The Magazine

The Anatomy of Melancholy


In Kingsley Amis's novel Stanley and the Women, the protagonist's son begins to exhibit strange, unpredictable, violent behavior. Stanley, the protagonist, consults a psychiatrist, who gives him some waffle about "affective disorders" and the like. Frustrated, Stanley goes for a second opinion to an old friend, a doctor who has seen everything and is old enough not to give a damn about keeping up with fashionable jargon. After examining the young man this oldster delivers his diagnosis: "Your son is mad."

Presumably that doctor was not a heavy user of the DSM. That is the Diagnostic and Statistical Manual of Mental Disorders, the authoritative text approved by the American Psychiatric Association for telling us when we need the attentions of APA members, and for what. This month, the APA announced proposed changes to the next issue of the DSM, due to come off the presses in 2013. This will be the fifth version of the DSM, the previous four having come out in 1952, 1968, 1980 (partially revised 1987), and 1994 (partially revised 2000).

You can read up outlines of the diagnostic categories and proposed changes on the DSM-5 website. I was surprised at how many categories there are -- nearly 300 in the DSM-IV. (The APA seems to have switched from Roman to Arabic numerals.) Some of them sound terrifying: Intermittent Explosive Disorder, for example -- "Several discrete episodes of failure to resist aggressive impulses that result in serious assaultive acts or destruction of property."

Others make one think of people one knows who in spite of it all, seem to function well enough without psychiatric help, like Muscle dysmorphia -- "The belief that one's body build is too small or is insufficiently muscular." Of course, all are qualified with the condition that they must "cause clinically significant distress . . . or impairment," but there's a lot of room for subjective judgment in that.

You'll be hearing a lot about the DSM-5 over the next three years. Diagnoses once in vogue will be thrown out or merged with others; new ailments will be identified and named. There is already a fuss about the proposal to fold Asperger's syndrome (awkwardness in dealing with people) into autism; "internet addiction" is being mooted as a distinct mental abnormality, to cheers from parents of teenage boys worldwide; so is "hypersexual disorder," though many of us don't really see where the "clinically significant distress . . . or impairment" kicks in with that one.

While I am sure that professional psychiatrists do their best, and study hard and long for those diplomas on their walls, it's hard to keep at bay the feeling that there is something dubious about the whole profession. Unlike other physicians, psychiatrists don't deal with disorders that present themselves in objective reality as lumps, fevers, rashes, fractures, or exsanguinations. There are no identifiable agents causing most of the conditions listed in the DSM -- no viruses, bacteria, or fungi.

There are a few physical conditions of which the same can be said -- migraine, for instance, which is real enough to sufferers, yet which displays nothing to an observer and has no known agent. Most physical ailments aren't like that, though; most psychiatric ones, are. If I have appendicitis, a broken leg, or psoriasis, I'm in no doubt about the matter. When do I know I have "generalized anxiety disorder"? When a shrink tells me so, after consulting his DSM.

The DSM seems to do what it can to inflame these suspicions of bogosity. Conditions drift in and out of its pages according to the whims of fashion (and, one cannot help suspect, the subtly-applied pressures of health insurers and drug companies). Fifty years ago you suffered from a disgusting and antisocial perversion; today you are "gay." Fifty years ago you were a heavy smoker; today you are in the fell grip of "Nicotine-Use Disorder." Quite large parts of the DSM seem to record not treatable conditions but the ebb and flow of lifestyle fads. This has undesirable consequences, especially in the area of making excuses for oneself. Recall that politician of my home state who pleaded "non-filer syndrome" when it was revealed that he hadn't paid his taxes for years.

Other parts, to be fair, describe real pathologies. An old friend of mine recently went mad, and has fled back to his home country to escape the legions of FBI and CIA agents who have been watching his every move and steaming open his mail. His unfortunate wife and child are left here without support in the ruins of their lives. Madness is real, and comes in different styles. There is some diagnosing and treating to be done, for sure. The problems arise from mission creep. If we can diagnose and treat the mad, why not the merely unhappy or eccentric?

Where is the borderline? Who is genuinely crazy, in real need of treatments that actually exist? Jack Nicholson tells his fellow psych-ward inmates in One Flew Over the Cuckoo's Nest: "You're no crazier than the average asshole out walking around on the streets," and by that point in the movie we're ready to agree with him. Society at large in fact did agree with him: that movie helped drive the larger movement of de-institutionalization that emptied out mental hospitals in the 1970s ... onto the streets. The results have not been happy.

Will the APA, this time around, get a better grip on the non-mad portion of psychiatry's clientele than they did in their previous four attempts? The sorry history of mental doctoring in the 20th century doesn't encourage confidence. There was, for example, the strange cult of psychoanalysis, which probably never cured anyone of anything, and which was elbowed aside at last by psychopharmacology, which seems to do little better.

Possibly I'm a jaundiced observer. In prolonged spells of unhappiness myself when young, I sought both kinds of relief, with no discernible effect. At last I came to the conclusion offered by Macbeth's doctor:


Canst though not minister to a mind diseased,

Pluck from the memory a rooted sorrow,

Raze out the written troubles of the brain,

And with some sweet oblivious antidote

Cleanse the stuffed bosom of that perilous stuff

Which weighs upon the heart?


