybooks.org
 In some nondynamic therapies the need for a
focus is also evident. In one of these, cognitive

therapy (Beck 1989), the therapist selects and then

maintains a focus on individually specified

dysfunctional attitudes. Until recently, the therapist

was not provided with a systematic method for

selecting these attitudes, but there are now systems

for deriving them (for example, Persons, 1989).

PROCEDURES FOR THERAPISTS TO LEARN

TO RELY ON THE CCRT IN PSYCHOTHERAPY

The general orientation given to a therapist who is

about to learn the CCRT method is that the
formulation of the CCRT by the therapist during each
therapy session has much in common with the
formulation of the transference; however, the CCRT

offers more explicit and precise guidelines for this

kind of inference making. These guidelines state that

the therapist listens to the patient’s communications,

especially to each relationship episode as it is told,

www.theipi.org 730
www.freepsychotherapybooks.org
 and notes the types of relationship components that

are most redundant across the episodes; these become

part of the CCRT. Then, from time to time, the
therapist uses pieces of the CCRT in the

interpretations. The therapist can also occasionally

prepare formulations of the CCRT from past sessions;
such reviews can ease the making of formulations
during subsequent sessions.

Some of the requirements of rigorous research

scoring of the CCRT are unnecessary for its clinical

use during psychotherapy sessions. As an example,

for clinical use of the CCRT there is no need to

restrict the choice of relationship episodes to the

complete ones used for research purposes, and it is

not necessary to have them all included. My strong

impression is that the incomplete relationship

episodes have similar, although unexpressed, CCRT
components to the complete ones, although this
comparison has not yet been studied systematically.

www.theipi.org 731
www.freepsychotherapybooks.org
Instruction in the CCRT During Supervision

Learning to use the CCRT in the course of

psychotherapy is best done during individual or group

training in psychotherapy. The instructor teaches by

showing example after example of how the therapist
formulates the central relationship pattern and then
how it is used in helping to shape the therapist’s
interpretative responses. In the course of supervision
sessions, when the therapist presents process notes

(preferably along with a sample of tape recordings or

videotapes), the supervisor can review the

formulation in terms of the CCRT and point out how
it is done. That kind of repeated instruction through
example is the mainstay of the clinical training in the
use of the CCRT.

A special format for intensive training in dynamic

psychotherapy that has proved to be satisfying to its

members consists of an hour and a quarter session,
and a 1-year participation in a four-therapist peer

www.theipi.org 732
www.freepsychotherapybooks.org
 supervision group (Luborsky, 1990c). Four therapists
are a good size for a training group: In 60 minutes
two therapists have 30 minutes each to present and to
have their presentation discussed. During the second
6 months of the year, the other two therapists present

their treatments for discussion. The aim of this group

training is for the therapist to learn how to carry out

supportive-expressive dynamic psychotherapy

following the manual by Luborsky (1984) with

related readings and to help supervise other therapists

in their learning. Each of the peer therapists-in-
training helps the presenting therapist in treating the
patient by following the methods of the treatment

manual by Luborsky (1984). A seasoned therapist

acts as a group leader, largely by filling the role of an

orchestrator of the group. The training, in contrast to

individual supervision, comes in large part from the

peer therapists, who provide their versions of the

formulation to compare with the presenting

www.theipi.org 733
www.freepsychotherapybooks.org
 therapist’s version and who suggest the kinds of
interpretations that would follow from the

formulation.

Scoring Practice Sets of Relationship Episodes

An efficient method of improving skills in scoring

the CCRT is through practice scoring of a graded
series of cases. Each of the practice cases consists of
a set of brief relationship episodes selected from a
session of a different patient. After each practice case
is scored, the therapist is given feedback about the

scoring until the therapist achieves an adequate level

of performance.

Practice With CCRT-Based Interpretations During

Tape Playbacks

A set of sessions can serve for practicing CCRT-

based interpretations during playback sessions. As in

the research method studied extensively by Strupp

(1973), when the tape is stopped, the therapist-in-

www.theipi.org 734
www.freepsychotherapybooks.org
 training must fill in the interruption with the most

appropriate CCRT-based interpretation. Therapists

find this method helpful because of its similarity to a

live session with its requirement to construct

interpretations immediately.

Training Through a Self-Reported and Self-Analyzed

CCRT

A more direct deeper appreciation of what is

measured by the CCRT can be gained by getting a

“free self-analysis”—an analysis of one’s own

narratives for their CCRT (Luborsky, 1980). Many

people who have tried the self-analysis say that it
gives them the surprise that comes with recognition
of the familiar: a quick re-viewing of the central

relationship pattern that they had become familiar

with during their personal intensive psychotherapy or

psychoanalysis. The reader can try the self-analysis

by writing at least 10 relationship episodes and

www.theipi.org 735
www.freepsychotherapybooks.org
 scoring them, following the instructions provided in
Figure 1.

Training in the CCRT During Psychiatric Residency: A

Survey of Therapists’ Benefits

Therapists have reported their impressions about

their own training experiences. They show
considerable consensus about the usefulness of the
CCRT in training for and in carrying out

psychotherapy. Their reports were derived from a

mail questionnaire study by Fried (1989) of 53

therapist respondents who had some exposure to the

CCRT in the course of their training. Most of these

were psychiatrists who had completed residency in

the previous 10 years. These are samples of the

therapists’ responses on a few of the questions:

In answer to a question about the extent to which

they use the CCRT in their psychotherapy
practice, on a scale of not at all, minimally,
somewhat, frequently, or continually, of 47
respondents, 28 chose frequently or continually
and 11 others chose somewhat.

www.theipi.org 736
www.freepsychotherapybooks.org
 Figure 1.

Instructions for Your Self-Analyzed CCRT

Write a series of narratives, each about an

interaction between you and another person.
Think of events that were meaningful to you,
either in a good or bad way, recently or in the
past. For each event tell when it occurred, whom
it was with, and what happened. For each event
give some of the conversation—what the other
person said and what you said. For writing each
event, 4 or 5 minutes should be enough. The
main other person might be father, mother, other
relatives, friends, people you work with,
anyone. It does not matter what events you
choose. Ten of these narratives about events
ought to be enough. After writing these, score
them by the CCRT procedures. After having
gone through this self-analytic process yourself,
it will be easier to see what is being tapped by
the CCRT.

Format for Self-Analyzed CCRT (underline

thought-units to be scored for wishes, responses
from others, and responses of self)

www.theipi.org 737
www.freepsychotherapybooks.org
 #1. Other person: ______
Relationship Episode:

#2. Other person: ______

Relationship Episode:

#3. Other person: ______

Relationship Episode:

etc.

Instructions and format for self-analyzed CCRT.

www.theipi.org 738
www.freepsychotherapybooks.org
 In answer to a question about when they think of
a patient’s CCRT, of 43 respondents, 34 said
during the session and 31 said while mulling
over a case.

In answer to a question about how the CCRT

compares with other ways to learn how to do
dynamic psychotherapy, of 43 respondents, 22
said it was very helpful and more than most of
the other ways of learning to practice
psychotherapy.

CONCLUSIONS

• The main advantage of CCRT-based clinical

formulations over unguided clinical formulations
comes from their guided, uniform format and,
consequently, the greater agreement obtained
among clinicians. The CCRT method helps with
the usual tasks facing dynamic psychotherapists:
(a) The CCRT is valuable for its assistance in
making formulations about the central
relationship patterns and as the centerpiece for
decision making about the treatment focus. The
therapist should use the CCRT as a basis for
fashioning appropriate interpretations because it
is desirable to have a convergence of the CCRT
with the interpretations derived from it (see

www.theipi.org 739
www.freepsychotherapybooks.org
 chapter 13, this volume), (b) Within the CCRT it
is desirable to concentrate on interpretations that
include the wish and the expected response from
others (see chapter 13). (c) The CCRT is helpful
in the timing of interpretations; it assists the
therapist in becoming aware of the central
relationship pattern so that interpretations can be
made when clinical indications are favorable, (d)
The CCRT is of special help when a prominent
symptom appears because the format of the
CCRT reveals the conflicts that are associated
with that symptom, (e) The patient’s CCRT
should be a routine qualifier to the DSM
diagnosis.

• A comparison was given of the value of relying on

the CCRT for finding the focus in dynamic
therapy versus relying on the interpretative
systems within other psychotherapies. My
conclusion is that reliance on the CCRT results in
a more patient-specific central relationship
pattern formulation for each patient, rather than
the relatively uniform formulations across
patients that are produced by overreliance on a
particular theory of therapy.

www.theipi.org 740
www.freepsychotherapybooks.org
 • The main methods of becoming skillful in the use
of the CCRT during psychotherapy sessions
include (a) the use of the four-therapist peer
supervision training group, (b) graded experience
in the scoring of CCRTs on practice cases, (c)
practice in making CCRT-based interpretations
during playback sessions, (d) a self-analysis by
the therapist of self-reported and self-analyzed
CCRTs, and (e) practical and theoretical
experience with the CCRT during professional
training.

www.theipi.org 741
www.freepsychotherapybooks.org
 20

ALTERNATIVE MEASURES OF THE

CENTRAL RELATIONSHIP PATTERN

LESTER LUBORSKY14

The CCRT measure is the first reliable central

relationship pattern measure when judged from

psychotherapy sessions. Such measures are in a class

that has expanded dramatically in the last dozen

years. In this chapter I describe each of the measures

to help potential users decide which might best meet

their research or clinical needs. The chapter ends with

(a) a review of controlled comparisons among these
observer-judged measures and (b) an evaluation of
the questionnaire methods that also claim to be

measures of transference patterns.

www.theipi.org 742
www.freepsychotherapybooks.org
 Seventeen measures, including the CCRT, now
make up the membership of this class of observer-

judged central relationship pattern measures. These

are named in Table 1, which also portrays their

almost yearly proliferation. Each of measures to be

included in the class must fulfill these criteria:

1. The database for scoring the measure must be a

sample of the person’s relationship interactions
selected from psychotherapy sessions or from
other interviews and based on either (a)
narratives or thought units about the interactions
or (b) actual behavioral samples of the
interactions.

2. The most central pattern, defined as the most

pervasive across relationship interactions, must
be extracted from these relationship interactions.

3. The extraction of this pattern must be derived

partly through clinical judgment and not be
limited to self-report questionnaires.

4. The reliability of the measure must have been

shown or preliminary research on reliability must
be in progress.

www.theipi.org 743
www.freepsychotherapybooks.org
 TABLE 1
Central Relationship Pattern Measures Based on Sessions

Year Researchers Method

1976 Luborsky Core Conflictual
Relationship Theme (CCRT)
1977 Weiss, Sampson, Plan Diagnosis (PD)
Caston, & Silberschatz;
Caston
1979 Benjamin Structural Analysis of Social
Behavior (SASB)
1979 M. Horowitz Configurational Analysis
(CA)
1981 Teller & Dahl Frame method (Frame)
1981 Carlson Tomkins’s Script Theory
1982 Gill & Hoffman Patient’s Experience of
Relationship With Therapist
(PERT)
1982 Schacht & Binder Cyclical Maladaptive Pattern
(CMP)
1984 Grawe & Caspar Plan Analysis (PA)
1985 Kiesler et al. Impact Message Inventory
(IMI)
1986 Bond & Shevrin Clinical Evaluation Team
1986 Maxim Seattle Psychotherapy
Language Analysis Schema
(SPLASH)
1987 Kiesler Psychotherapy and
Interpersonal Transactions
(CLOPT, CLOIT)
1989 Perry, Augusto, & Idiographic Conflict
Cooper Summary (ICS)
1989 L. Horowitz, Rosenberg, Consensual Response

www.theipi.org 744
www.freepsychotherapybooks.org
 Ureno, Kalehzan, & Formulation (CRF)
O’Halloran
1990 Crits-Christoph, Quantitative Analysis of
Demorest, & Connolly Interpersonal Themes
(QUAINT)
1992 Demorest & Alexander Personal Scripts

www.theipi.org 745
www.freepsychotherapybooks.org
 Here we give only a short sketch of each measure,
but we include references so the reader can find out

more about them. For about half of the methods,

longer accounts can be found in a process research

handbook (Dahl, Kächele, & Thomae, 1988), in a

guide to psychodynamic treatment research (Miller,
Luborsky, Barber, & Docherty, 1993), in a volume on
person schema studies (M. Horowitz, 1991), in an

evaluation of some of these measures (Barber &

Crits-Christoph, 1993), and in accounts of

comparisons of seven of these measures applied to a

specimen patient interview (Luborsky, Popp, Barber,

& Shapiro, 1994).

SKETCHES OF THE ALTERNATIVE METHODS

Plan Diagnosis

The Plan Diagnosis (PD) method grew out of a

particular psychoanalytic theory of therapy developed

by Weiss (1986) and empirically tested by Weiss et

www.theipi.org 746
www.freepsychotherapybooks.org
 al. (1986). The method has enabled clinicians to
develop comprehensive and reliable case

formulations that include these four components: the

patient’s goals for therapy, the inner obstructions

(pathogenic beliefs) that prevent or inhibit the patient

from attaining goals, the ways the patient is likely to
test the therapist to disconfirm pathogenic beliefs, and
the insights that will be helpful to the patient. The

method has been applied to the study of

psychoanalysis (Caston, 1977, 1986; Curtis &

Silberschatz, 1989) and a variety of brief

psychotherapies (Curtis & Silberschatz, 1989; Curtis,

Silberschatz, Sampson, Weiss, & Rosenberg, 1988;

Perry, Luborsky, Silberschatz, & Popp, 1989;

Rosenberg et al., 1986). The Plan Diagnosis method

has been studied as a measure of therapist accuracy

(Silberschatz, 1986; Silberschatz, Curtis, Fretter, &
Kelly, 1988; Silberschatz et al. 1986) and of therapy

process and outcome (Nathans, 1988; Norville et al.,

www.theipi.org 747
www.freepsychotherapybooks.org
 1996; Silberschatz, Curtis, & Nathans, 1989). These
studies have demonstrated the value of the Plan

Diagnosis method by showing that accurate

interventions lead to patient progress and to favorable

patient outcome.

In all of these studies, reliabilities (intraclass

correlations) have averaged in the .7 to .9 range for
each of the plan components: goals, obstructions,
tests, and insights (Curtis & Silberschatz, 1989;
Rosenberg et al., 1986). The method has also been

reliably used by investigators outside of the Mount

Zion Psychotherapy Research Group, for example by

Collins & Messer (1988), who came up with
somewhat different findings.

Structural Analysis of Social Behavior (SASB)

The essential Structural Analysis of Social

Behavior (SASB) model was presented by Benjamin

(1974); the first applications to psychotherapy

www.theipi.org 748
www.freepsychotherapybooks.org
 sessions came later (Benjamin, 1979). The model has
been applied to family interactions (Benjamin, 1977)

and to dyadic interactions (Benjamin, 1979), as well

as to interventions in psychotherapy (Benjamin,

1982).

The SASB can be used to trace the sequence of

the patient’s associations during a session (Benjamin,
1986b). It can track moment-to-moment changes in
associations, as well as provide a dynamic
formulation about conflicts. For this purpose sessions

are scored by trained SASB coders after the session

has been divided into codable units that are defined as

single thought units. Such thought units usually
consist of a subject and verb as well as any modifying
clauses. Each unit has only one speaker. The referent
is the “identified other,” usually another person.

Three types of judgments are then made: the focus of

the message, whether the message is friendly or

unfriendly, and the interdependence. The focus,

www.theipi.org 749
www.freepsychotherapybooks.org
 affiliation, and interdependence judgments are

combined to reach the SASB classification.

The SASB method is one of the oldest and

psychometrically most sophisticated methods; much

information is available about its validity and high

levels of reliability (Benjamin, 1994), As an example,

kappas for process codes of family therapy ranged

from .74 to .91 with a mean of .81 (Benjamin,

1986b).

Configurational Analysis

The Configurational Analysis (CA) method (M.

Horowitz, 1979, 1987) appears to estimate some of

the same basic relationship patterns as the CCRT, but

it involves a more encompassing method called the
Role Relationship Models Configuration (RRMC; M.
Horowitz, 1991). For this method, the data from
process notes and transcripts of sessions are

examined from three interrelated points of view:

www.theipi.org 750
www.freepsychotherapybooks.org
 states, relationship patterns, and information (M.

Horowitz, 1987, 1989, 1991). The point of view that

has most in common with the CCRT is the one for the
analysis of relationships, which includes the RRMC
approach. The CA approach offers a conceptual

model for intrapsychic conflict about relationships

and the scripts for interactions between self and other.

The five basic elements are (a) the roles and traits of
self schemas; (b) the schema of the object person; (c)

the aims from each toward the other, often beginning

as the wish for action or expressed emotion from the
self; (d) the response of the other; and (e) the

reactions of self. These have been illustrated and

compared with the CCRT (M. Horowitz, Luborsky, &
Popp, 1991). In the RRMC method, four types of role
relationship models are placed in a configuration
about a specified type of object relationship. Thus,

there are desired, dreaded, compromise-maladaptive,

and compromise-adaptive role relationship models.

www.theipi.org 751
www.freepsychotherapybooks.org
 Evidence has been provided for the satisfactory
reliability of the RRMC in four cases (M. Horowitz

& Eells, 1993) and later, in more detail, for two new
patients by independent configurational analysis

teams (Eells, Horowitz, Singer, Salovey, Daigle, &

Turvey, 1995; M. Horowitz, Eells, Singer, & Salovey,
1995).

Frame

This method is based on identification of

“frames.” A frame is a recurrent, structured sequence

of events that represents a person’s significant wishes

and beliefs (Teller & Dahl, 1981, 1986). The events

may include mental and other behaviors such as

acting, perceiving, believing, knowing, wishing, and

feeling. The most important relationship among the

events is their sequential order, for example,

expresses anger rarr; feels rejected rarr; withdraws.

Dahl (1988) proposed that frames (a) are represented
in the mind in a nonverbal code as described in

www.theipi.org 752
www.freepsychotherapybooks.org
 Bucci’s (1985) dual code system of mental

representations; (b) are structured sequences of

emotions and defenses (Dahl, 1978); (c) are the

residues of early object relations (Gedo, 1979); (d)

endure over time; (e) appear across conflicts, objects,

and situations; (f) can interact with each other; (g)
can account for a wide spectrum of repetitive,
neurotic, maladaptive behavior and, in principle,

normal, adaptive behavior; (h) permit specific

predictions of wishes and beliefs; and (i) provide the

framework for a theory of change that is independent

of any particular theory of how to bring about the
change (Dahl, 1988; Dahl & Teller, 1984, 1993).

Dahl and Teller (1993) described three methods

for identifying frames. In Method A judges use the

patient’s narratives first to construct prototypes and

second to find instantiations (repetitive examples)
both with different objects and in different situations.
Method B uses patients’ own inductive

www.theipi.org 753
www.freepsychotherapybooks.org
 generalizations about their behaviors as prototypes;
judges then search for instantiations as in Method A.

Method C (Leeds & Bucci, 1986) uses an objective

procedure to discover the repetitive sequences of

events. With this method Davies (1989) found frames

in the play of 3-year-olds that were consistent for
each child with two other children and reflected the
child’s interactions with his or her mother. Further

reliability studies are in progress.

Script Theory

Carlson (1981) drew from Tomkins’s (1987)

script theory of personality to identify particular

analyses that constitute a developing relationship

pattern measure. The theory posits that an enduring

set of relationship patterns are repeated throughout a

person’s life (Demos, 1995). Carlson described the

script as “the individual’s rules for predicting,

interpreting, responding to and controlling
experiences governed by a family of related scenes”

www.theipi.org 754
www.freepsychotherapybooks.org
 (1981, p. 502). Tomkins’s (1987) theory also

identifies one “nuclear scene,” and sometimes

several, that manifests these rules; this scene is
interpreted as a pattern-setter for later relationship

episodes. Carlson (1981) gave a cogent example of a

person’s nuclear scene that recurred after 30 years,
and Carlson’s (1986) follow-up provided empirical
study of analogues as reflected in transference

dreams. Reliability information is being developed.

Patient’s Experience of the Relationship with the

Therapist (PERT)

Gill and Hoffman (1982a, 1982b; Hoffman &

Gill, 1988a, 1988b; Gedo, 1993) provided a coding

scheme for studying transcripts of audio-recorded

psychotherapy sessions. The scheme includes codes

for several types of communications regarding what

the authors named the Patient’s Experience of the

Relationship With the Therapist (PERT). At the heart

of the scheme is the coding of disguised allusions to

www.theipi.org 755
www.freepsychotherapybooks.org
 the relationship in associations that are manifestly

about other matters. There is also a code for explicit

references to the relationship and one for readily

observable events in the interaction that are not
spoken about but that may affect the patient’s

experience of the immediate interaction. These

explicit references and unspoken events serve as

bases for the coding of disguised allusions to the

relationship. The system also has a component that

requires a rating of the degree to which the therapist’s

interventions deal with the main aspects of the
patient’s experience of the relationship, both latent
and manifest. The coding scheme emerges from a

conception of the therapeutic process in which the

therapist is viewed as a significant codeterminator of

the transference (Gill, 1982; Hoffman, 1983).

Hoffman and Gill (1988b) recently discussed their

view of the differences between the CCRT and the
PERT. They suggested that the PERT is “more geared

www.theipi.org 756
www.freepsychotherapybooks.org
 toward the tracking, not only of transference themes,

but also of resistance as it affects nuances of

communication during the course of the session’’ (pp.

92-93).

Gill and Hoffman (1982b) first reported some

preliminary reliability data. In a more recent study
employing an adapted version of the scheme
(Gabbard et al., 1988), reliability was demonstrated
for some of the therapist variables.

Cyclical Maladaptive Pattern (CMP)

This method offers guidelines for formulating the

pattern that provides a treatment focus for the

therapist’s interventions; the method was therefore

first called the Dynamic Focus method by Schacht

and Binder (1982) and Schacht et al. (1984). The shift
in label to the Cyclical Maladaptive Pattern (CMP)
was intended to stress the observation that the pattern

shows a self-perpetuating cycle. The system’s

www.theipi.org 757
www.freepsychotherapybooks.org
 components, as illustrated by Henry, Schacht, and

Strupp (1986), include (a) acts of self, (b)

expectations of others, (c) consequent acts of others

toward self, and (d) consequent acts of self toward

self. These components appear to be similar to the

components of the CCRT. Acts of self, for example,
include the wishes. The expectations of others and
consequent acts of others toward self are both

included in the responses from others in the CCRT

system. Consequent acts of self toward self are

similar to the responses of self in the CCRT system.

The CMP continues to be used in its original

form. But to increase reliability and theoretical

coherence, however, the CMP has another form that

includes the measurement methods of the Structural

Analysis of Social Behavior, which is called the

SASB-CMP. Consequently, the reliability should be
the same as that achieved by the SASB. This new-
generation system reorganizes the information into

www.theipi.org 758
www.freepsychotherapybooks.org
 three categories—(a) interpersonal acts, (b)
introjective acts, and (c) expectancies—using

procedures described by Schacht et al. (1984).

Plan Analysis (PA)

The Plan Analysis method is based on observable,

often verbal behavior as well as nonverbal behavior

(Grawe & Caspar, 1984). In its concern for nonverbal

behavior it differs from the other measures, including

the CCRT, which are usually based on verbal
behavior. The Plan Analysis method emphasizes the

interactional plans that are in conflict with each other.

It also includes intrapsychic elements of a client’s

functioning from an instrumental point of view. The

instrumental function of action is considered in terms

of these two questions: What is the behavior for?

Which means are used for a particular purpose? In

essence, the sources of information for the analysis

are (a) the behavior of the person, especially the
nonverbal behavior and interactions; (b) the emotions

www.theipi.org 759
www.freepsychotherapybooks.org
 and action tendencies that the patient triggers in other

persons; and (c) the behavior and the emotions

considered from a reactive perspective, that is,
negative emotions that arise when important plans are

threatened and positive emotions that arise when

plans are favored. Therapies based on Plan Analysis
have especially helpful therapeutic relationships
owing to the individualized interpretations made by

the therapist and richness of technical procedure

(Grawe, Caspar, & Ambühl, 1990).

Promising reliability studies have been done,

mainly in a descriptive qualitative mode (Caspar,

1989). Two studies have recorded agreement between

plan analyses. In each, videotapes were used with a

single patient with different judges judging the tapes

(Theus, 1987, as reported by Caspar, 1995). In both,

some degree of satisfactory case conceptualization
was shown.

Impact Message Inventory, Form IIA (IMI)

www.theipi.org 760
www.freepsychotherapybooks.org
 This measure, developed by Kiesler et al. (1985),
yields characterizations of the interpersonal behaviors

of interactants, including patients and therapists,

empirically derived from Lorr and McNair’s (1965)

version of the Interpersonal Circle (which provides

15 categories of behavior that overlap substantially
with the 1982 Interpersonal Circle’s 16 categories).
Scores are obtained from an interactant’s (B) report

of the feelings, action tendencies, and cognitive

attributions evoked in him or her during interactions

with another person (A); resultant scores characterize
the interpersonal behavior pattern of Person A. In the

psychotherapy context, Impact Message Inventories

(IMIs) filled out by therapists or observers (Bs) on

patients (As) use reports of objective

“countertransference” by Bs to characterize the

transference patterns of As. In particular, when the

measure is applied by clinical judges who are the

www.theipi.org 761
www.freepsychotherapybooks.org
 observers (Bs), the measure qualifies for membership

in the class of central relationship pattern measures.

