In Praise of Maintenance

She came by the habit professionally. For 22 years she worked as a mender at the woolen mill in Asterby, one of a row of women who sat under the tall north windows of the finishing room and went over every length of cloth before it left the building, finding each broken thread and slub and float and reweaving it so that the buyer’s inspector, running a thumb across the place, would feel nothing.2 Her job, in other words, was to make sure that nobody would ever know she had been there. She was very good at it. The mill’s wage book records her as “mender, first grade,” and pays her a little less, in every year I checked, than the men who loaded the finished bolts onto the trucks.^*

This is an essay in praise of that kind of work, the work of keeping things going, and I should admit at the outset that praise is the cheapest thing I could offer it. We are perfectly willing, as a culture, to admire maintenance in the abstract. What we will not do is notice it, fund it, credit it, or arrange our institutions around it. The builder gets the ribbon and the plaque; the keeper gets the budget line that is cut first in a lean year. We have lived inside that arrangement for so long that it has come to seem like weather. It is not weather. It is a choice, and an expensive one, and the bill has started to arrive.

The man with the ladder

The building I live in, in Halbridge, was finished in 1958. It is a three-story block of brown brick on the hill above the harbor, 18 apartments around a stairwell that smells faintly of floor polish, and in the lobby, beside the mailboxes, there is a brass plaque. It records that the Carrow Building was designed by one man, built by a second, and opened by a third, a member of the city council whose name is cut noticeably larger than the other two.3 I walked past that plaque twice a day for 11 years before it occurred to me that it was, in its small way, untrue. Those three men attended the building’s beginning. None of them has had anything to do with it since.

The person who has had everything to do with it is Teodor Vass, who has been its superintendent for 31 years. Teodor is 68, short and broad and unhurried, and he owns a 32-foot aluminum extension ladder that he carries, collapsed, on one shoulder, with the ease of a man carrying home a loaf of bread. Every October, before the first real storm comes in off the sea, he goes up that ladder and clears the gutters on all four sides of the building; then he clears the downpipes, then he goes down to the courtyard and rods out the drains, and then he waits for the rain. I asked him once how he knew when to start. He said he started when the plane trees on Carrow Hill had dropped about half their leaves. “Not all,” he said. “Half. You wait for all, it rains first.”

He keeps a daybook, a lined school exercise book with the dates written in the margin, and this summer he let me copy out a page from last October, the week before the first big storm.

I have read that page many times now, and the line I keep coming back to is the last one. More than an inch of rain fell on Halbridge that Friday night. The gutters on two nearby streets overflowed, and a basement on the next block flooded to the depth of a hand. In the Carrow Building, nothing happened. Teodor wrote it down because he writes everything down, but there was nothing to write, and so he wrote that. The nothing was the product of a week’s work. It is the only product maintenance ever makes.

The work that looks like nothing

Maintenance is the work whose success looks exactly like nothing happening. That is its central difficulty, and almost every other problem follows from it. When a builder succeeds, there is a building. When a maintainer succeeds, there is an absence: no flood, no outage, no collapse, no news. You cannot photograph an absence or cut a ribbon across it.4 The better the work, the more completely it disappears, so that the most accomplished maintainers—my grandmother at her north window, Teodor watching the plane trees—are precisely the ones whose labor is least likely to be seen.

The second difficulty is narrative. Stories, including the stories institutions tell about themselves, are built from beginnings and endings: a thing is founded, a thing is opened, a thing is lost. Maintenance is all middle. It has no climax and no conclusion; it is the same Tuesday, over and over, for 31 years. Nobody writes the history of a bridge’s 40th repainting. We write the history of its construction and, if things go badly, of its collapse, and the long decades in between, when someone was quietly scraping rust and replacing bearings, pass in a sentence, if they pass at all.

The third difficulty is accounting, and it does the most damage. In the budgets of most cities I know, including my own, building something new is classed as investment and keeping something old is classed as expense. Investment is a word that makes politicians stand up straighter. Expense is a word that makes them reach for a red pen. So upkeep is the first thing trimmed in a hard year, and the trimming does not show—that is the whole trouble—until, years later, it shows all at once. Engineers have a rule of thumb, repeated so often that nobody remembers who said it first, that every crown of maintenance deferred becomes four or five crowns of repair.^† “Deferred maintenance” is itself a small masterpiece of euphemism. It sounds like a postponed dentist’s appointment. It is a debt, compounding quietly, that appears on nobody’s balance sheet.

