Hostname: page-component-745bb68f8f-b95js Total loading time: 0 Render date: 2025-01-11T02:54:33.291Z Has data issue: false hasContentIssue false

Cutting without Cutting Connection: The Semiotics of Power Patrols in Urban Tanzania

Published online by Cambridge University Press:  01 January 2025

Michael Degani*
Affiliation:
Johns Hopkins University, USA
*
Contact Michael Degani at Department of Anthropology, 466 Mergenthaler Hall 3400 N. Charles Street, Baltimore, MD 21218, USA (mdegani1@jhu.edu).
Rights & Permissions [Opens in a new window]

Abstract

Before the total cutoff of a household for unpaid electrical bills, and in anticipation of the violence it might precipitate, national power utility inspectors in Dar es Salaam have recourse to a gradient of ‘soft’ disconnections designed to signal good faith and encourage repayment. Drawing on a mixture of semiotics and exchange theory, this article argues that such soft disconnections may be understood as moments when the shared grounds of the household-utility relation—both the physical power network and social commitments it embodies—become figured, modified, and ultimately preserved in the face of urban postcolonial strain. Whereas semiosis is often thought of as a kind of infrastructural bridge that links sign to interpretant (or object to sign), the inspectors’ reflexively phatic signaling of common grounds highlights the thoroughly semiotic nature of infrastructure itself. More generally, the very presumption and preservation of shared grounds is a salutary alternative to the anti-relational violence and extinction that characterizes much of the contemporary world.

Type
Articles
Copyright
Copyright © 2021 Semiosis Research Center at Hankuk University of Foreign Studies. All rights reserved.

One afternoon in 2011, I was conducting fieldwork on electricity use in Dar es Salaam, Tanzania’s major metropolitan city. Accompanied by my friend and Tanzanian research assistant Thierry, I was riding along with a Disconnection Team (DC team for short) dispatched by the national power company TANESCO. DC teams are responsible for inspecting meters for irregularities and for tracking down and cutting off indebted accounts. One day, the company truck dropped us off in a hilly, middle-class neighborhood called Sinza. I followed the two inspectors as they walked purposefully towards a large house perched on a gently descending slope. Outside, two young girls, perhaps around nine and fourteen years old, were washing clothes and seemed to be the only people around. As we got closer, the inspectors pointed out how the service line that extended from the nearest utility and tapered down to the house was not actually connected the wall-mounted meter, but rather directly to the household circuit.

One of the inspectors, a woman named Sabina, explained:

They were cut two days ago. They had a debt of 7 million shillings [~$2,500].Footnote 1 At least they could’ve reconnected through the meter—at the very least. But they’ve dodged the meter entirely. Better to get charged for having your power cut and reconnecting it, rather than getting charged for just stealing. Behold Tanzanian life. This is the problem. TANESCO is going down because of the stupidity, the ignorance of not paying one’s bills.

In cases of premeditated theft, the minimum response is usually to cut the wire itself, which Sabina’s team member, a middle-aged man named Chilambo, proceeded to do. However, perhaps due to the magnitude of the debt, or the audacity of stealing power just two short days after the patrol came through, Sabina decided to invoke legal measures. The plan was to pursue a kibali (license or warrant) and return with the police later that night to make an arrest. As Sabina explained these measures to us, the two young girls began inching back towards the house. Seeing this, Sabina called to the older one, playfully. She took out a small digital camera from her bag, lined up the two girls and snapped pictures of their faces, wide-eyed with fear. Laughingly, she taunted them:

Ala, c’mere! It was a sweet day for you renters today, you felt cool when using the easy power. Have you ever slept at the police station? The landlord’s name is what? You don’t know? Our police are young … you’re the one caught in the house, and so you have to go. Tonight they’ll be sucking on your breasts (leo watauynyonyosha matiti)!

Though I have explored this encounter elsewhere (Degani Reference Degani2017b), I have returned to it here in part because it was one of the most unsettling but also most unusual and therefore instructive scenes of my fieldwork. Sabina’s seemingly casual threat of violence threw into relief the fact that most of the time, inspectors were far more civil when disconnecting households. The circumstances of brazenly reconnecting to the grid—and what is more, bypassing the meter—had pushed her into a far more punitive and indeed violent stance. Clearly, the household had crossed a line, and this in turn had provoked Sabina to cross another.

Taking this scene as a starting point, this paper analyzes the workings of Dar es Salaam’s power infrastructure as a fundamentally semiotic process through which utility and household continuously interpret each other. For as power flows out through transmission and distribution lines to a bristling mass of consumers, and as payment flows back to TANESCO, a communicative circuit is formed, one that, because TANESCO is a state-owned monopoly, embodies a shared project of nation-building (kujenga taifa). And yet this circuit is fragile. Service interruptions, high tariffs, and aging materials all put pressure on the downward flow of current. In turn, consumer theft, non-payment, and vandalism both accommodate these shifts and put pressure on the reciprocal upflow of currency. Some of these perturbations are forgivably minor. Others reach a level of intensity that calls into question the reciprocal binding they presuppose. Indeed, at certain flashpoints, network breakdown threatens to erupt into arguments, protests, and even violence.

Most of the time, however, I found that a certain rhythm held. Chronic and cruddy though it may be, the circuit goes on respiring. In part, I have come to see, this owes to the way households and utility respond to each other in a way that is reflexively predicated on the ability to continue responding. Practically, this means inferring which stresses and strains their exchange can tolerate and which it cannot. Pace Sabina’s scorched-earth invocation of imprisonment and abuse, the better strategy was to modulate action, whether by allowing a bit of debt to slide or making sure not to get too out of hand in the first place.

And what happens if a household attached to the network does start to fall out of rhythm? To answer this question, I will ask the reader to picture the network as comprising two levels. The surface level is composed of the current/currency circuit I describe above. The deeper level is the infrastructural connection—the piping and presumptions—through which such flows circulate in the first place. These two levels relate to each other as figure to ground. And just as household and utility vary in their give and take of current and currency, they sometimes vary the ground of that give and take. Sometimes, that is, they modulate the very connectivity of the network itself so that the entire field of what is possible and permissible to give and take shifts.

In everyday life, people shift across grounds all the time. Interactions that would be appropriate and efficacious with friends are not the same as those that would be appropriate and efficacious with a mother-in-law. Speakers code-switch depending on whether they are at home or work, at the front entrance or in the backyard. However, it is also the case that grounds can shift within a given relation. That is, when something goes wrong between actors, they often have recourse to a kind of liminal mode that preserves the possibility of transitioning back (or forward) onto more normative grounds. These cool-down modes too are widespread in social life: penalty boxes (but not ejection), truce (but not peace), rationing (but not blackouts), moratoria (but not banishment), and, in our case, soft disconnections (but not cutoffs). As I will show, such modes are related to, and perhaps a variant of, phatic communication: ritualized empty speech, such as “hey” that suspends referential content in order to inaugurate (or repair, or maintain, or otherwise modify) the grounds of interaction.

