Contemporary Directions in Psychopathology: Scientific Foundations of the DSM-V and ICD-11 - PDF Free Download (2024)

Contemporary Directions in Psychopathology


Directions in Psychopathology

Scientific Foundations of the DSM-V and ICD-11

edited by

Theodore Millon Robert F. Kruegerâ•…â•…Erik Simonsen

THE GUILFORD PRESS New Yorkâ•…â•…London

©â•›2010 The Guilford Press A Division of Guilford Publications, Inc. 72 Spring Street, New York, NY 10012 All rights reserved No part of this book may be reproduced, translated, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or€otherwise, without written permission from the Publisher. Printed in the United States of America This book is printed on acid-free paper. Last digit is print number:â•… 9â•… 8â•… 7â•… 6â•… 5â•… 4â•… 3â•… 2â•… 1 Library of Congress Cataloging-in-Publication Data Contemporary directions in psychopathology : scientific foundations of the DSM-V and ICD-11 / edited by Theodore Millon, Robert F. Krueger, Erik Simonsen. â•…â•… p. cm. â•… Includes bibliographical references and index. â•… ISBN 978-1-60623-532-4 (hbk.) â•… 1.╇ Psychology, Pathological.â•… 2.╇ Mental illness—Etiology.â•… 3.╇ Mental illness— Diagnosis.â•… 4.╇ Mental illness—Classification.â•… I.╇ Millon, Theodore.â•… II.╇ Krueger, Robert F.â•… III.╇ Simonsen, Erik, 1949– â•… RC454.C647 2010 â•… 616.89—dc22 2009028186

About the Editors

Theodore Millon, PhD, DSc, is Dean and Scientific Director of the Institute for Advanced Studies in Personology and Psychopathology. He was Founding Editor of the Journal of Personality Disorders and inaugural president of the International Society for the Study of Personality Disorders (ISSPD). Dr. Millon has held full professorial appointments at Harvard Medical School, the University of Illinois, and the University of Miami. A prolific author, he has written or edited more than 30 books on theory, assessment, and therapy, as well as more than 200 articles and book chapters. Dr. Millon is the recipient of the 2008 American Psychological Foundation Gold Medal for Life Achievement in the Application of Psychology. Robert F. Krueger, PhD, is Professor of Psychology and Psychiatry at Washington University in St. Louis. His major research interests lie at the intersection of personality, psychopathology, psychometrics, and behavioral genetics. Dr. Krueger has received a number of awards, including the American Psychological Association’s Distinguished Scientific Award for Early Career Contribution and the American Psychological Foundation’s Theodore Millon Award for midcareer contributions to personality psychology. Erik Simonsen, MD, is Director of Psychiatric Research in Region Zealand, Denmark; Director of the Institute of Personality Theory and Psychopathology; and Associate Research Professor at the University of Copenhagen. He has published widely on personality disorders, first-episode psychosis, personality assessment, outcome of psychotherapy, and psychiatric classification. Dr. Simonsen is past president of the ISSPD and a recipient of the ISSPD Award. He has also served as president of the Section on Personality Disorders of the World Psychiatric Association.



Renato D. Alarcón, MD, MPH, Mood Disorders Unit and Department of Psychiatry, Mayo€Clinic College of Medicine, Rochester, Minnesota Steven R. H. Beach, PhD, Institute for Behavioral Research, University of Georgia, Athens,€Georgia Edward M. Bernat, PhD, Department of Psychology, Florida State University, Tallahassee,€Florida Roger K. Blashfield, PhD, Department of Psychology, Auburn University, Auburn, Alabama Sidney J. Blatt, PhD, Department of Psychiatry, Yale School of Medicine, New€Haven,€Connecticut Bekh Bradley, PhD, Trauma Recovery Program, Atlanta Veterans Affairs Medical Center, and Department of Psychiatry and Behavioral Sciences, Emory University, Atlanta, Georgia Jessica Combs, BA, Department of Psychology, University of Kentucky, Lexington,€Kentucky Rina Dutta, MRCPsych, Institute of Psychiatry, Kings College London, London,€United€Kingdom Nicholas R. Eaton, MA, Department of Psychology, Washington University in St. Louis, St.€Louis, Missouri Michael B. First, MD, Department of Psychiatry, Columbia University, New York, New€York Elizabeth Flanagan, PhD, Department of Psychiatry, Yale School of Medicine, New€Haven,€Connecticut




Joseph P. Gone, PhD, Department of Psychology, University of Michigan, Ann€Arbor,€Michigan Seth Grossman, PsyD, Counseling and Psychological Services Center, Florida International University, Miami, Florida William M. Grove, PhD, Department of Psychology, University of Minnesota, Minneapolis,€Minnesota Richard E. Heyman, PhD, Department of Psychology, Stony Brook University, State University of New York, Stony€Brook,€New York Nadine J. Kaslow, PhD, Department of Psychiatry, Emory University School of Medicine, Atlanta, Georgia Jared Keeley, MS, Department of Psychology, Auburn University, Auburn, Alabama Kenneth S. Kendler, MD, Virginia Institute for Psychiatric and Behavioral Genetics, Virginia Commonwealth University, Richmond, Virginia Laurence J. Kirmayer, MD, Division of Social and Transcultural Psychiatry, McGill University, and Institute of€Community and Family Psychiatry, Jewish General Hospital, Montreal, Quebec, Canada Helena Chmura Kraemer, PhD, Department of Psychiatry and Behavioral Sciences, Stanford€University, Stanford, California; Department of Psychiatry, University of€Pittsburgh, Pittsburgh, Pennsylvania Robert F. Krueger, PhD, Departments of Psychology and Psychiatry, Washington University in St. Louis, St. Louis, Missouri Mark F. Lenzenweger, PhD, Department of Psychology, State University of New York at€Binghamton, Binghamton, New York Mark R. Lukowitsky, MA, Department of Psychology, Pennsylvania State University, University Park, Pennsylvania Patrick Luyten, PhD, Department of Psychology, University of Leuven, Leuven, Belgium Mario Maj, MD, PhD, Department of Psychiatry, University of Naples, Naples, Italy Paul E. Meehl, PhD (deceased), Department of Psychology, University of Minnesota, Minneapolis, Minnesota Theodore Millon, PhD, DSc, Institute for Advanced Studies in Personology and€Psychopathology, Port Jervis, New York Robin M. Murray, MD, Institute of Psychiatry, Kings College London, London,€United€Kingdom Kile M. Ortigo, MA, Department of Psychology, Emory University, Atlanta, Georgia Joel Paris, MD, Department of Psychiatry, McGill University, and Institute of Community and€Family Psychiatry, Jewish General Hospital, Montreal, Quebec, Canada Christopher J. Patrick, PhD, Department of Psychology, Florida State University, Tallahassee, Florida Aaron L. Pincus, PhD, Department of Psychology, Pennsylvania State University, University€Park, Pennsylvania



Kristin Raley, MS, Department of Psychology, Auburn University, Auburn, Alabama David Reiss, MD, Yale Child Study Center, Yale University, New Haven, Connecticut Bryna Siegel, PhD, Department of Psychiatry, University of€California, San Francisco, San€Francisco, California Erik Simonsen, MD, Psychiatric Research Unit, Region Zealand, and Department of€Neuroscience and Pharmacology, University of€Copenhagen, Copenhagen, Denmark Andrew E. Skodol, MD, Sunbelt Collaborative and Department of€Psychiatry, University of Arizona College of Medicine, Tucson, Arizona Gregory T. Smith, PhD, Department of Psychology, University of Kentucky, Lexington,€Kentucky Susan C. South, PhD, Department of Psychological Sciences, Purdue University, West€Lafayette, Indiana Michael H. Stone, MD, Department of Psychiatry, Columbia University, and private practice, New York, New York Robert Tringone, PhD, Counseling Center, St. John’s University, Jamaica, New York Scott I. Vrieze, BA, Department of Psychology, University of Minnesota, Minneapolis,€Minnesota Jerome C. Wakefield, PhD, DSW, Silver School of Social Work and Department of€Psychiatry, New York University, New York, New€York Marianne Z. Wamboldt, MD, Department of Psychiatry, University of Colorado Denver School of Medicine, Aurora, Colorado Drew Westen, PhD, Departments of Psychology and Psychiatry and Behavioral Sciences, Emory University, Atlanta, Georgia Aidan G. C. Wright, MS, Department of Psychology, Pennsylvania State University, University Park, Pennsylvania Peter Zachar, PhD, Department of Psychology, Auburn University at Montgomery, Montgomery, Alabama



he present book is similar in concept to a well-received volume that one of us (Theodore Millon) edited with Gerald L. Klerman of Harvard University in 1986; it is, however, an almost entirely new book, with only one chapter carried over from the earlier work. Gerry and I were colleagues at the Stanley Cobb Psychiatric Laboratories of Massachusetts General Hospital, as well as active participants in the development of DSM-III. Our aim in the earlier volume was to describe substantive and innovative advances since the publication of DSM-III in 1980, and to emphasize themes we believed should be considered in the forthcoming DSMIV. We, the present editors, intend to do the same in this volume for DSM-IV(-TR) and for the forthcoming DSM-V and ICD-11. Numerous changes in the character of psychopathology have begun taking place in the past several decades. Slow though progress has been, there are inexorable signs that the study of mental disorders has advanced beyond its earlier history as an oracular craft. No longer dependent on the intuitive artistry of brilliant clinicians and theoreticians who formulated dazzling but often unfalsifiable insights, psychopathology has acquired a solid footing in the empirical methodologies and quantitative techniques that characterize mature sciences. Although the term “psychopathology” was used in the past as synonymous with “descriptive symptomatology,” it can now be justly employed to represent “the science of abnormal behavior and mental disorders.” Its methods of study comfortably encompass both clinical and experimental procedures. Among the many indices of continuing progress is the construction of psychometrically sound diagnostic tools that wed the quantitative and statistical precision typifying rigorous empirical disciplines with the salient and dynamic qualities characterizing the concerns of a clinical profession. Contributing to this precision is the introduction of comprehensive and comparable diagnostic criteria for each mental disorder—an advance that not only enhances the clarity of clinical communication, but strengthens the reliability of research, contributing thereby to the collection of reciprocal and cumulative data. Similarly, sophisticated multivari-


xii Preface

ate statistical methods now provide quantitative grounds for analyzing symptom patterns and constructing an orderly taxonomy. Theoretical formulations have also begun to take on a more logical and orderly structure. Whereas earlier propositions were often presented in haphazard form, with circular derivations and ambiguous or conflicting empirical consequences, contemporary theorists began to specify explicit criteria for their concepts, as well as to spell out objective procedures and methods for testing their hypotheses. Moreover, theorists have become less doctrinaire in their positions than formerly; that is, they no longer act and write as religious disciples of “theological purity.” A true “ecumenism” has emerged—an open-mindedness and sharing of views that are much more characteristic of disciplines with secure foundations. Thus erstwhile analysts have shed their former dogmatisms and have begun to incorporate findings such as those in the neurosciences and social psychology; similarly, once-diehard behaviorists have jettisoned their earlier biases and have integrated cognitive processes into their principles. On many levels and from several perspectives, the signs indicate consistently that psychopathology is becoming a full-fledged science. It is our intent in this book to draw attention to innovations that constitute continuations of these directions. The volume is not intended to be a comprehensive textbook, but many of its chapters provide thoughtful pedagogic reviews and heuristic recommendations that may prove useful to the forthcoming DSM-V and ICD-11. In this latter regard, we very much favor current efforts to construct further rapprochements between the American Psychiatric Association’s DSM and the World Health Organization’s ICD. Work on the new editions of both manuals is well underway, and we believe that a successful accommodation will come from the combination of careful theoretical and conceptual analyses, and the parallel acquisition of empirical data from well-designed research. This work not only reflects the current state of psychopathology as a science, but should help identify the issues and methods that can foster this important reconciliation. As noted, all but one of the 30 chapters in this volume is new. Only Paul E. Meehl’s classic chapter on “diagnostic taxa” is a repeat from the earlier book; it is one that Mark F. Lenzenweger reflects on and thoroughly reviews. Notable in this edition are several chapters that bring to the forefront the role played by social context and culture in the roots of numerous mental disorders. The book begins with an extensive historical survey that leads up to contemporary thinking. Here are traced the contributions of theorists and researchers from ancient times (e.g., Zang Zhongjing, Alcmaeon, Aretaeus), the many fruitful ideas of 19th-century clinicians (e.g., Esquirol, Griesinger, Kahlbaum), and the work of more modern scholars (e.g., Kraepelin, Freud, Beck). Classification matters are explored deeply in chapters that deal with philosophical issues underlying construct validity, syndromal comorbidity, and the clinical utility of categories versus dimensions. Innovative proposals are presented on such topics as the neuroscientific foundations of psychopathology and the use of evolutionary principles in articulating the development and composition of psychopathology. It is our hope that this volume will contribute further to the long and fruitful collaboration between the disciplines of psychology and psychiatry. Psychopathology needs “all the help it can get” if it is to fulfill its promise as a science. The best minds are not to be found in one school of thought or in one mental health profession. Different perspectives not only contribute to “rounding out” important areas of content and technique, but help spark fresh insights and ideas. We three editors have found collaborative work to be both stimulating and rewarding; we hope that this book will serve not only as a model of cooperation between our fields, but as an invitation to biochemists, epidemiologists, psychometricians, geneticists, sociologists, and professionals in other disciplines to join us in similar enterprises.



In closing, we should like to express our appreciation to the book’s contributors. Among them are numerous distinguished scholars, as well as well-respected and promising young investigators from psychology, psychiatry, philosophy, and sociology. In addition to being extremely pleased with the high quality and original thought that went into each chapter, we should note that both authorial and editorial royalties for this volume will be turned over to the World Health Organization, to provide it with additional resources to facilitate and expedite the development of ICD-11. Theodore M illon Robert F. K rueger Erik Simonsen


Part I.╇ Historical and Cultural Perspectives Chapter 1.

A Précis of Psychopathological History Theodore Millon and Erik Simonsen

Chapter 2.

Themes in the Evolution of the 20th-�Century DSMs


On the Wisdom of Considering Culture and€Context in€Psychopathology


Cultural Issues in the Coordination of DSM-V and ICD-11


Roger K. Blashfield, Elizabeth Flanagan, and€Kristin Raley Chapter 3.


Joseph P. Gone and€Laurence J. Kirmayer Chapter 4.

Renato D. Alarcón Chapter 5.

A Sociocultural Conception of the Borderline Personality€Disorder€Epidemic


Theodore Millon

Part II.╇ Conceptual Issues in Classification Chapter 6.

Philosophical Issues in the Classification of€Psychopathology


Classification Considerations in€Psychopathology and Personology


Peter Zachar and Kenneth S. Kendler Chapter 7.

Theodore Millon


xvi Chapter 8.


Diagnostic Taxa as Open Concepts: Metatheoretical and€Statistical Questions about Reliability and€Construct Validity in€the€Grand Strategy of Nosological Revision


Contemplations on Meehl (1986): The Territory, Paul’s Map, and€Our Progress in Psychopathology Classification (or, the Challenge of Keeping Up with a Beacon 30 Years Ahead€of the Field)


Paul E. Meehl

Chapter 9.

Mark F. Lenzenweger

Chapter 10. Issues of Construct Validity in€Psychological€Diagnoses Gregory T. Smith and Jessica Combs


Chapter 11. The Meaning of Comorbidity among Common Mental Disorders Nicholas R. Eaton, Susan C. South, and€Robert F. Krueger


Chapter 12. The Connections between Personality and€Psychopathology Susan C. South, Nicholas R. Eaton, and€Robert F. Krueger


Chapter 13. Is It True That Mental Disorders Are So€Common,


and So Commonly Co-Occur? Mario Maj

Chapter 14. Taking Disorder Seriously: A Critique of Psychiatric Criteria

for€Mental Disorders from€the€Harmful-�Dysfunction Perspective


Jerome C. Wakefield

Part III.╇ Methodological Approaches to€Categories, Dimensions, and€Prototypes Chapter 15. On the Substantive Grounding and Clinical Utility of Categories

versus Dimensions


William M. Grove and€Scott I. Vrieze Chapter 16. A Short History of a Psychiatric Diagnostic Category

That Turned Out to Be a Disease


Roger K. Blashfield and Jared Keeley

Chapter 17. Concepts and Methods for Researching Categories

and€Dimensions in€Psychiatric€Diagnosis


Helena Chmura Kraemer

Chapter 18. The Integration of Categorical and€Dimensional Approaches



Erik Simonsen

Chapter 19. Dimensionalizing Existing Personality Disorder Categories Andrew E. Skodol




Chapter 20. An Empirically Based Prototype Diagnostic€System

for DSM-V and ICD-11


Kile M. Ortigo, Bekh Bradley, and€Drew Westen Chapter 21. The Millon Personality Spectrometer: A Tool for Personality

Spectrum Analyses, Diagnoses, and€Treatments


Theodore Millon, Seth Grossman, and€Robert Tringone

Part IV.╇ Innovative Theoretical and Empirical Proposals Chapter 22. Neuroscientific Foundations of€Psychopathology Christopher J. Patrick and Edward M. Bernat


Chapter 23. Using Evolutionary Principles for Deducing Normal


and Abnormal Personality Patterns

Theodore Millon

Chapter 24. Biopsychosocial Models and€Psychiatric€Diagnosis Joel Paris


Chapter 25. Reactivating the Psychodynamic Approach to€the Classification



Sidney J. Blatt and Patrick Luyten

Chapter 26. A Life Course Approach to Psychoses:

Outcome and Cultural Variation


Rina Dutta and Robin M. Murray

Chapter 27. The Interpersonal Nexus of Personality and€Psychopathology Aaron L. Pincus, Mark R. Lukowitsky, and Aidan G. C. Wright


Chapter 28. Reconceptualizing Autism Spectrum Disorders as€Autism-�Specific


Learning Disabilities and€Styles Bryna Siegel

Chapter 29. Describing Relationship Patterns in DSM-V: A Preliminary Proposal Marianne Z. Wamboldt, Steven R. H. Beach, Nadine€J.€Kaslow, Richard€E.€Heyman, Michael€B.€First, and€David€Reiss


Chapter 30. On the Diversity of the Borderline Syndromes Michael H. Stone


Author Index


Subject Index


Part I



Cultural Perspectives

Chapter 1

A Précis of Psychopathological History Theodore Millon Erik Simonsen

efore we and our contributors undertake B a systematic analysis of current trends in psychopathology in this text, it may be useful to introduce the subjects with reference to its historical origins and evolution. Efforts to understand and resolve the problems that researchers studying mental disorders continue to face can be traced through many centuries in which solutions have taken unanticipated turns and have become enmeshed in obscure beliefs and entangled alliances, most of which have unfolded without the care and watchful eye of modern scientific thoughts and methods. Psychopathology remains today a relatively young science. We find that many techniques and theories of our time have long histories that connect current thinking to preexisting beliefs and systems of thought, many of which are intertwined in chance associations, primitive customs, and quasi-�tribal quests. The path to the present is anything but a simple and straight line; it has come to its current state through an involvement in values and customs of which we may be only partly aware. Many are the product of historical accidents and erroneous beliefs that occurred centuries ago, when mysticism and charlatanism flourished.

The traditions of psychopathology today are not themselves tight systems of thought in the strict sense of scientific theories; they certainly are neither closed nor completed constructions of ideas that have been worked out in their final details. Rather, they are products of obscure lines of historical development—Â� movements often subject to the confusions and misunderstandings of our remote past, when a disaffection with complexities typified life. Nevertheless, interest in ourselves, in our foibles as well as our achievements, has always been central to our human curiosity. The origins of interest in the workings of psychopathology were connected in their earliest form to studies of astronomy and spiritual unknowns. Even before any record of human thought had been drafted in written form, we humans were asking fundamental questions, such as why we behave, think, act, and feel as we do. Although primitive in their ideas, ancient people were always open to the tragic sources in their lives. Earliest answers, however, were invariably associated with metaphysical spirits and magical spells. Only slowly were more sophisticated and scientific ideas formulated. It was not until the 6th century B.C. that the actions, thoughts, and feelings of humans 3


were attributed to natural forces—that is, to sources found within ourselves. Philosophers and scientists began to speculate intelligently about a wide range of psychological processes; many of their ideas turned out to be remarkably farsighted. Unfortunately, much of this early imaginative and empirical work was forgotten through the centuries. Time and again, it was then slowly stumbled upon and rediscovered by careful or serendipitous efforts. For example, John Locke in the 17th century described a clinical procedure for overcoming unusual fears; the procedure he described is not very different from the systematic desensitization method developed this past century by Joseph Wolpe. Similarly, Gustav Fechner, founder of psychophysics in the mid-19th century, recognized that the human brain is divided into two parallel hemispheres that are linked by a thin band of connecting fibers (what we now term the corpus callosum). According to Fechner’s speculations, if the brain was subdivided, it would create two independent realms of consciousness—a speculation confirmed and elaborated in the latter part of this past century by Roger Sperry, in what has been referred to as “split-brain research.” Every historical period was dominated by certain beliefs that ultimately won out over previously existing conceptions while retaining elements of the old. As the study of mental science progressed, different and frequently insular traditions evolved to answer questions posed by earlier philosophers, physicians, and psychologists. Separate disciplines with specialized training procedures developed. Today divergent professional groups are involved in the study of the mind (e.g., the neuroscientifically oriented psychiatrists, with a clear-eyed focus on biological and physiological processes; the psychoanalytic psychiatrists, with an austere yet sensitive attention to unconscious or intrapsychic processes; the personological psychologists, with the tools and techniques for appraising, measuring, and integrating the mind; and the academic psychologists, with a penchant for empirically investigating the basic processes of behavior and cognition). Each group has studied the complex questions generated by mental disorders with a different focus and emphasis. Yet the central issues remain the same. By tracing the history of each of these and other conceptual traditions, we will

learn how different modes of thought today have their roots in chance events, cultural ideologies, and accidental discoveries, as well as in brilliant and creative innovations. From today’s perspective, it seems likely that future developments in the field will reflect recent efforts to encompass and integrate biological, psychological, and sociocultural approaches. No longer will any single and restricted point of view be prominent; each approach will enrich all others as one component of a synergistic whole. Integrating the disparate parts of a clinical science—Â� theory, nosology, diagnosis, and treatment— is the latest phase in the great chain of history that exhibits an evolution in mental science professions from ancient times to the new millennium. Intervening developments (both those that have been successful and those that have not) were genuine efforts to understand more fully who we are and why we behave the way we do. The challenge to know who we are is unending, owing to the complexity of human functioning. New concepts come to the fore each decade, and questions regarding established principles are constantly raised. Perhaps in this new century we will bridge the varied aspects of our poignant yet scientific understanding of psychopathology, as well as bring the diverse traditions of the past together to form a single, overarching synthesis.

Ancient History Primitive humans and ancient civilizations alike viewed the unusual and strange within a magical and mythological frame of reference. Behavior that could not be understood was thought to be controlled by animistic spirits. Although both good and evil spirits were conjectured, the bizarre and often frightening behavior of persons with mental disorders led to a prevailing belief that demon spirits must inhabit them. The possession of evil spirits was viewed as a punishment for failing to obey the teachings of the gods and priests. Fears that demons might spread to afflict others often led to cruel and barbaric tortures. These primitive “therapies” of shock, starvation, and surgery have parallels in recent history, although the ancients based them on the more grossly naive conception of demonology.

A Précis of Psychopathological History

What has been called the sacred approach in primitive times may be differentiated into three phases, according to Roccatagliata (1973): “animistic,” “mythological,” and “demonological.” These divergent paradigms shared one point of view: that psychopathology was the expression of transcendent magical action brought about by external forces. The animistic model was based on pre-Â�logical and emotional reasoning derived from the deep connection between primitive beings and the mysterious forces of nature. From this viewpoint, events happened because the world was peopled by animated entities driven by obscure and ineffable forces that acted upon human minds and souls. The second phase, that characterized by mythological beliefs, transformed the animistic conception so that indistinct and indefinable forces were materialized into myths. Every fact of life was imbued with the powers of a particular entity; every symptom of disorder was thought to be caused by a deity who could, if appropriately implored, benevolently cure it. In the third, or demonological phase, the transcendent mythological deities were placed into a formal theological system such as the Judeo-Â�Christian. In line with this latter phase, two competing forces struggled for superiority: one creative and positive, represented by a good father or God; the other destructive and negative, represented by the willful negation of good in the form of demonic forces of evil. These three conceptions followed each other historically, but they did overlap, with elements of one appearing in the others at times. Many aspects of prehistoric life could not be understood; magic and supernatural concepts helped early humans make sense out of the unfathomable and unpredictable. Weighted with life’s painful realities and burdensome responsibilities, these beliefs gave an order and a pseudo-logic to fears of the unknown—a repository of unfalsifiable assumptions in which the supernatural filled in answers for that which could not be understood. Ultimately, supernaturalism became the dominant world view in which the perplexing experiences of life could be objectified and comprehended. Priests and wizards became powerful, capitalizing on the fears and peculiarities of the populace to undo spells, “heal” those with physical illnesses, and “purify” those with mental dis-


tress. Within this worldview, eccentric or irrational individuals were assuredly touched by spirits who possessed superhuman powers to induce psychic pathology. Almost all groups permitted healing to fall into the hands of priests and magicians—a situation that still exists today in some societies. Living in a world populated with imaginary beings, these spiritual forces could often calm the worst human anxieties and expunge the ever-Â�present terrors of life. Despite extensive archeological analyses, however, our knowledge of primitive times is no more than fragmentary. Nevertheless, we may assume that primitive humans saw a world populated with spirits that were essentially illusions created by their own state of anguish and perplexity. India, Babylonia, and China Many contributions of the early Hindus are associated with the name of Susruta, who lived 100 years before Hippocrates. His works followed the traditional beliefs of his day regarding possible demonic possession. However, Susruta suggested that the passions and strong emotions of those mental disorders might also bring about certain physical ailments calling for psychological help (Bhugra, 1992). Anticipating the significance of temperament or innate dispositions, Hindu medicine proposed that three such inclinations existed: wise and enlightened goodness, with its seat in the brain; impetuous passions, the sources of the pleasure and pain qualities, with their seat in the chest; and the blind crudity of ignorance, the basis of more animalistic instincts, its seat located in the abdomen. A concern with mental health has long been a part of Indian cultures, which evolved various ways of attempting to understand and negotiate mental disorder and psychological problems. Indians have long been involved in constructing explanatory techniques. In the first formal system of medicine in India, Ayurveda (The Book of Life), physical and mental illnesses were not clearly demarcated. Caraka Samhita dealt with medical diagnoses and management possibly dating from 600 B.C. and was the foremost text of the ancient Indian medical system. Caraka defined ayu (life) as a state consisting of shareera (body), indriya (sens-


es), satva (psyche), and atma (soul). Soul could not be destroyed, and it underwent reincarnation. The mind was responsible for cognition, and it directed the senses, controlled the self, reasoned, and deliberated. The equilibrium between the self and mind was viewed as paramount to good health. Caraka used the general term doshas for the body fluids or humors, vata, pitta, and kapha. The theory of doshas may have developed independently of the Greek humoral theory, or possibly the Hindu system may have traveled to Greece. Types of food were thought to influence the mind, personality characteristics, and the interactions among the three doshas. Different personality types were described in detail as leading to mental illness, through either unwholesome diet or moral transgressions. In the Hindu system, mental disorders were seen as largely metaphysical, but different appearances of mental disorders (like unmada, insanity) were described as resulting from heredity, imbalanced doshas, temperament, inappropriate diet, and metapsychological factors. Caraka also contained many descriptions of possession states regarded as arising from supernatural agents—a belief that is still apparent in many parts of highly religious Indian society. Religious connotations and references to spiritual enlightenments were only challenged in the early 19th century by the emerging Western-Â�science-based medicine introduced by British rulers. In India, colonial medical institutions became brickand-Â�mortar symbols of Western intellectual and moral power, with European doctors even being taken as the sole excuse for empire. Indian magical practices and religious customs have been marginalized to some extent, but a variety of shamans—whose therapeutic efforts combine classical Indian alchemy, medicine, magic, and astrology with beliefs and practices from folk and popular traditions—are still present. In the Middle East was the ancient civilization of Babylonia; it was not only a vast geographical expanse, but the foundation of philosophical thought for most nations in the Mediterranean region. In fact, many of the traditions discussed among the Greeks and Romans can be traced to ideas generated initially in the Babylonian empire. Babylonians were oriented toward astronomical events; superstitions regarding the

stars produced many gods, a result largely of their intellectual leaders’ fertile imaginations. Help from the gods was often sought through magical rites, incantations, prayers, and the special powers of those who were physicians or priests. The Babylonians assigned a demon to each disease; insanity, for example was caused by the demon Idta. Each was to be exorcised through special medicines (primarily herbs and plants), confessions, and other methods to help restore a balance between conflicting supernatural forces. As the Babylonians saw it, invariable tensions existed among the different gods— but, more importantly, between a more or less rational, as opposed to a superstitious, explanation of psychic ailments. The first medical book in China, Neijing (The Canon of Internal Medicine), was compiled between 300 B.C. and 100 B.C. Organic syndromes, like epileptic seizures (dian) and delirium-like states, were also described, but with no clear distinction from the concepts of insanity and psychosis (kuang). The primary causes of psychiatric illness were suggested to be vicious air, abnormal weather, and emotional stress. The famous doctor Zang Zhongjing, the Hippocrates of China, introduced other concepts and syndromes, such as febrile delirium, globus hystericus, and puerperal psychosis, in his Jinkuiyaolue (A Sketchbook in a Golden Box). Chinese medicine has tended to explain pathology change by means of philosophical concepts, and this framework has undergone little change. It includes the notions of the complementary yin and yang; the five elements, gold, wood, water, fire, and earth; and the principle of Tao (i.e., the way), which has been considered as the ultimate regulator of the universe and the most desirable state of wellbeing and longevity achieved by integrating the individual self into the realm of nature. These ontological principles were described in The Yellow Emperor’s Classic of Internal Medicine some 20 centuries ago (Liu, 1981). Different personality types were portrayed as resulting from combinations of the five elements (e.g., the fiery type, the earthy type, the golden type, and the watery type). Phenomena occurring inside human beings were understood in terms of phenomena occurring outside in nature. Chinese medicine later became organ-Â�oriented; that is, every visceral organ was believed to have charge of

A Précis of Psychopathological History

a specific function. The heart was thought to house the mind, the liver to control the spiritual soul, the lung the animal soul, the spleen ideas and intelligence, and the kidney vitality and will. No attention was paid to the brain! For a long time psychiatric symptoms were interspersed with those of physical disease. The mind–body dichotomy was not a central theme. Mood disturbances and psychiatric symptoms attributed to menstrual irregularities tended to be expressed in somatic terms. In Chin-Yue’s Medical Book, the Chinese word for “depression” literally meant “stagnation,” implying obstruction of vital air circulation in the body. Case vignettes of patients with “deceiving sickness” (i.e., hysterical neurosis) were presented in the same book explaining symptom formation in people trapped in very difficult situations. In a similar way, sexual impotence was explained by excessive worry. In summary, psychiatric concepts of mental illness in China have undergone basically the same sequence as in the West: supernatural, natural, somatic, and psychological stages. However, Chinese medicine has been relatively less influenced by religious thoughts compared than early European medicine was; patients in Europe in the Middle Ages were declared by priests to be bewitched and were punished. Acupuncture, traditional Chinese medicine, folk herbs, and psychotherapy have been the most commonly used treatment approaches in China. Egypt, Greece, and Rome In Egypt, as in other early civilizations, there is evidence that the heart was thought to be the center of mental activity. Egyptians also had difficulty in separating prevailing supernatural beliefs from beliefs about things that could be observed and modified in nature. Astronomical phenomena were the primary objects of worship. “Natural” qualities were usually turned aside in favor of the mystical powers of the gods. Over the course of a century or two, Egyptian philosophers and physicians began studying the brain, ultimately recognizing it as the primary source of mental activity. Egyptians recognized that emotional disorders could be described in line with ideas proposed by the Greeks. Thus the set of disturbances the Greeks termed “hysteria” (using their word for


“uterus”) was caused, as the Egyptians saw it, by a wandering uterus that had drifted from its normal resting location; the task of the physician was to bring the uterus back to its normal setting. This explanation for hysteria continued until the late Middle Ages. In the earliest periods of Greek civilization, insanity was considered a divine punishment, a sign of guilt for minor or major transgressions. Therapy sought to combat madness by various expiatory rites that removed impurities, the causes of psychic disorders. Priests mediated an ill person’s prayers to the gods so as to assure his or her cure. Thus, with divine help, the person’s heart could be purified of its evil. Albeit slowly, Greek scholars realized that little of a rational nature characterized their way of thinking about mental pathology. To them, external but unseen agents could no longer serve as a logical basis for a genuine understanding of mentally troublesome phenomena. A fundamental shift began to take place, not merely in the manner in which different types of mental disorders might be described, but in the basis for thinking about ways to alter these aberrant behaviors. In order to “treat” mental disorders, the Greeks began to recognize the necessity of understanding how and why mental disorders were expressed in the natural world; only then could they successfully deal therapeutically with the tangible symptoms of everyday mental life. Instead of leaving the treatment of mental disorders to the supernatural and mystical, they began to develop a more concretely oriented perspective. This transition was led by a number of imaginative thinkers in the 5th and 6th centuries B.C. A central intellectual effort of Greek philosophers was the desire to reduce the vastness of the universe to its fundamental elements. Most proposed that complexities could be degraded to one element—be it water, air, or fire. Their task was to identify the unit of which all aspects of the universe were composed. Among the first philosopher-Â�scientists to tackle this task was Thales (652–588 B.C.). What little we know of Thales comes largely from the writings of later Greek philosophers, notably Aristotle, Plato, and the historian Herodotus. This nimble-Â�witted Greek proposed that the fundamental unit of the universe was a tangible and identifiable substance, water.


Though Thales was not the prime forerunner of a modern understanding of mental processes, he was a radical thinker who redirected attention away from mysticism, recognizing that psychic disorders were natural events that should be approached from a scientific perspective. As a pivotal figure in his time, he ushered in an alternative to earlier supernatural beliefs. Equally significant was Thales’s view that efforts should be made to uncover underlying principles on which overt phenomena were based. Oriented toward finding these principles in physical studies and “geometric proportions,” he turned to “magnetic” phenomena, convinced that the essential element of all life was its animating properties. To Thales, action and movement, based on balanced or disarrayed magnetic forces, was what distinguished human frailty. In this belief, he further derogated the view that external supernatural forces intruded on the psyche; rather, the source of pathology was inherent within persons themselves. Paralleling the views of Thales, Pythagoras (582–510 B.C.) reasserted the importance of identifying the underlying scientific principles that might account for all forms of behavior. He differed from Thales in that he retrogressively preferred to use ethics and religion as the basis for deriving his scientific principles. More progressively, however, he was the first philosopher to claim that the brain was the organ of the human intellect, as well as the source of mental disturbances. He adopted an early notion of biological humors (i.e., naturally occurring bodily liquids), as well as positing the concept of emotional temperament to aid in decoding the origins of aberrant passions and behavior. The mathematical principles of balance and ratio served to account for variations in human characterological styles (e.g., degrees of moisture or dryness, the proportion of cold or hot, etc.). Balances and imbalances among humoral fundamentals would account for whether health or disease was present. Possessing a deep regard for his “universal principles,” Pythagoras applied his ideas to numerous human, ethical, and religious phenomena. Though he believed in immortality and the transmigration of souls, this did not deter him from making a serious effort to articulate the inner “equilibrium” of human anatomy and health.

Pythagoras considered mental life as reflecting a harmony between antithetical forces: good–bad, love–hate, singular–Â� plural, limited–Â�unlimited, and so on. Life was regulated according to his conception of opposing rhythmic movements (e.g., sleep–Â� wakefulness, inspiring–Â�expiring). Mental disorders reflected a disequilibrium of these basic harmonies, producing symptoms of psychic impairment. To him, the soul could rise or descend from and to the body. The more the soul was healthy, in balance, and without psychic symptoms, the more it resembled solar energy. Pythagoras spoke of the soul as composed of three parts: reason, which reflected truth; intelligence, which synthesized sensory perceptions; and impulse, which derived from bodily energies. The rational part of the soul was centered in the brain; the irrational one, in the heart. Incidentally, Pythagoras coined the term “philosophy” by putting together the words philo, meaning “love,” and sophia, meaning “wisdom.” Ostensibly through his father, Apollo, Aesculapius (ca. 550 B.C.) gained his understanding of the nature of mental disorders through the divination of dreams, which he then transmitted to his sons, Machaon and Podaleirius. A series of followers, called Aesculapians, established long-Â�enduring “medical temples” and a distinguished cult. It is unclear historically whether Aesculapius actually existed or whether his ideas should properly have been attributed to Pythagoras. As the Aesculapian cult spread throughout the Greek empire, numerous temples were erected in the main cities of the Mediterranean basin, including Rome in 300 B.C. What may be best known about Aesculapiad temples today is the symbol of medical knowledge they employed: a serpent wrapped around a rod. Medicine gradually evolved into a branch of philosophy in the 6th and 7th centuries B.C. No one of that early period achieved the mythic stature of Aesculapius, however—the presumed founder of temple-based hospitals designed to execute the healing traditions in which he believed, notably a rest from life’s stressors and opportunities for positive mental growth. Located in peaceful and attractive settings, these temples were established to encourage patients to believe that there were good reasons to want to recover. Included

A Précis of Psychopathological History

among the temples’ treatment techniques were a balanced diet, a daily massage, quiet sleep, priestly suggestions, and warm baths, all of which were thought to comfort and soothe patients. Also of value during this early period was the work of Alcmaeon (557–491 B.C.), possibly a son or favorite student of Pythagoras, carried out in the 5th century B.C. Alcmaeon became a philosopher-Â�physiologist who asserted that the central nervous system was the physical source of mental activity, and that cerebral metabolism was based on the stability of “the humoral fluxes”; if these fluxes were imbalanced or unstable, they would create shifts in cerebral tissue functioning, leading then to various mental disorders. Metabolic fluxes were caused by a disequilibrium between the nervous system’s qualities of dry–moist and hot–cold. Most notable were Alcmaeon’s efforts to track the sensory nerves as they ascended to the brain. He articulated, as perhaps no one else before him had done, the structural anatomy of the body through methods of careful dissection. No less significant was his conviction that the brain, rather than the heart, was the organ of thought. As Aesculapius reportedly did, he also anticipated the work of Empedocles and Hippocrates, in that he believed that health called for a balance among the essential components of life—Â� coolness versus warmth, wetness versus dryness, and so on. The notion of fundamental elements in balance became a central theme in the work of Aesculapius and Alcmaeon; it also served to guide the views of their disciples. Alcmaeon’s “biological model” based on the concept of metabolic harmony, called “isonomy,” took the place of Greek’s early mythological theology and was an extension of the growing secular and democratic spirit of Greek’s 6th-Â�century B.C. culture. Empedocles (495–435 B.C.) adopted the homeostatic model generated in the work of Pythagoras, Aesculapius, and Alcmaeon. Most significant was his proposal that the basic elements of life (fire, earth, air, and water) interacted with two other “principles” (love vs. strife). Empedocles stressed that a balance among the four elements could be complicated by the fact that they might combine in either a complementary or a counteractive way. Love and strife represented human expressions of more elementary magnetic pro-


cesses such as attraction and repulsion. All of the elements/humors could be combined, but Empedocles wondered what the consequences would be if they were organized in different ways. He set out to weave the several threads of his theory and concluded that the force of attraction (love) would be likely to bring forth a harmonic unity, whereas repulsion (strife) would set the stage for a personal breakdown or social disintegration. To Empedocles, blood was a perfect representation of an equal mix of water, earth, air, and fire. He therefore suggested that persons with problematic temperaments and mental disorders would exhibit imbalances within their blood. Among his other contributions, Empedocles posited a rudimentary model of an evolutionary theory, anticipating Darwin’s by 2,000 years. As he phrased it, “creatures that survive are those whose blood elements are accidentally compounded in a suitable way,” whereas a problematic compounding will produce “creatures that will perish and die.” To him, nature created a wide variety of healthful and perishing blood configurations—that is, different ways in which the four elements combined. Some philosophers disagreed with the notion that the universe was composed of a simple and permanent element. Heracl*tus (530–470 B.C.), for example, proposed that all nature was made up of fire. He asserted, however, that the universe was composed of no lasting substance—Â�nothing stable, solid, or enduring. All real and tangible things would inevitably vanish, change their form, even become their very opposites. In a similar manner, Anaxagoras (500– 428 B.C.) asserted that a reduction to the basic elements could not explain the universe. He differed from Heracl*tus in that he did not believe the universe lacked an enduring substance. He asserted that an endless number of qualitatively different elements existed, and that the organization or arrangement of these diverse elements was central to the structure of the universe. Anaxagoras’s belief that the character of these constituents could not be explained except through the action of human thought was novel—a view similar to one asserted many centuries later by the phenomenologists and the gestaltists, who claimed that the structure of objective matter was largely in the interpretive eye of the perceiver.


Later the philosopher Democritus (460– 362 B.C.), following Leucippus (ca. 445 B.C.), proposed that the universe was made of variously shaped atoms—small particles of matter in constant motion, differing in size and form, but always moving and combining into the many complex components that comprise the universe as we know it. This innovative speculation endures to the present time. Extending the theme proposed a century earlier by Anaxagoras, Democritus stressed the view that all truths were relative and subjective. As noted, he asserted that matter was composed of numerous invisible particles called atoms. Each atom was composed of different shapes that combined and were linked in numerous ways; again, although this idea was based on pure speculation, it was highly innovative and is regarded as essentially correct to this day. The physical thesis of contemporary times known as the Heisenberg principle also has its origins in Democritus’s speculation. A contemporary of Democritus, born the same year, became the great philosopherÂ�physician who set the groundwork for sophisticated clinical medicine for the ensuing centuries. The fertility of this wondrous period of Grecian thought cannot be overestimated, ranging from the brilliant ideas of Democritus and Aristotle to the creative foundations of scientific medicine by Hippocrates. Hippocrates (460–367 B.C.; see Figure 1.1) was born on the island of Cos, the center of an ancient medical school. He was the son of an Aesculapian priest, from whom he acquired his first medical lessons and whose philosophy he would follow in his own future therapeutic efforts. In the work of Hippocrates—the inheritor of his father’s tradition and the humoral concepts of Pythagoras and Empedocles—Â�mental disorders progressed from the magical and mythical realm, and the demonological and superstitious therapeutic approaches of an earlier era, to one of careful clinical observation and inductive theorizing. He synthesized the practical and sympathetic elements of the Aesculapian cult with the more “biological” proposals of Pythagoras, blending these elements to elevate mental processes and disequilibria into a clinical science. Thus in the 5th century B.C., truly radical advances were made to supplant the super-

stitions of temple medicine. The astuteness and prodigious work of Hippocrates highlighted the naturalistic view that the source of all disorders, mental and physical alike, should be sought within the patient and not within spiritual phenomena. For example, the introductory notes to the Hippocratic book on epilepsy state: It seems to me to be no more divine and no more sacred than other diseases, but like other affections, it springs from natural causes. .€.€. Those who first connected this illness with demons and described it as sacred seem to me no different from the conjurers, purificators, mountebanks and charlatans of our day. Such persons are merely concealing, under the cloak of godliness, their perplexity and their inability to afford any assistance. .€.€. It is not a god which injures the body, but disease.

As a number of his progenitors had done, Hippocrates emphasized that the brain was the primary center of thought, intelligence, and emotions. It is only from within the brain, he asserted, that pleasures and joys and laughter arise, as well as sorrows, griefs, and tears. It is, he went on to say, this very same source that makes us mad or delirious, inspires us with dread and fear, and brings sleeplessness, inopportune mistakes, aimless anxieties, absentmindedness, and other acts contrary to the person’s habitual ways. All of these stem from the brain when it is not

Figure 1.1.╇ Hippocrates.

A Précis of Psychopathological History

healthy (i.e., as when an imbalance exists between hot and cold or moist and dry). Hippocrates’s approach was essentially empirical, despite the growing eminence of philosophical thought that characterized his time. He was a practical biologist stressing the role of bodily humors and focusing on the use of physical treatments (notably diet, massage, music, and remedies promoting sleep and rest) rather than philosophical ones. Central to the medical practices of Hippocrates and his followers was the crucial role given keen observation and fact gathering. Contrary to the work of Plato, who relied on abstract hypotheses and so-Â�called self-Â�evident truths, Hippocrates focused his attention on observable symptoms, their treatments, and their eventual outcomes. In this regard, Hippocrates modeled Aristotle’s empirical orientation, emphasizing facts rather than abstractions. As were a number of his forebears, Hippocrates was convinced that dreams could serve as indicators of health or illness. Mental pathology stemmed from a disparity between the content of dreams and that which existed in reality. Dream symbolism, as regarded by Hippocrates, led him to anticipate later hypotheses concerning the operation of “unconscious forces.” Hippocrates also established the tradition of carefully recording personal case history, detailing the course and outcome of the disorders he observed. These histories provide surprisingly accurate descriptions of such varied disorders as depression, phobias, convulsions, and migraine. With his associates at the Cos College of Medicine in Athens, he provided a logic for differentiating among various mental ailments—not only those we now label the DSM-IV-TR Axis I syndromes, but also the Axis II personality types, the latter of which were construed as abnormalities of temperament. Temperament was associated with the four-Â�humors model, which transformed earth, fire, water, and air into their parallel bodily elements. Individuals were characterized in terms of which one of the four elements predominated. Among other clinical syndromes differentiated were delirium, phobia, hysteria, and mania. Lacking precise observations of bodily structure, and prevented by taboo from performing dissections, Hippocratic physicians proposed hypothetical explanations of disease.


They adhered closely, however, to the first nonsupernatural schema that specified temperament dimensions in accord with the doctrine of bodily humors. Interestingly, history has come full circle, in that much of contemporary psychiatry continues to seek answers with reference to inner biochemical and endocrinological processes. Hippocrates identified four basic temperaments: the “choleric,” the “melancholic,” the “sanguine,” and the “phlegmatic.” These corresponded, respectively, to excesses in yellow bile, black bile, blood, and phlegm. As elaborated by a Roman, Galen, centuries later, the choleric temperament was associated with a tendency toward irascibility; the sanguine temperament prompted an individual toward optimism; the melancholic temperament was characterized by an inclination toward sadness; and the phlegmatic temperament was conceived as an apathetic disposition. Although the doctrine of humors has long been abandoned, giving way to studies on topics such as neurohormone chemistry, its archaic terminology still persists in contemporary expressions such as persons being “sanguine” or “good-Â�humored.” Hippocrates and his Cos associates were among the first to stress the need for a relationship between diagnosis and treatment. The mere description of a clinical disturbance was not sufficient for them, unless it provided a clear indication of the course that therapy should follow. Indeed, Hippocrates anticipated that much effort may be wasted in specifying diagnosis, unless followed by a consideration of its utility for therapeutic decisions. Although naive in conception and execution, Hippocrates’s approach to therapy followed logically from his view that disorders were of natural origin. To supplant the prevalent practices of exorcism and punishment, he recommended such varied prescriptions as exercise, tranquility, diet, venesection or bloodletting where necessary, and even marriage. Systematically (in a contemporaneous sense), Hippocrates and his colleagues devised a series of therapeutic regimens that they believed would reestablish the humoral balance thought to underlie most diseases; they also employed surgical techniques such as trephining to relieve purported pressure on the brain. Several themes relevant to the mind and its difficulties characterize Plato’s (429–347


work: (1) Powerful emotional forces could come to the foreground and overwhelm the everyday behavior typifying a person’s life; (2) conflicts could exist between different components of the psyche (e.g., the personal discord that often arises between an individual’s rational side—that which is desired—and the surge of emotional feelings); and (3) mental disorders did not result from simple ignorance, but from irrational superstitions and erroneous beliefs. To Plato, all humans were partly animal-like; hence all humans acted irrationally at times—some more, some less. He found evidence for these conclusions in dreams, where bizarre events invariably occur and unnatural connections among thoughts and images are dominant. Not to be overlooked was his contention that therapeutic efforts could modify any and all forms of mental illness. For Plato, the use of educational procedures could dispel ignorance and uncover “truth” through the application of fundamental principles. No less important with regard to therapy was Plato’s use of a dialectical model to change a patient’s cognitions and belief systems. In this regard, Plato’s philosophy provided a methodology for engaging in therapy, essentially the application of rational discussions to modify faulty cognitions (shades of contemporary cognitive therapies!). Plato had many distinguished students, the most eminent of whom was Aristotle (384–322 B.C.). Though he was Plato’s student for over 20 years, Aristotle turned sharply away from Plato and toward matters more realistic and tangible than abstract and idealistic. Some would say that Aristotle provided history’s first integrated and systematic accounts not only of psychological matters, but of astronomy, physics, zoology, and politics. The last of the great philosophers of the 4th century B.C., Aristotle was more scientist than philosopher. He gave special attention to the need for experimental verification and the use of sensory-based observable data; in fact, he was the first of the major philosophers to take an inductive and empirical approach in his writings. He was interested in the concrete observables of experience as registered through the senses. Although he admired the abstract rationalism of Plato, he was much more disposed to deal with the tangible world than with highorder abstractions or broad principles. He believed that data should be grounded in em-

pirical observables in order to minimize the risk of subjective misinterpretations. Despite these reservations, Aristotle believed that thought transcended the sensory realm. As he saw it, imagination could create thoughts of a higher order of abstraction than could sensations themselves. Yet not all matters were successfully brought within Aristotle’s purview. Despite growing evidence that the brain was the center of thought and emotion, Aristotle retained the erroneous belief that the heart served as the seat of these psychological experiences. He made keen and significant observations, however, in recognizing the psychological significance of cognitive processes, dreams, and emotional catharses. For example, it was Aristotle who said that events, objects, and people were linked by their relative similarity or their relative difference from one another. To Aristotle, things became “associated” if they occurred together; in this, he was clearly a forerunner of the associationist school of the 18th and 19th centuries. Aristotle viewed dreams as afterimages of the activities of the preceding day. Although he recognized that dreams might fulfill a biological function, he judged the content of dreams to be ideal gauges of potential pathology. He had a specific interest in how physical diagnoses could be deduced from dream content. Aristotle’s scope was exceptionally broad and inventive. It was he who wrote most perceptively of the intellectual and motivational features of the mind from the viewpoint of a natural scientist. Thus, in what might be termed a psychobiological theory, he outlined the basics of human perception and rational thought, stressing the importance and validity of sense impressions as the source for an objective form of experimental study. Along the same lines, Aristotle articulated a series of proposals concerning the nature of learning—a model based on the principles of association and reinforced by what we have come to term the “pleasure principle.” Similarly, he emphasized the importance of early experience and education in the acquisition of skills, and the role of habit and practice in the formation of psychological attitudes. To him, the processes of development were key themes in understanding human behavior. When Aristotle left Athens in the year 322 B.C., following the death of Alexander the Great, he arranged to have his associate

A Précis of Psychopathological History

Theophrastus (371–286 B.C.) succeed him as head of the Lyceum. Shortly thereafter, Aristotle, alone and despondent over the turn of political events in Athens, died in exile. Theophrastus was only a decade younger than Aristotle and had come to Athens to study with Plato. He and Aristotle had been friends, joined together in their travels and shared in their study of nature. Theophrastus remained head of the Lyceum for some 30 years. Perhaps most significant was the attention Theophrastus paid to the study of botany, establishing him as the true founder of that science, just as Aristotle’s works established the field of zoology. A prolific and sophisticated thinker, Theophrastus wrote no less than 220 treatises on a variety of different topics. Although this diversity of work was substantial, he became best known for a secondary aspect of his career, the writing of personality sketches he called “characters.” Each of these portrayals emphasized one or another psychological trait, providing a vignette of various personality “types” (e.g., individuals who were flatterers, garrulous, penurious, tactless, boorish, surly, etc.). Whether these portrayals were penetrating or poignant, Theophrastus (as well as later novelists) was free to write about his subjects without the constraints of psychological or scientific caution. Such lively and spirited characterizations most assuredly captured the interest of many, but they could also often mislead their readers about the true complexities of natural personality patterns. Although the beginning and ending of the Roman period cannot be sharply demarcated, it basically spanned a 12-century period from the 7th century B.C. to the 5th century A.D., when the last of the major Roman emperors was deposed. As a formal organization, the Roman Republic dated from the 5th century B.C. to the 3rd century A.D. The more cultured classes of Rome were determined to eliminate magic and superstition as elements in considering psychic processes. A mechanistic conception of mental disorders came to the foreground; it was fundamentally materialistic and opposed to all transcendental mythologies, which were regarded as superstitious beliefs that originated from fear and ignorance. Mental disorders were caused not by the action of mysterious forces, nor by biohumoral movements


or conflicts, but by the periodic enlargement or excessive tightening of the pores in the brain. In this corpuscular hypothesis, a derivative of the atomistic notions of Democritus of Greece, the task of the mental healer was to confirm and normalize the diameter of the pores. Persons with certain mental illnesses were seen as apathetic, fearful, and in a depressed mood, by what was called a laxum state. Those with other disorders presented an excited, delirious, and aggressive appearance; they were in a strictum state. If both sets of these symptoms co-Â�occurred, there was a mixtum state. A follower of the vitalist school of thought that adopted the concept of pneuma, the natural or animal spirit, as the physical embodiment of the soul, Aretaeus (30–90 A.D.) was little known in his time and was rarely quoted by fellow Roman scholars. This was probably owing to the fact that his works were written in the Ionic dialect rather than in Latin or Greek. Furthermore, his vitalistic philosophy, based on the fluidity of the soul’s nature, and adopted by Galen a century later, rivaled the more atomistic or solidistic corpuscular theory of his contemporary Roman thinkers. Scarcely familiar with the Greek language and its medical philosophies, Aretaeus was a born clinician who was retained as a physician for the ruling Roman classes. According to Aretaeus, the vicissitudes of the soul served as the basis of psychic disturbances. Specifically, the interconnecting linkages among “solid organs, the humours, and the pneuma” generated all forms of mental aberration. For example, anger and rage stirred the yellow bile, thereby warming the pneuma, increasing brain temperature, and resulting in irritability and excitability. Conversely, fear and oppression stirred black bile, augmenting its concentration in the blood, and thus leading to a cold pneuma and consequent melancholy. Disturbances of consciousness usually resulted from the sudden diminishing of the strength of the pneuma around the heart. Aretaeus’s descriptions of epilepsy were notably impressive. He spoke of its premonitory symptoms, such as vertigo and nausea, the perception of sparks and colors, and the perception of harsh noises or nauseating smells. Aretaeus also described the origins and characteristics of fanaticism; he formulated a primitive psychosomatic hypoth-


esis in stating that emotions could produce problematic effects on humoral metabolism, noting that “the black bile may be stirred by dismay and immoderate anger.” Similarly, he formulated what we speak of as cyclothymia in describing the alternation of depression with phases of mania. He stated, “Some patients after being melancholic have fits of mania .€.€. so that mania is like a variety of melancholy.” In discussing the intermittent character of mania, he recognized its several variants, speaking of one type as arising in subjects “whose personality is characterized by gayness, activity, superficiality, and childishness.” Other types of mania were more expansive in which the patient “feels great and inspired. Still others become insensitive .€.€. and spend their lives like brutes.” Perceptive observations by Aretaeus strengthened the notion of mental disorders as exaggerated normal processes. He asserted that a direct connection existed between an individual’s normal characteristics of personality and the expression of the symptoms the individual displayed when afflicted. His insightful differentiation of disorders according to symptom constellations (i.e., syndromes) was a striking achievement for his day. Although Hippocrates may have been the first to provide a medical description of depression, it was Aretaeus who presented a complete and modern portrayal of the disorder. Moreover, Aretaeus proposed that melancholia was best attributed to psychological causes (i.e., that it had nothing to do with bile or other bodily humors). As noted, he may have been the first to recognize the covariation between manic behaviors and depressive moods, antedating the views of many clinical observers in the 16th and 17th centuries. Aretaeus was also a major contributor to the humanistic school of thought in early Rome. Most notably, he introduced long-term follow-up studies of patients. He tracked their lifetime course, their periodic disease manifestations, and their return to a more normal pattern of behavior; in this regard, he anticipated the authoritative writings of Emil Kraepelin, who recognized the course of an illness as a key factor in discriminating a specific disorder from others of comparable appearance. He seriously studied the sequence and descriptive char-

acteristics of his patients, contending that a clear demarcation could be made between the basic personality disposition of a patient and the form in which a symptomatic and transient disorder manifested itself periodically. No less important was Aretaeus’s specification of the premorbid conditions of patients; he viewed these conditions as forms of vulnerability or susceptibility to several clinical syndromes. As Aretaeus phrased this, he found that persons disposed to mania are characteristically “irritable, violent, easily given to joy, and have a spirit for pleasantry or childish things.” By contrast, those prone to depression and melancholia were seen as characteristically “gloomy and sad often realistic yet prone to unhappiness.” In this manner, Aretaeus elaborated those essentially normal traits that make an individual susceptible to a clinical state. As Zilboorg and Henry (1941) have noted, the melancholia of Aretaeus is still observed in our time, although under different psychiatric labels. Owing to his observations of patients over extended periods of time, Aretaeus proposed a series of predictions about the general outcomes of different mental conditions. More than other physicians of his day, Aretaeus not only described psychological conditions with keen sensitivity and humane understanding, but (in a spirit more akin to recent scientific work) sought to compare various clinical syndromes and illuminate ways in which they could be differentiated. Claudius Galenus (Galen) (131–201; see Figure 1.2) was the last major contributor to adopt a psychological perspective in Rome. He preserved much of the earlier medical knowledge, yet generated significant new themes of his own. Galen lived more than 600 years after the birth of Hippocrates. A Greek subject of the Roman Empire, he was born in Asia Minor about 131 A.D. During his mature years, numerous radical political and cultural changes took place in Rome. Galen and his medical associates set out to synthesize primitive conceptions of disease with then-Â�modern methods of curing the sick. Following the ideas of Hippocrates, he stressed the importance of observation and the systematic evaluation of medical procedures, arguing against untested primitive and philosophical hypotheses in favor of those based on empirical test. As a follower

A Précis of Psychopathological History

Figure 1.2.╇ Galen.

of Aristotle as well as Hippocrates, Galen emphasized the data of experience, rather than logical hypotheses that were devoid of factual evidence. Unfortunately, he doubted that environmental and psychological factors could affect the course of human disease. Although Galen avoided philosophical themes concerning the nature of illness, he nevertheless proposed a principle termed spiritus anima, in which he asserted that humans possessed an extraphysical life-Â�giving force; this thesis was based on his efforts to distinguish organic from inorganic matter. Galen’s conception of psychic pathology was based on the physiology of the central nervous system. He viewed clinical symptoms as signs of dysfunctional neurological structures and characterized mental diseases as “a concourse of symptoms,” among which a specifically pathognomonic one could be isolated. According to his organicÂ�functional approach, mental symptoms originated from the pathogenic action of a toxic, humoral, vaporous, febrile, or emotional factor that affected the brain physically and then altered certain of its psychic functions. Consonant with the beliefs of his time, Galen believed that the activities of the mind were prompted by animal spirits that carried out both voluntary and involuntary actions. These animalistic spirits (pneuma) were divided by Galen into two groups: those that controlled sensory perceptions and motility,


whose damaging effects would cause neurological symptoms; and those that had more directive functions, such as coordinating and organizing imagination, reason, and memory. To him, most psychiatric symptomatology stemmed from alterations of the second group of functions. In describing catatonic psychosis, Galen suggested a paralysis of the animal spirits in which the imaginative faculty was “blocked or incomplete.” As far as the syndrome of hysteria was concerned, he disagreed strongly with Hippocrates’s uterocentric view. Galen asserted that hysteria, on the basis of his own clinical examinations, could not be a disease that reflected the uterus “wandering agitated in the body.” As he saw it, hysterical symptoms were provoked by the toxic action of vapors that formed in the normal uterus and vagin*; it arose from the stagnation of sem*n, owing to a lack of sufficient sexual intercourse. The disease therefore signified a lack of sexual hygiene. Galen’s stature grew over the next millennium—so much so that his views were thought to be sacrosanct. His writings were summarized and commented on by many lesser physicians, most of whom were recognized as being wrong-Â�headed; indeed, their books were often referred to as “wretched treatises.” Some of these post-Galen compilations were not based on his work at all, but dishonestly carried his name for its ability to promote the sale of untenable or alien ideas. Although many of his notions were diluted by the passage of time or refuted by empirical knowledge, his vast contributions must be considered significant, in that no other figure in history exercised so extended an influence on the course of medicine. Later in Roman history, there emerged an organized theology known as Christianity, including faith healing, magic, and superstition. The doctrine of the early Christian church became the dominant approach to thought, medicine, and mental healing in the Western world until the 17th century. Most of the populace remained illiterate during this period. Education was religious, otherwise inchoate, and of dubious value. The idea of a scientific basis for understanding mental disorders barely appeared on the scene. Faith was the all-Â�powerful guide. During the first two to three centuries A.D., a distinction was made between psy-


chologically normal individuals who doubted the dogma of the church’s ideology, and those whose “peculiar” beliefs arose not out of opposition, but out of a mental affliction. Nevertheless, both groups were considered guilty of heresy and subjected to punishment. In a similarly irrational twist, others’ implausible or nonsensical behavior ostensibly demonstrated their fervent adherence to church authorities and their dogma. Such persons were venerated. It was not long thereafter that the works of Aristotle and other Greek philosophers were condemned. Christianity in the 3rd century led physicians to assume a moralistic and judgmental approach to psychic pathology. Unable to escape the growing spirit of superstitiousness, they proposed that mental cases were definitely the products of mystical events that could not be understood in the natural world. More seriously, they adopted the ancient belief that demons often appeared under the guise of confused humans, and that it was the job of physicians to identify and to “eliminate” them. In this and other similar matters, they laid the groundwork for a return to the age of supernaturalism and superstitions; they were nevertheless thought well of until the close of the 17th century. Aurelius Augustine (354–430) was a key figure in the transition from early Roman thought to the Middle Ages. Better known as St. Augustine of Hippo, we can see in his writings an effort to synthesize the Greek and the new Christian perspectives on mental maladies. Perhaps the most influential philosopher of his time, Augustine set the foundation and tone of Christian intellectual life for centuries to come. To him, all knowledge was based on the belief that only God could provide the ultimate truth, and that to know God was the ultimate goal. To think otherwise, as Augustine averred, would not only be vain, but would assuredly lead to error and corruption. Individuals, as children of God, would in their faith begin to understand the very nature of life, and thereby would be able to lead a life of grace and honor. The Early Muslim World Three major medical figures from the Muslim world of the Middle East around the

end of the first millennium A.D. are worthy of note: Rhazes, Unhammad, and Avicenna. Each proposed helpful ideas that came to represent a fresh and innovative point of view concerning mental illness. Rhazes (860–930) lived during the late 9th and early 10th centuries and wrote textbooks dealing with a wide variety of medical, psychological, philosophical, and religious subjects. In contrast to the predominant religious orientation of Baghdad, Rhazes strongly argued against the notion of a demonological concept of disease and the use of arbitrary authority to determine what is scientific and what is not. He attacked the superstitious religious beliefs of his contemporaries and was strongly in favor of developing a rational schema for understanding all disorders. Empirically oriented, he nonetheless subscribed to the theory of the four elements originally developed by Empedocles and Hippocrates. Unhammad (870–925) was a contemporary of Rhazes who provided intelligent descriptions of various mental diseases. The observations he compiled of his patients resulted in a nosology that was the most complete classification of mental disorders in its day. Unhammad described nine major categories of mental disorders, which, as he saw it, included 30 different diseases. Among the categories was an excellent description of anxious and ruminative states of doubt, which correspond in our thinking today with compulsions and obsessions. Other categories of mental disease were judged by Unhammad to be degenerative in their nature; a few were associated with the involutional period of a man’s life. The term used by the Greeks for mania was borrowed to describe states of abnormal excitement. Another category, most closely associated with grandiose and paranoid delusions, manifested itself, according to Unhammad, in the mind’s tendency to magnify all matters of personal significance, often leading to actions that proved outrageous to society. A most significant and influential philosopher and physician of the Muslim world was Avicenna (980–1037), often referred to as the “Galen of Islam,” largely as a consequence of his vast and encyclopedic work called the Canon of Medicine. The Canon became the medical textbook chosen throughout European universities from the 10th through

A Précis of Psychopathological History

the 15th centuries. However, Avicenna was not regarded as a highly original writer, but rather as a systematizer who encompassed all knowledge from the past that related to medical events. Similar to Galen, Avicenna noted the important connection between intense emotions and various medical and physiological states, although he fully accepted Hippocrates’s humoral explanations of temperament and mental disorder. To his credit as a sophisticated scholar of the brain, Avicenna speculated that intellectual dysfunctions were in large part the results of deficits in the brain’s middle ventricle, and asserted that common sense and reasoning were mediated by the frontal areas of the brain.

The Middle Ages The enlightened ideas of Hippocrates were submerged for centuries after the death of Galen and the fall of the Roman Empire. During the thousand years of the so-Â�called Dark Ages, superstition, demonology, and exorcism returned in full force and were further intensified by sorcery and witch burning. With few dissenting voices during this period, the naturalism of the Greco-Roman period was all but condemned or distorted by notions of magic. Only in the Middle East did the humane and naturalistic aspects of Hippocratic thought remain free of the primitivism and demonology that overcame Europe. Signs for detecting demonic possession became increasingly indiscriminate in the Christian world. During epidemics of famine and pestilence, thousands wandered aimlessly until their haggard appearance and confusion justified the fear that they were cursed. The prevalent turmoil, the fear of one’s own contamination, and the frenetic desire to prove one’s spiritual purity led widespread segments of the populace to use these destitute and ill roamers as convenient scapegoats. As the terrifying uncertainties of medieval life persisted, fear led to wild mysticism and mass pathology. Entire societies were swept up simultaneously. Epidemic manias of raving, jumping, drinking, and wild dancing were first noted in the 10th century. Referred to as “tarantism” in Italy, these epi-


demic manias spread throughout other parts of Europe, where they were known as St. Vitus’s Dance. During the early Middle Ages, before later catastrophes of pestilence and famine, few people with mental illnesses were totally destitute. Monasteries served as the chief refuge for such individuals, providing prayer, incantation, holy water, relic touching, and mild exorcism as prescriptions for cure. As the turmoil of natural calamity grew more severe, mental disorders were equated increasingly with sin and Satanic influence. Significant advances were made in agriculture, technology, and architecture during the Middle Ages, but the interplay between changing theological beliefs and naturalistic catastrophe speeded acceptance of the belief that “madness” and “depravity” were the devil’s work. At first, it was believed that the devil had seized mentally ill individuals against their will, and such individuals were treated with established exorcistic practices. Soon, however, the afflicted were considered willing followers of Satan; classed now as witches, they were flogged, starved, and burned. Among the major tenets of this medieval mythology was a belief that an international conspiracy, based on Satanic forces, was bent on destroying all forms of Christianity. The agents of this widespread conspiracy were witches, who not only worshipped Satan at secret meetings, but attempted to desecrate Christian symbols and beliefs, as well as to engage in murder, cannibalism, and sexual orgies. The ideas of a demonic and Satanic conspiracy existed first and foremost in the imagination of the religious leaders of the day. It was Pope Gregory IX who established the Inquisition in 1233 to root out witches, heretics, and all other agents of Satan, who he asserted were setting out rapidly to destroy the clerical and political orders of the Church. Those with an administrative status possessed the legal right to judge which aspects of Satanic witchcraft would be deemed demonic. It was not only higher-order religious leaders who conveyed this dogma; the common people took these belief systems to heart, as well. From the 15th through the 17th century, demonic possession and exorcism became common phenomena among the masses. In the postmedieval period, both Catholics and Protestants believed that


witches, fueled by Satanic forces, would send demons to possess those judged to be undesirable. It was the task of religious authorities to coerce those possessed by demons to admit that they were witches. These individuals could justly be arrested and tortured, especially if they “confessed” to their involvement in these nonexistent Satanic conspiracies. “Witch finders” soon became prominent guardians of the faith, prompted by religious authorities who sought to undo the political powers of their ostensible “enemies.” Encouraged by the 1484 Summis Desiderentes Affectibus, in which Pope Innocent VIII exhorted the clergy to use all means for detecting and eliminating witchcraft, two inquisitional Dominicans, Heinrich Kramer and Johann Sprenger, issued a notorious manual titled Malleus Malefacarum (The Witches’ Hammer). Published between 1487 and 1489, this “divinely inspired” text set out to prove the existence of witchcraft, to describe methods of identification, and to specify the procedures of examination and legal sentencing. Malleus Malefacarum reflected the spirit of its time, even though it was published in the early stages of the Renaissance and at the threshold of the Reformation. Here the conflict between paganism and Christianity, between magic and a monotheistic outlook, had not ceased to be a burning issue (in more than one sense of the word). As the ancient idols and deities were torn down from their pedestals, demons nevertheless retained their grip on the minds of the ordinary people. Idols and deities were relegated to the role of fallen angels, but devils and evil demons continued to reside in the human unconscious, and belief in them continued to be widely embraced. With torture recommended as a means of obtaining confession, and with feelings of guilt and hopeless damnation characteristic of many of the afflicted, the inevitable consequence for most persons with mental illnesses was strangulation, beheading, or burning at the stake. Unredeemed by good sense and wise judgment, this barbaric epidemic swept Protestant and Catholic countries alike, including several American colonies. Although the last execution of a witch occurred in 1782, the bewildering notion that mentally ill individuals were in league

with the devil persisted in popular thought well into the 19th century. It was in the 15th century that the medieval period began to be gradually transformed into what we view today as the modern world. Slowly but persistently, the importance of human emotions and strivings became a significant element to guide intellectual thought, ultimately replacing the medieval belief that the revelation of deeper human truths were beyond human capabilities. Psychological processes became increasingly humanized; opportunities to study human beings as biological rather than purely spiritual organisms permitted these processes to be considered as aspects of natural rather than metaphysical science. Christianity had begun to lose its spirit and vitality; although the supernatural world still existed in human minds, it had lost much of its power, increasingly ruled by static and rigid belief systems and symbols. Desiderius Erasmus (1465–1536) was a sincere churchman who asserted a new humanism. He attacked the formalism and the corruption of the church, which he judged as sterile and possessed of rituals that were divested of their purpose and humanism. As Robinson (1976) has noted, his “psychology” was both practical and wise, expressed with verve and clarity in Erasmus’s Colloquies. Here he pricked vanity, exorcised exorcism, lamented superstition gleefully, and guided individuals to their duty to adhere to the simple and humane lessons of Christ’s life, rather than to behaviors that would obscure or deceive his worthiness. In his essays and letters, Erasmus, neither scientist nor formal philosopher, addressed the everyday world, seeking to expose its vanities, follies, charlatans, and warmongers. His was the attitude of a Renaissance humanist, with a fine mind and a sympathetic heart. So, too, was the humanistic outlook of the Spanish Jew Juan Louis Vives (1492–1540), who contributed fundamentally to educational reforms and evinced a passionate concern for the welfare of those with mental illnesses, who were routinely incarcerated and maligned.

The Renaissance and Beyond Gradually, the horrors of the Inquisition were left behind. In the 16th century, the

A Précis of Psychopathological History

work of intelligent and humanistic thinkers slowly awakened humankind from its long slumber. Zilboorg and Henry (1941) have written that for several centuries, philosophers had repeatedly stated that human beings should be studied and not their souls; these scholars slowly convinced themselves by listening to their own voices. These scholars were not physicians, because physicians had turned their attention to the new anatomy and physiology of animals and cadavers, rather than the emotions and natural states of living humans. The waning of medieval supernaturalism and the advent of the liberating Renaissance era had numerous effects upon the emergence of psychological thought. The Renaissance broke the hold of medieval dogma upon the mind of early clinicians. It also opened up new nonphilosophical pathways for purely psychological ventures and inquiries into the general character of human nature, as well as the substantive nature of mental disorders. The Theories of Paracelsus Paracelsus (1493–1541) would have been an extraordinary person in any historic age, but, given his time, he looms as a strange if not rare blend of the mysticism of the past with the practicality of his day. Paracelsus’s actual name was Theophrastus Bombastus Von Hohenheim. Perhaps in anticipation that he would be a courageous and intrusive battler all of his life, he shortened his name to simply Paracelsus, even if its selection was rather pretentious. Specifically, he adopted this name to suggest that his views were superior to those of Celsus, the chief medical authority of ancient Rome. Others have suggested that he gave himself this name in order to show that he was surpassing the encyclopedists and medical methodists of his time—in other words, that he intended to blaze a new trail by his adventurous approach to mental disorders. Like most thinkers of his day, Paracelsus was a believer in divination from the stars and in the healing powers of such preparations as powdered Egyptian mummy. As such, he was both an astrologer and an alchemist. For some of his disciples, Paracelsus was the towering medical figure of the Renaissance period, comparable to such other con-


temporary luminaries as Leonardo da Vinci, Copernicus, and Shakespeare. To most historians of today, he is regarded as an imaginative adventurer, if not a charlatan, and most are inclined to view his contributions to scientific medicine as modest at best. Among his works were efforts to test the effects of various chemical agents to treat several medical conditions (attempts not unlike the activities of pharmaceutical firms today). Although he made no lasting discoveries, he was an inventive and creative pioneer. Nevertheless, the whimsies he had proposed were consigned largely to the rubbish heap. Despite a number of sound insights, he dissipated much of his energy combating colleagues who did whatever they could to make his life unbearable. When Paracelsus interrupted his mystical flights of fantasy to deal with his medical opportunities, he spoke in a voice akin to that of a seeker of scientific truth, despite his rebellious defiance of ancient traditions and scholastic dogma. Most notably, he denounced the cruelties of the Inquisition, stating that “there are more superstitions in the Roman Church than in all these poor women and presumed witches.” In his rejection of the views of the clergy regarding the sources of mental disorders, Paracelsus (1567/1941) wrote: In nature there are not only diseases which afflict our body and our health, but many others which deprive us of sound reason, and these are the most serious. While speaking about the natural diseases and observing to what extent and how seriously they afflict various parts of our body, we must not forget to explain the origin of the diseases which deprive man of reason, as we know from experience that they develop out of man’s disposition. The presentday clergy of Europe attribute such diseases to ghostly beings and threefold spirits; we are not inclined to believe them.

Paracelsus was the first physician to lay out a systematic classification of disorders that abandoned the habit of categorizing disorders by beginning with the head, then working down step by step to the feet. His mental health classification was outlined in a treatise titled “On the Diseases Which Deprive Men of Health and Reason.” Here whole groups of persons with mental disorders were identified, notably lunatici, insani,


vesani, and melancholici. Lunatici suffered from disorders stemming from their reactions to the phases of the moon. Insani suffered from disorders identifiable at birth and clearly derived from family heritage. Vesani were poisoned or contaminated by food or drink. Melancholici, by virtue of their temperament, lost their ability to reason accurately. In addition to these four forms of mental illness, Paracelsus identified others as obsessi—that is, obsessed by the devil. In this latter formulation, Paracelsus was dissenting from the dogmatic view of earlier centuries in which a devil obsession lay at the heart of all mental disorders. As he perceived it, numerous sources of mental dysfunction existed, only one of which could be traced to demonic preoccupations; he saw the other disorders as problems of defective thought processes, rather than as consequences of supernatural powers. Physiognomy and Phrenology before, during, and after the€Renaissance “Physiognomy,” the art of interpreting people’s psychological characteristics from aspects of their physical characteristics, was present in ancient times, reaching its peak of study in the 2nd century A.D. Advocates of physiognomy assumed that inner traits of people are expressed in their outer physical features, especially the face. The great thinkers of Greece made formal efforts to systematically interpret physiognomic characteristics—for example, in Pythagoras’s 6th-Â�century B.C. writings, and later in Aristotle’s Analytica Priora (Tredennick, 1967) and Historia Animalium, where he wrote: “Persons who have a large forehead are sluggish, those who have a small one are fickle; those who have a broad one are excitable, those who have a bulging one, quick tempered” (Peck, 1965, I, VIII, 891b, p.€39). Physiognomica (Hertt, 1936)—also attributed to Aristotle, but more likely written by his followers—Â�examined parallels between the physiques of men and animals, to compare different ethnic groups, and to investigate the relationship between bodily characteristics and temperamental dispositions. Among the useful signs recorded were the movements, shapes, and colors of the face; the growth of hair; the smoothness of

skin; the condition of the flesh; and the general structure of the body. Sluggish movements denoted a soft disposition, quick ones a fervent temperament; a deep voice denoted courage, a high one signified cowardice. The writers were wise enough to note that it would be foolish to base a judgment on any one of these signs. Centuries later, Leonardo da Vinci (1452–1519) made similar physiognomic proposals in his Treatise on Painting, in which he explored relationships between emotional states and overt facial expressions. In the 16th century, Giovanni Battista della Porta (1535–1615) published a book titled De Humana Physiognomia (1586), derived from Aristotle’s writings, which included many drawings designed to show similarities between humans and animals. For example, a person who looked leonine ostensibly possessed the courage, strength, and will of a lion. This book proposed the theory that every person’s head resembled a specific animal’s head, thereby suggesting that the person possessed the same personal characteristics as that animal. Another work by Porta, Natural Magick (1558/1957), outlined similar speculations by a number of his contemporary colleagues. No less speculative was Porta’s Phytognomonica (1588), in which he addressed matters of vegetable physiognomy—that is, the art of determining the inner nature of plants on the basis of their exterior appearance. In his five-Â�volume work, Les Charactères des Passions (1640), eminent French physician Marin Cureau de la Chambre (1594– 1669) wrote: the resemblance Man has with other Creatures .€.€. teacheth us that those who have any part like to those of beasts, have also their inclinations .€ .€ . that men who have anything of a feminine beauty, are naturally effeminate; and that those women who have any touch of a manly beauty, participate also of manly inclinations.

Burdened with the prejudices of his day, de la Chambre (1640) was nevertheless a highly insightful physiognomist, addressing in detail the significance to be found “in the motions of the eyes, the inflection of the voice, the color of the lips,” and so on. Unfortunately, he could not help but draw

A Précis of Psychopathological History

upon astrological influences, speculating on the power, especially of the moon upon the brain, “causing it to increase or decrease in volume upon whether the moon is waxing or waning.” A distinguished philosopher and jurist, Christian Thomasius (1655–1728) helped inaugurate the period of German enlightenment, founded the University of Halle, and asserted that philosophy should concern itself with practical matters of everyday life. A prolific author, Thomasius wrote only briefly on physiognomy, drafting an essay entitled Recent Proposals for a New Science for Obtaining a Knowledge of Other Men’s Minds (1692). Basing his ideas on the work of de la Chambre, Thomasius recommended that observation can be most useful when obtained through personal conversation with one’s subject; he also cautioned that observers must distinguish between genuine and affected emotions. A theorist of physiognomy in the late 18th century, Johannes Kaspar Lavater (1741– 1801), asserted unequivocally the existence of a relationship between fixed aspects of the body’s surface and a person’s character. In his well-Â�received book Essays on Physiognomy (1789), published in four lavish volumes, Lavater claimed that physiognomy was truly a science because it offered law-like regularities and depended on empirical observation. In characterizing the trait of obstinacy, Lavater wrote: The higher the forehead, and the less the remainder of the countenance, the more knotty the concave forehead, the deeper sunken the eye, the less excavation there is between the forehead and the nose, the more closed the mouth, the broader the chin, the more perpendicular the long profile of the countenance— the more unyielding the obstinacy: the harsher the character.

Though similar in many respects to classical approaches in physiognomy, a new “scientific” model known as “phrenology” emerged in the late 18th century. Both approaches drew inferences about character and personality from external bodily features—Â�physiognomy from facial structure and expression, phrenology from external formations of the skull. Their underlying assumptions, however, were quite different.


Physiognomists believed that a person’s inner feelings and characteristics were expressed in facial features, voice, and so on. Phrenologists made no assumptions as to the external expression of varied dispositions. Their two fundamental assumptions were unusual for their era: First, that different mental functions were located in different regions of the brain; and, second, that the skull’s external topography reflected the magnitude of these functions. This was the first “scientific” effort made to analyze the underlying brain structure from which character and personality might be derived. Despite its discredited side, phrenology, as Franz Joseph Gall (1758–1828) proposed it, was an honest and serious attempt to construct a neurological substrate in the brain to undergird a science of character depiction. Although numerous writers such as Vesalius, Willis, and Stensen in medieval times (Millon, 2004) had speculated and explored brain structures as the center of mental functioning, Gall took this view in an original direction. Most early characterologists conceived the brain as a locale where the immaterial soul might influence bodily activities. Gall asserted not only that the brain was the mind in an explicitly material sense, but that different regions subserved different dispositions. Gall identified 27 different “organs” in the brain that undergirded separate psychological tendencies. Through “reading” the skull (usually by running one’s hands over the head), one could identify different enlarged organs. Gall went to prisons and lunatic asylums to read skulls and collect data on correlations between protuberances in certain locations and personality traits. Gall referred to his research on brain physiology as “organology” and “crainoscopy,” but the term “phrenology,” which his younger associate Johann Spurzheim coined, came to be its popular designation. As noted, the rationale that Gall presented for measuring contour variations of the skull was not illogical. In fact, his work signified an important advance over the naive and subjective studies of physiognomy of his time, in that he sought to employ objective and quantitative methods to deduce the inner structure of the brain. He concluded, quite reasonably, that both the intensity and character of thoughts and emotions would correlate with varia-


tions in the size and shape of the brain or its encasem*nt, the cranium. That this gross expression of personality proved invalid is not surprising when we think of the exceedingly complex structure of neuroanatomy. Views from the Later Renaissance to€the 17th Century A chronically depressed clergyman and reclusive scholar, Robert Burton (1576–1640), wrote a single major work of extraordinary insight and sensitivity in 1621, titled The Anatomy of Melancholy. Burton’s introspective accounts of his moods contained a wealth of impressive clinical analyses. He also sought to record the behavior and emotions of others, recognizing patterns similar to his own moodiness and eccentricity. This great volume of work, despite rambling irrelevancies and inaccuracies, makes fascinating reading today, as may be judged from the following excerpt: It is most absurd and ridiculous for any mortal man to look for a perpetual tenure of happiness in this life. Nothing so prosperous and pleasant, but it hath some bitterness in it, some complaining, some grudging; it is all a mixed passion, and like a chequer table, black and white men, families, cities, have their falls and wanes; now trines, sextiles, then quartiles and oppositions. We are not here as those angels, celestial powers and bodies, sun and moon, to finish our course without all offence, with such constancy, to continue for so many ages: but subject to infirmities, miseries, interrupted, tossed and tumbled up and down, carried about with every small blast, often molested and disquieted upon each slender occasion, uncertain, brittle, and so is all that we trust unto. (Burton, 1621, p.€261)

Although his perspective was limited, Burton did establish a classification system, one that differentiated melancholy from madness—a distinction akin to our differentiation of neuroses from psychoses. He outlined the following general categories: (1) diseases emanating from the body; (2) diseases of the head (primarily the brain); (3) madness (mania); and (4) melancholy, for which Burton further distinguished melancholy of the head, the body, or the bowels, and identified the major sources of melancholy (e.g., excessive love, excessive study, intense preoccupation with religious themes).

Burton’s introspective awareness of his own personal sadness and depression led him to recognize the sources of his own melancholy. He recognized guilt as a major element, despite his exemplary lifestyle. Other causes of melancholy included bodily deterioration and old age; bad diets; sexual excesses; idleness; solitariness; and an overpreoccupation with imagination, fears, shame, and malice. Burton clearly stated that melancholy could be engendered by a wide range of human frailties and life circ*mstances. Among the many topics that Burton included in his book on melancholy, he touched on a variety of mental aberrations that we recognize today as obsessions and compulsions. Thus he wrote of an individual who dared not to go over a bridge, come near a pool, rock, steep hill, lie in a chamber where cross-beams were, for fear he’d be tempted to hang, drown, or precipitate himself. In a silent auditorium, as at a sermon, he was afraid he shall speak aloud at unawares, something indecent, unfit to be said. (1621, p.€253)

Burton anticipated what ultimately became the core of modern psychotherapy— that is, engaging a patient in a dialogue with a trusted and sympathetic outsider. But because he was not part of the medical establishment, his proposals had little effect on the course of mental health study of his time, despite the brilliance of his book. A man of great intensity and imagination, Thomas Willis (1621–1675) was the originator of the term “neurology”; he also generated the term “psychology” to designate the study of the so-Â�called “corporeal soul.” Arguably the most significant founder of what came to be referred to as “biological psychiatry,” he considered most ailments to be disorders of nerve transmission, rather than diseases of the blood vessels. He is perhaps best known by the circuit of arteries located at the base of the brain, known today as the “circle of Willis.” In 1664, Willis published a major book on the history of the brain sciences, entitled Cerebri Anatome. It was a work of considerable scope and insight, and was for many decades thereafter without equal in the field. The title suggested that the book was limited to anatomy. However, Willis, a thoroughly educated Oxford physician, concerned himself not only with brain functions, but with

A Précis of Psychopathological History

their behavioral consequences. Willis proposed also that vital and involuntary systems existed in the brain that were mediated not by the higher centers of the brain, but by the “cerebellum.” The detailed articulation of the functional segments of the brain, grounded in comparative anatomic precision, was enriched by his clinical observations. Drawing ideas from existing theories, his work, both speculative and empirical, stimulated many another neuroanatomist. In his clinical work, Willis (1664/1978) reported his observation of a sequence in which “young persons who, lively and spirited, and at times even brilliant in their childhood, passed into obtuseness and hebetude during adolescence.” Thus Willis anticipated by two centuries an idea more fully developed by Benjamin Morel, who termed this behavioral course “dementia praecox.” To his credit, Willis rejected the idea of a “wandering womb” that ostensibly led to the syndrome of hysteria. In his view, the brain functioned as the center of all mental disturbances, and the various nerves emanating from the brain served to connect this overarching organ to the rest of the body. Willis, like most others of his time, spoke of processes generated by “animal spirits”—that is, the soul, which somehow or other could be sucked out of the brain. Also of note was Willis’s observation that melancholia and mania frequently coexisted within the same person, who would shift erratically from an excited state to one of depression. This observation contributed to what we now refer to as bipolar disorder and/or manic–Â�depressive psychosis. Willis’s clinical observations were uncontaminated by formal theories. His accurate inferences were based on repeated observations of patients over time—that is, on the long-term course of their difficulties. Included in Willis’s classification system were some 14 categories, of which several were primarily neurological. His system, published in De Anima Brutorum (1672/1971), specified three major impairments: morosis, mania, and melancholia (each encompassing several subcategories). It also encompassed a number of neurological disorders, such as headache, insomnia, and vertigo. Thomas Sydenham (1624–1689), a colleague of the philosopher John Locke, held strongly to the view that hypotheses should be set aside in favor of closely observing all forms of natural phenomena, such as vari-


ous medical diseases. As he put it, too many writers had saddled fairly distinct diseases with excessive features that stemmed from their overblown interpretations. Sydenham did not trust books, believing only what he could see and learn from his own bedside observations (Comrie, 1922). Locke preached that all reliable knowledge came from observation. In his work, Sydenham came to typify the 17th-Â�century empiricist emphasis in England. Especially informative were Sydenham’s contributions to the description of hysteria. His observations of hysterical patients enabled him to recognize the variations of conversion symptoms among patients with paralysis and pain, as well as to speculate on the operation of intense but unconscious emotions. The precision of his descriptions of hysterical phenomena was so comprehensive that little can be added today to what he said over three centuries ago. He recognized that hysteria was among the most common of chronic diseases, and observed that men exhibited the symptom complex no less than women. He averred that hysterical symptoms could simulate almost all forms of truly organic diseases; for example, he noted that a paralysis of the body might be caused by stroke, but could also be found in a hysterical hemiplegia “from some violent commotion of the mind.” He spoke of hysterical convulsions that resembled epileptic attacks, psychogenic palpitations of the heart, and hysterical pain that could be mistaken for kidney stones; he also suggested that differential diagnosis between real biological diseases and those generated by the mind could only be made if the patient’s psychological state could be thoroughly known. He was among the most successful in illustrating that emotions can generate and simulate physical disorders. In his efforts to formulate a syndromal pattern for numerous disorders, he extended the range of his observations to include not only the patient’s dispositions, emotions, and defenses, but the family context within which they arose. In this way, he sought to determine the overall pathogenesis of certain syndromes, largely through the use of both physical and psychological phenomena. What was most informative was Sydenham’s recognition that a syndromal picture rarely developed from a single pathogenic agent, be it a humoral imbalance or a systemic disturbance of the


body. In fact, to Sydenham, multiple influences operated simultaneously on a patient, each of which took a somewhat different turn and produced a somewhat different appearance in the same disease process. He strongly believed in syndrome complexes rather than in a distinct or singular expression of a disorder. As a consequence, all physicians were eventually trained to consider a wide range of elements, which together play a partial role in generating disease. However, Sydenham believed that hypotheses and philosophical systems should be set aside to ensure that pathological phenomena were observed with reliability and accuracy. Particularly notable was Sydenham’s belief in nature’s own healing processes. These natural remedies of the body would not invariably solve a problem because they were often delayed or displaced. Included among the healing processes of nature, according to Sydenham, were a variety of wellÂ�established “excretions, eruptions, and fevers.” Sydenham’s speculations were based on comprehensive observations, which comprise the most modern methods for investigating mental illness and diagnosing specific clinical syndromes. He also emphasized the importance of identifying the antecedent emotional factors that may lead to the development of mental disorders. Insightfully, he observed the interplay between personal emotions and social pressures. Born in Germany, Georg Ernst Stahl (1660–1734) wrote his doctoral dissertation in his early 20s. In it, he expressed the view that the then-Â�prevalent theory of animal spirits was essentially incorrect, and that the various processes of the mind stemmed from a life-Â�giving force, to which he applied the term “soul.” However, Stahl’s soul was not the supernatural phenomenon that characterized ancient and medieval thinking; it represented the source of energy of all living organisms, both human and animal. Stahl’s life force was not notably different from Freud’s conception of the libido. It was the sum total of the nonmaterial side of humans and animals, which, together with nature, had the power to effect desired cures. Hence Stahl’s “soul,” which in many ways is equivalent to our “psyche,” was able to perform a variety of functions that could either bring on or stave off various diseases. Many present-day scholars consider Stahl the originator of the distinction between or-

ganic and functional mental disorders. To him, mental disorders were the result of neither physical, mechanical, nor supernatural forces, but were in fact essentially psychogenic. Stahl was appalled by the sharp demarcation of body and mind. Not only did he judge this dichotomy to be unjustified, in that it hindered a fundamental understanding of disease unity, but it was especially problematic in understanding the complexity of forces involved in mental diseases. He advocated a synthesis of physical and mental phenomena.

The 18th and 19th Centuries As clinics and hospitals began to record case histories and detail observations, physicians could identify syndromal groupings (i.e., clusters of symptoms) and classify them into disease entities. The success with which botanical taxonomists had systematized their field by the 18th century provided additional impetus to the trend toward categorizing symptom clusters into a formal psychiatric taxonomy or nosology. A second major trend within biological medicine—the view that mental disorders might result from organic pathology—can be traced to the early writings of Hippocrates, Aretaeus, and Galen. With the advent of valid anatomical, physiological, and biochemical knowledge in the early 18th century, and the discovery in the 19th century of the roles played by bacteria and viruses, the disease concept of modern medicine (including the view of mental illness as a disease) was firmly established. Efforts at developing somatic (e.g., electrical, chemical, surgical) treatment methods followed naturally. Although these three stages—Â�diagnostic classification, biological causation, and somatic treatment—Â�rarely proceeded in a smooth or even logical fashion, they characterized progress in psychopathology and continue today to guide neuroscientists who follow the medical and biological tradition. These scientific and medical activities, however, presuppose a classification system (i.e., a taxonomy) that is not only logical but valid. Unfortunately, physicians classified diseases long before they understood their true nature. Such nosologies have persisted because of widespread or authoritative use; however, they rested most often

A Précis of Psychopathological History

upon unfounded speculations or, at best, judicious but essentially superficial observations. Criticism of premature nosological schemes is justified, given the frequent slavish adherence to them. On the other hand, there is no reason to overlook the potential value of a taxonomy, or to abolish a sound classification system that may serve many important purposes. With the waning role of supernaturalism and the advent of liberating thought during the Renaissance, several enlightened thinkers of the 16th and 17th centuries began to explore ideas related to a realistic classification of mental disorders, as reported above. Perhaps the leading taxonomist of the 18th century was François Boissier de Sauvages (1706–1767). He had completed his dissertation at the age of 20, defending the teachings of the faculty at his medical school, Montpelier. Conservative in mind but clearly distant from the demonological prejudices that were prevalent in most academic circles of his day, he thought that all mental diseases were located in distinct anatomical regions. Moreover, he believed that the “will” had much to do not only with the generation of mental aberrations, but also with their ultimate treatment. He also believed that physicians had a responsibility to shape or guide individual behaviors; otherwise, there would be no social compact or personal justice. de Sauvages followed Linnaeus in seeking to create an encyclopedic framework for the many categories of mental disorder. He outlined 10 classes, 295 genera, and 24,000 species, spending the better part of his life immersing himself in the large body of medical knowledge that had accumulated from early times. His urge to catalog the bewildering and scattered array of human disorders can be viewed as an effort to surmount the rather spotty and supernatural beliefs that typified earlier thought. Besides being a physician, de Sauvages was a botanist. Most of his colleagues spent their time arranging plants and animals in a clearly articulated and “evolutionary” system; the latter was a new departure that did not achieve its fullest impact until the work of Darwin a century later. The details of de Sauvages’s presentations were first published in a small book, Treatise de Nouvelles Classes de Maladies (1731). Included in his broad classification were such illnesses as fevers, inflammations, spasms, breathing


disturbances, weaknesses, pains, and dementias. Dementias, which comprised the bulk of mental diseases in this book, were organized into four types: those of extraÂ� cerebral origin, disturbances of the instinctual and emotional life, disturbances of the intellectual life, and irregular eccentricities and follies. de Sauvages completed the three-Â�volume Nosologie Methodique (1771) late in life, and it was published several years after his death. In this work, de Sauvages made available to others the complete model he had constructed; this model was used as an orderly classification for decades, if not centuries, to come. In this comprehensive volume, de Sauvages organized all forms of mental illness. For example, he grouped the syndrome of melancholia into numerous species (e.g., religious, imaginary, extravagant, vagabonding, enthusiastic, and, sorrowful). In the late 1770s and early 1780s, a distinguished physician and professor at the University of Edinburgh, William Cullen (1710–1790), became a most influential nosologist; he drew upon the work of de Sauvages, but extended the Linnean themes even more comprehensively. The Frenchman Philippe Pinel, who played a well-Â�publicized role in the movement toward humane mental treatment, used Cullen’s nosology as the basis for his “scientific” teachings. In contrast to most of his colleagues, Cullen became a popular educator because he refused to lecture in esoteric Latin and spoke in the vernacular. In his first major work, the four-Â�volume First Lines of the Practice of Physick (1777), Cullen made an effort to categorize all the then-known diseases (both psychological and physical) in line with the symptoms they displayed, the methods by which diagnoses were generated, and the therapy that might best be applied. Cullen was a notable pioneer of neuropathology and, in keeping with his orientation, believed that most pathological conditions of the mind should be attributed to diseases of the brain. Despite this orientation, he recognized that life experiences often influenced the character in which these biologically grounded diseases were expressed. Cullen proposed the term “neuroses” to represent neurologically based diseases. Most etiologically obscure mental illnesses were labeled neuroses, ostensibly to represent diseases of nerves that were inflamed and


irritable. As he perceived it, neuroses were affections of sense or motion that stemmed from a disharmony of the nervous system. Into the general category of neuroses, Cullen subcategorized four variants: those representing a diminution of voluntary motion, those representing a diminution of involuntary activity, those representing disturbances in the regular motions of the muscles or muscle fibers, and disorders of judgment. Along with Cullen, Robert Whytt (1714– 1766) played a large role in providing Scottish physicians of his day with a classification system of “neurotic” individuals. Cullen and Whytt proposed somewhat different schemas of mental disorders, although each adhered to a physiological grounding for these disturbances. Whytt attended to the less severe mental conditions of his time, categorizing them into three broad syndromes: hysteria, hypochondriasis, and nervous exhaustion—the last of which was subsequently referred to by George Beard (1839–1883) as neurasthenia. This classification does not deviate much from our current diagnostic manual, although Whytt’s ideas were not based on detailed psychological observations. Whytt’s basic theory was similar to Cullen’s: He posited that disturbed motility within the nervous system produced nervous disorders. The selection of the term “neuroses” made good sense, as both Cullen and Whytt assumed that different sensibilities of the nerves could be the foundation upon which certain problematic behaviors might be based. This belief continued for at least another century, anticipating ideas that were explored in greater depth first by Charcot, and later by Janet and Freud. John Haslam (1766–1844), a British psychiatrist, is perhaps best known for the diligence and astuteness of his clinical observations. As Zilboorg and Henry (1941) noted, “Through the sheer effort of keen observation of minute, seemingly unrelated details .€.€. and orderly arrangements of these details .€.€. a coherent clinical picture of the disease came to the fore” (p.€ 303). More careful than his predecessors, Haslam provided the first clinical description of various forms of paralysis, most notably general paresis. Alert to the epidemic of venereal disease that spread across Europe in the early 19th century, he wrote:

A course of debauchery long persisted would probably terminate in paralysis .€.€. frequently induces derangement of mind. Paralytic affections are a much more frequent cause of insanity than has been commonly supposed, and they are also a very common effect of madness; more maniacs die of hemiplegia and apoplexy than from any other disease. (1809, p.€209)

Also of great significance was Haslam’s recognition that states of excitement and depression alternated in the same individual—an observation recorded by Aretaeus 17 centuries earlier. Importantly, it also recognized the significance of the course of a disease as a factor in classifying mental syndromes, thereby laying the groundwork for Kraepelin’s central rationale for his nosological model almost a century later. In his 1809 book, Observations on Madness and Melancholy, Haslam described a number of cases that would subsequently be classified as dementia praecox or schizophrenia. In the following year, he published an innovative text, Illustrations of Madness, which presented a detailed examination of an individual with diverse paranoid features. No less significant was Haslam’s sophistication in matters of nomenclature and semantics. In his 1809 text he wrote: Madness is therefore not a complex idea, as has been supposed, but a complex term for all the forms and varieties of this disease. Our language has been enriched with other terms expressive of this affliction. .€.€. Instead of endeavoring to discover an infallible definition of madness, which I believe will be found impossible, [I will] attempt to comprise, in a few words, the wide range and mutable character of a Proteus disorder. (1809, pp.€5–6)

Note should be made in this chronological sequence of the important contributions of Jean Esquirol (1772–1840), the great humanistic reformer and associate of Philippe Pinel. Among Esquirol’s diagnostic proposals was the attention he gave to a patient’s dispositions and deficits of affect and impulse in his concept of lypemanie, by which he meant a deficiency in the capacity to feel or desire—a feature seen in patients whom many would speak of today as depressed. Esquirol grouped the several variants of mental disorder into five broad classification syndromes: lypemanie, monomanie, manie,

A Précis of Psychopathological History

dementia, and imbecility/idiocy—a series of distinctions utilized in France for over a century. Esquirol also made significant contributions to the clarification of delusions and hallucinations. He wrote: In hallucinations there is no more sensation or perception than in dreaming or somnambulism, when no external object is stimulating the senses. .€.€. In fact, [a] hallucination is a cerebral or psychological phenomenon that takes place independently from the senses. The pretended sensations of the hallucinated are images and ideas reproduced by memory, improved by the imagination, and personified by habit. (1838, pp.€191–192)

His description of a hallucination as essentially a variant of a delusion differentiated it from simple sensory errors such as illusions and brought it into the realm of the patient’s personality dysfunctions. Also notable were the contributions of Jean-Â�Pierre Falret (1794–1870), another humane reformer and student of Esquirol, who articulated notions similar to his mentor’s regarding delusions. He specified several factors instrumental in their formation—Â� notably, the state of the brain, the character of the patient, the circ*mstances surrounding the time the delusion began, and concurrent internal and external sensations. He expressed his conception of delusions as follows: Delusions may reflect the most intimate preoccupations and emotions of the individual. Indeed, the features of delusions may help us recognize what aspects of the subject’s organization are suffering the most. Practitioners should give attention to relationships between delusions and the character of the subject. (1862, p.€357)

Falret also contributed an early and insightful series of papers that further detailed the variable character of mania and melancholy, which he called forme circulaire de maladie mentale, consisting of periods of excitation followed by longer periods of weakness. Presenting this theme as a facet of his 1851 lectures at the Salpêtrière Hospital, he subsequently elaborated these views in a book published in 1854; similar ideas were proposed almost concurrently by Jules Baillarger.