Elsewhere in this issue there is a good deal of evidence for what that debt looks like when it is called in: a payroll system in Corran older than most of the people who keep it running, an estuary bridge wired with 1,400 sensors that nobody has time to read, a seed vault in Norhaven that stays alive only as long as somebody keeps testing it.5 What these stories share is not a villain. Nobody decided that the bridge should go unwatched. It is simply that at every decision point, for decades, the new thing won the argument, because the new thing could be photographed and the old thing could only be kept.

What the darning knows

Philosophers have had less to say about maintenance than about almost any other universal human activity, which tells you something about who has historically had the leisure to be a philosopher. But the ideas are there if you look, mostly at the edges. An old monastic rule, written for a small community of men who had to feed, clothe, and house themselves for life, instructs that the community’s ordinary tools—its spades and kettles and kitchen knives—be treated with the same reverence as the vessels on the altar.^‡ I think it is one of the most radical sentences I know about work. It refuses the hierarchy outright. The chalice and the cooking pot are equally holy, because both are in daily use, and both will fail if no one minds them.

There is an older puzzle, too, about a ship whose planks are replaced one at a time as they rot, until none of the original timber remains, and which asks whether it is still the same ship.6 It is usually taught as a riddle about identity. Read by a maintainer, it is not a riddle at all. Of course it is the same ship. It is the same ship because somebody kept replacing the planks; its sameness belongs not to the wood but to the work. A maintained thing has a continuous identity in the way a person does—not because nothing about it changes, but because someone has been paying attention to the changes.

Paying attention: that, I have come to think, is where the real philosophy of maintenance lives. A more recent strand of moral thought begins not with rules or consequences but with an inconvenient fact, which is that every one of us was once fed, washed, and kept alive by someone else, for years, before we could do a single useful thing in return. On this view care is not a sentiment that some people happen to have. It is the ground floor of human life, the work on which every other kind of work stands, and it has been undervalued for exactly as long as it has been handed to people with too little power to insist on its value.^§ Maintenance is care extended to things. It is attention, paid in installments.

The trouble with praise

I want to complicate my own argument here, because an essay in praise of maintenance risks becoming the very thing it describes: a pleasant routine that changes nothing. And the praise of upkeep goes wrong more often than its admirers like to admit. For a start, not everything deserves to be maintained. My grandmother darned socks; she did not darn everything. She had a rule for that, too—“Three darns and it’s a duster”—and she applied it without sentiment. Maintenance can become a way of refusing to decide, of keeping a failing thing alive long past the point where keeping it serves anyone. Institutions are especially prone to this. They will tend a broken process devotedly for 20 years rather than admit that it should be retired. An ethic of maintenance has to include the judgment of ending, or it is not an ethic at all, just a habit.

Praise can also be a substitute for pay. There is a genre of sentimental writing—I have contributed to it—that honors the humble caretaker, the faithful janitor, and the selfless nurse in language so warm that nobody notices it costs nothing. The hierarchy of building over keeping does not float free of the other hierarchies; it maps onto them with depressing exactness. Darning was women’s work. Mending was paid below loading. The people who clean offices at night, tend boilers, reset breakers, and carry the ladders are disproportionately people whom the rest of the economy had not bothered to see in the first place.7 To call their work noble without changing what it pays is to compliment them on their invisibility.

The third trouble is subtler, and it changed my mind more than the other two. The opposition between builders and maintainers is partly false, because the best builders build for maintenance. Teodor has a grievance against the architect of the Carrow Building, and it is a specific one. The downpipe at the northwest corner was set behind a projecting bay, where it can be reached only from the full extension of the long ladder, at an angle he describes as “an opinion about my life.” That single decision, made at a drawing board in 1957 by a man who never had to climb to it, has cost the building a dangerous afternoon every October for nearly seven decades. Every monument is a maintenance schedule in disguise. The only question is whether the people who design it ever read the schedule they are writing.