One reason such phatic modes of suspension interest me is that they highlight the dynamics of how to properly be in relation to something. Eduardo Viveiros de Castro (Reference Viveiros de Castro2014, 170) observes that the relation is the “master concept in anthropology,” though more recently it seems fair to say that relations have become an ethnographic object in their own right. It is not hard to see why, given that degradation is our default political and ontological state. Escalating climate change, pandemics, mass violence, extractivism: these horrors are both root and fruit of an anti-relational way of being in the world predicated on the power to exhaust and extinguish. Against its gravitational pull, anthropologists have come to reappraise even the most ordinary relations as a fragile achievement: how, in this era of extinction, can (and do) we make kin of each other and of non-humans? How can we relate to technology, commodities, and natural resources in ways that allow for sustainability and repair, rather than punishing cycles of innovation and planned obsolescence (Jackson Reference Jackson2014)?

Rather than romanticizing being in relation as inherently harmonious, the best of this work is admirably frank about the knotted processes of conflict and cooperation that constitute them—about the possibility that relations might grow corrosive, suffocating, toxic, or debilitating. Any relation is a kind of living bridge across difference; the tricky question is always what parts to repair (and when) and what parts to decompose (and how). Or, as Eduardo Kohn (Reference Kohn2013, 218) puts it, citing Donna Haraway’s provocative discussion of animals (Reference Haraway2008): “[K]illing … is not the same as killing a relation. And killing may actually permit a kind of relationship.”

For utility inspectors, cutting off a household’s power may indeed actually permit a kind of relationship, because cutting, this article will suggest, is not quite the same as cutting connection. Along the way, the article contributes to a broad disciplinary concern with relations by triangulating two theoretical traditions that are not often, or at least not explicitly, in conversation with each other.Footnote 2 One is that of Peircean semiotics, mostly filtered through the work of Paul Kockelman (Reference Kockelman2017). The other is broadly concerned with gifts, money, and exchange (Sahlins Reference Sahlins1972; Parry and Bloch Reference Parry1989; Mauss Reference Mauss2016). Both traditions are attuned to the importance of what Kockelman (Reference Kockelman2017, 9) calls the “phatic commons,” the shared grounds that relate sender to receiver, giver to taker, self to other, origin to destination. In any and all of these dyads, there is always, however implicitly, a “third” (Kockelman Reference Kockelman2017, 148) that connects them and that, in complicated ways, is irreducible to the relata it relates. And the presence of this third means that, between the full connection and the full disconnection of a service line, there is a range of interesting and productive intermediate possibilities.

Gifts, Infrastructure, Thirds

As is well known, Peirce’s definition of semiosis is fundamentally triadic, in contrast to Sassurean binaries: a sign stands for an object insofar as it gives rise to an appropriate and efficacious interpretant. In his own wide-ranging work, Kockelman emphasizes that same triadic logic applies to all manner of relations. For instance, sign-object-interpretant is equivalent to object-code-sign, input-algorithm-output, signer-infrastructure-interpreter, or (most broadly) origin-path-destination. In each case, the middle term functions as the ground that allows the object-sign, input-output, or origin-destination pairing to emerge as figure. The shared grounds that link two people in a face-to-face interaction partake of a bit of each of these: if I offer up a wave to my friend (qua input/origin) and she responds with a wave in turn (qua output/destination) it is our shared grounds qua ensemble of social conventions (who we are to each other), codes (waving means hello), and technical channels (we can see and hear each other) that opens up the pathway between my sign and her interpretant, allowing them to correspond. Put another way, it is the grounding of a third (even if it is only imagined or assumed) that ensures the sign-interpretant dyad is semiotic (i.e., meaningful), that the sequence of our handwaves is something more than what Peirce calls the mere “secondness” of contingency or coincidence (Hénaff and Doran Reference Hénaff2013, 75).

The main interaction between a household and TANESCO can be framed in just this way. Let electrical current flowing into households be a sign and currency flowing back out to TANESCO be its interpretant. The object that grounds this exchange is comprised of infrastructure, codes, and conventions: the technical channels linking household and utility, the tariff that codes the cost of a kilowatt hour, and the interlocking rights and responsibilities household and utility hold for each other. Thus, every successful completion of this sign-interpretant sequence is, to consumer or utility, a figure that indexes its ground: a shared sense of equivalence, a shared capacity to communicate, and a shared social commitment to each other.

It is interesting to observe that a broadly substantivist tradition of economic anthropology begins with the comparable intuition that material exchanges are figures whose grounds are relations. Marshall Sahlins (Reference Sahlins1972, 185–86), for example, asserts that “a material transaction is usually a momentary episode in a continuous social relation. The social relation exerts governance: the flow of goods is constrained by, is part of, a status etiquette.” There is a telling ambiguity in that last clause: is the flow of goods merely constrained by, or does it directly express, status etiquette? In a more complex formulation, Jonathan Parry and Maurice Bloch (Reference Parry1989) account for both possibilities through what they call long-term and short-term “transactional orders.” As Keith Hart (Reference Hart2001, 271) observes, the theory is Durkheimean: “everywhere there are circuits of social life: one, the everyday, is short-term, individuated and materialistic; the other, the social, is long-term, collective and idealised, even spiritual.” Kula and Gimwali (Malinowski Reference Malinowski[1922] 2002); marriage and market (Bohannan Reference Bohannan1955); wintertime communion and summertime dispersal (Mauss [1979] Reference Mauss2004); in each case one can find a short-term realm in which exchanges more or less freely flow across certain grounds and a long-term realm of status in which exchanges performatively constitute the grounds themselves.

Crucially, then, such grounds are not just there in any simplistic sense; they must be created and affirmed. Thus, in his reading of Marcel Mauss as a political theorist of the social contract, Sahlins (Reference Sahlins1972, 169–70) suggests that gift exchange is a way of inaugurating a relation of trust between strangers or otherwise unrelated people or groups, signaling they can get down to the ordinary business of exchanging substantive words or things. From this perspective, such ritualized exchange is phatic. It is a metacommunicative exchange of messages that betokens and thus performatively constitutes the very possibility of exchanging messages. In a marvelous essay (Reference Elyachar2010), for example, Julia Elyachar analyzes the daily “phatic labor” of Cairene women who visit each other for tea to exchange pleasantries. These small social graces nourish relations that can deliver actual substantive goods (favors, information, a word to the right person in government) should the occasion arise. Such visits, then, are a kind of maintenance work, ritually lingering upon and attendant to the channel/relation as such. The mutual overlap with more stereotypical infrastructural channels should be clear: to extend water pipes or electrical wires is not to transmit water, power, or data, but to extend the conditions of possibility for those vital flows, the delimiting ground against which a figure can emerge. They are real channels, given and received. But like empty speech, they are hollow in the middle. Their meaning is virtual, lying in the future meanings they make possible.

