This is not an article about how the world is breaking down. We all see it, of course: the sudden collapse of dams and bridges; the slow deterioration of power grids and sewer systems; the hacked data, broken treaties, rigged elections. Infrastructures fail everywhere, all the time. Some people will even tell you that it’s okay if the Carnegie- and Roosevelt-era foundations of America crumble. Rather than fix the systems we have, we can stand by for the imminent rollout of autonomous vehicles and blockchain-based services (and let Amazon take over the public libraries). 1 Values like innovation and newness hold mass appeal — or at least they did until disruption became a winning campaign platform and a normalized governance strategy. Now breakdown is our epistemic and experiential reality.
Maintenance has taken on new resonance as a theoretical framework, an ethos, a methodology, and a political cause.
What we really need to study is how the world gets put back together. I’m not talking about the election of new officials or the release of new technologies, but rather the everyday work of maintenance, caretaking, and repair. Steven Jackson’s now-classic essay “Rethinking Repair,” written in the before-time — way back in 2014 — proposes that we “take erosion, breakdown, and decay, rather than novelty, growth, and progress, as our starting points” in considering relations between society and technology. His sober exercise in “broken world thinking” is matched with “deep wonder and appreciation for the ongoing activities by which stability … is maintained, the subtle arts of repair by which rich and robust lives are sustained against the weight of centrifugal odds.” 2
In many academic disciplines and professional practices — architecture, urban studies, labor history, development economics, and the information sciences, just to name a few — maintenance has taken on new resonance as a theoretical framework, an ethos, a methodology, and a political cause. This is an exciting area of inquiry precisely because the lines between scholarship and practice are blurred. To study maintenance is itself an act of maintenance. To fill in the gaps in this literature, to draw connections among different disciplines, is an act of repair or, simply, of taking care — connecting threads, mending holes, amplifying quiet voices.
This is necessarily a collective endeavor. In 2016, the historians of technology Andrew Russell and Lee Vinsel roused a research network called The Maintainers. Playing off Walter Isaacson’s book, The Innovators: How a Group of Hackers, Geniuses and Geeks Created the Digital Revolution, the Maintainers adopted a humorous tagline: “how a group of bureaucrats, standards engineers, and introverts made digital infrastructures that kind of work most of the time.” They held two celebrated conferences and published essays in Aeon and The New York Times, which in turn inspired dozens of journal articles, conference panels, exhibitions, dissertations, and workshops. At the first Festival of Maintenance, held recently in London, speakers addressed topics like social housing, facilities management, self-care, tool libraries, and the emotional labor of volunteer work. 3
Before maintenance can challenge innovation as the dominant paradigm, we’ll need to build a bigger public stage.
Maintenance may be a timely subject, but it isn’t new. Ancient humans had to fix their aqueducts and mud-brick dwellings. Karl Marx was concerned with “the maintenance and reproduction of the working class” as a condition of capitalism. 4 And Russell and Vinsel identify maintenance as “a near-constant topic in the prescriptive literature that arose between the 1870s and 1920s around new technology,” from telephones to roads. 5 As we pick up the theme, we have to recognize that maintenance and repair have always been shaped by the political, social, cultural, and ecological contexts of technology (and, more broadly, techne or craft). More than that: we have to know the history of what we’re up against. Russell and Vinsel trace a genealogy of fetishized innovation, from 19th-century industrialism through the age of invention, postwar consumer tech, Cold War R&D labs, and the 1980 Bahy-Dole Act — which enabled federally-funded researchers to patent their inventions — and on to today’s Silicon Valley. 6
Before maintenance can challenge innovation as the dominant paradigm, we’ll need to build a bigger public stage. The current discourse is tilted toward economists, engineers, and policymakers — and they’re a pretty demographically homogeneous group. 7 Given the degree of brokenness of the broken world (and the expense of fixing it), we need all maintainers to apply their diverse disciplinary methods and practical skills to the collective project of repair. Jackson proposes that repair-thinking be considered a distinct epistemology. Fixers, he says, “know and see different things — indeed, different worlds — than the better-known figures of ‘designer’ or ‘user.’” Breakdown has “world-disclosing properties.” 8 Similarly, Stephen Graham and Nigel Thrift identify breakdown and failure as “the means by which societies learn to reproduce,” because the repair of broken systems always involves elements of “adaptation and improvisation.” 9
So what can we learn about how these concepts have been taken up in various fields? How can science and technology scholars build more bridges with architects, librarians, and other professionals engaged with stewardship? I’d say that if we want to better understand and apply maintenance as a corrective framework, we need to acknowledge traditions of women’s work, domestic and reproductive labor, and all acts of preservation and conservation, formal and informal. At the same time, we have to avoid romanticizing maintenance and repair. We can learn from feminist critiques of the politics of care (particularly the reliance on poorly paid immigrants and people of color) and look to maintenance practices outside the Western world.