The widening sphere of utilization of narratives as

a basic unit in psychotherapy sessions is illustrated by

McMullen and Conway (1997). They scored

relationship episodes for 20 cases of short-term
dynamically oriented psychotherapy involving the
self and others for 2 early, 2 middle, and 2 late
sessions for 20 cases. These narratives were coded on

Kiesler’s (1983) version of the Interpersonal Circle.

The most successful cases were those that showed
increasing friendly—submissive portions of the

circumplex in their portrayal of self. In contrast, the

least successful cases were in the hostile half of the

circumplex.

Internal consistency reliabilities for the 15 scales

are high, with coefficients tending to be .80 or higher.

Additional information on reliability and results of

www.theipi.org 762
www.freepsychotherapybooks.org
 more than 40 empirical studies can be found in

Kiesler’s research manual (1987b). A more recent

light-scale circumplex version (Kiesler & Schmidt,

1993) has excellent psychometric properties and is

routinely recommended to researchers and clinicians.

Clinical Evaluation Team

In this system (Bond & Shevrin, 1986b), the

database includes transcripts of diagnostic interviews
and psychological tests. In making the relationship
pattern formulation, the clinical judge is not tied to a

confined system. Indeed, clinicians have the wide

latitude usually involved when they are asked to

produce a relatively free-form diagnostic formulation.

The only constraints placed on clinicians are that they

are asked to describe (a) the patient’s conscious

experience and understanding of his or her presenting

symptom (usually a social phobia) and (b) their
inferences as to the unconscious conflict that might
underlie and cause the symptom. Although most

www.theipi.org 763
www.freepsychotherapybooks.org
 systems rely on frequency, the Clinical Evaluation
Team method allows much more freedom for

weighting information according to clinically inferred

salience, not only in terms of frequency but also in

terms of what seems omitted, what seems especially

significant because of vivid associations, and other
signs. Each clinical judge has an opportunity to
interact with the group of other judges to arrive at a

consensus formulation. Data on reliability are not yet

available.

The Seattle Psychotherapy Language Analysis Schema

This method, called the SPLASH (Maxim, Straus,

& Rosenfarb, 1986; Maxim, 1986), analyzes verbal

texts in terms of short units. Each unit of the patient

and the therapist is coded by speaker and text line

number. The interpersonal message between speaker

and listener and the frame of reference (that is, the

view that Person X has of Person Y and a particular
frame of reference for Person X) are also coded. In

www.theipi.org 764
www.freepsychotherapybooks.org
 each unit, five main variables are coded: (a) object,

viewpoint, and frame of reference under discussion;

(b) affects; (c) impulses; (d) coping strategies; and (e)
interpersonal message. It takes about 40 hours of

practice to achieve about 90% accuracy. It takes

about 35 hours to score a 30-page session, which is a
little over an hour per page. Clearly this is a system
for research and not easily adapted to everyday

clinical use. The method offers a description of the

interaction, but its main purpose is to illuminate

change across different therapeutic situations. A

companion coding scheme, Metacommunication of
Interactive Sequences in Therapy (Maxim & Sprague,
1989), analyzes the knowledge that is
metacommunicated by the spoken utterance. A
therapy session is divided into discussion topics of its

manifest content. A maladaptive patient belief is

identified for each discussion topic and represents a

specific version of the session’s CCRT. Knowledge

www.theipi.org 765
www.freepsychotherapybooks.org
 structure categories of plans, strategies, and

interactive processes are coded for patient and

therapist as they are used to address the patient’s

belief. This additional coding takes 20 more minutes
per page.

Coders-in-training have to achieve a minimum

kappa of .7 on each category code before they start

coding text for research. After coding one session,

each coder is given a separate five-page section of

text to code to test for drift. One or two training

sessions are sufficient to correct for drift in a

particular category type.

The Check List of Psychotherapy Transactions-

Revised (CLOPT-R) and the Check List of
Interpersonal Transactions-Revised (CLOIT-R)

These measures by Kiesler (1987a), called the

CLOPT-R and the CLOIT-R, are derived from the
latest version of the Interpersonal Circle (Kiesler,
1983). They are used as a self-report for interactants’

www.theipi.org 766
www.freepsychotherapybooks.org
 (patients’ and therapists’) ratings of each other and
for observer ratings of psychotherapy sessions. These

observer ratings qualify the method as a central

relationship pattern measure. It yields a profile of

interpersonal behavior as represented by the 16

categories of the Interpersonal Circle, as well as

indexes of the degree of complementarity present in

patient-therapist dyads. Important clinical

applications have been made to analysis of

prototypical interpersonal behaviors of DSM-III-R

personality disorders and to complementarity and
therapeutic alliance in outpatient therapy dyads.

Although reliabilities vary for the different

versions of the measure, internal consistency
reliabilities for the 16 scales are moderately high,

tending to range from .50 to .80 with a median in the

mid-.60s. Additional information on reliability and

validity can be found in the work of Kiesler,

Goldston, and Schmidt (1991), who also

www.theipi.org 767
www.freepsychotherapybooks.org
 recommended that researchers and clinicians

routinely score a psychometrically improved light-

scale version of this circumplex measure.

Idiographic Conflict Formulation (ICF)

This method (Perry, 1994; Perry & Cooper, 1989)

includes the assessment of four components (along

with a statement of the evidence for the assessment):

conscious and unconscious wishes and fears,

symptomatic and avoidant outcomes resulting from

conflicting wishes and fears, specific stressors to

which the patient is vulnerable, and the patient’s best

level of adaptation to the conflict. A standardized list
has been made of 40 wishes and 39 fears that can be
used for scoring the first two components of the
Idiographic Conflict Formulation (ICF) method.

These wishes and fears are arranged according to the

Eriksonian hierarchy of psychological development.

Because each motive is placed within one of the eight

developmental stages (e.g., Stage 1, trust versus

www.theipi.org 768
www.freepsychotherapybooks.org
 mistrust), an overall developmental score can be

obtained by weighting each wish or fear by its stage,

then taking an overall weighted average. This
calculation yields one number, which represents the
mean developmental stage for the person’s wishes (or

fears). Comparing assessments across several points

in time then allows the detection of change in an

individual’s developmental level with treatment or

with time.

The method offers reliability evidence based on

two independent formulations of 20 cases, using

paired comparisons for the similarity of correctly

matched versus mismatched pairs of formulations.

The mean similarity of correctly matched pairs,

assessed by a 7-point scale, was 4.41, significantly

higher (p < .001) than mismatched formulation pairs

either with the same diagnoses (3.05) or with
different diagnoses (2.91).

Consensual Response Formulation

www.theipi.org 769
www.freepsychotherapybooks.org
 In this method (L. Horowitz et al., 1989), a
videotape of an evaluation interview is presented to a
group of clinicians, each of whom writes a dynamic
formulation. Then the formulations are divided into
thought units. The most frequent thought units across

clinicians are collected into a composite formulation

called a Consensual Response Formulation (CRF).

The focus of the method, therefore, is on the

clinicians’ consensual observations and inferences.

One validity study found that formulations with a

higher proportion of interpersonal content were
associated with greater improvement (L. Horowitz et
al., 1989). In another validity study, naive clinicians,

reading only the consensual formulations, were able

to anticipate correctly the interpersonal problems that

were discussed in the treatment, achieving a mean

chi-square of 22.2 (p < .001). These results confirmed

the earlier finding (L. Horowitz et al., 1988) that

www.theipi.org 770
www.freepsychotherapybooks.org
 patients with primarily interpersonal problems are
especially suitable for brief dynamic psychotherapy.

The replicability of the method was established

by having another group of clinicians repeat the

entire Consensual Response Formulation procedure;

corresponding formulations had an 80% overlap in

content. In addition, 100% of the judges were able to

match the replicated formulation correctly to the

original formulation of the same case.

Quantitative Analysis of Interpersonal Themes

(QUAINT)

The Quantitative Analysis of Interpersonal

Themes (QUAINT) method employs the CCRT
structure of wishes, responses from other, and
responses of self. However, the QUAINT method

differs from the CCRT in that (a) it uses a broad

vocabulary of reliable categories derived from the

Structural Analysis of Social Behavior (Benjamin,

1974, 1986a, 1986b), which covers interpersonal

www.theipi.org 771
www.freepsychotherapybooks.org
 behaviors defined across dimensions of affiliation,

interdependence, and activity-passivity; (b) the

method assesses the patient’s narratives separately

and in random order, rather than in the context of

other narratives; and (c) the method uses a cluster

analysis approach to determine the coherent, multiple
themes apparent across each patient’s narratives.

A precursor to the method is presented by Crits-

Christoph, Demorest, and Connolly (1990), and the

fully developed method, with associated reliabilities,

is presented by Crits-Christoph, Demorest, Muenz,
and Baranackie (1994). The QUAINT has been used

to examine the degree of consistency in themes

across narratives (Crits-Christoph et al., 1994) and to

explore the nature of themes with therapist versus

themes with other people (Connolly et al., 1996).

Personal Scripts

www.theipi.org 772
www.freepsychotherapybooks.org
 A clinical-quantitative script method derived from
Tomkins’s script theory is reflected in the work of

Demorest and Alexander (1992), who outlined a

method for deriving personal scripts from narrative

reports. Emotional experiences or scenes are first

identified, and then scripts are extracted from them by
a two-step process of abstraction and sequencing.
Abstraction involves translating the literal elements

of a specific scene into the abstract form of a generic

script (e.g., “father” becomes “intimate male

authority”); sequencing involves deriving the order of

events that the script seeks to predict (e.g., the self

approaches with interest &rarr; the other rejects with

disgust &rarr; the self withdraws with sadness).

Using this method, they found that people display the

same scripts within two different types of narrative

imagery: autobiographical reports and projective test
responses. Reliability has been established for both

the abstraction and sequencing tasks of script

www.theipi.org 773
www.freepsychotherapybooks.org
 translation (Demorest & Alexander, 1992; Demorest
& Siegel, 1996).

COMPARISONS AMONG METHODS OF

MEASURING CENTRAL RELATIONSHIP
PATTERNS

Because most central relationship pattern

measures have come on the scene in the last 2

decades, the measures differ widely in the quality of

their reliability and validity data. The oldest, the
CCRT, has been one of the pacesetters; it is among
the most advanced psychometrically in terms of

information about reliability and validity. The

Structural Analysis of Social Behavior (Benjamin,
1979) is also one of the most advanced, for it
represents years of ingenious research development.

Significant advances also have been shown by the

PD, Frame, CLOPT-R and CLOIT-R, and QUAINT

methods.

www.theipi.org 774
www.freepsychotherapybooks.org
 Most of these diverse methods appear to have

commonalities in their basic categories. One of these

is in the broad duality of impulse versus executive

functions, for example, on one side, wishes, needs,
and goals, and on the other side, responses from

others and responses of self.

Quantitative research on the commonalities

among central relationship pattern measures has

begun, with studies comparing one measure with

another measure, usually on a single case, by the

paired-comparisons method. These studies are by

Luborsky (1988a; CCRT, PERT, Frame); M.

Horowitz et al. (1991; RRMC and CCRT); Johnson,

Popp, Schacht, Mellon, and Strupp (in press; CCRT

and CMP); Kächele, Luborsky, & Thomae (1988;

CCRT and PERT); and Perry, Luborsky, et al. (1989;

CCRT, ICF, and PD). The consistent finding from

these comparisons of measures is that there are
significant similarities among the measures. A larger

www.theipi.org 775
www.freepsychotherapybooks.org
 scale study (Mackenzie, 1989) compared 12 patients

on four measures: the Inventory of Interpersonal

Problems (L. Horowitz, Weckler, & Doren, 1983), the

Relationship Anecdotes Paradigms interview (see
chapter 7, this volume), the Structural Analysis of

Social Behavior, and the Repertory Grid (Kelly,

1955). The results from all four methods were

translated into uniform terms on the basis of the

SASB. Each method was shown to have uniqueness

as well as areas of overlap.

A more recent and more complete set of

illustrations and comparisons of most of these

measures applied to the same patient interview has

been reported (Luborsky, Popp, Luborsky, & Mark,

1994); it includes the CCRT, Configurational

Analysis, the Plan Formulation method, the SASB-

CMP, the Consensual Response Formulation (CRF),

Idiographic Conflict Formulation method, and
Frames. All of these central relationship pattern

www.theipi.org 776
www.freepsychotherapybooks.org
 measures were compared with each other by a paired-
comparisons method by Luborsky, Popp, and Barber

(1994), who showed the methods to be moderately

similar; the most similar of all were the CRPF, the

SASB-CMP, the CCRT, and the Frame methods.

The decision to use a particular measure may

depend on a practical matter—the time it takes to use

the measure. In terms of time for scoring, the

methods generally are expensive. The CCRT based

on relationship episodes drawn from psychotherapy

sessions or on the RAP interviews takes about 1.5–3

hours per session to score properly, using the

combined tailor-made plus standard categories
procedure described in chapter 2, this volume. New
scoring systems have been developed for the CCRT

that cut down on the time required, such as a decrease

in the number of forms needed, but the time estimates

for the CCRT still place it among the least time

consuming of the observer-judged central relationship

www.theipi.org 777
www.freepsychotherapybooks.org
 pattern measures. Of course, the CCRT can be

applied by the therapist in everyday practice in the

course of a session with no extra time taken (using
procedures described in chapter 19).

QUESTIONNAIRE MEASURES OF
TRANSFERENCE PATTERNS

The questionnaire methods in this review were

intended to be measures of the transference pattern.

To achieve an operational version of the concept of

transference, Chance (1952) developed a
questionnaire measure of the similarity between the
patient’s description of a significant parent and the

patient’s description of the therapist. In Fiedler’s

studies (e.g., Fiedler & Senior, 1952), transference

was defined in terms of a comparison of the patient’s

description of the ideal person with the patient’s

prediction of the therapist’s self-description and by
both of these with similar measures completed by the
therapist.

www.theipi.org 778
www.freepsychotherapybooks.org
 In Apfelbaum’s (1958) method, transference was
intended to be tapped by a Q-sort questionnaire on
the patient’s expectations about the qualities of the
therapist who would later be assigned to the patient.
The patients were grouped in terms of three types of

expectations reported in their preassignment Q sort:

Cluster A (therapist will give nurturance), Cluster B

(therapist will be a model), and Cluster C (therapist

will be a critic). Each of these expectations tended to

be maintained until the end of treatment, as indicated

by high test-retest reliability. Such stability was

considered by Freud (1912/1958a) to be a
characteristic of transference.

Rawn (1958, 1981) developed Q-sort-based

scales and applied them to four sessions of one
patient’s analysis. He took the unusual further step of

comparing these results with those from clinical

observations and noted signs of convergence. Crisp
(1964a, 1964b, 1966) continued this line of research

www.theipi.org 779
www.freepsychotherapybooks.org
 with Q sorts of questionnaire items. The items rated
were about father and therapist figures; the estimate
of transference was based on a comparison of these
ratings. One finding indicated that attitudes toward
the therapist, as measured by the questionnaires,

tended to change with or to precede changes in

symptoms. Subotnick (1966a, 1966b) developed the

method further using two separate sets of Q sorts:

attitudes toward parents and attitudes toward the

therapist at various points in therapy. Similarity was

found between the attitudes toward parents and
therapist (there were high loadings on the factors
common among the Q sorts).

These six sets of studies used the questionnaire

approach, usually in the form of the Q-sort method;
they are a good sample of this type of study. Studies

based on the questionnaire approach, however, suffer

from questionable validity: The questionnaires may

not measure the same construct measured by the

www.theipi.org 780
www.freepsychotherapybooks.org
 clinically inferred transference pattern. It is not clear

that a person’s responses on a questionnaire reveal

the nature of the transference pattern in the way that a
clinical judge would assess it from sessions or other
interview data. Consequently, this oldest line of

quantitative transference research has not gained

much acceptance. Measures derived from

questionnaires need to be compared with measures of

transference based on psychotherapy sessions. The

research reported in chapter 15, this volume, has

suggested that there is some degree of association of

the two. In the event that the two approaches turn out

to agree substantially, my colleagues and I will have

to acknowledge that the advent of objective
transference measures was much earlier than was
recognized in the review of the field by Luborsky and
Spence (1978). In fact, Barber (1993) is constructing

a questionnaire measure of the central relationship

www.theipi.org 781
www.freepsychotherapybooks.org
 pattern that holds promise for bridging the gap

between the early and present research.

SUMMARY, DISCUSSION, AND CONCLUSIONS

• This chapter summarized the qualities of each of 16

alternative measures that appeared after the
CCRT method was fashioned (Luborsky, 1976,
1977b) and that fit the class of central
relationship pattern measures.

• Alternative measures have appeared steadily since

1976, with a new one every few years since then.
However, the steady progression may have come
to a stop in 1992 with the work of Demorest and
Alexander; whether this is a stop or just a long
pause remains to be seen.

• Many of the measures have common elements with

the CCRT. Clearly, when the CCRT method is
being considered, a researcher has two options:
to use it or to reject it and try to devise a better
measure. Most researchers have used an existing
measure, but an impressive number of
researchers have decided to develop their own
measure.

www.theipi.org 782
www.freepsychotherapybooks.org
 • The few comparisons of the different measures
with each other have suggested a moderate
overlap among them and implied that they
indeed are alternative measures.

• Although most of these alternative measures

require more psychometric development, each of
them is thought to have special virtues.

• One of the virtues claimed for more and more of

these measures in the course of their
development is that they can help the therapist
during the therapy to make an accurate
transference formulation and, on well-timed
occasions in the therapy, they are of pivotal help
in guiding the therapist’s interventions. The
measures that claim these benefits include the
CCRT, Patient’s Experience of the Relationship
With the Therapist, Plan Diagnosis, Structural
Analysis of Social Behavior, Configurational
Analysis, Cyclical Maladaptive Pattern, and
Quantitative Analysis of Interpersonal Themes.

• Another genre of alternative approaches to

measuring central relationship patterns is that of
questionnaires to measure the transference
pattern. Work in this genre appears to have

www.theipi.org 783
www.freepsychotherapybooks.org
 started in about 1952 and continued until now,
with several promising findings uncovered.
Although questionnaires have the appearance of
simplicity, what they measure needs to be shown
to be similar to what is measured by the session-
based clinical judgment method.

Note

[14] This chapter is a revised version of an article by L.

Luborsky, P. Crits-Christoph, and J. Mellon (1986) from
the Journal of Consulting and Clinical Psychology, 54, pp.
39-47. Reprinted with permission.

www.theipi.org 784
www.freepsychotherapybooks.org
 IV

WHAT’S NOW AND

WHAT’S NEXT

www.theipi.org 785
www.freepsychotherapybooks.org
 21

THE CONVERGENCE OF FREUD’S

OBSERVATIONS ABOUT
TRANSFERENCE WITH THE CCRT
EVIDENCE

LESTER LUBORSKY

Freud’s many observations about transference

have never before been brought together in one place.

In this chapter, 23 observations—apparently all that

he made—are assembled. Most of these observations

are given in his 1912 paper (1958a), a few in his

earlier work in the postscript to the Dora case (1901-

1905/1953a), some even before that in his “Studies in

Hysteria” (1895/1955b), and some spread over his

works in 1915, 1917, and 1937. From the wide range
of these dates it is obvious that Freud had a career-
long wish to solve the puzzles inherent in his

www.theipi.org 786
www.freepsychotherapybooks.org
 transference concept and to explore its clinical
applications.

One by one Freud’s observations are compared in

this chapter with evidence from the Core Conflictual

Relationship Theme method. Most of that evidence

has been assembled from the University of
Pennsylvania Center for Psychotherapy Research;
some of it has drawn on collaborations with
researchers at the University of California at San
Francisco’s MacArthur Foundation program for the

study of conscious and unconscious mental

processes; and some of it has come from longer

collaborations with faculty at the University of Ulm
(Germany), Department of Psychotherapy.

This chapter is the latest in the succession of ever

more complete reports on the convergence of Freud’s

observations with CCRT evidence; the first was by

Luborsky (1977b); the next was by Luborsky et al.

www.theipi.org 787
www.freepsychotherapybooks.org
 (1985); and the next three were by Luborsky et al.
(1986); Luborsky and Crits-Christoph (1990); and

Luborsky, Crits-Christoph, Friedman, Mark, and

Schaffler (1991). The achievement of the present

chapter is that it surveys all findings that fit the topic

of Freud’s observations about transference and, most
comprehensively of all of the chapters, helps in
understanding transference as illuminated by findings

from the CCRT method.

Freud’s concept of transference became more

differentiated over time. His early use of the term

transference, as in his Dora case (1901-1905/1953a),

concentrated on what was directly implied by the idea
of transference: the transfer of attitudes and behaviors
derived from early parental relationships to the
current one with the therapist. In his Dora case report

Freud gave an example of a formulation of the

transference pattern derived from his famous

psychotherapy with Dora, an 18-year-old woman he

www.theipi.org 788
www.freepsychotherapybooks.org
 diagnosed as having hysterical symptoms. Freud
believed that her transference pattern was based on

her relationship with her father and with her father’s

close “friends,” Frau K and Herr K—Frau K and her

father were lovers. Dora’s central relationship pattern

(in simplified clinically derived CCRT form) was “I
wish for love from my father, but I see that father’s
love is for Frau K rather than for me” (meaning that

his response is that he rejects me for Frau K and he

also rejects me by throwing me to Herr K). Dora’s

responses of self to this negative response from father

included her hysterical symptoms and an inclination

to take revenge on men by cutting herself off from

them. Only belatedly did Freud recognize her

inclination; it came at the end of the psychotherapy

when the patient abruptly broke off the treatment with

him. Because this case was so instructive to Freud
about the nature of transference, we draw on it to

explain further his concept.

www.theipi.org 789
www.freepsychotherapybooks.org
 By the time of his 1912 paper, Freud had
observed a wide range of characteristics of what he

called then the transference “template”; this word is

more easily understood as a translation of the German

word than as “stereotype plate,” rendered by Strachey

in his translation of Freud (1912/1958a). In that
article Freud surveyed his many observations about
this template much as a naturalist would in describing

a natural phenomenon. One of his main conclusions

was that the transference template is a central

relationship pattern that serves as a prototype, or a

schema for guiding, shaping, and conducting

subsequent relationships.

FREUD’S OBSERVATIONS COMPARED WITH

CCRT EVIDENCE

This chapter reviews the whole range of Freud’s

observations about the transference template—

numbered 1 through 23 in Table 1—mostly from his

1912 article but also from his other articles on

www.theipi.org 790
www.freepsychotherapybooks.org
 TABLE 1
Freud’s “Transference Template” Observations and the CCRT
Evidence

CCRT
Freud’s Observation
Evidence
1. Wishes toward people are prominent +
2. Wishes conflict with responses from other +
and of self
3. Especially evident in erotic relationships +?
4. Partly out of awareness +?
5. Originates in early parental relationships +
6. Comes to involve the therapist +
7. May be activated by the therapist’s R
perceived characteristics
8. May distort perception R
9. Consists of one main pervasive pattern +?
10. Subpatterns appear for family members +?
11. Distinctive for each person +?
12. Remains consistent over time +
13. Changes slightly over time +
14. Shows short-term fluctuations in activation R
15. Accurate interpretation changes expression +
of pattern
16. Level of insight is associated with change in +?
the pattern
17. Can serve as resistance R
18. Symptoms may emerge during its activation +?
19. Is expressed in and out of therapy +
20. Positive vs. negative patterns are +
distinguishable

www.theipi.org 791
www.freepsychotherapybooks.org
21. Is similar in different modes (dreams and +
narratives)
22. Improvement means greater mastery of the +
pattern
23. Innate disposition plays a part R

Note. + = study with positive results; +? = preliminary study with

positive results; R = remains to be studied.

www.theipi.org 792
www.freepsychotherapybooks.org
 transference. Under each observation, two pieces of
information are given: (a) the essence of the

observation and (b) the CCRT evidence that is

consistent or inconsistent with it.

1. The “instincts," “aims," and “impulses” that the

person wishes to satisfy in relation to other
people are prominent in the pattern. This
observation is about the nature of a main drive
component in Freud’s transference template.
Along with the preceding terms, Freud used the
term “libidinal cathexes,” which are to be
satisfied in relations with other people during the
“conduct of the erotic life” (1901-1905/1953a, p.
116).

Although Freud’s terms are not specifically

defined in his articles, his uses of them implies that

the CCRT’s “wishes, needs, and intentions” are

concrete versions of his terms. It is consistent with
his observation that clinical judges who apply the
wish categories to the narratives find these categories

to be very prevalent; for example, as reported in

www.theipi.org 793
www.freepsychotherapybooks.org
 chapter 10, this volume, the main wishes are

pervasive in more than 60% of the narratives both

early and late in psychotherapy. This pervasive

category of wishes is also reliably judgeable (see
chapter 6). Our conclusion is that if wishes, needs,

and intentions are similar to the terms Freud used, we

have found considerable evidence for this parallel.

2. Wishes to others conflict with responses from

others and responses of self. The arousal of a
wish tends to become part of a conflict in which
the responses from others and responses of the
self become active. The three components of
Freud’s transference concept clearly emerge from
a review of the examples of transference that he
provided. In one of his early examples
(1895/1955b), a woman’s transference is
described as including a thought about her wish
that Freud give her a kiss. This thought is
followed by her responses of self of extreme
anxiety, sleeplessness, and inability to work. In a
later example from the Dora case (1901-
1905/1953a), the transference pattern began with
a wish for love from her father and from Herr K,

www.theipi.org 794
www.freepsychotherapybooks.org
 followed by her response of feeling rejected
because father’s love was not for her but for Frau
K, followed by responses from herself of feeling
rejected and then rejecting men and experiencing
dissociative symptoms.