In Dar es Salaam, extending a service line is both a gift and a phatic communique. Its installation does not determine how often or how much power a given household will consume. It simply opens a channel, elevating the household into a long-term transactional order with the state power company. As Tanja Winther aptly put it in her ethnography of the electrification of Zanzibar, “the costly and solid physical connection between consumer and utility binds the two parties together like spouses in marriage: tied to each other through the good times and the bad times and with a high cost in the case of separation” (Reference Winther2008, 40). Like matrimonial vows, a service line also has ritual formalities and proscriptive constraints. While residents must pay for the parts and installation of the service line, they do not actually own the wires, poles, or meters that comprise them. Electricians often spoke to me of the meter as a “gate” or border beyond which is TANESCO’s property and responsibility; households cannot repair or alter service lines nor take them if they move, and they certainly cannot sell them. As Mauss (Reference Mauss[1925] 2016) famously theorized of the gift, in a service line one receives not just the object but something of the giver as well.

To be sure, not all relations are wholly desirable. Critical theory has its share of primal scenes in which phatic communication lays the grounds for power and subjection. This is the ruse of the cop’s “hey you” that constitutes Louis Althusser’s (Reference Althusser2006) primal scene of ideological interpellation: once you have turned around, it’s too late to have done anything else; you were, retroactively, always the guilty party. For Jacques Lacan, the gift of language is a trap, a “trojan horse” that seems innocuous but ends up, once you have accepted its premises, colonizing what you can say (and think, and do) in relation to the society that gave it to you (Lacan Reference Lacan2007, 275). In a more historicist vein, this is a central dynamic of what Michel Foucault (Reference Foucault1983) called the “conduct of conduct.” The art of governance often involves the laying of grounds—channels and codes, infrastructures and institutions (roads, social media, prisons, languages)—that, despite their significant start-up costs, are ultimately more efficient than mere despotism (Mann Reference Mann2008). Conducting the conduct of dogs, algorithms, children, men, women, workers, colonized subjects, and so forth involves enticing or enclosing them within certain grounds such that only particular interactions with authority become technically possible or socially permissible. Over time, subjects may internalize that enclosure, but then again they may well refuse to be addressed, subject to, or connected under existing terms (Simpson Reference Simpson2014).

The possibility of refusal (or at least frictive resistance) points to a key set of criticisms of a broadly Durkheimian account of thirdness as channeling solidaristic, functionally cohesive interactions. When anthropologists emphasize the noncontractual element of the contract, the consensual over the conflictual, they are in danger of simply turning utilitarianism on its head. In Pericean terms we might say that by overemphasizing grounds they are ironically in danger of recapitulating secondness—not the secondness of brute actuality but of brute efficiency, the secondness of the programmer, planner, or priest, all aspiring to a kind of computational always-already in which sign-intepretant relations never deviate from their objective. But of course the best laid plans of mice and men go awry: what Michel Serres and Lawrence Schehr (Reference Serres1982) called parasites—lags, interference, interception, enemies—mark the fact that “even the most conservative and hegemonic of social orders is ultimately unstable and history-riven” (Keane Reference Keane2018, 71). One consequence, as I explore below, is that internal frictions within a relation can drive actors to upshift and downshift across different modes: ruler and ruled might shift from command and obedience to a war footing of insurgency and counterinsurgency. Intimates might become extimates; revelers might become rioters (Singh Reference Singh2014); diplomatic friends might become enemies—or frenemies (Weichselbraun Reference Weichselbraun2019). Terms of engagement can drift backwards, forwards, or sideways.

This much was certainly evident at the household in Sinza and on field patrols more generally. After 20 years of broadly neoliberal reform, the current-currency circuit has come under severe strain. In the 1990s, when the power sector’s largely hydropower fuel base was hit by drought, a series of expensive and in some ways unnecessary emergency power contracts locked the system into a pattern of ratcheting prices and physical deterioration (Degani Reference Degani2017a). In response to these pressures and to unpredictable rhythms of money circulation that characterize African urban life more broadly, many residents elected to indulge in what is sometimes called a “culture of nonpayment,” which includes racking up debt as well as varieties of tampering, theft, and other non-technical losses.Footnote 3 In turn, TANESCO has stepped up its inspections and disconnection campaigns, a squeeze which further disincentivizes consumers to play by the rules. A discourse of mutual recrimination prevails, with state officials castigating non-paying citizens as irresponsible “parasites” (wanyonaji) and citizens countering that the fault lies in the venality and greediness of TANESCO, and by extension the government, itself. Against this backdrop of a relation potentially spiraling out of control, an interesting ethnographic question is how consumer and utility find ways to accommodate these strains in ways that preserve their collective interdependence. That is, how do they find ways to still be in relation?

Modality, Strain, Inference

Answering this question requires shifting the focus away from the ritual work of setting up grounds and toward what happens when the semiotic processes that take place upon them are set into motion. When TANESCO provides current for the month (qua sign/input/origin), the consumer’s return of currency (qua interpretant/output/destination) is also simultaneously a sign that confirms to TANESCO that they share the same channel/code/convention, and this confirmation of shared grounds in turn licenses another round of current provision. This is the intrinsic circularity of “phatic commons,” whereby the object of exchanging (the thing the sign-interpretant refers to, its purpose or objective) is nothing less than the grounds of the exchange itself. And while one could imagine a perfectly even, balanced circuit, it wobbles with variation as responses to contingencies accumulate.

This wobbling owes firstly to the fact that, a sign usually has more than one interpretant that would be appropriate and efficacious in the context of the object. For example, as far as TANESCO is concerned, a household could pay its entire balance at the beginning of the payment period, or it could pay in installments across the pay period as the money becomes available; it could pay via a mobile finance app or via cash at the office. These permutations are why Kockelman defines the object as a “correspondence preserving projection” out from all the interpretants of a given sign (Reference Kockelman2017, 131). Shared grounds are an actor’s projection of a kind of virtual possibility space, or mode, that delimits a range of possible sign-interpretant correspondences. Akin to a style, kind, nature, or sensibility, it exists insofar as it “preserves” those correspondences but is not itself a correspondence (Kockelman Reference Kockelman2017, 125–31).

