MY NEIGHBOR KEPT A 40-YEAR-OLD COMBINE THAT EVERYBODY LAUGHED AT. WHEN THE KILL SWITCH ACTIVATED AND MY FIELDS WERE DROWNING, HE SHOWED UP WITHOUT SAYING A WORD. WHY DID HE KEEP IT RUNNING?

The rain didn’t stop that afternoon, but it slackened just enough for a man to work without drowning. Gerald Harms backed the 6600 off the flatbed trailer with the kind of slow, precise movements that come from doing something ten thousand times. The old diesel rattled to life, a sound that was all mechanical clatter and no computer hum, and he drove it straight into the East Field like he’d been waiting for the invitation.

I stood at the edge of the field and watched him make that first pass. The soybeans were bent and heavy with water, the ground soft enough that the tires left tracks that filled up as soon as they were made. Gerald didn’t rush. He ran the header low, the reel turning at a speed that looked too slow to do anything useful, and he watched the crop feed into the machine the way a man watches something he’s done since before I was born.

After a hundred feet he stopped the machine, climbed down, and walked back along the row. I walked out to meet him, the rain soaking through my jacket, and found him crouched in the stubble counting beans on the ground.

— How’s it look? I asked.

He held up a hand, palm open, showing me the few beans he’d gathered. “Could be better. Concave’s a little tight for this moisture. Give me a minute.”

He walked back to the machine, reached into a toolbox mounted on the side, and pulled out a wrench that had seen so much use the metal was polished smooth in the grip. The concave adjustment on a 6600 is a physical lever connected to a linkage, and he worked it the way you’d adjust a stubborn window latch, a quarter inch at a time, listening to the change in the machine’s note while he did it. Then he climbed back into the cab and made another pass.

I walked behind again. This time the ground was cleaner. A few beans here and there, nothing that would change the bottom line. Gerald came down again, checked the sample coming out of the clean grain elevator, rolled a handful of beans between his fingers, and nodded.

— That’ll do. You want to run it, or you want me to?

I looked at the machine. I’d never run anything that old in my life. My whole operation was built around the kind of equipment that needed a password to start. But I wasn’t going to stand on the edge of my own field and watch a seventy-one-year-old man harvest my crop while I did nothing.

— Talk me through it.

He did. Not in a manual way, not like a teacher standing at a whiteboard. He just showed me. The throttle here, the hydrostatic lever there. The cylinder speed set at this RPM for these conditions. The concave adjustment where he’d left it, but if the rain picked up again, back it off a quarter turn. The cleaning fan at this setting, the chaffer and sieve at these openings. He pointed to the analog gauges, the ones that told you oil pressure, coolant temperature, and engine speed, and said, “If the oil pressure drops below here, shut it down. Anything else, you’ll hear it before it becomes a problem. The machine will tell you. You just got to listen.”

I climbed into the cab of a combine that was older than my son. The steering wheel was thin and hard, the seat worn through in places, the windshield streaked with forty years of fieldwork and one morning of fresh rain. I put my hand on the hydrostatic lever, pushed it forward, and the 6600 moved into the soybeans.

The first hundred feet I was tense, waiting for something to go wrong, waiting for a warning light that didn’t exist on a machine that had no computer to flash one. But the machine just worked. The header fed the crop in, the cylinder threshed it, the straw walkers shook the grain free, and the clean beans poured into the tank. It was not fast. Four miles an hour felt like crawling compared to the modern combine sitting dead in my shed. But it was steady, and the sample in the tank looked good, and I was harvesting.

I called my son, Kyle, from the cab. He was out on the far side of the farm doing fall tillage, and when I told him what was happening, there was a long pause on the other end of the line.

— Dad, are you telling me you’re running Gerald Harms’s museum piece?

— I’m telling you we’re getting the beans out. Come over and pull the International out of the shed. The 1480. We’re running two machines.

The 1480 was our old combine, the one we’d kept around for small fields and tight corners but hadn’t used in three years because the newer machine was bigger and faster and had all the precision features. It was from the early ’80s, still a conventional machine, still no computer to speak of. Kyle had learned to run it when he was sixteen, and he hadn’t touched it since we’d bought the new combine. But he pulled it out of the shed, changed the fuel filter that had gummed up from sitting, checked the belts and the oil, and had it in the field by late afternoon.

For the next four days, the two of us ran those machines through the rain-dampened soybeans. Gerald stayed the first day, walking between the two combines, checking the loss behind each one, making small adjustments to the settings on the 6600 until he was satisfied I could manage it myself. By the second day, he trusted me enough to go back to his own place, but he made me promise to call if anything changed.

— If the rain stops and the sun comes out and those beans dry down, you’re going to need to tighten the concave up. Not much. Just a little. You’ll feel it.

I nodded like I understood, and to my surprise, I did. By the third day I could tell the difference between a machine that was set right and one that was a quarter turn off, just from the sound of the cylinder and the look of the grain sample. It was not knowledge I’d learned from a manual. It was knowledge I’d learned by doing it wrong once and seeing the difference.

