Outcome studies point to the importance of dealing with the full complexity of emotional and social patterns. Numerous researchers (e.g., Blatt, this volume; Norcross 2002; Wampold 2001) have concluded that the nature of the psychotherapeutic relationship, reflecting interconnected aspects of mind and brain operating together in an interpersonal context, predicts outcome more robustly than any specific treatment approach per se . Westen, Novotny, and Thompson-Brenner (2004 and this volume) have presented evidence that treatments that focus on isolated symptoms or behaviors (rather than personality, emotional, and interpersonal patterns ) are not effective in sustaining even narrowly defined changes. Shedler and Westen, Dahlbender and colleagues, Blatt, and others (this volume) have developed reliable ways to measure complex patterns of personality, emotion, and interpersonal processes that constitute the active ingredients of the psychotherapeutic relationship. A number of recent reviews (e.g., Fonagy's and Leichsenring's in this volume) demonstrate that in addition to alleviating symptoms, psychodynamically based therapeutic approaches improve overall emotional and social functioning.

The PDM was created though a collaborative effort of the major organizations representing psychoanalytically oriented mental health professionals; namely, the American Psychoanalytic Association, the International Psychoanalytical Association, the Division of Psychoanalysis (39) of the American Psychological Association, the American Academy of Psychoanalysis, and the National Membership Committee on Psychoanalysis in Clinical Social Work. Their presidents formed a steering committee and recommended members to serve on work groups to construct this classification system.

The diagnostic framework formulated by the PDM work groups systematically describes:

• Healthy and disordered personality functioning ;
• Individual profiles of mental functioning, including patterns of relating, comprehending and expressing feelings, coping with stress and anxiety, observing one's own emotions and behaviors , and forming moral judgments ;
• Symptom patterns, including differences in each individual's personal, subjective experience of symptoms.

The Psychodynamic Diagnostic Manual adds a needed perspective to existing diagnostic systems. In addition to considering symptom patterns described in existing taxonomies, it enables clinicians to describe and categorize personality patterns, related social and emotional capacities, unique mental profiles, and personal experiences of symptoms. It provides a framework for improving comprehensive treatment approaches and understanding both the biological and psychological origins of mental health and illness.

Rationale for the PDM

A clinically useful classification of mental health disorders must begin with an understanding of healthy mental processes. Mental health comprises more than simply the absence of symptoms. It involves a person's overall mental functioning, including relationships; emotional depth, range, and regulation; coping capacities; and self-observing abilities. Just as healthy cardiac functioning cannot be defined simply as an absence of chest pain, healthy mental functioning is more than the absence of observable symptoms of psycho pathology. It involves the full range of human cognitive, emotional, and behavioral capacities.

Any attempt to describe and classify deficiencies in mental health must therefore take into account limitations or deficits in many different mental capacities, including ones that are not necessarily overt sources of pain. For example, as frightening as anxiety attacks can be, an inability to perceive and respond accurately to the emotional cues of others-a far more subtle and diffuse problem-may constitute a more fundamental difficulty than periodic episodes of unexplained panic. A deficit in reading emotional cues may pervasively compromise relationships and thinking and may itself be a source of anxiety.

That a comprehensive conceptualization of health is the foundation for describing disorder may seem self-evident, and yet the mental health field has not developed its diagnostic procedures accordingly. In the last two decades, there has been an increasing tendency to define mental problems primarily on the basis of observable symptoms, behaviors, and traits, with overall personality functioning and levels of adaptation noted only secondarily. There is increas ing evidence, however, that both mental health and psychopathology involve many subtle features of human functioning, including affect tolerance, regulation, and expression; coping strategies and defenses; capacities for understanding self and others; and quality of relationships. Mounting evidence from neuroscience and developmental studies supports the position that mental functioning, whether optimal or compromised, is highly complex. To ignore mental complexity is to ignore the very phenomena of concern. After all, our mental complexity defines our most human qualities.

Over the past 30 years or so, in the hope of developing an adequate empirical basis for diagnosis and treatment, the mental health field has progressively narrowed its perspective, focusing more and more on isolated symptoms. The whole person has been less visible than the various disorder constructs on which researchers can find agreement. Recent reviews of this effort have raised the possibility that such a strategy was misguided. Ironically, emerging evidence suggests that oversimplifying mental health phenomena in the service of attaining consistency of description (reliability) and capacity to evaluate treatment empirically (validity) may have compromised the laudable goal of a more scientifically sound understanding of mental health and psychopathology. Most problematically, reliability and validity data for many disorders are not as strong as the mental health community had hoped they would be. Allen Frances, Chair of the DSM-IV American Psychiatric Association Task Force , recently acknowledged (Spiegel 2005) that the desired reliability has not been obtained.