A series of novel classifications also gained prominence in Germany. They were based on a threefold distinction among the “faculties of the mind” (volition, intellection, and emotion), as well as a number of “morbid” processes (e.g., exaltation and depression). Among the early promoters of this schema was Johann Christian Heinroth (1773–1843)—perhaps the first physician to occupy a chair in psychiatry, that at Leipzig University in 1811. He subdivided one of the major categories of mental disorder, vesania, into several orders, genera, and species. Designing a complex matrix combining the major faculties on one dimension with the morbid processes on the other, he proposed a classification system comprising subtypes that became the basis of several variations throughout Germany and England in the ensuing century. Heinroth also developed a theory of mind with a tripartite structure. The basic or undergirding layer was characterized by the animalistic instinctual qualities of human beings; the intermediary layer reflected consciousness, including both intelligence and self-Â�awareness; and, finally, a superior layer consisted of what we would call conscience. Presaging ideas proposed later by Freud, Heinroth also proposed the notion of conflict when two layers became opposing forces—such as the instinctual impulses of sin on the one hand, and the conscience’s sense of moral correctness on the other. Especially insightful was Heinroth’s (1818) recognition of the significance of the patient’s affect, or passions. He specified these insights in the following passage: The origin of the false notions in patients are erroneously attributed to the intellect. The intellect is not at fault; it is the disposition which is seized by some depressing passion, and then has to follow it, and since this passion becomes the dominating element, the intellect is forced by the disposition to retain certain ideas and concepts. But it is not these ideas or concepts which determine the nature of the disease.

Heinroth recognized a deep connection between the human qualities of mind and the more fundamental vegetative or animal passions that are fundamental to mental disorders, notably those of melancholy and rage. Heinroth also conceived of a term akin to what today we call “psychosomatics,” in


which he took exception to Descartes’s contention of a dualism between mind and body. In his view, health reflected harmony between these two components when they acted as a singular entity. He not only recognized a unity between mind and body, but considered that each person was composed of the same elements that made up the rest of nature. Heinroth traced the term “paranoia” some 2,000 years back in the medical literature. The word had disappeared from the medical lexicon in the 2nd century B.C. and was not revived until Heinroth, following the structure of Kantian psychology, employed the term in 1818 to represent a variety of disorders. He termed disturbances of the intellect “paranoia”; he called disturbances of feeling “paranoia ecstasia.” He also proposed the parallel concepts of Wahnsinn and Verrucktheit (the latter term is still in use as a label for paranoia in modern-day Germany). Griesinger, to be discussed shortly, picked up the term Wahnsinn in 1845 to signify pathological thought processes and applied it to cases of expansive and grandiose delusions. In 1863, Kahlbaum, also discussed later in this chapter, suggested that paranoia be the exclusive label for delusional states. British alienist James Cowles Prichard (1786–1848), credited by many as the first to formulate the concept of “moral insanity,” was in fact preceded in this realization by several theorists; nevertheless, he was the first to label it as such and to give it wide readership in English-Â�speaking nations. Although he accepted Pinel’s notion of manie sans délire, he dissented from Pinel’s morally neutral attitude toward these disorders and became the major exponent of the view that these behaviors signified a reprehensible defect in character that deserved social condemnation. He also broadened the scope of the original syndrome by including under the label “moral insanity” a wide range of previously diverse mental and emotional conditions. All of these patients ostensibly shared a common defect in the power to guide themselves in accord with “natural feelings”—that is, a spontaneous and intrinsic sense of rightness, goodness, and responsibility. In Prichard’s opinion, those afflicted by this disease were swayed, despite their ability to intellectually understand the choices before them, by overpowering “af-

fections” that compelled them to engage in socially repugnant behaviors. A major figure in extending the ideas of Esquirol and Falret at the Salpêtrière Hospital, Felix Voisin (1794–1872) was also a strong adherent of the phrenological speculations of Gall. His particular expertise was related to the linkage between the brain and the sexual organs; he stressed the importance of the nervous system as causally involved in generating various disorders of sexual desire. Placing special attention on the pathologies of nymphomania and satyriasis, especially as they were related to hysteria, Voisin articulated a progression in these disorders from their early stages to their more severe forms, contributing to the idea that disease course was central to clinical diagnostics. In his major work, The Analysis of Human Understanding (1851), Voisin specified three major faculties of human functioning: moral, intellectual, and animal. This division predated and paralleled Freud’s subsequent formulation of the mind’s structure of superego, ego, and id. Also notable was Voisin’s contribution to the moral treatment of persons with mental retardation at the Bicêtre Hospital. Influenced by Prichard, Voisin delved briefly in his later years into the problems of criminal and forensic pathology, speaking of criminals as products of lower-class origins and of their inevitable moral degeneration— a theme addressed elsewhere by Cesar Lombroso and Benedict Morel. Another contributor to French thinking of the day was Paul Briquet (1796–1881), who focused primarily on problems of hysteria and their ostensive connection to female maladies. In his extensive monograph, Traite Clinique et Therapeutique a l’Hystérie (1859), he took exception to the notion posited by Plato and Hippocrates that hysteria was a consequence of sexual incontinence. Briquet specified with great clarity the multiple, exaggerated gastrointestinal, sexual, and other complaints that typified the symptoms presented by his “hysterical” patients. Such symptoms are labeled “somatization disorder” in official nosologies today, as well as occasionally referred to as “Briquet’s syndrome.” He recorded, in contrast to prior beliefs, that married women were no more inclined to hysteria than were unmarried women; that numerous cases appeared before puberty; and, most significantly, that an

A Précis of Psychopathological History

active sexual life was no assurance that one would not develop such symptoms. Going beyond the assumptions of many of his contemporaries, Briquet rejected the view that men could not develop symptoms of hysteria. He also pointed to numerous psychological influences that often contributed to the symptomatological expression of the disorder, noting painful emotional states (such as sadness and fear) as elements in precipitating the syndrome. Moreover, he speculated on a variety of untoward developmental and life experiences as playing a pathogenic role (e.g., parental mistreatment, spousal abuse, unfavorable employment circ*mstances or business failures). Recognizing that only a small subset of those subjected to these psychosocial experiences developed the hysterical syndrome, Briquet proposed the concept of “predispositions” as pathogenic factors. Aware that life circ*mstances often troubled his patients, he suggested that many would benefit from speaking to an empathic counselor or physician who might serve as a confidant. Briquet showed great sensitivity in going beyond the crude medications of the day to employ a psychotherapeutic approach to his patients’ difficulties. Ernst von Feuchtersleben (1806–1849) may have been the first Austrian psychiatrist to gain a distinguished status in European circles during the mid-19th century. His one major publication, The Principles of Medical Psychology, published in 1847, probably had a significant influence on Freud and his many disciples in Vienna. A strong critic of those who supported the Cartesian mind– body dichotomy, Feuchtersleben (like many in the 20th-Â�century psychosomatic movement) considered the mind and the body to be a unitary phenomenon, essentially indivisible. An exponent of the role of personality qualities in the life of mental patients, Feuchtersleben wrote with great sensitivity on the psychic sources of mental disorders. In describing those inclined to the development of depressive diseases, Feuchtersleben said: Here the senses, memory, and reaction give way, the nervous vitality languishes at its root, and the vitality of the blood, deprived of this stimulant, is languid in all its functions. Hence the slow and often difficult respiration,

29 and proneness to sighing. .€.€. When they are chronic, they deeply affect vegetative life, and the body wastes away. (1847, p.€135)

Moreover, in what may have been the first purely psychological description of what is now referred to as histrionic personality disorder, Feuchtersleben depicted women disposed to hysterical symptoms as being sexually heightened, selfish, and “over-Â�privileged with satiety and boredom.” Attributing these traits to the unfortunate nature of female education, he wrote: “It combines everything that can heighten sensibility, weaken spontaneity, give a preponderance to the sexual sphere, and sanction the feelings and impulse that relate to it” (1847, p.€ 111). Chauvinistic as this judgment may be regarded today, Feuchtersleben at least recognized and was sensitive to the limitations Victorian society placed upon women in his time. Moreover, he asserted the important role that psychological factors could play in helping patients understand the origins of their difficulties. He also espoused a hopeful therapeutic attitude and recommended opportunities for patients to acquire a second education in life. As noted earlier in this chapter, the great English neurologist Thomas Willis (1664/1978) reported having observed a pathological sequence in which “young persons who, lively and spirited, and at times even brilliant in their childhood, passed into obtuseness and hebetude during adolescence.” Better known historically, however, are the texts of Belgian psychiatrist Benedict-Â�Augustin Morel (1809–1873), who described the case of a 14-year-old boy who had been a cheerful and good student, but who progressively lost his intellectual capacities and increasingly became melancholy and withdrawn. Morel considered such cases to be irremediable and ascribed the deterioration to an arrest in brain development that stemmed from hereditary causes. He named the illness “dementia praecox” (demence precoce), to signify his observation that a degenerative process began at an early age and progressed rapidly. After Morel became chief physician at St. Yon Asylum in 1856, he continued to lecture and write on the inevitable sequence of deterioration, which he considered to be an inexorable course in all mental disorders.


He judged this “incessant progression” of degeneration to be human destiny. Speaking of those subjected to hereditary mental disorders, he wrote: The degenerate human being, if he is abandoned to himself, falls into a progressive degradation. He becomes .€.€. not only incapable of forming part of the chain of transmission of progress in human society, he is the greatest obstacle to this progress through his contact with the healthy portion of the population. (1857, p.€46)

Although his work secured him a niche in the history of psychiatry, Morel’s views contributed to the pessimistic attitude regarding mental illness that was then pervasive in the European public at large—a view that unfortunately gained a horrendous following a century later in Nazi Germany. In 1854, Jules Baillarger (1809–1892) and Jean-Â�Pierre Falret summarized the results of their independent work with depressed and suicidal persons. They reported that a large proportion of these patients showed a course of extended depression, broken intermittently by periods of irritability, anger, elation, and normality. The terms la folie circulaire (Falret, 1854) and folie à double forme (Baillarger, 1853) were applied to signify this syndrome’s contrasting and variable character. Baillarger contributed to a wide range of psychopathological conditions beyond the syndrome known today as bipolar disorder, notably in his ideas on hallucinations and delusions, neurohistology, epilepsy, and general paralyses. With regard to delusions, he sought to describe the perceptual basis of this disorder by stating that delusions were based on false interpretations of normal sensations, whereas illusions were distortions at the sensory rather than the ideational level. Similarly, he explored the question of whether hallucinations were sensory or psychological phenomena. He proposed two types: psychosensory hallucinations, which stemmed from the interaction of both sensory and imaginal distortions, and psychological hallucinations, which were independent of any sensory involvement. Although born and educated in Germany, Richard von Krafft-Ebing (1810–1874) became a close follower of Morel, whose concept of degeneration struck a resonant chord

in his work and practice at the Illenau Asylum in Baden. Krafft-Ebing was convinced that the Morelian process of degeneration was the primary cause not only of mental disorders, but also of criminality and sexual pathology. He wrote: “Madness, when it finally breaks out, represents only the last link in the psychopathic chain of constitutional heredity, or degenerate heredity” (1879, p.€439). Moving to Graz, Austria, he became a professor of psychiatry at the university there and the director of its provincial asylum. In his major work, Lehrbuch der Psychiatrie (1879), he referred to the problem of progressive sexual degeneration as follows: “It is specially frequent for sexual functioning to be .€ .€. abnormally strong, manifesting itself explosively and seeking satisfaction impulsively, or abnormally early, stirring already in early childhood and leading to masturbation” (p.€ 424). By the mid-1880s, Krafft-Ebing assumed the chair at the University of Vienna and wrote his most famous book, entitled Psychopathia Sexualis (1882/1937), in which he spoke of the pervasive pathology of all variants of sexual activity (i.e., those differing from the approved and “proper” behavior of Victorian times). The label “masochism” was proposed by Krafft-Ebing as a new concept in his catalog of sexual perversions. In a manner similar to the creation of “sadism” from the name of the Marquis de Sade, the “masochism” label was created from the name of a well-known writer of the time, Leopold von SacherÂ�Masoch. In Sacher-Â�Masoch’s novel Venus in Furs (1870), the hero suffers torture, subjugation, and verbal abuse from a female tormentor. Krafft-Ebing asserted that flagellation and physical punishment were necessary elements in the perversion, but were less significant than a personal relationship that included enslavement, passivity, and psychological serfdom. Hence, from its first formulations, the concept of masochism (although centrally sexual in nature) included the need to experience suffering, not just physical pain. The growth of knowledge in anatomy and physiology in the mid-18th century strengthened the trend toward organically oriented disease classifications. Wilhelm Griesinger (1817–1868; see Figure 1.3), a young German psychiatrist with little direct patient experience, asserted the disease concept in his

A Précis of Psychopathological History

Figure 1.3.╇ Wilhelm Griesinger.

classic text Mental Pathology and Therapeutics, published in 1845 when he was barely 28 years of age. His statement “Mental diseases are brain diseases” shaped the course of German systematic psychiatry for the next 40 years. Griesinger’s contention that classifications should be formed on the basis of underlying brain lesions was not weakened by the fact that no relationship had yet been established between brain pathology and mental disorders. In fact, Griesinger’s own system of categories—Â�depression, exaltation, and weakness—did not parallel his views regarding the importance of brain pathology. Nevertheless, he convinced succeeding generations of German neurologists, led by Thomas Meynart and Carl Wernicke, that brain diseases would be found to underlie all mental disturbances. Griesinger was born in Stuttgart, Germany, and completed his medical studies in Zurich and Tübingen. There he learned to view medicine as a science based on the direct observation of patient experiences and behaviors rather than on historical speculations and philosophy. He began his formal career in psychiatry at the Winnenthal Asylum in Stuttgart. Assuming that he had gathered sufficient expertise in a 3-year span, he penned his classic 1845 text. To him, the study of mental illness was integral to the study of general medicine. He conceived of mental disorders as chronically progressive,


like most medical diseases. Thus he regarded depression as beginning with a minor level of cerebral irritation, leading next to a chronic and irreversible degeneration, and ending ultimately in pervasive dementia—a path of deterioration that became a central theme of Kraepelin’s belief that the course of a mental disorder was its most crucial characteristic. It was not until 1861 that Griesinger revised his 1845 text, following which he returned to his work in psychiatry at the University of Berlin. Here he both lectured and practiced at its Charité Clinic, where he divided his patients into those with routine nervous diseases and those with nervous diseases that also exhibited psychiatric symptoms. He also initiated and assumed the editorship of a new journal, the Archives for Psychiatry and Nervous Diseases. In its first volume, Griesinger wrote: Psychiatry has undergone a transformation in its relationship to the rest of medicine. .€.€. This transformation rests principally on the realization that patients with so-Â�called mental diseases are really individuals with diseases of the nerves and the brain. .€ .€ . Psychiatry .€ .€ . must become an integral part of general medicine and accessible to all medical circles. (1868, p.€12)

Although the work of Griesinger and his followers regarding the role of the brain in mental disorders soon dominated continental psychiatry, a different emphasis regarding the basis of classification was developing concurrently. Jean Esquirol, Pinel’s distinguished associate, had often referred to the importance of age of onset, variable chronicity, and deteriorating course in understanding pathology. This idea was included as a formal part of classification in 1856 when German psychiatrist Karl Ludwig Kahlbaum (1828–1899) extended Esquirol’s idea by developing a classification system in which disorders were grouped according to their course and outcome. It became the major alternative system to the one Griesinger proposed. Kraepelin, noting his indebtedness to Kahlbaum’s contributions, stated that “identical or remarkably similar symptoms can accompany wholly dissimilar diseases while their inner nature can be revealed only through their progress and termination” (Kraepelin, 1920, p.€116).


Kahlbaum wrote of how useless attempts had been to group disorders on the basis of the similarity of their overt symptomatology, as if such superficial symptom collections would themselves expose something essential concerning the underlying diseases. He commented as follows: It is futile to search for the anatomy of melancholy or mania, because each of these forms occurs under the most varied relationships and combinations with other states, and they are just as little the expression of an inner pathological process as the complex of symptoms we call fever. (1874, p.€2)

Kahlbaum turned his attention in 1857 to psychoses that were typical of young adolescents, focusing on the sudden emergence of mental disorientation and rapid disintegration—a pattern not unlike that described by Morel a decade or two earlier. Similarly, reading the work of Falret and of Jules Baillarger, he also directed his attention to the problems of patients whose mood disorders appeared to follow a sequential course from mania to depression and back. In a series of monographs and books published between 1863 and 1874, Kahlbaum not only established the importance of including longitudinal factors in psychiatric diagnosis, but described newly observed disorders that he labeled “hebephrenia” and “catatonia,” as well as coining the modern terms “symptom complex” and “cyclothymia.” Kahlbaum, together with his disciple Ewald Hecker, introduced the term “hebephrenia” to represent conditions that began in adolescence, usually starting with a quick succession of erratic moods, followed by a rapid enfeeblement of all functions, and finally progressing to an unalterable psychic decline. The label “catatonia” was introduced to represent “tension insanity” in cases where the patient displayed no reactivity to sensory impressions, lacked “selfwill,” and sat mute and physically immobile. These symptoms ostensibly reflected deterioration in brain structure. It was Kahlbaum also who, in 1882, clearly imprinted current thinking on the fixed covariation of mania and melancholia, known today as bipolar disorder. Although he regarded them as facets of a single disease, which he termed “dysthymia” (following a

label introduced two decades earlier by Carl Flemming), the disease actually manifested itself in different ways at different times—Â� occasionally euphoric, occasionally melancholic, and occasionally excitable or angry. It was the primacy of the former two emotions that rigidified future conceptions of the syndrome and redirected thinking away from its more typical affective instability and unpredictability. He termed a milder variant of the illness, notable for its frequent periods of normality, “cyclothymia.” A more severe and chronic form of the same pattern was designated by Kahlbaum as vesania typica circularis. Henry Maudsley (1835–1918) was admitted to London’s University College at age 15; here he proved to be a brilliant student, completing his medical degree at age 21. Unsure about his future and lacking the means to follow an early interest in surgery, he entered the East India Company’s service, spending the better part of a year as medical officer at the Wakefield Mental Asylum. Owing to his high intelligence and vigorous appearance, at age 23 he was appointed medical superintendent of the Cheadle Royal Hospital, despite a total lack of administrative experience or formal psychiatric training. Shortly thereafter, he became superintendent of the newly opened Manchester Royal Lunatic Hospital. His fame grew throughout England, and at age 27 he became the editor of the country’s major psychiatric publication, the Journal of Mental Science. He was appointed to a professorship at the University College Hospital in 1870. In his major text, Physiology and Pathology of Mind (1876), Maudsley attempted to redirect the philosophical inclinations typical of British clinicians and sought to anchor the subject more solidly within the biological sciences. He vigorously asserted that mind and body comprised a unified organism, “each part of which stirs the furthest components, [and] which then acts upon the rest and is then reacted on by it. .€.€. Emotions affect every part of the body and [are] rooted in the unity of organic life.” He also wrote, consistent with comparable views expressed by Griesinger in Germany, that “mental disorders are neither more nor less than nervous diseases in which mental symptoms predominate” (1876, p.€ 41). Despite this view, Maudsley had asserted earlier

A Précis of Psychopathological History that there is no boundary line between sanity and insanity; and the slightly exaggerated feeling which renders a man “peculiar” in the world differs only in degree from that which places hundreds in an asylum. .€.€. Where hereditary predisposition exists, a cause so slight as to be inappreciable to observers is often efficient to produce the disease. (1860, p.€14)

The Japanese first met with Europeans in the middle of the 16th century, but they were highly ambivalent toward European influences and remained isolated until the latter half of the 19th century. Western medicine (including Western psychiatry in Britain and Germany) was introduced, and Japanese psychiatry became strongly organically oriented. The writings of Maudsley, who viewed insanity as a bodily disease, and of Griesinger, whose approach has been described as “psychiatry without psychology,” were among the most important influences. Shuzo Kure’s (1865–1932) visit to Europe brought back not only the ideas behind Kraepelin’s descriptive psychiatry, but also the emerging interest in psychoneuroses and psychotherapy. Also introduced was the term “neurasthenia,” in which a wide variety of bodily symptoms were explained as exhaustion of the central nervous system under the influence of physical and social stressors. Kure’s pupil Shoma Morita (1874–1938), having personal experiences with neurasthenia, developed a psychogenic theory and treatment of neurosis. Subsequently labeled “Morita therapy,” his approach deserves a closer description as an example of how Western and Eastern thinking met. (See Goddard, 1991, for a full discussion.) Morita translated Binswanger’s work Fundamentals of Treatment for Mental Illness; Binswanger recommended a strictly regulated 5-week timetable for “life normalization,” which included intellectual and manual activities. Morita was also influenced by American neurologist Silas Weir Mitchell (1829–1914), who invented a regimen of bed rest, isolation, rich diet, massage, and electrostimulation for neurasthenia, and by neuropathologist and psychotherapist Paul Charles Dubois (1848–1918), who believed that the therapist’s task was to convince the patient that his or her neurotic feelings, thoughts, and behaviors were irrational. By integrating these ideas of Western scientists


with thoughts from Zen Buddhism, Morita developed Morita therapy, which for a century has been widely used not only in Japan, but also in China and in some Western societies. Morita’s term for neurosis was shinkeish*tsu. Everyone is born with sei no yokubo, the desire to live, but this drive may be hindered by oversensitivity to oneself and one’s limitations. Morita did not regard patients as sick persons, but as healthy persons obsessed by their own anxieties and fears. He described three different kinds of shinkeish*tsu: (1) the ordinary type, which resembles what is now labeled somatization disorder; (2) the obsessive/phobic type or taijin kyofusho, which includes symptoms of present-day agoraphobia, specific phobia, social phobia, and obsessive–Â�compulsive disorder; and (3) the paroxymal neurosis type, which includes symptoms of agoraphobia and generalized anxiety disorder. Morita therapy followed a strict time schedule, starting with complete isolation in a private, familial, homelike room to give a feeling of security for a week; this was followed by light activities and the keeping of a diary, which each patient discussed with a therapist. Then followed a stage of work such as gardening, and finally a stage where the patients were turning toward realities with their families and society. The goal for the treatment was not necessarily the disappearance of symptoms, but the ability to function normally and productively despite the symptoms. Morita endorsed the concept of arugamama, meaning “things are as they are,” which was the mental attitude patients were encouraged to show toward their symptoms. The emphases on body–mind–Â�nature monoism, affirming and accepting worldly passion and desires, and the practice of daily life were clearly borrowed from Japanese Shintoism, Zen Buddhism, and Asian psychology, but they also owed much to Japanese cultural patterns (e.g., the meaning of and devotion to work, the acceptance of reality, persistence, and dependency). Throughout the 19th century, German psychiatrists abandoned what they considered to be the value-laden theories of the French and English alienists of the time and toward what they judged to be empirical or observational research. Among this group was J. A. Koch (1841–1908), who


proposed that the label “moral insanity” be replaced by the term “psychopathic inferiority,” which included “all mental irregularities whether congenital or acquired which influence a man in his personal life and cause him, even in the most favorable cases, to seem not fully in possession of normal mental capacity” (1891, p.€ 67). Koch used the word “psychopathic”—a generic label employed to characterize all personality diagnoses until recent decades—to signify his belief that a physical basis existed for these character impairments. Thus he stated: “They always remain psychopathic in that they are caused by organic states and changes which are beyond the limits of physiological normality. They stem from a congenital or acquired inferiority of brain constitution” (1891, p.€54).

Descriptive Psychopathology in€the 20th Century Kraepelin’s comprehensive textbooks at the turn of the 20th century served as one of psychiatry’s two major sources of inspiration; the other consisted of Freud’s innovative psychoanalytic contributions. As the preeminent German systematist, Emil Kraepelin (1856–1926; see Figure 1.4) bridged the diverse views and observations of Greisinger and Kahlbaum in his outstanding

Figure 1.4.╇ Emil Kraepelin.

texts, revised from a small compendium in 1883 to an imposing four-Â�volume eighth edition in 1913. Kraepelin constructed a system that integrated Kahlbaum’s descriptive and longitudinal approach with Greisinger’s somatic disease view. By sifting and sorting prodigious numbers of well-Â�documented hospital records, and directly observing the varied characteristics of patients, he sought to bring order to symptom pictures and, most importantly, to patterns of onset, course, and outcome. Kraepelin felt that syndromes based on these sequences would be best in leading to accurate identification and distinction among the different conditions that differentiated and caused these disorders. Psychiatric historian Ray Porter (2002) has summarized this contribution of Kraepelin as follows: He approached his patients as symptomÂ�carriers, and his case histories concentrated on the core signs of each disorder. The course of psychiatric illness, he insisted, offered the best clue to its nature. .€.€. Kraepelin’s commitment to the natural history of mental disorders led him to track the entire life histories of his patients in a longitudinal perspective which privileged prognosis (likely outcome) as definitive of the disorder. (pp.€184–185)

Kraepelin was born in Germany in the same year as Sigmund Freud. A serious and diligent student, Kraepelin was exposed in medical school to several professors who were instrumental in shaping his style of thinking and research for the rest of his career. Most notable among these was Wilhelm Wundt, the founder of experimental psychology. Wundt himself had been trained by Hermann von Helmholtz, the great physiological theorist. Owing to visual difficulties that deterred him from research with microscopes, Kraepelin began to pursue psychological research, becoming one of Wundt’s most distinguished students. Nevertheless, Wundt advised him to pursue medicine rather than psychology, which was then a fledgling science with limited career opportunities. In 1882, Kraepelin began the initial drafts of his first textbook, which later became the standard for educating psychiatrists. His first text, a 300-page volume titled Compendium of Psychiatry, was so successful that it led to several subsequent editions published under the general title Short Text-

A Précis of Psychopathological History

book of Psychiatry. By the sixth edition of what he subsequently called his Lehrbuch or Textbook of Psychiatry at the turn of the century, Kraepelin was known throughout the Continent and the English-Â�speaking world. In 1904, he became chairman of the Psychiatric Clinic and Laboratory at the University of Munich—a distinguished department where he was able to bring along with him from Heidelberg such promising young researchers as Alois Alzheimer and Franz Nissl, both already known for their excellent neurohistological studies. At the time of his death in 1926 at age 70, Kraepelin was actively working on a ninth edition of his textbook, which had expanded to four volumes and more than 3,000 pages. Kraepelin did not set out initially to create the nosology for which he became so famous. Although he proposed a series of revolutionary ideas concerning the nature of clinical syndromes, the astuteness of his observations and the clarity of his writing were what proved to be central to the success of his work. Kraepelin wrote very little about how classification should be organized; that is, he utilized no formal set of principles to rationalize how a nosology should be structured. It was the implicit structure of his books (i.e., their basic table of contents) that served as his classification system. Not to be dismissed was the logic that he presented for organizing syndromes on the basis of clinical symptomatology, course, and outcome. Perhaps it was the input of his mentor Wundt’s keen observation and analysis of the behavior of his subjects in his research studies that taught him to provide such richly descriptive characterizations of his patients. Moreover, Kraepelin focused on the overt psychological manifestations of mental disorders, in contrast to his more organically and physiologically oriented contemporaries. The following paragraphs touch on only a few of his conceptions regarding the major forms of psychoses and the syndromes now termed “personality disorders.” Kraepelin constantly revised his diagnostic system, elaborating it at times, simplifying it at others. In the sixth edition of 1899, he established the definitive pattern of two modern major disorders: “manic–Â�depressive psychosis” (now known as bipolar disorder) and “dementia praecox” (now known as schizophrenic disorders). These were clini-


cally vivid syntheses of previously independent concepts that Morel and Kahlbaum had formulated. Within the manic–Â�depressive group, he brought together the excited conditions of mania and the hopeless melancholia of depression, indicating the periodic course through which these moods alternated in the same patient. To be consistent with his disease orientation, he proposed that this disorder was caused by an irregular metabolic function transmitted by heredity. As recorded previously, many of Kraepelin’s predecessors viewed mania and melancholia as a single disease that manifested itself in different forms and combinations over time. Kraepelin borrowed heavily from these formulations, but separated the “personality” and “temperament” variants of the disorder from the clinical state of the disease. Nevertheless, in the fifth edition of his text, he proposed the name “maniacal–Â�depressive insanity” for “the whole domain of periodic and circular insanity”; it included such diverse disturbances as “the morbid states termed melancholia and certain slight colorings of mood, some of them periodic, some of them continuously morbid” (1896, p.€161). Like Kahlbaum, Kraepelin viewed “circular insanity” as a unitary illness. Moreover, he believed that every disorder that featured mood disturbances—Â�however regular or irregular and whatever the predominant affect, be it irritability, depression, or mania—was a variant or “rudiment” of the same basic impairment. To Kraepelin, the common denominator for these disturbances was an endogenous metabolic dysfunction that was “to an astonishing degree independent of external influences” (1896, p.€173). Four varieties of the cyclothymic disposition identified by Kraepelin were termed “hypomanic,” “depressive,” “irascible,” and “emotionally unstable.” He described the hypomanic type as follows: They acquire, as a rule, but scant education, with gaps and unevenness, as they show no perseverance in their studies, are disinclined to make an effort, and seek all sorts of ways to escape from the constraints of a systematic mental culture. The emotional tone of these patients is persistently elated, carefree, selfÂ�confident. Toward others they are overbearing, arbitrary, impatient, insolent, defiant. They mix into everything, overstep their prerogatives, make unauthorized arrangements,

36 HISTORICAL AND CULTURAL PERSPECTIVES as they prove themselves everywhere useless. (1913, p.€221)

In describing the depressive personality type, Kraepelin (1921) wrote: There are certain temperaments which may be regarded as rudiments of manic–Â�depressive insanity. They may throughout the whole of life exist as peculiar forms of psychic personality, without further development; but they may also become the point of departure for a morbid process which develops under peculiar conditions and runs its course in isolated attacks. Not at all infrequently, moreover, the permanent divergencies are already in themselves so considerable that they also extend into the domain of the morbid without the appearance of more severe, delimited attacks. (p.€118)

Typically, Kraepelin considered this type to be characterized by an inborn temperamental predisposition to “a permanent gloomy emotional stress in all experiences in life” (p.€118). According to him, “the morbid picture is usually perceptible already in youth, and may persist without essential change throughout life” (p.€123). The irascible type was ostensibly endowed simultaneously with both hypomanic and depressive inclinations. According to Kraepelin, “They are easily offended, hot-Â�headed, and on trivial occasions become enraged and give way to boundless outbursts of energy. Ordinarily the patients are, perhaps, serene, self-Â�assertive, ill-Â�controlled; periods, however, intervene in which they are cross and sullen” (1921, p.€222). The emotionally unstable variant presumably also possessed both hypomanic and depressive dispositions, but manifested them in an alternating (or, as Kraepelin viewed it, true cyclothymic) pattern. He described these patients as follows: It is seen in those persons who constantly swing back and forth between the two opposite poles of emotion, now shouting with joy to heaven, now grieved to death. Today lively, sparkling, radiant, full of the joy of life, enterprise, they meet us after a while depressed, listless, dejected, only to show again several months later the former liveliness and elasticity. (1921, p.€222)

Kraepelin had considered hebephrenia, the diagnosis of adolescent psychosis, and

dementia praecox to be synonymous prior to the sixth edition of his psychiatric text. In his original treatise, he concluded that the diverse symptom complexes of catatonia and hebephrenia, as well as certain paranoid disturbances, displayed a common theme of early deterioration and ultimate incurability. As he conceived them, each of these illnesses was a variation on Morel’s concept of dementia praecox. By subsuming the disparate symptoms of these formerly separate syndromes under the common theme of their ostensible early and inexorable mental decline, Kraepelin brought a measure of order and simplicity to what had previously been diagnostic confusion. In line with the traditions of German psychiatry, Kraepelin assumed that a biophysical defect lay at the heart of this new coordinated syndrome. In contrast to his forebears, however, he speculated that sexual and metabolic dysfunctions were the probable causal agents, rather than the usual hypothesis of an anatomical lesion. Among the major signs that Kraepelin considered central to these illnesses, in addition to the progressive and inevitable decline, were discrepancies between thought and emotion; negativism and stereotyped behaviors; wandering or unconnected ideas; hallucinations and delusions; and a general mental deterioration. Kraepelin believed that the “autistic” temperament served as the constitutional soil for the development of dementia praecox. Of particular note was Kraepelin’s observation that children of this temperament frequently “exhibited a quiet, shy, retiring disposition, made no friendships, and lived only for themselves” (1921, p.€109). They were disinclined to be open and become involved with others, were seclusive, and had difficulty adapting to new situations. They showed little interest in what went on about them, often refrained from participating in games and other pleasures, seemed resistant to influence (but in a passive rather than active way), and were inclined to withdraw increasingly into a world of their own fantasies. Among the “morbid” personalities, Kraepelin included a wide range of types disposed to criminal activities. As early as 1905, he identified four kinds of persons with features akin to what we speak of today as Cluster B personality disorders. First were the “morbid liars and swindlers,” who were glib and

A Précis of Psychopathological History

charming but lacked an inner morality and sense of responsibility to others. They made frequent use of aliases, were inclined to be fraudulent con artists, and often accumulated heavy debts that were invariably unpaid; this type proves to be descriptively similar to those we might classify today as having narcissistic personality disorder. The second group included “criminals by impulse”— individuals who engaged in crimes such as arson, rape, and kleptomania, and were driven by an inability to control their urges; they rarely sought material gains for their criminal actions. The third type, referred to as “professional criminals,” was neither impulsive nor undisciplined; in fact, such persons often appeared well mannered and socially appropriate, but were inwardly calculating, manipulative, and self-Â�serving. The fourth type consisted of the “morbid vagabonds,” who were strongly disposed to wander through life, never taking firm root, lacking both self-Â�confidence and the ability to undertake adult responsibilities. Although less successful in influencing nosological thinking in the latter half of the 19th and early 20th centuries than Kraepelin, several other distinguished thinkers deserve recognition. Philippe Chaslin (1857–1923) was a great French theorist whose life’s work overlapped with Bleuler’s in Switzerland, Kraepelin’s in Germany, and Freud’s in Austria. A philosopher and linguist at heart, he spent the majority of his professional career at the Salpêtrière Hospital in Paris, where he wrote on a wide range of topics (including history, linguistics, and mathematics, as well as psychiatry). Among his central formulations was the concept of “discordance,” a notion he used to describe and explain dementia praecox; Bleuler, who originated the term “schizophrenia” in his 1911 treatise on the subject, stated later that he might have preferred “discordant insanity” as an alternative label had he known of it earlier. In his major work, Elements de Seminologie et de Clinique Mentale, written in 1912, Chaslin conveyed a series of ideas similar to those formulated concurrently by Freud, but with special reference to psychotic delusions. For example, he wrote: Delusional ideas seem to have their source in the emotions of the patient of which they are symbolic representations. .€ .€ . One could

37 illustrate the origins of delusions by recollecting the mechanisms of dreaming. Propensities, desires, and feelings from the waking state reappear in dreams in symbolic scenes. (1912, p.€178)

Chaslin devoted much of his theoretical writing to articulating different variants of delusions and states of confusion. He spoke of the several ways in which delusions presented themselves—Â�sometimes in isolation, sometimes combined with hallucinations; occasionally incoherent, but also at times systematic and logical, as in paranoid conditions. Regarding confusional states, Chaslin asserted that these temporary periods signified a loosening of intellectual, affective, and motivational functions; he concluded, for example, that the distinctions between confusion and dementia were modest and reflected an assumption that dementia possessed a chronic and deteriorating course. Chaslin was also concerned, as were many philosophers of the day, with the failure of psychiatric language to adequately represent the nature of the disorders they diagnosed and treated. In describing the difficulties of psychopathological terminology, Chaslin exclaimed: I believe that the imprecision of terms is due to the imprecision of our ideas, but I also think that the inexactitude of a language may cause further inexactitude in our ideas. .€ .€ . If [the terminology] only helped to combat factual imprecisions, but the opposite is the case; it is often imagined that progress has been made simply because fancy names have been given to old things. (1912, p.€18)

Eugen Bleuler (1857–1939) is universally recognized for his description of what is presently known as “schizophrenia,” the term he coined to replace the historic diagnostic label “dementia praecox.” The label “schizophrenia” is now judged by many to be unfortunate, suggesting a splitting between segments of the mind—a concept then prevalent in French circles, and a notion Janet had proposed as an alternative to Freud’s conception of three levels of consciousness. As evidence now indicates, patients diagnosed with schizophrenia do not suffer any form of splitting, but rather are characterized by disordered thinking leading to delusions and hallucinations.


In 1898 Bleuler took over the headship of the Burgholzli Mental Hospital, an already distinguished center for the clinical study of mental illness. Bleuler daily spent hours talking with his patients, often in their own unusual dialects, searching to gain an understanding of the psychological meaning of their seemingly senseless verbalizations and delusions. Most importantly, he urged his students and residents to be open-Â�minded and to establish an emotional rapport with their patients; he believed that doing so would enable them to track the meaning of the words their patients used, as well as the word associations that might give meaning to their utterings. It was in this regard that he saw the utility of Freud’s new freeÂ�association methods, and it was on these grounds also that he instilled an interest in his young associate, Carl G. Jung, in Freud’s early psychoanalytic concepts. Bleuler’s studies of word associations led to his theory of schizophrenia. That is, the “loosening” or disintegration in patients’ capacity to associate ideas and emotions reflected their ostensible inability to connect their thoughts with their feelings, and hence the presumed “split” between these two core psychic processes. Following upon ideas that were then emerging in the writings of both Freud and Janet, Bleuler asserted that his patients would display secondary symptoms that derived from the primary or fundamental thought–Â�feeling disconnection—Â� symptoms that evidenced themselves in an autistic separation from reality, in repetitive psychic ambivalences, and in verbal behaviors akin to dreaming. Although committed to Kraepelin’s view that dementia praecox was primarily an organic disease, Bleuler emphasized the presence of psychological ambivalence and disharmony in this impairment, to signify the intellectual–Â�emotional split he believed he observed in these patients. Bleuler’s conception of schizophrenia also encompassed a wider range of syndromes than Kraepelin’s notion of dementia praecox. He included several acute disturbances that Kraepelin previously judged to be independent disease entities. Moreover, Bleuler believed that those displaying acute schizophrenic symptoms could recover readily with proper intensive care before their condition devolved into a more chronic state.

Observing hundreds of patients diagnosed with dementia praecox in the early 1900s led Bleuler to conclude that it was misleading to compare the type of deterioration they evidenced with that found among patients suffering from metabolic deficiencies or brain degeneration. Moreover, he judged his patients’ reactions and thoughts to be qualitatively complex and often highly creative, contrasting markedly with the simple or meandering thinking that Kraepelin observed. Furthermore, not only did many of his patients display their illness for the first time in adulthood rather than in adolescence, but a significant proportion evidenced no progressive deterioration, which Kraepelin considered the sine qua non of the syndrome. Thus Bleuler viewed the label “dementia praecox” as misleading, in that it characterized an age of onset and a course of development not supported by the evidence. As noted, schizophrenia’s primary symptoms, in Bleuler’s view, were disturbances in the associative link among thoughts, a breach between affect and intellect, ambivalence toward the same objects, and an autistic detachment from reality. The several varieties of patients that displayed these fragmented thoughts, feelings, and actions led Bleuler to term their disorders “the group of schizophrenias.” Nevertheless, he retained the Kraepelinian view that the basic impairment in these diverse disorders stemmed from a unitary disease process that was attributable to a basic physiological pathology. As he saw it, this shared neurological ailment produced their common primary symptoms. Bleuler ascribed the content of secondary symptoms to the patients’ distinctive life experiences and to their efforts to adapt to their basic disease. Psychogenic factors shaped the unique character of each patient’s impairment, but Bleuler was convinced that experience did not itself cause the ailment. Bleuler recognized that some dispositions left untreated might ultimately evolve into a clinical schizophrenic state, which he termed schizoidie. In his initial formulation of the schizophrenia concept in 1911, he also provided one of the first portrayals that approximates what we now call avoidant personality disorder. Discussing several of the contrasting routes that often led to the psychotic syndrome, Bleuler recorded the early phase of certain patients as follows:

A Précis of Psychopathological History There are also cases where the shutting off from the outside world is caused by contrary reasons. Particularly in the beginning of their illness, these patients quite consciously shun any contact with reality because their affects are so powerful that they must avoid everything which might arouse their emotions. The apathy toward the outer world is then a secondary one springing from a hypertrophied sensitivity. (1911/1950, p.€65)

Bleuler spoke of other personalities as being “irritable of mood“(reizbare Verstimmung), as Aschaffenburg (1922) did later in describing them as “dissatisfied personalities” who went through life as if they were perpetually wounded. Applying the label “amphithymia,” Hellpach (1920) also depicted a similar pattern of “fussy people” who tended to be of a sour disposition, constantly fretted over whatever they did, and made invidious and painful comparisons between themselves and those of a more cheerful inclination (whose simpler and brighter outlook was both envied and decried). Adolf Meyer (1866–1950), like Bleuler, was born in Switzerland. He completed his medical training in 1892 at the age of 26, following several predoctoral years in France, England, and Germany. A student of Forel at the University of Zurich, he decided to emigrate to the United States shortly after receiving his medical degree, having heard that Chicago was a city with numerous opportunities for young physicians. Meyer eventually served as a staff pathologist at the Illinois Eastern Hospital for the Insane, remaining there from 1893 to 1895. For the next 7 years, he was director of clinical research laboratories at the Worcester Insane Hospital and was associated with Clark University, both in Massachusetts. Increasingly recognized as a major contributor to neuropathology, as well as a lecturer known for his detailed history taking, interviews, and note taking, Meyer was appointed director of the New York Pathological Institute in 1902, as well as professor of psychiatry at Cornell University Medical School, where he continued autopsied brain research, teaching, and administrative activities until 1910. Along with Freud and Jung, he was awarded an honorary doctoral degree at Clark University in 1909. Owing to his distinguished achievements, Meyer later became chair of


a new Department of Psychiatry and director of the Henry Phipps Psychiatric Clinic at Johns Hopkins University in Baltimore, where the aim was to blend scientific research and clinical practice. He remained at Johns Hopkins for over 30 years, building a German-style psychiatric clinic akin to Kraepelin’s in Munich; in the process, he became the most influential psychiatrist in the United States and a mentor to an entire generation of both academic and clinical psychiatrists. Meyer introduced the concept of a “constitutionally inferior” type into American literature at the turn of the century, shortly after his arrival from Germany. Although following Koch’s ideas in the main, Meyer sought to separate psychopathic from psychoneurotic disorders, both of which were grouped together in Koch’s “psychopathic inferiorities” classification. Meyer was convinced that the etiology of the neuroses was primarily psychogenic—that is, colored less by inherent physical defects or by constitutional inferiorities. Meyer later became disillusioned with both Kraepelin’s and Koch’s approaches, particularly their fatalistic views of illness and their strictly deterministic prognosis and outcome for those of a problematic temperament. Meyer turned to a view increasingly shared by psychoanalysts—that is, discarding the disease model and viewing psychiatric disorders not as fundamentally organic conditions, but rather as consequences of environmental factors and life events. Although initially sympathetic to Freud’s theories, Meyer soon became critical of the mystic and esoteric nature of psychoanalysis; despite his break from Freud’s metapsychology, however, he shared Freud’s view regarding the role of life experiences as central to the emergence of all psychiatric disorders. As early as 1906, Meyer espoused the view that a true understanding of patients could be derived only by studying the individuals’ total reaction to their organic, psychological, and social experiences. Although Meyer was the most prominent psychiatrist to introduce the Kraepelinian system in this country, he believed that these disorders were not disease entities, but “psychobiological reactions” to environmental stress. Through his work, Meyer bridged the physiological


orientation of the late 19th century and the psychodynamic orientation of the 20th. For example, in 1912 Meyer asserted that dementia praecox was not an organic disease but a maladaptive way of reacting to stress, fully understandable in terms of a patient’s constitutional potentials and life experiences. To him, these maladaptive reactions led to what he called “progressive habit deteriorations,” which reflected “inefficient and faulty attempts to avoid difficulties” (1912, p.€98). He regarded symptoms of mental illness as the end products of abortive and selfÂ�defeating efforts to establish psychic equilibrium. His well-Â�reasoned “psychobiological” approach to schizophrenia, which he called “parergasia” to signify its distorted or twisted character, was the most systematic recognition of his interactive and progressive view of the nature of pathogenesis. Of special note also was Meyer’s view that parergasia could be present in dilute and nonpsychotic form—that is, without delusions, hallucinations, or deterioration. He considered the classic psychotic symptoms to be advanced signs of a potentially, but not inevitably, evolving habit system that might stabilize at a prepsychotic level. In its nonclinical state, parergasia could be detected from a variety of attenuated “soft signs” that merely suggested the manifest psychotic disorder. Meyer’s proposal of a self-Â�defeating and maladaptive reaction system (personality) that paralleled schizophrenia in inchoate form was a highly innovative, but unheeded, notion. Karl Jaspers (1883–1969) was undoubtedly an influential pioneer of phenomenological and existential psychiatry, though, oddly enough, he did not consider himself a phenomenologist. His system of mental illness approached classification in a unique way; that is, it sought to describe each patient’s true subjective experience and how he or she faced mental illness, rather than simply describing overt psychological syndromes as observed by the therapist. To this end, Jaspers made distinctions such as that between “feelings” and “sensations”; he described the former as emotional states of the individual, and the latter as part of the individual’s reactions to and perceptions of the environment. The ultimate goal of this system was to enable the therapist to be as sensitive and empathic as possible with the patient. It was Jaspers’s contention that the

inexhaustibly infinite depth and uniqueness of any single individual—Â�whether mentally ill or healthfully functioning—could not be completely understood and objectified, but that the medical/psychological practitioner must strive for as close an understanding as possible. This existential view of humankind was what set this system apart from the traditional means of diagnosis and treatment. In contrast with the psychoanalysts, who attempted to probe beneath the surface of patients’ verbal reports to uncover their unconscious roots, Jaspers focused on patients’ conscious self-Â�description of feelings and experiences, believing that their phenomenological reports were the best routes to achieving a true understanding of their world. Together, Meyer’s notion of reaction types, Jaspers’s existential phenomenology, and Bleuler’s focus on cognitive and emotional experience reshaped Kraepelin’s original system into a more contemporary psychiatric nosology. In their classifications, Kraepelin’s clinical categories were retained as the basic framework, and Meyer’s, Jaspers’ and Bleuler’s psychological notions provided guides to patients’ inner processes and social reactions.

The Rise and Fall of 20th-Â�Century Psychoanalytic Psychopathology Many consider Jean-Â�Martin Charcot (1825– 1893; see Figure 1.5) the father of clinical neurology. Open-Â�minded, deeply curious, and capable of observing subtle clinical details of his patients’ behaviors, Charcot was an extraordinarily astute observer of physical defects and dysfunctions. Charcot was a senior physician at the Bicêtre, and later at the deteriorated Salpêtrière women’s hospital, where Pinel had carried out his humane activities earlier in the century. In 1862, along with another young physician of exceptional ability, Edme F. A. Vulpain, Charcot studied the chronically ill women housed in its decaying wards. These two highly motivated and skilled physicians quickly recognized that more than half of those for whom they were responsible had been incorrectly diagnosed, most having been lumped indiscriminately into one or two categories. Charcot’s first discoveries were related to multiple sclerosis (MS), a significant neuro-

A Précis of Psychopathological History

Figure 1.5.╇ Jean-Martin Charcot.

logical disorder that was unrecognized as a distinct disease in the 1860s. Collaborating with Vulpain, he demonstrated the classic disintegration of the myelin sheath—the basic anatomical feature of the disorder. Also important was Charcot’s recognition of the visual problems typical of those with MS, as well as his patients’ tendency to exhibit extreme fluctuations in symptomological intensity over time. Another important contribution was his distinction between MS and the “shaking palsy,” or what came to be called Parkinson’s disease. Charcot identified features of the latter that Parkinson overlooked, such as patients’ blank stares, motionless and stolid expressions, and periodic and involuntary oscillation of hand movements. Owing to Charcot’s distinguished work, the Salpêtrière was granted substantial funds to develop laboratory facilities for clinical research and for weekly lectures by “the master.” These lectures were prepared in great detail and with careful thought, although their public presentation appeared to be spontaneous. Charcot had already achieved considerable recognition in France; his work was now quickly recognized throughout the Continent, attracting disciples and students from far and wide. Of special note in his later years was an interest in “hysteria,” a label used in his day for patients with clinical signs of pathology that could not be cor-


related with underlying anatomical or neurological diseases. Because this category was generally considered a catch-all—a place to assign those who could not be properly diagnosed in one or another class of standard disorders—Â�Charcot made a valiant effort to subdivide the variants of those so categorized. He differentiated subgroups still in use, such as those with defective memories, peculiar or inexplicable losses of sensitivity, apparently (false) motoric seizures that simulated epilepsy, and so on. It was Charcot’s contention that all patients with hysteria suffered from a “weak” constitution; that is, they possessed neurological vulnerabilities that made them highly susceptible to ordinary life conditions, such as work-Â�related stresses. Among Charcot’s assertions were that these constitutionally weak patients could be readily hypnotized. In fact, Charcot believed that only patients with hysteria could be hypnotized, as they were impressionable individuals whose neurologically weak minds could be readily swayed by the suggestions of others. Worthy of note, however, was Charcot’s recognition that hysteria could be found in men as well as women, although he asserted that secondary psychological features typically differentiated the genders. Charcot’s stature and ideas concerning hysteria attracted the young Sigmund Freud, a neurologist in training from Vienna, who came to study with him during the winter of 1885. So impressed was Freud with Charcot’s lectures that he set out to translate the professor’s writings for German-Â�reading neurologists. After this, Freud progressed in his own innovative direction, disagreeing fundamentally with Charcot’s neurological assertions regarding hysteria. Three classes of experience were stressed by Freud and his psychoanalytic colleagues as conducive to psychopathology: (1) the extent to which the earliest and most basic needs of a young child’s nurturance and protection are frustrated; (2) the conflicts with which children must deal as they develop; and (3) the general parental attitudes and familial settings in which children’s experiences occur and are learned. The emphasis the psychoanalytic theorists placed on early childhood experience represented their view of disorders in adulthood as direct products of the continued


and insidious operation of past events. For them, knowledge of the past should provide information indispensable to understanding adult difficulties. To the question “What is the basis of adult disorders?”, they would answer: “The anxieties of childhood and the progressive sequence of defensive maneuvers that were devised to protect against a recurrence of these feelings.” According to psychoanalysis, therefore, adult patterns of behavior are not the results of random influences, but arise from clear-cut antecedent causes. For the most part, these causes persist out of awareness; that is, they are kept unconscious because of their troublesome character—Â�notably the stressful memories and emotions they contain, and the primitive nature of the child’s youthful defenses. Central also to the analytic viewpoint is the concept of psychic conflict. In this notion, behavior is considered to result from competing desires and their prohibitions, which are expressed overtly only through compromise and defensive maneuver, and often in disguised form. Furthermore, all forms of behavior, emotion, or cognition are likely to serve multiple needs and goals; that is, they are “overdetermined.” Behavioral expressions and conscious cognitions emerge as surface manifestations of several hidden forces that reside in the unconscious. The concept of the unconscious—inner thoughts and feelings beyond immediate awareness—was brought to the fore through the dramatic methods of an Austrian physician, Franz Anton Mesmer (1734–1815). Borrowing Paracelsus’s notion of a physically based planetary magnetism, Mesmer believed that many forms of illness resulted from imbalances of universal magnetic fluids. These imbalances, he concluded, could be restored either by manipulating magnetic devices or by drawing upon invisible magnetic forces that emanated from one person to another. By the late 19th century, both magnetism and hypnotism, a method developed by James Braid (1795–1860), had begun to fall into disrepute as therapeutic procedures. A modest physician working in a rural region near Nancy in France had heard of James Braid’s work at a lecture and decided to explore its possibilities in his limited practice. Well regarded in his local community,

Ambroise-Â�Auguste Liébault (1823–1904) utilized a simple method of inducing sleep by suggesting to patients that they look into his eyes while he spoke to them in quiet tones. In 1866, Liébault published a small book titled Du Sommeil et des États Analogues (Sleep and Analogous States), in which he stressed that the power of suggestion not only was central to successful hypnotism, but was the primary vehicle of therapeutic efficacy. Liébault was generally considered a simpleton, if not a quack, by his colleagues. Nevertheless, rumors of his therapeutic successes came to the attention of a wellÂ�regarded professor of medicine at the Nancy School of Medicine, Hippolyte-Marie Bernheim (1840–1919), a young Jewish physician who had recently been appointed to this new medical institution. Bernheim had been treating a patient with sciatica for 6 years with minimal success. He referred this patient to Liébault, who utilized his methods of suggestive sleep and succeeded within 6 months in fully relieving the patient of the disorder. As a result, Bernheim decided to experiment with Liébault’s radical hypnotic methods in his own clinic. We have just discussed Charcot’s signal importance in developing methods of clinical neurology. By contrast, his role in fostering a psychoanalytically oriented psychiatry stems less from the intent or the originality of his work than in the incidental part he played in stimulating the ideas of others, notably Freud and Janet. As noted earlier, Charcot studied the diverse and confusing symptoms of hysteria at the Salpêtrière. Because of his neurological orientation, he viewed trances, memory losses, and bodily anesthesia as diagnostically difficult cases of an underlying nervous system disease. It was not until his associates demonstrated that the symptoms of hysteria could be induced by hypnotic procedures that Charcot reconsidered his views of this puzzling ailment. His inability to differentiate between hypnotized and naturally produced paralyses, as well as the frequently noted migration or disappearance of symptoms and the anatomically impossible location of many of the paralyses he saw, convinced him that hysteria could not be a product of a simple injury or local disease of the nervous system. Despite suggestive evidence to the contrary, Charcot could

A Précis of Psychopathological History

not abandon his biological perspective. To accommodate his observations, he proposed that hysteria resulted from a wide-Â�ranging and congenital neurological deficiency, and that hypnosis merely served as a precipitant of the inborn defect. Charcot presented his neurological thesis regarding hypnotism at the French Academy of Sciences in the early 1880s. Shortly thereafter, Bernheim brought to the world’s attention Liébault’s alternative interpretation concerning the role of suggestion in the hypnotic technique. First, Bernheim wrote, hypnosis could be employed with a variety of ailments; second, its effects stemmed from the power of suggestion; and third, all humans were susceptible to suggestion in varying degrees. Although Bernheim was an internist and not a neurologist or psychiatrist, he vigorously disagreed with Charcot—Â�maintaining that hysteria was primarily a state of heightened self-Â�suggestion, and that hypnosis was an equivalent state induced by others. Moreover, Bernheim advanced the view that hysteria was essentially a psychogenic disorder, and applied the term “psychoneurosis” to this and similar puzzling symptom syndromes. His belief that unconscious self-Â�suggestion might underlie the symptoms of many mental disorders played a significant role in influencing Freud’s thinking. In developing the concept of psychoneurosis, Bernheim sought to parallel the medical tradition of seeking underlying biological causes for the disorder with a comparable notion of underlying psychological causes. Josef Breuer (1842–1925) was born in Vienna, where his father was a well-known teacher and author of Jewish thought. He helped Freud financially in his early years. Even more importantly, he whetted Freud’s curiosity about both hysteria and hypnosis in discussing a young patient of his, later to become famous under the pseudonym of Anna O. The case of Anna O. was described to Freud in 1880; it involved a classical example of hysteria, which followed a period when the young woman had nursed her father through a major illness. Breuer employed a hypnotic technique to encourage his patient to voice her experiences and thoughts at the time her symptoms had emerged. The memories that Anna O. recalled under hypnosis were accompanied by intense outbursts of emotion


that she had been unable to vent at the time of her symptoms. Moreover, she became intensely attached to Breuer; uncomfortable with her affectionate feelings toward him, Breuer withdrew from the case. Some years thereafter, Freud traveled to Paris and later to Nancy, where he observed the methods that Breuer had utilized—both those of Charcot and, later, those of Bernheim. Upon his return from these travels in the late 1880s, Breuer and Freud continued their discussions with a series of new cases employing the methods of hypnosis and the stirrings of emotional catharses. This work ultimately led to a series of papers and the publication of a major book, entitled Studies on Hysteria, in 1895. Freud and Breuer formulated their idea in this text that patients with hysteria suffered from repressed memories of emotionally traumatic events—Â�events so distressing that the emotions they aroused could not be faced consciously at the time they occurred. It was Freud and Breuer’s contention that the technique for curing hysteria was to unblock the repressed and pentup emotions that were “kept secret” in the unconscious. Pierre Janet’s (1859–1947) career was an unusual one for a psychiatrist. Janet first taught philosophy at a small college, the Lyceum in Chateau Roux, in the rural province of Berry, and later at the Lyceum in Le Havre, where he remained for over 6 years. He began his early clinical work at the Le Havre mental hospital, where he was assigned the task of examining all incoming women who were deemed to have hysteria. Most of Janet’s patients at Le Havre were young, fresh, and unsophisticated, unlike the usual inmates at the major institutions of France, such as the Salpêtrière, who had typically been examined numerous times by scores of physicians and students. By the mid-1880s, Janet had turned to the highly esteemed studies of Jean Charcot, as well as those of other scholars engaging in what was known as “psychical” research. Janet might have been considered the most original thinker about psychoanalytic processes, had he not been overshadowed by the unusually courageous and innovative Freud. Janet evolved a theory in which neuroses resulted from an inability to integrate coÂ�occurring psychic processes; this thesis foreshadowed, and may have led Bleuler to, the


concept of dementia praecox (schizophrenia) as a split between thought and emotion. As did Freud, Janet observed that painful experiences and undesirable impulses could not be tolerated by his patients. In developing his concept of “dissociation,” Janet speculated that intolerable thoughts and feelings might take on an independent existence within a person and manifest themselves in amnesia, multiple personality, hysterical fits, and/or conversion paralyses. In this formulation, Janet recognized that different systems of thought could become pathologically separated, with one or another part lost to consciousness. This strengthened the idea that unconscious processes might persist unmodified within the person. Despite his capacity to describe his patients and their frequent exotic behaviors and complaints, Janet did not display Charcot’s and Freud’s relentless curiosity, or their courage in exploring the outer reaches and deeper roots of their patients’ psyches. He seemed overly cautious and circ*mspect, unable to plumb the depths of psychic conflict and sexual pathology. As some have characterized him, he was a “neat and well-Â�stocked pantry, with everything in its proper place.” Sigmund Freud (1856–1939; see Figure 1.6) was arguably the most influential psychologist and physician of the 20th century. His reinterpretation of the observations first made by Charcot and Bernheim initiated an

Figure 1.6.╇ Sigmund Freud.

intellectual and cultural revolution of worldwide proportions. His theories have been both extravagantly praised and intensely castigated. Venerated by some and condemned by others, Freud has been spoken of at times as one of history’s greatest scientists, and at others as a fraudulent cult leader. Numerous historians refer to him as the greatest psychologist of all time, the profoundest of all human scientists. Others are convinced that the unconscious never existed except in Freud’s mind, and that his theories were baseless and aberrational. Some speak of him as a false prophet; others depict him as a courageous fighter for the truth. His most condemning detractors describe him as a neurotic egotist who propounded irrational and fantastic theories. More balanced historians aver that Freud’s discoveries merely crystallized previously diffuse ideas of his many predecessors, such as those described in previous pages. Personally and professionally, Freud was a man of divergent dispositions. A militant atheist and radical theorist, he espoused liberated attitudes toward sexuality; at the same time, he was politically conservative, usually somber and unsmiling, impeccably dressed, invariably anxious about finances, clearly suffering in his middle years from assorted psychosomatic symptoms, and fearfully hesitant about modern contrivances. He always felt that he was an outsider. “A godless Jew” and free thinker, yet conservative in personal behavior, prissy, and formalistic, he did not leave his home city until forced to do so following the Nazi takeover of Austria. Freud devoted his long and fruitful life to the development and elaboration of his theories and techniques. Unlike his great German contemporary Kraepelin, who sought to classify broad groups of disorders with a common course and symptoms, Freud stressed the brightly etched inner memories, the feverish imaginations, and the unique attributes of each patient. And unlike Janet, his French contemporary, who viewed neuroses as the results of an underlying constitutional deficiency, Freud set out to trace the perplexing ambiguities, the afflicted emotional palette, the convoluted psychogenic origins, and the primitive passions that he perceived and explored as the unconscious source and undergirding force of each manifest disor-

A Précis of Psychopathological History

der. It was not only the dense interplay of refracted realities in his findings that proved so epochal; the ever-Â�dividing and sprawling new lines of his individualistic philosophy and his orientation toward the implausible and desultory character of life’s realities, as well as the odds and ends of its rarefied energies, all served as a foundation for the 20thÂ�century understanding of humankind’s complicated and intriguing nature. According to Freud, each stage of psychosexual development would produce a distinctive set of anxieties and defenses resulting from instinct frustration and conflict. Symptoms and character traits would arise from the persistence into adulthood of childhood anxieties and defenses. Freud’s early disciples, notably Karl Abraham (1877–1925) and Wilhelm Reich (1897–1957), differentiated the oral psychosexual period into two phases: the “oral-Â�sucking” phase, in which food was accepted indiscriminately, followed by the “oral-Â�biting” period, in which food was accepted selectively, occasionally rejected, and aggressively chewed. In their view, excessive gratifications, conflicts, or frustrations associated with each of these phases could establish different patterns of adult personality. For example, an overly indulgent sucking stage might lead to imperturbable optimism and naive self-Â�assurance. An ungratified sucking period might lead to excessive dependency and gullibility; for example, deprived children might learn to accept anything in order to ensure that they will get something. Frustration experienced at the biting stage might lead to the development of aggressive oral tendencies such as sarcasm and verbal hostility in adulthood. At a later period in his exploration of the disorders of personality, Freud speculated that character classification could be based on his threefold structural distinction of “id,” “ego,” and “superego.” Thus, in 1931, he sought to devise character types in accord with which psychoanalytic structure was dominant. First, he proposed an “erotic” type—Â�persons whose lives were governed by the instinctual demands of the id. “Narcissistic” individuals were so dominated by the ego that neither other persons nor the demands of id or superego can affect them. “Compulsive” persons were so tightly regulated by the strictness of the superego that all other functions were dominated. Lastly,


Freud identified a series of “mixed” types in which combinations of two of the three characterological structures outweighed the third. Freud’s compulsive character type has been well represented in the literature, but only in the past 30 years have his proposals for a narcissistic personality disorder gained attention (Millon, 1981, 1996). Alfred Adler (1870–1937), founder of the school of individual psychology, became an outspoken critic of Freud’s views on infantile sexuality shortly before Jung did in 1911. On the basis of his own clinical observations, Adler concluded that superiority and power strivings were more fundamental to pathology than sexuality was. Although many of his patients were not overtly assertive, he observed that their disorder enabled them to dominate others in devious and subtle ways. Phobias and hypochondriasis, for example, not only excused patients from disagreeable tasks, but allowed them to control and manipulate others. Adler hypothesized that these strivings for superiority were consequences of the inevitable and universally experienced weakness and inferiority in early childhood. In this conception, Adler attempted to formulate a universal drive that would serve as an alternative to Freud’s universal sexual strivings. According to Adler, basic feelings of inferiority led to persistent and unconscious compensatory efforts. These were manifested in pathological struggles for power and triumph if individuals experienced unusual deficiencies or weaknesses in childhood. Among healthier personalities, compensation accounted for strivings at self-Â�improvement and interests in social change and welfare. These compensatory struggles or strivings, acquired by all individuals as a reaction to the restrictions imposed by their more powerful parents, led to general patterns of behavior that Adler called “styles of life.” Although chosen by Freud as his heir apparent, Carl Gustav Jung (1875–1961) did not agree with Freud’s emphasis on the sexual nature of development and motivation, and established his own system of analytic psychology in 1913. Jung expanded the notion of “libido,” Freud’s concept for the basic sexual energies, to include all life-Â�propelling forces. The concept of “racial memories,” later termed the “collective unconscious,” was proposed to suggest that instinctual


forces were more than seething animalistic impulses; according to Jung, these forces contained social dispositions as well. These primitive dispositions were often expressed in folklore and mystical beliefs. When no acceptable outlet could be found for them in societal life, they took the form of symptoms such as phobias, delusions, and compulsions. Jung’s belief in unconscious social dispositions led also to his formulation of two basic personality types, the “extrovert” and the “introvert.” Despite these and other original contributions, Jung’s views had a minimal impact upon the mainstream of psychodynamic theory and practice. Karen Horney (1885–1952) contended that neurotic disorders reflected cultural trends learned within the family; she minimized biological determinants and stressed interpersonal relationships. She believed that anxiety and repressed anger were generated in rejected children and led to feelings of helplessness, hostility, and isolation. As these children matured, they developed an intricate defensive pattern of either withdrawal, acquiescence, or aggression as a means of handling their basic anxiety. Although Horney felt that adult patterns resulted largely from early experience, she argued, in contrast with Freud, that therapy should focus on its adult form of expression. First, she averred that the intervening years between childhood and adulthood caused important changes in adaptive behavior. And, second, present-day realities had to be accepted, and the goals of therapy had to take them into account. Horney’s descriptive eloquence was without peer; nevertheless, difficulties arose when she summarized what she referred to as the major “solutions” to life’s basic conflicts. Although her primary publications were written over a short period, she utilized different terms to represent similar concepts (Horney, 1937, 1939, 1942, 1945, 1950). Faced with the insecurities and inevitable frustrations of life, Horney identified three emergent modes of relating: “moving toward” people, “moving against” people, or “moving away” from them. In her 1945 book, Horney formulated three character types to reflect each of these three solutions: Moving toward was found in a “compliant” type; moving against, in an “aggressive” type, and moving away, in a “detached” type. In 1950, Horney reconcep-

tualized her typology in line with the manner in which individuals solve intrapsychic conflicts; she termed these solutions “selfÂ�effacement,” “expansiveness,” and “neurotic resignation.” Although these sets of three do not match perfectly, they do correspond to the essential themes of Horney’s characterology. Several major thinkers from Great Britain began to formulate what is referred to as the “object relations” approach to psychoanalytic theory in the 1940s and 1950s. Most inventive of these was Melanie Klein (1882–1960), one of the originators of child psychoanalysis (along with Anna Freud, with whom she vigorously differed and contended for leadership in the British analytic community). Klein’s views met with intense opposition in the wider psychoanalytic world, and fierce battles raged within British analytic circles over her inventive concepts. Although she was a vigorous critic of more orthodox psychoanalytic thought, she believed that emphasizing the very earliest and most primitive stages of development was a natural extension of Freud’s original formulations. In the United States since the mid1960s, Otto Kernberg (1928– ) has sought to develop a synthesis of drive reduction and object relations frameworks—an approach that has brought considerable attention to modern analytic thought, as well as generating considerable controversy. Owing to numerous pragmatic considerations at the time—not the least of which were the advent of effective psychopharmacological medications and the emergence of the sophisticated community mental health movement—the balance of power within American psychiatry shifted slowly but surely away from psychoanalysis in the 1970s. The wider culture had also reconsidered the high repute in which it had formerly held for psychoanalysts: They were no longer seen as wise, generous, and kindly, but were depicted increasingly as irrelevant stumblebums. In a review of what was wrong with psychoanalysis, Richard Weber (1995) stated that concepts such as infantile sexuality were more than objects of disbelief, not so much disproven as incapable of disproof; in his view, they should be relegated to the same scientific status as astrology. And eminent English psychologist Hans Eysenck (1985) asserted that just as chemistry had had to

A Précis of Psychopathological History

unshackle itself from the fetters of alchemy, and the brain sciences had had to disengage themselves from phrenology, so too must psychology and psychiatry abandon the pseudoscience of psychoanalysis. Questions have been raised as to whether or not scientific concepts can be founded on unconscious data. Psychoanalytic theories have been criticized as unscientific mixtures of metaphorical analogies, speculative notions, and hypothetical constructs because their data are anchored so tenuously to the observable world. Added to this rather harsh judgment is the equally critical view that the methods of collecting unconscious data are both unreliable and imprecise. How can concepts of the unobservable unconscious be empirically anchored? Can one accept what a patient says without having it corroborated by external evidence? Is the patient an unbiased judge, or is he or she motivated to agree with the all-Â�knowing therapist? These and many other questions have been raised about the subjective and methodologically uncontrolled procedures used for the development of psychoanalytic theories. To critics, the ingenious speculations of psychoanalytic theorists are at best a starting point—a preliminary set of propositions requiring reformulation as clearly specified hypotheses that can be confirmed or disproved. Despite these criticisms, psychoanalytic processes may be a necessary part of the study of humankind’s pathological functioning. These processes may be difficult to formulate according to the tenets of scientific objectivity, but their existence cannot be denied or overlooked. Efforts to unravel them may fall prey to theoretical obscurity and methodological difficulties, yet the search should be mandatory. To the deeply inquiring and instinctively insightful thinker, the intricate themes of mental life articulated by analysts have a richness and unquestionable accuracy about them.

Current Trends In the latter decades of the 20th century, several major theorists appear to have developed a strong foundation of ideas that may influence the future course of psychopathology’s history. We describe some of them briefly here.


Aaron Timothy Beck (1921– ) has been a prominent and insightful contributor to cognitive therapy, especially as applied to a wide range of the Axis I clinical syndromes. More recently, he and his associates have addressed the subject of personality, articulating “cognitive schemas” that shape the experiences and behaviors of numerous personality disorders. Beck focused his early research efforts largely on testing psychoanalytic theories of depression, but when his studies failed to support his hypotheses, he explored a more cognitive explanation of the disorder. He found that most depressed patients had broad negative views of themselves, of the world at large, and of their own future. Beck reasoned that these negative “cognitive distortions,” as he termed them, could be reoriented to accord with reality through the application of logic and the rules of evidence. He eventually applied these cognitive investigations to a broad range of disturbances, from anxiety to substance use to personality disorders. Cognitive approaches to the treatment of mental disorders have become more than merely the mainstream of “talking therapies” today. More than one-third of all therapists speak of themselves as cognitive in orientation; the others employ cognitive techniques periodically. C. Robert Cloninger (1945– ) has formulated a recent model of personality dispositions, drawing upon genetic and neurobiological substrates. Cloninger’s complex theory is based on the interrelationship of several heritable characteristics or functional dispositions, notably “novelty seeking,” “harm avoidance,” and “reward dependence.” Each of these is associated with different neurobiological systems (dopamaninergic, serotonergic, and noradrenergic, respectively). More specifically, novelty seeking is hypothesized to dispose individuals toward exhilaration or excitement in response to novel stimuli; it leads to the pursuit of potential rewards, as well as an active avoidance of both monotony and punishment. Harm avoidance reflects a disposition to respond strongly to aversive stimuli, leading individuals to inhibit behaviors to avoid punishment, novelty, and frustrations. Reward dependence is hypothesized as a tendency to respond to signals of reward (e.g., verbal signals of social approval), and to resist extinction of behaviors previously associated with re-


wards or relief from punishment. To extend the theme of novelty seeking, for example, individuals high on this dimension but average on the other two dimensions would be characterized as impulsive, exploratory, excitable, quick-Â�tempered, and extravagant—Â� likely to seek out new interests, but inclined to neglect details and to become quickly distracted or bored. Larry Siever’s (1950– ) theoretical model has also attempted to link neurotransmitters’ properties to the various personality disorders. Siever has developed a dimensional model that has major clinical syndromes at one extreme and the milder personality disorders at the other end. He proposes four major dimensions: “cognitive/perceptual organization,” “impulsivity/aggression,” “affective instability,” and “anxiety/inhibition.” For example, schizophrenic disorders are viewed as disturbances of a cognitive/ perceptual nature, exhibiting themselves in thought disorders, psychotic symptoms, and social isolation; schizotypal personality disorder would serve as the prototype among the Axis II disorders. Disorders of impulsivity/aggression are hypothesized as resulting in poor impulse control, particularly as evident in aggressive actions. In the more distinct clinical syndromes, Siever suggests the presence of impulsivity/aggression in explosive disorders, pathological gambling, or kleptomania. When this dimension is more pervasive and chronic, it may be seen in persistent self-Â�destructive behaviors, such as those characteristic of borderline and antisocial personality disorders. Problems of affective instability are most clearly observed in the intensity and dysregulation of mood disorders. When this inclination is more sustained over time, it may interfere with the development of stable relationships and selfimage, as may be manifested in borderline and histrionic personality disorders. Lastly, the anxiety/inhibition dimension appears to be related to the Axis I anxiety syndromes (e.g., social phobia, compulsive rituals); when it is present at a low threshold over extended periods of development, avoidant, compulsive, or dependent personality disorder may result. Eric Kandel (1930– ) left Austria for the United States with his family in 1939. He attended Harvard University to study the hu-

manities. Intrigued by his reading of Freudian literature, Kandel went to medical school to pursue a career as a psychoanalyst. Early in his training, Kandel undertook work in neurophysiology, in the hope of gaining a clearer understanding of how memory and emotions are biologically generated and intertwined. His early research led him to explore the hippocampus as the primary source of memory formation, but he soon turned to a simpler neurosystem for his intensive analysis. Utilizing the Aplysia sea slug, a creature with only 20,000 nerve cells, as an experimental animal, Kandel identified a number of the biochemical changes that accompany memory formation—Â�explicating how shortterm memory involves just a minor modulation of the synapses, whereas long-term memory requires new synaptic linkages. More specifically, he showed that a protein termed CREB helps the nervous system retain a memory or a learned skill for a long period of time rather than just briefly. This work earned Kandel, together with Paul Greengard and Arvid Carlsson, the Nobel Prize in Physiology or Medicine in 2000. Together, they began to decode how the vast numbers of synaptic connections in the brain (at least 100 trillion) are able to communicate continually, as well as to alter their strength, flexibility, and function. Although Paul Meehl’s (1920–2003) biologically oriented social learning model is limited to schizophrenia, it is notable for both its elegance and specificity. He hypothesized that only a certain class of people, those with a particular genetic constitution, have any liability to schizophrenia. Meehl suggested that the varied emotional and perceptual–Â�cognitive dysfunctions people with schizophrenia display are difficult to explain in terms of single-Â�region disorders. The widespread nature of these dysfunctions suggested to Meehl the operation of a more diffuse integrative neural defect. Although a combination of different neurological disturbances can account for this defect, he opted for an explanation in terms of deficits in synaptic control. More specifically, he believed that the major problem in schizophrenia lies in a malfunctioning of the two-way mutual control system between perceptual–Â� cognitive regions and the limbic motivation center. Meehl proposed that integrative neu-

A Précis of Psychopathological History

ral defects are the only direct phenotypic consequences produced by the genetic disorders; these consequences, given the label “schizotaxia,” are all that can properly be spoken of as inherited. The imposition of certain social learning histories on schizotaxic individuals results in a personality organization that Meehl called the “schizotype.” Four core behavior traits—Â�namely, anhedonia, cognitive slippage, interpersonal aversiveness, and ambivalence—are not innate. However, Meehl postulated that schizotaxic individuals universally learn them, given any existing social learning regimen (from the best to the worst). If the social environment is favorable and a schizotaxic person has the good fortune of inheriting a low anxiety readiness, physical vigor, and a general resistance to stress, the personality organization will remain a well-Â�compensated schizotype, and the individual may never manifest symptoms of clinical schizophrenia. In England, two child-Â�oriented analysts, Michael Balint (1896–1970), and John Bowlby (1907–1990), contributed to an understanding of developmental vicissitudes. Balint’s concept of the “basic fault” was derived from studies of patients whose borderline characteristics appeared to be consequences of having missed something during the first year or two of life. In Balint’s view, such a fault can lead to one of two extreme reactions. In the so-Â�called “ocnophile” adaptation, infants deal with the experience by clinging excessively to others; in the “philobat” adaptation, children learn to distance themselves from others and rely entirely on themselves. Bowlby stressed “attachment learning,” especially that resulting from the loss of a significant early relationship. He spoke of children suffering maternal loss as passing through three phases: protest, despair, and detachment. In the first stage, children evidence anger at their loss; in the second, children begin to lose hope that the mother will ever return; finally, despair turns to detachment (i.e., the children become depressed and unresponsive). Sharing Melanie Klein’s object relations model, Bowlby asserted that the manner in which children deal with affectional deprivation will determine how they will react in later life to problematic relationships with loved ones.


In the 1980s, sharp criticisms were raised against the dominance of the DSM and ICD systems, both based on a Kraepelinian vision (Klerman, 1986). Some eminent psychiatrists have asserted that, except for organic disorders, a classificatory diagnosis is less important than a psychodynamic study of personality; that is, rather than fitting a patient’s symptoms into a fixed classificatory scheme, a clinician should seek to understand the person in terms of his or her distinctive life experiences (American Psychoanalytic Association, 2006). Others have noted that too much research time is wasted and too many errors are perpetuated because investigators cling to an outdated classification. Proponents, on the other hand, have explained the viability of a Kraepelinian schema by the “fact” that there is a considerable amount of truth contained in the system and the practical implications associated with its labels; that is, they are more than merely sufficient when compared with the power of competing concepts. One of the major problems facing a field as inchoate and amorphous as mental health is its susceptibility to subjective values, cultural biases, and chance events. Were the field a “hard” science, anchored solidly in readily verified empirical fact, progress would presumably derive from advances of a tangible and objective nature. Unfortunately, that is not the case. Nevertheless, the field has endeavored to standardize, as much as possible, the language conventions and classification rules for diagnosing mental disease categories. To say the least, effective communication among clinical centers was seriously compromised, as were useful records for epidemiological statistics and research. Theodore Millon (1928– ) has come to believe that the widespread desire among theorists to unify science should not be limited to explicating physics; that is, it should be possible in all fields of nature that have been subdivided by habit, tradition, or pragmatics (e.g., economics, sociology, geology). He believes unification to be a worthy goal even within the newer sciences, such as personology. Efforts to coordinate the separate realms that constitute the field of the mind and, more specifically, the field of mental disorders should be particularly useful. Rather than independently developing autonomous


and largely unconnected professional activities and goals, a truly mature mental science should embody, and create a synergistic bond among, five explicit elements: 1. Universal scientific principles that are grounded in the ubiquitous laws of nature; despite their varied forms of expression, these principles may provide an undergirding framework for constructing narrow-based subject-Â�oriented theories. 2. Subject-Â�oriented theories, or explanatory and heuristic conceptual schemas of the mind and mental illness. These theories should be consistent with established knowledge in both their own and related sciences, and should enable reasonably accurate propositions concerning all clinical conditions to be both deduced and understood, enabling thereby the development of a formal classification system. 3. Classification of personality styles and pathological syndromes, or a taxonomic nosology that has been derived logically from the theories. The taxonomy should provide a cohesive organization within which its major categories can readily be grouped and differentiated, permitting thereby the development of coordinated assessment instruments. 4. Personality and clinical assessment instruments, or tools that are empirically grounded and sufficiently sensitive quantitatively to enable the theories’ propositions and hypotheses to be adequately investigated and evaluated. Hence the clinical categories constituting the nosology should be able to be readily identified (diagnosed) and measured (dimensionalized), thus specifying target areas for interventions. 5. Integrated therapeutic interventions, or planful strategies and modalities of treatment. These interventions should accord with the theories and be oriented to modify problematic clinical characteristics, consonant with professional standards and social responsibilities. Perhaps the only realistic and significant question to be posed in appraising a new taxonomy or nomenclature is not whether it mirrors the state of the science perfectly, or whether it provides answers to all possible questions professionals within the dis-

cipline may ask, but whether it represents an advance over preceding nosological systems and whether it will be employed with greater clinical accuracy and facility by future practitioners and researchers. Having participated over two intense 5-year periods as a member of the DSM-III and DSM-IV committees, Millon is considerably more charitable than he once was about the purposes and success with which these task forces met their responsibilities. He has no illusion, however, that the task was completed. As he wrote nearly two decades ago, Classifying mental illness must be an outgrowth of both psychology and medicine. As such, efforts to construct a taxonomy must contend with the goals, concepts, and complications inherent in both disciplines (e.g., context moderators, definitional ambiguities, overlapping symptomatologies, criterion unreliabilities, multidimensional attributes, population heterogeneities, instrument deficits, and ethical constraints). (Millon, 1991, p.€245)

Thus the profession remains unsure today whether to conceive depression as a taxon (category) or an attribute (symptom); whether to view it as a dimension (with quantitative degrees of severity) or as a set of discrete types; or whether to conceive it as a neuroendocrinological disease or as an existential problem of life. Although debates on these issues often degenerate into semantic arguments and theoretic hairsplitting, it is naive to assume that metaphysical verbiage and philosophical word quibbling are all that are involved. Nevertheless, the language we use, and the assumptions such language reflects, are very much a part of our scientific disagreements. This volume addresses these substantive and philosophical issues as they may apply to DSM-V and ICD-11. In addition to reviewing the history of psychopathology, this chapter has sought to illustrate that philosophical issues and scientific modes of analysis must be considered in directing the future of mental illness classification. The many recommendations made in this book will not in themselves achieve clear resolutions to all nosological quandaries. It is more likely that their role will be to unsettle prevailing habits and thereby force progress, if only by challenging cherished beliefs and assumptions.

A Précis of Psychopathological History

Acknowledgment Portions of this chapter have been adapted and updated from Millon (2004). Copyright 2004 by John Wiley & Sons. Adapted by permission.

References American Psychoanalytic Association. (2006). Psychodiagnostic manual. Silver Spring, MD: Author. Aschaffenburg, G. (1922). Constitutional psychopathies. In G. Aschaffenburg, Handbook of medical practice (Vol. 4). Leipzig: Barth. Baillarger, J. (1853). De la melancholie avec stupeur. Annales Medico-Â�Psychologiquies, 5, 251. Bhugra, D. (1992). Psychiatry in ancient Indian texts: A review. History of Psychiatry, 3, 167– 186. Bleuler, E. (1950). Dementia praecox oder gruppe der schizophrenien. Leipzig: Deuticke. (Original work published 1911) Briquet, P. (1859). Traite clinique et therapeutique a l’hystérie. Paris: Baillière. Burton, R. (1621). The anatomy of melancholy: What it is, with all the kinds, causes, symptoms, prognostickes, and severall cures of it. Oxford, UK: Lichfield for Cripps. Chaslin, P. (1912). Elements de seminologie et de clinique mentale. Paris: Asselin et Houzeau. Comrie, J. D. (1922). Selected works of Thomas Sydenham, M.D. New York: William Wood Co. Cullen, W. (1777). First lines of the practice of physick for the use of students in the UniÂ� versity of Edinburgh (4 vols.). Edinburgh, UK: Creech. de la Chambre, M. C. (1640). Les charactères des passions (Vol. 1). Paris. de Sauvages, F. B. (1731). Treatise de nouvelles classes de maladies. Paris: Avignon. de Sauvages, F. B. (1771). Nosologie methodique dans la quelle les maladies sont rangées par classes. Paris: Herrisant. Esquirol, J. (1838). Des maladies mentales (2 vols.). Paris: Baillière. Eysenck, H. J. (1985). Decline and fall of the Freudian empire. London: Penguin. Falret, J. P. (1854). De la folie circulaire. Bulletin de l’Academie Medicale, 19, 382–394. Falret, J. P. (1862). Traite pratique des maladies mentales. Paris. Feuchtersleben, E. (1847). The principles of med-

51 ical psychology: Being the outlines of a course of lectures. London: Sydenham Society. Freud, S. (1950). Libidinal types. In Collected papers (Vol. 5). London: Hogarth Press. (Original work published 1931) Freud, S., & Breuer, J. (1895). Studien über hysterie. Vienna. Goddard, K. (1991). Morita therapy: A literature review. Transcultural Psychiatry, 28, 93–115. Griesinger, W. (1845). Mental pathology and therapeutics. London: New Sydenham Society. Griesinger, W. (1868). Introductory comments. Archives for Psychiatry and Nervous Diseases, 1, 12. Haslam, J. (1809). Observations on madness and melancholy. London: Callow. Haslam, J. (1810). Illustrations of madness: Exhibiting a singular case of insanity. London: G. Hayden. Heinroth, J. C. (1818). Lehrbuch der Störungen des Seelenlebens. Leipzig: Thieme. Hellpach, W. (1920). Amphithymia. Zeitschrift für die Gesamte. Neurologie und Psychiatrie, 19, 136–152. Hertt, W. S. (Ed. & Trans.). (1936). Physiognomics. In Aristotle: Minor works. Cambridge, MA: Harvard University Press. Horney, K. (1937). The neurotic personality of our time. New York: Norton. Horney, K. (1939). New ways in psychoanalysis. New York: Norton. Horney, K. (1942). Self-Â�analysis. New York: Norton. Horney, K. (1945). Our inner conflicts. New York: Norton. Horney, K. (1950). Neurosis and human growth. New York: Norton. Kahlbaum, K. L. (1874). Die Katatonie, oder das Spannungsirresien. Berlin: Kirschwald. Koch, J. A. (1891). Die psychopathischen Minderwertgkeiten. Ravensburg: Maier. Klerman, G. L. (1986). Historical perspectives on contemporary schools of psychopathology. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV. New York: Guilford Press. Kraepelin, E. (1896). Psychiatrie: Ein Lehrbuch für Studirende und Aertze (5th ed.). Leipzig: Barth. Kraepelin, E. (1899). Psychiatrie: Ein Lehrbuch (6th ed.). Leipzig: Barth. Kraepelin, E. (1913). Psychiatrie: Ein Lehrbuch für Studirende und Aerzte (8th ed., 4 vols.). Leipzig: Abel.

52 HISTORICAL AND CULTURAL PERSPECTIVES Kraepelin, E. (1920). Hundert jahre psychiatrie. Berlin: Springer-Â�Verlag. Kraepelin, E. (1921). Manic–Â�depressive insanity and paranoia. Edinburgh, UK: Livingstone. Krafft-Ebing, R. von. (1879). Lehrbuch der Psychiatrie auf klinischer Girundlage. Stuttgart: Erike. Krafft-Ebing, R. von. (1937). Psychopathia sexualis. New York: Physicians & Surgeons Books. (Original work published 1882) Kramer, H., & Sprenger, J. (1928). Malleus malefacarum [The witches’ hammer] (M. Summers, Trans.). London: Rodker. (Original work published 1487–1489) Lavater, J. C. (1789). Essays of physiognomy: for the promotion of knowledge and the love of mankind (T. Holcroft, Trans.). London: Robinson, Paternoster-Row. Liébault, A. A. (1866). Du sommeil et des états analogues consideres surtout au point de vue de l’action du moral sur le physique [Sleep and analogous states]. Paris: Masson. Liu, X. (1981). Psychiatry in traditional Chinese medicine. British Journal o9f Psychiatry, 138, 429–433. Maudsley, H. (1860). Annual board report at Cheadle Royal Hospital. Unpublished manuscript. Maudsley, H. (1876). Physiology and pathology of mind. New York: Appleton. Meyer, A. (1906). Fundamental conceptions of dementia praecox. British Medical Journal, 2, 757–760. Meyer, A. (1912). Remarks on habit disorganizations in the essential deteriorations. Nervous€ and Mental Disease Monographs, 9, 95–109. Meyer, A. (1912). Remarks on habit disorganizations in the essential deteriorations. Nervous and Mental Disease Monographs, 9, 95–109. Millon, T. (1981). Disorders of personality: DSM-III, Axis II. New York: Wiley. Millon, T. (1991). Classification in psychopathology: Rationale, alternatives, and standards. Journal of Abnormal Psychology, 100, 245– 261. Millon, T. (2004). Masters of the mind. Hoboken, NJ: Wiley. Millon, T. (with Davis, R. D.). (1996). Disorders

of personality: DSM-IV and beyond (2nd ed.). New York: Wiley. Morel, B. A. (1857). Traité de dégénérescences physiques, intellectuelles et morales de l’espèce humaine. Paris: Baillière. Paracelsus. (1941). Von den Krankheiten, so die Vernufft berauben (G. Zilboorg, Trans., & H. Sigerist, Ed.). Baltimore: Johns Hopkins University Press. (Original work published 1567) Peck, A. L. (Ed. & Trans.). (1965). Aristotle: Historia animalium. Cambridge, MA: Harvard University Press. Porta, G. B. della. (1957). Natural magic (D. J. Price, Ed.). New York: Basic Books. (Original work published 1558) Porta, G. B. della. (1586). De humana physiognomia. Apud J. Cacchium: Vici Aequensis. Porta, G. B. della. (1588). Phytognomonica. Naples: Saluianum. Porter, R. (2002). Madness. New York: Oxford University Press. Robinson, D. R. (1976). An intellectual history of psychology. New York: Macmillan. Roccatagliata, G. (1973). Storia de la psichiatria antica. Milan: Ulrico Hoepli. Sacher-Â�Masoch, L. (1870). Venus in furs. Dresden: Dohrn. Thomasius, C. (1692). Recent proposals for a new science for obtaining a knowledge of other men’s minds. Halle, Germany: Christopher Salfeld. Tredennick, H. (Ed. & Trans.). (1967). Aristotle’s prior analytics. Cambridge, MA: Harvard University Press. Voisin, F. (1851). The analysis of human understanding. Weber, R. (1995). Why Freud was wrong: Sin, science and psychoanalysis. New York: Basic Books. Willis, T. (1971). De anima brutorum [Two discourses concerning the soul of brutes]. Gainesville, FL: Scholar’s Facsimiles & Reprints. (Original work published 1672) Willis, T. (1978). Cerebri anatome [The anatomy of the brain and nerves]. Birmingham, AL: Classics of Medicine Library. (Original work published 1664) Zilboorg, G., & Henry, G. W. (1941). A history of medical psychology. New York: Norton.

Chapter 2

Themes in the Evolution of the 20th-�Century DSMs Roger K. Blashfield Elizabeth Flanagan Kristin Raley

recent book by Wedding, Boyd, and NiA emiec (2005) is titled Movies and Mental Illness: Using Films to Understand Psy-

chopathology. Wedding and colleagues start their book with a quotation from a screenwriter for A Beautiful Mind, in which the writer makes an analogy between watching movies about psychopathology and going to a zoo. In both instances, the viewer is presented with instances of wondrous diversity. Yet an organized zoo, in many ways structured around the principles of biological classification, can illustrate how this diversity can both be explained and demonstrate regularities that a person might miss when first looking at the wild, scary, and beautiful creatures of the zoo. Consistent with the quotation that they use to start their book, Wedding and colleagues (2005) base each chapter in their book on a family of mental disorders as represented in the current, official edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM). Discussions of the movies are placed in the book according to the diagnostic concept that each movie illustrates, at least in the authors’ opinion. By using DSM in this way, Wedding et al. are drawing on

one of the basic purposes of a classification: It is a systematic organization of concepts describing the diversity of mental disorders. According to Blashfield and Draguns (1976), the five purposes of a classification of psychopathology are as follows: to serve (1) as a nomenclature, or a list of accepted nouns that mental health professionals use to talk about the patients that they see; (2) as a basis for information retrieval, so that novices can search for existing knowledge about mental disorder categories; (3) as a descriptive system, in which the name of each mental disorder summarizes the behaviors, thoughts, and emotions of individuals with that disorder; (4) as a predictive system, in which classificatory concepts allow both professionals and nonprofessionals to know which treatments are likely to lead to the best outcomes and to know what is most likely to happen in the course of each disorder if it is untreated; and finally (5) as a basis for a theory (or theories) of psychopathology, which will allow scientists, professionals, and laypeople to understand these disorders. Sadler (2005) has performed a detailed philosophical analysis of the current classifications of psychopathology as they are 53


embodied in the most recent editions of DSM (DSM-IV-TR; American Psychiatric Association [APA], 2000a) and of the International Classification of Diseases (ICD-10; World Health Organization [WHO], 1993). Sadler focuses on the implicit values that have shaped these classifications. In his analysis, he comments repeatedly on an inherent tension within modern psychiatric classifications between what he labels “pragmatic” values (i.e., values associated with clinical practice) and “scientific” values (i.e., values associated with research). Pragmatic values include such things as the utility of classificatory concepts both for making treatment decisions and for communicating to patients and the public; the acceptance of these concepts across languages and cultures; and the ease with which they fit the way in which humans view psychopathology. Scientific values, in contrast, are more epistemic. Included within scientific values are such issues as diagnostic reliability; the level of empirical support for classificatory concepts (validity); and the relationship between classificatory concepts and the development of scientific theories about the phenomena being classified. This chapter is a short overview of the history of changes in classification systems during the middle and late 20th century. As we present this overview, the tension between clinical/pragmatic values and scientific values is a repeated theme. However, the reader will also note, as Sadler (2005) does, that other value systems (e.g., economic and political values) have also had an impact on these changes in classification.

A Brief History of the ICDs In the medical field, the official classification of medical disorders is currently known as the International Statistical Classification of Diseases and Related Health Problems— or, more briefly, as the International Classification of Diseases or ICD (WHO, 1992b). Although we devote most of this chapter to discussing the evolution of the DSMs, we begin with a very brief account of how the ICDs have been developed. This classification system was first adopted internationally in 1900 under the name The Bertillon Classification of Causes of Death, in honor of

Jacques Bertillon, head of the International Statistical Institution at the time of its inception. Between 1909 and 1938, this document went through four additional iterations. After World War II, the WHO met once again to revise the system for the sixth time. It was decided at this time to expand the classification to all diseases, thereby including causes of morbidity as well as mortality. The revised classification became the International Classification of Diseases, Injuries, and Causes of Death (ICD-6; WHO, 1948). With the inclusion of all other diseases came a section entirely devoted to mental disorders. The ICD-6 section on mental disorders was not widely accepted, however; instead, most countries, including the United States, had their own unique classification systems (Stengel, 1959). The WHO periodically and systematically attempted to update its classification of all medical diseases. Initially, the goal was to make these revisions every 10 years. However, since medical knowledge has been growing exponentially, this task has overwhelmed the WHO, which faces a number of other and more immediate tasks. Thus the last major revision to the ICDs occurred in the early 1990s and was published as ICD10. The section of ICD-10 devoted to mental disorders was also published separately, in two forms: a green version for use by general clinicians (WHO, 1992a), and a blue version with diagnostic criteria that could be used in research on psychopathology (WHO, 1993). These two different versions reflect the WHO’s attempt to resolve the competing tensions between pragmatic/clinical values and scientific/research values.

DSM-I By the late 1940s, American psychiatry was undergoing change. Before World War II, the American Psychiatric Association (APA), which was the outgrowth of an earlier organization titled the Superintendents of Insane Asylums, had been dominated by psychiatrists who focused on severe mental illnesses and who believed that most mental disorders were biological diseases (Shorter, 1997). During World War II, an innovative, psychoanalytically oriented physician named William Menninger became the head

Themes in the Evolution of the 20th-�Century DSMs

of military psychiatry. Through Menninger, the Army developed a new classification system called Medical 203 (Houts, 2000). Given the concerns that developed from frustration with the American Psychiatric Association’s earlier classification of mental disorders, they formed a Committee on Nomenclature and Statistics. This committee was charged with the task of integrating and revising the existing systems; the goal was to develop a classification system that would encompass all areas of psychological disturbance and would address problems of the emerging outpatient population. The resulting document, DSM-I, was published by the American Psychiatric Association in 1952. According to Houts (2000), DSM-I was quite similar structurally to the Medical 203 system created by Menninger and his colleagues. DSM-I contained prose descriptions of mental disorders, and the disorders therein were organized into two major categories: disorders with organic causes and disorders of psychogenic origin. The latter disorders were labeled “reactions,” reflecting the contemporary thought in psychiatry that mental disorders were the results of interactions among biological, psychological, and environmental forces. Three features of this earliest DSM are worth noting. The first is that the major goal of DSM-I was to create a classification system that all American psychiatrists could accept and use. Thus this system was not intended to be a revolutionary document that would alter thinking about mental disorders; instead, the values governing its development were pragmatic. The classification was written for psychiatrists and was intended to be used only by psychiatrists. Second, DSM-I essentially organized mental disorders into three broad groups: the psychoses (many of which were then placed into categories based on etiology); the neuroses (i.e., disorders that could be treated by psychoanalysts); and the character disorders (i.e., psychological disorders that were largely untreatable and often involved forensic decisions). Third, the DSM-I prose definitions of categories were intentionally vague. These definitions were added to increase the clinical utility of this classification, but they were intended for use as diagnostic guidelines, rather than as strict rules that could be used to differentiate between separate disorders.


From a political perspective, the development of DSM-I was stimulated by a group of psychiatrists who were distressed with the status quo in American psychiatry after World War II. At that time, psychiatry was becoming increasingly psychoanalytic in its focus (Grob, 1991; Shorter, 1997). However, the American Psychiatric Association still tended to be dominated by old-line psychiatrists working in state hospitals, who focused on the psychoses and tried to find biologically based treatments for their patients. William Menninger, who had helped develop and write Medical 203, became politically active within psychiatry and gained recognition in a collection of “young Turks” named the Group for the Advancement of Psychiatry (GAP) (Grob, 1991). The GAP pushed for the creation of DSM-I and influenced its clinical/ pragmatic approach to classification, which was consistent with the social-Â�psychiatric focus of its psychoanalytic members.