We lost beans. There was no way around it. The rain had swollen the pods, and when the sun finally broke through on the afternoon of day three and started drying things down, some of those pods split open before we could get to them. The moisture dockage at the elevator was real. The harvest loss from running two smaller, slower machines instead of one large modern one was real. When I sat down at my kitchen table after the last load was hauled in and ran the numbers, I figured my total loss at somewhere around twenty-two thousand dollars.

It was a real number. It hurt. But it was not the number that kept me awake that night.

What kept me awake was the number I’d heard about the Hennessey brothers.

The Hennesseys farmed about three miles east of me. Two brothers, Mike and Travis, who’d taken over the operation from their father a few years back. They’d been among the first in the county to buy into the new precision agriculture platform, and both of their machines, S770s they were still making payments on, had been locked out the same morning mine was. They had waited for the patch.

Their logic had been sound. The manufacturer was a reputable company. The engineers were working on it. The patch would come. Spending money on a custom harvester when the fix might arrive any hour seemed like throwing good money after bad. So they waited on day one, and they waited on day two, and by the time day three came and the first patch didn’t fix their machines, they started making calls. But by then, every custom harvester within a hundred miles was already booked. The ones who weren’t booked couldn’t get there for another week. And the rain was falling, and the beans were swelling, and the window was closing.

They found a guy eventually, a custom harvester from three counties away who ran a different brand of combine that hadn’t been affected by the update. He got there on day six, the same day the final patch arrived. By then, the beans had been through four days of rain and two days of partial drying, and the moisture content was a mess. The custom harvester charged premium rates because he knew he was the only option. On top of the moisture dockage, on top of the shrink, on top of the harvest loss from running a machine that wasn’t dialed in for their specific conditions the way Gerald’s had been for mine, the Hennessey brothers’ total loss came in somewhere north of forty-one thousand dollars.

Mike Hennessey told me that number himself, standing in the parking lot of the co-op a week after harvest was over. He said it the way a man says something he still can’t quite believe has happened to him.

— Forty-one thousand, Dale. That’s a down payment on a new machine. That’s a year of college for my daughter. That’s gone.

I didn’t know what to say to him. What was there to say? That I’d gotten lucky because I had a neighbor who’d refused to throw away an old machine? That didn’t seem right. It wasn’t luck. It was Gerald. And Gerald hadn’t been lucky. He’d been deliberate.

A few days after the harvest was finished, I drove over to his place to thank him properly. It was a Saturday morning, the kind of flat gray November day that Iowa does so well, the sky the color of old dishwater, the fields bare and brown and looking like they were resting. I pulled into his yard and found him in the machine shed, the 6600 parked in its usual spot, its panels open, Gerald underneath it on a creeper checking the feeder house chain.

I stood in the door of the shed for a minute, just watching him work. He didn’t notice me at first, or if he did, he didn’t stop what he was doing. His hands moved the way a man’s hands move when they’ve been doing the same work for fifty years, certain and unhurried. He tightened a bolt, slid out from under the machine, and looked up at me without any particular expression.

— Dale.

— Gerald. I wanted to say thank you.

He picked up a rag and wiped the grease off his hands. He didn’t wave away the thanks. He didn’t say it was nothing. He just nodded once, like he was accepting a fact that happened to be true.

— How’d the numbers shake out?

I told him. The moisture dockage, the harvest loss, the twenty-two thousand. He listened without interrupting, and when I was done, he thought about it for a moment.

— Twenty-two is not nothing. But it’s better than forty.

— That’s what Mike Hennessey lost. Forty-one.

Gerald’s face didn’t change. He’d known the Hennessey boys since they were children. He’d known their father before them. But he didn’t say anything about what they’d lost. He just nodded again, the same slow nod, and looked at the 6600 sitting in the shed.

— Machine did what it’s supposed to do.

He said it the way a man says something that is obvious to him and might not be obvious to anyone else. Then he put the rag down on the workbench and turned back to me.

— You got coffee at your place, or you want some here?

I stayed for an hour. We drank coffee in his kitchen, a room that hadn’t been remodeled since the ’70s, with cabinets made of real wood and a Formica countertop that was worn through in places. He asked me about the rest of my operation, about the corn that was still standing, about whether I’d thought about what I was going to do differently next year. I said I wasn’t sure. He didn’t push. He just refilled my coffee cup and let the silence sit.

When I got up to leave, he walked me out to my truck. The wind had picked up, the kind of November wind that finds every gap in your coat, and I stood there with my hand on the door handle for a moment before getting in.

— Gerald, can I ask you something?

— You can ask.

— Ten years ago, when you sold your last modern combine and kept the 6600 instead. Everybody thought you were stubborn. Did you know something we didn’t?

He looked at me for a long moment. Then he looked past me, out at the bare fields stretching toward the horizon, the same fields he’d been looking at his whole life.

— I didn’t know anything, Dale. I just didn’t trust anything I couldn’t fix myself. That’s not the same thing.

