THE LOST GRAMMAR OF MAINTENANCE — PART 2 OF 3 This is the second essay in a series about what we lost when we stopped maintaining things, and who profited from the loss. Part 1: Make Do and Mend (Loss) | Part 2: Right to Repair (Legal Response) | Part 3: The Disappearance of Domestic Science (Root Cause)
I. The Sticker
There is a small label on the back of the device you are reading this on — or if not this device, the one charging in the other room, or the one in the drawer that no longer holds a charge. You have seen the words a hundred times. You have never heard what they say.
No user-serviceable parts inside.
It is printed in a typeface designed for authority. It appears on the back panels of televisions, inside the battery compartments of radios, on the underside of routers and the chassis of laptops. It is so familiar it has become invisible — a piece of text that we absorb without processing, the way we absorb "do not remove under penalty of law" on mattress tags. Background noise. A warning we long ago stopped reading as a warning.
But read it now. Read it slowly.
No user-serviceable parts inside.
The sentence tells you two things. First: this product contains parts. Components, circuits, mechanisms — things that work, and that will eventually stop working. Second: you are not the person who will fix them. Not because you lack intelligence. Not because the repair is genuinely dangerous. But because the product has been designed so that the word "user" and the word "serviceable" do not belong in the same sentence. The parts are inside. You are outside. The sticker marks the boundary.
This was not always the boundary.
In 1947, a company called Heathkit began selling electronic kits from Benton Harbor, Michigan. The premise was simple and, to modern ears, faintly surreal: by investing the time to assemble a Heathkit, the purchaser could build a radio, a television, an oscilloscope — and if it malfunctioned, could repair it themselves.1 The New York Times noted in 1992, when the company finally closed, that "before there were nerds, before there was a Silicon Valley, there were Heathkits, which let tens of thousands of ambitious amateurs and aspiring engineers build their own radios, televisions and other electronic equipment."1 Commercial electronics of that era used generic, discrete components — vacuum tubes, tube sockets, capacitors, resistors — mostly hand-wired using point-to-point construction.1 A tube burned out, you pulled it and plugged in a new one. The product assumed your presence inside it. The product assumed you belonged.
Then the components changed. Vacuum tubes gave way to transistors in the 1950s and 1960s. Transistors gave way to integrated circuits. Integrated circuits gave way to surface-mounted microprocessors soldered to multi-layer boards, sealed under adhesive, locked behind proprietary fasteners that require tools no consumer owns. At each step, the interior of the product became less legible, less accessible, less yours. The transition happened gradually enough that no one marked the moment the default shifted — from "you are welcome inside" to "no user-serviceable parts."
The sticker arrived somewhere in that transition. No one can point to the first product that carried it, the first engineer who drafted the text, the first legal department that approved the phrasing. It emerged the way enclosures always emerge — incrementally, rationally, until the new arrangement feels as though it was always there.2
You do not need a right to do something that is ordinary. You need a right only when the ordinary has been made exceptional.
Which brings us to the law that says you can ignore the sticker.
II. The Right
On 13 June 2024, the European Parliament adopted the Right to Repair Directive — Directive (EU) 2024/1799.3 It entered into force on 30 July 2024, with Member States required to transpose it into national law by 31 July 2026.3 The Directive is the most ambitious repair legislation in history. It covers washing machines, dishwashers, refrigerators, electronic displays, vacuum cleaners, mobile phones, tablets, e-bikes, and e-scooters. Manufacturers must repair goods at a "reasonable price and within a reasonable period." They must provide spare parts and tools "at a price that does not deter repair." They cannot restrict repairs through contractual clauses, hardware, or software. Choosing repair over replacement extends the warranty by twelve months.3
In the United States, all fifty states have now introduced Right to Repair bills, with six — New York, California, Minnesota, Oregon, Colorado, and California again for powered wheelchairs — enacting comprehensive legislation.4 Oregon's law, effective January 2025, is the strongest: it is the first jurisdiction to explicitly ban parts pairing, the practice of coupling components via digital serial number so that a replacement part — even a genuine one — triggers a warning or loses functionality unless validated through the manufacturer's own systems.4
The United Kingdom, post-Brexit, has no equivalent. Its 2021 Ecodesign regulations mirror some EU provisions for specific appliances but stop well short of the Directive's scope.5
The legislation is real. We should honour what it represents.
