The dead bodies depicted as data in Gould’s graph orient mortality, too, though shorn of fleshy references. But the stories of those who died before or after the eight-month median—those in some way described by the graph—dissipate into the universal, timeless curve. The stories lent to the prognosis will come to be inhabited by other people—others who will wear those stories in their own ways, leave their own imprints. The search for oneself in this chart will always end in disappointment, for numbers are not shoes. A number will not mold to your arches; it will not record the shape of your life.
The graph abstracts the lives it represents, painting Gould as a victor against the odds rather than as one who literally vanquished those who landed to his left. In reading the graph, we can all hope that we might find ourselves on the right side of the graph, even though we know this is logically impossible. Yet justifying one’s own life in the numeric death of the collective makes a dangerous bedrock for hope. Fickle adulterers, numbers make love with the generations who move through them. These data have no allegiance.
Statistics render another sort of violence by abstraction. Gould’s disease is virtually always caused by asbestos exposure; according to historians, the disease exists only because of a massive, decades-long cover-up by the asbestos industry. In different circumstances, mesothelioma might easily have never existed, which would have led to a different curve entirely (a flat one). The spread of the disease was enabled, arguably, by the impersonality of aggregates—it is as if a gun was shot into a crowd, and fifty years later someone from that crowd keeled over and died. Given this cloak of anonymity (who was it who had the gun all that time ago?), a would-be assassin might well be more likely to shoot.
Gould’s graph offers a seemingly objective view of the natural course of a cancer, rather than a glimpse into the politics of diagnosis—a politics that could easily fill a museum in the nation’s capital. Ovarian cancer, for example, is known as a particularly aggressive form of cancer because women often die relatively soon after diagnosis. But like most cancers, life chances have to do with how far the cancer has advanced at diagnosis, and so the label aggressive masks the fact that patients and doctors may have ignored subtle symptoms until the cancer advanced to a stage at which it was no longer treatable. In other words, skipping over the causes of cancer gives it an apolitical mystique. Statistical aggregations provide a logic through which bodies become interchangeable numbers for which nothing need be felt, neither guilt, nor pleasure, nor horror. They enable prediction.
Donne’s bell can neither notice nor toll for a statistic. Donne can’t rationalize survivorship. Gould aims to comfort us with the possibility that in the coin of life in prognosis, we could each flip tails, even if some of those in a group of one hundred will invariably stare at the nickeled eyes of Thomas Jefferson. The Holocaust shoe project refuses statistical logic altogether; it’s not about the six million who died, but about each one of those people who died.
Built of the dead—people we’ve never met nor could meet—survival prognoses contain homogeneous units with only one variable: alive or dead. These Frankenstein numbers do more than scare each of us. They become something sinister: they feed on our friends’, acquaintances’, and enemies’ deaths, and they will feed one day on each of our deaths, just as they feed now on our lives.16 The statistics that offer the promise of beating the odds also evacuate the politics of prognoses.
STAND UP AND BE COUNTED
After my treatment, I went to the hospital to see if I carry a cancer gene. The genetic counselor congratulated me on my negative result; I had won the genetic roulette and could avoid a horrid conversation with my offspring about what I had done to them. But a strange chat still ensued. The genetic counselor told me she was pretty certain I am a carrier of something, they just don’t know what. She then showed me a chart that detailed my two sisters’ patterns of cancer risk, which increased a couple of percent each year until the chart ended when each turned seventy-nine. How weird to see my little sisters’ lives as a bar chart on the desk of a genetic counselor who knows nothing, absolutely nothing, else about them. I tried to picture my younger sister at seventy-nine. Would she still live in Vancouver? Would I get to see her? Would she still be my little sister? If I died, would she still be my sister? Then I chided myself for my narcissism.
My other sister, younger still, has an even higher risk for cancer. I couldn’t get my head around it until I realized that it is all about time: the older sister has lived cancer-free for eight extra years, and so has weighed in on one side of the calculated risk, while the younger sister has to live through those still risky years. Irony ensued when my oncologist told me that even at age 110 I will have a higher risk for cancer than the “general population.” Even my most doddering imagined future carries a threat.
In projecting a misleading solidity, the numbers don’t count only what’s already out there. They become a basis of evidence for arguments about cancer by virtue of the preset categories for data collection.17 Numbers can seem equivalent and then tradable. Before you know it, you can exchange lives for other things, especially money, forgetting that the numbers once represented real people, with real communities and real histories and complex genealogies. Taking an objective count can be as misleading as it is illuminating.
I don’t particularly want to join the head-counting tribe, but since numbers so often define this disease, it’s worth examining them.
As the numbers stand now, one in two American men, and one in three American women, will be diagnosed with an invasive form of cancer during their lifetimes. Each day, over 1,500 Americans die of cancer, and a quarter of all Americans will eventually die from this disease. While more men will ultimately develop cancer, under the age of 39, women are significantly more likley to develop invasive cancers.18 Cancer has been the leading cause of death for Americans under 85 since 2001, and is the largest killer of women aged 34–70 and of men aged 60–79.19 Of all diseases, leukemia is the biggest killer for men under 40; after 40 it’s lung and bronchus cancers. Breast cancer is the main killer, period, of women aged 20–59.20
Currently, more than thirteen million cancer survivors live in the United States.21 Overall, cancer death rates are slightly declining: between 2004 and 2008, death rates decreased 1.3 percent per year.22 Some people consider the falling death rate the result of decreasing smoking rates, others attribute it to the success of early detection, and still others consider the decline meaningless given its minuscule size and the wide spread of sundry diseases it covers.
Different cancer registries use different categories to collect data, including the site at which the cancer first presents; stage at diagnosis; the patient’s age, race, and education; and the geographic location of treatment. The American Cancer Society estimates absolute numbers of cancer deaths each year as follows: lung and bronchus: 160,340 (with a median age at death of 72); colon: 51,690 (median age, 74); breast: 39,510 (female), 410 (male); prostate: 28,170.23 Cancer incidence rates, as opposed to death rates, offer quite a different lens. For example, the lung and bronchus cancer incidence rate, with 226,160 diagnoses annually, is about 41 percent higher than the death rate, while there are nearly three times the number of colon cancer diagnoses (143,460) than deaths each year. Breast cancer incidence is about six times the annual death rate for both men (with 2,190 diagnoses) and women (226,870); the prostate cancer incidence rate (241,740 diagnoses) is nearly ten times the annual death rate. About 360 men a year die of testicular cancer, with a median age of forty-four. Over two million Americans a year are diagnosed with nonmelanoma skin cancer, a disease with fewer than a thousand deaths annually; meanwhile, the 76,250 cases of melanoma each year correlate to about 9,180 deaths a year.24
Although the numbers vary from year to year, certain trends emerge. For example, testicular cancer incidence rates have increased by at least 75 percent since 1975 (although death rates have decreased to less than a third), and over the same timespan rates of brain cancers and central nervous system cancers have doubled for those aged 65 and over. Mortality rates