He said it without pride, without any sense of I-told-you-so. Just as a fact. Then he turned and walked back into the machine shed, and I got in my truck and drove home.

The dealership conversation happened two days later, and I only know about it because Brett Sellers told me himself, months afterward, when he drove out to the farm to apologize.

I’d gone into the dealership on day two, after the first patch had failed. I’d been desperate and angry and not really thinking straight, and I’d mentioned, almost as an afterthought, that Gerald Harms had offered to bring his old 6600 over to run the beans. I wasn’t asking Brett’s permission. I was just talking, the way a man talks when he’s trying to figure out his options.

Brett’s face had done something complicated. It wasn’t anger, and it wasn’t dismissal. It was something closer to bewilderment, the expression of a man who has spent his entire career believing that newer is better and has just heard something that doesn’t fit inside that belief. He’d told me the 6600 would shatter my beans. He’d said it with the confidence of a man who had sold a lot of combines and knew their specifications. And I’d nodded and said I’d think about it, and then I’d left.

What Brett didn’t know, and what I didn’t know at the time, was what happened in the dealership about two seconds after I walked out the door.

Brett turned to his parts manager, a guy named Rick who’d worked there for ten years, and said, “Gerald Harms and that old 6600. Jesus.” He said it the way you say something you find both sad and slightly ridiculous. The way you talk about a man who’s too attached to the past to see the present clearly.

Rick smiled and shook his head. Neither of them said anything else about it. They moved on to the next call, the next crisis, the next farmer whose machine was locked out and couldn’t be fixed.

That conversation sat in the back of Brett’s mind for the next two weeks. While the patches rolled out and the rains continued and the harvest limped along, he kept thinking about the look on my face when he’d told me the old machine would ruin my crop. He’d said it with such certainty. He’d believed it completely. And then, gradually, through the grapevine of a small farming community, he’d started to hear what had actually happened.

He’d heard that Gerald had dialed the 6600 in by hand, that the loss behind the machine was within normal parameters, that the grain sample was clean, that I’d gotten my crop out while the Hennessey brothers were bleeding money waiting for a patch that came too late. He heard it from three different people before he decided he needed to see for himself.

On a Saturday morning in late November, about two weeks after harvest was done, Brett Sellers got in his pickup, the one with the dealership logo on the door, and drove out to Gerald Harms’s farm.

He found Gerald in the machine shed, same as I had, going over the 6600, checking everything the harvest season had used, replacing what needed replacing, lubricating what needed lubricating. Gerald was on his back under the feeder house when Brett pulled up, and he didn’t notice he had company until he slid out from under the machine and saw Brett standing in the doorway of the shed.

Gerald sat up slowly. He didn’t say anything. He just looked at Brett with the patient, unreadable expression of a man who has seen a lot of things and is willing to see one more.

Brett said, “I owe you an apology, Gerald, and I owe Dale one, too.”

Gerald picked up the rag he’d been using and wiped his hands. He didn’t say that Brett didn’t owe him anything, the way some men would have. He didn’t wave it off. He just waited.

Brett took a breath. You could see he’d been thinking about this moment for a while, and now that it was here, he wasn’t quite sure how to say what he’d come to say.

— I told Dale that machine would shatter his beans. I was wrong.

He paused. Gerald still didn’t say anything.

— I’ve been selling combines for fifteen years. I’ve never seen a 6600 set up the way yours is. I didn’t think it could do what it did.

Gerald looked at the machine for a moment. The 6600 sat in the shed the way it always sat, enormous and quiet. The paint worn in the places where forty years of use had worn it. The cylinder housing scratched and dented in the particular way that working equipment gets scratched and dented. It did not look impressive. It looked like what it was, an old machine that had been kept running by someone who understood it.

— It’s not magic, Gerald said. It’s just set right. You set any machine right, it’ll do what it’s supposed to do.

Brett nodded. He looked at the 6600, and then he looked at Gerald, and he asked a question that told me later he’d been turning over in his mind since the whole thing started.

— The thing I keep thinking about is the software. The fact that it locked everybody out. Does that bother you? That they can do that?

Gerald considered this for a moment. He stood up, his knees taking a moment to straighten, the way old knees do when they’ve spent too long on concrete.

— It bothers me that people bought machines they can’t fix themselves. The locking out is just what happens after that. You give somebody else control of something you depend on, you got to live with the decisions they make.

He said it the way he said everything, calm and matter-of-fact, without heat. Then he tilted his head toward the house.

— I got coffee inside if you want it.

Brett Sellers stayed for an hour. He drank two cups of coffee in Gerald’s kitchen, at that same Formica countertop, and he asked Gerald about the 6600. About the cylinder speed settings for different moisture conditions. About how Gerald knew where to set the concave when the beans were wet versus when they were dry. About how he checked his loss and what he looked for in the grain sample.

Gerald answered every question with the patience of a man who has spent his life learning things and does not mind sharing them. He did not lecture. He did not make Brett feel foolish for not knowing. He just explained, the way a man explains something he understands completely and has no reason to be secretive about.