DSM-II Although DSM-I gave American psychiatry a common nomenclature, international communication about mental disorders was still a veritable Tower of Babel. To better define the problem at hand, the WHO recruited a British psychiatrist named Erwin Stengel to conduct an overview of the different classification systems in use throughout the world. Stengel’s (1959) review was a comprehensive assessment of the variations in the terminology used in each system. His conclusions were that many diagnostic concepts contained their own etiological implications, and that the terms used represented different theories regarding psychopathology (e.g., psychoanalytic, biological, etc.), thus drawing lines among theoretical camps. Stengel’s observations led to an eighth revision of the mental disorders section of ICD (WHO, 1967); however, his suggestion of operational definitions was apparently ignored. The American Psychiatric Association was involved in this revision and subsequently published its own version of ICD-8 known as DSM-II (APA, 1968). DSM-II did not differ greatly from DSMI. The most significant difference was the removal of the term “reaction” from most of the diagnoses. This change could be viewed


as an attempt to define mental disorders as actual entities—in other words, a return to a more biological theoretical stance. However, those involved in the revision process denied that this was the case, offering the explanation that the removal of the term was an effort to rid the classification of theoretical biases toward etiological explanations for disorders that did not have known causes (Spitzer & Wilson, 1968). Although the authors of DSM-II asserted that the revised terms were theory-Â�neutral, the use of names like “schizophrenia” rather than “schizophrenic reaction” did seem to reify the disorders in accordance with a disease model (Millon, 1986). The main goals of DSM-II were primarily sociopolitical in nature. DSM-II was an attempt by American psychiatry to unite with psychiatrists around the world in agreeing upon a nomenclature that all psychiatrists would use. Thus DSM-II and the mental disorders section of ICD-8 were virtually the same. However, differences did exist. ICD-8 was strictly a nomenclature (i.e., simply a list of names of acceptable diagnoses, with no definitions for those names). Physicians using ICD-8 were expected to know the meanings of diagnoses within their specialty as a result of their medical training. The authors of DSM-II, in order to make this system more clinically useful, did offer short, broadly worded definitions of its categories. Although the goal of DSM-II’s developers was to create a document that would be accepted by consensus, both in the United States and around the world, its publication was not without controversy. For instance, William Menninger’s brother, Karl Menninger, thought that the deletion of the term “reaction” was a major mistake, and he objected loudly by labeling DSM-II “sheer verbal Mickey Mouse” (Menninger, 1969). More importantly, the gay and lesbian communities united around their dissatisfaction with the inclusion of hom*osexuality as a mental disorder in the classification. In 1971, gay activists demanding a change created major disruptions at the annual convention of the American Psychiatric Association (Kutchins & Kirk, 1997). A young associate member of the DSM-II committee named Robert Spitzer was instrumental in suggesting a change that could be acceptable to both organized psychiatry and to the gay and les-

bian communities. Bayer (1987) provides a detailed historical analysis of the controversies within the American Psychiatric Association about the status of hom*osexuality as a diagnosis. Another political development stimulated by the publication of DSM-II was concern about the vagueness of the definitions of diagnostic categories within this system. A few years earlier, a philosopher named Carl Hempel (1965) had presented an influential paper at a meeting of research-Â�oriented psychiatrists, in which he argued that the science of the field could not improve until psychiatry had better and more specific definitions of its central categories (Kendell, 1975; Sadler, 2005). Hempel’s criticism fit with a series of research studies published in the 1950s and 1960s expressing concern with the poor reliability of psychiatrists and psychologists when making diagnoses (for overviews of this literature, see Blashfield, 1984; Kreitman, 1961; Matarazzo, 1978). Two additional publications served as disturbing demonstrations of this concern. One was a controversial study by David Rosenhan (1973) published in the prestigious journal Science. Rosenhan and his colleagues went to inpatient facilities requesting admission; all reported a vague symptom suggesting a possible auditory hallucination. Otherwise, all personal information they reported was accurate. All were admitted, and all but one were diagnosed as having schizophrenia. This study led to a major outcry in both the professional and lay public literatures about the validity of psychiatric diagnoses as represented in DSM-II (Farber, 1975; Rosenhan, 1975; Spitzer, 1975). Interestingly, the controversy centering around this paper has continued into the 21st century (Slater, 2004). The other important study was a collaborative paper in which British and American psychiatrists diagnosed patients based on videotaped interviews (Kendell et al., 1971). British psychiatrists used a range of diagnoses for these patients. American psychiatrists, however, persistently diagnosed all of the patients as having schizophrenia. American psychiatrists were embarrassed by this demonstration of their diagnostic sloppiness. The development of the first two editions of DSM provided American psychiatry with the common nomenclature that it so desperately needed; however, in many ways,

Themes in the Evolution of the 20th-�Century DSMs

DSM-II offered little more than names. The classification’s value in terms of description was questionable; the concepts were poorly defined and often overlapped. Moreover, the descriptive prose used to define these concepts contained subjective phrases that necessarily led to disagreements among psychiatrists as to the meanings of the concepts, and therefore disagreement as to the appropriate diagnosis for a given patient (Ward, Beck, Mendelson, Mock, & Erbaugh, 1962). Diagnostic unreliability became a prominent criticism of psychiatric diagnosis (Wilson, 1993; Zubin, 1967). However, some authors have argued that the focus on reliability as a necessary area of scientific improvement in DSM-II was actually a smokescreen for political change by biologically oriented psychiatrists attempting to gain control of the classification system (Kirk & Kutchins, 1992). Although the first two DSMs met, to some degree, the sociopolitical need for a unified classification system that would be applicable in outpatient clinical settings, critics of these systems expressed concerns about the social impact of the diagnostic labels provided therein. These critics tended to view mental illness and other forms of deviant behavior as largely politically defined and reinforced by social factors and agencies (Foucault, 1965; Schacht, 1985). Psychiatric diagnoses were considered to be self-Â�fulfilling prophecies, in which patients adopted the behaviors implied by the labels. Another social concern, the negative stigma associated with many psychiatric diagnoses, was also used as a criticism of official diagnostic labels (Matza, 1968). In the Rosenhan (1973) study mentioned above, the individuals who faked the auditory hallucinations and were admitted to psychiatric hospitals reported that their normal behavior was often misconstrued as pathological by the psychiatric staff.

DSM-III In the 1970s, concern over the diagnostic unreliability and conceptual ambiguity of DSM-II led to an increasing interest in developing more reliable diagnostic categories for the purpose of research in the mental health field. At Washington University in St. Louis, a group of psychiatrists took on the task of


outlining what they considered to be the necessary steps to creating such a category: (1) description of the symptoms, (2) laboratory studies of the disorder, (3) distinguishing the category from other disorders, (4) follow-up studies of the disorder, and (5) family studies (Robins & Guze, 1970). These researchers emphasized that psychiatry is a branch of medicine, and that mental illnesses are discrete disease entities, the etiologies of which can be discovered through rigorous scientific study. This approach to psychopathology was viewed as being similar to Emil Kraepelin’s late-19th-Â�century conceptualization of mental disorders as biological diseases whose underlying essences could be understood through strict observations of the symptoms and course of each disorder. Because of this similarity, the Washington University group and those who ascribed to their ideas became known as the “neoÂ�Kraepelinians” (Blashfield, 1984; Klerman, 1978); however, others have rejected this title (see Andreasen, 2007). Adhering to Robins and Guze’s (1970) proposed paradigm for developing diagnostic categories, the group at Washington University produced a paper in which they outlined diagnostic criteria for 14 mental disorders that they believed had sufficient empirical evidence supporting their scientific validity (Feighner et al., 1972). The Feighner criteria, so named for the first author of this collaborative work, did not set forth diagnostic descriptions in the prose form characteristic of the first two DSMs. Instead, the criteria for a syndrome were listed in terms of specific signs and symptoms. Each of the disorders proposed by Feighner and colleagues (1972) was considered to be a discrete group with set boundaries (i.e., a patient either met criteria for the diagnosis or did not). Feighner et al. recognized that not all symptoms of a disorder will be present in any one individual; therefore, their criteria stipulated a requisite number (e.g., five of nine) as necessary for a given diagnosis. Feighner and colleagues stated that a symptom could be considered present if a patient (1) sought the help of a physician for the symptom, (2) was sufficiently disabled by the symptom that normal functioning was disrupted, (3) took medication for the symptom, or (4) exhibited the symptom to such a degree that the examiner believed it to be clinically sig-


nificant. Three years after the publication of the Feighner criteria, the list of 14 disorders was expanded to 25 in an effort termed the Research Diagnostic Criteria (RDC; Spitzer, Endicott, & Robins, 1975). The Feighner criteria and the RDC served as responses to criticisms of psychiatric diagnoses as unreliable and perhaps even invalid. Robert Spitzer, lead author of the RDC, provided evidence across multiple studies suggesting that for most diagnostic categories the RDC offered better reliability than DSMII did (Meier, 1979). This movement represented a paradigmatic shift from an emphasis on psychoanalytic conceptualization, which had dominated North American psychiatry for the better part of a century, to a medical model of mental illness. Although the practice of psychiatric diagnosis had been the topic of much debate, the neo-Â�Kraepelinians emphasized the importance of a universally accepted classification system as a means for conducting research and gaining knowledge in the area of psychopathology (Klerman, 1978). Having made a name for himself in the politics of the American Psychiatric Association through his work on DSM-II and his contributions to the resolution of the debate over the inclusion of hom*osexuality as a mental disorder, Spitzer was appointed to the position of chair of the Task Force on Nomenclature and Statistics, and was given the onerous task of organizing the third edition of DSM. By taking the lead in the development of DSM-III (APA, 1980), Spitzer was in charge of choosing the members of the task force. Just as a U.S. president chooses a cabinet of like-Â�minded individuals who will support and enact his decisions, the task force that created DSM-III consisted of individuals who adhered to Spitzer’s ideology. Many of these individuals came from the Washington University group or had been trained at Washington University. Thus the task force could be understood as an “invisible college”—a group of individuals at diverse locations who were associated professionally and ideologically (Blashfield, 1984; Price, 1963). This is important to note because the final product, DSM-III, was intended for use by all mental health professionals; yet it was formulated around the views held by a particular group of individuals. One of Spitzer’s goals for DSM-III was that it would be devoid of

the psychoanalytic underpinnings present in the first two editions of the manual (Bayer & Spitzer, 1985). However, by adhering to the neo-Â�Kraepelinian ideals of psychiatry as a branch of medicine, the DSM-III task force was using an implicit theoretical blueprint to construct its classification. DSM-III (APA, 1980) differed from its predecessors on a number of levels. First, unlike DSM-I and DSM-II, DSM-III was an attempt at a strictly scientific classification system based on observable evidence rather than on a consensus of clinicians (Kendler, 1990). In particular, there was little attempt to merge the creation of DSM-III with changes in the mental disorders section of ICD-9. As a result, DSM-III was viewed by some in the international community as a typically American creation, in which U.S. psychiatrists assumed that they knew best and the rest of the world would follow after them (Stone, 1997). Because DSM-III did not focus on consensus, the very process of writing the manual changed dramatically. In creating DSM-I and DSM-II, committees of experts had taken a rational approach to defining categories by basing their decisions on experience, precedent, and the utility of the final product. At that time, the scientific literature on mental disorders was scarce and had little impact on the DSMs. In contrast, the authors of DSMIII, in their effort to create an empirically based classification, attempted to organize the system around the existing literature. However, many of the debates that arose during this endeavor could not be solved by referring to the literature or scientific observation; thus some decisions were necessarily political in nature, reflecting the fact that science is not immune to political processes (Schacht, 1985). The second major departure from DSM-I/ DSM-II was in the actual presentation of the disorders. Instead of prose descriptions of syndromes that made implicit reference to the underlying “essence” of a disorder, DSM-III described disorders in terms of specific diagnostic criteria, following the model created by Feighner and colleagues (1972). This shift was an effort to improve diagnostic reliability, which had been less than optimal in the first two editions of the manual. Although the diagnostic criteria used in DSM-III were more specific than the vague-

Themes in the Evolution of the 20th-�Century DSMs

ly worded prose definitions in DSM-I and DSM-II, these new definitions did not quite provide clear, operational specifications of when a symptom was or was not present. The actual operationalization of DSM-III diagnoses occurred through the development of a cottage industry that appeared in the 1980s and 1990s—the creation and publication of semistructured diagnostic interviews to assess patients (see APA, 2000b). These semistructured interviews were focused primarily on patients’ verbal reports about their symptoms and were based to some extent on observations of patients’ behaviors, eventually becoming the “gold standards” by which research diagnoses were made (see, e.g., Sohler & Bromet, 2003). The interclinician reliability of diagnoses made with these interviews was a vast improvement over diagnoses made using the earlier DSMs. However, semistructured interviews, particularly those focused on individual mental disorders, require special training and are timeÂ�consuming to administer. Therefore, they are rarely used in standard clinical practice. DSM-III’s third distinction was that it was a multiaxial classification (Williams, 1985). The categories in DSM-I and DSM-II were not based on a single organizing principle; instead, the organizing principles varied within parts of these classifications. Thus the organization of organic brain syndromes in the DSM-I was structured according to etiology. In contrast, the psychotic disorders were largely separated by views of the course of those disorders, and the DSM-I organization of mental retardation was a simple nominal scaling of the single dimension of intellectual functioning. The confusion inherent in a system without a consistent organizing principle led the authors of DSM-III to develop a diagnostic system in which patients would be diagnosed on five axes. This allowed clinicians to describe the patient’s presentation in terms of psychopathological symptoms (Axis I), personality style and/or mental retardation (Axis II), relevant medical disorders (Axis III), environmental factors (Axis IV), and overall role impairment (Axis V). Interestingly, although this fiveaxis structure has been maintained across the subsequent editions of DSM, neither clinicians nor researchers seem particularly enamored of it (Fabrega, Ahn, & Mezzich, 1991).


Such drastic changes to the classification of psychopathology did not occur without criticisms. During the period of more than 5 years during which DSM-III was conceived, controversy surfaced over the exclusion of the term “neurosis” in the diagnostic categories. Spitzer and the neo-Â�Kraepelinians recognized the term as stemming from psychoanalytic theory and therefore avoided it. The psychoanalysts lost the battle, and their influence in psychiatry since then has continued to diminish. In an attempt to parry criticisms by psychoanalysts that DSM-III was being created by a group of anti-Â�psychoanalytic, pro-Â�biological-model psychiatrists (which it was), Spitzer strenuously argued that DSM-III was intended to be theory-Â�neutral and should only contain diagnostic terms that could be accepted by a broad spectrum of psychiatrists, regardless of their theoretical orientations (Bayer & Spitzer, 1985).

DSM-III-R DSM-III was truly a revolutionary classification in both its scope and its impact (Stone, 1997). The changes made in DSM-III relative to its predecessors were large. DSM-I, for instance, was created by a committee of 7 people and was 130 pages long (including appendices). DSM-II had 10 members on its committee and was a total of 134 pages long. The DSM-III had a central task force of 19 members, with 13 auxiliary committees ranging in size from 4 to 18 members, and the entire manual was an astonishing 494 pages in length. The number of categories in the three editions increased from 106 in DSM-I to 182 in DSM-II to 265 in DSMIII. The impact of DSM-III was also revolutionary. Within 5 years, it had been published in 16 other languages. Mezzich, Fabrega, Mezzich, and Coffman (1985) showed that DSM-III was used as frequently as, if not more frequently than, ICD-9 in a number of countries around the globe. Also, since the diagnostic criteria for many of the DSM-III categories had been created without any research to test them, a large number of studies appeared in the early 1980s that analyzed the DSM-III diagnostic criteria and made cogent comments on problems


with particular criteria (Skodol & Spitzer, 1987). Only 7 years after the publication of DSM-III, the American Psychiatric Association published a revised version of the manual, known as DSM-III-R (APA, 1987). The original goal of this revision was to update the diagnostic criteria by integrating recent research findings, and the changes were intended to be minor. However, the modifications to the manual extended beyond criteria updates. Names of disorders were changed (e.g., paranoid disorder became delusional disorder); specific criteria were altered (e.g., the criteria for histrionic personality disorder); and some categories were reorganized (e.g., panic disorder was explicitly linked to agoraphobia). Six diagnostic categories, including ego-�dystonic hom*osexuality, were dropped from the classification entirely, while approximately 33 new categories were added (e.g., body dysmorphic disorder, trichotillomania, and a whole new set of sleep disorders). Despite the limited goals of DSM-III-R, major controversies developed as this system was being written. This time, the antagonists within the controversy were not psychoanalysts, but feminists, who attacked DSM-IIIR as a male-�generated classification that could have a negative impact on women. The focus of their concerns were three disorders: premenstrual syndrome (PMS), masoch*stic personality disorder, and paraphilic rapism (Blashfield, 1998). The arguments in this controversy were largely based on politics rather than science. The argument against PMS as a diagnostic category was that it blamed female anatomy for emotional states, and that if it were in fact a disorder, it would be more appropriately categorized as a gynecological syndrome than as a psychiatric one (Gallant & Hamilton, 1988; Spitzer, Severino, Williams, & Parry, 1989). masoch*stic personality disorder was attacked as a way for psychiatry to blame victims of abuse for being abused (Caplan, 1985). Paraphilic rapism was thought to provide a psychiatric excuse for the behavior of serial rapists. Spitzer and his colleagues attempted to change the names of PMS and masoch*stic personality disorder to periluteal phase dysphoric disorder and self-�defeating personality disorder, respectively, in order to deflect these attacks. They also proposed the addition of sad*stic personality disorder, to counter

the argument that self-Â�defeating personality disorder would be used to blame the victim. Paraphilic rapism was deleted altogether. The feminists were ingenious in their political attacks. Rather than attempt to persuade particular subcommittees in the DSM-III-R process, the feminists waited until DSMIII-R was being considered by the American Psychiatric Association’s Board of Directors. They had gained allies in this setting, particularly among some individuals who had become alienated by the entire DSM-III-R process. In the end, the American Psychiatric Association’s Board of Trustees came up with a compromise that angered both sides: The three controversial disorders—Â�sad*stic personality disorder, self-Â�defeating personality disorder, and late luteal phase dysphoric disorder—were all moved to a new appendix in DSM-III-R for “disorders needing further study.”

DSM-IV While ICD-10 was still in progress, the American Psychiatric Association decided to revise DSM-III once more, with the intention of coordinating this revision with ICD-10. The result, DSM-IV (APA, 1994), contained 354 categories (as compared to the 297 in DSM-III-R) and was 886 pages long. Not only did the classification provide diagnostic criteria, but supplementary material on the different disorders was also included. For example, in addition to the new criteria for histrionic personality disorder, DSM-IV contained three pages of ancillary information—Â�including diagnostic features; associated features and disorders; specific culture, age, and gender features; prevalence; and differential diagnosis. The organization of the document was the same as that of DSM-III. Some diagnostic criteria were changed or expanded upon, but this iteration was explicitly designed to address the criticism that although the authors of DSM-III had sought to make the classification a purely scientific endeavor, in the end decisions still depended heavily on expert consensus. The revision process was structured in such a way as to limit the use of expert consensus and to maximize the impact of scientific inquiry. However, it was recognized that values are necessarily involved in diagnosis, and the authors of DSM-IV did

Themes in the Evolution of the 20th-�Century DSMs

not overlook the important role of the sociopolitical function of classification. Also of concern was the clinical utility of the classification (Frances, Widiger, & Pincus, 1989). DSM-III was designed to provide a basis for more reliable research, but the DSMs are used in clinical practice and must be serviceable in that sense. The steering committee, comprising 27 members (4 of whom were psychologists), oversaw the progress of 13 work groups, each with 5–16 members and usually at least 20 advisors. The guidelines for how the work groups were to go about their tasks were strictly laid out, in an effort to limit personal biases as much as possible (Blashfield, 1998). Interested parties may also look to the five-Â�volume companion set of sourcebooks (which contain edited papers by members of the work groups) for a scholarly foundation for comprehending the decisions that were made. The work groups approached their tasks in three major steps. First, each group conducted extensive literature reviews of the disorders in their domain. Then the work groups requested access to descriptive data from researchers in the field. These data were reanalyzed by each work group, and the results of their analyses led to decisions regarding which criteria were in need of revision. Finally, a series of field trials were conducted for specific topics, the results of which affected decisions to alter diagnostic criteria. The political process in the development of DSM-IV was quite different from what had occurred with DSM-III and DSM-IIIR. In those earlier editions, Spitzer and the neo-Â�Kraepelinians who worked with him had a firm hand in the work of the individual committees. In DSM-IV, a new psychiatrist, Allen Frances, became the chairperson (Frances, Pincus, Widiger, Davis, & First, 1990). Spitzer became an advisor to the DSM-IV process, but he had relatively little impact on its resulting changes. Frances delegated much more responsibility to the individual work groups that oversaw the details of the changes in DSM-IV. From a political standpoint, this division of labor probably contributed to the acceptance of the resulting classifications, but there were also problems with this process. Dividing psychopathology into discrete segments is not reasonable. Despite this, different work groups were reluctant to give up their rights to a particular domain, even when it might


be better categorized elsewhere. Thus, for example, thoughtful decisions about whether schizotypal disorder should be viewed as a schizophrenia spectrum disorder or whether it should be classified as a personality disorder were hard to make. In the same way, making a decision about whether posttraumatic stress disorder (PTSD) should be classified as an anxiety disorder or as a form of an adjustment disorder (it is classified as the latter in ICD-10) were harder to make in this work-group-based organization. Another concern about the DSM-IV work groups surfaced after the manual was published. Paralleling the changes in the DSMs with the creation of DSM-III was a major political shift within psychiatry away from psychoanalysis and toward a biologically based approach to mental health issues. In particular, the standard form of treatment offered by most psychiatrists by the end of the 1990s was the use of medications to relieve targeted symptoms and syndromes. Implicitly, the drug-�producing companies had large amounts of money invested in treatments for particular mental disorders. Thus changes in the definitions or status of the diagnoses for which drug companies produced targeted drugs were major issues. In 2006, Cosgrove, Krimsky, Vijayaraghavan, and Schneider documented the extent to which members of the various DSM-IV work groups had received financial support from drug companies. According to their analysis, over half (56%) of these members had at least one financial tie to pharmaceutical companies. In fact, every member of the Mood Disorders Work Group and the Schizophrenia and Other Psychotic Disorders Work Group was associated financially with at least one drug company (Cosgrove et al., 2006).

DSM-IV-TR Another issue that generated some controversy regarding the creation of the DSMs, especially after the publication of DSM-III, was the rapidity with which new editions were published. For researchers, particularly those relying on semistructured interviews to make diagnoses, these rapid changes presented serious problems (Zimmerman, 1988). With each new edition of DSM, changes in the diagnostic criteria necessitated changes in the semistructured interviews.


Often these changes, even when they seemed small, could have a substantial impact on decisions about who did or did not meet the definitions for the various mental disorders (Blashfield, Blum, & Pfohl, 1992). From DSM-III to DSM-III-R to DSM-IV, psychiatric diagnoses were talked about as being “moving targets” that impeded careful, systematic research on specific disorders. In response to this criticism, Frances and others involved with DSM-IV promised that the American Psychiatric Association would wait for a longer period of time before attempting to create DSM-V. As this chapter is being written (early 2008), the work groups for DSM-V have been formed and are starting to work. The intent is for DSM-V to appear in May 2012 (APA, 2008). However, like politics, economics play a role in the production of the DSMs. The American Psychiatric Association made substantial amounts of money with the publication of DSM-III and the subsequent editions. All American mental health professionals and mental health settings must have copies of this classification, as must all major college and university libraries. The DSMs have represented a major product line for the American Psychiatric Association that this organization cannot afford to ignore. DSM-IV-TR was different from the other editions in that it included no major, substantive changes from DSM-IV, in terms of either the categories or their definitions. Because the time between DSM-III-R and DSM-IV was so short, each work group primarily focused on revising diagnostic criteria and dealing with any major controversies of relevance to the diagnoses within its purview (e.g., whether self-Â�defeating personality disorder should be moved from the appendix to the personality disorders section). Work on the supporting textual material describing the various mental disorders was not completed until later. These changes were incorporated into DSM-IV-TR. In many ways, DSM-IV-TR does not deserve its designation as a separate edition; however, that designation did lead to substantial new sales of DSM.

above. One goal for DSM-V is to expand the research base for the diagnoses, and in particular to develop a priori classificatory questions that can be researched, rather than having to answer post hoc classificatory questions with existing data sets. To this end, the American Psychiatric Association published A Research Agenda for DSM-V (Kupfer, First, & Regier, 2002), which included a set of white papers outlining research plans for neuroscience; developmental science; the personality and relational disorders; mental disorders and disability; and culture and psychiatric diagnosis. This volume was followed by the more theoretical Advancing DSM: Dilemmas in Psychiatric Diagnosis (Phillips, First, & Pincus, 2003), which addressed various issues: disorder versus nondisorder; the DSM diagnostic groupings; laboratory testing and neuroimaging; schizotaxia and schizophrenia; subthreshold mental disorders; multiaxial assessment; personality disorders; and relationship disorders. Even the foci of these chapters highlight the values of DSM-V: neuroscience, personality disorders, and the underlying mechanisms of disorders. From 2004 to 2007, invited conferences were held to discuss the current research for individual groups of disorders (e.g., mood disorders, personality disorders, and anxiety disorders). In April 2006, David Kupfer and Darrel Regier were appointed as Chair and Vice-Chair of the DSM-V task force; in July 2007, DSM-V work group chairs were appointed. Work group members were appointed in April 2008. In light of the steadily increased focus on the scientific values of classification across the DSMs as documented above, we would expect that scientific issues will remain the foci of the DSM-V process. First and colleagues (2004) noted that a major change in the DSM-V should be an increased emphasis on clinical utility. To promote this more pragmatic focus, we comment on two types of research studies that might be considered in future efforts to improve the clinical utility of the DSMs.

DSM-V and Beyond

Making the Diagnostic and Statistical Manual of Mental Disorders useful to clinicians has been designated as an important goal by the American Psychiatric Association. The

The fifth edition of the DSM is currently expected to be published in May 2012, as noted

Clinical Utility

Themes in the Evolution of the 20th-�Century DSMs

introduction to DSM-IV-TR (APA, 2000) indicates, “Our highest priority has been to provide a helpful guide to clinical practice” (p. xxiii). In addition, in one chapter of A Research Agenda for DSM-V, the joint American Psychiatric Association–Â�National Institute of Mental Health committee charged with outlining a research agenda for DSM-V suggested that examining clinicians’ conceptualizations of DSM-IV disorders is important for creating a reliable and valid DSM (Rounsaville et al., 2002). Furthermore, the American Psychiatric Association’s ad hoc committee on psychiatric diagnosis and assessment has called for studies investigating DSM’s clinical utility (First et al., 2004). Defining clinical utility, developing its theoretical basis, and designing a research agenda to investigate it, however, have proven difficult. First and colleagues (2004) have taken the lead in defining clinical utility as “the extent to which the DSM assists clinical decision makers in fulfilling the various clinical functions of a psychiatric classification system” (p.€ 947), and have defined these functions as including conceptualizing diagnostic entities, communicating clinical information, using diagnostic categories and criteria sets in clinical practice, choosing effective interventions to improve clinical outcomes, and predicting future clinical management needs. First and colleagues have also outlined how changes to clinical utility can be assessed: “Improvements to clinical utility can be measured in terms of 1) their impact on the use of DSM, 2) their enhancement of clinical decision making, and 3) whether they lead to improvement in clinical outcomes” (2004, p.€953). One Method of Developing Utility A few researchers have been applying principles from cognitive psychology to understand how clinicians use diagnostic concepts when discussing patients. For instance, we have been involved in a set of studies looking at the discrepancies between the ways in which clinicians think about mental disorders and the actual structure of the DSM (Flanagan & Blashfield, 2006, 2007; Flanagan, Keeley, & Blashfield, 2008). In this research, 76 psychologists and psychiatrists were asked to sort 67 DSMIV mental disorders into groups of “similar diagnoses” or diagnoses that had “similar


treatments.” The exact disorders used in this research are listed in Table 2.1. Then these clinicians were asked to indicate relationships among the groups of diagnoses by either making progressively larger and then smaller groups of disorders (i.e., creating hierarchical “nodes”), or placing the groups of diagnoses next to each other in two-Â�dimensional space. The amount of distance between each pair of diagnoses was then represented by the hierarchical node at which two diagnoses were first put in the same group, or the number of inches between each pair of diagnoses. These distances were then assembled into a dissimilarity matrix and submitted to two types of hierarchical agglomerative data analysis: single-Â�linkage cluster analysis and Ward’s method of cluster analysis. The goal of this research was to compare the similarities between clinicians’ sortings of disorders and other published suggested changes to DSM. In the absence of a more rigorous empirical investigation of clinical utility, this research could show the similarities among the proposed suggestions and the way clinicians think about mental disorders. The Ward’s method analysis of clinicians’ sortings of DSM disorders appears in Table 2.1. In this table, there are three primary groupings: The first group has a strong medical component; the second group includes the more traditional serious mental illnesses; and the third group includes the more commonly diagnosed disorders in psychiatry. What is most interesting about this organization is that these groupings are more similar to the structure of DSM-I (APA, 1952) or DSM-II (APA, 1968) than they are to that of more recent versions of DSM. The older versions of DSM were organized by simulating a clinician’s decision-Â�making process. First the organic disorders were divided from the nonorganic disorders; then the nonorganic disorders were divided into the psychotic disorders versus the neuroses; and then the personality disorders were separated from the neuroses. Similarly, when making a diagnostic decision, a clinician would first decide whether a disorder was organically based, then whether it had a psychotic component, and then whether the disorder had a neurotic or a personality basis. Notice that the clinicians’ view of mental disorders does not include the Axis I–Axis II distinction or the separation of adult from childhood dis-

64 HISTORICAL AND CULTURAL PERSPECTIVES TABLE 2.1.╇Clinicians’ Groupings of DSM-IV Disorders According to Ward’s Method I. Organic mental disorders A. Disorders often seen in children 1. Encopresis 2. Enuresis 3. Sleepwalking disorder 4. Sleep terror disorder 5. Nightmare disorder B. Medically related disorders 1. Substance-induced sleep disorder 2. Psychotic disorder due to a medical condition 3. Sleep disorder due to a medical condition 4. Circadian rhythm sleep disorder 5. Primary insomnia 6. Primary hypersomnia C. Neurological disorders 1. Delirium 2. Dementia II. Serious mental illnesses A. Substance use disorders 1. Substance dependence 2. Substance abuse 3. Substance-induced disorder 4. Substance-induced psychotic disorder B. Schizophrenia-related disorders 1. Disorders with psychotic features a. Paranoid b. Schizotypal c. Schizoid 2. Schizophrenic disorders a. Delusion disorder b. Brief psychotic disorder c. Schizophrenia d. Schizoaffective disorder e. Shared psychotic disorder C. Mood disorders 1. Cyclothymia 2. Dysthymia 3. Major depressive disorder 4. Depression or mania with psychotic features 5. Bipolar I 6. Bipolar II III. Commonly treated psychiatric disorders A. Externalizing disorders 1. Childhood forms a. Tourette’s disorder or other tic disorders b. Attention-deficit/hyperactivity disorder c. Oppositional defiant disorder d. Conduct disorder 2. Adult forms a. Intermittent explosive disorder b. Pathological gambling c. Pyromania d. Kleptomania e. Antisocial personality disorder


Themes in the Evolution of the 20th-�Century DSMs


TABLE 2.1.╇ (cont.) B. Internalizing disorders 1. Personality-related disorders a. Altered-mental-state disorders (1) Dissociative amnesia (2) Amnestic disorders (3) Depersonalization disorder (4) Dissociative identity disorder b. Personality disorders (1) Avoidant personality disorder (2) Obsessive–compulsive personality disorder (3) Dependent personality disorder (4) Histrionic personality disorder (5) Narcissistic personality disorder (6) Borderline personality disorder 2. Anxiety-related disorders a. Body-focused disorders (1) Bulimia nervosa (2) Anorexia nervosa (3) Focus on fear of disease (hypochondriasis, body dysmorphic disorder) (4) Focus on physical symptoms (somatization disorder, conversion disorder, pain disorder) b. Anxiety disorders (1) Trichotillomania (2) Obsessive–compulsive disorder (3) Generalized anxiety disorder (4) Phobias (5) Panic disorder (6) Posttraumatic stress disorder (7) Separation anxiety disorder Note. Gender identity disorder, paraphilias, sexual dysfunctions, and adjustment disorder were removed from this cluster analysis, since they were deemed to be outliers by single-linkage cluster analysis. From Flanagan, Keeley, and Blashfield (2008, p.€695). Copyright 2008 by the American Psychological Association. Adapted by permission.

orders, both of which are part of the hierarchical structure of DSM-IV. Also of note in the clinicians’ sortings is that at the lower levels of the hierarchy, the DSM groupings were preserved. For instance, the DSM-IV mood disorders were grouped together (as were the dissociative disorders, the anxiety disorders, and the psychotic disorders). However, there were interesting exceptions to this trend. Antisocial personality disorder was grouped with other acting-out disorders rather than with the Cluster B personality disorders. Separation anxiety disorder was included with the other anxiety disorders rather than with the childhood disorders. The Cluster A personality disorders (i.e., paranoid, schizoid, schizotypal) were grouped with the psychotic disorders rather than with the other personality disorders, and the DSM-IV eating disorders and somatoform disorders were grouped together into a “body-Â�focused disorders” group.

These groupings are at variance with suggestions in the current literature for reorganizing the mental disorders for DSM-V. For instance, based on genetic and structural/ comorbidity data, Watson (2005) suggests reorganizing the DSM-IV mood and anxiety disorders into three groups: “bipolar disorders,” “distress disorders,” and “fear disorders.” In that regrouping, major depressive disorder, dysthymia, PTSD, and generalized anxiety disorder are classified as distress disorders. However, clinicians in the study mentioned above kept the DSM-IV mood and anxiety disorders in separate groups. These results suggest that a reorganization such as Watson’s would not have high clinical utility. Another suggestion in the current literature is designating an obsessive–Â�compulsive spectrum of disorders (Castle & Phillips, 2006; Hollander, 2006), which would include neurological disorders (e.g., autism), preoccupations with bodily sensations or ap-


pearance (e.g., body dysmorphic disorder), and impulsive disorders (e.g., trichotillomania) in addition to obsessive–Â�compulsive disorder. In the clinicians’ groupings described above, trichotillomania was grouped with obsessive–Â�compulsive disorder, but autism and body dysmorphic disorder were not. These results suggest that an obsessive–Â� compulsive spectrum of disorders would also not have high clinical utility. A further suggestion in the literature is to return to Achenbach’s (1966) original distinction between externalizing and internalizing disorders. Krueger and Markon (2006) used meta-Â�analytic methods to group DSM disorders along this continuum. Clinicians’ grouping of the DSM-IV impulse control disorders, childhood acting-out disorders, and antisocial personality disorder into an externalizing spectrum of disorders is similar to current suggestions. However, the clinicians’ inclusion of the DSM-IV dissociative, personality, somatoform, and eating disorders on the internalizing spectrum does not match current suggestions. Despite this latter finding, these results suggest that organizing at least some of the DSM disorders along an externalizing–Â�internalizing spectrum has the potential to increasing the utility of the DSMs for clinicians. A Second Method of Developing Utility Another innovative and potentially informative approach to a classification of psychopathology is to use patients’ descriptions of their subjective experiences of these disorders to refine or perhaps redefine the diagnostic criteria, so that they more accurately reflect the phenomenology of mental disorders. This change could help clinicians to better understand the true experiences of their patients and refine ways to help them. In order to illustrate our argument, we use borderline personality disorder as an example. There is evidence that the DSM-IV criteria for borderline personality disorder do not match people’s subjective experience of the disorder. Miller (1994) investigated the subjective experiences of 10 individuals who met DSM-III-R diagnostic criteria for this disorder. Participants were given a minimally structured interview in which they were asked to tell the researcher about themselves

and their everyday lives. Through this methodology, Miller was able to obtain a rich understanding of the ways in which each person’s symptoms affected his or her life, and to assess the extent to which these experiences matched the depiction of the disorder in the DSM. Despite the lack of structure in Miller’s interview and the heterogeneity of symptoms of people with borderline personality disorder, this study found remarkable similarities across participant narratives. In DSM-IV-TR, people with borderline personality disorder are described as having an “identity disturbance: markedly and persistently unstable self-image or sense of self” (p.€ 710). In Miller’s (1994) study, however, participants described a cohesive identity and sense of self, although the explicit diagnosis of borderline personality disorder was notably absent from their self-Â�descriptions. Across approximately six 90-minute interviews, participants were able to describe clearly and consistently their preferences and who they were or wanted to be. They made distinctions between those times when they were able to be themselves and those times when they were not, due to trying to please others or to hide their insecurities. They further described themselves as estranged from others and inadequate in relation to social standards. As Miller concluded, “rather than having an impaired sense of self, they seemed to have a sense of themselves as impaired” (p.€1216). In response to these feelings, participants would invoke a variety of strategies for feeling better, which to an observer might look like identity diffusion and impulsivity (e.g., making changes in appearance, adopting different lifestyles). These failed strategies were compounded by a reluctance to reveal themselves, especially to therapists, and involuntarily choosing to be thought of as having a lack of identity rather than to show their flawed identities. Central to their narratives were themes of struggling through life and attempting to assuage feelings of despair. Subsequently, perhaps a more valid diagnostic criterion for borderline personality disorder would be “sense of self as estranged and inadequate” rather than “identity disturbance: markedly and persistently unstable self-image or sense of self.” This contrast suggests that there may be important, accuracy-Â�increasing differences

Themes in the Evolution of the 20th-�Century DSMs

between patients’ subjective experiences and the perceptions of these same experiences by others (i.e., clinicians or researchers). Another of Miller’s (1994) findings relates to conceptions of the emotional lives of people with borderline personality disorder. According to DSM-IV-TR, their emotional experience is characterized by either “chronic feelings of emptiness” or “affective instability due to a marked reactivity in mood (e.g., intense episodic dysphoria, irritability, or anxiety usually lasting a few hours and only rarely more than a few days)” (p.710). Rather than reporting feelings of emptiness or marked changes in mood, participants in Miller’s study described relentless feelings of despair. Each person described an ever-Â�present wish not to be alive and made at least one reference to spending hours in bed crying. Miller further commented that the pain felt by persons with this disorder was hard to capture in words: “the sense of emotional pain conveyed by these patients was overwhelming .€.€. much of the impact [of their words] may be lost without hearing the pain in their voices or experiencing the redundancy of such comments in the narratives” (p.€ 1217). This study suggests that what may appear to an observer as “emptiness” or “affective instability” may be experienced subjectively by a person with the disorder as chronic emotional pain and despair. These examples contrasting the diagnostic criteria of borderline personality disorder with the subjective experiences of people who have this disorder suggest that in order to validly represent mental disorders, and to help clinicians diagnose and treat people with mental disorders, DSM must accurately reflect the inner experiences of people with mental disorders and offer clinicians a window into these experiences. Paradoxically, then, using patient-Â�subjective based criteria could also address our other primary concern for DSM-V—increasing the clinical utility of the manual. An important, although perhaps unintended, effect of including patient-Â�subjective criteria could be an increase in empathic “bridges” between clinicians and their patients. If clinicians can arrive at a more accurate understanding of patients’ own experiences of mental disorders, they may be more likely to achieve the degree of empathic reflection that has


been found to be one of the most important factors predicting positive treatment relationships and outcomes (Gehrs & Goering, 1994; Horvath & Symonds, 1991; Neale & Rosenheck, 1995).

Postscript We have begun this chapter by discussing the early ICDs and DSMs, along with the clinical/pragmatic goals that dominated these early classifications. As we have moved through discussions of more recent editions of DSM, the story has shifted to discussions of empirical research, political struggles, economics, and changes occurring both within and between the mental health professions. The three of us who are writing this chapter are clinical psychologists. With different levels of involvement, all three of us have spent considerable portions of our time functioning as clinicians—a role in which our primary responsibility is the care of our clients. Although we have all devoted considerable energy to looking at the classificatory issues in psychopathology, none of us as clinicians are happy with the current, official DSM-IV(-TR) system. To us as clinicians, the DSMs seem mostly irrelevant to what we do. In our various clinical roles, we are required to make diagnoses, and we do so. But we rarely find that process informative. Our dissatisfaction with what the DSMs have evolved to become can be summarized by referring to a famous 1960s movie titled Who’s Afraid of Virginia Woolf? This movie is about one night in the lives of two couples who meet and interact with each other. The emotional intensity of Who’s Afraid of Virginia Woolf? is almost unequaled by any other American-made movie. The level of anger, interpersonal destructiveness, and relentless psychological assault is hard to explain. An important psychological puzzle in this movie is how the main characters in the movie can maintain some of level of homeostasis in their dance of anger with each other, so that the interactive pattern continues on and on. The Wedding and colleagues (2005) book mentioned at the start of this chapter, about the representation of psychopathology in


movies, discusses Who’s Afraid of Virginia Woolf? as depicting “other psychotic disorders.” The reason for this diagnostic assignment is that the main couple in the movie (played by Elizabeth Taylor and Richard Burton who, in their nonacting lives, were married twice to each other and then divorced twice) appear to hold a joint delusional belief in the existence of a son who was never born. Although this diagnosis may be defensible, this diagnosis (or any other DSM label) seems irrelevant to the profound psychological issues represented in the interactive pattern among the four characters in the movie. This short historical overview of the DSMs has begun with a comment on the goals of psychiatric classification. We have listed five purposes of such classification: (1) nomenclature, (2) information retrieval, (3) description, (4) prediction, and (5) concept formation for a theory (or theories) of psychopathology. Forty years ago, when one of us (Roger K. Blashfield) wrote about these purposes, he viewed them as being hierarchical (i.e., lower-order purposes had to be met in order for the higher-order purposes to be met). His opinion at the time was that DSM-II functioned satisfactorily as a nomenclature and as a basis for information retrieval. The major scientific problem of the time was to create a classification that was descriptively useful. Now our opinion has changed. The DSMs have failed to meet any of these goals, starting with the most fundamental of being a nomenclature. To understand this claim, think again of Who’s Afraid of Virginia Woolf? and the multiaxial diagnoses for the central characters in this movie. If you have seen this film, are those what you remember about it—the characters’ diagnoses? Probably not. In the same way, what is striking about a movie like Ordinary People is that the clinician in the movie spends little or no time trying to come up with a DSM diagnosis. And even if the clinician were to decide on a diagnosis, we don’t think that such a diagnosis would have much (if any) influence on the process of treatment depicted in that movie, or on its resolution as shown in the last therapy scene of the movie. Psychiatric diagnoses have become institutionalized, and they serve as crude, but somewhat useful, leads for scientific litera-

tures. However, diagnostic concepts trivialize the complexity of people, their lives, and what might help them—just as movies like Who’s Afraid of Virginia Woolf? and Ordinary People are trivialized when thought of simply in terms of the diagnostic concepts they potentially represent. Psychopathology and its manifestations in people are not trivial matters. People with psychopathology and the clinicians who treat them deserve better help than the editions of the DSM have offered. References Achenbach, T. M. (1966). The classification of children’s psychiatric symptoms: A factorÂ�analytic study. Psychological Monographs, 80(7, Whole No. 615). American Psychiatric Association (APA). (1952). Diagnostic and statistical manual of mental disorders. Washington, DC: Author. American Psychiatric Association (APA). (1968). Diagnostic and statistical manual of mental disorders (2nd ed.). Washington, DC: Author. American Psychiatric Association (APA). (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association (APA). (1987). Diagnostic and statistical manual of mental disorders (3rd ed., revised). Washington, DC: Author. American Psychiatric Association (APA). (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association (APA). (2000a). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. American Psychiatric Association (APA). (2000b). Handbook of psychiatric measures. Washington, DC: Author. American Psychiatric Association (APA). (2008). DSM-V: The future manual. Washington, DC: Author. Retrieved from MainMenu/Research/DSMIV/DSMV.aspx. Andreasen, N. (2007). DSM and the death of phenomenology in America: An example of unintended consequences. Schizophrenia Bulletin, 33, 108–112. Bayer, R. (1987). hom*osexuality and American psychiatry: The politics of diagnosis. Princeton, NJ: Princeton University Press. Bayer, R., & Spitzer, R. L. (1985). Neurosis,

Themes in the Evolution of the 20th-Â�Century DSMs psychodynamics and DSM-III: A history of a controversy. Archives of General Psychiatry, 42, 187–196. Blashfield, R. K. (1984). The classification of psychopathology: Neo-Â�Kraepelinian and quantitative approaches. New York: Plenum Press. Blashfield, R. K. (1998). Diagnostic models and systems. In A. S. Bellack & M. Hersen (Eds.), Comprehensive clinical psychology (pp.€ 57– 80). New York: Pergamon Press. Blashfield, R. K., Blum, N., & Pfohl, B. (1992). The effects of changing Axis II diagnostic criteria. Comprehensive Psychiatry, 33, 245–252. Blashfield, R. K., & Draguns, J. G. (1976). Toward a taxonomy of psychopathology: The purpose of psychiatric classification. British Journal of Psychiatry, 129, 574–583. Caplan, P. J. (1985). The myth of women’s masochism. New York: Dutton. Castle, D. J., & Phillips, K. A. (2006). Obsessive–Â� compulsive spectrum of disorders: A defensible construct? Australian and New Zealand Journal of Psychiatry, 40, 114–120. Cosgrove, L., Krimsky, S., Vijayaraghavan, M., & Schneider, L. (2006). Financial ties between DSM-IV panel members and the pharmaceutical industry. Psychotherapy and Psychosomatics, 75(3), 154–160. Fabrega, H., Ahn, C., & Mezzich, J. (1991). Can multiaxial diagnosis predict future use of psychiatric hospitalization? Hospital and Community Psychiatry, 42, 920–924. Farber, I. E. (1975). Sane and insane: Constructions and misconstructions. Journal of Abnormal Psychology, 84(6), 589–620. Feighner, J. P., Robins, E., Guze, S. B., Woodruff, R. A., Winokur, G., & Munoz, R. (1972). Diagnostic criteria for use in psychiatric research. Archives of General Psychiatry, 26, 57–63. First, M. B., Pincus, H. A., Levine, J. B., Williams, J. B. W., Ustun, B., & Peele, R. (2004). Clinical utility as a criterion for revising psychiatric diagnoses. American Journal of Psychiatry, 161, 946–954. Flanagan, E. H., & Blashfield, R. K. (2006). Do clinicians see Axis I and Axis II as different kinds of disorders? Comprehensive Psychiatry, 47, 496–502. Flanagan, E. H., & Blashfield, R. K. (2007). Clinicians’ folk taxonomies of mental disorders. Journal of Philosophy, Psychiatry, and Psychology, 14, 249–270. Flanagan, E. H., Keeley, J., & Blashfield, R. K. (2008). An alternative hierarchical organiza-

69 tion of the mental disorders of DSM-IV. Journal of Abnormal Psychology, 117, 693–698. Foucault, M. (1965). Madness and civilization. New York: Vintage Books. Frances, A., Pincus, H. A., Widiger, T. A., Davis, W. W., & First, M. B. (1990). DSM-IV: Work in progress. American Journal of Psychiatry, 147, 1439–1448. Frances, A., Widiger, T. A., & Pincus, H. A. (1989). The development of the DSM-IV. Archives of General Psychiatry, 46, 373–375. Gallant, S. J., & Hamilton, J. A. (1988). On a premenstrual psychiatric diagnosis: What’s in a name? Professional Psychology, 19, 271–278. Gehrs, M., & Goering, P. (1994). The relationship between the working alliance and rehabilitation outcomes of schizophrenia. Psychosocial Rehabilitation Journal, 18, 43–54. Grob, G. N. (1991). Origins of DSM-I: A study in appearance and reality. American Journal of Psychiatry, 148, 421–431. Hempel, C. G. (1965). Aspects of scientific explanation and other essays in the philosophy of science. New York: Free Press. Hollander, E. (2006). Obsessive–Â�compulsive disorder and spectrum across the life span. International Journal of Psychiatry in Clinical Practice, 9, 79–86. Horvath, A. O., & Symonds, D. (1991). Relation between working alliance and outcome in psychotherapy. Journal of Counseling Psychology, 38, 139–149. Houts, A. C. (2000). Fifty years of psychiatric nomenclature. Journal of Clinical Psychology, 56, 935–967. Kendell, R. E. (1975). The role of diagnosis in psychiatry. Oxford, UK: Blackwell Scientific. Kendell, R. E., Cooper, J. E., Gourlay, A. J., Copeland, J. R. M., Sharpe, L., & Gurland, B. J. (1971). Diagnostic criteria of American and British psychiatrists. Archives of General Psychiatry, 25, 121–130. Kendler, K. S. (1990). Toward a scientific psychiatric nosology: Strengths and limitations. Archives of General Psychiatry, 47, 969–973. Kirk, S. A., & Kutchins, H. (1992). The selling of DSM. New York: Aldine de Gruyter. Klerman, G. L. (1978). The evolution of a scientific nosology. In J. C. Shershow (Ed.), Schizophrenia: Science and practice. Cambridge, MA: Harvard University Press. Kreitman, N. (1961). The reliability of psychiatric diagnosis. Journal of Mental Science, 107, 878–886. Krueger, R. F., & Markon, K. E. (2006). Under-

70 HISTORICAL AND CULTURAL PERSPECTIVES standing psychopathology: Melding behavior genetics, personality, and quantitative psychology to develop an empirically-based model. Current Directions in Psychological Science, 15, 113–117. Kupfer, D. J., First, M. B., & Regier, D. A. (Eds.). (2002). A research agenda for DSM-V. Washington, DC: American Psychiatric Association. Kutchins, H., & Kirk, S. A. (1997). Making us crazy. New York: Free Press. Matarazzo, J. D. (1978). The interview: Its reliability and validity in psychiatric diagnosis. In G. G. Wolman (Ed.), Clinical diagnosis of mental disorders. New York: Plenum Press. Matza, D. (1968). On becoming deviant. Englewood Cliffs, NJ: Prentice-Hall. Meier, A. (1979). The Research Diagnostic Criteria: Historical background, development, validity, and reliability. Canadian Journal of Psychiatry, 24, 167–178. Menninger, K. (1969). Sheer verbal Mickey Mouse. International Journal of Psychiatry, 7, 415. Mezzich, J. E., Fabrega, H., Mezzich, A. C., & Coffman, G. A. (1985). International experience with the DSM-III. Journal of Nervous and Mental Disease, 173, 738–741. Miller, S. G. (1994). Borderline personality disorder from the patient’s perspective. Hospital and Community Psychiatry, 45, 1215–1219. Millon, T. (1986). On the past and future of the DSM-III: Personal recollections and projections. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV (pp.€29–70). New York: Guilford Press. Neale, M. S., & Rosenheck, R. A. (1995). Therapeutic alliance and outcome in a VA intensive case management program. Psychiatric Services, 46, 719–721. Price, D. S. (1963). Little science, big science. New York: Cambridge University Press. Phillips, K. A., First, M. B., & Pincus, H. A. (Eds.). (2003). Advancing DSM: Dilemmas in psychiatric diagnosis. Washington, DC: American Psychiatric Association. Robins, E., & Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric illness: Its application to schizophrenia. American Journal of Psychiatry, 126, 107–111. Rosenhan, D. L. (1973). On being sane in insane places. Science, 179, 250–258. Rosenhan, D. L. (1975). The contextual nature of psychiatric diagnosis. Journal of Abnormal Psychology, 84(5), 462–474.

Rounsaville, B. J., Alarcon, R. D., Andrews, G., Jackson, J. S., Kendell, R. E., & Kendler, K. S. (2002). Basic nomenclature issues for DSMV. In D. J. Kupfer, M. B. First, & D. A. Regier (Eds.), A research agenda for DSM-V. Washington, DC: American Psychiatric Association. Sadler, J. Z. (2005). Values and psychiatric diagnosis. New York: Oxford University Press. Schacht, T. E. (1985). DSM-III and the politics of truth. American Psychologist, 40, 513–521. Shorter, E. (1997). A history of psychiatry. New York: Wiley. Skodol, A. E., & Spitzer, R. L. (1987). An annotated bibliography of DSM-III. Washington, DC: American Psychiatric Press. Slater, L. (2004). Opening Skinner’s box. New York: Norton. Sohler, N. L., & Bromet, E. J. (2003). Does racial bias influence psychiatric diagnoses assigned at first hospitalization? Social Psychiatry and Psychiatric Epidemiology, 38, 463–472. Spitzer, R. L. (1975). On pseudoscience in science, logic in remission, and psychiatric diagnosis: A critique of Rosenhan’s “On being sane in insane places.” Journal of Abnormal Psychology, 84(5), 442–452. Spitzer, R. L., Endicott, J., & Robins, E. (1975). Research Diagnostic Criteria (RDC) for a selected group of functional disorders. New York: New York State Psychiatric Institute. Spitzer, R. L., Severino, S., Williams, J. B. W., & Parry, B. L. (1989). Late luteal phase dysphoric disorder and DSM-III-R. American Journal of Psychiatry, 146, 892–897. Spitzer, R. L., & Wilson, P. T. (1968). A guide to the new nomenclature. In Diagnostic and statistical manual of mental disorders (2nd ed., pp.€ 120–134). Washington, DC: American Psychiatric Association. Stengel, E. (1959). Classification of mental disorders. Bulletin of the World Health Organization, 21, 601–663. Stone, M. H. (1997). Healing the mind: A history of psychiatry from antiquity to the present. New York: Norton. Ward, C. H., Beck, A. T., Mendelson, M., Mock, J. E., & Erbaugh, J. K. (1962). The psychiatric nomenclature: Reasons for diagnostic disagreement. Archives of General Psychiatry, 7, 198–205. Watson, D. (2005). Rethinking the mood and anxiety disorders: A quantitative hierarchical model for DSM-V. Journal of Abnormal Psychology, 114(4), 522–536. Wedding, D., Boyd, M. A., & Niemiec, R. (2005).

Themes in the Evolution of the 20th-Â�Century DSMs Movies and mental illness: Using films to understand psychopathology (2nd ed.). Cambridge, MA: Hogrefe & Huber. Williams, J. B. W. (1985). The multiaxial system of DSM-III: Where did it come from and where should it go? I. Its origins and critiques. Archives of General Psychiatry, 42, 175–180. Wilson, M. (1993). DSM-III and transformation of American psychiatry: A history. American Journal of Psychiatry, 150, 399–410. World Health Organization (WHO). (1948). Manual of international statistical classification of diseases, injuries, and causes of death (6th rev.). Geneva: Author. World Health Organization (WHO). (1967). Manual of the international statistical classification of diseases, injuries, and causes of death (8th rev.). Geneva: Author.

71 World Health Organization. (1992a). The ICD10 classification of mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva: Author. World Health Organization (WHO). (1992b). International statistical classification of diseases and related health problems (10th rev.). Geneva: Author. World Health Organization (WHO). (1993). The ICD-10 classification of mental and behavioural disorders: Diagnostic criteria for research. Geneva: Author. Zimmerman, M. (1988). Why are we rushing to publish DSM-IV? Archives of General Psychiatry, 45, 1135–1138. Zubin, J. (1967). Classification of behavior disorders. Annual Review of Psychology, 18, 373–406.

Chapter 3

On the Wisdom of Considering Culture and€Context in Psychopathology Joseph P. Gone Laurence J. Kirmayer

ince the advent of the “new cross-Â�cultural S psychiatry” in the late 1970s (Kirmayer, 2006; Kleinman, 1977; Littlewood, 1990), psychiatric researchers have distinguished between the historical enterprise of exporting conventional categories of “mental disorder” throughout the world’s diverse cultural communities for the purposes of comparative study, and the more recent commitment to examining the cross-Â�cultural viability and coherence of such categories within culturally local frameworks of distress, illness, and dysfunction (Kirmayer, 2007a). Such reflexive awareness concerning the Western cultural foundations of the categories and constructs used in cross-Â�cultural studies of psychopathology parallels a broader conceptual and methodological revolution in the social sciences (Rabinow & Sullivan, 1987). This reflexive stance has encouraged attention to the ways in which science and technical practices are embedded in local and international systems of power and knowledge, and has urged caution in generalizing or applying dominant approaches to disparate cultures and communities. And yet, as helping professions rooted in an understanding of the human condition, 72

psychiatry and psychology aim for theories of psychopathology that can be used across social and cultural contexts. An international diagnostic nosology should provide a common language allowing psychiatrists everywhere to exchange knowledge about specific patients, have ready access to current technical approaches, and contribute to the advance of psychiatric science. Unfortunately, this project of a global scientific psychiatry tends to view culture as a distraction from the project of developing a body of universal knowledge. That is, cultural diversity becomes an obstacle to scientific research and delivery of care, or else a matter of trivial differences—of “window dressing” on the essential core of universal human experience that might ground a universal nosology. This more dismissive view of the relevance of culture for world psychiatry is part of the legacy of European empire, in that it assumes that the pertinent categories, concepts, principles, and practices—Â�constructs that emerged almost exclusively from certain subpopulations or social strata within a handful of European and North American societies—Â�constitute a universal, transcen-

On the Wisdom of Considering Culture and Context

dent, ahistorical, and “culture-free” basis for recognizing “natural kinds” (i.e., the categories that are immediately given to perception or that can be readily discerned by “carving nature at its joints”) within the domain of psychopathology. This ethnocentrism is also evident in the way non-Â�Western cultures are frequently construed by Westerners: There is a tendency to dichotomize self and other, to view the world as “us and them.” We have knowledge, while they have beliefs; we see things as they truly are, while they are deluded by their stubborn traditions and superstitions. The imperialist roots of this thinking are evident in the resultant asymmetries in valuing truth claims: European American epistemological practices yield transcendent technical knowledge, while other epistemological traditions yield mere folk knowledge comprising beliefs rather than truths about the world. In actual fact, however, the concepts and categories of contemporary psychiatry are not transcendent, culture-free outcomes of objective observation and scientific research; instead, they carry forward the legacy of their own cultural histories (Gaines, 1992; Mezzich et al., 1999; Mezzich, Kleinman, Fabrega, & Parron, 1996; Young, 1995). To explore this further, we briefly consider the neo-Â�Kraepelinian nosological project before turning to questions about the place of culture within mainstream scientific work on psychopathology.

The Neo-Â�Kraepelinian Vision and€the Nature of Contemporary Nosology The historically contingent nature of the reigning nosology—and thus its salience as a “cultural artifact” (i.e., a creation or product that emerges within a unique time and place)—remains evident despite its increasing circulation and influence around the world. More specifically, modern versions of the Diagnostic and Statistical Manual of Mental Disorders (DSM; American Psychiatric Association [APA], 1980, 1987, 1994, 2000) are the products of the neoÂ�Kraepelinian movement that emerged in the 1960s at Washington University in St. Louis and that has dominated psychiatry since the 1970s (Wilson, 1993; Woodruff, Goodwin, & Guze, 1974; Young, 1991).


The Neo-Â�Kraepelinian Vision Klerman (1978) outlined several commitments of this neo-Â�Kraepelinian movement, including the conviction that the study of psychopathology and its treatments belongs properly within the field of medicine; that mental illnesses/disorders are discrete entities with etiologies that can be discovered principally within the realm of disordered biology (as opposed to the previously dominant mode of explanation, derived from psychoanalysis, that privileged intrapsychic dynamics); that psychiatric research on psychopathology should depend principally on statistical inference in the context of modern scientific methodology; that the progress of science in the context of understanding psychopathology requires extensive concern with the standardization of diagnostic concepts and categories for implementation in research and treatment settings in reliable and valid ways; and that the relationships among and between discrete psychiatric disorders should be represented in scientifically valid classification schemes, with explicit diagnostic criteria for the disorders so classified (see Blashfield, 1984, for discussion and amplification of these commitments). The neo-Â�Kraepelinian concern with standardized categories of disorder (consisting of detailed and relatively explicit criteria with accompanying decision rules for determining category membership) was central to the revolutionary reformulation of psychiatric nosology codified in DSM-III (APA, 1980) and its descendants. There can be little doubt that cumulative scientific progress in understanding the origins, outcomes, and treatments of psychiatric distress requires some semblance of standardization, in order for independent research findings to coalesce and be built upon in useful ways. In addition, it seems sensible to specify and represent the relationships among and between various kinds of “mental disorders” within some heuristic classification scheme. The challenge is how to select among the almost infinite range of principles that could serve as organizational bases for structuring such a taxonomy (Millon, 1991). Given the fact that psychiatric scientists typically conduct their research within medical contexts, and thus tend to privilege the biological foundations of medi-


cal practice, psychiatric research has emphasized the empirical specification of “underlying pathophysiology” for various conditions and disorders as the sine qua non of the scientific project to advance our understanding of psychopathology. It is important to note, however, that many psychopathologists—Â� especially researchers trained in clinical psychology and the remaining “psy-” disciplines—Â�ardently contest the commitment of the neo-Â�Kraepelinians to biological reductionism, arguing instead that psychopathology may result from emergent psychological or social processes that are not simply reducible to biology (Henningsen & Kirmayer, 2000; Kirmayer & Young, 1999; see also Beutler & Malik, 2002, for recent critiques). Nevertheless, the pursuit of underlying pathophysiology reflects the dominant trend in psychopathology research, as evidenced by the distribution of research funds, journal citations, and professional prestige. Within the dominant frame of contemporary psychiatry, then, psychopathology is concerned with the fundamental biological processes (i.e., “basic” genetic, anatomical, and physiological processes in complex interactions that are “influenced” by the organism’s environment) that go awry in instances of illness or disorder. Such pathophysiology is characterized as “underlying” because the precise means and mechanisms within the brain and body that are presumed to culminate in the reported symptoms or observable signs of psychiatric disorder are elusive and (in almost every instance) unknown. Thus the empirical pursuit of “endophenotypes” and other correlates of underlying pathophysiologies for a wide variety of mental disorders represents the dominant paradigm in the scientific investigation of psychopathology. Finally, within the traditions of scientific medicine, it is presumed that the empirical identification of etiology in the form of distinctive pathophysiology will ultimately define the disorders in question—in much the same fashion that modern medicine currently understands Down’s syndrome, general paresis, or phenylketonuria in terms of their underlying pathophysiologies. If the etiological pursuit of underlying pathophysiology characterizes the dominant paradigm in psychopathology research,

then the obvious implication for developing a classificatory strategy is to organize the nosology in terms of kinds of pathophysiologies. The dilemma, of course, is that psychopathologists have yet to empirically identify any pathognomonic features of a purported mental disorder, much less its definitive pathophysiology. In the interim, the reigning taxonomic strategy depends on grouping taxa by similarities in “phenomenology”1 (in most instances, based on clusters of symptoms at the syndromal level; see DSM-IV-TR [APA, 2000] for discussion). It is important to recognize that no credible psychopathologist—Â�including the psychiatric scientists who developed the various revisions of DSM—would suggest that the disorders currently classified within DSM are validated constructs that warrant much scientific confidence. Nevertheless, the majority of psychopathologists are confident in the validation strategies described by Robins and Guze (1970) and elaborated by Kendell (1989) for empirically evaluating the merits of purported disorders as viable “hypothetical constructs” (Morey, 1991, drawing upon MacCorquodale & Meehl, 1948). These and other closely related strategies constitute normative science within a neo-Â�Kraepelinian psychiatry. In sum, although psychopathology researchers acknowledge that the hypothetical constructs contained in even the most recent version of DSM are “splendid fictions” (Millon, 1991, p.€ 246), most imagine that systematic inquiry within the paradigm just described will one day yield a much less arbitrary nosology that more closely approximates “carving nature at its joints.” In response, we simply observe that the promise of the neo-Â�Kraepelinian pursuit of distinctive pathophysiologies for the wide variety of mental disorders remains a matter of professional faith. The currently authorized nosological categories, consisting of nearly 300 of Millon’s splendid fictions, reflect numerous political, aesthetic, and pragmatic commitments that yield abundant evidence of prescientific or nonscientific arbitrariness— all of which arise as expressions of cultural processes and practices (as indeed does scientific inquiry itself). To illustrate the impact of such processes and practices, we briefly consider the cultural history of one hypothetical construct within the domain of con-

On the Wisdom of Considering Culture and Context

temporary psychopathology—Â�namely, posttraumatic stress disorder (PTSD). The Cultural Construction of PTSD The advent of DSM-III saw a movement to attribute a vast array of problems to trauma exposure, gathered together under the umbrella of PTSD. The ensuing years have seen the expansion of this category, which to some extent has absorbed other conditions formerly linked to adverse life events. Conventional histories of this construct suggest that PTSD has afflicted survivors of psychological trauma for millennia (Herman, 1992; Trimble, 1985), and that it has merely awaited discovery in the reports and behaviors of its sufferers by intrepid psychopathologists. In contrast to these seamless accounts, cultural historians have contended that the current conceptualization of PTSD, as both a category of clinical attention and a kind of crippling experience, is rather newly arrived on the historical stage (Hacking, 1996; Lerner, 2003; Leys, 1996, 2000; Young, 1995, 1996a, 1996b). These cultural analysts argue that PTSD, instead of possessing the timeless universality and intrinsic unity assumed in our “received” notions of the disorder, has only recently been “glued together” (Young, 1995, p.€ 5) from fragmentary shards of theory, politics, and practices spanning more than a century. Analytic attention to cultural processes and practices is therefore relevant not just for charting the varieties of PTSD experience among the world’s diverse peoples, but also for grounding our conceptual understanding of the PTSD construct itself as one increasingly prominent category within the early21st-Â�century classification of psychopathology. The construct of PTSD singles out the health consequences of the adaptive conditioned fear response to life-Â�threatening danger. Although exposure to the threat of violence, pain, and injury readily gives rise to the specific forms of conditioned emotional response and avoidance learning held to underlie PTSD, this captures only a small part the human response to trauma and virtually never exists in isolation (Kirmayer, Lemelson, & Barad, 2007). The same traumatic events that give rise to PTSD have a wide range of other personal and social effects on


biobehavioral systems involving fear, attachment, coherence, hope, identity, and sense of justice (Silove, 1999). The act of singling out a single biobehavioral response as a discrete disorder creates a measure of diagnostic clarity and precision—but, despite a considerable body of research, the extent to which this response should be framed as psychopathology and the degree of correspondence with the actual experience of suffering individuals both remain contested (Konner, 2007). Even a casual inspection of recent versions of DSM indicates some evolution of psychiatric thought relative to PTSD since its initial incorporation into the official nosology of DSM-III in 1980. For example, the Criterion A definition of the qualifying stressor has changed between versions; the number of symptoms required for clinical inference of the disorder has increased; and the duration of symptoms necessary for diagnosis has ultimately been fixed at 1 month. More illuminating still are the questions and controversies that occupied the PTSD sub-Work Group of the DSM-IV Task Force during preparation of DSM-IV (Davidson et al., 1996). These included how narrowly to define the Criterion A stressor; what duration of symptoms to adopt for distinguishing between normative and pathological reactions to trauma; which subtypes and course specifiers to include; what number of avoidance symptoms to require for a diagnosis; whether to classify PTSD as an anxiety disorder or a dissociative disorder, or within a new class of trauma-Â�related disorders; how to make sense of the high rates of comorbidity between PTSD and other mental disorders; and whether to include a new form of pathological posttraumatic response indicated by “extensive characterological changes” (p.€592) attributed to repeated and prolonged trauma. In some instances, the associated empirical investigation known as the PTSD Field Trial (Kilpatrick et al., 1998) obtained data related to these questions (e.g., neither broad nor narrow definitions of the qualifying stressor appeared to significantly alter sample prevalence of the disorder), but in most instances such data, even when obtained, were insufficient to resolve these questions (e.g., reduction of the required avoidance symptoms increased PTSD prevalence, and yet the overall implications of this


increase remained unclear in the face of delimiting sample characteristics). For the purposes of our argument, the most important acknowledgment is that data alone will never be adequate to resolve these and many other similar questions and controversies surrounding PTSD (see McNally, 2003, for an elaboration), as long as independent and reliable measures of distinctive pathophysiology or specific etiology remain unestablished for the disorder. In fact, as the cultural historians suggest, it seems debatable whether meaningful human responses to “traumatic” experience are even of the “natural kind” variety that might be amenable to scientific demonstrations of distinctive pathophysiology or confirmations of specific etiology. As a result, for a hypothesized syndrome currently without any pathognomonic indicators that might unify diverse patient profiles, decisions regarding conceptualization of the disorder at this stage of inquiry are rendered largely by expert consensus (sometimes with recourse to data, but often not). Such consensus is made and unmade in cultural terms, through enduring and recognizable “logics” of expertise, argument, inquiry, and influence. Nowhere was this consummation of expert consensus more evident than in the historical events surrounding the initial inclusion of PTSD within DSM-III in 1980. According to Scott (1990), PTSD was born of an unusual political alliance between psychiatrists such as Robert Lifton and Chaim Shatan and activists affiliated with Vietnam Veterans Against the War (VVAW) beginning in the late 1960s. This alliance’s escalating campaign for medical recognition of “post-Â�Vietnam syndrome” found footing in a 1975 meeting with psychiatrist Robert Spitzer, the architect of DSM-III, at the Anaheim convention of the American Psychiatric Association. As Scott recounts, Spitzer there dismissed the alliance’s proposal for inclusion of the new syndrome, explaining that psychiatric researchers John Helzer and Lee Robins at Washington University had demonstrated with their data that the problems of returning Vietnam veterans were already subsumed under existing disorders (major depression, substance use disorder, etc.). Spitzer challenged the alliance to provide contradictory evidence. Later that year, following additional lobbying, Spitzer agreed

to form a task force on the issue and invited alliance members Lifton, Shatan, and Jack Smith (a VVAW activist) to join himself and two other psychiatrists on the official Advisory Committee on Reactive Disorders. With Spitzer’s attention frequently drawn elsewhere, according to Scott, the alliance members reasoned that he could be most effectively persuaded to include the syndrome if fellow committee member Nancy Andreasen, then a specialist in treating patients with burns, was first to be convinced of the merits of their cause. Despite ongoing opposition from the Washington University researchers, Andreasen and Spitzer eventually accepted that combat veterans were probably suffering from a distinct psychiatric illness. Given the limited empirical literature available at the time, this recognition depended principally on a series of compelling case studies, presented by outsiders to the psychiatric establishment, including the only member of any DSM-III advisory committee not to have obtained a graduate degree. In sum, this alliance of “radical” psychiatrists and retired soldiers obtained official recognition of PTSD “because they were better organized, more politically active, and enjoyed more lucky breaks than their opposition” (Scott, 1990, p.€ 308). Of course, the implications of this watershed historical moment would be difficult to overemphasize. As Scott has observed, official recognition by the American psychiatric establishment accorded PTSD the status of “objective knowledge,” which in turn undergirds what people experience as “objective” reality: “each new clinical diagnosis of PTSD, each new warrantable medical insurance claim, each new narrative about the disorder reaffirms its reality, its objectivity, its ‘just thereness’â•›” (p.€308). Such reaffirmations of objectivity are possible only if our perspective is fundamentally ahistorical and deeply inattentive to cultural processes and practices, the result of which is the reification of a provisional psychiatric construct for which pathophysiology and etiology remain unknown. And, in an ironic twist, such instances of unwarranted reification actively create culture by virtue of prescribing novel forms of illness experience, even as they disavow the relevance of culture for the nosological project.

On the Wisdom of Considering Culture and Context

Unwarranted reification represents a significant liability for any scientific endeavor, including the empirical opportunities and theoretical possibilities that remain unacknowledged and unexplored, owing to the premature foreclosure of conceptual alternatives. Such conceptual alternatives to contemporary PTSD are again suggested in the work of the cultural historians, who chart the rise of modern notions of “trauma” alongside late-19th-Â�century investigations of hysteria, dissociation, and hypnosis. Ruth Leys (2000) asserts that our conceptualization of trauma and its pathologies continues to vacillate between two historical paradigms: mimetic theory, in which the symptoms of trauma are held to involve a kind of unconscious imitation of the original traumatic event, generally through dissociative mechanisms; and antimimetic theories, in which the posttraumatic symptoms are more or less direct consequences of the violent threat or assault. The dilemma is that mimesis can reflect pathological processes mediated by mechanisms of repression or dissociation in memory, or instead can be created factitiously by similar processes of recollection and recall (Young, 2007). Antimimetic theories circumvent this problem by positing psychophysiological effects of trauma (unmediated by the sufferer’s own agency or unconscious dynamics) on subsequent symptoms. The so-Â�called “memory wars” of the 1990s perhaps most clearly illustrated this paradigmatic tension (Crews, 1995). Both theories are represented within DSM-IV by the inclusion of PTSD (with a decided “antimimetic” deemphasis of attendant dissociative phenomena) and dissociative identity disorder (a mimetic pathology with a clear emphasis on purported traumatic etiology [Gleaves, 1996; Hacking, 1995b]); the former diagnosis has garnered current respectability within scientific psychiatry, while the latter has not. In addition, Allan Young (1996a) observed that since their inception, the posttraumatic pathologies reported by survivors of 19thÂ�century railway accidents were difficult for physicians to differentiate from neurological insult, neurotic disposition (leading to trauma-Â�related “hysteria” and other “functional” disorders), or malingering (in pursuit of monetary damages). Young then raised the uncomfortable question of whether trau-


matic experience in fact caused the PTSD symptoms experienced by the combat veterans in his study, or whether these veterans only later attributed the cause of their long-Â�standing symptoms to previous trauma in post hoc fashion as a direct response to treatment discourse. These concerns continue to trouble the field, insofar as psychopathologists have come to acknowledge the etiological importance of some preexisting “phenotypic expression of vulnerability” in PTSD (Yehuda & McFarlane, 1995) and to “worry” about recent evidence suggesting that a substantial proportion of Vietnam veterans—Â�perhaps as many as 75%, as reviewed by McNally (2003)—have received disability payments for PTSD or have taken part in research studies as “cases” of PTSD, even though they may never have actually experienced combat (Frueh et al., 2005). Even such brief attention to the cultural history of trauma and its pathologies serves to remind us that, far from the timeless universality and intrinsic unity frequently ascribed to the diagnostic entity, PTSD is a construct of rather recent invention. Certainly humans throughout history have responded to extremely distressing events with extreme distress. Such distress was undoubtedly evidenced through posttraumatic changes in individual cognition, emotion, and behavior. In our particular historical context, it would appear that Westerners (and increasingly the rest of the world as well, especially individuals making bids for international asylum to escape war, torture, and oppression) experience such distress in the increasingly popular genre of PTSD (Pole, Gone, & Kulkarni, 2008). The notion of PTSD serves these social and political functions, which in turn reinforce its coherence as a discrete entity, but this coherence is purchased at the expense of attention to a wide range of other individual responses to trauma. For every nightmare, flashback, amnesia, and exaggerated startle response currently assessed in traumatized patients, the avolition, weakness, headache, nausea, giddiness, photophobia, palpitations, paraesthesias, paralyses, double vision, altered posture, unsteady gait, feeble pulse, pressured speech, loss of appetite, and shortness of breath that characterized 19th-Â�century pathological responses to psychological trauma (Kinzie & Goetz, 1996; Young, 1996a) have fallen by


the wayside. And yet many of these symptoms may continue to be prominent features of posttraumatic distress in diverse cultural settings (Kirmayer, 1996). In short, PTSD as currently configured is a malady of our time, emergent from and dependent upon the same cultural processes and practices that actively constitute contemporary life.

ral kinds as opposed to “human kinds”— that is, intentional categories that emerge from our social institutions, knowledge, and practices?

Implications for Cultural Analysis

In 1980, anthropologist and psychiatrist Arthur Kleinman (1986) conducted a landmark study in Hunan, China, that shed light on the universality of categories of common mental disorders. Kleinman studied a group of patients who had received the diagnosis of neurasthenia (shenjing shuairuo)—a syndrome marked by somatic complaints such as headache, fatigue, dizziness, and muscle tension, which was a common form of distress routinely diagnosed by Chinese psychiatrists in clinical settings. Neurasthenia (or “nervous weakness”) was originally described by the American neurologist George Beard in the late 1800s and soon became a common diagnosis worldwide (Beard, 1869). After the 1920s, the popularity of neurasthenia waned in the West as it was gradually replaced by other construals of psychopathology, most recently clinical depression. Nevertheless, neurasthenia persisted as a professional diagnostic label and a mode of illness experience throughout China up to the 1990s (Lee, 1998). Applying the diagnostic criteria from the DSM to 100 Chinese patients diagnosed with neurasthenia, Kleinman (1986) determined that the vast majority of these individuals met criteria for major depressive disorder. When they were treated with tricyclic antidepressants, most patients showed some improvement in their symptoms of depression; however, many continued to see themselves as suffering from neurasthenia, pointing to symptoms of depleted energy and other somatic symptoms or difficulties in their lives, which they attributed to their catastrophic experiences during the Cultural Revolution. Instead of concluding that Chinese neurasthenia was identical to clinical depression, Kleinman argued that neurasthenia and depression were in fact distinct forms of distress that did not always co-occur among Chinese patients. Subsequent work has borne this out (Zheng et al., 1997). Nevertheless, drawing on Eisenberg’s (1977) differentia-

This brief foray into the origins of PTSD as a nosological category and pathological construct raises questions about its claim to be a timeless, culturally universal entity. While scientific methods hold the prospect of refining our knowledge of how the world works, at any point in time scientifically derived knowledge remains an approximation that incorporates culturally and historically contingent features reflecting the origins of our constructs and the contexts of their use (Collins & Pinch, 1993). In the case of psychiatric nosology, we might consider that this cultural and historical embedding is not a defect or limitation of current scientific knowledge, but a necessity, since psychiatric distress, like all human experience, takes shape from cultural particulars. Psychiatric disorders reflect the outcome of interactions between biological processes and a social surround mediated by psychological mechanisms over the developmental trajectory of a human lifespan. The notion that a comprehensive or complete nosology can be created without regard to culture and context, therefore, can be sustained only by adopting a reductionist perspective that minimizes or ignores the fact that human beings are fundamentally social and cultural beings. Nevertheless, such reductionism is frequently embraced and promoted in the name of a scientific psychopathology, based on the assumption that modern psychiatry pursues transcendent understanding of disorders that exist in the world as natural kinds. In other words, the contemporary recognition of flashbacks and amnesia as symptoms of PTSD, as opposed to photophobia and double vision, is justified on the basis that recent systematic investigations have yielded historical progress in our approximation of the natural kind known as PTSD. But what are the conceptual grounds for presuming that PTSD or any other DSM disorders are natu-

Natural Kinds and Intentional Categories

On the Wisdom of Considering Culture and Context

tion between subjective illness experience and objective disease process, Kleinman assumed that both neurasthenic and depressive syndromes were superficially divergent expressions of the same underlying disease: a “universal core depressive disorder” (1986, p.€66). In short, for Kleinman, neurasthenia was a somatized form of an “underlying” depressive disease. As a folk and professional category, and as a cultural “idiom of distress,” neurasthenia had its own sociomoral uses and implications. Kleinman emphasized this sociomoral dimension of experience (though he diplomatically downplayed the continuing role of political repression), but was largely uninterested in neurasthenia as a psychopathological construct for which one might seek to understand underlying mechanisms. In a critique of Kleinman’s report, psychological anthropologist Richard Shweder (1988) noted an unresolved tension in the study’s conclusions between a positivist and a constructivist perspective on the diagnostic problem at hand. According to Shweder, neopositivists remain interested in discovering “natural kinds,” those phenomena that “exhibit a causation independent of what they mean to us, independent of our involvement with them, independent of our experience with them or evaluation of them, independent of our aesthetic or emotional response to them” (p.€488). In contrast, he continued, constructivists remain interested in discovering “intentional categories,” those phenomena that “exhibit whatever causation they may have by virtue of what they mean to us, by virtue of our conceptions and representations of them and reactions to them” (p.€488). Natural kinds thus include such phenomena as trisomy 21 and dopamine. Intentional categories may include such phenomena as psychopathic deviance or la belle indifférence.2 While Kleinman explicitly adopted a constructivist perspective for many of his analyses of Chinese neurasthenia, Shweder worried about Kleinman’s characterization of this syndrome as somatized depression: Could not depression just as easily be construed as a psychologized form of neurasthenia? Shweder wondered what rationale might be offered in support of Kleinman’s clear preference for a depression-Â�centered discourse: “If a disease process is different from an illness experience and if depression


is a disease process (as well as an illness experience), then what precisely is that depressive disease process that is other than an illness experience, and how do we know that neurasthenia is a somatized version of it?” (1988, p.€ 494). Here Shweder laid bare the fundamental problem of cross-Â�cultural analysis in psychopathology research—Â�namely, the challenge of determining how we might reconcile divergent frames of reference, modes of representation, and modalities of experience that give rise to diverse patterns of dysfunctional or disordered experience and expression within and between culturally distinctive communities throughout the world. One implication of the distinction between natural kinds and intentional categories is that while attention to cultural processes and practices throughout the diverse regions of the world may be helpful and illuminating for investigation of disorders of the “natural kinds” variety, such attention is absolutely indispensable for investigation of disorders of the “intentional categories” variety. In other words, if human pathological reactions to traumatic experiences are indeed widely contingent on time, locale, and ethos, then conceptualizing, classifying, investigating, and treating such reactions are heavily dependent on the historically and culturally contingent frameworks of meaning that mediate such pathological experiences. But what are the grounds for imagining that such cultural frameworks of meaning might actually mediate the experience of many forms of psychopathology?

The Depth and Sweep of€Culture: The Case of Emotional Experience A robust cross-Â�cultural psychopathology takes as its point of departure the recognition of the co-Â�constitution of mind and culture. More specifically, cultural psychologists and psychiatrists are concerned with the manner in which human beings—and the cultures they dynamically and interactively construct and reproduce—give rise to “culturally constituted persons” who are both producers and products of the intentional worlds they inhabit (Shweder, 1991). For our purposes, “culture” may be understood as the socially patterned and historically reproduced systems of semiotic practices that both facili-


tate and constrain human meaning making (Geertz, 1973; Gone, Miller, & Rappaport, 1999). Culture is social (and often public) because such systems must be shared; there is no culture of one. Culture is patterned because such systems are organized and utilized systematically in order to be intelligible to others; they are not randomly recreated with each usage. Culture is historically reproduced, in that successive generations are socialized into using the intelligible systems of their communities (which is not to argue that culture is simply “transmitted” from one generation to the next, as innovations and modifications are constantly introduced both in the process of socialization and as subsequent generations adapt to novel circ*mstances). Finally, cultural practices are symbolic, in that they allow for the ascription and communication of meaning or “intelligibility” to others. In other words, culture comprises shared patterns of activity, interaction, and interpretation. Perhaps the most salient example of culture is language, which serves as the primary semiotic system available to human beings for achieving mutual intelligibility, as well as the principal medium of intergenerational cultural reproduction. The study of mind and mentality within enculturated human communities makes it clear that cultural meanings and practices are just as central to realizing personhood as biological mechanisms or processes are. That is, human experience is crafted, constituted, or constructed from the complex and divergent ways in which culture and biology come together to render such experience possible. Thus, obviously, there is no culture without human biology—but, similarly, biology in the absence of culture is neither recognizable nor sustainable as human experience (Kirmayer, 2006; Wexler, 2006). Our point here is simply that, contrary to the evident commitments of the neo-Â�Kraepelinians and the disciplinary traditions of psychiatry and psychology (which routinely refer to culture as rather superficially “shaping” or “influencing” putatively more basic biological processes),3 there is no compelling reason to routinely privilege biology as more fundamental than culture to many of the constructs of interest within psychopathology. One example of the reductionist bias toward psychological processes that are in fact

co-Â�constituted by both biology and culture occurs routinely in the psychological study of emotion, one of the most basic constituents of psychopathological experience. Most forms of psychopathology are accompanied by troubling emotions, and specific kinds of emotional experience provide the phenomenological basis for two of the most prevalent DSM diagnostic classes (at least as surveyed in the affluent West): the “mood” and “anxiety” disorders. As a result, the psychology of emotion figures prominently in the study of psychopathology. The Dominant Approach to€Emotion€Research Oatley and Jenkins (1992) traced the conceptual paradigms that have guided emotion research in the discipline back to Darwin (1872/1965) and James (1890). Whereas Darwin emphasized the biological and evolutionary significance of emotional processes and James emphasized the phenomenology of emotional experience, both writers conceptualized emotions as primarily intrinsic biological or physiological properties of the organism. The Darwinian tradition in particular inspired research by Ekman (1984) into the cross-Â�cultural prevalence of emotion. Drawing on the presumed evolutionary significance of facial expression in the communication of internal emotional states to other members of one’s species, Ekman discovered that respondents from many of the world’s cultures expressed consistent associations of certain facial expressions with comparable emotion terminology, suggesting the universality of at least six basic or core emotions. For Ekman, the cultural and linguistic diversity encountered in these investigations was less interesting than the search for affective universals. Cognitive investigations of emotional experience by psychologists have also tended to assume a universal biological core to emotion (Oatley & Jenkins, 1992). Building on the early idea of James that an emotion was the “feeling of the reaction to an event” (Oatley & Jenkins, 1992, p.€ 58), the neoÂ�Jamesian tradition declared that “emotion was perception of a generalized arousal plus an attributional label” (p.€58). Although this idea represents a step beyond the view of emotion as fundamentally a biological pro-

On the Wisdom of Considering Culture and Context

cess, it suggests that a cognitive attributional label has been overlaid on the physiological core of emotion. Focusing on the cognitive mechanisms involved in emotional experience, contemporary psychology tends to emphasize the specificity and function of emotions, including their effects on attention and memory as well as their communicative roles in social interactions. These investigations have acknowledged that “the conditions that elicit an emotion distinguish it from other emotions” (Oatley & Jenkins, 1992, p.€ 60), and a growing body of work has examined the social determinants and consequences of emotion. Although these relatively recent developments in psychology seem to be conceptual moves in the right direction, there remains a conceptual bias toward viewing emotions as a set of biophysical and intrapsychic states. In this view, emotions are natural kinds, and culture is relegated to the role of configuring the situations that elicit emotions and shaping their outward expression. But is culture really so peripheral to the psychology of emotional experience? In the past two decades, philosophers, cultural historians, cross-Â�cultural psychologists, and anthropologists have trained critical attention on the dominant conceptualization of emotional experience and expression. Grounded in constructivist approaches to the study of self, personhood, and social relations, these scholars have proposed a reconceptualization of affect that transcends the Western notion of emotions as primary physiological processes with secondary cognitive, social, or cultural overlays. The result has been a new paradigm for emotion research that acknowledges biology, but that also gives serious attention to the cultural construction of experience (Abu-Â�Lughod & Lutz, 1990; Griffiths, 1997; Gross, 2006; Harré, 1986; Kitayama & Markus, 1994; Leavitt, 1996; Lutz & White, 1986; Reddy, 2001; Rosaldo, 1984; Shweder, 1993; White, 1993). The Constructivist Alternative for€Emotion Research The primary challenge facing the new paradigm for emotion research is overcoming the Cartesian dualism evident in most Western academic traditions (Leavitt, 1996) that gives rise to a familiar series of concep-


tual dichotomies (e.g., natural vs. cultural) that shape Western discourse. Several such dichotomies are evident in both scientific and Western folk discourse about emotions: mind versus body, cognition versus affect, thinking versus feeling, rational versus emotional, conscious versus unconscious, intentional versus unintentional, controlled versus uncontrolled, and so forth (Kirmayer, 1988; White, 1993). These conceptual oppositions are deeply ingrained in Western thinking and have resulted in “two-layer” theories (Lutz & White, 1986) or “dualÂ�process” models (White, 1993) of emotion that conceptualize affect as “psychobiological processes that respond to cross-Â�cultural environmental differences but retain a robust essence untouched by the social or cultural” (Abu-Â�Lughod & Lutz, 1990, p.€ 2). Thus, with regard to the study of emotions, “any phenomenon acknowledged to be culturally variable (e.g., the language available for talking about emotion) is treated as epiphenomenal to the essence of emotion” (Lutz & White, 1986, p.€408). Instead of replicating such dualisms, cross-Â�cultural researchers with serious commitments to examining the individual as an embodied agent in a sociocultural context must transcend such thinking. Leavitt (1996) described an appropriate outcome with regard to the study of affect: We would have to see emotions as primarily neither [cultural] meanings nor [psychobiological] feelings, but as experiences learned and expressed in the body in social interactions though the mediation of systems of signs, verbal and nonverbal. We would have to see them as fundamentally social rather than simply as individual in nature; as generally expressed, rather than as generally ineffable; and as both cultural and situational. But we would equally recognize in theory what we all assume in our everyday lives: that emotions are felt in bodily experience, not just known or thought or appraised. (p.€526)

Although Leavitt was perhaps a bit too dismissive of the private, inchoate, and sometimes inexpressible qualities of emotions, his larger point is clear: The dominant characterization of emotions as fundamentally individual, interior, biological events must be counterbalanced with attention to their cultural, social, and expressive dimensions.


What concretely, then, does all of this imply for the study of emotional experience—and, by extension, to the study of psychopathology as well? First, as a research construct, emotions must be understood to include biological, psychological, linguistic, social, and cultural processes that are unified in the embodied person engaged in situated and meaningful action. Second, claims regarding the uniformity of emotional experience across cultures (at least in any nuanced sense) seem implausible. An affective experience that is substantively constituted by its semiotic context cannot possibly be universal (i.e., mean the same thing) across all cultural communities of the world (see Wierzbicka, 1999, for numerous examples). Finally, the meanings of emotional experience, as facilitated and constrained by linguistic practices in particular, are situated within wider conceptual webs of cultural meaning regarding personhood, social relations, spirituality, the moral order, and so on (Harré, 1986; Lutz & White, 1986; Shweder, 1993; White, 1993). Of particular interest here is the manner in which such local webs of meaning inform and construct emotional experience for the person. Thus a systematic exploration of local ethnopsychology (i.e., theories of mind, self, and personhood) must be central to studies of emotional experience and psychopathology across cultures. An illustration of these issues is found in the work of anthropologist Theresa O’Nell (1996) on depression among the Salish Indians of the Flathead reservation in northwestern Montana. Similar to Kleinman (1986) in his investigations of Chinese neurasthenia, O’Nell discovered that depression on the Flathead Indian reservation was explicitly associated with community experiences of colonial conquest and historical oppression, as exacerbated by ongoing contention with European American racism. Most importantly, O’Nell determined that depressive-like experiences among the Salish were explicitly cast in relational terms (e.g., these were characterized by feelings of interpersonal loneliness rather than intrapsychic sadness). The relational orientation of this sociocentric society thus gave rise to three persistent states of being that shared symptoms of DSM major depression: feeling bereaved, feeling aggrieved, and feeling worthless. Of these, only the third condition

was at all likely to lead to suicide, while the first was in fact esteemed as a mark of maturity among elderly Salish tribal members, who were seen to grieve appropriately for the many losses experienced by members of Flathead society over the previous century and more. The lesson here is that forms of psychopathology that are characterized by distressing or disordered emotional experience may be configured quite differently for individuals from societies that construe the person in more egocentric or individualistic terms and from those that are more sociocentric (Kirmayer, 2007b). Implications for Cross-Â�Cultural Psychopathology Cross-Â�cultural work on emotions has shown that most complex feelings are tied to specific developmental experiences and social scenarios, which depend in turn on social structure and cultural knowledge and practice. If culture thus has the depth and sweep to actively co-Â�constitute the varieties of emotional experience around the world, then human emotions are best understood not as “natural kinds” but instead as “human kinds,” born of an interaction between biological processes and cognitive and social construals (Griffiths, 2004; Hacking, 1995a, 1999; Hinton, 1999). This interactional, biosocial view points to a way to integrate our understanding of the embodied substrate of emotion with the complex social and cultural practices that give meaning and import to emotional experience as they unfold through development. Neo-Â�Kraepelinian psychiatry—with its commitment to biological reductionism and the accompanying presumption that “real” psychiatric disorders are natural kinds—Â�cannot do justice to this complex interaction.

DSM and the Problem of Cultural Imperialism As we have already observed, the publication of DSM-III (APA, 1980) was a landmark historical, scientific, and political achievement, signaling the advent of neo-�Kraepelinian psychiatry in the United States. Owing to standardized criterion sets with explicit application algorithms, modern versions of

On the Wisdom of Considering Culture and Context

DSM afford reliable psychiatric diagnosis, and thereby permit a cumulative science of psychopathology. Of course, construct validity for the hundreds of postulated disorders within DSM remains elusive; instead, psychopathologists employ DSM under the optimistic assumption that over time, accumulating evidence from research studies using standardized diagnostic criteria will enable them to “bootstrap” their way to diagnostic validity. Nevertheless, DSM has come to dominate the ways in which mental health professionals in the United States and in many other countries classify and diagnose psychiatric illness—and, as a consequence, to suffuse the ways in which patients (and the broader public) make sense of their distress and dysfunction. That is, in everyday clinical practice, the hypothetical constructs classified within DSM take on a privileged ontological status in the lives of patients, professionals, and institutions through routine processes of reification. Indeed, it is through clinical praxis that the scientific and clinical conjectures codified in DSM become accepted as authorized knowledge and authoritative discourse. In actuality, then, DSM simultaneously serves two different purposes that are potentially at odds with one another: namely, as a provisional scientific taxonomy for facilitating empirical research on the one hand, and as an institutionalized professional manual for guiding clinical practice on the other (Gone, 2003b). The tensions between these functions (and epistemic stances) are greatly exacerbated in cross-Â�cultural applications of DSM, especially those in which enduring asymmetries in cultural capital and political power lend themselves to the unwarranted hegemony of Western psychiatric discourse. Psychiatric Services and Western Cultural Proselytization Contemporary views of culture recognize that most individuals have access to multiple cultural systems, and that the “culture” of specific communities is actually made up of many competing and contesting streams or positions. Acknowledging the importance of cultural difference is not simply a matter of taking account of variations in developmental experiences, social contexts, and com-


mitments. Cultures are unequally accorded or invested with power and authority. The power attached to specific cultural systems and communities arises from a specific history of domination and control that may continue to exert effects on ways of thinking long after the machinery of domination has been challenged or dismantled. In psychiatric research concerned with the mental health status of historically oppressed ethnic/racial minority communities in the United States, for example, psychopathologists must recognize that the “culture” of the clinic is not the “culture” of the community. More specifically, the assumptions, assertions, aspirations, and attributions that mental health professionals routinely rely on are grounded in the categories and conventions of Western therapeutic discourse, including those contained within DSM. Such discourse has emerged historically from northern European and European American sensibilities regarding normative and disordered psychological, emotional, and behavioral functioning (Gaines, 1992). As a result, the therapeutic discourse that anchors mainstream clinical activity undertaken in many non-Â�Western cultural contexts may diverge in substantial ways from local assumptions and expectations of wellness, health, and “the good life” (as we have already seen in the context of emotional experience and expression). Moreover, for much of the history of psychiatry, the profession has worked in cooperation or collusion with the powers of colonial domination (Bhugra & Littlewood, 2001; Jackson, 2005; Keller, 2007; McCulloch, 1995; Sadowsky, 1999). More specifically, the privileging of Western theories of psychopathology and therapeutic discourse has been associated with longÂ�standing efforts by European Americans to express or achieve cultural dominance over other peoples through processes of colonization, and to maintain dominance through racialized hierarchies of power and authority. This historical bid for European American cultural dominance frequently involved the explicit devaluation, disruption, and displacement of these alternate frames of reference, modes of representation, and modalities of experience. A small group of psychiatric thinkers and practitioners has challenged this collusion, rejecting the rac-


ist ideologies that rationalized colonial violence, and supporting the political struggles that have sought to transform or overthrow colonial regimes (Fanon, 1982). The result is an important literature that has examined the impact of colonial systems of racism and oppression on the identities, personalities, and psychological well-being of colonized subjects, as well as the possibilities for liberatory psychiatric practice (Vergès, 1996). This literature would benefit from contemporary reconsideration in light of the changing forms of structural violence (Gilroy, 2004). Nevertheless, despite these occasional (and politically marginalized) efforts, both cultural divergences and asymmetries in power render the provision of conventional psychiatric services to historically oppressed communities a politically suspect activity that may advance Western cultural proselytization in the guise of therapeutic knowledge and activity.4 A Postcolonial Discourse of Distress We can illustrate the kinds of ideological dangers we have in mind with reference to two of the most prevalent forms of DSM psychopathology: alcohol dependence and major depression. In an ethnographic investigation on a northern Plains Indian reservation, Gone (2007, 2008c) identified a prototypical “discourse of distress” concerning problematic drinking and depression in contemporary Native American tribal life. According to one especially instructive respondent (pseudonymously named “Traveling Thunder”), these problems could be traced to disrupted ceremonial tradition in the context of historical dominance by European Americans. More specifically, Traveling Thunder identified four historical epochs in his characterization of the causes of pathological depression and drinking on the reservation. The first epoch was the era of “Paradise,” a precolonial existence in which such pathologies were largely unknown, owing to the perfect harmony and balance wrought by community adherence to the strict observation of social custom and sacred ritual. The second was the era of “Conquest,” or the colonial encounter in which the genocidal and assimilative activities of European Americans led to the annihilation of custom and ritual. The third epoch was the era of

“Loss,” in which the postcolonial effects of the annihilation of custom and ritual led to anomie, and in turn to substance abuse, depression, and sometimes suicide. Finally, the current epoch is the era of “Revitalization,” in which the Creator has “pitied” the people enough to facilitate a communal reclamation of indigenous custom and ritual. In one particularly illuminating moment during the interview, Gone (2007) asked Traveling Thunder to reflect on the conditions under which he would refer a distressed loved one to the mental health professionals at the local reservation clinic. His reply lacked any trace of ambivalence: That’s kind of like taboo. You know, we don’t do that. We never did do that. .€.€. If you look at the big picture, you look at your past, your history, where you come from .€.€. and you look at your future where the Whiteman’s leading you, I guess you could make a choice. Where do I want to end up? And I guess a lot of people .€.€. want to end up looking good to the Whiteman. .€.€. Then it’d be a good thing to do: go [to the] white psychiatrists .€.€. in the [reservation clinic] and say, .€.€. “Go ahead and rid me of my history, my past, and brainwash me forever so I can be like a Whiteman.” (p.€294)

Thus, for Traveling Thunder, the activity of “white psychiatrists” on the reservation was explicitly marked as an extension of the colonizing project, in which indigenous selfhood remains a site of neocolonial engagement and resistance. As an alternative, Traveling Thunder proposed the reclamation of indigenous selfhood through the reestablishment of ritual practice. Such practice serves to link the human self to other-than-human Persons5 in the respectful offering of gifts and prayers in exchange for the compassionate outpouring of prosperity and blessings. In the process, alienation and anomie are simultaneously (but secondarily) resolved through the establishment of a robust cultural identity (Gone, 2006a, 2008b, in press-a). A central feature of Traveling Thunder’s discourse of distress was its reliance on observations, inferences, and insights drawn from the sociohistorical and spiritual levels of experience and analysis. From this perspective, mental health problems—Â�including the anomie, demoralization, depression, substance abuse, and suicide found on the reservation—were understood as direct

On the Wisdom of Considering Culture and Context

consequences of the European American colonial encounter that disrupted ritual relationships and community responsibilities to powerful other-than-human Persons. It follows that the most effective remedy for pathological drinking and depression within the community would be a restoration and return to individual and collective ceremonial practice (Gone, 2007). In sum, this contemporary ethnopsychological discourse configures wellness (i.e., life lived “in a good way”) quite differently from the “mental health” of psychiatry and the associated professions, and posits quite different etiologies for serious distress (Gone, in press-c). For Traveling Thunder, pathological drinking and depression were functions of culture, history, and identity, contrasting sharply with the reigning psychiatric emphasis on genetic predispositions, chemical imbalances in the brain, and other biologically reductionist explanations as fundamental to these disorders. As we have already noted, the concepts, categories, principles and practices of neoÂ�Kraepelinian psychiatry—Â�including the codifications of DSM—remain cultural artifacts, the meanings and mechanisms of which emerge from and depend on their cultural intelligibility within a shared discursive frame. As a result, casually embracing DSM in one’s cross-Â�cultural professional activity risks irrelevance at best, or an often subtle (but sometimes overt) Western cultural proselytization in the guise of therapeutic progress at worst. Certainly the ideological hazards of this nearly invisible “West is best” cultural imperialism in postcolonial societies and contexts such as Traveling Thunder’s reservation homeland remain worrisome and require serious consideration and redress. Decolonizing Psychiatry “Postcolonial” is a term that has been used to characterize the struggles for liberation undertaken by formerly colonized peoples as they assert their social, political, and cultural autonomy. Such struggles, however, have not eliminated structures of domination established during colonial eras or prevented the emergence of new strategies of exploitation rooted in national or ethnic interests. These structures and strategies have the effect of maintaining inequalities, with profound consequences for the quality of life of


formerly colonized peoples in postcolonial societies. In addition, recent processes of globalization have facilitated shifts in strategies of domination toward systems of power structured by consumer capitalism and the interests of multinational corporations and their associated economic institutions. For this reason, the prefix “post-” in “postcolonial” probably warrants scare quotes to denote the fact that many oppressive features of colonization have not ended, but instead have mutated or gone underground, only to reemerge in powerful new forms. Indeed, the increasingly global influence of Western psychiatry—Â�accompanied by its material and discursive power to undermine or displace local notions of self, personhood, identity, emotion, social relations, spirituality, distress, wellness, and healing around the world—would seem to require a great deal more ethical attention to the role of psychiatric services as vehicles to export specific cultural values, particularly those of secularism and especially individualism. By virtue of their creation, utilization, and dissemination by psychiatrists, the psychopathological constructs classified within DSM are generally cast in terms that locate the “disorder” within an individual. This reflects a “causal attributional bias” that may result in blaming the person for his or her affliction. In a now-Â�classic article, Caplan and Nelson (1973) criticized “the tendency to hold individuals responsible for their own problems” (p.€199)—first, by focusing on “person-Â�centered” characteristics while downplaying or ignoring situationally relevant factors; and, second, by attributing causal significance to any person-Â�centered variables found to be statistically associated with the social problem in question. Caplan and Nelson reviewed a sample of published articles indexed in Psychological Abstracts to demonstrate that in research with African Americans, psychologists invested “disproportionate amounts of time, funds, and energy in studies that lend themselves, directly or by implication, to interpreting the difficulties of black Americans in terms of personal shortcomings” (p.€204), rather than in terms of situational factors or systemic inequalities. They identified several social and political functions served by such construals of social problems, and concluded that “person-blame interpretations are in everyone’s


interests except those subjected to analysis” (p.€210). In light of these observations, let us return to the alternative presented by Traveling Thunder, who observed that the epidemic of distress in his reservation community appeared to have emerged hand in hand with the ravages of colonization. Traveling Thunder’s account emphasized situational factors and systemic inequalities rather than “person-Â�centered” biogenetic or intrapsychic factors. Accordingly, Traveling Thunder asserted that the community rather than the individual ought to be the focus of therapeutic attention and intervention, and that the problems faced by individuals and the community might best be characterized as an existential and spiritual crisis. Like the expressions of depression among the Salish recounted by O’Nell, and like the current appropriations of the term “historical trauma” among many other indigenous peoples and communities (Brave Heart & DeBruyn, 1998; Gone, 2008b, in press-a), Traveling Thunder’s discourse embeds psychopathological experience in the larger meanings of collective experiences of longstanding European American subjugation. This focus on social and historical context as a way of characterizing individual suffering is in marked contrast to the dominant ideological commitments of neo-Â�Kraepelinian psychiatry, in which mental disorders are presumed to be natural kinds that afflict individuals through presently unknown pathophysiological processes. To the extent that they employ this decontextualized view of psychiatric disorders, mental health services in the reservation context cannot help engaging in the sort of “person blaming” decried by Caplan and Nelson. The basic remedy for this unfortunate state of affairs is to resituate individual and social suffering in its cultural and historical contexts. This has a political dimension, insisting on the importance of the interpretive frames and perspectives of the culturally diverse subjects of psychiatry theory. But the development of situated theory in psychopathology is not simply a matter of “political correctness.” It requires a vibrant program of cross-Â�cultural research on various forms of psychopathology—a program that seriously engages “emic” (local or emergent) frames of reference, in addition to the “etic” (exter-

nal and imposed) models of psychology and psychiatry. Such research would not ignore conventional approaches to the investigation of psychopathology, but would recognize that the relationship between local and external models and frames of reference requires systematic study through open-ended empirical work that does not assume that either framework will provide all the answers. In some instances, the compelling validity of local understandings may directly challenge the constructs of DSM, demonstrating their inapplicability or irrelevance to local forms of suffering (i.e., Kleinman’s [1988] “category fallacy”). In other cases, emic constructs may lead to models of wider applicability and so themselves become etic constructs. This systematic empirical project is based on the conviction that many forms of psychopathology (including some of the most popular and prevalent diagnoses) are “human kinds” best approached through careful investigation of the local, lived meanings of experience, rather than “natural kinds” that can be adequately characterized in terms of universal biological mechanisms and corresponding categories of experience.

