Sadness is one of the small number of human emotions that have been recognized in all societies and in all time periods. Some of the earliest known epics, such as “The Iliad” and “Gilgamesh,” feature protagonists’ intense sadness after the loss of close comrades. Likewise, anthropological work across a great range of societies clearly describes emotions of sadness that develop in response to frustration in love, humiliation by rivals or the inability to achieve valued cultural goals.

Even primates display physiological and behavioral signs after losses that are unmistakably similar to sadness among humans. There is little doubt that evolution designed people to have a propensity to become sad after such situations.

Depressive mental disorders also have been known for as long as written records have been kept. Writing in the fifth century B.C., Hippocrates provided the first known definition of melancholia (what we now call depression) as a distinct disorder: “If fear or sadness last for a long time it is melancholia.” The symptoms that Hippocrates associated with melancholic disorder – “aversion to food, despondency, sleeplessness, irritability, restlessness” – are remarkably similar to those contained in modern definitions of depressive disorder.

Like Hippocrates, physicians throughout history recognized that the symptoms of normal sadness and depressive disorder were similar. Depressive disorders differed from normal reactions because they either arose in the absence of situations that would normally produce sadness or were of disproportionate magnitude or duration relative to whatever cause provoked them.

Such conditions indicated that something was wrong with the individual, not with his environment. Traditional psychiatry thus adopted a contextual approach to diagnosing a depressive disorder. Whether a condition was diagnosed as disordered depended not just on the symptoms, which might be similar in normal sadness, and not just on the condition’s severity, for normal sadness can be severe and disordered sadness moderate, but on the degree to which the symptoms were an understandable response to circumstances.

The distinction between contextually appropriate sadness and depressive disorders remained largely unchanged for two and a half millennia. But the psychiatric profession abandoned this distinction in 1980, when it published the third edition of its official diagnostic manual, the DSM-III.

The definition of major depressive disorder became purely symptom-based. All conditions that display five or more of nine symptoms – including low mood, lack of pleasure, sleep and appetite difficulties, inability to concentrate, and fatigue – over a two-week period are now considered depressive disorders.

The sole exception is “uncomplicated” grief-related depression. Symptoms otherwise meeting the DSM criteria are not considered disorders if they arise after the death of an intimate, do not last more than two months, and do not include certain particularly severe symptoms. Yet comparable symptoms that arise after, say, dissolution of a romantic relationship, loss of a job, or diagnosis of a life-threatening illness are not excluded from diagnosis of disorders.

The DSM-III’s confusion of normal intense sadness and depressive mental disorder, which persists to the present, emerged inadvertently from psychiatry’s response to challenges to the profession during the 1970s. A powerful group of research psychiatrists was dissatisfied with the definitions of depression and other common mental disorders in the earlier, psychoanalytically influenced diagnostic manuals.

These earlier definitions separated feelings of sadness proportionate to contextual loss from those excessive to their contexts, and defined only the latter as disordered. But they also assumed that unconscious, unresolved psychological conflicts caused depression. In order to abolish this unwarranted psychoanalytic assumption, the researchers abandoned the attempt to distinguish natural from disordered conditions by context or etiology and assumed that all conditions that met the symptom-based criteria were disordered.

The new definition of depression has resulted in extensive medicalization of sadness. Parents whose child is seriously ill, spouses who discover their partners’ extramarital affairs, or workers unexpectedly fired from valued jobs are defined as suffering mental disorders if they develop enough symptoms to meet the DSM criteria. This is so even if the symptoms disappear as soon as the child recovers, the spouses reconcile or a new job is found.

The medicalization of sadness has proven to be of tremendous benefit to the mental health and medical professions. Millions of people now seek professional help for conditions that fall under the medicalized, overly inclusive definition of depression. Indeed, depression is now the most commonly diagnosed condition in outpatient psychiatric treatment.

The medicalization of depression has proven to be even more profitable for pharmaceutical companies, whose sales of anti-depressant medications have soared. While it is impossible to know what proportion of these people are experiencing normal sadness that would go away with the passage of time or a change in social context, it is almost certainly very high.

It would not be hard for psychiatry to develop a more adequate definition of depressive disorder that de-medicalizes natural emotions of sadness. The diagnostic criteria could simply extend the current bereavement exclusion to cover conditions that develop after other losses and that are not especially severe or enduring.

Such a change would acknowledge what humans have always recognized: intense sadness after loss is a painful and perhaps inevitable aspect of the human condition, but it is not necessarily a mental disorder.

Allan V. Horwitz is professor of sociology at Rutgers University. Jerome C. Wakefield is professor of social work and professor of the conceptual foundations of psychiatry at NYU School of Medicine. Their column is printed here courtesy of Project Syndicate,a voluntary, member-based institution;