Cluster Bomb

The facts won't stop people from blaming cancer on synthetic chemicals.

After an $8 million study, we now know who and what to blame for high rates of breast cancer on Long Island: nobody and nothing. Or, at least, not the usual suspects, chemical companies and their products.

National Cancer Institute (NCI) researchers, in the most exhaustive study of its kind, could find no link between increased breast cancer rates and exposure to chemicals such as the pesticides DDT and chlordane or PCBs used as coolants in electrical transformers. This study is one more in a long line that could find no link between breast cancer and exposure to synthetic chemicals.

By coincidence, around the time the NCI's findings were reported, The New York Times Magazine ran a story on coincidences and the deep human need to find causes, to identify someone or something to blame for an occurrence. Human brains are adapted to be pattern recognition machines. Being able to recognize a tiger's stripes in a sun-dappled bamboo forest enhances one's chances of escaping its jaws.

But it turns out that we over-recognize patterns. Our brains can find patterns in anything, which is why tarot cards and astrology remain popular. Researchers have shown that people will consistently claim to identify patterns in tables of randomly generated numbers. Just as our brains succumb to optical illusions, they fall victim to causality illusions.

The New York Times Magazine article noted the "pattern" in the 9/11 atrocities. The numbers 9/11 equal 11 (9 + 1 + 1 = 11), and American Airlines Flight 11 was the first plane to hit the twin towers. There were 92 people aboard (9 + 2 = 11). September 11 is the 254th day of the year (2 + 5 + 4 = 11). There are 11 letters each in Afghanistan, New York City, and The Pentagon. The twin towers themselves took the form of the number 11. Of course, this "pattern" leaves out the other flight numbers, the numbers of passengers on the other planes, the fact that Giuliani and World Trade Center aren't composed of 11 letters, and so on.

Another point to keep in mind is that low-probability events do occur. Even if there is only a one in a million chance of something occurring to someone, with 6 billion people on the earth it will occur 6,000 times somewhere. As the Internet reaches its tentacles further into human society, these odd occurrences have ever greater chances of being marveled at by wider audiences.

From a public policy perspective, one of the most common and problematic misperceived patterns is cancer clusters. Every year we are treated to reports of communities that purportedly suffer more than their share of cancer. We all know the script by heart: Ten people in a small town are diagnosed with cancer--say, leukemia--within five years. Victims, reporters, regulators, and trial lawyers frantically search for a corporation manufacturing some allegedly toxic chemical nearby on which to pin the blame. The victims' suffering, it's assumed, must be the result of corporate greed.

This is exactly the script that has been playing out on Long Island for at least a decade. Under orders from Congress, the NCI conducted the most searching inquiry ever into an alleged cancer cluster and came up with exactly nothing. The NCI's Deborah Winn told The New York Times the data "were very, very conclusive" that the synthetic chemicals studied "are not associated with breast cancer."

Interestingly, another study, reported just a month earlier, did find a very high correlation between a risk factor and a woman's chances of getting breast cancer. That study confirmed that the longer a woman breast-fed, the lower her risk of getting breast cancer. Could it be that long-term breast-feeding was not fashionable among suburban moms in the comfortable purlieus of Long Island two or three decades ago?

Fate or bad luck are not acceptable to us pattern-searching humans. What we once blamed on the malevolence of witchcraft, we now blame on the malevolence of corporations. In a sense, we are still in hot pursuit of witches.

The latest witch hunt was launched by the Pew Charitable Trusts a couple of years ago, when it set up its Health Track program. One of the chief goals of Health Track is to to uncover correlations between diseases and exposure to environmental toxins. Health Track boils down to a computerized system for identifying cancer clusters.

As the Long Island breast cancer study illustrates, most of the cancer clusters Health Track will uncover will have no identifiable causes. By pointing the finger of suspicion at modern witches, the program will unnecessarily frighten communities and spark useless regulations and lawsuits.

As our deepening knowledge of biology is revealing, much of the human body's resources is aimed at keeping our cells' natural tendency to become cancerous at bay long enough for us to reproduce successfully. The plain, unavoidable fact of life is that our bodies' defenses against cancer break down as we age. It is true that high, prolonged exposures to some synthetic chemicals (or, in the case of cigarettes, natural chemicals), can cause cancer. But as the NCI epidemiologists found in Long Island, trace exposures to "environmental toxins" generally can't be linked to cancer. Nevertheless, our built-in drive to identify patterns will guarantee that we will continue to seek someone to blame for our ills and that we will suffer through many more expensive, unnecessary, and self-defeating witch hunts for a long time to come.