Making Light Work: Infrastructures and Their Many Publics

Introduction

This article explores the infrastructural making of publics through a multi-sited case study of street lighting in the United Kingdom undertaken in 2013/14. We focus on the ways in which particular publics were enacted as innovations in light emitting diode (LED) technologies and central management systems (CMSs) made lighting infrastructures newly visible. These technologies are now somewhat routine: England's National Highways, for instance, is committed to replacing the majority of its lighting with LED lanterns by 2027, and they have been adopted by most local authorities (UK RLG n.d.). However, in 2013/14, there was little data on how such installations fared in the real world. Many professional lighting engineers were grappling with the technical and financial uncertainties of newly installed, or planned, LED components and the CMSs that enabled remote, smart control. As one local authority lighting engineer explained:

The changes are rapid. And it's what knocks any attempt we make at a whole life cost analysis completely out the window. LEDs [haven't] been around long enough to do a whole life cost analysis. If you’d done a whole life cost analysis five, six years ago…all you had was the purchase and installation costs. There was no awareness that drivers would fail. You may not get five years out of your driver… It's in such an infancy, we are learning about it as we go along. (Lighting professional 8, London)

In the decade preceding our study, LED technologies had become (in the laboratory at least) more efficient, longer lasting, and offered a greater range of color temperatures, making them available for lighting rather than display purposes. By calibrating the color of light, LED lanterns can emit “whiter” light temperatures, with potential advantages for the visibility of detail at lower power usages. The greater efficiency—lumens produced per watt of electricity—offered potential for carbon, cost and light pollution reductions. By around 2008, LED technologies had developed sufficiently to enable workable public street lighting systems. Further, decreases in costs began to gather pace as more manufacturers started producing luminaires, and developing the CMSs that enable remote monitoring and management (including the ability to dim lanterns) through intelligent, centrally controlled “smart” systems. Engineering companies made offers to cash-strapped local authorities in the UK, often with an outsourced lighting management contract, to replace existing infrastructure with what promised to be more efficient and responsive street lighting as a cost-saving measure, with the bonus of also reducing carbon emissions (Shaw 2014; Schulte-Römer et al. 2019).

By the early 2010s, the proliferation of both new technologies and providers was bewildering. At one conference for (public sector) lighting engineers, a large hall full of commercial stallholders advertised their wares. Each had glossy brochures of lighting installations, lists of awards won for public lighting schemes, and technical details from laboratory studies of their system components. One delegate, whose company had a number of existing contracts with UK local authorities, enthusiastically described this vibrant competitive commercial field as “the Wild West at the moment!” (Fieldnotes, June 2014). This market had emerged from a panoply of socio-material circumstances in addition to technological developments, as financial, safety, sustainability and environmental factors all drove demand from urban planners (Shaw 2014).

The promises made by commercial infrastructure providers extended beyond cost-saving. Many also explicitly offered to “make” new publics through re-engineering the contemporary “smart” city. For instance, Philips Lighting (n.d.) promised its software would enable a “smarter and more livable” city:

Interact City IoT lighting software works with our connected luminaires to provide you with a robust infrastructure to improve city services, improve citizen safety, beautify public spaces, engage with citizens and encourage civic pride.

Our aim in this article is to trace how changes to street lighting infrastructure did “make” some publics. We begin by looking at the literature on infrastructures and publics. From this emerges a picture of infrastructures as multiplicitous and in-process, but a picture in which the role of publics could be better articulated. We then consider the framing of this study, which was commissioned in the context of uncertainties such as those expressed by the lighting engineer above. Analyzing data from this study, we show how the public took many forms, which recursively served and affected the processes of infrastructuring.

Infrastructures, Lighting, and Publics

It has long been argued that one quality of infrastructures is their abiding invisibility: if they work, they remain largely unnoticed (e.g., Star 1999; Bowker and Star 2000; Edwards et al. 2009). However, infrastructures themselves are profoundly heterogeneous both in their constitution and their function (e.g., Karasti and Blomberg 2018); are subject to maintenance that is continuous and multiplicitous (e.g., Harvey and Knox 2012; Blok, Nakazora, and Winthereik 2016); incorporate particular patterns of knowledgeability and ignorance (Anand 2015); and entail a topology of relations that can—crucially for the present article—include the “public.” In sum, infrastructures are not monolithic but in a state of constant adaptation or “oscillation” (Richardson 2016) as they align (or mis-align) with multiple, heterogeneous actors, and respond to new challenges and opportunities.