Here I aim to show how these different disciplinary approaches converge across four scales of maintenance. In “Rust,” we’ll look at the repair of large urban infrastructures, from transportation systems to social networks. In “Dust,” we consider architectural maintenance alongside housework and other forms of caretaking in the domestic and interior realms. In “Cracks,” we study the repair of objects, from television sets to subway signs to cell phones. Finally, in “Corruption,” we turn to the curators who clean and maintain data — a resource that fuels the operation of our digital objects, our networked architectures, and our intelligent cities.
People and data work across these scales of maintenance, and they do so within particular cultures and geographies, and through different subjectivities. Throughout the essay, I’ll highlight work by artists who can help us see these other perspectives and imagine how maintenance makes itself apparent within the world. 10
Rust: Urban Repair
Every four years the American Society of Civil Engineers releases an “infrastructure report card,” which reliably generates a wave of headlines about the poor condition of our public works. In 2017 the United States earned a disappointing-but-not-surprising D+ overall. 11 Water systems scored a D (six billion gallons of treated water are lost every day); dams, a D (seventeen percent are highly hazardous); and roads, a D (one out of every five miles is in poor condition). Transit earned a D- (in part, for the $90 billion backlog of maintenance projects). Why such neglect? At a forum hosted by the Brookings Institution (naturally!), economist Larry Summers gave the usual explanation: “All of the incentives for all the actors are against maintenance. Nobody ever named a maintenance project, nobody ever got recognized for a maintenance project, nobody ever much got blamed for deferring maintenance during the time while they were in office.” His interlocutor, Edward Glaeser (see, it’s always the economists!), agreed: “you get a lot of press for a new project. … You don’t get a lot of press for maintaining the HVAC system in the school, even though that’s more socially valuable.” 12
Yet this macroeconomic view obscures the phenomenal reality that the world is being fixed all around us, every day. Window washers work high above the street and cable layers below it. Bridge painters combat salt air and exhaust fumes. “Modern urban dwellers are surrounded by the hum of continuous repair and maintenance,” Thrift observes. We hear the chatter of pneumatic drills, the drone of street sweepers, and, in the city’s peripheral zones, the clang and hydraulic hiss of auto repair and waste management. 13 Even the cacophony of a construction site — a new building going up on a vacant lot — can be a sign of repair. Planner Douglas Kelbaugh proposes that we think of infill construction as a mending of the urban fabric. 14
Meanwhile, caregivers, therapists, clergy, social workers, and other outreach agents attend to the city’s social infrastructures. Sociologists Tom Hall and Robin James Smith regard these “carers” as instruments of “urban kindness,” but we should be wary of conflating care and altruism. Geographer Jessica Barnes warns against the romanticism inherent in the revival of maintenance studies. Scholars have elevated certain types of underappreciated work and have framed repair contra consumption and waste, but in many settings, especially outside the post-industrialized West, the motivations behind urban and ecological maintenance are more complex. 15
Where infrastructures are absent or unreliable, the gaps are filled by illegal water taps, grafted cables, pirate radio stations, backyard boreholes … Many regions have their own distinctive ‘repair ecologies.’
Around the world, many formal infrastructures are products of colonialism, and imperial legacies persist through global financing. “Rehabilitation” efforts funded by the World Bank and IMF reflect a “tendency for neglected maintenance expenditures to be capitalized through ‘new build’ projects.” 16 Maintenance is thus entangled with plans to open or protect access to markets or resources. Some development projects are stalled by local resistance or administrative problems; others leave marginalized and disenfranchised people off the grid. And where infrastructures are absent or unreliable, the gaps are filled by illegal water taps, grafted cables, pirate radio stations, backyard boreholes, shadow networks, and so forth. Many regions have their own distinctive “repair ecologies,” like the underground market in Cuba for el paquete semanal, a weekly supply of new digital content circulated offline, via hard drive, in order to circumvent the nation’s insecure internet. 17 This, too, is a kind of maintenance. Graham and Thrift argue that in the Global South, “the very technosocial architectures of urban life are heavily dominated by, and constituted through, a giant system of repair and improvisation.” 18 Developing regions also become offshore “back lots” for wealthier nations’ abject maintenance work, like breaking up rusty ships and processing e-waste. As Jackson puts it, some places are “more on the receiving end of globalization than others.” 19
Outsiders sometimes make the mistake of focusing on the rusty bridges and broken pipes — the “defective objects” themselves — whereas local fixers are more concerned with “the social and political relationships in which [those objects are] embedded.” Barnes reports that Egyptian farmers in the Nile Valley maintain irrigation ditches not just to keep the water flowing, but also to “sustain communal ties with other farmers.” 20 Similar protocols prevail in Nikhil Anand’s Hydraulic City, in which the anthropologist shows how the maintenance of water infrastructures binds residents, plumbers, engineers, and politicians in an (uneven) system of “hydraulic citizenship.” 21 We should always ask: what, exactly, is being maintained? “Is it the thing itself,” Graham and Thrift ask, “or the negotiated order that surrounds it, or some ‘larger’ entity?” 22 Often the answer is all of the above. Maintenance traverses scales.
