Two Soldiers

On July 12, 2006, Hezbollah fighters crossed into northern Israel near the village of Zar'it, ambushed a border patrol, killed three soldiers, wounded three more, and dragged two — Ehud Goldwasser and Eldad Regev — back into Lebanon. A botched rescue that afternoon killed five more Israeli soldiers when a tank hit a mine. By nightfall the incident had a name and a trajectory. By the next day, Israeli aircraft were bombing the runways of Beirut's international airport.

Thirty-four days later, a UN ceasefire ended what the ambush had started. In between: more than 1,100 dead in Lebanon, the overwhelming majority civilians; about 160 dead in Israel, most of them soldiers; a million people displaced; a country's infrastructure returned to rubble. All of it downstream of a raid whose stated objective was to capture two soldiers and trade them.

This is a specific machine, and it is very old. Call it the casus belli engine. A small, local, tactical event — a raid, a shot, a wrong turn — becomes the authorization for a strategic response. The event does not cause the war. It licenses it. But licensing comes in degrees, and the degree is the whole moral question. At one end the war is already decided — printed, chambered, timetabled — and the trigger merely supplies the date. At the other, only the plan exists, and the trigger is what converts a standing option into a chosen act. The engine runs at both ends. What changes is how much was settled before the morning arrived.

Sarajevo, June 1914: a driver takes a wrong turn, a nineteen-year-old fires two shots, and a continent that had spent decades loading itself finally goes off. The archduke did not start the First World War. He unlocked it. The powers had their mobilization timetables printed and their alliances chambered; they needed a morning, and they got one. Gulf of Tonkin, 1964: an ambiguous naval skirmish — one engagement real, one likely imagined — becomes a congressional blank check that was drafted before the ships ever reported contact. Those are the pre-decided end of the line, the drawer already open before the trigger arrives.

Where does 2006 sit on that line? Honestly, further toward the ambiguous end than the anniversary sermons admit. Israel had contingency plans for a Hezbollah campaign — but every military keeps contingency plans, and a plan in a drawer is not a decision on a calendar. What July 12 supplied was not a printed timetable; it was a plan meeting a government primed to reach for it, and a morning that made reaching look like answering. That is enough for the engine. It does not require that the war be scheduled. It only requires a slope and a pebble.

What makes the pattern grimly reliable is the asymmetry between the size of the cause and the size of the effect. Two soldiers on one side of a line. Twelve hundred coffins on both. The engine takes a tactical pebble and returns a strategic avalanche, and afterward everyone stands in the debris insisting the pebble was to blame. It was not. The slope was to blame. The pebble was just the thing that happened to move first.

Here is the detail that should be carved somewhere permanent. Ehud Goldwasser and Eldad Regev — the two men whose capture licensed the whole thing — were never traded alive. They came home in July 2008, in coffins, in exchange for prisoners and the bodies of the dead. It is not clear they survived the ambush that took them. So it is possible, and the evidence leans toward it, that the war fought to recover two living soldiers was fought to recover two men who were already dead on the first morning.

And here is the engine's cleverest function, the one that keeps it in service: it does not only authorize the war, it absolves it. The trigger converts an elective war into a forced one in the only account that ends up mattering — the public one. Nobody chose this; we were attacked, we answered, what else was there to do. The raid launders the decision. It takes an act that was reached for and reissues it as an act that was suffered, and by the time anyone counts the coffins the choosing has been edited out of the story. That is the real product. Not the war. The innocence.

Twelve hundred people died over an equation whose two variables were, by the end, both zero. The engine does not audit its own inputs, and it never returns a verdict of guilt — that is the point of it. It waits for a morning it can use, launders the morning into a grievance, and hands back a war no one has to admit they wanted. There is always another morning.