With this in mind, infrastructures are routinely oriented toward the public, and can be instrumental in the making of publics (Michael 2020). As consumers, as citizens, as users, as representations and representatives, publics are enacted by infrastructures in multiple ways. While there are various other routes through which publics emerge, infrastructures play a major role in shaping certain types of publics. In their usual invisibility, their routine operation and maintenance, and their occasional transformation and breakdown, infrastructures address, consult and mobilize publics through various means. Through bills and enclosures, through advertising and public announcements, through consultation, engagement and surveying, and through the socio-material provision of relevant services, publics come to be performed through infrastructures. Indeed, scholarship on public lighting covers how assemblages of lighting technologies, distribution systems and forms of governance have made publics, from Schivelbusch's (1987) account of how centralized lighting in Paris in the seventeenth century made an urban public realm, visibly governed by the king, through to Martland's (2002) example of how early electrification enabled city governors in South America to use urban lighting to “civilize” local populations in the nineteenth century.

However, this enactment is not a straightforward process. In an echo of the multifariousness of infrastructures, so too there are many versions of publics that circulate in relation to infrastructures who, complexly and multifariously, can be variously enacted as, for instance, citizens, consumers, users, partners and adversaries (or some combination of these). Such enactments are mediated through an array of methods such as advertising and public announcements, through processes of consultation and engagement, through modes of surveying and sometimes even experimentation. For instance, publics will be viewed as by and large “ignorant” of the processes and promises of Danish or German infrastructures (such as emerging smart grid infrastructures, e.g., Schick and Winthereik 2013), or they might be “recruited” as service-oriented communities who effectively mediate the operation of an Indonesian water infrastructure (Barker 2017), or they might resist Austrian biobank infrastructures on the basis of the “inappropriateness” of their proposals and engagement procedures (Goisauf and Durnová 2019). Put simply, in what might be regarded as a messy recursive process, publics can “oscillate” between representations (where, as parts of imaginaries, they affect the ways infrastructures run) and representatives (where, as political actors, they come to impact on those infrastructures).

Like other mundane infrastructures, those that provide public light at night do not usually register among users until there is some disruption—from a less than smooth upgrade through localized breakdowns, to dramatic system failures. Fox (2020), for instance, details the introduction of charges for public lighting in a former East German city, which surfaced tensions between the uncertain borderlands of private and public provision, and nostalgia for socialist abundance and social collectivism of decades past. New lighting technologies are a disruption that makes visible the work of infrastructuring, acting as a potential point of resistance, and bringing into being and mobilizing new publics. In a nuanced ethnography of new LED street lighting installations in Lyon and Berlin, Schulte-Römer (2015) describes these interventions as public experiments: conducted in public, largely funded with public money, and with reference to heterogeneous publics which might, or might not, have a view. Implementing new technologies can enact some publics as subjects (or “representatives” who contribute to negotiating its operation) while others are the objects of its operation (who are converted into “representations”). With regard to the latter, in terms of more explicit governance strategies, public experiments in lighting can be targeted at particular “problematic” locations and settings to gauge their effectiveness in, say, reducing the incidence of crime or road injuries. For example, Chalfin et al. (2022) found that randomly installing large floodlights around housing projects in New York City with continued high crime rates led to more safety. By contrast, Steinbach et al. (2015) found evidence that reducing light at night in UK streets reduced violent crime. These seemingly conflicting findings remind us that any effects do not attach to inherent properties of public light at night, but to pre-existing and emergent relations between citizens and governance in specific places. As such, publics both shape and are shaped by the specificities of change to light provision: to neglect this is to “repeatedly overlook certain elements” (Shaw 2014, 2234) of locations and settings, and the contingencies these introduce into any public experiment.

The Project: A Note on Framing, Methodology, and Ignorance

If lighting engineers in 2013 were facing uncertainties around the technical and financial implications of new technologies, other actors were grappling with uncertainties about their health implications. A paucity of evidence on human health effects of light at night had curtailed the ability of public health departments to advise on proposed changes prompted by the new availability of LEDs and CMSs (Vohra 2013). This article draws on data from a project commissioned by the UK's National Institute for Health Research in response to these uncertainties (Green et al. 2015; Steinbach et al. 2015). Given lighting professionals’ own awareness of their uncertainties, 62 local authorities in England and Wales volunteered to collaborate with the study, and to provide access to their data on lighting installations to assess their impact on road injury and crime (Steinbach et al. 2015).

