There has been controversy lately over the new edition of the American Psychiatric Association’s new version of their Diagnostic and Statistical Manual (DSM 5). Among a number of critiques, the most-forwarded on the Internet seeems to be that of the National Institute of Mental Health, which recommended that researchers applying for grants not use the DSM 5 – because, rather than classify diseases and diagnoses based on etiology, it does so based on symptomatology.

The narrative that the psychiatrists are trying to write is a triumphal progression from anecdote to empiricism, from Freud’s couch to the colorful images of the fMRI. 

Unfortunately, the example psychiatrists would like to follow – medicine – is not as simple as all that. When we diagnose, we often apply a heuristic which relies not just on results of tests and images but also on the patients’ symptomatic report. Call it “bacterial pneumonia” all you want, but a patient without symptoms can’t be diagnosed with pneumonia even if there’s an infiltrate on the X-ray – and even if a patient is symptomatic, we don’t go and culture the bacteria from the lung tissue except in relatively serious circumstances. The point here is not the details of what makes a pneumonia, merely that we do not have a magic on-off switch within the body that we can examine when making a diagnosis. Psychiatrists know this as much as we do, which is why (to internists’ secret shame) we don’t have anything like a unified diagnostic and statistical manual – and if we did, it would be just as controversial as the DSM.