WHEN PRESIDENT BUSH set aside the Pentagon's strategy for war with Iraq and ordered an attack on Saddam Hussein and his inner circle, it created shock and awe in the media and perhaps in a few offices of Bush's own administration. It shouldn't have. The president behaved, without much ado, as a commander in chief who intends to make the major war decisions himself and not reflexively defer to the expertise of military brass. Bush relied on his own judgment and instincts--and thus acted in the best tradition of wartime leadership.

Wars have political as well as military goals, and it's left to the president to mix the two in the right combination. This is perilous business. Still, it requires boldness rather than caution. Bush delivered. He delayed the heaviest bombing in a bid to decapitate the Iraqi leadership, prompt mass surrender of Iraqi forces, minimize civilian casualties, and create postwar conditions favorable to the emergence of an Iraqi democracy. He later stepped up the bombing as originally planned, but only after American and British ground troops had penetrated far into Iraq--another unexpected sequence of events with political implications.

Bush has wisely read and imbibed the advice of military strategist Eliot Cohen in "Supreme Command," a study of successful wartime leaders from Lincoln to Churchill. The common thread among those leaders was their insistence on taking charge, personally and aggressively, and not leaving decisions to the military. More often than not, the generals tended to be cautious in the extreme, more inclined to

But political leaders shouldn't simply dictate to the military, Cohen writes. What existed in the best cases was "an unequal dialogue--a dialogue, in that both sides expressed their views bluntly, indeed, sometimes offensively, and not once but repeatedly." In the end, though, the leader must be willing to impose his judgment, even on operational matters. And this is what happened at the outset of America's war with Iraq.

Just as the 48-hour deadline for Saddam to flee Iraq was nearing, the administration learned the dictator, his two sons, and several top aides were gathered in a bunker on the outskirts of Baghdad. CIA director George Tenet discussed the matter with defense secretary Donald Rumsfeld and military officials at the Pentagon, then took the intelligence to the White House. Tenet advised the president that it was "worth a shot" to try to eliminate Saddam and his entourage. Bush agreed.

This was no minor decision and was hardly risk-free. The press had been informed no hostilities would occur that evening. Attempts to hit Saddam, reputed to sleep in a different bed every night, had famously failed during the Gulf War. By suddenly changing the script, Bush delayed the Pentagon strategy, long in place, while the attack on Saddam went ahead with cruise missiles and precision-guided munitions. Meanwhile, negotiations to bring about mass surrenders of Iraqi soldiers went forward. Why bother with all this, since a quick military victory was inevitable? The reasons are political. Killing Saddam would shorten the war and minimize casualties. The fewer civilian deaths, the more humane and popular the war. The less destruction of Iraq's infrastructure, the easier the task of rebuilding the country. And so on.