When the keepers go

If the hierarchy cost only money, it would be bad enough. It also costs knowledge, which is harder to replace. Maintainers know things that builders never knew and no drawing records: which pipe knocks in February and why that doesn’t matter, which one knocks in March and why that does; that the flashing by the chimney stack is “holding,” in Teodor’s word, which means it will fail within five years and should be watched. Such knowledge lives in a person, accumulates over decades, and leaves when the person does. My colleague Ruth Hollis, elsewhere in this issue, sits with retired lighthouse keepers in Port Ellery who are racing to write down a vanished job before the last of them forgets it. They are doing, at the very end, what nobody thought to ask of them while the lights were still theirs to keep.

Teodor retires next spring. The company that manages the building has written to tell residents that he will not be replaced. Instead there will be an app, connected to a contractor service, through which we can report problems as they arise. As they arise. I read the letter twice to be sure I had understood it, and I had. The building is moving from a regime of maintenance to a regime of repair—from someone who prevents the flood to someone who comes after it. On paper it will be cheaper. The company’s accountants are not wrong about the first year, or the second. They are wrong about the 10th, and by the 10th year they will be working somewhere else.

How to honor upkeep

So what would it mean to take maintenance seriously—not to praise it, but to arrange things around it? I have three modest proposals, offered in the spirit of someone who has only recently learned to darn and distrusts anyone who claims to have solved the sock.

The first concerns budgets. No public building, road, or system should be opened without its upkeep being announced at the same ceremony, to the same crowd, in the same size type. If a council member wants to cut the ribbon on a new library, let her also read aloud what it will cost to keep that library dry, lit, and staffed for 30 years, and let that sum sit in a fund that cannot be raided for the next ribbon. Some of the districts my colleague Tiago Mendes visits in this issue have gone further and refused to build anything new until the old is in repair; the results, as he reports, are not simple. But even the modest version would change what a ribbon means. It would stop being the end of a story and become the first entry in a daybook.

The second concerns honors. We give prizes for new buildings, new bridges, new software, and new books. We could give them for duration. Since 2019 the Halbridge Civic Trust has awarded a small and slightly eccentric prize each year to one building that has reached its 50th birthday in good condition and continuous use, and the medal goes not to the architect but to whoever kept the place up.^‖ Imagine the habit spreading: an award for the bridge still sound at 80, the program still running at 40, the hospital ward whose infection rate has stayed low for a decade because someone cleaned it properly every single night. These are achievements. They are harder than opening day.

The third concerns names. My grandmother’s title was “mender,” a precise word and a proud one, which said what she did. Teodor’s contract, which he showed me, calls him a “facilities operative.” In the job listings I read while writing this essay, the people who keep things going were variously technicians, support staff, operatives, and resources, and almost never anything that suggested what they kept or for whom.8 Language is not everything, but it is where an institution tells you what it thinks it is paying for. Keeper is a good old word. So are warden and steward and mender. We could use them again, and mean them, and pay accordingly.

None of this will happen quickly, and some of it will not happen at all. But it begins, I think, where my grandmother began: in the good wool. You do not start with the hole—the collapsed bridge, the crashed system, the flooded basement—because by then there is nothing left to do but patch. You start with what is still sound, and with the people who are keeping it sound, and you anchor yourself there before you reach across the gap.

A few weeks ago I asked Teodor whether, when he retires, he would like his name added to the plaque in the lobby, under the architect and the builder and the councilman. I meant it half as a joke and half as a campaign; I was ready to write letters. He considered it for a while, standing by the mailboxes with his hands in his jacket pockets, and then he shook his head. “More brass,” he said. “And who’s going to polish it?”

It was only on the stairs that I remembered the line in his daybook, the one I had copied out in my own hand months before and read straight past. Thursday, between a loose step and a weather forecast: Plaque polished. He has done it every Thursday since 1995. In 11 years of walking past those three names, I had never once wondered why the brass was so bright. It was bright because someone had spent 31 years making sure that nothing happened to it.