Culture, Context, and Experience in Psychiatric Science and€Clinical€Practice In line with the reflexive stance central to contemporary social studies of science, we have so far approached the importance of culture for psychiatric nosology through its impact on the nosological enterprise itself. This framework targets not the ethnocultural characteristics of patients per se, but instead addresses the cultural embedding of diagnostic theory and practice, especially as it pertains to everyday clinical concerns. The model discussed here is quite general and, in the context of preceding observations and insights, argues for professional recognition of the pervasive effects of culture and context on every aspect of the psychiatric enterprise. The Uses of Psychiatric Nosology and the Impact of Diagnosis Psychiatry covers a broad domain of human problems. Mental illness is not one thing,

On the Wisdom of Considering Culture and Context

but a congeries of heterogeneous problems—Â� including forms of brain dysfunction, psychopathological processes that result from various forms of learning, problems that reside in interpersonal interaction, and problems that consist of incoordination or contradiction among these different levels of organization (Kirmayer & Young, 1999). These problems are related to one another by family resemblances, so that there is no common essence or single characteristic shared by every psychiatric disorder, except at a very high level of abstraction. Although, as we have argued, biological, psychological, and social factors contribute to all of these problems, the relative importance of causal and aggravating factors varies for each type of problem as well as for each individual, episode, and situation. As a result, no one solution to the structure and function of psychiatric nosology will work. In particular, neither genetics research nor neuroimaging will tell us what to include in a nosology unless we decide to redefine the domain of psychiatry narrowly in terms of these technologies (Robert, 2007; Robert & Plantikow, 2005). The construction of a nosology and related diagnostic instruments and techniques reflects specific goals or purposes. Earlier, we have discussed the tensions between the use of psychiatric nosology as a provisional scientific classification of psychopathology and as a manual for professional practice. Diagnostic systems have additional uses in other domains, including the determination of health care policy and the regulation of other social institutions. The scope and content of a diagnostic system may have profound effects on the design and function of health care systems, including resource allocation and access to care. In the wider social context, diagnoses serve to position individuals by assigning them the sick role, and thus identifying the persons as legitimately distressed or disabled and deserving of help, compensation, or support. Diagnosis also has implicit functions. For the clinician, assigning a diagnostic label serves to name and contain the confusion and threat presented by the suffering patient (Kirmayer, 1994). For patients, a diagnostic label and its connotations are used to draw out the implications of an illness and, when the condition is chronic, to (re)construct aspects of personal


identity. These implicit meanings of diagnosis may also have powerful social implications, conferring stigma or prompting other practices of exclusion. A psychiatric nosology, then, is not simply a systematic ordering of categories found in nature, but constitutes a map and charter of a social world. Nosology provides a map, in that it marks off specific domains and establishes borders and boundaries whose crossing makes a difference to individuals’ social status. Nosology also functions as a social charter because this act of mapping creates an “official” reality and authorizes the architects and users of the diagnostic system to exercise specific forms of social power. In the context of the clinic, diagnosis is part of constructing a problem list, identifying the issues that require some form of help or clinical attention. Clinical problem lists commonly go well beyond the specific entities of diseases or disorders to include social problems, interpersonal conflicts, and existential dilemmas—all of which figure in patients’ suffering, and which may influence the appropriate intervention for specific disorders or may be primary foci of concern in their own right. Clinical Epistemology and the Place of Culture in Psychopathology Scientific research and clinical assessment involve different epistemological assumptions. Clinical knowledge is constrained by the temporal frame of the clinical encounter and its specific goals for problem identification and solution. The focus is on signs and symptoms, and on what can be identified through history taking, systematic interviewing and observation, physical examination, and laboratory tests. The aim is to use this information to infer the underlying disorder that accounts for a patient’s distress and that can then be targeted for intervention. The implicit theory of medical semiotics views symptoms simply as more or less veridical reports of bodily events or physiological perturbations (Kirmayer, 1994). Although the mapping from pathophysiology to symptom may be many-to-one (nonspecific symptoms may result from many different forms of pathology) or one-to-many (a single pathology may have variable clinical manifestations), in practice medical semiotics commonly as-


sumes a one-to-one mapping or isomorphism from physiological disturbances to bodily experience and from bodily experience to symptom report. Hence symptoms are taken as indicators of underlying pathophysiological processes. Given the lack of independent biomarkers for psychiatric disorders, an assumption is also made that clusters of symptoms (syndromes) are sufficient to identify distinct forms of pathology. Furthermore, it is tacitly assumed that the diagnostic nosology identifies all the clinically significant forms of pathology that can occur. The accuracy and completeness of the nosological map are therefore matters of great importance for science, clinical care, and policy. Problems that fall outside the nosology are not accorded the same level of interest, status, or priority by researchers, clinicians, and policymakers. This makes the nosology an important regulator of psychiatric science and practice. This is an especially important issue in cross-Â�cultural work because of the epistemological problem identified by Kleinman (1988) as the “category fallacy.” Efforts to apply a set of diagnostic categories developed in one cultural context in a different setting may obscure important cultural differences. Although it may be possible to identify people who fit the diagnostic criteria, this does not ensure the local validity of the category; nor does it rule out the possibility that individuals with related forms of suffering are not captured by the diagnostic criteria. Local categories of illness may yield better indicators of distress, and better predictors of prognosis and treatment outcome. Testing this possibility requires specific research methods (Canino, Lewis-Â�Fernandez, & Bravo, 1997). An additional epistemological problem arises from what Hacking has called “the looping effect of human kinds” (Brinkmann, 2005; Hacking, 1995a, 1999)—an elaboration on our prior consideration of “natural kinds” and “intentional categories.” Hacking recognized intentional categories (or human kinds) as those that depend on specific ways of construing experience. In such instances, the very act of diagnosing a given pathology in an individual harbors the potential to alter that individual’s experience of the pathology, as well as the subsequent scientific and professional construals of that individual’s behavior. These ways of con-

struing behavior circulate in the larger society, becoming social and cultural norms, models, and practices that alter other individuals’ interpretations of their own experiences. Hence changes in cultural assumptions or cognitive models will lead to new conceptual categories that are reified and stabilized by recursive social processes of dissemination and enactment. The changing forms of “trauma” reviewed earlier, as well as the evolution of “hysteria,” provide clear examples of this phenomenon (Hacking, 1995b, 1998). Although we have previously employed Shweder’s (1988) distinction between natural kinds and intentional categories (or human kinds), many psychiatric categories are best thought of as what Hacking terms “interactive kinds,” in which there is a transaction between natural distinctions and culturally constructed concepts. Cognitive theory would suggest that panic disorder and major depression are two examples of a specific version of interaction that Hacking calls “biolooping,” in which (culturally mediated) modes of construing experiences of the body and the self lead to physiological disturbances (Hinton & Hinton, 2002; Hinton, Hinton, Pham, Chau, & Tran, 2003). To the extent that these disturbances follow a biologically dictated final common pathway, the disorders may be viewed as ultimately independent of our construals, and hence as natural kinds (Cooper, 2004). For example, if there were a core syndrome of neurasthenia or depression involving a state of physiological depletion, we might fix on this aspect to define a category of pathology independently of how an individual arrives at that state (Kirmayer & Jarvis, 2005). However, there may be forms of psychopathology in which the culturally and cognitively mediated modes of construal are essential to defining the problem. In the case of “intentional” behavior, which is distinguished by the fact that the person can give reasons for their action (and that the reasons are causally implicated in the action), there is a loop that depends on the distinctively human capacity for self-Â�awareness. For this type of problem, there is no way to define the pathology without characterizing the nature of the disturbances in self-Â�awareness, self-Â�representation, and self-Â�control, which in turn depend on particular cultural con-

On the Wisdom of Considering Culture and Context

cepts of self and personhood, and on larger systems of values, social institutions, and discursive practices (Kirmayer, 2006). None of these distinctions means that we must dispense with constructing categories, but we must recognize that the larger social contexts of psychiatry—Â�including cultural notions of personhood and affliction— loop back at multiple levels into our nosology; into the process of clinical assessment and diagnosis; and into the vicious circles of attention, attribution, and behavior that constitute many forms of psychopathology (Kirmayer & Sartorius, 2007). Psychiatric nosologies, therefore, do not simply describe problems out there in the world, but actively contribute to the ways in which people construe and experience their distress. That is, as we have seen, psychiatric nosologies actively create culture even as they reflect cultural processes and practices. Implications for Research To the extent that social processes of meaning construction and positioning are central to the cause, course, and outcome of various forms of psychopathology, research must include systematic attention to the range of variables reflecting cultural variations in human experience. Given the marked heterogeneity within ethnocultural groups, this must go beyond mere comparisons of individuals on the basis of their ethnic identity, to examine the impact of specific knowledge, behaviors, or practices that can be linked to putative psychopathological processes. This type of research would decompose “culture” and “ethnicity” into explicit components or dimensions (e.g., specific practices associated with the body, concepts of personhood, explanations of affliction, techniques of healing) that can be studied in interaction with other biological, psychological, and social processes. However, a basic insight of anthropology is that the components of culture are not arbitrarily arranged, but constitute coherent systems (even if they contain tensions and contradictions), so that the interactions between different components must be studied to understand the tradeoffs that may occur for individuals following one or another illness trajectory. Beyond this incorporation of culture as sets of interrelated components, “factors,”


or parameters that configure human biology, psychology, and the social contexts that govern behavior (whether pathological or adaptive), we have argued that the refinement of theory in psychopathology requires systematic attention to the social, cultural, and historical dimensions of human suffering and of the conceptual systems we devise to categorize, explain, and intervene. These deserve critical analysis not only for political reasons, since they have served as instruments of oppression or exclusion, but also because, in the nature of human experience, our conceptual categories shape our lives in ways that can give rise to new types of problems and solutions. Study of these “intentional kinds” and social looping effects requires different methods from those that currently dominate psychopathology research, including the critical and interpretive strategies of the social sciences, but also empirical studies of the social, economic, and political shaping of psychiatric knowledge and practice (Healy, 2004; Horwitz & Wakefield, 2007). This social analysis is not only a corrective to the tendency to promote specific models that serve special interests; it also opens a space for fresh thinking about the nature of psychopathology and well-being. Implications for Clinical Practice Although psychiatric diagnoses serve as a form of explanation, they differ in important ways from the biographical accounts common in personal narratives (McHugh & Slavney, 1988). Psychiatric nosologies contain generic information on postulated diseases and disorders. The act of diagnosis maps a patient’s idiosyncratic story and clinical presentation onto a general set of categories. It does this by abstracting the essential characteristics of the patient’s history and illness experience, paring away the irrelevant details, and seeing through the obscuring masks of style of narration and illness behavior to uncover the essence of a prototypical disorder. This, at least, is how disease categories and nosologies are constructed as systems of ideal types. Some concession to individual variability in illness manifestations occurs in the construction of polythetic categories, in which a case needs only a certain number of symptoms from a list to meet diagnostic criteria.


In contrast to this abstracting, decontextualizing, and essentializing process in the construction of disease categories, clinical explanation moves in the opposite direction. To convey a meaningful diagnosis to a patient and plan an appropriate clinical response, the clinician must particularize, qualify, and contextualize illness explanations. Often, however, clinicians simply present a generic story to patients modeled on the disease prototype. Unfamiliar symptoms or problems are reinterpreted to fit a specific prototype, or discounted and ignored as minor and irrelevant. Insofar as a patient’s experience does not fit the template, the discrepancies are viewed as irrelevant or the patient is viewed as a poor historian, oblivious to or misinterpreting the true nature of his or her condition (Kirmayer, 1988, 1994). This stripping down of illness experience to fit the diagnostic paradigm is justified on the basis of the notion that diagnostic entities have essential biological characteristics, and that what is crucial about the patient’s condition can be typified by these core features. However, this does not address the basic mandate of medicine. People bring symptoms and predicaments to their doctors, not just diseases or disorders. These predicaments may contribute to the cause, course, and outcome of specific disorders. Because these predicaments are socially constituted, they will differ across social and cultural contexts, giving rise to potentially important differences in the nature of psychopathology. Strategies for including social and cultural context in clinical assessment of psychopathology include the cultural formulation presented in an appendix to DSM-IV-TR (APA, 2000). This was introduced by a working group of psychiatrists, psychologists, and anthropologists, to provide a minimal list of the sorts of contextual factors to be considered in assessing psychopathology (Mezzich et al., 1999). The list includes identity, illness explanations, functioning, family or social supports, and the relationship with the clinician. Since its introduction, many case studies using the cultural formulation have been published, but there has been no systematic assessment of its utility. Given all of the arguments we have adduced above, it would seem that something akin to the cultural formulation is crucial to provide balance to the decontextualized view of problems inherent

in DSM. Clinical experiences with cultural consultation clearly demonstrate the potential of the cultural formulation to identify errors in diagnosis and produce more comprehensive and culturally appropriate assessment and treatment plans (Kirmayer, Groleau, Guzder, Blake, & Jarvis, 2003). Much further work is needed to elaborate the cultural formulation, evaluate its utility, and give it more prominence as a way to foster clinical thinking that moves from abstract category to lived reality.

Conclusion In this chapter, we have considered what is at stake in the assimilation of local discourses of distress into the increasingly global discourse of neo-Â�Kraepelinian psychiatry— with its overt construal of various forms of psychopathology as “natural kinds” arising from distinctive, underlying pathophysiologies. We have outlined an alternative perspective that gives due weight to culture in the study of psychopathological experience. This view centers on the importance of meaningful human predicaments as a way to understand the interaction of biological, psychological, and social processes in the emergence of distinct (though not discrete) forms of psychopathology—forms that depend to varying degrees on social context for their shape, content, and “natural history.” In this view, the diagnostic entities found in psychiatric nosologies may not reflect natural kinds (occurring in nature independently of our cognitive and cultural construals), but are the outcomes of social-Â�interactional and historical processes that include our culturally mediated ways of understanding and interpreting human suffering. For the most part, culture functions as a taken-for-Â�granted background that sustains our common sense and tacit knowledge of the social world, as well as our clinical models, institutions, and practices. We recognize culture only at the margins, in the encounter with those we view as different or “other.” Confronting our own cultural assumptions through encounters with others has been commonplace throughout human history, but most often cultural diversity has been subordinated to a single set of categories, concepts, and values imposed by dominant

On the Wisdom of Considering Culture and Context

groups, which work to devalue and disqualify alternative frameworks for experience. Psychiatry itself has practiced this form of conceptual imperialism, and challenges to this hegemonic view are few and far between. In recent years, however, migration and telecommunications have brought new levels of cultural diversity into clinical settings in many parts of the world. Culture then presents itself as a daily problem of recognizing and addressing diversity in the clinical application of psychiatric nosology (Kirmayer & Minas, 2000). This diversity cannot be addressed with theories of psychopathology and a psychiatric nosology based on research conducted in only one or a few cultural contexts (Alarcón et al., 2002). Enlarging the study of psychopathology by emphasizing the contextual shaping of psychiatric problems holds the prospect of generating a more accurate view of the sources of suffering and the mechanisms of psychopathology. Attention to culture, then, is not only a matter of serious ethical, political, and pragmatic issues in the delivery of mental health care, but a basic requirement for a science of psychopathology that seeks to understand our nature as cultural beings. Human biology is cultural biology. The wisdom—we should say the necessity— of attending to culture in the development of psychopathological theory and in the preparation of future nosologies therefore emerges on the grounds of both scientific and political aspirations. Acknowledgment Portions of this chapter are from Kirmayer (2005). Copyright 2005 by S. Karger AG, Basel. Adapted by permission.

Notes 1. We use scare quotes here to acknowledge the fact that the “phenomenology” of DSM-III (APA, 1980) and DSM-IV (APA, 1994) gives scant recognition to the realm of inner experience explored by several generations of phenomenologically oriented philosophers and psychologists working within a Continental tradition. Instead, “phenomenology” in the DSM system means discrete symptoms, signs,

91 and behaviors that can be reliably measured by an external observer. 2. In fact, the distinction between natural and intentional kinds (like that between positivist and constructivist epistemologies) is overstated, and the examples themselves point to the difficulty of making a sharp contrast. Ian Hacking (1999) has described the wide range of uses of the notion of social construction, and has also provided some compelling examples of the social construction of psychiatric disorders (Hacking, 1995b, 1998). However, most examples of intentional categories in the area of psychiatry are what Hacking has called “interactive kinds,” built out of an interaction between more or less obdurate features of the natural world (including our own physiology and psychology) and socially mediated responses. 3. Historically, in the study of culture and psychopathology, this has been framed as a contrast between “pathogenesis,” usually assumed to involve biological processes or physical interactions with the environment, and “pathoplasticity,” the cultural shaping of the expressions of more basic pathogenic processes. 4. One of us has argued this claim in more detail in a series of papers (see Gone, 2003a, 2004a, 2004b, 2006b, 2008a, 2009, in press-b; Gone & Alcántara, 2007). 5. In the traditions of many indigenous peoples—Â� particularly those who were hunters—Â�animals and other “natural” beings were seen to possess some of the same qualities of human personhood (e.g., autonomy, intentionality, and so forth) and hence are best termed “otherthan-human Persons.” The capital P serves to convey respect for their often sacred status.

References Abu-Â�Lughod, L., & Lutz, C. A. (1990). Introduction: Emotion, discourse, and the politics of everyday life. In C. A. Lutz & L. Abu-Â�Lughod (Eds.), Language and the politics of emotion (pp.€ 1–23). Cambridge, UK: Cambridge University Press. Alarcón, R. D., Bell, C. C., Kirmayer, L. J., Lin, K.-H., Ustun, T. B., & Wisner, K. L. (2002). Beyond the funhouse mirrors: Research agenda on culture and psychiatric diagnosis. In D. J. Kupfer, M. B. First, & D. A. Regier (Eds.), A research agenda for DSM-V (pp.€219–289).

92 HISTORICAL AND CULTURAL PERSPECTIVES Washington, DC: American Psychiatric Association. American Psychiatric Association (APA). (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association (APA). (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. American Psychiatric Association (APA). (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association (APA). (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. Beard, G. (1869). Neurasthenia, or nervous exhaustion. Boston Medical and Surgical Journal, 3(13), 217–221. Beutler, L. E., & Malik, M. L. (2002). Rethinking the DSM: A psychological perspective. Washington, DC: American Psychological Association. Bhugra, D., & Littlewood, R. (Eds.). (2001). Colonialism and psychiatry. New Delhi: Oxford University Press. Blashfield, R. K. (1984). The classification of psychopathology: Neo-Â�Kraepelinian and quantitative approaches. New York: Plenum Press. Brave Heart, M. Y. H., & DeBruyn, L. M. (1998). The American Indian Holocaust: Healing historical unresolved grief. Journal of the National Center, 8(2), 60–82. Brinkmann, S. (2005). Human kinds and looping effects in psychology. Theory and Psychology, 15(6), 769–791. Canino, G., Lewis-Â�Fernandez, R., & Bravo, M. (1997). Methodological challenges in crossÂ�cultural mental health research. Transcultural Psychiatry, 34(2), 163–184. Caplan, N., & Nelson, S. D. (1973). On being useful: The nature and consequences of psychological research on social problems. American Psychologist, 28(3), 199–211. Collins, H., & Pinch, T. (1993). The Golem: What everyone should know about science. Cambridge, UK: Cambridge University Press. Cooper, R. (2004). Why Hacking is wrong about human kinds. British Journal for the Philosophy of Science, 55, 73–85. Crews, F. (1995). The memory wars: Freud’s legacy in dispute. New York: New York Review of Books. Darwin, C. (1965). The expression of the emotions in man and the animals. Chicago: Uni-

versity of Chicago Press. (Original work published 1872) Davidson, J., Foa, E. B., Blank, A. S., Brett, E. A., Fairbank, J., Green, B. L., et al. (1996). Posttraumatic stress disorder. In T. A. Widiger, A. J. Frances, H. A. Pincus, R. Ross, M. B. First, & W. W. Davis (Eds.), DSM-IV sourcebook (Vol. 2, pp.€577–605). Washington, DC: American Psychiatric Association. Eisenberg, L. (1977). Disease and illness. Culture, Medicine, and Psychiatry, 1, 9–23. Ekman, P. (1984). Expression and the nature of emotion. In K. Scherer & P. Ekman (Eds.), Approaches to emotion (pp.€319–343). Hillsdale, NJ: Erlbaum. Fanon, F. (1982). Black skin, white masks (1st Evergreen ed.). New York: Grove Press. Frueh, B. C., Elhai, J. D., Grubaugh, A. L., Monnier, J., Kashdan, T. B., Sauvageot, J. A., et al. (2005). Documented combat exposure of US veterans seeking treatment for combat-Â�related post-Â�traumatic stress disorder. British Journal of Psychiatry, 186, 467–472. Gaines, A. D. (1992). From DSM-I to III-R: Voices of self, mastery and the other: A cultural constructivist reading of U.S. psychiatric classification. Social Science and Medicine, 35(1), 3–24. Geertz, C. (1973). The interpretation of cultures. New York: Basic Books. Gilroy, P. (2004). Postcolonial melancholia. New York: Columbia University Press. Gleaves, D. H. (1996). The sociocognitive model of dissociative identity disorder: A reexamination of the evidence. Psychological Bulletin, 120, 42–59. Gone, J. P. (2003a). American Indian mental health service delivery: Persistent challenges and future prospects. In J. S. Mio & G. Y. Iwamasa (Eds.), Culturally diverse mental health: The challenges of research and resistance (pp.€ 211–229). New York: BrunnerÂ�Routledge. Gone, J. P. (2003b, November). DSM dilemmas: Ruminations on the culture of classification and the discourse of diagnosis. In J. P. Gone & T. D. O’Nell (Co-Â�Chairs), Decolonizing psychiatry: Revisiting the role of DSM criteria in Native American mental health care. Symposium conducted at the annual meeting of the American Anthropological Association, Chicago. Gone, J. P. (2004a). Keeping culture in mind: Transforming academic training in professional psychology for Indian country. In D. A.

On the Wisdom of Considering Culture and Context Mihesuah & A. Cavender Wilson (Eds.), Indigenizing the academy: Transforming scholarship and empowering communities (pp.€124– 142). Lincoln: University of Nebraska Press. Gone, J. P. (2004b). Mental health services for Native Americans in the 21st century United States. Professional Psychology: Research and Practice, 35(1), 10–18. Gone, J. P. (2006a). Mental health, wellness, and the quest for an authentic American Indian identity. In T. Witko (Ed.), Mental health care for urban Indians: Clinical insights from Native practitioners (pp.€ 55–80). Washington, DC: American Psychological Association. Gone, J. P. (2006b). Research reservations: Response and responsibility in an American Indian community. American Journal of Community Psychology, 37(3–4), 333–340. Gone, J. P. (2007). “We never was happy living like a Whiteman”: Mental health disparities and the postcolonial predicament in American Indian communities. American Journal of Community Psychology, 40(3–4), 290–300. Gone, J. P. (2008a). Introduction: Mental health discourse as Western cultural proselytization. Ethos, 36(3), 310–315. Gone, J. P. (2008b). The Pisimweyapiy Counselling Centre: Paving the red road to wellness in northern Manitoba. In J. B. Waldram (Ed.), Aboriginal healing in Canada: Studies in therapeutic meaning and practice (pp.€ 131–203). Ottawa, ON, Canada: Aboriginal Healing Foundation. Gone, J. P. (2008c). “So I can be like a Whiteman”: The cultural psychology of space and place in American Indian mental health. Culture and Psychology, 14(3), 369–399. Gone, J. P. (2009). Encountering professional psychology: Re-Â�envisioning mental health services for Native North America. In L. J. Kirmayer & G. G. Valaskakis (Eds.), Healing traditions: The mental health of Aboriginal peoples in Canada (pp.€419–439). Vancouver: University of British Columbia Press. Gone, J. P. (in press-a). A community-based treatment for Native American historical trauma: Prospects for evidence-based practice. Journal of Consulting and Clinical Psychology. Gone, J. P. (in press-b). “I came to tell you of my life”: Narrative expositions of “mental health” in an American Indian community. In M. Aber, K. Maton, & E. Seidman (Eds.), Empowering settings and voices for social change. New York: Oxford University Press. Gone, J. P. (in press-c). Psychotherapy and tradi-

93 tional healing for American Indians: Exploring the prospects for therapeutic integration. The Counseling Psychologist. Gone, J. P., & Alcántara, C. (2007). Identifying effective mental health interventions for American Indians and Alaska Natives: A review of the literature. Cultural Diversity and Ethnic Minority Psychology, 13(4), 356–363. Gone, J. P., Miller, P. J., & Rappaport, J. (1999). Conceptual self as normatively oriented: The suitability of past personal narrative for the study of cultural identity. Culture and Psychology, 5(4), 371–398. Griffiths, P. E. (1997). What emotions really are: The problem of psychological categories. Chicago: University of Chicago Press. Griffiths, P. E. (2004). Is emotion a natural kind? In R. C. Solomon (Ed.), Thinking about feeling: Contemporary philosophers on emotions (pp.€233–249). New York: Oxford University Press. Gross, D. M. (2006). The secret history of emotion: From Aristotle’s rhetoric to modern brain science. Chicago: University of Chicago Press. Hacking, I. (1995a). The looping effect of human kinds. In D. Sperber, D. Premack, & A. J. Premack (Eds.), Causal cognition: A multidisciplinary debate (pp.€ 351–383). Oxford, UK: Oxford University Press. Hacking, I. (1995b). Rewriting the soul. Princeton, NJ: Princeton University Press. Hacking, I. (1996). Memory sciences, memory politics. In P. Antze & M. Lambek (Eds.), Tense past: Cultural essays in trauma and memory (pp.€67–87). New York: Routledge. Hacking, I. (1998). Mad travelers: Reflections on the reality of transient mental illnesses. Charlottesville: University Press of Virginia. Hacking, I. (1999). The social construction of what? Cambridge, MA: Harvard University Press. Harré, R. (Ed.). (1986). The social construction of emotions. Oxford, UK: Blackwell. Healy, D. (2004). Let them eat Prozac: The unhealthy relationship between the pharmaceutical industry and depression. New York: New York University Press. Henningsen, P., & Kirmayer, L. J. (2000). Mind beyond the net: Implications of cognitive neuroscience for cultural psychiatry. Transcultural Psychiatry, 37(4), 467–494. Herman, J. (1992). Trauma and recovery. New York: Basic Books. Hinton, A. (Ed.). (1999). Biocultural approaches

94 HISTORICAL AND CULTURAL PERSPECTIVES to the emotions. Cambridge, UK: Cambridge University Press. Hinton, D., & Hinton, S. (2002). Panic disorder, somatization, and the new cross-Â�cultural psychiatry: The seven bodies of a medical anthropology of panic. Culture, Medicine, and Psychiatry, 26(2), 155–178. Hinton, D., Hinton, S., Pham, T., Chau, H., & Tran, M. (2003). “Hit by the wind” and temperature shift panic among Vietnamese refugees. Transcultural Psychiatry, 40(3), 342– 376. Horwitz, A. V., & Wakefield, J. C. (2007). The loss of sadness: How psychiatry transformed normal sorrow into depressive disorder. New York: Oxford University Press. Jackson, L. (2005). Surfacing up: Psychiatry and social order in colonial Zimbabwe, 1908– 1968. Ithaca, NY: Cornell University Press. James, W. (1890). The principles of psychology. New York: Holt. Keller, R. C. (2007). Colonial madness: Psychiatry in French North Africa. Chicago: University of Chicago Press. Kendell, R. E. (1989). Clinical validity. Psychological Medicine, 19, 45–55. Kilpatrick, D. G., Resnick, H. S., Freedy, J. R., Pelcovitz, D., Resick, P., Roth, S., et al. (1998). Posttraumatic stress disorder field trial: Evaluation of the PTSD construct—Â�Criterion A through E. In T. A. Widiger, A. J., Frances, H. A. Pincus, R. Ross, M. B. First, W. W. Davis, et al. (Eds.), DSM-IV sourcebook (Vol. 4, pp.€803–838). Washington, DC: American Psychiatric Association. Kinzie, J. D., & Goetz, R. R. (1996). A century of controversy surrounding posttraumatic stress-Â�spectrum syndromes: The impact on DSM-III and DSM-IV. Journal of Traumatic Stress, 9(2), 159–179. Kirmayer, L. J. (1988). Mind and body as metaphors: Hidden values in biomedicine. In M. Lock & D. Gordon (Eds.), Biomedicine examined (pp.€57–92). Dordrecht, The Netherlands: Kluwer. Kirmayer, L. J. (1994). Improvisation and authority in illness meaning. Culture, Medicine, and Psychiatry, 18(2), 183–214. Kirmayer, L. J. (1996). Confusion of the senses: Implications of ethnocultural variations in somatoform and dissociative disorders for PTSD. In A. J. Marsella, M. J. Friedman, E. T. Gerrity, & R. M. Scurfield (Eds.), Ethnocultural aspects of post-Â�traumatic stress disorders: Issues, research and clinical applications

(pp.€ 131–164). Washington, DC: American Psychological Association. Kirmayer, L. J. (2005). Culture, context, and experience in psychiatric diagnosis. Psychopathology, 38, 192–196. Kirmayer, L. J. (2006). Beyond the “new crossÂ�cultural psychiatry”: Cultural biology, discursive psychology and the ironies of globalization. Transcultural Psychiatry, 43(1), 126–144. Kirmayer, L. J. (2007a). Cultural psychiatry in historical perspective. In D. Bhugra & K. Bhui (Eds.), Textbook of cultural psychiatry (pp.€ 3–19). Cambridge, UK: Cambridge University Press. Kirmayer, L. J. (2007b). Psychotherapy and the cultural concept of the person. Transcultural Psychiatry, 44(2), 232–257. Kirmayer, L. J., Groleau, D., Guzder, J., Blake, C., & Jarvis, E. (2003). Cultural consultation: A model of mental health service for multicultural societies. Canadian Journal of Psychiatry, 48(3), 145–153. Kirmayer, L. J., & Jarvis, G. E. (2005). Depression across cultures. In D. Stein, A. Schatzberg, & D. Kupfer (Eds.), Textbook of mood disorders (pp.€ 611–629). Washington, DC: American Psychiatric Press. Kirmayer, L. J., Lemelson, R., & Barad, M. (Eds.). (2007). Understanding trauma: Integrating biological, clinical, and cultural perspectives. New York: Cambridge University Press. Kirmayer, L. J., & Minas, H. (2000). The future of cultural psychiatry: An international perspective. Canadian Journal of Psychiatry, 45(5), 438–446. Kirmayer, L. J., & Sartorius, N. (2007). Cultural models and somatic syndromes. Psychosomatic Medicine, 69(9), 832–840. Kirmayer, L. J., & Young, A. (1999). Culture and context in the evolutionary concept of mental disorder. Journal of Abnormal Psychology, 108(3), 446–452. Kitayama, S., & Markus, H. R. (1994). Emotion and culture: Empirical studies of mutual influence. Washington, DC: American Psychological Association. Kleinman, A. (1977). Depression, somatisation, and the new “cross-Â�cultural psychiatry.” Social Science and Medicine, 11, 3–10. Kleinman, A. (1986). Social origins of distress and disease: Depression, neurasthenia, and pain in modern China. New Haven, CT: Yale University Press.

On the Wisdom of Considering Culture and Context Kleinman, A. (1988). Rethinking psychiatry. New York: Free Press. Klerman, G. L. (1978). The evolution of a scientific nosology. In J. C. Shershow (Ed.), Schizophrenia: Science and practice (pp.€ 99–121). Cambridge, MA: Harvard University Press. Konner, M. (2007). Trauma, adaptation, and resilience: A cross-Â�cultural and evolutionary perspective. In L. J. Kirmayer, R. Lemelson, & M. Barad (Eds.), Understanding trauma: Biological, psychological and cultural perspectives (pp.€ 300–338). New York: Cambridge University Press. Leavitt, J. (1996). Meaning and feeling in the anthropology of emotions. American Ethnologist, 23(3), 514–539. Lee, S. (1998). Estranged bodies, simulated harmony, and misplaced cultures: Neurasthenia in contemporary Chinese society. Psychosomatic Medicine, 60, 448–457. Lerner, P. (2003). Hysterical men: War, psychiatry, and the politics of trauma in Germany, 1890–1930. Ithaca, NY: Cornell University. Leys, R. (1996). Traumatic cures: Shell shock, Janet, and the question of memory. In P. Antze & M. Lambek (Eds.), Tense past: Cultural essays in trauma and memory (pp.€ 103–145). New York: Routledge. Leys, R. (2000). Trauma: A genealogy. Chicago: University of Chicago Press. Littlewood, R. (1990). From categories to contexts: A decade of the “new cross-Â�cultural psychiatry.” British Journal of Psychiatry, 156, 308–327. Lutz, C., & White, G. M. (1986). The anthropology of the emotions. Annual Review of Anthropology, 15, 405–436. MacCorquodale, K., & Meehl, P. E. (1948). On a distinction between hypothetical constructs and intervening variables. Psychological Review, 55, 95–107. McCulloch, J. (1995). Colonial psychiatry and “the African mind.” Cambridge, UK: Cambridge University Press. McHugh, P. R., & Slavney, P. R. (1998). The perspectives of psychiatry (2nd ed.). Baltimore: Johns Hopkins University Press. McNally, R. (2003). Progress and controversy in the study of posttraumatic stress disorder. Annual Review of Psychology, 54, 229–252. Mezzich, J., Kirmayer, L. J., Kleinman, A., Fabrega, H., Jr., Parron, D. L., Good, B. J., et al. (1999). The place of culture in DSM-IV. Journal of Nervous and Mental Disease, 187(8), 457–464.

95 Mezzich, J., Kleinman, A., Fabrega, H., Jr., & Parron, D. (Eds.). (1996). Culture and psychiatric diagnosis. Washington, DC: American Psychiatric Press. Millon, T. (1991). Classification in psychopathology: Rationale, alternatives, and standards. Journal of Abnormal Psychology, 100(3), 245–261. Morey, L. C. (1991). Classification of mental disorder as a collection of hypothetical constructs. Journal of Abnormal Psychology, 100(3), 289–293. Oatley, K., & Jenkins, J. M. (1992). Human emotions: Function and dysfunction. Annual Review of Psychology, 43, 55–85. O’Nell, T. D. (1996). Disciplined hearts: History, identity, and depression in an American Indian community. Berkeley: University of California Press. Pole, N., Gone, J. P., & Kulkarni, M. (2008). Posttraumatic stress disorder among ethnoracial minorities in the United States. Clinical Psychology: Science and Practice, 15(1), 35–61. Rabinow, P., & Sullivan, W. M. (Eds.). (1987). Interpretive social science: A second look. Berkeley: University of California Press. Reddy, W. M. (2001). The navigation of feeling: A framework for the history of emotions. New York: Cambridge University Press. Robert, J. S. (2007). Gene maps, brain scans, and psychiatric nosology. Cambridge Quarterly of Health Care Ethics, 16, 209–218. Robert, J. S., & Plantikow, T. (2005). Genetics, neuroscience and psychiatry. Psychopathology, 38(4), 215–218. Robins, E., & Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric illness: Its application to schizophrenia. American Journal of Psychiatry, 126, 983–987. Rosaldo, M. Z. (1984). Toward an anthropology of self and feeling. In R. A. Shweder & R. A. LeVine (Eds.), Culture theory: Essays on mind, self, and emotion (pp.€137–157). Cambridge, UK: Cambridge University Press. Sadowsky, J. H. (1999). Imperial bedlam: Institutions of madness in colonial southwest Nigeria. Berkeley: University of California Press. Scott, W. (1990). PTSD in DSM-III: A case in the politics of diagnosis and disease. Social Problems, 37(3), 294–310. Shweder, R. A. (1988). Suffering in style. Culture, Medicine, and Psychiatry, 12, 479–497. Shweder, R. A. (1991). Thinking through cultures: Expeditions in cultural psychology. Cambridge, MA: Harvard University Press.

96 HISTORICAL AND CULTURAL PERSPECTIVES Shweder, R. A. (1993). The cultural psychology of the emotions. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions (pp.€417–431). New York: Guilford Press. Silove, D. (1999). The psychosocial effects of torture, mass human rights violations, and refugee trauma. Journal of Nervous and Mental Disease, 187(4), 200–207. Trimble, M. R. (1985). Post-Â�traumatic stress disorder: History of a concept. In C. R. Figley (Ed.), Trauma and its wake (pp.€ 5–14). New York: Brunner/Mazel. Vergès, F. (1996). To cure and to free: The Fanonian project of “decolonized psychiatry.” In L. R. Gordon, T. D. Sharpley-Â�Whiting, & R. T. White (Eds.), Fanon: A critical reader (pp.€85–99). Oxford, UK: Blackwell. Wexler, B. E. (2006). Brain and culture: Neurobiology, ideology, and social change. Cambridge, MA: MIT Press. White, G. M. (1993). Emotions inside out: The anthropology of affect. In M. Lewis & J. M. Haviland (Eds.), Handbook of emotions (pp.€29–39). New York: Guilford Press. Wierzbicka, A. (1999). Emotions across languages and cultures: Diversity and universals.€ Cambridge, UK: Cambridge University Press. Wilson, M. (1993). DSM-III and the transformation of American psychiatry: A history. American Journal of Psychiatry, 150(3), 399–410. Woodruff, R. A., Goodwin, D. W., & Guze, S. B. (1974). Psychiatric diagnosis. New York: Oxford University Press.

Yehuda, R., & McFarlane, A. C. (1995). Conflict between current knowledge about posttraumatic stress disorder and its original conceptual basis. American Journal of Psychiatry, 152(12), 1705–1713. Young, A. (1991). Emil Kraepelin and the origins of American psychiatric diagnosis. In B. Pfleiderer & G. Bibeau (Eds.), Curare: Vol. 7. Anthropologies of medicine: A colloquium on west European and North American perspectives (pp.€175–181). Dordrecht: Association of Ethnomedicine. Young, A. (1995). The harmony of illusions: Inventing posttraumatic stress disorder. Princeton, NJ: Princeton University Press. Young, A. (1996a). Bodily memory and traumatic memory. In P. Antze & M. Lambek (Eds.), Tense past: Cultural essays in trauma and memory (pp.€89–102). New York: Routledge. Young, A. (1996b). Suffering and the origins of traumatic memory. Daedalus, 125(1), 245– 260. Young, A. (2007). Bruno and the Holy Fool: Myth, mimesis and the transmission of traumatic memories. In L. J. Kirmayer, R. Lemelson, & M. Barad (Eds.), Understanding trauma: Integrating biological, clinical, and cultural perspectives (pp.€339–362). New York: Cambridge University Press. Zheng, Y. P., Lin, K. M., Takeuchi, D., Kurasaki, K. S., Wang, Y., & Cheung, F. (1997). An epidemiological study of neurasthenia in Chinese-Â�Americans in Los Angeles. Comprehensive Psychiatry, 38(5), 249–259.

Chapter 4

Cultural Issues in the Coordination of DSM-V and ICD-11 Renato D. Alarcón

ne of the most interesting epistemologiO cal developments in contemporary psychiatry is the increasing recognition of the

role played by culture and cultural factors in all aspects (clinical, diagnostic, therapeutic, and prognostic) of the entities we know as mental disorders. This realization is even more impressive if we consider that the renewed acceptance of cultural components is taking place in the midst of an undeniable predominance of neurobiological research in the field (Lopez-Ibor, Gaebel, Maj, & Sartorius, 2002). Many reasons have been invoked to explain the phenomenon: the need for a multidisciplinary contribution to the delineation and expression of symptoms, syndromes, and nosological labels; the pervasive findings of differences in the actual experience, description, and expression of clinical manifestations across different societies, historical periods, or geographic regions; and the impact of globalization, both as a social occurrence with complex characteristics, and as a political and economic event requiring better explanations and fairer applications in today’s world (Bauman, 1998; Stiglitz, 2002). Globalization is nourished by massive migrations and displace

ments of peoples and communities to and from entirely different milieus, for a variety of reasons (poverty, war, natural disasters, political exile, etc.). In turn, it leads to the realities of multiculturalism and of interfaces between cultures across the world—other sources of stress, as well as of hope (Simmons, 2002). The actual impact of globalization, migration, and multiculturalism is thus reflected in the unique context of the cultural background of each and every individual. If and when such experiences evolve into clinically detectable features, a cultural approach becomes essential in the description and characterization of the resulting diagnoses (Mastrogianni & Bhugra, 2003). The need to study and systematize the cultural aspects of all psychiatric conditions has been one of the main driving forces in the field of “cultural psychiatry.” Defined as the discipline that examines, among other topics, the context and meaning of morbid (abnormal or “pathological”) emotional behaviors along the lines of a true biopsychosocial model (Engel, 1977), cultural psychiatry uses a set of unique variables (language, religion, traditions, beliefs, ethnicity, gender, sexual orientation, etc.) to ascertain 97


the real (as opposed to stereotyped) nature of mental suffering. Thus culture definitely shapes a variety of clinical dimensions and plays a variety of roles in the assessment of any given patient; it is an interpretive/explanatory tool, a pathogenic/pathoplastic factor, a diagnostic/nosological instrument, a therapeutic/protective intervention, and a service/management component (Alarcón, Westermeyer, Foulks, & Ruiz, 1999). Cultural psychiatry therefore aims at preserving the essential integration of the human condition. It is the search for a genuinely monistic, truly comprehensive clinical assessment of each patient. The growth of cultural psychiatry, and its implications for the field in general, make it also necessary to ascertain what the discipline is not. Cultural psychiatry is not a psychiatric subspecialty; a “rehash” of old ideas; a political ploy; the study of just minorities, immigrants, or refugees; a collection of rhetoric; the observations of exotic people in distant lands; or an anti-Â�biological-Â�psychiatry movement (Alarcón, 1998, 2007). In parallel with its professed comprehensiveness, however, cultural psychiatry strives to define its boundaries, while maintaining that all human beings always show either hints or massive evidence of their cultural makeup—quite particularly while experiencing an illness, be that physical or mental (a distinction that unfortunately still permeates significantly the whole field of medicine). As such, it cannot evolve in isolation from other considerations of disease as a peculiar human status, especially those pertaining to mental or emotional conditions. The preceding concepts are useful in the examination of the role of culture in psychiatric diagnosis. There are no doubt significant differences between the purely nosological and the cultural perspectives in regard to psychiatric diagnosis. The former perspective is disease-Â�centered and individualized; the latter is person-Â�centered but also focuses on social networks. Nosology is “essentialist,” focused on specific, “purified” symptoms or diagnostic categories; culture is “contextualized,” integrative, interconnected (Kleinman, 1988). The nosological perspective tends to be based on biomedical technology, whereas the cultural perspective may be more oriented toward psychosocial components and interventions. Furthermore, most

of the medical specialties and subspecialties claim that in addition to specific symptomatology, they can count on generally reliable laboratory tests, technically known as “biomarkers,” to confirm or rule out any given diagnostic option under consideration (Follette & Houts, 1996). Clinical psychiatry, as is well known, is lacking in pathognomonic symptoms, and is still looking for solid biological markers; therefore, its approach is still mostly based on a detailed, more or less rigorous gathering of chronological history, collection of information from sources other than the patient him- or herself, determination of main and collateral symptoms, examination of clinical course, and the use of some measurement instruments. The concepts of “cultural competence” and “cultural fluency” must inform every attempt to introduce culture into the process of psychiatric diagnosis (Johnson, Saha, Arbelaez, Beach, & Cooper, 2004; Tseng & Strelzer, 2004). Although more frequently used in the field of treatment and provision of care, these terms point to skills and abilities that enhance communication, as well as to the context and background in which the diagnostic encounter takes place. Culturally oriented clinical inquiries lie behind clinical descriptions, biological measurements, and so-Â�called “socioenvironmental” factors, all operating as well in the diagnostic arena. Of all the areas of interaction between culture and psychiatry mentioned above, the one centered on diagnosis and nosology may be most in need of competent and serious attention from clinicians, researchers, and administrators alike. Although the notions of explanatory models of mental illness (Kleinman, 1988), behavioral styles influenced by both family microculture and broad environmental (cultural) factors, cultural competence on the part of providers (Tseng & Streltzer, 2004), and the presence of culturally based risk and protective indices (Kuh, Power, Blane, & Bartley, 1997) have been more or less accepted for a long time, the practical details of how culture affects diagnosis, and of how psychiatric diagnosticians can recognize and identify the cultural roots of their patients’ ailments, have only recently become topics of interest. It is generally agreed that only with the inclusion of cultural considerations, the “cultural formulation,” and the glossary of “culture-bound

Cultural Issues in the Coordination of DSM-V and ICD-11

syndromes” in the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association [APA], 1994) was the importance of cultural concepts in psychiatry duly recognized. Moreover, the actual utilization and implementation of these concepts since then have been insufficient, generating claims of “tokenism” (Kleinman, 2008) on the one hand, and misuse or plain neglect of a promising research and practice area on the other (Kirmayer & Young, 1999). After briefly reviewing the historical sources of diagnostic endeavors in psychiatry, this chapter examines the evolution and current status of the connections between culture and psychiatric diagnosis, followed by an assessment of inevitable political and organizational aspects of the equation. The chapter then delineates the different aspects of the process in need of coordination for both the APA’s DSM-V and the mental disorders section of the 11th revision of the World Health Organization’s (WHO’s) International Classification of Diseases (ICD-11). Outlining a diagnostic structure that includes a desirable cultural component, and key cultural issues that should be part of the overall diagnostic enterprise, is another objective of the chapter. The last section summarizes the main points of the preceding analyses.

Main Sources of the Psychiatric Diagnosis Process Historically, psychiatric diagnosis has relied on different bodies of knowledge, or study approaches, at different times. This has resulted from both the dominance (some would call it “fashionability”) of specific schools of thought at a given moment in history, and the political climate among leading national or international institutions in charge of this aspect of psychiatric work (Berrios, 1996). From this vantage point, it seems possible to identify up to five sources on which psychiatric diagnosis has relied to articulate different systems throughout the last 150–200 years—a period that most consider the “coming of age” of psychiatry as a well-defined field of study and clinical practice (Leighton, 1982). These epistemological sources are as follows:


1.╇ Phenomenology. A primary reliance on descriptive narratives of what the patient reports as his or her personal experiencing of “strange,” “unusual,” or “different” subjective phenomena has been probably the oldest and most widely used approach to diagnosis. Initially an exclusively intuitive modality, this process began to evolve when clinicians started to ascribe technical names to these occurrences, based on the different areas of psychological or behavioral functioning seemingly affected: perception, mood, thought, instincts or drives, personality, and so forth (Delgado, 1967). This was then followed by the systematic grouping of the symptoms so described, leading to syndromes and eventual nosological categories. The phenomenological approach to diagnosis had its philosophical origin in Husserl’s and other existential works (Spiegelberg, 1972), but its clinical application was ultimately free of theoretical links to this school. A naturalistic, neutral observation of the clinical picture was at the core of this type of diagnosis, first utilized in France, Italy, and Germany in the late 18th and early 19th centuries, and cogently promulgated by Jaspers’s General Psychopathology (published in Germany in 1913, and translated into English only in 1968). It seems only fair to acknowledge the phenomenological flavor of the versions of APA’s DSM from DSM-III onward. 2.╇ Psychodynamics. The need to “explain” symptoms has always been a driving force of every clinical endeavor, and an important part of the diagnostic task. Explanations would confer a “scientific” seal on otherwise esoteric and often terrifying experiences, thus contributing to a relative assuaging of symptom-Â�related fears. The failure of early explanatory attempts ranging from animistic or heavily religious to mechanistic ones (the latter were nourished by the strength of positivistic thinking during the “scientific revolution”; Delgado, 1947) shaped the psychodynamic approach of psychoanalytic authors inspired by Freud’s findings, their speculative, unproven nature notwithstanding (Chessick, 2003). The historical context of the Freudian contribution, particularly its “liberating” nature in the middle of a rigid Victorian era, was a factor as important in its success as the brilliance of its author’s literary style. It is


well known that Freud professed the hope that his explanations of subjective/unconscious phenomena would sooner or later be superseded by the findings of chemical and other biological methods of inquiry. In fact, psychodynamic “explanations” are, more precisely, interpretations of unexplained phenomena—a historically important ingredient of a diagnostic process with aspirations of true comprehensiveness. 3.╇ Epidemiology. The search for a better and more solid methodological base for scientific inquiries led to the growth and variety of epidemiological tools, which carried the banner of quantitative (and therefore, irrefutable) precision. Diagnosis jumped from an individual focus to a massive, populationÂ�oriented scale in research attempting to quantify phenomenologically obtained data on symptoms and their frequency, to define combinations of symptoms delineating clinical patterns, and to provide more consistent characterizations of syndromes and categories. Such research also seriously attempted to eliminate subjectivity in the description of symptoms. Although it sometimes paid the price of potential simplifications and/or premature generalizations, epidemiology provided useful perspectives on such important diagnostic components as gender, socioeconomic status, educational background, and the like (Richerson & Boyd, 2005). One of its soundest accomplishments was the demonstration of symptomatic hom*ogeneity of schizophrenia across cultures, as in the International Pilot Study of Schizophrenia (WHO, 1973), even though such findings were later subjected to solid critiques (Halliburton, 2004). As the “basic science” of public health, epidemiology also timidly but decisively hinted at the importance of context—a major aspect of the cultural approach (Van Ommeren, 2003). No diagnostic system or syndromic grouping nowadays can exist without substantial epidemiological justifications. 4.╇ Neurobiology. With the spectacular advances in multiple areas of neurobiological research in the last four or five decades, many now do not seem to doubt that a psychiatric diagnosis will only be both reliable and valid if and when it has a laboratory test, a genetic profile, neuroimaging documentation, or all of these together, to substantiate it. Nevertheless, this has been an

elusive objective so far, in spite of the overwhelming progress in genetics, biochemistry, pharmacology, neurophysiology, and other basic sciences, as well as the conceptual and heuristic elaborations of neuroplasticity, neurocircuitry, brain microbiology, and molecular biology (Guze, 1992; Kandel, 2006; Lopez-Ibor et al., 2002). There are no firmly established “biomarkers” in clinical psychiatry and psychiatric diagnosis so far, and the debates in the field reflect some of the frustrations as well as the high expectations generated by its undeniable progress. The field of genetics is touted as the one that will provide the information needed, but the search seems to be moving from candidate genes to a more or less broad compositum of genotypes whose precision will be based on better-Â�defined phenotypes or on reemerging endophenotypes (Eisenberg, 2005; Hasler, Drevets, Manji, & Charney, 2004). 5.╇ Informatics. With an extraordinary and ever-Â�growing number of measurement instruments of different natures, versions, characteristics, and patterns of symptoms as its initial endowment, the field of “informatics” may be the youngest source of a modern, more sophisticated process of psychiatric diagnosis (World Health Organization, 2001). Defined as a massive set of analyzable data and data banks, working at an almost quantum-like velocity and precision, informatics can perform meta-Â�analyses, correlating tasks, inferential associations, and many more operations. It utilizes both mathematical tools and operations, as well as logical deductive procedures, providing therefore both quantitative estimates and qualitative assessments. Although it cannot be used as the only source of diagnosis, informatics will be an essential component of diagnostic work in the not-so-Â�distant future. Reliance on the different sources for diagnosis outlined here entails richness and unmitigated promise. Nevertheless, they all have in common the omission of a clearly stipulated cultural component as part of psychiatric diagnosis. Although it may be tacitly included, or assumed to be included, the fact that it is not specifically dealt with may be a form of denial, a sign of arrogance, or a demonstration of ignorance or distraction. Let us hope that this neglect of cultural factors will be overcome in the diagnostic

Cultural Issues in the Coordination of DSM-V and ICD-11

and classification systems of the 21st century and beyond.

The Role of Culture in Psychiatric Diagnosis The history and development of the two main diagnostic systems in the world— APA’s DSM and WHO’s ICD—have similarities and differences, primarily related to the structure, nature, and institutional dynamics of their sponsoring organizations. The APA is a national professional organization that (due to a number of sociopolitical, financial, and demographic factors inherent in the position of the United States in the world) enjoys global influence, shares in the country’s scientific accomplishments, attracts the largest number of attendees to its annual meetings, leads the psychiatric publishing market, and has DSM as one of its most impressive signs of power. That English is the de facto scientific language of the world certainly does not hurt. The APA is politically and financially stable; its leadership changes in an organized fashion; it conducts its business in a transparent way; and, in spite of inevitable internal discrepancies and debates, it works within a predictable, hom*ogeneous frame. The WHO, on its side, is an international institution. Its work spans the globe but must respond to regional, national, and local bureaucracies and diverse political ideologies; its decision-Â�making processes are the results of complex and unavoidable negotiations; and its finances are supported by uneven contributions from its member nations. It has regional offices in different continents, and mental health in many cases is only part of broader divisional or sectional structures. Because of all these circ*mstances, the WHO frequently resorts to international consultants, advisory groups, or working committees. With the development of diagnostic and classification systems being among the main responsibilities of both organizations, it is clear that the APA conducts a more focused, more consistent, and better-Â�financed process. The WHO’s work in this area covers all kinds of diseases, and diagnoses for mental disorders constitute only one ICD section. Furthermore, the public health perspective is closer to the core of the WHO’s


(and therefore ICD’s) work, whereas DSM may have more clinical and actuarial purposes, including its attention to and use in legal and insurance-Â�related areas. A paradoxical reality seems to be that for the last 20–30 years, DSM (a national manual) has been more widely recognized and utilized in most countries of the world, even though political obligations dictate that disease and diagnostic codes follow the international nomenclature embodied in ICD. This has made it necessary to work on equivalencies or compatibilities between the two systems—a process that is certainly beneficial, and may have led to closer collaboration between the APA and the WHO’s Mental Health Division. Actually, a series of international conferences mostly funded by APA and sponsored by both organizations took place between 2005 and early 2008. The memberships of the APA’s DSM-V Committee and Work Groups have also been announced (APA, 2004, 2008). Interestingly, both organizations have ultimately relied on “expert consensus” for the elaboration of their respective systems, even though the APA supported some field trials prior to the publication of DSM-IV. A feature common to both nomenclatures is their “benign neglect” of the cultural perspective. It must be made clear that the WHO’s international vantage point is not necessarily cultural in its scope, structure, and substance. International psychiatry is not cultural psychiatry—Â�honest attempts to portray it as such, or political twists or spins, notwithstanding (Kirmayer, 2006). Incidence or prevalence data do not say much about cultural characteristics, explanatory models, or idioms of distress. On the APA’s side, DSM-IV was the first version of the manual that included cultural concepts, over 40 years after the publication of DSMI. And even in DSM-IV, the cultural formulation and culture-bound syndromes were relegated to Appendix I, at page 843 of the 886-page volume (APA, 1994). Under these circ*mstances, criticisms of both systems seem justified (Hughes, 1998; Kirmayer & Young, 1999; Kleinman, 2008). What are the reasons for this “lip service” that two otherwise relevant classification systems pay to the cultural cause? Why are cultural factors seen as collateral or negligible in diagnostic systems that call them-


selves “comprehensive” or “integrating”? The reasons are as complex in nature and origin as the questions themselves. First, the aim of any diagnostic catalog is to seek and emphasize hom*ogeneity, uniformization, generalization, and subsequent universal acceptance. Cultural approaches, epistemologically based on the principle of relativism, are seen as purposely searching for differences, uniqueness, or singularities in countries, regions, or societies across the world—Â�therefore preventing generalizations perceived as necessary, and nourishing heterogeneity, disagreement, and conflict (Bains, 2005). Second, it is not easy to articulate what aspects of the complex cultural realities of patients, families, and communities should be included in a multiaxial, descriptive, narrative, or multidimensional diagnostic system. Third, culture is considered by some either as epiphenomenal and superficial (because “it is there”)—more a motionless background curtain than a main actor in the diagnostic play—or as a politically inspired and therefore divisive issue. Fourth, research on cultural topics is regarded as “soft,” more the province of social sciences such as anthropology or sociology, and therefore lacking in clinical or diagnostic relevance. Fifth, psychiatrists and other mental health professionals interested in cultural issues (and, more specifically, in culture and diagnosis) have not produced research as abundant as that in other fields; have published it in a limited number of specialized journals; and have only recently strengthened their professional presence through national and international organizations, scientific meetings, and dissemination of their research efforts (Tseng, 2007). Some of the above-Â�described views certainly have elements of truth; others stem from the same types of arrogance, misunderstanding, or ignorance mentioned earlier. Some are the results of others’ actions or ideas; others, unfortunately, represent self-Â�inflicted damage. There seems to be agreement, however, that DSM-IV’s cultural formulation (APA, 1994) is the most satisfactory currently existing tool for the assessment of cultural topics in psychiatric diagnosis (Lewis-Â�Fernandez, 1996). Because it includes concrete information about cultural characteristics, cultural assumptions, explanatory models, style of interpersonal

transactions, and an overall evaluation of the patient’s cultural identity and culturally based behaviors, the cultural formulation seems to be a valuable tool, conceived as clinically useful and valid. The unfortunate reality, however, is that its utilization in clinical and teaching settings has been minimal; research on it has been equally scarce; and its value and even its existence have (again) been denied, minimized, or plainly ignored (Lewis-Â�Fernandez & Diaz, 2002). Its narrative (ethnography-based) structure and lack of quantifiable components have been pointed out as heuristic disadvantages, but the lack of consistency in its everyday use in clinical settings has been the main reason for this situation. Comparisons with the fate of the “psychodynamic formulation” of the 1960s and 1970s are not totally fair, since the latter’s popularity was related to the dominance of the psychodynamic school of thought in American psychiatry at the time—Â�something that cultural psychiatry has never sought or had. By the same token, the demise of the dynamic formulation was closely related to the emergence of DSM-III, which reintroduced a phenomenological, neo-Â�Kraepelinian approach to psychiatric diagnosis (Ghaemi, 2003). Although for entirely different reasons, both biological psychiatry and cultural psychiatry currently appear to be enjoying prominent positions in the preparation of new versions of DSM and ICD. Public pronouncements by leaders of both the APA and the WHO acknowledge the need to pay attention to cultural issues, and both organizations have in fact conducted international conferences to provide mutual input into both perspectives and the documents that will result from them (Phillips, First, & Pincus, 2003). It may be that DSM-V will become more deliberately “global” (i.e., culturally flexible) than its predecessors—Â� especially in view of the growing diversity of the U.S. population—and that the WHO will agree to incorporate more DSM-like structural and terminological components into ICD-11. In 2002, the APA published a volume titled A Research Agenda for DSMV, which expressly included a chapter on cultural issues (Alarcón et al., 2002). And the WHO is, by definition, naturally more receptive than a national organization to true cultural inputs.

Cultural Issues in the Coordination of DSM-V and ICD-11

Political Implications The pertinent organizational issues in both the APA and the WHO have already been mentioned. Their audiences, decisionÂ�making rules, views, and perspectives may differ, but if the purpose is to produce compatible documents, negotiations and agreements will ideally favor this goal. For obvious reasons, the APA seems to be moving more decisively; as noted above, it has established a DSM-V Committee or Task Force, Work Groups, and Study Groups, and it has also announced a timetable for the publication of the manual (APA, 2008). With the assistance of the U.S. National Institute of Mental Health, and the collateral participation of the World Psychiatric Association, the APA is demonstrating vigor in this process. Its powerful networks and communication instruments (plus global electronic systems) assure rapid, agile exchange of information. This is indeed a promising process. Political disagreements, ideological differences, and historical animosities will require both the APA and the WHO to make goodfaith efforts at understanding each other’s positions and constituencies, to exhibit flexibility in recognizing differences, to acknowledge diversity, to set aside ill-Â�conceived or narrow “-isms” (including nationalisms), and to arrive at a consensus resulting from both clinical/research-based and valuebased evidence (Baca-Â�Baldomero & Lázaro, 2005; Ruiz, Alarcón, Lolas, Lázaro, & Baca-Â�Baldomero, 2008). As noted earlier, the aims and roles of the two organizations are different (but ideally complementary). The APA provides a pragmatic vision, the WHO a social/public health perspective; the APA has substantial resources (both human and financial), while the WHO has a wellÂ�established global setting. The APA works in a country that still enjoys superpower status (with all its positive and negative implications), while dealing with the diversity generated by migration and globalization; the WHO deals with a complex and at times chaotic world audience in the unquestionably altruistic pursuit of public health in general and public mental health in particular. Both groups—again perhaps for different reasons—agree, however, that culture is important, and that it has to be accepted and included both conceptually and practi-


cally in any new psychiatric diagnostic classification. Other factors to keep in mind include the phenomenon of globalization, as well as the political changes of the last 30–40 years throughout the world. Globalization includes, among its multiple implications, an emotional impact of drastic proportions related to massive displacements of people, acculturation processes, economic opportunities gained and lost, new notions of family structure and relationships, new communication patterns, and so forth (Bauman, 1998; Lacroix & Shrad, 2004; Stiglitz, 2002). Cultural factors cannot be ignored in the evaluation of individuals and communities showing psychiatric symptoms that adopt different shapes and expressions from those usually seen in mainstream Western societies (Kirmayer & Young, 1999; Kleinman, 1988; Kuh et al., 1997). Political changes determined by any of numerous factors— the rise or fall of ideologies, religious wars, neocolonialism, the wealth–Â�poverty chasm in most countries of the world, natural disasters, or environmental conflicts—all have an undeniable cultural stamp as well: They are culturally based occurrences inducing, in turn, somewhat unpredictable cultural changes. And, once again, the psychological impact of these changes is undeniable. The relativism inherent to a cultural assessment of everything human may be an excellent tool in the elaboration of the new diagnostic systems in psychiatry. If diagnoses are going to be accurate, integrated, and thorough, culture has to be present. If the diagnoses are going to include etioÂ�pathogenesis, cultural factors must be included. If the diagnoses are going to determine comprehensive treatment strategies, cultural settings must be anticipated. If the diagnoses are going to predict or intuit outcomes, cultural assertions have to be used. Finally, if the diagnoses are going to summarize the experience of human beings, culture—the most human of products—must be part of their texture. Those responsible for the new systems will have to adjust their working calendars to these considerations. DSM-V and ICD-11 offer unique opportunities for a genuine inclusion of cultural factors in the generation of each and every psychiatric diagnosis. Together with a substantial dose of goodwill,


clinical and research strategies and verifications must be well coordinated with the scope and structure of the new nosological catalogs.

Coordination Areas and Issues If two new diagnostic systems are going to exist (or start to coexist) within the next 4–5 years, an intense dialogue will have to continue between the two organizations involved in the process. In this section, I examine the main areas or topics in more or less urgent need of discussion and coordination, still from the cultural perspective. The main goals will be recognizing these areas; understanding their definitions, scope, and boundaries; manualizing and instrumentalizing these areas; and adapting them to the overall structure of the system. These areas and their related issues include the following: 1.╇ Settings. The need for a universal scope for a diagnostic system cannot ignore the fact that countries, geographic regions, communities, and societies across the world differ in a variety of ways. Therefore, it is inevitable that different versions or adaptable arrangements will have to be considered. The differences in health structures and health care systems must be major factors in these considerations. 2.╇ Diversity. The people in these different settings carry their cultural legacies through generations, and any diagnostic system looking for relevance and cogency has to address this human diversity. “People” refers, of course, to both patients and care providers— who either share a culture, or come from different cultures but attempt to find common ground in the description, explanation, and understanding of clinical realities. 3.╇ Clinical variables. Issues of crucial clinical importance (other than the symptoms themselves) that are to be part of the process of diagnosis in psychiatry should include the following: •• Informant or informants (including relationship with the patient, educational level, reliability, etc.). •• Help-Â�seeking pattern or modality (including level and manner of information

regarding the health care setting, circ*mstances of contact, way of reporting clinical manifestations, expectations, body language, etc.) (Rogler, Malgady, & Constantino, 1987). •• Compliance history (in the case of previous clinical and treatment contacts) or compliance potential (in the case of a first visit). •• Perception of severity of symptoms and other clinical manifestations. •• Impact of clinical condition on the patient’s family group or other surrounding relationships or settings (neighborhood, workplace, church, friendships, community organizations, etc.). •• Coping style (including resilience, vulnerabilities, individual and socially based risk and protective factors, characterological strengths and weaknesses, survival strategies, etc.). •• Overall organization of the clinical report, with emphasis on eventual cultural correlations of the gathered material. 4.╇ Specific cultural variables. Coordinating efforts must be intense and sustained to make the primary cultural component of both DSM-V and ICD-11 a solid and wellÂ�accepted feature (Group for the Advancement of Psychiatry, 2002). In close connection with the above-Â�described clinical variables (already embodying clearly defined cultural qualities), the efforts must put special emphasis on the following variables: •• Demographics. Age, gender, educational level, occupation, and socioeconomic status are demographic items of undeniable cultural relevance, as is the setting in which the evaluation encounter takes place. All of them should be adequately described and utilized in the diagnostic assessment and subsequent clinical steps. •• Race and ethnicity. Although also “demographic” in nature, these two items have a unique cultural value: race as an expression of inadequate groupings of human beings on the basis of variable, often ambiguous, and highly politicized physiognomic characteristics; ethnicity as cultural legacy, anchor of membership and participation in larger social systems, and core feature of individual and collective identities (Beneduce & Martelli, 2005).

Cultural Issues in the Coordination of DSM-V and ICD-11

•• Language and terminology. Because language is the most distinctive tool of human communication, all its variations and components (dialects, jargon, rules, accent, inflections, complexity, expressiveness, meaning, etc.) offer the most direct route toward culturally charged information. In the clinical realm, there is sometimes a special terminology that clinicians must learn to understand in order to make thorough cultural sense out of clinical occurrences. •• Religion. Together with language, religion is probably the cultural variable with the deepest, most distinctive roots. It addresses collective concepts about transcending ideas and beliefs about the origin and end of life and nature, perception of virtue and vice, spiritual purity and sinfulness, a set of “explanations about the unexplainable,” and management of intriguing or enigmatic situations or entities, including health and illness. It provides abundant information about the relationship between the individual and his or her environment, on the basis of rules and traditions that may be used to justify a variety of both normal and pathological behaviors. It also entails the “mentality” of groups and communities (Alarcón et al., 2002; Group for the Advancement of Psychiatry, 2002; Hutton, 1981). •• History gathering. Aspects of this include informants and their style of providing data, areas of emphasis, spontaneous causal explanations, levels of closeness and/or identification with the patient’s plight, and so on. •• Context. This includes data on the physical and human/personal characteristics of the clinical setting, where and how the clinical history and symptoms evolved, sequence of events, initial reactions or management of the situation, and so forth. Context also covers the very important areas of locus of control and expressed emotions (Sontag, 1978). •• Meaning. Any culturally oriented clinical interview must inquire about the personal, unique, and intimate meaning of the experience of illness and its different manifestations, as well as the significance and implications of the illness for life, future, family, and personal fate. Symptoms as physical or subjective occurrences,


and€the words used to describe them, do have a potent cultural message to be deciphered. •• Explanatory model(s). In many cases, the patient and other informants provide their own versions of events, the reasons for the appearance of symptoms, and even opinions about what to do to correct them, on the basis of the causes invoked (which may range from the quite Westernized “chemical imbalance” to explanations based on exotic animistic phenomena or divine punishment) (Alarcón, 1990; Kleinman, 1988; Villaseñor, 2008). If explanatory models are not mentioned spontaneously, the clinician must ask specifically about them, as their cultural content may be decisive. Needless to say, a conflict in explanatory models between the provider and the patient or family should be avoided or negotiated, to prevent diminished rapport, lack of support, or stigmatization. •• Dimensions of suffering. These refer not only to areas of physical, emotional, or behavioral pain, but to the deeper and more meaningful dimensions of authentic human suffering (reflected in changes in self-image, self-Â�perception, quality of life, relationship with family, God and religion, values and traditions, fate issues, etc.). A related concept is “social desirability,” which is inversely related to individual suffering: The less desirable a given symptom or behavior is within an individual’s culture, the more likely he or she is to be stigmatized and ostracized (Weatherhill, 1991). •• Manuals and instruments. As part of all clinical evaluations, the cultural assessment should be substantiated by the use of specific measurement tools and instruments. These measures can cover some of the above-Â�described demographic/statistical variables, but should also include such areas as acculturation, stress levels, personality factors, and quality of life. Fortunately, various cultural measurement instruments are now available in today’s mental health and clinical settings. •• Treatment implications. Like any other type of clinical evaluation, a culturally oriented psychiatric evaluation must include a variety of treatment implications. Information on race, ethnicity, language, religion, and other cultural variables will


lead to appropriate therapeutic indications that take them into account: individual, group, family, and/or couple therapy with cultural emphasis, and even biological modalities if newly developed techniques such as pharmacogenomic testing (with growing recognition of ethnic differences) are applied systematically and judiciously (Alarcón & Mrazek, 2007). 5.╇ Culture-bound syndromes. The importance of this group of disorders—Â� reflected in the vast existing literature, and even in the inclusion of an abbreviated list in DSM-IV—makes them deserving of special comment. “Culture-bound syndromes” are defined as unique clinical expressions of cultural idiosyncrasies, with mostly (if not exclusively) cultural determinants, playing a critical role in the context of individual, family, community, and social tapestries. The culture-bound syndromes have had a stormy history among official or institutional nosologists and diagnosticians, while being enthusiastically hailed by culturally oriented clinicians (Hughes, 1996; Tseng, 2006). Although such syndromes were initially restricted to non-Â�Western societies, some authors now claim that they may indeed also exist in the West, particularly in the form of eating disorders or “new” somatoform disorders (Johnson et al., 2004; Littlewood & Lipsedge, 1986). Whether or not this is the case—and whether culture-bound syndromes can be absolutely (and smoothly) incorporated into existing or updated nomenclatures (without falling into what Kleinman [1988] calls “category fallacy”), or whether they still constitute a Â�special class of disorders—they indicate how much work remains to be done in the field.

Overall Structure Efforts to coordinate the two main diagnostic systems in the world will demand more than simply agreement on terms, categories, and criteria. Such efforts must also extend to the full integration of cultural aspects into the structures of DSM-V and ICD-11. Acceptance and recognition of cultural content across symptoms and syndromes, spectra, or well-�defined clinical entities must go beyond the mere declarative nature of political

or bureaucratic agreements. The inclusion of cultural elements in psychiatric diagnosis could be all too easily accomplished with the inclusion of a “cultural axis” in any future multiaxial or multilayered diagnosis. Nevertheless, objections to this type of addition are strong, growing, and well justified. The main one is the isolation in which cultural factors would find themselves if restricted to a single axis; the risks of neglect, minimization, or weakness would be quite realistic. Such a restriction would also go against the cardinal ecumenical feature of what is cultural: Everything human is cultural, and any symptom experienced by any human being in any part of the world has a cultural ingredient that is important to consider as part of the “whole” person, not just to note within a separate compartment. Furthermore, the risks of scarcity or excess of cultural data to be considered relevant for diagnosis would be ever-Â�present; boundaries would be difficult to ascertain or reinforce. A cultural dimension, if well delineated and quantifiable (or measurable), could serve better the stated purposes of recognition, value, and clinical usefulness of a diagnosis. As suggested earlier in this chapter, the interpretive/explanatory, pathogenic/pathoplastic, therapeutic/protective, and service/ management roles of culture (Alarcón et al., 1999) can assist in consolidating a cogent diagnostic/nosological role for it. If agreements could be reached about key cultural variables—such as context, meaning, explanatory models, and impact of the morbid condition, as well as language, eventual migration or acculturative history, religion, and management implications—a dimension of cultural influence could then be measured in terms of length and depth, thus providing pragmatic information about treatment emphases and possible outcome predictions (Kraemer, 2007; Lopez, Compton, Grant, & Breiling, 2007). This would allow also an estimate of the cultural “weight” of specific symptoms, syndromes, or diagnostic categories (if a sort of mixed dimensionalÂ�categorical model ends up being accepted by the makers of the systems), with many positive implications for future classifications (Alarcón & Foulks, 1995). Although the assertion that “everything is cultural” still retains value, a dimensional approach would provide a realistic assessment of the

Cultural Issues in the Coordination of DSM-V and ICD-11

relevance of cultural factors in any individual diagnosis. The suggestion of a cultural dimension does not mean that there are no objections to the dimensional approach. The same objectives can be made to this approach as to the idea of a single cultural axis, although less strongly. Furthermore, some commentators prefer using the term “environmental” to mean “cultural and everything else that is not biological”; however, this runs an even greater risk of neglect or oblivion. Conversely, the dimensional view can exaggerate, or even distort, the impact of cultural factors in psychopathology through excessive scoring or measurements, overemphasis on specific variables, or biasing of clinicians’ perceptions. The old labels of “irrelevant,” “generic,” “too subjective,” or “soft” can be applied to cultural research data. There are good reasons to agree that different clinical conditions vary in terms of their “cultural load.” Schizophrenia, bipolar disorders, and autism are justifiably considered the most biologically rooted disorders in modern psychiatry. This does not mean that cultural factors do not play a role in the pathogenesis, symptom expression, or clinical course of these conditions; actually, such factors are crucial in their management, outcome, and ultimate prognosis. On the other hand, it is fair to say at this point that most of the mental diagnostic conditions included in DSM-IV or ICD-10 (and, obviously, in their forthcoming versions) do include significant cultural components in their origins, and therefore must be subjects of culturally aimed diagnostic efforts. Such conditions as somatization disorder, and personality disorders, depression and its variants (e.g., bereavement), some anxiety disorders, and others have already been the topics of studies in different parts of the world; the results are revealing, rich, and varied, going beyond the simplified findings of large epidemiological inquiries (Gaw, 2001; Tseng, 2001). Whatever the cultural content of DSM-V and ICD-11 turns out to be, it is important to keep in mind that the inclusion of cultural factors does not guarantee their universal acceptance or practical use in clinical work. For both conceptual and pragmatic reasons, it will be important to keep the cultural formulation of DSM-IV in the future editions of the manual, and to include some equiva-


lent in ICD-11. Criticisms that the use of the formulation may result in fragmentation, dichotomization, simplification, stereotyping, and impersonalization have been, in a few cases, well justified. A few studies have attempted to demonstrate the cultural formulation’s usefulness and relevance, while standardization, user-Â�friendliness, and enlarged databases have been called for (Group for the Advancement of Psychiatry, 2002; Lewis-Â�Fernandez, 1996; Lewis-Â�Fernandez & Diaz, 2002). The text and content of the formulation could be enhanced, taking into account the rich variety of items examined here and in other publications; a quantitative version could also be devised. Above all, however, concerted efforts to publicize the formulation and to utilize it with all kinds of clinical populations (generating, in turn, concomitant useful research data sets) would be important steps toward guaranteeing a healthy presence of cultural factors in the field of psychiatric diagnosis.

Conclusions There is ample evidence for a growing understanding and acceptance of the value of culture and cultural factors in the generation of a comprehensive and accurate psychiatric diagnosis. Even biologically oriented clinicians and researchers acknowledge this value and admit the need to add a cultural perspective to the diagnostic process. Indeed, there is a lot of truth in Kirmayer’s (2006) assertion that psychiatric diagnosis is in fact a cultural interpretation of a person’s (patient’s) experience (symptoms). Despite this new level of respectability, however, the inclusion of cultural factors in diagnosis still encounters obstacles in areas of definition and terminology, efforts to operationalize such inclusion, and limited (until recently) coordination of initiatives by responsible organizations. The examination of all the political implications of such coordinated efforts should be followed by well-Â�delineated work, paying attention to pervasive clinical and culturally proper variables and perspectives. Discussions have included (and should continue to include) addressing issues of cultural validity, threshold definitions, compatibility, and a consistent focus on cultural “fitness” of the forthcoming systems (Harper, 2001;


Kirmayer & Minas, 2000; Knapp & Jensen, 2006). In turn, cultural variables should be part of a more dimensional approach to clinical conditions; cultural factors should be measured in an objective manner; and tools such as DSM-IV’s cultural formulation should be improved. Although research in cultural psychiatry requires further and more vigorous development, the current state of knowledge is promising. In the area of psychiatric diagnosis, field studies should go beyond the descriptive or comparative modus operandi to utilize cultural approaches that embrace epidemiological and anthropological principles, while creating a descriptive account of the illness experience, its meaning, and the context and cultural determinants of occurrence, course, and outcome (Van Ommeren, 2003)—all of them diagnostic elements of the first order. Through the combined use of clinical, ethnographic, and even experimental studies, cultural case ascertainment and culturally standardized assessment instruments (including adapted psychological tests) would help in dimensionalizing a culturally based diagnosis. The coordinated efforts to make DSM-V and ICD-11 more culturally relevant—Â�substantiated by joint research efforts with adequate cross-Â�cultural methodology—will undoubtedly result in a better diagnostic process, and therefore in more appropriate treatments and solid outcomes for persons with mental illnesses across the world. References Alarcón R. D. (1990). Identidad de la psiquiatría Latinoamericana: Voces y exploraciones en torno a una ciencia solidaria. México City: Siglo XXI Editores. Alarcón, R. D. (1998). What cultural psychiatry isn’t. Psychline, 2, 27–28. Alarcón, R. D. (2007). Psiquiatría folklorica, etnopsiquiatría o psiquiatría cultural?: Examen crítico de las perspectivas de C. A. Seguín. Revista Latinoamericana de Psiquiatria, 5, 8–15. Alarcón, R. D., Bell, C. C., Kirmayer, L. J., Lin, K. M., Ustun, B., & Wisner, K. L. (2002). Beyond the funhouse mirrors: Research agenda on culture and psychiatric diagnosis. In D. J. Kupfer, M. B. First, & D. A. Regier (Eds.), A

research agenda for DSM-V (pp.€ 219–289). Washington, DC: American Psychiatric Association. Alarcón, R. D., & Foulks, E. F. (1995). Personality disorders and culture: Contemporary clinical views (Part B). Cultural Diversity and Mental Health, 1, 79–91. Alarcón, R. D., & Mrazek, D. A. (2007). Fármacogenomica psiquiátrica: Actualización y perspectivas. In H. Silva (Ed.), Genética y farmacogenómica en psiquiatría (pp.€59–79). Santiago de Chile: Ediciones SONEPSYN. Alarcón, R. D., Westermeyer, J., Foulks, E. F., & Ruiz, P. (1999). Clinical dimensions of contemporary cultural psychiatry. Journal of Nervous and Mental Disease, 187, 465–471. American Psychiatric Association (APA). (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association (APA). (2008). Announcement of the establishment of the DSM-V Committee and Work Groups [Press release)]. Washington, DC: Author. Baca-Â�Baldomero, E., & Lázaro, J. (Eds.). (2005). Hechos y valores en psiquiatría. Madrid: Triacastela. Bains, J. (2005). Race, culture, and psychiatry: A history of transcultural psychiatry. History of Psychiatry, 16, 139–154. Bauman, Z. (1998). Globalization: The human consequences. New York: Columbia University Press. Beneduce, R., & Martelli, P. (2005). Politics of healing and politics of culture: Psychiatry, identity, and migration. Transcultural Psychiatry, 42, 367–393. Berrios, G. E. (1996). History of mental symptoms. Cambridge, UK: Cambridge University Press. Chessick, R. (2003). Psychoanalytic peregrination: VI. The effect on countertransference of the collapse of stabilization. Journal of the American Academy of Psychoanalysis, 31, 541–562. Delgado, H. (1947). Acerca de la psicopatología general de Jaspers. Revista de NeuroÂ�Psiquiatría, 10, 413–421. Delgado, H. (1967). Curso de psiquiatría (6th€ ed.). Barcelona: Editorial CientíficoÂ�Medica. Eisenberg, L. (2005). Are genes destiny?: Have adenine, cytosine, guanine and thymine replaced Lachesis, Clotho and Atropos as the weavers of our fate? World Psychiatry, 4, 3–8. Engel, G. L. (1977). The need for a new medical

Cultural Issues in the Coordination of DSM-V and ICD-11 model: A challenge for biomedicine. Science, 196, 129–132. Follette, W., & Houts, A. (1996). Models of scientific progress and the role of theory in taxonomy development: A case study of the DSM. Journal of Consulting and Clinical Psychology, 64, 1120–1132. Gaw, A. C. (2001). Concise guide to crossÂ�cultural psychiatry. Washington, DC: American Psychiatric Press. Ghaemi, S. (2003). The concepts of psychiatry: A pluralistic approach to the mind of mental illness. Baltimore: Johns Hopkins University Press. Group for the Advancement of Psychiatry, Committee on Cultural Psychiatry. (2002). Cultural assessment in clinical psychiatry. Washington, DC: American Psychiatric Press. Guze, S. E. (1992). Why psychiatry is a branch of medicine. New York: Oxford University Press. Halliburton, M. (2004). Finding a fit: Psychiatric pluralism in South India and its implications for WHO studies of mental disorder. Transcultural Psychiatry, 41, 80–98. Harper, G. (2001). Cultural influences on diagnoses in children and adolescents. Psychiatric Clinics of North America, 24, 711–728. Hasler, G., Drevets, W. C., Manji, H. K., & Charney, D. S. (2004). Discovering endophenotypes for major depression. Neuropsychopharmacology, 27, 1–17. Hughes, C. C. (1996). The culture-bound syndromes and psychiatric diagnoses. In J. E. Mezzich, A. Kleinman, H. Fabrega, & D. L. Parron (Eds.), Culture and psychiatric diagnoses: A DSM-IV perspective (pp.€289–305). Washington, DC: American Psychiatric Press. Hughes, C. C. (1998). The glossary of “culturebound syndromes” in DSM-IV: A critique. Transcultural Psychiatry, 35, 413–421. Hutton, P. H. (1981). The history of mentalities: The new map of cultural history. History and Theory, 20, 237–259. Jaspers, K. (1968). General psychopathology. Chicago: University of Chicago Press. (Original work published 1913) Johnson, R. L., Saha, S., Arbelaez, J. J., Beach, M. C., & Cooper, L. A. (2004). Racial and ethnic differences in patient perceptions of bias and cultural competence in health care. Journal of General Internal Medicine, 19, 101–110. Kandel, E. (2006). In search of memory. New York: Norton.


Kirmayer, L. J. (2006). Beyond the “new crossÂ�cultural psychiatry”: Cultural biology, discursive psychology and the ironies of globalization. Transcultural Psychiatry, 43, 126–144. Kirmayer, L. J., & Minas, H. (2000). The future of cultural psychiatry: An international perspective. Canadian Journal of Psychiatry, 45, 438–446. Kirmayer L. J., & Young, A. (1999). Culture and context in the evolutionary concept of mental disorder. Journal of Abnormal Psychology, 108, 446–452. Kleinman, A. (1988). Rethinking psychiatry: From cultural category to personal experience. New York: Free Press. Kleinman, A. (2008). The normal, the pathological, and the existential [Editorial]. Comprehensive Psychiatry, 49, 111–112. Knapp, P., & Jensen, P. S. (2006). Recommendations for DSM-V. In P. S. Jensen, P. Knapp, & D. A. Mrazek (Eds.), Toward a new diagnostic system for child psychopathology: Moving beyond the DSM (pp.€ 162–181). New York: Guilford Press. Kraemer, H. C. (2007). DSM categories and dimensions in clinical and research contexts. International Journal of Methods in Psychiatric Research, 16(Suppl. 1), S8–S15. Kuh, V., Power, C., Blane, D., & Bartley, M. (1997). Social pathways between childhood and adult health. In D. Kuh & Y. Ben-Â�Shlomo (Eds.), A life course approach to chronic disease epidemiology (pp.€ 169–200). Oxford, UK: Oxford University Press. Lacroix, M., & Shrad, G. E. (2004). Introduction: Globalization and community mental health. Canadian Journal of Community Mental Health, 23, 5–21. Leighton, A. H. (1982). Caring for mentally ill people: Psychological and social barriers in historical context. Cambridge, UK: Cambridge University Press. Lewis-Â�Fernandez, R. (1996). Cultural formulation of psychiatric diagnosis. Culture, Medicine, and Psychiatry, 20, 133–144. Lewis-Â�Fernandez, R., & Diaz, N. (2002). The cultural formulation: A method for assessing cultural factors affecting the clinical encounter. Psychiatric Quarterly, 73, 271–295. Littlewood, R., & Lipsedge, M. (1986). The “culture bound syndromes” of the dominant culture: Culture, psychopathology and biomedicine. In J. L. Cox (Ed.), Transcultural psychiatry (pp.€ 253–273). London: Croom Helm.

110 HISTORICAL AND CULTURAL PERSPECTIVES Lopez, M. F., Compton, W. N., Grant, B. F., & Breiling, J. P. (2007). Dimensional approaches in diagnostic classification: A critical appraisal. International Journal of Methods in Psychiatric Research, 16(Suppl. 1), S6–S7. Lopez-Ibor, J. J., Gaebel, W., Maj, M., & Sartorius, N. (Eds.). (2002). Psychiatry as a neuroscience. New York: Wiley. Mastrogianni, A., & Bhugra, V. (2003). Globalization, cultural psychiatry, and mental distress. International Journal of Social Psychiatry, 49, 163–165. Phillips, K. A., First, M. B., & Pincus, H. A. (Eds.). (2003). Advancing DSM: Dilemmas in psychiatric diagnosis. Washington, DC: American Psychiatric Association. Richerson, P. J., & Boyd, R. (2005). Not by genes alone: How culture transformed human evolution. Chicago: University of Chicago Press. Rogler, L. H., Malgady, R. G., & Constantino, G. (1987). What do culturally sensitive mental health services mean? American Psychologist, 49, 565–570. Ruiz, P., Alarcón, R. D., Lolas, F., Lázaro, J., & Baca-Â�Baldomero, E. (2008). Role of Spanish speaking psychiatry in a globalized world. Manuscript under revision. Simmons, A. (2002). Mondialisation et migration international: Tendencies, interrogations et modeles theoriques. Montréal: Cahiers Quebecois de Demographie. Sontag, S. (1978). Illness as a metaphor. New York: Farrar, Straus & Giroux. Spiegelberg, H. (1972). Phenomenology in psy-

chology and psychiatry: A historical introduction. Evanston, IL: Northwestern University Press. Stiglitz, J. (2002). Globalization and its discontents. London: Penguin Books. Tseng, W. S. (2001). Handbook of clinical psychiatry. San Diego, CA: Academic Press. Tseng, W. S. (2006). From peculiar psychiatric disorders through culture-bound syndromes to culture-Â�related specific syndromes. Transcultural Psychiatry, 43, 554–576. Tseng, W. S. (2007). Role of the World Association of Cultural Psychiatry. Paper presented at the inaugural meeting of the World Association of Cultural Psychiatry, Beijing. Tseng, W. S., & Streltzer, J. (Eds.). (2004). Cultural competence in clinical psychiatry. Washington, DC: American Psychiatric Publishing. Van Ommeren, M. (2003). Validity issues in transcultural epidemiology. British Journal of Psychiatry, 182, 376–378. Villaseñor, S. J. (2008). Apuntes para una etnopsiquiatria Méxicana. Guadalajara, México: Imprenta Universitaria. Weatherhill, R. (1991). The psychical realities of modern culture. British Journal of Psychotherapy, 7, 268–277. World Health Organization (WHO). (1973). The international pilot study of schizophrenia. Geneva: Author. World Health Organization Commission on Macroeconomics and Health. (2001). Macroeconomics and health: Investing in health for economic development. Geneva: Author.

Chapter 5

A Sociocultural Conception of the Borderline Personality Disorder Epidemic Theodore Millon

he interface between psychopathology Tcomplex and sociocultural contexts is extremely and variable. Numerous biologi-

cal, psychological, and cultural factors give shape to the development of personality characteristics and general psychopathology. Symptomatological pictures presented by patients are multidetermined in an intricately interwoven pattern of causality. The ways in which social and cultural influences give rise to the traits and disorders of personality are many, complicated, and divergent. Sociocultural processes can be sources of psychic stress themselves, as well as giving form and quality to the nature of life experiences. In addition, social influences furnish a framework for how individuals will learn to cope with the nature of their distress. The manner in which clinical symptoms will be interpreted is also a product of social and cultural value systems. No less relevant is the influence of culture in guiding how patients seek assistance therapeutically, as well as the modes of treatment that characterize the orientations of society’s healers. The value systems and nosological schemas generated in Western societies, such as those seen in the ICD-10 (World Health Organization, 1992) and the DSM-IV (Ameri-

can Psychiatric Association, 1994), reflect particular models of thought that may be at variance with the models of numerous cultures and subcultures around the world. More specifically, the Western schemas are oriented specifically to an individual’s personal experiences and are largely grounded in a medical model of infectious disease—a model driven by contemporary biomedical technologies and, more particularly, those of pharmacological therapies. Western cultural perspectives contrast with numerous cultural orientations that are centered more on social contexts and relational networks, as well as interpersonal methods of intervention that reflect broad-Â�ranging social health systems. Care must be taken, therefore, not to generalize findings generated in Western concepts and research to those of other cultures (Alarcón, Foulks, & Vakkur, 1998). The mind is shaped by the institutions, traditions, and values that make up the cultural context of societal living. Methods by which social rules and regulations are transmitted between people and from generation to generation are often emotionally charged and erratic, entailing persuasion, seduction, coercion, deception, and threat. Feelings of anxiety and resentment are generated as a 111


function of environmental stress, leaving pathological residues that linger and serve to distort future relations. Bear in mind, however, that cultural and social conditions do not directly shape the mind or “cause” disorders; rather, they serve as a context for the more direct and immediate experiences of interpersonal and family life. They not only color but may degrade personal relationships, establishing maladaptive styles of coping and pathogenic models for imitation. In this chapter, I address in part forces that characterize “society as the patient,” as Frank (1948) suggested over six decades ago: Instead of thinking in terms of a multiplicity of so-Â�called social problems, each demanding special attention and a different remedy, we can view all of them as different symptoms of the same disease. That would be a real gain even if we cannot entirely agree upon the exact nature of the disease. If, for example, we could regard crime, mental disorders, family disorganization, juvenile delinquency, prostitution and sex offenses, and much that now passes as the result of pathological processes (e.g., gastric ulcer) as evidence, not of individual wickedness, incompetence, perversity or pathology, but as human reactions to cultural disintegration, a forward step would be taken. (p.€42)

In much the same way as the paired strands of the DNA double helix unwind, and each strand then selects environmental nutrients to duplicate its discarded partner, so too does each culture fashion its constituent members to fit an existing template. Those societies whose customs and institutions are fixed and traditional will generally produce a psychically structured and formal citizenry, and those societies whose values and practices are highly fluid and inconsistent are likely to evolve so that their citizens develop deficits in psychic solidity and stability. Though it would be naïve to state that sociocultural factors are directly responsible for individual pathology, responsible scientists would be lax if they ignored the role of cultural dynamics in the well-being of individuals. Whether or not one is inclined to attribute the increased incidence in youthful depressive disorders in Western countries, for example, to the waxing and waning of population parameters, it is both intuitively

and observationally self-�evident that sweeping cultural changes can affect innumerable social practices. Among those social practices most evidently affected by cultural dynamics are those of an immediate and personal nature, such as patterns of child nurturing and rearing, marital affiliation, family cohesion, leisure style, entertainment content, diminution of the role of organized religion, and so on. Numerous demographic, political, and international changes, as well as advances in science and technology in recent decades, have stimulated a growing awareness of the role of sociocultural factors in public health and illness prevention (Brody, 1990; Kleinman, 1980, 1988a, 1988b; Littlewood & Lipsedge, 1987; Mezzich, Kleinman, & Fabrega, 1996). I focus in this chapter, however, on some of the sociocultural elements of experience affecting the lives of those who have been diagnosed as having borderline personality disorder (BPD).