In eight of these areas, we also undertook rapid ethnographic research to look in more depth at “what was going on” as local authorities grappled with new installations. In contrast to more traditional forms of ethnography that involve a lone researcher spending a long time in one bounded field, rapid ethnography draws on shorter periods of focused fieldwork, with intense data collection (Knoblauch 2005). In our case, fieldwork was done by a team of researchers (including Authors JG, RS and PE) spending 2 to 3 days in each site, doing walk-arounds, conducting informal and intercept interviews in public spaces like shopping centers, transport hubs, and pubs (N = 61), as well as pre-arranged more formal in-depth and group interviews with lighting professionals (N = 16) and residents or workers (N = 57), focusing on those likely to be traveling at nighttime (police officers, transport workers, hospitality workers). Additional data included observation at national and regional events for lighting professionals, documents (policy, technical and commercial materials; submissions to formal consultations; local media coverage), and a household survey of experiences and attitudes in affected and non-affected streets. The study was approved by the London School of Hygiene & Tropical Medicine (LSHTM) Ethics Committee (no. 6341).

Our research design shared in the mess that is typical of social scientific research (Law 2004): it necessarily entailed a patchwork engagement with the empirical world. Three particular contingencies are important to framing the analysis below. First, given its funding, organizational and methodological parameters, our study was also embroiled in the process of infrastructuring and the making of publics. To an extent, an “ignorant” public was pre-empted by the steer of the funders. Given uncertainty about public views about changes, they advised we focus on the “general” public rather than what were framed as “activist” publics: indeed, an initial bid for an action research project collaborating with local authority consultations was rejected as likely to only include publics with “vested interests.” Thus, a “pure” public was envisaged in the framing of the research; a public unpolluted by relevant knowledge, views or politics. This expectation is not atypical: as several scholars in public engagement with science and technology (PEST) have remarked, publics who are disengaged from the issue at stake are often regarded by funders as the ideal public-in-general (Michael 2009) from which impartial data can be derived (Lezaun and Soneryd 2007).

Second, lighting professionals had a dual role, as both partners in the research and its “subjects.” The analysis below draws on interviews with them as key informants, but we were also a conduit through which knowledge about publics was fed back to them, and they were also of course themselves engaging with publics. In this respect they, and we, were practitioners of PEST. Public engagement takes many forms and is implemented by many different sorts of actors. In brief we can note that PEST is generally understood to aim at enabling citizens to enter into the process of scientific and technological decision-making through invites to participatory events of various sorts in which they can voice their views. Critical commentary on this broad endeavor argues that PEST actually enacts public voices—shaping, silencing and deflecting—in ways that diffuse what is often performed as an exercise in democracy (e.g., Irwin 1995; Irwin et al. 2013; Chilvers and Kearnes 2016). In this view, PEST can be grasped as a repertoire of techniques that compose participants as particular sorts of citizens (e.g., Lezaun and Soneryd 2007).

The “uncertainty” of professional expertise was also prefigured by the volunteer collaboration of local authorities aware of their own needs for evidence. Nonetheless, the matter of ignorance loomed large in our data in other ways: how ignorance is ascribed and denied is of enormous importance in the infrastructuring of street lighting, operating at many levels from acknowledged (lack of) knowledge about the effects of LED lighting systems through (lack of) knowledge about comporting oneself as an “expert,” to (lack of) knowledge about how to “do” social science. Although the initial findings of this commissioned study were oriented toward reducing uncertainty, continued debates about the impacts of street lighting changes have since brought into focus the irreducibility of uncertainty, and prompted a retrospective analysis of how this folded into the making of publics. This will be a central focus of our analysis.

Third, our data are now around a decade old. We do not offer this analysis as an empirical description of contemporary lighting infrastructures in England and Wales, but rather of how publics were enacted at a point when infrastructures were in flux, and thus visible. The specific publics will have shifted in the last ten years, yet we note in a brief update that the general point about mobilizing knowledge and ignorance remains important in the making of publics, as is the fact that those publics are always enacted vis-à-vis other entities.

