Six Degrees of Uncertainty

In the aftermath of the World Trade Center attack, it seemed that if you didn't know anyone who had seen the towers fall, you know someone who knew someone who did. Our society is so tightly knit, it seemed, that a tragedy that affected 6000 people in New York would affect everybody.

Proof for the popular notion that there are "six degrees of separation" between every individual in the country? Not necessarily. A new paper argues that the famous 1967 research paper that gave rise to that phrase is seriously flawed. In now-classic experiments, Yale psychologist Stanley Milgram gave people in Kansas a letter and asked them to attempt to forward it to a particular person in Massachusetts. If they didn't know the target person, they were to send it to someone they thought might know him. Surprisingly, he reported that it only took five jumps, on average, for the letters to reach their destination.

But Milgram didn't tell the whole story, says Judith Kleinfeld, a psychologist at the University of Alaska, Fairbanks. Working in the Yale archives, she discovered that the successful trips may have been brief, but only 30% of letters reached their destination. What's more, in a pilot study using less-official looking letters, just 5% made it. By not counting the uncompleted chains, Milgram skewed the results, she writes in a forthcoming paper in the journal Society. "I came out profoundly upset," Kleinfeld says. "One of my heroes had fallen."

Small-world expert Steven Strogatz, a mathematician at Cornell, says that Kleinfeld's findings raise valid concerns about whether people can perceive their ties with each other. But he says mathematical models show that such ties exist--even if people are not aware of them.

Related sites

The Stanley Milgram Web site
Six degrees among Hollywood stars