The CCRT’s conceptual categories are congruent

with Freud’s basic categories for what he referred to
as the mental apparatus. The CCRT is based on a

similar conflictual dichotomy: wishes (wishes, needs,

intentions), which conflict with responses (responses

from others and of self). The counterpart to our

categories might be considered to be Freud’s

conflictual dichotomy: id impulses (wishes, drives,

instincts), which conflict with ego responses (the
executive functions of defense and action). It is
difficult to construct a precise test of the degree of

this congruence between Freud’s dichotomy and that

of the CCRT. It is easier to demonstrate that the

CCRT scoring system components of wishes and

responses typically have a high frequency of

www.theipi.org 795
www.freepsychotherapybooks.org
 association and that the association is typically
conflictual. The conflicts among the wishes and

responses were especially evident in the special

analysis, in chapter 8, this volume, of the sequences

of wishes and responses for the wish to be close

versus the wish to be independent.

3. The central relationship pattern is especially

evident in erotic relationships. Freud
(1912/1958a) stated that the pattern applies to the
“conduct of ... erotic life” (p. 99). As an obvious
example, this observation fits his account of the
relationship pattern in the Dora case (1901-
1905/1953a).

No systematic study of this distinction between

erotic and nonerotic relationships has yet been done.

In the few patients we have examined, it is clear that
a CCRT is found in both erotic and nonerotic
relationships, but there appears to be a greater

concentration of pervasive components of the CCRT

within the erotic relationships, for example, within

www.theipi.org 796
www.freepsychotherapybooks.org
 the Dora case and within the Mr. Howard case (see

chapter 5, this volume).

4. The central relationship pattern is partly out of

awareness. The transference concept is used by
Freud (1912/1958a) in two senses that imply
states of reduced awareness: (a) as a template
consisting of largely unconscious memory
systems of past relationships and (b) as the
activations of these memory systems in the
experience of the relationship with the therapist,
implying that before the activation they had been
in reduced awareness. In both instances a large
portion of the pattern is considered to be out of
awareness. In the second instance, for example in
the Dora case, Freud (1901-1905/ 1953a)
referred to the expected activation of the
transference in the treatment in blaming himself
for not having dealt with the patient’s reduced
awareness of her pattern: “I neglected the
precaution of looking out for the first signs of
transference” (p. 118).

This observation appears to fit with the CCRT

data, but more systematic methods are needed to

www.theipi.org 797
www.freepsychotherapybooks.org
 define degrees of lack of awareness. Some beginning
methods are already launched. In one method (see

chapter 15), there is a comparison of judgments that

are based on the CCRT with those based on the

patient’s self-interpretation of narratives; it reveals

that the clinician using the CCRT has a more
differentiated focus than the patient about the central
relationship pattern. Another method (Luborsky &

Popp, 1989) for recognizing unconscious conflicts

within the CCRT also shows promise: the provision

of a set of inference cues to point to ideas that are out

of awareness.

5. The central relationship pattern originates in the

early relationships with parental figures. This
observation, repeatedly presented throughout
Freud’s writings, implies that there should be a
parallel between the early relationship patterns
with the parents and the current ones with other
people. The parallel can also be between a later
representation of the earlier parental figures

www.theipi.org 798
www.freepsychotherapybooks.org
 (such as Herr K in the Dora case) and the
therapist.

There is some evidence for this observation

(Luborsky et al., 1985) based on a comparison of

CCRTs from relationship episodes involving

memories of events about the early parental figures

with CCRTs from relationship episodes about other

people in the present. The similarity of the patterns

across these two lifetime eras was evident (mean

similarity rating on a 7-point scale was 6.4); as would
be expected, the similarity for purposely mismatched
CCRTs was much lower (mean similarity was 3.6).

The degree of similarity of the early and late

relationship episodes does not prove causality but is
consistent with the supposition that the later pattern
was prefigured by and may have originated in the

earlier one.

The current CCRT pattern, when traced back in

time, may be found to show concrete

www.theipi.org 799
www.freepsychotherapybooks.org
 correspondences with early traumatic scenes in terms

of the recurrence of theme components as well as of

fragments of the traumatic scenes (Carlson, 1981;

Reiser, 1984). Further evidence of a parallel between
the very early relationship patterns with the parents

and the current relationships is seen in the findings of

Main and Goldwyn (1984). They used the Adult

Attachment Interview to measure adults’

recollections of their own childhood attachment to

their mother. They found a parallel between a

mother’s early childhood relationships and her
current relationships with her own children: Mothers
who distorted their recollections of their own

childhood relationship with their parents had

attachments to their own children with basic

similarities to the ones they experienced with their

own parents; in contrast, mothers who realistically

recalled poor relationships and who were able to
forgive their mothers had secure attachments with

www.theipi.org 800
www.freepsychotherapybooks.org
 their own children. As a whole, however, research on

the influence of very early relationship events on later

personality development reveals a deficiency of

evidence, according to a review by Kagan (1996).

6. The central relationship pattern affects the

relationship with the therapist. It is basic to
Freud’s (1912/1958a) concept of transference
that with the start of therapy the therapist
becomes “attached to” one of the templates, and
the patients’ relationship pattern comes to
involve the patient’s “perceptions” of the
therapist.

An operational CCRT translation of Freud’s

observation about transference requires that the main

pattern in relation to other people also be found

within the relationship episodes about the therapist.

The study of this clinically expected convergence of
the pattern with other people and with the therapist
reported in chapter 11, this volume, yielded the first

www.theipi.org 801
www.freepsychotherapybooks.org
 systematic confirmation of this observation based on

psychotherapy sessions.

7. The central relationship pattern can be activated

by similarities the patient perceives in the current
relationship in the therapy. This observation
deals with two conditions for the activation of
the patient’s pattern. In Condition 1, some
aspects of the transference are just a substitution,
a “new impression or reprint,” as Freud (1901-
1905/1953a) called it. Such “reprints” may be
merely the distortion of the perception of the
current relationship, which is experienced as a
replica of the past relationship. In Condition 2,
other activations of the template may be aroused
and “may even become conscious” on the basis
of finding a similarity in the therapist’s “person
or circumstances” (p. 116). He added to this
another basis for such therapist-stimulated
arousal of transference, “some detail in our
relations” (p. 118). It may be a detail, but its
meaning to the patient is not unimportant.

No studies with the CCRT or with any of the

alternative methods discussed in chapter 20 have

www.theipi.org 802
www.freepsychotherapybooks.org
 focused exactly on these two conditions for the
arousal. However, the most pertinent collection of

data derives from a version of Condition 2: instances

of “negative fit” (Singer & Luborsky, 1977), in which

the patient’s behavior actually stimulates the

behavior of the therapist in ways that fit the patient’s
negative transference expectations. These are
common patient-stimulated countertransference

experiences, as described by Hoffman and Gill

(1988a). In the Dora case the correspondences of the

patient’s transference template and the therapist’s

behavior were thought by Freud to have been

coincidental (Condition 1) rather than based on

behaviors of the therapist that were elicited by the

patient’s transference.

8. The central relationship pattern may distort

perception. It is clear that Freud (1895/1955b)
thought of the activation of transference as
setting off a distortion in perception. Other
observations have the same implication:

www.theipi.org 803
www.freepsychotherapybooks.org
 Observation 6, that the process involves the
therapist; Observation 7, that the therapist’s
characteristics may activate it; and even
Observation 15, that interpretations of the pattern
can benefit the patient by correcting the
distortion. Observation 7 states that the distortion
involves perceiving the therapist as having an
attribute that the therapist does not have or
magnifying an attribute that the therapist does
have. In both cases the distortion also involves
identifying the therapist as similar to an earlier
person to whom the patient is attached (Freud,
1895/ 1955b or 1905/1960).

It is not easy to construct an operational measure

of the distortion of perception in relation to the

activation of transference. One possible measure of

distortion might be the degree of parallel between the

patient’s perception of the therapist and the patient’s

perception of other people, although as stated in
chapter 11, this volume, not all such parallels can be
justified as distortion.

www.theipi.org 804
www.freepsychotherapybooks.org
 9. The concept of the transference template
emphasizes that there is one main relationship
pattern. Freud (1912/1958a) said there was only
one main pattern. However, he must not have
been completely sure because he added, “or
several such.”

We reported evidence in chapter 10 for the

existence of a highly pervasive central relationship

pattern across each session especially in terms of the

wishes. But to determine whether there is one main

pattern requires operational definitions. In CCRT

terms it might be that for each patient there is one

high-frequency theme and other themes with much

lower frequency. This translation implies that for the
one high-frequency theme the components of the self-
other interactions are highly pervasive across

different relationship episodes. We found this to be

true in a sample of 8 patients (Luborsky et al., 1985,

1986); for example, the wish with the highest

pervasiveness was much more pervasive across the

www.theipi.org 805
www.freepsychotherapybooks.org
 narratives than its runner-up in the order of

frequency. Another example of stepped-down

rankings of frequencies of CCRTs appeared in the

rankings of the pervasiveness of each cluster for each

child at Age 3 and 5, reported in chapter 16. Most

children had high pervasiveness for their top two
clusters and much lower pervasiveness in their
remaining six clusters.

This emphasis both in Freud’s observations and in

the CCRT data on one main pattern needs to be

considered in relation to the work of others, who
posit multiple schemas (e.g., Crits-Christoph &

Demorest, 1991; M. Horowitz, 1987; J. Singer,

1985). However, the emphasis of the CCRT on one

main pattern is only relative; the CCRT analysis

typically turns up several other, less frequent patterns

along with the main one. The observation that the
CCRT has one main pattern is often misunderstood
by people who have not used the CCRT extensively.

www.theipi.org 806
www.freepsychotherapybooks.org
 Such people at times ask, “How can that be? I often

see other themes.” Further experience with the CCRT

would make it plain to them that although the scoring

reveals a variety of themes, only the most frequent

one is designated the CCRT.

Another factor that needs to be considered is the

method of CCRT scoring of the narratives. Was it the
usual method as spelled out in chapter 2, this volume,
or was it another method? More precisely, was it a

system that allows for scoring the relationship

episodes in the context of other relationship episodes
in the session, or was it a scoring system that

evaluates each relationship episode one by one after

they have been randomized? In the latter case, it is

likely that more variety in CCRTs would be

discovered, because the clinical judge is not given the

usual opportunity to know the larger context of other
relationship episodes in which each relationship
episode occurs. The function of the other relationship

www.theipi.org 807
www.freepsychotherapybooks.org
 episodes may be to allow the judge to see a
component that is revealed only opaquely in the

relationship episode being judged and may not be

noticed without the other relationship episodes. (In

addition, even with the usual CCRT method,

described in chapter 2, it is important to know the
degree to which the judge followed the
recommendation for scoring both Phase 1 and Phase

2, and then Phase 1' and Phase 2', that is, the degree
to which the judge considered and then reconsidered

the scoring, taking into account the whole set of

relationship episodes.)

10. Specific subpatterns appear for each family

member. In the same article referred to in the
previous section, Freud (1912/ 1958a) also noted
that the pattern is not tied to a “particular
prototype.” Instead, there are several common
“prototypes”; the principal ones are the “father-
imago,” “mother-imago,” and sometimes
brother- or sister-imago. It follows from the
nature of these prototypes (although it is not

www.theipi.org 808
www.freepsychotherapybooks.org
 directly stated in the 1912 paper) that each
specific prototype forms from the memories of
experiences in relation to a significant family
figure. In the Dora case, Freud (1901-
1905/1953a) thought of the main prototype in the
transference pattern as the father and a father
figure (Herr K). He noted from time to time that
there was another pattern in Dora’s relationships:
the one involving Frau K and Dora’s mother.

Evidence for specific CCRT prototypes has been

extensively examined in one case (Crits-Christoph &

Demorest, 1991) in which there were indications of

subpatterns. It would be helpful if researchers

collected narratives about each of the family

members to permit separate CCRTs for each. My

strong impression at this time is that there is both a

main CCRT, which is general across different other

people, and some partially specific subpatterns for a
few significant other people.

11. The central relationship pattern is distinctive for

each narrator. Freud (1912/1958a) stated that

www.theipi.org 809
www.freepsychotherapybooks.org
 “people have their own special transference
pattern....each individual...has acquired a specific
method of his own in the conduct of his erotic
life” (p. 99).

Evidence for the distinctiveness of each person’s

main pattern compared with that of other people is

based on the CCRTs in a small sample of 8 patients

(Luborsky et al., 1985, 1986). The distinctiveness is

especially striking when the entire CCRT pattern,

based on all three components, is compared among

people, rather than when each component is

compared among people. A larger study of the degree
of distinctiveness of each person’s pattern is needed.

12. The central relationship pattern tends to be

consistent over time. Freud (1912/1958a)
referred to the pattern as “constantly repeated—
constantly reprinted afresh—in the course of the
person’s life.” The term template fits best with
this observation, for it implies that there is a
pattern serving as a prototype for shaping
replicas in later editions.

www.theipi.org 810
www.freepsychotherapybooks.org
 There is no evidence about the degree of
consistency over the course of the person’s life, but

there is evidence for the consistency of the pattern

during approximately 1 year of treatment (see chapter

10, this volume). For example, for 33 patients whose

pattern was measured early in treatment, on the

average the main wish was evident across 66% of the

early relationship episodes and across 62% of the late

relationship episodes. For the future, a longitudinal

study extending the studies of Dengler (1990) and

Luborsky, Luborsky, et al., (1995; also see chapter
16) of children’s narratives at ages 3 and 5, and

continued into later life, could help establish the

degree of consistency of the CCRT.

13. The central relationship pattern changes slightly

over time. In Freud’s (1912/1958a) words, the
pattern is “certainly not entirely insusceptible to
change” (p. 100). The wording of Freud’s
observation implies that he believed the pattern
had considerable stability but also some
plasticity over the long term.

www.theipi.org 811
www.freepsychotherapybooks.org
 In fact, it was found (see chapter 10) that whereas

wishes had considerable stability in their

pervasiveness across the person’s narratives over
time, they also showed a slight (nonsignificant)

decrease from the early to the late period in

psychotherapy. Even larger changes were shown in
pervasiveness of the negative responses from others
(from 41% to 29%) and negative responses of self

(42% to 23%). These decreases were expected

because most of the patients had improved, and the

degree of improvement was associated with the

degree of decrease in these two negative components.

14. The central relationship pattern shows short-

term fluctuations in activation. This observation
involves the very short-term changes in
activation of the pattern, often during a session,
as distinguished from the long-term changes
referred to in Observation 13. This Observation
14 is based partly on the effects of Observation
7, that the therapist’s characteristics may activate
the pattern.

www.theipi.org 812
www.freepsychotherapybooks.org
 Changes in activation of the pattern are

commonly observed within sessions as well as across

sessions. Examples of such changes are not difficult
to find: For Mr. Howard, within Session 3 (see
chapter 5, this volume), the relationship episode

about the therapist reflects greater activation of the

pattern and reflects a within-session increase in

activation. The therapist response to that relationship

episode implies that the therapist was aware of the

activation. When the patient began to be markedly

unresponsive, distant, and tense during the session,

the therapist offered the interpretation that the patient

expected the therapist to be more responsive. The

interpretation was timed to coincide with the greatest
activation of the transference pattern in the session.
There has been little operational translation of this

basic observation about within-session activations.

Only the work of Gill and Hoffman (1982b) has

shown an attempt to score such changes in activation.

www.theipi.org 813
www.freepsychotherapybooks.org
 15. Interpretation changes the expression of the
central relationship pattern. Through
interpretations that focus on the transference
pattern, the pattern becomes altered. An absence
of such interpretation was the “defect” Freud
(1901-1905/ 1953a) decided was responsible for
the premature breaking off of treatment by Dora.
He wrote the following:

I did not succeed in mastering the transference in

good time. … when the first dream came, in
which she gave herself the warning that she had
better leave my treatment just as she had
formerly left Herr K’s house, I ought to have
listened to the warning myself. ‘Now,’ I ought to
have said to her, ‘it is from Herr K that you have
made a transference onto me. Have you noticed
anything that leads you to expect evil intentions
… similar to Herr K’s?’ (p. 118)

There is some evidence that is exactly on target in

supporting the observation: Interpretations with

greater focus on the pattern reflected in the CCRT

—“accurate” interpretations—were more strongly

associated with benefit to the patient (see chapter 13,

this volume). There is also some evidence that bears

www.theipi.org 814
www.freepsychotherapybooks.org
 generally on this observation; for example, a greater
number of interpretations that combine present and

past relationships have been reported in a review of

studies (Luborsky et al., 1988) to be positively related

to outcome of therapy. Although the number of

interpretations is a factor in changing the central
relationship pattern, as shown by Piper, Azim, Joyce,
and McCallum (1991), the accuracy of the focus of

the interpretations is also important.

16. Insight into the central relationship pattern can

benefit the patient. Accurate interpretations help
the patient to gain insight into the pattern, which
in turn leads to a reduction of the intensity of
transference or to greater mastery of it and to
ultimate benefits to the patient. “One must allow
the patient time to become conversant with this
resistance ... to work through it, to overcome it”
(Freud, 1914/1958e, p. 155).

The evidence from the CCRT research shows

mixed results about the benefits of insight (see
chapter 14, this volume). The measure of self-

www.theipi.org 815
www.freepsychotherapybooks.org
 understanding or insight used was the convergence of

(a) the patient’s self-descriptions with (b) the

independently established CCRT. The assumption

within this operational measure is that the more these

two converge, the greater the self-understanding. The

first result was consistent with the theory: The level
of self-understanding in an early session was
significantly correlated with the patient’s benefits

from the treatment. The second finding was not

confirmatory of the theory: The gain in self-

understanding was not significantly correlated with

the patient’s benefits from the treatment. However,
the operational measure in the study that was chosen
for examination of gains—that is, change from
Session 3 to Session 5—did not seem to fit well with
the theory; therefore, better operational measures are

needed.

Operational measures of this observation, and of

Observation 15 on the benefits of interpretation of the

www.theipi.org 816
www.freepsychotherapybooks.org
 pattern, have also been successfully tried by the
Mount Zion psychotherapy research group (Norville

et al., 1996; Silberschatz et al., 1986) but with a focus

on the immediate impact of convergent

interpretations. The authors defined self-

understanding operationally as the convergence of the
interpretations with their measure of the Plan
Diagnosis (as described in chapter 20, this volume).

In contrast to our study, which was based on benefits

received by the time of the termination of the

treatment, their measure of impact on the patient was

an immediate one; that is, a sample of the session just

after the interpretation as rated by independent

judges. They found that the greater the convergence

of the interpretation with the Plan Diagnosis, the

more beneficial was the immediate impact on the

patient, in terms of greater experiencing of affects
and greater insight.

www.theipi.org 817
www.freepsychotherapybooks.org
 17. The central relationship pattern can serve as a
resistance. Freud (1912/1958a) considered it
surprising and “a puzzle why in analysis
transference emerges as the most powerful
resistance to the treatment” (p. 101). “Every
single association, every act of the person under
treatment must reckon with the resistance and
represents a compromise between the forces that
are striving towards recovery and the opposing
ones [from the transference]” (p. 103). For
example, when Dora viewed Freud (1901-
1905/1953a), her therapist, as acting like her
father, and this piece of the transference was not
interpreted, the transference served as a
resistance, and the treatment became ineffectual
and was aborted by the patient.

Empirical studies must be designed to examine

the hypothesis that the transference is the most
powerful resistance. But the puzzle might become

somewhat less puzzling if one considered that the

pattern reflects traitlike attitudes in the relationship
patterns to self and to others; in that light, it is not

surprising that there is resistance to changing them.

www.theipi.org 818
www.freepsychotherapybooks.org
 18. Symptoms may emerge when the pattern is
activated. The Dora case (Freud, 1901-
1905/1953a) provides examples of the
observation that the conflicts that are active
during the telling of a patient’s memories
constitute a context conducive to the appearance
of symptoms. Dora’s hysterical symptoms
emerged following the activation of her central
relationship patterns, in particular of the
erotically related ones.

In the CCRT research, we have found many

examples that fit Freud’s observation. The conflictual
relationship pattern, expressed in terms of the CCRT
format, forms the context for the appearance of the

symptoms, with the symptoms typically included in

the response of self component of the CCRT. The
appearance of a symptom during a psychotherapy
session can be greeted, therefore, as an opportunity to

examine its context and learn more about its

meanings, as shown systematically for the first time

by Luborsky (1996).

www.theipi.org 819
www.freepsychotherapybooks.org
 19. The pattern expressed within therapy is similar
to the pattern expressed outside of therapy. The
point of this observation is that the template is a
general pattern and not confined to expression in
the treatment. Freud’s (1912/1958a) observation
was exactly that: “It is not a fact that transference
emerges with greater intensity and lack of
restraint during psychoanalysis than outside” (p.
101).

There is evidence for the expression of the pattern

outside of the treatment from a small-sample study of

narratives that were told during an interview
conducted by a person other than the therapist

compared with narratives that were extracted from

treatment sessions (van Ravenswaay et al., 1983).
Following the proposed plan by Luborsky (1986a), a
larger, more controlled study (Barber et al., 1995)
showed this parallel: A RAP interview before

treatment conducted by a person other than the

therapist revealed similar CCRTs to those found in

early therapy sessions.

www.theipi.org 820
www.freepsychotherapybooks.org
 20. Positive and negative patterns are
distinguishable. Freud (1912/ 1958a) generally
used the term positive to refer to “affectionate”
feelings and negative to mean “hostile” feelings.
Positivity was further divisible into transference
of (a) “friendly or affectionate” feelings and (b)
feelings that have erotic sources (p. 105). Yet in
the most general terms, Freud commonly used
this distinction of positive or negative to mean
the expectation or nonexpectation of a
“frustration of satisfaction” (p. 103).

It is difficult to come up with exactly specified

operational translations of the terms positive and

negative, yet what Freud meant is clear enough to

guide in making the distinction between positive and

negative instances in sessions. My colleagues and I
(see chapter 4, this volume) tried an operational
measure of a CCRT component that is related to

Freud’s definition: the positive or negative responses

from other and the positive or negative responses of
self. These appear to be appropriate measures,

because when they are positive or negative they

www.theipi.org 821
www.freepsychotherapybooks.org
 brand the whole pattern as positive or negative. We

found that when each judge identified the same

response from other or the same response of self,
there was 95% agreement on whether it was positive
or negative. A more precise scoring of positive and

negative, allowing two degrees of severity for

positive and two degrees for negative, also showed

high agreement between judges and evidence of

validity (see chapter 4).

One of the side benefits of having scored the

CCRT components for their positive or negative

quality was the knowledge acquired about their

relative frequency. For virtually every patient, the

number of negative responses far exceeded the

number of positive ones (see chapters 4, 7, and 16).

This is true both for the narratives told spontaneously

in the course of psychotherapy and for the narratives
told on request as part of the RAP interview. Possibly
the high frequency of negative responses in narratives

www.theipi.org 822
www.freepsychotherapybooks.org
 derives from the need to remember and to talk about
negative or traumatic events or the need to master
negative or traumatic events (as discussed in chapter
22).

21. The pattern is expressed similarly through

different expressive modes. A similar pattern can
appear in a variety of modes of expression: in
behavior, in narratives about relationships, and in
dreams. Although this observation was not
explicitly stated, it was clearly exemplified in
Freud’s 1912 paper (1958a) and in the Dora case
(1901 — 1905/1953a), in which two dreams
were analyzed to show the patient’s main
relationships to parents and others.

In my initial study (Luborsky, 1977b), the main

CCRT was found to appear similarly in both dreams

and narratives. In chapter 12, independently scored

CCRTs within dreams and within narratives were

found to be significantly similar in three preliminary

cases, and the trend was the same in an expanded

www.theipi.org 823
www.freepsychotherapybooks.org
 study of the dreams and narratives of 13 patients in
psychoanalysis.

22. Greater improvement in dealing with the pattern

implies greater mastery of the pattern, although
the pattern itself remains evident. Several of
Freud’s observations dealt with the pervasive
therapeutic aim to give the patient greater
freedom, control, and even mastery over neurotic
conflicts. Among these are the following:
Observation 4 concerns lack of awareness of
aspects of the transference, which needs to be
overcome by the treatment. Observation 8 deals
with distortions of perception, which need to be
overcome in the course of treatment. Observation
15 deals with the use of interpretation to
overcome deficits in knowledge of the negative
transference. Observation 16 fits with the aim of
the treatment to give greater insight into the
conflicts, especially those related to the
transference. In Freud’s writings, there are
numerous instances in which treatment was
aimed at recovering memories in the service of
mastery. For example, Freud (1920/ 1955a, p.
35) discussed the way that neurotic patients try to
achieve resolution of their conflicts through

www.theipi.org 824
www.freepsychotherapybooks.org
 “mastering or binding.” At the end of one of his
successful cases, Freud said, “In these last
months of his treatment he was able to reproduce
all the memories and to discover all the
connections which seemed necessary for
understanding his early neurosis and mastering
his present one" (Freud, 1937/1964, p. 217).

Luborsky (1977b, 1984) noted that when patients

improve, there is a shift toward greater mastery of the

pattern. This finding was much more fully examined

by Grenyer and Luborsky (1996) and is discussed in

chapter 18, this volume, where it is revealed to be a

predictor of outcome in psychotherapy.

23. Innate disposition plays a part. Freud

(1912/1958a) observed that the factor of
constitution plays a conjoint role with the factor
of early experience: “We assume that the two
sets regularly act jointly in bringing about the
observed result” (p. 99). The interaction differs
in individual cases because experience and
constitution operate jointly in determining the
pattern.

www.theipi.org 825
www.freepsychotherapybooks.org
 Freud’s observation seems applicable on the basis

of modern knowledge of genetics, but relevant data

on heritability of the CCRT are not available. A

collection of RAP interviews (see chapter 7) from a

source like the Danish adoption registry, with data

from adoptive and biological parents and children,
probably could make such a contribution.