Outsiders mistakenly focus on the rusty bridges and broken pipes — the ‘defective objects’ — whereas local fixers are more concerned with social and political relationships.
And if we reverse the lens, we see how the multi-scalar nature of broken systems impedes repair. Consider the New York City subway. Mayor Bill DeBlasio and Governor Andrew Cuomo have famously fought for years about whether the city or the state is responsible for fixing the subterranean mess. Nobody wants to pay, so one of the world’s great transportation systems falls into disrepair. Historian Scott Gabriel Knowles proposes that we think of the “deferred maintenance” of public infrastructures as slow-motion disasters, which sustain the oppression of marginalized and underserved populations. 23 Summers, meanwhile, emphasizes the “debt burden on the next generation,” since the cost of fixing the world compounds over time. A chorus of economists says that infrastructure maintenance has positive effects on economic growth and productivity. 24 And yet here we are, waiting for an overcrowded 7 train.
Dust: Spaces of Labor and Care
If even the economists and engineers can’t rally public funding for urban maintenance, what chance do the rest of us have? ASCE’s “report card” gives no grades for public housing or mental health clinics, and it doesn’t recognize the infrastructures built and maintained by librarians and domestic workers and data managers. Fortunately, maintenance researchers take a wider view of the repairspace. In some cases, there are other evaluative bodies that can assess infrastructural conditions.
The civil engineers’ ‘report card’ doesn’t recognize the infrastructures built by librarians and domestic workers and data managers.
Consider the New York City Housing Authority, which oversees more than 2,400 buildings in 325 developments, sheltering about five percent of the population. On average those buildings are around 60 years old, and their systems periodically break down, leaving residents without heat or hot water. NYCHA’s latest Public Needs Assessment describes leaky roofs, windows, and pipes, which have caused mold and other extensive damage to walls and ceilings. A proposal to renovate kitchens and bathrooms could cost $31.8 billion over the next five years. 25 And that was before federal sanctions. After housing officials were caught submitting false reports about lead paint, NYCHA has been placed under federal monitoring and must spend at least $1 billion more on repairs. These kind of negative feedback loops are typical of deferred maintenance: federal budget cuts lead to local neglect which leads to federal sanctions. 26
We can zoom in from that regulatory apparatus to see all the other labor involved in maintaining a building. At the architectural scale, maintenance involves a wide spectrum of professional expertise: “preservation, material science, development, policy, insurance law, and building codes,” and more, as Hilary Sample explains. Different styles of buildings — from pre-modern dwellings to modernist airports — call for different modes of upkeep, preservation, or conservation. In recent decades, architects have used post-occupancy evaluations to assess how their buildings are performing and make adjustments. They can also anticipate maintenance needs and design for them, by choosing durable materials and conducting lifecycle cost analyses and environmental impact studies. 27
Building maintenance is sometimes legible from the street. We see work permits in foyer windows and repair vans parked out front. A recent exhibition on “Scaffolding” at the Center for Architecture showed how this seemingly utilitarian structure — none too popular in New York, given its tendency to compress pedestrian traffic and create street-level dungeons — can serve as a social infrastructure, as a tool for improvisatory construction, and even as a platform for performance. 28 Inside the building, the cast of maintainers widens. For Urban Omnibus, Juliette Spertus and Valeria Mogilevich interviewed building superintendents and showed how they serve as “educators, enforcers, and innovators in maintenance.” Sociologist Christopher Henke has studied how physical plant mechanics negotiate amongst themselves, and, in a separate paper, he recommends that repair be considered integral to sustainable building practices. And of course we can’t forget the work tenants and owners do to maintain their own homes. For over a century, engineers, management consultants, and efficiency experts — many of them women — have been studying the mechanics of housework. They launched the “domestic science” movement and the field of home economics. 29
When women entered the workforce in great numbers in the 1960s, scholars and activists (drawing inspiration from early feminists) began thinking differently about the maintenance those women had long been doing at home without compensation. As Silvia Federici puts it, “after two world wars … the lures of domesticity and the prospect of sacrificing our lives to produce more workers and soldiers” — of reproducing the labor force necessary to maintain a productive economy — “had no hold on our imagination.” 30 Mierle Laderman Ukeles pioneered the genre of “Maintenance Art,” performing the mundanity of this exhausting work, while granting it (and herself, a wife and mother) visibility and value within the civic realm. 31 Yet much early thinking about “reproductive labor” among Marxist feminists ignored the fact that women of color, poor women, and immigrants “had been engaged in paid market work in large numbers for many decades.” As they cooked, cleaned, and nannied for affluent families, they were often less available to care for their own. 32
Nowadays, social scientists are more likely to focus on the socioeconomic dynamics of reproductive labor, particularly on shifting gender (im)balances, the rights of domestic workers, and the “global care chains” transferring maintenance labor from the Global South to the north. 33 Critics and activists have validated a greater range of (re)productive activities, to include all the mental, manual, and emotional work necessary to “maintain existing life and to reproduce the next generation.” 34 Maintaining life — that’s a big job. In a foundational article from 1992, Evelyn Nakano Glenn listed some of those responsibilities: “purchasing household goods, preparing and serving food, laundering and repairing clothing, maintaining furnishings and appliances, socializing children, providing care and emotional support for adults, and maintaining kin and community ties.” Today, we might add tech support and digital filtering. 35
Contemporary theorists and activists are also talking a lot about ‘care,’ which has more to do with the ethos and affect of maintenance than with its (re)productive capacities.
