By study­ing the footage from an uniden­ti­fied UK city’s CCTV cam­eras, psy­chol­o­gist Mark Levine is find­ing that a num­ber of the­o­ries about crowd psy­chol­ogy pre­vi­ously taken as gospel may be incor­rect, includ­ing the bystander effect (some­times referred to as the Kitty Gen­ovese effect) and the idea that crowds are inclined to be unruly and violent.

Dr Levine per­suaded the author­i­ties in one British city to allow him to look at their CCTV footage of alcohol-fuelled con­flict in pub­lic places. […] He analysed 42 clips of inci­dents that oper­a­tors in a con­trol room had judged had the poten­tial to turn vio­lent, though only 30 of them actu­ally did so. He recorded ges­tures he labelled either “esca­lat­ing”, such as point­ing and prod­ding, or “de-escalating”, such as con­cil­ia­tory open-handedness. […]

Judg­ing the fight to begin with the aggressor’s first point­ing ges­ture towards his tar­get, the researchers found that the first inter­ven­tion usu­ally involved a bystander try­ing to calm the pro­tag­o­nist down. Next, another would advise the tar­get not to respond. If a third inter­ven­tion rein­forced crowd sol­i­dar­ity, send­ing the same peace­ful mes­sage, then a vio­lent out­come became unlikely. But if it did not—if the third bystander vocally took sides, say—then vio­lence was much more likely.

via Mind Hacks