Expertise and Uncertainty

Lighting professionals, as we have noted, were highly attuned to the contingencies of publics and infrastructures. These had been brought into sharp focus by the possibilities of novel LED and CMS technologies:

One of the things we’ve done, we replaced some quite high-energy carpark lighting with LEDs …We’re doing that sort of hand-in-glove with our colleagues over in CCTV, because [of] the fear of crime. … Are they still getting footage that's as good or better? …We just haven't come up with a dimming profile yet because we’ve got to understand pedestrian flow, their nighttime activity [then] we’ll go and sit down with the designers and our people and [ask] what's the flow through there before we start dimming them? (Lighting professional 4, north England)

Most lighting professionals were acutely aware of the chronic partiality of such experimentation. Reducing uncertainty in one area (say, the impact of dimming on CCTV visibility) revealed new gaps in knowledge in another (the impact on pedestrian activity). Across the eight local authority areas we studied, lighting professionals spoke of a range of specific unknown “technical” potentials of new lighting technologies. These included the longevity of LEDs and other components, thermal management, surge protection and the impact on safety, with the hard light cut-off that created contrast between lit and unlit areas.

More fundamentally though, uncertainties around specific components rendered the behavior of the whole system (and its interface with other systems) potentially unknowable “in the wild.” These uncertainties around how new technologies would interact with publics were crucial to experts’ thinking about what would best serve the lighting infrastructure. In what follows, we build this picture of lighting professionals as acknowledging their own ignorance by exploring patterns of knowledgeability and ignorance, and how these might be embodied in diverse versions of the public. Importantly, these emergent publics were enacted in relation to other actors or entities. That is, the term “publics” applies to any actor that can be contrasted to other actors who putatively possess expertise: publics are enacted specifically vis-à-vis other entities.

Supra-Publics Vis-à-Vis Residents

Lighting professionals were much exercised by their relations to their publics, and how they saw themselves as mapping onto the public “good.” While they identified as technical experts, their accounts in interviews also gave a strong sense of “knowing” the civic spaces for which they provided lighting:

I’ve got over 30 years invested in street lighting at various levels. Yeah sad isn't it? [laughs] Even, if you really want to be bored…my great, great grandfather…put…the original Swan neck columns up on [seaside town] seafront. (Lighting professional 2, London)

So, despite the technocratic expertise drawn on by lighting professionals (concerning not just lighting technologies, but also regulatory frameworks and standards, contracting issues, and statistical relationships between lighting and crime, traffic, health, etc.) an expertise in lighting-as-governance was also pertinent: that is, local knowledge of how resident publics interact with other components of the infrastructure:

[we] switch the lights out or just have feature lighting. If you light the parks, people think they can walk in…they’d be full of homeless people. (Lighting professional 2, London)

These quotes point toward to a sensitivity to local publics. That sensitivity around the public as embedded within a techno-social assemblage of light at night was summed up by professionals as having skills in assembling this system of components to provide “the right light, in the right place, at the right time.” This was a slogan quoted by many over the course of our fieldwork. In this respect, our data suggest that professional identity and pride were predicated on enacting technocratic public service through managing the recursive relationships between publics and the lighting infrastructure.

In mobilizing both technocratic and embedded local knowledge, lighting professionals could enact themselves as something akin to “supra-publics” vis-à-vis local residents. They were, in other words, much better placed to understand what the public want than the general public itself. As such, successful infrastructuring both enacted and required a comparatively ignorant public. A job well done was one where lighting changes made for the broader public good—to save expenditure on lighting, to foster the night-time economy, to improve security, or to exclude certain people from certain areas—would ideally be effected in ways that did not draw attention to the infrastructure's operation. One proudly recounted: “I took the councilors out to see the streets we’d dimmed—they couldn’t spot the dimming” (Lighting professional 15, north England). In some ways, this was business as usual—ensuring the ignorance of publics to ensure smooth operation but also development of the infrastructure. In the latter case, in installing the new LED lights with a visibly different color temperature, one common tactic was deliberately installing too bright or unshielded lights initially, so that they could then be turned down or shielded when residents complained. Although the public were sensitive to the changing brightness, they were ignorant of the tactic (and the relatively smooth transition to LED lighting). This could of course be framed in Foucauldian terms as an occlusion of the operation of power, or as the strategic deployment of ignorance. And yet, the lighting professionals were simultaneously enacting themselves as supra-publics vis-à-vis the local residents for whom they provided services. In this respect, they are “superior” because they can legitimately and more accurately represent the authentic desires, needs and claims of those residents.