— The manual gives you starting points, he said. But the manual doesn’t know your field. It doesn’t know the weather you’ve had or the variety you planted or what happened to those beans in June when it was too wet and in August when it was too dry. The machine will tell you what it needs if you pay attention. The manual can’t do that.

Brett wrote some of it down. Not on his phone, not on a tablet, but on a small notepad he kept in his pocket, the way a man writes things down when he cares about remembering them. When he finally left, he sat in his pickup for a full minute before starting the engine. He looked at the machine shed, at the 6600 visible through the open door. Then he started the truck and drove back to Harlan.

That visit changed something in Brett. I didn’t know it at the time, but I would find out later, over the course of the next few years. He never stopped selling modern equipment. He couldn’t have, even if he’d wanted to; that was the business he was in, and in most conditions and most years, the new machines were genuinely better. Faster, more efficient, more precise. But he started asking different questions when he sat down with customers.

— What’s your backup plan if the primary machine goes down during harvest? he would ask. Have you thought about it? Do you have old iron on the farm? Something that doesn’t need a satellite connection to start?

Some of his customers thought it was an odd question for a dealer to ask. A few of them even said so out loud. But then Brett would tell them about what happened in the fall of 2019, about the firmware update and the licensing verification module and the four days of rain while the machines sat paralyzed in their sheds, and they would understand.

The winter after that harvest was a long one. The fields froze and the wind came down from the Dakotas and there was nothing to do but maintain the equipment and think about what had happened and plan for the year ahead. I spent a lot of that winter in my own machine shed, looking at my modern combine, the JDS790 that had sat dead while an old 6600 harvested my beans.

It was a magnificent machine. Even sitting still, even shut down for the winter, you could see the engineering that had gone into it. Every system was optimized, every component was designed for maximum efficiency. In good conditions, in the conditions it was built for, it could harvest more acres in a day than the 6600 could manage in a week. I wasn’t going to sell it. It was the right machine for my operation in most years and most conditions.

But I couldn’t stop thinking about what Gerald had said. About what it meant to own a machine you couldn’t fix yourself. About what it meant to depend completely on something that someone else could shut off with a software update you never asked for.

In January, I drove over to Gerald’s place again. The snow was thick on the ground, and the machine shed door was closed against the cold, but the lights were on inside and I could hear the sound of a radio playing old country music. I knocked on the side door and Gerald let me in, and I stood there in the warmth of his shed, surrounded by tools and spare parts and the particular smell of diesel and grease and old metal that hangs in the air of any space where a man has worked on machines for decades.

— I’m thinking about buying a backup machine, I said. Something old. Something conventional. Something I can fix myself if I have to.

Gerald didn’t look surprised. He just nodded and pulled a shop stool over for me to sit on.

— You got something in mind?

— I was thinking a 7720. Maybe early ’80s. I’ve heard good things about them.

— Good machine, Gerald said. The 7720 is just a bigger 6600 with a few improvements. The feeder house is better, the cab is a little more comfortable, but the threshing system is the same basic design. You can set it up the same way.

— Will you help me find one?

He said yes without hesitation, the way he’d said yes about the harvest. And over the next few weeks, Gerald and I looked at five different machines before we found the right one. It was a 1982 John Deere 7720, parked in a barn about an hour south of us, owned by a retired farmer who’d stopped row-cropping and was selling off his equipment. The machine had sat for two years, but it had been kept indoors, and the engine turned over on the first try after Gerald put a new battery in it.

I paid eleven thousand dollars for it. Not nothing, but not a fortune either. I spent another three thousand getting it into reliable running condition, replacing belts and filters and hoses, going through the threshing system, checking every bearing and chain and hydraulic line.

Gerald spent two Saturdays in my machine shed that spring, going through the machine with me, showing me the concave adjustment, the cylinder speed, the cleaning shoe settings. He showed me how to listen to the machine, how to tell from the sound of the separator whether it was working right or struggling, how to check the loss by walking behind the combine and counting beans on the ground.

Kyle, my son, was there for both of those Saturdays. He was twenty-two that year, fresh out of Iowa State with a degree in Ag Systems Technology, and he’d grown up around the kind of equipment that was controlled by touchscreens and GPS. But he sat on a shop stool and watched Gerald work and asked questions the entire time.

— How do you know when to tighten the concave? Kyle asked on the second Saturday, as Gerald was showing him how to adjust the cylinder bars.

Gerald thought about the question for a moment. Not because he didn’t know the answer, but because he was deciding how to explain something he’d known for so long that explaining it had become difficult.

— You check the sample, he said. If you’re cracking beans, the cylinder speed is too high or the concave is too tight. If you’re leaving beans in the pods, the concave is too loose or the cylinder is too slow. You adjust one thing at a time and you check again. After a while, you’ll know from the sound of the machine before you even check the sample. It’ll hum when it’s right and it’ll rattle when it’s wrong. You’ll learn.