In a recent commentary in the Journal of the American Medical Association , Paul McHugh (2005) pointed out that medicine has moved beyond simply describing symptoms to categorizing disorders according to the nature of the functional impairment and etiological factors (if known). Contending that the classification of mental health disorders may have gone too far in a purely descriptive direction (overlapping categories, excessive co-morbidities, etc.), thereby compromising efforts to improve our understanding and treatment of psychopathology, he recommended that the classification of mental disorders also re
flect the quality and degree of functional impairment and, where possible, etiology.

The mental health field has a long history of describing symptom patterns. As in the development of many fields, this effort began with pioneers who made meticulous observations and discovered common clusters of patient complaints, symptoms, and behaviors. In attempting to make progress in describing naturally occurring patterns, however, much recent research has been a mixed blessing. On the one hand, carefully constructed questionnaires and structured interviews have led to more reliable judgments about symptom patterns and have facilitated research into what belongs in a pattern, including its antecedents and course. On the other hand, fixed definitions (often made by clinical consensus) and incomplete data have impeded the improvement of descriptions of naturally occurring patterns.

A patient may experience a number of symptom patterns. Many such patterns have long been observed to overlap. In the DSM and ICD systems, the use of fixed definitions and strict criteria (e.g., four out of six, not three out of six, items on a diagnostic checklist) forces an artificial separation of conditions that are frequently related. Symptoms that may be etiologically, phenomenologically, or contextually interconnected are described as co-morbid conditions, as if these discrete problems coexist more or less accidentally in the same person, much as a sinus infection and a broken toe might coexist. Assumptions about discrete, unrelated, co-morbid conditions are rarely justified by compelling data such as clear genetic, biochemical, and neurophysiological distinctions between syndromes. The cut-off criteria for diagnosis are often arbitrary decisions of committees rather than conclusions drawn from the best scientific evidence.

The development of the PDM reflects our concern that mental health professionals may have uncritically and prematurely adopted methods from other sciences instead of developing empirical procedures appropriate to the complexity of the data in our field. The intent of those who moved the DSM and ICD classifications in the direction of specifying discrete, externally observable disorders was to build a stronger foundation for the diagnosis and treatment of psychopathology. This was a worthy project. Now, however, it is time to take a hard look at the phenomena with which mental health professionals regularly deal and adapt the methods to the phenomena rather than vice versa (see Part III, review essays by Blatt, Shedler, Westen).

The PDM attempts to do this. Because the current terminology for symptoms and their groupings comes from a long and intellectually serious tradition, we employ the descriptions of symptoms and patterns of symptoms used in the currently prevailing taxonomies, the DSM-IV-TR and ICD-10, systems that represent a valuable history of careful observation and description. In the most recent versions of the DSM and ICD systems, however, some of the more subtle features of many basic symptom patterns have been lost. Most notably, despite the fact that it is usually the patient's subjective suffering that brings him or her to treatment, a full description of the patient's internal experience of the symptoms is often absent.

All approaches to assessment and treatment rely at least in part on patients' reports of their thoughts, feelings, and behaviors. (Does the patient feel depressed? Anxious? Does the patient hear voices? Think about suicide?). Therefore, despite the fact that mental health professionals are always inevitably dealing with the elusive world of subjectivity, we require a fuller description of the patient's internal life to do justice to understanding his or her distinctive experience. We are hoping that with more elaborated depictions, we can make more progress on understanding naturally occurring patterns. The rapidly advancing neuroscience field, including genetic studies, can only be as useful as our understanding of the naturally occurring basic patterns of mental health and pathology. We cannot expect our colleagues in genetics to separate the apples and oranges for us. If we do not properly separate them, we will continue to frustrate the search for underlying biological pathways and common experiential etiologies.

Even in general medicine, instances in which etiological factors are fully understood are rare. Most commonly, we are at the level of functional rather than etiological explanation. Neoplastic disorders, for example, are often thought to be understood etiologically, but we are still searching for the causes of many malignancies, as we attempt to comprehend the relationship between genetic, environmental, and, in some instances, viral and other infectious processes. We are nonetheless able to describe various malignancies in detail in terms of their functional characteristics. Both in general medicine and in mental health, progress in understanding the functional nature of disorders should eventually facilitate a greater understanding of etiological factors. Functional and etiological understanding together provide the fullest basis for diagnosis and treatment.

In general, there is a healthy tension between the goals of capturing the complexity of clinical phenomena (functional understanding) and developing criteria that can be reliably judged and employed in research (descriptive understanding). It is vital to embrace this tension by pursuing a step-wise approach in which complexity and clinical usefulness influence operational definitions and inform research. A scientifically based system begins with accurate recognition and description of complex clinical phenomena and builds gradually toward empirical validation. Relying on oversimplification and favoring what is measurable over what is meaningful do not operate in the service of good science .