Alternative Views Concerning the BPD Epidemic Despite its exegetic brilliance, I am concerned that the Talmudic habit of intricate and abstruse argument, most notably within the psychoanalytic community, has drawn us into recondite intellectual territories, leading us to overlook the impact of palpable social forces generative of BPD. Though less labyrinthine and tortuous, the database for biological conjectures favoring constitutional origins is similarly equivocal, if not controvertible. Unfortunately, the riveting effects of internecine struggles between and within competing analytic and biological theories keep us fixated on obscure and largely unfalsifiable etiological hypotheses, precluding thereby serious consideration of sociocultural notions that may possess superior logic and validity. It is toward this latter goal that the present chapter is addressed. By no means is the role of constitution or early experience to be dismissed. Nevertheless, the logical and evidential base for currently popular hypotheses is briefly examined and left wanting. Supplementary proposals that favor the deductive and probative primacy of social and cultural factors are posited. They point to a wide army of

A Sociocultural Conception of the Borderline Personality Disorder Epidemic

influences that either set in place or further embed those deficits in psychic cohesion that lie at the heart of the disorder. Specifically, the view has been advanced that our contemporary epidemic of BPD can best be attributed to two broad sociocultural trends that have come to characterize much of Western life this past half century: first, the emergence of social customs that exacerbate rather than remediate early, errant parent–child relationships; and, second, the diminished power of formerly reparative institutions to compensate for these ancient and ubiquitous relationship problems. To raise questions about either the validity or adequacy of one or another theoretical interpretation is not to take issue with all aspects of its formulations; much of what has been proposed concerning the nature and origins of the BPD epidemic has both substantive merit and heuristic value. One may offer a compelling, if not persuasive, alternative and yet be entirely sympathetic to the major tenets of that which has been critically examined. Such is the case in what follows; the alternative proposed is more in the nature of an addendum than a supplantation. Hence I begin the discussion with references to views shared rather than disputed. Before noting themes in common, I should say a few additional words of précis about the faddistic character of BPD. It is, in my judgment, overly diagnosed, having become a “wastebasket” for many patients who demonstrate the protean constellation of multiple symptoms that characterize the syndrome. Exhibiting almost all of the clinical attributes known to descriptive psychopathology, BPD and related conditions lend themselves to a simplistic if not perverse form of diagnostic logic: That is, patients who display a potpourri of clinical indices, especially where symptomatic relationships are unclear or seem inconsistent, must perforce have BPD. Having stated my dismay over contemporary diagnostic fashions and fallibilities, let me record my firm belief in both the clinical existence and the significantly increased incidence of BPD and related conditions these past three or four decades. Whether they are voguish or not, I believe that the emergence of these disorders is a “real” phenomenon—Â� albeit, in my view, not for the reasons advanced in contending analytic and biological theories.


Agreement on the Features of€BPD Overdiagnosed and elusive as the syndrome has been—Â�approached from innumerable and diverse analytic perspectives (e.g., Eriksonian, Kernbergian, Mahlerian, Kohutian), as well as clothed in an assortment of novel conceptual terms (identity diffusion, self–Â�object representations, projective identification)—there remain, nevertheless, certain shared observations that demonstrate the continued clinical astuteness and heuristic fertility of the analytic mindset. Whatever doubts one may have with respect to either the logical or methodological merit of conjectures posed by our modem cadre of third-Â�generation analytic thinkers (Eagle, 1984), they deserve more than passing commendation, not only for their willingness to break from earlier and perhaps anachronistic etiological notions, but for their perspicacity in discerning and portraying the key features of a new and major clinical entity. Similarly, contemporary biogenic theorists (e.g., Akiskal & Akiskal, 1992; Klein, 1971) are to be congratulated for the care with which they have adduced empirical data favoring their views. Overlooking for the present seemingly intractable conflicts among and within analytic, biological, and social learning schools of thought (Millon, 1987), let me note what I believe to be the key features of BPD that contemporary theorists of each orientation appear to judge salient and valid (Millon, 1996). There is a reasonable consensus that pervasive instability and ambivalence intrude constantly into the everyday lives of persons with BPD, resulting in fluctuating attitudes, erratic or uncontrolled emotions, and a general capriciousness and undependability. Such persons are impulsive, unpredictable, and often explosive; it is difficult for others to be comfortable in their presence. Both relatives and acquaintances feel “on edge,” waiting for these patients to display a sullen and hurt look or become obstinate and nasty. In being unpredictably contrary, manipulative, and volatile, people with BPD often elicit rejection rather than the support they seek. Displaying marked shifts in mood, they may exhibit extended periods of dejection and apathy, interspersed with spells of anger, anxiety, or excitement.


Dejection, depression, and self-Â�destructive acts are common. These patients’ anguish and despair are genuine, but they are also means of expressing hostility, a covert instrumentality to frustrate and retaliate. Angered by the failure of others to be nurturant, persons with BPD employ moods and threats as vehicles to “get back” at them or to “teach them a lesson.” By exaggerating their plight and by moping about, these individuals avoid responsibilities and place added burdens on others, causing their families not only to care for them, but to suffer and feel guilt while doing so. In the same way, cold and stubborn silence may function as an instrument of punitive blackmail, a way of threatening others that further trouble is in the offing. Easily nettled, offended by trifles, persons with BPD are readily provoked into being sullen and contrary. They are impatient and irritable, unless things go their way. Cognitively capricious, these persons may exhibit rapidly changing and often antithetical thoughts concerning themselves and others, as well as the odds and ends of passing events. They voice dismay about the sorry state of their lives, their sadness, their resentments, their “nervousness.” Many feel discontented, cheated, and unappreciated; their efforts have been for naught; they have been misunderstood and are disillusioned. The obstructiveness, pessimism, and immaturity that others attribute to them are only reflections, they feel, of their “sensitivity” and the inconsiderateness that others have shown. But here again, ambivalence intrudes; perhaps, they say, their own unworthiness, their own failures, and their own “bad temper” are the causes of their misery and the pain they bring to others. The affective and interpersonal instability of persons with BPD may be traced in great measure to their defective psychic structures—their failure to develop internal cohesion and hierarchical priorities. Both a source and a consequence of this lack of inner harmony is their uncertain sense of self, the confusions they experience of either an immature, nebulous, or wavering sense of identity. Hence the deeper structural undergirding for intrapsychic regulation and interpersonal processing provides an inadequate scaffolding for both psychic continuity and self-Â�integration. Segmented and fragmented,

subjected to the flux of their own contradictory attitudes and enigmatic actions, their very sense of being remains precarious. Their erratic and conflicting inclinations continue as both causes and effects, generating new experiences that feed back and reinforce an already diminished sense of wholeness.

A Supplementary Sociocultural€Thesis Although the logic of biogenic conjectures justifies critical examination, the following comments are limited to the philosophical and empirical grounding of psychoanalytic theories, owing to their far greater currency among contemporary clinicians. The premise that early experience plays a central role in shaping personality attributes is one shared by analytic and social learning theorists. To say the preceding, however, is not to say that these two groups agree on which specific factors during these developing years are critical in generating particular attributes; nor is it to note any agreement that known formative influences are either necessary or sufficient. Analytic theorists almost invariably direct their etiological attention to the realm of early childhood experience. Unfortunately, they differ vigorously among themselves (e.g., Kernberg, Kohut, Mahler, Masterson, Erikson) as to which aspects of nascent life are crucial to development. The preceding critique of the epistemic and probative foundations of our discipline’s theoretical assertions may seem unduly severe, or perhaps even discourteous to analytic colleagues whose views have been singled out for critical note. Despite appearances, however, I consider myself a loyal if peripheral “follower”—one not only sympathetic to both historic and contemporary psychoanalytic formulations, but one whose own views have been informed by and continue to rest on the foundations they provide. Thus, despite retaining aspects of my earlier conceptions of the “borderline levels” of personality pathology (Millon, 1969, 1981, 1996), I have restructured my ideas significantly in recent years, owing to the thoughtful contributions of numerous contemporary analytic writers. What follows is not intended, therefore, to supplant the core notions proposed by pres-

A Sociocultural Conception of the Borderline Personality Disorder Epidemic

ent-day analytic theorists concerning the experiential background of persons with BPD, despite these theorists’ own divergences and the unfalsifiability of their propositions. Rather, it should be seen as an addendum— a proposal that societal customs that served in the past to repair disturbances in early parent–child relations have declined in their efficacy, and have been “replaced” over the past several decades with customs that exacerbate these difficulties, contributing thereby to what may be termed our contemporary BPD “epidemic.” Two central questions guide this commentary. First, what are the primary sources of influence that give rise to the symptoms distinguishing BPD—namely, an inability to maintain psychic cohesion in realms of affect, self-image, and interpersonal relationships? Second, which of these sources has had its impact heightened over the past few decades, accounting thereby for the rapid and marked increase in the incidence of the disorder? The first question is not elaborated here, other than to note that well-Â�reasoned yet contending formulations have been posited by constitutional, analytic, and social learning theorists, and that despite important divergences, there is a modest consensus that biogenic, psychogenic, and sociogenic factors each contribute in relevant ways. It is the second question that calls for explication. It relates to which of these three etiological factors productive of the diffuse or segmented personality structure characteristic of BPD—constitutional disposition, problematic early nurturing, or contemporary social changes—has shown a substantial shift in recent decades. Is it some unidentified yet fundamental alteration in the intrinsic biological makeup of present-day youngsters? Is it some significant and specifiable change in the ways in which contemporary mothers nurture their infants and rear their toddlers? Or is it traceable to fundamental and rapid changes in Western culture that have generated divisive and discordant life experiences, while reducing the availability of psychically cohering and reparative social customs and institutions? Despite the fact that tangible evidence favoring one or another of these possibilities is not accessible in the conventional sense of empirical “proof,” it is my contention that


the third “choice” is probatively the most sustainable and inferentially the most plausible. Toward these ends, two sociocultural trends generative of the segmented psychic structures that typify persons with BPD are elucidated. Although they are interwoven substantively and chronologically, these trends are separated for conceptual and pedagogical purposes. One adds to the severity of psychic dissonance; it appears to have been on the upswing. The other has taken a distinct downturn, and its loss also contributes to diminished psychic cohesion.

Increasingly Divisive Social€Customs We are immersed deeply in both our time and our culture. This immersion obscures our ability to discern many profound changes that may be underway in our society’s institutions—changes often generative of unforeseen psychic and social consequences, Tom Wicker, a distinguished columnist for the New York Times, portrays sequential effects such as these in the following graphic passage: When a solar-Â�powered water pump was provided for a well in India, the village headman took it over and sold the water, until stopped. The new liquid abundance attracted hordes of unwanted nomads. Village boys who had drawn water in buckets had nothing to do, and some became criminals. The gap between rich and poor widened, since the poor had no land to benefit from irrigation. Finally, village women broke the pump, so they could gather again around the well that had been the center of their social lives. (1987, p.€23)

Not all forms of contemporary change can so readily be reversed. “Progress” wrought by modem-day education and technology is too powerful to be turned aside or nullified, much less reversed, despite conservative efforts to revoke or undo their inexorable effects. Over two decades ago, Klerman (1987), reviewing the increased incidence of depressive disorders among young people, contended that an as-yet-Â�unidentified cohort effect (which he termed “Agent Blue”) might be operative. Inclined to ascribe this pathological drift to Easterlin and Crimmins’s (1985)


thesis that the baby boom had created a deterioration in economic possibilities among contemporary youth, Klerman noted that this group was raised in relative affluence and is physically the most healthy population in history; nevertheless, they paradoxically were (and are) suffering an epidemic of depressive and other mental disorders. Whether or not one assigns the increased incidence of youthful depressive disorders to the waxing and waning of population parameters, it is both intuitively and observationally self-Â�evident that sweeping cultural changes can affect innumerable social practices—Â�including those of an immediate and personal nature, such as patterns of child nurturing and rearing, marital affiliation, family cohesion, leisure style, entertainment content, and so on. I turn next to contemporary changes such as these, narrowing the focus to those cultural transitions conducive to the formation of psychic diffusion and division. Mirrored Social Discordance It would not be too speculative to assert that the organization, coherence, and stability of a culture’s institutions are in great measure reflected in the psychic structure and cohesion of its members. As noted earlier, in a manner analogous to the DNA double helix—in which each paired strand unwinds and selects environmental nutrients to duplicate its jettisoned partner—so too does each culture fashion its constituent members to fit an extant template. In societies whose customs and institutions are fixed and definitive, the psychic composition of its citizenry will likewise be structured; and where a society’s values and practices are fluid and inconsistent, so too will its residents evolve deficits in psychic solidity and stability. This latter, more amorphous cultural state, so characteristic of our modern times, is clearly mirrored in the interpersonal vacillations and affective instabilities that typify BPD. Central to our recent culture have been the increased pace of social change and the growing pervasiveness of ambiguous and discordant customs to which children are expected to subscribe. Under the cumulative impact of rapid industrialization, immigration, urbanization, mobility, technology, and mass communication, there has been

a steady erosion of traditional values and standards. Instead of a simple and coherent body of practices and beliefs, children find themselves confronted with constantly shifting styles and increasingly questioned norms whose durability is uncertain and precarious. No longer do youngsters find the certainties and absolutes that guided earlier generations. The complexity and diversity of everyday experience play havoc with simple, “archaic” beliefs, and render them useless as instruments to deal with contemporary realities. Lacking a coherent view of life, maturing youngsters find themselves groping and bewildered, swinging from one set of principles and models to another, unable to find stability either in their relationships or in the flux of events. At few times in history have so many children faced the tasks of life without the aid of accepted and durable traditions. Not only does the strain of making choices among discordant standards and goals beset them at every turn, but these competing beliefs and divergent demands prevent them from developing either internal stability or external consistency. And no less problematic in generating such disjoined psychic structures is the escalation of emotionally capricious and interpersonally discordant role models. Schismatic Family Structures Although transformations in family patterns and relationships have evolved fairly continuously over the past century, the speed and nature of transitions since World War II have been so radical as to break the smooth line of earlier trends. Hence children today typically no longer have a clear sense of either the character or the purpose of their fathers’ work activities, much less a detailed image of the concrete actions constituting that work. Beyond the little they know about their fathers’ daily routines to model themselves after, mothers of young children have shifted their activities increasingly outside the home, seeking career fulfillments or needing dual incomes to sustain family aspirations. Not only are everyday adult activities no longer available for direct observation and modeling, but traditional gender roles, once distinct and valued, have become blurred and questionable. Today, little that

A Sociocultural Conception of the Borderline Personality Disorder Epidemic

is rewarded and esteemed by the larger society takes place where children can see and emulate it. What “real” and “important” people do cannot be learned from parents who return from a day’s work too preoccupied or too exhausted to share their esoteric activities. Lost, then, are the crystallizing and focusing effects of identifiable and stable role models, which give structure and direction to maturing psychic processes. This loss contributes significantly to the maintenance of the undifferentiated and diffuse personality organization so characteristic of many persons with BPD. With the growing dissolution of the traditional family structure, there has been a problematic increase in parental separation, divorce, and remarriage. Children subject to persistent parental bickering and family restructuring not only are exposed to changing and destructive models for imitative learning, but develop the internal schisms characteristic of BPD. The stability of life, so necessary for the acquisition of a consistent pattern of feeling and thinking, is shattered when erratic conditions or marked controversies prevail. There may be an everÂ�present apprehension that a parent will be totally lost through divorce; dissension may also lead to the undermining of one parent by the other, and a nasty and cruel competition for the loyalty and affections of children may ensue. Constantly dragged into the arena of parental schisms, a child not only loses a sense of security and stability, but is subjected to paradoxical behaviors and contradictory role models. Raised in such a setting, the child not only suffers the constant threat of family dissolution, but in addition is often forced to serve as a mediator to moderate conflicts between the parents. Forced to switch sides and divide loyalties, the child cannot be a coherent individual, but must internalize opposing attitudes and emotions to satisfy antagonistic parental desires and expectations. The different roles the child must assume to placate the parents are markedly divergent: As long as the parents remain at odds, the child persists with behaviors, thoughts, and emotions that are intrinsically irreconcilable. For many children, divorce not only undermines their sense that they can count on things to endure, but it often dislodges formerly secure and crucial internaliza-


tions within their psychic selves, upsetting the fusions and integrations that evolved among once-Â�incorporated parental models and standards. Because they are alienated from parental attachments, as well as often disillusioned and cynical, these internalized structures may now be totally jettisoned. Moreover, the confidence that they can depend in the future on a previously internalized belief or precept may now be seriously undermined. Devoid of stabilizing internalizations, such youngsters may come to prefer the attractions of momentary and passing encounters of high salience and affective power. Unable to gauge what to expect from their environment, how can they be sure that things that are true today will be there tomorrow? Have they not experienced capriciousness when things appeared stable? Unlike children who can predict their fate (good, bad, or indifferent), such youngsters are unable to fathom what the future will bring. At any moment, and for no apparent reason, they may receive the kindness and support they crave; at any other moment, and for equally unfathomable reasons, they may be the recipient of hostility and rejection. Having no way of determining which course of action will bring security and stability, such youngsters vacillate—Â�feeling hostility, guilt, compliance, assertion, and so on, and shifting erratically and impulsively from one tentative action to another. Unable to predict whether their parents will be critical or affectionate, they must be ready for hostility when most might expect commendation, or assume humiliation when most would anticipate reward. Because they are eternally “on edge,” their emotions build up and become raw to the touch. They are ready to react impulsively and unpredictably at the slightest provocation. Capricious TV Models Other “advances” in contemporary society have stamped deep and distinct impressions on children’s psyches as well—ones equally effectively loaded, erratic, and contradictory. The rapidly moving, emotionally intense, and interpersonally capricious character of TV role models, displayed in swiftly progressing 30- or 60-minute vignettes that encompass a lifetime, add to the impact of disparate, highly charged, and largely dam-


aging value standards and behavior models. What is incorporated is not only a multiplicity of selves, but an assemblage of unintegrated and discordant roles, displayed indecisively and fitfully, especially among those youngsters bereft of secure moorings and internal gyroscopes. The striking images created by our modern-day flickering parental surrogate have replaced all other sources of cultural guidance for many; by age 18, the typical American child will have spent more time watching TV than in going to school or relating directly to his or her parents. TV may be nothing but simple pablum for those with comfortably internalized models of real human relationships, but for those who possess a world of diffuse values and standards, or one in which parental precepts and norms have been discarded, the impact of these “substitute” prototypes is especially powerful, even idealized and romanticized. And what these TV characters and story plots present to vulnerable youngsters are the stuff of which successful “life stories” must be composed to capture the attention and hold the fascination of their audiences—Â� violence, danger, agonizing dilemmas, and unpredictability, each expressed and resolved in an hour or less. These are precisely the features of social behavior and emotionality that come to characterize the affective and interpersonal instabilities of persons with BPD. Mind-Â�Blurring Drugs Adding to this disorienting and cacophonous melange are aggravations consequent to drug and alcohol involvements. Although youth is a natural period for exploratory behaviors, of which many are both socially adaptive and developmentally constructive, experimental substance use entails high risks and can have severe adverse consequences in both the short and long run. However, from the perspective of youngsters who see little in life that has proven secure or desirable, the risks of using these all-too-Â�accessible substances appear minimal. Although many adolescents may consider their substance use casual and recreational, the psychic effects can be quite hazardous, especially among youth who are already vulnerable. Thus for those prone to BPD, the impact of these sub-

stances will only further diminish the clarity and focus of their feeble internalized structures, as well as dissolving whatever purposefulness and aspirations they may have possessed to guide them toward potentially reparative actions. Together, these mind�blurring effects exacerbate already established psychic diffusions.

Decreases in Reparative Social€Customs As noted previously, the task of attaining personal integration is not an easy one in a world of changing events and practices. What is best? What is right? How shall I handle this or think about that? Questions such as these plague growing children at every turn. On top of these everyday perplexities, how can children subjected additionally to parental hostility or indifference acquire guidelines for a well-Â�integrated and socially approved system of beliefs and actions? From what source can a consistent, valued, and effective set of internalized feelings, attitudes, and relationships be consolidated? The fabric of traditional and organized societies not only comprises standards designed to educate and socialize the young; it also provides “insurance,” if you will—Â�backups to compensate for or repair system defects and failures. Extended families, church leaders, schoolteachers, and neighbors provide nurturance and role models through which children experiencing troubling parental relationships can find means of support and affection, enabling them thereby to become receptive to society’s established body of norms and values. Youngsters subject to any of the diffusing and divisive forces described previously must find one or another of these culturally sanctioned sources of surrogate modeling and sustenance to give structure and direction to their emerging capacities and impulses. Without such bolstering, maturing potentials are likely to become diffuse and scattered. Without admired and stable roles to emulate, such youngsters are left to their own devices to master the complexities of their varied and changing worlds, to control the intense aggressive and sexual urges that well up within them, to channel their fantasies, and to pursue the goals to which

A Sociocultural Conception of the Borderline Personality Disorder Epidemic

they may aspire. Many become victims of their own growth, unable to discipline their impulses or find acceptable means for expressing their desires. Scattered and unguided, they are unable to fashion a clear sense of personal identity, a consistent direction for feelings and attitudes, a coherent purpose to existence. They become “other-Â�directed” persons who vacillate at every turn—Â�overly responsive to fleeting stimuli, shifting from one erratic course to another. Ultimately, without the restitutive and remedial power of beneficent parental surrogates, they fail to establish internalized values to anchor themselves and to guide their future. This aimless floundering and disaffiliated stagnation may be traced in part to the loss in contemporary society of various meliorative and reparative customs and institutions, which once “made sure” that those who had been deprived or abused would have a second chance by being exposed to compensatory sponsors and institutions exhibiting values and purposes around which social life could be focused and oriented. I next turn to these losses. Decline of Consolidating Institutions The impact of much of what has been described previously might be substantially lessened if concurrent or subsequent personal encounters and social customs were compensatory or restitutive—that is, if they repaired the intrapsychically destabilizing and destructive effects of problematic experiences. Unfortunately, the converse appears to be the case. Whereas the cultural institutions of most societies have retained practices that furnish reparative stabilizing and cohering experiences, thereby remedying disturbed parent–child relationships, my thesis in this chapter is that the changes of the past several decades have not only fostered an increase in intrapsychic diffusion and splintering, but have also resulted in the discontinuation of psychically restorative institutions and customs, contributing thereby to both the incidence and exacerbation of features that typify BPD pathology. Without the corrective effects of undergirding and focusing social mentors and practices, the diffusing or divisive consequences of unfavorable earlier experience take firm root and


unyielding form, displaying their structural weaknesses in clinical signs under the press of even modestly stressful events. One of the by-Â�products of the rapid expansion of knowledge and education is that many of the traditional institutions of our society—such as religion, which formerly served as a refuge to many, offering “love” for virtuous behavior and caring and thoughtful role models—have lost much of their historic power as a source of nurturance and control in the contemporary Western world. Similarly, and in a more general way, the frequency with which families in our society relocate has caused a wide range of psychically diffusing problems. Families not only leave behind stability, but with each move jettison a network of partially internalized role models and community institutions, such as those furnished in church, school, and friendships. Undone thereby are the psychic structure and cohesion that could have been solidified to give direction and meaning to otherwise disparate elements of existence. Not only do children who move to distant settings feel isolated and lonely in these unfamiliar surroundings, and not only are they deprived of opportunities to develop a consistent foundation of social customs and a coherent sense of self, but what faith they may have had in the merits of holding to a stable set of values and behaviors can only have been discredited. Disappearance of Nurturing€Surrogates The scattering of the extended family, as well as the rise of single-Â�parent homes and shrinkage in sibling number, adds further to the isolation of families as they migrate in and out of transient communities. Each of these undermines the once-Â�powerful reparative effects of kinship support and caring relationships. Contemporary forms of disaffection and parent–child alienation may differ in their particulars from those of the past. But rejection and estrangement have been and are ubiquitous, as commonplace as rivalry among siblings. In former times when children were subjected to negligence or abuse, they often found familial or neighborly parental surrogates—Â�grandmothers, older sibs, aunts or uncles, or even a kind


childless couple down the street who would, by virtue of their own needs or identifications, nurture or even rear them. Frequently more suitable to parental roles, typically more affectionate and giving as well as less disciplinary and punitive, these healing surrogates have historically served not only to repair the psychic damage of destructive parent–child relationships, but to “fill in” the requisite modeling of social customs and personal values that youngsters so treated are receptive to imitate and internalize. In the past several decades, however, estranged and denigrated children have much less often found nurturing older sibs, grandparents, or other relatives; nor are the once accessible and nurturing neighbors now available. With increased mobility, kinship separation, single parenting, and reduced sibling numbers, our society has few surrogate parents to pick up the pieces of what real parents may have fragmented and discarded, much less to restore the developmental losses engendered thereby. Reemergence of Social Anomie For some, the question is not which of the changing social values they should pursue, but whether there are any social values that are worthy of pursuit. Youngsters exposed to poverty and destitution; provided with inadequate schools; living in poor housing set within decaying communities; raised in chaotic and broken homes; deprived of parental models of success and attainment; and immersed in a pervasive atmosphere of hopelessness, futility, and apathy cannot help questioning the validity of the “good society.” Children reared in these settings quickly learn that there are few worthy standards to which they can aspire successfully. Whatever efforts they may make to raise themselves above these bleak surroundings run hard against the painful restrictions of poverty, the sense of a meaningless and empty existence, and an indifferent if not actually hostile world. Moreover, and in contrast to earlier generations whose worlds rarely extended beyond the shared confines of ghetto poverty, the disparity between everyday realities and what is seen as so evidently available to others in enticing TV commercials and bounteous shopping malls is not only

frustrating, but painfully disillusioning and immobilizing. Why make a pretense of accepting patently “false” values or seeking the unattainable goals of the larger society, when reality undermines every hope, and social existence is so pervasively hypocritical and harsh? Nihilistic resolutions such as these leave youngsters bereft of a core of inner standards and customs to stabilize and guide their future actions, exposing them to the capricious power of momentary impulse and passing temptation. Beyond being merely “anomic” in Durkheim’s sense of lacking socially sanctioned means for achieving culturally encouraged goals, these youngsters have incorporated neither the approved customs and practices nor the institutional aspirations and values of our society. In effect, they are both behaviorally normless and existentially purposeless—Â�features seen in manifest clinical form among individuals with prototypal BPD. Loss of Compelling Causes As with the good fortunes that were anticipated upon the arrival of the Indian village’s solar-Â�powered water pump, so too have the “blessings” consequent to our modern-day standard of living produced their share of troublesome sequelae. Until a generation or two ago, children had productive, even necessary economic roles to fill within the family. More recently, when the hard work of cultivating the soil or caring for the home was no longer a requisite of daily life, youngsters were encouraged to advance their family’s fortunes and status via higher education and professional vocations. Such needed functions and lofty ambitions were not only internalized, but gave focus and direction to a child’s life—Â�creating clear priorities among values and aspirations, and bringing disparate potentials into a coherent schema and life philosophy. Coherent aspirations are no longer commonplace today, especially among the children of the middle classes. In contrast to disadvantaged anomic youngsters, they are no longer “needed” to contribute to their families’ economic survival; on the other hand, neither can upwardly mobile educational and economic ambitions lead them to readily

A Sociocultural Conception of the Borderline Personality Disorder Epidemic

surpass the achievements of already successful parents. In fact, children are seen in many quarters today as economic burdens, not as vehicles to a more secure future or to a more esteemed social status. Parents absorbed in their own lives and careers frequently view children as impediments to their autonomy and narcissistic indulgences. But even among children who are not overtly alienated or rejected, the psychically cohering and energizing effects of being needed or of fulfilling a worthy family aspiration have been lost in our times. Without genuine obligations and real purposes to create intent and urgency in their psychosocial worlds, such youngsters often remain diffused and undirected. At best, they are encouraged to “find their own thing,” to follow their own desires and create their own aims. Unfortunately, freedoms such as these translate for many into freedom to remain in flux, to be drawn to each passing fancy, to act out each passing mood, to view every conviction or ethic as being of equal merit— and ultimately to feel evermore adrift, lost, and empty. Satisfying each momentary wish, consuming pleasures once shrouded in mystery, today’s youngsters have nonetheless been deeply deprived not of material wants, but of opportunities to fulfill both the minor daily chores now routinely managed by modern technology and the more distant goals that kept the minds of yesterday’s children centered around a value hierarchy and oriented toward ultimate achievements. Although many of this generation have their bearings in good order, some have submerged themselves in aimless materialism. Others remain adrift in disenchantment and meaninglessness, a state of disaffected malaise. Some have attached themselves to naive causes or cults that ostensibly provide the passion and purpose they crave to give life meaning, but even these solutions too often prove empty, if not fraudulent. In earlier times, the disenchanted and disenfranchised, often dislocated from burdened homes or cast out as unwelcome troublemakers, joined together in active protest and rebellion. Problematic economic, social, and political conditions were shared by many profoundly dissatisfied individuals, who formed philosophical movements such as the German Sturm und Drang of


the late 18th century or the Wandervogels of the late 19th. The 20th century as well has witnessed similar though more benign movements stemming from an antipathy to parental ideals and cultural norms, such as the “beat” and “hippie” generations of the second half of the century. Children in today’s Western societies, however, are “rebels without a cause.” Whereas earlier generations of disaffected youth were bound together by their resentments, opposed to economic or political oppression, or motivated by other discernible and worthy common causes that provided both group camaraderie and a path to action, today’s middle-class youngsters have no shared causes to bring them together. Materially well nourished and clothed, unconstrained in an open society, they can follow their talents and aspirations freely. The purposelessness and emptiness they experience are essentially internal matters, private rather than collective affairs, with no external agents against whom or which they can join with others to take to the streets. I contend that these rebels without a cause—Â� unable to forgo the material comforts of home and ineffective in externalizing their inner discontents upon the larger scene, yet empty and directionless—Â�comprise a goodly share of today’s population with BPD. Were it not for the general political, economic, and social well-being for many until society’s very recent downturn, a good number would band together, finding some inspiration or justification to act out in concert. With advances in modern education, we have seen a marked growth in our populace’s psychological-Â�mindedness, sufficient to encourage the parents of youngsters such as those described here to turn to our profession for guidance in solving the perplexing character of their children’s emotional and social behaviors (e.g., “I don’t understand him; he has everything a young person could want”.) What in other times might have taken root as a social movement of disaffected young radicals has taken the form of an epidemic of materially prosperous youth who possess the freedom to pursue abundance and contentment, but who feel isolated, aimless, and empty, and whom we “treat” for a deeply troubling psychological disorder.


Conclusions The preceding pages have described a number of the elements contributing to the broad mosaic of BPD-disposing influences in our times. Although I have largely bypassed their specifics, there are also salient elements of a biogenic nature in this multifactorial mosaic of determinants. Similarly, and prior critiques notwithstanding, this mosaic should be seen as encompassing the psychogenic role of adverse early nurturing and rearing. What is troubling to those who seek an “ecumenical” synthesis among rival etiological models is not the observation that some biogenic and analytic authors view this constitutional proclivity or that ordeal of early life as crucial to the development of a particular personality disorder. Rather, it becomes troubling when claimants couple empirically unproven or philosophically untenable assumptions with the assertion that they alone possess the sole means by which such etiological origins can be revealed. Perhaps it is too harsh to draw parallels, but presumptions such as these are not unlike Biblical inerrantists who claim their construals of the Bible to be “divine” interpretations, or conservative jurists who assert their unenlightened views to correspond to the “original intent” of the U.S. Constitution’s framers. So too do many of our more self-Â�righteous interpreters practice blindly what some have judged our “hopelessly flawed craft” (Grunbaum, 1984)—one in which we demonstrably can neither agree among ourselves, nor discover either the data or methods by which a coherent synthesis may be fashioned among our myriad conjectures. Accordingly, I hope the reader will recognize that the sociocultural thesis presented here assumes that individuals who will develop BPD in the future are likely to possess troublesome constitutional proclivities and/ or to have been subjected to early and repetitive experiences of a psychically diffusing or divisive nature. By contrast, youngsters endowed with an emotionally sturdy disposition and/or reared in a uniform, dependable, and stable manner are not likely candidates for BPD, whatever their encounters may have been with the social forces described herein. The present thesis is conceived best, then, as an addendum—one that seeks to describe

the final set of elements in the trio of biopsychosocial influences that coalesce to form BPD. Acknowledgment Portions of this chapter have been adapted and updated from Millon (1987a). Copyright 1987 by The Guilford Press. Adapted by permission.

References Akiskal, H. S., & Akiskal, K. (1992). Cyclothymic, hyperthymic and depressive temperaments as subaffective variants of mood disorders. In A. Tasman & M. B. Riba (Eds.), Annual review of psychiatry (Vol. 11, pp.€43–62). Washington, DC: American Psychiatric Press. Alarcón, R. D., Foulks, E. F., & Vakkur, M. (1998). Personality disorders and culture: Clinical and conceptual interactions. New York: Wiley. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Brody, E. B. (1990). Editorial: Biomedical science and the changing culture of clinical practice. Journal of Nervous and Mental Disease, 178, 279–281. Eagle, M. (1984). Recent developments in psychoanalysis: A critical evaluation. New York: McGraw-Hill. Easterlin, R. A., & Crimmins, E. M. (1985). The fertility revolution. Chicago: University of Chicago Press. Frank, L. K. (1948). Society as the patient: Essays on culture and personality. New Brunswick, NJ: Rutgers University Press. Grunbaum, A. (1984). The foundations of psychoanalysis: A philosophical critique. Berkeley: University of California Press. Klein, D. F. (1971). Approaches to measuring the efficacy of drug treatment of personality disorders: An analysis and program. In Principles and problems in establishing the efficacy of psychotropic agents (Publication No. 2138, pp.€187–204). Washington, DC: U.S. Department of Health, Education and Welfare, Public Health Service. Kleinman, A. (1980). Patients and healers in the context of culture. Berkeley: University of California Press. Kleinman, A. (1988a). The illness narratives:

A Sociocultural Conception of the Borderline Personality Disorder Epidemic Suffering, healing and the human condition. New York: Basic Books. Kleinman, A. (1988b). Rethinking psychiatry: From cultural category to personal experience. New York: Free Press. Klerman, G. L. (1987, January). The current age of youthful melancholia: Agent Blue. Lecture presented to the Royal College of Psychiatrists, London. Littlewood, R., & Lipsedge, M. (1987). The butterfly and the serpent: Culture, psychopathology and biomedicine. Culture, Medicine, and Psychiatry, 11, 289–336. Mezzich, J. E., Kleinman, A., & Fabrega, H. (Eds.). (1996). Culture and psychiatric diagnosis: A DSM-IV perspective. Washington, DC: American Psychiatric Press. Millon, T. (1969). Modern psychopathology. Philadelphia: Saunders.


Millon, T. (1981). Disorders of personality: DSM-III, Axis II. New York: Wiley. Millon, T. (1987a). On the genesis and prevalence of the borderline personality disorder: A social learning thesis. Journal of Personality Disorders, 1, 354–372. Millon, T. (198b7). Social learning models. In R. Michels (Ed.), Psychiatry: Personality disorders and the neuroses. New York: Basic Books. Millon, T. (with Davis, R. D.). (1996). Disorders of personality: DSM-IV and beyond (2nd ed.). New York: Wiley. Wicker, T. (1987, September 24). The pump on the well. New York Times, p.€23. World Health Organization. (1992). The ICD-10 classification of mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva: Author.

Part II

Conceptual Issues



Chapter 6

Philosophical Issues in the Classification of€Psychopathology Peter Zachar Kenneth S. Kendler

umerous approaches to a chapter on N philosophical issues in classification are possible. Philosophers of mind might

ask whether psychiatric disorders are best considered to be physical or mental entities. Epistemologists could analyze what it means for a diagnosis to be “evidence-based,” and what kind of “evidence” counts. Philosophers of science might ask whether psychiatry and clinical psychology should aspire to discover natural and law-Â�governed categories like those used in physics and chemistry. Political and moral philosophers might be interested in exploring how gendered norms are subtly written into some diagnostic categories (e.g., borderline personality disorder). Like Dewey (1938) and Quine (1969), we construe philosophical inquiry as naturally integrated with, not separate from, science. We do not, therefore, claim some special authority to inform readers about what count as the “important” philosophical problems in the classification of psychopathology. Our own list of important issues consists of topics commonly encountered in the scientific literature. This disparate list includes ques-

tions about dimensional versus categorical approaches to nosology; problems about clinical significance; a query about the role that pathological processes should play in conceptualizing psychiatric disorders; and the exploration of validity. That these are important problems is a matter of widespread consensus. Our goal is to draw out and examine the underlying philosophical issues that are already there. The potential advantage of a philosophically oriented approach to the contemporary classification problems of psychopathology is that those problems might be seen somewhat differently than they would be otherwise. A philosophically oriented approach might also make the failings of proposed solutions more evident, so that any solution is less likely to be taken literally. The appropriate philosophical framework could serve as a conceptual resource that functions as an “antidote” to literalism and one of its most important symptoms—the reification of diagnostic constructs. The framework we propose in this chapter has such a goal in mind. 127


Should Classification Be Dimensional or Categorical? One of the more visible issues is whether psychiatric disorders should be modeled as categories or dimensions (First et al., 2002; Krueger, Watson, & Barlow, 2005; Regier, 2007). Three alternative perspectives are used to conceptualize the categories versus dimensions problem. Our descriptions are idealized, and a perspective’s advocates will have varying degrees of commitment. The first is an “essentialist” perspective; the second is an “empiricist” perspective; and the third is the perspective of “pragmatism.” Each of these perspectives has strengths and weaknesses, but we argue that the proponents of the empiricist and pragmatic perspectives are least vulnerable to the charms of diagnostic literalism. The Essentialist Perspective “Essentialism” is the view that true classifications optimally depict the inherent structure of the world in itself. How the world is, and not our interests, should guide classification (Kripke, 1980; Putnam, 1975). For example, the periodic table of the elements is often considered to represent a discovery of the inherent, true structure of the world. Essentialism is often associated with the sorting of classifications into categories such as “natural” versus “artificial/arbitrary” (Ahn, Flanagan, Marsh, & Sanislow, 2006; Haslam, 2000). Scientific classifications are supposed to be natural. One form this perspective takes is to ask whether psychiatric disorders can be naturally segregated into real classes or whether they are really dimensional (Lilienfeld & Waldman, 2004; Livesley, 2003). Real “classes” are groups for which there are definitive criteria for deciding membership. Examples include even numbers, U.S. Senators, and people who are HIV-positive. Categorical distinctions of this type are binary: One either is or is not a member of a class, and degree of membership is equally shared by all members (i.e., 4 is not more of an even number than 100). In contrast to a class, a “dimension” refers to a construct of interest that can be mea-


sured and on which everyone has a value. Examples of dimensions are height and weight. Dimensions are also “graded,” meaning that some people will be assigned lower and others higher values on the dimension. Categorical distinctions such as “tall” and “short” require dividing the dimension into discrete subgroups. From an essentialist perspective, such categorical distinctions are considered to be arbitrary with respect to the dimension itself. In philosophical parlance, they do not “carve nature at its joints.” One of the important virtues of the essentialist perspective is its insistence that we should classify the world based on what it is actually like, and not based on what we want it to be like. For example, if it could be demonstrated that psychiatric disorders really are dimensional rather than categorical, proponents of the essentialist approach would unequivocally state that they should be modeled dimensionally—Â�because in accurately reflecting the way the world really is, both the science and the practice of psychiatry/psychology would benefit in the long run. The Empiricist Perspective The “empiricist” perspective on the categories versus dimensions issue focuses not on what is really there, but on what claims can be justified by the evidence. As a general rule, scientific empiricists adopt a skeptical stance, holding that our current beliefs and theories could turn out to be mistaken (van Frassen, 1980). By long-Â�standing tradition, empiricists are suspicious of unobservable (or latent) constructs, such as quarks, dark matter, implicit memory, and neuroticism. They consider unobservable constructs to be tools or instruments that can help us predict how the world will appear in the future, but that do not describe its essence. Empiricists claim that these instruments have no reality beyond the observations themselves. They do not believe that the world comes to us prepackaged in neat categories, but neither would they license the inferential leap from “Dimensional models fit the data better” to “The world is really dimensional.” As far as they are concerned, all that can be legitimately said is that “Dimensional models fit the data better.”

Philosophical Issues

They also might add, “ .€.€. fit the data better, considering what measurement instruments are currently available.” From this skeptical perspective, empiricists believe that categorical and dimensional classifications each have advantages and disadvantages. They would be reluctant to endorse metaphysical claims about what is really there. Empiricists espouse “antirealism.” They argue that theoretical constructs such as neuroticism typically come to be defined with respect to other theoretical constructs (such as temperament, schema, and identity), and therefore inevitably become increasingly distant from observation. As this process of theoretical elaboration continues, metaphysical attributions about their “reality” become increasingly tenuous. This is as true for the five-Â�factor model of personality as it is for the Freudian metapsychology. As a result, empiricists tend to be suspicious of the metaphor of carving nature at its joints, especially if it is used to support an essentialist perspective. The categorical versus dimensional issue is an empirical question to the extent that theories about structure can be tested. Various mathematical models have been developed to test whether psychiatric disorders can be segregated into classes or dimensions, with factor analysis, latent-class analysis, and taxometrics being the best known (Goldberg & Velicer, 2006; Haslam & Williams, 2006; Markon & Krueger, 2006). Because the conditions listed in psychiatric diagnostic manuals represent a heterogeneous collection and do not form a single, hom*ogeneous kind, it should not be surprising to find that the results of the various statistical methodologies vary (Haslam, 2007). There is good evidence that major depressive disorder (MDD), generalized anxiety disorder (GAD), and the personality disorders have a dimensional structure (Krueger, 1999; Krueger, Caspi, Moffitt, & Silva, 1998; Widiger & Samuel, 2005). One can also find subtypes within dimensions (such as melancholia within depression) and dimensional structures within categories (such as the positive, disorganized, and negative symptom dimensions of schizophrenia). Categories can be nested in dimensions, and vice versa. Being concerned with all the evidence rather than with preferred models


of reality, empiricism is consistent with the suggestion that diagnostic manuals should utilize both categorical and dimensional perspectives where appropriate, although how to integrate the two is a matter of some debate (Helzer, Kraemer, & Krueger, 2006; Kessler, 2002). The Perspective of Pragmatism A group of thinkers referred to as “pragmatists” also emphasize the instrumental or practical value of classifications. They contend that classifications should be chosen with respect to their ability to help us achieve our disciplinary goals (Brendel, 2006; James, 1907, 1909/1975). Pragmatists claim that the phenomena we classify are inherently complicated and that rich, and that any classification inevitably involves highlighting some features and deemphasizing others (Goodman, 1978). As a result, they are also suspicious of talk about carving nature at its joints (Zachar, 2006; Zachar & Bartlett, 2001). An important philosophical difference between empiricists and pragmatists is that empiricists tend to be antirealists about theoretical constructs, whereas pragmatists need not be. For pragmatists, metaphysical inferences about “reality” are another set of philosophical tools, and it is sometimes informative to distinguish between “real” and “not real” (Dewey, 1929/1958; Hacking, 1999). For example, in biology it has not been possible to discover a single, optimal model of what counts as a species. There exist several species concepts, such as the “phenetic species” concept, the “biological species” concept, and the “phylogenetic species” concept (Mayr, 1988, 1991; Mishler & Brandon, 1987; Ridley, 1986). No single species concept works for all classificatory purposes. The phenetic species model can occasionally contradict evolutionary history; the biological species concept does not apply to many plant species; and the phylogenetic species concept classifies only branching, not divergence. This complication inclines empiricists to say that species categories are not real, while those with pragmatist leanings are willing to say that species are real, but there are multiple ways to define a good species (Ereshefsky, 1992; Mayr, 1982).


How does the categories versus dimensions issue look from the standpoint of pragmatists? To repeat, essentialist-Â�leaning thinkers are willing to claim that psychiatric disorders really are dimensional, not categorical (or vice versa), whereas empiricists would claim only that dimensions fit the current data better (or vice versa). Pragmatists tend to be “pluralists.” Pluralism is the belief that no single classification system can answer all the classificatory goals and purposes we adopt; therefore, different goals may require different classifications, each of which can have validity. As noted elsewhere (Zachar & Kendler, 2007), relevant goals for a classification system include selecting treatment, increasing etiological hom*ogeneity, maximizing true positives and true negatives, being measurable, being consistent with genetics and physiology, being culturally congruent, and being clinically informative. No single kind (or dimension) is likely to achieve all the valid goals we can articulate. For example, if we want to find out how many people in a country were diagnosed with schizophrenia in a single year or to learn about the age of onset and course of schizophrenia across different countries, some kind of a categorical model with inclusion and exclusion criteria will be important. If, on the other hand, we want to understand the relationship between level of disorganization and social functioning in the past 6 months, we will want to measure the full range of each variable (from low to high). Pragmatists therefore ask what purposes are being served with either dimensional or categorical approaches to classification. Because they are more or less appropriate to specific kinds of goals, choosing dimensions or categories is not only about describing what is really there. In what follows, we explore some of these diverse goals.

Dimensions and Categories in Research Many psychologists, especially those with a psychometric focus, advocate dimensional models as a function of being trained in correlational research designs. In correlational statistics, taking a continuous variable such as weight measured in kilograms, and transforming it into a dichotomous variable such as “light versus heavy,” represents an important loss of information. The same is true for


dichotomizing mood into “depressed versus not depressed.” Classifying all the numerical information is a statistical virtue of dimensional models. In contrast, those with backgrounds in epidemiology tend to have a greater appreciation of the virtues of categorical analyses. In practical settings, it may be more clinically informative to know that the Rorschach Perceptual Thinking Index (PTI) can accurately detect 85% of patients who are diagnosed with a psychotic disorder than to know that scores on the PTI explain 41% of the variance in psychotic symptoms (Dao, Prevatt, & Horne, 2007). Studying deaths from a disease, premorbid status, ages of onset, and number of recurrences requires categorization. Coding occurrences in terms of 1-month and lifetime prevalence is also categorical. Such concepts as “caseness,” “sensitivity,” “positive predictive value,” and “base rate” are geared toward categorically defined outcomes. Needless to say, if true discontinuities exist, categorical approaches have even greater virtues.

Dimensions and Categories in€Clinical€Practice Dimensions can also serve purposes more specific to practitioners. For example, dimensional models such as those proposed by Simms and Clark (2006) or Livesley (2006) should more comprehensively model the phenotypic space of personality disorders than do the current ICD and DSM categories. There will always be clinically important data not accounted for in the classification system, but the overreliance on the “not otherwise specified” diagnosis for Axis II conditions suggests that too many data are not being modeled. A more comprehensive model of the domain offers the possibility of identifying new targets for intervention, or distinguishing between those parts of the phenomena that are amenable to change and those that are not. What purposes might categories serve for practitioners? For one, decisions about whether or not to treat are usually categorical. Such binary decisions will not be eliminated in a dimensional system. In DSM and ICD, describing conditions and stating the need for treatment is lumped into a single act of diagnosis. Dimensional models would

Philosophical Issues

disentangle these descriptive and prescriptive tasks. Another purpose of categories is to group together cases that share something in common. Groups of cases of this sort are called “kinds.” In DSM and ICD, MDD, schizophrenia, and anorexia nervosa are kinds. Psychiatrists and clinical psychologists learn more about MDD, schizophrenia, and anorexia nervosa by studying what people with these conditions have in common. In theory, the same strategy is possible with groups sharing the same dimensional profile (e.g., the group that is high on narcissism and callousness, low on diffidence and compulsivity). The purpose of grouping cases is to make generalizations about group members, but for pragmatists there is nothing sacred about a grouping. If groups can be combined or decomposed in such a way as to enable new and better generalizations, then they should. An important question about kinds is this: How good are the generalizations? With very hom*ogeneous groups, such as “copper things,” we can make robust generalizations (e.g., “All copper conducts electricity”). With very heterogeneous groups, such as “white things,” few or no useful generalizations can be made. Groups of patients with similar profiles of psychiatric symptoms are neither as hom*ogeneous as copper nor as heterogeneous as white things. The generalizations are good to the extent that they help us achieve a variety of disciplinary goals. We conclude this section on dimensional versus categorical approaches by observing that the starkest choice for classifiers is between the realism of the essentialists and the antirealism of the empiricists. As noted, empiricists are no friends to the view that the world comes preorganized into a set of real categories waiting to be discovered, but neither do they seek to discover the true list of dimensions that are really out there. Among the various available perspectives, advocates of empiricism are the least vulnerable to the problems of reification. Along with the pragmatists, empiricists caution against overexuberance regarding the nosological momentum that dimensional models clearly have. Like anything else, once a new conceptual strategy is institutionalized, it may expand the scope of psychiatric nosology, but then itself may


become another barrier to progress because there is a tendency to take it too literally. The same kind of reification of diagnostic categories that occurred after the publication of DSM-III would be likely to occur with respect to any “official” dimensions included in DSM-V or ICD-11. Advocates of a more essentialist approach to dimensions, intentionally or not, subtly enable this unfortunate outcome.

What Counts as Clinically€Significant? An assumption of essentialism is that mental disorders are out there in rerum natura, and we just have to work hard to discover them. Diagnostic literalism signifies great respect for the authority of diagnostic manuals. This may refer to either a current manual or a future ideal manual. Attributing extensive authority to a current manual is associated with a tendency to believe that the inherent structure has been successfully classified (Kendler & Zachar, 2008). On the flip side of the same coin, evidence that a classification is flawed will incline others to say that the classification is scientifically primitive and needs to be eliminated in favor of a more “optimal” approach. We agree that if they are appropriately classified, disorders are out there. However, we also believe that developing a good classification system requires more than discovery of facts; it also requires that we make decisions based partly on goals and values. Values can be defended, but are not easily classified as either true or false. The authority that accrues to a classification manual regarding issues of value is not the same as the authority that accrues to it with respect to statements of fact. In this section, we argue that considerations of clinical significance highlight the role played by values in the classification of psychopathology. It turns out that very few thinkers claim that psychiatric nosology should be purged of value judgments. In agreement with the mainstream philosophical opinion on this issue, we claim that rational discussion and deliberation about values is not only possible; in some cases, it should be articulated more explicitly than it has been in the past.


The Role of Evaluation “Values” refer to judgments of things as good or bad, beneficial or harmful. When the word “values” is used, people often think of moral values such as honesty, fairness, and concern for others, or loaded terms such as “family values,” but values are not limited to the domain of the moral. Reliable cues to value terms include use of the words “should” and “ought.” There are also political values, such as habeus corpus and freedom of speech. Objectivity, validity, and peer review are scientific values. Training in the mental health disciplines partly involves teaching students, interns, and residents what to value and to disvalue. For example, they learn to recognize compulsivity, emotional lability, and inflexibility as negatively valued (pathological) conditions. Bentall’s (1992) proposal to classify happiness as a mental disorder offers an amusing introduction to the role of values in classification. Bentall notes that happiness is a coherent syndrome with behavioral, affective, and cognitive components. As proponents of depressive realism might contend, happiness is also associated with impaired reality testing (Alloy, Albright, Abramson & Dykman, 1990). Bentall states that there is good anatomical and neurochemical evidence that happiness is biologically based, and may even have a genetic component. From an epidemiological perspective, persistent happiness is statistically deviant. It also has a higher prevalence in people of higher socioeconomic status (SES), suggesting that the high-SES population is more exposed to the “risk factors” for happiness. Furthermore, happiness seems to be better modeled as a dimensional construct, perhaps a subthreshold version of hypomania. So why is happiness not a mental disorder? According to Bentall (1992), it is not a disorder because it is not negatively valued. There are two ways to be negatively valued. First, happiness is not a “negatively valued experience,” whereas panic attacks are. Second, happiness is not negatively valued by being an “object of therapeutic concern.” The two kinds of negative evaluations do not always coincide. Grief is a negatively valued experience, but it is not usually an object of therapeutic concern. Hypomania may not be


a negatively valued experience, but it is typically an object of therapeutic concern. The question is this: Should such values considerations bear on the objective facts of the matter? Bentall’s conclusion is that yes, they should. In his own words, “only a psychopathology that openly declares the relevance of values to classification could persist in excluding happiness from the psychiatric disorders” (1992, p.€97). A synonym for “object of therapeutic concern” is “clinical significance.” About half of the disorders in DSM-IV require that there be clinically significant impairment or distress in social, occupational, or other important areas of functioning. The clinical significance criterion (CSC) was introduced because finding high prevalence rates in epidemiological studies that used strict translations of DSM criteria raised fears of a falseÂ�positive problem (Regier, 2007). A concern about the legitimacy of such disorders as social phobia and oppositional defiant disorder was also an issue (Spitzer & Wakefield, 1999). The CSC also allows professionals to attribute mental disorder status to conditions that do not reach diagnostic thresholds (i.e., to correct for false negatives). In many cases, the CSC is a strategy for resolving the boundary problems that occur at diagnostic thresholds. Impairment, Distress, and Value By what standards does one decide whether impaired social or occupational functioning is present? Is this a purely factual matter? In some cases, there is the observation of “decline in functioning,” but how much decline counts as clinically significant? A related concept is “loss of control.” The change in functioning is factual, but assessing clinically significant change in functioning requires making evaluative distinctions among “optimal,” “normal,” “compromised,” and “impaired.” Such evaluations will vary across individual patients, developmental stages, social roles, cultures, and historical epochs. For example, is a professional baseball player in his prime who successfully competes but is no longer elected to the All-Star team significantly impaired because he has a level of drug and alcohol use that interferes with his training? What if the substance use

Philosophical Issues

interferes with his friendships and parenting as well—even though he still has good friends and is a better parent than many? Even with something as observable as decline in functioning, to identify a disorder there must be a value judgment of clinically significant impairment (or inability to do what one normally should be able to do). To some extent, clinical significance is a categorical judgment (present or absent); however, it is not only a matter of observing present-at-hand impairment because, according to DSM-IV, impairment need not be actually present. The manual allows the diagnosis of “threat to health” conditions, which are mental disorders that are analogous to hypertension. One could examine the baseball player, decide that there is a potential impairment if he continues his current level of substance use, and therefore diagnose a disorder. With potential impairment, a prediction is being made that “compromised” functioning could become “impaired,” leading to the additional value judgment that preventative steps would be beneficial. Impairment is clearly a value-laden concept. Next, we turn to distress. Distress or unpleasantness can be considered inherently negative. Even so, deciding whether distress is clinically significant distress is not a purely factual matter. Most of the heat generated by debates about medicalizing subthreshold conditions revolves around disagreements about when distress should be an object of therapeutic concern. Some mental health professionals would claim that the distress associated with active psychotherapy or the normal distress of loss is potentially healthy rather than pathological, even if it is intense (Horwitz & Wakefield, 2007). What if a person in grief were to feel better after being prescribed an antidepressant or an anxiolytic? Some suggest that an improvement in functioning does justify writing a prescription because a good deal of misery could be averted if subthreshold conditions were taken more seriously (Kendler, 2008b; Kessler et al., 2003; Pincus, McQueen, & Elinson, 2003). Making evaluative distinctions among “impaired,” “compromised,” and “normal” has a palatable degree of ambiguity. The ambiguity about norms specifying what should and should not be suggests that the offi-


cial listing of psychiatric disorders may be subject to modification and redecision over longer periods of time. The declassification of hom*osexuality as a mental disorder is an example of this process. One of the problems with reification is that it takes complicated decisions about what conditions count as psychiatric disorders and translates those decisions into official lists that people take to be only matters of scientific fact. We suggest that accepting that evaluative complexities are inherent to good psychiatric classification is a more “realistic” strategy than seeking to discover a purely natural classification system. Values, Psychiatric Classification, and Rational Deliberation Thomas Szasz and his followers are generally considered to believe that values should play no role in demarcating true medical conditions. Szaszians tend to be out of the mainstream on this issue, however. Wakefield (1992, 1999a), Sadler (2005), and Fulford (1989) all argue that values are inevitable in nosology. As we discuss in the next section, Wakefield believes that classifying valid psychiatric disorders requires identifying the potential presence of an objective dysfunction and making value judgments regarding harm. In contrast, Sadler and Fulford do not favor a strict fact–value separation. They see values as fully penetrating psychiatric nosology, even where Wakefield sees facts. One of Sadler’s and Fulford’s main points is that that professionals can come to rational agreement on which set of values should guide both the science and practice of psychiatry/psychology. In an earlier article, we identified “objectivism versus evaluativism” as one of the important philosophical issues inherent to psychiatric classification (Zachar & Kendler, 2007). We stated that a condition such as schizophrenia might be considered objective in the sense that describing what is broken (see below) could be considered more of a factual matter, but we also noted that its “objectivity” rests on the fact that there is widespread agreement about the relevant values issues. For example, hallucinations and delusions tend to violate epistemic norms. When there is intersubjective consen-


sus about “normality,” medical evaluations can gain a degree of objectivity. Examples of the “minimally evaluative” include forestalling death when there is still potential quality of life and reducing intense physical pain. The relevant value judgments needed for a psychiatric nosology will have varying degrees of consensus. There will be more concurrence with psychotic disorders, but less with personality disorders, sexual dysfunctions, and substance use disorders. Anxiety and depressive conditions will lie somewhere in between. Interestingly, although eating disorders are associated with death and serious physical deterioration and can be as medically relevant as any psychiatric disorder, members of the “pro-ana” movement assiduously define anorexia nervosa as a lifestyle choice and not a disorder (Udovitch, 2002). Dimensional models, which lack boundaries between normal and abnormal, are least able to avoid the value-laden problem of assessing clinical significance. We are tempted to suggest that the more disagreement that is likely to exist about particular value judgments (“shoulds” and “oughts”), the more explicit the DSM and ICD committees should be in articulating the justifications for making them. At the same time, we urge caution. To give value judgments regarding impairment and distress a more explicit presence in DSM and ICD would increase the probability of subjecting those values to reification.

What Role Should Pathological Processes Play in Understanding Psychiatric Disorders? In this section, we argue that psychiatrists and psychologists should seek to develop models of psychiatric disorders that help explain why these conditions come about in terms of underlying pathological processes. These processes can be more psychological or more biological—or, ideally, both. One of our main claims is that discovering underlying pathological processes is not the same as discovering what disorders are out there. Natural Dysfunctions Jerome Wakefield (1999a, 1999b) claims that a legitimate psychiatric disorder has


two essential aspects: (1) the presence of an objective dysfunction, and (2) the evaluation that the dysfunction is harmful (impaired/ maladaptive). Wakefield claims that a dysfunction is present whenever something is broken. “Broken” in this context means that part of an organism is not able to function as it was designed to function in evolution. For example, eukaryote cells were designed during evolution to use adenosine triphosphate (ATP) as a source of energy, and furthermore to obtain ATP from a type of prokaryote organism called mitochondria. Lack of mitochondria is a fatal dysfunction. In order to conceptualize dysfunction, Wakefield and others have reintroduced the Aristotelian idea of natural function, which in the post-Â�Darwin world now has a historical–Â�evolutionary interpretation. “Natural function” refers to how the part or process in question contributed to a species’s survival, and therefore also explains why that part/process was selected during evolution. For example, it is likely that detecting light increased one of our ancestors’ ability to survive and reproduce, which in turn established a selection pressure for light detection that eventually resulted in eyes. The important point is that if this historical story is true, then it is a matter of fact that eyes evolved because seeing conferred an adaptive advantage. Seeing counts as an eye’s natural function. Whether or not an eye performs this natural function is also a factual matter. According to Wakefield, value judgments of the type “An eye should be able to see” can potentially be translated into factual statements about evolutionary history. As matters of fact and not matters of value, dysfunctions are supposedly things we can discover. An advantage of Wakefield’s harmfulÂ�dysfunction model is that it captures our common-sense notion of a disorder (i.e., that the attribution of a disorder means that something is broken and needs to be fixed). Dysfunctions can pertain not only to organs and organ systems, but to more abstract processes such as psychological functions (e.g., attention and emotional regulation). The problem with the harmful-Â�dysfunction model is that it is not amenable to empirical confirmation. In contrast to Wakefield, we claim that underlying “pathological” processes are discoverable, but attributions

Philosophical Issues

of dysfunction in practice have an intuitive element. Nominating natural functions requires speculation that goes far beyond the available evidence because we cannot say for sure what selection pressures existed in human evolutionary history or what the targets of selection were. Other processes in addition to natural selection, such as genetic drift and molecular drive, may also have played roles in evolution. That is, not everything that has evolved had to have a natural function. Let us explore this problem further by examining reading disorder. What is broken in an inability to read despite having a high level of intelligence and an appropriate education. Reading is not a natural function because the cognitive mechanisms used in reading were no more designed by natural selection for reading than three-Â�dimensional visual–Â�spatial abilities were designed for neurosurgery. Hunters and gatherers did not need to read. Wakefield (1999a) claims that although reading is not a natural function, brains that are functioning as designed can typically read. Ability to read is a beneficial side effect of one or more other natural functions. Wakefield may be correct. It is, however, just speculation. No discovery of fact has been made. It is possible that in a future scientifically advanced society, not being able to master calculus will be maladaptive. The inference that a normally functioning brain should be able to learn calculus will be likely to occur, and understandably so, but this inference will not occur independently of the perception of harm/impairment. Analogously, given its importance in modern society, and the very high base rate of people who can read, most will agree that being unable to read is an impairment. As with calculus, the inference that something is broken or not functioning properly may depend on a prior value judgment—in this case, that people should be able to read. Here is the important point. If cognitive psychologists were to identify some underlying attentional mechanisms that helps explain reading difficulties, what would make those mechanisms objectively dysfunctional in Wakefield’s sense, rather than normal variations in design that have recently become harmful? Looking at a targeted individual


difference and claiming that it is a design failure rather than a maladaptive variation in design does not constitute a simple task. Again, hom*osexuality provides a good example. Would a scientifically correct model of the biological genesis of an exclusively hom*osexual orientation describe a dysfunction or a natural variation? We suggest that one’s intuitions about this issue will be correlated with one’s previous value judgments about how sexuality should develop. It is certainly true that some psychiatric disorders are objective dysfunctions that are harmful, especially when there is evidence of trauma or of a developmentally unexpected decline in functioning from a previous level of adaptation. Broken bones, autism, and disorganized schizophrenia are good examples. Impulsivity and executive function deficits contingent on a traumatic brain injury are also arguably objective dysfunctions. Other disorders, however, might be harmful variations. Narcissistic and obsessive–Â�compulsive personality disorders may represent normal variations that are first judged to be maladaptive, and subsequently conceptualized as being broken. As Wakefield (2006) has noted, one of the ways a culture works is to make its own values seem natural. For these reasons, we claim that nosologists should be cautious in speculating about what counts as an unnatural function (i.e., a design failure). Paul Meehl (1993) perceptively wrote that scientists should adopt the ethic of trying not to fool themselves and not to fool others. The high risk of fooling ourselves while speculating about what counts as unnatural is one of the reasons for being cautious about the medicalization of both subthreshold Axis I conditions and individual differences in personality. Few are more cautious in this respect than Wakefield. Dysfunctions and Causal Roles Let us assume for the moment that Wakefield’s critics are correct regarding the difficulty or even impossibility of discovering natural functions and dysfunctions. Even if they are correct, we suggest that one should not conclude that a concept of dysfunction has no important role to play in psychopathology, even if it does not conform to Wakefield’s stipulations.


In agreement with Ken Schaffner (1993), we claim that the concept of function does not have to be historical–Â�evolutionary. Rather, it can refer to whatever role a particular part or process plays in the overall functioning of the organism. For example, a function of the heart is to pump blood; whether pumping blood really is the natural function is irrelevant. The same is true for the cognitive mechanisms that enable reading (or calculus). If reading is considered to be an important ability, whatever parts or processes can be used to explain failure to read can be identified as dysfunctional with respect to reading. In concluding this section, we assert that the labeling of underlying mechanisms as dysfunctional is sometimes dependent on having already made a disorder attribution. Although the mechanisms are potentially objectively present, a complicated and shifting set of indicators have to be consulted in nominating disorders, and different indicators may be more or less relevant depending on the disorder in question. The decline in functioning indicator is relevant in autism and schizophrenia, but not in attention-Â�deficit/hyperactivity disorder (ADHD). It has also recently been claimed that liberalism is a mental disorder—a claim that some people take seriously (Rossiter, 2006; Savage, 2005). Similar claims have been made about racism (Bell, 2004). Justifying ADHD’s status as a psychiatric disorder, while rejecting that of liberalism and racism, is a worthy conceptual problem—one that requires careful thinking. Discovery of fact is crucial, but such problems cannot be reduced to the kind of straightforward empirical questions favored by those inclined toward diagnostic literalism.

What Is Validity? Rather than conceptualizing validity as a single quality, we advocate “validity pluralism.” It is not just a question of whether a diagnosis is valid per se, but what kind of validity it has and how good the validity is. We review four perspectives on validity. Although the perspectives are distinguishable, many psychiatrists and psychologists will endorse them simultaneously. The four perspectives are “entity-based approaches,”


“information-based approaches,” “construct validation,” and “explanatory validity.” Entity-Based Approaches It is well known in the classification literature that beginning with DSM-III, nosologists became concerned with reliability, yet paid limited attention to validity. But what is validity? Usually when mental health professionals ask, “Is schizophrenia a valid disorder?”, they mean either “Is it a legitimate mental illness?” or “Is it really a disorder?” Some believe that legitimate mental illnesses are diseases (Kendell, 1975; Robins & Guze, 1970), whereas others claim that they are better considered disorders than diseases (Wakefield, 2000). Kendell and Jablensky (2003) define valid psychiatric diagnoses as those for which natural boundaries between cases and noncases can be established. They state that valid disorders should be qualitatively different from normality and also should not overlap with other disorders. According to Kendell and Jablensky, valid diagnoses will eventually be shown to possess a combination of features that exist only when the disorder is present. In highlighting the importance of discreteness, Kendell and Jablensky advocate the “disease realism” approach to validity. Its most famous proponents, Eli Robins and Samuel Guze (1970), proposed five criteria for confirming the presence of a disease, which they termed “validity indicators.” These criteria are clinical description, laboratory tests, differential diagnosis, follow-up studies, and family studies. Later, Kendler (1980) expanded the number of potential validators and added a temporal dimension, distinguishing among antecedent, concurrent, and predictive validators. Disease realists believe that mental illnesses have as-yet-Â�undiscovered pathological processes. In the absence of a confirmed etiology, the goal is to look for integrated syndromes and fill in the etiologies as they are uncovered They astutely point out that one does not have to know the etiology of a mental illness for it to be valid, any more than 18th-Â�century physicians had to know the etiology of tuberculosis for it to be valid. In both cases, diseases can be inferred if the appropriate indicators (validators) are present. To this view, Kendell and Jablensky (2003)

Philosophical Issues

have added the validator of a qualitative difference between normal and abnormal. Arguments that involutional melancholia is not a valid syndrome exemplify the disease realism approach. Involutional melancholia was a DSM-II syndrome that referred to first-Â�episode psychotic depressions among women. It included an agitated presentation, as well as symptoms that were not supposed to be reactions to stress, but were contingent on menopause (i.e., the involutional period). In the 1970s, it was discovered that a large number of the identified cases did have precipitants and that an agitated presentation was related to past histories of depression. Furthermore, rates of depression for women are not higher during menopause (Becker et al., 2001; Berrios, 1991; Weissman, 1979). The obvious conclusion is that the syndrome referred to by the construct of involutional melancholia is not really there. An important problem for entity-based approaches is that of systematics. In biology, “systematics” is the study of the relationship between species. As defined by Zachar (2008), “psychiatric systematics” studies what relationships, if any, exist between various disorders. A more absolute form of validity would ask whether a disorder is a coherent entity, but it is also important to understand a disorder relative to others disorders. For example, are MDD and GAD two disorders or one? If they are two overlapping disorders, how can they best be differentiated? If MDD segregates with GAD, then what is its relationship to bipolar disorder? Is there also a relationship between GAD and bipolar disorder? How should nosologists respond if the validators offer conflicting answers to such questions? The issues of overlap, differentiation, and hierarchical structure are complicated aspects of classification that deserve more empirical and conceptual attention. Information-Based Approaches Favoring a more data-based, quantitative conceptualization, proponents of empiricism reject the metaphysical predilections of disease realists. For empiricists, validity refers not to whether a disease is really there, but to what kind of inferences one legitimately can make about a patient on the basis of the diagnosis. Whereas for disease realists valid-


ity is a property that diagnoses have or do not have, for empiricists inferences, not diagnoses, are validated. If someone is diagnosed with MDD, “validity” refers to how much evidence exists for the inferences that mental health professionals can make about etiology, treatment, and outcome. Inferential power will vary depending on a host of factors. A diagnosis of MDD would suggest some inferences about time course, but if the patient is also experiencing a third episode and there is a family history, different inferences can be legitimately made. In contrast to the validity of the disease realists, the validity of the empiricists is not an either–or phenomenon. It exists in degrees. A reasonable approximation to a more empirical approach is the quantitative taxonomy of Achenbach and his colleagues (Achenbach, 1995, 2001; Achenbach, Bernstein, & Dumenci, 2005a, 2005b). Rather than using clinical observation to discover real syndromes with different natural histories, they use statistical procedures to detect syndromes defined as symptom patterns—or empirical regularities. The symptom patterns they have identified include anxious depression, withdrawn depression, rule breaking, and aggressive behavior. Instead of seeking to discover what is really there, they note that the observed patterns for any individual may differ depending on the data collected, such as self-Â�report versus other-Â�report. Variations in reporting may be weighted relative to each other, but not necessarily reduced to a single, true representation. Their multitaxonomic approach suggests that partialing the data into different patterns may increase the amount of information that is available to the clinician. As noted previously, empiricists tend to more attracted to antirealism. For them the goal of science is to be able to make connections between observable phenomenon. They want to be able to make such claims as “If X happens, then we can predict Y,” or “If you do A, then you can expect B to follow.” Empiricists consider nonobservable entities such as electrons and diseases to be “inference tickets.” In empiricist psychopathology, the kinds of inferences that are important include testable inferences about etiology, natural history, treatment response, outcome, and so on.


The difference between disease realism and empiricism can be seen in how they would conceptualize the historical shift from explaining phenomena such as hearing voices and thought insertion by reference to “demonic possession” to explaining them with respect to a syndrome named “schizophrenia.” A disease realist will say that psychiatry has learned that demons do not exist. Instead of demonic possession, hearing voices is a psychotic episode that results from an underlying biopsychological dysfunction. For empiricists, arguments about whether demons exist are metaphysical superfluities. What is important is that the evidence for the inferences one would make with the demonic possession theory are not as valid as those one would make with the schizophrenia theory. Inferences about sinfulness as an etiological factor, the role of punishment from God, and the efficacy of exorcism all lack supporting evidence. In contrast, inferences about family history, time course, and response to medication have been better justified (supported by evidence). It is important to emphasize that although the empiricists do not favor the validity of the disease realists, disease realists readily accept validity as it is defined by the empiricists. The information-based emphasis on the validity of inferences is a consensus view on validity among the scientific community, serving as the lowest common denominator. Whether scientific validity can or should refer to more than the validity of inferences is an important philosophical problem. Construct Validation Another approach to validity, introduced to psychology by Paul Meehl in the 1950s, is called “construct validation” (Cronbach & Meehl, 1955). Construct validation represents an evolution of the rigorous empiricism that dominated the philosophy of science in the early 20th century. The proponents of the revised empiricism, such as Herbert Feigl, were close colleagues of Meehl. The early empiricists were committed operationalists. That is, they believed that abstract concepts had to be explicitly defined in terms of observable data, and that the methods used to make the observations had to be explicit. For example, in psychological testing, a score of 65 or greater on a depression


scale operationally defines “depression.” For strict operationalists, what is meant by “depression” is the measurement. Meehl and his colleagues questioned such a firm commitment to operationalism. As we learn more about the correlates of the depression scale, we may discover that “depression” is associated with other constructs, such as “anxiety” (Kendler, 1996); that it involves negative thoughts about the self, the world, and the future (Beck, 1976); or that being depressed leads to interpersonal rejection, which sets up a feedback loop that maintains the depression (Joiner & Coyne, 1999). A construct is explicitly defined in term of observables, but implicitly defined in terms of other constructs. The bearer of this surplus meaning above and beyond the operational definition is a theoretical construct—Â�termed “depression.” In Meehl’s evolved empiricism, depression is construed realistically, not instrumentalistically. Let us focus on the differences between the strict operationalist approach of the classical empiricists and Meehl’s approach based on theoretical constructs/latent variables. For operationalists, a theoretical term such as “depression” is a verbal label for referring to what is being measured. The goal is to use the measure to infer observable consequences. In contrast, for Meehl the primary object of scientific interest is the latent variable called “depression,” and our measure is considered to be a fallible indicator of that latent construct. Different measures of depression are conceptualized to be assessing the same thing, but each one is at best a partial measure. Observable evidence is used to triangulate on the construct and better understand its full nature. Applied to psychiatric classification, the diagnostic criteria for MDD are also fallible indicators. Each indicator samples only a part of the domain of depression, and they all work better together to represent the whole domain. These indicators can be assessed for diagnostic validity, but the construct of depression cannot be reduced to its indicators. For example, in the development of DSM and ICD, some potentially relevant characteristics of MDD (such as increased experience of panic, somatic complaints, and irritability) were considered poor diagnostic criteria because they muddled differential diagnosis with the anxiety disorders, but from

Philosophical Issues

a construct validity framework they are still indicators of depression. An indicator can be sensitive but not specific. A complete description of depression at the criteria level is sacrificed in order to facilitate differential diagnosis. The intricacies of such tradeoffs are another reason not to be too literalist about official diagnostic criteria. Construing a stripped-down DSM or ICD category literally also has the unfortunate consequence of limiting what kind of features are considered relevant in developing models of underlying pathological processes. For these reasons, psychologists, who are schooled in the construct validity tradition, are uncomfortable using DSM or ICD as a textbook of psychopathology.

Validation, Theory, and Law As noted earlier, the construct validation approach was inspired by an evolution within the empiricist philosophy of science. One of the things the revised empiricism focused on was laws (Feigl, 1970; Hempel, 1966). Laws describe patterns or regularities in nature that must occur. Knowing about the regularities allows one to predict what will occur, and also to explain why events did occur. An example of a law is “All copper conducts electricity.” If the law is valid, then whenever presented with an individual piece of copper, we know that it has to conduct electricity. This law-like or “nomological” approach to science would seek to confer on “Bill will respond to treatment X because he has schizophrenia” the same degree of certainty that we attribute to statements such as “This metal will conduct electricity because it is copper.” Conducting electricity and responding to treatment X would be lawful consequences of the physical nature of copper and schizophrenia. In this tradition, a theory is considered to be an interconnected network of theoretical terms and observations. Laws and hypotheses relate observations to theoretical terms. Laws also relate theoretical terms to each other. The whole edifice is called a “nomological network.” For example, observable behaviors related to facial, vocal, and postural cues might be associated with theoretical constructs such as anger, sadness, and fear, which are then


combined to infer an even more abstract construct—Â�namely, “neuroticism.” A vulnerability relationship between neuroticism and other theoretical constructs, such as phobias and depression, can be formulated. The construct of depression has also been correlated with loss. By patiently tracking all the regularities, we can use constructs to predict and explain observable phenomena. To illustrate, we might predict imminent depression if a person with neuroticism experienced a major loss or humiliation in the past year, or explain a current depressive episode by referring to that prior loss. Validating the construct consists in looking for evidence that either supports or refutes the various connections. As Meehl himself observed, his notion of validating a construct was based on the logical empiricists’ notion of validating theories (Cronbach & Meehl, 1955). The different kinds of psychometric validity, such as concurrent and predictive validity, are strategies for testing theories in a logical empiricist framework. Developments in the philosophy of science subsequent to Cronbach and Meehl’s (1955) proposal have cast doubt on the logical empiricists’ views of these issues, as also acknowledged by Meehl (1990). For example, the relationship between observation/ evidence and a theory is surprisingly complicated. In testing theories, we make a conjecture about the implications that the evidence will have for the theory, but it turns out that it is often possible to reinterpret those implications. The idea that scientific theories bear a complex relationship with the evidence is called the “Duhem–Quine thesis.” When the evidence does not correspond to our predictions, we know something is wrong, but we do not know where in the network the error lies. The theory could be mistaken or incomplete; the evidence could be unreliable; or one of the background assumptions on which the theory rests could be mistaken. In Meehl’s terms, how we cope with this problem is a matter of scientific strategy. Kendler’s (1990) discussion of validation in light of constructs can also be understood from this perspective. On the basis of their disease theory, Robins and Guze (1970) listed a set of validators or evidence that would confirm disease status. For example, hypothesizing that schizophrenia is a major mental


illness with a genetic etiology indicates that increased family prevalence would be a good validator. What happened when biological psychiatrists discovered that what runs in the families of patients diagnosed with schizophrenia is not only the Feighner criteria’s major mental illness “schizophrenia,” but a wider range of conditions (including certain personality disorders)? Did they conclude that schizophrenia does run in families as predicted, but that it is actually a broad category similar to the DSM-II concept? They could have adopted this decision-Â�making strategy, but did not. Instead, they revised their conjecture about the relationship between evidence and theory by altering an auxiliary genetic hypothesis, rather than altering the “Schizophrenia is a major mental illness” hypothesis. The revised auxiliary hypothesis says that there is a genetic vulnerability to schizophrenia, and that schizophrenia is the most pathological outcome of that vulnerability, but not the only possible outcome. A category called the “schizophrenia spectrum” was introduced, and with it the construct of “schizotypal personality.” The biological psychiatrists protected the hypothesis that schizophrenia is a major mental illness. A theory consists of a network of assumptions, and a number of those assumptions can be modified to make the theory consistent with the evidence. Let us give another example. Diseases were once construed as altered states of functioning associated with pain and/or impairment. Did discovering hypertension and then not being able to confirm the presence of pain or impairment mean that hypertension was not a disease? No. Instead, physicians added a new component hypothesis to the disease theory—Â�specifically, that a disease may raise the probability of pain or impairment in the future. Similar conditions, such as coronary artery disease, could subsequently be called “diseases.” It is often possible to modify our conceptualizations about what implications empirical findings have for a theory being tested. Kendler (1990) has stated that these fundamentally nonempirical issues cannot be eliminated from classification. Meehl (1986; reprinted in Chapter 8, this volume) initially touched upon such complications with respect to his notion of “open concepts” (not fully validated). When we try


to validate a particular measure of schizophrenia, there is no gold standard against which we can evaluate it because any other standard (such as an ICD diagnosis) is also an open concept. Modifying the nomological network is also a possible response to a failed hypothesis, but Meehl was skeptical of the Duhem–Quine thesis because he believed that over time we can triangulate on the roughly more correct theory—by which he meant we would have good evidence for the basic laws and/or generalizations that make up the theory. With a similar focus, the philosopher Charles Peirce (1878/1968) said that “the opinion which is fated to be ultimately agreed to by all who investigate, is what we mean by the truth, and the object represented in this opinion is the real” (p.€77; original emphasis). Inquiry will cease when we come to a point at which there are no longer any disagreements. Everyone should agree that a good fit between theory and observation is quite difficult to achieve. Histories of past successes and abilities to predict the unexpected provide theories such as the central dogma in genetics and natural selection in evolution with a level of gravitas. The Duhem–Quine thesis, however, is quite radical. One of its consequences is that no matter how solid it appears, any construct is potentially revisable in the light of new evidence, and it is impossible for all the evidence ever to be in. It also claims that constructs can sometimes be saved from invalidation by altering another part of the theoretical network. Whenever there are multiple options for achieving a good fit, the facts alone cannot prescribe how the fit is to be decided. Therefore, we should not be too literal about any fit on which we settle. We are left with a difference in emphasis between Meehl’s rigorous approach (which focuses on constructs’ being temporarily open, with the goal being to make them incrementally tighter) and the Duhem–QuineÂ�inspired view (in which constructs will always be potentially open). We suggest that Meehl’s approach articulates an important practical goal on which progress can be made; yet the grain of truth in Duhem–Quine radicalism is not likely ever to be eliminated from the classification of the complicated, multilevel constructs of psychopathology.

Philosophical Issues

Explanatory Validity Philosopher Dominic Murphy (2006) is critical of the construct validation perspective because he claims that it focuses on discovering “lawful regularities” rather than pathological processes. The notion that science explains with respect to laws is widely believed by philosophers to be inappropriate for biology, or even physics. Cronbach and Meehl’s (1955) law-based account of theories is somewhat outdated. Murphy claims that validity should refer to a model’s ability to explain successfully how it is that putative psychiatric disorders come about. It is not disease entities, inferences, or constructs that are valid; rather, models are valid when they track genuine causal structures. There is nothing new about focusing on causality, as both disease realists and logical empiricists did so. The disease realists took their inspiration from the infectious-Â�disease model in biology, whereas the logical empiricists were inspired by the explanatory laws of physics. What is new about model-based approaches is that they have a more pluralistic view of cause, and they also undermine the descriptive–Â�etiological dualism that has driven a good deal of nosological bickering for 30 years. The infectious-Â�disease approach motivated psychopathologists to search for “necessary causes” (Guze, 1992). A necessary cause is one that must occur for the target effect to occur. If one can prevent or alter such a cause, then one can also eliminate the effect. Many thinkers have claimed that valid psychiatric disorders should also be sorted with respect to etiology, implicitly understood to be some kind of necessary cause (Andreasen, 1984; Kendell & Jablensky, 2003). Infectious diseases and Mendelian genetic conditions have both casual and sorting value, but they have not been fruitful models for understanding other conditions. One reason given for this failure is to claim that the causes are there, but our basic science has not yet advanced enough for us to discover them. Some also claim that the syndromes identified in the current nosology are causally heterogeneous entities, and as a result function as barriers to the search for causes (Charney et al., 2002; Livesley, 2003; Parshall & Priest, 1993).


Another strategy is to adopt a broader notion of “cause”—one that includes risk factors and vulnerabilities. Risk factors include biological, psychological, and sociocultural variables that each contribute to producing an effect, but none of which is sufficient. Disorders are explained as a result of multifactorial packages, but any one of the factors by itself has limited causal force. As argued by Kendler (2005) with respect to genetics, the problem is not that psychiatric disorders lack a genetic basis, or that most of the important genes remain hidden; the problem is that any single gene on average accounts for only a small proportion of the variance of most disorders. Many different genes raise the risk of developing psychiatric disorders, but usually not by much. Genes also have nonspecific effects, raising the risk for several psychiatric disorders rather than having a 1:1 relationship with a particular disorder. Other genes may have protective effects, lowering one’s risk. New philosophical approaches to understanding scientific explanations that emphasize models offer good reasons for believing that the conventional causal theory of explanation in psychiatric nosology can be improved upon (Giere, 1988; Harré, 1986). Models are simplified and idealized representations of phenomena that are not literally true in all respects, but that still tell us what phenomena are like. A famous example of an idealized model is Freud’s model of id, ego, and superego. Other examples of models include the central dogma in molecular genetics, the model of allopatric speciation in evolutionary theory, and neural network models in cognitive science. Models only represent the information that is deemed relevant. To make this point, philosophers such as Ronald Giere (1999) compare models to maps. A road map is a model of a geographical area that is realistic to a greater or lesser extent, but usually a partial representation. It does not model the vegetation, the diversity of animal life, the uphill and downhill sections of the road, the style of the buildings, or the population density. Although they are not intended to be literally true, models should have what philosophers call “verisimilitude,” or approximate truth. An example of approximate truth is Carnot’s explanation of heat engines with


reference to a fluid called “caloric.” Although caloric fluid does not exist, Carnot’s description can be considered the first formulation of the second law of thermodynamics. According to Murphy (2006), DSM and ICD syndromes are like maps as well. They are more or less realistic, but, like maps, they represent information relative to a particular set of interests, rather than being literally true. Instead of viewing diagnostic categories as “natural kinds,” Murphy thinks they should be seen as idealized representations called “exemplars.” In doing so, he tentatively joins others (e.g., Ghaemi, 2003; Zachar, 2000a, 2000b, 2008) who are suspicious of using the philosophical concept of natural kind to understand psychiatric classification. Murphy notes that an exemplar should itself be mapped with a typical causal profile, or even families of profiles. The casual profiles will themselves be idealizations that become more or less realistic as details are added and subtracted, depending on one’s explanatory purposes. In psychopathology, models can potentially provide coarse-Â�grained accounts of a disorder in general, or can be made so fineÂ�grained that they attempt to model the development and maintenance of a disorder in a single individual. One model of MDD for 3 million patients risks being a gross approximation, but 3 million idiographic models of depression would be impractical, and neither is remotely the best model for all purposes. Hesse (2000) notes that philosophers have proposed a range of model-based views for understanding scientific explanation. An approach that is particularly relevant for psychiatric nosology, and is often used in biology, is called the “mechanistic model.” Rather than basing explanations on general laws and nomological networks, mechanistic models seek to explain by understanding how systems work, usually by decomposing them into parts and proposing theories about how those parts work together. They are often described as attempts to understand how causal processes are actually implemented in the world. Constructing such models involves more than the search for necessary causes. For example, in developing a model of an internal combustion engine, we would not say that the fuel injector causes the car to go. Instead,


the model seeks to describe what role the fuel injector plays in the overall functioning of the system. Whether these roles will be construed causally depends on the question being asked. In asking about why a car is overemitting pollutants, if we find out that the injector is delivering too much fuel to the cylinders, then we would assign the injector a causal role. Most of the time, however, a fuel injector is just seen as a part of the engine with a certain functional role. Most of the modeling work involves understanding functional roles rather than seeking causes. As Kendler (2008a) notes, mechanistic models begin as proposals or sketches of how a system might work. Over time, the details are filled in and the sketch is updated. Models are tested until things start to fit together; in this way, they are evidence-based. Once the description reaches a certain level of detail, it can be used to do explanatory work. Bechtel and Abrahamsen (2005) say that mechanistic models explain why by describing how. One final aspect about models is that they can be formulated at different levels of generality. Murphy (2006) believes that the primary explanatory level for a scientific psychopathology will be that of cognitive neuroscience, but he and most others assume that the better explanations will be multilevel. The genetic, computational, psychological, social-�psychological, and cultural levels will all work together, and no single model will fill in all the details. What is new is an expectation that the details of how cultures and brains interact should be modeled in terms of empirically supported causal structures, rather than left at the level of slogans.

Literalism and Science We bring this chapter to a close with some thoughts about diagnostic literalism. It may have occurred to readers that one of the goals of science is to offer literally true rather than metaphorical accounts of natural phenomena. For example, scientists believe that the evolutionary account of organic diversity is literally true, but that the creation story in Genesis is not. They also believe that the Copernican theory is literally true, whereas the Ptolemaic theory was not. Shouldn’t nosology aspire to be like Darwin and Copernicus? Our response is “Yes, of course it should.”

Philosophical Issues

The current scientific understanding of the solar system is a good choice for exploring literalism. The literalist story is that the sun occupies the center of the solar system, and the planets revolve around the sun. The Copernican model is well supported. There are no good competitors to it, nor does it seem likely that there will ever be. From the sun in the center outward, the planets are in this order: Mercury, Venus, Earth, Mars, Jupiter, Saturn, Uranus, Neptune, and Pluto. This model has been scientific orthodoxy for over 75 years. But wait. Are there literally 9 planets? Or are there 8, or 10? One of the amazing things about Pluto is that its existence and location were predicted by Percival Lowell in the early 1900s. Astronomers discovered that the observed orbit of Uranus was not what the scientific calculations predicted, but deduced that if there were another planet beyond Neptune, the predicted orbit would more closely match the observed orbit (Sobel, 2005). Lowell made some calculations, and Clyde Tombaugh found the planet Pluto in 1930, just where Lowell said it would be. In 2005, astronomers at the Palomar Observatory discovered an object beyond Pluto that later appeared to be a 10th planet, now named Eris. One of the reasons for considering Eris a 10th planet is that it is 20–30% larger than Pluto and also has at least one moon (Ridpath & Tirion, 2008). An interesting feature of Eris is that it has a highly eccentric orbit. Other planets orbit the sun in roughly the same plane; their orbits mostly differ by being farther away from the sun (Millennium House, 2007). The orbit of Eris, on the other hand, is tilted 44 degrees relative to the orbital plane of Earth. This made some astronomers reluctant to call Eris a real planet. It turns out that Pluto has a similar problem. Its orbit is tilted 17 degrees relative to that of Earth. Furthermore, for 20 years of its 250-year orbit around the sun, Pluto is actually closer to the sun than is Neptune (Sobel, 2005). It also seems to be a member of a large swarm of objects that orbit the sun beyond Neptune, called the “Kuiper belt.” Many “Kuiper belt objects” (KBOs) also have their own moons. Echoing the American Psychiatric Association’s 1973 decision to declassify hom*osexuality as a mental disorder, in 2006 the


International Astronomical Union voted to demote Pluto from a real planet to a dwarf planet (Millennium House, 2007). At the same time, Ceres, which lies between Mars and Jupiter, was promoted from an asteroid to a dwarf planet. An alternative proposal would have made Ceres a planet. There is apparently continuing disagreement about these classifications, especially regarding the scientific criteria for “valid” planets. For example, there was a concern that if Pluto and Eris were accepted as planets, several other KBOs would also be planets, and 30 or more planets is too many. Compare this to claims that a 30–50% lifetime prevalence for experiencing a mental disorder is too high (Mechanic, 2003; Narrow, Rae, Robins, & Regier, 2002). A typical solution to such conundrums is to choose a polythetic set of distinctions that also respect our intuitions—for example, defining “planet” so that Mercury through Uranus are included but Pluto is excluded. Some believe that such nonempirical considerations are scientifically unseemly, but such complexities are less of a problem for those who construe classification as an ongoing deliberative activity. Quite likely, there are astronomical thinkers with tendencies toward literalism who want to keep Pluto a planet. We would hope that there are not any who want to keep the number of planets at 9. Should we call Pluto a planet, a dwarf planet, or a KBO? Closer to home, is chronic dysthymia better termed depressive personality disorder? Is social phobia really avoidant personality disorder? Should we call minor depression a psychiatric disorder, or is it a subthreshold condition that is occasionally a focus of clinical attention? Some of the arguments about minor disorders in psychiatry and minor or dwarf planets in astronomy are quite similar. Reasons can be given for making classificatory distinctions, and facts about the world are among those reasons. There is a close relationship between facts (such as the nature of orbits or the presence of moons) and nonempirical preferences about what role astronomers want planets to play in classification. The same is true for facts about genetics and cognitive–Â�emotional processes and what mental health professionals want a diagnosis to do. In psychopathology,


once there is an agreement on what we want a diagnosis to do (e.g., to identify etiologically hom*ogeneous groups), then science can address that problem. Science can, however, address multiple diagnostic purposes, and there is no good reason to believe that all scientifically supported classifications will be consistent with each other. Studying history also helps occasionally. For example, when Pluto was discovered and named a planet, it was assumed that in order to influence Uranus’s orbit, it had to be much larger than it was eventually discovered to be. More importantly, the astronomers had made a mistake when predicting the orbit of Uranus. In their calculations, they plugged in an incorrect size for Neptune’s mass. If they had plugged in the correct size, the difference between the predicted and the observed orbits of Uranus would not have been so great (Sobel, 2005). It was mere accident that Tombaugh found a big object where Lowell said it should be. So are there literally 8, 9, or 10 planets? As noted, if there are 10, the number of planets is likely to be greater than 10 because there are a lot of big spherical objects with moons out there. Many readers are probably thinking that there are 8, but what if Earth-like “planets” in other solar systems have orbits like those of Pluto or Eris? We hope it has occurred to readers that the taxonomic problem in astronomy is not about developing either a literal or a metaphorical classification of the solar system. The 9-planet model was not a metaphor, and a commitment to its literal truth would be scientifically harmful. Returning to the more specific topic of philosophical problems in the classification of psychopathology, taxonomists should reject the assumption that in order to be as scientific as the classification systems of the natural and the biological sciences, psychiatric nosology should be construed literally. Seeking the literal truth can be an important motivator, but also, as just noted, has the potential to do harm.

Conclusion In deciding which philosophical issues to explore in this chapter, we have selected conceptual problems that are internal to the


scientific study of psychopathology. With respect to dimensional and categorical models, we have advocated a pragmatic perspective that prefers to utilize the strengths of both approaches. We have argued that conceptualizing scientific classification as an attempt to “carve nature at its joints” helps the field aspire to classify the world objectively, but also makes diagnostic literalism and reification of classifications harder to avoid. With respect to the concept of “clinical significance,” we have stated that making distinctions among “optimal,” “normal,” “compromised,” and “impaired” requires not just the discovery of facts, but also evaluative judgments. In some cases these judgments can be rationally articulated, although the level of consensus on values will vary from disorder to disorder. We have further argued that it would be better if disorders were conceptualized in terms of pathological processes, rather than being solely descriptive. Wakefield’s harmfulÂ�dysfunction model provides a quite helpful conceptual definition of “disorder,” but we have suggested that objective dysfunctions may not be subject to empirical confirmation in many cases. Describing four different perspectives on the validity of psychopathological constructs, we do not believe that any of those perspectives should be considered to be the correct approach to validity. In our view, a “validity-of-Â�inferences model” represents a consensus view on validity. We have suggested that a philosophically updated version of the construct validity approach emphasizing local explanatory models—Â� models that are calibrated to the purposes one has for classifying—Â�offers a promising way forward. Acknowledgment Andrea Solomon and Steve LoBello provided helpful commentary on an earlier version of this chapter.

References Achenbach, T. M. (1995). Empirically based assessment and taxonomy: Applications to clinical research. Psychological Assessment, 3, 261–274.

Philosophical Issues Achenbach, T. M. (2001). Empirically based assessment and taxonomy across the life span. In J. E. Helzer & J. J. Hudziak (Eds.), Defining psychopathology in the 21st century (pp.€155– 168). Washington, DC: American Psychiatric Press. Achenbach, T. M., Bernstein, A., & Dumenci, L. (2005a). DSM-oriented scales and statistically based syndromes for ages 18 to 59: Linking taxonomic paradigms to facilitate multitaxonomic approaches. Journal of Personality Assessment, 84, 49–63. Achenbach, T. M., Bernstein, A., & Dumenci, L. (2005b). A snark or a boojum?: Exploring multitaxonomic possibilities and building on Widiger’s commentary. Journal of Personality Assessment, 84, 66–69. Ahn, W., Flanagan, E. H., Marsh, J. K., & Sanislow, C. A. (2006). Beliefs about essences and the reality of mental disorders. Psychological Science, 17, 759–766. Alloy, L. B., Albright, J. S., Abramson, L. Y., & Dykman, B. M. (1990). Depressive realism and nondepressive optimistic illusions: The role of the self. In R. E. Ingram (Ed.), Contemporary psychological approaches to depression: Treatment, research, and theory (pp.€71–86). New York: Plenum Press. Andreasen, N. C. (1984). The broken brain: The biological revolution in psychiatry. New York: Harper & Row. Bechtel, W., & Abrahamsen, A. (2005). Explanation: A mechanistic alternative. Studies in History and Philosophy of the Biological and Biomedical Sciences, 36, 421–441. Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: International Universities Press. Becker, D., Lomranz, J., Pines, A., Shmotkin, D., Nitza, E., BennAmitay, G., et al. (2001). Psychological distress around menopause. Psychosomatics, 42, 252–257. Bell, C. (2004). Racism: A mental illness. Psychiatric Services, 55, 1343. Bentall, R. P. (1992). A proposal to classify happiness as a psychiatric disorder. Journal of Medical Ethics, 18, 94–98. Berrios, G. E. (1991). Affective disorders in old age: A conceptual history. International Journal of Geriatric Psychiatry, 6, 337–346. Brendel, D. (2006). Healing psychiatry. Cambridge, MA: MIT Press. Charney, D. S., Barlow, D. H., Botteron, K., Cohen, J. D., Goldman, D., Gur, R. E., et al. (2002). Neuroscience research agenda to guide

145 development of a pathophysiologically based classification system. In D. J. Kupfer, M. B. First, & D. A. Regier (Eds.), A research agenda for DSM-V (pp.€31–83). Washington, DC: American Psychiatric Press. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Dao, T. K., Prevatt, F., & Horne, H. L. (2007). Differentiating psychotic patients from nonpsychotic patients with the MMPI-2 and Rorschach. Journal of Personality Assessment, 90, 93–101. Dewey, J. (1938). Logic: The theory of inquiry. New York: Holt. Dewey, J. (1958). Experience and nature. New York: Dover. (Original work published 1928) Ereshefsky, M. (1992). Eliminative pluralism. Philosophy of Science, 59, 671–690. Feigl, H. (1970). The “orthodox” view of theories: Remarks in defense as well as critique. In M. Radner & S. Winokur (Eds.), Minnesota studies in the philosophy of science (Vol. 4, pp.€3–16). Minneapolis: University of Minnesota Press. First, M. B., Cuthbert, B., Krystal, J. H., Malison, R., Offord, D. R., Reiss, D., et al. (2002). Personality disorders and relational disorders. In D. J. Kupfer, M. B. First, & D. A. Regier (Eds.), A research agenda for DSM-V (pp.€ 123–129). Washington, DC: American Psychiatric Association. Fulford, K. W. M. (1989). Moral theory and medical practice. Cambridge, UK: Cambridge University Press. Ghaemi, S. N. (2003). The concepts of psychiatry. Baltimore: Johns Hopkins University Press. Giere, R. N. (1988). Explaining science: A cognitive approach. Chicago: University of Chicago Press. Giere, R. N. (1999). Science without laws. Chicago: University of Chicago Press. Goldberg, L. R., & Velicer, W. F. (2006). Principles of exploratory factor analysis. In S. Strack (Ed.), Differentiating normal and abnormal personality (pp.€209–237). New York: Springer. Goodman, N. (1978). Ways of worldmaking. Indianapolis, IN: Hackett. Guze, S. (1992). Why psychiatry is a branch of medicine. New York: Oxford University Press. Hacking, I. (1999). The social construction of what? Cambridge, MA: Harvard University Press.

