Law of Error
Most Americans take for granted that our lives are etched in data. The sums of our vital statistics, stored electronically, have become what criminologists and direct marketers call "identity profiles." When President George W. Bush promises to catch terrorists by monitoring databases, it's obvious he means our moral dispositions are part of the information traces we leave behind. Credit card transactions at the grocery store, locations of cell phone calls, and medical records are the variables in an equation that, when solved, is supposed to reveal our hidden intentions, our most profound character traits, and our predilections for crime.
It all sounds very 'five minutes into the future.' We could never have conceived of something like John Poindexter's Total Information Awareness plan before the mid-1990s, when an explosion of digital media put seemingly everything online. But today's high-tech identity profiles are not a product of the Internet boom. Their origins should really be traced back to the strange physiognomic obsessions of a 19th-century Belgian mathematician named Adolphe Quetelet.
In 1844, after years of research, Quetelet published an astonishing article analyzing the chest sizes of 5,000 Scottish soldiers. His theory was simple: some people are more normal than others. Borrowing a premise called the law of error from astronomy, Quetelet asserted that the more human specimens you measure, the less your margin of error will be in determining what is truly normal. Out of thousands of Scottish body parts, Quetelet created the "average man." And this became the first identity profile against which all deviants, criminals, and undesirables could be measured and found wanting. For Quetelet, abnormal bodies were invariably linked to criminal behavior. He argued that prostitutes were "enervated" by the immorality of their lives, often giving birth to stunted, doomed babies who rarely lived beyond childhood. The "average man" thus defined his degenerate counterpart.
Quetelet's work revolutionized the emerging field of population studies, affecting everything from criminology to evolutionary biology. He was, for example, responsible for encouraging a young French police officer, Alphonse Bertillon, to fanatically measure the ears, legs, fingers, and skulls of thousands of criminals for the purposes of developing, in 1882, the world's first database devoted to identity tracking. We can also see Quetelet's ideas coupled with social Darwinism in the work of Francis Galton, founder of the eugenics movement, who advocated tracking the inheritance of behaviors he associated with certain physical characteristics. For example, a tendency toward thievery might be passed along to your children along with dark skin. According to Galton, if you could just track the emergence of dark skin in a population, you could predict the rise in crime rates.
As the 20th century wore on, psychological tests replaced Bertillon's caliper compass measurements and Galton's eugenics, but the underlying principle remained the same. What unifies the theories of these men, as well as their 21st-century successors, like Poindexter, is the notion that data about human lives can be recorded, categorized, and analyzed to predict future behavior. If the government could just get enough information about you, so the logic goes, Central Intelligence Agency analysts could write a report about your next move. It might read something like this: Trip took a trip to Malaysia, regularly attended his local mosque, bought some books on terrorism, donated some money to a Palestinian activist group, and took a class on chemistry at the Massachusetts Institute of Technology; now it's likely Trip will become a terrorist.
But of course this report would be based entirely on identity profiles -- personalities fashioned out of baby blocks and 19th-century French finger measurements. It would be measured against an idea of "normal" first developed by a mid-19th-century Belgian mathematician and later updated by a series of pseudoscientists and law enforcement agents. There is no room in identity profiling for the ambiguities, arbitrary impulses, and inexplicable bursts of hope that so often rule our actions.
More than 150 years ago, 5,000 Scottish soldiers became some of the first unwitting data points in a vast information-aggregating enterprise devoted to tracking people whose identities deviate from the average. A potent mix of bogus rationalism and mythology led us to believe humans could be profiled and categorized into types. And these dubious classification systems have become the foundation for current theories of terrorist behavior and identification. Profiling may have gone high tech, but it still isn't good science.
Annalee Newitz ( email@example.com) is a surly media nerd whose identity is now available in a beta version but may require frequent security updates. Her column also appears in Metro, Silicon Valley's weekly newspaper.