The Levee and the Letter

McPhee had a rolled-up Army Corps of Engineers survey map tucked under his arm when he arrived at the diner outside Baton Rouge. The map was from 1963, and he’d been carrying it in the trunk of his car since a reporting trip fifteen years earlier, waiting for a reason to unroll it. He smoothed it flat across the table, anchoring one corner with the sugar dispenser and another with his coffee cup, and began pointing out the places where the Mississippi River had tried to change course.

“Here,” he said, tapping a spot near the Old River Control Structure. “And here. And here, forty miles south, where the Atchafalaya takes half the flow during a flood and the Corps has to throttle it back. The river wants to go west. It has wanted to go west for fifty years. The entire infrastructure of southern Louisiana — the refineries, the ports, the city of New Orleans — exists because the Army Corps of Engineers has been winning an argument with a river. They will not win it forever.”

I was sitting across from him, trying to match the map to what I’d seen driving in: the chemical plants along River Road, the sugarcane fields behind the levees, the towns that existed in the floodplain because the floodplain had been declared, by engineering fiat, to be dry land. Baldwin was late. His flight from New York had been delayed, and when I’d called his hotel, a woman at the front desk said he’d gone out walking and hadn’t said where.

“The thing about a levee,” McPhee said, still studying the map, “is that it solves one problem by creating a different problem in a different place. You build a levee on the east bank, the water rises higher on the west bank. You build a levee on the west bank, the water rises higher everywhere. The river carries silt, and when you confine it between levees, the silt drops in the channel instead of on the floodplain, and the channel rises, and you build the levees higher, and the channel rises higher still. The system is self-defeating. Every year of successful flood control brings you closer to the catastrophic flood that will overwhelm the control.”

“How long has this been going on?” I asked.

“Since 1717. When Bienville founded New Orleans, he put it on a crescent of relatively high ground — five feet above sea level, maybe six in places. The first levee was three feet high and a mile long. By the 1850s, the levee system was six hundred miles long. By the 1920s, after the great flood of 1927, it was redesigned to handle everything the river could produce. The engineers were certain. They used that word — certain. The 1927 flood was the worst case. They had built for the worst case.”

“And then?”

“And then the river kept being a river.”

Baldwin arrived during this exchange, entering the diner with the particular alertness of a man who has just been walking through a place that disturbs him. He was wearing a light jacket despite the heat, and he stood inside the door for a moment, surveying the room — the white waitress, the Black cook visible through the kitchen pass, the farmers at the counter, the rolled-out map with its blue contour lines — before sitting down.

“I walked along the levee,” he said. “From the hotel. Maybe two miles. On the river side, the water was brown and high and there were barges. On the land side, there were houses. Small houses. Shotgun houses, some of them. Mostly Black families, as far as I could see. Children in the yards. Laundry on lines. And between those houses and that water, a wall of earth that someone built and someone maintains and someone decided was sufficient.”

McPhee looked up from his map. “The levee at that point is approximately twenty-two feet above the surrounding grade. The design flood stage — ”

“I’m not asking about the design,” Baldwin said. “I’m asking about the decision. Who decided those families could live there? Who decided the wall was high enough? And when the wall fails — because you’ve just been telling him it will fail — who will be on the wrong side of it?”

This was the first minute of their conversation, and I understood already that it would not be a negotiation between compatible sensibilities. McPhee deals in material facts — the tonnage of alluvium, the cubic feet per second of discharge, the precise angle at which a revetment meets a current. Baldwin deals in moral facts — who benefits, who is exposed, who decided, and what it costs them to pretend they don’t know.

“I can tell you who decided,” McPhee said, carefully. “The Mississippi River Commission was established in 1879. The Army Corps of Engineers took primary responsibility after the Flood Control Act of 1928. The decisions about levee placement, levee height, and land use within the floodplain are made by a combination of federal engineers, state agencies, and local levee boards. The levee boards in Louisiana have historically been — ”

“Political,” Baldwin said.

“Political. Yes. Patronage positions. The people who decide where the levees go are not the people who live behind them.”

“Now you’re talking about something,” Baldwin said. He ordered coffee, black, and turned to look at the map. He studied it differently than McPhee had — not reading the contour lines and channel depths, but tracing the relationship between the levee system and the towns it enclosed. His finger followed the river south from Baton Rouge, past Plaquemine, past Donaldsonville, down through the parishes where the land gets lower and wetter and the names on the map shift from French to something older.

“This is a map of a promise,” Baldwin said. “That’s what a levee is. It’s a promise that the land behind it is safe. And every year the promise is kept, the people behind it invest more — they build houses, they raise children, they plant gardens, they bury their dead in the ground because the ground is supposed to be dry ground. The longer the promise holds, the more it costs when it breaks. A levee that fails after ten years destroys property. A levee that fails after a hundred years destroys a civilization. The infrastructure isn’t protecting those people. It’s trapping them.”

McPhee took a slow drink of his coffee. “That’s a way of seeing it. Another way of seeing it is that without the levees, most of southern Louisiana would be uninhabitable. The entire alluvial valley of the Mississippi, from Cairo, Illinois, to the Gulf, would flood regularly. Millions of people would be displaced. The levees aren’t a trap. They’re the reason the place exists.”