Contemporary theorists and activists are also talking a lot about “care,” which has more to do with the ethos and affect of maintenance than with its (re)productive capacities. Joan Tronto and Berenice Fisher define care as “everything that we do to maintain, continue, and repair ‘our world’ so that we can live in it as well as possible. That world includes our bodies, ourselves, and our environment, all of which we seek to interweave in a complex, life sustaining web.” Maria Puig de la Bellacasa argues that caring involves an “ethico-political commitment” to the neglected and oppressed and a concern with the affective dimensions of our material world. We care for things not because they produce value, but because they already have value. 36
It’s hard to argue with that. Who doesn’t care for care? Yet care, like maintenance, is easily romanticized. Historian Michelle Murphy argues that the “politics of care” promoted by 1970s feminists were “conditioned by white privilege [and] capitalism.” 37 In correcting for these oversights, theorists and activists have turned their attention to the rehearsed, compulsory care performed by female workers — stewardesses, receptionists, nurses, waitresses, customer service reps — in the expanding service industries. Knowledge workers sometimes face similar concerns. Fobazi Ettarh argues that librarians are conditioned to accept low pay, low status, and expanding workloads because librarianship is regarded as a calling — a care-centric vocation that is “inherently good and sacred, and therefore beyond critique.” Yet historians have shown that libraries, for all their goodness, are built upon protocols and policies rooted in colonialism and privilege. And as other social services in cities are starved for funding, librarians are often left to pick up the slack. 38
Now consider healthcare, which is not only unevenly accessible, but is also meted out through policies that exacerbate inequality and through treatments that benefit insurance and pharmaceutical companies. Historically, medical research has been downright exploitative at times. Women are also marginalized in contemporary care practices, where their self-reported ailments may be dismissed as psychosomatic. 39 Black women are particularly disenfranchised. The New York Times recently reported that extreme racial disparities in prenatal care are getting worse: “Black infants in America are now more than twice as likely to die as white infants.” 40 This is happening today, in 2018.
Murphy doesn’t dismiss the importance of care, but “in a moment when so many scholars” — and, I’d add, policymakers, activists, artists, and designers — “are turning to affect and care to re-imagine politics,” she wants us to reckon with its troubling histories and administrative structures. She wants us to consider recuperative strategies that don’t normalize care as inherently virtuous and good-feeling. Aryn Martin, Natasha Myers, and Ana Viseu propose that a critical practice of care would “pay attention to the privileged position of the caring subject, wary of who has the power to care, and who or what tends to get designated the proper or improper objects of care.” 41 We could extend these questions to every scale of maintenance work — from transit networks and school systems to homes and objects.
Going further, we could imagine physical infrastructures that support ecologies of care — cities and buildings that provide the appropriate physical settings and resources for street sweepers and sanitation workers, teachers and social workers, therapists and outreach agents. How can we position “care” as an integral value within the city’s architectures and infrastructures of criminal justice, designing systems and spaces for restoration rather than retribution? Urban Omnibus has been exploring these questions in a remarkable series, “The Location of Justice.” If we apply “care” as a framework of analysis and imagination for the practitioners who design our material world, the policymakers who regulate it, and the citizens who participate in its democratic platforms, we might succeed in building more equitable and responsible systems. 42 We should also remember that the preservation of our world — the human one — is sometimes at odds with caring for the ecological context. Perhaps not every road should be repaired. Geographer Caitlin DeSilvey encourages us to embrace entropy within the built world, to ask ourselves for whom we engage in preservation, and to consider cultivating an acceptance of “curated decay” where appropriate. 43
How can we position ‘care’ as an integral value within the city’s architectures and infrastructures of criminal justice, designing systems and spaces for restoration rather than retribution?
At home, infrastructure is personal. In the 1990s, newspaper publisher Jean François Lemoine, bound to a wheelchair, decided to build a material world for himself that didn’t emphasize his disability. “I want a complex house,” he reportedly told the architect Rem Koolhaas, “because the house will define my world.” Koolhaas designed a three-level structure organized around a writing desk on a 10 x 11.5-foot elevator platform, which blends into the house’s floors as it moves up and down, allowing Lemoine to bring the rest of the house to him. As Koolhaas explained to architecture critic Nicolai Ouroussoff, the building “reassert[s] the position of the French male within the family.” 44 Ouroussoff observes that the “children’s rooms are pointedly difficult to reach” from this moving platform, which suggests the distance between this French male and his childcare duties.