The Restart Project — a London-based repair advocacy organisation — has logged data from over 200,000 repair attempts across thirty-one countries over twelve years.6 In the most recent twelve-month period, the broader Open Repair Alliance network documented nearly 70,000 repair attempts, with over 2,000 tonnes of CO2 equivalent saved cumulatively.6 The Repair Cafe International Foundation reports approximately 3,800 Repair Cafes operating globally, with roughly 450 in the United Kingdom — a twentyfold increase from 22 in 2016.7 iFixit has published over 100,000 repair guides and grown its community by 200,000 users in a single year.4 These are real people doing real work. The legislation gives them legal ground to stand on. It matters. It is progress.
And we need to look at it more carefully.
The word "repair" entered English in the mid-fourteenth century via Old French reparer, from the Latin reparare.8 Re- meaning "again." Parare meaning "to make ready, to prepare." To repair is to make something ready again. The word assumes a cycle: the object was ready, it became unready, it can be made ready again. The word assumes return.
But return to what?
The first essay in this series established that the capacity to maintain — the skills, the material knowledge, the intergenerational grammar of repair — was not lost through carelessness. It was enclosed. Systematically, on three fronts, over five decades, the commons of practical competence was fenced and privatised until disposal felt like nature rather than design.
Reparare assumes there is something to return to. A state of readiness that preceded the damage. A population that was once prepared, became unprepared through neglect or wear, and can be made prepared again.
The legislation makes the same assumption. Grant the right. Provide the parts. Remove the restrictions. The population will return to repair.
A law that grants permission to do what no one remembers how to do is not a solution. It is a monument to the depth of the problem.
III. The Gap
In May 2022, Sean Hollister, a senior technology editor at The Verge, decided to replace the battery in his iPhone.9
He had every right to do so. Apple's Self Service Repair programme, launched the previous month, offered genuine parts, official manuals, and rental tools — the infrastructure of repair, delivered to your door. Hollister ordered the kit. Apple shipped him two Pelican cases containing seventy-nine pounds of industrial-grade tools to replace a battery weighing 1.1 ounces. They placed a twelve-hundred-dollar hold on his credit card for the tool rental. The battery itself cost sixty-nine dollars — the same price Apple charges to replace it for you at the Apple Store. With the forty-nine-dollar tool rental fee, the total came to a hundred and eighteen dollars: more than seventy percent more expensive than walking into Apple and handing over the phone.9
After performing the repair — which he completed, because he is a senior technology editor — his iPhone displayed a message: "Unknown Part." A genuine Apple battery, purchased from Apple, installed following Apple's own manual, and the phone did not recognise it as genuine. He had to call Apple's logistics partner post-repair to validate the part remotely before the warning would clear.9
Hollister's verdict: the programme is "the perfect way to make it look like the company supports right-to-repair policies without actually encouraging them."9
You have to admire the engineering, if nothing else. The right exists. The tools exist. The parts exist. And the experience is designed to make you wish you had never tried.
If a senior technology editor — a person whose profession is understanding devices — found the process absurd, consider the experience for the person the legislation is supposed to serve: someone with no technical background who simply wants their phone to work. A separate assessment published by MacRumors confirmed that a person with no repair experience found the process "not worth the time and money invested."9
This is what I mean by the gap. Not a gap in the law. A gap between the right and the ability to exercise it. The law says you may. And the gap says: even so.
The gap has three layers.
The first layer is permission. The legal right exists. The cultural capacity does not. UK surveys suggest that one in nine young Brits cannot change a light bulb; two-thirds of the population feel they lack basic DIY competence.10 These are the same statistics that appeared in the first essay of this series, and their repetition is deliberate: they have not changed between essays because they will not change between laws. The legislation grants the right to repair. It does not — cannot — grant the thirty years of intergenerational knowledge transmission that would make the right usable.
Martin Seligman identified learned helplessness in 1967: organisms exposed to inescapable negative outcomes eventually stop attempting to escape, even when escape becomes possible.11 In 2016, Seligman and his colleague Steven Maier revised the theory. They now argue that passivity is not learned — it is the default. What must be learned is control. Agency. The belief that your actions can change your circumstances.11 This revision makes the problem worse, not better. If passivity is the default, then a population that has never experienced successful repair does not need to learn helplessness. It needs to learn agency. And you cannot learn agency from a legal document. You learn it from doing — from the experience of opening a device, diagnosing a fault, performing a repair, and discovering that you are capable of more than you were told.
