restricted access Should Clinicians’ Views of Mental Illness Influence the DSM?
In lieu of an abstract, here is a brief excerpt of the content:

Should Clinicians’ Views of Mental Illness Influence the DSM?

clinicians, DSM, values, psychopathology, science

The relationship between clinicians and the DSM is complex. Clinicians are the primary intended audience of the DSM. However, as Widiger (2007) pointed out in his commentary, there is a tension associated with trying to meet the clinical goals of the DSM and also trying to optimize the scientific goals of the DSM. Most commentators would agree that the recent editions of the DSM (from the DSM-III onward) have favored scientific values (e.g., Sadler 2005). Some commentators have taken sides and aligned themselves with either clinicians or with the science of classification. For instance, Garb (2005) has argued that clinicians should receive incentives to align themselves more closely with the science of the DSM. In contrast, Westen and Shedler (1999) have favored a clinician-based approach to the DSM and have argued that an ideal classification should be based on the prototypes that clinicians use to conceptualize diagnostic categories. Perhaps an extreme suggestion on how to deal with this tension between the values of clinicians and the values of scientists is to adopt the ICD-10 solution and create two versions of the next DSM. One version would be written for clinicians and another for researchers (Phillips et al. 2003).

Widiger (2007) wondered whether the two of us as authors are on the fence about how the views of clinicians should affect the DSM. As clinical psychologists who are both clinicians and researchers, we understand his comment, especially in light of our attempt in our article not to overgeneralize from the time-consuming, labor-intensive, yet methodologically limited study we reported in our major paper.

Nonetheless, our answer to the question asked in the title of this response, “Should clinicians’ views of mental illness influence the diagnostic manual?” is a clear and unequivocal yes. In our view, too much of the research literature stimulated by the recent DSMs has primarily used the diagnostic criteria in these DSMs to specify patient samples and then perform various experiments to study the DSM categories. We favor a different approach to research which is consistent with a comment once made by the philosopher of science, Karl Popper: “scientific theories originate from myths, and . . . a myth may contain important anticipations of scientific theories” (1963, 39). We believe that clinicians, with their years [End Page 285] of experience dealing with the ambiguities, the emotionality, and the chaos of psychopathology, have formed complex conceptual systems to try to understand the mysteries of abnormal behavior. We strongly support increased research attempts that will use clinicians as subjects and that will attempt to understand how clinicians use diagnostic concepts to understand patients.

Although research on clinicians has not attracted a large amount of attention in the field, there is a long history of such research studies. For instance, in the 1970s, Overall used multivariate techniques to create graphical maps of how clinicians represented relationships among diagnoses (Overall et al. 1977). Widiger has used a similar strategy in some of his research (Widiger et al. 1987). In the early 1980s, a view of concept formation from cognitive psychology was used by Cantor et al. (1980) to look at psychiatric diagnoses as representing prototypes. Horowitz and colleagues extended this research. Blashfield and his colleagues had a grant to use the prototype model to study personality disorders and performed research investigating both case histories (Blashfield et al. 1985) and analyses of diagnostic criteria (Blashfield and Breen 1989; Davis, Blashfield, and McElroy 1993). Westen and colleagues have attempted to reconceptualize DSM personality disorder categories using the prototype model (Westen and Shedler 1999). Ahn and her colleagues (Ahn et al. 2006; Kim and Ahn 2002) have taken more recent applications of ideas from cognitive psychology to study how clinicians use and understand diagnostic categories. Finally, Flanagan and colleagues (Flanagan and Blashfield 2006; Flanagan, Keeley, and Blashfield, in press; Flanagan, Miller, and Davidson, submitted) have used qualitative and quantitative data analysis techniques to try to understand the diagnostic concepts held by experienced clinicians.

Besides empirical studies of how clinicians use diagnostic concepts, we also support the increased interest in...