Vicksburg — what fell was not a town but a line

By 1863 the Civil War was in its third year. The Mississippi was the spine of the continent's freight and, at the same time, the line that held the Confederacy's two halves together. Cattle, grain and men came across from Texas, Arkansas and western Louisiana¹⁴.

The Union already held New Orleans at the mouth and Memphis upstream. Two stoppers remained: Vicksburg, and Port Hudson below it. Vicksburg did not matter as a town. It mattered because, while that one point was corked, the river could not be used as a line.

And the stopper could not be pulled frontally. The town sits on a bluff at a bend, its batteries firing down from some 60 metres above the water; the approaches from the north are swamp and bayou. Since the previous winter Grant had failed at nearly everything, including an attempt to dig a canal and move the river itself¹³.

Which sets the question. How does a fortress everyone agrees cannot be taken frontally fall without a frontal attack?

The campaign is built out of three acts of giving something up.

First, the fleet's line of retreat. On the night of 16 April, Porter's 12 or so vessels passed directly under the batteries and came out to the south. Ships that ran past could not fight the current back up. It was one-way, and failure would have cost the fleet. Success meant a means of crossing and supply below the town¹³.

Second, the supply line. Grant marched down the far bank, crossed to the east side at Bruinsburg on 30 April, and then cut loose from his line to the north. Supplies stop arriving, and the line that had to be guarded stops existing. The troops moved living largely off the country¹³.

Third, the objective itself. He did not head straight for the town. He went east to Jackson, drove off the force there and wrecked the railway and the stores, and only then turned back towards Pemberton. He gave up "reach the town quickly" in exchange for "keep the two enemy forces apart".

5 actions in 17 days: Port Gibson, Raymond, Jackson, Champion Hill, Big Black River. None of them is a battle on the way to the town; each is a battle to keep two Confederate forces from joining.¹³

The result was that Pemberton was pushed into the town and Johnston was left outside it. The assaults of 19 and 22 May both failed. So Grant stopped assaulting and started digging. Trenches crept forward, mines went under the works, guns were emplaced, the fleet fired from the river. 47 days later the garrison ran out of food and surrendered².

This is the core of the structure. The object of the campaign was never the town. It was the river the town was corking. The town was the stopper, and a stopper can be dealt with by hitting it or by working round it. What worked was the second.

So the pattern this case carries is a design of supply. The decision comes not from what the enemy holds now but from whether things keep reaching him. Which means the work before fighting is to make his line the battlefield, and, if necessary, give up your own for a while.

The garrison was not alone in the town. Some 3,000–5,000 civilians stayed. To get out of the shelling they cut into the hillsides and lived in an estimated 500 caves through the heat and the bombardment. By June the food had gone: mule meat appeared in the market, and rations of peas and ground corn were issued as a substitute for bread².

This is not a sad aside; it is part of the pattern. A design of supply cannot starve only the enemy's army. Stop the line and everybody inside the line pays the same price. The reason a siege is the dull, reliable method is that the reliability is being charged to somebody.

The same thing was written at Stalingrad. A design decides whether it can be won, not whether it will be paid for lightly. Reading something as structure means putting the effect and the price on the same page.

Reading Pemberton as an incompetent who shut himself in a town and lost an army is not accurate. He was being given different orders by two superiors.

Johnston, commanding the department, said it repeatedly: abandon the town, save the army, join me. President Davis said: hold Vicksburg. Both were superiors and both had a case. Johnston was militarily right and Davis was politically right, because the town was also the proof that the Confederacy was holding¹⁴.

And Pemberton was Pennsylvania-born, a Confederate general whose origins were held against him. A man already under suspicion finds it hard to choose the option that would seem to confirm it. So he took a little of each, stayed half-committed to both the town and the army, and lost both. Splitting the difference is not a decision.

Kutuzov at Austerlitz, Paullus at Cannae, Gravina at Trafalgar, Paulus at Stalingrad, Boucicaut at Agincourt, Gérard at Waterloo — and now Pemberton. A correct reading losing to a structure makes its seventh appearance on this site, this time in the variant "two superiors who say different things" — the same failure as two consuls commanding on alternate days at Cannae, reappearing in 1863.

