6c
Igor Weinberg
First, I would like to thank Conway (this volume) and Hallquist and colleagues (this volume) for their thoughtful commentaries. Hallquist and colleagues introduce two principles – the principle of information compression and the principle of information fidelity. The balance between the principles creates a “space” that optimally conceptualizes the psychopathological phenomena, such as personality pathology. Such conceptualization has a pragmatic nature of informing clinical thinking and intervention.
Conway introduces a comprehensive model – an alternative to the categorical one – that arguably better aligns with the dimensional nature of markers of pathology and with putative risk factors. The model incorporates five dimensions: internalizing, externalizing, somatoform, thought disorder, and detachment. The argument for the dimensionality of markers associated with personality disorders (lack of pathognomonic signs) is by no means unique to personality disorders. No condition in psychiatry is associated with such markers either. The search for “X-ray” tests for psychiatric disorder still continues and is a valuable avenue of study, but no objective pathognomonic signs have been identified yet. Similarly, putative etiological risk factors related to personality disorders are not specific and many people with such factors do not go on to develop personality disorders. This is not unlike other conditions in psychiatry. Many people experience traumatic events, but only a minority goes on to develop posttraumatic stress disorder. Loss is a tragic and ubiquitous part of human existence, but only some people develop a major depressive episode in response to it. In medicine, many people suffer from hypertension, but only a fraction of them suffers a heart attack. This fact, however, never questions the validity of the heart attack as a medical diagnosis.
The appeal of the dimensional model is in its capacity to represent or cover all aspects of the personality pathology. This is consistent with the principle of information fidelity, described by Hallquist and colleagues (this volume). The richness of data introduces the possibility of mathematic modeling and identification of implicit dimensions that lend themselves to concise theoretical model of pathology. Such models have statistical strength and the capacity to explain variability of the psychopathology. In fact, compared to the binary nature of categorical diagnoses, dimensional data are a better statistical predictor. This fact, at least in part, explains why dimensional models sometimes outperform the categorical models on some dimensions of prediction. However, the appeal of abstract dimensions has to be considered against the loss of data resulting from transformation of data and blurring the boundaries between clinically distinct phenomena that statistically load on the same dimension.
From a categorical perspective, the transformation required in a dimensional perspective moves description away from the clinical complexity and the experience-near phenomena into the realm of abstract, theoretical constructs. Statistical techniques that unify variables into one dimension are based on the assumption that variables that correlate with each other are conceptually related. In such a way conceptually unrelated phenomena that correlate with each other might load on the same dimension, creating an illusion of common factors. For instance, distancing is common to schizoid, avoidant, and narcissistic personality disorders. However, the nature of distancing in each disorder is different. Statistically speaking, this invites identification of moderating variables.
Both approaches have their strengths and weaknesses. As reviewed earlier in the chapter, both models have their utility in predicting slightly different aspects of functioning (Morey et al., 2007, 2012). The heated debate as to which of the models has the greatest advantages reminds one of the rift between the Ptolemaic model (Earth-centric system) of the solar system and the model proposed by Copernicus (Solar-centric system). As the history of the personality disorders shows, early categorical models implicitly used dimensions, suggesting that there is a room for a creative integration. In other words, personality disorder models are somewhat akin to the definition of light in physics, which behaves as both particle and wave. Initially, Newton published his famous treatise on optics describing light as particles (Newton, 1988). Even though he cited numerous and quite convincing experiments to support his theory, it was later replaced by a combined one.
Considering the strengths of both, categorical models are more sensitive to unique and specific characteristics of the disorder, while the dimensional models typically are concerned with identification of dimensions that are common to a number of disorders. One way to combine these models is to define dimensions clinically and conceptually, as opposed to statistically. The challenge is to identify dimensions that are close to the clinical experience. For example, the currently proposed alternative model with dimensions of interpersonal relations and self-definition is one such definition. Categorical disorders can be conceptualized in terms of interaction between these universal dimensions and disorder-specific moderators. For example, dependence on others is related to a sense of helplessness and the self-image of being a helpless child in people with dependent personality disorder, the need for self-validation from admiring others in narcissistic personality disorder, and fear of abandonment in patients with borderline personality disorder.
Another challenge is to account for the presence of seemingly contradictory traits that constitute part and parcel of personality disorders. Patients with narcissistic personality disorder present with both grandiose and inferior self-image and some patients with schizoid personality disorder present with aloofness and secret longing for connection. Therefore, rather than thinking in terms of dimensions that connect polar opposites (e.g., superior/inferior, connected/disconnected, dependent/independent), it is important to conceptualize those as orthogonal dimensions. With such conceptualizing, categorical disorders can be organized around these dimensions, moderated by disorder-specific variables. A personality disorder not otherwise specified (NOS) can “fill in the space” around them. Not surprisingly, PDs NOS can be more meaningfully understood in terms of their dimensional representation (Widiger & Mullins-Sweatt, 2005).
The search for clinically meaningful dimensions requires a clinical theory. Statistical analyses are not able to give meaning to clinical phenomena. They can only organize data in a systematic way. One such example of theory is description of personality pathology using the object relation model (Caligor, Kernberg, Clarkin, & Yeomans, 2018). Such an approach incorporates dimensional and categorical models and it incorporates meaningful clinically derived dimensions. This is a promising example, demonstrating the importance of integration of clinical observation with a theory of personality. One can only hope that further efforts will be made to propose similar integrations that conceptualize personality disorders in a meaningful way that is close to the clinical experience of patients and clinicians.
References
Caligor, E., Kernberg, O. F., Clarkin, J. F., & Yeomans, F. E. (2018). Psychodynamic Therapy for Personality Pathology: Treating Self and Interpersonal Functioning. Washington, DC: American Psychiatric Publishing.
Morey, L. C., Hopwood, C. J., Gunderson, J. G., Shea M. T., Skodol, A. E., Grilo, C. M., … McGlashan, T. H. (2007). A comparison of personality disorder models. Psychological Medicine, 37, 983–994.
Morey, L. C., Hopwood, C. J., Markowitz, J. C., Gunderson, J. G., Grilo, C. M., McGlashan, T. H., … Skodol, A. E. (2012). Comparisons of alternative models of personality disorders, II: 6-, 8-, and 10-year follow-up. Psychological Medicine, 42, 1705–1713.
Newton, I. (1988). Opticks: or, a treatise of the reflexions, refractions, inflexions and colours of light. Also two treatises of the species and magnitude of curvilinear figures. Commentary by Nicholas Humez (Octavo ed.). Palo Alto, CA: Octavo. (Opticks was originally published in 1704.)
Widiger, T. A., & Mullins-Sweatt, S. N. (2005). Categorical and dimensional models of personality disorders. In J. M. Oldham, A. E. Skodol, & D. S. Bender (Eds.), The American Psychiatric Publishing Textbook of Personality Disorders (pp. 35–53) Washington, DC: American Psychiatric Publishing.