However, developing a sense of shared grounds often comes through improvised live testing. Modal auxiliary verbs like may and must, can, or should give a sense of this inferentially emergent process, expressing the socially normative requirements (i.e., deontic modality) and physically causative requirements (i.e., epistemic modality) that would make an interpretant—and thus retroactively its sign—count in relation to some object, or grounds (Kockelman Reference Kockelman2007b, 152). For example, a customer may infer that she should probably play it safe with her local utility office by saying please and thank you to the clerk even if she is frustrated with their service or by paying her bills on time even if money is tight. But she could act short with the clerk or elect to short them for the month, though perhaps at the cost of the clerk being less sympathetic to her in future encounters, or at the cost of future strains on her ability to pay. Each possible choice has consequences, and those consequences expand or contract the conditions for the next choice. Like a top spinning on a table, this self-propelling cycle explores its own conditions of possibility, perhaps meandering around the middle or spinning out dangerously to the threshold.

In particular, the different possible interactional dynamics that consumer and utility might assume are so many attempts to distribute the strain generated by the inherent contradictions of coupling present and future, self and other, short-term advantage and long-term sustainability. These are central themes in the anthropology of exchange, spanning Durkheim’s duplex theory of social life to Bloch and Parry’s account of dynamically interdependent transactional orders, perhaps best exemplified in what Claude Levi-Strauss (Reference Lévi-Strauss1969, 496) called the “law of exchange”—the fact that one cannot “gain without losing [or] enjoy without sharing.” To what degree may and must one live it up, so to speak, in the moment at the expense of others who sustain us into the future, and to what degree may or must one commit to others at the expense of our present? Any turn of a reciprocal exchange is a hypothesis. It is both interpretant and sign simultaneously, and thus inherently unstable: is it more oriented to the needs of the present (qua interpretant of a prior sign) or more to the needs of the future (qua sign committed to a subsequent interpretant)?

Finding a self-sustaining balance (or the upper hand, or an escape route) thus requires a certain attentiveness to the other to whom one is coupled (and a certain attentiveness to the other’s attentiveness, or lack thereof). To use Kockelman’s (Reference Kockelman2017, 158–60) cybernetic/ecological/informatic idiom, the coupling forms an “envorganism.” Each agent constitutes the other’s “environment” that it must learn, and in some sense internalize, in order to navigate. Whether it is algorithms coupled to users, protestors coupled to police, or humans coupled to companion species, these envorganismic spirals of reciprocal adjustment can be chaotic and stochastic, or they can lock into an escalating feedback loop. But they can also achieve a kind of patterned stability if the agents are sufficiently attuned (or felicitously mis-attuned) to each other. Such reflexive attention to being in relation is the hallmark not only of symbolic interactionism and its Theory of Mind (Bloch Reference Bloch2015); it also seems to inform the perspectivism required of “living with others” more generally (Haraway Reference Haraway2008; Hénaff and Doran Reference Hénaff2013; Kohn Reference Kohn2013). In the next two sections of this article, I explore this adaptive process as it plays out in questions of debt in a postcolonial metropolis. Officially, inspectors were obligated to cut off households for any amount of debt from the previous month. And yet in practice they seemed to have a tacit understanding that at least some debt was appropriate or efficacious insofar as it sustained the circuit in the medium- to long-term. If ongoingness is a kind of destination that both consumer and utility must continuously arrive at, this constituted a new way to reach it.

And yet other sequences of exchanges reached thresholds beyond what inspectors could tolerate. One of the most striking things about these situations was how consumers and inspectors griped—often to each other—that they could not, evidently, see themselves from the other’s perspective. As Sabina’s encounter with the household in Sinza suggests, this miscoordination could lead to conflict and violence. Still, as I explore in the final two sections, this was hardly the predetermined outcome. Responses to outstanding debt in fact began by exploiting the affordances of the service line to contrive a graded, temporary disconnection, one that offered the possibility of restoring their original dynamic or at least forestalled the possibility of it deteriorating further. Such soft disconnections shifted household and utility into a kind of reflexively liminal, phatic suspension, in which what is being offered is nothing more and nothing less than the future possibility of exchanging itself. This was a different, more conflictual mode of consumer-utility interaction, but it was also the continuation of a give-and-take within an expanded sense of ongoingness.

Urban Grounds

With its sunlit glass windowpanes and vast open ceiling, the Kinondoni North branch office of TANESCO provides a spacious, climate-controlled alternative to the equally bright, but hot and dusty, world outside. As part of a broader traineeship that cycled me through various TANESCO departments and field crews, I spent about three weeks working at the branch office’s billing kiosk printing out and handing bills to residents whose households had conventional, that is, postpaid meters (see fig. 1). After snagging their bills, they headed laterally to a long, snaking line up to the payment window. The flow of people from billing kiosk to payment window mirrors the current-currency circuit: the amount of power consumed is the input, and the amount paid is the output. Ideally, there is correspondence: one uses Tsh. 50,000 worth of current for private consumption or economic development (and not, for instance, as a commodity for resale) at a D1 tariff, and one pays that much to TANESCO (and not, say, to the touts and scammers that hung out outside). But in any given month, hundreds of residents never showed up to the office. Bills were never printed and payments never rendered. In such cases a journey might look more like Tsh. 50,000 consumed and zero paid. Does this sign-interpretant correspondence still make sense in relation to the object, that is, to the shared grounds between TANESCO and household?

Figure 1. The author hands out electricity bills at a TANESCO branch office

From TANESCO’s perspective, such an input-output is a strained one indeed. As both workers and managers repeated ad infinitum, TANESCO cannot continue to provide service if not enough money comes in. To be sure, the reality of TANESCO’s finances owes more to a mixture of structural problems and poor contracts at the generation level, and it is routinely bailed out with government and donor support (Degani Reference Degani2017a). Nevertheless, there is a kind of moral imagination that bristles at this so-called culture of nonpayment among residents, rooted in the colonial construction of urban-coastal “Swahili” people as indolent spender-types who pursue superfluous luxuries and leisure (Lewinson Reference Lewinson2006). Inheriting these ideas in the 1960s and 70s, the newly independent socialist state sought to discipline its populace, demanding austerity and hard work in exchange for collective development. Such ideas persisted, if transfigured, into the postsocialist 1990s, when, as one TANESCO manager told me, urban residents suffered from a socialist hangover where they expected everything yet gave little in return (“even a thief could go to the hospital and get treated for free!”).

