At the annual meeting last month of the American Association for Advancement of Science, Paul Slovic, a psychology professor at the University of Oregon, recommended a review and overhaul of the 1948 Genocide Convention. He offered two related reasons. The first is that it has been completely ineffective, and the second is that it doesn't accord well with our human tendency to be moved by dramatic individual tragedies and unmoved by mass killings.
The sentiment is not new. Stalin famously noted, "One death is a tragedy; a million is a statistic."
What is new are a couple of experiments that elucidate this unfortunate tendency. Slovic remarks, "We have to understand what it is in our makeup -- psychologically, socially, politically and institutionally -- that has allowed genocide to go unabated for a century. If we don't answer that question and use the answer to change things, we will see another century of horrible atrocities around the world."
An important ingredient in our makeup is that we respond more emotionally to individuals than we do to groups, even small ones.
One Slovic study (done with colleagues Daniel Vastfjäll and Ellen Peters) focused on respondents who were shown a picture of a starving African girl along with text detailing her individual plight. A second group of respondents was shown a picture of a starving African boy with accompanying text, and a third group was shown a picture with the children together and accompanying text describing their situation.
The researchers made some measure of the sympathy each photo elicited and of the amount of money people were willing to donate to the girl alone, to the boy alone, or to both. The individual photos and text elicited approximately equal sympathy and donations, but the joint photo and text elicited less.
Even if the number of people in a group is only two, it seems that our capacity to feel begins to decline. Other studies show it declines more precipitously for groups of five or 10, and the difference in our responses to, say, 57 or 58 deaths is indistinguishable.
If our moral feeling begins to decline at 2, it's not surprising that at 250,000 the feeling is reduced to a gray "Isn't that too bad" before we switch to a discussion of a flamboyant celebrity's personal situation, say Anna Nicole Smith and her love life.
Slovic offers an evolutionary explanation for this decline in sympathy. Humans evolved in an environment where looking out for themselves and their families was their paramount concern.
"There was no adaptive or survival value in protecting hundreds of thousands of people on the other side of the planet," he says. "Today, we have modern communications that can tell us about crises occurring on the other side of the world, but we are still reacting the same way as we would have long ago."
I would add that a somewhat similar explanation accounts for the inability of many to think statistically.
Other instances of moral failure are furthered by a simple lack of arithmetical skills. Many people are more likely to make charitable contributions, for example, when the outcome of their contributions can be characterized as largely positive, even when an outcome that is likely to be more negatively characterized does much more good. Thus, they're more likely to contribute to a project that promises to save 90 percent of 100 lives than to one that's likely to save only 10 percent of 5,000 lives, despite the fact that the latter intervention would save 410 more lives.
Or what about an almost cost-free intervention, such as the bombing of a few train lines during World War II that might have saved "only" 1 percent of those who died in concentration camps? (Of course, a reluctance to look at the numbers and a readiness to be swayed by marginally relevant personal stories is pervasive and extends far beyond the issue of genocide.)
Another study by Slovic (together with Loewenstein and Small) demonstrates the power of individual pictures. Researchers gave people the opportunity to contribute up to $5 to Save the Children and showed a group of them a picture of a starving African girl with text describing her dire conditions. The researchers showed a second group of people a set of general statistics about starving children in Africa, and they showed a third group both the picture and facts of an identifiable life as well as a statistical description of the general conditions.
The first group contributed the most money, the second and third groups substantially less. Astonishingly, adding the statistical description to the picture and personal information actually decreased the amount contributed.
One inference from this work is that for most people a compelling picture of an individual is worth a thousand statistics. The thought occurs that this might be an opportunity for some paparazzi to redeem themselves. Forget celebrities' rehab centers; go to Darfur's refugee camps.
John Allen Paulos, a professor of mathematics at Temple University, has written such best-sellers as "Innumeracy" and "A Mathematician Plays the Stock Market." His "Who's Counting?" column on ABCNEWS.com appears the first weekend of every month.