Right to Repair says: you may try. But the trying is where the learning happens. And the trying requires an infrastructure that no longer exists.
The second layer is infrastructure. Approximately 450 Repair Cafes serve the United Kingdom's sixty-seven million people — one cafe for every 150,000 residents.7 The growth is real: from 22 in 2016 to 450 in 2025, a twentyfold increase.7 And the scale mismatch is also real. The Restart Project's research at a London recycling centre found that thirty-six percent of small electrical items heading for recycling were still working, and an additional ten percent needed only simple, low-cost repairs — an estimated 30,000 usable products recycled every week across the country.12 The infrastructure of repair operates at artisanal scale. The infrastructure of disposal operates at industrial scale. One does not correct the other by growing tenfold. It would need to grow a hundredfold — and staff itself with skills that the education system stopped teaching a generation ago.
When repairs are attempted, the RepairMonitor data for 2024 tells us what happens: sixty-two percent succeed.13 Of those that fail, spare parts are the primary barrier — twenty-five percent of failed repairs occur because parts are unavailable, and a further eighteen percent because parts are too expensive.13 Combined: forty-three percent of repair failures come down to parts. The legislation mandates that manufacturers provide spare parts at a reasonable price. "Reasonable" is undefined in the Directive.3 The EU left it to individual product-category regulations. The gap between the mandate and the definition is the space where the Infrastructure Gap lives.
In the United Kingdom, approximately one hundred Libraries of Things — broader than tool libraries, lending kitchen equipment, camping gear, and repair tools — have opened in recent years, led by projects like the Edinburgh Tool Library, which makes over 1,500 tools available for community use.14 These are genuine infrastructure. They are also approximately one hundred and fifty thousand short of what a functioning repair ecosystem would require.
The third layer is design. The legislation mandates access to parts and manuals. The products resist repair at the material level.
For years, according to iFixit's direct testing, nearly half of iPhone parts were coupled via digital serial numbers — a practice called parts pairing that meant a replacement component, even a genuine one, would trigger warnings or lose functionality unless validated through Apple's own systems.15 Under legal pressure from Oregon's ban on the practice, Apple has begun loosening these restrictions on its newest models. The rollback is partial. Implementation is, in iFixit's assessment, "unacceptably buggy," causing boot loops and calibration failures.15 MacBooks remain locked behind parts pairing with no repair assistant tool available.15
Samsung presents the same problem through different engineering. When iFixit ended its Samsung repair partnership in May 2024, the reason was blunt: Samsung's devices are "frustratingly glued together," forcing the sale of batteries and screens in pre-glued bundles that raised repair costs.16 Samsung required independent repair shops to provide customers' names, contact information, phone identifiers, and complaint details. It imposed a limit of seven parts per shop per quarter.16 Best Buy ended its Samsung partnership around the same time.16
In January 2025, the Federal Trade Commission, joined by the attorneys general of Illinois, Minnesota, Arizona, Michigan, and Wisconsin, sued John Deere for restricting farmer repairs through software lockouts.17 Deere's diagnostic software — Service ADVISOR — is available only to authorised dealers. Farmers who purchase the customer version pay a hundred and ninety-five dollars per year per machine and receive a tool that redacts basic error code descriptions and omits troubleshooting guides. Independent repair professionals pay $5,995 per year for up to ten local downloads.17 The U.S. Public Interest Research Group and Farm Action estimate that Deere's restrictions cost American farmers $4.2 billion annually — three billion in equipment downtime alone.17
And then there is the lobby. Companies collectively valued at over ten trillion dollars — including Apple, Tesla, Johnson & Johnson, AT&T, Medtronic, Caterpillar, John Deere, and General Electric — have been identified by PIRG as lobbying against Right to Repair legislation.18 In New York State alone, according to PIRG's analysis of state lobbying records, companies retained $366,634 in lobbyists opposing the Fair Repair Act between January and April 2024.18 In December 2025, bipartisan provisions for military Right to Repair were stripped from the National Defense Authorization Act during the conference process — despite support from Republicans, Democrats, the White House, and all three service Secretaries — after the National Defense Industrial Association argued the provisions would "hamper innovation."19 If the United States military cannot secure the right to repair its own equipment against industry lobbying, the scale of the opposition facing a consumer with a broken phone becomes a matter of structural arithmetic, not personal initiative.
The three layers are the gap. Permission without capacity. Infrastructure without scale. Rights without the material conditions to exercise them.