“The place existed before the levees,” Baldwin said. “The Choctaw lived here. The Houma. They didn’t need the Army Corps of Engineers to tell the river where to go. They lived with the river’s schedule. They moved when it flooded. The levees didn’t make the place habitable. The levees made the place exploitable. Cotton. Sugar. Oil. You can’t run a plantation on land that floods every spring. You need the water controlled so the crop can grow and the labor — the enslaved labor, let’s not be delicate — can be fixed in place. The first levees in Louisiana were built by enslaved people. They were digging their own prison walls.”

I opened my mouth to say something and then closed it. I had been thinking about this essay as a nature piece — a McPhee-style examination of a physical system, the Mississippi’s hydrology, the engineering that constrains it, the geological forces that will eventually overwhelm the engineering. Baldwin was pulling the floor out from under that frame. The levee wasn’t a feature of the landscape. The levee was an instrument of power. The nature in this story was not the river. The nature in this story was the human arrangement built on top of the river, and the refusal to see that arrangement as temporary.

“Can it be both?” I asked. “An engineering story and a moral story?”

“It is both,” McPhee said. “Everything I’ve written about the control of nature is both. The Old River Control Structure is an engineering marvel and a monument to human arrogance. I don’t think those things contradict each other. The engineers I interviewed in Louisiana were intelligent, serious people who understood the river’s geology better than anyone alive, and who also understood that they were engaged in a project that would, over geological time, fail. They knew. They just didn’t have a better option. You can’t un-build New Orleans.”

“You can’t un-build it,” Baldwin agreed. “That’s the American situation in a sentence. You can’t un-build the thing you built on top of the thing you stole on top of the people you destroyed. You can only maintain it. And maintenance is a kind of forgetting. Every time you repair the levee, you are choosing not to ask why the levee is there, who it serves, what it replaced. Maintenance is the way America avoids looking at what it has done. We call it upkeep. We call it infrastructure spending. We call it the responsible thing. But what we are maintaining is the arrangement — the specific arrangement of who is safe and who is not — and every dollar spent on maintenance is a dollar not spent on asking whether the arrangement is just.”

The waitress refilled McPhee’s coffee. He thanked her and watched Baldwin with the patient attention of a man who has spent decades interviewing people who think differently than he does and has learned that the best reporting happens when he stops steering.

“I don’t disagree with any of that,” McPhee said, after a pause. “But I don’t think the moral argument changes the hydrology. The river doesn’t care about justice. It doesn’t care about the arrangement. It is moving sediment from the interior of the continent to the Gulf of Mexico, and it will take whatever path requires the least energy, and when the levees prevent it from taking that path, it builds pressure, and pressure finds weakness, and weakness becomes failure. That process is indifferent to everything you’re describing. The injustice of who lives behind the levee and the physics of what the levee is holding back are two different problems.”

“They are the same problem,” Baldwin said, and his voice shifted into something harder, more cadenced — the voice I recognized from his essays, the one that builds its sentences like a prosecutor’s closing argument. “They are the same problem because the people who are most at risk from the physics are the people who were placed at risk by the politics. Lower Ninth Ward. Plaquemines Parish. The communities downriver, where the levees are lower and the land is sinking and the salt water is coming in. These are not random populations. These are the people who were given the worst land because they had no power to refuse it, and now the river is coming for the worst land first, and the people with power — the people upriver, behind the higher levees, on the higher ground — will describe what happens as a natural disaster. A natural disaster. As though nature selected its victims at random.”

McPhee was quiet. He rolled the edge of the map between his fingers.

“You’re right about the distribution of risk,” he said. “The levee system is not equal. The protection is not equal. The areas that flood first in an overtopping event are the areas with the lowest levee crests, which are the areas with the least political influence, which are — I take your point. The geography of vulnerability in the Mississippi Valley is not an accident. It was engineered.”

“By the same people who engineered the levees,” Baldwin said.

“By many of the same institutions. Yes.”

I asked McPhee about the Old River Control Structure — the specific piece of infrastructure where the Army Corps prevents the Mississippi from abandoning its current channel and joining the Atchafalaya. I’d read his account of it, the 1973 flood that nearly destroyed the structure, the frantic night when the engineers thought they were losing the river. I wanted to know what happens if they actually lose it.

“New Orleans loses its water supply,” McPhee said, matter-of-factly. “The refineries between Baton Rouge and New Orleans lose their cooling water. The port of New Orleans — the largest port complex in the Western Hemisphere — becomes a port on a tidal estuary instead of a river. Morgan City, on the Atchafalaya, gets the full discharge of the Mississippi and is destroyed. The economic damage is incalculable. The word the Corps uses is ‘unacceptable.’ The scenario is simply categorized as unacceptable, which means it is not planned for, which means if it happens — ”

“You have no plan for the unacceptable,” Baldwin said. “That is America’s autobiography. We decide what is unacceptable and then we refuse to imagine it. Slavery was unacceptable to confront, so we had a war. Reconstruction was unacceptable to complete, so we abandoned it. Segregation was unacceptable to defend, so we renamed it. And the river — the river changing course is unacceptable, so we pretend it can’t happen. We pour concrete and drive sheet piling and build wing dams and tell ourselves we have controlled the thing. But we haven’t controlled it. We have postponed it. And postponement is not the same as safety, though it looks identical from the inside.”

