Illness as Metaphor: Sontag Against the Cancer Personality

Illness as Metaphor is a long essay Susan Sontag published in 1978. Her current publisher's page dates it to that year and notes that she returned to the subject almost a decade later, with the outbreak of a new stigmatised disease, in AIDS and Its Metaphors. The two are usually printed together now.

It is not a memoir, and it is not about her. Sontag was being treated for breast cancer while writing it, and deliberately kept herself out of the text. There is no diagnosis scene, no treatment account, no illness narrative.

Said plainly. Sontag died in 2004, of a blood cancer, some years after this essay and a second cancer diagnosis. She wrote nothing here about her own case, and this page does not fill that in beyond what she made public.

What she wrote instead is a work of criticism with a practical target. Her argument is that the metaphors a culture attaches to a disease become part of what patients suffer, and that in the case of cancer those metaphors were actively harmful.

There are no chapters in the usual sense. The essay runs in numbered sections, short and cumulative, closer to a sequence of propositions than to a structured argument with headings.

It opens with the framing that has outlived everything else in the book: the kingdom of the well and the kingdom of the sick, and dual citizenship in both.

The long comparative section sets tuberculosis against cancer. In the nineteenth century, tuberculosis was romanticised. It was a disease of sensitivity, of passion, of interesting people. It happened in the lungs and in the imagination. Cancer, in the twentieth, was moralised in the opposite direction: a disease of repression, of feelings not expressed, of people who are dull rather than passionate.

She then turns on the psychological theories directly. The claim she takes apart is that cancer is caused by a personality type, or by suppressed emotion, or by grief and resignation. She traces the literature that promoted it and shows what it does to a patient: it makes the illness a confession. A patient who accepts the theory has to account for her own tumor, and no honest accounting is possible, because the theory admits no evidence that would refute it.

The military metaphors get their own treatment. Cancer cells invade, colonise and lay siege; treatment bombards. The patient is a battlefield and then a combatant, and the outcome becomes a verdict on their effort. She notes the same vocabulary running in the other direction, in public rhetoric about wars on cancer, and argues that the two reinforce each other.

The essay closes on what she wants instead: the disease described as a disease, without meaning added. She is not asking for optimism. She is asking for accuracy, and treating the two as different things.

It was right, and it was right early.

In 1978, the psychosomatic account of cancer was not fringe. It appeared in reputable books and in clinics, and patients were told that their cancers reflected something they had failed to express. Sontag attacked that at a moment when the science could not yet fully settle it, and the science has since settled it in her favour.

She is also unusually clear about what the harm actually is. It is not that a metaphor is inaccurate. It is that a metaphor of this kind transfers responsibility to the patient, who then has to manage guilt on top of the disease. That mechanism is easy to recognise and very hard to unsee once named. Once a reader has it, they start hearing it everywhere. In the question about whether a diagnosis followed a stressful year. In the compliment about staying positive.

The battle-language critique has aged into common sense. It is now routine to hear patients say they do not want to be described as fighting, losing, or brave, and Sontag is why that vocabulary exists.

And her instinct about mystification was correct in a way she could not have predicted. She argued that the metaphors would recede as the disease became better understood and more treatable. That is broadly what happened. Cancer is now discussed openly, named in obituaries and on television, and described in federal publications in molecular terms. The mystery that made the metaphors possible has thinned considerably.

The cancer described in this essay is the cancer of the 1970s. Her recurring premise is that cancer is largely a mystery, largely fatal, and largely unspeakable. None of those hold in the same way now. NCI states that the likelihood of dying from cancer has dropped steadily since the 1990s, that 5-year survival for breast, prostate and thyroid cancers is 90 percent or better, and that combined 5-year survival is around 69 percent.

She also overstates her case in places. Her position is that illness should carry no meaning at all. That is a rhetorical maximum rather than a workable rule. She never really separates metaphors imposed on a patient from metaphors a patient picks up willingly. Some people do find fighting language useful, and there is nothing wrong with that when it is their own.

The essay is difficult reading. It assumes familiarity with nineteenth-century European literature and moves fast through it. Readers who come to it expecting an illness memoir often stop early.

And it offers nothing practical. There is no information about cancer here, and it was never intended to be a source of any.

This is the substance of the disagreement, and current federal language is unambiguous.

NCI describes cancer as a disease in which some of the body's cells grow out of control and spread to other parts of the body. It then says something plainer. Cancer is a genetic disease, caused by changes to the genes that control how cells work. It names three routes for those changes: errors that occur as cells divide, damage to DNA from harmful substances in the environment such as the chemicals in tobacco smoke, and changes inherited from parents. It identifies proto-oncogenes, tumor suppressor genes and DNA repair genes as the gene types involved, and calls changes in them drivers of cancer. It also notes that cancer-causing genetic changes most often accumulate slowly as a person ages.

NCI's list of risk factors is equally specific: age, alcohol, cancer-causing substances, chronic inflammation, diet, hormones, immunosuppression, infectious agents, obesity, radiation, sunlight and tobacco. Personality, repression and emotional history are not on it.

On attitude, NCI's myths page is explicit. No convincing scientific evidence links a person's attitude to their risk of developing or dying from cancer. It does add that social connection and physical activity may help people cope.

On stress, NCI is careful rather than dismissive. Chronic stress can lead to many health problems. Whether it is linked to cancer is not clear, and studies have produced conflicting results. Where an association does appear, the relationship could be indirect: people under chronic stress may smoke or drink more, and both raise risk on their own. Research on breast cancer specifically, NCI notes, has found no association between perceived stress levels and risk.

That is a fair summary of where the evidence stands, and it is close to what Sontag was arguing for: describe the disease, and stop describing the patient. See what causes cancer, the suppressed anger myth and positive thinking and cancer.

Sontag's essay is also a template for a skill readers still need, because the psychosomatic theories she attacked have not disappeared. They have moved online and changed vocabulary.

The pattern is consistent. A weak claim explains a cancer by something about the person: their unexpressed grief, their negativity, their failure to detoxify. It offers a mechanism in metaphor rather than in biology. It is unfalsifiable, because any patient can be described retrospectively as having repressed something. And it implies that the reader can prevent or reverse cancer through an attitude.

A strong claim looks different. It names a specific exposure or gene. It says how much the risk changes. It admits uncertainty where uncertainty exists. And it comes from a source willing to say what it does not know. NCI's stress page is a good example: it declines to answer the question rather than answering it wrongly.

The cost of the weak version is the one Sontag identified. It gives a frightened person something to do and someone to blame, and the someone is usually themselves. On language specifically, see why some patients dislike warrior and battle language. Sontag wrote it while being treated for breast cancer. That fact is what gives the argument its edge.

It suits readers who want to understand where cancer's vocabulary came from, and readers who have been told, however gently, that their illness says something about who they are. It suits people who prefer argument to narrative.

It does not suit anyone looking for information about a diagnosis. It is a poor choice for the first week after one. The book is abstract and cold by design. It also rests on a 1970s picture of cancer, which would be frightening if mistaken for the current one. Readers who dislike literary criticism will find it hard going, and there is no reason to persist. It is a better book to read a year after treatment than during it.

• Illness as Metaphor and AIDS and Its Metaphors — publisher's page (Macmillan)
• NCI — What Is Cancer?
• NCI — Risk Factors for Cancer
• NCI — Common Cancer Myths and Misconceptions
• NCI — Psychological Stress and Cancer

This page discusses Illness as Metaphor for education. It is not medical advice, and nothing here is a judgement of anyone's real medical care.