The other man deserves a look too. Johnston never let go of the order of priority: a town can be given up, an army cannot. Judging that the force he had could not break the siege, he did not attempt a relief he could not deliver¹². Militarily that is right.

So this case has two people whose correct reading did not carry. Pemberton could not decide; Johnston decided and was not heeded. The first is the consequence of a forked chain of command, the second of a sound military judgment colliding with a political necessity. There is more than one way for a structure to break.

Vicksburg surrendered on 4 July and Port Hudson downstream on 9 July, and the whole Mississippi passed into Union hands. The Confederacy was cut in two, and the cattle, grain and men that had come from the western bank stopped coming¹⁴.

At Cannae, Trafalgar, Stalingrad and Agincourt this site has repeatedly watched the rule that a decisive victory is decisive only inside its own arena, and at Waterloo its inverse. Vicksburg is one of the few exceptions.

Why it is an exception is itself structural. What happened here was not the outcome of a battle but a permanent rearrangement of logistics. The outcome of a battle can be recovered in the next battle. Control of a river has to be taken back by another campaign, and the Confederacy did not have one to spare. If a victory tells outside its arena, it is not because it was won but because it rewrote the conditions.

The turning point of the Civil War was Gettysburg: that is the most widely circulated account. Line the dates up and it gets strange. Gettysburg ended on 3 July and Vicksburg surrendered on 4 July. They are the same week.

And they told differently. Gettysburg stopped Lee's invasion of the north, but Lee's army survived and the war ran on for nearly two more years. Vicksburg halved the Confederacy's logistics permanently and deleted an army and 172 guns. It is still Gettysburg that is remembered⁴.

The reason is structural, and that is the interesting part. Gettysburg was a dramatic battle that took three days, happened on the eastern front where the capital and the newspapers were, had Lee in it, and could be summed up in one scene: that charge. Vicksburg took three months, happened in the west, was decided by digging, and has no scene to sum it up.

What separates the famous battle from the one that told is often how easily it makes a drama. This is the same mechanism as the frozen ponds at Austerlitz, the fateful five minutes at Midway and "a near-run thing" at Waterloo — but running the other way. Those are stories in which the contents of an event were inflated; this is one in which an entire event dropped out of view.

The images confirm it. The Kurz & Allison lithograph above shows a general assault; what actually took the town was digging. A picture cannot show digging. What cannot be pictured tends to fall out of the memory unpictured.

A: if the fleet had not got past on 16 April. No means of crossing or supplying below the town, and the move to the east bank does not happen at all. The hinge of this campaign is not a battle but one irreversible night passage.

B: if Jackson had not been struck first. Johnston and Pemberton join, and the army trying to besiege is caught between them. "Hurry to the town" was, in this case, wrong.

C: if the supply line had been kept. Troops are detached to guard it and the pace drops; 5 actions in 17 days is not available. What was given up was not supplies but the obligation to defend.

Each is an untestable thought experiment, not a claim about what was. Only by removing a factor does its weight become visible.

Summarise Vicksburg as a long siege and you miss the core. That it could not be taken frontally was known from the start. What changed was the object. Not take the town but open the river the town was corking. The moment it was put that way, the one-way passage of the fleet, the cutting loose from supply and the strike to the east all connected into a single design.

And the decision came not from an assault but from 47 days of unglamorous work. Not requiring the victory to look impressive is, here, a technique. The price of that was losing the memory to a more dramatic victory won the same week.

What this case adds to the seven already here is an eighth type. Austerlitz was a design of space, Cannae of momentum, Midway of time, Trafalgar of command, Stalingrad of attachment, Agincourt of frontage, Waterloo of convergence. Vicksburg is a design of supply — make his line the battlefield, and give up your own for a while if you must.

One last thing. This site has repeatedly found that a decisive victory is decisive only inside its own arena. Vicksburg is an exception, and the reason it is an exception explains the rule. What happened was not the outcome of a battle but a rewriting of how things move. If a victory tells outside its arena, it is not a victory that was won but one that changed the conditions.