DISCUSSION

The Nature of Freud’s Observations

What kinds of observations are these 23 that led

Freud to the concept of the transference template?

They are a varied bunch, with some overlap. They
deal with the origin, function, and activating stimuli

of the transference template, as well as with measures

that may reduce or contain the expressions of the

template. This diversity makes it difficult to

characterize the observations as a set. Rather than
having been theoretically inferred, most of the

www.theipi.org 826
www.freepsychotherapybooks.org
 observations appear to have been empirically based

and derived from Freud’s own experience in making

inferences about patients’ relationship patterns. They
reflect Freud’s penchant for carefully collecting
observations about mental phenomena (Holt, 1965).

He showed this observational style early in his career

when, working in Breucke’s physiology laboratory,

he described the structure and function of specimens

he viewed under the microscope. The basic

observations he made about the transference template

deserve special attention for they form the “durable
core of Freud’s empirical science,... it is the
[psychoanalytic] method, the observations and the

immediate inferences drawn from them that count in

the end ... it provides a solid and dependable base

upon which to build” (Reiser, 1986, p. 8).

Freud’s style was to marshal his observations

from time to time so that they might lead to concepts
about the operation of deep structures. Although in

www.theipi.org 827
www.freepsychotherapybooks.org
 the course of theorizing he recurrently resolved to

restrain his attraction to concepts that were too

inferential and too metapsychological, he often gave
in to the temptation (Holt, 1989). Fortunately, with
the transference template he came to a middle-level,

clinically grounded concept (1912/1958a) of a mental

representation of knowledge about patterns of

relating to other people that guides the conduct of

relationships.

In this chapter the focus is on the diversity in

Freud’s concept of transference, as stated in his

specific observations and as illustrated by the
transference formulations in his case examples. But
much work on his observations remains to be done to
fulfill the research agenda as well as to deal with

comparisons with the long line of clinical and

theoretical definitional clarifications by others written

over the years; among these authors are Arlow and

Brenner (1964), Curtis (1983), Waelder (1936), Spitz

www.theipi.org 828
www.freepsychotherapybooks.org
 (1956), Greenacre (1954), and Nunberg (1951), as
well as many others.

The Correspondence of Freud’s Observations With

CCRT Evidence

To compare Freud’s observations about his

transference template with parallel evidence from the

CCRT, each of Freud’s observations had to be

expressed in an operational form that could be

examined using CCRT data derived from

psychotherapy or psychoanalytic sessions. So far,

researchers have been able to make this comparison
for 18 of the 23 observations (see Table 1). For 11 of

the 18 observations, studies showed a good

correspondence between Freud’s observations and the
CCRT evidence (marked with a plus sign):
Observations 1, 2, 5, 6, 12, 13, 15, 19, 20, 21, and 22
(the five underlined have the most supportive
studies). For 7 of the 18, there is pilot data with

results that are mixed but look promising (marked

www.theipi.org 829
www.freepsychotherapybooks.org
 with a plus sign and question mark). For a few of the

5 observations that remain to be studied (marked with

R), it probably will be possible to work out

operational measures and proceed with a study. The
overall conclusion for now is that researchers have

found a degree of correspondence for 18 of the 23

observations and that performance implies some

success for the comparisons.

The 23 observations can be subgrouped on the

basis of two broad principles: (a) the breadth of the

pattern and (b) the stability of the pattern. Four of

Freud’s observations relate to the breadth of the

pattern. These observations include the fact that there

is one main pattern (Observation 9); that after

treatment is begun, the pattern soon involves the

therapist (Observation 6); that the pattern is present

both in and out of therapy (Observation 7); and that

the pattern also emerges in different modes of

www.theipi.org 830
www.freepsychotherapybooks.org
 expression, including waking narratives and dreams
(Observation 21).

The second broad principle in the observations is

stability over time, a principle that qualifies

transference as a personality structure. According to

Rapaport (1951), a personality structure is an entity
with a slow rate of change. The observations about
transference clearly fit this principle: It has some
consistency over time (Observation 12); it also shows
consistency from the time of its early parental origins

to late in life (Observation 5); and its stability is

especially localized in the remarkable long-term

persistence of the wishes toward people (Observation
1).

Another encompassing generalization about the

23 characteristics of Freud’s concept of transference

is the degree to which these characteristics are like

the general characteristics of psychological schemas

www.theipi.org 831
www.freepsychotherapybooks.org
 of the kind that Knapp (1991) defined as “an enduring

symbolic framework that organizes constellations of

thought, feeling, memory, and expectation about self

and others” (p. 94). Such self-other person schemas

are considered by Knapp to have (a) pervasiveness,

(b) concreteness, (c) tenacity, (d) urgency, (e)
simultaneity of existence of many elements of
patterning, and (f) plasticity of outward

manifestations, including shifts in awareness. All of

these appear to be represented in Freud’s observations

about the concept, although by slightly different
names.

Is there at this point enough evidence that the

transference template and the CCRT can be thought

of as similar conceptually? Mostly, yes. Although

more evidence should be and will be assembled, the
evidence pulled together in this chapter shows that
the two concepts have much in common.

www.theipi.org 832
www.freepsychotherapybooks.org
 A further principle for making conclusions about

the commonality of the concept of the transference

and the CCRT evidence has emerged from the
congressional debates in 1988 and 1989 about
whether each of various proposed tax measures was

really a disguised tax. The opinion of the opposition

was often: “If it looks like a duck and it talks like a

duck, then it is a duck!” Applying this principle then,

is it proper to say about the CCRT that if it looks like

transference and talks like transference, then it is

transference? Yes, almost, but not exactly. As the

diagram in Figure 1 shows, the similarities and

differences can be simply stated: (a) It is fitting to say

that a clinician’s transference formulation is the
clinician’s unguided estimate of the central
relationship pattern, (b) A clinician’s formulation in

CCRT terms is probably a largely overlapping but

guided version of the central relationship pattern.

CONCLUSIONS

www.theipi.org 833
www.freepsychotherapybooks.org
 Figure 1

The central relationship pattern in the transference

and in the CCRT.

www.theipi.org 834
www.freepsychotherapybooks.org
 • In the course of Freud’s perennial attempts to
specify his transference concept, he made 23
observations about its nature.

• A key element in the strategy of our research was

the translation of each of Freud’s observations
about transference into its most tenable
operational definition to simplify and objectify
the comparison with the CCRT evidence.

• For 18 of these 23 observations, research with the

CCRT has found some convergence of the two
versions of the central relationship pattern:
Freud’s transference template and the CCRT.
Because researchers have found a reasonable
operational measure for the majority of Freud’s
observations, it has been possible to examine
convergence with the evidence from the CCRT,
and the results of the comparison have
demonstrated a meaningful kind of validity for
the CCRT.

• The benefit from using the CCRT method is that it

enables clinicians to achieve consensus in their
session-based judgments about this complex
concept, a sought-after feat that had repeatedly

www.theipi.org 835
www.freepsychotherapybooks.org
 suffered defeat before the advent of the CCRT
method.

www.theipi.org 836
www.freepsychotherapybooks.org
 22

WHERE WE ARE IN UNDERSTANDING

THE CCRT

LESTER LUBORSKY

Just tell a set of narratives about events that have

happened to you in relation to other people, either

recently or in the past. Then call on this guide to the

CCRT method to help you trace through the set of

narratives to find the red threads of the central

relationship patterns within them. The pattern that

will be revealed is much like the one typically called
the transference by Freud (1912/1958a) and other
psychoanalysts. That, in brief, is a satellite’s snapshot
of the large contours of what is presented in this

book.

www.theipi.org 837
www.freepsychotherapybooks.org
 This chapter re-views what has been seen in this
book so far but with a limited focus on where to build

around the two largest formations supporting the

CCRT in its present, mature state: the areas of

reliability and validity. Construction should be on the

dozen or so highest priority topics that must be
developed to promote the continued maturation of the
CCRT method. Roughly, the first half dozen of these

topics are in the area of reliability and the second half

dozen are in the area of validity.

HOW MUCH MORE RELIABLE IS THE CCRT

METHOD THAN THE USUAL CLINICAL
METHOD?

For an assessment method like the CCRT to be

useful, clinicians and trained research judges must be

able to follow its procedures with at least moderate

interjudge agreement. As chapter 1, this volume,

details, it took years of trying this and that and a dash

of luck to come up with a reliable system. The now

www.theipi.org 838
www.freepsychotherapybooks.org
 considerable evidence that the CCRT can be judged
reliably (see chapter 6) makes it the first clinically

judged measure of a central relationship pattern

derived from psychotherapy that shows satisfactory

levels of reliability. The secret of my success in

creating a reliable measure of this complex concept
was the reliance on the principle of guided clinical
judgment: The CCRT provides all judges with the

same basic system for scoring so that all judges can

follow it in making inferences about the central

relationship pattern. Therefore, the CCRT method is

consistent with Holt’s (1978) principle that guided

clinical judgments have a better chance than

unguided ones of achieving both interjudge

agreement and successful predictive performance.

In summary, the reliability of the CCRT method is

superior to that of the unguided clinical method, as

found in the research by Seitz (1966) and DeWitt et

al. (1983). Yet a more exact comparison of the two

www.theipi.org 839
www.freepsychotherapybooks.org
 approaches is needed. One study in this area is
already in progress (Friedman & Luborsky, 1996); it

compares a set of cases evaluated by the same

clinicians first using the usual clinical judgment

method and then using the CCRT method. The

authors already can see the expected advantage, but
they hope to be able to specify the amount and kinds
of advantages.

DO TAILOR-MADE OR STANDARD
CATEGORIES WORK BETTER?

The mainstay of the original form of the CCRT

was the tailor-made scoring system, in which each of

the clinical judges makes inferences about each
patient’s narratives using descriptive categories that
fit each patient best. The main asset of the tailor-made

method is that categories are fashioned to fit each

person. But some of its categories, therefore, are

likely to be unique to each person and some overlap

ambiguously with each other. Naturally, with this

www.theipi.org 840
www.freepsychotherapybooks.org
 system researchers continually came up against
perplexities in computing interjudge agreement for

categories.

To deal with the problem, my colleagues and I

tried scoring systems with sets of standard categories,

as we report in chapter 3, this volume. In these
systems each time a judge wishes to make a particular
kind of inference about a thought unit, the judge is
instructed to use an appropriate word in the standard
category list. In that way, each judge uses the same

words for the same concept as the other judges.

Judges were able to do this task relatively rapidly and

with significant agreement with each other (see
chapter 6). In effect, the main assets of the standard
category system are that (a) it relies on the use by
each judge of the same categories, and (b) it is easy to

find the degree of agreement among judges.

www.theipi.org 841
www.freepsychotherapybooks.org
 More comparisons of the tailor-made and
standard category methods are needed. These studies

should teach researchers for which special purposes it

is better to use the tailor-made categories first and

then translate them into standard categories versus for

which purposes it is justified to use the shorter
procedure of omitting tailor-made categories and
inferring standard categories directly. After

experience with both the tailor-made and the standard

category methods (see chapters 2 and 5), our opinion

is that, because of their different assets, both methods

used in succession contribute more than either system

by itself; use of the tailor-made system should be

followed by application of the standard categories.

There is a problem that lessens the adequacy of

the translation from tailor-made to standard

categories: From time to time there appear to be

meanings in the tailor-made categories that are not

captured well in the existing standard categories.

www.theipi.org 842
www.freepsychotherapybooks.org
 Research efforts will in time move toward expanding

the standard category list to lessen the frequency of

its shortfalls in meaning.

MIGHT THE TRANSFERENCE BE A MERE

PRODUCT OF SUGGESTION BY THE
THERAPIST?

The possibility of the therapist’s contaminating

the patient’s transference pattern through suggestion

is a crucial defect of reliance on sessions according to

Grunbaum (1984), that prevents research on

psychotherapy sessions from being able to prove

anything. In essence, he argued that because the
content of the session is vulnerable to the therapist’s

influence, research can prove nothing using data from

sessions. This argument reflects the all-or-nothing
reasoning of a logician. In contrast, this now age-
worn issue was examined by Freud (1914/1958e),
who believed that there were ways to minimize the

possibility of the therapist’s suggestion effects and

www.theipi.org 843
www.freepsychotherapybooks.org
 ways to judge the therapist’s distortion of the

patient’s transference; for example, when a patient

begins treatment with a behavior in the session that is
remembered as a repetition of any early experience,
the therapist is not likely to have suggested it.

Despite this controversy, there is a huge and

growing body of empirical studies of psychotherapy

sessions focusing on the many initial factors that

influence their outcomes (Luborsky et al., 1988;

Orlinsky, Grawe, & Parks, 1994). The findings of

these studies support the view that even though the

therapist might have the power to influence the
patient through suggestion and other means, there are
aspects of the patient that are relatively stable and
thus less subject to influence. The CCRT may be one

instance of a facet of personality that is resistant to

change, with the wishes within the CCRT the most

constant of the CCRTs components over time.

www.theipi.org 844
www.freepsychotherapybooks.org
 A demonstration of the stability of the CCRT
would involve getting narratives from the patient

before the patient even meets the therapist and

comparing these with narratives obtained after a few

early sessions with the therapist, as was done by

Barber et al. (1995). The CCRTs were extracted from
each set independently to determine the degree of
similarity of the data sets from before with those from

after the therapy started. We found considerable

consistency between these data sets regardless of the

other types of changes that were stimulated by the

therapy.

Another design also provides results that bear on

the question: use of data from a patient who had two
different therapists in sequence so that a comparison
can be made of the CCRTs with each therapist. Only

one such example with one patient, Ms. Apfel (see

chapter 12, this volume), was available (Luborsky,

1988b, 1996). This patient’s analysis was interrupted

www.theipi.org 845
www.freepsychotherapybooks.org
 by the illness of her first analyst, and she continued

with a second analyst. An examination of the context

of her momentary forgettings showed a similar CCRT

context for the first and second analyst.

IS THE CCRT FROM THERAPY NARRATIVES

SIMILAR TO THE CCRT FROM RAP
NARRATIVES?

Although clinicians are inclined to believe that

narratives told during psychotherapy sessions are
more revealing than those told as part of a special
narrative-telling interview, such as the Relationship

Anecdotes Paradigm (RAP) interview, the two data

sources appear to give similar results in terms of the

CCRT. My colleagues and I studied a group of

depressed patients (Eckert et al., 1990) from whom

we obtained both spontaneously told narratives from
their psychotherapy and requested narratives from a
RAP interview. As was expected by Luborsky

(1986a), the CCRTs from RAP narratives told to

www.theipi.org 846
www.freepsychotherapybooks.org
 someone other than the therapist before treatment
started were not significantly different from the

CCRT from early session narratives (Barber et al.,

1995). This study contributed to the knowledge of the

reliability of the RAP, as well as to allaying the

concern that some have expressed (e.g., Grunbaum,
1984) that the content of the transference pattern is a
product of suggestion on the part of the

psychotherapist or psychoanalyst.

ARE RAP NARRATIVES ABOUT “REAL”

EVENTS SIMILAR TO “STORIES” FROM THE
TAT?

The Thematic Apperception Test (TAT), used

extensively in personality assessment for the last 60
years (Murray, 1938), entails asking the person to

make up stories about a standard set of pictures. My

colleagues and I rely instead on RAP (Relationship

Anecdote Paradigms) narratives that are intended to

be about real events and are not stimulated by a set of

www.theipi.org 847
www.freepsychotherapybooks.org
 pictures. Some similarity of TAT and RAP narratives

should be present, but it will be only slight. It is

easier for people to tell about events that happened to
them than to compose stories in response to pictures.

In addition, narratives about real events offer more

information to the evaluator revealing what the
patient knows about his or her relationships and also
provide the therapist with more reliable information

about the patient’s life.

WOULD BEHAVIORAL ENACTMENTS OF

RELATIONSHIPS REVEAL MORE THAN THE
USUAL NARRATIVES ABOUT
RELATIONSHIPS?

Narratives about relationship events naturally

tend to be told from the perspective of the teller of the

narrative. Direct observations of actual relationship

events would be expected to yield additional or

different information from narratives about these
events. The differences among viewers of events can
be so great at times as to warrant their being called a

www.theipi.org 848
www.freepsychotherapybooks.org
 “Rashomon” experience, meaning that the

perceptions of the viewers differ drastically (Mintz,

Auerbach, Luborsky, & Johnson, 1973). It is because

of our group’s interest in this comparison that we
have collected direct observations of relationship

interactions by means of examining “enactments.”

These are brief behavioral relationship interactions

with the therapist that occur in psychotherapy

sessions (described in chapter 2, this volume). They

are identified within the session by the same judge

who identifies the narratives. They are a special
category of the relationship episodes with the
therapist in which the therapist and patient engage in

a discrete, delimited enactment of an episode of

interaction. They can also be used as part of the

database for the CCRT, although there are not many

included so far. When enough enactments have been

collected, we will compare these with the narratives
told about the relationship with the therapist.

www.theipi.org 849
www.freepsychotherapybooks.org
WHAT WOULD BE LEARNED FROM SCORING
SEQUENCES OF CCRT COMPONENTS?

In chapter 2, I offered another scoring option

from Luborsky (1984): to record the exact sequence

of the CCRT components as they are presented within

each narrative. This option might give a better picture

of the interactional sequences than the current method

of simply counting all wishes, responses from other,

and responses of self. Only a little information has
been systematically collected on this option. In the
one comparison we did (see chapter 9) of the actual

sequence method with the usual CCRT method, we

found that the two methods yielded similar CCRTs.

But of course more evidence is needed because the

issue is so basic to the usual CCRT method. In

another study, Mitchell (1995) has shown that
sequences that entail a larger percentage of
interactions with people indicate a higher degree of
psychological health.

www.theipi.org 850
www.freepsychotherapybooks.org
ARE DEFENSES CLASSIFIABLE WITHIN THE
RESPONSES OF SELF?

A frequent question from audiences after

presentations about the CCRT is, “How does the

CCRT deal with defenses?” The answer is that it

could but it does not; there is not yet a formal place

for defenses in the scoring system, although it is clear

that the response of self component is the logical

place for including judgments of defenses. It is
noteworthy that clinical judges, even when using the
tailor-made category system, virtually never list

defenses. Perhaps they would if they were instructed

on the desirability and method of doing so. If one is

to examine categories of defenses in relation to the

CCRT, one should include a sample of the most

frequent and recognizable defenses within the
standard categories list (drawing from Perry &
Cooper, 1989, or from Vaillant, 1977). Good
candidates for such categories are denial, projection,

www.theipi.org 851
www.freepsychotherapybooks.org
 and isolation, because they are likely to be reliably
recognizable.

One study has been done, however, on the CCRT

in relation to a repressive defensive style. Luborsky,

Crits-Christoph, and Alexander (1989) found that

people who are classed as repressors tend to express
wishes having to do with closeness; patients who are
classed as isolators tend to express wishes that have
to do with independence and autonomy. These
findings for defensive style suggest the potential

value of the inclusion of categories in the CCRT that

measure defenses.

HOW CAN THE OBSERVATION BE TESTED

THAT THE TRANSFERENCE IS “PARTLY OUT
OF AWARENESS”

My colleagues and I began studying the

awareness of the CCRT with the work reported in

chapter 15, this volume, about the patients’ versus the

clinicians’ interpretations of the narratives. Our

www.theipi.org 852
www.freepsychotherapybooks.org
 reasoning was that patients might be less aware of

some aspects of their CCRT than the clinicians were.

Surprisingly, there is considerable parallel between
the patients’ understanding of their own CCRT and

the clinician’s understanding of that CCRT. But the

parallel primarily appears in terms of the recognition
of many of the same types of components by the
patients and the clinicians. What the clinicians

apparently do better than the patients is to identify the

relative degrees of importance of the types of

components; all this and more is discussed in chapter

15.

This question is examined in another way as well

in chapter 15: by the first of the questionnaire

methods for assessing the CCRT. The patient fills out

a CCRT relationship questionnaire that contains the

same categories that are used for standard scoring of
CCRTs. An improved version of the CCRT

www.theipi.org 853
www.freepsychotherapybooks.org
 relationship questionnaire is being tried (Barber,
1993).

This kind of exact comparison of the relationship

questionnaire and the narrative-based CCRT needs to

be continued. The CCRT method takes a lot of time;

in contrast, the questionnaire methods are relatively

brief. Questionnaires conceivably might be developed

that can identify the kind of pattern identified by the

CCRT and do the job more easily. Another candidate

among the questionnaire methods is the Inventory of

Interpersonal Problems (L. Horowitz et al., 1983,

1988; also see chapter 20). This questionnaire

consists of about 90 self-report items about the
patient’s main interpersonal problems. The
questionnaire even asks the patient to pick the 5 items

that are most pressing: It is these 5 problems that

might be most related to the CCRT.

www.theipi.org 854
www.freepsychotherapybooks.org
 I have in progress another, very different

approach to measuring reduced awareness: a method

of identifying unconscious conflicts within the CCRT

(Luborsky, 1987). The method provides the judge

with (a) the CCRT, (b) the session transcripts on

which the CCRT is based, and (c) a set of principles
about how to make inferences about the type and
degree of reduced awareness for each aspect of the

CCRT. For example, the first of these principles is

that the most unconscious wishes may be found to be

the opposite of the most expressed wishes in the

CCRT. An instance of this principle in the CCRT for
Mr. Howard follows: The most expressed (and
conscious) wish is “to be close”; a more unconscious,
opposite wish is “to be distant.” The larger set of
principles that serve as a guide to making inferences

of this kind is helpful to the judges and appears to be

responsible for some of the agreement in their

www.theipi.org 855
www.freepsychotherapybooks.org
 inferences about the aspects of the CCRT that are in

reduced awareness.

WHY RELY ON NARRATIVES ABOUT

RELATIONSHIPS AS THE BASIS FOR THE
CCRT?

The decision to use narratives as the database for

the CCRT was a crucial choice and a blessed event

for the CCRT; it has demonstrated that narratives

offer a viable road to both the conscious and
unconscious basic conflictual relationship patterns.
The narrative can now claim to be worthy of sharing

with the dream the classic title of “the royal road”

(Freud, 1900/ 1953b); as Lyman Wynne (personal

communication, 1988) noted, “The CCRT method by

relying on the pattern derived from narratives about

episodes in relationships represents the first
systematic use of narratives that establishes them as
on a par with the systematic use of dreams.” We even

have been able to show, by direct comparison (see

www.theipi.org 856
www.freepsychotherapybooks.org
 chapter 12, this volume), that there is a parallel in
terms of CCRT content between narratives and

dreams. Narratives are, therefore, an informative

mode of communicating to the therapist the nature of

one’s central relationship pattern and the conflicts

within it. Beyond that discovery, the focus on the
narrative has led to an entirely new kind of
observation about the frequency of such narratives in

the course of psychotherapy sessions (see chapter 9).

The choice of narratives for deriving the CCRT

was based on a lucky observation. The first chapter of

this book tells the story of how my idea for the CCRT
took shape, at least for the part of the choice that was
in my awareness. “New” ideas reflect more of the
intellectual atmosphere than we know, it may be more
than just good luck, because the fashion for the study

of narratives could have played a part. The

atmosphere in the last two decades has been

increasingly charged with enthusiasm about the

www.theipi.org 857
www.freepsychotherapybooks.org
 properties of narratives, and that ambience could

have had some undercover influence on my attraction

to narratives. Although it is true that what is absorbed
and used in making discoveries shows signs of

passive intake of these trends of the times, that is not

the whole story. The rest of the story is, as Pasteur
observed, that Fortune favors the prepared mind.

To demonstrate that the present era has had a

permeating preoccupation with narratives, a brief
sample follows of recent works on four facets of

narratives.

Characteristics of Narratives

Narratologists, such as Labov (1972) and

Chapman (1980), agree that narratives have the

following characteristics: (a) two kinds of time are
distinguishable in a narrative: the time sequence of
plot events (the story time) and the presentation time
of the story in the text (the discourse time); (b) the

www.theipi.org 858
www.freepsychotherapybooks.org
 events have a time sequence, which means that there

is more than one event in the narrative; and (c) the

subject of the narrative is implicated in several

different events in the negative. Another contribution
on narratives (Toolan, 1989) explored the whole

range of narrative types—written and spoken, literary

and nonliterary—and showed what systematic

attention to the language can reveal about the

narratives themselves, their tellers, and their

audience.

The Truth Value of Narratives

This is a highly active frontier of exploration

about narratives. The truth value of narratives told

during psychotherapy is of concern to many writers

on the topic. Their views fall on two sides of a

controversy. On the one side are the “empirical

optimists,” who believe that parallels in such
narratives with actual events can be found out and are
worth trying to find out, for example, Edelson (1984,

www.theipi.org 859
www.freepsychotherapybooks.org
 1988), Reiser (1984), and my research group. On the
other side are those who might be called the

“hermeneutists,” who believe that the truth is not

easily found out or that trying is not worth the effort.

The kind of truth that matters, according to this

viewpoint, is the presence of an inner consistency of

the meanings within the narratives. Spence’s book

(1983b), Narrative Truth and Historical Truth, takes

up this issue and resolves it in favor of the “narrative

truth” alternative. Spence (1983a) wrote that “we are

all the time constructing narratives about our past and
our future, and … the core of our identity is really the

thread that gives meaning to our life” (p. 480).