In 2002, under pressure by the World Bank and other donors to implement market discipline in the power sector, TANESCO signed a private management contract with South African company NetGroup Solutions. As Rebecca Ghanadan (Reference Ghandan2009) describes, NetGroup raised tariffs for residential and light commercial consumers by 39 percent while lowering industrial tariffs by as much as 28 percent. The residential “lifeline” subsidy for poor households was cut from 100 to 50 kilowatt-hours per month, and overall residential electricity bills tripled by 2005. Small-scale residential and commercial consumers bore the brunt of an aggressive disconnection campaign, while middle- and upper-class domestic consumers were serviced with sophisticated prepaid metering systems, also imported from South Africa. Overall, this was a new neoliberal mode of consumer-utility relation, an imposition of new grounds regulating what was possible and permissible.

And yet this new mode was itself in many respects infelicitous. While it increased revenue in the short term, doubling collection rates in the first two years of the contract, it also threatened to undermine the long-term legitimacy of the utility as a public good. Though the public appreciated forcing high-volume consumers like government ministries and institutions to pay their bills, they resented the toll on Dar es Salaam households. Indeed, one of the more odious things NetGroup did was to establish 2 percent interest on unpaid debts, thereby creating a potentially runaway feedback loop in which consumers could never get out of the red. TANESCO’s subsequent interpretants of debt (disconnections) in turn created an overwhelming demand to ease the strains of austerity, and the period saw an efflorescence of so-called vishoka (hatchets; singular, kishoka). In sprawling, largely informal uswahilini (Swahili) neighborhoods, street electricians would surreptitiously reconnect residents to the grid after the TANESCO DC teams rolled through. Outside branch offices, touts would offer to help resolve people’s debts at a discount, sometimes absconding with the money.

In light of these difficulties, after a few years the utility changed tack again, and while much of this heightened conflict between household and utility remains, the intensity has downshifted somewhat. Though there are still substantial numbers of conventional meters, particularly in older neighborhoods nearer the built-up core of the city, all new connections are now prepaid—a technology widely preferred precisely for the way it obviates debt build-up and reduces the fines, delays, and general transaction costs of reconnection. Moreover, while tariffs have generally continued to rise, NetGroup’s contract was canceled in 2006, in part due to what was seen as its excessively mercenary focus on debt recovery and revenue collection. In the year following, TANESCO introduced three- and six-month schedules in which households were permitted to pay down their debt along with that month’s bill while retaining power; if the household could pay off the entire debt, the interest would be dropped. This policy could be combined with a meter upgrade policy in which a household could obtain a prepaid meter on the condition that it continue to pay down the debt on a given schedule.

In sum, these policy changes constituted shifts in the codes, channels, and conventions that link consumer and utility. Such shifted grounds created new strains, which provoked readjustments in turn. This cybernetic back and forth highlights the “unstable and history-riven” (Keane Reference Keane2018, 71) nature of any relation, the ways that sign-interpretant sequences always threaten to exceed the object meant to enclose or channel them, prompting agents to revise their understanding of (and thus performatively reconstitute) the object in turn. As I detail in the next section, such excesses and adaptive revisions take place not only at the level of the utility and consumer population, but also at the branch level of utility personnel and individual households. Indeed, because inspectors and clerks formed an important interface with consumers (i.e., they were often literally face-to-face), they were continuously confronted with and forced to accommodate feedback.

Debating Debt

While “infrastructure” may conjure up visions of material substrates, ethnographers have highlighted the ways such systems are always peopled (Star Reference Star1999). Laborers in particular are a vital part of waste management, power provision, and other postcolonial public services. Their activitiy routines are just as important as capital inputs such as meters, databases, or trucks, and indeed often compensate for the latter’s deterioration. And yet, as in many postcolonial situations, TANESCO civil servants and workers I spoke with were by their own estimation overworked and underpaid. This can be traced back to the economic crises of the 1980s, which saw real wages in Dar es Salaam plummet and a black market for government-controlled goods flourish. The government helped stabilize this condition by liberalizing cheap commodity imports and easing restrictions on private wealth. The latter in effect licensed salaried workers (most of whom were employed by the state) to “pay themselves,” whether by moonlighting in private business, or, more tacitly, through forms of entrepreneurial pilfering or extortion (Tripp Reference Tripp1997). This was exacerbated in the 2000s when NetGroup negotiated a voluntary retrenchment of the 1,753 TANESCO workers, or 21 percent of the staff (Ghanadan 2009, 413). The diminishing institutional thickness of infrastructural labor thus nourishes what Achille Mbembe (Reference Mbembe2001, 128–29) calls a postcolonial “conviviality,” wherein the relation between state and society is one of collaboration and complicity interwoven with resistance and oppression. This is particularly true for employees of parastatals like TANESCO, as opposed to the army or security forces. Inspectors are, as one once described to me, the “police wing” of the utility. But, after I spent weeks wandering around with them in various informal settlements, it was clear to me that they are hardly able to unilaterally enforce utility policy, nor are they necessarily convinced that they must always do so according to the terms laid out by upper management.

Consider debt enforcement. The English versions of TANESCO’s customer service charter and distribution engineering instruction manual refer to customers’ responsibilities for “outstanding debts” and TANESCO’s rights to discontinue service until the debt is paid. The Kiswahili version of the customer service charter refers merely to debt in general (deni). What constitutes an (outstanding) debt? Many residents I spoke with assumed TANESCO has always reserved the right to cut off customers for any unpaid amount. A few recalled a policy of forwarding unpaid debt to the next month so long as it did not exceed 50,000 shillings, while others could remember no such thing. There seems to have been significant variation in policy, both formal and informal, from branch to branch and region to region. In 1992, for example, the World Bank commissioned a Power Loss Reduction Study for Tanzania, that, much like my own fieldwork, included interviews and participant-observation among various sectors of the utility. The authors wrote that “the head office in Dar es Salaam stated that when each new bill is issued, disconnection orders are issued for accounts that are one month’s payment in arrears. The Morogoro office reported that disconnections are undertaken on accounts that are three months in arrears. The observed facts did not support either assertion. Many cases were found where active accounts were much more than three months in arrears” (ESMAP 1998, 6). Along these same lines, during our first or second day on field patrols, Thierry suddenly recalled a story from his childhood in the 1990s, when he was living with his sister’s family in the northern city of Arusha. One afternoon, TANESCO inspectors arrived at their household and disconnected their meter for around Tsh. 150,000 of unpaid debts. At the branch office a few day later, the clerk “advised” his brother-in-law that he could be reconnected if he just brought the balance down below Tsh. 100,000. Very roughly speaking, then, we might say that debt is assessed in intervals of 50,000 shillings, which is equal to about one to two months’ worth of average consumption, with anything beyond that threshold probably considered outstanding.

