Nikhil Rao in his blog OK so I’m not really a cowboy has an interesting article about the perils of taking diagnosis too seriously. He makes his point by noting that his tendancies to show schizoid and schizotypal traits is more a function of his chronic pain than a reflection of his personality.

He argues that diagnosis should not be about social acceptability or conformity but should reflect an more meaningful underlying process. In other words, diagnosis should be in part based on widely accepted constructs about internal psychological processes that reflect on etiology and cognitive and neurological function as well as behavior and symptoms.

The problem with schizoid and schizotypal PD I think is shared with anti-social PD. The concepts originated in psycho-dynamic theory representing internal states. The first two have been thought by some to be traits related to schizophrenia. Schizoid PD relates to a preferance for social isolation. Schizotypal PD relates to systematic errors in thought process like ideas of reference and magical thinking that appear to be a simple more common pattern similar to a thought disorder. Anti-social PD is most closely related to criminal attitudes and behavior.

When looking at a group of people in prison, most would have a number of traits suggesting anti-social traits. However, there are many reasons for a person to behave anti-social. Some grow up in families or subcultures that accept criminal behavior. Traditionally, the concept of anti-social comes from the psychopathy. Psychopaths are thought to have no or at least an unconventional conscience. Neurological studies suggest these people have underdeveloped or "immature" brains. The concept of anti-social PD adds so many people to the catagory that it deflects attention away from the causes of criminality towards the behavior itself. The value of concepts is largely rooted in their contribution to understanding. I see anti-social PD as muddying the waters.

The American Psychiatric Association adopted the Diagnostic and Statistical Manual (DSM) as it's diagnositic standard. It has steadily moved diagnosis towards a set of measurable criteria that is necessarily behavioral. Cognitions are not measurable in the traditional scientific sense. By taking us to “measurable”, the DSM de-emphasized the causes and internal processes that created the diagnostic construct. At the same time, the DSM effectively added a significant number of people who would not fit the constructs internal process and etiology. I would argue that many now fit a particular diagnosis that do not belong there.

The risk here is that we forget we are also working with internal states rather than just behavior and symptoms. Diagnosis should have some relationship to theory and methods of treatment. Based on DSM, we treat symptoms and change behavior which may not be addressing the underlying problem that caused the behavior. If diagnostics is taught solely within DSM nosology, we turn out diagnosticians who know little more about how to help someone, not qualified mental health professional who have an in depth understanding of psychology and the art of treatment.

Diagnoses don't reflect real phenomena, they represent conceptual models that are by and large still, in a way, experimental. Theoretically, diagnosis should lead us to treatment methods. While an experienced clinician sees some guidance in the diagnosis, it's still more about art than science. The DSM tries to make diagnosis more scientiific but in the end guts important content and understanding.