That local residents were successfully rendered as ignorant of the operation of lighting infrastructure was amply demonstrated by the intercept and informal interviews we conducted in public spaces. Across all eight case study areas, when asked about local changes to street lighting, most reported having not noticed the change or if they had, did not much care (one London couple's response “We haven’t really noticed it have we?” was echoed in the Midlands by “We haven’t noticed, really,” and in Wales “to be honest, I haven’t got an opinion”). In a household survey of 500 residents, we found almost no difference in views between residents surveyed in unchanged roads and those where lights had been turned off at night (Green et al. 2015). Put crudely, there was a general tenor of “don’t know, don’t care” among local residents—a finding reflective of what Anand (2017, n.p.) has called the dullness of infrastructure. Street lighting is simply one of the “mundane and banal features” of local environments, whose best management is appropriately delegated to technocratic expertise.

However, public providers of infrastructure are not invariably legitimate as supra-publics: when public ignorance is disrupted, such claims to legitimacy potentially unravel. A critical challenge faced by the lighting professionals was that, in many areas, they no longer made and managed decisions. With cost-cutting and reorganizations, many professionals with technical expertise had been replaced by general facilities managers who brought in external lighting consultants when needed. Knowledge of the local area and knowledge of technical systems were no longer necessarily co-located:

Those making decisions may not know enough about local contexts to make decisions about street lighting. “No two roads are the same:” detailed knowledge, by those who are familiar with locality, is needed. (Fieldnotes, workshop discussion with lighting engineers, September 2014)

In some ways these general facilities managers, sometimes outsourced to external providers, comprised failed supra-publics, lacking expertise in assembling “the right light, in the right place, at the right time” and, crucially, expertise in enrolling publics as integral components of the system. They lacked the skills in maintaining public ignorance about street lighting, and thus risked occasioning new, resistant publics. As an example, below we present three lighting professionals recounting what seems to be a well-known recent incident when business-as-usual public ignorance was catastrophically disrupted once the infrastructure of street lighting became visible, when a council switched off alternate street lights. This story was presented as demonstrating the almost caricatured ignorance of a failed supra-public in the essential skills of enrolling compliant publics by keeping infrastructure invisible:

R8: Whereas in [names local authority], they found their budgets from the government were slashed and somebody high up in the council said, right we’ll just switch every other street light off. Which is kind of the worst way of doing it! …And what we saw was public meetings in, in every town and, and city in [county] marching on—

In sum, the changing personnel structure of local authorities born of funding cuts reshaped the lighting infrastructure, introducing failed attempts at enacting a supra-public of general managers and outside contractors. The ignorance that informed the actions of these failed supra-publics, as contrasted to the local knowledge of lighting professionals qua supra-publics, precipitated public outcry.

An Update: Publics in Flux

Since our fieldwork, the terrain of knowledge and uncertainty about the impacts of street lighting infrastructure has shifted. Our own study (Steinbach et al. 2015), found no evidence that street lighting changes were associated with increased road injury or crime, and there is now considerable data in the public domain about the costs associated with LED and CMS installations (Shaw 2020; UK RLG n.d.). As Shaw (2020) notes, in times of austerity, cost data can trump other concerns. LED installations in public lighting are now routine. Recent years have seen the backgrounding of debates on the projected economics and safety implications that so exercised lighting engineers and research funders at the time of our fieldwork, but uncertainty overall has not reduced. Rather, as LED installations have become routine, new uncertainties have come to the fore (Schulte-Römer et al. 2019), regarding their contributions to light pollution (Kyba et al. 2023), and (crucially) the impact of the typically “bluer” light on circadian rhythms, nocturnal species and human health (AMA 2016; Tähkämö, Partonen, and Pesonen 2019).