In addition , we are learning that when therapists apply manualized treatments to selected symptom clusters without addressing the complex person who experiences the symptoms and without attending to the therapeutic relationship that supports the treatment, therapeutic results are short-lived and rates of remission are high (Westen, Novotny, et al. 2004; Hilsenroth, Ackerman, et al. 2003; Baumann, Hilsenroth, et al. 2001; Stiles, Agnew-Davies, et al. 1998). A recent meta-analysis of outcomes of manualized treatments for targeted symptoms found that symptomatic improvement often did not persist and that fundamental psychological capacities involving the depth and range of relationships, feelings, and coping strategies did not show evidence of long-term change. In a number of studies these critical areas were not even measured (Westen, Novotny, et al. 2004).

At the same time, process-oriented research has demonstrated that essential characteristics of the psychotherapeutic relationship as conceptualized by psychodynamic models (the working alliance, transference phenomena , and stable characteristics of patient and therapist) are more predictive of outcome than any designated treatment approach per se . Most dynamically oriented clinicians pay careful attention to the therapeutic relationship, noting interpersonal patterns, feelings, coping strategies, and other indicators of mental processes. Although under-researched for decades, several recent meta-analyses and reviews reveal evidence of the efficacy of psychodynamically based treatments (see Part III review essays by Fonagy and Leichsenring, as well as Hilsenroth, Ackerman, et al. 2003; Leichsenring & Leibing 2003) .

Although the psychoanalytic tradition, or depth psychology, has a long history of examining overall human functioning in a searching and comprehensive way, the diagnostic precision and usefulness of psychodynamic approaches have been compromised by at least two problems. First, until fairly recently, in attempts to capture the range and subtlety of human experience, psychoanalytic accounts of mental processes have been expressed in competing theories and metaphors that have, at times, inspired more dis
agreement and controversy than consensus. Second, there has been difficulty distinguishing between speculative constructs on the one hand, and phenomena that can be observed or reasonably inferred on the other. Where the tradition of descriptive psychiatry has had a tendency to reify "disorder" categories, the psychoanalytic tradition has tended to reify theoretical constructs.

In recent years, however, having developed empirical methods to quantify and analyze complex mental phenomena, depth psychology has been able to offer clear operational criteria for a more comprehensive range of human social and emotional conditions (see (Lingiardi, Shedler, et al. 2005, and Part III review essays by Blatt, Dahlbender, Westen, Shedler). The current challenge is to systematize these advances with a growing body of rich clinical experience in order to provide a widely usable framework for understanding and specifying complex and subtle mental phenomena.

A psychodynamically based system highlights the processes that contribute to emotional and social functioning. Early in its history, psychodynamic theories speculated about etiological factors. As in all fields of medicine, however, clinicians and researchers quickly learned that the etiologies of psychological disorders are more complex than initial observations and theory had suggested. Consequently, psychodynamic models have moved toward functional understanding of psychopathologies, with the expectation that such understanding will guide the identification of etiological patterns.

In light of all this, the PDM addresses the full range of mental functioning. Its approach to personality disorders identifies patterns that capture the quality and degree of impairment in such basic capacities as forming and sustaining relationships; regulating affects, moods, and impulses; and carrying out essential human functions in family, educational, and work settings. Its profile of mental functioning specifies components of these functional patterns. Its approach to symptom patterns is to add to the DSM descriptions an understanding of the patient's unique internal experience of his or her problems.

The PDM uses a multi dimensional approach to describe the intricacies of the patient's overall functioning and ways of engaging in the therapeutic process. It begins with a classification of the spectrum of personality patterns and disorders, then offer s a "profile of mental functioning" covering in more detail the patient's capacities, and finally considers symptom patterns, with emphasis on the patient's subjective experience.

Dimension I: Personality Patterns and Disorders

The PDM classification of personality patterns takes into account two areas: the person's general location on a continuum from healthier to more disordered functioning, and the nature of the characteristic ways the individual organize s mental functioning and engage s the world.

This dimension has been placed first in the PDM system because of the accumulating evidence that symptoms or problems cannot be understood, assessed, or treated in the absence of an understanding of the mental life of the person who has the symptoms. For example, a depressed mood may be manifested in markedly different ways in a person who fears relationships and avoids experiencing and expressing most feelings and in an individual who is fully engaged in all of life's relationships and emotions. There is not just one clinical presentation of the artificially isolated phenomenon known as depression.

Dimension II: Mental Functioning

The second PDM dimension offers a more detailed description of emotional functioning-the capacities that contribute to an individual's personality and overall level of psychological health or pathology. It takes a more microscopic look at mental life, systematizing such capacities as information processing and self-regulation; the forming and maintaining of relationships; experiencing, organizing, and expressing different levels of affects or emotions; representing, differentiating, and integrating experience; using coping strategies and defenses; observing self and others; and forming internal standards.

Dimension III: Manifest Symptoms and Concerns