In the opening sequence of the 2008 documentary Koolhaas Houselife, directed by Ila Bêka and Louise Lemoine, we focus not on the father but his housekeeper, Guadalupe Acedo, poised on the platform amidst a tableau of buckets, mops, and vacuum. She ascends to the tune of Johann Strauss the Younger’s Acceleration Waltz, an ironic choice because the house’s mechanics and structure seem insistent on slowing her down, causing friction. 45 Rather than cutting swiftly from scene to scene, the filmmakers laboriously trace the circuitous routes of circulation that Acedo and the other maintenance workers follow. As they climb ladders, wind through labyrinthine hallways, and descend steep outdoor trails, we feel their burden. At one point, the long pole of Acedo’s swimming pool net bonks the camera. The impact — and the absurdity of it — rings in our own heads.
Acedo has learned to improvise, to negotiate the space’s idiosyncratic orientations and accept its inefficiencies. In one sequence, she lugs a mop, bucket, and vacuum up a spiral staircase, then resourcefully deploys these long linear tools in a space of tight curves. We marvel at her patience in accommodating a structure that clearly aims to make life difficult for its caregivers. Yet when he saw the film, Koolhaas was disappointed that the architecture hadn’t inspired more innovative maintenance:
I am kind of surprised by the fact that someone who has such a daily involvement [with the building] is so insistent on a kind of generic technique of cleaning something so exceptional. I can easily imagine if I were a cleaner — maybe this is something we should have thought of — [I would have devised] some sort of protocol of what is convenient to be done by hand and what is convenient to be done by machine. I am completely surprised that something that is as harsh and exceptional as the spiral staircase is treated with a Hoover. It is completely insane. 46
This probably isn’t what Stewart Brand had in mind when he wondered how a building might “teach good maintenance habits.” 47
Acedo, a link in the global care chain, sleeps overnight in the staff quarters and thus dedicates more time to this home than to her own. And she cares not just for the house but also for “the negotiated order that surrounds it” (to reprise Graham and Thrift), including the family that inhabits it and the tourism industry that feeds upon it. We don’t see much of the family in the film; we instead see the traces of their presence: messy stacks of books, dirty dishes. Lemoine died not long after the building was finished, making its central mechanical conceit obsolete. Koolhaas told The New Yorker’s Daniel Zalewski that “the elevator had become a monument to his absence.” 48 And the house itself is a monument, designated a landmark just three years after its completion. Yet if Acedo’s cleaning is an act of preservation, this obstinate house seems to resist care. Its leaks and deteriorating concrete core, both discussed in the film, suggest a hastening toward the end. Even monuments turn to dust.
Cracks: Fixing Objects
In Houselife, the rags, mops, and pool skimmers serve to extend Acedo’s limbs and mediate her interactions with the building. Yet, as Ruth Schwartz Cowan observes, household labor is increasingly “performed with tools that can be neither manufactured nor understood by the workers who use them.” 49 Maintenance machines like vacuum cleaners and washing machines often demand specialized maintenance themselves. The same is true at the office, where workers rely on the expertise of copy-machine technicians and IT staff. We can perform various actions on broken objects — “mending, repairing, fixing, restoring, preserving, cleaning, recycling, up-keeping, and so on” — yet these objects, much like architectures, vary in their “openness and capacity to be taken care of.” 50 Some devices are designed for obsolescence and thwart any attempts at repair; others are modular, open to upgrades; while still others, like the mop and bucket, are timeless. Yet the lifespan of an object also depends on context. While in the West a cracked screen can mean death; elsewhere, it opens up possibilities for reuse.
Today, few people would imagine fixing their own washing machines or refrigerators, but that wasn’t always the case.
Today, few people would imagine fixing their own washing machines or refrigerators, but that wasn’t always the case. Land-grant colleges in the U.S. have long offered practical training in farm- and housework, including programs that emphasize technical mastery and maintenance. In 1929, Iowa State College (now University) launched an undergraduate major in household equipment, where women learned about the chemistry and physics of laundry, food science, child development, home management, household economics, electronics, and a variety of other domestic sciences. The school offered a parallel graduate program through the 1940s. As historian Amy Sue Bix explains, students were required to “take apart and reassemble machinery in order to appreciate details of its construction, operation, and repair.” These exercises were meant to “educate self-reliant homemakers who would confidently accept active responsibility for their kitchen equipment rather than cultivate attitudes of feminine helplessness.” 51 Some graduates even parlayed that expertise into careers at appliance manufacturers or utility companies. The presumption was that, in maintaining these machines, women could also maintain a household and nurture a family — and, by extension, a society.
