Saturday, 18 January 2014

Diagnosing the DSM

Rachel Cooper
I’m a philosopher working mainly on conceptual problems surrounding the DSM (the main classification of mental disorders used by psychiatrists). This post looks at how a certain type of epistemic imperfection – ambiguity – can be the strategically useful, by considering the definition of mental disorder in the most recent edition of the DSM (DSM-5, published 2013).  

During its development, two distinct definitions of “mental disorder” were drafted for the DSM-5. The first was an iteration of the previous DSM definition and took mental disorder to be a value-laden concept, i.e. it claimed that disorders are necessarily harmful. The second definition characterized mental disorder as mental dysfunction, and aimed to offer a value-free account. The working groups said that a decision between these two definitions would be made at a later date. Note that philosophers of medicine generally hold that value-laden and value-free accounts of disorder are competing accounts. To be consistent one has to opt for one or the other.

However, when the DSM-5 was finally published, the definition of ‘disorder’ sought a compromise between the two draft definitions. According to the DSM-5,

"A mental disorder is a syndrome characterized by clinically significant disturbance in an individual’s cognition, emotion regulation, or behaviour that reflects a dysfunction in the psychological, biological, or developmental processes underlying mental functioning. Mental disorders are usually associated with significant distress or disability in social, occupational, or other important activities…" (emphasis added, DSM-5, p. 20)

This definition bears no resemblance to the sorts of definition via necessary and sufficient conditions that philosophers tend to propose; in any particular case it is unclear whether harm is needed prior to diagnosis or not. Philosophers tend to think that clarity is a good thing, and from such a stance the DSM-5 definition of mental disorder looks to be a very bad definition.

However, a different tradition has it that in certain contexts ambiguity may be useful, and rather than being refined away, may be retained, or even created. In The Concept of Law (1961), H.L.A. Hart discusses the “open texture” of language, and how it enables judges to employ sensible discretion in the application of rules. The rule may say “No vehicles in the park” but the ambiguity of “vehicle” means that the judge is left free to make sensible and context-specific decisions about bikes, wheelchairs, skateboards and so on. Specifying in advance exactly what would count as a “vehicle” would not be helpful.

The definition of “mental disorder” included in the DSM-5 can perhaps best be understood as answering such goals. In failing to provide necessary and sufficient criteria for a condition to count as a disorder it places no constraints on what might be added to the classification in the future. And, from a certain point of view this might be considered a useful feature. Those who speak ambiguously now, leave themselves free to claim what they wish later.

This article has been posted on behalf of Rachel Cooper, who is a senior lecturer in philosophy at Lancaster University.

Her new book Diagnosing the Diagnostic and Statistical Manual of Mental Disorders will be published by Karnac later this year.


  1. Interesting article! Doesn't the DSM-V warn against using its symptom(s) criteria for the purposes of diagnosis and treatment without accompaniment by "clinically significant distress" in the guidelines to the manual, though?

    In my view, and from my experience working with DSM-V, the "usually associated with" perhaps ought to be taken as flavour text. What is "mental disorder" according to the DSM-V seems to be fulfilment of symptom(s) criteria coupled with distress sufficient to bring it to the attention of health professionals in the first place. In full, mental "abnormality" and "dysfunction" only count as mental disorder if they produce this distress, and distress only counts as mental disorder if it's accompanied by "abnormality" and "dysfunction".

    This chimes with the rejection of attenuated psychosis syndrome (APS) for DSM-V. APS would be diagnosed in absence of "clinically significant distress" (but with presence of some sub-threshold psychotic symptoms) in order to intervene early and reduce distress later on. This doesn't fit with the distress + abnormality model, so it does not occupy a proper category in the DSM-V.

    Would love to hear your thoughts!

  2. This reply has been posted by the Epistemic Innocence team on behalf of Rachel Cooper.

    Thanks for the comment!

    You're right to say that the text that counts for health professionals is that contained in the actual DSM diagnostic criteria, rather than the introductory material. Many sets of diagnostic criteria do specify that the diagnosis can only be made if there is clinically significant distress or impairment (this is true of the criteria for depression and autism spectrum disorder, for example, so no-one should receive these diagnoses in the absence of harm). Still, some diagnostic criteria, for example for tic disorders, do not require distress or impairment.

    Although few health professionals read the introductory material, the definition of mental disorder that's included in the definition does play a role in debates about possible revisions to the DSM. In the past those proposing a new addition would have to argue that the condition caused harm. With the revised and ambiguous definition it's unclear whether this is any longer necessary. Someone advocating for the inclusion of a condition that isn't clearly linked to harm can argue that the revised definition of disorder doesn't require harm in all cases. Someone arguing against such an inclusion can point out that most conditions recognised as disorders are linked to harm. The debate could go either way; the revised definition places no restrictions on new proposals.

  3. "Although few health professionals read the introductory material, the definition of mental disorder that's included in the definition does play a role in debates about possible revisions to the DSM. In the past those proposing a new addition would have to argue that the condition caused harm."

    Bang on the money. I think this is really important - thanks for highlighting this!

  4. For readers who are interested in Rachel's book, a review has recently appeared on Metapsychology Online.


Comments are moderated.