The Revelations of Narratives About Modes of

Thought

Narratives about oneself have been Bruner’s

(1987) focus of investigation in the current phase of

his multifaceted career. He described two distinct

modes of thought: a narrative mode and a logical

www.theipi.org 860
www.freepsychotherapybooks.org
 argument mode. The narrative mode uses stories

about oneself, which are seen as guides for

structuring experience, and thus narratives are a
fundamental form of communication with others and

with the self. In contrast, the logical argument mode

generates a view of the mind as computerlike. That
mode leaves out what is retained in narratives about
the self: beliefs, desires, intentions, expectations, and

emotions.

The Interpretability of Narratives

For at least 3 decades some influential literary

critics have been carrying on an enthusiastic romance
with a deconstructionist view of narratives. Their
view is a rejection of what they believe is the aim of
older literary critics, of trying to find the meanings in

narratives. These current critics believe there is no

meaning in narratives—that meaning is only assigned

by the reader. This radical view is an extension of

Husserl’s (1960) concept of the “phenomenological

www.theipi.org 861
www.freepsychotherapybooks.org
 attitude” (and related philosophers’ concepts, as

reviewed by Murphy, 1938), in which one tries to see

a phenomenon for what it is in itself, through an
unprejudiced phenomenological reduction. This
current view has been extended and disseminated by

Derrida (1977).

Our findings with the CCRT have a lesson to

teach the literary critics who hold the current view of
the inherent nonmeaningfulness of narratives. The
lesson is that their basic premise is much too broadly

applied. CCRT research has shown that readers can,

in fact, agree fairly well with each other about certain

major meanings of narratives; one of these is the

central relationship patterns within the narratives. The

deconstructionist’s assumption that readers do not

agree may have some factual basis only when readers

provide interpretations of meanings without any prior
agreed-on guidelines for making their inferences. As
in the time-worn narrative about the blind men

www.theipi.org 862
www.freepsychotherapybooks.org
 examining the elephant, the unguided readers may be
looking at different parts of the narrative or looking at
the same part but judging according to different
guidelines. The findings based on the CCRT method
imply that reader agreement increases considerably

when readers are looking at the same part and guided

by the same method; for the CCRT, the raters are

directed by the relationship episode markers to the

exact location of the relationship episode.

Beyond that basic lesson, the CCRT findings also

have limitations that favor a restricted form of

deconstructionism! There are central parts of

narratives that allow inferences about their central

relationship pattern, but undoubtedly there are crucial
parts of narratives that cannot be judged with
agreement. These usually are parts that require high

levels of inference, such as, at what points in the play

is Hamlet mad, and at what points is he feigning

madness? How do we understand Lear’s last words?

www.theipi.org 863
www.freepsychotherapybooks.org
 With these parts of narratives, even guided systems
such as my own would have a hard time yielding

agreement among different judges. So, it is this

distinction between what can and what cannot be

agreed on that is lost sight of in deconstructionist

circles.

HOW DID EACH PERSON’S CCRT ORIGINATE

AND WHY IS IT SO PERSISTENT?

This book started with a discussion of the history

of how the CCRT measure came to be born. Now I

begin to end it with how the theme in the CCRT

originates and becomes a pervasive schema—a much

more mysterious origin to trace. This venture turns
into an adventure—a search for the factors that
explain the origins and persistence of the CCRT. So

far I have managed remarkable restraint in resisting

being drawn into meta-psychological speculation, a

temptation that Freud himself had a hard time with.

Even the recapitulation of the research results in the

www.theipi.org 864
www.freepsychotherapybooks.org
 previous chapter was limited to evidence about the

CCRT in comparison with Freud’s observations about

transference.

But in this section, I give in to a bit of boldness

by taking off on a speculative flight that allows more

perspective on the sources of the CCRT. Although the
flight will be over the mostly familiar terrain of
Freud’s observations about the transference versus
the CCRT evidence, from that great height one can

make out the contours of four or more formative and

maintenance factors for the central relationship
pattern and for that concrete version of it, the CCRT.

Source Factor 1: Learning the Pattern From Parental

Figures

A central relationship pattern is discernible very

early in a child’s development, and the pattern may

owe much to learning from the parental figures. The

earliest systematic examination of such a pattern has

been in narratives told by 3-year-old children who

www.theipi.org 865
www.freepsychotherapybooks.org
 have become 5-year-olds, reported in chapter 16, this
volume. Sroufe’s (1983) method discerned such

patterns even earlier through behavioral observations.

There is not yet much data on how early the central

relationship patterns begin as evaluated through

narratives and on the consistency of these patterns
through the later years (Kagan, 1996). The
expectation, on the basis of Freud’s observation 12

(see chapter 21), is that there is considerable

consistency of the pattern from early childhood until

late in life. So far, the CCRT evidence is based

mainly on the narratives told in adulthood that

contain the earliest memories. It has been found that

these are thematically consistent with narratives told

about current events.

The transference concept implies that the early

relationship patterns owe their start to the interactions

with the parents and other early caregivers and

clearly must become generalized to other people. In

www.theipi.org 866
www.freepsychotherapybooks.org
 fact, both Freud’s Observation 9, that the pattern is a
general one, and the similar CCRT findings

emphasize that the pattern is expressed to and

probably has generalized to a variety of other people.

Rereading the narratives from any of the sessions

about a variety of different people reminds one of this
fact (see chapter 5).

How the early learning of the pattern comes about

deserves more investigation. The pattern appears to
be acquired through four modes, the first three of

which follow: (a) Repeated experiences in

interactions with the parents set up expectations about

parental contingent responses (Stern, 1985; Tronick,
1982); (b) the parents directly teach some aspects of
the pattern; and (c) the child learns some aspects of
the pattern through identification with the parents. In

these modes of acquisition of the pattern, motives are

the wish to please as well as the wish to avoid

displeasing the parents; both wishes are widely

www.theipi.org 867
www.freepsychotherapybooks.org
 prevalent, as emphasized by Weiss, Sampson, and the

Mount Zion Psychotherapy Research Group (1986);

these wishes are based partly on love and partly on

fear and guilt. The three modes of acquisition appear
to be built on and interact with a fourth one, the more

biological capacities, such as for empathy, as shown

by Brothers (1989), and for temperament, as shown

by Kagan (1989).

Source Factor 2: Needing to Gratify Certain Pressing

Wishes

Each person’s central relationship pattern includes

the recurrent inclination to gratify certain wishes
toward other people and to the self. These wishes

have an urgency to be expressed and satisfied, given

half a chance, as stressed in George Klein’s (1970)

observations about “peremptory ideation.” The

persistence of such wishes is increased because of

hoped-for positive responses and fears of negative
responses from other people and from the self. The

www.theipi.org 868
www.freepsychotherapybooks.org
 findings about the pervasiveness of each person’s

central wishes (see chapter 10, this volume) indicate

that each person has certain high-frequency wishes

expressed in many kinds of relationship interactions.
For each of the three patients discussed in chapter 5,

the two highest frequency wishes were as follows:

For Ms. Smyth the two most urgent and related

wishes were to end nonsupportive relationships and

to get support and caring; for Mr. Howard the two

most pressing wishes were to be close and receive

affection; for Ms. Cunningham the two most
pervasive wishes were to be in control and to be
reassured. For each of these patients, these wishes

were both prominent initially and evident at the end

of treatment.

Source Factor 3: Repeating of Traumatic Ideas and

Scenes

After observing the strong inclination for the

repetition of traumatic ideas and scenes, Freud

www.theipi.org 869
www.freepsychotherapybooks.org
 (1920/1955a, 1914/1958e) hit on the descriptive and
yet explanatory concept of the “repetition

compulsion,” the need to repeat, reexperience, and

reenact traumatic memories. There are many reliable

and recurrent reports of observations that are

consistent with this concept, for example, the soldier

who has a close call with death and then repeatedly

dreams of the event. The concept is applicable to data

from a variety of investigators of the repetition of

traumatic scenes: Reiser (1984), Tomkins (1987), M.

Horowitz (1986), Loevinger (1976), and Marmar and
Horowitz (1988).

There are bases for repetition of themes that fit

with the concept of the repetition compulsion because
they imply an automatic component to the repetition.

One of these is repetition based on the presence of a

prior schema. Cognitive psychologists, such as Fiske

and Dyer (1985), have reported research findings

about the carryover of schemas; these findings show

www.theipi.org 870
www.freepsychotherapybooks.org
 that a schema, once learned, sets the stage for

interpreting later events in a similar way. The schema

gradually becomes more unconscious and
generalizes. The person shows some learning about
when to apply the schema and when not, but much of

the application to later experiences is automatic, as

noted in the even broader brain-based schema theory

reviewed by Arbiv (1995).

The repetition of themes may also be based on the

schema’s getting its start through emotion-laden

events. Emotion-laden events are especially prone to

repetition when events of a similar nature are

expected. There may even be an arrangement in the

brain such that emotion-laden memory is triggered

with a minimum of conscious control. The research

of Le Doux, Romanski, and Xagoraris (1989)

suggests that a focal point for cognition, the
hippocampus, can be involved in the activation of
emotions before cognitive processes take place. Their

www.theipi.org 871
www.freepsychotherapybooks.org
 research suggests that there exist alternative nerve
pathways leading from the thalamus to the amygdala
that deal with emotions without going through the
cortex. Their resultant point of view is that because
the original schema was emotion-laden, arousal and

repetition without conscious control can more readily

occur.

Source Factor 4: Repeating in the Service of Mastery

The inclination to repeat ideas or scenes that fit

into the pattern could serve as more than the need to

repeat the memory of traumatic events (Source Factor

3); the repetition could also be part of an effort to find

ways to master traumas (as suggested by Mayman,

1978). Competence in and control over one’s life and

relationship problems is a vital human need (White,

1952). The conclusion is consistent with the work of

Seligman (1980; and Seligman et al., 1984) on

explanatory style in response to negative events: A

www.theipi.org 872
www.freepsychotherapybooks.org
 pessimistic style is associated with decreased mastery
(helplessness) and consequent depressive symptoms.

The need to repeat in the service of gaining

mastery may be related to our findings about the

CCRT, specifically, the relative frequency of negative

versus positive components of the CCRT, noted in

relation to Freud’s observation about positive and
negative transference patterns (see chapters 4 and 21,
this volume). The existence of a higher frequency of
negative than positive components of the CCRT

could be fueling the person’s preoccupation with the

need to solve the negative, and therefore upsetting,

relationship conflicts. The negative components point

to where the source of threat lies, and much of the
content of the narratives can be thought of as an effort
to rehearse the event and find ways of coping with it.

In positive components there is less threat; they tend

to portray situations in which one’s coping has

worked out well. The shift from the negative to the

www.theipi.org 873
www.freepsychotherapybooks.org
 positive components of the CCRT during the course

of successful psychotherapy also could be taken to

imply that the level of masterful coping has increased
(see chapters 4 and 18).

Some evidence for the potency of the need for

mastery can also be found from studies of the effects

of interpretation on the transference (Observation 15).
Interpretations, especially of the negative
transference, tend to have the effect of changing the
pattern in the direction of greater mastery. From this

point of view, the effects of accurate interpretations

can be thought of as giving more support to the side

of the patient that is attempting to master the

traumatic aspects of the CCRT. Freud’s (1901-

1905/1953a) reflections about the premature

interruption of treatment by his patient Dora

contained his hypothesis that if he had interpreted
Dora’s negative transference with its fearful
expectation that his behavior toward her would be

www.theipi.org 874
www.freepsychotherapybooks.org
 like Herr K’s, she would have been able to continue

her therapy with Freud. Without the interpretation,

the patient was merely repeating and not mastering

her pattern. But through the interpretation, Freud

expected that she would have been able to see her

fear-inducing misidentification of Freud with Herr K
and, through this awareness, would have been able to
go beyond repetition to mastery.

In summary, my aim of launching on a wide-

ranging search for the origins of the CCRT and of the

transference template has brought out four powerful
sources and maintainers of the central self-other

relationship pattern. These four are the learning of the

pattern from the parents, the persistence of wishes to

gratify certain impulses, the need to continue to

repeat the traumatic parts of the pattern, and the need

to master the conflicts within the pattern. A fifth
factor must also play a significant role in maintaining
the pattern: the person’s wish to avoid risking

www.theipi.org 875
www.freepsychotherapybooks.org
 displeasing the parent or parent figure by altering the

old pattern because an alteration would hurt the

parent or the parent might retaliate. There is much
evidence for this factor; for example, Weiss,
Sampson, and the Mount Zion Psychotherapy Group

(1986) stressed the need to please the parent by

maintaining the pattern. Also, a related theory by

Benjamin (1994) stressed that development of the

symptom can serve to take account of the parent’s

needs; that is, it is based on consideration of the

parent’s feelings. A sixth factor may also have a
significant role: The primacy of a once-established
schema may have a part in maintaining the pattern in

its original form (Fiske & Dyer, 1985). And a seventh

factor may have a shaping role: a person’s birth order

(Sulloway, 1996). In fact, all of these sources and

maintainers of the schema may collaborate in starting

and then maintaining the pervasive expression of the

central relationship pattern as measured by the CCRT.

www.theipi.org 876
www.freepsychotherapybooks.org
 In further fact, as this exposition is completed, my

appetite to fill out this sketch of source factors

continues to grow; already the many factors have
begun to form the germinal cell of a personality
theory about the nature of the construction of the

inner and interpersonal world of each person

(Luborsky, 1997).

HOW MUCH DOES THE CCRT CORRESPOND

WITH THE TRANSFERENCE?

For the CCRT to fit in the same family as the

transference pattern requires that we reexamine the
CCRT’s validity. Validity is the extent that a measure

measures what it is supposed to measure. For the

CCRT measure, a useful expectation is that the CCRT

will be found to be associated with phenomena that

are meaningfully related to the concept of a

transference-like central relationship pattern
following the logic of the nature of clinical inference
as presented by Benjamin B. Rubinstein (Holt, 1997).

www.theipi.org 877
www.freepsychotherapybooks.org
 It was stated in chapter 21, this volume, that there is
no single definition of transference but rather a

variegated assemblage of at least 23 qualities

associated with the concept. A measure of validity,
therefore, is the number of correspondences between

these qualities and the CCRT evidence. The

following is a partial listing of some of the already

confirmed transference-based meaningful predictions

about the CCRT measure:

1. The CCRT should reveal a general pattern across

relationships with different types of people. In
fact, the CCRT is pervasive across narratives
about different other people and it maintains its
pervasiveness late in treatment as well (see
chapter 10).

2. The CCRT should have a prominent emotional

dimension. Such a dimension is strongly
represented by the positive and negative
dimension. In fact, that dimension is reliably
scorable and it changes during treatment in a
small, although clinically meaningful, way (see
chapter 4).

www.theipi.org 878
www.freepsychotherapybooks.org
 3. The CCRT, as implied by the results of Number 1
in this list, should show a parallel pattern for the
relationship with the therapist and for the
relationship with other people: A careful study
has shown that this is true (see chapter 11).

4. The CCRT should appear in similar form in

different modes of expression. In fact, the pattern
is similar in dreams reported in sessions and in
waking narratives (see chapter 12).

5. The CCRT should show changes in response to

interpretations so that the more fully the therapist
focuses on it in the interpretations, the more
benefit the patient might receive from the
therapy; in fact, my group found just that (see
chapter 13).

6. The CCRT should appear not only within

psychotherapy but outside of psychotherapy as
well. The only study that gives solid evidence for
this is one by Barber et al. (1995; see chapter
16), which shows a significant parallel of the
CCRT from RAP narratives told before the
treatment starts with the CCRT from narratives
told in the early sessions.

www.theipi.org 879
www.freepsychotherapybooks.org
 7. A CCRT-like pattern should appear in young
children, and it should show stability over time
(see chapter 16).

8. The CCRT should show some results in common

with other central relationship pattern measures.
In every instance in which this has been
examined, patterns were discovered for
components that were in common among the
different measures (see chapter 20).

9. The CCRT should behave in ways that fit Freud’s

observations that led him to his concept of the
transference template. In fact, when my
colleagues and I were able to translate Freud’s
observations into operational terms in the CCRT,
we found the parallel was confirmed; the trend so
far confirms a convergence of the majority of
Freud’s 23 observations about transference with
the CCRT evidence—at least for the 18
observations for which a translation seemed
possible (see chapter 21).

These confirmations should lead to more serious

attention in psychology, psychiatry, and social work
to Freud’s concept of the transference pattern. But it

www.theipi.org 880
www.freepsychotherapybooks.org
 is not clear to what degree this will happen. The

rejection and acceptance of psychoanalytic concepts

such as transference have a curious history. This

history may have parallels with the response to other

revolutionary concepts like Darwin’s theory of

evolution. As the embryologist Karl Ernst von Baer
noted (Gould, 1977), every major theory that wins
acceptance goes through three stages: It is first

condemned as untrue; it is then branded as against

religion; and ultimately it is dogmatically embraced.

The history of acceptance of the concept of
transference approximately fits these stages. It did go
through a stage in which it was considered untrue,
and some people have remained in that stage, caught
up in the current fashion of rejection of dynamic

concepts. There were some people who considered it

contrary to religion; they are not a very active

opposition now. Finally, for some psychoanalytic
practitioners it has become dogma; for some of these

www.theipi.org 881
www.freepsychotherapybooks.org
 people, it would even be considered a violation of the

concept to construct a measurement method for it.

But in time the evidence of the CCRT’s validity
reported in this book should gamer even greater
recognition of the usefulness of the concept and even

greater acceptance of the discoveries achieved

through its operational measurement. More and more

clinicians will then use the CCRT in treating their

patients, and more and more studies will examine it.

The inclination to accept the findings of the

research, at least for the dynamic audience, should be

fortified by the strategy of our research style. The

essential attribute of the style is its reliance on data

from psychoanalytic and psychotherapy sessions for

examining Freud’s basic observations about the

transference pattern. This examination of sessions

through operational measures contrasts with the style
of much past research on Freud’s theories, especially
that which relies on analogue or experimental

www.theipi.org 882
www.freepsychotherapybooks.org
 recreations of the phenomena suggested by Freudian

theory (as reviewed by Fisher & Greenberg, 1977,

1996); the obvious weakness of such analogues is
feebleness in showing that they in fact capture the
intended phenomena. The obvious strength of relying

on psychoanalytic and psychotherapy sessions is the

appropriateness of the database: It is the milieu from

which the basic clinical concepts such as transference

were generated, as stressed by Rapaport (1960),

Schlesinger (1974), Holzman (1985), and Eagle and

Wolitsky (1989).

CONCLUSIONS

• The essential narrative in this book is about the

successful translation of a key clinical concept
into a key clinical-quantitative operational
measure of the clinical concept. The concept is
that there is a schema for the central relationship
pattern that can be found in a series of narratives
told by persons about themselves in relation to
other people or in relation to themselves.

www.theipi.org 883
www.freepsychotherapybooks.org
 • The Core Conflictual Relationship Theme is the
reliable measure of this schema; the measure can
be reliably scored in a set of narratives from each
person.

• About a dozen areas for additional research that

will continue the maturation of the CCRT
method are listed in this chapter and chapter 21.

• A wealth of validity data implies that the Core

Conflictual Relationship Theme has much in
common with Freud’s concept of the transference
template, as reported in this chapter and in
chapter 21.

• I have taken a flier into speculation about the

factors that contribute to the origin and
maintenance of the CCRT: (a) the learning of the
pattern from parental figures, (b) the push of
persistent wishes, (c) the need for repetition of
the traumatic parts of the pattern, and (d) the
need to master conflicts. Three other factors also
have a role: (e) the fear of displeasing and the
wish to please parental figures, (f) the
confinement by the rut of the preestablished
schema, and (g) the influence of birth order.

www.theipi.org 884
www.freepsychotherapybooks.org
 • The CCRT method has a host of effective clinical
applications in the conduct of psychotherapy and
in diagnosis.

• Like all “better mouse trap” discoveries, the CCRT,

as discussed in this much revised volume, has
both contributed to our knowledge about the
central relationship pattern and raised an array of
new questions, some of which will soon be
answered through the more than 150 ongoing
CCRT studies that have spread worldwide in the
last 7 years.

www.theipi.org 885
www.freepsychotherapybooks.org
 REFERENCES

Ainsworth, M., Blehar, M., Waters, E., & Wall, S. (1978).

Patterns of attachment. Hillsdale, NJ: Erlbaum.

Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster

analysis. Beverly Hills, CA: Sage.

American Psychiatric Association. (1980). Diagnostic and

statistical manual of mental disorders (3rd ed.).
Washington, DC: Author.

American Psychiatric Association. (1990). Diagnostic and

statistical manual of mental disorders (4th ed.).
Washington, DC: Author.

Apfelbaum, B. (1958). Dimensions of transference in

psychotherapy. Berkeley: University of California
Press.

Arbiv, M. (1995). Schema theory. In M. Arbiv (Ed.), Brain

theory and neural networks (pp. 830-834).
Cambridge, MA: MIT Press.

Arlow, J. (1961). Ego psychology in the study of

mythology. Journal of the American Psychoanalytic
Association, 9, 371-393.

www.theipi.org 886
www.freepsychotherapybooks.org
 Arlow, J. (1969a). Fantasy, memory, and reality testing.
Psychoanalytic Quarterly, 38, 28-51.

Arlow, J. (1969b). Unconscious fantasy and disturbances

of conscious experience. Psychoanalytic Quarterly,
38, 1-27.

Arlow, J., & Brenner, C. (1964). Psychoanalytic concepts

and the structural theory. Madison, CT: International
Universities Press.

Aron, B. (1949). A manual for analyses of the Thematic

Apperception Test. Berkeley, CA: Berg.

Auerbach, A., & Childress, A. (1988). Diagnosis in

assessment for psychotherapy, Current Opinion in
Psychiatry, 1, 293-298.

Auerbach, A., & Luborsky, L. (1968). Accuracy of

judgments of psychotherapy and the nature of the
“good hour.” In J. Shlien, H. F. Hunt, J. P. Matarazzo,
& C. Savage (Eds.), Research in psychotherapy (Vol.
3, pp. 155-168). Washington, DC: American
Psychological Association.

Baguet, J., Gerin, P., Sali, M., & Marie-Cardine, M.

(1984). Evolution des themes transferentials
individuels dans une psychotherapie de groupe
[Evolution of individual transferential themes in a
psychotherapy group]. Psychotherapies, 2, 43-49.

www.theipi.org 887
www.freepsychotherapybooks.org
 Barber, J. P. (1989). Analyzing CCRT sequences for its
most frequent wishes. Center for Psychotherapy
Research, University of Pennsylvania. Unpublished
manuscript.

Barber, J. P. (1993). The Central Relationship Pattern

Questionnaire. Unpublished Version 4-5. University
of Pennsylvania Medical School, Philadelphia.

Barber, J. P., & Crits-Christoph, P. (1993). Advances in

measures of psychodynamic formulations. Journal of
Consulting and Clinical Psychology, 61, 574-585.

Barber, J. P., Crits-Christoph, P., & Luborsky, L. (1990). A

guide to the standard categories and their
classification. In L. Luborsky & P. Crits-Christoph
(Eds.), Understanding transference: The CCRT
method (pp. 37-50). New York: Basic Books.

Barber, J. P., Crits-Christoph, P., & Luborsky, L. (1996).

Effects of therapist adherence and competence on
patient outcome in brief dynamic therapy. Journal of
Consulting and Clinical Psychology, 64, 619-622.

Barber, J. P., Luborsky, L., Crits-Christoph, P., & Diguer,

L. (1995). A comparison of core conflictual
relationship themes before psychotherapy and during
early sessions. Journal of Consulting and Clinical
Psychology, 63, 145-148.

www.theipi.org 888
www.freepsychotherapybooks.org
 Bargh, J. A., Chaiken, S., Govender, R., & Felicia, P.
(1992). The generality of the automatic attitude
activation effect. Journal of Personality and Social
Psychology, 62, 893-912.

Battle, C., Imber, S., Hoehn-Saric, R., Stone, A. R., Nash,

C., & Frank, J. D. (1966). Target complaints as
criteria of improvement. American Journal of
Psychotherapy, 20, 184-192.

Beck, A. T. (1967). Depression: Clinical, experimental,

and theoretical aspects. New York: Hober.

Beck, A. T. (1970). Depression: Clinical, experimental,

and theoretical aspects (Rev. ed.). New York: Hober.

Beck, A. T. (1971). Cognitive patterns in dreams and

daydreams. In J. Masserman (Ed.), Science and
psychoanalysis: Vol. 19. Dream dynamics. New York:
Grune & Stratton.

Beck, A. T. (1989). Cognitive therapy for depression and

panic disorder. Western Journal of Medicine, 151, 9-
89.

Beebe, B., & Lachmann, F. (1988). Mother-infant mutual

influence and precursors of psychic structure. In A.
Goldberg (Ed.), Frontiers of self-psychology:
Progress in self-psychology (Vol. 3, pp. 3-26).
Hillsdale, NJ: Analytic Press.

www.theipi.org 889
www.freepsychotherapybooks.org
 Beliak, L. (1954). The Thematic Apperception Test and the
Children's Apperception Test in clinical use. New
York: Grune and Stratton.

Benjamin, L. S. (1974). Structural analysis of social

behavior. Psychological Review, 81, 392-425.

Benjamin, L. S. (1977). Structural analysis of a family in

therapy. Journal of Consulting and Clinical
Psychology, 45, 391-406.

Benjamin, L. S. (1979). Use of structural analysis of social

behavior (SASB) and Markov chains to study
dynamic interactions. Journal of Abnormal
Psychology, 88, 303-319.

Benjamin, L. S. (1982). Use of structural analysis of social

behavior (SASB) to guide intervention in
psychotherapy. In D. Kiesler & J. Anchin (Eds.),
Handbook of interpersonal psychotherapy. New
York: Pergamon Press.

