Phrenology and phenomenology

I recently received a request from a sympathetic but concerned correspondent that I had neglected good papers on phenomenology and needed to put this right in the Journal. I appreciate the concern and would like to stress that we are not averse to receiving papers on phenomenology provided they satisfy all our other exacting requirements. But I was reminded when I received this letter of one of my first experiences in the subject. I was working in a medical post and a man with strange symptoms of abdominal pain and bizarre experiences had been admitted as an emergency. During the night he had woken many times in fear and claimed he could see men who were determined to kill him brandishing knives, but when the nurses reassured him he quickly settled, only to wake again with similar experiences. On the ward round I, as the junior doctor interested in psychiatry, was asked what I made of these symptoms. I waxed eloquently on the important differences between hallucinations, illusions and pseudo-hallucinations, and between autochthonous and secondary delusions. There was a short silence, after which the down-to-earth ward sister (they were still called ‘sisters’ then) said, ‘that's as maybe, all I know is that 'is urine's a funny colour’. Before long it was established the patient had acute porphyria, a diagnosis that explained all his symptoms, and my little seminar on psychiatric phenomenology was seen as no more than a puff of self-important steam. This exemplifies the tension we now all feel in psychiatry about our craft. Are we skirting round the edges of our subject waiting for ‘proper doctors’ to find the causes and take over, or are we doing something important that no other specialists have the skills to do? These skills could now include full awareness of neurocognitive function and brain scan abnormalities. So depression, surely the most common of all mental disorders, might require in the future a diagnostic assessment derived from a neurocognitive test or an MRI scan, and several papers in this (Cole et al, pp. 33–39; Firbank et al, pp. 40–45; Sexton et al, pp. 46–51; Penton-Voak et al, pp. 71–72) and earlier issues ^{Reference Tsopelas, Stewart, Savva, Brayne, Ince and Thomas1–Reference Colloby, Vasudev, O'Brien, Firbank, Parry and Thomas3} hint at this possibility. Are we really suggesting that psychiatrists of the future should be using these approaches, called the ‘new phrenology’ by a sceptical colleague, rather than clinical skills in deciding how to treat their patients? Mary Phillips (pp. 1–3) thinks we should, and her suggestion that machine learning may help at the level of the individual case is something about which we may hear much more.

But old-fashioned phenomenology still has a place, and we can see it in this issue too. The growing evidence that psychosis is like most other mental disorders and lies on a continuum is being challenged by powerful voices ^{Reference David4,Reference Lawrie, Hall, McIntosh, Owens and Johnstone5} and phenomenology may have a part in its resolution. Kelleher et al (pp. 26–32), although noting that some children with psychotic symptoms do develop psychotic disorders later, also note the important finding that these symptoms are more predictive of common mental disorders such as anxiety and depression. The assessment of symptoms is not a straightforward tick-box exercise, as Murray & Jones (pp. 4–6) point out in their perceptive editorial: ‘clinicians should view psychotic symptoms in the same way as they view depressive symptoms: psychological states that require assessment but that, in themselves, do not signify any particular diagnosis or any specific course of action’. The clinician's skills lie in putting symptoms into context, not in putting context into symptoms, as the DSM strategy is prone to do. The importance of context is also emphasised by the review by Shaw et al (pp. 11–19); the now abundant evidence that ethnic density protects against psychosis in particular shows the importance of the environment in the generation and persistence of symptoms. Porphobilinogens and your next-door neighbours have more in common than you might think.

When do labels misinform?

In a reappraisal article Bryant (pp. 9–10) argues the case for prolonged grief to be regarded as a formal psychiatric disorder, and here again phenomenology has a central place. He presents the arguments for the distinction between prolonged grief and depression and on balance Bryant thinks the combination in the proposed DSM-5 definition of ‘intense yearning, emotional pain and preoccupation with the death’ is strong enough to justify the diagnosis. But Arthur Kleinman, one of our most respected correspondents, has argued the opposite in a graphic account of his grief symptoms after the death of his wife. ‘My grief, like that of millions of others, signalled the loss of something truly vital in my life. This pain was part of the remembering and maybe also the remaking. It punctuated the end of a time and a form of living, and marked the transition to a new time and a different way of living’. ^{Reference Kleinman6} For Arthur, prolonged grief was not in any way an illness, but for others it may be; this is where the clinician has to think hard before acting. ‘Medicalising grief, so that treatment is legitimised routinely with antidepressants, for example, is not only dangerously simplistic, but also flawed’, argued the Lancet ⁷ in an editorial linked to Kleinman's article. Of course it is, but there is no reason why a diagnostic label should legitimise any specific treatment. Prolonged grief is no different from depression in labelling terms, as no diagnostic system for depression is watertight ^{Reference Maj8} and persistence of symptoms may indicate incipient bipolar disorder, ^{Reference Forty, Smith, Jones, Jones, Caesar and Cooper9,Reference Smith and Craddock10} another diagnostic exercise that involves soothsaying as much as science. ^{Reference Parker11} Good psychiatric practice needs the bricks of phenomenology for support, but these alone never construct a ‘disorder’; they only create a working structure that needs to be refined and reworked by the clinician, sometimes over and over, before focused intervention, if any, is initiated.