Prof. Rachel Cooper of the Lancaster University Department of Politics, Philosophy & Religion delivered a presentation on the struggle to define the concept of disorder to the university’s Language, Ideology and Power Research Group.

Cooper began by declaring that she was there for help. She had hit a wall in her research and was hoping that by presenting her problem to a multidisciplinary group of researchers, with their different methodologies, she may be able to break through. She then described her two main research goals: the use of conceptual analysis in the history of medicine; and the problems caused by shifts in the meaning of concepts over time.

She explained conceptual analyis as a tool used by philosophers to establish a basis for categorisation. Regarding the concept of disorder, she said that it acted as an umbrella concept for a variety of things: injuries, disease, disability (though recent disability activism had challenged the latter’s inclusion). Conceptual analysis begins with a first stab at detailing the necessary and sufficient conditions for the application of a given term. This is followed by testing with actual and hypothetical cases in order to expose flaws in the concept. For example, a first stab might list a condition for disorder as is treated by doctors, but the example of doctors also delivering babies would expose that as not sufficient.

In another example, Cooper asked us to consider what knowledge means. Beginning with a condition of true belief, she gave the example of someone who reads tea leaves to predict the result of an election, but who gets lucky and coincidentally is correct. We would not say that that person had knowledge of the result of the election from their tea leaves, so additional qualifiers are needed. The aim, said Cooper, of conceptual analysis is to be descriptive; to unearth a tidied-up account of a pre-existing concept that, when explained to a layperson would result in them agreeing that that was what they meant.

Cooper delineated two prevailing views on the concept of disorder:

• Christopher Boorse’s value-free view of disorder as biological dysfunction, harmful or otherwise; and
• the value-laden view, subscribed to by Cooper, that dysfunction must be detrimental to be a disorder.

The test case of these two views, she said, was homosexuality. Whilst we do not yet know the exact causes, it is possible that it results from a biological dysfunction. However, we generally accept that it is not harmful, so in her view it would not be considered a disorder. The American Psychological Association voted in line with this view in 1973, and from 1980–2011 all psychologists subscribed to the value-laden view of disorder.

This changed in 2013, with publication of the DSM-5. In 2011 a draft had included two definitions of mental disorder—one value-free, one value-laden—and a note stating that a decision on which to adopt would be made later. The two definitions were the result of the two working groups whose work is combined to create the DSM coming from differing traditions: one from the tradition of homosexuality being removed as a disorder, the other from disability studies and its view that many disorders are not necessarily harmful but are still disorders. The aim, suggested Cooper, was to align the definition of mental disorder more with that of physical disorder, and the two views on disorder could be summed up as one of impairment and one of disability. As another example, Cooper highlighted the shift in discussions around tic disorders, usually benign and often not even noticed by those with them, being described as such.

This shift presented a problem for conceptual analysis. Terms change meaning over time, and the unearthed necessary requirements might not always be necessary and are therefore weak connections. Cooper said there had been three types of academic response to this. The first was to argue against seeking a definition of disorder entirely. The second, embodied by the work of Sally Haslanger, was a new revisionist project that aimed for concepts to be prescriptive, rather than descriptive. That is, concepts should aim for better science or better politics. The example was given of the definition of parent, which has shifted in recent times to mean a child’s guardian more than exclusively their biological antecedants (e.g., school parents’ evenings that invite carers where parents are absent, adoptive parents, etc.). Haslanger’s view is that philosophical study can help to facilitate a linguistic shift already underway.

Cooper’s preference, option three, is to develop methods that are robust in the face of future conceptual changes. This means supporting a claim with multiple, redundant conditions, such that the invalidation of one does not invalidate the entire concept. She cited the example of mountain climbers on unstable rocks using multiple anchor points, and said that in many cases these additional conditions were already present, but implicit. For example, whilst a directly-stated condition for disorder to apply might be that is is harmful, indirect ones could be that someone with a disorder is unlucky (e.g., what is happening to them does not happen to most people) and that it is internal (e.g., the disorder is not a result of social or environmental pressures).

During the Q&A, many (perhaps to be expected, considering the research group hosting the presentation) fixated on the power dynamics of the DSM, which both describes and affects concepts of mental illness, as well as the impact of such definitions on people’s ability to access treatment, exam extensions, etc., as well as the effect on their own self-image. Someone asked why we don’t just have two concepts, harmful and non-harmful disorders, and Cooper replied that this represents the revisionary project she had touched on, which she had concerns about—the idea of best tends to depend on the project itself and the politics of who is doing it, she said, and these sorts of things are very hard to predict.

Prompted by her mention of caffeine addiction in a prior edition of the DSM, I asked if she thought there might be a value in the DSM cataloguing all non-expected modes of existence, even benign ones, so that practitioners could still find what is up with their patients and see that it is no cause for concern. She replied that there was a huge range of possible non-normal functioning that a person could experience, so such a catalogue would be prohibitively large. I’m not sure I’m convinced; the DSM-5 is already almost 1,000 pages long, and there’s nothing stopping it being split into volumes. I also asked what her definition of harmful was with the following example: the disorder is a harmful dysfunction definition could be accepted by two psychiatrists and yet still lead to one considering homosexuality a disorder and not, based on whether they thought it was harmful. Additionally, three psychiatrists could be of the opinion that it is harmful for different reasons: one might think it is a sin, one might think it increases the risk of certain STIs and one might think it leads to social exclusion.

Cooper replied that harmful is a term that is very difficult to define in the abstract, but easy to define in practice. This willingness to selectively use duck typing seemed strange from someone’s whose entire project is based around avoiding such a technique, but she did offer a valuable way of assessing different reasonings: consistency. For example, for someone to claim that homosexuality is a sin and therefore harmful, they would have to demonstrate that they adhere to all the other commandments of their religion. That said, this does still seem to lead to a situation in which homosexuality (to continue the example) is not just considered to be a disorder in one culture and not another, but actually is simultaneously a disorder and not a disorder, and in which somebody can be cured of their disorder by moving to a different country. I suppose that political construction of disorder is what Foucault was trying to get at.