Benjamin, L. S. (1986a). Adding social and intrapsychic

descriptors to Axis I of DSM-III. In T. Millon & G.
Klerman (Eds.), Contemporary directions in
psychopathology (pp. 599-638). New York: Guilford
Press.

Benjamin, L. S. (1986b). Operational definition and

measure of dynamics shown in the stream of free
associations. Psychiatry, 49, 104-129.

www.theipi.org 890
www.freepsychotherapybooks.org
 Benjamin, L. S. (1994). SASB: A bridge between
personality theory and clinical psychology.
Psychological Inquiry, 5, 273-316.

Bernal, G., & Baker, J. (1979). Toward a

metacommunicational framework of couple
interactions. Family Process, 18, 293-302.

Bibring, E. (1954). Psychoanalysis and the dynamic

psychotherapies. Journal of the American
Psychoanalytic Association, 2, 745-770.

Block, J. (1971). Lives through time. Berkeley, CA:

Bancroft Books.

Block, J. (1977). Advancing the psychology of

personality: Paradigmatic shift for improving the
quality of research. In D. Magnussen & N. S. Endler
(Eds.), Personality at the crossroads (pp. 37-63).
New York: Wiley.

Blos, P. (1941). The adolescent personality: A study of

individual behavior for the commission on secondary
school curriculum. New York: Appleton-Century-
Crofts.

Bonanno, G., & Singer, J. (1990). Repressive personality

style: Theoretical and methodological implications
for health and pathology (pp. 435-470). In J. Singer
(Ed.), Repression and dissociation: Defense

www.theipi.org 891
www.freepsychotherapybooks.org
 mechanisms and personality styles. Chicago:
University of Chicago Press.

Bond, J., Hansell, J., & Shevrin, H. (1987). Locating a

reference paradigm in psychotherapy transcripts:
Reliability of relationship episode location in the
Core Conflictual Relationship Theme (CCRT)
method. Psychotherapy, 24, 736-749.

Bond, J., & Shevrin, H. (1986a). The Clinical Evaluation

Team method. Unpublished manuscript.

Bond, J., & Shevrin, H. (1986b). Similarities and

differences between two methods of formulating a
patient’s psychodynamically relevant schemata.
Unpublished manuscript, University of Michigan,
Ann Arbor.

Book, H. (in press). How to practice brief psychodynamic

psychotherapy: The CCRT method. Washington, DC:
American Psychological Association.

Bowlby, J. (1969). Attachment and loss: Vol. I.

Attachment. New York: Basic Books.

Bowlby, J. (1973). Attachment and loss: Vol. 2.

Separation. New York: Basic Books.

Bretherton, I. (1995). A communication perspective on

attachment relationships and internal working

www.theipi.org 892
www.freepsychotherapybooks.org
 models. Monographs of the Society for Research in
Child Development, 60(2-3), 310-329.

Bretherton, I., Prentiss, C., & Ridgeway, D. (1990).

Family relationships as represented in a story-
completion task at thirty-seven and fifty-four months
of age. In J. Bretherton & M. W. Watson (Eds.),
Children’s perspective on the family: New directions
for child development (Vol. 48, pp. 85-105). San
Francisco: Jossey-Bass.

Bretherton, I., Ridgeway, D., & Cassidy, J. (1990).

Assessing internal working models of the attachment
relationship—An attachment story completion task
for 3-year-olds. In M. Greenberg, D. Cichetti, & E.
M. Cummings (Eds.), Attachment during the pre-
school years (pp. 273-308). Chicago: University of
Chicago Press.

Brothers, L. (1989). A biological perspective on empathy.

American Journal of Psychiatry, 146, 10-19.

Bruner, J. (1987). Actual minds, possible worlds.

Cambridge, MA: Harvard University Press.

Bucci, W. (1985). Dual coding: A cognitive model for

psychoanalytic research. Journal of the American
Psychoanalytic Association, 33, 571-607.

Buchsbaum, H., & Emde, R. (1990). Play narratives in 36-

month-old children: The portrayal of early moral

www.theipi.org 893
www.freepsychotherapybooks.org
 development and family relationships. In A. J. Solnit,
P. Newbauer, S. Abrams, & A. S. Dowling (Eds.),
The psychoanalytic study of the child (Vol. 45, pp.
129-155) New Haven, CT: Yale University Press.

Carlson, R. (1981). Studies in script theory: I. Adult

analogs of a childhood nuclear scene. Journal of
Personality and Social Psychology, 40, 501-510.

Carlson, R. (1986). After analysis: A study of transference

dreams following treatment. Journal of Consulting
and Clinical Psychology, 54, 246-252.

Caspar, F. (1989). Einfuhrung in die Plananalyse

[Introduction to plan analysis]. Berne, Switzerland:
Huber.

Caspar, F. (1995). Plan analysis: Toward optimizing

psychotherapy. Seattle, WA: Hogrefe & Huber.

Caston, J. (1977). Manual on how to diagnose the plan. In

J. Weiss, H. Sampson, J. Caston, & G. Silberschatz
(Eds.), Research on the psychoanalytic process: Vol.
1. A comparison of two theories about analytic
neutrality (pp. 15-21). San Francisco: Psychotherapy
Research Group, Department of Psychiatry, Mount
Zion Hospital and Medical Center.

Caston, J. (1986). The reliability of the diagnosis of the

patient’s unconscious plan. In J. Weiss, H. Sampson,
& the Mount Zion Psychotherapy Research Group

www.theipi.org 894
www.freepsychotherapybooks.org
 (Eds.), The psychoanalytic process: Theory, clinical
observations, and empirical research (pp. 241-255).
New York: Guilford Press.

Chance, E. (1952). The study of transference in group

therapy. International Journal of Group Therapy, 2,
40-53.

Chapman, S. (1980). Story and discourse. Ithaca, NY:

Cornell University Press.

Cierpka, M., Zander, B., Krannich, S., Reich, G., Ratzke,

K., Hombburg, H., Staats, H., & Seide, L. (1992,
June). Differences in conflictual relationship themes
of male and female students. Paper presented at the
annual meeting of the Society for Psychotherapy
Research, Berkeley, CA.

Cohen, J. (1960). A coefficient of agreement for nominal

scales. Educational and Psychological Measurement,
20, 37-46.

Cohen, J. (1968). Weighted kappa: Nominal scale

agreement with provision for scaled disagreement on
partial credit. Psychological Bulletin, 70, 213-220.

Cohen, J., & Cohen, P. (1975). Applied multiple

regression/correlational analysis for the behavioral
sciences. Hillsdale, NJ: Erlbaum.

www.theipi.org 895
www.freepsychotherapybooks.org
 Collins, W., & Messer, S. (1988, June). Transporting the
Plan Diagnosis method to a different setting.
Reliability, stability, and adaptability. Paper
presented at the annual conference of the Society for
Psychotherapy Research, Santa Fe, NM.

Connolly, M. B., Crits-Christoph, P., Demorest, A.,

Azarian, K., Muenz, L., & Chittams, J. (1996). The
varieties of transference patterns in psychotherapy.
Journal of Consulting and Clinical Psychology, 64,
1213-1221.

Crisp, A. (1964a). An attempt to measure an aspect of

transference. British Journal of Medical Psychology,
37, 17-30.

Crisp, A. (1964b). Development and application of a

measure of transference. Journal of Psychosomatic
Research, 8, 327-335.

Crisp, A. (1996). Transference symptom emergence and

social repercussion in behavior therapy: A study of
54 treated patients. British Journal of Medical
Psychology, 39, 179-196.

Crits-Christoph, P. (1986). Assessing conscious and

unconscious aspects of relationship themes from self-
report and naturalistic data. Paper presented at the
MacArthur Workshop on Person Schemata, Palo Alto,
CA.

www.theipi.org 896
www.freepsychotherapybooks.org
 Crits-Christoph, P. (1987). Development of a set of rating
scales to assess CCRT dimensions of wishes,
responses from other and responses of self.
Unpublished manuscript.

Crits-Christoph, P., Barber, J. P., & Kurcias, J. S. (1993).

The accuracy of therapists’ interpretations and the
development of the therapeutic alliance.
Psychotherapy Research, 3, 25-35.

Crits-Christoph, P., Barber, J. P., Miller, N., & Beebe, K.

(1993). Evaluating insight. In N. Miller, L. Luborsky,
J. P. Barber, & J. Docherty (Eds.), Psychodynamic
treatment research: A handbook for clinical practice
(pp. 407-422). New York: Basic Books.

Crits-Christoph, P., Cooper, A., & Luborsky, L. (1988).

The accuracy of therapists’ interpretations and the
outcome of dynamic psychotherapy. Journal of
Consulting and Clinical Psychology, 56, 490-495.

Crits-Christoph, P., & Demorest, A. (1988). List of

standard categories (Edition 2). Unpublished
manuscript, University Pennsylvania School of
Medicine, Philadelphia.

Crits-Christoph, P., St Demorest, A. (1991). Quantitative

assessment of relationship theme components. In M.
Horowitz (Ed.), Person schemas and maladaptive

www.theipi.org 897
www.freepsychotherapybooks.org
 interpersonal behavior. Chicago: University of
Chicago Press.

Crits-Christoph, P., Demorest, A., & Connolly, M. B.

(1990). Quantitative assessment of interpersonal
themes over the course of a psychotherapy.
Psychotherapy, 27, 513-521.

Crits-Christoph, P., Demorest, A., Muenz, L., &

Baranackie, K. (1994). Consistency of interpersonal
themes for patients in psychotherapy. Journal of
Personality, 62, 499-526.

Crits-Christoph, P., & Luborsky, L. (1990). Changes in

CCRT pervasiveness during psychotherapy. In L.
Luborsky & P. Crits-Christoph (Eds.), Understanding
transference: The CCRT method (pp. 133-146). New
York: Basic Books.

Crits-Christoph, P., Luborsky, L., Dahl, L., Popp, C.,

Mellon, J., & Mark, D. (1988). Clinicians can agree
in assessing relationship patterns in psychotherapy:
The Core Conflictual Relationship Theme Method.
Archives of General Psychiatry, 45, 1001-1004-

Crits-Christoph, P., Luborsky, L., Popp, C., Mellon, J., &

Mark, D. (1990). The reliability of choice of
narratives and of the CCRT measure. In L. Luborsky
& P. Crits-Christoph (Eds.), Understanding

www.theipi.org 898
www.freepsychotherapybooks.org
 transference: The CCRT method (pp. 93-101). New
York: Basic Books.

Curtis, H. (1983, May). Toward a metapsychology of

transference. Paper presented to the annual meeting
of the American Psychoanalytic Association, New
York.

Curtis, J. T., & Silberschatz, G. (1989). The Plan

Formulation method: A reliable procedure for case
formulation. Manuscript submitted for publication.

Curtis, J. T., Silberschatz, G., Sampson, H., Weiss, J., &

Rosenberg, S. E. (1988). Developing reliable
psychodynamic case formulations: An illustration of
the Plan Diagnosis method. Psychotherapy, 25, 256-
265.

Dahl, H. (1978). A new psychoanalytic model of

motivation: Emotions as appetites and messages.
Psychoanalysis and Contemporary Thought, 1, 373-
408.

Dahl, H. (1988). Frames of mind. In H. Dahl, H. Kächele,

& H. Thomae (Eds.), Psychoanalytic process
research strategies (pp. 51-66). New York: Springer-
Verlag.

Dahl, H., Kächele, H., & Thomae, H. (Eds.). (1988).

Psychoanalytic process research strategies. New
York: Springer-Verlag.

www.theipi.org 899
www.freepsychotherapybooks.org
 Dahl, H., & Teller, V. (1984). The characteristics,
identification, and application of frames.
Psychotherapy Research, 4, 253-276.

Dahl, H., & Teller, V. (1993). Characteristics and

identification of frames. In N. Miller, L. Luborsky, St
J. Docherty (Eds.), Doing research on psychodynamic
therapy. New York: Basic Books.

Dahlbender, R. (1992, June). Intra- and inter-subjectivity

in RAP interviews of young women. Paper given at
the annual meeting of the Society for Psychotherapy
Research, Berkeley, CA.

Dahlbender, R., Albani, C., Pokomy, D., & Kachele, H. (in

press). The central relationship pattern (CRP): A
structural version of the CCRT. Psychotherapy
Research.

Dahlbender, R., Kurth, R., Stubner, S., Kalmykova, K., &

Porkomy, D. (in press). Das zentrale
beziehungskonflikt thema in der selbst-beurteilung:
Der zentrale Beziehungs-muster (ZBM) Fragebogen
[The central relationship conflict theme in self-
assessment: The Central Relationship Pattern
Questionnaire]. Gruppenpsychotherapie und
Gruppendynamik.

Dahlbender, R., Volkert, M., Torres, L., Pokomy, D.,

Frevert, G., Reichert, S., & Kachele, H. (1992). Intra-

www.theipi.org 900
www.freepsychotherapybooks.org
 and inter-subjectivity in Relationship Anecdotes
Paradigm (RAP) interviews. Paper presented at the
annual meeting of the Society for Psychotherapy
Research, Berkeley, CA.

Davanloo, H. (1978). Basic principles and techniques in

short-term dynamic psychotherapy. New York:
Spectrum.

Davanloo, H. (1980). A method of short-term dynamic

psychotherapy. In H. Davanloo (Ed.), Short-term
dynamic psychotherapy. New York: Jason Aronson.

Davies, J. (1989). The development of emotional and

interpersonal structures in three-year-old children.
Unpublished doctoral dissertation, Demer Institute
for Advanced Psychological Studies, Adelphi
University, Garden City, NJ.

Demorest, A. P., & Alexander, I. E. (1992). Affective

scripts as organizers of personal experience. Journal
of Personality, 60, 645-663.

Demorest, A. P., & Crits-Christoph, P. (1989). A

quantitative comparison of interpersonal themes in
fantasy versus reality. Paper presented at the meeting
of the Society for Psychotherapy Research, Toronto,
Canada.

Demorest, A. P., Crits-Christoph, P., Hatch, M., &

Luborsky, L. (1997). A comparison of interpersonal

www.theipi.org 901
www.freepsychotherapybooks.org
 scripts in clinically depressed versus non-depressed
individuals. Manuscript submitted for publication.

Demorest, A. P., & Siegel, P. F. (1996). Personal

influences on professional work: An empirical case
study of B. F. Skinner. Journal of Personality, 64,
243-261.

Demos, V. (Ed.). (1995). Exploring affect. Cambridge,

England: Cambridge University Press.

Dengler, D. (1990). Anwendung des zentralen

beziehungskonfliktsthemas auf narrative von
dreijahrigen und suche riach zusammenhangen mit
fahigheit zur problemsldsung [Use of central
relationship themes from narratives of 3-year-olds
and a search for relationship with ability in problem
solving]. Dissertation zu Erlangung des Doktorgrades
der Medizin der Facultat fur Klinische Medizin der
Universitat Ulm.

Derogatis, L. R., Lipman, R. S., Covi, L., Rickels, K., &

Uhlenhuth, E. H. (1970). Dimensions of outpatient
neurotic pathology: Comparison of a clinical vs. an
empirical assessment. Journal of Consulting and
Clinical Psychology, 34, 164-171.

Derrida, J. (1977). Of grammatology. Baltimore: Johns

Hopkins University Press.

www.theipi.org 902
www.freepsychotherapybooks.org
 DeWitt, K. N., Kaltreider, N., Weiss, D. S., & Horowitz,
M. J. (1983). Judging change in psychotherapy:
Reliability of clinical formulations. Archives of
General Psychiatry, 40, 1121-1128.

Diguer, L., Barber, J. P., & Luborsky, L. (1993). Three

concomitants: Personality disorder, psychiatric
severity, and outcome of dynamic psychotherapy of
major depression. American Journal of Psychiatry,
150, 1246-1248.

Eagle, M. N. (1983). Recent developments in

psychoanalysis. In R. S. Cohen & L. Laudan (Eds.),
Physics, philosophy, and psychoanalysis (pp. 31-56).
Dordrecht, The Netherlands: D. Reidel.

Eagle, M. N., & Wolitsky, D. (1989). The idea of progress

in psychoanalysis. Psychoanalysis and Contemporary
Thought, 12, 27-72.

Eckert, R., Luborsky, L., Barber, J. P., & Crits-Christoph,

P. (1990). The CCRT in patients with major
depression. In L. Luborsky & P. Crits-Christoph
(Eds.), Understanding transference: The CCRT
method (pp. 222-234). New York: Basic Books.

Edelson, M. (1984). Hypothesis and evidence in

psychoanalysis. Chicago: University of Chicago
Press.

www.theipi.org 903
www.freepsychotherapybooks.org
 Edelson, M. (1988). Psychoanalysis: A theory in crisis.
Chicago: University of Chicago Press.

Eells, T. D., Horowitz, M., Singer, J., Salovey, P., Daigle,

D., & Turvey, C. (1995). The role relationship models
method: A comparison of independently derived case
formulations. Psychotherapy Research, 5, 161-175.

Ekstein, R. (1956). Psychoanalytic techniques. In D.

Brower & L. E. Abt (Eds.), Progress in clinical
psychology (Vol. 2, pp. 79-99). New York: Grune &
Stratton.

Endicott, J., & Spitzer, R. L. (1978). A diagnostic

interview: The Schedule for Affective Disorders and
Schizophrenia. Archives of General Psychiatry, 35,
837-844.

Engel, G., Jr., & Reichsman, F. (1956). Spontaneous and

experimentally induced depression in an infant with
gastric fistula: A contribution to the problem of
depression. Journal of the American Psychoanalytic
Association, 4, 428-452.

Erikson, E. H. (1959). Identity and the life cycle.

Psychological Issues (Monograph No. 1).

Fiedler, F., & Senior, K. (1952). An exploratory study of

unconscious feeling reactions in fifteen patient-
therapist pairs. Journal of Abnormal and Social
Psychology, 47, 446-453.

www.theipi.org 904
www.freepsychotherapybooks.org
 Fisher, S., & Greenberg, R. (1977). The scientific
credibility of Freud's theories and therapy. New
York: Basic Books.

Fisher, S., & Greenberg, R. (1996). Freud scientifically

reappraised: Testing the theories and the therapy.
New York: Wiley.

Fiske, S., & Dyer, L. (1985). Structure and development of

social schemata: Evidence for positive and negative
transfer effects. Journal of Personality and Social
Psychology, 48, 839-852.

Frank, J. (1979). Present status of outcome studies.

Journal of Consulting and Clinical Psychology, 47,
310-316.

French, T., & Wheeler, D. R. (1963). Hope and

repudiation of hope in psychoanalytic therapy.
International Journal of Psychoanalysis, 44, 304-
316.

Freud, S. (1953a). Fragment of an analysis of a case of

hysteria. In J. Strachey (Ed. and Trans.), The standard
edition of the complete psychological works of
Sigmund Freud (Vol. 7, pp. 15-122). London:
Hogarth Press. (Original work published 1901-1905)

Freud, S. (1953b). The interpretation of dreams. In J.

Strachey (Ed. and Trans.), The standard edition of the
complete psychological works of Sigmund Freud

www.theipi.org 905
www.freepsychotherapybooks.org
 (Vols. 4 and 5). London: Hogarth Press. (Original
work published 1900)

Freud, S. (1955a). Beyond the pleasure principle. In J.

Strachey (Ed. and Trans.), The standard edition of the
complete psychological works of S. Freud (Vol. 18,
pp. 7-64). London: Hogarth Press. (Original work
published 1920)

Freud, S. (1955b). Psychotherapy of hysteria. In J. Breuer

& S. Freud, Studies on hysteria. In J. Strachey (Ed.
and Trans.), The standard edition of the complete
psychological works of Sigmund Freud (Vol. 2, pp.
255-305). London: Hogarth Press. (Original work
published 1895)

Freud, S. (1955c). Studies on hysteria. In J. Strachey (Ed.

and Trans.), The standard edition of the complete
psychological works of Sigmund Freud (Vol. 2., pp.
3-305). London: Hogarth Press. (Original work
published 1893-1895)

Freud, S. (1958a). The dynamics of the transference. In J.

Strachey (Ed. and Trans.), The standard edition of the
complete psychological works of Sigmund Freud
(Vol. 12, pp. 99-108). London: Hogarth Press.
(Original work published 1912)

Freud, S. (1958b). Observations on transference-love. In J.

Strachey (Ed. and Trans.), The standard edition of the

www.theipi.org 906
www.freepsychotherapybooks.org
 complete psychological works of Sigmund Freud
(Vol. 16, pp. 431-447). London: Hogarth Press.
(Original work published 1915)

Freud, S. (1958c). On beginning the treatment (further

recommendations on the technique of psycho-
analysis). In J. Strachey (Ed. and Trans.), The
standard edition of the complete psychological works
of Sigmund Freud (Vol. 12, pp. 121-144). London:
Hogarth Press. (Original work published 1913)

Freud, S. (1958d). Recommendations to physicians

practicing psychoanalysis. In J. Strachey (Ed. and
Trans.), The standard edition of the complete
psychological works of Sigmund Freud (Vol. 12, pp.
111-120). London: Hogarth Press. (Original work
published 1912)

Freud, S. (1958e). Remembering, repeating and working

through. In J. Strachey (Ed. and Trans.), The standard
edition of the complete psychological works of
Sigmund Freud (Vol. 12, 145-156). London: Hogarth
Press. (Original work published 1914)

Freud, S. (1960). Jokes and their relation to the

unconscious. In J. Strachey (Ed. and Trans.), The
standard edition of the complete psychological works
of Sigmund Freud (Vol. 8, pp. 9-258). London:
Hogarth Press. (Original work published 1905)

www.theipi.org 907
www.freepsychotherapybooks.org
 Freud, S. (1964). Analyses terminable and interminable. In
J. Strachey (Ed. and Trans.), The standard edition of
the complete psychological works of Sigmund Freud
(Vol. 23, pp. 216-253). London: Hogarth Press.
(Original work published 1937)

Fried, D. (1989). The benefits of the CCRT for training in

psychotherapy: A survey of practitioners.
Unpublished manuscript.

Fried, D., Crits-Christoph, P., & Luborsky, L. (1990). The

parallel of narratives about the therapist with the
CCRT for other people. In L. Luborsky & P. Crits-
Christoph, Understanding transference: The CCRT
method (pp. 147–157). New York: Basic Books.

Fried, D., Luborsky, L., & Crits-Christoph, P. (1992). The

first empirical demonstration of transference in
psychotherapy. Journal of Nervous and Mental
Disease, 180, 326-331.

Friedman, S. H., & Luborsky, L. (1996). Enhancing

agreement of dynamic formulations: Comparison
before versus after training in the Core Conflictual
Relationship Theme (CCRT) method. Unpublished
manuscript.

Gabbard, G., Horowitz, L., Frieswyk, S., Allen, J., Colson,

M. D., Newsome, G., & Coyne, J. L. (1988). The
effect of therapist interventions on the therapeutic

www.theipi.org 908
www.freepsychotherapybooks.org
 alliance with borderline patients. Journal of the
American Psychoanalytic Association, 36, 697-727.

Gedo, J. (1979). Theories of object relations: A

metapsychological assessment. Journal of the
American Psychoanalytic Association, 27, 361-373.

Gedo, R (1993). A micro analytic method for tracing

transference references: Gill and Hoffman’s PERT
coding scheme and its elaborations. In N. Miller, L.
Luborsky, & J. Docherty (Eds.), Doing research on
psychodynamic therapy. New York: Basic Books.

George, C., Kaplan, N., & Main, M. (1985). The

attachment interview for adults. Unpublished
manuscript, University of California, Berkeley.

Ghannam, J. (1987, June). Representation of self and other

from clinical transcripts: A quantitative approach for
characterizing schemas. Paper presented at the
meeting of the Society for Psychotherapy Research,
Ulm, West Germany.

Gill, M. (1982). The analysis of transference: Vol. 1,

Theory and technique. Psychological Issues
(Monograph No. 54).

Gill, M., & Hoffman, I. (1982a). Analysis of transference:

Studies of nine audio-recorded psychoanalysis
sessions. Psychological Issues (Monograph No. 54).

www.theipi.org 909
www.freepsychotherapybooks.org
 Gill, M., & Hoffman, I. (1982b). A method for studying
the analysis of aspects of the patient’s experience of
the relationship in psychoanalysis and psychotherapy.
Journal of the American Psychoanalytic Association,
30, 137-167.

Gilligan, C. (1982). In a different voice: Psychological

theory and women's development. Cambridge, MA.
Harvard University Press.

Goldhirsh, M. (1961). Manifest content of dreams of

convicted sex offenders. Journal of Abnormal and
Social Psychology, 63, 643-645.

Gottschalk, L. A. (1985). How to understand and analyze

your own dreams. Corona Del Mar, CA: Art
Reproductions Press.

Gottschalk, L. A., & Gleser, G. (1969). The measurement

of psychological states through the content analysis
of verbal behavior. Berkeley: University of California
Press.

Gottschalk, L. A., Winget, C. N., & Gleser, G. C. (1969).

Manual instructions for using the Gottschalk-Gleser
Content Analysis Scale: Anxiety, hostility, and social
alienation—Personal disorganization. Los Angeles:
University of California Press.

Gould, S. J. (1977). Ever since Darwin: Reflections in

natural history. New York: Norton.

www.theipi.org 910
www.freepsychotherapybooks.org
 Graff, H., & Luborsky, L. (1977). Long-term trends in
transference and resistance: A quantitative analytic
method applied to four psychoanalyses. Journal of
the American Psychoanalytic Association, 25, 471-
490.

Grawe, K., & Caspar, F. (1984). Die Plan Analyse als

Konzept und Instrument fur die Psychotherapie
Forschung [Plan analysis as a concept and instalment
for psychoanalytic research]. In U. Bauman (Ed.),
Psychotherapie: Makro-und mikro Prespectiven (pp.
177-197). Gottingen, Germany: Hogrefe.