Therein the patient

Must minister to himself.

However, if you lift up your eyes from the bickering over DSM classifications and the feebleness of psychiatric medicine, something remarkable and stirring comes into view. The APA's struggles with diagnostic terms of art are one small aspect of the great human-sciences project of our age: the project to obtain true understanding of ourselves, of the human organism, and of the causative factors that make us what we are.

The first phase in this tremendous project is to map the human phenome, the collection of all our phenes. A phene (from Greek phainein, "to show") is a single trait with some likely genetic input. Examples of phenes are height, eye color, or susceptibility to diabetes. The human organism in all its generality is a collection of phenes like these. The actual value of each phene, the phenotype, varies among individuals -- tall or short, eyes blue or brown, more or less susceptible to diabetes. The collection of all your phenotypes is . . . you.

How many phenes are there? How do they vary? Why do they vary? That's what we are striving to understand. It will be the work of decades, but it's well under way.

The attempt to answer that first question has in fact been under way for tens of thousands of years. For all that time, human beings have been observing each other and separating out different characteristics -- different phenes -- for scrutiny and evaluation. The ability to do this observation well -- to know when to challenge a rival, when to back down, who'll be a steady companion on the hunt, who can be trusted to look after your kids -- has obvious survival value. When civilization came up, philosophers and religious thinkers took up the challenge, trying to add to our understanding of ourselves.

Some of this accumulated understanding has been incorporated into the modern human sciences. The "Big Five" model of human personality, for example, developed from studies of the adjectives we use to describe each other. Researchers trawled through dictionaries gathering up words descriptive of personality and putting them into broad common categories, drawing on the observational wisdom of ages, as encoded in our very language.

For all that intellectual effort and all those millennia of sifting by natural selection, the results are still pretty meager. Merely physical phenes are no problem: we can all agree on someone being fat, cross-eyed, or hairy, and the underlying biological causes are yielding to methodical inquiry. It's the BIPs -- Behavior, Intelligence, Personality -- that are tricky. Some gross features we can all agree on: he's irritable, she's shy, he's smart, she's arrogant. As the wrangling over the DSM shows, though, we have no general agreement on when a BIP phenotype needs professional attention.

The ultimate objective of all the research is to link as much of the phenome as possible to the genome. That is going to be immensely difficult. The gene-phene link is certainly not one-one, nor even one-many; it is many-many. Even quite ordinary phenes -- height, for example -- are influenced by large numbers of genes. Contrariwise, a single genetic abnormality can cause (for example) both sickle-cell anemia and resistance to malaria.

And then, in the path from gene to phene, the environment kicks in. Gather up a population of adults with similar life histories and measure their heights. The differences are almost entirely genetic. So are most differences at the population level: the shortest population in the world, the pygmies of the Ituri forest in northeast Congo, live only a few hundred miles from the tallest, the Nilotics of Sudan, in much the same climate. The difference is genetic. But look: North and South Koreans have exactly the same population genetics, yet North Koreans are on average shorter, because of poor nutrition.

It's Genetic!

Figure 1: Ituri Pygmies -- It's genetic

Sometimes the pathway is so indirect that it's hard to know whether to allow genetics into the causal chain at all. Consider the phenotype "one-leggedness." No doubt there are rare genetic abnormalities that cause a person to be born with only one leg, but the great majority of such conditions are caused by accidents. My house fell on me in an earthquake, and I lost a leg -- where is the genetic contribution to that?

On the other hand, suppose I lost a leg as a result of falling asleep on the railroad tracks. That seems to be an equally gene-free situation. But what if the reason I was sleeping on the tracks is that I was stone drunk, as I am every night, on account of being a chronic alcoholic? Alcoholism is strongly suspected of having some genetic input. Or what if I had in fact been on the wagon for years, but fell off it when my wife left me? What is the cause of my exhibiting the one-leggedness phenotype? My genes? My wife? The Long Island Railroad? The Jack Daniels Corp.? My own dumb self?

From such knotty tangles of nature and circumstance, science must somehow extract the component of nature. The news in so far is that there is no escaping nature. Work on that Big Five model of personality, for example (it's not the only such model, by the way) suggest that all five main clusters of personality phenes are heritable at about the 50 percent level -- not a thing that will surprise anyone who has raised children, but another boot in the rear end for the "psychodynamic" model ("I'm like this because of the way I was raised!") that underlay psychoanalysis, and earlier editions of the DSM.

A few decades further on it may be that our grandchildren, when applying for a job, or seeking a marriage partner, will have to present a chip containing their phenome -- all the traits and peculiarities that together comprise the uniqueness of an individual human being. Possibly a second chip containing the genome will also be required, but I doubt it. Phenes are what matter, not genes. It is phenes, for example, not genes, that are "visible" to natural selection.

The ruminations of the DSM compilers are, in a crude and approximate way, helping to inch us forward to that improved understanding, to a full mapping of the human phenome. "The proper study of mankind is man." For millennia that study was the province of art, philosophy, and politics. Now the biologists and statisticians are making some real inroads. If the DSM compilers will take in what has been learned, after a couple more revisions we may have a book worth paying $70 for.