Women had a different relationship with the equipment in their mid-century living rooms. Technical knowledge about gadgets — about their mechanical and electrical operation, as well as their installation and repair — was “something that came from outside the home.” As media scholar Lisa Parks explains, the TV repairman made house calls that both reinforced and upended the gendered roles around maintenance. By entering the home, the repairman “challenged the authority of the family patriarch,” created “opportunities for unseen interactions” between married women and men who weren’t their husbands, and allowed female consumers to “engage with the more complex aspects of television on their own terms.” 62 In the future, as the domestic sphere envelops more smart technologies — flat screens, black-boxed sensors, and voice assistants networked to a home operating system and the cloud beyond — there’ll be fewer opportunities for such engagement. Breakdowns might require complicated systemwide diagnostics involving both on-site fixers and technicians in remote call centers. (The phone, we should note, has long been an instrument of relational maintenance, where operators are expected to assume a consistent “positive affect” of care, regardless of how they’re treated by customers.) 53
What happens to our broken laptops and Alexas? We can still find old radios and film projectors at flea markets and thrift shops. But rarely does one find a used iPhone there. Some technological objects are refurbished and resold, and some are disassembled for scrap. Scholars in various fields have turned their attention to “discard studies,” including flows of electronic waste. Sociologist Jenna Burrell, writing in 2012, describes the internet cafes of Accra, Ghana, where teenagers chat and play games on old computers cast out of schools and offices in North America and Europe. While Western media has commonly portrayed Ghana as a node in the “shadowy industry” of e-waste disposal, Burrell sees the country and its diasporic communities as networks of entrepreneurial refurbishment and secondhand trade, where workers have opportunities to develop technical skills. Ghana’s contribution to the tech world, Burrell argues, is not in designing new machines, but instead in “finding opportunities for agency and innovation” in their provisioning, repair, and distribution. 54
The open-air shop is a space of public pedagogy, an ‘operating theater’ where the repairman demonstrates technical skills.
When those machines have lived out their second domestic lives in Ghanaian cafés or homes, they move out into the city, where scrap collectors, processors, and traders — most of them first-generation immigrants — spirit the machines away, accelerating their decomposition into copper, aluminum, iron, and circuit boards. These processes of transformation typically occur in the city’s marginal zones and, as we see in many National Geographic-style photo essays, present serious health risks. The constituent parts are then redistributed — some domestically, some to Nigeria or China — and reassembled into new objects. This “ecosystem of distribution, repair, and disposal” is, Burrell argues, a “fact of life in everyday places marked by scarcity.” 55
Parks describes public performances of repair on the city streets in Macha, Zambia. In open-air shops, “repair not only extends the use value of objects but becomes a mechanism of social interaction.” People gather around, watch, and chat. The shop is a space of public pedagogy, an “operating theater” where the repairman opens gadgets, demonstrates technical skills, and perhaps encourages observers to mend rather than discard their own broken things. The collectivity of repair is a kind of social infrastructure; in fixing a phone, one also creates the context for a temporary public. 56
Lest we believe that only the developing world yields such wisdom, we can follow Steven Bond, Caitlin DeSilvey, and James Ryan through southwestern England, on a tour of repair shops for everything from typewriters and tools to books and bicycles. 57 Or we can visit the Parisian metro with Jérôme Denis and David Pontille, who argue that the repair of signage not only serves to maintain social order by directing commuters to the right platforms and exits, but also “teach[es] us something about material ordering processes, about the ordinary life of … objects, and about the role of the people in charge of them.” Damage to a seemingly fixed object like a subway sign reminds us that the world is fragile and that we all bear some responsibility for attending to it. We participate in systems of distributed maintenance. 57
Shanzhai, jugaad, and gambiarra become the focus of Western design studios and workshops — as if we can ‘learn from’ them as we did from Las Vegas.
That responsibility — or right — is at the heart of ongoing debates over consumers’ “right to repair” their own electronic devices. iFixit is a wiki “that teaches people how to fix almost anything”; users can upload manuals with instructions for repairing cars, game consoles, and hundreds of things in between. Parks suggests that, while such sites “extend the social circulation of technical knowledge, they sometimes link repair to a heroic masculinity … preoccupied with restoring order or turning the insides of machines into spectacles.” 58 The Restart Project has a slightly different ethos. This U.K.-based organization hosts parties and a podcast and collaborates with schools to teach people how to repair their devices. In a similar vein, some public libraries in the U.S. have opened “U-Fix-It” clinics and “repair cafes,” which are a natural extension of the recent proliferation of library makerspaces. In Brazil, artists practice gambiarra, the making of makeshifts, the kludging of found electronics. Media scholar Jennifer Gabrys proposes that we mix these ethical and aesthetic strains of repair and embrace “salvage,” which in her formulation involves transforming the “exhausted and wasted into renewed resources,” as well as exhuming the stories behind objects’ production and use, and imagining other “possibilities for adaptation.” 59
One result of all this attention to e-waste and supply chains is that examples of non-Western improvisation are now often adapted, appropriated, or fetishized in the West. Designers and artists express fascination with “informal” and “entrepreneurial” design practices, or with favela “bricolage” and marginalized “maker” cultures. This can lead to the idealization of repair, the romanticization of strategies of survival, and even the recasting of austerity as a form of intellectual or moral prosperity. Shanzhai, jugaad, and gambiarra become the focus of Western design studios and workshops — as if we can “learn from” them as we did from Las Vegas. Sometimes, it seems, repair entails complicity with capitalism and colonialism. 60 The “protean capabilities” of the “Third-World bricoleur,” Ginger Nolan argues, can be used to “enable and justify the perpetuation of economic instability.” 61 So we need to be aware of how these stories of maintenance traverse geographies and scales, and take care in mining them for ethnographic insight, morality tales, aesthetic inspiration, and design solutions.