Patrols corroborated this impression. In our own experience, Thierry and I found that inspectors did not concern themselves with debts below Tsh. 100,000. Each day they were given a billowing dot-matrix printout sheet with reams of addresses listing outstanding debts. Inspectors of a given branch could never hope to actually complete the disconnection sheet, which spanned thousands of houses. On a particularly hard-driving day, a team might work from 9 a.m. to 5 p.m. in the afternoon, often under the hot sun, and hit 70 houses. More likely it was 9 a.m. to 2 p.m. and an average of 40 hofuses. The houses they did hit tended to have debts of anywhere from Tsh. 200,000 to 1,000,000 or more, with special days reserved at the end of the month for targeted cutoffs of truly outstanding debts. Small fry were left to keep swimming.

Under-equipped inspectors prioritizing high-level debt makes sense from a cost-recovery perspective, of course. If one has only a limited number of resources, it is best to allocate them in a way that maximizes returns. But to some degree, what is at stake with such thresholds is the grounds of the household-utility relation—the broader context of postcolonial strain, in which some small amount of non-payment is in fact an appropriate and efficacious interpretant of state provision.

Consider an exchange between Thierry and Sabina during rounds, after they had just disconnected a household with a debt of around Tsh. 200,000—that is, at the lower end of the disconnectable debt range. Musing on the rather quiet demeanor of the woman at the house, Sabina laughed with satisfaction and said, “that’s the right, because the bosses [i.e., Thierry and myself] are around.” This was certainly untrue in a literal sense—Thierry and I were officially “trainees,” and thus regarded as gophers and ladder-haulers. However, my presence as a white man, alongside Thierry, whose collared shirt and pressed slacks contrasted with everyone’s green TANESCO shirts and coats, was certainly unusual and meant I was probably coded as a boss (i.e., a donor, a consultant, a “loss reduction specialist,” etc.). Thierry, following up on this, asked: “So do you think that mama that we’ve just cut off—if [Author] wasn’t here what kind of words do you think she’d use?” The exchange continued:

Sabina: Kashfa [slurs]. “You’re all thieves, you just want bribes, you’re all hungry” and whatnot—all sorts of words.

Thierry: [musing] But her debt wasn’t that big you know? Have there ever been customers that [in the past] you’ve decided to cut a break [i.e., not disconnect] but it turns out they don’t pay and the debt continues to build?

Sabina: Exactly. And [when you do disconnect a household] those with small debts they slur you worse than those with big debts!

Thierry: [chuckling] But why’s that?

Sabina: He sees that he just owes a little bit while others owe much more!

Sabina’s exchange with Thierry highlights three important points about the relatively accommodating stance of inspectors out in the field. First, despite my unusual presence on patrols, I still caught wind of such kashfa—in part, perhaps, because many residents were not aware that I understood Kiswahili (see, e.g., Degani Reference Degani2018). In conversation, inspectors relayed war stories of having rocks thrown or even, in one case, a gun pulled on them. Second, as Sabina passingly acknowledged, some customers were indeed given reprieves. At one house, a woman insisted she had money for payment but could not get to the office that day because she was caring for someone elderly and sick, while in another, a man deferentially apologized and promised he would go pay in the morning. In both cases, the inspectors simply moved on.

The third and most important point was Sabina’s observation that customers who tend to get angrier about disconnection are those with lower debts. Indeed, wealthy households with millions of shillings of debt tended to be sheepish, solicitous, or even playful (Degani Reference Degani2018), rather than angry per se. On the other hand, when small fry were caught and cut for a measly few hundred thousand shillings—debts that were large enough to put them months in arrears but minor compared to the outré levels of debt that could be found in other households—the scene could be tense.

Consider another exchange, this time between an inspector named Mary and two elder “mamas” who were renting in a household that had a listed debt of Tsh. 130,000. As Mary’s colleague went about disconnecting their meter, the mamas protested angrily, explaining the situation. The previous owners had racked up a large debt and never paid it off before selling the house. Because debt is technically tied to the meter, not any particular person, the new tenants had unwittingly inherited it. The new tenants successfully applied for a new prepaid meter on the condition that they would pay down the old debt on installment, which is exactly what they had been doing. Mary replied to all of this explanation with a mixture of sympathy and restraint:

Mary: If you were under Tsh. 100,000, we would not disconnect you, but you are under schedule, sisters.

Mama 1: What is the amount, my sister (ndugu yangu)?

Mary: You owe Tsh. 133,000.

Mama 2: Do you know what the debt used to be? 500,000?

Mama 1: Something like Tsh. 600,000—here is the bill.

Mary: Yes indeed.

Mama 2: Not all in the house have, we are the ones that paid.

Mary: Listen, nothing has been destroyed, go and reduce it a little, at least.

Mama 2: Just before yesterday we reduced it! It’s not like we pick money on the ground, we have to search for it!

Mama 1: Everyone knows how difficult money is. The owners of the house have died and left this debt. But we renters who have heart have decided to pay it so we can live in peace. Do you think anyone else would have paid it? We are planning to finish the money by the end of this month but you’ve still cut off this electricity. It’s not right (sio haki).

Mary: Yes, we often hear about such issues about renters. “The other renters have left a debt,” or “the previous renter has left a debt and you have just moved in.” [But] these are things we cannot know. We go by the debt of the household.

Here we see an illustration of Sabina’s observation that households with relatively small debts tend to curse and argue as well as the underlying threshold principle at work. If the residents had been under Tsh. 100,000, they would have been “on schedule” with their repayments and would have thus avoided disconnection—though it still remains unclear to me whether that was an official threshold pegged to a formal repayment schedule set up at the office, or whether Mary was suggesting an informal reprieve. It is also unclear to me just what Mary meant by urging them to reduce the debt “a little, at least.” Would simply getting under 100,000 shillings qualify them for reconnection for the month? Was this simply something she said to placate them?

In any case, such arguments were moments when the shared grounds linking consumer and utility were reflexively figured and, however inchoately, debated. The mamas’ protestations point to the fact that “Swahili”-style households often feature multiple renters with irregular income flows who share a single meter, move in and out, and have contentious or ill-defined relationships with landlords. Along with the dangers of a postpaid meter that allows a household to consume on credit, the strain of such a hard life (maisha magumu) means that some amount of debt is basically inevitable. For the two mamas, getting disconnected for a mere Tsh. 33,000 over 100,000—that is, for about one interval above the threshold—highlighted Mary’s perhaps willful misunderstanding of the strained basis on which renters were making payments in the first place. By contrast, Mary insists (with the help of a modal auxiliary verb, notably) that these broader grounds are something that TANESCO “cannot know” and thus cannot consider. That is, inspectors are permitted to see a given amount of debt only as an (inappropriate) interpretant of the current provided to the household, not as an indexical trace of tenants’ broader struggles to get by.