The franchise of permission is the enclosure's most elegant product. It converts a commons into a right, charges admission, and calls it progress.
IV. The Right We Did Not Need
For most of human history, no one needed a right to repair. The capacity was assumed. The objects were legible. The infrastructure was ordinary. A cobbler did not require legislation to resole a shoe — the shoe was built to be resoled. A village blacksmith did not need a parliamentary directive to fix a hinge — the hinge was forged from materials that spoke to the person holding the hammer. The Japanese potter practising kintsugi did not petition the government for access to spare parts. The Finnish student working wood in sloyd class did not need a European directive to use a chisel. The capacity was embedded — in the objects, in the education, in the philosophy, in the thousand small assumptions that made maintenance ordinary.
We are the first civilisation to need a law that says we may fix our own possessions.
Sit with that for a moment.
We built devices we cannot open, passed laws saying we may open them, and then discovered we have forgotten what "open" means. We are, it turns out, exactly as ingenious at creating problems as we are at failing to notice we created them.
The opposition's case is not frivolous, and we should not pretend it is. Rights have preceded capacity before. Compulsory education laws preceded mass literacy. Accessibility legislation preceded accessible architecture. In each case, the law created the conditions for the capacity to develop. The Right to Repair optimist says: give it time. The repair cafes are growing. The YouTube tutorials are multiplying. Fairphone — a company that designs phones to be repaired — returned to profitability in 2024 with revenue of fifty-four million euros. By the third quarter of 2025, its spare parts sales had grown forty-one percent year on year.20 The ecosystem is self-correcting.
The numbers are real. And the scale tells a different story. Fairphone's 100,000 units sold represent less than 0.01 percent of the 1.2 billion smartphones sold annually.20 This is not a market correction. It is a proof of concept searching for a market. And the comparison to compulsory education is structurally flawed. Compulsory education created new institutions — schools, teacher training, curricula — to deliver a new capacity. It did not merely grant a right to be literate and hope that literacy emerged. Right to Repair removes restrictions. It does not build the institutions — the training, the trade pathways, the cultural infrastructure — that would make the right exercisable. The closer analogy is a right to farm granted to people who have been cleared off the land. The right is real. The fields are fenced. And no one remembers how to plough.
The "No user-serviceable parts inside" sticker is still there. Peel it off if you like. The legislation says you can. Underneath, you will find adhesive where screws once were. Paired serial numbers where interchangeable components once sat. A diagnostic architecture designed to exclude you. A product that speaks — but not to you.
The right to repair is the first legislation in history that requires an entire population to relearn what it once knew by instinct. That it exists at all is evidence that the enclosure has been noticed. That it is insufficient is evidence of how deep the enclosure goes. Both truths. Neither comfortable. The legislation is a beginning. But a beginning is not yet the work — and the work is not the work of parliaments. It is the work of hands.
Somewhere in a kitchen right now, someone is holding a device that has stopped working. The law says they may open it. The sticker says they should not. The product says they cannot. And everything in their education, their experience, and their sense of themselves says they are not the kind of person who fixes things.
That last part is not true. It was placed there. And the distance between the right they now hold and the ability they were never given — that distance is not a policy gap. It is the measure of everything that must be rebuilt. Not legislated. Not granted. Rebuilt — with hands, with curricula, with communities, with the slow and patient work of teaching a population that they are capable of more than they were designed to believe.
The right is necessary. The ability is the real project. And we have not yet begun.
Dominic Vale writes about why we are the way we are.
What Changes
This is not a piece about what to do. It is a piece about what to see. But seeing changes things, and three things are worth seeing clearly.
The sticker is a claim, not a fact. The next time you see "No user-serviceable parts inside," read it as a sentence that says more about the manufacturer's intentions than about your capacities. The sticker is not a description of you. It is a description of how the product was designed — and who it was designed to exclude.
The right is real but it is not the whole story. Right to Repair legislation is genuine, hard-won progress. It is also the first step in a journey that requires institutions, education, infrastructure, and a cultural shift that no single law can deliver. Supporting the legislation and recognising its limits are not contradictory positions. They are the same position, held honestly.
The gap is the thing to watch. The distance between the legal right and the lived ability to repair — that gap is the clearest indicator of whether the legislation is producing change or producing the appearance of change. Watch the gap. If it closes, the legislation is working. If it persists, the next question — what would actually restore the capacity? — is the one worth asking. That question is where we are going next.