Tet: the verdict was returned off the battlefield

At the start of 1968 there were about 500,000 American troops in Vietnam. There was no front. There was ground you took and gave back the following month. So the yardstick adopted for progress was the number of enemy dead. Kill faster than he can replace and eventually he breaks. The logic is sound.

And on the numbers, the war was being won. In November 1967 Westmoreland told reporters the end was coming into view²⁴.

But the yardstick carried an unwritten premise. Who looks at these numbers and rules that the war is won? What decided whether the war in South Vietnam continued was not Saigon and not Hanoi. It was Washington, 13,000 km away, and the people who voted there.

This case turns on that single point.

On the night of 30–31 January 1968, about 77,000 troops attacked more than 100 towns simultaneously: 36 of 44 provincial capitals, 72 of 245 district towns, and Saigon. It fell inside the Tet holiday truce, with much of the South Vietnamese army on leave¹.

Nothing on that scale or that simultaneity had happened in the war. But the attacks were not the centre of the plan.

The operation's own name was General Offensive, General Uprising. Strike the cities together, the South Vietnamese army breaks, the population rises, the regime falls — the plan handed its entire second half to the civilian population.³

The uprising did not come. The urban population did not go over, and the South Vietnamese army did not break. Half the plan stayed blank while the other half was executed.

Within weeks most of the committed force was gone. The southern insurgency in particular was effectively spent in this one operation, and the war from then on was carried by North Vietnamese regulars¹³.

Militarily, Tet was a defeat for the side that launched it. Not one objective was held, the losses were an order of magnitude apart, and the uprising never came. There is little dispute about this.

And yet the direction of the war changed afterwards. What changed it was not the battlefield.

In the United States, approval of Johnson's handling of the war fell from 39% to 26%. Those calling themselves hawks fell from 60% to 41%. And on 31 March the President announced a curtailment of the bombing and, at the end of the speech, that he would not seek re-election.

Here is the eleventh type: a design of the verdict — the people who rule on winning are not the people on the field. Decide who they are before you fight.

Do not confuse it with the deception at Normandy. That manufactured what the opposing commander believed. This manufactured an event that reached the third party who was keeping score — and did so not with a lie but with real attacks. What was built was not the enemy's belief but the material of the verdict.

Nine days before the offensive, North Vietnamese forces surrounded the base at Khe Sanh in the far north-west. About 6,000 Marines held it, and 114,810 tons of bombs fell around it in the first three months. A repeat of Điện Biên Phủ, where the French had been surrounded and forced to surrender fourteen years earlier — that is how the Americans read it.

And not only the commander in the field. A terrain model of Khe Sanh was set up in the White House Situation Room, and the President asked the Joint Chiefs for a written assurance that there would be no second Điện Biên Phủ²⁴. The judge's attention was collected on this one point.

And after the simultaneous attacks on the cities began, Westmoreland went on classifying them as a diversion from Khe Sanh. The same shape that the German command fell into at Normandy — with one difference. There, an enemy manufactured the error. Here it was made in-house.

To be honest about it, the question is still open. Whether Khe Sanh was a feint for the cities or a serious offensive in its own right is disputed among historians¹, and Westmoreland maintained afterwards that Tet had been the diversion from Khe Sanh.

But the real point is not which answer is right. It is that answering it correctly would not have decided anything. Khe Sanh held. The attacks on the cities were thrown back. Get both battlefield questions right, and the verdict is still returned 13,000 km away.

Reading Westmoreland as incompetent does not fit the record. In a war with no front and no ground that stayed taken he built a measure of progress, sustained it, and threw the Tet attacks back in weeks. On his own yardstick he was winning.

What broke was not the judgement but the premise underneath it: that he would be scored on the yardstick he was using. The scoring was done in Washington, and Washington was not looking at loss ratios.

There is a deeper hole in attrition itself. Attrition is an instrument for measuring the limit of what the enemy will absorb. He measured that diligently. But the same clock is also counting your own limit. There was a ceiling on the human and political price America would pay, and nobody had put a number on that one. A two-way clock, read from one side only.