Of course, the very presence of such debates highlights that despite inspectors’ claims about what they cannot see, and despite TANESCO management’s insistence on timely payment, consumers have pushed inspectors to expand the possibility space of debt, or rather have regimented the utility’s attempts to shrink this space. Indeed, if assessments of small debts that hovered just an interval or two above the Tsh. 100,000 range as “outstanding” were bitterly contested, we can imagine how pursuing disconnections at even lower levels would make it all the more difficult for inspectors to enter into residential neighborhoods to do their job—to share literal common ground with residents. Under-equipped and overworked, inspectors generally focused on larger debts. And when they judged it reasonable to cut households with relatively small debts, they prepared themselves for kashfa.

On one hand, one can plainly see the logic of residents’ anger. Yes, households must pay for their power; but if every household with any debt whatsoever were cut off, the current-currency circuit itself would grind to a halt. Allowing a little bit of slack by focusing on households with bigger rather than smaller debts is thus felicitous to being in relation. It signals that inspectors understand something about the households with which they are interacting and that those households can be trusted to honor the good faith and good sense of letting small debts slide by paying them down. On the other hand, it is clear that inspectors could never arrive at the right cutoff point given their diminished enforcement capabilities, because there would always be other customers getting away with far more. The only truly consistent position would be not to cut at all; short of this, the best bet was to draw a line at some level of slack and endure the anger of those households that were relatively near it, but not quite below it. Working along that delicate threshold was tricky; tempers flared as the slack tightened.

Soft Disconnections

However tacitly or situationally they defined thresholds for disconnections, inspectors inevitably arrived at households that were past them. But what does it actually mean to disconnect? We can start answering this question by observing that a service line is generally comprised of three components: the utility pole, the cable, and the meter. While utility poles are pegged to an area of consumers (anyone within thirty meters of the pole is eligible to connect to it), cables and meters are unique to individual households. In cases of extreme debt, inspectors will remove the cable and levy a schedule of fines and reconnection fees. But in most cases, disconnection means that when a DC team arrives at the house, one inspector will fill out a disconnection report in duplicate as she explains to the customer, often over their begging or arguing, that they should head to their local TANESCO branch office and pay all arrears and reconnection fees, while the other inspector yanks out a small black cutout fuse with a satisfying pop, sticks it in her bag, and moves on to the next house.

This sort of disconnection is not just a riposte within an existing household-utility relation, but rather a reflexive, metacommunicative riposte that alters the infrastructural media—and thus shifts the very grounds—of that relation. Relevant here is Kockelman’s (Reference Kockelman2017) insightful definition of media as any prosthetic extension of an agent which functions to shape quali-signs, that is, the virtual realm of all possible signs that could be sensed or instigated. Examples include “gloves, camouflage, drones, erasers, ice-picks, and sunglasses,” all of which “augment, intensify, buffer, mollify, and mitigate the sensory and instigatory capabilities of semiotic agents” (Reference Kockelman2017, 124). In the same way, a de-fused meter alters the range of quali-signs that the household can physically sense (namely the arrival of some quantity and quality of current), and thus alters the range of quali-interpretants it can subsequently instigate (namely some quantity and quality of payment).

Phrased this way, it becomes clear that a de-fused meter is a kind of buffering media. First, there are other infrastructural channels through which consumer and utility can communicate, notably via face-to-face exchanges with the field teams and office clerks. Household and utility are no longer directly touching each other, as it were, but they can still hear and see one another. Second, relatedly, removing the cutout is a reflexively soft disconnection—a buffering media that is itself relatively bufferable. As TANESCO well knows, already disgruntled households have even less incentive to pay their debts if they cannot eventually resume consuming electricity. If DC teams remove the cable or even the meter, the household must in effect reapply for the entire service line, straining TANESCO’s already backlogged application queue and, more importantly, increasing the likelihood that the debt will remain unrecovered without legal action. Popping a cutout back in is less labor intensive than reinstalling a meter or even scaling up a pole to reconnect a cable. Indeed, the denouement to Thierry’s story was that once his brother-in-law brought the balance down to Tsh. 55,000, the clerk simply dumped a box of cutout fuses onto the counter, asked for Tsh. 5,000 in majji ya kunywa (drinking water, i.e., a bribe or tip) and had him choose the one that matched their meter’s model. In other words, reinserting a fuse is so easy that residents could do it themselves.

What is interesting here is the phatic way that modifying what can or cannot happen physically (epistemic modality) also functions as a sign that communicates what could or should happen socially (deontic modality). For example, if I am arguing with my friend and I walk away before I say something I regret, so to speak, I physically prevent myself from instigating a response to her sign, implying that the problem is that we have simply drifted off course and thereby offering a reset. Likewise, removing the cutout rather than the entire cable or meter, modifies the shared grounds in a way that reflexively includes and indeed invites the possibility of reconnection. Its very softness neither repairs nor resets the phatic commons (though those may be the end results) but simply forestalls its continued degradation. And it is in this forestalling that the ongoingness between consumer and utility is preserved, if ratcheted up a level. Rather than giving and taking current and currency, they are giving and taking the possibility of giving and taking. The shift into the suspended mode is itself a sign that the household senses and whose subsequent interpretant is either a shift towards one overarching mode of relation (say, full reconnection), or toward another (de facto full disconnection). In other words, it is a kind of modest gift or last chance on the part of TANESCO. It is a grace period, a suspension, a limbo: it cuts without cutting connection.

There is an interpersonal quality to this grace as well. Sabina’s cruel threat in the opening vignette was remarkable for the way it contrasted with inspectors’ generally neutral if not appeasing comportment, exemplified in Mary’s encounter with the two mamas. Usually, inspectors tried to be gentle but firm, executing their mandate while at the same time avoiding escalation in face-to-face interactions. “You can’t fight the customer,” they told me over and over again. Properly suspending a relation involves the tricky art of doing nothing less than—but also, crucially, nothing more than—cutting where you need to.

Soft Disconnections of Soft Disconnections

Once a cutout has been removed, the utility-household relation has shifted into a kind of reflexively suspended mode. For the household in question, the possibility space of sign-interpretant sequences includes at least three options. First, they might simply decide to pay up, reapply for service as instructed (qua sign), and get reconnected (qua interpretant) and thus participate in yet another reflexively phatic sequence that commences the resumption of normal relations. As discussed earlier, this is the sequence that TANESCO implicitly anticipates and encourages via the moderated punishment of merely removing the cutout and via their patience and forbearance in verbal interaction. A soft disconnection is premised on the idea that the consumer will recognize it as a gesture of relative good faith, and act appropriately.