Over the last decade, the publics we delineated vis-à-vis other entities have inevitably morphed in relation to these new domains of knowledge and uncertainty. Some occasioned publics fade into obscurity once infrastructure stabilizes again into a mundane backdrop. Others mobilize as activist publics. Whereas the impact of the light color on human well-being was barely raised by participants, residents or professionals in our fieldwork in 2013/14, this has now become a topic around which activists organize, with a charity established in 2015 to campaign for the rights of those particularly sensitive to white light. ¹ Similarly, activist networks campaigning for dark skies—often instigated by amateur astronomers—have proliferated in the UK and elsewhere (Zielińska-Dabkowska, Xavia, and Bobkowska 2020), even though in our fieldwork the impact of lighting reductions on light pollution was not raised in urban settings (Green et al. 2015). Other actors also force awareness of infrastructural affordances: one result of a ransomware cyberattack on a UK local authority in 2024 was the visibility of new management systems, as street lights remained on through the day while the CMS awaited an IT reset. ²

Discussion

We have described a range of ways in which publics are made through the processes of infrastructuring street lighting, and enrolled successfully (or not) in the operation of that infrastructure. First, in the provision of public infrastructure, providers can and do make claims for being a supra-public who can act in the best interests of a resident public. The legitimacy of this claim rests on a particular combination of knowledge and ignorance: that is, technical expertise coupled with expertise in local publics that can forge a stable assemblage of taken-for-granted light at night. This is tempered with professional awareness of the inherent uncertainties of technological change, particularly around how the (then new) technical components of lighting infrastructure would perform “in the wild.” In tracing the politics of water infrastructure in Mumbai, Anand (2015) points out that professionalism relied on ignorance as much as knowledge: water engineers’ efficacy was predicated not on more accurate monitoring of inherently leaky and unstable infrastructure, but on knowing when leaks could be ignored. In our case, too, managing intersections of knowing, not-knowing and uncertainty were key to professional legitimacy as supra-publics. Where there are failures in assembling these components of knowledge and ignorance, claims to legitimacy fail. Here, lighting professionals can position themselves as publics vis-à-vis other functionaries in, or contracted by, local authorities. The broader lesson here is that those who operate infrastructures can situate themselves as supra-publics, and it is their patterning of knowledge and ignorance that grounds their legitimacy with user publics. However, as noted from the outset, infrastructures are heterogeneous and fluid assemblages, so organizational changes (including where other authorities intervene inappropriately) can undermine the status of supra-publics by shifting this balance.

Second, we identified publics vis-à-vis occasions who are (often inadvertently) brought into being as the operation and work of infrastructuring becomes visible during change or failure. Here occasions of engagement by infrastructural actors, which may be more or less explicit, occasion publics who might be more or less ignorant, more or less knowledgeable, about light at night, in terms of its governance effects, esthetics, or impacts on everyday life. Lighting professionals must tread with care in making such occasioned publics: consultations make visible the hidden work of infrastructuring, and risk resistance and unravelling. Yet insufficient consultation also risks undercutting the legitimacy of supra-public claims. This is further complicated by the occasion of research. Research processes such as ours also make publics by rendering private views into public knowledge, and by offering the potential of representing those views to the publics’ others (relevant authorities, for instance). The wider implication here is that publics are occasioned by various forms of engagement (consultative and research in the present case) but there are others and admixtures of these. The point is to nurture attunement to this complex and fluid occasioning, and to its ramifications for analysis of the intersections of publics and infrastructures.

Finally, publics vis-à-vis governance can be successfully enrolled in infrastructuring on the basis of their experiential expertise in living with lighting. However, this can unravel when knowledge claims encroach on those of experts, and where there is insufficient skill to delineate appropriate mobilizations of uncertainty and ignorance. This highlights the need to be analytically sensitive to the form as well as the substance of lay expertise and its impact on how publics come to be regarded in relation to the operation of infrastructures.

We are aware of the partiality of our analysis. In addressing infrastructures, we acknowledge that they are highly situated and that what counts as an infrastructure varies: Russian residual planned-economy urban infrastructures (Rudenko and Zemnukhova 2017) and privatized Indonesian water infrastructures (Barker 2017) are unlike forms of infrastructure found in UK local government. Our findings may reflect the specific kinds of publics made when local authorities have some (however minimal) obligations to democratic process, where public services are reasonably well-funded and delivered by those with technocratic and local expertise, and where public views are a legitimate research topic for both providers and research funders.