He said it with such calm certainty that I could see Kyle believing him. And by the end of the second Saturday, Kyle could set the machine himself. Not perfectly, not with Gerald’s fifty years of instinct, but well enough to run it through a field and get clean grain at the other end.

The spring of 2020 came, and with it came planting season, and for a few months the whole firmware incident faded into the background. There were seeds to put in the ground, fields to spray, the thousand daily tasks that make up a farming season and leave no room for thinking about anything else. But I didn’t forget. None of us did.

The manufacturer, to its credit, had not been indifferent to what happened. In the months after the 2019 harvest, they had issued a revised firmware update policy. Any mandatory update that could affect machine operability during active harvest season now came with a seventy-two-hour operator opt-out window. They established a dedicated harvest season support line with guaranteed four-hour response times. They created a compensation program for the operators who had been affected, though the payments, when they came, were not generous by any reasonable measure. They covered maybe a third of what most farmers had lost, and the paperwork required to claim them was enough to make a man wonder if it was worth the trouble.

The company’s engineers were not bad people. The executives in the boardroom in Moline were not villains twirling mustaches. They had made a mistake in a complicated system, a licensing verification module with a bug that had been deployed through an automated update process that no one had fully thought through, and the consequences had landed on a handful of farming operations in western Iowa with a weight that those engineers and executives would never have to carry themselves. They tried to fix it. They made real changes. The problem did not repeat itself in 2020 or 2021.

But in Shelby County, Iowa, something else had happened, too.

Two other farms in the area had done similar things to what I’d done. Not all of them had bought conventional machines. One of them, a younger guy named Eric who farmed about five miles north of me, had bought a used combine of a different brand entirely, specifically to avoid single-brand dependency. If one manufacturer’s software locked his primary machine out, he reasoned, he wanted a backup that ran on a completely different system. Another farmer, a woman named Linda who’d taken over her father’s operation, had bought a model year old enough to predate the precision agriculture software architecture that had caused the problem. She’d paid next to nothing for it at auction, and she’d put in several winter months of work getting it running again.

None of them abandoned their modern equipment. They weren’t making a statement about technology or progress or the evils of automation. They were making an insurance decision, the way a farmer makes insurance decisions, quietly and practically and without drama. You pay a premium every year for something you hope you never need, and if you never need it, you don’t think of the premiums as wasted. You think of them as the price of knowing that if the worst happens, you won’t be standing at the edge of your field watching your crop walk away from you.

Gerald Harms noticed all of this and said nothing about it. He did not take credit for the changes his neighbors were making. He did not point out that he had been doing this for ten years already, that he had kept the 6600 running while everyone else called him stubborn and sentimental and out of touch. He just noticed it the way he noticed everything, carefully, without comment, filing it away.

The harvest of 2020 was uneventful, and the harvest of 2021 was the same. The manufacturer’s revised protocols held. The modern machines ran as they were supposed to, harvesting more acres per day than the old machines could have dreamed of. My 7720 sat in the shed, maintained and ready, and there were days when I walked past it and wondered if I’d wasted my money.

Gerald would have told me I hadn’t. I didn’t need to ask him to know. The money wasn’t wasted because the machine was there, and because it was there, I slept better during harvest season than I’d slept in years. It’s hard to put a price on sleep, but if you’ve ever lost a crop because your equipment failed at the wrong moment, you know exactly what it’s worth.

The fall of 2022 was a different kind of test. There were no software issues that year. The manufacturer’s systems held, and the 2019 incident did not repeat itself. But the weather in September and October of 2022 brought its own kind of trouble, the kind of trouble that doesn’t come from a boardroom but from the sky.

Week after week of alternating rain and partial drying kept field conditions marginal and pushed harvest deep into November. The soybeans that came out in that window were high moisture, tough and damp and hard to thresh cleanly. The modern combines in the county, which are extraordinarily capable machines in the right conditions, were struggling. The precision harvest systems were doing their job, adjusting automatically to the changing moisture levels, but the conditions were at the edge of what the machines were designed to handle. Rotor losses were elevated. Grain quality was marginal. Every operation in the county was fighting the weather and fighting the clock.

I ran both machines that fall, my modern combine and the 1982 7720. Kyle ran the 7720 on the days when the conditions were worst, when the moisture was highest and the rotor losses on the modern machine were climbing past the point I was willing to accept. He’d pull ahead of me in the field, running the older machine through the toughest sections while I held back and waited for conditions to improve.

It was not a romantic arrangement. It was not a statement about old versus new or man versus machine or any of the things that city journalists like to write about when they visit farm country for a weekend and think they understand it. It was two machines doing what each of them did best, managed by people who understood both of them.

The 7720 is a conventional design, and the conventional design has characteristics that become advantages in certain conditions. In high moisture soybeans, the cylinder and concave system, set correctly, is more forgiving than a rotor. It’s less aggressive. It handles the tough, damp crop with a kind of mechanical patience that the higher-speed rotor system sometimes cannot match.

