But the U.S. government has countered with its own "green" strategy, as in greenbacks to rebuild the whole of Lebanon.

Amir Taheri, in an Aug. 25 Wall Street Journal essay, has made the most cogent argument that Hezbollah has actually lost the war. StrategyPage.com (which carries this column) and its editor, James F. Dunnigan, started making the case for Hezbollah's looming defeat in late July.

Taheri argues that Hezbollah is on the edge of a huge political defeat within Lebanon. "The leaders of the March 14 movement," Taheri writes, "which has a majority in the Lebanese Parliament and government, have demanded an investigation into the circumstances that led to the war, a roundabout way of accusing Hezbollah of having provoked the tragedy."

StrategyPage.com noted Hezbollah's political and military failure as it occurred. StrategyPage wrote on July 26: "Hezbollah knows, however, that as long as they can launch at least one rocket a day, they can claim victory. This is because Arabs no longer expect to ... defeat Israel militarily, so that if the Arab force is still fighting, it is considered a victory. While ludicrous, this attitude has been widely accepted throughout the Middle East. However, this twisted logic is beginning to fray, and an increasing number of Arabs are questioning it. But in the short term, it still works."

StrategyPage is arguing that what happens on the battlefield, in the neighborhood and on the street eventually trumps "media perception" generated by propagandists and the sensationalist press.

Most presciently, StrategyPage noted on July 25: "While Hezbollah has been able to muster public support throughout Lebanon and the Arab world, they know that in the aftermath of all this, despite declaring a victory, they are already being blamed for causing a disaster, and will suffer substantial losses in the aftermath of this war."

We're in the aftermath. Hezbollah experienced a moment of media glory, but that glory has faded. For Hezbollah, the "continuing aftermath" is anything but promising.