Grawe, K., Caspar, F., & Ambühl, H. (1990). Die Berner

therapie-vergleichs studie [The Berner comparisons
of therapies study]. Zeitschrift fur Klinische
Psychologie, 19,362-376.

Greenacre, P. (1954). The role of transference. Journal of

the American Psychoanalytic Association, 2, 671
-684.

Greenberg, R. (1987). On the importance of wearing two

hats: Dream researcher and psychoanalyst. Journal of
the American Academy of Psychoanalysis, 15, 321-
330.

Greenberg, R., & Pearlman, C. (1975). A psychoanalytic

continuum: The source and function of dreams.
International Review of Psychoanalysis, 2, 441-448.

www.theipi.org 911
www.freepsychotherapybooks.org
 Grenyer, B. F. (1994). Mastery Scale I: A research and
scoring manual. Wollongong, Australia: University
of Woolongong.

Grenyer, B. F., & Luborsky, L. (1996). Dynamic change in

psychotherapy: Mastery of interpersonal conflicts.
Journal of Consulting and Clinical Psychology, 64,
411-416.

Grunbaum, A. (1984). The foundations of psychoanalysis:

A philosophical critique. Berkeley: University of
California Press.

Grunbaum, A. (1986). Precis of The Foundations of

Psychoanalysis: A Philosophical Critique.
Behavioral and Brain Sciences, 9, 217-228.

Guitar-Amsterdamer, H., Stahli, R., Schneider, H., &

Berger, E. (1988). Konnen komponenten konfliktiver
beziehungsmuster in einem psychotherapeutischen
gesprach reliabel identiziert werden? [Can
components of conflictual relationships in a
psychotherapy session be reliably identifed?]. In L.
Luborsky & H. Kachele (Eds.), Der Zentrale
Beziehungskonflict—Ein Arbeitsbuch [The CCRT —A
workbook] (pp. 60-78). Ulm, Germany: PSZ
Publications.

Gustafson, J. (1986). The complex secret of brief

psychotherapy. New York: Norton.

www.theipi.org 912
www.freepsychotherapybooks.org
 Hall, C. S., & Van de Castle, R. L. (1966). The content
analysis of dreams. New York: Meredith.

Heider, F. (1958). The psychology of interpersonal

relations. New York: Wiley.

Henry, W. P., Schacht, T. E., (St Strupp, H. H. (1986).

Structural analysis of social behavior: Application to
a study of interpersonal process in differential
psychotherapeutic outcome. Journal of Consulting
and Clinical Psychology, 54, 27-31.

Hobson, J. A., St Stickgold, R. (1994). Dreaming: A

neurocognitive approach. Consciousness and
Cognition, 3, 1-15.

Hoffman, I. The patient as interpreter of the analyst’s

experience. Contemporary Psychoanalysis, 19, 389-
422.

Hoffman, I., & Gill, M. (1988a). Critical reflections on a

coding scheme. International Journal of
Psychoanalysis, 69, 55-64.

Hoffman, I., & Gill, M. (1988b). A scheme for coding the

patient’s experience of the relationship with the
therapist (PERT): Some applications, extensions, and
comparisons. In H. Dahl, H. Kächele, & H. Thomae
(Eds.), Psychoanalytic process research strategies.
New York: Springer-Verlag.

www.theipi.org 913
www.freepsychotherapybooks.org
 Holt, R. R. (1965). Freud’s cognitive style. American
Imago, 22, 163-179.

Holt, R. R. (1967). Diagnostic testing: Present situation

and future prospects. Journal of Nervous and Mental
Disease, 144, 444-465.

Holt, R. R. (1978). Methods in clinical psychology: Vol. 2.

Prediction and research. New York: Plenum Press.

Holt, R. R. (1989). Freud reappraised. New York:

Guilford Press.

Holt, R. R. (Ed.). (1997). Psychoanalysis and the

philosophy of science—Collected papers of Benjamin
B. Rubinstein. New York: Guilford Press.

Holzman, P. (1985). Psychoanalysis: Is the therapy

destroying the science? Journal of the American
Psychoanalytic Association, 33, 725-770.

Horowitz, L. M., Rosenberg, S. E., Baer, B. A., Ureno, G.,

& Villasenor, V. S. (1988). Inventory of interpersonal
problems: Psychometric properties and clinical
applications. Journal of Consulting and Clinical
Psychology, 56, 885-892.

Horowitz, L. M., Rosenberg, S., Ureno, G., Kalehzan, B.,

& O’Halloran, P. (1989). Psychodynamic
formulation: Consensual Response Method and

www.theipi.org 914
www.freepsychotherapybooks.org
 interpersonal problems. Journal of Consulting and
Clinical Psychology, 57, 599-606.

Horowitz, L. M., & Vitkus, J. (1986). The interpersonal

basis of psychiatric symptoms. Clinical Psychology
Review, 6, 443-469.

Horowitz, L. M., Weckler, D., & Doren, R. (1983).

Interpersonal problems and symptoms: A cognitive
approach. In P. C. Kendall (Ed.), Advances in
cognitive-behavioral research and therapy (Vol. 2,
pp. 82-127). New York: Academic Press.

Horowitz, M. J. (1979). States of mind: Analysis of change

in psychotherapy. New York: Plenum Press.

Horowitz, M. J. (1986). Stress response syndromes.

Northvale, NJ: Jason Aronson.

Horowitz, M. J. (1987). States of mind: Analysis of change

in psychotherapy (2nd ed.). New York: Plenum Press.

Horowitz, M. J. (1989). Relationship schema formulation:

Role relationship models and intrapsychic conflict.
Psychiatry, 52, 260-274-

Horowitz, M. J. (Ed.). (1991). Person schemas and

maladaptive interpersonal behavior. Chicago:
University of Chicago Press.

www.theipi.org 915
www.freepsychotherapybooks.org
 Horowitz, M. J., & Eells, T. D. (1993). Case formulations
using role-relationship model configurations: A
reliability study. Psychotherapy Research, 3, 56-68.

Horowitz, M. J., Eells, T. D., Singer, J., & Salovey, P.

(1995). Role relationship models for case
formulation. Archives of General Psychiatry, 53, 627-
632.

Horowitz, M., Luborsky, L., & Popp, C. (1991). A

comparison of the Role Relationship Models
Configuration and the Core Conflictual Relationship
Theme. In M. J. Horowitz (Ed.), Person schemas and
maladaptive interpersonal behavior (pp. 197-212).
Chicago: University of Chicago Press.

Husserl, E. (1960). Cartesian mediations: An introduction

to phenomenology (D. Cairns, Trans.). Dordrecht,
The Netherlands: Martinus Nijhoff.

Johnson, M. E., Popp, C., Schacht, T. E., Mellon, J., &

Strupp, H. S. (in press). Converging evidence for
identification of recurrent relationship themes:
Comparison of two methods. Psychiatry.

Kächele, H., Luborsky, L., & Thomae, H. (1988).

Obertragung als structur und verlaufsmuster—zwei
methoden ziirfassung dieser aspekte [Transference as
structure and process pattern—Two method versions
of these aspects]. In L. Luborsky & H. Kächele

www.theipi.org 916
www.freepsychotherapybooks.org
 (Eds.), Der zentrale beziehungskonflikt. Ulm,
Germany: PSZ-Verlag.

Kagan, J. (1989). Temperamental contributions to social

behavior. American Psychologist, 44, 668-674-

Kagan, J. (1996). Three pleasing ideas. American

Psychologist, 51, 901-908.

Karasu, T. B., & Skodol, A. (1980). Vlth Axis for DSM-

III: Psychodynamic evaluation. American Journal of
Psychiatry, 197, 607-610.

Kelly, G. A. (1955). The psychology of personal

constructs (Vol. 1). New York: Norton.

Kernberg, O., Burnstein, E., Coyne, L., Applebaum, A.,

Horowitz, L., & Voth, H. (1972). Psychotherapy and
psychoanalysis: Final report of the Menninger
Foundation’s Psychotherapy Research Project.
Bulletin of the Menninger Clinic, 36, 1-275.

Kiesler, D. J. (1983). The 1982 interpersonal circle: A

taxonomy for complementarity in human
transactions. Psychological Review, 90, 185-214.

Kiesler, D. J. (1987a). Check List of Psychotherapy

Transactions—Revised (CLOPT-R) and Check List of
Interpersonal Transactions—Revised (CLOIT-R).
Richmond: Virginia Commonwealth University.

www.theipi.org 917
www.freepsychotherapybooks.org
 Kiesler, D. J. (1987b). Research manual for the Impact
Message Inventory. Palo Alto, CA: Consulting
Psychologists Press.

Kiesler, D. J., Anchin, J. C., Perkins, M. J., Chirico, B. M.,

Kyle, E. M., & Federman, E. J. (1985). The Impact
Message Inventory: Form II. Palo Alto, CA:
Consulting Psychologists Press.

Kiesler, D. J., Goldston, C. S., & Schmidt, J. A. (1991).

Manual for the Check List of Interpersonal
Transactions—Revised and the Check List of
Psychotherapy Transactions—Revised. Richmond:
Virginia Commonwealth University.

Kiesler, D. J., & Schmidt, J. A. (1993). The Impact

Message Inventory: Form IIA Octant Scale Version.
Palo Alto, CA: Mind Garden.

Kiresuk, T. (1973). Goal attainment scaling at a county

mental health service. Evaluation, 1, 12-18.

Klein, G. (1970). Perception, motives and personality.

New York: Knopf.

Knapp, P. (1991). Self-other schemas: Core organizers of

human experience. In M. J. Horowitz (Ed.), Person
schemas and maladaptive interpersonal behavior
patterns. Chicago: University of Chicago Press.

www.theipi.org 918
www.freepsychotherapybooks.org
 Koss, M., & Butcher, J. (1986). Research on brief
psychotherapy. In S. Garfield & Bergin (Eds.),
Handbook of psychotherapy and behavior change
(pp. 627–670). New York: Wiley.

Kubie, L. (1952). Problems and techniques of

psychoanalytic validation and progress. In E.
Pumpian-Mindlin (Ed.), Psychoanalysis as science
(pp. 46-124). Stanford, CA: Stanford University
Press.

Labov, W. (1972). Language in the inner city.

Philadelphia: University of Pennsylvania Press.

Lambert, M. J., & Bergin, A. E. (1994). The effectiveness

of psychotherapy. In A. E. Bergin & S. L. Garfield
(Eds.), Handbook of psychotherapy and behavior
change (pp. 143-149). New York: Wiley.

Landis, J., & Koch, G. (1970). The measurement of

observer agreement for categorical data. Biometrics,
33, 159-174.

LeDoux, J., Romanski, L., & Xagoraris, A. (1989).

Indelibility of subcortical emotional memories.
Journal of Cognitive Neuroscience, I, 238-243.

Leeds, J., & Bucci, W. (1986). A reliable method for the

detection of repetitive structures in a transcript of an
analytic session. Paper presented at the meeting of

www.theipi.org 919
www.freepsychotherapybooks.org
 the Society for Psychotherapy Research, Wellesley,
MA.

Lefebvre, R., Diguer, L., Morissette, E., Rousseau, J. P., &

Normandin, L. (1996). The borderline and their Core
Conflictual Relationship Themes. Paper presented at
the annual meeting of the Society for Psychotherapy
Research, Amelia Island, FL.

Levine, F. J., & Luborsky, L. (1981). The Core Conflictual

Relationship Theme method: A demonstration of
reliable clinical inferences by the method of
mismatched cases. In S. Tuttman, C. Kaye, & M.
Zimmerman (Eds.), Object and self: A developmental
approach (pp. 501-526). Madison, CT: International
Universities Press.

Liberman, B. L. (1978). The role of mastery in

psychotherapy: Maintenance or improvement and
prescriptive change. In J. Frank, R. Hoen-Saric, S.
Imber, B. Liberman, & A. Stone (Eds.), Effective
ingredients of successful psychotherapy (pp. 35-72).
New York, Brunner/Mazel.

Loevinger, J. (1976). Ego development. San Francisco:

Jossey-Bass.

Lorr, M., & McNair, D. M. (1965). Expansion of the

interpersonal behavior circle. Journal of Personality
and Social Psychology, 2, 823-830.

www.theipi.org 920
www.freepsychotherapybooks.org
 Luborsky, E. (1987). Stability and transformation of
coping and affect from 12 to 23 months of age.
Unpublished doctoral dissertation, New York
University, New York.

Luborsky, L. (1962). Clinicians’ judgments of mental

health: A proposed scale. Archives of General
Psychiatry, 7, 407-417.

Luborsky, L. (1967). Momentary forgetting during

psychotherapy and psychoanalysis: A theory and
research method. In R. R. Holt (Ed.), Motives and
thought: Psychoanalytic essays in honor of David
Rapaport. Madison, CT: International Universities
Press.

Luborsky, L. (1975). Clinicians’ judgments of mental

health: Specimen case descriptions and forms for the
Health-Sickness Rating Scale. Bulletin of the
Menninger Clinic, 35, 448-480.

Luborsky, L. (1976). Helping alliances in psychotherapy:

The groundwork for a study of their relationship to its
outcome. In J. Claghom (Ed.), Successful
psychotherapy (pp. 92-116). New York:
Brunner/Mazel.

Luborsky, L. (1977a). Curative factors in psychoanalytic

and psychodynamic psychotherapies. In J. P. Brady, J.
Mendels, M. T. Orne, & W. Rieger (Eds.),

www.theipi.org 921
www.freepsychotherapybooks.org
 Psychiatry: Areas of promise and advancement (pp.
187-203). New York: Spectrum.

Luborsky, L. (1977b). Measuring a pervasive psychic

structure in psychotherapy: The Core Conflictual
Relationship Theme. In N. Freedman & S. Grand
(Eds.), Communicative structures and psychic
structures (pp. 367-395). New York: Plenum Press.

Luborsky, L. (1978a). A measure of self-understanding

based on self-interpretation of one’s own relationship
anecdotes. Unpublished manuscript.

Luborsky, L. (1978b). The Relationship Anecdotes

Paradigm (RAP) interview: A TAT-like method using
actual narratives. Unpublished manuscript.

Luborsky, L. (1980). A free self-analysis: The therapist's

own narratives analyzed by the CCRT. Unpublished
manuscript.

Luborsky, L. (1984). Principles of psychoanalytic

psychotherapy: A manual for supportive-expressive
treatment. New York: Basic Books.

Luborsky, L. (1986a). Evidence to lessen Professor

Grunbaum’s concern about Freud’s clinical inference
methods. [Review and commentary on the book The
foundations of psychoanalysis: A philosophical
critique]. Behavioral and Brain Sciences, 9, 247-249.

www.theipi.org 922
www.freepsychotherapybooks.org
 Luborsky, L. (1986b). A set of standard categories for the
CCRT, Edition 1. Unpublished manuscript.

Luborsky, L. (1987). A method of inferring unconscious

conflicts: An extension beyond the CCRT.
Unpublished manuscript.

Luborsky, L. (1988a). A comparison of three transference-

related measures applied to the same specimen hour.
In H. Dahl, H. Kächele, & H. Thomae (Eds.),
Psychoanalytic process research strategies. New
York: Springer-Verlag.

Luborsky, L. (1988b). Recurrent momentary forgetting: Its

content and context. In M. Horowitz (Ed.),
Psychodynamics and cognition (pp. 223-251).
Chicago: University of Chicago Press.

Luborsky, L. (1988c). Take a moment to really look at the

little lawful world of momentary forgetting: A reply
to Spence. In M. Horowitz (Ed.), Psychodynamics
and cognition (pp. 265-268). Chicago: University of
Chicago Press.

Luborsky, L. (1990a). A guide to the CCRT method. In L.

Luborsky & P. Crits-Christoph (Eds.), Understanding
transference: The CCRT method (pp. 15-36). New
York: Basic Books.

Luborsky, L. (1990b). The Relationship Anecdotes

Paradigms (RAP) interview as a versatile source of

www.theipi.org 923
www.freepsychotherapybooks.org
 narratives. In L. Luborsky & P. Crits-Christoph
(Eds.), Understanding transference: The CCRT
method (pp. 102-116). New York: Basic Books.

Luborsky, L. (1990c). Theory and technique in dynamic

psychotherapy: Curative factors and training
therapists to maximize them. Psychotherapy and
Psychosomatics, 53, 50-57.

Luborsky, L. (1996). The Symptom-Context method:

Symptoms as opportunities in psychotherapy.
Washington, DC: American Psychological
Association.

Luborsky, L. (1997). A CCRT-based personality theory of

the inner and the interpersonal world. Unpublished
manuscript, University of Pennsylvania Medical
School, Philadelphia.

Luborsky, L. (in press). The Core Conflictual Relationship

Theme (CCRT): A basic case formulation method. In
T. D. Eells (Ed.), Handbook of psychotherapy case
formulations. New York: Guilford Press.

Luborsky, L., & Auerbach, A. H. (1969). The Symptom-

Context Method: Quantitative studies of symptom
formation in psychotherapy. Journal of the American
Psychoanalytic Association, 17, 68-99.

Luborsky, L., & Bachrach, H. (1974). Factors influencing

clinicians’ judgments of mental health. Eighteen

www.theipi.org 924
www.freepsychotherapybooks.org
 experiences with the Health-Sickness Rating Scale.
Archives of General Psychiatry, 31, 292-299.

Luborsky, L., Barber, J. P., & Diguer, L. (1992). The

meanings of the narratives told during psychotherapy:
The fruits of a new operational unit. Psychotherapy
Research, 2, 277-290.

Luborsky, L., Barber, J. P., & Schaffler, P. (1989). The

assessment of the CCRT: Comparison of tailor-made
with standard category rating scales on a specimen
case. In J. C. Perry (Chair), Progress in assessing
psychodynamic functioning: A comparison of four
methods on a single case. Symposium conducted at
the meeting of the Society for Psychotherapy
Research, Toronto, Canada.

Luborsky, L., & Crits-Christoph, P. (1990). Understanding

transference: The CCRT method. New York: Basic
Books.

Luborsky, L., Crits-Christoph, P., & Alexander, K. (1990).

Repressive style and relationship patterns: Three
samples inspected. In J. Singer (Ed.), Repression and
dissociation: Implications for personality theory,
psychopathology and health. Chicago: University of
Chicago Press.

Luborsky, L., Crits-Christoph, P., Alexander, L., Margolis,

M., & Cohen, M. (1983). Two helping alliance

www.theipi.org 925
www.freepsychotherapybooks.org
 methods for predicting outcomes of psychotherapy: A
counting signs versus a global rating method. Journal
of Nervous and Mental Disease, 17, 480-492.

Luborsky, L., Crits-Christoph, P., Friedman, S., Mark, D.,

& Schaffler, P. (1991). Freud’s transference template
compared with the Core Conflictual Relationship
Theme (CCRT): Illustrations by the two specimen
cases. In M. Horowitz (Ed.), Person schemas and
maladaptive interpersonal behavior (pp. 167-195).
Chicago: University of Chicago Press.

Luborsky, L., Crits-Christoph, P., & Mellon, J. (1986). The

advent of objective measures of the transference
concept. Journal of Consulting and Clinical
Psychology, 54, 39-47.

Luborsky, L., Crits-Christoph, P., Mintz, J., & Auerbach,

A. (1988). Who will benefit from psychotherapy?
Predicting therapeutic outcomes. New York: Basic
Books.

Luborsky, L., Diguer, L., Barber, J., Cacciola, J., Moras,

K., Schmidt, K., & DeRubeis, R. (1996). Outcomes
of short term dynamic psychotherapy for chronic
versus non-chronic major depression. Journal of
Psychotherapy Research and Practice, 5, 152-159.

Luborsky, L., Diguer, L., Luborsky, E., McLellan, A. T.,

Woody, G., & Alexander, L. (1993). Psychological

www.theipi.org 926
www.freepsychotherapybooks.org
 health as a predictor of the outcomes of
psychotherapy. Journal of Consulting and Clinical
Psychology [Special section], 61, 542-548.

Luborsky, L., Graff, H., Pulver, S. S., & Curtis, H. (1973).

A clinical-quantitative examination of consensus on
the concept of transference. Archives of General
Psychiatry, 29, 69-75.

Luborsky, L., & Kächele, H. (1988). Der Zentrale

Beziehungskonflikt: Ein Arbeitsbuch [The CCRT: A
workbook]. Ulm, Germany: PSZ.

Luborsky, L., Kächele, H., & Dahlbender, R. (Eds.).

(1997). The CCRT Newsletter. University of
Pennsylvania.

Luborsky, L., Luborsky, E. B., Diguer, L., Schaffler, P.,

Schmidt, K., Dengler, D., Faude, J., Morris, M.,
Buchsbaum, H., & Emde, R. (1995). Extending the
Core Conflictual Relationship into childhood. In G.
Noam & K. Fisher (Eds.), Development and
vulnerability in close relationships. Hillsdale, NJ:
Erlbaum.

Luborsky, L., Mark, D., Hole, A. V., Popp, C., Goldsmith,

B., & Cacciola, J. (1995). Supportive-expressive
dynamic psychotherapy of depression: A time-limited
version. In J. P. Barber & P. Crits-Christoph (Eds.),
Psychodynamic psychotherapies for psychiatric

www.theipi.org 927
www.freepsychotherapybooks.org
 disorders (Axis I) (pp. 13-42). New York: Basic
Books.

Luborsky, L., Mellon, J., & Crits-Christoph, P. (1985a). An

aid to reliability studies of the CCRT: Preset scoring
categories. Unpublished manuscript, University of
Pennsylvania.

Luborsky, L., Mellon, J., St Crits-Christoph, P. (1985b,

July). Stalking a social phobia armed with the CCRT
and the symptom-context methods. Paper presented at
a conference sponsored by the John D. and Catherine
T. MacArthur Foundation, Stanford, CA.

Luborsky, L., Mellon, J., van Ravenswaay, P., Childress,

A., Cohen, K. D., Hole, A. V., Ming, S., Crits-
Christoph, P., Levine, F. J., & Alexander, K. (1985).
A verification of Freud’s grandest clinical hypothesis:
The transference. Clinical Psychology Review, 5,
231-246.

Luborsky, L., & Popp, C. (1989). A method of inferring

unconscious conflicts: An extension beyond the
CCRT. Unpublished manuscript.

Luborsky, L., Popp, C., & Barber, J. P. (1994). Common

and special factors in different transference-related
measures. Psychotherapy Research, 4, 277-286.

Luborsky, L., Popp, C., Barber, J. P., & Shapiro, D.

(1994). Editor’s introduction. Psychotherapy

www.theipi.org 928
www.freepsychotherapybooks.org
 Research, 4, 151.

Luborsky, L., Popp, C., Luborsky, E., & Mark, D. (1994).

The Core Conflictual Relationship Theme.
Psychotherapy Research, 4, 172-183.

Luborsky, L., & Schimek, J. (1964). Psychoanalytic

theories of therapeutic and developmental change:
Implications for assessment. In P. Worchel & D.
Byrne (Eds.), Personality change (pp. 73-99). New
York: Wiley.

Luborsky, L., & Spence, D. (1978). Quantitative research

on psychoanalytic therapy. In S. L. Garfield & A. E.
Bergin (Eds.), Handbook of psychotherapy and
behavior change: An empirical analysis (2nd ed., pp.
331-368). New York: Wiley.

Luborsky, L., Stuart, J., Friedman, S., Seligman, D. A.,

Pulver, S., & Woody, G. (1997). The Penn collection
of tape-recorded psychoanalyses as a research
resource. Unpublished manuscript.

Luborsky, L., van Ravenswaay, P., Ball, W., Steinman, D.,

Sprehn, G., & Bryan, C. (1993). Come centrare il
trattamento in ambiente psichiatrico. Uso del metodo
CCRT-FIT (Trattamento ospedaliero centrato) [How
to focus psychiatric hospital treatment: Use of the
CCRT-FIT method (Focused In-patient Treatment)].

www.theipi.org 929
www.freepsychotherapybooks.org
 In Prospective Psicoanalitiche nel Lavoro
Instituzionale, 11, 9-16.

Mackenzie, R. (1989, May). Comparing methods to assess

patients for therapy. Paper presented at the meeting
of the American Psychiatric Association, San
Francisco.

Main, M., & Goldwyn, R. (1984). Predicting rejection of

her infant from mother’s representation of her own
experiences: Implications for the abused-abusing
intergenerational cycle. Child Abuse and Neglect, 8,
203-217.

Main, M., & Goldwyn, R. (1985). Adult attachment

classification system. Unpublished manuscript,
University of California, Berkeley.

Main, M., Kaplan, N., & Cassidy, J. (1985). Security in

infancy, childhood, and adulthood: A move to the
level of representation. Growing Points of Attachment
Theory and Research: Monographs of the Society for
Research in Child Development, 50(1-2, Serial No.
209), 66-104.

Malan, D. (1963). A study of brief psychotherapy.

Springfield, IL: Charles C Thomas.

Malan, D., Bacal, H., Heath, E. S., & Balfour, F. H. G.

(1968). A study of psychodynamic changes in

www.theipi.org 930
www.freepsychotherapybooks.org
 untreated neurotic patients. British Journal of
Psychiatry, 114, 525-551.

Mann, J. (1973). Time-limited psychotherapy. Cambridge,

MA: Harvard University Press.

Marmar, C., & Horowitz, M. J. (1988). Diagnosis and

phase-oriented treatment of post-traumatic stress
disorders. In J. Wilson (Ed.), Human adaptation to
extreme stress: From the Holocaust to Vietnam. New
York: Brunner/Mazel.

