One blueprint for such an approach has been proposed recently by the foreign affairs columnist of the Wall Street Journal, Bret Stephens, whose former position as editor-in-chief of the Jerusalem Post gave him an on-site education into what world politics in the real world looks like — and how it can kill you, and any hope for a measure of decency in international affairs, if you don't bother to pay attention.

In his new book, America in Retreat: The New Isolationism and the Coming Global Disorder (Sentinel, £24), Stephens dismisses out of hand the basic premises that inform Democratic/neo-isolationist thinking, which are the premises on which Obama's foreign policy has been based. There is no self-constructing or self-maintaining world order and there never has been, Stephens argues. What order there is in world politics is an achievement, not a given. That achievement, which is always fragile, is realised and sustained only through leadership. And there is no other plausible candidate for taking the lead in stabilising the world and maintaining a modicum of order in world affairs than the US. Anyone who denies these self-evident truths, he suggests, ought not be trusted with the power and authority of the American presidency.

As for the strategic vision that ought to guide American leadership in reconstructing a measure of world order from the disorder Obama will leave behind, Stephens suggests taking a cue from American social science, and specifically from new theories of policing that have had a major impact on reducing crime in American cities over the past several decades. The origins of that theory may be found in a 1982 article by a Rutgers criminologist, George Kelling, and the man who would become, before his death in 2012, the most respected social scientist in America, Harvard's James Q. Wilson. Kelling and Wilson's article had the deceptively simple title "Broken Windows" and it drew on an experiment conducted years before by a Stanford psychologist, Philip Zimbardo. But let Stephens tell the fascinating tale:

Dr Zimbardo parked a car on a street in the Bronx, with the hood [bonnet] up and without licence plates. Within ten minutes, vandals began to pick the car clean of its valuables: battery, radiator, tires. By the next day, people began destroying the car, ripping up pieces of upholstery and smashing windows. Then, Zimbardo conducted the same experiment in tony Palo Alto, California, near the Stanford campus. This time, the car — also with hood up and the licence plates removed — sat untouched for several days. So Dr Zimbardo smashed a window with a sledgehammer. "Soon, passers-by by were joining in," wrote Drs Kelling and Wilson. "Within a few hours the car had been turned upside down and utterly destroyed." . . . What to conclude? "Disorder and crime are usually inextricably linked, in a kind of developmental sequence,'' Drs Kelling and Wilson argued. It had long been known that if one broken window wasn't replaced, it wouldn't be long before all the other windows were broken, too. Why? Because, they wrote, "one unrepaired broken window is a signal that no one cares, and so breaking windows costs nothing."