Kyle had learned to set the machine from Gerald, and he’d learned well. He adjusted the concave for every change in moisture, backed the cylinder speed down when the beans were tough, checked the loss monitor by walking the rows behind the machine. It was slower. It was always slower. But in those conditions, clean mattered more than fast, and the sample coming out of the 7720 was cleaner than what the modern machine could manage.

Gerald, of course, was harvesting too. He was seventy-four years old that fall, and he’d been running the 6600 through wet soybeans for fifty years. The conditions didn’t bother him. He’d seen worse. He’d seen everything. He ran slower than anyone, adjusted more frequently than anyone, checked his loss more carefully than anyone, and his grain sample was the cleanest in the county. That’s not local legend. That’s what the elevator manager told me when I hauled in a load of Gerald’s beans for him one afternoon after he’d had a problem with his truck.

— I don’t know how he does it, the manager said, looking at the sample. Every year, no matter what the weather does, Gerald Harms’s beans come in cleaner than anybody else’s. What’s he running, a magic machine?

— A 1979 6600, I said.

The manager looked at me like I was joking. I wasn’t.

One afternoon in October of that year, Brett Sellers drove past my field on his way to a service call at another farm. He slowed down when he saw the two machines working, the big modern combine and the old 7720 running parallel passes through the wet beans. Kyle was in the 7720, running a section near the road, and I was further back in the field with the modern machine, waiting for a patch of drier ground.

Brett pulled over. He didn’t get out of his truck. He just sat there in the cab with the engine idling, watching the two machines move through the rain-flattened soybeans, the old one managing the tough stuff, the new one resting until conditions improved. He’d been thinking since the fall of 2019 about the conversation he’d had with Gerald in the machine shed. He’d been thinking about what Gerald had said, that the locking out is just what happens after you give somebody else control of something you depend on. He’d thought about it in the context of his own business, about what it meant to sell machines his customers could not repair themselves, about what the terms of use actually said about who owned what.

He hadn’t come to any dramatic conclusions. He was not about to stop selling modern equipment. He had a job and a mortgage and the modern equipment was genuinely better in most conditions and most years. But he had started asking different questions when he sat down with customers. And watching my two machines work that afternoon, the old and the new side by side, each doing what they did best, something clicked into place in his mind.

He told me about it later, at the co-op over coffee. He said he’d realized that the question wasn’t which machine was better in the abstract. The question was which machine was better in these conditions, right now, and whether you had the option to choose. The farmers who had options, who had kept old iron around or bought a different brand or maintained a backup plan, were the ones who could adapt when conditions changed. The farmers who had bet everything on one machine, the most modern machine, the most capable machine under ideal conditions, were the ones who struggled when conditions stopped being ideal.

— That’s not a technology problem, Brett said. That’s a resilience problem. And I’ve been part of the resilience problem for fifteen years without realizing it.

He said it with the expression of a man who has organized his professional life around a particular set of beliefs and has just realized that those beliefs have a hole in them. It wasn’t a comfortable realization, but he didn’t run from it either. To his credit, he started talking about it with his customers. Not as a sales pitch, not as a way to move used equipment, but as a genuine question. What’s your backup plan? What do you do if your primary machine goes down during the critical window? Have you thought about having something on the farm that you control completely?

Some of his customers thought he’d lost his mind. Others understood immediately, especially the ones who remembered the fall of 2019.

Gerald Harms died in the spring of 2024.

It happened on a Tuesday morning in April, a day when the air was just starting to warm up after a long winter and the fields were beginning to dry enough for the first tillage passes. He was in the machine shed, as he always was in April, going through the 6600, getting it ready for the season that was coming. He’d been doing this every spring for as long as anyone could remember. Change the oil, check the belts, go through the threshing system, replace whatever had worn during the previous harvest, prepare the machine for the work that was ahead.

His son Paul found him there, beside the machine. A heart attack, the doctors said. Quick and massive and without warning. That’s how Gerald would have wanted it, if you’d asked him. Not in a hospital bed, not fading slowly over months of illness, but in his machine shed, doing what he’d done his whole life, with the 6600 close by.

The family buried him in the cemetery outside Earling, the same cemetery where his parents were buried and his grandparents before them. The funeral was packed. Farmers from four counties came, and so did people from town, and so did Brett Sellers, who stood in the back of the church in a suit that didn’t quite fit and looked like he was thinking about things that were hard to put into words.

After the service, I walked over to Paul Harms and shook his hand. He was forty-eight years old and he’d been farming alongside his father for twenty-five years, and you could see in his face that he was carrying the weight of everything that had just changed and everything that was now on his shoulders.

— Your dad was a good man, I said. The kind of good that’s hard to explain to people who didn’t know him.

Paul nodded. He’d been hearing variations of that all day, and he was handling it with the kind of steady composure that reminded me of Gerald.

— He talked about you, Paul said. After that fall, 2019. He never said much about it, but I could tell it mattered to him. Being able to help.

— He saved my crop, Paul. More than that. He changed how I think about a lot of things.

