Forty-five years ago, some social psychological experiments posed story problems that assessed people's willingness to take risks (for example, what odds of success should a budding writer have in order to forego her sure income and attempt writing a significant novel?). To everyone's amazement, group discussions in various countries led people to advise more risk, setting off a wave of speculation about group risk taking by juries, business boards, and the military.
Alas, some other story problems surfaced for which group deliberation increased caution (should a young married parent with two children gamble his savings on a hot stock tip?).
Out of this befuddlement—does group interaction increase risk, or caution?—there emerged a deeper principle of simple elegance: group interaction tends to amplify people's initial inclinations (as when advising risk to the novelist, and caution in the investing).
This "group polarization" phenomenon was then repeatedly confirmed. In one study, relatively prejudiced and unprejudiced students were grouped separately and asked to respond—before and after discussion—to racial dilemmas, such as a conflict over property rights versus open housing. Discussion with like-minded peers increased the attitude gap between the high- and low-prejudiced groups.
Fast forward to today. Self-segregation with kindred spirits is now rife. With increased mobility, conservative communities attract conservatives and progressive communities attract progressives. As Bill Bishop has documented, the percentage of landslide counties—those voting 60 percent or more for one presidential candidate—nearly doubled between 1976 and 2008. And when neighborhoods become political echo chambers, the consequence is increased polarization, as David Schkade and colleagues demonstrated by assembling small groups of Coloradoans in liberal Boulder and conservative Colorado Springs. The community discussions of climate change, affirmative action, and same-sex unions further diverged Boulder folks leftward and Colorado Springs folks rightward.
Terrorism is group polarization writ large. Virtually never does it erupt suddenly as a solo personal act. Rather, terrorist impulses arise among people whose shared grievances bring them together. In isolation from moderating influences, group interaction becomes a social amplifier.
The Internet accelerates opportunities for like-minded peacemakers and neo-Nazis, geeks and goths, conspiracy schemers and cancer survivors, to find and influence one another. When socially networked, birds of a feather find their shared interests, attitudes, and suspicions magnified.
Ergo, one elegant and socially significant explanation of diverse observations is simply this: opinion segregation + conversation → polarization.