146 Harré, R. (1986). Varieties of realism. Oxford, UK: Blackwell. Haslam, N. (2000). Psychiatric categories as natural kinds: Essentialist thinking about mental disorder. Social Research, 67, 1032–1058. Haslam, N. (2007). The latent structure of mental disorders: A taxometric update on the categorical vs. dimensional debate. Current Psychiatry Reviews, 3, 172–177. Haslam, N., & Williams, B. (2006). Taxometrics. In S. Strack (Ed.), Differentiating normal and abnormal personality (pp.€ 284–308). New York: Springer. Helzer, J. R., Kraemer, H. C., & Krueger, R. F. (2006). The feasibility and need for dimensional psychiatric diagnoses. Psychological Medicine, 36, 1671–1680. Hempel, C. G. (1966). Philosophy of natural science. Englewood Cliffs, NJ: Prentice-Hall. Hesse, M. (2000). Models and analogies. In W. H. Newton-Smith (Ed.), A companion to the philosophy of science (pp.€299–307). Oxford, UK: Blackwell. Horwitz, A. V., & Wakefield, J. C. (2007). The loss of sadness. New York: Oxford University Press. James, W. (1975). Pragmatism and The meaning of truth. Cambridge, MA: Harvard University Press. (Original works published 1907, 1909) Joiner, T., & Coyne, J. C. (Eds.). (1999). The interactional nature of depression. Washington, DC: American Psychological Association. Kendell, R. (1975). The concept of disease. British Journal of Psychiatry, 127, 305–315. Kendell, R., & Jablensky, A. (2003). Distinguishing between the validity and utility of psychiatric disorders. American Journal of Psychiatry, 160, 4–12. Kendler, K. S. (1980). The nosologic validity of paranoia (simple delusional disorder): A review. Archives of General Psychiatry, 37, 699–906. Kendler, K. S. (1990). Toward a scientific psychiatric nosology. Archives of General Psychiatry, 47, 969–973. Kendler, K. S. (1996). Major depression and generalised anxiety disorder: Same genes, (partly) different environments—Â�revisited. British Journal of Psychiatry, 168(Suppl. 30), 68–75. Kendler, K. S. (2005). “A gene for .€.€. ”: The nature of gene action in psychiatric disorders. American Journal of Psychiatry, 162, 1243– 1252. Kendler, K. S. (2008a). Explanatory models for

CONCEPTUAL ISSUES IN CLASSIFICATION psychiatric illness. American Journal of Psychiatry, 165 , 695–702. Kendler, K. S. (2008b). Review of The loss of sadness: How psychiatry transformed normal sorrow into depressive disorder. Psychological Medicine, 38, 148–150. Kendler, K. S., & Zachar, P. (2008). The incredible insecurity of psychiatric nosology. In K. S. Kendler & J. Parnas (Eds.), Philosophical issues in psychiatry: Explanation, phenomenology and nosology (pp.€ 368–383). Baltimore: Johns Hopkins University Press. Kessler, R. C. (2002). The categorical versus dimensional assessment controversy in the sociology of mental illness. Journal of Health and Social Behavior, 43, 171–188. Kessler, R. C., Merikangas, K. R., Berglund, P., Eaton, W. W., Koretz, D. S., & Walters, E. E. (2003). Mild disorders should not be eliminated from the DSM-V. Archives of General Psychiatry, 60, 1117–1122. Kripke, S. (1980). Naming and necessity. Cambridge, MA: Harvard University Press. Krueger, R. F. (1999). The structure of common mental disorders. Archives of General Psychiatry, 56, 921–926. Krueger, R. F., Caspi, A., Moffitt, T. E., & Silva, P. A. (1998). The structure and stability of common mental disorders (DSM-III-R): A longitudinal–Â�epidemiological study. Journal of Abnormal Psychology, 107, 216–227. Krueger, R. F., Watson, D., & Barlow, D. H. (2005). Introduction to the special section: Toward a dimensionally based taxonomy of psychopathology. Journal of Abnormal Psychology, 114, 491–493. Lilienfeld, S. O., & Waldman, I. D. (2004). Comorbidity and Chairman Mao. World Psychiatry, 3, 26–27. Livesley, W. J. (2003). Diagnostic dilemmas in classifying personality disorder. In K. A. Phillips, M. B. First, & H. A. Pincus (Eds.), Advancing DSM (pp.€153–189). Washington, DC: American Psychiatric Association. Livesley, W. J. (2006). Dimensional assessment of personality pathology (DAPP) approach to personality disorder. In S. Strack (Ed.), Differentiating normal and abnormal personality (pp.€401–429). New York: Springer. Markon, K. E., & Krueger, R. F. (2006). Latent variable modeling. In S. Strack (Ed.), Differentiating normal and abnormal personality (pp.€239–256). New York: Springer. Mayr, E. (1982). The growth of biological

Philosophical Issues thought: Diversity, evolution, and inheritance. Cambridge, MA: Harvard University Press. Mayr, E. (1988). Toward a new philosophy of biology: Observations of an evolutionist. Cambridge, MA: Harvard University Press. Mayr, E. (1991). One long argument: Charles Darwin and the genesis of modern evolutionary thought. Cambridge, MA: Harvard University Press. Mechanic, D. (2003). Is the prevalence of mental disorders a good measure of the need for services? Health Affairs, 22, 8–20. Meehl, P. E. (1986). Diagnostic taxa as open concepts: Metatheoretical and statistical questions about reliability and construct validity in the grand strategy of nosological revision. In T. Millon & G. Klerman (Eds.), Contemporary directions in psychopathology (pp.€215–231). New York: Guilford Press. Meehl, P. E. (1990). Appraising and amending theories: The strategy of Lakatosian defense and two principles that warrant it. Psychological Inquiry, 1, 108–141. Meehl, P. E. (1993). Philosophy of science: Help or hindrance. Psychological Reports, 72, 707– 733. Millenium House. (2007). Astronomica. Elanora Heights, Australia: Author. Mishler, B. D., & Brandon, R. N. (1987). Individuality, pluralism, and the phylogenetic species concept. Biology and Philosophy, 2, 397–414. Murphy, D. (2006). Psychiatry in the scientific image. Cambridge, MA: MIT Press. Narrow, W. E., Rae, D. S., Robins, L. N., & Regier, D. A. (2002). Revised prevalence estimates of mental disorders in the United States. Archives of General Psychiatry, 59, 115–123. Parshall, A. M., & Priest, R. G. (1993). Nosology, taxonomy and the classification conundrum of the functional psychoses. British Journal of Psychiatry, 162, 227–236. Peirce, C. S. (1968). How to make our ideas clear. In P. Kurtz (Ed.), American philosophy in the twentieth century (pp.€ 62–79). New York: Macmillan. (Original work published 1878) Pincus, H. A., McQueen, L. E., & Elinson, L. (2003). Subthreshold mental disorders: Nosological and research recommendations. In K. A. Phillips, M. B. First, & H. A. Pincus (Eds.), Advancing DSM (pp.€129–144). Washington, DC: American Psychiatric Association. Putnam, H. (1975). The meaning of “meaning.” In H. Putnam, Philosophical papers: Vol. 2.

147 Mind, language and reality (pp.€ 215–271). Cambridge, UK: Cambridge University Press. Quine, W. V. (1969). Ontological relativity and other essays. New York: Columbia University Press. Regier, D. A. (2003). Mental disorder diagnostic theory and practical reality: An evolutionary perspective. Health Affairs, 22, 21–27. Regier, D. A. (2007). Dimensional approaches to psychiatric classification: Refining the research agenda for DSM-V: An introduction. International Journal of Methods in Psychiatric Research, 16(Suppl. 1), S1–S5. Ridley, M. (1986). Evolution and classification: The reformation of cladism. New York: Longman. Ridpath, I., & Tirion, W. (2008). Stars and Â�planets. Princeton, NJ: Princeton University Press. Robins, E., & Guze, S. (1970). Establishment of diagnostic validity in psychiatric illness: Its application to schizophrenia. American Journal of Psychiatry, 126, 983–987. Rossiter, L. H. (2006). The liberal mind: The psychological causes of political madness. St. Charles, IL: Free World Books. Sadler, J. Z. (2005). Values and psychiatric diagnosis. Oxford, UK: Oxford University Press. Savage, M. (2005). Liberalism is a mental disorder. Nashville, TN: Nelson Current. Schaffner, K. F. (1993). Discovery and explanation in biology and medicine. Chicago: University of Chicago Press. Simms, L. J., & Clark, L. A. (2006). The Schedule of Nonadaptive and Adaptive Personality (SNAP): A dimensional measure of traits relevant to personality and personality pathology. In S. Strack (Ed.), Differentiating normal and abnormal personality (pp.€ 431–450). New York: Springer. Sobel, D. (2005). The planets. New York: Penguin. Spitzer, R. L., & Wakefield, J. W. (1999). DSMIV diagnostic criteria for clinical significance: Does it help solve the false positive problem? American Journal of Psychiatry, 156, 1856– 1864. Udovitch, M. (2002, September 8). The way we live now: 9-08-02: Phenomenon; A secret society of the starving. New York Times. Retrieved May 30, 2008, from www.nytimes. com Wakefield, J. C. (1992). The concept of mental disorder: On the boundary between biological

148 facts and social values. American Psychologist, 47, 373–388. Wakefield, J. C. (1999a). Evolutionary versus prototype analyses of the concept of disorder. Journal of Abnormal Psychology, 108, 374– 399. Wakefield, J. C. (1999b). Mental disorder as a black box essentialist concept. Journal of Abnormal Psychology, 108, 465–472. Wakefield, J. C. (2000). Aristotle as sociobiologist. Philosophy, Psychiatry, and Psychology, 7, 17–44. Wakefield, J. C. (2006). Personality disorder as harmful dysfunction: DSM’s cultural deviance requirement reconsidered. Journal of Personality Disorders, 20, 157–169. Weissman, M. M. (1979). The myth of involutional melancholia. Journal of the American Medical Association, 242, 742–744. Widiger, T. A., & Samuel, D. B. (2005). Diagnostic categories or dimensions? A question for the Diagnostic and Statistical Manual of Mental Disorders–Fifth Edition. Journal of Abnormal Psychology, 114, 494–504.

CONCEPTUAL ISSUES IN CLASSIFICATION van Frassen, B. C. (1980). The scientific image. Oxford, UK: Oxford University Press. Zachar, P. (2000a). Folk taxonomies should not have essences either. Philosophy, Psychiatry, and Psychology, 7, 191–194. Zachar, P. (2000b). Psychiatric disorders are not natural kinds. Philosophy, Psychiatry, and Psychology, 7, 167–182. Zachar, P. (2006). The classification of emotion and scientific realism. Journal of Theoretical€ and Philosophical Psychology, 26, 120– 138 Zachar, P. (2008). Real kinds, but no true taxonomy: An essay in psychiatric systematics. In K. S. Kendler & J. Parnas (Eds.), Philosophical issues in psychiatry: Explanation, phenomenology and nosology (pp.€ 327–355). Baltimore: Johns Hopkins University Press. Zachar, P., & Bartlett, S. (2001). Basic emotions and their biological substrates. Consciousness and Emotion, 2, 189–221. Zachar, P., & Kendler, K. S. (2007). Psychiatric disorders: A conceptual taxonomy. American Journal of Psychiatry, 164, 557–565.

Chapter 7

Classification Considerations in€Psychopathology and Personology Theodore Millon

thoughts I expressed over 40 years ago Ilon,nconcerning the character of theory (Mil1969), I voiced my chagrin that nature was not made to suit our need for a tidy and well-Â�ordered universe. Quite evidently, the complexity and intricacy of the natural world make it difficult not only to establish clear-cut relationships among phenomena, but to find simple ways in which these phenomena can be classified or grouped. In our desire to discover the essential order of nature, we find it necessary to concern ourselves with only a few of the infinite number of elements that could be chosen; in this selection, we narrow our choice only to those aspects of nature that we believe best enable us to answer the questions we have posed. Moreover, the elements we choose are labeled, transformed, and reassembled in a variety of ways, but we must bear in mind that these labels and transformations are not “realities.” The various concepts and categories that we construct as scientists are only optional tools to guide our observation and interpretation of the natural world; different concepts and categories may be formulated as alternative approaches to the understanding of the same subject of inquiry. These tools are especially necessary when the terrain we

face is as uncharted as the taxonomy of psychopathology, and the materials of which it is composed are as intractable as they are. The subject areas which subdivide the natural world differ in the degree to which their phenomena are inherently differentiated and organized. Some areas are “naturally” more articulated and quantifiable than others. To illustrate, the laws of physics relate to highly probabilistic processes in many of its most recondite spheres, but the features of our everyday physical world are highly ordered and predictable. Theories in this latter realm of physics (e.g., mechanics, electricity) serve largely to uncover the lawful relationships that do in fact exist in nature: it was the task of physicists at the turn of the last century to fashion a network of constructs that faithfully mirrored the universal nature of the phenomena they studied. By contrast, probabilistic realms of physical analysis (e.g., short-lived elementary particles) or systems of recent evolutionary development (e.g., human interactions) are inherently weakly organized, lacking either articulated or invariant connections among their constituent elements. In knowledge domains that relate to these less ordered spheres of nature (the “softer” sciences), classifiers and theorists 149


find it necessary to impose a somewhat arbitrary measure of systematization; in so doing, they construct a degree of clarity and coherence that is not fully consonant with the “naturally” unsettled and indeterminate character of their subject. Rather than equivocate strategically, or succumb to the “futility of it all,” these workers make noble or pretentious efforts to arrange and categorize these inexact and probabilistic elements so that they simulate a degree of precision and order transcending that which they intrinsically possess. For instance, in fields such as economics and psychopathology, categories and classifications are in considerable measure splendid fictions, compelling notions, or austere formulas devised to give coherence to their inherently imprecise subjects. The logic, substance, and structures created or imposed as a means of giving order to the phenomena of psychopathology are the principal topics of this chapter. It may serve as either a pedagogical introduction or a comprehensive review, depending on the reader’s starting point. Humans developed reliable and useful classifications long before the advent of modern scientific thought and methods. Information, skill, and instrumentation were achieved without “science” and its symbolic abstractions and techniques of research. If useful classifications could be acquired through intelligent observation and common sense alone, what special values are derived by applying the complicated and rigorous procedures required in developing explicit criteria, taxonic hom*ogeneity, and diagnostic efficiency? Are rigor, clarity, precision, and experimentation more than compulsive and picayunish concerns for details, more than the pursuit of the honorific title of “science”? Are the labors of differentiating attributes or exploring optimal cutoff scores in a systematic fashion worth the time and effort involved? There is little question in the “age of science” that the answer would be yes. But why? What are the distinguishing virtues of precision in terminology, the specification of observable conceptual referents, the analysis of covariant attribute clusters? What sets these procedures apart from everyday methods of categorizing knowledge? Because the number of ways we can observe, describe, and organize the natural


world is infinite, the terms and concepts we create to represent these activities are often confusing and obscure. For example, different words are used to describe the same behavior, and the same word is used for different behaviors. Some terms are narrow in focus, others are broad, and some are difficult to define. Because of the diversity of events to which we can attend, or the lack of precision in the language we employ, different processes are confused and similar events get scattered in hodgepodge fashion across a scientific landscape; as a consequence, communication gets bogged down in terminological obscurities and semantic controversies. One of the goals of formalizing the phenomena that constitute a scientific subject is to avoid this morass of confusion. Not all phenomena related to the subject need be attended to at once. Certain elements may be selected from the vast range of possibilities because they seem relevant to the solution of a specific question. And to create a degree of reliability or consistency among the efforts of those interested in a subject, its elements are defined as precisely as possible and classified according to their core similarities and differences (Dougherty, 1978; Tversky, 1977). In a subject such as psychopathology, these classes or categories are given specific labels, which serve to represent them. This process of definition and classification is indispensable for systematizing observation and knowledge. Are conceptual definition and classification possible in psychopathology? Can these most fundamental of scientific activities be achieved in a subject that is inherently inexact, of only modest levels of intrinsic order—one in which even the very slightest variations in context or antecedent conditions (often of a minor or random character) produce highly divergent outcomes (Bandura, 1982)? Because this “looseness” within the network of variables in psychopathology is unavoidable, are there any grounds for believing that such endeavors could prove more than illusory? Persuasive answers to this question of a more philosophical nature must be bypassed in this all-too-Â�concise chapter; those who wish to pursue this line of analysis would gain much by reading, among others, Pap (1953), Hempel (1965), and Meehl (1978). Let us touch, albeit brief-

Considerations in Psychopathology and Personology

ly, on a more tangible and psychologically based rationale for believing that formal classification in psychopathology may prove to be at least a moderately fruitful venture.

Why May Formal Classification Be Useful? There is a clear logic to classifying “syndromes” in medical disorders. Bodily changes wrought by infectious diseases and structural deteriorations repeatedly display themselves in reasonably uniform patterns of signs and symptoms that “make sense” in terms of anatomical structures’ and physiological processes’ alterations and dysfunctions. Moreover, these biological changes provide a foundation not only for identifying the etiology and pathogenesis of these disorders, but also for anticipating their course and prognosis. Logic and fact together enable us to construct a rationale to explain why most medical syndromes express themselves in the signs and symptoms they do, as well as the sequences through which they unfold. Can the same be said for psychopathological classifications? Is there a logic, perhaps evidence, for believing that certain forms of clinical expression (e.g., behaviors, cognitions, affects, mechanisms) cluster together as medical syndromes do—that is, not only covary frequently, but “make sense” as coherently organized and reasonably distinctive groups of characteristics? Are there theoretical and empirical justifications for believing that the varied features of personality display a configurational unity and expressive consistency over time? Will the careful study of individuals reveal congruency among such attributes as overt behavior, intrapsychic functioning, and biophysical disposition? Are this coherence and stability of psychological functioning valid phenomena—that is, not merely imposed upon observed data by virtue of clinical expectation or theoretical bias? There are reasons to believe that the answer to each of the preceding questions is yes. Stated briefly and simply, the observations of covariant patterns of signs, symptoms, and traits may be traced to two facts: People possess relatively enduring biophysical dispositions that give a consistent color-


ation to their experience; and the range of experiences to which people are exposed throughout their lives is both limited and repetitive (Millon, 1969, 1981). Given the limiting and shaping character of these biogenic and psychogenic factors, it should not be surprising that individuals develop clusters of prepotent and deeply ingrained behaviors, cognitions, and affects that clearly distinguish them from others of dissimilar backgrounds. Moreover, once several components of a particular clinical pattern are identified, knowledgeable observers are able to trace the presence of other, unobserved, but frequently correlated features seen in that pattern. If we accept the assumption that most people do display patterns of internally consistent characteristics, we are led next to the question of whether groups of patients evidence commonality in the patterns they display. The notion of clinical categories rests on the assumption that there are some such shared covariances—for example, regular groups of diagnostic signs and symptoms that can confidently be used to distinguish certain classes of patients. (However, this assumption does not negate the fact that patients classified into categories display considerable differences as well—Â�differences we routinely observe with medical diseases.) Although grievances itemizing the inadequacies of both our current and historic systems of psychopathology have been voiced for years, as have suggestions that endeavors to refine these efforts are fussy and misdirected, if not futile and senseless pretensions that should be abandoned, the presence of categorical systems is both unavoidable (owing to our human linguistic and attribution habits) and inevitable (owing to our need to differentiate and to record, at the very least, the most obvious of dissimilarities among psychologically impaired individuals). Given the fact that one or another set of categories is inevitable—or, as Kaplan (1964, p.€279) once phrased it, “it is impossible to wear clothing of no style at all”—it would appear both sensible and fitting that we know the explicit basis upon which such distinctions are to be made, rather than have them occur helter-Â�skelter in nonpublic and nonverifiable ways. Furthermore, if psychopathology is to evolve into a true science, its diverse phenomena must be subject to


formal identification, differentiation, and quantification procedures. Acts such as diagnosis and assessment presuppose the existence of discernible phenomena that can be recognized and measured. Logic necessitates, therefore, that psychopathological states and processes be distinguished from one another, being thereby categorizable in some degree before they can be subjected to identification and quantification. The number of categories that can be distinguished in a classification schema will depend in part on the incisiveness with which diagnosticians make their clinical observations and the creative inferences they draw from them. As has been discussed in earlier pages, classification data may legitimately be derived both from concrete observations and from abstract inferences. In spite of a long history of brilliant cogitations, psychopathological nosology still resembles Ptolemy’s astronomy of over 2,000 years ago: Our diagnostic categories describe, but they do not really explain. Like so many crystalline spheres, each lies in its own orbit, for the most part uncoordinated with the others. We do not know why the universe takes its ostensible form. There is no law of gravity that undergirds and binds our psychopathological cosmos together. In fact, the word “cosmos” implies an intrinsic unity, a laudable ideal, which is not appropriate in its usage here: Our “star charts,” our DSMs and ICDs, remain aggregations of taxa, not true taxonomies. Because of their reliability but dubious validity, our field possesses the illusion of science but not its substance. Such a state of affairs is simply unscientific. Our most radical (albeit reactional) alternative would be to discard taxonomies altogether. This, of course, would be impossible, as a taxonomy serves indispensable clinical and scientific functions. Clinically, it provides a means of organizing pathological phenomena—the signs and symptoms or manifestations of mental disorder. By abstracting across persons, a taxonomy formalizes certain clinical commonalities and relieves the clinician of the burden of conceptualizing each patient sui generis, as an entity so existentially unique it has never been seen before, nor ever will be seen again. For psychopathology to be practiced at all, there cannot be as many groups as individu-


als. Even if the formal categories that constitute a taxonomy are but convenient fictions of dubious reality, some groups are better than no groups at all. Despite any existential disenchantment, personological taxonomists may take a lesson from mystics, both as a point of contrast and as a point of departure. This particular metaphor helps us realize and assert our goals: In short, we want what mystics have (or say they have). We want a clear vision; we want freedom from confusion. Our greatest dream is one of almost mystical insight, wherein our representational blinders are removed and the inner essences of reality are revealed—for the purposes of this chapter, the substantive structural and functional variables that constitute personology and its nexus with psychopathology. Our scientific sensibilities, however, inform us that the actual mystical experience may not be all that we wish for. It almost invariably resists representation, perhaps actively so, proving ultimately too numinous and ineffable to articulate. Science, of course, cannot afford the luxury of being numinous and ineffable. Science depends on self-Â�conscious knowledge. What is numinous to mystics is vague to scientists. Whereas mystics comprehend nature in its totality as a radically open system of seamless unity, scientists must create artificially closed relational systems. We are wedded to representational systems, including taxonomies—so wedded, in fact, that the abandonment of all representational schemas would be an abandonment of knowledge itself. In the best of all possible worlds, of course, we would have both the experience of true seeing and a representational system with which to articulate it. Such is the holy grail of a taxonomy of psychopathology or personology, and only such a taxonomy “should be viewed as having objective existence in nature” (Hempel, 1965)—that is, as carving nature at its joints or affording a sense of communion that goes beyond intervening variables and construct systems. Such a sense of communion is the only true validity, that which comes from an intuition of nature as it is. A scientist born thinking within such a taxonomy might never become conscious of the representational aspect; the taxonomy would be completely transparent. Whether such a taxonomy exists, or

Considerations in Psychopathology and Personology

whether it must remain an ideal which all actual taxonomies will fall short of in various degree, only just such a taxonomy will prove ultimately scientifically satisfying for psychopathology and personology, and ultimately satisfying to researchers and clinicians who must work inside it. The metatheoretical logic and scientific study of classification are of relatively recent origin and have been given their most important impetus in several major works, notably those by Simpson (1961) and by Sokal and Sneath (1963). Both these volumes were oriented to the application of quantitative methods in biological taxonomies by setting forth explicit principles and procedures to achieve scientific goals, such as interjudge reliability and external validity. Of no less importance was a seminal article by Hempel (1961) in which he specifically addressed psychopathologists; with unerring logic, it served not only to raise conceptual considerations involved in developing productive taxonomies, but to alert clinicians to the key role that theoretical clarity and empirical synthesis must play. Before these splendid and influential contributions, psychopathological classification reflected belief systems that were based on impressionistic clinical similarities; most were not grounded in quantifiable data, used unrepresentative populations, and were devoid of a cohering theory. Psychopathology is an outgrowth of both psychology and medicine. As such, efforts to construct a taxonomy must contend with the goals, concepts, and complications inherent in both disciplines (e.g., context moderators, definitional ambiguities, overlapping symptomatologies, criterion unreliabilities, multidimensional attributes, population heterogeneities, instrument deficits, and ethical constraints). As already noted, the current state of psychopathological nosology and diagnosis resembles that of medicine a century ago. Concepts remain overwhelmingly descriptive. To illustrate, the third edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-III; American Psychiatric Association, 1980) was not only formulated to be atheoretical, but addressed itself exclusively to observable phenomena. It is not that inferences and theory have failed in the past to provide useful knowledge, but that seg-


ments of the mental health profession have not been convinced of their scientific utility, at least not sufficiently to use this knowledge for a nosology. On the other hand, despite the fact that much observation and experimentation have been done, the significance of their products lacks an adequate consensus. Thus we remain unsure today whether to conceive depression as a taxon (category) or an attribute (symptom), whether to view it as a dimension or as a set of discrete types, or whether to conceive it as a neuroendocrinological disease or as an existential problem of life. Although debates on these issues often degenerate into semantic arguments and theoretic hairsplitting, it is naive to assume that metaphysical verbiage and philosophical quibbling about words are all that these debates involve. Nevertheless, the language we use, and the assumptions it reflects, are very much a part of our scientific disagreements. In this chapter, I hope to illustrate at least one point—Â�namely, that philosophical issues and scientific modes of analysis must be considered in formulating a psychopathological taxonomy. Arguably, the following considerations will not in themselves reveal clear resolutions to all nosological quandaries. Their more likely role will be to unsettle prevailing habits and thereby to force us to progress, if for no other reason than because our cherished beliefs and assumptions have been challenged. Logic necessitates that psychopathological states and processes be distinguished from one another and be thereby categorizable to some degree before they can be subjected to identification and quantification.

Conceptual Issues What exactly is a clinical attribute? That is, what constitutes a taxon, and on what grounds shall we decide that it exists? To pose this question is not mere sophistry. For example, a serious issue demanding the attention of responsible philosophers, psychiatrists, and psychologists (Medin, Altom, Edelson, & Freko, 1982; Smith & Medin, 1981) is the very nature of the construct “disease entity.” Noting the complexities inherent in conceiving so universal a concept as “disease,” Kendell (1975) wrote that

154 contemporary writers still make frequent reference to “disease entities”—almost invariably without defining their meaning. Like “disease” itself, “entity” has become one of those dangerous terms which is in general use without ever being defined, those who use it fondly assuming that they and everyone else knows its meaning. (p.€65)

Feinstein (1977) noted the circular nature of many definitions in the following illustration: The complex situation we have been contemplating can be greatly simplified if only we are allowed a bit of circular reasoning. At the root of the difficulty is the problem of deciding what is a diagnosis. We can dispose of that difficulty by defining a diagnosis as the name for a disease. We are now left with defining disease, which we can call a state of abnormal health. Abnormal health is readily defined as a departure from normal health. And then, completing the circle, normal health can be defined as the absence of disease. (p.€189)

As Feinstein went on to say, “disease” is an abstract and multifaceted concept that remains so even if one prefers to represent it with another term, such as “ailment,” “sickness,” “illness,” or “disorder.” Those who have more than a tangential interest in the nature of psychopathological classification cannot help being apprehensive over the profound problems that arise in merely defining its constituents, no less in their specification, criteria, and measurement. I now survey the extent to which constructs such as “taxa” and “attributes” have tangible empirical referents. The basic issue is whether these concepts are anchored to observable or to inferred phenomena. The constituents of a nosological classification are represented by a set of terms or labels—that is, a language by which members of a clinical group communicate about a subject. These terms serve two functions. First, they facilitate the manipulation of ideas. Clinical or theoretical concepts are systematically (if implicitly) linked; it is through their interplay that meaningful clinical ideas are formulated and deductive scientific statements are proposed. Second, most concepts possess an empirical significance; that is, they are linked in some way to the observable world. Although some may


represent processes or events that are not apparent, they can be defined by or anchored with reference to the explicit and tangible. It is this translatability into the empirical domain that allows nosologists to test their schemas in the clinical world. Ideally, all of the concepts that constitute a nosology ought to be empirically anchored (i.e., to correspond to observable properties in clinical practice); this minimizes confusion about the attributes of which a taxon is composed (Schwartz, 1991). Furthermore, nosological labels must be more precise and clinically descriptive than the words of ordinary language. Although everyday language has relevance to significant real-world events, it gives rise to ambiguity and confusion because of the varied uses to which conventional words are often put. Taxonic concepts must be defined as precisely and with as much clinical relevance as possible, in order to assure that their meaning is clear and pertinent. Empirical precision can be achieved only if every defining feature that distinguishes a taxon is anchored to a single and observable phenomenon; that is, a different datum will be used for every difference that can be observed in the clinical world. This goal is simply not feasible, nor is it desirable (for reasons to be noted shortly). Classificatory terms do differ, however, in the extent to which they achieve empirical precision. There are points along a gradation from conceptual specificity to conceptual openness that may be identified; doing so may aid readers to recognize certain features that distinguish among both attributes and taxa. Some concepts are defined literally by the procedures that measure observable events, and they possess no meaning other than the results obtained in this manner. They reflect what Bridgman (1927) termed “operational definitions.” The meaning of an operationally defined concept becomes synonymous with how we measure it, not with what we say about it. Its scientific advantage is obvious: Taxa and the attributes of which they are composed are unambiguous, and diagnostic identifications associated with these taxa are translatable directly into the clinical attributes they represent. Useful as these definitions may be, they present several problems (Schwartz & Wig-

Considerations in Psychopathology and Personology

gins, 1986; Spitzer, 1990). Diagnostic terms must be generalizable; that is, they ought to enable clinicians to include a variety of measures and observations as gauges of a taxon. Operational definitions are too restrictive. They preclude extensions to new situations that are even slightly different from the original defining condition. One of the primary goals of a classification is to integrate diverse observations with a minimum number of terms. A strict operational approach floods clinicians with an infinite number of attributes and taxa, and clutters their thinking with largely irrelevant distinctions. Intrapsychic processes and dispositional traits are nebulous and concealed, hidden from the observable world, and hence only to be inferred. These unobservable mediating structures and processes are not merely useful, but may be necessary elements in constructing an effective psychopathological nosology. Because of their abstract and hypothetical character, these indeterminate constructs are known in philosophy as “open concepts” (Pap, 1953). Some are defined by, and largely reducible to, a set of diverse empirical events. For example, the concept of “projection” may be gauged by observing persons ascribe their own traits to others, by the presence of certain scores on a psychological test, by a history of litigious actions, and so on. Although the term “projection” implies a concealed intrapsychic process within a person that cannot itself be observed, its existence may be inferred from a variety of observables. Many open and fully speculative concepts are formulated with minimal or no explicit references. Their failure to be anchored to the realm of observables has led some to question their suitability in scientific contexts. No doubt clarity gets muddled, and deductions are often tautological, when a diagnosis is explained in terms of a series of such constructs. For example, such statements as “In an individual with borderline personality disorder, the mechanisms of the ego become diffused when libidinous energies overwhelm superego introjections” are at best puzzling. Postulating connections between one set of open concepts and another may lead to facile but often confusing clinical statements, as any periodic reader of contemporary psychoanalytic literature can


attest. Such use results in formulations that are difficult to decipher because one cannot specify observables by which the formulations can be anchored or evaluated. Open concepts usually assume their meaning with reference to a theoretical network of variables and constructs within which they are embedded. The significance they derive thereby accounts for both their weaknesses and their strengths as components of taxonomy (Spitzer, 1990). Because classification is a human artifact, not every one of its terms needs to be linked to observable events, especially if its purpose is to extend the generalizability of knowledge (Dougherty, 1978). Unrealistic standards of empirical anchorage, particularly in the early stages of taxonomic construction, often discourage the kind of imaginative speculation necessary to decode and to integrate elusive phenomena. Vague and risky as open concepts may be, they prove among the most useful tools available in developing a productive classification.

Clinical Attributes What data from the stream of ongoing clinical events and processes ought to be selected to serve as the basic units of taxa? Must we restrict ourselves only to observables, or will inferred processes be permitted? Ought the data to consist of behaviors only, or will self�reports or physiological signs be admissible? What of past history or of socioeconomic or situational factors? Is everything grist for the taxonic mill? Must the attributes of diagnostically comparable syndromes be uniform (i.e., consist of the same class of data), or will biological indices be included only in some, cognitive processes only in others, and so on? What structural framework shall we use to organize the components of the taxonomy, and what rules will govern the taxa into which its defining features will be placed? Shall its overall architecture be horizontal, vertical, or circular in format, and shall its elements constitute categories or dimensions? Shall we construct or select its elements on the basis of formal statistical techniques of analysis, such as numerically derived clusters, or shall we turn to theory and choose an element on the basis


of logically deduced constructs? Such questions as these, and the issues and alternatives they raise, are central to the taxonomic enterprise—Â�considerations of a formal character that have only recently been examined in constructing psychopathological nosologies. Psychopathology has been studied from many vantage points; it has been observed and conceptualized in legitimately different ways by behaviorists, phenomenologists, psychodynamicists, and biochemists. No point of observation or conceptualization encompasses all of the complex and multidimensional features of psychopathology. Clinical processes and events have been described in terms of conditioned habits, reaction formations, cognitive expectancies, or neurochemical dysfunctions. These domains cannot be arranged in a hierarchy, with one level viewed as reducible to another (Millon, 1990; Sartorius, 1990). Neither can they be compared in terms of some objective truth value. Alternative substantive domains are merely different; they facilitate the observation and conceptualization of different clinical attributes, and lead therefore to different taxa. The point is that taxa may be differentially composed in accord with the kinds of clinical data (e.g., etiology, symptoms, and treatment response) they include as their basic constituents. Choices are often pragmatic, and questions of comparative utility cannot be determined a priori. However, irrelevant controversies and needless confusions can be avoided if the class of attributes from which taxa are composed has been specified clearly. When this has been done properly, clinicians and researchers can determine whether two taxonomies are comparable, whether the same diagnostic label refers to different clinical phenomena, whether different taxa encompass the same attributes, and so on. Panzetta (1974) stated: As one begins to consider the variety of starting-off points then we quickly appreciate that the first step in the nosologic process is inherently arbitrary. I would insist that we begin by acknowledging that the arbitrary focus is not, per se, a deficiency but rather a reality which flows naturally from the tremendous complexity of human behavior. It is useless to try to develop the “correct” initial focus. There is no correct focus, only several alternatives. (p.€155)


Sneath and Sokal (1973) recorded the wide array of attributes available in every discipline that can be potentially useful in forming a taxonomy. The task is that of identifying which are likely to be most relevant and optimally productive. The history of psychopathology provides guidelines that may prove fruitful—for example, the well-Â�established and important distinction between longitudinal and concurrent attributes. The former represent the progression of various clinical phenomena across time and circ*mstance; the latter represent the diverse ways in which these phenomena manifest themselves contemporaneously and simultaneously across a number of expressive dimensions. Longitudinal Attributes Data that concern causal factors as clinical attributes for taxa would be extremely useful, if only such knowledge were available. Unfortunately, etiological data are scanty and unreliable. Moreover, they are likely to remain so because of the obscure, complex, and interactive nature of influences that shape psychopathological phenomena. In great measure, etiological attributes are conjectures that rest on tenuous empirical grounds; most reflect schools of thought that emphasize their developers’ favorite hypotheses. These speculations are best construed as questions that deserve empirical testing on the basis of taxa formed by less elusive data. Among the few etiological schemas that lend themselves to taxonomic goals is Zubin’s (1968) differentiation of six categories: sociocultural (ecological), developmental, learning, genetic, internal environment, and neurophysiological. Systematic and orderly though models such as Zubin’s may be, they do not fully address issues associated with alternate levels of analysis; nor do they trace the intricate and varied causal chains that unfold ultimately into a clinical state. Beyond the issue of identifying which of several levels of an interactive causal system may be selected as etiological attributes, there are questions of a more philosophical and methodological nature concerning what exactly is meant by “etiology” and in what precise way it may be gauged. Meehl (1972) addressed this matter succinctly:

Considerations in Psychopathology and Personology A metatheoretical taxonomy of causal factors and a metataxonomy of causal relations (such as “necessary but not sufficient condition,” “interaction effects,” “threshold effects,” and the like) are badly needed. In medicine, we recognize several broad etiologic classes such as deficiency diseases, autoimmunity diseases, disease due to microorganisms, hereditaryÂ�degenerative diseases, developmental anomalies, diseases due to trauma. .€.€. The concept “specific etiology” .€.€. appears to have half a dozen distinguishable and equally defensible meanings (e.g., sine qua non, critical threshold, uniformly most powerful factor) that might be useful under various circ*mstances. (p.€22)

The yearning among taxonomists for a neat package of etiological attributes simply cannot be reconciled with the complex philosophical and methodological issues and the difficult-to-Â�disentangle networks of both subtle and random influences that shape our mental disorders. It also makes understandable the decision of the DSM-III Task Force to set etiological and course variables aside as clinical grist for its taxonic mills. When we turn from the antecedent to the consequent side of the clinical course, logic argues that the nature of a mental disorder must be at least partially revealed by its response to treatment. The data available on this matter, however, provide little that goes beyond broad generalizations. This contrasts with medicine at large, in which a variety of interventions are specific to particular disorders. Notable in medicine are a variety of challenge or stressor tasks that serve to identify patient vulnerabilities (e.g., introducing allergens to uncover susceptibilities or to elicit reaction sensitivities). Considerations of ethics and human sensibilities preclude adopting parallel strategies in psychopathology. Even where the intent is clearly beneficial, as in psychological treatment interventions, the problems encountered in discerning and differentiating optimal clinical attributes are many—for example, the inevitability of spontaneous remissions and unanticipated life events. Even if we could partial out the effects of these confounding events, could we identify the ingredients that account for beneficial reactions in a heterogeneous patient population? Furthermore, what conclusions can be drawn about a therapy whose efficacy is wide-Â�ranging (i.e., demonstrable among disorders whose origins and expres-


sion are highly diverse)? Complicating matters further is the fact that every technique of therapy claims high efficacy across a wide band of sundry diagnostic classes. We may then conclude that treatment response, as with etiology, may prove ultimately to be a useful taxonic attribute; for the present, however, reliable data are not in hand. Concurrent Attributes Given the problematic nature of longitudinal attributes, what remain to constitute psychopathological taxa are coexistent attributes of a contemporaneous nature—Â�notably, objective signs on the one hand, and subjectively reported symptoms on the other. To these two classical indicators of disorder may be added the essentially inferred attributes of personality traits. Information about these clinical features may be derived from four conceptually and methodologically distinct data sources: namely, the biophysical, intrapsychic, phenomenological, and behavioral.

Signs “Signs” consist of more or less objectively recorded changes in state or function that indicate both the presence and character of clinically relevant processes or events. Two sources provide the main body of clinical signs: biophysical markers and behavioral acts. As for the first data set, biophysical markers, there are few anatomical, biochemical, and neurophysiological gauges among the standard diagnostic taxa, despite their tangible, objective, and quantitative nature. This failure is not only surprising but disconcerting, given the vast number of psychopathological studies over the years that have hypothesized and investigated potential biophysical markers obtained from a multitude of measures (urine or blood analysis, diverse muscle appraisals, skin gauges, cardiac assessments, eye movements, metabolic indices, electroencephalographic rhythms, etc.). That so few of these biophysical measures have aided the definition of clinical attributes can be traced to a number of factors. Most cannot be arranged into clearly discriminable categories, and with few exceptions, normative distributions on relevant clinical populations are either unavailable or incon-


sistent. For many, reliability measures are lacking or indicate high levels of variability over time and across settings. To complicate interpretive efforts, low intercorrelations are typically found among measures that ostensibly represent the same basic functions. Moreover, the expense of technical equipment is high, and the availability of needed expertise is often so scarce as to place many of these procedures out of reach for all but well-Â�funded research investigators. As for explicit behavioral acts, the second of the major objective indices, it has been the methodological goal of behavioral purists to avoid drawing inferences about internal or subjective processes. Hence they seek to use techniques that ostensibly bypass explicit dependence on subjective symptomatic data. As promising as overt behavior may be as a source for clinical attributes, numerous concerns about its utility have been registered, not the least of which is its unpredictability—that is, its high variability across setting and time (e.g., the recording of verbal behaviors fails to generate reliable normative data across diverse circ*mstances). This is in contrast with data produced by well-Â�constructed rating instruments. Most of these are reasonably reliable, succeed in discriminating among relevant patient groups, and possess adequate normative data for differential or comparative purposes. Rather commonplace behaviors of potential diagnostic significance have come under systematic scrutiny and analysis. Buss and Craik (1983, 1987) and Livesley (1985, 1991) have attempted to provide a descriptive (i.e., nonexplanatory) basis for diagnostically relevant features; these investigators have developed lists of familiar acts observed in the course of everyday life that may typify certain clinical characteristics. Despite these promising advances, behavioral methods are long on neat measures of a rather trivial character, with tangential or limited substantive significance in the realm of psychopathological taxonomy (Block, 1989).

Symptoms In contrast to clinical signs, “symptoms” are subjective in nature, represented by reports from patients of their conscious recollections and recorded experiences (e.g., moods, feelings, perceptions, memories, attitudes). To-


gether with clinical signs, symptoms focus on phenomenological processes and events that relate directly to diagnostic matters. It is here where the psychopathologist has an advantage over the physicist or the biologist, for neither can ask the objects of their study to reflect on their experience, no less to communicate it in articulate or meaningful ways. Strauss (1986) framed it thus: To some extent, the field has been discouraged by previous claims and promises regarding the understanding of subjective experiences such as these. Although the claims were often overstated, and the assessment of these processes is complex, neither reason is adequate for avoiding a major attempt to develop creative ways for looking at the role of such subjective experiences in the course of psychopathology and their relevance to diagnosis. (p.€262)

Contributing to the utility of symptomatic data is their ease of evocation. Moreover, structured assessments, such as interview schedules and self-�report inventories, minimize potential sources of distortion. Nevertheless, these methods are subject to difficulties that can invalidate data (e.g., one cannot assume that subjects will interpret questions in the same way, that they possess sufficient self-�knowledge to respond informatively, or that they may not be faking or dissembling). Although the substantive contents of phenomenological symptoms are elusive and often unreliable and are fraught with philosophical and methodological complexities, taxonomists cannot afford the luxury of bypassing them. Symptoms lie at the very heart of all psychopathological inquiries. The events they portray are real and represent facets of experience far richer in scope and diversity than concrete observables.

Traits In contrast to signs, which represent objective biological measures or behavioral acts, and symptoms, which are phenomenologically reported recollections and experiences, “traits” include inferred psychological habits and stable dispositions of broad generality and diverse expression. This long-Â�established psychological construct has been used in two ways. First, it encompasses various characteristic habits, moods, and attitudes; second,

Considerations in Psychopathology and Personology

through inference it identifies dispositions to act, feel, and think in certain ways. Traits can be considered to be both more and less than signs and symptoms. For example, several traits can coalesce to form the expression of a single behavioral sign. Conversely, several different specific symptoms may be the upshot of a single trait. Furthermore, each trait may express itself in diverse signs and symptoms. Clearly, there is no one-toone correspondence between traits and signs or symptoms. Traits are often inferred rather than observed, generalized rather than specific, and dispositional rather than consequential. They are assumed to be enduring and pervasive. However, only certain traits of a person display this durability and pervasiveness; that is, only some of them prove to be resistant to the influences of changing times and circ*mstances. Other forms of behavior, attitude, and emotion are presumably more transient and malleable. It is noteworthy that the traits exhibiting consistency and stability in one person may not be the same as those in others. These qualities are most prominent among characteristics that are central to maintaining a person’s overall psychological balance and style of functioning. To illustrate, the interpersonal conduct trait of significance for some is that of being agreeable, never differing or having conflict; for others, it may be interpersonally important to maintain one’s distance from others so as to avoid rejection or humiliation; for a third group, the influential interpersonal trait may be that of asserting one’s will and dominating others. The sources used to identify clinical traits are highly diverse. They range from methods designed to uncover intrapsychic processes, such as free association, dream analysis, hypnosis, and projective techniques, to such phenomenological methods as structured interviews and self-Â�report inventories, and to behavioral methods of observation and rating (be they systematic or otherwise). It is no understatement to say that the rich vein of clinical attributes uncovered by dispositional and intrapsychic traits has been a boon to clinical theory, but a source of perplexity and despair to taxonomists. More than any other domain, dispositional data and methods produce information fraught with complexities and obscurities that can


bewilder the most sophisticated of classifiers. Part of the difficulty stems from the fact that the identification of hidden traits is highly inferential. Because the dispositional structure and processes that make up traits can be only partially observed and take different manifest forms in different contexts, it is difficult to identify them reliably, and hence to assign them a standard place in a taxonomy. Matters are made more difficult by the absence of intrapsychic normative and base rate data, as well as by the lack of strong validational support.

Structural Models Whatever features are chosen to provide the substantive body of a taxonomy, decisions must be made about the structural framework into which the taxonomy will be cast, the rules that will govern the taxa into which its clinical attributes and defining features will be placed, and the compositional properties that will characterize these attributes and features. These are problems of the essential architecture of the taxonomy: whether it should be organized horizontally, vertically, or circularly; whether all or only a limited and fixed subset of features should be required for taxonic membership; whether its constituents should be conceived as categories or dimensions; and which of a host of other differentiating characteristics one should choose. I discuss several modern structural designs, and the options available among them—a task of no simple proportions, given that nothing is logically self-Â�evident and that there is no traditional format or contemporary consensus to guide selections among these alternatives. Taxonomic Structure Ought the various attributes that make up the substantive data of psychopathology to be listed more or less randomly, or should they be ordered into a series of logical or functional groups that attempt to mirror the inherent nature of psychopathology? The obvious answer is the latter. Several frameworks for structuring psychopathology have been formulated in recent years, and they are not mutually exclusive. From a design viewpoint, they can


be described as having vertical, horizontal, or circular structures. The vertical or “hierarchical” framework organizes the various taxa of psychopathology (e.g., depressive disorders or schizophrenic disorders) in a series of echelons in which lower tiers are subsumed as subsets of those assigned higher ranks. The second or “horizontal” framework is known as the “multiaxial” schema; it orders different classes of attributes (e.g., symptoms or etiologies) in a series of aligned or parallel categories. Since the publication of DSM-III, the DSMs have encompassed both hierarchical and multiaxial structural forms, albeit with modest logic and success. The circular framework is referred to as the “circumplical” model. It has not received official recognition, although it has gained considerable currency among theorists who emphasize the role of interpersonal attributes.

Hierarchical Models Hierarchical models are typically arranged in the form of taxonic decision trees. Once a particular branch (i.e., a higher-order diagnosis) has been chosen, subsequent taxonic choices are limited to the several branches and twigs that constitute subdivisions. To illustrate, once it has been decided with DSM-IV-TR (American Psychiatric Association, 2000) criteria that a patient is exhibiting a mood disorder, the clinician may further differentiate the disturbance as either a bipolar disorder or a depressive disorder. If the choice is bipolar, the clinician may move further down the hierarchy to select among bipolar I disorder (manic, hypomanic, mixed, or depressed episode), bipolar II disorder (hypomanic or depressed episode), cyclothymic disorder, or bipolar disorder not otherwise specified. A consequence of so carefully fashioned a sequential chain of categories is that successive taxa in the classification are invariably more specific and convey more precisely differentiated information than those that precede them. This increasing distinctness and exactitude—Â�necessary ingredients in a successful hierarchical schema—Â�assures that each successive category possesses authentic clinical features not found in categories previously listed.


Sequential patterns of the decision tree type would be a remarkable achievement for any hierarchical nosology, if they were naturally or logically justified (Millon, 1983). Not only is there no inherent structure to psychopathology that permits so rigorous an arrangement, but the various DSMs, for instance, impose only a modest degree of sequential rigor on the taxonomic organization. The problems of pursuing a hierarchical method for differential diagnosis are compounded by the fact that modern DSMs not only permit but encourage multiple diagnoses—a problem aggravated further by the manuals’ standard multiaxial framework. Not only does the hierarchical goal of orderly and successive diagnostic choice points run hard against the structural character of DSM taxa, but its formalism and sequential requirements are undermined repeatedly by the multidiagnostic aims and intrinsic multiaxial schema of the DSMs.

Multiaxial Models The multiaxial format, the second of the overarching structural models, encounters few of the logical difficulties and assumptions found in hierarchical systems. The formal adoption of the multiaxial schema in DSM-III and ICD-10 (World Health Organization, 1992) approached a paradigm shift (Millon, 1983). It reflected a distinct turn from the traditional infectious-Â�disease model, in which the clinician’s job is to disentangle distracting symptoms and to clear away confounding situational problems so as to pinpoint the underlying or true pathophysiological state. By contrast, the multiaxial model (Essen-Â�Moller & Wohlfahrt, 1947; Mezzich, 1979; Williams, 1985a, 1985b) not only recognizes that distracting and confounding circ*mstances are aspects worthy of attention, but encourages recording them on their own representative axes as part of an interactive complex. The multiaxial structure aligns many of the potentially relevant factors that can illuminate the nature of a clinical condition, and it provides a means of registering their distinguishing attributes. In contrast to the more traditional hierarchical model, in which a single class of attributes (signs or etiologies) is differentiated, the multiaxial format per-

Considerations in Psychopathology and Personology

mits multiple classes of data (again, signs and etiologies) and thereby encourages diagnostic formulations that include several facets of information relevant to clinical decision making. The very comprehensiveness of the multiaxial model can prove to be its undoing, however. Such systems provide a more thorough picture than do schemas of unitary axes— but they are therefore also more complicated and demanding to implement, and require a wider band of data and a greater clarity and diversity of judgments than clinicians are accustomed to performing. They often impose a procedural complexity on an otherwise expedient process. From a pragmatic view, a fully comprehensive multiaxial assessment may be an unnecessary encumbrance in routine diagnostic work, impractical for everyday decision making, and abhorrent to clinicians accustomed to the diagnostic habit of intuitive synthesis.

Circumplical Models Circumplical models have been used in the arrangement of both taxa and attributes. In neither case have they been recognized in formal psychopathological taxonomies; rather, their primary use has been as a structural tool for ordering interpersonal traits (Benjamin, 1974, 1986, 2006; Lorr, 1966), most notably in conjunction with personality processes and disorders (Kiesler, 1983, 1996; Leary, 1957; Pincus & Wiggins, 1990; Plutchik & Platman, 1977; Sim & Romney, 1990; Strack, 2005; Strack, Lorr, & Campbell, 1990). Circumplical models are structured so as to locate similar taxa in adjoining or nearby segments of a circle; taxa located diametrically on the circle are considered psychologically antithetical. Plutchik and Conte (1985) provided evidence that emotions, traits of personality, and personality disorders line up in parallel ways on a circumplex, which suggests that this structure can arrange diverse concepts into a common framework and lead thereby to the identification of relations that may otherwise not be recognized. As interesting as this formulation may be for organizing conceptual categories, the circumplex appears at present to be essentially an academic tool of theoretical rather than


clinical value, despite indications that promise the latter as well. Taxonic Structure Taxonic units in psychopathology may be monothetic or polythetic in structure. All of the attributes that constitute a monothetic taxon must be in evidence for a diagnosis to be correctly made. In polythetic taxa, various and different optional subsets of the full attribute list can suffice to justify a diagnosis.

Classical versus Prototypal Taxa Classical taxa comprise categories made up of discrete entities that are hom*ogeneous with respect to their defining features—that is, arranged in a restrictive, monothetic format (Cantor & Mischel, 1979; Rosch, 1978). Failures to identify all of the attributes of a taxon can result from obscuring and confounding conditions and/or from deficits in observational technology and skill. Frances and Widiger (1986) characterized the major features of and difficulties with classical taxa as follows: The classical model of categorization conceives of disorders as qualitative, discrete entities and assumes that the defining features are singly necessary and jointly sufficient, that the boundaries between categories are distinct, and that members are hom*ogeneous with respect to the defining features. .€.€. Although the classical model works well for abstract categories (e.g., “square”), it fails to do justice to the complexity of naturally occurring taxonomic problems. All squares share the features of having four equal sides joined at right angles .€.€. [but] actual objects, plants, animals, and persons, however, often fail to share a set of singly necessary and jointly sufficient features. .€.€. If a classical typology is an inappropriate model for classification of objects, birds, and plants, it is clearly inappropriate for psychiatric diagnosis. (p.€392)

Horowitz, Post, French, Wallis, and Siegelman (1981) described the contrasting prototypal construct succinctly: A prototype consists of the most common features or properties of members of a category and thus describes a theoretical ideal or stan-

162 dard against which real people can be evaluated. All of the prototype’s properties are assumed to characterize at least some members of the category, but no one property is necessary or sufficient for membership in the category. Therefore, it is possible that no actual person would match the theoretical prototype perfectly. Instead different people would approximate it to different degrees. The more closely a person approximates the ideal, the more the person typifies the concept. (p.€575)

The prototypal structure assumes a measure of taxonic heterogeneity, and hence is likely to require a polythetic format. Its open and permissive taxonic structure is more consonant with the natural fuzziness of conceptual boundaries (Cantor & Genero, 1986; Osherson & Smith, 1981; Schwartz, Wiggins, & Norko, 1989; Wittgenstein, 1953), as well as the inherent inexactness of natural reality (Meehl, 1978; Millon, 1969, 1987, 1990). Referring to the classical approach, Widiger and Frances (1985a, 1985b) have written that once patients are placed in the same taxon, there is a tendency to exaggerate their similarities, ignore their differences, and focus on the stereotypic features that distinguish the category, all at the expense of bypassing disconfirming traits and downplaying idiosyncratic behaviors. By contrast, they have stated that polythetically constructed taxa limit stereotyping, permit diagnostic flexibility, and encourage within-�groups variability. On the other hand, no pathognomonic signs will be present; persons similarly diagnosed will vary in their degree of prototypicality; and defining features will differ in their diagnostic efficiency.

Categorical versus Dimensional Taxa My discussion of structural alternatives has progressed from choices among three taxonomic configurations (hierarchical vs. multiaxial vs. circumplical) to a selection between two taxonic structures (classical vs. prototypal). I continue this progression toward increasing specifics, addressing the choices that need to be made with concern for the compositional character of taxa and attributes. Here the issue is raised as to whether taxa and their attributes ought to be conceived qualitatively (categorically) or quantitatively (dimensionally). Thoughtful papers


on the issue will be found in Chapters 15, 17, and 18 later in this book. For monothetic (classical) taxa, authors have asked whether clinical syndromes (e.g., dysthymia) ought to be conceived as qualitatively discrete categories, or whether they ought to be conceived on a quantitative dimension of severity (Frances, 1982). The issue among those who arrange polythetic rules (usually prototypal) is whether different combinations of defining features can be conceived as forming quantitative variations of the same qualitative category. This latter approach focuses not on the taxa themselves, but on their clinical attributes and defining features. For this issue, each attribute is conceived as a quantifiable dimension along such lines as salience or severity. What is categorized is not the taxon, but the several variants of a clinical attribute (e.g., if interpersonal conduct is a relevant attribute for diagnosing personality disorder taxa, then choices may be made first among the several interpersonal defining feature options, such as aversive, seductive, or secretive). Second, once the interpersonal options have been chosen (qualitative categorization), each may be given a score to represent its degree of salience or severity (quantitative dimensionalization). The first of the two approaches to the issue is related to categorical versus dimensional taxa. The issue applies especially to clinical attributes, for which the trend toward prototypal models and polythetic definitions is more central. The taxonic issue may be stated in the form of a question: Ought taxa to be conceived and organized as a series of dimensions that combine to form distinctive profiles for each person, or should certain characteristics found commonly in clinical populations be selected to exemplify and classify taxa (Livesley, 1991)? Dimensional conceptions emphasize quantitative gradations among persons, rather than qualitative, discrete, all-or-none class distinctions. To illustrate, Kendell (1968) proposed that a single dimension might suffice to represent the continuum he found between neurotic and psychotic depressions. By contrast, and consistent with the clear boundaries between taxa expected with a categorical schema, Paykel (1971) found minimal overlapping among four classes—

Considerations in Psychopathology and Personology

those of psychotic, anxious, hostile, and depression in young persons. Several advantages to dimensional models may be noted. First, they combine several clinical attributes (or their defining features) in a single configuration. This comprehensiveness results in a limited loss of information, and no single attribute is given special significance, as is the case when only one distinctive characteristic is brought to the forefront. Dimensional profiles also facilitate the assignment of unusual or atypical cases. In categorical formats odd or mixed conditions are often excluded because they fail to fit the prescribed criteria. Given the idiosyncratic character of many clinical conditions, a dimensional system permits representation and assignment of interesting and unique cases without forcing them into Procrustean categories for which they are ill suited. A major advantage of the dimensional model is that the strength of its constituent features is gauged quantitatively, wherein each characteristic extends into the normal range. As a consequence, normality and abnormality are construed as points on a continuum rather than as distinct and separable phenomena. Despite these advantages, dimensional taxa have not fared well in psychiatric classifications (Gunderson, Links, & Reich, 1991). Numerous complications have been noted in the literature; for example, there is little agreement among theorists about the number of dimensions necessary to represent psychopathological phenomena. Menninger (1963) contended that a single dimension suffices; Eysenck (1960) asserted that three are needed; Cattell (1965) claimed to have identified as many as 33 and believed there to be many more. In fact, theorists appear to invent dimensions in accord with their expectations, rather than to discover them as if they were intrinsic to nature, merely awaiting scientific detection. The dimensions required to assess psychopathological phenomena appear to be determined not by the ability of research to disclose some inherent truth, but rather by researchers’ predilections for conceiving their investigations and interpreting the findings. Categorical models are the traditional form used to represent clinical conditions. There are several reasons for this preference. First, most taxa neither imply nor are


constructed to be all-or-none categories. Certain features are given prominence, but others are not overlooked; rather, they are merely assigned lesser significance. The success of categorical taxa may be traced to the ease with which clinicians can use them in making rapid diagnoses with numerous briefly seen patients. Although clinical attention is drawn to only the most salient patient attributes, other, less conspicuous characteristics are often observed, suggested, or inferred. The quality of intimating characteristics beyond the immediately observed contributes to the value of established categorical taxa. The categorical approach’s power of extending its scope to associated attributes contrasts with the tendency of dimensional schemas to segment, if not fractionate, persons and disorders into separate components. Categories restore the unity of a patient’s pathology by integrating seemingly diverse elements into a single, coordinated configuration. Well-Â�established categorical taxa often provide a standard of reference for clinicians who are otherwise faced with reconstructions or de novo diagnostic creations (Gunderson et al., 1991). There are objections to the use of categorical taxa. For example, they contribute to the fallacious belief that psychopathological processes constitute discrete entities, even medical diseases, when in fact they are merely concepts that help focus and coordinate our observations. Moreover, categories often fail to identify or include significant aspects of behavior because of the decision to narrow their list to a set of predetermined characteristics. This discarding of information is not limited to categories; dimensional schemas also select certain attributes to the exclusion of others. The problem, however, is that certain categorical schemas give primacy only to one or two attributes. A related criticism is that both the number and diversity of categories in most taxonomies are far less than the clinically significant individual differences observed in everyday practice. Not only are there problems in assigning many patients to the limited categories available, but clinicians often claim that the better they know patients, the greater the difficulty they have in fitting them into a category. Issues of categoricality versus dimensionality are more properly the province of


attributes than of taxa. For example, the Axis I taxon of depressive disorders really represents a clinical attribute; the distinction between its two major subcategories, major depression and dysthymia, may essentially be a matter of quantitative severity and hence may reflect dimensionality. Furthermore, the distinction between bipolar disorders and depressive disorders may be best conceived as variations in two clinical attributes, not taxa. The former reflects the operation of two mood attributes, mania and depression—each of which may vary as a single dimension, although both may be found in certain persons (bipolar disorders), whereas only one may be exhibited in others (depressive disorders). The fact that some taxa are composed essentially of a single clinical attribute, whereas others encompass several distinct attributes, has not only confounded discussions of categoricality versus dimensionality; it has contributed a share of confusion to theory, research, and practice as well. To restore order to some aspects of the problem, Skinner (1986) elaborated several hybrid models that integrate elements of these ostensibly divergent schemas. In what Skinner termed the “class–Â�quantitative approach,” efforts are made to synthesize quantitative dimensions and discrete categories. I described an endeavor of this nature in a couple of articles (Millon, 1984, 1986). I proposed that an essential aspect of integrating a mixed categorical–Â�dimensional model for personological taxa will be the specification of a distinctive defining feature for each clinical attribute of each personality disorder. If the clinical attribute of expressive mood is deemed of diagnostic value in assessing personality disorders, then a specific defining feature will be identified to represent the distinctive manner in which each personality disorder manifests its emotional feelings. To further enrich the qualitative categories (the several defining features that compose the clinical range of each attribute) with quantitative discriminations (numerical intensity ratings), clinicians will not only identify which features (e.g., distraught, hostile, labile) of a clinical attribute (e.g., expressive mood) best characterize a patient, but also record a number (e.g., 1–10) to represent the degree of prominence or pervasiveness of the chosen defining features. Clinicians will be


encouraged in such a prototypal schema to record and quantify more than one defining feature per clinical attribute (e.g., if suitable, to note both distraught and labile moods, if their observations and inferences so incline them). Such a procedure as this illustrates that categorical (qualitative distinction) and dimensional (quantitative distinction) taxonic models need not be framed in opposition to, or be considered as exclusive of, each other.

Construction Methods How have psychiatric taxonomies come into being? In the main, traditional classifications have been the product of a slowly evolving accretion of clinical experience (Menninger, 1963), which has been fostered and formalized periodically by the systematizing efforts of respected clinician-Â�scholars such as Kraepelin (1899). It may be expected that empirical data or theoretical advances on matters of causality or structure will serve as a primary heuristic impetus, but such has not been the case. With but a few exceptions (e.g., the DSM-I and DSM-II [American Psychiatric Association, 1952, 1968] psychoanalytic explication of neurotic disorders, which was subsequently expunged as an organizing construct in DSM-III), theoryÂ�generated or research-Â�grounded taxonomies have fared rather poorly. Another spur to developing classifications has originated in a series of quantitative methods known as “cluster analyses” (Sneath & Sokal, 1973). Time will tell whether these mathematical tools will generate taxa of sufficient consequence to gain acceptance in the clinical world. Describing taxonomic advances in the biological sciences, Sokal (1974) wrote: In classification, theory has frequently followed methodology and has been an attempt to formalize and justify the classificatory activity of workers in various sciences. In other instances, classificatory systems have been set up on a priori logical or philosophical grounds and the methodology tailored subsequently to fit the principles. Both approaches have their advantages and drawbacks; modern work tends to reflect an interactive phase in which first one and then the other approach is used, but in which neither principles nor methodology necessarily dominate. (p.€115)

Considerations in Psychopathology and Personology

Taxonomic methods in psychopathology are much less advanced than those in the biological sciences, but they are approaching the threshold at which some of the same controversies that occurred in their biological forerunners are likely to arise (Krueger & Tackett, 2006; Strack, 2006). Although each of the alternative construction methods to be discussed shortly may prove fruitful, psychiatric taxonomists are already engaged in debates as to which is best. It is important to recognize that there is no correct choice, and that no rules can be found in nature to confirm which are best or likely to be profitable. To make sense of and give order to the taxonomies they use, clinicians must know what approach to construction was followed and what attributes constituted their database. With the techniques and building blocks of these methods clearly in mind, clinicians may assess them intelligently and judge their relevance to the questions they pose. A few words must be said about the similarities between methods of taxonomic formation and procedures for developing psychometric tools. Skinner (1981, 1986) has drawn on the logic outlined by Loevinger (1957) for sequentially validating diagnostic tests and applied it creatively to the composition of taxonomies. The mutually reinforcing strength achieved by a combination of Loevinger’s three validation strategies may be kept in mind as I elaborate each of the three construction options. In Loevinger’s seminal article, she recommended progress from the theoretical to the statistical (internal) to the clinical (external)—a sequence especially suitable to the validation of diagnostic tests, and one that both Jackson (1971) and I (Millon, 1977) followed in fashioning our psychometric inventories. I reverse this sequence to accord with the historical order in which taxonomies have and are likely to continue to be composed. Clinically Derived Categories Until recently, psychiatric taxonomies were formed solely on the basis of clinical observation—the witnessing of repetitive patterns of behavior and emotion among a small number of carefully studied mental patients. Hypotheses were generated to give meaning to these patterns of covariance (e.g., Hip-


pocrates anchored differences in observed temperament to his humoral theory, and Kraepelin distinguished two major categories of severe pathology, dementia praecox and manic–Â�depressive disease, in terms of their ostensibly divergent prognostic course). The elements of these theoretic notions were post hoc, however—Â�imposed after the fact, rather than serving as a generative source for taxonomic categories. The most recent examples of a clinical taxonomy, tied explicitly to phenomenal observation and constructed to be both atheoretical and nonquantitative, are of course the DSMs. Spitzer, chairperson of the DSM-III Task Force, stated in DSMIII (American Psychiatric Association, 1980) that “clinicians can agree on the identification of mental disorders on the basis of their clinical manifestations without agreeing on how the disturbances came about” (p.€7). Despite assertions to the contrary, the recent DSMs are products of implicit causal or etiological speculation. Nevertheless, the DSM-III Task Force sought to eschew theoretical or pathogenic notions, adhering to as strict an empiricist philosophy as possible. Only those attributes that could be readily observed or consensually validated were to be permitted as diagnostic criteria. Numerous derelictions from this epistemology were notable, however, especially among the personality disorders, whose trait ascriptions called for inferences beyond direct sensory inspection. By no means do all who draw their philosophical inspiration from an empiricist mindset restrict themselves to the mere specification of surface similarities (Medin et al., 1982). Not only those who formulate theoretically generated nosologies succumb to the explanatory power and heuristic value of pathogenic, dynamic, and structural inferences. Feinstein (1977), a distinguished internist, provided an intriguing illustration of how one clinician’s factual observations may be another’s inferences: In choosing an anchor or focus for taxonomy, we can engage in two distinctly different types of nosologic reasoning. The first is to form names, designations, or denominations for the observed evidence and to confine ourselves exclusively to what has actually been observed. The second is to draw inferences from the observed evidence, arriving at inferential

166 titles representing entities that have not actually been observed. For example, if a patient says, “I have substantial chest pain, provoked by exertion, and relieved by rest,” I, as an internist, perform a denomination if I designate this observed entity as angina pectoris. If I call it coronary artery disease, however, I perform an inference, since I have not actually observed coronary artery disease. If a radiologist looking at a coronary arteriogram or a pathologist cutting open the coronary vasculature uses the diagnosis coronary artery disease, the decision is a denomination. If the radiologist or pathologist decides that the coronary disease was caused by cigarette smoking or by a high fat diet, the etiologic diagnosis is an inference unless simultaneous evidence exists that the patient did indeed smoke or use a high fat diet. (p.€192)

In great part, clinically based taxa gain their importance and prominence by virtue of consensus and authority. Cumulative experience and tradition are crystallized and subsequently confirmed by official bodies. Specified criteria are denoted and articulated, and they acquire definitional if not stipulative powers, at least among those who come to accept the attributes selected as infallible taxonic indicators. Numerically Derived Clusters Clinically based categories stem from the observations and inferences of diagnosticians; as such, they constitute, in circular fashion, the very qualities that clinicians are likely to see and deduce. Categories constructed in this manner will not only direct future clinicians to mirror these same taxa in their patients, but may lead future nosologists away from potentially more useful schemes with which to fathom less obvious patterns of attribute covariation. Toward the end of penetrating beneath the sensory domain to more latent commonalities, taxonomists have been led to turn either to numerical methods or to theoretical principles. Andreasen and Grove (1982) enumerated the advantages of what they termed “empirical” or “numerical” methods for computing patient similarities: First, the empirical method gives an opportunity for the observed characteristics of the subjects to determine the classification and perhaps to lead to a classification that the cli-

CONCEPTUAL ISSUES IN CLASSIFICATION nician was unable to perceive using clinical judgment alone. Second, the empirical method allows a great deal of information on the subjects to enter into the genesis of the classifications; human beings can keep in mind only a relatively small number of details concerning a case at any given time, but the empirical approach can process very large sets of measurements. Third, empirical or numerical approaches can combine cases in more subtle ways than can clinicians; combinations of features too complex to grasp intuitively may yield better classifications than simple combinations. (p.€45)

There has been a rapid proliferation of powerful mathematical techniques for both analyzing and synthesizing vast bodies of clinical data. This expansion has been accelerated by the availability of inexpensive computer hardware and software programs. Unfortunately, this growth has progressed more rapidly than its fruits can be digested. As Kendell (1975) said, early in this development, “most clinicians .€ .€ . have tended to oscillate uneasily between two equally unsatisfactory postures of ignoring investigations based on these techniques, or accepting their confident conclusions at face value” (p.€106). This growing and diverse body of quantitative methods can be put to many uses, of which only a small number are relevant to the goal of taxonomic construction. Some statistical techniques relate to the validation of existent nosologies (e.g., discriminant analyses) rather than to their creation. Among those used for taxonomic development, some focus on clinical attributes as their basic units, whereas patients themselves are the point of attention for others (Grove & Tellegen, 1991). For example, factor analysis condenses initially diverse sets of clinical attributes and organizes them into potential syndromic taxa. Cluster analysis, by contrast, is most suitable for sorting patient similarities into personological taxa. Reviews of these two numerical techniques, as well as other mathematical procedures for taxonomic construction and evaluation (e.g., latent-class analysis, log-Â�linear analysis, discriminant analysis, and multivariate analysis of variance), may be examined in a number of useful publications (Blashfield, 1984; Grove & Andreasen, 1986; Hartigan, 1975; Kendell, 1975).

Considerations in Psychopathology and Personology


Although cluster algorithms have begun to mirror broad diagnostic classes, these slender advances do not answer the question of whether cluster analysis produces categories that resemble the natural structure of psychopathology any better than those of our more traditional or clinically based nosologies. Nor is there any evidence that they provide more accurate predictions of such nonstructural concerns as prognosis and treatment response. Several authors have summarized the current state of affairs, and have addressed the problems that are likely to persist in the use of numerical construction procedures. For example, Skinner and Blashfield (1982) noted:

sification together and gives it both its scientific and its clinical relevance. In Hempel’s (1965) discussion of classificatory concepts, he wrote that

Clinicians have at best given only a lukewarm reception to such classifications. They have been skeptical about the value of clustering methods to identify “naturally” occurring subgroups. Furthermore, the classifications generated by these methods have not seemed Particularly meaningful or relevant to everyday clinical practice. (p.€727)

As Hempel (1965) stated, mature sciences progress from an observationally based stage to one that is characterized by abstract concepts and theoretical systemizations. Contemporary philosophers of science believe that classification alone does not make a scientific taxonomy, and that similarity among attributes does not necessarily constitute a scientific category (Smith & Medin, 1981). The characteristic that distinguishes a latent scientific classification is its success in grouping its elements according to theoretically consonant explanatory propositions. These propositions are formed when certain attributes that have been categorized have been shown or have been hypothesized to be logically or causally related to other attributes or categories. The latent taxa that undergird a scientific nosology are not therefore mere collections of overtly similar attributes or categories, but linked or unified patterns of known or presumed relations among them. These theoretically grounded patterns of relations provide the foundation of a scientific taxonomy. Several benefits of systematizing clinical data in a theoretical fashion are not readily available from either clinical or numerical procedures (Wright & Murphy, 1984). Given the countless ways of observing and analyzing a set of data, a system of explanatory propositions becomes a useful guide to clinicians as they seek to comprehend the stream of amorphous signs and chaotic symptoms they normally encounter. Rather than shifting from one aspect of behavior, thought, or emotion to another, according

Kendell’s (1975) comment of more than three decades ago, on reviewing the preceding 20-year period, is unfortunately no less apt today than it was then: Looking back on the various studies published in the last twenty years it is clear that many investigators, clinicians and statisticians, have had a naive, almost Baconian, attitude to the statistical techniques they were employing, putting in all the data at their disposal on the assumption that the computer would sort out the relevant from the irrelevant and expose the underlying principles and regularities, and assuming all that was required of them was to collect the data assiduously beforehand. (p.€118)

Theoretically Deduced Constructs In the early stages of knowledge, the categories of a classification rely invariably on observed similarities among phenomena (Tversky, 1977). As knowledge advances, overt similarities are discovered to be an insufficient, if not false, basis for cohering categories and imbuing them with scientific meaning (Smith & Medin, 1981). As Hempel (1965) and Quine (1977) have pointed out, theory provides the glue that holds a clas-

the development of a scientific discipline may often be said to proceed from an initial “natural history” stage .€.€. to subsequent more and more “theoretical” stages. .€.€. The vocabulary required in the early stages of this development will be largely observational. .€.€. The shift toward theoretical systematization is marked by the introduction of new, “theoretical” terms .€.€. more or less removed from the level of directly observable things and events. These terms have a distinct meaning and function only in the context of a corresponding theory. (pp.€139–140)


to momentary impressions of importance, theoretically guided clinicians may be led to pursue in a logical and perhaps more penetrating manner only those aspects that are likely to be related (Dougherty, 1978). In addition to furnishing this guidance, a theoretically anchored taxonomy may enable diagnosticians to generate insights into clinical relations they may not have grasped before. Furthermore, it ought to enlarge the sensitivity and scope of knowledge of observers by alerting them to previously unseen relations among attributes, and then guiding these new observations into a coherent body of knowledge. Taxonomic theories need be neither fully comprehensive nor extensively supported to inspire and guide the early phases of taxonomic development. Meehl (1972) addressed these points with relevance to his concept of the schizophrenia taxon: I would not require that a genuinely integrated theory explain everything about schizophrenia, a preposterous demand, which we do not customarily make of any theory in the biological or social sciences. At this stage of our knowledge, it is probably bad strategy to spend time theorizing about small effects, low correlations, minor discrepancies between studies and the like. Being a neo-Â�Popperian in the philosophy of science, I am myself quite comfortable engaging in speculative formulations completely unsubstantiated by data. To “justify” concocting a theory, all one needs is a problem, plus a notion (I use a weak word advisedly) of how one might test one’s theory (subject it to the danger of refutation). (p.€11)

The reader may be taken aback by Meehl’s seemingly tolerant views and conclude that theory can lead to scientific irresponsibility—a conclusion that would justify taking a rigorous atheoretical stance. As I have implied previously, however, the belief that one can take positions free of theoretical bias is naive, if not nonsensical (Heelan, 1977; Hempel, 1965; Kukla, 1989; Leahey, 1980; Weimer, 1979). Those who claim to have eschewed theory are likely to have (unknowingly) subscribed to a position that gives primacy to such experience-near data as overt behaviors and biological signs, as opposed to experience-Â�distant data that require a greater measure of inference. The positivist (em-


piricist) position may once have held sway in philosophy, as it still does in some psychiatric and psychological quarters (Schwartz & Wiggins, 1986); however, it is difficult, as Meehl (1978) noted, “to name a single logician or a philosopher (or historian) of science who today defends strict operationism in the sense that some psychologists claim to believe in it” (p.€815). What distinguishes a true theoretically based taxonomy from one that merely provides an explanatory summary of known observations and inferences? Essentially, the answer lies in its power to generate new attributes, relations, or taxa—that is, ones other than those used to construct it. This generative power is what Hempel (1961) termed the “systematic import” (p.€ 6) of a scientific classification. In contrasting what are familiarly known as “natural” (theoretically guided and based in deduction) and “artificial” (conceptually barren and based in similarity) classifications, Hempel (1965) wrote: Distinctions between “natural” and “artificial” classifications may well be explicated as referring to the difference between classifications that are scientifically fruitful and those that are not; in a classification of the former kind, those characteristics of the elements which serve as criteria of membership in a given class are associated, universally or with high probability, with more or less extensive clusters of other characteristics. Classification of this sort should be viewed as somehow having objective existence in nature, as “carving nature at the joints” in contradistinction of “artificial” classification, in which the defining characteristics have few explanatory or predictive connection with other traits. In the course of scientific development, classifications defined by reference to manifest, observable characteristics will tend to give way to systems based on theoretical concepts. (pp.€146–148)

Evaluative Standards Feinstein (1977) commented that a classification system “can be a product of sheer speculation or arbitrary caprice” (p.€196). Hence once a taxonomy has been constructed, be it comprehensive or circ*mscribed, it behooves clinicians and scientists to examine it as a

Considerations in Psychopathology and Personology

unit—that is, to evaluate its constituent taxa, as well as the specific attributes and defining features that constitute each taxon. To ensure a minimum of “sheer speculation or .€.€. caprice,” developers of taxonomies must keep in mind several principles or standards that may optimize both the validity and utility of their creations. Such guidelines may prove especially useful in the formulation and construction phases of a taxonomy and may serve to orient developers in ways that may enhance their system’s ultimate efficacy I now note a few of the principles or standards that may guide both the construction and evaluation of taxonomic taxa and attributes. More extensive discussions of these and other standards may be found elsewhere (Millon, 1987, 1991; Sneath & Sokal, 1973). For pedagogical purposes, I make a distinction between standards more applicable to the diagnostic attributes that constitute taxa and those more relevant to the structure of the taxonomy. Optimal Attribute Standards What are some of the properties of clinical attributes that enable them to serve as a secure base for diagnostic criteria? A few are worth noting. •• Feature comparability. One method of refining diagnostic comparisons is to spell out a series of defining features for every relevant clinical attribute associated with a set of parallel diagnostic taxa. For example, as I have noted earlier, if a clinical attribute— this time, let us say interpersonal conduct— is deemed of diagnostic value in identifying and differentiating personality disorder taxa, then a distinctive description must be written to represent the characteristic or singular manner in which persons with each personality disorder conduct their interpersonal lives. A format of this nature furnishes symmetry among the taxa that constitute a taxonomy and enables investigators to systematically compare each taxon’s diagnostic validity (e.g., sensitivity and specificity), as well as the relative diagnostic efficiency (e.g., positive and negative predictive power) for each relevant attribute (e.g., interpersonal conduct). •• Empirical reference. As I have also mentioned previously, attributes that de-


pend on higher-order inferences contribute to diagnostic unreliability. Whenever this is feasible, the features that constitute diagnostic criteria ought to be assigned properties in the observable world. Problems arise when one seeks to represent intrinsically unobservable processes (e.g., defense mechanisms), or when one attempts to balance the desire for generality or openness among attributes with the standard of empirical precision. Can attributes be empirically anchored (and thereby the ambiguity in language be minimized), while the attributes are simultaneously freed to encompass wide-Â�ranging phenomena, including those that reflect interior processes? •• Quantitative range. Clinical features usually express themselves as matters of degree rather than of simple presence or absence (e.g., severity of depression and level of anxiety). It is useful, therefore, if the clinical attributes that constitute taxa permit the registration of a wide range of intensity or frequency differences. This psychometric property of quantitative gradation is one of the notable strengths of psychological tests, but it is not limited to them. Optimal Structural Standards Unless a taxonomy is easy to understand and use, it is quite unlikely to gain adherents in the clinical world, no matter how well formulated and scientifically sound it may otherwise be. •• Clinical relevance. Some historic taxonomies were shrouded in a cloak of words and dense concepts. The structure of others was so opaque that assumptions were concealed, principles difficult to extract, and consistent connections to the clinical world impossible to establish. In short, the structure and language of the taxonomy and its taxa were formulated more complexly and obscurely than necessary. Relevance and simplicity suggest that a taxonomy should depend on a minimum number of assumptions, concepts, and categories. Alone, these standards neither eliminate taxonomic opaqueness nor validate the clinical utility of a taxonomy’s derivations. They merely suggest that excess and misguided baggage should be eliminated, so that the features of clinical relevance in the system can be seen more clearly.


•• Representative scope. If a taxonomy is too narrow in its range of applicability, failing to encompass disorders for which clinicians have diagnostic responsibility, then its level of utility and acceptance will be markedly diminished. Ideally, the number of taxa and attributes that a taxonomy subsumes ought not to be limited. However, it is wise to recognize that a disparity will exist between the potential range of a taxonomy’s applicability and its actual range of empirical support. •• Concurrent robustness. A question arises as to whether taxonic groupings will retain their membership composition under new conditions and with attributes other than those used to construct them initially. For example, monothetic (hom*ogeneous) taxa based on a single source of data (e.g., test scores), on one type of attribute (e.g., interpersonal conduct), or on one class of patients (e.g., inpatients) may fail to crossÂ�generalize—that is, to remain stable, distinct, and uniform when based on parallel yet unidentical sources of data (e.g., structured interviews), attributes (e.g., cognitive style), or populations (e.g., outpatients).

Closing Comments Although this chapter is long, it is but a brief sketch of a burgeoning field. Perhaps this introduction will tempt the reader to examine the subject in greater detail. At present, there are no unequivocal answers to the many questions posed in psychopathological taxonomy—be they the matters of selecting attributes, choosing structures, or opting for one construction method or another. In my critical remarks, I have not meant to imply that the philosophies and techniques of classification today are irrelevant, or that the theoretical or diagnostic underpinnings of contemporary practice are valueless. Rather, I wish to encourage current taxonomic conceptualizers to step back and reflect more deeply on established assumptions and formulations. On the one hand, taxonomies in psychopathology must not reduce the richness of the natural clinical world to a series of conflicts among competing and abstruse speculations; nor must this world be passively shaped by the dehumanized and arcane methods of mathematical analysis.


Protected from convention, vogue, presumption, or cabalism, taxonomic psychopathology will become neither dogmatic, trivial, and formalistic nor devoid of a substantive life of its own. Prevailing frameworks must continue to be challenged, and imaginative alternatives encouraged. Acknowledgment Portions of this chapter have been adapted and updated from Millon (1991). Copyright 1991 by the American Psychological Association. Adapted by permission.

References American Psychiatric Association. (1952). Diagnostic and statistical manual of mental disorders. Washington, DC: Author. American Psychiatric Association. (1968). Diagnostic and statistical manual of mental disorders (2nd ed.). Washington, DC: Author. American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. Andreasen, N., & Grove, W. (1982). The classification of depression: A comparison of traditional and mathematically derived approaches. American Journal of Psychiatry, 139, 45–52. Bandura, A. (1982). The psychology of chance encounters and life paths. American Psychologist, 37, 747–755. Benjamin, L. S. (1974). Structural analysis of social behavior. Psychological Review, 81, 392–425. Benjamin, L. S. (1986). Adding social and intrapsychic descriptors to Axis I of DSM-III. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV (pp.€599–637). New York: Guilford Press. Benjamin, L. S. (2006). Interpersonal reconstructive therapy: An integrative, personalitybased treatment for complex cases. New York: Guilford Press. Blashfield, R. K. (1984). The classification of psychopathology. New York: Plenum Press. Block, J. (1989). Critique of the act frequency ap-

Considerations in Psychopathology and Personology proach to personality. Journal of Personality and Social Psychology, 56, 234–245. Bridgman, P. W (1927). The logic of modern physics. New York: Macmillan. Buss, D. M., & Craik, K. H. (1983). The act frequency approach to personality. Psychological Review, 90, 105–126. Buss, D. M., & Craik, K. H. (1987). Act criteria for the diagnosis of personality disorders. Journal of Personality Disorders, 1, 73–81. Cantor, N., & Genero, N. (1986). Psychiatric diagnosis and natural categorization: A close analogy. In T Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV (pp.€ 233–256). New York: Guilford Press. Cantor, N., & Mischel, W. (1979). Prototypes in person perception. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 12). New York: Academic Press. Cattell, R. B. (1965). The scientific analysis of personality. Chicago: Aldine. Dougherty, J. W. D. (1978). Salience and relatively in classification. American Ethnologist, 5, 66–80. Essen-Â�Moller, E., & Wohlfahrt, S. (1947). Suggestions for the amendment of the official Swedish classification of mental disorders. Acta Psychiatrica Scandinavica, 47(Suppl.), 551–555. Eysenck, H. J. (1960). The structure of human personality. London: Routledge & Kegan Paul. Feinstein, A. R. (1977). A critical overview of diagnosis in psychiatry. In V. M. Rakoff, H. C. Stancer, & H. B. Kedward (Eds.), Psychiatric diagnosis (pp.€189–206). New York: Brunner/ Mazel. Frances, A. (1982). Categorical and dimensional systems of personality diagnosis: A comparison. Comprehensive Psychiatry, 23, 516–527. Frances, A., & Widiger, T. A. (1986). Methodological issues in personality disorder diagnosis. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV (pp.€ 381–400). New York: Guilford Press. Grove, W. M., & Andreasen, N. C. (1986). Multivariate statistical analysis in psychopathology In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV (pp.€347–362). New York: Guilford Press. Grove, W. M., & Tellegen, A. (1991). Problems in the classification of personality disorders. Journal of Personality Disorders, 5, 31–41.

171 Gunderson, J. G., Links, P. S., & Reich, J. H. (1991). Competing models of personality disorders. Journal of Personality Disorders, 5, 60–68. Hartigan, J. A. (1975). Clustering algorithms. New York: Wiley. Heelan, P. A. (1977). The nature of clinical science. Journal of Medicine and Philosophy, 2, 20–32. Hempel, C. G. (1961). Introduction to problems of taxonomy. In J. Zubin (Ed.), Field studies in the mental disorders (pp.€3–22). New York: Grune & Stratton. Hempel, C. G. (1965). Aspects of scientific explanation. New York: Free Press. Horowitz, L. M., Post, D. L., French, R. de S., Wallis, K. D., & Siegelman, E. Y. (1981). The prototype as a construct in abnormal psychology: 2. Clarifying disagreement in psychiatric judgments. Journal of Abnormal Psychology, 90, 575–585. Jackson, D. W. (1971). The dynamics of structured personality tests. Psychological Review, 78, 229–248. Kaplan, A. (1964). The conduct of inquiry. San Francisco: Chandler. Kendell, R. E. (1968). The classification of depressive illness. London: Oxford University Press. Kendell, R. E. (1975). The role of diagnosis in psychiatry. Oxford, UK: Blackwell. Kiesler, D. J. (1983). The 1982 Interpersonal Circle: A taxonomy for complementarity in human transactions. Psychological Review, 90, 185–214. Kiesler, D. (1996). Contemporary interpersonal theory and research: Personality, psychopathology, and psychotherapy. New York: Wiley. Kraepelin, E. (1899). Psychiatrie: Ein Lehrbuch (6th ed.). Leipzig: Barth. Krueger, R. F., & Tackett J. L. (Eds.). (2006). Personality and psychopathology. New York: Guilford Press. Kukla, A. (1989). Nonempirical issues in psychology. American Psychologist, 44, 785–794. Leahey, T. (1980). The myth of operationism. Journal of Mind and Behavior, 1, 127–143. Leary, T. (1957). Interpersonal diagnosis and personality. New York: Ronald Press. Livesley, W. J. (1985). The classification of personality disorder: 1. The choice of category concept. Canadian Journal of Psychiatry, 30, 353–358. Livesley, W. J. (1991). Classifying personality dis-

172 orders: Ideal types, prototypes or dimensions. Journal of Personality Disorders, 5, 55–59. Loevinger, J. (1957). Objective tests as measurements of psychological theory. Psychological Reports, 3, 635–694. Lorr, M. (1966). Explorations in typing psychotics. New York: Pergamon Press. Medin, D. L., Altom, M. W., Edelson, S. M., & Freko, D. (1982). Correlated symptoms and simulated medical classification. Journal of Experimental Psychology: Learning, Memory, and Cognition, 8, 37–50. Meehl, P. E. (1972). Specific genetic etiology, psychodynamics, and therapeutic nihilism. International Journal of Mental Health, 1, 10–27. Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology, 46, 806– 834. Menninger, K. (1963). The vital balance. New York: Viking. Mezzich, J. E. (1979). Patterns and issues in the multiaxial psychiatric diagnosis. Psychological Medicine, 9, 125–137. Millon, T. (1969). Modern psychopathology: A biosocial approach to maladaptive learning and functioning. Philadelphia: Saunders. Millon, T. (1977). Millon Clinical Multiaxial Inventory manual. Minneapolis, MN: National Computer Systems. Millon, T. (1981). Disorders of personality: DSM-III, Axis II. New York: Wiley. Millon, T. (1983). The DSM-III: An insider’s perspective. American Psychologist, 38, 804– 814. Millon, T. (1984). On the renaissance of personality assessment and personality theory. Journal of Personality Assessment, 48, 450–466. Millon, T. (1986). Personality prototypes and their diagnostic criteria. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV (pp.€671–712). New York: Guilford Press. Millon, T. (1987). On the nature of taxonomy in psychopathology. In C. G. Last & M. Hersen (Eds.), Issues in diagnostic research (pp.€3 –85). New York: Plenum Press. Millon, T. (1990). Toward a new personology: An evolutionary model. New York: WileyÂ�Interscience. Millon, T. (1991). Classification in psychopathology: Rationale, alternatives, and standards. Journal of Abnormal Psychology, 100, 245– 261.

CONCEPTUAL ISSUES IN CLASSIFICATION Osherson, D. N., & Smith, E. E. (1981). On the adequacy of prototype theory as a theory of concepts. Cognition, 9, 35–58. Panzetta, A. E. (1974). Towards a scientific psychiatric nosology: Conceptual and pragmatic issues. Archives of General Psychiatry, 30, 154–161. Pap, A. (1953). Reduction-Â�sentences and open concepts. Methodos, 5, 3–30. Paykel, E. S. (1971). Classification of depressed patients: A cluster analysis derived grouping. British Journal of Psychiatry, 118, 275–288. Pincus, A. L., & Wiggins, J. S. (1990). Interpersonal problems and conceptions of personality disorders. Journal of Personality Disorders, 4, 342–352. Plutchik, R., & Conte, H. R. (1985). Quantitative assessment of personality disorders. In J. O. Cavenar (Ed.), Psychiatry (Vol. 1). Philadelphia: Lippincott. Plutchik, R., & Platman, S. R. (1977). Personality connotations of psychiatric diagnoses: Implications for a similarity model. Journal of Nervous and Mental Disease, 165, 418–422. Quine, W. V. O. (1977). Natural kinds. In S. P. Schwartz (Ed.), Naming, necessity and natural groups. Ithaca, NY: Cornell University Press. Rosch, E. (1978). Principles of categorization. In E. Rosch & D. B. Lloyds (Eds.), Cognition and categorization. Hillsdale, NJ: Erlbaum. Sartorius, N. (1990). Sources and traditions of psychiatric classification: Introduction. In N. Sartorius, A. Jablensky, D. A. Regier, J. D. Burke, Jr., & R. M. A. Hirschfeld (Eds.), Sources and traditions of classification in psychiatry (pp.€1–6). Bern, Switzerland: Hogrefe & Huber. Schwartz, M. A. (1991). The nature and classification of the personality disorders: A reÂ�examination of basic premises. Journal of personality Disorders, 5, 25–30. Schwartz, M. A., & Wiggins, O. P. (1986). Logical empiricism and psychiatric classification. Comprehensive Psychiatry, 27, 101–114. Schwartz, M. A., Wiggins, O. P., & Norko, M. A. (1989). Prototypes, ideal types, and personality disorders: The return of classical psychiatry. Journal of Personality Disorders, 3, 1–9. Sim, J. P., & Romney, D. M. (1990). The relationship between a circumplex model of interpersonal behaviors and personality disorders. Journal of Personality Disorders, 4, 329–341. Simpson, G. G. (1961). Principles of animal taxonomy. New York: Columbia University Press.

Considerations in Psychopathology and Personology Skinner, H., & Blashfield, R. (1982). Increasing the impact of cluster analysis research: The case of psychiatric classification. Journal of Consulting and Clinical Psychology, 50, 727–735. Skinner, H. A. (1981). Toward the integration of classification theory and methods. Journal of Abnormal Psychology, 90, 68–87. Skinner, H. A. (1986). Construct validation approach to psychiatric classification. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology: Toward the DSM-IV (pp.€307–329). New York: Guilford Press. Smith, E. E., & Medin, D. L. (1981). Categories and concepts. Cambridge, MA: Harvard University Press. Sneath, P. H. A., & Sokal, R. R. (1973). Numerical taxonomy. San Francisco: Freeman. Sokal, R. R. (1974). Classification: Purposes, principles, progress, prospects. Science, 185, 1115–1123. Sokal, R. R., & Sneath, P. H. A. (1963). Principles of numerical taxonomy. San Francisco: Freeman. Spitzer, M. (1990). Why philosophy? In M. Spitzer & B. A. Maher (Eds.), Philosophy and psychopathology. New York: Springer-Â�Verlag. Strack, S. (Ed.). (2005). Handbook of personology and psychopathology. Hoboken, NJ: Wiley Strack, S. (Ed.). (2006). Differentiating normal and abnormal personality. New York: Springer. Strack, S., Lorr, M., & Campbell, L. (1990). An evaluation of Millon’s circular model of personality disorders. Journal of Personality Disorders, 4, 353–362. Strauss, J. S. (1986). Psychiatric diagnoses: A reconsideration based on longitudinal processes. In T. Millon & G. L. Klerman (Eds.), Con-

173 temporary directions in psychopathology: Toward the DSM-IV (pp.€257–264). New York: Guilford Press. Tversky, A. (1977). Features of similarity. Psychological Review, 84, 327–352. Weimer, W. (1979). Notes on the methodology of scientific research. Hillsdale, NJ: Erlbaum. Widiger, T. A., & Frances, A. (1985a). Axis II personality disorders: Diagnostic and treatment issues. Hospital and Community Psychiatry, 36, 619–627. Widiger, T. A., & Frances, A. (1985b). The DSMIII personality disorders: Perspectives from psychology. Archives of General Psychiatry, 42, 615–623. Williams, J. B. W. (1985a). The multiaxial system of DSM-III. Where did it come from and€ where should it go?: I. Its origins and critiques. Archives of General Psychiatry, 42, 175–180. Williams, J. B. W. (1985b). The multiaxial system of DSM-III. Where did it come from and where should it go?: II. Empirical studies, innovations, and recommendations. Archives of General Psychiatry, 42, 181–186. Wittgenstein, L. (1953). Philosophical investigations. Oxford, UK: Blackwell. World Health Organization. (1992). International statistical classification of diseases and related health problems (10th rev.). Geneva: Author. Wright, J. C., & Murphy, G. L. (1984). The utility of theories in intuitive statistics: The robustness of theory-based judgments. Journal of Experimental Psychology: General, 113, 301–322. Zubin, J. (1968). Biometric assessment of mental patients. In M. M. Katz, J. O. Cole, & W. E. Barton (Eds.), Classification of psychiatry and psychopathology (pp.€353–376). Washington, DC: U.S. Public Health Service.

Chapter 8

Diagnostic Taxa as Open Concepts Metatheoretical and Statistical Questions about Reliability and€Construct Validity in the Grand Strategy of Nosological Revision Paul E. Meehl

ince I find it hard to conceive that a raS tional mind could think otherwise, I presuppose that, ceteris paribus, careful de-

lineation of the signs, symptoms, and course of a disorder (I cannot interest myself much in the semantic hassle over whether to call it “disease”) so as to increase the reliability of classifying clients or patients is desirable. While reliability and validity are not the same thing, it is a psychometric truism that the former bounds the latter, although it is worth mentioning that the bound is the square root of the reliability, so validity can theoretically be larger. Usually the operative validity (net attenuated construct validity) runs far below that upper bound set by the square root of the reliability coefficient. Hence alterations in the format of assessment or in the content sampled, which might under some circ*mstances reduce reliability, could nevertheless increase the net attenuated construct validity. Similarly, changes in content or format that increase reliability may theoretically decrease validity. For instance, an alteration in the open-ended, unstructured format of Rorschach administration (as was attempted during World War II to make it possible to test large numbers of individuals and score reliably without in-


quiry) seemed to eliminate whatever slight validity the instrument had as usually administered. There is no mystery about this, although it is paradoxical at first look. We may be concerned about the reliability of behavior sampling by two different samplers (“interjudge agreement”) or with the trustworthiness of a sample as drawn by an individual judge (how many marbles do we draw from the urn, and how do we draw them?). In either case, the point is this: Whether an interview behavior or a psychological test item is viewed primarily as “sample” or as “sign” (Cronbach & Meehl, 1955), there are kinds of alterations in the examining situation and in the procedure of response classification that can alter qualitatively the intrinsic construct validity of the sample in such a way as to reduce its net validity—Â�despite reliability, in either of the two senses mentioned, having been enhanced. I am not arguing that such has occurred in the process of improving our old Mental Status Examination or in the construction of DSM-III, but merely that this methodological point should be kept in mind when discussing reliability/validity questions.