Paul nodded again. He looked out at the parking lot, at the line of pickups stretching down the gravel road, at the brown April fields waiting for the planters to start rolling.

— He left me the 6600, Paul said. Not in the will, I mean he left it to me in the shed. It’s still there. I’m going to keep it running.

— Good, I said. That’s good.

Paul did keep it running. That summer after Gerald died, he rebuilt the cylinder himself, following the notes his father had kept in a spiral notebook that lived on the shelf above the workbench in the machine shed. The notes were not formal. They were the kind of notes a man writes for himself. Reminders, measurements, observations about what worked and what didn’t, dates of repairs and parts used. Page after page of Gerald’s handwriting, small and neat, recording fifty years of accumulated knowledge about one machine.

The notes were detailed enough that Paul, who had worked alongside his father for decades but had never done a cylinder rebuild on his own, could follow them step by step. Torque specifications, clearance measurements, the particular order in which things had to be disassembled and reassembled. And in the margins, little observations that were not strictly mechanical but were somehow just as important. “Wet fall, ran concave loose, worked better.” “Dry beans, cylinder speed 600 RPM max or you’ll crack them.” “Replaced sieve, aftermarket brand didn’t fit right, went back to OEM.”

Paul told me about the notebook one evening when I ran into him at the gas station in Earling. He pulled it out of the glove box of his pickup to show me, and I stood there under the fluorescent lights and paged through it. It was like holding a piece of Gerald’s mind in my hands, all that knowledge written down not for anyone else but just so it wouldn’t be lost.

— The last thing he said to me about the 6600, Paul said, was sometime that winter before he died. I was thinking about maybe trading it in, getting something newer. He didn’t get mad. He just said, “Don’t let anybody talk you into selling it because it’s old. Old isn’t the same as wrong. Old just means it’s been tested longer.”

Paul put the notebook back in the glove box and closed the door.

— I’m not selling it, he said. Not ever.

The fall of 2024 came, and with it came a problem that closed the circle of this story in a way I don’t think anyone could have predicted.

It was a Thursday afternoon in October, and Paul Harms was in the cab of the 6600, combining his soybeans. The weather was marginal, not quite as bad as 2019 but not good either, and he was making slow progress through a field of wet beans when his phone rang.

It was my son, Kyle. He was twenty-six now, farming with me full-time since he’d graduated from Iowa State, and he’d grown up enough in the last few years that his voice on the phone had the kind of steadiness that comes from handling problems you didn’t expect.

— Paul, you got a minute?

Paul shut down the header and let the engine idle. The old diesel chuffed away beneath him, the same sound it had been making since before either of them was born.

— I got a minute.

Kyle told him about Travis Hennessey. The younger Hennessey brother, the one who had taken over the operation from his father a few years ago. Travis’s combine was down. Not a firmware issue this time, thank God for small mercies, but something mechanical. A rotor failure, the kind of catastrophic failure that happens without warning and takes a machine out of commission for days. His dealer had told him five days minimum for the repair, and that was if the parts came in on time.

— He’s got four hundred acres of soybeans standing, Kyle said. Wet conditions. The forecast looks like it did in ’19. He’s called the dealership and they don’t have a loaner. He’s called three custom harvesters and none of them can get there in under a week. He’s pretty well stuck.

Paul didn’t hesitate. He didn’t have to think about it. He didn’t make a mental calculation about what was owed or what was fair or whether Travis Hennessey deserved help after his family had lost forty-one thousand dollars in 2019 while the Harms machine ran through the rain. He just said it.

— Tell him to pull in the yard whenever he’s ready.

He said it the way his father would have said it. Without drama, without making Travis feel the weight of history that hung between their two families. He said it because that is what you do in Shelby County when a neighbor needs help, and because the 6600 was in the shed with a full tank of diesel, and because the soybeans were not going to wait.

Travis Hennessey pulled into the Harms yard that evening. It was nearly dark, the headlights of his pickup cutting through the gray October twilight, and Paul was waiting for him in the machine shed with the door open and the lights on. The 6600 sat in its usual spot, cleaned and maintained and ready to go, the same machine that had harvested my beans in the rain five years earlier.

Travis got out of his truck and walked over to the shed. He looked at the 6600 the way I had looked at it in 2019, with the particular expression of a man who is grateful and uncertain in equal measure. The machine was old. It didn’t look impressive. It looked like a museum piece that someone had forgotten to retire. But Travis had been there in 2019. He knew what the machine had done. He knew what his own family had lost while it ran.

— Paul, I don’t know what to say.

— You don’t got to say anything. Machine’s full of fuel. Let me walk you through the settings.

Paul showed him the 6600 the way Gerald had shown me. The concave adjustment, the cylinder speed, the cleaning shoe. He showed him how to check his loss by walking the rows behind the machine, how to listen to the sound of the separator, how to know when the settings were right and when they needed adjusting. He showed him the spiral notebook on the shelf above the workbench, the pages of his father’s handwriting, and told him that if he had questions about settings for different conditions, the answers were probably in there.