The same shape appears in another direction on Giáp's side. He was sceptical of a simultaneous urban uprising, and that caution was removed before the political decision was taken. A correct reading without the standing to carry it.

A correct reading losing to a structure now appears here for the tenth time: Kutuzov at Austerlitz, Paullus at Cannae, Gravina at Trafalgar, Paulus at Stalingrad, Boucicaut at Agincourt, Gérard at Waterloo, Pemberton at Vicksburg, Mitsunari at Sekigahara, Rundstedt at Normandy. This time the variant is "he was scoring the match with his own yardstick".

Read this far and it looks as though Hanoi designed the verdict correctly. That is not what happened.

The judge the plan aimed at was the population of South Vietnam. Strike the cities together and they rise, and the regime falls from the inside: that is what "General Uprising" meant. That judge did not move. The verdict was actually returned by a body that appears nowhere in the plan — the American electorate¹³.

So this case contains the type's success and failure at once. The reading that the verdict comes from off the battlefield was right. The reading of who the judge was, was wrong. What made it land was not accuracy of aim but the fact that there was more than one judge off the field.

Which makes the failure conditions plain. Believing there is only one judge. And never checking that the judge you aimed at is the judge who actually scores.

There is no room in this case for the reading that the winner got off cheaply.

The side that gained the strategic result paid with the near-destruction of the southern insurgency. The underground that came into the open to fight in the cities exposed itself and was rolled up. From 1968 the war was carried by North Vietnamese regulars, and the NLF barely survived as a political force into the post-war government¹³.

The civilian bill was heavy: 14,300 killed and 630,000 made homeless in the first phase alone. In Huế civilians were killed systematically by communist forces while the street fighting destroyed the city around them⁵. A way of fighting whose judge sits far away drops the bill where the judge is not.

The war ended seven years later. What was settled in these weeks was not who won, but that the premise "America keeps adding" was gone.

The popular memory is that America lost at Tet. Militarily the side that lost was the side that attacked: no objective held, losses an order of magnitude apart, the southern organisation destroyed.

There is a mirror-image correction too — that the war was being won and was lost by the press. That does not stand either. The request for 206,000 men came from the army, not from a newsroom, and support had been falling before Tet.

Both stories make the same mistake. They assume winning can be measured on one scale. This battle had at least three scorecards — loss ratios, ground, and the patience of the American electorate. Only the third had the authority to decide whether the war continued.

In the site's taxonomy of story corrections this is the fifth kind. There is the inflated kind (the frozen pond at Austerlitz), the missing kind (Vicksburg), the compressed kind (Sekigahara) and the substituted kind (the rubber tanks at Normandy). Tet is the unit kind: the result is argued about while nobody has agreed what counts as winning.

A: no attack on the embassy. Militarily nothing changes. But the event that happened closest to the judge disappears. In this case that is plausibly the single most effective point of the whole offensive.

B: no request for reinforcements. The account of having thrown the attack back at least stays internally consistent. Asking for more men days after declaring victory is what broke it.

C: no American election in 1968. The judge loses his deadline. The shift in opinion is the same, but the speed at which it converts into policy falls. This type borrows the judge's clock.

These are unverifiable thought experiments, not claims about what happened. Removing one element is how you see what it weighed.

Read Tet as "the battle America lost" and you miss the centre. On the battlefield the attacker lost. The direction of the war changed anyway, because the authority to rule on winning was not on the battlefield.

And Hanoi could not name that authority either. The judge it aimed at did not move; a judge it had not aimed at did. The value of this case is not that it is a clean example of the type but that it shows, on the same page, what happens when you get the type wrong.

That makes eleven. Austerlitz — space; Cannae — momentum; Midway — time; Trafalgar — command; Stalingrad — fixation; Agincourt — frontage; Waterloo — junction; Vicksburg — supply; Sekigahara — defection; Normandy — deception. Tet shows a design of the verdict: who won is usually decided by somebody who was not fighting.

One last thing. Unlike the other ten, this type asks a question back of whoever uses it. Reaching the judge also means making the bill fall where the judge is not. The people standing in front of the rubble in Cholon appear on none of this battle's scorecards.