Psychiatric diagnosis: Nominalism versus realism

To the Editor

Allen Frances has made a persuasive case that our addiction to diagnostic inflation has spawned an ever-growing assortment of classifications of questionable clinical and research validity, culminating in the DSM-5 (Frances, 2013, 2014). He has also drawn attention to our professional drift towards diagnostic bracket creep, wherein we attempt to drag the ‘slings and arrows’ of normal human experience into the realms of psychopathology (Frances, 2013).

In his letter ‘How many psychiatric diagnoses fit on the head of a pin’ (Frances, 2014), Dr Frances claims that in relation to psychiatric diagnoses, history tells us that the realists got things wrong because we “…construct our realities rather than just sensing them”. In my view, discarding realism in favour of nominalism risks throwing the baby out with the bathwater. When Dr Frances claims that “ ‘Schizophrenia’ is just a name, not a thing” (Frances, 2014), he implies that schizophrenia is really just a construct of language, and by extension does not exist outside of our minds. The logical consequence of nominalism is not only the denial of the reality of delusions and hallucinations, but also of the very existence of neurotransmitters, synapses, and, ultimately, of the brain itself. But by just using the term ‘schizophrenia’, Dr Frances is attempting to communicate an idea about which we have a shared understanding (and about which our patients have a lived experience), notwithstanding the subjectivity of our own judgements and the heterogeneous presentations of the illness. Are we as likely to entertain nominalist explanations for neuropsychiatric disorders that possess more clearly measurable and localising biological correlates, such as the dementias and other neurodegenerative conditions? I think that the answer is no.

Despite apparently endorsing a nominalist worldview, however, Dr Frances concludes by hoping that “The mental disorders … will be replaced by closer approximations of reality as we develop better tools and more knowledge” (Frances, 2014). His call to align our observations with what really exists is consistent with the Aristotelian theory of correspondence and is perhaps an unwitting concession to realism. Regardless, I am sure that he would agree that future diagnostic criteria should be reduced in number, simplified, and wherever possible anchored in empiricism and common sense.