Corruption: Cleaning Code and Data
Many manufacturers aim to keep their wares out of repair and remix economies, and they carefully control the evolutionary lifecycle of their products. That’s especially true for smartphones and laptops, which “live and die by the update,” as Wendy Chun says. 62 But code increasingly plays a critical role in the functioning of responsive architectures and networked cities. Maintaining buildings and public infrastructures now involves attending to their underlying software. In so-called smart cities, that’s going to get expensive. Historian Nathan Ensmenger reports that “from the early 1960s to the present, software maintenance costs have represented between 50 and 70 percent of all total expenditures on software development.” 63 For all the talk about innovation and disruption in the tech industry, most coders are actually busy fixing stuff.
For all the talk about innovation and disruption in the tech industry, most coders are actually busy fixing stuff.
Computer historians have shown that the maintenance of hardware and social infrastructures are often intertwined. Bradley Fidler and Andrew Russell demonstrate that Arpanet lived beyond its original “demo” function thanks to the maintenance work performed by “sponsors, liaisons and bureaucrats who labored to sustain and link … organizations and technologies.” Elizabeth Losh highlights the work of Mina Rees, a computing infrastructure planner at the Office of Naval Research who conducted “home evaluations” at universities and research centers in the 1940s and ’50s to determine which had the appropriate mix of “funding, personnel, equipment, supply chains, policies, and social dynamics” to become major computing hubs. “Although [Rees] often cast herself in a supporting role,” Losh writes, “her awareness of administrative nuances was clearly essential to care and repair of infrastructure.” 64
Paul Edwards has more fully incorporated software in his pioneering analyses of computational systems, including climate modeling and military command-and-control projects. His work inspired David Ribes and Thomas Finhold to examine big cyber-infrastructure projects — and the unfortunate lack of attention and prestige attached to their maintenance. Ribes and Finhold argue that designers of those systems should be planning for the “long now.” They need to ask, “How can the perseverance of the infrastructure project be ensured, in the face of changing technologies, emerging standards, and uncertain institutional trajectories? How can the continued commitment of participants be secured?” In other words, how can we maintain technical systems and communities of practice and, by extension, the larger research enterprises they serve? 65
Just like buildings and cities, most software applications and platforms and portals would break down quickly were it not for the maintenance workers who keep them in good working order. There are systems administrators, whom Gabriella Coleman profiles as “part plumber, part groundskeeper, and part ninja, fixing problems, maintaining the system, and fending off attacks.” And there are content moderators who screen sites for illegal or inappropriate content. Sarah T. Roberts reports that this work “is almost always done in secret for low wages by relatively low-status workers, who must review, day in and day out, digital content that may be pornographic, violent, disturbing, or disgusting.” These workers have a high rate of burn-out, so most companies hire contract labor, and a large part of the workforce lives in the Philippines, where tech companies can find Internet-savvy, English-speaking contractors who are familiar with Western culture. They, like Acedo, sacrifice their own well-being in order to make the internet a “clean,” wholesome, “safe for work” environment. 66
The first Festival of Maintenance celebrated ‘repair, custodianship, stewardship, tending and caring’ in both the physical and digital spheres.
Then there are open-source communities. Lots of software relies on free, public code maintained by volunteer developers. Being “open,” Christopher Kelty writes, carries responsibilities: it “means not only sharing the ‘source code’ … but devising ways to ensure the perpetual openness of that content, that is, to create a recursive public devoted to the maintenance and modifiability of the medium or infrastructure by which it communicates.” Yet, predictably, there’s little financial support for this work, and the volunteer publics are stretched thin. When Nadia Eghbal surveyed tech workers and volunteers who maintain open-source projects she found widespread “stress and exhaustion.” 67
In the last couple years, a group of self-described “sustainers” — people “concerned with the fragile state and future of highly-used and impactful open-source projects” — have met twice IRL to sustain one another in their often-invisible labor, and to develop recommendations to make their community stronger. 68 Introduced above, the first Festival of Maintenance celebrated “repair, custodianship, stewardship, tending and caring” in both the physical and digital spheres, recognizing that the maintenance of open-source software, online communities, co-ops, and datasets is not unlike the maintenance of natural environments, infrastructures, industries, cultural heritages, and material resources. 69
Further, we should acknowledge the human “data processors” and “curators” who “clean up” structural problems and reformat data sets used in research or marketing. As Jean-Christophe Plantin learned through his ethnographic work at a social science data archive, the data set “must look pristine at the end of its processing.” The internal maintenance work isn’t supposed to be visible to end users, who tend to like the idea that they’re working with “raw” data. Yet “data never come as raw,” Plantin observes. “Multiple interventions are always needed before data can be reused.” 70
Digital archivist Hillel Arnold laments the “invisibility problem” in his profession. Archivists are often portrayed as “save-iors” who “erase [their own] labor” so that researchers who use the archive can “discover” “lost” treasures. Meanwhile, “vocational awe” compels archivists to do more with less; for example, to rely on temporary labor and unpaid internships rather than paid staff. Inspired by the rise of The Maintainers, Arnold calls upon archivists to demand the labor conditions and resources necessary to properly care for their collections, themselves, their colleagues, and their patrons. 71 Some funders are even beginning to come around. NEH Digital Humanities Advancement Grants now cover “revitalizing and/or recovering existing digital projects,” and the Sloan and Ford foundations offer a new funding program for sustaining digital infrastructure.