Diagnostic Taxa as Open Concepts

It requires neither psychometric nor philosophical expertise to see that the reliability/ validity helps and tradeoffs can be somewhat complicated, and especially so when the aimed-at diagnostic construct itself (category or dimension) is an open concept, lacking a definitive “operational” criterion, specified implicitly (“contextual definition”) by presumably fallible indicators. In that kind of knowledge situation, we subtly alter meanings as we discover facts; we amend theoretical definitions as we revise indicator weights. The basic point can be better brought out by considering the decision to include an unreliable indicator in a standard examination for any “disease.” A general medical examination always includes blood pressure and not anthropometric determination of wrist width, despite the mediocre reliability of the former and r = .98 for the latter. We do not find this evidentiary preference puzzling; we simply say, “Blood pressure unreliably measured is a stronger indicator of more different and important conditions than wrist width reliably measured.” Similarly, a psychotherapist who employs dream interpretation (with the manifest–Â�latent content model) would not seriously consider substituting reliably scorable multiple-Â�choice inquiry for free association under the Fundamental Rule, despite the grave reliability problems posed by the classical procedure. One might prefer to avoid Freud’s technique altogether, and partly because of unreliability considerations (cf. Meehl, 1983); what one would almost surely not do is retain Freud’s core idea and its entailed technique, while substituting a multiple-Â�choice inquiry in the service of reliability. The reasons for desiring diagnostic reliability are well known. The most important reason is generalizability of research findings by other investigators thinking about their research and by practitioners in applying research findings to clinical decision making. The easiest way to understand the former is in terms of the number of pairwise relationships of input and output variables involved in a decision-Â�making process, whether of a theoretical or practical clinical nature (Meehl, 1959). If a set of behavior data (history and current status, interview, ward behavior, neurological, psychometric) permits us to classify patients in some “rational” (ultimately “causal”?) way, it is not


necessary that each of the possible n output variables (e.g., treatment of choice, second choice, prognosis, employability, response to group therapy, suicide risk, genetic risk to offspring) has to be correlated singly pairwise with all the input variables—a process which would require studying mn relationships, where mn is in the thousands (Meehl & Golden, 1982, pp.€130–131). Instead, we can first relate the m input variables to the diagnostic dimension or rubric and then relate the diagnostic dimension or rubric to the several output variables of interest. Hence only (m + n) correlations need to be studied. But that process cannot be carried out with any confidence if the relation of some of the input variables to dimension X or categorical rubric C as found at the University of Texas has only a little better than chance relationship between (only partially overlapping) relationships as reported by investigators in Milwaukee. The pulling together of research data to give a coherent interpretation of an alleged psychiatric entity, whether taxonomic or dimensional in nature, presupposes the possibility of scanning the research literature with at least some reasonable confidence that patients called “schizophrenic” by one investigator are like those called “schizophrenic” by another. Similarly, suppose a clinician reads a research report claiming that a certain drug is efficacious for [patients with] paranoid schizophreni[a], except when they have a history of an episode, when much younger, of acute catatonic excitement. That report is not helpful to the clinician who lacks rational belief that the investigator was looking at the same indicators of paranoid and catatonic schizophrenia that he or she can now look at in his or her clinical decision making. It is an interesting question whether one can ever lose by improving reliability, except in the (rare? I don’t know) sense discussed above. The main respect in which some workers, and I gingerly include myself here, seem to worry about it is that research aimed at improving, correcting, or—in the extreme case—Â�refuting views implicit in the DSM-III conceptual system will somehow be cramped by an overly enthusiastic view of it, which sometimes takes the form of a dogmatic insistence upon its merits throughout. I have heard research-Â�oriented clinicians express concern about this, but it is difficult to


track down persuasive examples where, for instance, an otherwise admirable research proposal was rejected by the peer reviewers on the grounds that it did not employ “official categories” approved by DSM-III. While people talk about this, and one sometimes hears it alleged that it has occurred, I do not myself know of any clear cases. Admittedly, it would be hard to ascertain whether a subtle kind of social process, of the kind that the Supreme Court likes to call a “chilling effect,” is taking place. Some researchers might be otherwise disposed to advocate a mild Feyerabendian “proliferation of theories” (Feyerabend, 1970), which he advocates even for cases when the going theories are extremely powerful and well corroborated and, a fortiori, for theories in such primitive fields as psychopathology. Some of them might not be getting research grant money because they have timidly avoided challenging the establishment category system. Here again, I have no affirmative evidence that such things happen. If they do, it would appear quite easy to find a way around it, and whether it failed would hinge upon whether some peer reviewers have become overidentified with the present product. For example, suppose I am interested in studying people with a cyclothymic personality makeup who have very mild ups and downs on an endogenous (genetic/biochemical) basis, but who at no time become diagnosably psychotic or even semipsychotic. The psychiatric tradition has connected endogenousness with severity, which there is no strong theoretical reason for insisting upon, although there is a correlation empirically. I don’t see why a clinical investigator, behavior geneticist, or neurochemist should in any way be hampered by the received rubrics. He or she can be careful in adhering to criteria for diagnosing manic–Â�depressive disorder as given by DSM-III. It may well be that the only available rubric for some of the other people he or she wants to study is “normal,” or even perhaps some other piece of nonmanic–Â�depressive terminology as specified in DSM-III. Nothing prevents the investigator from saying, in writing up a grant proposal, “It is my empirical conjecture that there are persons who don’t manage quite to squeak through the conditions for diagnosing a manic–Â�depressive attack (because of extreme damping in their cyclothymic cycle). But all


of my classifications are indicated, and all of the correlations of them with all of the other things I studied, whether psychometric or genetic or familial or whatever, are clearly indicated, so that other investigators may rely on the fact that I stuck literally to the received criteria for making that diagnosis. I have also listed, however, the set of special criteria, together with their time sampling and interjudge reliabilities, that I used to demarcate my special subgroup of individuals that do not fit the official rubrics.” At no point does this investigator have to depart from the semantics of DSM-III; nor does he or she have to do any inordinate amount of work in order to include the DSM-III criteria as available for investigators who want to examine his or her data critically. It is, I suppose, imaginable that somebody might want to do something where the task of “double diagnosis” (i.e., according to his or her conjectured criteria for entities or dimensions not in the official list along with the received one) will be a considerable amount of excess work, but I am not aware of any clear [evidence] showing that that has happened. The diagnostic criteria for DSMIII simply do not involve that much additional work, and most of the overload will arise from his or her idiosyncratic system. Despite the fact that my own views on many categories are quite heterodox, when there is an adoption by an empowered body of clinicians and scientists as to a certain terminology, I think one is not unduly burdened or imposed upon by some extra scientific or clinical toil when the investigator chooses to deviate from it in his or her own research. An interesting statistical question arises in the “context of discovery” (Reichenbach, 1938, pp.€ 6–7), where a plausible case—I do not urge that it is more than plausible— can be made for concern about increased difficulty of detecting subtle relationships. I mean by “detection” the development of a clinical hunch and, in a more formalized research context, the problem of the statistical power function failing to detect something that is there. Consider the following: By tightening up the diagnostic criteria, we have increased reliability and, hence (almost certainly), the net attenuated construct validity in identifying the whole class of patients called “schizophrenic.” In the course of so doing, we have been forced to elimi-

Diagnostic Taxa as Open Concepts

nate some signs and symptoms that some clinicians have been relying on. Perhaps we ourselves had been doing so, but we are willing to pay this price. We are even willing to pay the price of dropping something that was considered fundamental by the master himself—as, for instance, DSM-III does not include Bleuler’s ambivalence or his autism; or, to take an instance closer to my heart, Rado’s anhedonia (Meehl, 1962, 1964, 1974–1975, 1975). Less counterconventional, one thinks of the pan-Â�anxiety considered extremely important—Â�perhaps the most important single symptom—in the “pseudoneurotic schizophrenia” syndrome described by Hoch and Polatin (1949). The latter two examples are of course controversial; but as to the former, it is hard to believe that we should omit two of Bleuler’s cardinal signs unless this choice is dictated by difficulty objectifying them in the interest of reliability. I repeat that I am not here disputing the claim that the net attenuated construct validity for identifying the whole class of [patients called] “schizophrenic” has been increased by the tightening process, and I am not at the moment concerned with the efficacy of clinical handling, but I am attending to the research context. It is surely possible, and to a statistically and philosophically sophisticated person not even paradoxical, that some subset of patients sharing underlying etiology and psychopathology (genetics, biochemistry, [central nervous system] fine structure, and psychodynamics or “personality structure”) with the core group of [patients who have] schizophrenias but who, because of modifying genes and normal-range individual Â�differences factors (Meehl, 1975) as well as life history experiences, do not develop the signs and symptoms that have remained in the selected list of DSM-III, or—Â�equally possible despite average heightened reliability—do not have them in sufficient quantity to be clear instances. Such a state of affairs is not only consistent with, but is probabilistically inferable from either the medical model, classical psychometrics, genetics, learning theory, or ordinary trait theory. The point is that clinicians trained to classify patients with the reduced highÂ�reliable list of criteria will not be psychologically disposed to consider the subset of peripheral or borderline cases as belonging


to the schizophrenic group (as they should not in applying the objectified criteria). In the context of discovery, this could sometimes operate adversely, since the way you categorize your world, as we all know, will in considerable part determine what you are capable of noticing. But suppose a perceptive clinician does notice something about these borderline cases and undertakes a systematic research study of something middling complicated and not easy to discern—say, for example, a second-order interaction between phenothiazines and a certain mode of psychotherapeutic intervention (e.g., [rational–Â�emotive therapy]). Now if the polygenic modifiers or environmental factors that make the [patients with] atypical schizophrenias show a different kind of interaction effect from the core group, that will not be detected statistically, even having been noticed clinically by a gifted clinician, because such cases will only rarely (and mostly due to carelessness in applying the new criteria!) be included in the study. If one believes (as I do) that the psychiatric treatment of the future will involve complicated kinds of actuarial grounds for selecting and sequencing the treatment of choice (Meehl, 1972a, pp.€ 135–137), early research progress along such lines could be hampered in this way. I want to emphasize that I’m not here invoking some kind of vague clinical intuitions about “patterns.” I am making a simple point about research statistics—that is, that you can’t detect a trend that makes a subset of subjects different from the other subjects in a certain group if there aren’t any of the subset present in the study. Furthermore, as we move into higher-order interaction effects, such as Drug × Psychotherapy × Subdiagnosis patterns, the degrees of freedom shrink so that errors of Type II begin to preponderate due to marked reduction in statistical power. It might be argued that while this may impose an irksome hurdle in the context of discovery at the intuitive stage for the clinician trained in the use of DSM-III, and thinking more or less automatically that way, it will not have any long-run bad effect because the cases not included in such studies will be detectable in studies focusing on some other diagnostic rubric. I think that is an optimistic view because it implies that


some sort of massive research network of all possible combinations of everything with everything is going to take place in psychiatry and clinical psychology, which it is not, both for economic and professional interest reasons. Furthermore, what kind of thing is detected will depend on what initial overall rubric is being studied. If these borderline cases were subsumed under “anxiety state” rather than “borderline schizophrenia,” the interaction effect between an antipsychotic drug and cognitive therapy will not be a likely subject matter of investigation. Finally, what is perhaps the more serious statistical point, such people will not be found in any one rubric if misdiagnosed because of the tight criteria (by misdiagnosed, I of course mean subsumed to the wrong specific etiological group [Meehl, 1972b, 1977] in the eyes of Omniscient Jones), hut are likely to be dispersed. When one disperses a group of people who are heterogeneous in some respects, but hom*ogeneous in some core feature of high causal relevance, into a number of heterogeneous diagnostic categories, the best bet is that they will simply get lost in the shuffle. While I do not claim to know that this is a serious problem, it is not a silly consideration that can be dismissed out of hand without thorough mathematical analysis. Moving away from what one might call the “political–Â�social–economic” impact of DSM-III, it is worthwhile to examine at a more philosophical level the ways in which a practitioner or researcher may view its categories and dimensions. I can see three (although not sharply demarcated), one of which is admirable, one of which is criticizable but fairly harmless, and only the third of which is scientifically malignant. The first is to view the delineation of a syndrome as an empirically observed (clinically or statistically!) cluster, a syndrome plus course, that suggests to us some kind of underlying causal hom*ogeneity in the subjects who show it—Â�although we may, depending on our theoretical predilections, sit quite loosely to this etiological promissory note. Its justification is mainly communicative and pragmatic, together with whatever degree of faith we have from the history of medicine (and genetics, and psychometrics) that future research will give us a more detailed understanding of whatever historical and “latent” (inner) current processes and structures are


at work to produce the covariation of the signs, symptoms, aspects of course, prognosis, and response to treatment. Covariation is the essence of descriptive science and the touchstone of scientific thinking, whether we read such diverse writers as Freud, Skinner, Allport, Murray, Eysenck, Thurstone, or Cattell—Â�strange bedfellows indeed, whose unanimity on this point should surely tell us something about how to study the mind! Ceteris paribus again, the more standardized the examination can be made, the more objectively described the classification of the responses, and as a result, the greater interjudge agreement by different examiners, and the more striking the empirical “tightness” of the cluster, the better we like the syndrome as an entity. As already stated, it is hard to understand why a rational mind would object to any approach that enhances these desirable properties. Second, one may believe that DSM is the best that can be achieved, at least in the foreseeable future, and may be suspicious or even antagonistic to deviations from it, for either clinical or research purposes. This attitude troubles me, but I should think it can be adequately buffered by the practice I suggested above—that is, that investigators have a responsibility to employ it until it is officially revised by some “culturally empowered” group such as those who constructed it in the first place. But we do not pressure researchers or punish them financially or otherwise, once they have met these conditions in their semantics, for delineating some further conjectural entities or dimensions of their own, hoping to persuade the profession on the basis of clinical experience of better evidence that they are right. It is the third attitude which I think is malignant, partly because of its potential chilling effect, but mainly because it is philosophically so terribly mistaken. It says not merely that “this is a good thing so far as it goes, and should not be lightly discarded or whimsically amended.” This third view claims it is the truth, as a matter of some kind of rigorous definition process. The extreme (simplistic, “vulgar operationist”) form of this view is that the very meaning of the concepts is contained, exhaustively and explicitly, in the “operational definitions” provided by DSM. It would be hard to find one single logician or historian of sci-

Diagnostic Taxa as Open Concepts

ence today (or for that matter, since around 1935!) who would countenance the conception of scientific method enshrined in this view. I find it puzzling that physicians, or for that matter, psychologists, unless they are of the most dogmatic behaviorist kind, should adopt this position when neither the history of organic medicine, nor of genetics (I don’t mean here merely behavior genetics), nor of traditional trait theory in academic psychology, nor of classical psychometrics gives any support to it. It is simply not true that diseases in organic medicine are “defined by” the syndrome or by the syndrome and course together. Organic diseases are defined by a conjunction of their etiology and pathology when these are known, and otherwise— with much less scientific assurance—as syndromes remaining to be researched so as to be medically understood. A disease entity, as delineated in the early stages of clinical experience and scientific study, at the level of mere syndrome description when there is as yet no (or minimal and conjectural) knowledge of the etiology or pathology underlying it, is an open concept (Cronbach & Meehl, 1955; Meehl, 1972b, 1977; Meehl & Golden, 1982; Pap, 1953, 1958, Chap.€ 11). It is neither philosophically rigorous nor scientifically sophisticated to make a literal identification of a disease entity with its currently accepted signs and symptoms. Corresponding to organic medicine’s pathology (in a more extended sense than that envisaged by Virchow) is personality structure (genotypic traits, psychodynamics). Corresponding to etiology are, except for an environmentalist fanatic, the genetic predispositions not only to specific mental disorders, but to “temperamental genotypic traits” generally, such as anxiety conditionability, rage readiness, hedonic capacity, general intelligence, and the like, and the learning history imposed on an organism whose varied behavior acquisition functions are characterized by such-and-such inherited parameters. Our problem in psychopathology of the so-Â�called functional behavior disorders is obvious—to wit, that we do not possess an equivalent to the pathologist’s and microbiologist’s report telling us the “right answer” at the conclusion of a clinicopathological case conference (Meehl, 1973, pp.€ 284–289). If I make a psychodynamic inference, it is not possible for me to ask the psychopathologist whether


his [or her] stained slides showed the patient’s psyche had holes in the superego. To a thoughtful clinician with philosophical sophistication, it is perfectly obvious that disease syndromes are inherently open concepts, as mentioned above. Nothing but dogmatism on the one hand, or confusion on the other, is produced by pretending to give operational definitions in which the disease entity is literally identified with the list of signs and symptoms. Such an operational definition is a fake. If somebody does not like the medical model (and if that’s the case, one wouldn’t be taking the DSM—concocted by a group of psychiatrists for medical purposes—Â�seriously to begin with), he [or she] should be reminded that in classical psychometrics (such as factor analysis) or in more recent developments (such as multidimensional scaling), we cannot even write the basic equations, let alone the embedding interpretative text required to give empirical meaning to the variables in those equations, unless a clear distinction is already made between the manifest behavior indicators and the inferred (latent, causal) factors. The same is true of biophysical trait theory as classically elaborated by Allport (1937), Murray (1938), Cattell (1946), and others. Obviously, the great breakthrough in genetics with Mendel, and the rediscovery of Mendel’s concepts at the turn of the [20th] century, hinged upon the distinction between the genotype and the phenotype. This distinction forced theoretical recognition that under many circ*mstances or available pedigrees, the weakly stochastic relationship between the two made an inference to genotype impossible. One simple-Â�minded mistake that I am surprised to find physicians making is to think that if, in a given concrete instance (single case, not class), we do not have a touchstone for testing whether a certain inferred construct property such as a latent disease is present or absent, that lack means that it is scientifically meaningless to ask the question—a view that the logician Carnap, a strongly positivist and tough-Â�minded philosopher of science, refuted definitively almost a half century ago! The same is true of most variants of learning theory—the old-Â�fashioned kind (Tolman, Hull, or Guthrie) as well as the souped-up developments in mathematical learning theory, information processing, and


cognitive processes generally that took place subsequently. The only plausible exception to the genotypic–Â�phenotypic, inner–outer, inferred–Â�observed distinction in learning theory is strict Skinnerian learning theory, which is almost entirely dispositional, although not as “pure” in this respect as some of its adherents like to think when they talk metatheory about it. I am fond of referring clinical psychology students to a little known two-page article published many years ago by the late T. A. Peppard (1949), a reputedly brilliant diagnostician who practiced internal medicine in Minneapolis for many years. He made a statistical study of the source of his diagnostic mistakes, using very strict criteria against postmortem findings. Errors of omission (well known to be commoner than errors of commission in medical diagnosis) sometimes occurred because he failed to look for something, other times because he looked for it but didn’t give it the proper weight, other times because he made an “error” on a judgment call, and so on. But the interesting thing is that 29% of the errors of omission were attributable, even by very tight standards imposed on himself, to the factor he called “symptoms and signs not found.” Of course, all physicians know the concept of “silent disease” such as an undiagnosed staghorn kidney or an early Pick’s frontal lobe atrophy, not to mention subjects with an epileptic brain wave who never have a fit and would not be discovered except for being the monozygotic twin of somebody who has a clinically recognizable convulsive disorder. I repeat that I find it strange that one must remind physicians about the distinction between the construct “disease” and its presently accessible symptom picture, although it is not so surprising that some psychologists confuse them. Finally, of course the most obvious example, which would still be persuasive to some of my generation, is psychodynamics, whose essence consists in the distinction between the easily observed manifest behavior or self-Â�awareness and the “hidden, latent, underlying source” of some aspect of observable covariation. Since neither psychodynamics, classical psychometrics, taxometrics, organic medicine, genetics, learning theory, nor trait theory has proceeded by explicit identification


between theoretical entities and their indicators, it would be strange to hold that rational use of DSM-III requires us to consider its syndromes as literally definitive and totally noninferential. It might be argued that if the builders of DSM had achieved consensus on constructing a purely descriptive (atheoretical, noninferential) “phenomenological” taxonomy, they should have proceeded by applying an appropriate formal cluster algorithm to a huge batch of carefully gathered clinical data, “letting the statistics do the whole job for them,” which would have saved a lot of conference time as well as generating a more objective scientific product. This sounds plausible to a psychologist, and maybe to some statisticians, but the main trouble with it is that there is no “accepted” cluster algorithm which is known to be sufficiently powerful to be used in this way (cf. Meehl, 1979). Even if there were such an agreedupon cluster analysis algorithm, one doubts that the committee could have proceeded in that way. The fact is that different clinicians do not share an equally “operational” view, partly for the reasons I have given and partly because of certain clinical (perhaps one could even say ideological) identifications—for example, between organicists and psychoanalysts, biotropes and sociotropes, scientists interested in genes and psychotherapists interested in battle-ax mothers. I am inclined to think that the next DSM development round ought to at least settle on some way of deciding when the orientation should be taxonomic versus dimensional. But that would hinge upon having a sufficiently well-Â�trusted algorithm for determining whether the latent order of a syndrome or dimension should be thought of as taxonic or nontaxonically factorial. Another possibility, which again seems simplistic and arbitrary until you ask what are the reasons for doing it another way, would be to collect all of the information or input kinds of variables, including life history data and the like, that go into diagnosis, and all of the output dispositions that are clinical reasons for making a diagnosis, such as differential response to psychotropic drugs, response to individual and group therapy, danger of acting out, suicide risk, and long-term employability. Absent cogent reasons for giving higher weight to some of these output ones

Diagnostic Taxa as Open Concepts

than others, it is arguable that the proper statistical model should be canonical correlation, in which we simultaneously optimize the predictability of the most predictable composite on the output side by optimal weights on the input variables. If the various output consequences of clinical importance are not prima facie very different in “importance,” if they are, so to speak, qualitatively of equal significance to us in decision making, then the difference in the weights they get might best be to weight them so as to make them collectively most predictable. The justification for defining a syndrome (or a nontaxonic factor) by some subset of input and output considered jointly would be that the canonical correlation between the two sets reaches a certain minimum size. It would be interesting, by the way, to ascertain whether such a distribution of candidate canonical correlations would show, if not an actual break, at least some tendency to bimodality, suggesting that some syndromes are “real” and others are more or less arbitrary carvings out by the clinician of regions of slightly greater densification in the multivariate descriptor space (but see Murphy, 1964). My own research interests are such that I consider that the initial distinction between whether one should proceed taxometrically or factorially should be given very great priority in the next revision. The question as to the desirability of adopting a fixed-rule approach to diagnostic criteria involves a complicated mix of statistical, philosophical, and clinical issues that are beyond the space limitations of this chapter and about which I myself have formed no definite opinion. This question has been aired recently in papers by Finn (1982, 1983) and Widiger (1983) (see also Meehl & Rosen [1955], comment by Cureton [1957], and Rorer, Hoffman, LaForge, & Hsieh [1966]). In thinking about this difficult question, it is necessary first to distinguish between issues regarding base rate fluctuations in different clinical or research populations and the separate but intimately related issues of clinical utility in treatment and prognosis. In saying these are distinct but intimately related, I mean to emphasize that from the standpoint of scientific realism (surely the implicit assumption of organic medicine, whether medical researchers or practitioners use the philosopher’s terminology for it


or not!), one does not wish to conflate the probability or corroboration of a diagnostic statement as a factual claim with the seriousness of a mistake. As Widiger worried about in his exchange with Finn, we do not want to adopt a decision rule based on a policy of systematically misdiagnosing patients on the grounds that correctly diagnosing a subset of them would, in certain pragmatic contexts, be too costly or risky or have too many side effects or make them more uncomfortable than the disease makes them or whatever. Crudely put, the first business of a diagnostic assertion is to be right! We cannot make use of differential utilities and disutilities of clinical errors without at least some crude assessments of diagnostic confidence, whereas we can investigate the optimality of a diagnostic procedure with regard to truth value without referring to any utilities other than the “cognitive utility” of being correct in our assertions. It would seem best, if it can be done and is psychologically acceptable to practitioners, to optimize the diagnoses by some suitable adjustment for known or guesstimated base rates in a given clinical population, and subsequently to raise the question of the various utilities involved in adopting a certain treatment plan or making predictions to the patient, court, employer, insurer, family, or whatever. In that mode of reasoning, the best inferable diagnostic statement is made first and the utilities are plugged in afterward. But this of course doesn’t take care of the base rate problem. Theoretically we know that both the cutting score on a variate which is an indicator of the disease entity and any formal or informal weighting of the scores or way of combining them into a pattern, as in Bayes’s formula, should not be done independently of the base rates. In ordinary clinical medicine, practitioners who never heard of the Reverend Thomas Bayes or the subsequent controversy about his ideas (this use of the formula itself is, of course, hardly controversial) make implicit use of it. They know that if you diagnose syphilis in Puerto Rico on the grounds of a positive Wassermann, you are likely to fall into errors that you would not make in Minnesota because of the geographic epidemiology of lues versus yaws. Every general practitioner at times says to the patient, “Well, I think you’ve got the winter crud; there’s a lot of


that going ground these days,” an informal Bayesian inference. It is an unsettled question how much the explicit and formalized inverse probability machinery of the statistician should become part of the decision making by a busy doctor. Of course, even given a certain diagnosis, perhaps tentatively arrived at with the intention to be flexible about revising it should the predicted results of a therapeutic intervention fail to materialize in the usual fashion, it is common practice, within the category of patients who meet the diagnostic criteria, to pay attention to the pattern of symptoms that is relevant to treatment choice and to include in this those “extraneous” characteristics (e.g., age, family, income, unrelated concurrent illness) that themselves did not enter into the diagnostic decision proper. There is nothing either wrong or particularly complicated about any of this. The only question is the extent to which formalization improves or impairs certain of these generally accepted clinical practices. Unfortunately, the behavior of a Bayes-Â�theorem-Â�computed inverse probability depends in somewhat complicated ways upon the distribution of sign validities, the relationship between valid and false-Â�positive rates, the extent to which the independence assumption of the signs pairwise is not satisfied, how robust the inferred diagnostic p value is with respect to departures from those assumptions, differential responsiveness of error rate at different regions of the base rate continuum, and the like. It would seem that some rather largescale but also intensive research by statisticians and clinicians would be in order. I do not think it is safe to assume that because such actuarial refinements are not part of the everyday mental habits of practitioners in organic medicine, then we don’t have to worry about it in psychopathology. There are probably important differences in the latter area. Furthermore, we still do not know the extent to which ordinary clinical practice of organic medicine commits more diagnostic errors than need be because of the extent to which the mathematics of clinical reference is not explicitly employed by the practitioner (Blois, 1980; Dawes, 1979; Dawes & Corrigan, 1974; Engelhardt, Spicker, & Towers, 1979; Goldberg, 1970, 1976; Gough, 1962; Holt, 1970, 1978; Kahneman, Slovic, & Tversky, 1982; Kleinmuntz, 1982;


Meehl, 1954, 1956a, 1956b, 1956c, 1957, 1960, 1967; Sawyer, 1966; Sines, 1970). Finally, how one thinks about this and what kinds of research are conducted depend on how confident we are that the underlying psychopathology is intrinsically taxonic (categorical, “typal”) versus nontaxonically multidimensional, where class concepts and qualitative predicates are only handy rubrics for roughly designating regions in an ontologically continuous descriptor hyperspace. While the very title of this volume orients us toward revision, one hopes that the intellectual fretfulness of primates and the availability of taxpayer dollars will not induce us to attempt substantial revisions until a large mass of evidence, including experimental research, clinical trials, quantitative analysis of clinical file data, and exchange of experience by seasoned practitioners of various persuasions, puts us in a position to do something more than speculate or nitpick. A tremendous amount of work by able people and a lot of taxpayer money went into generating the DSM product, and it is foolish to tinker with it very much, let alone undertake a complete overhaul, because it isn’t perfect, or because the results of an unavoidable compromise are not located precisely where one might prefer [them], given [one’s] own theory and practice. Sometimes the best advice is that of the Baptist preacher, “Leave it lay where Jesus flang it.” I bethink myself of how difficult it is for me, after 40 years on the Minnesota faculty, to interest myself in interminable discussions about how we should revise the written preliminary examination for the PhD so as to get a better assessment, reduce student anxiety, or whatever. A half century of observation (if I include my student days) reveals mostly primate meddlery, irrational optimism, a disinclination to consult the past, and the Hegelian swing of even short-term history! It goes without saying that the most important developments one can anticipate that would make it rational to revise are substantive advances in our understanding of mental disorder. But there are also, I think, several metaquestions that it would be desirable to have “settled” (if not exactly solved) before the next round of major revision. The reader will discern that the “answers” to these metaquestions involve a mix of mathematical development of statistical

Diagnostic Taxa as Open Concepts

methods especially suitable for taxonomic problems, the usual impact of substantive developments upon methodology (no contemporary philosopher of science conceives of methodology as entirely prior to theory), and considerations of clinical utility. I repeat that it is a grave mistake to conflate this last class of questions with questions regarding the intrinsic science—that is, factual validity—of any proposed concept. There are four metaquestions that should meanwhile be addressed by high-Â�competence investigators so that we will be in good methodological shape when the time for major revision arrives. Without dogmatism, I might go so far as to say that in my judgment until these four are answered, at least in the sense of a fairly high consensus among qualified individuals (there is no point in absolute democracy in a field like this!), we are probably not in a cognitive position that warrants a major revision being attempted. First, what role should a conjectural etiology, when moderately to strongly corroborated, play in the taxonomic strategy? Here one must avoid a simplistic division into “known” and “unverified” etiology, assuming a sharp dividing line where none exists even in organic medicine, genetics, or other fields of knowledge. It is obvious on mere inspection of the present list of rubrics that etiological factors partially understood, and in which varying degrees of “strong influence” as causal factors (Meehl, 1972b, 1977) must have been taken into account at least behind the scenes, have been unavoidable. It will be necessary to have a uniform standard of proof rather than a double standard of methodological morals such as prevails in some quarters today. For example, there are clinicians in the medical and psychological professions who resist recognizing the genetic influences in major mental disorders or, while reluctantly recognizing them, would not want to split the nosology of affective disorders into unipolar and bipolar, despite the strong evidence available presently as to the reality of that distinction genetically and its correlates with certain aspects of the syndrome, course, and so on. Yet some of these same clinicians, while justly pointing out that an absolute hammer-blow unavoidable demonstration (there is no such thing as this in empirical science, of course) has not been given for the unipolar–Â�bipolar distinction,


will in their own diagnostic thinking rely upon highly speculative psychodynamics, or family factors, or other alleged causal influences, whose degree of evidentiary support at the present time is nowhere in the running with that for the biological distinctions made. This parallels some clinical psychologists who, because of hostility to medicine (or simply poor training at a second-rate school?), continue to decry all psychiatric diagnosis as “mere labeling” or “completely unreliable,” refusing to read the quantitative evidence of diagnostic reliability developed in recent years, and then by some obscure mental process (which I confess myself quite unable to understand) proceed to substitute for such “unreliable” psychiatric nosology a batch of unproved, politicized social determiners, or flimsy psychodynamics inferred from an instrument with as low reliability and validity as the Rorschach! That is the sort of thing I mean by a double standard of epistemological morals. Second, the strategic distinction between thinking in terms of dimensions and categories (types, species, taxa, disease entities) remains with us. While one can get by with a kind of compromise between these, the basic theoretical claim of a classification system should be methodologically clear, even if a sizable proportion of patients are not clearly sortable into one or the other (a different question). Sooner or later we should get clear about which of our nosological rubrics are intended to be rough designations of persons’ location in a multidimensional descriptor space (whether phenotypic or genotypic, psychodynamic or genetic, that’s not the point) and which rubrics have a genuine typological (taxonomic) theoretical intent. Thus, for instance, the very meaning of some standard terms in epidemiology and psychometrics, such as “false positive” and “base rate,” which can be made tolerably clear on a taxonomic model, becomes fuzzy and—if the point is pressed—Â�hardly interpretable on a nontaxonomic model. An adequate understanding of the philosophical and statistical aspects of this in relation to substantive theories of causation might properly lead us to abandon the idea of rubrics entirely for some subsets of conditions. For example, when I used to teach clinical psychology, in order to make this point I sometimes pushed the following (doubtless exaggerated) doc-


trine: There are several major mental disorders (e.g., schizophrenia, manic–Â�depression, unipolar depression, delirium tremens, Alzheimer’s disease) that are truly taxonomic in nature, and for which category rubrics are semantically strictly appropriate, not merely as rough ways of delineating regions in a continuous descriptor space. There is also, in my opinion, a true entity of the solid-gold essential psychopath[y] (sociopathic personality, asocial [type], amoral type). But when we get to the so-Â�called neuroses and psychophysiological disorders of the neurotic kind, there is only one rubric (with the possible exception of the textbook obsessional neurosis)—namely, “psychoneurosis, mixed,” a term no longer found in the official nomenclature. The distinctions within that mixed category are quantitative only; they are merely differing degrees of anxiety, depression, somatization, and defense mechanisms in the neurotic mixture. In the long run, it may be worth the trouble to teach clinicians to think more dimensionally than categorically and mold their verbal and inferential habits in those directions. Third, we should get clearer than we presently are about the matter of sliding cuts on various indicators of an entity in relationship to base rates and various clinical populations in geographic, social classes, and the like, and the relevance of Bayes’s theorem. [See the MCMI (Millon, 1977) for a promising beginning in the rise of base rates in identifying personality disorders of different prevalences.] In matters where extremely asymmetrical likelihoods exist for the combination of a small number of high-valid signs, the importance of the base rate, except for the most extreme values, is considerably reduced, and it is probably statistical pedantry to push some kind of Bayes’s theorem algorithm onto working clinicians under such circ*mstances. I think that more mathematical analysis in relationship to the diagnostic habits of practitioners is in order here before altering the character of a psychiatrist’s or clinical psychologist’s education in this regard. Nobody acquainted with my writings would suspect me of being even faintly “antistatistical” in my biases; but I believe we should think like behavioral engineers in considering ourselves and others as clinical practitioners, taking into account what kinds of psychological disruptions in


diagnostic cognitive activity could take place that might reduce net efficiency, even though the underlying mathematical model makes it look like an improvement. Finally, at the risk of projecting my own current research interests, I would say that a desideratum for the next major revision is agreement upon the general taxometric problem as such, which I see as having two elements: (a) Is a taxometric procedure in psychopathology aimed at anything more than identifying phenotypic clusters; and, if it is, (b) which of the available formal taxometric methods (if any!) have shown themselves capable of detecting an underlying causal structure (whatever its biological or social nature), being meanwhile free of any appreciable tendency to detect taxonic structures that aren’t there (Meehl, 1979)? I think it not unduly optimistic to opine that we will have a pretty clear answer to the second question before the end of this decade (Grove & Andreasen, 1986; Meehl & Golden, 1982; Sneath & Sokal, 1973). Acknowledgment This chapter is reprinted from Meehl (1986, pp.€ 215–231). Copyright 1986 by The Guilford Press. Reprinted by permission.

References Allport, G. W. (1937). Personality: A psychological interpretation. New York: Henry Holt. Blois, M. S. (1980). Clinical judgment and computers. New England Journal of Medicine, 303, 192–197. Cattell, R. B. (1946). Description and measurement of personality. New York: World Book Company. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Cureton, E. E. (1957). Recipe for a cookbook. Psychological Bulletin, 54, 494–497. Dawes, R. M. (1979). The robust beauty of improper linear models in decision making. American Psychologist, 34, 571–582. Dawes, R. M., & Corrigan, B. (1974). Linear models in decision making. Psychological Bulletin, 81, 95–106.

Diagnostic Taxa as Open Concepts Engelhardt, T. H., Spicker, S. F., & Towers, B. (Eds.). (1979). Clinical judgment: A critical appraisal. Boston: Reidel. Feyerabend, P. K. (1970). Against method: Outline of an anarchistic theory of knowledge. In M. Radner & S. Winokur (Eds.), Minnesota studies in the philosophy of science: Vol. 4. Analysis of theories and methods of physics and psychology. Minneapolis: University of Minnesota Press. Finn, S. E. (1982). Base rates, utilities, and DSMIII: Shortcomings of fixed-rule systems of psychodiagnosis. Journal of Abnormal Psychology, 91, 294–302. Finn, S. E. (1983). Utility-Â�balanced and utilityÂ�imbalanced rules: Reply to Widiger. Journal of Abnormal Psychology, 92, 499–501. Goldberg, L. R. (1970). Man versus model of man: A rationale plus some evidence for a method of improving on clinical inferences. Psychological Bulletin, 73, 422–432. Goldberg, L. R. (1976). Man versus model of man: Just how conflicting is that evidence? Organizational Behavior and Human Performance, 16, 13–22. Gough, H. G. (1962). Clinical versus statistical prediction in psychology. In L. Postman (Ed.), Psychology in the making. New York: Knopf. Grove, W. M., & Andreasen, N. C. (1986). Multivariate statistical analysis in psychopathology. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology (pp.€153–168). New York: Guilford Press. Hoch, P., & Polatin, P. (1949). Pseudoneurotic forms of schizophrenia. Psychiatric Quarterly, 3, 248–276. Holt, R. R. (1970). Yet another look at clinical and statistical prediction. American Psychologist, 25, 337–339. Holt, R. R. (1978). Methods in clinical psychology: Vol. 2. Prediction and research. New York: Plenum Press. Kahneman, D., Slovic, P., & Tversky, A. (1982). Judgment under uncertainty: Heuristics and biases. London: Cambridge University Press. Kleinmuntz, B. (1982). Computational and noncomputational clinical information processing by computer. Behavioral Science, 27, 164–175. Meehl, P. E. (1954). Clinical versus statistical prediction: A theoretical analysis and a review of the evidence. Minneapolis: University of Minnesota Press. Meehl, P. E. (1956a). Clinical versus actuarial

185 prediction. In Proceedings of the 1955 Invitational Conference on Testing Problems (pp.€ 136–141). Princeton, NJ: Educational Testing Service. Meehl, P. E. (1956b). The tie that binds. Journal of Counseling Psychology, 3, 163–164. Meehl, P. E. (1956c). Wanted—a good cookbook. American Psychologist, 11, 263–272. Meehl, P. E. (1957). When shall we use our heads instead of the formula? Journal of Counseling Psychology, 4, 268–273. Meehl, P. E. (1959). Some ruminations on the validation of clinical procedures. Canadian Journal of Psychology, 13, 102–128. Meehl, P. E. (1960). The cognitive activity of the clinician. American Psychologist, 15, 19–27. Meehl, P. E. (1962). Schizotaxia, schizotypy, schizophrenia. American Psychologist, 17, 827–838. Meehl, P. E. (1964). Manual for use with checklist of schizotypic signs. Minneapolis: Psychiatry Research Unit, University of Minnesota Medical School. Meehl, P. E. (1967). What can the clinician do well? In D. N. Jackson & S. Messick (Eds.), Problems in human assessment (pp.€594–599). New York: McGraw-Hill. Meehl, P. E. (1972a). Reactions, reflections, Â�projections. In J. N. Butcher (Ed.), Objective personality assessment: Changing perspectives (pp.€ 131–189). New York: Academic Press. Meehl, P. E. (1972b). Specific genetic etiology, psychodynamics and therapeutic nihilism. International Journal of Mental Health, 1, 10–27. Meehl, P. E. (1973). Why I don’t go to case conferences. In P. E. Meehl, Psychodiagnosis: Selected papers. Minneapolis: University of Minnesota Press. Meehl, P. E (1974–1975). Genes and the unchangeable core. Voices: The Art and Science of Psychotherapy, 38, 25–35. Meehl, P. E. (1975). Hedonic capacity: Some conjectures. Bulletin of the Menninger Clinic, 39, 295–307. Meehl, P. E. (1977). Specific etiology and other forms of strong influence: Some quantitative meanings. Journal of Medicine and Philosophy, 2, 33–53. Meehl, P. E. (1979). A funny thing happened to us on the way to the latent entities. Journal of Personality Assessment, 43, 563–581. Meehl, P. E. (1983). Subjectivity in psychoanalyt-

186 ic inference: The nagging persistence of Wilhelm Fliess’s Achensee question. In J. Earman (Ed.), Minnesota studies in the philosophy of science: Vol. 10. Testing scientific theories. Minneapolis: University of Minnesota Press. Meehl, P. E. (1986). Diagnostic taxa as open concepts: Metatheoretical and statistical questions about reliability and construct validity in the grand strategy of nosological revision. In T. Millon & G. L. Klerman (Eds.), Contemporary directions in psychopathology (pp.€215– 231). New York: Guilford Press. Meehl, P. E., & Golden, R. (1982). Taxometric methods. In P. Kendall & J. Butcher (Eds.), Handbook of research methods in clinical psychology. New York: Wiley. Meehl, P. E., & Rosen, A. (1955). Antecedent probability and the efficiency of psychometric signs, patterns, or cutting scores. Psychological Bulletin, 52, 194–216. Millon, T. (1977). The Millon Clinical Multiaxial Inventory manual. Minneapolis, MN: National Computer Systems. Murphy, E. A. (1964). One cause? Many causes?: The argument from a bimodal distribution. Journal of Chronic Diseases, 17, 301–324.

CONCEPTUAL ISSUES IN CLASSIFICATION Murray, H. A. (1938). Explorations in personality. London: Oxford University Press. Pap, A. (1953). Reduction sentences and open concepts. Methodos, 5, 3–30. Pap, A. (1958). Semantics and necessary truth. New Haven, CT: Yale University Press. Peppard, T. A. (1949). Mistakes in diagnosis. Minnesota Medicine, 32, 510–511. Reichenbach, H. (1938). Experience and prediction. Chicago: University of Chicago Press. Rorer, L. G., Hoffman, F. J., LaForge, G. E., & Hsieh, K. E. (1966). Optimum cutting scores to discriminate groups of unequal size and variance. Journal of Applied Psychology, 50, 153–164. Sawyer, J. (1966). Measurement and prediction, clinical and statistical. Psychological Bulletin, 66, 178–200. Sines, J. O. (1970). Actuarial versus clinical prediction in psychopathology. British Journal of Psychiatry, 116, 129–144. Sneath, P. H. A., & Sokal, R. R. (1973). Numerical taxonomy. San Francisco: Freeman. Widiger, T. A. (1983). Utilities and fixed diagnostic rules: Comments on Finn (1982). Journal of Abnormal Psychology, 92, 495–498.

Chapter 9

Contemplations on Meehl (1986) The Territory, Paul’s Map, and Our Progress in Psychopathology Classification (or, the Challenge of Keeping Up with a Beacon 30 Years Ahead of the Field) Mark F. Lenzenweger

aul E. Meehl (1920–2003; see Figure 9.1) Pleading was recognized by many as perhaps the intellectual light in psychological

science of the past century. The breadth and depth of his interests, as well as his contributions to the scientific psychological corpus, remain peerless. The numerous tributes

Figure 9.1.╇ Paul E. Meehl, PhD (1920–2003). Photograph provided by and reproduced with permission of Leslie J. Yonce, PhD.

written in his honor attest to his impact, creativity, and generativity for psychology writ large, as well as for psychiatry, philosophy of science, and law (see the special tribute issues of the Journal of Clinical Psychology [2005, 61(10)], the Journal of Abnormal Psychology [2006, 115(2)], Applied and Preventive Psychology [2004, 11(1)], and Developmental Psychopathology [2003, 15(3)]). Simply stated, a Mozart comes along rarely. We were fortunate that Paul devoted his considerable intellectual energies to psychological science (particularly psychopathology) as he helped to chart the course for the discipline in a way that remains timeless. Thus, when Theodore Millon called to ask me to reflect on Paul’s contribution to the 1986 Contemporary Directions in Psychopathology volume, I hesitated and wondered, “What can be said that Paul Meehl did not already mention in the verdant original work?” Ted engaged in some gentle arm twisting and suggested I comment freely on the chapter. I reread and pondered the 1986 piece; I thought a bit more about Ted’s invitation; and here we are. One of the hallmarks of Meehl’s thinking was that it was typically well ahead of its time (and the field) by a good 30 years 187


or so. Consider just one example: his seminal model of schizotaxia, schizotypy, and schizophrenia, published in 1962. At a time when the field was still consumed with ideas regarding the psychosocial origins of schizophrenia (e.g., the schizophrenogenic mother; marital schism and skew) and was caught in its own double bind of sorts with competing psychoanalytic models of schizophrenia, Meehl proposed a genetically based, neurodevelopmental model of the origins and pathogenesis of schizophrenia. What is remarkable is that it took the field 30 years to realize that most of the major components in his model of the illness were quite on target, though debate still rages over the precise nature/structure of the genetic liability for the illness (mixed vs. oligogenic vs. latenttrait vs. multifactorial polygenic-Â�threshold models vs. genetic structural variants). For example, the neural basis of what Meehl (1962) called “synaptic slippage” is the focus of modern conceptualizations of the neural dysfunction in the illness, though it has been given different names by different theoreticians (e.g., “cognitive dysmetria,” “reduced developmental synaptic connectivity,” “intermittent degradation”). Paul liked to work in territory that was poorly mapped—or, if a map existed, he decidedly recharted it (e.g., clinical vs. statistical prediction, Meehl, 1954; construct validity, Cronbach & Meehl, 1955; see also Maher & Gottesman, 2005). His contribution to the 1986 volume had the same quality as these prior substantive efforts. The chapter was titled “Diagnostic Taxa as Open Concepts: Metatheoretical and Statistical Questions about Reliability and Construct Validity in the Grand Strategy of Nosological Revision” (it is reprinted as Chapter 8 of the present volume),1 and it was (and remains) a treasure trove of intellectual nuggets. In my view, Meehl’s (1986) chapter should be read and pondered by current (and aspiring) DSM architects if the classification business in psychopathology is ever to move forward appreciably in terms of validity and predictive power. 2 What I have done, therefore, in this chapter is to take Ted Millon at his word: I contemplate the issues raised by Meehl (1986) in light of where, in my view, we currently stand in the field—Â�explicating some issues in greater detail to underscore their meaning and importance, as well as point-


ing to areas where progress is still needed. In addition, I address the four “metaquestions” that Meehl left for us at the end of his 1986 chapter; these, he argued, were “big questions” that needed to be resolved (or at least “settled”) before the next big revision of the nomenclature. As many readers will realize, and as I discuss below, these metaquestions were certainly not settled prior to the publication of DSM-IV (American Psychiatric Association, 1994). But what of DSM-V and the four metaquestions? I would now like to reflect on a number of the points raised by Meehl (1986) in his chapter. I must confess that my survey consists of what grabbed my eye in the original piece; therefore, my commentary is necessarily selective and nonexhaustive in content and scope. I have also sought to highlight items that I believe may benefit contemporary psychiatry and clinical psychological science, particularly as nearly 25 years’ worth of PhDs and MDs have earned their degrees since the original publication appeared. One of the wonderful things about Meehl’s writing is that it needs to be read and reread because new insights and ideas emerge from each reading. This view is not idiosyncratic to yours truly, but has been the collective impression of long-time readers of Meehl’s works (see Waller, Yonce, Grove, Faust, & Lenzenweger, 2006). Let us begin with the reliability and validity issue that Meehl raised early in his chapter. He suggested that in general, the effort after reliability was a productive venture, in that it probably helped in most instances to increase the net construct validity of the disorders listed in the diagnostic manual (hereafter DSM). He spent some time explaining the complex relationships that can exist between reliability and validity, which bear restatement and examination here.

High Reliability Is Not Always Your Friend The reliability and validity issues, as well as the “attenuation paradox” (see Loevinger, 1954), were addressed by Meehl (1986); however, this part of the text was highly concentrated, and the profound implications of his remarks may have eluded many on a first reading. It is generally assumed that reliabil-

Contemplations on Meehl (1986)

ity is a good thing to seek in the area of diagnostic assessment. Although long known in the field of measurement and psychometrics, the value of reliability made itself known in the psychopathology diagnostic process after a series of studies in the early to mid-1960s highlighted difficulties with the execution of psychiatric diagnosis. Clearly things improved in terms of diagnostic reliability after the advent of DSM-III (American Psychiatric Association, 1980), with its explicit criteria and diagnostic guidelines. However, appreciable gains in the reliability of diagnosis since DSM-III-R (American Psychiatric Association, 1987) have been modest and limited to just a few areas (e.g., personality disorders). Meehl alerted readers to the situation whereby the validity of assessments may actually decrease if efforts at maximizing reliability in assessment or measurement are pushed too hard. Contemporary workers should continue to keep this admonition in mind, as it does happen in the real world, so to speak. For example, as the definition of schizophrenia was narrowed excessively to positive symptomatology in the DSM-III criteria set, one saw an important criterion of validity (i.e., familial schizophrenia) squeezed right out of the construct, as evidenced by two studies (Abrams & Taylor, 1983; Pope, Jonas, Cohen, & Lipinski, 1982). Moreover, we have seen that training raters of psychopathological phenomena to achieve high levels of reliability does not in any way guarantee validity. This was demonstrated by the clever study of Hooley and Richters (1991), which showed that the ratings of expressed-�emotion indicators by intelligent undergraduate raters (Harvard students) possessed high reliability, but little validity. The moral here is straightforward: Questing after high reliability may not always result in increases in validity. 3

The Notion of “Open Concepts” The notion of “open concepts” (Pap, 1953) ran through and through Meehl’s chapter; it was mentioned specifically over a dozen times. In short, if one got nothing else out of Meehl (1986) other than an appreciation for open concepts in relation to psychiatric diagnosis, then the chapter could be viewed as a pedagogical success by any measure. (See


Meehl, 1978, as well as his rich 1977 paper on specific etiology.)4 In his classic “two knights” paper of 1978 (the paper’s title refers to Sir Karl Popper and Sir Ronald Fisher), Meehl distinguished among three kinds of openness in constructs: (a) openness arising from indefinite extensibility of our provisional list of operational indicators of the construct; (b) openness associated with each indicator singly, because of the empirical fact that indicators are only probabilistically, rather than nomologically, linked to the inferred theoretical construct; and (c) openness due to the fact that most of our theoretical entities are introduced by an implicit or contextual definition, that is, by their role in the accepted nomological network, rather than their inner nature. (p.€815)

With respect to psychiatric diagnosis, both points (a) and (b) are rather central and critical. The joint meaning of the two kinds of openness with respect to diagnosis implies that (a) the list of indicators for a given disorder is necessarily incomplete and could be extended/expanded (thus a given criteria set is not definitive); and (b) the indicators of a disorder are linked to the underlying disease or disorder construct via stochastic or probabilistic relations (i.e., the signs and symptoms of a disorder are not directly linked to the underlying construct, but are merely [fallibly] associated with the construct). The third kind of openness as defined in (c) above speaks to our field’s lack of understanding of the inner nature of the constructs that we specify, but our need to embed them in our theories, models, and analytic approaches. Whether the topic of a theoretical discussion or an open circle in a path diagram depicting latent variables (which prompted Meehl to call this type of openness “Little Orphan Annie” eyes as suggested by the comic strip character of yesteryear). It is important to admit that we still have precious little understanding of the pathogenesis and pathology involved in most forms of psychopathology. This state of affairs is reflected, therefore, in our lack of a definitive understanding of the mode of action of most psychiatric medications. 5 Our primary constructs in psychopathology are defined in an external sense or by their implicit roles in our models; stated differently, the conceptual context in which


they are embedded defines them. Ultimately, our full understanding of constructs (gained through empirical research) will emerge, and the inner nature of such open concepts will be illuminated—in much the same way as the inner nature of the gene emerged with the discovery of DNA, although prior to this discovery the notion of a gene existing was highly plausible and helpful in explanatory networks.6

Don’t Let the DSM Rubrics Get€You Down Meehl always encouraged psychopathologists to think outside the proverbial box when it came to identifying meaningful parsings in the realm of mental disorder. Remember that this was the fellow who taught the field to conceive of latent constructs (e.g., love, anxiety, schizotypy, and perfectionism are all latent constructs) and construct validity at a time when most psychological scientists were only willing to think in terms of observable behaviors.7 Thus it should come as no surprise that he advised workers not to let themselves be excessively hampered by received rubrics in the search for order in psychopathological phenomenology. Experimental psychopathology and clinical psychological science, it seems, have been more comfortable exploring the manifestations of severe psychopathology that fall outside what appears in print on the pages of DSM. Such concepts as schizotypy (the liability for schizophrenia), psychopathy (as distinct from the sociodemographically defined antisocial personality disorder), and borderline personality organization (e.g., Kernberg’s [1984] phenomenological organizing framework for personality pathology) serve as only a few examples. Each of these theoretical concepts has spawned rich research literatures that have illuminated important aspects of more traditionally defined psychiatric conditions. For example, the laboratory study of schizotypy has generated a remarkable corpus of findings that support schizotypic psychopathology as an alternative manifestation of schizophrenia liability. To work outside the realm of DSM disorders in psychopathology research requires persistence and tenacity, however, as many reviewers and funding agencies only feel com-


fortable dealing with papers and proposals seeking to investigate “established” (if you will) entities. This is not completely unexpected, as review processes (including National Institute of Mental Health [NIMH] study sections) are characterized by an inherently conservative spirit and, according to some long-time observers, “groupthink” processes, which limit the adventurousness of all concerned. Along the lines of advocating that we not let established rubrics get us down, I would like to suggest to the psychopathology research community (which includes, of course, psychiatry and the DSM-V microcommunity) that we consider moving away from the 100-year-old “signs and symptoms” phenomenological approach to defining psychopathology, and move toward organizing our understanding of psychopathology in terms of neural circuits, neurobehavioral systems, and validated environmental stressors.8 In short, should we abandon the phenotypic-Â�indicator-based approach to diagnosis and talk about assessing neural systems? For example, would it make sense to move away from discussing the many varied anxiety disorders—which may or may not correspond to the way that nature is really organized—and speak of the neural circuits known to be related to anxiety and negative affect systems? Such progress would necessarily depend on advances in our understanding of the neural circuitry of the brain, as well as of the neurobiology of complex human behaviors, but this strikes me as an achievable goal. There are, of course, those who object to the “hobbyhorse” nature of much neuroimaging research, as well as those who argue cogently that present-day neuroimaging techniques are simply too coarse (resolution is limited; important neural events happen at the level of cells) and too slow (discrete mental events of greatest interest at the level of greatest interest are long past by the time the hemodynamic response catches up in the form of a blood-Â�oxygen-level-Â�dependent [BOLD] signal). Nonetheless, I think it would behoove the field to attempt a classification system based on the pathology of neural circuits and neurobehavioral processes, with relevant environmental inputs (e.g., exposure to tetrahydrocannabinol in those carrying a liability for schizophrenia). This is some-

Contemplations on Meehl (1986)

thing the DSM-V committee should actively consider encouraging as a “research system in need of further study.”

The “Context of Discovery” versus Anxiety in Bethesda Meehl strongly supported exploratory thinking, as well as empirical research carried out within the “context of discovery.” He advocated generating ideas from the armchair and probing one’s data well beyond a priori hypotheses, and he had enormous respect for the creative process in scientific research. So, one might ask, just what exactly is the “context of discovery”? Hans Reichenbach (1938) proposed that this context, as distinct from the “context of justification,” represented the part of the scientific creative process that was to some extent outside the purview of the logician and could be carried on in a highly nonlinear and emergent manner, offering few insights into its inner nature. Stated differently, a scientist can engage in all sorts of mental activities—Â�hypothesis generation; theory formulation; juxtaposition of constructs, processes, and temporal parameters in a divergent manner—in the search for knowledge in the context of discovery. The context of discovery does often, of course, result in testable and falsifiable conjectures (the work of the context of justification), but in Reichenbach’s view, the discovery context is the playground of intellectual ideas and creativity. He felt that this intellectual “free-for-all” should not be criticized, but rather should be viewed as an important part of the scientific enterprise. Meehl’s enthusiasm for the context of discovery reflects a certain spirit—a certain zeal for the exploratory and creative process in psychological science. In my view, contemporary psychopathology research, including efforts to improve the classification system, could use an infusion of this zeal for inquiry. Clearly, creative insights and proposals can be developed a priori, based entirely on theory or conceptual models, and/or they can be gleaned from post hoc analyses of existing data sets. Notions such as “expanded phenotypes” for certain forms of psychopathology (e.g., schizophrenia), “endophenotypes” (Gottesman & Gould, 2003), and deficits in smooth-Â�pursuit eye movements in schizo-


phrenia (Holzman, Proctor, & Hughes, 1973) have grown out of the rich matrix of the context of discovery, to name just three interesting advances consistent with Meehl’s enthusiasm for the context of discovery. At this juncture in considering the “context of discovery,” the issue of post hoc analysis and its inherent value requires explicit comment. Whereas many psychologists and psychiatrists have been taught that in empirical research one needs to formulate a hypothesis, gather data, run a statistical test, reject the null hypothesis, and write up the results, those of us who are experienced researchers know that this is not really how the process works. We do begin with ideas (we can even “conjecture from the armchair,” as Meehl was fond of saying); we do collect data; but then we must deeply probe the data that took considerable time, effort, and resources to collect. That considerable gold, in the form of findings, can be mined in post hoc analyses is well known; that manuscript and grant proposal reviewers are made anxious by post hoc analyses of data is also well known. The moral here should be self-Â�evident, and I believe Meehl would have endorsed this gentle admonition: Allow yourself the freedom to dive into the context of discovery in your work. Do not be hampered by the desire to hew to the line of NIMH interests in your applications,9 or by the publication preferences (biases) of journal editors when reporting on the results of post hoc analyses in the “Results” or “Discussion” sections of manuscripts. It is well known that most research grants are for the support of what Kuhn (1970) called “normal science,” the work of filling in the blanks and connecting dots of existing models. But thinking that potentially reflects paradigm shifts—the “big” leaps forward— is not likely to be received well (at least initially) by the scientific masses, or likely to be funded by conservative funding agencies. How can the spirit of Meehl’s enthusiasm for the “context of discovery” be built into the funding process—to fix, in part, what has become thought of as the “broken pipeline” (National Institutes of Health, 2008)? In the spirit of innovation, I would suggest that the NIMH should consider creating “context of discovery” grant awards, whereby worthy recipients would be identified and unrestricted awards would be made to them—Â�something


like a Career Development (K) Award without the requirement of concurrent R0-1 (or other mechanism) funding, or a Method to Extend Research in Time (MERIT)-Awardlike mechanism without the prerequisite of numerous years of prior NIMH involvement. Perhaps a blue-Â�ribbon panel of high-Â�impact researchers could serve to advise the NIMH about potential recipients of such “context of discovery” awards; perhaps the top 10 publications of a potential recipient would serve as the basis for evaluation (not total publications, not total number of prior awards). In regard to post hoc analysis, I think reviewers and journal editors should routinely ask investigators, “What interesting things did you learn in the post hoc analysis of your data? Would you consider including some description of these findings, with appropriate caveats (i.e., pending replication and so on), in your ‘Discussion’ section as food for thought for others?”

The Study of Subsyndromal (Fringe) Cases Just as we should not let the established DSM rubrics limit our research interests or sense of creativity, we should be sure to recruit subsyndromal cases of a disorder (construct) of interest for our studies, as well as to identify them in a data set in order to use the leverage provided by their group membership to advance knowledge. Meehl (1986) was clear to note the advantages associated with the enhanced attention to detail that came with DSM-III, and he suggested that there would probably be a net increase in construct validity associated with this increase in orderliness and explicit description (e.g., tightening up the diagnostic definition of schizophrenia would mean that the construct of schizophrenia would benefit in terms of increased construct validity, as noted above). That said, Meehl advocated the study of subsyndromal or “fringe” cases that inhabit the borders of established constructs. For example, consider the large number of individuals who inhabit the realm of schizotypy, but are not clinically psychotic or manifesting fulminant schizophrenic illness. Study of such individuals has genuinely advanced our understanding of schizophrenia as well as supplied an empirical basis for methodological


decisions, such as the inclusion of this group in genetic analyses focused on schizophrenia. The important points here are that (1) subsyndromal cases should be studied, and (2) they need to be identified in data sets. These points seem rather mundane, but they are nontrivial. If we do not study (i.e., collect data on such folks) and identify the subsyndromal cases in databases and, by implication, statistical analyses, then nothing can be learned about them. If a cell for them does not exist in an analysis of variance (or if it contains no subjects), it is not going to tell us much about phenomena that might be of great potential interest to us. Thus, if we are only working on DSM-defined constructs in our research, and subsyndromal variants are not assessed and included in analyses, we are likely to miss a valuable opportunity to advance our knowledge in psychopathology. An important corollary here is that the boundaries of most (if not all) psychopathology entities have some degree of fuzziness (some are fuzzier than others), and thus the issue of subsyndromal cases really concerns the boundaries of the constructs at issue in DSM or other nomenclatures.10

Just What Do You Believe about€the Constructs in DSM? One of the more engaging sections of Meehl’s (1986) chapter focused on what he described as possible intellectual views of the DSM constructs. He regarded one as “admirable,” one as “criticizable but fairly harmless,” and only the third as “scientifically malignant” (p.€220). Although I suspect that most research-Â�oriented psychiatrists and psychologists as well as other psychopathology researchers would not subscribe to Meehl’s third point of view on the DSM constructs, it is nonetheless possible to encounter some (perhaps more likely among practitioners) who do. DSM is a powerful professional–Â�social– political document. This claim does not speak to the validity of the document; rather, it notes how the document affects the conduct of research, research proposal development, the treatment of patients, and public health planning. Meehl (1986) was careful to note that the manner in which one regards the document can have profound effects on

Contemplations on Meehl (1986)

research and clinical progress, as well as the emergence of new ideas and themes for revision. My experience, based on many discussions with researchers, practitioners, and students over the years, suggests that Meehl’s analysis is worth repeating here. Thus Meehl argued that there are those who view the diagnostic constructs explicated in DSM as representing symptomatically defined syndromes that (1) hang together (descriptively or statistically), (2) have some unknown or unspecified etiology (which is likely to be shared by those diagnosed with the conditions), and (3) have some communicative value. This reveals a reasonable intellectual approach to the thorny problem represented by the classification and treatment challenges represented by psychopathology. There are two other viewpoints, speaking coarsely (there is probably some gradient across these viewpoints), that may be seen among those working with the DSM system. The more common one is reasonably harmless, although intellectually impoverished. The other is, as Meehl (1986) noted, “scientifically malignant” (p.€220)”; it is the view that somehow DSM articulates the truth, as it were. The view that is relatively harmless is that the constructs defined in DSM are somehow correct and the best we can do for now, with an implied openness to revision down the line. This view is OK, so to speak, just as long as those who think that DSM is the best we can do for now do not get in the way of those seeking to explore other entities or models of psychopathology. In other words, with respect to new roads and avenues of exploration in alternative approaches to psychopathology, adherents of this view cannot use this view as a basis for blocking progress. As Bob Dylan sings, “Your old road is rapidly agin’,/Please get out of the new one, if you can’t lend your hand” in “The Times They Are A-Changin’.” The mistaken and intellectually indefensible view that the DSM constructs represent the truth is indeed bad news. Not only can such a misguided view get in the way of progress; it is also grounded in a view of entities and operational definitions that has long been abandoned by informed philosophers of science and psychological science researchers. What is meant here? If someone believes that the meaning of the conditions in DSM is defined by the list of signs


and symptoms (or DSM-defined diagnostic rules), then one is implicitly subscribing to an “operationism” (so-Â�called operational definitions) that has long been discounted in philosophy and psychology. Conditions (or diseases) in traditional organic medicine are defined not merely by signs and symptoms; rather, they represent (implicitly) information regarding etiology, pathophysiology, and so on. If information about etiology and pathophysiology is absent from such constructs in organic medicine, then, at a minimum, the construct in question is viewed as in need of further research. In short, the constructs defined explicitly in DSM11 do not represent the truth, so to speak, and they do not represent the intellectual end of the line. For readers who might think that this is self-Â�evident stuff, I might only remind them of the frequent precursor to the modal question that follows many conference talks or colloquia—Â�namely, “According to DSM-IV, yadda, yadda, yadda.” I think that a routine rejoinder to such a preface to commentary after talks should be “Just what do you believe about the constructs in DSM?”

Densifications in the Multivariate Descriptor Hyperspace If one conceives of phenotypic (or endophenotypic) indicators as defining a multivariate hyperspace, this space may be characterized by a topography that resembles a uniform surface (imagine a blanket) with little to no variation or perhaps a steady gradient of variation, on the one hand; or by a variety of quasi-Â�independent clumpings (imagine satellites floating in space); or by all manner of possibilities in between. The clumpings may represent what Meehl (1986) referred to as “densifications.” Our work has yet to determine the precise nature of these densifications: Do they represent quantitative variation along continua, or dimensions, or genuine qualitative discontinuities? Meehl suggested that before the next big revision of DSM, the system’s architects should decide “when the orientation should be taxonomic versus dimensional.” This recommendation remains essentially as valid today as it was in 1986. Despite the flourishing taxometric literature and what to some seems to be the “bandwagon science” quality of taxometric


exploration, DSM remains fundamentally committed to a categorical model of psychopathology. Well-Â�designed taxonomic studies using taxometric methodology appropriately, as well as other approaches (e.g., finiteÂ�mixture modeling, latent-Â�growth-mixture modeling), remain sorely needed to generate the empirical data needed to satisfy the demand of Meehl’s recommendation (see Lenzenweger, 2004).

There Are No “Operational” Definitions in DSM Although I have alluded to this in note 11, it is worth stressing explicitly here that there are no “operational definitions” in DSM. There were none in DSM-III, DSM-III-R, DSM-IV, or DSM-IV-TR, and there will not be any in DSM-V. The manual does not specify the operations that one uses to make a diagnosis in any strict sense. Though we do not seek a set of Bridgman-Â�approved (Bridgman, 1927) operational definitions in DSM (as this would reflect adherence to an abandoned philosophical/scientific position), more guidance from the architects of DSM as to the actual diagnostic process would be useful. By this I do not mean simply narrative in the text that n of N symptoms are required for a given disorder. Rather, can more specification of the diagnostic operations to be followed be provided in future diagnostic systems? In this context, it is essential to note that the signs and symptoms of the disorders in DSM do not constitute the entities themselves. Meehl (1986) emphasized that such an intellectual error would only serve to propagate further confusion and/or dogmatism. As noted above, Meehl devoted considerable time to developing the notion of a “latent construct” (MacCorquodale & Meehl, 1948)—not surprisingly, since he “wrote the book,” as it were, on construct validity and the nomological net (Cronbach & Meehl, 1955; see also Hempel, 1952)—and contemporary psychopathologists may need to be reminded of the necessity of the latentÂ�construct formulation. To hypothesize the existence of a latent construct in psychopathology is not just some sort of high-flown intellectual endeavor or escapade; rather, it is an essential theoretical refinement for


understanding psychopathological entities. I realize that I am probably “preaching to the choir” for many readers, but the notion of a latent construct is still not well understood by many practicing psychologists and psychiatrists. Simply stated, this distinction (latent construct vs. observable indicators) is needed, and it promotes further substantive and statistical analysis: For example, factor analysis, taxometric analysis, and other multivariate techniques require the assumption of a latent entity or construct. Just as genetics and genomics have found the phenotype–Â� genotype distinction to be not only useful but a scientific necessity, so should psychopathology research more fully embrace the notion of latent constructs. In short, defining a disorder solely in terms of signs and symptoms is just really poor thinking. As Meehl said, “Such an operational definition is a fake” (p.€ 222). The moral here is straightforward: Everything in DSM is a hypothetical entity and thereby represented by a latent construct, and one should not conflate the signs/symptoms of a disorder with the underlying entity. One way to improve the validity of the DSM approach is to continue to encourage discovery of what will eventually come to be regarded as criteria of validity for a disorder, and such criteria of validity (à la Cronbach & Meehl, 1955; see also Maher & Gottesman, 2005) should come in the form of endophenotypes (Gottesman & Gould, 2003) and neural processes underpinning behavioral and/or psychological processes that are discovered through laboratory studies (cf., Lenzenweger & Hooley, 2003).

There Is Always a Role to Be Played by Clinical Observation Meehl’s (1986) chapter clearly had an air of quantitative sophistication, and the value of quantitative thinking was clearly implied throughout (see Waller et al., 2006). However, many people mistakenly think that Meehl only advocated a statistical approach to organizing information regarding psychopathology. Nothing could be further from the truth: He strongly advocated the value of clinical observation. His manual of schizotypic signs grew out of years of observations drawn from clinical practice; he maintained a psychotherapeutic prac-

Contemplations on Meehl (1986)

tice well into the later years of his career; and, believe it or not, he had a psychoanalytic couch in his office at the University of Minnesota. Clearly, the impressive observational skills of the clinician were not lost on Meehl. In this context, I should merely like to note that clinical observation remains a crucial vantage point for potential additions to or revisions of the DSM system. A caveat is needed here to avoid a potential misreading of the foregoing text. In short, we should not confuse how we collect data (e.g., clinical observation) with how we combine those collected data for the purposes of prediction or whatever. The value of clinical observation is not to be underestimated, especially in the age of advanced technological methods (e.g., neuroimaging) and complex and sophisticated statistical methods. No data collection method or statistical procedure or test is self-�interpreting; it cannot tell us where to look; it cannot resolve problems. We need only consider the classic blood pressure controversy (quantitative vs. qualitative variation) to gain an appreciation for the value of clinical observation in conjunction with complex statistical analyses of data (see Swales, 1985; cf. Lenzenweger, McLachlan, & Rubin, 2007).

Where Does Neuroimaging Fit€In? At the time Meehl (1986) wrote his original chapter, few would have envisioned the manner in which neuroimaging would have taken off as an approach to the study of human brain functioning. Not only were the early technologies of neuroimaging (computed tomography and positron emission tomography) only just beginning to be applied to psychopathology; there was little in the way of a guiding strategic approach for neuroimaging itself in the study of human brain function (e.g., “region of interest” approaches were essentially nonexistent). Thus Meehl did not really address this emerging technology. However, with the benefit of nearly 25 years’ progress, and with the clear-cut relevance of neuroimaging now well established in psychopathology research, we might wonder how the data derived from this powerful family of methodologies should be used in the explication of disorders in DSM. Not-


withstanding the concerns of some who view neuroimaging techniques as inherently too slow and too coarse to really get at the discrete neural events of greatest interest to psychopathologists, we might consider a bold (no pun intended) proposal: With the help of knowledge gained from neuroimaging studies, might it be possible to abandon the phenotypic (signs and symptoms) approach to diagnosis altogether (as implied by the proposal above regarding neural circuits as a basis for classification/description)? A shorter-term goal for neuroimagers might be to understand the basis of specific signs or symptoms (eschewing the need to organize these by categorical phenotypes in the DSM); by way of a longer-term goal, disturbances in established and well-�understood neural circuits (which cut across any number of DSM-defined constructs) would become the bases for classification. Eschewing the established disorder phenotypes in such work would be desirable, as the phenotypes are noisy and heterogeneous, particularly when there is minimal organization of the phenotypic indicators. Thus, as suggested previously, we might actually come to define pathologies in terms of disturbances in the reward circuitry, or the fear circuitry, or the nonaffective constraint system (see Depue & Lenzenweger, 2005; Epstein, Isenberg, Stern, & Silbersweig, 2002). I am fully aware that this would be a long-term strategic shift, especially given that to this day, DSM contains not a single biological criterion for use as an inclusion or exclusion indicator in the landscape of psychopathology.

The Weightless Environment of€DSM Space The world of DSM remains a “weightless” environment, and this deep-space quality only serves to limit the utility of the methodological approach embodied in it. The diagnostic criteria for the various disorders remain simply listed in a numbered format, and they are not weighted in terms of their diagnostic importance in any compelling manner. By retaining a fundamentally unweighted approach (in terms of either predictive power or what Dana and Dawes, 2004, would call simple “improper” linear modeling) to the organization of diagnostic


criteria, DSM only serves to propagate the definitions of disorders that will maintain or enhance heterogeneity across cases that presumably have the same form of psychopathology. For example, if one only needs any five of nine criteria for the diagnosis of borderline personality disorder (PD), then there are many, many (126, actually—order does not matter) ways for a patient to present as having borderline PD. We must ask: Is this the best way to define borderline PD? What does such heterogeneity do to the communication value of the diagnosis or the unit of analysis for research studies? Is this the best way to define and diagnose psychopathology in general? Does a one-Â�methodological“size” approach fit all? Related to the problem of an unweighted system is the manner in which the definitions of disorder contained within DSM may serve to structure relations among symptoms in a manner that limits (even precludes) studies seeking to illuminate a latent organization among the signs and symptoms that could advance classification efforts. For example, consider the diagnosis of schizophrenia. As it stands, even though the diagnostic criteria for the disorder are unweighted in any formal sense, it is essentially the case that a patient needs to display either hallucinations or delusions from the start to receive the diagnosis of schizophrenia. Why does this matter? It matters because the relations observed among the various phenotypic signs and symptoms of schizophrenia when derived from, say, the DSM-IV definition of schizophrenia will be conditioned on the presence of these positive symptoms. In terms of multivariate statistical procedures, the covariance matrix for the symptoms under investigation will bear the stamp of this conditioning. Thus all statistical studies of the latent organization of schizophrenia symptoms/signs will be affected by this implicit organization imposed by DSM. If one, therefore, wants to study the latent organization of phenomenology in schizophrenia, then one will need to be comfortable in departing from the narrative rules provided in DSM for the diagnosis of schizophrenia (see Lenzenweger & Dworkin, 1996). Not unrelated to the issue of the weightless nature of the DSM nomenclature is the problem that the criteria are still largely consistent only with, at best, what could be


termed “one-way pathognomonicity”—that is, the notion that the presence of a sign/ symptom suggests the presence of an illness. However, it should be stressed that the signs and symptoms listed in DSM often fall short of true one-way pathognomonicity; for example, the presence of “unstable interpersonal relations” does not unequivocally predict the presence of borderline PD. As far as I can tell, there is no symptom or sign in the DSM that has true “two-way pathognomonicity,” in that its presence indicates the presence of the disorder and (importantly) its absence indicates the absence of the disorder. If the diagnostic nomenclature is to move forward, it must begin to resolve these issues related to one-way and two-way pathognomonicity. It is important to note that even if one aspires to a “dimensional” basis for the definition of psychopathology, where levels and cutoff scores will be used to characterize pathology, one still needs to know whether putative cutoff scores or levels are associated with one-way or twoway pathognomonicity. For example, let us consider nonaffective constraint measured dimensionally, and ask ourselves whether or not specific levels on such a dimension suggest the presence or absence of pathology in either a one-way or two-way manner. The issue of one-way versus two-way pathognomonicity received considerable attention in Meehl’s thinking, and the interested reader is referred to his papers on the topic (see Waller et al., 2006).

The Pitfall of Heterogeneity Finally, there remains a crucial problem to be dealt with not only in the revision of definitions of disorders in the current diagnostic system, but also in the analysis of data gleaned from nearly any studies of DSM constructs in psychopathology research. This problem, simply stated, concerns heterogeneity—heterogeneity among patients owing to the unweighted, polythetic diagnostic system embodied in DSM, but also heterogeneity in performance on laboratory and other measures used in the study of psychopathology. Not all patients present the same symptom profile, owing to the nature of the rules of diagnosis in the DSM system, as well as the reality of polythetically

Contemplations on Meehl (1986)

defined constructs. How does heterogeneity affect the research enterprise? First, heterogeneity in the composition of patient samples, despite a common diagnostic label, serves to add noise to the unit of analysis. For example, patients with borderline PD drawn from one setting may be quite different from patients with borderline PD drawn from another setting, despite both samples being described as suffering from borderline PD (Korfine & Hooley, 2009). Such variation across patient samples is clearly going to be a source of heterogeneity in performance on measures in laboratory and other studies. However, the heterogeneity issue is not merely a function of variations in symptom profiles across groups of patients sharing the same diagnosis. Even within relatively well-�characterized subject samples, where the inclusion criteria have been set up so that all subjects under study are relatively hom*ogeneous, one also sees considerable heterogeneity of performance on laboratory measures. For example, in my own laboratory, where we have defined schizotypy in a consistent manner across all subjects, we find that (1) not all schizotypic individuals are deviant on all laboratory tasks and, moreover, (2) different subjects show different degrees of deviance on different measures (Lenzenweger, 1998). For example, one schizotypic subject may show deviance on eye tracking, working memory, and negative priming tasks, but not on other measures in a protocol; another subject may display thought disorder and sustained attentional deficits, but may not reveal deficits on other measures in the same protocol. The issue of heterogeneity looms large for psychopathology research, and resolution of this issue represents a research challenge that must be embraced if progress is to be made. The interested reader is referred to Maher (2003) and Lenzenweger, Jensen, and Rubin (2003); an older reference, but a wonderful one, with reference to the resolution of heterogeneity is Blashfield (1984).

Let’s Consider Meehl’s Four Metaquestions Meehl ended the 1986 chapter with what he termed “four metaquestions,” which he suggested represented crucial issues in need of


resolution before any large-scale revision of the diagnostic nomenclature was undertaken. Needless to say, these four metaquestions were not resolved prior to the publication of the DSM-III-R, DSM-IV, and DSM-IV-TR. Despite the best-Â�seller status of all these versions of DSM, the nomenclature embodied in them remains in need of improvement along the lines discussed above, as well as in terms of Meehl’s metaquestions. Although the committee wheels for DSM-V are turning already, let us consider where we are with respect to the metaquestions from 1986. Conjectural Etiology “First, what role should a conjectural etiology, when moderately to strongly corroborated, play in the taxonomic strategy?” (Meehl, 1986, p.€227). At this point, it appears that conjectural etiology still plays essentially no role in the taxonomic enterprise as embodied in the DSM approach. By this I mean that corroborated causal factors have not found their way into the diagnostic system in any genuine manner. Thus the information garnered from years of genetic research in schizophrenia, trauma research in borderline PD, and so on does not play a role in the definition of these disorders or in the diagnostic criteria. The “endophenotype” concept (Gottesman & Gould, 2003), long known in experimental psychopathology (Gottesman & Shields, 1972; Lenzenweger & Loranger, 1989), has begun to be picked up by many researchers and may eventually serve to organize information that can be included in a future diagnostic nomenclature. However, endophenotype information has not been included in past versions of the nomenclature—and, to the best of my knowledge, will not be included as a basis for diagnostic decisions in the forthcoming DSM-V. We have been through various “decades of this and that” research foci; yet it seems we have miles to go before genetic, neural, neurobehavioral, or other brain-based indices are included in a taxonomic scheme. As a senior observer of the field once said to me, “The brain isn’t just a place holder in the skull. It is high time we understand it and take that information into the diagnostic system.” Some of the slowness in the incorporation of etiological factors into the


diagnostic scheme may be due to the fact that psychiatry was slow to advance to a research basis, after decades of fascination with psychoanalysis. Moreover, despite initial lightning strikes of luck in the psychopharmacological arena from the early 1950s to the early 1960s (e.g., imipramine, a depression treatment, worked well for panic; chlorpromazine, a medication for operative and anesthetic shock, worked well for psychosis; lithium, a medication for gout, worked well for mania), psychiatry was very slow to advance to the laboratory. As noted by Epstein and colleagues (2002), psychiatry lacked the tools, both conceptual and methodological (especially statistical), for this advance. I would like to make it clear that with respect to the linkage of the brain with the grand traditions of psychological testing and assessment, clinical psychology has been similarly slow to move into the laboratory as well. The subdiscipline of experimental psychopathology (see Lenzenweger & Hooley, 2003) has been a force in moving both psychiatry and clinical psychology into the laboratory. Let us hope that information gathered from the laboratory regarding conjectural etiology and pathology will eventually find its way into future taxonomic systems. Traits versus Types “Second, the strategic distinction between thinking in terms of dimensions and categories (types, species, taxa, disease entities) remains with us” (Meehl, 1986, p.€ 228). Those who are familiar with Meehl’s thinking, especially as it relates to the development of coherent cut kinetics (taxometric analysis), will quickly recognize that this was a major interest of his—Â�namely, distinguishing taxa from quantitative dimensions. In short, the answer to the question “Is it quantitatively or qualitatively structured at the latent level?” remains unanswered for the most part, although progress has been made with respect to methods (see Lenzenweger, 2004; Lenzenweger et al., 2007; Waller & Meehl, 1998). For example, it is now generally recognized that schizotypy, or the underlying liability for schizophrenia, is probably structured in a qualitative fashion (see Lenzenweger & Korfine, 1992; Lenzenweger et al., 2007). Whether the qualitative


discontinuity reflects a complete discontinuity in schizophrenia liability or a severe step function remains an area of active study (see Lenzenweger et al., 2007). Other areas of inquiry about latent structure are active and generating new findings rapidly, such as latent-Â�structure studies of depression, posttraumatic stress disorder, psychopathy, and so on. Thus, although this second metaquestion is not resolved, methods exist for its resolution and we are making progress. In this context, it is essential to note that this issue, as properly conceived, has addressed itself to the underlying structure of nature and not to the 1980s hassle about how to assess psychopathology, especially PDs. Rather, Meehl’s metaquestion concerned which form of psychopathology had a quantitative latent structure versus a taxonic (categorical) latent structure. Meehl (1995) made it clear that this profound issue could not be settled by methods and viewpoints reflective of guild preferences (i.e., psychologists are trained to think dimensionally, and psychiatrists are trained to think categorically). Rather, it would require considerable time and effort; careful theoretical analysis; and dedicated taxonomic research using proper methods, such as taxometric analysis (Waller & Meehl, 1998),12 latent-class analysis, and finite-Â�mixture modeling (McLachlan & Peel, 2000; see also Lenzenweger et al., 2007; Lenzenweger, Clarkin, Yeomans, Kernberg, & Levy, 2008), among others. Along these lines it is important to note that psychiatrists (particularly the architects of DSM-V) have begun to discover the utility of a dimensional approach to measurement, as well as the methodological tools needed to explore such an approach (largely factor analysis). Thus, after years of being psychoanalyzed, psychiatry is now likely to go through an extended factor analysis! Indeed, the research agenda for DSM-V has been defined in part as a dimensional enterprise (Helzer et al., 2008). The enthusiasm for factor analysis as a panacea for the classification problems we face, however, must be tempered, and the early experience of psychology with the technique provides a bracing whiff of reality (i.e., despite initial enthusiasm for the power of factor analysis, many in the field now regard the methods as simply data reduction techniques). I would merely urge caution in the adoption of a

Contemplations on Meehl (1986)

wholesale factor-Â�analysis-based, dimensional approach to the measurement and classification of psychopathology. I cannot help having a sense of déjà vu when pondering the current fascination with dimensional models for psychopathology (especially the internalizing vs. externalizing scheme that has been with us since the 1960s), bearing in mind the extensive corpus of such prior efforts reviewed by Blashfield (1984). I would recommend a balanced approach as we move forward, incorporating both dimensional and taxometric approaches to the resolution of the dimensions versus types issue in relation to psychopathology. I say this, although I am mindful that the volume A Research Agenda for DSM-V (Kupfer, First, & Regier, 2002) does not even include the word “taxometric,” “construct validity” appears once, and “endophenotype” appears only three times. In this context, this issue of traits versus taxa and their relative frequency in psychopathology deserves comment. I have frequently been asked, as one who has done taxometric research and reviewed dozens of taxometric manuscripts over the past 20 years, “How many taxonic entities will be found in psychopathology?” My answer has typically been “Who knows?” My response is not guided by cynicism or the desire to be a wise guy. I simply do not know how many taxa are out there in psychopathology. I should note that I normally want to have a good substantive model in play—one that hypothesizes the existence of a latent taxon in a particular form of psychopathology. Frankly, not that many models or theories argue for the existence of latent, qualitative classes a priori in psychopathology. Is this merely my view? An informal polling of and discussions with well-Â�informed taxometricians suggest that many do not believe that numerous taxa will be discovered.13 In this context, it is just as important to point out that very few truly (genuine) dimensional entities will be found. Few data sets in the real world conform faithfully to a factor model; thus the notion of a dimensional model of psychopathology is something of a Platonic ideal (just as some think the notion of taxa represents an ideal). The reality is that most data sets in psychopathology, when analyzed carefully, will produce results regarding latent structure that will fall between dimensional and taxonic ideals. Moreover, within


the dimensional framework, it is important to realize that even though data can be treated as if they are dimensional and may be factor-Â�analyzable, this provides no assurance that there still might be interesting things going on out in the tails of putative continua that would be of interest to classificationists and taxometricians. Finally, it is important for all who want to push the “factor analysis” (or, more typically, “principal components”) button on their statistical software program to realize that factors (continua) will always emerge from factor analysis. But this statistical reality does not mean that nature is truly dimensional at the latent level. The Reverend Bayes, Base Rates, and Cutoff Scores “Third, we should get clearer than we presently are about the matter of sliding cuts on various indicators of an entity in relation to base rates and various clinical populations in geographic, social classes, and the like, and the relevance of Bayes’s theorem” (Meehl, 1986, p.€ 229). Upon rereading Meehl’s chapter nearly 25 years later, I cannot help saying, upon reflecting on the state of psychiatric diagnosis and classification, “Yes, Paul, you are right; you were right in 1986, and you are right today.” It is clearly the case that the underlying mathematical basis for taking base rates and sliding cuts into account in the classification problem is well grounded, is based on a firm statistical foundation, and has a true elegance that should be considered carefully in the context of nosological revision (as well as everyday practice). One of Meehl’s concerns, however, was not whether or not taking base rates and the like into account was relevant, but whether it was worth the time and effort. Would it make sense to push Bayes’s theorem into the hearts and minds of diagnosticians of all sorts, in research as well as clinical settings (the latter was a true concern of Meehl’s)? In short, he wanted the adjustment of the nomenclature and diagnostic practice, based on a consideration of base rates and sliding cuts, to be an effective and validity-Â�enhancing enterprise. In 1986, Meehl felt that considerable work needed to be done on this issue before we went ahead and began to overhaul medical and clinical


science training to reflect such an approach in the diagnostic process. In short, that work is still needed, and this metaquestion is far from resolved (early 1980s discussion of this issue notwithstanding). The Taxonic Question and€Taxometric Problem Finally, at the risk of projecting my own current research interests, I would say that a desideratum for the next major revision is agreement upon the general taxometric problem as such, which I see as having two elements: (a) Is a taxometric procedure in psychopathology aimed at anything more than identifying phenotypic clusters; and, if it is, (b) which of the available formal taxometric methods (if any!) have shown themselves capable of detecting an underlying causal structure (whatever its biological or social nature), being meanwhile free of any appreciable tendency to detect taxonic structures that aren’t there? (Meehl, 1986, p.€229)

This is an area where activity continues apace. The taxometric and related latentÂ�structure work in schizotypy (Lenzenweger & Korfine, 1992; Lenzenweger et al., 2007), for example, has shown that endophenotypes for schizotypy (i.e., schizophrenia liability) have a qualitative latent structure—Â� transcending the realm of phenotypic indicators reflected in signs and symptoms. Importantly, this latent structure has been found to be associated with important criteria of validity, such as increased rates of familial schizophrenia among the members of an identified schizotypy-Â�positive class (without, importantly, an abundance of other forms of psychopathology in the same family members) (Lenzenweger et al., 2007). As regards the concern expressed in the second part of Meehl’s taxometric metaquestion—the identification of methods that do not falsely identify latent classes—this also remains an area of active inquiry. The methodological morals guiding the field of taxometric research have received scrutiny (Lenzenweger, 2004), and alterations of the basic taxometric research strategy claiming to address this issue have been appropriately scrutinized and greeted with caution (Beach, Amir, & Bau, 2005). What must be stressed in this discussion of taxometric analysis is that if it is to serve a useful purpose, it should


be conducted in a manner that conforms faithfully to Meehl’s description, especially in the ample use of consistency tests. It is remarkable how many published taxometric reports have not conducted consistency tests in a manner consistent with Meehl’s writings and intentions. It is worth revisiting Meehl’s (1995) recommendations on the necessity of consistency tests: “It is a mistake to think of consistency tests in bootstrap taxometrics as a sort of luxury, pleasantly reassuring if one is so lucky as to get it. Consistency tests are an absolute necessity in bootstrap taxometrics, and the more available the better” (p.€272).14 Consistent with a central theme of this chapter, it would be particularly useful if future taxometric work in psychopathology focused more on laboratory-Â�assessed, ratioscale data, particularly within the domain of endophenotypes (i.e., if it moved away from the taxometric analysis of phenotype-level values) (see Lenzenweger, 2004).

Coda Paul E. Meehl was a thinker who walked many trails in the intellectual high country of the philosophy of science, psychopathology, assessment and prediction, schizophrenia, taxometrics, and (yes) psychoanalysis and beyond (see Waller et al., 2006). The breadth of his thinking, as well as its rigor, continues to be a source of inspiration for many and still has much to offer current and would-be architects of the official diagnostic nomenclature (i.e., DSM-V and beyond). Unfortunately, one still encounters relatively advanced researchers in psychiatry (and, to a lesser extent, clinical psychology) who are fundamentally unfamiliar with many of Meehl’s central contributions that have shaped the field of psychopathology and, if understood and applied, will continue to help the field advance (Maher & Gottesman, 2005). This suggests to me that there are still intellectual and professional (guild) walls in psychopathology that need to come down for the field to advance in the nosological enterprise—and, importantly, to fill in the open constructs. The song “Pancho and Lefty by the late Townes Van Zandt includes the lines “He wore his gun outside his pants,/for all the honest world to feel.” Paul Meehl wore his scientific values and propos-

Contemplations on Meehl (1986)

als for an improved science of psychopathology very openly; he left no doubt where he stood on most important substantive and methodological issues. In doing so, he left us an excellent map in the 1986 chapter and elsewhere, as well as an important scientific attitude. We need to use these contributions as a basis for continuing to probe the territory of psychopathology—to better organize our classification system, discern the latent structure of psychopathology, and move in the direction of illuminating specific etiology. Acknowledgments I thank Irving I. Gottesman, Brendan A. Maher, Lauren Korfine, David A. Silbersweig, Niels G. Waller, and Theodore P. Beauchaine for useful discussions during the preparation of this chapter. I also thank Niels G. Waller for useful comments on an earlier version of the chapter. Preparation of this chapter was supported in part by sabbatical leave funds provided by the State University of New York at Binghamton.

Note â•⁄ 1. Meehl (1986) is also available as a PDF download (see pubs.htm). â•⁄ 2. A colleague, who is by any measure a worldclass psychopathologist, suggested that all members of the DSM-V revision groups should receive a copy of Meehl’s (1986) chapter, should be told to read it, and subsequently should be quizzed on its contents before making any changes in the current nomenclature. â•⁄ 3. Given that some of the constructs in the DSM system are defined in a very narrow fashion, one must also be alert to the possibility that measures associated with these constructs are at risk for being “bloated specific” measures (Cattell, 1966). Such measures often lack theoretical utility, are rarely linked to broader and more established measures of substantive interest, and have only limited predictive power. â•⁄ 4. Both the 1977 and 1978 Meehl papers can be found in the recently published A Paul Meehl Reader (Waller et al., 2006). The Meehl (1978) paper can also be downloaded from

201 the Meehl website, which is maintained by Leslie J. Yonce ( pubs.htm). â•⁄ 5. One need only read the boxed warning medication inserts for many psychiatric medications to come across the ubiquitous statement “The mechanism of action of [drug name], as with other drugs having efficacy in [name of disorder], is unknown.” Statements such as these represent the third type of openness referred to by Meehl (1978). â•⁄ 6. A concept that Meehl frequently discussed in connection with open concepts in psychopathology was “specific etiology” (Meehl, 1977). Although this concept was not developed in any great detail in the 1986 chapter, the reader is referred to Meehl (1977) for a rich discussion of this important concept, as well as to those papers in Waller and colleagues (2006) that address specific etiology. â•⁄ 7. The notion of latent constructs and their utility remains a conceptual issue that has been difficult for psychiatry to grasp fully and embrace (it is less so in psychology). â•⁄ 8. I do not wish to suggest that a “signs and symptoms” approach has no value whatsoever. In fact, I think that a considerable amount of work still needs to be done on parsing the psychopathology domain to improve phenotypic-Â�indicator-based classification systems. Rigorous and well-Â�conducted multivariate work still, in my view, has much to offer the “signs and symptoms” approach. â•⁄ 9. A well-known psychiatrist/researcher in schizophrenia noted in a keynote address at a meeting of the American Psychopathological Association a few years ago that the NIMH tends to be about 10 years behind the curve when it comes to supporting innovative thinking. Although I have no way of knowing whether this is a precise estimate (and one can imagine that such a suggestion would surely not sit well in Bethesda), the spirit of the comment, made in the presence of numerous top-notch researchers, is well worth considering. 10. One could easily extend this line of thinking to the issue of quantitative variation and the perennial question of “where to drop the cut” on a dimension of interest. Moving the cutoff score around on a dimension of interest is one straightforward way of studying variation at a boundary or border. This issue could be the focus of an article in its own right and is not developed further here.

202 11. It is essential to note that there are no “operational definitions” in DSM. That is, the operations by which the disorders are related to the signs and symptoms defining the disorder are not specified; nor are the operations needed to make the diagnosis specified in any strict sense of the term “operational” (see the following text section). 12. Given that many students are likely to read this volume, I think it important to stress that Waller and Meehl (1998) should serve as the foundational reference for taxometric method in addition to the original papers written specifically on the topic. Many “how-to” taxometric books are beginning to appear on the scene, and Waller and Meehl remains the classic and error-free exposition on taxometrics. 13. I have heard some editors and editorial board members at reputable journals say that those who conduct taxometric analyses are “committed to a taxonic model” of psychopathology. This is rather crude thinking, and simply quite far from the truth in my experience. In my observations, particularly in the schizophrenia and PD research domains (where I have my primary research and clinical experience), I believe that the energetic pursuit of dimensional models in psychopathology has far outstripped what appears to be the case for taxonic models. Consider that over 500 citations for the keywords “dimensional and personality disorders” emerge from a PsycInfo search, whereas 19 citations for the keywords “taxonic or taxa and personality disorders” emerge for the same time period. 14. My colleague Niels G. Waller, at Minnesota, tells me that there are at least 35 identifiable consistency tests for use in taxometric analysis. However, the vast majority of taxometric reports do not really follow Meehl’s (1995) stern guidance on the use of consistency tests in their taxometric research efforts.

References Abrams, R., & Taylor, M. A. (1983). The genetics of schizophrenia: A reassessment using modern criteria. American Journal of Psychiatry, 140, 171–175. American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association. (1987). Diag-

CONCEPTUAL ISSUES IN CLASSIFICATION nostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Beach, S. R. H., Amir, N., & Bau, J. (2005). Can sample specific simulations help detect low base rate taxonicity? Psychological Assessment, 17, 446–461. Blashfield, R. K. (1984). Classification of psychopathology: Neo-Â�Kraepelinian and quantitative approaches. New York: Plenum Press. Bridgman, P. W. (1927). The logic of modern physics. New York: Macmillan. Cattell, R. B. (1966). Psychological theory and scientific method. In R. B. Cattell (Ed.), Handbook of multivariate experimental psychology (pp.€1–18). Chicago: Rand McNally. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Dana, J., & Dawes, R. M. (2004). The superiority of simple alternatives to regression for social science predictions. Journal of Educational and Behavioral Statistics, 29, 317–331. Depue, R. A., & Lenzenweger, M. F. (2005). A neurobehavioral model of personality disturbance. In M. F. Lenzenweger & J. F. Clarkin (Eds.), Major theories of personality disorder (2nd ed., pp.€ 391–453). New York: Guilford Press. Epstein, J., Isenberg, N., Stern, E., & Silbersweig, D. (2002). Toward a neuroanatomical understanding of psychiatric illness: The role of functional imaging. In J. E. Helzer & J. J. Hudziak (Eds.), Defining psychopathology in the 21st century: DSM-V and beyond (pp.€57– 70). Washington, DC: American Psychiatric Press. Gottesman, I. I., & Gould, T. D. (2003). The endophenotype concept in psychiatry: Etymology and strategic intentions. American Journal of Psychiatry, 160, 636–645. Gottesman, I. I., & Shields, J. (1972). Schizophrenia and genetics: A twin study vantage point. New York: Academic Press. Helzer, J. E., Kraemer, H. C., Krueger, R. F., Wittchen, H.-W., Sirovatka, M. S., & Regier, D. A. (Eds.). (2008). Dimensional approaches in diagnostic classification: Refining the research agenda for DSM-V. Arlington, VA: American Psychiatric Association. Hempel, C. G. (1952). Fundamentals of concept formation in empirical science. Chicago: University of Chicago Press.

Contemplations on Meehl (1986) Holzman, P. S., Proctor, L. R., & Hughes, D. W. (1973). Eye-Â�tracking patterns in schizophrenia. Science, 181, 179–181. Hooley, J. M., & Richters, J. E. (1991). Alternative measures of expressed emotion: A methodological and cautionary note. Journal of Abnormal Psychology, 100, 94–97. Kernberg, O. F. (1984). Severe personality disorders. New Haven, CT: Yale University Press. Korfine, L., & Hooley, J. M. (2009). Detecting individuals with borderline personality disorder in the community: An ascertainment strategy and comparison with a hospital sample. Journal of Personality Disorders, 23, 62–75. Kuhn, T. S. (1970). The structure of scientific revolutions (2nd ed.). Chicago: University of Chicago Press. Kupfer, D. J., First, M. B., & Regier, D. A. (Eds.). (2002). A research agenda for DSM-V. Washington, DC: American Psychiatric Association. Lenzenweger, M. F. (1998). Schizotypy and schizotypic psychopathology: Mapping an alternative expression of schizophrenia liability. In M. F. Lenzenweger & R. H. Dworkin (Eds.), Origins and development of schizophrenia: Advances in experimental psychopathology (pp.€93–121). Washington, DC: American Psychological Association. Lenzenweger, M. F. (2004). Consideration of the challenges, complications, and pitfalls of taxometric analysis. Journal of Abnormal Psychology, 113, 10–23. Lenzenweger, M. F., Clarkin, J. F., Yeomans, F. E., Kernberg, O. F., & Levy, K. N. (2008). Refining the borderline personality disorder phenotype through finite mixture modeling: Implications for classification. Journal of Personality Disorders, 22, 313–331. Lenzenweger, M. F., & Dworkin, R. H. (1996). The dimensions of schizophrenia phenomenology?: Not one or not two, at least three, perhaps four. British Journal of Psychiatry, 168, 432–440. Lenzenweger, M. F., & Hooley, J. M. (Eds.). (2003). Principles of experimental psychopathology: Essays in honor of Brendan A. Maher. Washington, DC: American Psychological Association. Lenzenweger, M. F., Jensen, S., & Rubin, D. B. (2003). Finding the “genuine” schizotype: A model and method for resolving heterogeneity in performance on laboratory measures in experimental psychopathology research. Journal of Abnormal Psychology, 112, 457–468.

203 Lenzenweger, M. F., & Korfine, L. (1992). Confirming the latent structure and base rate of schizotypy: A taxometric analysis. Journal of Abnormal Psychology, 101, 567–571. Lenzenweger, M. F., & Loranger, A. W. (1989). Detection of familial schizophrenia using a psychometric measure of schizotypy. Archives of General Psychiatry, 46, 902–907. Lenzenweger, M. F., McLachlan, G., & Rubin, D. B. (2007). Resolving the latent structure of schizophrenia endophenotypes using expectation-Â�maximization-based finite mixture modeling. Journal of Abnormal Psychology, 116, 16–29. Loevinger, J. (1954). The attenuation paradox in test theory. Psychological Bulletin, 51, 493– 504. MacCorquodale, K., & Meehl, P. E. (1948). On a distinction between hypothetical constructs and intervening variables. Psychological Review, 55, 95–107. Maher, B. A. (2003). Psychopathology and delusions: Reflections on methods and models. In M. F. Lenzenweger & J. M. Hooley (Eds.), Principles of experimental psychopathology: Essays in honor of Brendan A. Maher (pp.€ 9–28). Washington, DC: American Psychological. Maher, B. A., & Gottesman, I. I. (2005). Deconstructing, reconstructing, and preserving Paul E. Meehl’s legacy of construct validity. Psychological Assessment, 17, 415–422. McLachlan, G., & Peel, D. (2000). Finite mixture models. New York: Wiley. Meehl, P. E. (1954). Clinical versus statistical prediction: A theoretical analysis and a review of the evidence. Minneapolis: University of Minnesota Press. Meehl, P. E. (1962). Schizotaxia, schizotypy, schizophrenia. American Psychologist, 17, 827–838. Meehl, P. E. (1977). Specific etiology and other forms of strong influence: Some quantitative meanings. Journal of Medicine and Philosophy, 2, 33–53. Meehl, P. E. (1978). Theoretical risks and tabular asterisks: Sir Karl, Sir Ronald, and the slow progress of soft psychology. Journal of Consulting and Clinical Psychology, 46, 806– 834. Meehl, P. E. (1986). Diagnostic taxa as open concepts: Metatheoretical and statistical questions about reliability and construct validity in the grand strategy of nosological revision. In T. Millon & G. L. Klerman (Eds.), Contempo-

204 rary directions in psychopathology (pp.€215– 231). New York: Guilford Press. Meehl, P. E. (1995). Bootstraps taxometrics: Solving the classification problem in psychopathology. American Psychologist, 50, 266–275. National Institutes of Health. (2008). A broken pipeline?: Flat funding of the NIH puts a generation of science at risk. Washington, DC: Author. Pap, A. (1953). Reduction sentences and open concepts. Methodos, 5, 3–30. Pope, H. G., Jonas, J. M., Cohen, B. M., & Lipinski, J. F. (1982). Failure to find evidence of schizophrenia in first-Â�degree relative of schizophrenic probands. American Journal of Psychiatry, 139, 826–828.

CONCEPTUAL ISSUES IN CLASSIFICATION Reichenbach, H. (1938). Experience and prediction: An analysis of the foundations and the structure of knowledge. Chicago: University of Chicago Press. Swales, J. D. (1985). Platt versus Pickering: An episode in recent medical history. London: Keynes Press. Waller, N. G., & Meehl, P. E. (1998). Multivariate taxometric procedures: Distinguishing types from continua. Newbury Park, CA: Sage. Waller, N. G., Yonce, L. J., Grove, W. M., Faust, D. A., & Lenzenweger, M. F. (2006). A Paul Meehl reader: Essays on the practice of scientific psychology. Mahwah, NJ: Erlbaum.

C hap te r 10

Issues of Construct Validity in€Psychological€Diagnoses Gregory T. Smith Jessica Combs

ver the last century, there has been O enormous progress in “validation theory,” which (as its name indicates) describes

how to validate psychological theories and the measures of constructs described in those theories (Kane, 2001; Smith, 2005; Smith & Zapolski, 2009). There has also been remarkable progress in researchers’ understanding of psychopathology and in their ability to describe it cogently, reliably, and validly, as reflected in the contents of this volume. Progress in these two areas has been closely linked. As psychologists learned how to validate measures of unobservable constructs, and theories relating constructs to one another, it became possible to define and test progressively more sophisticated theories of psychopathology. At the same time, advances in understanding psychopathology probably spurred the advances in validation theory: Cronbach and Meehl’s (1955) classic introduction of construct validity followed considerable pressure to develop sound means to test theories, rather than just to predict criteria. The same process of reciprocal influence between the two fields is operative today.