Maslow, A. H. (1970). Motivation and personality. New

York: Harper & Row.

Maxim, P. (1986). The Seattle Psychotherapy Language

Analysis Schema. Seattle: University of Washington
Press.

Maxim, P., & Sprague, M. (1989). Metacommunication of

interactive sequences in therapy. Seattle: University
of Washington Press.

Maxim, P., Straus, M., & Rosenfarb, I. (1986). The Seattle

Psychotherapy Language Analysis Schema:
Reliability and validity. Unpublished manuscript.

Mayman, M. (1968). Early memories and character

structure. Journal of Projective Techniques, 32, 303-
316.

www.theipi.org 931
www.freepsychotherapybooks.org
 Mayman, M. (1978). Trauma, stimulus barrier, ego
boundaries and self-preservation: Ego psychology in
Beyond the pleasure principle. In S. Smith (Ed.), The
human mind revisited: Essays in honor of Karl A.
Menninger (pp. 141-158). Madison, CT: International
Universities Press.

Mayman, M., & Faris, N. (1960). Early memories as

expressions of relationship paradigms. American
Journal of Orthopsychiatry, 30, 507-520.

McLaughlin, J. T. (1987). The play of transference: Some

reflections on enactment in the psychoanalytic
situation. Journal of the American Psychoanalytic
Association, 35, 557-582.

McMullen, L., & Conway, J. B. (1997). Dominance and

nurturance in the narratives told by clients in
psychotherapy. Psychotherapy Research, 7, 83-98.

Meichenbaum, D., & Gilmore, J. B. (1984). The nature of

unconscious processes: A cognitive-behavioral
perspective. In K. Bowers & D. Meichenbaum (Eds.),
The unconscious reconsidered (pp. 273-298). New
York: Wiley.

Merritt, J. M., Stickgold, R., Pace-Schott, E., Williams, J.,

& Hobson, J. A. (1994). Emotion profiles in the
dreams of men and women. Consciousness and
Cognition, 3, 46-60.

www.theipi.org 932
www.freepsychotherapybooks.org
 Miller, J. B. (1970). Waking and dreaming
conceptualization in depression. In J. Masserman
(Ed.), Depression: Science and psychoanalysis: Vol.
17. Depression: Theories and therapies. New York:
Grune & Stratton.

Miller, N., Luborsky, L., Barber, J. P., & Docherty, J.

(Eds.). (1993). Psychodynamic treatment research: A
handbook for clinical practice. New York: Basic
Books.

Mintz, J. (1981). Measuring outcome in psychodynamic

psychotherapy. Archives of General Psychiatry, 38,
503-506.

Mintz, J., Auerbach, A., Luborsky, L., & Johnson, M.

(1973). Patients’, therapists’, and observers’ views of
psychotherapy: A “Rashomon” experience or a
reasonable consensus? British Journal of Medical
Psychology, 46, 83-89.

Mintz, J., & Kiesler, D. J. (1982). Individualized measures

of psychotherapy outcome. In P. C. Kendall & J. N.
Butcher (Ed.), Handbook of research methods in
clinical psychology (pp. 491-534). New York: Wiley.

Mischel, W. (1968). Personality and assessment. New

York: Wiley.

Mitchell, J. (1995). Coherence of the relationship theme:

An extension of Luborsky’s Core Conflictual

www.theipi.org 933
www.freepsychotherapybooks.org
 Relationship Theme Method. Psychoanalytic
Psychology, 12, 495-512.

Morgan, R., Luborsky, L., Crits-Christoph, P., Curtis, H.,

& Solomon, J. (1982). Predicting the outcomes of
psychotherapy by the Penn Helping Alliance Rating
method. Archives of General Psychiatry, 39, 397-402.

Murphy, G. (1938). Historical introduction to modern

psychology. New York: Harcourt Brace.

Murray, H. (1938). Explorations in personality. A clinical

and experimental study of fifty men of college age.
Oxford, England: Oxford University Press.

Nathans, S. (1988). Plan Attainment: An individualized

measure for assessing outcome in psychodynamic
psychotherapy. Unpublished doctoral dissertation,
California School of Professional Psychology at
Berkeley.

Nisbett, R. E., & Wilson, T. D. (1977). Telling more than

we can know: Verbal reports on mental processes.
Psychological Review, 84, 231-259.

Noam, G. (1991). Beyond Freud and Piaget: Biographical

worlds—interpersonal self. In T. Wren (Ed.), The
moral domain (pp. 360-399). Cambridge, MA: MIT
Press.

www.theipi.org 934
www.freepsychotherapybooks.org
 Norville, R., Sampson, H., & Weiss, J. (1996). Accurate
interpretations and brief psychotherapy outcome.
Psychotherapy Research, 6, 16-29.

Nunberg, H. (1951). Transference and reality.

International Journal of Psychoanalysis, 32, 1-9.

O’Malley, S., Suh, C., & Strupp, H. (1983). The

Vanderbilt Psychotherapy Process Scale: A report on
the scale development and a process-outcome study.
Journal of Consulting and Clinical Psychology, 51,
581-586.

Orlinsky, D., Grawe, K., & Parks, B. (1994). Process and

outcome in psychotherapy—Noch ein mal. In A.
Bergin & J. S. Garfield (Eds.), Handbook of
psychotherapy and behavior change (4th ed., pp.
270-378). New York: Wiley.

Orlinsky, D., & Howard, K. (1978). The relation of

process to outcome in psychotherapy. In S. Garfield
& A. Bergin (Eds.), Handbook of psychotherapy and
behavior change: An empirical analysis (2nd ed., pp.
283-329). New York: Wiley.

Orlinsky, D., & Howard, K. (1986). Process and outcome

in psychotherapy. In S. Garfield & A. Bergin (Eds.),
Handbook of psychotherapy and behavior change:
An empirical analysis (3rd ed., pp. 311-381). New
York: Wiley.

www.theipi.org 935
www.freepsychotherapybooks.org
 Perry, J. C. (1993). Assessing psychodynamic patterns
using the Idiographic Conflict Formulation (ICF)
Method. In N. Miller, L. Luborsky, & J. Docherty
(Eds.), Dynamic psychotherapy research: A
handbook for clinical practice. New York: Basic
Books.

Perry, J. C. (1994). Assessing psychodynamic patterns

using the Idiographic Conflict Formulation Method.
Psychotherapy Research, 4, 239-252.

Perry, J. C., Augusto, F., & Cooper, S. H. (1989).

Assessing psychodynamic conflicts: I. Reliability of
the Idiographic Conflict Formulation Method.
Psychiatry, 52, 289-301.

Perry, J. C., & Cooper, S. H. (1989). An empirical study of

defense mechanisms: I. Clinical interview and life
vignette ratings. Archives of General Psychiatry, 46,
444-452.

Perry, J. C., Luborsky, L., Silberschatz, G., & Popp, C.

(1989). An examination of three methods of
psychodynamic formulation based on the same
videotaped interview. Psychiatry, 52, 302-323.

Persons, J. (1989). Cognitive therapy in practice: A case

formulation approach. New York: Norton.

Peterson, C., Bettes, B., & Seligman, M. E. (1985).

Depressive symptoms and spontaneous casual

www.theipi.org 936
www.freepsychotherapybooks.org
 attributions: Content analysis. Behavior Therapy and
Research, 23, 379-382.

Peterson, C., Luborsky, L., & Seligman, M. E. (1983).

Attributions and depressive mood shifts: A case study
using the symptom-context method. Journal of
Abnormal Psychology, 92, 96-103.

Peterson, C., & Seligman, M. E. (1984). Content analysis

of verbatim explanations: The CAVE technique for
assessing explanatory style. Unpublished manuscript.

Pfeffer, A. (1963). The meaning of the analyst after

analyses: A contribution to the theory of therapeutic
results. Journal of the American Psychoanalytic
Association, 11, 229-244.

Piotrowski, C., & Keller, J. (1984). Psychodiagnostic

testing in APA approved clinical psychology
programs. Professional Psychology: Research and
Practice, 15, 450-456.

Piper, W. E., Azim, H. F., Joyce, A. S., & McCallum, M.

(1991). Transference interpretations, therapeutic
alliance, and outcome in short-term individual
psychotherapy. Archives of General Psychiatry, 4,
946-953.

Popp, C., Diguer, L., Luborsky, L., Faude, J., Johnson, S.,
Morris, M., Schafer, N., Schaffler, P., & Schmidt, K.
A. (1996). Repetitive relationship themes in waking

www.theipi.org 937
www.freepsychotherapybooks.org
 narratives and dreams. Journal of Consulting and
Clinical Psychology, 64, 1073-1078.

Popp, C., Luborsky, L., & Crits-Christoph, P. (1990). The

parallel of the CCRT from therapy narratives with the
CCRT from dreams. In L. Luborsky & P. Crits-
Christoph (Eds.), Understanding transference: The
CCRT method (pp. 37-50). New York: Basic Books.

Pulver, S. E. (1987). The manifest dream in

psychoanalysis: A clarification. Journal of the
American Psychoanalytic Association, 35, 99-118.

Rapaport, D. (1951). Organization and pathology of

thought. New York: Columbia University Press.

Rapaport, D. (1960). The structure of psychoanalytic

theory: A systemizing attempt. Psychological Issues
[Monograph No. 6].

Rapaport, D., & Gill, M. (1967). The points of view and

assumptions of metapsychology. In D. Rapaport
(Ed.), Collected papers (pp. 795-811). New York:
Basic Books.

Rapaport, D., Gill, M., & Schafer, R. (1968). Diagnostic

psychological testing. Madison, CT: International
Universities Press.

Raskin, N. J. (1949). An analysis of six parallel studies of

the therapeutic process. Journal of Consulting

www.theipi.org 938
www.freepsychotherapybooks.org
 Psychology, 13, 206-221.

Rawn, M. (1958). An experimental study of transference

and resistance phenomena in psychoanalytically-
oriented psychotherapy. Journal of Clinical
Psychology, 14, 418.

Rawn, M. (1981). A note on unwitting replication:

Quantitative studies of transference and resistance
twenty years apart. Journal of Clinical Psychology,
37, 782.

Reiser, M. (1984). Mind, brain, body: Toward a

convergence of psychoanalysis and neurobiology.
New York: Basic Books.

Reiser, M. (1986, November). The durable core of Freud's

empirical science. The 38th A. A. Brill lecture, The
New York Psychoanalytic Society.

Rhode, A. B., Geller, J. D., & Farber, B. A. (1992).

Dreams about the therapist: Mood, interactions, and
themes. Psychotherapy, 29, 536-544.

Rice, L., & Greenberg, L. (Eds.). (1984). Patterns of

change. New York: Guilford Press.

Rosenbaum, M., Friedlander, J., & Kaplan, S. (1956).

Evaluation of results of psychotherapy.
Psychosomatic Medicine, 18, 113-132.

www.theipi.org 939
www.freepsychotherapybooks.org
 Rosenberg, S. E., Horowitz, L. M., Hanks, S., Hartley, D.,
Lebenson, H., Schulman, T., & Skuja, A. (1984). The
consensual psychodynamic formulation: Part II.
Application to case of Ms. Smithfield. Psychotherapy
Research, 4, 234-238.

Rosenberg, S. E., Silberschatz, G., Curtis, J., Sampson, H.,

& Weiss, J. (1986). A method for establishing
reliability of statements from psychodynamic case
formulations. American Journal of Psychiatry, 143,
1454-1456.

Ryle, A. (1990). Cognitive-analytic active participation

and change: A new integration in brief
psychotherapy. Chichester, England: Wiley.

Ryle, A. (1991). Object relations theory and activity

theory: A proposed link by way of the procedural
sequence model. British Journal of Medical
Psychology, 64, 307-316.

Sachs, J. S. (1983). Negative factors in brief

psychotherapy: An empirical assessment. Journal of
Consulting and Clinical Psychology, 51, 557–564.

Schacht, T. B., & Binder, J. (1982). Focusing: A manual

for identifying a circumscribed area of work for time-
limited dynamic psychotherapy (TLDP). Unpublished
manuscript, Vanderbilt University.

www.theipi.org 940
www.freepsychotherapybooks.org
 Schacht, T. B., Binder, J., & Strupp, H. (1984). The
dynamic focus. In H. Strupp & J. Binder (Eds.),
Psychotherapy in a new key: A guide to time limited
dynamic psychotherapy (pp. 65-109). New York:
Basic Books.

Schafer, R. (1983). Narration in the psychoanalytic

dialogue. Critical Inquiry, 12, 29-53.

Schaffer, N. (1982). Multidimensional measures of

therapist behavior as predictors of outcome.
Psychological Bulletin, 92, 670-681.

Schaffer, N. (1983). Methodological issues of measuring

the skillfulness of therapeutic techniques.
Psychotherapy: Theory, Research, and Practice, 20,
486-493.

Schlesinger, H. J. (1974). Problems of doing research on

the therapeutic process in psychoanalysis. Journal of
the American Psychoanalytic Association, 22, 3-13.

Schlessinger, N., & Robbins, F. (1975). The

psychoanalytic process: Recurrent patterns of conflict
and changes in ego. Journal of the American
Psychoanalytic Association, 23, 761-782.

Seitz, P. (1966). The consensus problem in psychoanalytic

research. In L. Gottschalk & A. Auerbach (Eds.),
Methods of research in psychotherapy (pp. 209-225).
New York: Appleton-Century-Crofts.

www.theipi.org 941
www.freepsychotherapybooks.org
 Seligman, M. (1975). Helplessness: On depression,
development and death. San Francisco: Freedman.

Seligman, M. (1980). A learned helplessness point of

view. In L. Rhem (Ed.), Behavior therapy for
depression (pp. 123-142). New York: Academic
Press.

Seligman, M. (1991). Learned optimism. New York:

Pocket Books.

Seligman, M., Kamaen, L., & Nolen-Hoeksema, S. (1988).

Explanatory style across the life span: Achievement
and health. In E. Hetherington, R. Lerner, & M.
Perlmutter (Eds.), Child development in lifespan
perspective (pp. 91-114). Hillsdale, NJ: Erlbaum.

Seligman, M., Peterson, C., Kaslow, N., Tannenbaum, R.,

Alloy, L., & Abramson, L. (1984). Attributional style
and depressive symptoms among children. Journal of
Abnormal Psychology, 93, 235-238.

Shultz, L. H., Hauser, S. T., & Allen, J. P. (1990).

Autonomy and relatedness development in early adult
close peer relationships. Paper presented at the Fifth
Biennial International Conference on Personal
Relationships, Oxford, England.

Sifneos, P. (1979). Short term dynamic psychotherapy:

Evaluation and technique. New York: Plenum Press.

www.theipi.org 942
www.freepsychotherapybooks.org
 Silberschatz, G. (1986). Testing pathogenic beliefs. In J.
Weiss, H. Sampson, & the Mount Zion
Psychotherapy Research Group (Eds.), The
psychoanalytic process: Theory, clinical
observations, and empirical research (pp. 256-266).
New York: Guilford Press.

Silberschatz, G., Curtis, J. T., Fretter, P. B., & Kelly, T.

(1988). Testing hypotheses of psychotherapeutic
change processes. In H. Dahl, H. Kächele, & H.
Thomae (Eds.), Psychoanalytic process research
strategies (pp. 129-145). Berlin: Springer-Verlag.

Silberschatz, G., Curtis, J. T., & Nathans, S. (1989). Using

the patient’s plan to assess progress in psychotherapy.
Psychotherapy, 26, 40-46.

Silberschatz, G., Fretter, P., & Curtis, J. (1986). How do

interpretations influence the process of
psychotherapy? Journal of Consulting and Clinical
Psychology, 54, 646-652.

Singer, B., & Luborsky, L. (1977). Countertransference:

The status of clinical vs. quantitative research. In A.
Gurman & A. Razin (Eds.), The therapist's handbook
for effective psychotherapy: An empirical assessment
(pp. 431-448). New York: Pergamon Press.

Singer, J. (1985). Transference and the human condition:

A cognitive-affective perspective. Psychoanalytic

www.theipi.org 943
www.freepsychotherapybooks.org
 Psychology, 2, 189-219.

Singer, J., & Salovey, P. (1991). Organized knowledge

structures and personality: Person schemas, self-
schemas, prototypes and scripts. In M. Horowitz
(Ed.), Person schemas and maladaptive interpersonal
patterns. Chicago: University of Chicago Press.

Spence, D. (1983a). Narrative persuasion. Psychoanalysis

and Contemporary Thought, 6, 457-481.

Spence, D. (1983b). Narrative truth and historical truth.

New York: Norton.

Spitz, R. (1956). Transference, the analytic setting and its

prototype. International Journal of Psychoanalysis,
57, 380-385.

SPSS/PC & Advanced Statistics [Computer software].

(1986). Chicago, IL: SPSS Inc.

Sroufe, L. A. (1983). Infant-caregiver attachment and

patterns of adaptation in pre-school: The roots of
maladaptation and competence. In M. Perlmutter
(Ed.), Minnesota Symposium on Child Psychology
(Vol. 16, pp. 41-81). Hillsdale, NJ: Erlbaum.

Sroufe, L. A., & Fleeson, J. (1986). Attachment and the

construction of relationships. In W. Hartug & Z.
Rubin (Eds.), Relationships and development (pp. 51-

www.theipi.org 944
www.freepsychotherapybooks.org
 71). Cambridge, England: Cambridge University
Press.

Sroufe, L. A., <Si Waters, E. (1977). Attachment as an

organizational construct. Child Development, 48,
1184-1199.

Stein, K., & Beall, L. (1971). Externalizing-internalizing

symptoms and psychotherapeutic outcome.
Psychotherapy: Therapy, Research and Practice, 8,
269-272.

Stern, D. (1985). The interpersonal world of the infant.

New York: Basic Books.

Stern, D. (1989). The representations of relational

patterns: Development considerations. In A. Sameroff
& R. Emde (Eds.), Relationship disturbances in early
childhood (pp. 52-69). New York: Basic Books.

Stickgold, R., Pace-Schott, E., & Hobson, J. A. (1994). A

new paradigm for dream research: Mentation reports
following spontaneous arousal from REM and
NREM sleep recorded in a home setting.
Consciousness and Cognition, 3, 16-29.

Strupp, H. (1973). Psychotherapy: Clinical, research, and

theoretical issues. Northvale, NJ: Jason Aronson.

Strupp, H., & Binder, J. (1984). Psychotherapy in a new

key: A guide to time-limited dynamic psychotherapy.

www.theipi.org 945
www.freepsychotherapybooks.org
 New York: Basic Books.

Strupp, H., Chassan, J., & Ewing, J. (1966). Toward the

longitudinal study of the psychotherapeutic process.
In L. Gottschalk & A. Auerbach (Eds.), Methods of
research in psychotherapy (pp. 361-400). New York:
Appleton-Century-Crofts.

Subotnick, L. (1966a). Transference in child therapy: A

third replication. Psychological Record, 16, 265-277.

Subotnick, L. (1966b). Transference in client-centered

play therapy. Psychology, 3, 2-17.

Sulloway, F. J. (1996). Born to rebel: Birth order, family

dynamics, and creative lies. New York: Pantheon
Books.

Teller, V., & Dahl, H. (1981). The framework for a model

of psychoanalytic inference. Proceedings of the
Seventh International Joint Conference on Artificial
Intelligence, 1, 394-400.

Teller, V., & Dahl, H. (1986). The microstructure of free

association. Journal of the American Psychoanalytic
Association, 34, 763-798.

Thome, A. (1989). Conditional patterns, transference, and

the coherence of personality across time. In D. Buss
& N. Cantor (Eds.), Personality psychology: Recent

www.theipi.org 946
www.freepsychotherapybooks.org
 trends and emerging directions (pp. 150-159). New
York: Springer-Verlag.

Thome, A. (1995a). Developmental truths in memories of

childhood and adolescence. Journal of Personality,
63, 139-163.

Thome, A. (1995b). Juxtaposed scripts, traits, and the

dynamics of personality. Journal of Personality, 63,
593-616.

Thome, A., & Michaelieu, Q. (1996). Situating adolescent

gender and self-esteem with personal memories.
Child Development, 67, 1374-1390.

Tomkins, S. (1947). The Thematic Apperception Test: The

theory and technique of interpretation. New York:
Grune & Stratton.

Tomkins, S. (1979). Script theory: Differential

magnification of affects. In H. E. Howe, Jr., & R. A.
Diensbier (Eds.), Nebraska Symposium on Motivation
(Vol. 26, pp. 201-236). Lincoln: University of
Nebraska Press.

Tomkins, S. (1987). Script theory. In V. Aronoff, A. A.

Rabin, & R. Zucker (Eds.), The emergence of
personality (pp. 147-216). New York: Springer-
Verlag.

www.theipi.org 947
www.freepsychotherapybooks.org
 Toolan, M. (1989). Narrative: A critical linguistic
introduction. London: Routledge & Kegan Paul.

Tronick, E. (1982). Social interchange in infancy.

Baltimore: University Park Press.

Vaillant, G. (1977). Adaptation to life. Boston: Little,

Brown.

van Ravenswaay, P., Luborsky, L., & Childress, A. (1983,

July). Consistency of the transference in versus out of
psychotherapy. Paper presented at the meeting of the
Society for Psychotherapy Research, Sheffield,
England.

Wachtel, P. (1993). Therapeutic communication:

Principles and effective practice. New York: Guilford
Press.

Waelder, R. (1936). The principle of multiple function:

Observations on overdetermination. Psychoanalytic
Quarterly, 5, 45-62.

Waldinger, R. (1997a). Continuities and discontinuities in

relationship themes from adolescence to young
adulthood. Manuscript in preparation.

Waldinger, R. (1997b). Relationship narratives in 5-year-

olds: A comparison of abused and neglected children
with normal children. Manuscript in preparation.

www.theipi.org 948
www.freepsychotherapybooks.org
 Wallerstein, R. (1986). Forty-two lives in treatment: A
study of psychoanalysis and psychotherapy. New
York: Guilford Press.

Wallerstein, R., Robbins, L., Sargent, H., & Luborsky, L.

(1956). The psychotherapy research project of the
Menninger Foundation: Rationale, method and
sample use. Bulletin of the Menninger Clinic, 20,
221-280.

Weinberger, D. (1990). The construct validity of the

repressive copying style. In J. L. Singer (Ed.),
Repression and dissociation. Chicago: University of
Chicago Press.

Weiss, D. S., DeWitt, K. N., Kaltreider, N. B., &

Horowitz, M. J. (1985). Proposed method for
measuring change beyond symptoms. Archives of
General Psychiatry, 42, 703-708.

Weiss, J. (1986). Theory and clinical observations. In J.

Weiss, H. Sampson, & the Mount Zion
Psychotherapy Research Group (Eds.), The
psychoanalytic process: Theory, clinical
observations, and empirical research (pp. 3-138).
New York: Guilford Press.

Weiss, J., Sampson, H., Caston, J., & Silberschatz, G.

(Eds.). (1977). Research on the psychoanalytic
process. San Francisco: Psychotherapy Research

www.theipi.org 949
www.freepsychotherapybooks.org
 Group, Department of Psychiatry, Mount Zion
Hospital and Medical Center.

Weiss, J., Sampson, H., & the Mount Zion Psychotherapy

Research Group. (1986). The psychoanalytic process:
Theory, clinical observations, and empirical research.
New York: Guilford Press.

White, R. (1952). Lives in progress. Hinsdale, IL: Dryden

Press.

Wilson, A., Passik, S., Morral, A., Turner, A., & Kuras, M.
(1994). An epigenetic-TAT approach to the
assessment of modes of organization (EPI-TAT).
Manuscript in preparation.

Winget, C., & Kramer, M. (1979). Dimensions of dreams.

Gainesville: University of Florida Press.

Witkin, H. A. (1949). Sex differences in perceptions.

Transactions of the New York Academy of Sciences,
12, 22-26.

Wolpert, L., & Richards, A. (1988). A passion for science.

Oxford, England: Oxford University Press.

Zander, B., Strack, M., Cierpka, M., Reich, G., Steats, H.,
assisted by Biskey, J., Homburg, H., Krannich, S.,
Ratzke, K., & Seide, L. (1995). Coder agreement
using the German edition of Luborsky’s CCRT

www.theipi.org 950
www.freepsychotherapybooks.org
 method in videotaped or transcribed RAP interviews.
Psychotherapy Research, 5, 231-236.

Zolik, E. S., & Hollon, T. W. (1960). Factors characteristic

of patients responsive to brief psychotherapy
[Abstract]. American Psychologist, 15, 387.

www.theipi.org 951
www.freepsychotherapybooks.org
 ABOUT THE AUTHORS

There is one main theme in Lester Luborsky’s

research over the past five decades: the translation of

clinically useful concepts, now 36 of them, into

clinically useful clinical-quantitative measures. His

favorites include the Health-Sickness Rating Scale,

the Core Conflictual Relationship Theme, the Helping

Alliance, and the symptom onset state, including

helplessness and hopelessness. After receiving a PhD

at Duke University in 1945, he has been working at

the University of Pennsylvania for the past 38 years,

having turned out eight books, with six of those in the

past dozen years, along with 370 articles and

chapters.

Paul Crits-Christoph, PhD, is Associate Professor

of Psychology in Psychiatry, School of Medicine,

University of Pennsylvania, where he is also Director

www.theipi.org 952
www.freepsychotherapybooks.org
 of the Center for Psychotherapy Research. He is
coeditor, with Jacques P. Barber, Phd, of the

Handbook of Short-Term Dynamic Psychotherapy

(1991) and Dynamic Therapies for Psychiatric

Disorders (1995), both published by Basic Books. In

addition, he has published over 100 articles and
chapters.

www.theipi.org 953
www.freepsychotherapybooks.org