This specificity raises a question about the intentionality of ignorance (Croissant 2014) and its politics (McGoey 2020). Public ignorance is often problematized in questions of participatory democracy. When ignorance or uncertainty have been strategically manufactured by polluting industries, or where areas of scientific research have been systematically under-resourced, through the structural generation of epistemic inequalities, such ignorance is a challenge to democratic governance. However, in this case, a focus on strategic ignorance is perhaps misleading. It is not that there are no commercial or political vested interests in this field: certainly, corporate players have an interest in profiting from new technologies, and neoliberal austerity policies have had a major impact on both the capacity of lighting engineers (Shaw 2014) and relationships between local governments and their citizens. In this case, though, the strategic uses of ignorance and non-knowledge were perhaps less salient than other mobilizations and attributions of non-knowledge.

To be sure, lighting professionals did intentionally produce public ignorance—indeed, an invisible infrastructure was an aim of their work, as a marker of a technocratic job well done. There is a strategic element to their work as local authority functionaries who deploy lighting as governance, in demarcating public space and time for use or exclusion, policing normative expectations of movement through space, and curtailing or enabling rights to light at night. This was, largely, a role accepted by residents as citizens (“don’t know, don’t care”), conscious of themselves as publics vis-à-vis technical experts. When routine ignorance was disrupted and lighting infrastructure became visible, residents could become aware of infringements on the right to public light at night. Yet, largely, the deficit that mattered was not technical expertise: most publics mobilized not a right to technical knowledge, but a right to have their experiential expertise (of esthetics, of safety) incorporated. One reading is that of democratic virtue: here, perhaps, was evidence of public engagement as a functioning check and balance for the over-reach of supra-publics. However, McGoey (2020) cautions that accepting the usefulness of ignorance risks underplaying the social relations of knowing: that the ability to observe, comment, and challenge are not equally distributed. In this study we have not traced the sociodemographic dimensions of knowledge and ignorance. Our methods could not access the unhoused, made less welcome in parks due to lack of lighting, or those less likely to be in the public sphere accessed through our walk-about methods because of gendered, racialized or other structural exclusions (Pain 2001). The kinds of exclusion wrought by lighting infrastructure as governance are clearly cut across with questions of power and social division (see for instance Chikowero 2007 on electrification, settler colonialism, and public lighting). We have also not attended to whether changes implemented were aligned with other governance projects, or how they played out in terms of relations of social power.

Within the bounds of this limitation, we have nevertheless traced several ways in which publics featured in the operation of a particular infrastructure. The key points are that these publics are multiple, emergent as part of the recursive process of infrastructuring street lighting; and that publics play a key role in making light work. Much research on publics in relation to techno-science and environmental infrastructure is oriented to how issues become matters of concern for publics, and more normatively, on how public knowledge and values should be incorporated into decision-making (Fiorino 1990; Ottinger 2013). Drawing on pragmatist philosophy and insights from science and technology studies, Marres (2005, 2007) discusses how “matters of concern” arise for publics from the co-constitution of issues and publics in relation to the uneasy hybrids of techno-science. That is, publics both emerge from contestations over issues (in our example, light at night), and shape the emergence of those issues as political, as well as technical or scientific, problems. As Fox (2020, 654) noted with regard to the introduction of charges for street lighting in a German city, “Infrastructures constitute a key point of interface between citizens and government,” and transformations in that infrastructure are moments when tacit understandings are surfaced, and relationships between citizens and administrations can potentially become matters of concern. There were, in our study, instances where issues did become politically framed, as with Callum's organization to protest street lighting changes. But these citizen publics which arise when issues are not being dealt with by the business-as-usual of politics (Marres 2005) were not the only publics enacted by infrastructuring.

Conclusion

Taking a case study of a point when novel technologies had unsettled the mundane invisibility of street lighting in the UK, we have delineated some of the various publics that were enacted in relation to infrastructuring. These publics were constituted vis-à-vis other actors: supra-publics vis-à-vis a (less technologically knowledgeable) local citizenry; publics vis-à-vis occasions of consultation; and citizen publics vis-à-vis governance. The careful deployment and attribution of expertise and ignorance are crucial to the making of these publics. As the specific domains of expertise and ignorance in relation to lighting technologies have shifted in recent years, new publics have emerged. Our categories of public are neither comprehensive nor mutually exclusive, and they are certainly in constant flux. However, our general point—that the publics that circulate within, through and around infrastructures such as street lighting are a core component of the necessary assemblages for “making light work”—holds.