“That’s the sentence,” I said, before I could stop myself. “Postponement is not the same as safety, though it looks identical from the inside. That’s the essay.”

Baldwin looked at me with an expression I couldn’t quite read — not pleased, not displeased, something more like wariness. “Don’t write the essay around a sentence. Write the essay around the levee. A specific levee. One you can stand on and look at the water and look at the houses and feel the dirt under your feet and know that the dirt was put there by human hands and that it is not enough.”

“Which levee?” I asked.

McPhee pulled the map closer. “The Mississippi River levee south of Donaldsonville, where it runs through Assumption Parish. The land behind it is sinking at approximately ten millimeters per year due to subsidence. The levee was designed for conditions that no longer exist. The river stage is higher than it was when the levee was built because the channel has aggraded. The Gulf is closer than it was because the delta is eroding. The community behind the levee is predominantly Black, predominantly poor, and predominantly unaware of the subsidence rate.”

“They know the water is getting closer,” Baldwin said. “They don’t need a subsidence rate to know that. They can see it. The yards that didn’t used to flood. The roads that wash out now when it rains. The insurance companies pulling out. The knowledge is in the body. It’s in the foundation cracks and the mold in the walls and the way the children have learned which streets to avoid after a storm. You don’t need an Army Corps report to tell you your ground is sinking. Your ground is sinking.”

McPhee nodded. “The data and the lived experience are converging. The question is whether anyone will act on either before the convergence becomes a catastrophe.”

“The question,” Baldwin said, “is whether the catastrophe is the flood or the century that preceded the flood. I keep coming back to this. You want to write about the moment the levee fails. The drama of the breach. The water pouring in. But the breach is not the catastrophe. The catastrophe is the hundred years of decisions that put people behind an inadequate wall and called it civilization. The water is just the bill arriving.”

I sat between them, feeling the two gravitational pulls — McPhee’s patient accumulation of physical fact, Baldwin’s insistence that every physical fact is also a moral fact — and I was not sure I could hold both. The essay McPhee would write would be a masterpiece of engineering narrative, the kind that makes you understand a system so completely you can feel it straining. The essay Baldwin would write would be a sermon, the kind that doesn’t let you sit comfortably in your knowledge because it keeps asking what your knowledge is for. The essay I needed to write had to be both, and neither man seemed convinced that both was possible.

“You’re going to have to choose,” McPhee said, reading my hesitation. “At some point in the essay, you will arrive at the levee, and you will look one direction and see the river, and you will look the other direction and see the houses, and the prose will turn. It will turn toward one or the other. It cannot face both ways.”

“It has to face both ways,” Baldwin said. “That’s what a levee does. It has two sides. The essay should be built like the thing it describes.”

McPhee almost smiled. “A levee isn’t built to face both ways. A levee is built to face the river. The houses face away.”

“Exactly,” Baldwin said. “The houses face away. The people in the houses face away. The entire arrangement is designed so that you can live your life without looking at the thing that will end it. That’s the architectural principle of American denial. Build the wall. Turn your back. Tend your garden. And when the water comes over the top, call it an act of God.”

The map was still spread across the table, the blue lines of the river branching and re-branching through the delta, the levees drawn as confident black lines on either side, the whole system looking, from above, like a long wound held closed with sutures. McPhee began rolling it up, carefully, corner to corner, the way you handle a document that has been unrolled too many times.

“Walk the levee,” he said to me. “Before you write anything, walk it. Both sides. Measure the freeboard — the distance between the water surface and the levee crest. Talk to the people who maintain it. Talk to the people who live behind it. Get the soil composition, the compaction data, the seepage rates. Understand the object. Then write about it.”

“And when you walk it,” Baldwin said, “ask yourself who you don’t see. Who used to live there. Where they went. Why they left. The levee is not just holding back the river. It’s holding a story in place. And the story is coming apart whether anyone writes about it or not.”

I paid the check. McPhee folded the map into a precise rectangle and tucked it back under his arm. Baldwin was already standing at the door, looking out at the parking lot, which was flat and paved and had been, not long ago in geological terms, the bottom of a river.

“One more thing,” McPhee said. “The levee at Donaldsonville. It was overtopped in 2016 during an unnamed storm — not a hurricane, not even a tropical system. Just rain. Enough rain, sustained long enough, and the river rose past the design capacity, and the water came over, and the Corps called it an anomalous event. Anomalous. From the Greek — irregular, uneven, departing from the general rule.”

“The general rule,” Baldwin said from the doorway, “was never general. That’s what I’ve been trying to tell you.”

McPhee didn’t answer. He was looking at the map again, though it was folded and there was nothing to see.