— He wrote down everything, Paul said. Fifty years of running this machine through every kind of weather Iowa can throw at you. If you run into something you don’t know how to handle, it’s in there.

Travis looked at the notebook. He opened it to a random page and read for a moment. I don’t know what he saw there, what particular note Gerald had written down years ago that was now going to help a Hennessey harvest his beans, but his expression changed. He closed the notebook carefully and put it back on the shelf.

— Thank you, he said. I mean it.

Paul nodded. He didn’t make a speech. He didn’t say anything about what had happened in 2019 or the money the Hennessey family had lost or the way the circle had come around. He just pointed to the fuel tank and told Travis where the extra diesel was if he needed it, and then he went back to the house and let Travis get to work.

Travis harvested his four hundred acres over the next four days. He ran the 6600 through wet soybeans and damp ground and the kind of marginal conditions that Iowa falls are famous for. The machine was not fast. It was never fast. But it was steady, and the grain sample was clean, and the loss behind the machine, when Travis walked the rows and checked, was within acceptable parameters.

On the last day, after the final load had been hauled to the elevator, Travis found Paul in the machine shed. Paul was under the 6600, checking the feeder house chain the way his father had always done after a stretch of heavy use.

— I want to pay you for this, Travis said.

Paul slid out from under the machine and sat up. He wiped his hands on a rag, the same gesture his father had made a thousand times, and he looked at Travis with an expression that was not unkind.

— Neighbors don’t charge neighbors, he said. If you want to do something, you can come help with the Harms corn when the time comes.

— I’ll be here, Travis said. He meant it.

He drove home that evening thinking about the 6600. He thought about the fact that it had no computer, no GPS, no satellite connection, no software license that could be revoked by an update pushed overnight from a boardroom a thousand miles away. He thought about the fact that Paul Harms could fix anything that went wrong with it using tools in the machine shed and knowledge in a spiral notebook. He thought about his own combine, the magnificent modern machine parked in his shed, faster and more capable and more precise in almost every measurable way, and about the terms of use agreement he had signed without reading carefully that gave the manufacturer certain rights regarding software updates and remote diagnostics and licensing verification.

He wasn’t going to sell his modern combine. It was the right machine for his operation in most years and most conditions. But he was going to think differently about what it meant to depend on it completely. He was going to think about what a backup plan looked like. What old iron, properly maintained, was worth. What it meant to have a machine that you controlled entirely, that nobody could lock out with a firmware update delivered while you slept.

He was going to think about the spiral notebook on the shelf above the workbench in the Harms machine shed, and what it represented. Not just a record of repairs, but a record of understanding. One man’s accumulated knowledge about one machine, written down so that it would not be lost when the man was gone.

Some things don’t become obsolete. They just wait for people to remember why they were built.

The 6600 sits in the Harms machine shed right now, tank full, concave set, the spiral notebook on the shelf above the workbench with Gerald’s handwriting still clear on every page. Paul runs it every fall, not because he has to, but because it does what it was built to do and because doing so keeps his father close in a way that the cemetery outside Earling can’t match.

The engineers in Moline will tell you that the modern machines are better, and in most ways and most conditions and most years, they are right. The new combines are marvels of efficiency and precision, capable of things that Gerald’s generation could never have imagined. But Paul Harms will tell you that his father’s machine does what it was built to do, and in every way that matters, he is also right.

And in Shelby County, Iowa, when the firmware locks and the rain is coming and the soybeans are standing in the field, the question is not which answer is correct. The question is which machine is running.

That’s the thing about knowledge that’s been tested for fifty years. It doesn’t need a software update. It just needs someone who paid attention and wrote it down, and a son who was paying attention enough to keep the tank full.

I think about Gerald often now. I think about him every time I walk past the 7720 in my shed, every time I check the concave adjustment before harvest, every time I hear a machine that’s running right and remember what it sounds like when it isn’t. I think about what he said to Brett Sellers in his kitchen, that giving someone else control of something you depend on means you have to live with the decisions they make. I think about what he said to Paul, that old isn’t the same as wrong, that old just means it’s been tested longer.

And I think about the morning in October of 2019 when I stood at the edge of my East Field and watched thirty-six thousand dollars walk away from me, and about the sound of Gerald’s pickup pulling into my yard with the 6600 on the trailer behind it. He didn’t make a speech. He didn’t tell me he’d been right all along. He just asked if I wanted to run my beans through his machine.

That’s what neighbors do in Shelby County, Iowa. But Gerald Harms did something more than that. He showed us that the knowledge that matters most can’t be downloaded in a firmware update. It has to be learned by doing, preserved by writing it down, and passed on to whoever is paying enough attention to receive it.

The 6600 is still running. The spiral notebook is still on the shelf. And somewhere in the machine shed on the Harms farm, if you listen close enough on a quiet afternoon, you can almost hear Gerald’s voice, patient and certain, explaining the concave adjustment one more time to anyone who’s willing to learn.