On winking at the facts, and losing one's Hare: value pluralism and the harmful dysfunction analysis

Disorder is partly a value concept; conditions that do not cause harm (e.g., benign angiomas, dyslexic dysfunction in a preliterate society where reading is neither taught nor valued) are not disorders. However, disorders are not merely disvalued conditions. Most negative conditions, from ignorance and lack of talent to shortness in an aspiring basketball player, are not disorders. Nor does need for treatment imply disorder, as in abortion and delinquency (1).

What is the factual criterion that distinguishes disorders from other disvalued conditions? Fulford et al never address this question. By merely winking at the facts in disorder judgments rather than opening their eyes to them, the authors fail to correctly apply Hare's analysis. According to Hare, a value concept like "good strawberry" has three elements: first, a factual domain to which the value is applied ("strawberry"); second, a value term ("good"); and third, a factual basis for the value (e.g., taste and color), which may vary. The authors, focusing on values, ignore the domain specification.

I analyze disorder as harmful dysfunction (HD), with "harmful" a value term based on social judgments, and "dysfunction" a factual term meaning failure of biologically designed functioning (2). The HD analysis fits Hare's model well. Dysfunction is the factual domain, harm is the value applied to the domain, and there are implicit cri- teria for harm (e.g., suffering, disability) that may vary from culture to culture.

However, "harmful" reflects social, not individual, values. For example, in a literate society, a person who does not value reading still has a dyslexic disorder if incapable of learning to read due to a brain dysfunction; and, in a society valuing reproductive capacity, a sterile individual has a disorder even if he or she does not want children.

The authors argue that values are more diverse regarding mental than physical conditions, so mental disorder is more value-laden. They propose that diagnosis be negotiated with the patient based on the patient's values. In effect, they make "harm" relative to individual rather than social values in response to growing value pluralism that casts doubt on the very notion of social values.

The proposal inflates the value ladenness of mental disorder and mischaracterizes its consequences. Once the factual domain of disorder is specified as dysfunction, value divergence narrows considerably. For example, people may differ in how much they value joy or hate sadness in response to life's vicissitudes or as philosophical attitudes, but there is much less difference in how they feel about true depressive disorder in which something has gone wrong with the mind so as to continually generate painful sadness unrelated to actual losses or philosophical insights.

Moreover, the effects of value variation on diagnosis are severely limited because of the dysfunction requirement. Values can disqualify a dysfunction from being a disorder, but values cannot make a non-dysfunction into a disorder.

The authors' proposal erodes the distinction between disorder, anchored in biological facts, and disvalued conditions in general. The HD analysis suggests an alternative approach. First, limit diagnosis to dysfunctions based on facts, setting aside values to the extent possible. Second, further refine the current approach of using individual values in treatment decisions. Patient values may dictate not treating a dysfunction or treating a non-dysfunction (e.g., cosmetic surgery, abortion).

Mental medicine faces a dilemma. As society becomes more pluralistic, the assumption of shared social values underlying disorder's "harm" component becomes problematic; yet adopting the authors' individual - values approach to harm undermines the distinctiveness and usefulness of the concept of disorder. The answer to pluralism, I suggest, is to make diagnosis more scientific and make treatment decisions more explicitly value laden. Helpfully, the value ladenness of disorder turns out to be much less than the authors suggest once its factual component is understood.