Libraries’ digital resources require maintenance, too. In an essay on “Broken-World Vocabularies,” Daniel Lovins and Dianne Hillmann describe the challenges of retrofitting and reconciling metadata as technologies advance. That includes maintaining the bibliographic vocabularies that librarians use to describe their collections: “Anyone who has participated in library standards committees knows how much effort is required to keep MARC, RDA, LCSH, etc. in stable condition. This is partly from internal inconsistencies born of compromise, and partly because the world around descriptive vocabularies is itself constantly breaking” and changing. Science advances, new disciplines emerge, understandings of human identity evolve. Librarians can’t “fix” the external world — they can’t hold it steady, nor should they want to — but they maintain the informational systems that help patrons better understand it. 72
Data maintenance is particularly consequential in medicine, and thus caring for medical sites, objects, communities, and data has been recognized as an important part of caring for patients. As Ribes explains, clinical trials involve maintenance activities like calibrating instruments, cleaning data, preserving specimens and data, retaining participants, and stewarding sites and communities (or “biomes”) of data collection. Study participants, and especially patients with chronic illnesses, sometimes adopt what Laura Forlano calls “broken body thinking” — “actively participating in, maintaining, repairing, and caring for multiple medical technologies”: pumps, sensors, monitors, needles, and vials. 73 And what happens when those things aren’t cared for? Brittany Fiore-Gartland reports that NGOs often fund the launch of ambitious digital health initiatives in developing countries, then abandon programs that “never make it to scale,” or allow them to “burn out at the end of the … funding cycle.” Yet participants in pilot studies often become dependent on the tools and services provided. When resources are withdrawn, caregivers improvise to fill in the gaps. Fiore-Gartland finds that the “organizational structures” that emerge around abandoned technologies “must be repaired and re-articulated.” 74
Even if we build an army of repair robots and AIs, their hardware and software will still require upkeep. Labor is essential to maintenance.
This is reactive care, responsive care. Benjamin Sims argues for a more proactive model of “anticipatory repair,” informed by “modeling and simulation, user studies, trials, testing, and other methods that aim to instigate breakdowns in a controlled setting before a technology is deployed for general use, thus affording an opportunity to fix problems before they affect end users.” Engineers now use artificial intelligence to predict and prevent infrastructural snafus — to scan for idiosyncrasies within high-performance computing systems, power plants, or financial markets, and to act preemptively to thwart disaster. Carefully managed data are essential in domains like preventive healthcare and building maintenance. As Jes Ellacott proposes on The Maintainers’ blog, emphasizing the innovative potential of repair — highlighting the fact that cutting-edge technologies can improve maintenance — could “change how we see, value, and reward” our maintainers. 75
Perhaps. Yet even if we build an army of repair robots (a longtime sci-fi dream) and maintenance AIs, their hardware and software will still require upkeep. They’ll still depend upon well-maintained, interoperable technical infrastructures. They’ll still require cleaning staff — “industrial hygienists” — to maintain pristine conditions for their manufacture. 76 We’ll need curators to clean the data and supervisors for the robot cleaning crew. Labor is essential to maintenance.
As Jay Owens reminds us,
There will be dust. There is always dust. By that I mean there is always time, and materiality, and decay. Decomposition and damage are inescapable. There is always the body, with its smears and secretions and messy flaking bits off. There is always waste and it always has to be dealt with, and shipping it out of sight overseas to the developing world does not change the fact this work has to be done (and it is dirty, dangerous work that demands its pound of flesh). 77
That’s true whether we’re talking about ditch-digging or dam-building or data-diving. Maintenance at any particular site, or on any particular body or object, requires the maintenance of an entire ecology: attending to supply chains, instruments, protocols, social infrastructures, and environmental conditions.
Across the many scales and dimensions of this problem, we are never far from three enduring truths: (1) Maintainers require care; (2) caregiving requires maintenance; and (3) the distinctions between these practices are shaped by race, gender, class, and other political, economic, and cultural forces. Who gets to organize the maintenance of infrastructure, and who then executes the work? Who gets cared for at home, and who does that tending and mending? Agreements about what things deserve repair — and what “good repair” entails — are always contingent and contextual. If we wish to better support the critical work performed by the world’s maintainers, we must recognize that maintenance encompasses a world of standards, tools, practices, and wisdom. Sometimes it deploys machine learning; other times, a mop.