We believe that this process has produced recent advances in both psychopathology theory and validation theory that are profoundly important. Specifically, validation theorists have come to recognize that single scores that combine multiple dimensions of functioning necessarily lack clear theoretical meaning (Edwards, 2001; McGrath, 2005; Smith, Fischer, & Fister, 2003; Smith, McCarthy, & Zapolski, 2009). They have concluded that unidimensional traits and unidimensional symptoms are the proper objects of theory and measure validation. Virtually simultaneously, psychopathology researchers have been applying advanced statistical techniques to disaggregate psychological disorders that have long been recognized as heterogeneous (American Psychiatric Association, 1994), with the goal of identifying hom*ogeneous dimensions of dysfunction (see Brinkley, Newman, Widiger, & Lynam, 2004; Jang, Livesley, Taylor, Stein, & Moon, 2004; Lynam & Widiger, 2007; Watson & Wu, 2005). In the bulk of this chapter, we focus on those advances and their implications. The advances are likely to lead to a fundamentally different, and more valid, means 205


of describing psychological dysfunction. Perhaps they have been made in time to influence the organizational structure of diagnostic systems, such as the DSM and ICD approaches. To make this argument, we first describe aspects of the last century’s worth of advances in psychopathology description and validation theory, which paved the way for the recent progress to which we refer. We then describe this recent progress, doing so in terms of the conclusions drawn by validation theorists and in terms of psychopathology research programs that have fruitfully disaggregated heterogeneous disorders. We touch briefly on a new generation of reaggregation approaches as well, and we consider evidence for the validity, parsimony, and utility of these new approaches. Once we have made this argument, we turn to a second recent concern of validation theorists: how to conduct sound validation tests in the absence of well-Â�developed, lawful networks of relationships among psychopathology constructs. When such lawful networks are absent, the meaning of construct validation has not always been clear. As a result, validation efforts sometimes appear to have an ad hoc, opportunistic quality, in which any correlation is considered evidence of construct validity (Cronbach, 1988; Kane, 2001). By focusing on how informative a validation test is, given the current state of knowledge, researchers can conduct sound, rigorous tests even in the absence of wellÂ�developed lawful networks of relationships among constructs. We turn to this issue later in this chapter.

Theoretical Advances over€the€Last Century: A€Very€Brief€History Early Efforts One of the early attempts to codify what was known about psychopathology was provided by the Woodworth Personal Data Sheet (WPDS). The WPDS was developed in 1919 to assist in screening out U.S. Army recruits who might be psychologically vulnerable to the stress of war; it was thought to measure emotional stability (Garrett & Schneck, 1928; Morey, 2003). To construct the test, Woodworth created 116 dichotomous items


on both rational grounds (he drew on case histories of identified patients for item content) and empirical grounds (he deleted items endorsed by 50% of more of a normal comparison group) (Garrett & Schneck, 1928). Although the items appear to have been developed by plausible methods, the scale did not perform very well. The total WPDS score did not differentiate between college freshmen and so-Â�called “avowed psychoneurotics” (Garrett & Schneck, 1928); nor did it correlate with teacher ratings of students’ emotional stability (Flemming & Flemming, 1929). To understand this failure, researchers focused on the item content and concluded that a total score appeared to combine multiple, different forms of dysfunction (Garrett & Schneck, 1928; Laird, 1925). Indeed, sample WPDS items included “Have you ever lost your memory for a time?”, “Can you sit still without fidgeting?”, “Does it make you uneasy to have to cross a wide street or an open square?”, and “Does some particular useless thought keep coming into your mind to bother you?” In noting the diversity of mental complaints that were summed to yield an overall score, Garrett and Schneck (1928) concluded: It is this fact, among others, which is causing the present-day trend away from the concept of mental disease as an entity. Instead of saying that a patient has this or that disease, the modern psychiatrist prefers to say that the patient exhibits such and such symptoms. (p.€465)

These authors thus emphasized the identification of specific symptoms, rather than aggregations of those symptoms. Even in this early study, Garrett and Schneck recognized the need to avoid combining individuals with different symptom pictures. They appear to have anticipated the current construct validity focus on hom*ogeneous constructs and on the inadequacy of single scores to reflect multidimensional entities. In fact, the subsequent history of psychopathology theory has reflected movement in precisely the direction advocated by Garrett and Schneck. The first step in this direction was to describe dysfunction in terms of syndromes, or constellations of symptoms thought to stem from a common cause (Kraepelin, 1883/1981). This approach has been predominant for most of the 20th century and into the 21st. The dif-

Issues of Construct Validity

ferentiation of syndromes from one another represented an important advance in specificity, as we describe below. More recently, researchers have begun to describe psychological dysfunction in a new, different way. They have come to describe dysfunction in terms of basic dimensions; this approach appears to provide even greater precision and clarity than did the traditional syndrome approach. We next consider the syndrome approach, its obvious generativity, and its relationship to advances in validation theory. Syndromes of Dysfunction In surveying the broad category of “mentally disturbed” individuals, Kraepelin (1883/1981) famously perceived that certain groups of symptoms tended to occur together. Such groups of symptoms were called “syndromes,” and he understood them to have a common biological cause, just as medical diseases often involve a cluster of symptoms stemming from a single cause. The use of the medical perspective to identify separate syndromes of psychological dysfunction has proven enormously generative; it has helped generations of astute clinicians and theoreticians define different classes of dysfunction that probably had different etiologies and different consequences for individuals. Each newly developed version of the DSM and the ICD has reflected the productive differentiation of different forms of dysfunction. From a theoretical standpoint, with each new differentiation, it becomes possible to investigate ever more precise, better-Â�delineated theories of dysfunction. For example, the recognition that disorders of personality (i.e., chronic dysfunction in characteristic modes of perceiving the world and behaving in it) are distinct from other forms of dysfunction has been enormously important (Millon, 1996). Together with the awareness of the stability of human temperament and personality across the lifespan (Caspi & Roberts, 2001; Roberts & DelVecchio, 2000), it has led to both theories and treatments specific to personality dysfunction (Linehan, 1993; Millon, 1996; Widiger & Boyd, 2009; Widiger, Simonsen, Krueger, Livesley, & Verheul, 2005). The progressive differentiation among different psychological disorders has also


proven highly beneficial to persons suffering from these disorders. Increasingly, psychologists have identified different treatments for different forms of dysfunction. For example, the basic description of the tripartite model of emotional distress (Clark & Watson, 1991) has led to different interventions for those high only in negative affect and for those who are both high in negative affect and low in positive affect. The former pattern is understood to characterize anxiety, and treatments involve exposure to the specific stimuli that elicit the distress. The latter pattern is understood to characterize depression, and an important focus of treatment is to increase positive affect through behavioral activation: Individuals return to active engagement in activities that were previously reinforcing, and the reinforcement and positive mood follow (Barlow, Allen, & Choate, 2004; Dimidjian et al., 2006). A second example is the emergence of dialectical behavior therapy as an effective treatment for borderline personality disorder. Before the development of such a treatment was possible, researchers had to identify the disorder as distinct from other forms of dysfunction.

Syndromes of Dysfunction and Test Validation Procedures In the first half of the 20th century—Â�perhaps in part due to the failures of rationally based test construction, and perhaps in part due to philosophical objections to the invocation of unobservable entities (Blumberg & Feigl, 1931)—test validity came to be understood as a test’s ability to predict a criterion (Kane, 2001). In fact, many validation theorists explicitly rejected the idea that scores on a test mean anything beyond their ability to predict an outcome. As Anastasi (1950) put it, “It is only as a measure of a specifically defined criterion that a test can be objectively validated at all. .€.€. To claim that a test measures anything over and above its criterion is pure speculation” (p.€67). In a great deal of psychopathology research at the time, the criterion to be predicted was psychiatric diagnosis, or presence of a syndrome. The criterion validity approach led to the criterion-Â�keying test construction approach, in which one selects items entirely on the basis of whether the items predict the criterion. This method was used in the


construction of two of the most prominent measures of personality and psychopathology, the Minnesota Multiphasic Inventory (MMPI) (Cox, Weed, & Butcher, 2009) and the California Psychological Inventory (CPI) (Megargee, 2009). Typically, items were selected for scales on the basis of whether they differentiated between individuals thought to have a certain syndrome and individuals thought not to have the syndrome. Substantively, psychopathology scales constructed with criterion-Â�keying methods have afforded important benefits to clients. For example, the MMPI-2 has provided valid clinical assessments and has proven useful for treatment planning (Butcher, 1990; Greene, 2006; Nichols & Crowhurst, 2006; Perry, Miller, & Klump, 2006). The CPI has also validly predicted a wide range of criteria (Gough, 1996). The substantive success of criterion-keyed tests meant that clinical psychologists, and others, could make important diagnostic decisions that affected people’s lives with improved validity. From today’s perspective, there are two obvious problems with this approach. The first is that it presumes the validity of the criteria. As Bechtoldt (1951) put it, reliance on criterion-Â�related validity “involves the acceptance of a set of operations as an adequate definition of whatever is to be .€ .€ . [predicted]” (p.€ 1245; original emphasis). Thus the criterion validity approach, though useful for identifying predictors of psychopathological syndromes, generally did not involve direct evaluations of the validity of the syndrome descriptions themselves. In short, the criterion validity approach tended to presume, rather than test, the validity of the syndrome characterization of psychological dysfunction. The second problem with exclusive reliance on the criterion validity approach is this: When tests are developed only to predict a circ*mscribed criterion, and when they are only validated with respect to that predictive task, the process adds little to basic theory. Criterion-Â�validated tests are not well suited for use in testing theories describing relationships among psychological processes. To test such theories, one often needs tests representing psychological constructs that cannot be captured by a single criterion (Cronbach & Meehl, 1955). Today it seems that an exclusive focus on predict-


ing narrow, syndromal criteria retarded theory development; thus comprehensive theories of personality functioning (Costa & McCrae, 1992), or models identifying core dimensions of psychopathology (Krueger & Markon, 2006) or personality dysfunction (Widiger et al., 2005; Widiger & Trull, 2007), were simply not available as objects of inquiry. However, this limitation needs to be understood within its historical context. In the first half of the 20th century, the state of knowledge in psychopathology simply did not permit the development of sound measures based on well-�developed theory. The predictive failure of the WPDS and other instruments reflects that reality. Perhaps the reliance on the criterion-�related validity approach was necessary at the time. In fact, despite its limitations for theory testing, it did in fact lead to a vast growth in knowledge that made possible many of the theoretical advances that have taken place since then. As researchers came to understand which criterion-keyed test items or subscales identified which disorders, it became possible to extrapolate from successful prediction to the development of theories of dysfunction. Perhaps ironically, the success of the criterion-based validation method led to its ultimate replacement with construct validity theory. Increasingly sophisticated, integrative theories of psychopathology that relied on the operation of unobservable psychological entities could not be validated by using the criterion approach. Thus there was a need for advances in validation theory to make it possible to test theories that included unobservable entities. Construct validity theorists developed an impressive theoretical and methodological perspective to meet this need (Campbell & Fiske, 1959; Cronbach & Meehl, 1955; Loevinger, 1957). We discuss construct validity theory later in this chapter.

Limitations of the Syndrome Approach The syndrome approach has two fundamental limitations. First, historically, decisions concerning which sets of symptoms go together have not been made on fully empirical grounds; in fact, the statistical and computational tools necessary for readily investigating symptom covariation were developed

Issues of Construct Validity

after symptoms were initially assigned to syndromes. It has only recently become possible to test empirically whether certain sets of symptoms do in fact covary in ways that justify their combination into a syndrome. As we describe below, there is now good reason to doubt that many syndromes actually reflect a uniquely high degree of covariation among their element symptoms. Second, many symptoms are common across numerous putative syndromes (Barlow et al., 2004; Brown, Chorpita, & Barlow, 1998; Clark & Watson, 1991; Watson, 2005; Widiger & Costa, 2002; Widiger et al., 2005; Widiger & Trull, 2007; Whiteside & Lynam, 2001). If symptoms are frequently common across syndromes, one has reason to question the value of the syndrome organization. Perhaps, instead, hierarchical models can be organized along dimensions of dysfunction, thus better representing the empirical covariation of symptoms among humans. The movement from describing psychological dysfunction in terms of syndromes to describing it in terms of hom*ogeneous dimensions of dysfunction is the core advance we discuss in this chapter. To provide a full appreciation for the methodological and substantive basis for this movement, we proceed in two steps. First, we review the basis for the conclusion by validation theorists that the study of unidimensional entities provides the foundation for sound theory testing. We then review a sample of recent programs of psychopathology research that have identified basic dimensions of dysfunction.

Construct Validation and Theory€Testing: The Importance of Unidimensionality Psychopathology theory involves, in part, specifying the nature of relationships among different psychological attributes (causes, correlates, mediators, moderators, etc.). One of the fascinating challenges of psychological science is that the psychological entities we study cannot be directly observed (Cronbach & Meehl, 1955); researchers must infer their existence. Researchers do so in order to best approximate their understanding of real psychological phenomena (Borsboom, Mellenbergh, & van Heerden, 2003).


Their doing so has clear utility for helping us understand human behavior, differences among individuals, and dysfunction (Smith, 2005; Smith et al., 2007). Psychopathology researchers test a theory by developing measures of the inferred entities and testing whether the measures relate to measures of other inferred entities as specified by the theory. Repeated findings consistent with the theory produce increasing confidence in both the theory and the measures used to represent the constructs of interest. Findings inconsistent with the theory raise questions about both the theory and the measures used to test it. The indeterminate nature of this process is clear and has been described many times before (Smith, 2005). It follows that when we refer to “construct validity investigations” of psychopathology, we are necessarily referring to simultaneous tests of theories and of measures (Cronbach & Meehl, 1955; Smith, 2005). The process of construct validation requires sound definitions of target constructs and a clear statement of anticipated relationships among these constructs. Tests of the validity of construct measures must inevitably be tests of theories specifying relationships among the constructs. Working from this perspective, validation theorists have drawn an important conclusion concerning test construction, test validation, and theory validation. To the degree that one uses a single score from a target measure that includes multiple dimensions (such as a measure of posttraumatic stress disorder [PTSD] thought to include four factors or a measure of neuroticism thought to have six facets), one’s construct validation/theory test has theoretical uncertainty built in. Such a test is likely to have reduced scientific value (Edwards, 2001; Hough & Schneider, 1996; McGrath, 2005; Paunonen, 1998; Paunonen & Ashton, 2001; Schneider, Hough, & Dunnette, 1996; Smith et al., 2003; Smith & McCarthy, 1995; Smith et al., 2009; Smith & Zapolski, 2009). If one correlates a total score of a multidimensional measure with a criterion, one builds two sources of uncertainty into one’s test. The first source of uncertainty is that with a single score, one cannot know the nature of the different dimensions’ contribution to that score. Conceivably, an over-


all correlation could reflect the same magnitude of relation between each dimension and the criterion, but it need not. In fact, it is more likely that such a correlation reflects a kind of average of strong and weak relationships between different dimensions and the Â�criterion (Smith et al., 2003; Smith & McCarthy, 1995). Mathematically, one cannot know the meaning of a single score representing a multidimensional measure (Borsboom et al., 2003; McGrath, 2005; Smith et al., 2009). Psychometricians have been making this point in various ways for the last 10–15 years. Edwards (2001) noted that researchers have long appreciated the need to avoid heterogeneous items; if such an item predicts a criterion, one will not know which aspect of the item accounts for the covariance. The same reasoning extends to tests: If a test includes multiple dimensions, one cannot know which dimensions account for the test’s covariance with measures of other constructs. If one uses single scores from multidimensional tests, one has simply moved the heterogeneity problem from the item level to the scale level (Smith et al., 2003). Hough and Schneider (1995), McGrath (2005), Paunonen and Ashton (2001), and Schneider and colleagues (1996), among others, have all noted that using scores of broad measures often obscures predictive relationships. Indeed, studies that have compared prediction using specific facets of broad personality dimensions with prediction using scores on the dimensions themselves show that prediction is improved when one represents each facet individually (Paunonen, 1998; Paunonen & Ashton, 2001). Essentially, one gives oneself the chance to study the separate and incremental roles of each dimension involved in one’s measures, rather than averaging across the different dimensions before predicting. Concerning the second source of uncertainty, it is not just that a composite score averages the functioning of separate constructs in its association with measures of other constructs. The problem is more severe than that. The same composite score will tend to reflect different combinations of construct scores for different individuals in a sample. For example, imagine two individuals with the same overall neuroticism score on the NEO Personality Inventory—Â�Revised (NEO PI-R) measure of the five-Â�factor model (FFM) of personality (Costa & McCrae,


1992). Two of the six facets of neuroticism in that measure are angry hostility and anxiety. One person could be high in angry hostility but low in anxiety, and the other could be low in angry hostility but high in anxiety. This possibility is not just hypothetical; individuals exhibiting psychopathy appear to be high in angry hostility but low in anxiety (Lynam & Widiger, 2007). The same overall neuroticism score could easily be obtained by a person with psychopathy and a highly internalizing individual suffering from anxiety. Indeed, the two traits correlated r = .47 in the standardization sample, meaning that they shared only 22% of their variance (Costa & McCrae, 1992). It follows that covariation of an overall neuroticism score with another variable lacks clear meaning. It is not just that when one correlates neuroticism with another variable, one cannot know whether the correlation was “carried” by, in this case, angry hostility or anxiety. It is that the same score could reflect angry hostility elevations for some individuals and anxiety elevations for others. For these reasons, the central construct validation process should be to test hypothesized relationships among what are thought to be hom*ogeneous, precisely defined constructs. In the present instance, one should study either angry hostility or anxiety, but not a composite that obscures their different roles. Of course, the determination of unidimensionality is itself a validation/theoryÂ�testing enterprise. Unidimensionality is established through analyses both internal to a measure, such as factor analysis, and external to a measure, such as tests of convergent and discriminant validity. Psychopathology research has already proceeded in the pursuit of hom*ogeneous dimensions of dysfunction; we next review programs of research that have sought to identify such dimensions within multidimensional syndromes. As we will show, these efforts indicate that many putative syndromes are in fact characterized by multiple dimensions that are only modestly interrelated, that sometimes appear to have different etiologies, and that are sometimes differentially responsive to a given treatment. When this is true, the use of the syndrome descriptions appears not to be indicated, and may even be counterproductive. We use the current Axis I–Axis II distinction to present this research.

Issues of Construct Validity

The Disaggregation of Mental Disorders: Axis I Schizophrenia Schizophrenia is now well understood to be a “heterogeneous disorder with diverse history, course, and symptoms” (Bell, Lysaker, Beam-Â�Goulet, Milstein, & Lindenmayer, 1994, p.€295). One approach to understanding the nature and implications of the diversity of schizophrenia symptoms has involved the development of the Positive and Negative Syndrome Scale (PANSS; Kay, Fiszbein, & Opler, 1987). Although there is some debate about the precise number of factors on the PANSS, a five-Â�factor structure is commonly found (Bell et al., 1994; Kay et al., 1987; Levine & Rabinowitz, 2007; Nakaya, Suwa, & Ohmori, 1999). A typical description of the five dimensions includes negative symptoms, positive symptoms, disorganization, anxiety/depression, and symptoms of excitement (Levine & Rabinowitz, 2007). The five dimensions are not highly correlated: Levine and Rabinowitz found intercorrelations ranging from –.14 to .41, and the intercorrelations reported by Nakaya and colleagues (1999) ranged from –.09 to .52. Clearly, individuals can have elevations in one dimension without having elevations in other dimensions. An overall schizophrenia symptom count, whether treated as an interval scale variable or as a basis for DSM classification, would combine endorsem*nts of clearly separable constructs. Attempts to clarify the implications of the heterogeneous processes included in the “schizophrenia” label have included attempts to “deconstruct” schizophrenia by studying presumably unidimensional endophenotypes. The rationale for this approach is that there appear to be multiple, moderately related underlying processes involved in “schizophrenia,” and each such process may be influenced by different etiological factors (Braff, Freedman, Schork, & Gottesman, 2007). By studying specific, unidimensional endophenotypes, researchers hope to be able to describe the different neurobiological and genetic architectures of patients with schizophrenia who exhibit different symptom pictures (Braff et al., 2007). It seems increasingly clear that the dimensions of schizophrenia, not the overall diagnosis, are the proper objects of theoretical inquiry.


Often the positive symptoms are understood to be the indicators of psychosis (Buchanan & Carpenter, 1994; Serretti & Olgiati, 2004). However, the experience of psychosis itself has been shown to be multidimensional by numerous authors (see Mizrahi et al., 2006, for a review). Typical dimensions identified include conviction in the psychotic experience, cognitive preoccupation with the psychotic experience, behavioral impact of the experience, emotional involvement with the experience, and external perspective about the experience. A comparative factor analysis indicated that a five-Â�factor structure of a measure of positive symptoms fit far better than did a single-Â�factor solution (Mizrahi et al., 2006). Perhaps most interestingly, antipsychotic medications appear to influence different dimensions of psychosis differently: In one recent study, medication produced a 32% improvement on the behavioral impact dimension, but only a 6% improvement in the degree of conviction about the psychotic experience and no improvement in external perspective about the experience (Mizrahi et al., 2006). It is not the case that antipsychotic medications influence the disorder as a whole; rather, they influence specific dimensions of functioning. In addition, there is increasing recognition that different psychotic syndromes actually share common dimensions of functioning (Serretti & Olgiati, 2004). These authors, studying a mixed group of patients with psychotic disorders, identified five dimensions of “psychosis” (broadly construed): mania, positive symptoms, disorganization, depression, and negative symptoms. They observed that several of these dimensions are shared across syndromes; perhaps the dimensions, not the constructed syndromes, should be the basis for understanding psychopathology (Serretti & Olgiati, 2004). Depression Jang and colleagues (2004) studied the factor structure of depression. Using several symptom lists, they identified 14 subfactors. Examples of these subfactors included feeling blue/lonely, insomnia, positive affect, loss of appetite, and psychom*otor retardation. Strikingly, intercorrelations among the factors ranged from .00 to .34, and the factors were differentially heritable, with heritability coefficients ranging from .00


to .35. It therefore seems that (1) some of the dimensions of depression do not covary substantially, and (2) some have a heritable basis and others do not (i.e., their etiologies appear to differ). McGrath (2005) provides interesting examples of the heterogeneity of depression symptom items. If the factors of depression share between 0% and 12% of their variance, then two individuals with the same depression score can easily be experiencing very different symptoms. And if the factors have different etiologies, then the sources of the difficulties the two individuals face may also be very different. It thus appears to be the case that depression is not a hom*ogeneous psychological construct; it may not even represent a sound hierarchical aggregation of hom*ogeneous constructs. Use of overall depression scores as a variable in psychopathology research is likely to be problematic. For example, testing whether stressful events are a risk factor for depression is imprecise. Are they a risk factor for each construct subsumed within the overall label? Are they a risk factor for only one construct, or for some subset of constructs? For example, do they tend to reduce positive affect, but not influence negative affect? Or do they increase negative affect but not relate to positive affect? Do they influence both? The imprecise test yields imprecise results. Obsessive–Â�Compulsive Disorder Many authors have separated obsessive–Â� compulsive disorder (OCD) into several dimensions. Watson and Wu (2005) identified obsessive checking, obsessive cleanliness, and compulsive rituals as separate and only moderately related constructs (sharing between 25% and 31% of their variance), and concluded that OCD may be both phenotypically and genotypically heterogeneous. Mathews, Jang, Hami, and Stein (2004) identified four very similar factors: contamination, repeating/doubts, checking/detail, and worries/just right. They found that the different dimensions had different external correlates. For example, trait anxiety correlated .44 with checking/detail and .15 with worries/just right, whereas depression correlated .30 with worries/just right, but only .04 with checking/detail. Leckman and colleagues (1997) reported similar findings.


The putative syndrome appears to have multiple dimensions, which share only a moderate amount of variance with each other. Clearly, individuals can be high on one dimension without being high on another dimension (e.g., elevation in obsessive checking does not necessitate elevation in hoarding). The dimensions have different external correlates, and they may have different etiologies. It thus appears that OCD scores combine different psychological entities; therefore, the use of overall OCD scores or OCD diagnoses is problematic in psychopathology research. Posttraumatic Stress Disorder The symptoms of PTSD have been shown to fall on four factors (intrusions, avoidance, dysphoria, and hyperarousal) by Simms, Watson, and Doebbeling (2002). The four�factor model fit far better than did a model describing a single PTSD dimension, and the four factors had different external correlates. King, Leskin, King, and Weathers (1998) also found that a four-�factor model (reexperiencing, effortful avoidance, emotional numbing, and hyperarousal) fit their 17-symptom clinical interview better than did any other model, including a single-�factor model or a hierarchical model, in which an overall PTSD factor was thought to underlie the four factors. Most recently, Palmieri, Weathers, Difede, and King (2007) also found four factors with different external correlates. Clearly, identical PTSD symptom counts can refer to very different symptom pictures. There is little evidence for the existence of an overall construct of PTSD. Thus there is good reason to question the use of the PTSD syndrome; use of the four dimensions instead appears to involve a more valid approximation of individual differences in dysfunction within this domain.

The Disaggregation of Mental Disorders: Axis II Psychopathy The disaggregation of the many components of psychopathy has received considerable research attention (Brinkley et al., 2004; Cooke & Michie, 2001; Harpur, Hakstian,

Issues of Construct Validity

& Hare, 1988; Harpur, Hare, & Hakstian, 1989; Lynam & Widiger, 2007). In one classic modern description of psychopathy, Hare’s (2003) Psychopathy Checklist—Â� Revised (PCL-R) identified two separate factors—one representing the callous and remorseless use of others, and the other representing a deviant and antisocial lifestyle. In the PCL-R, the two factors share only 25% of their variance (Hare, 2003; Harpur et al., 1988), and they have numerous different correlates (Harpur et al., 1989). In addition, the factors may have different etiologies. Based on a series of findings, Patrick and colleagues suggested that high scores on the dimension of callous, emotional detachment may reflect a deficit in normal responsivity to aversive stimuli, whereas high scores on the antisocial behavior dimension may reflect a deficit in the higher-order processes governing goal setting and delay of gratification (Patrick, Bradley, & Lang, 1993; Patrick, Cuthbert, & Lang, 1994). If the two dimensions share only a moderate amount of variance, have different external correlates, and have different etiologies, then a single score combining the two obscures important differences in psychological processes. More importantly, these findings question the value of combining the two dimensions and referring to a single syndrome of psychopathy. Others have disaggregated the syndrome differently. Cooke and Michie (2001) identified three factors, described as (1) arrogant and deceitful interpersonal style, (2) deficient affective experience, and (3) impulsive and irresponsible behavioral style. Furthermore, it appears to be the case that the PCL-R does not include all of the dimensions of the classic description of psychopathy provided by Cleckley (1941); for example, low anxiousness is not represented (Lynam & Widiger, 2007; Rogers, 1995). It may be that the putative syndrome of psychopathy includes several dimensions of dysfunction. Most recently, Lynam and Widiger (2007) took advantage of the hierarchical FFM of personality to describe the psychopathy construct across all 30 basic personality traits identified in the NEO PI-R five-Â�factor measure. For each trait, they identified whether the trait was related to psychopathy and whether high or low trait scores reflected the psychopathy construct. The result was a


placement of psychopathy along each of the 30 hom*ogeneous dimensions of personality; in this view, psychopathy is understood to represent a multidimensional combination of constructs, rather than a coherent theoretical entity in and of itself. Interestingly, Lynam and Widiger’s view of psychopathy made extensive use of distinctions between personality facets on the same broad personality domain. For example, persons exhibiting psychopathy were understood to be high on impulsiveness and angry hostility, and low on anxiety and self-Â�consciousness; all four of those traits are placed on the neuroticism domain of the FFM. Schizotypal Personality Disorder Fossati and colleagues (2005) compared several different factor structures for the schizotypal personality disorder criteria, and found that a three-Â�factor model (cognitive–Â� perceptual, interpersonal, and disorganization) fit best. Intercorrelations among the three factors ranged from .14 to .63, again indicating substantial unshared variance in each factor. Here, too, individuals were often high on one factor but not on another. Again in this case, the same quantitative symptom count could reflect very different dysfunctional experiences. Narcissism Emmons (1987) found four scales in the Narcissistic Personality Inventory (NPI; Raskin & Hall, 1979): leadership/authority, self-Â�absorption/self-Â�admiration, superiority/ arrogance, and exploitiveness/entitlement. Intercorrelations among these four factors ranged from .16 to .57. Use of the total NPI score obscures differences among the four factors and can lead to lack of empirical clarity. For example, Emmons (1987) found the correlation between the total NPI score and the Narcissistic Personality Disorder Scale (NPDS) to be nonsignificant and small (r = .12), perhaps suggesting a lack of convergent validity between the two measures. However, this value essentially reflected an average of different correlations between individual NPI subscales and the NPDS: Superiority/ arrogance correlated –.04 with the NPDS, but exploitiveness/entitlement correlated .32 with the measure. The four scales also had


different correlations with external criteria. Perhaps, then, narcissism is not a coherent psychological construct, but rather an aggregate of constructs. If so, then theory testing should be conducted separately on the dimensions. We have sampled from research programs seeking to identify the hom*ogeneous dimensions of putative psychiatric syndromes. Disaggregation research is currently underway with respect to other disorders as well; some research suggests that certain syndromes are valid ways to describe dysfunction, but other research does not. For example, Sanislow and colleagues (2002) identified three dimensions to borderline personality disorder (affective dysregulation, behavioral impulses, and disturbed relatedness), but intercorrelations among the three were all greater than .90, suggesting hom*ogeneity to the disorder. Fossati and colleagues (1999) also found evidence for hom*ogeneity. On the other hand, intercorrelations among the DSM-IV criteria for borderline personality disorder range from .01 to .34 (Sanislow, Grilo, & McGlashan, 2000), and individuals with certain clusters of symptoms are not very likely to have other symptom clusters (Rusch, Guastello, & Mason, 1992). It does seem that individuals diagnosed with borderline personality disorder can have very different symptom pictures, and hence very different psychological experiences.

Implications of the Advances in Validation Theory and€of€Disaggregation Research The advances in validation theory and in the disaggregation of putative syndromes offer three important advantages to psychopathology research: theoretical clarity, parsimony, and utility. We consider each in turn. Theoretical Clarity Because it is now clear that single scores representing multiple dimensions average the different effects of different psychological processes, and because it is increasingly clear that many putative syndromes combine different dimensions of dysfunction with different etiologies, theoretical clarity requires


a focus on hom*ogeneous dimensions of dysfunction. For multidimensional syndromes, the use of symptom counts or overall scores in theory or validation studies cannot be defended. For many disorders, then, the use of diagnostic status or a disorder score as either a predictor or a criterion will tend to produce unclear results. A score on depression, or a depression diagnosis, reflects scores on several different constructs. To test a theory that experience x is a risk factor for depression is therefore to be imprecise. It may be the case that experience x is a risk factor for one factor within depression, but not for other factors. A proper test of that possibility requires assessment of the separate components of depression and examination of the association between experience x and the target factor. If one were to use overall depression scores instead, and if none of the other factors were related to experience x, one would risk missing the association altogether. Perhaps more problematically, in such a situation one has to assume that the symptom count score reflects the same psychological experience for each person, but this does not appear to be true. Individuals may have very similar symptom counts, but very different patterns of scores on individual constructs (McGrath, 2005; Widiger & Trull, 2007). When that is the case, the symptom count does not refer to a coherent theoretical entity, and its correlation with measures of other constructs has unclear meaning. In short, to continue using single scores to represent multidimensional entities is to risk retarding the rate of advance in understanding psychopathology. Certainly the hetereogeneity of disorders, such as depression, is no surprise to practicing clinicians. Clinicians know full well that they do not fully understand a client’s symptom picture simply because they understand the client to be depressed. Clinicians proceed from that diagnosis to try to understand whether the client is experiencing sad affect, anhedonia, loss of appetite, sleep disruption, or other forms of dysfunction. They tailor their interventions to the symptoms that are present. Parsimony The approach we advocate does not mean endless fractionation. In fact, the opposite

Issues of Construct Validity

appears to be true: By focusing on dimensions of dysfunction, psychopathology researchers are likely to find fewer such dimensions of importance than there are constellations of symptoms called syndromes. Perhaps the most well-Â�developed body of relevant research concerns dimensional descriptions in the domain of personality disorders. A recent international conference indicated that researchers appear close to a consensus on describing the domains of dysfunction relevant to personality disorders in terms of four basic dimensions of personality (Widiger & Simonsen, 2005; Widiger et al., 2005). That is, patterns of elevations across four personality dimensions and their underlying facets can be used to describe the dysfunction currently described by the full set of personality disorders. There appears to be an advance in parsimony from targeting basic dimensions of personality dysfunction, rather than trying to delineate multiple syndromes. In fact, integrative work holds the promise of identifying dimensions of dysfunction shared by personality disorders and Axis I clinical disorders (Krueger, 2002, 2005; Widiger & Simonsen, 2005). Krueger (2005) has suggested that the two groups of disorders share basic dimensions of normal and abnormal personality functioning. This suggestion is supported by the findings of Markon, Krueger, and Watson (2005), which constituted a compelling empirical integration of measures thought to reflect normal and abnormal personality functioning. It appears that the same five factors summarize both domains, and that what is understood to be abnormal personality functioning simply represents extremes of normal dimensions (Clark, 2007; Widiger & Trull, 2007). Again, the demands of parsimony indicate the value of focusing on basic dimensions of dysfunction, rather than multiple putative syndromes that sometimes reflect slightly different combinations of those dimensions. Serretti and Olgiati’s (2004) identification of basic dimensions of psychosis that apply across current diagnostic distinctions suggests parsimony in the dimensional description of psychosis. Recognition of the two broad dimensions of positive affect and negative affect (Clark & Watson, 1991) has shed light on dimensions of dysfunction common to multiple disorders and contributed to the development of a unified model for the treat-


ment of emotional disorders (Barlow et al., 2004); using this approach, one addresses the relevant dimension of functioning rather than specific diagnostic syndromes. Utility For several reasons, descriptions of psychopathology in terms of dimensions of dysfunction are likely to be far more clinically useful than descriptions using categories of syndromes have been. First, validity is a necessary condition for utility. Investigations of the causes, correlates, and consequences of unidimensional psychological constructs produce interpretable results: The associations observed refer to a definable dimension of functioning. But investigations of the causes, correlates, and consequences of multidimensional constructs represented by a symptom count or diagnosis produce unclear results: One cannot know which dimensions of a construct influence the associations observed and which do not. Thus the matrix of validity evidence for such constructs rests on weaker grounds and so can provide clinicians with ambiguous or misleading results. Consider the finding that behavioral activation therapy reduces depression (Dimidjian et al., 2006). This finding is very important, but it may also lack important precision. Suppose it were the case that behavioral activation increases positive affect but has no influence on negative affect. If so, then after behavioral activation treatment, high levels of negative affect may remain to be treated. The conclusion that the treatment alters depression levels does not provide the clinical precision necessary; it would be more useful to clinicians to know that a given treatment influences one symptom but not others. Verheul (2005), reviewing evidence for clinical utility between syndromal and dimensional models of psychopathology, concluded that “overall, the categorical system has the least evidence for clinical utility, especially with respect to coverage, reliability, subtlety, and clinical decisionÂ�making” (p.€295). The syndrome model involves heterogeneity among individuals with a common diagnosis, a lack of precision in description, diagnostic co-Â�occurrence, and arbitrary diagnostic boundaries. Each of these characteristics complicates the diag-


nostic and treatment tasks facing clinicians (Widiger & Lowe, 2008). There is also good reason to believe that a descriptive system based on hom*ogeneous dimensions would prove more feasible for clinical use than the syndrome-based system has. Widiger and Lowe (2008) have recently offered one practical approach to personality disorder assessment, in which individuals are assessed across 26 facets of personality that are thought to be clinically relevant. They note that assessment along these dimensions takes about half the time that a DSM-IV assessment takes. The existing diagnostic syndromal categories generally do not lead to identification of treatments specific to disorders (Kupfer, First, & Regier, 2002). In the domain of personality disorders especially, an approach based on dimensions of functioning appears to lead more directly to treatment implications. To earn a DSM-IV personality disorder diagnosis, one must be experiencing “clinically significant distress or impairment in social, occupational, or other important areas of functioning” (American Psychiatric Association, 1994, p.€633). The FFM of personality appears quite useful in identifying dimensions of functioning that are directly related to these concerns (e.g., extraversion–Â� introversion and agreeableness–Â�antagonism pertain specifically to social functioning). Widiger and Lowe (2008) offer specific suggestions tying personality profiles to appropriate interventions. Empirical demonstrations of improved utility for dimensional descriptions are underway. Samuel and Widiger (2006) showed that descriptions of clinical cases using the 30 NEO PI-R facets of the FFM were significantly more useful to clinicians than were DSM-IV diagnoses. The detailed descriptions of patients along each of 30 unidimensional components of personality were rated as more useful in providing global personality descriptions of clients, in communicating information to clients, in describing clients’ important personality difficulties, and in formulating effective treatment interventions. As those authors noted, the detailed information provided by the use of all 30 facets, rather than just the five broad personality scales, probably improved the model’s clinical utility significantly. Clinical psychologists can now use numerous detailed


and coherent constructs to describe clients’ characteristic functioning, and doing so has proven quite useful. Widiger and Trull (2007) address this issue in depth. Having noted these advantages in regard to utility, we should also recognize the practical difficulties associated with any shift of this magnitude. First (2005) observed that this change would complicate record keeping; cause significant administrative problems; require retraining; disrupt research; and disrupt clinicians’ ability to (1) integrate prior research into clinical care and (2) communicate effectively with other mental health practitioners. Of course, as First recognized, none of these issues concern the validity of the science involved. As he noted, if the change provided improved validity and clinical utility, researchers might conclude that it would be worth the disruption. We agree. Surely disruptions of these kinds can be tolerated as by-Â�products of an improved descriptive system.

Dimensions and Taxons Before we conclude this chapter, we need to clarify our use of terms. We have repeatedly referred to “hom*ogeneous dimensions of functioning”; we chose the term in order to emphasize the concept of “hom*ogeneity.” Our intent has been to argue that different constructs should not be combined to produce a single score or to represent what is thought of as a single entity. A different question concerns whether a given psychological entity should be understood to exist along a continuum or as a category/taxon. The arguments we have made here, concerning the need to avoid combining multiple entities, apply whether the entities are taxa or continua. We have used the language of “dimensions,” rather than “taxa,” for ease of communication. Although we do feel confident that most aspects of psychopathology exist along continua (Widiger & Trull, 2007), this is a separate matter. The core new contribution of validation theory concerns hom*ogeneity and so applies in either case. To finish our discussion of the implications of current construct validity theory for psychopathology description, we turn to a different issue altogether. We identify a

Issues of Construct Validity

current difficulty with the implementation of construct validation procedures, and we offer a practical solution to that difficulty.

Informative Validation Tests In their treatment of construct validity, Cronbach and Meehl (1955) relied on the concept of a “nomological network,” which refers to a set of lawful relations among entities. As they described it, the process of construct validation involved specifying the lawful relations between an inferred construct and other constructs, and then testing whether one’s measure of the inferred construct produced the results specified in the nomological network. Unfortunately, the idea that one can define constructs by their place in a lawful network of relationships assumes a theoretical precision that is generally not present in psychopathology research specifically, and in the social sciences generally. Psychopathology researchers are often faced with the task of validating their measures and theories despite the absence of a set of expected lawful relations among constructs. When this is true, the meaning of “construct validity,” and of what counts as validation evidence, is ambiguous. Cronbach (1988) addressed this issue by contrasting strong and weak programs of construct validity. Strong programs depend on precise theory that leads to specific predictions. They represent what appears to have been meant by construct validation using a nomological network of relationships. Given the typical lack of precision in psychological science, strong programs should perhaps be understood to represent an ideal to which researchers aspire. Weak programs, on the other hand, stem from less fully articulated theories and construct definitions. With weak validation programs, there is less guidance as to what counts as validity evidence (Kane, 2001). One result can be approaches in which almost any statistically significant correlation between a target measure and another measure, of any magnitude, can be described as validation evidence (Cronbach, 1988). In the absence of a commitment to precise construct definitions and specific theories, validation research can have an ad hoc, opportunistic quality (Kane, 2001), in which sets of correlations with readily


available measures are cobbled together as evidence of construct validity. The results of this kind of approach do not tend to provide strong bases for confidence in the validity of measures. If strong programs of construct validity tend to be aspirational rather than real, and if weak programs tend to provide little information, how should psychopathology researchers approach the validation of their measures? Fortunately, researchers are not stuck between an unattainable ideal and weak, ad hoc theory testing. Rather, there is an iterative process in which tests of partially developed theories provide information that leads to theory refinement and elaboration, which in turn provide a sounder basis for subsequent theory validation research and hence more precise validation tests. Cronbach and Meehl (1955) referred to this “bootstrapping” process and to the inductive quality of construct definition and theory articulation. This process has proven effective; it is evident throughout this book that striking advances in clinical research have provided clear benefits to the consumers of clinical services. We suggest a standard for evaluating validation research that acknowledges the iterative nature of theories and empirical tests of them. Psychopathology researchers should consider whether their theoretical statements and validation tests are informative, given the current state of knowledge (Smith, 2005). The informative nature of these tests depends on their satisfactory answers to the following questions: To what degree does a hypothesis involve direct criticism of a theory, or direct comparison between two alternative theoretical explanations? To what degree does a hypothesis involve a direct response to a criticism of a theory? To what degree does a hypothesis involve a claim that, if supported, would undermine criticism of one’s theory? To what degree does a hypothesis involve a claim that, if not supported, would cast real doubt on one’s theory? Given the state of development of any one theory, tests of this kind may or may not constitute strong programs of construct validation, but they do not represent weak validation programs. They address questions that will clarify the validity of theories and the measures used to test them. Because of the iterative nature of theory and mea-


sure development and validation, theoretical tests of this kind are likely to provide useful information. One characteristic of informative theory tests is that they evaluate, as directly as possible, specific claims made for a theory. In psychopathology risk factor research, if one holds that trait A is a risk factor for problem B, one must generate tests of the claim that, if not supported, would undermine the most important components of the claim. Demonstration of a positive cross-�sectional correlation between A and B does provide information (the absence of a correlation would pose serious problems for a risk theory), but its information value is limited because it is not a direct test of the risk factor hypothesis. In contrast, a longitudinal study in which trait A predicts the onset of problem B, and other possible explanations for the onset of B have been controlled, are more informative. A positive result from such a test is greater reason for confidence in both the theory and the measures of the constructs specified by the theory. Tests of mediation in which the putative cause predicts subsequent changes in the putative mediator, and the mediator then predicts still later changes in the putative consequence (e.g., Fried, Cyders, & Smith, 2008; Stice, 2001), are informative because they are more direct tests of a mediational risk process than are cross�sectional correlations; they have ruled out several possible explanations that compete with the mediational theory. Tests comparing alternative theoretical explanations of the same data (e.g., Bartusch, Lynam, Moffitt, & Silva, 1997) can also be informative because they hold multiple theories up to critical examination, both individually and in comparison to each other. Doing so is an effective way to provide information to researchers and clinicians. Research that advances understanding of psychopathology is research that involves direct tests of claims made by psychopathology researchers. To relate this concern to the main focus of this chapter, let us consider the challenge of determining whether a diagnosis represents single or multiple dimensions of functioning. To make that determination, it is necessary to compare alternative factor structures of the diagnosis directly. Confirmatory factor-�analytic techniques permit direct comparisons of the degree to


which different hypothesized factor structures accurately represent the covariances in the data. If a single factor represents the data as well as multifactor solutions do, then parsimony requires adoption of the single factor and rejection of the claim of multidimensionality. If, on the other hand, a multifactor solution best represents the data, the research enterprise is not yet complete. It is then necessary to demonstrate that the different factors relate differently to variables external to the measure; establishing that two factors represent different constructs involves demonstrating that they play different roles in the theoretical structure of psychopathology. An informative demonstration of multidimensionality requires analyses both internal and external to the diagnosis in question.

Summary In this chapter, we have noted that advances in construct validation theory and advances in psychopathology theory have been closely linked; advances in one field spur advances in the other. One dimension along which advances in both fields have taken place concerns the need to differentiate accurately among different forms of psychopathology. The current system, which relies on syndrome descriptions, has been in place for a century or more. Advances in both fields now indicate that the syndrome approach may compromise the validity of psychopathology descriptions by combining different entities and describing them with a single name and a single score. We have reviewed the validation theory argument against doing so, and we have described research programs that have identified separate dimensions of dysfunction within many currently used syndromal diagnoses. We have argued that a shift to describing psychopathology in terms of dimensions of functioning will increase validity, parsimony, and utility. We have also briefly considered another advance important for psychopathology researchers. To avoid conducting weak programs of construct validation (Cronbach, 1988; Kane, 2001), researchers should develop studies designed to be informative; that is, researchers should develop construct validation and theory validation tests that

Issues of Construct Validity

resolve important outstanding questions. Doing so will facilitate subsequent advances in theory and understanding. The recent history of psychopathology research is one of considerable success. The last century has seen dramatic increases in knowledge, with accompanying increases in clinicians’ ability to intervene and ameliorate psychological distress. As researchers apply the recent advances in validation theory discussed in this chapter, their understanding of psychopathology will only continue to increase. References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Anastasi, A. (1950). The concept of validity in the interpretation of test scores. Educational and Psychological Measurement, 10, 67–78. Barlow, D. H., Allen, L. B., & Choate, M. L. (2004). Toward a unified treatment for emotional disorders. Behavior Therapy, 35, 205– 230. Bartusch, D. R. J., Lynam, D. R., Moffitt, T. E., & Silva, P. A. (1997). Is age important?: Testing a general versus a developmental theory of antisocial behavior. Criminology, 35, 13–48. Bechtoldt, H. P. (1951). Selection. In S. S. Stevens (Ed.), Handbook of experimental psychology (pp.€1237–1266). New York: Wiley. Bell, M. D., Lysaker, P. H., Beam-Â�Goulet, J. L., Milstein, R. M., & Lindenmayer, J. (1994). Five-Â�component model of schizophrenia: Assessing the factorial invariance of the Positive and Negative Syndrome Scale. Psychiatry Research, 52, 295–303. Blumberg, A. E., & Feigl, H. (1931). Logical positivism. Journal of Philosophy, 28, 281–296. Borsboom, D., Mellenbergh, G. J., & van Heerden, J. (2003). The theoretical status of latent variables. Psychological Review, 110, 203–219. Braff, D. L., Freedman, R., Schork, N. J., & Gottesman, I. I. (2007) Deconstructing schizophrenia: An overview of the use of endophenotypes in order to understand a complex disorder. Schizophrenia Bulletin, 33, 21–32. Brinkley, C. A., Newman, J. P., Widiger, T. A., & Lynam, D. R. (2004). Two approaches to parsing the heterogeneity of psychopathy. Clinical Psychology: Science and Practice, 11, 69–94.

219 Brown, T. A., Chorpita, B. F., & Barlow, D. H. (1998). Structural relationships among dimensions of the DSM-IV anxiety and mood disorders and dimensions of negative affect, positive affect, and autonomic arousal. Journal of Abnormal Psychology, 107, 179–192. Buchanan, R. W., & Carpenter, W. T. (1994). Domains of psychopathology: An approach to the reduction of heterogeneity in schizophrenia. Journal of Nervous and Mental Disease, 182, 193–204. Butcher, J. N. (1990). Use of the MMPI-2 in treatment planning. New York: Oxford University Press. Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait–Â�multimethod matrix. Psychological Bulletin, 56, 81–105. Caspi, A., & Roberts, B. W. (2001). Personality development across the life course: The argument for change and continuity. Psychological Inquiry, 12, 49–66. Clark, L. A. (2007). Assessment and diagnosis of personality disorder: Perennial issues and an emerging reconceptualization. Annual Review of Psychology, 58, 227–257. Clark, L. A., & Watson, D. (1991). Tripartite model of anxiety and depression: Psychometric evidence and taxonomic implications. Journal of Abnormal Psychology, 100, 316–336. Cleckley, H. (1941). The mask of sanity. St. Louis, MO: Mosby. Cooke, D. J., & Michie, C. (2001). Refining the construct of psychopathy: Towards a hierarchical model. Psychological Assessment, 13, 171–188. Costa, P. T., Jr., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Â�Factor Inventory (NEO-FFI) professional manual. Odessa, FL: Psychological Assessment Resources. Cox, A. C., Weed, N., & Butcher, J. (2009). The Minnesota Multiphasic Personality Inventory—MMPI-2. In J. N. Butcher (Ed.), Oxford handbook of personality assessment (pp.€250– 276). New York: Oxford University Press. Cronbach, L. J. (1988). Five perspectives on validation argument. In H. Wainer & H. Braun (Eds.), Test validity (pp.€3–17). Hillsdale, NJ: Erlbaum. Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52, 281–302. Dimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., Kohlenberg, R. J., Addis, M. E.,

220 et al. (2006). Randomized trial of behavioral activation, cognitive therapy and antidepressant medication in the acute treatment of adults with major depression. Journal of Consulting and Clinical Psychology, 74, 658–670. Edwards, J. R. (2001). Multidimensional constructs in organizational behavior research: An integrative analytical framework. Organizational Research Methods, 4, 144–192. Emmons, R. A. (1987). Narcissism: Theory and measurement. Journal of Personality and Social Psychology, 52, 11–17. First, M. B. (2005). Clinical utility: A prerequisite for the adoption of a dimensional approach in DSM. Journal of Abnormal Psychology, 114, 560–564. Flemming, E. G., & Flemming, C. W. (1929). The validity of the Mathews’ revision of the Woodworth Personal Data Questionnaire. Journal of Abnormal and Social Psychology, 23, 500–506. Fossati, A., Citterio, A., Grazioli, F., Borroni, S., Carretta, I., Maffei, C., et al. (2005). Taxonic structure of schizotypal personality disorder: A multiple-Â�instrument, multisample study based on mixture models. Psychiatry Research, 137, 71–85. Fossati, A., Maffei, C., Bagnato, M., Donati, D., Caterina, N., & Novella, L. (1999). Latent structure analysis of DSM-IV borderline personality disorder criteria. Comprehensive Psychiatry, 40, 72–79. Fried, R. E., Cyders, M. A., & Smith, G. T. (2008, June). Longitudinal validation of the acquired preparedness model. Paper presented at the annual meeting of the Research Society on Alcoholism, Washington, DC. Garrett, H. E., & Schneck, M. R. (1928). A study of the discriminate value of the Woodworth Personal Data Sheet. Journal of General Psychology, 1, 459–471. Gough, H. G. (1996). CPI manual. Palo Alto, CA: Consulting Psychologists Press. Greene, R. L. (2006). Use of the MMPI-2 in outpatient mental health settings. In J. Butcher (Ed.), MMPI-2: A practitioner’s guide (pp.€ 253–272). Washington, DC: American Psychological Association. Hare, R. D. (2003). PCL-R technical manual. North Tonawanda, NY: Multi-Â�Health Systems. Harpur, T. J., Hakstian, A. R., & Hare, R. D. (1988). Factor structure of the Psychopathy Checklist. Journal of Consulting and Clinical Psychology, 56, 741–747.

CONCEPTUAL ISSUES IN CLASSIFICATION Harpur, T. J., Hare, R. D., & Hakstian, A. R. (1989). Two-Â�factor conceptualization of psychopathy: Construct validity and assessment implications. Psychological Assessment, 1, 6–17. Hough, L. M., & Schneider, R. J. (1996). Personality traits, taxonomies, and applications in organizations. In K. R. Murphy (Ed.), Individuals and behavior in organizations (pp.€31–88). San Francisco: Jossey-Bass. Jang, K., Livesley, W., Taylor, S., Stein, M., & Moon, E. (2004). Heritability of individual depressive symptoms. Journal of Affective Disorders, 80, 125–133. Kane, M. T. (2001). Current concerns in validity theory. Journal of Educational Measurement, 38, 319–342. Kay, S. R., Fiszbein, A., & Opler, L. A. (1987). The Positive and Negative Syndrome Scale (PANSS) for schizophrenia. Schizophrenia Bulletin, 13, 261–276. King, D., Leskin, G., King, L., & Weathers, F. (1998). Confirmatory factor analysis of the Clinician-Â�Administered PTSD Scale: Evidence for the dimensionality of posttraumatic stress disorder. Psychological Assessment, 10, 90– 96. Kraepelin, E. (1981). Clinical psychiatry (A. R. Diendorf, Trans.). Delmar, NY: Scholar’s Facsimiles and Reprints. (Original work published 1883) Krueger, R. F. (2002). Psychometric perspectives on comorbidity. In J. E. Helzer & J.J. Â�Hudziak€ (Eds.), Defining psychopathology in the 21st century: DSM-V and beyond (pp.€41–54). Washington, DC: American Psychiatric Press. Krueger, R. F. (2005). Continuity of Axes I and II: Toward a unified model of personality, personality disorders and clinical disorders. Journal of Personality Disorders, 19, 233–261. Krueger, R. F., & Markon, K. (2006). Reinterpreting comorbidity: A model-based approach to understanding and classifying psychopathology. Annual Review of Clinical Psychology, 2, 111–133. Kupfer, D., First, M., & Regier, D. (Eds.). (2002). A research agenda for DSM-V. Washington, DC: American Psychiatric Association. Laird, D. A. (1925). A mental hygiene and vocational test. Journal of Educational Psychology, 16, 419–422. Leckman, J. F., Grice, D. E., Boardman, J., Zhang, H., Vitale, A., Bondi, C., et al. (1997). Symptoms of obsessive–Â�compulsive disorder.

Issues of Construct Validity American Journal of Psychiatry, 154, 911– 917. Levine, S. Z., & Rabinowitz, J. (2007). Revisiting the five dimensions of the Positive and Negative Syndrome Scale. Journal of Clinical Psychopharmacology, 27, 431–436. Linehan, M. M. (1993). Cognitive-Â�behavioral treatment of borderline personality disorder. New York: Guilford Press. Loevinger, J. (1957). Objective tests as instruments of psychological theory. Psychological Reports, Monograph Supplement 3, 635– 694. Lynam, D. R., & Widiger, T. A. (2007). Using a general model of personality to identify the basic elements of psychopathy. Journal of Personality Disorders, 21, 160–178. Markon, K. E., Krueger, R. F., & Watson, D. (2005). Delineating the structure of normal and abnormal personality: An integrative hierarchical approach. Journal of Personality and Social Psychology, 88, 139–157. Mathews, C. A., Jang, K. L., Hami, S., & Stein, M. B. (2004). The structure of obsessionality among young adults. Depression and Anxiety, 20, 77–85. McGrath, R. E. (2005). Conceptual complexity and construct validity. Journal of Personality Assessment, 112–124. Megargee, E. (2009). The California Psychological Inventory. In J. N. Butcher (Ed.), Oxford handbook of personality assessment (pp.€323– 335). New York: Oxford University Press. Millon, T. (with Davis, R. D.). (1996). Disorders of personality: DSM-IV and beyond (2nd ed.). New York: Wiley. Mizrahi, R., Kiang, M., Mamo, D. C., Arenovich, T., Bagby, R. M., Zipursky, R. B., et al. (2006). The selective effect of antipsychotics on the different dimensions of the experience of psychosis in schizophrenia spectrum disorders. Schizophrenia Research, 88, 111–118. Morey, L. (2003). Measuring personality and psychopathology. In Weiner (Series Ed.) & J. Schinka & W. Velicer (Vol. Eds.), Handbook of psychology: Vol. 2, Research methods in€psychology (pp.€377–406). New York: Wiley. Nakaya, M., Suwa, H., & Ohmori, K. (1999). Latent structure underlying schizophrenic symptoms: A five-Â�dimensional model. Schizophrenia Research, 39, 39–50. Nichols, D. S., & Crowhurst, B. (2006). Use of the MMPI-2 in inpatient mental health settings. In J. N. Butcher (Ed.), MMPI-2: A prac-

221 titioner’s guide (pp.€ 195–252). Washington, DC: American Psychological Association. Palmieri, P. A., Weathers, F. W., Difede, J., & King, D. W. (2007). Confirmatory factor analysis of the PTSD Checklist and the ClinicianÂ�Administered PTSD Scale in disaster workers exposed to the World Trade Center ground zero. Journal of Abnormal Psychology, 116, 329–341. Patrick, C. J., Bradley, M. M., & Lang, P. J. (1993). Emotion in the criminal psychopath: Startle reflex modulation. Journal of Abnormal Psychology, 102, 82–92. Patrick, C. J., Cuthbert, B. N., & Lang, P. J. (1994). Emotion in the criminal psychopath: Fear image processing. Journal of Abnormal Psychology, 103, 523–534. Paunonen, S. V. (1998). Hierarchical organization of personality and prediction of behavior. Journal of Personality and Social Psychology, 74, 538–556. Paunonen, S. V., & Ashton, M. C. (2001). Big Five factors and facets and the prediction of behavior. Journal of Personality and Social Psychology, 81, 524–539. Perry, J. N., Miller, K. B., & Klump, K. (2006). Treatment planning with the MMPI-2. In J. Butcher (Ed.), MMPI-2: A practitioner’s guide (pp.€ 143–164). Washington, DC: American Psychological Association. Raskin, R., & Hall, C. S. (1979). A Narcissistic Personality Inventory. Psychological Reports, 45, 590. Roberts, B. W., & DelVecchio, W. F. (2000). The rank-order consistency of personality traits from childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126, 3–25. Rogers, R. (1995). Diagnostic and structured interviewing. Odessa, FL: Psychological Assessment Resources. Rusch, K. M., Guastello, S. J., & Mason, P. T. (1992). Differentiating symptom clusters of borderline personality disorder. Journal of Clinical Psychology, 48, 730–738. Samuel, D. B., & Widiger, T. A. (2006). Clinicians’ judgments of clinical utility: A comparison of the DSM-IV and five factor models. Journal of Abnormal Psychology, 115, 298–308. Sanislow, C. A., Grilo, C. M., & McGlashan, T. H. (2000). Factor analysis of the DSM-III-R borderline personality disorder criteria in psychiatric inpatients. American Journal of Psychiatry, 157, 1629–1633.

222 Sanislow, C. A., Grilo, C. M., Morey, L. C., Bender, D. S., Skodol, A. E., Gunderson, J. G., et al. (2002). Confirmatory factor analysis of DSM-IV criteria for borderline personality disorder: Findings from the Collaborative Longitudinal Personality Disorders Study. American Journal of Psychiatry, 159, 284–290. Schneider, R. J., Hough, L. M., & Dunnette, M.€ D. (1996). Broadsided by broad traits: How to sink science in five dimensions or less. Journal of Organizational Behavior, 17, 639–655. Serretti, A., & Olgiati, P. (2004). Dimensions of major psychoses: A confirmatory factor analysis of six competing models. Psychiatry Research, 127, 101–109. Simms, L., Watson, D., & Doebbeling, B. (2002). Confirmatory factor analyses of posttraumatic stress symptoms in deployed and nondeployed veterans of the Gulf War. Journal of Abnormal Psychology, 111, 637–647. Smith, G. T. (2005). On construct validity: Issues of method and measurement. Psychological Assessment, 17, 396–408. Smith, G. T., Fischer, S., Cyders, M. A., Annus, A. M., Spillane, N. S., & McCarthy, D. M. (2007). On the validity and utility of discriminating among impulsivity-like traits. Assessment, 14(2), 155–170. Smith, G. T., Fischer, S., & Fister, S. M. (2003). Incremental validity principles in test construction. Psychological Assessment, 15(4), 467–477. Smith, G. T., & McCarthy, D. M. (1995). Methodological considerations in the refinement of clinical assessment instruments. Psychological Assessment, 7, 300–308. Smith, G. T., McCarthy, D. M., & Zapolski, T. C. B. (2009). Elemental constructs and descriptive hierarchies: Toward increased validity of measurement and theory. Manuscript submitted for publication. Smith, G. T., & Zapolski, T. C. B. (2009). Construct validation of personality measures. In J. N. Butcher (Ed.), Oxford handbook of personality assessment (pp.€ 81–98). New York: Oxford University Press.

CONCEPTUAL ISSUES IN CLASSIFICATION Stice, E. (2001). A prospective test of the dualÂ�pathway model of bulimic pathology: Mediating effects of dieting and negative affect. Journal of Abnormal Psychology, 110, 124–135. Verheul, R. (2005). Clinical utility for dimensional models of personality pathology. Journal of Personality Disorders, 19, 283–302. Watson, D. (2005). Rethinking the mood and anxiety disorders: A quantitative hierarchical model for DSM-V. Journal of Abnormal Psychology, 114, 522–536. Watson, D., & Wu, K. (2005). Development and validation of the Schedule of Compulsions, Obsessions, and Pathological Impulses (SCOPI). Assessment, 12, 50–65. Whiteside, S. P., & Lynam, D. R. (2001). The five factor model and impulsivity: Using a structural model of personality to understand impulsivity. Personality and Individual Differences, 30, 669–689. Widiger, T. A., & Boyd, S. (2009). Assessing personality disorders. In J. N. Butcher (Ed.), Oxford handbook of personality assessment (pp.€336–363). New York: Oxford University Press. Widiger, T. A., & Costa, P. T. (2002). Five factor model personality disorder research. In P. T. Costa & T. A. Widiger (Eds.), Personality disorders and the five factor model of personality (2nd ed., pp.€ 59–87). Washington, DC: American Psychological Association. Widiger, T. A., & Lowe, J. R. (2008). A dimensional model of personality disorder: Proposal for DSM-V. Psychiatric Clinics of North America, 31(3), 363–378. Widiger, T. A., & Simonsen, E. (2005). Alternative dimensional models of personality disorder: Finding a common ground. Journal of Personality Disorders, 19, 110–130. Widiger, T. A., Simonsen, E., Krueger, R., Livesley, W. J., & Verheul, R. (2005). Personality disorder research agenda for the DSM-V. Journal of Personality Disorders, 19, 317–340. Widiger, T. A., & Trull, T. J. (2007). Plate tectonics in the classification of personality disorder: Shifting to a dimensional model. American Psychologist, 62, 71–83.

C h a p t e r 11

The Meaning of Comorbidity among Common Mental Disorders Nicholas R. Eaton Susan C. South Robert F. Krueger

ental disorders are typically concepM tualized as discrete, dichotomous entities, and the current nosological system,

the DSM-IV-TR (American Psychiatric Association, 2000), loosely groups them under such rationally derived headings as “mood disorders” and “anxiety disorders” (Watson, 2005). However, high levels of comorbidity between individual disorders— and even across putatively distinct broader groupings—Â�highlight the interrelatedness of many manifestations of psychopathology (Krueger & Markon, 2006). For example, major depressive disorder, panic disorder, and generalized anxiety disorder co-occur more frequently than would be expected by chance alone (Krueger, 1999; Krueger, Caspi, Moffitt, & Silva, 1998; Mineka, Watson, & Clark, 1998). This diagnostic comorbidity raises questions about why certain disorders seem to group together empirically. Do they share at least a portion of their etiologies (e.g., environmental circ*mstances or genetic predispositions)? Can they therefore be better conceptualized as distinguishable manifestations of more general underlying spectra? If so, what factors determine whether a latent spectrum will primarily manifest itself as one type of disorder (e.g.,

depression) instead of another (e.g., anxiety) at a specific point in an individual’s life? This chapter reviews research that attempts to answer these questions. We begin by briefly discussing issues relevant to comorbidity and statistical modeling. The early structural literature regarding the two principal factors that underlie common mental disorders—Â�internalizing (INT) and externalizing (EXT)—are reviewed, followed by replications and expansions of these models. Next, we discuss the link between the INT and EXT factors. Several studies that have examined the invariance of the INT–EXT model in diverse populations (e.g., cultural invariance, gender invariance) are described. We also review the literature on the longitudinal stability of INT–EXT. Finally, we address some future directions for research on structural models of comorbidity.

Representations of Comorbidity There are several ways to think about comorbidity, and it is worthwhile to discuss them briefly because they form the basis of the conceptualization and associated statistical modeling that follow. More in-depth dis223



cussions of comorbidity are presented elsewhere for the interested reader (e.g., Krueger & Markon, 2006; Lilienfeld, Waldman, & Israel, 1994). In general, for our purposes, these representations of comorbidity differ in terms of the type of variables analyzed and corresponding conceptual models. Three types of variables are discussed in terms of comorbidity modeling: categorical variables, continuous variables, and categorical variables that represent dichotomizations of underlying continua. We then explore the difference between co-�occurrence and correlation of disorders in terms of comorbidity, which is illustrated with a real-world example.

study may be given a structured clinical interview to determine whether or not each diagnostic criterion for major depression and generalized anxiety disorder is present or absent in their lives. The numbers of depression symptoms and of generalized anxiety symptoms assessed to be present are totaled separately for each individual, and a covariance between the two symptom counts is calculated. The size of this covariance—or, alternatively, the standardized covariance (i.e., correlation) between the two disorders—can be thought of as the degree of symptomatic comorbidity these two disorders show (e.g., Krueger, Chentsova-Â�Dutton, Markon, Goldberg, & Ormel, 2003).

Comorbidity of Putatively Distinct Categorical Disorders

Comorbidity of Categorical Variables Modeled Continuously

The first way to conceptualize comorbidity is in terms of co-�occurring diagnoses that are putatively distinct. Under the traditional medical model, disorders are typically conceived of as distinct categorical entities, with distinct etiopathophysiologies. For instance, a patient either has a tumor or does not, and that same patient may also either have diabetes or not (Feinstein, 1970). Mental disorder comorbidity is commonly conceptualized in this way, such that a person meets diagnostic criteria for two or more disorders at the same time.

One final means of understanding and modeling comorbidity germane to our purposes should be addressed. This approach combines the categorical diagnostic and dimensional approaches discussed above. Categorical yes–no diagnoses of disorders can be treated statistically as continuous dimensions by utilizing tetrachoric correlations. Tetrachoric correlations are indices of association that assume a liability threshold model (i.e., at a certain threshold point on the liability continuum of a disorder, the disorder switches from being “absent” to being “present”; see Kendler, 1993). Many recent studies examining the comorbidity between, and underlying structures of, common major mental disorders utilize tetrachoric correlations (e.g., Krueger, 1999; Krueger et al., 1998), and therefore an understanding of the use of tetrachoric correlations is beneficial. On the most basic level, the key concept underlying the analysis of tetrachoric correlations is that manifest dichotomous variables (e.g., a yes–no categorical diagnosis of major depression) can be modeled in such a way that they reflect latent dimensions. For example, a researcher might assume that an underlying distribution of disease X symptomatology is continuously distributed. Individuals who do not receive a categorical diagnosis of disease X would fall below a certain diagnostic threshold on this continuum, and individuals who fall at or above the dimension’s diagnostic threshold would receive a categorical diagnosis.

Comorbidity of Continuous Disorders Although the categorical disease system has prevailed for many years in the psychiatric community, more recent research has demonstrated the benefits of moving toward dimensional models of mental disorders (Helzer et al., 2008; Krueger & Piasecki, 2002; Widiger & Samuel, 2005). Under a dimensional system, comorbidity can be thought of differently than in the categorical disease model, and it is conceptualized in this dimensional way in some of the research to be discussed henceforth. Instead of calculating the proportion of individuals who either have major depression, generalized anxiety disorder, or both, dimensional models of comorbidity typically utilize covariation between continuous symptom counts for each disorder. For example, all individuals in a

The Meaning of Comorbidity


Comorbidity as Co-Â�Occurrence versus Correlation “Comorbidity” can be defined in a variety of ways. Two definitions of comorbidity with different implications for understanding disorder overlap are “co-Â�occurrence” and “correlation” (Krueger & Markon, 2006). The simultaneous presentation of two (or more) disorders in one individual does not necessarily indicate that the disorders are related. Instead, individuals experiencing one disorder may have the other disorder by chance. This scenario, co-Â�occurrence, simply implies that, due to base rates of each disorder, a particular number of individuals with one disorder will probably experience the other. The second scenario, correlation, is seen when two disorders relate more strongly than chance (i.e., their base rates) would dictate. Thus, when comorbidity is thought of in terms of disorder–Â�disorder correlation, mental disorders are present simultaneously because of some association between them. We focus on this correlational view of comorbidity in this chapter because the evidence supports the existence of correlations among disorders (i.e., disorders do tend to co-occur more frequently than would be expected by chance).

The following example illustrates the difference between comorbidity as coÂ�occurrence and as correlation; it utilizes the basic categorical diagnosis conceptualization of comorbidity discussed above. Under a categorical nosological system, comorbidity between dichotomous (i.e., yes–no) diagnoses of two disorders can be thought of in terms of two-by-two contingency tables, such as those depicted in Table 11.1. The two tables presented represent the difference between co-Â�occurrence and correlation as bases for comorbidity between major depressive disorder and generalized anxiety disorder. Prevalence rates of the two disorders and observed cases of comorbidity in these tables are based on data from 7,108 individuals in the national probability sample of the Midlife Development in the United States (MIDUS) study (see Brim, Ryff, & Kessler, 2004). These rates have been rescaled to a sample of 1,000 individuals for simplicity of illustration and rounded to the nearest whole number. The upper table represents the overlap of major depression and generalized anxiety disorder if the disorders showed no interrelation, and thus represents comorbidity conceptualized as co-Â�occurrence (and not correlation). The number of individuals in each

TABLE 11.1.╇Two-by-Two Contingency Tables with Data Representing Frequency of€Major Depression and Generalized Anxiety Disorder Diagnoses in Scenarios Where the Disorders Are Independent (Top Panel) and Correlated (Bottom Panel), Based on Rates from a National Probability Sample Generalized anxiety disorder and major depression as independent disorders Generalized anxiety Depression

Absent Present Marginals

Absent 844 129 973

Present 23 â•⁄ 4 27

Marginals 867 133 N = 1,000

Generalized anxiety disorder and major depression as correlated disorders Generalized anxiety Depression

Absent Present Marginals

Absent 858 115 973

Present 10 17 27

Marginals 867 133 N = 1,000

Note.╇ Independent-presentation table represents expected values calculated from population prevalence rates of major depression and generalized anxiety disorder in 7,108 individuals from the Midlife Development in the United States (MIDUS) study, which utilized a national probability sample. Correlated-presentation table represents observed values of MIDUS disorder comorbidity. Values were scaled to N€=€1,000 for simplicity of presentation.


cell are expected values, however, based on the prevalence rates of major depression and generalized anxiety disorder in the sample. For example, we see from the marginal values that 13.3% (i.e., [133/1,000] × 100 = 13.3%) of individuals in the sample experienced major depression, and 2.7% of individuals experienced generalized anxiety disorder. The expected frequencies of individuals in each cell were calculated by using these prevalence rates. We would expect to see approximately 4 individuals out of 1,000 who experienced comorbid major depression and generalized anxiety, on the basis of only these prevalence rates (i.e., .133 × .027 × 1,000 = 3.591, which rounds up to 4 individuals). Remember, this expected value is calculated under the assumption that major depression and generalized anxiety disorder have no association and simply co-occur by chance alone. The lower table represents disorders that are comorbid not only because they coÂ�occur, but also because they correlate; they are associated with one another at greater than chance levels. The frequencies in the lower table, unlike those in the upper table, are observed values, and thus each cell represents the actual frequencies of individuals seen in the MIDUS study. The marginal rates remain the same in the lower table, but the cell values differ from those in the upper table. Of most importance for our purposes are the cells with values in boldface: the numbers of individuals observed who experienced both major depression and generalized anxiety disorder. As mentioned above, we would expect four individuals to experience comorbidity of these disorders if they were unrelated. However, we see 17 individuals who have both disorders; this is more than 400% of the number of individuals with comorbidity we would expect if the disorders were in fact not associated. This marked increase in observed over expected values suggests that the disorders are correlated to some extent, and thus are seen in tandem more frequently than chance levels would dictate because of a relation between them (e.g., they may both be manifestations of the same latent construct, one disorder may “cause” the disorder, and etc.; see Klein & Riso, 1993; Neale & Kendler, 1995). Once we have established, as we have in the example above, that disorders are comor-


bid because of a correlation (and not simply a co-�occurrence level due to base rates), we begin to ask why this correlation exists. Numerous factors could account for this sort of comorbidity. It could be the case that one disorder commonly causes another, and this etiological pattern could result in comorbidity. HIV infection is commonly seen in tandem with AIDS-related medical complications, for instance, because the HIV infection leads to suppressed immunity, which allows for the proliferation of the medical complications. In terms of psychopathology, we are unaware of compelling data for most major mental disorders that indicate a clean, causal etiological pathway from one disorder to the other (Krueger & Markon, 2006). Thus another hypothesis is needed to account for mental disorder comorbidity. A potentially compelling explanation for correlations between disorders is that the disorders are linked by a common latent spectrum. This hypothesis begins to take hold as more and more disorders are shown to interrelate to one another, and thus the presence of a psychologically meaningful underlying factor can be posited. We explore this line of thinking in the next section.

Comorbidity and Common€Factors As discussed above, many mental disorders show observed comorbidity levels that are higher than one would expect by chance alone. A hypothesis to account for the observed comorbidities between many forms of psychopathology is that seemingly distinct mental disorders may be manifestations of common underlying spectra. That is, an unobserved latent factor, or factors, would account for the observed covariations between disorders. This is an application of the common-Â�factor model (Thurstone, 1947), which states that related observed variables are linear functions of one or more common factors and one unique factor per observed variable (these unique factors being typically understood as “psychometric error” or “uniqueness”). This common-Â�factor hypothesis can also be tested statistically; indeed, results of such analyses are the primary focus of this chapter.

The Meaning of Comorbidity


The common-�factor model is depicted in Figure 11.1. In this hypothetical example, there are three manifest (i.e., observed) variables, depicted as rectangles by convention, and one latent common factor that links the three, depicted as ovular. A unique factor loading (denoted l) links each manifest variable to the latent variable. Each observed variable also has a unique factor (denoted e) that accounts for its specific variance, which is the variance in the observed variable not accounted for by the latent factor. Each individual has a score on the latent factor (commonly represented as h; not included in the figure). The observed scores for individuals on any given manifest variable is a linear composite of the factor loading and the unique factor for that variable. For example, the level of major depression observed (which we might denote y MD) is y MD = λMD h + eMD. The common-�factor model utilizes observed variables, such as a symptom count for major depression, in exploratory or confirmatory factor analyses to clarify the way disorders relate to one another and the presence of any latent factors that could connect them. If a researcher had collected data on major depression, generalized anxiety disorder, and panic disorder and noticed strong interrelations among the three disorders, he or she might be curious about what was driving this observed covariation. The model shown in Figure 11.1 depicts an answer to this question. If the model shown in this figure represented the true state of nature, it would

Common factor



Major depression

Generalized anxiety



LPD Panic disorder


Figure 11.1.╇ A theoretical example of the common-factor model.

imply that all individuals have a standing on a latent factor. This latent factor is related to each of the researcher’s three variables in different ways (as represented by the three factor loadings); these three variables’ relations to the common factor are what account for their observed comorbidity with one another. The researcher could then posit different models (e.g., a model with two latent factors) and, by comparing model fit indices, could determine which model best accounted for the observed covariances between disorders in a parsimonious way. This process, known as “confirmatory factor analysis” (as opposed to the more common “exploratory factor analysis”), has been used in the majority of the research we discuss subsequently. The interested reader is referred to Brown (2006) for a solid introduction to the theory and application of these techniques.

The Structure of Common Mental€Disorders: INT and EXT Researchers have long considered the structure of mental health problems, especially in light of the levels of comorbidity between certain disorders. Studies of the structure of psychopathology have an especially strong history in the area of child mental health research. Indeed, this child-Â�oriented research posited the notion that two factors could account well for the comorbidity between many common psychopathological syndromes. Achenbach and Edelbrock (1978, 1984) have reviewed this early thought about the structure of childhood psychopathology in detail. Building upon the foundation of the child psychopathology literature, confirmatory factor analyses of the type described above have recently been applied to questions of psychiatric disorder comorbidity in adults. The results of these studies, by and large, have also indicated that the presence of two broad superordinate factors—INT and EXT—accounts best for the observed covariances between many common major mental disorders (Krueger, 1999; Krueger et al., 1998; see also Achenbach & Edelbrock, 1978, 1984). INT includes such disorders as major depression, generalized anxiety disorder, agoraphobia, panic disorder, and obsessive–Â�compulsive disorder. EXT con-


sists of such disorders as antisocial personality disorder, conduct disorder, and alcohol and drug dependence. It is worth noting that the following analyses have focused on epidemiological data, and thus only common mental disorders are typically included in the statistical models. Disorders with low base rates (e.g., psychotic disorders) are not easily amenable to such analyses.

The General Structure of INT–EXT The structures of INT and EXT are the bestÂ� understood aspects of the higher-order factors that account for comorbidity between syndromes. Krueger and colleagues (1998), drawing on previous research (e.g., Achenbach & Edelbrock, 1978), examined the relations between 10 common DSM-III-R (American Psychiatric Association, 1987) psychiatric disorders (i.e., major depressive episodes, dysthymia, generalized anxiety disorder, agoraphobia, social phobia, simple phobia, obsessive–Â�compulsive disorder, conduct disorder, marijuana dependence, and alcohol dependence). Categorical diagnostic data were available for individuals followed longitudinally at age 18 (n = 930; we refer to this as time 1) and age 21 (n = 937; time 2). Due to the categorical nature of the data, the researchers adopted a liability threshold model, and thus analyzed tetrachoric correlations between the disorder diagnoses. Because this study has been cited by subsequent research in this area, was a relatively early contribution to the literature on this topic, and serves as a model for the reader of most studies to be reviewed subsequently, we discuss it here at some length. Krueger and colleagues (1998) used confirmatory factor analysis to fit three different models to the data at times 1 and 2. Figure 11.2 illustrates several types of the common structural models tested in this and subsequent studies. Because studies frequently differ in the particular disorders (and number of disorders) included in the models that are tested, eight of the disorders examined by Krueger and colleagues were selected to illustrate the models in Figure 11.2. Simple phobia and obsessive–Â�compulsive disorder were not included for the purpose of simplifying these illustrated models. The first model fitted by Krueger and colleagues was


defined as all 10 of the disorders included in their study loading on a single common factor (i.e., a “general psychopathology” factor). Model fit indices indicated that this model fit the data reasonably well but left noticeable room for improvement. (Due to space limitations, these indices of model fit are not explored in detail; interested readers are referred to the original Krueger et al. [1998] paper, as well as the treatment of this topic by Brown [2006]. Briefly, however, fit indices gauge the degree to which a researcherÂ�defined model reproduces the observed relations between disorders, and several of these indices favor more parsimonious models as well.) The second model fit was a two-Â�factor model in which major depressive episodes, dysthymia, generalized anxiety disorder, agoraphobia, social phobia, simple phobia, and obsessive–Â�compulsive disorder were indicators of one latent factor (i.e., INT), and conduct disorder, marijuana dependence, and alcohol dependence were indicators of a second latent factor (i.e., EXT). This model fit the data very well according to fit indices, and it also provided a markedly better fit than did the one-Â�factor model; Figure 11.2 includes a simplified representation (“TwoÂ�factor INT–EXT model”) to illustrate this model’s two factor structure. The time 1 data were fitted to one final model by Krueger and colleagues (1998): a four-Â�factor model (see Figure 11.2’s simplified “Four-Â�factor model”). As mentioned above, DSM-IV-TR places disorders into rationally derived subgroups, and this model represented this approach. In this model, major depressive episodes and dysthymia made up one latent factor (i.e., “affective disorders”); generalized anxiety disorder, obsessive–Â�compulsive disorder, social phobia, agoraphobia, and simple phobia made up another latent factor (i.e., “anxiety disorders”); alcohol and marijuana dependence made up a third latent factor (i.e., “substance dependence”); and conduct disorder served as an indicator for the fourth latent factor (i.e., “antisocial behavior”). Although this model fit the data well, it was clearly overparameterized, due to its large, unfavorable change in fit statistics versus the two-Â�factor model. These results taken together highlighted that a two-Â�factor INT-EXT model accounted most parsimoniously for the observed comorbidity between disorders.

The Meaning of Comorbidity

229 One-factor “general psychopathology” model General psychopathology









Two-factor INT–EXT model











Four-factor model Affective disorders


Substance dependence

Anxiety disorders






Antisocial behavior



Hierarchical INT–EXT model













Figure 11.2.╇ Simplified representations of commonly modeled structures. Ovals represent latent variables; rectangles represent observed variables; straight arrows represent factor loadings; and curved arrows represent factor correlations. Specific variances have been omitted for simplicity. MDE, major depressive episode; Dys, dysthymia; GAD, generalized anxiety disorder; Agor, agoraphobia; Soc, social phobia; Alc, alcohol dependence; Marij, marijuana dependence; CD, conduct disorder.


The time 2 data of individuals at age 21 years showed a similar pattern. The same two-Â�factor model fit the data well. The one-Â�factor model showed a worsening of fit compared to the two-Â�factor model. The four-Â�factor model again yielded a good fit, but fit indices again indicated that it was overparameterized and thus increased fit at the cost of parsimony. Thus the two-Â�factor model was again preferred at this later time point. It was noteworthy that at both time 1 and time 2, the latent variables in the four-Â�factor model tended to mimic the bestÂ�fitting, two-Â�factor model. At time 1, the correlation between the anxiety and affective factors was estimated at 1.00, and the correlation between the antisocial behavior and substance dependence factors was estimated at .89. These correlations were .90 and .72 at time 2, respectively. Thus, even when INT and EXT were split into two separate factors (i.e., when INT was split into anxiety and affective factors, and EXT was split into antisocial behavior and substance dependence factors), those factors tended to correlate very highly together; this indicates the presence of two even higher-level factors to account for these factor–Â�factor correlations (i.e., INT and EXT). This finding can be taken as further support for the hypothesis that the INT and EXT factors account for the observed comorbidity between numerous major mental disorders. A study published the following year (Krueger, 1999) utilized diagnostic data from 8,098 individuals from the National Comorbidity Survey. Unlike the previous study, in which participants were in late adolescence and early adulthood, these individuals ranged in age from 15 to 45 years. Ten disorders were again modeled, and they were for the most part the same as those from the Krueger and colleagues (1998) study; however, panic disorder replaced obsessive–Â� compulsive disorder, antisocial personality disorder replaced conduct disorder, and drug dependence replaced marijuana dependence. Four models were fitted to these disorders, three of which were the same as those in the earlier study (i.e., a one-Â�factor, a two-Â�factor, and a four-Â�factor model, which divided both the INT and EXT factors in two). Exploratory factor analyses of the data, however, had revealed the presence of two subfactors for the INT factor: (1) an “anxious–Â�misery”


factor (referred to as “distress” in more recent literature; e.g., Krueger & Markon, 2006), with indicators of major depressive episodes, dysthymia, and generalized anxiety disorder; and (2) a “fear” factor, with indicators of social phobia, simple phobia, agoraphobia, and panic disorder (see Figure 11.2 for a simplified representation: “Hierarchical INT–EXT model”). The results of model-Â�fitting analyses indicated that the three-Â�factor model best balanced fit and parsimony in the total sample by a wide margin. The sample was then divided randomly in half, and the three-Â�factor model fit best in both halves. Krueger (1999) took these results (and other subsample results to be discussed below) collectively to indicate the superiority of the three-Â�factor model in general—Â�especially due to the fact that this model provided the best fit in five groups, including the large total sample. Overall, this study provided strong evidence for the three-Â�factor model, wherein EXT is unitary and INT is bifurcated into distress and fear factors. We refer to this model hereafter as the “hierarchical INT–EXT model.”

Replications of the INT–EXT Model The two studies described above converged on a two-Â�factor model of psychopathology: The observed covariances between disorders could be conceptualized adequately, but also parsimoniously, by means of the INT and EXT latent factors. Although these studies had utilized different data drawn from different populations, further replication of this finding was warranted. Additional studies on this topic were also necessary to clarify whether INT was best conceptualized as a unitary factor or as having a hierarchical structure, with a higher-order factor consisting of two subfactors (i.e., distress and fear). Several research groups have replicated and extended the INT–EXT model in the past decade. These replications further support the hypothesis that the INT and EXT factors link particular disorders. Vollebergh and colleagues (2001) modeled the structure of nine mental disorders in a community sample from the Netherlands. Diagnoses of DSM-III-R disorders (occurring in the previ-

The Meaning of Comorbidity

ous 12 months, and thus not lifetime diagnoses) were analyzed via tetrachoric correlations and fitted to four models. The first was a one-Â�factor, “general psychopathology” model, in which comorbidity between all nine disorders was accounted for by a single latent dimension. The second model reflected a two-Â�factor structure of INT and EXT. A third model tested the structure outlined by Krueger (1999)—to wit, the hierarchical INT–EXT model. The fourth model placed conceptually similar disorders together. The authors of the study tested the fit of this fourth model by placing disorders into three DSM-based groups: mood disorders, anxiety disorders, and substance use disorders. Due to the longitudinal nature of their data (discussed below), these models were fitted to several subsets of the data. The results of this study served as strong replication for previous INT–EXT research and the hierarchical INT–EXT model. The one-Â�factor model tended to fit worst across all the analyses. Also providing a relatively poor fit was the three-Â�factor, DSM-based model (i.e., mood disorders, anxiety disorders, and substance use factors). The twoÂ�factor INT–EXT model fit better. This INT– EXT model had a significantly better fit than the one-Â�factor model, and did not show a worse fit than the more parameterized DSM-based three-Â�factor model. However, the superior model was the three-Â�factor hierarchical INT–EXT model. Across analyses, this model had the most favorable results on several fit statistics. Other studies have also replicated the INT–EXT structure. Slade and Watson (2006) utilized 12-month DSM-IV and ICD10 (World Health Organization, 1992) diagnoses collected in an Australian community sample. The four models they fitted were the same as those utilized by Vollebergh and colleagues (2001) discussed above, and the results were largely the same. For DSM-IV and ICD-10 disorders, the one-Â�factor model provided the worst fit according to a variety of fit indices. The DSM-based three-Â�factor model fit somewhat better, but generally not as well as the two-Â�factor INT–EXT model. Again, in both diagnostic systems, the hierarchical INT–EXT model showed the most favorable fit index statistics. The individual results of these studies are compelling, but a compilation could prove


even more helpful. Indeed, Krueger and Markon (2006) undertook a meta-Â�analysis of these findings. The authors utilized tetrachoric correlation matrices reported in five major studies of psychopathological comorbidity (Kendler, Prescott, Myers, & Neale, 2003; Kessler, Chiu, Demler, & Walters, 2005; Krueger, 1999; Krueger et al., 1998; Vollebergh et al., 2001), which represented a total of 23,557 participants. Several models were fitted to these data, and these analyses produced results largely congruent with those of previous studies: The one-Â�factor “general psychopathology” model had a poor fit; a two-Â�factor INT–EXT model fit better; and the best fit overall was provided by the hierarchical INT–EXT model. This study, with its very large sample size and aggregation of different data sets, provides the strongest evidence yet for the latent INT– EXT structure. In addition, these results support the bifurcation of the INT factor into two subfactors: distress and fear. Several conclusions can be drawn from the convergence of the studies discussed up to this point. First, the overall latent structure of INT–EXT has been replicated by a number of independent research teams. When compared with other models, such as a one-Â�factor “general psychopathology” model and a three-Â�factor DSM-based model, INT–EXT has been found to account best for the observed covariances between many major mental disorders. Second, the structure of the INT factor seems to be hierarchical. The majority of studies have found that the hierarchical INT–EXT model fits the data best, which indicates that INT is best conceptualized as a higher-order INT factor with distress and fear subfactors.

The Expansion of the INT–EXT Model The results of structural modeling depend to a large degree on the disorders included for analysis. The earlier work on the INT–EXT dimensions (e.g., Krueger, 1999; Krueger et al., 1998) found that a particular set of disorders could be accounted for by INT and EXT. The DSM-III-R disorders that related to INT were major depressive episodes, dysthymia, generalized anxiety disorder, social phobia, simple phobia, agoraphobia, panic


disorder, and obsessive–Â�compulsive disorder. The disorders that related to EXT were alcohol dependence, drug (e.g., marijuana) dependence, antisocial personality disorder, and conduct disorder. Subsequent research has expanded the disorders (DSM-IV and ICD-10) and syndromes associated with the INT factor. Krueger and colleagues (2003) parsed anxiety into “anxious worry” and “anxious arousal,” both of which are associated with INT. They also found that neurasthenia, somatization, and hypochondriasis loaded on the INT factor. Slade and Watson (2006) demonstrated that posttraumatic stress disorder was related to the INT distress subfactor, and they replicated the inclusion of neurasthenia (also on the distress subfactor). Recent work has also indicated that the syndrome of bulimia/ binge-Â�eating disorder loads on the INT factor (Kramer, Krueger, & Hicks, 2008). Bulimia/binge-Â�eating disorder may not be what some readers would intuitively identify as an internalizing syndrome. This study demonstrates that bulimia/binge-Â�eating disorder has an INT factor loading similar in magnitude to that for panic disorder, which is about half of the loading of more traditionally conceived internalizing disorders (e.g., generalized anxiety disorder and social phobia); however, the loading of bulimia/bingeÂ�eating disorder is notably larger than that of hypochondriasis, and more than three times as large as the loading for obsessive–Â� compulsive disorder in this study. Finally, in children and adolescents, it appears that separation anxiety disorder is related to INT (Lahey et al., 2008). The disorders and syndromes included in EXT have remained largely similar for the past decade, and fewer have been identified for EXT than for INT. This is due in part to a greater focus of diagnostic systems on more internalizing-Â�related disorders, as well as inclusion of fewer externalizing disorders in the data sets utilized for the statistical modeling. Adult antisocial behavior has been linked to EXT (Kramer, Krueger, & Hicks, 2008). In children and adolescents, inattention, hyperactivity–Â�impulsivity, and oppositional defiant disorder seem related to EXT (Lahey et al., 2008). A study of categorical versus continuous liability models of externalizing disorders elaborated on the previously identified EXT substance-Â�related


disorders. This study found an externalizing-based interrelation between various forms of substance dependence: nicotine, alcohol, marijuana, cocaine, and other substances (Markon & Krueger, 2005). Finally, some externalizing-�specific behaviors have been linked to EXT, such as relational, destructive, and physical aggression; boredom proneness; low empathy; alcohol, marijuana, and drug use; blame externalization; feelings of alienation; problematic impulsivity; low planful control; impatient urgency; theft; fraud; low honesty, irresponsibility; low dependability; rebelliousness; and excitement seeking (Krueger, Markon, Patrick, Benning, & Kramer, 2007). These studies taken together indicate that EXT, like INT, is a broad factor that underlies many temperamental and psychopathological constructs.

The Association between INT and EXT As we have seen, numerous studies have uncovered and replicated the INT–EXT structure. This supports the notion that INT and EXT account for the comorbidity between many commonly diagnosed major mental disorders. However, up to this point, we have only discussed the covariances between manifest variables (e.g., between measured major depression symptom counts and generalized anxiety disorder symptom counts) and have largely neglected possible covariances between INT and EXT. It is not necessary that these factors be orthogonal (uncorrelated); they can be modeled obliquely (allowed to correlate). As mentioned above in passing, the factors in the four-Â�factor model by Krueger and colleagues (1998) correlated highly. There is more to discuss on the topic of factor correlation: the relation between INT and EXT themselves. Many studies that have replicated the INT–EXT model have reported an association between INT and EXT. Krueger and colleagues reported correlations (standardized covariances) between INT and EXT of .454 and .417 when their participants were 18 and 21 years old, respectively. Other studies have indicated similar degrees of relationship between INT and EXT. For example, Krueger (1999) found an INT–EXT correlation of .51

The Meaning of Comorbidity

in a different sample, which closely mirrored the correlation of .50 found via a large metaÂ�analysis (Krueger & Markon, 2006). Correlations between INT and EXT estimated from two assessment waves in the Netherlands taken 2 years apart (.56 and .66) are consistent with the previously reported estimates as well (Vollebergh et al., 2001). Although the first studies of INT–EXT tended to use DSM-III-R diagnostic criteria, converging INT–EXT relationships have been found in research using other classification systems as well. Slade and Watson (2006) found INT–EXT correlations of .65 when the factors were assessed via DSM-IV diagnostic criteria. The correlation between INT and EXT was almost identical (.61) when the ICD-10 criteria were utilized for assessment. Thus, even across classification systems, the INT–EXT correlation holds at relatively the same level of association. To summarize, across studies, samples, diagnostic systems, and nations, correlations between INT and EXT have converged on a relatively small range of values. This moderate correlation has been reported to range from about .42 to about .66, with .50 seeming to be a reasonable compromise. This result indicates that approximately 25% of the variance (i.e., .502 = .25) in INT is accounted for by variance in EXT, and vice versa. It should be noted, however, that the common statistical parlance of “accounted for” should not be interpreted to mean “accounted for causally,” but instead is a comment on the level of covariation between INT and EXT. In other words, this shared variance does not necessarily demonstrate that one factor causes the other. This moderate relationship has several implications for research, theory, and practice regarding the INT and EXT latent factors of psychopathology. Most clearly, this correlation indicates that individuals with mental disorders do not always (or necessarily even usually) fall into an internalizing or an externalizing group. Indeed, the INT–EXT relation suggests that individuals who have internalizing psychopathology are likely to have some externalizing psychopathology; alternatively, individuals with externalizing behaviors or disorders are likely to have some internalizing ones as well. It is again important to note that this is not necessarily a causal relationship. For clarification of


what this correlation might mean, consider the following example. Externalizing individuals might commit crimes and frequently use alcohol and drugs. This lifestyle, with its associated impulsivity, health problems, and mistreatment of others, could easily leave these individuals feeling alienated. These feelings of alienation and abandonment by family and friends could lead to feelings of depression. In addition, these individuals might have some anxiety arising out of fear of capture and incarceration. Thus, in this case, the internalizing symptoms (mood and anxiety) would “arise from” externalizing behaviors. This is only one interpretation, but it implies a longitudinal sequence of events (EXT problems antedating INT problems) that could be tested to help clarify the nature of the INT–EXT association.

Personality and Temperament as€the Link between INT and EXT One compelling hypothesis to account for the INT–EXT relationship incorporates a role for individuals’ personality and temperament. If particular traits or predispositions are associated with both INT and EXT, these personality and temperamental constructs could serve as the bridge that links the two factors correlationally. The notion that personality and temperament are related to psychopathology has a long history in clinical thought and nosology. Previous editions of the DSM, as well as the most recent edition, have included personality disorders in one form or another. Although personality disorders are often considered to be distinct from other forms of psychopathology (hence their placement on a separate axis from other psychopathology in DSM-IV), there is little compelling evidence to support this division. Indeed, reviews of the literature indicate that personality disorders are typically more similar to other mental disorders than they are different (Krueger, 2005). Normal (nonpathological) personality traits have also been linked to mental disorders (e.g., Trull & Durrett, 2005). Thus the hypothesis that personality and temperament may show associations with the two factors of psychopathology, and in fact may link them, is not without a priori empirical support.


Several studies have tested the role of personality and temperament in internalizing and externalizing disorders. This research has noted that both INT and EXT are associated with negative emotionality (or neuroticism), and that EXT is further linked with disinhibition (e.g., low conscientiousness, low control, high novelty seeking, high disagreeableness; Clark, 2005; Krueger & Markon, 2006). These findings suggest that negative emotionality may account for a large part of the relation between INT and EXT and may predispose one to internalizing and externalizing general psychopathology. The presence or absence of disinhibition, on the other hand, may play a role in determining whether this underlying propensity for psychopathology is manifested as externalizing (i.e., disinhibition present) or internalizing (i.e., disinhibition absent). In addition, several personality traits account for some lower-level disorder–Â�disorder comorbidity. In one study, neuroticism appeared to account, to a strong degree, for the interrelations between the internalizing disorders; neuroticism and novelty-Â�seeking trait levels accounted for a good deal of the observed comorbidity between the externalizing disorders (Khan, Jacobson, Gardner, Prescott, & Kendler, 2005).

The Genetic and Environmental Bases of INT–EXT A number of studies have indicated that the hierarchical INT–EXT model accounts best for observed comorbidity between a variety of psychological disorders, and these factors seem to be related to some personality- and temperament-Â�related predispositions. However, until this point, we have not discussed any studies that addressed the etiology of the INT and EXT factors. Indeed, the previous studies were observational in nature, and did not evaluate the risk factors that account for the observed comorbidity (Kendler et al., 2003). A determination of the origin of these disorders’ comorbidity and higher-order factors could have implications for both conceptualization and treatment. Etiological questions of this sort can be tested empirically by utilizing appropriate models and particular samples (e.g., twins). For readers unfamiliar with behavior genetic


methodology, a brief description may prove helpful. Behavior genetic models, typically conducted with samples of identical (monozygotic) and fraternal (dizygotic) twins, parse variance in the observed variables into genetic, shared environmental, and nonshared environmental factors. Shared environment encompasses nongenetic factors that are shared by twins as they grow up, such as familial socioeconomic status, which serve to make the two twins more similar to one another. Nonshared environment is composed of nongenetic factors that differ between twins, such as one twin playing baseball while the other participates in a school orchestra, that serve to make the two twins less similar to one another. Error is also included in nonshared environment. Finally, it is important to note that the variance accounted for by genes, shared environment, and nonshared environment is estimable in these studies because monozygotic and dizygotic twin pairs differ in the proportion of genes shared between the twins (100% and 50%, respectively); monozygotic and dizygotic twins who were reared together do not differ in the amount of shared and nonshared environment between twins in a twin pair (all twins who were reared together, regardless of zygosity, share 100% of the shared environment and 0% of the nonshared environment, by definition). Kendler and colleagues (2003) investigated the role of genetic, shared environmental, and nonshared environmental factors in the risk for developing many common disorders in a large sample (N = 5,600) of same-sex twin pairs. The authors modeled two different combinations of 10 disorders (major depression, generalized anxiety disorder, animal phobia, situational phobia, panic disorder, alcohol dependence, drug abuse/ dependence, adult antisocial behavior, and conduct disorder) to address the origins of disorders within both INT and EXT, as well as between the distress and fear subfactors of INT. Their analyses utilized an independent�pathway model (a full description of which is beyond the scope of this chapter; interested readers are referred to the original paper). On a basic level, however, this model allowed for the estimation of the effects of genes, shared environment, and nonshared environment both (1) as higher-order, common factors that conferred risk for all relat-

The Meaning of Comorbidity

ed disorders; and (2) as unique effects that conferred risk for each disorder separately. This could be illustrated in Figure 11.1, if the common factor were, for instance, “genetic effect,” and each of the unique variances of the observed variables were disorder-Â�specific genetic effects. The results of this study indicated a strong genetic effect common to INT and EXT. In addition, genes also showed notable specific effects on alcohol dependence and drug abuse/dependence. Shared environment had effects for adult antisocial behavior and conduct disorder. Nonshared environment tended to show strong unique effects for each disorder. These findings