The second option is a soft reconnection: a household might hire a kishoka (hatchet) or take it upon themselves to surreptitiously pop another cutout fuse back into the meter. In fact, this is exactly what Thierry’s brother-in-law did back in the 1990s: that night after the inspectors left, he contrived a jerry-rigged fuse out of a bent wire and jimmied it into the receptor. This allowed the family to continue using lights and television for a few nights until he could go and settle the debt at the branch office.

A reinserted fuse can thus function as kind of stop-gap measure predicated on the idea that a household will eventually pay. Alternatively, a household might elect to run up the bill and let the unpaid debt ripen into a kind of retroactive theft. Still, as Sabina observed, such “theft” is recorded on the meter, thereby at least preserving the possibility that it might be paid back. In a sense, this tactic mirrors and redoubles the phatic logic of the original cutout removal, whose virtue is that it disconnects the flow of current and currency but not its future possibility. In a direct parallel, a reinsertion disconnects the disconnection—but not its future possibility. Though faint and indirect, it can be taken as a sign of good intentions, that the household is still willing to go on with TANESCO. Indeed, Sabina says as much, grumbling that if a household is going to illegally reconnect, they should at least do it through the meter.

Finally, the third and most aggressive option is this: the household might decide to blow off TANESCO and pay a local kishoka electrician to bypass the meter, as was the case in Sinza. There are presumably any number of reasons for this decision. As we have seen, renters may have difficult relations with landlords, or they may plan on moving out. Householders may feel aggrieved or suspect that because they have already been disconnected, inspectors will not be checking up on them anytime soon. Whatever the reason, a full bypass ruins the fragile détente and propels the household-utility relation into a fully antagonistic mode: it constitutes an evasion of the law of exchange, an unambiguous taking without giving. By bypassing the meter, the offending household reduces its connection, with its complex tissue of reflexively signaled intentions, to the utility, reducing the connection itself to a merely phenomenological set of material affordances: a series of lines and circuits that can be rewired, social dimensions be damned.

For Sabina, tellingly, this purely physical connection is fundamentally stupid. As far as she is concerned, any household that contrives a bypass is semiotically impaired: morally wrong, yes, but, more to the point, deaf and dumb to anything beyond itself, reduced to brute material secondness, in that it can sense only whether it has electricity, not how the utility will respond to its manner of having electricity. It is a kind of municipal “soul-blindness” (Kohn Reference Kohn2013, 117), a breakdown of intersubjectively coordinated action that allows two agents to become caught up in an ongoing relation—in this case, a circuit. Such naked theft cannot really be justified by maisha magumu, and thus can only result in self-liquidation; a household with a bypassed meter may live it up in the short-term, “feeling cool enjoying the nice AC,” but in so doing they scorch the grounds of any future enjoyment. The household in Sinza, moreover, had not simply bypassed the meter; it bypassed the meter after the reprieve of a soft disconnection just a few days earlier. The residents had actively rebuffed the utility’s implicit good faith, and I suspect that it was this insult in particular that propelled Sabina into such an aggressive confrontation with the two young girls.

Sabina’s assertion of stupidity has an interesting relation to children. In interviews, some handymen and contractors described to me how they got interested in electricity by sometimes surreptitiously tinkering with appliances and wires as young boys. In (but not entirely of) the household, they “saw” the flow of electricity in a relatively pre-social manner, before understanding how it is embedded within and substantiates social relations. The ugly irony of Sabina’s interaction with the two young girls is that she was punishing them for their guardians’ stupid (i.e., childish) actions. The violence of her threat lay not only in the possibility it described (the night in central booking, sexual assault) but the fact that it was directed at those who simply happened to be there at the moment, who were not necessarily a part of the household as a unit of accountability. Sabina had in effect reduced culpability to mere guilt by association and thus reduced semiosis, the meaningful relation between sign and interpretant, to purely mechanical secondness of random cause and effect with no real reason (qua object) mediating it. We might say the household had in effect been reduced to homo infrasacer, banished to a realm beyond infrastructural grounds where nothing is shared and all bets are off, with all the violence and freedom entailed therein.

Or had it? Walking away, Chilambo laughed that it was highly unlikely that they would be coming back with the police anytime soon. For that kind of assistance, the TANESCO security office would be required to pay so-called gas money for the police to come at night, when the responsible parties were likely home. Even then, as the Kinondoni branch’s head security officer later explained, the resident might be able to bribe police or the courts to run the case aground. Absent enforcement, Chilambo observed, there was “no better way to compel a man to pay his bills than to threaten his daughter.” Sickening as it was, and though she clearly took a cruel pleasure in delivering it, Sabina’s threat can also be interpreted as a kind of strategic gambit to keep the household in relation.

Conclusion

In his Malinowski Lecture, David Graeber perceptively drew out the relation between violence and stupidity. “Pretty much any other way one might try to influence another’s actions, one at least has to have some idea who they think they are, who they think you are, what they might want out of the situation, and what their aversions and proclivities are. Hit them over the head hard enough and all of this becomes irrelevant” (Reference Graeber2012, 116). Violence was an ambient presence on rounds, but even the physical act of brute disconnection (and reconnection) was graded and thus interwoven with communicative intention. Household and utility attempted to read each other and, what is more, often phatically signaled that they were offering each other the respect of trying to read each other. Thus, while the strain of being in relation could intensify, there was, ideally, an intersubjectively coordinated regulation of that intensity, a controlled upshifting or downshifting. I have tried to reconstruct this modal logic, how it attempts to strike the tricky balance of responding to the past in a way that holds open a “living future” (Kohn Reference Kohn2013, 191).

While I stand by this account, questions about my own interpretive grounds still trouble me. Could it be that my presence on rounds chastened not just residents but inspectors, who otherwise might not have been so diplomatic about cutting off households? Can one, in fact, “fight the customer”? Even more sickeningly, might my presence have emboldened Sabina to threaten those girls, inspiring her to invoke the power of a violent state with all of its white colonial overtones and histories?

I write this conclusion during an explosive summer of Black Lives Matter protests in the United States, where it is easy to see how police and protestors are testing their shared grounds—often literal grounds such as streets, plazas, and bridges, as well as the reciprocal rights and obligations of citizens and state—as they exchange move and counter-move in real time. This is an intersubjective space in which agents are not just dyadically clashing with each other but are interpreting each other in ways that reflexively propel the nature of the encounter forward and backward across the threshold of violence.

Still, maybe all this mind-reading is the wrong thing to focus on, given that no such interpretive labor surrounds the systematic imprisonment and murder of black lives. Indeed, police often seem to revel in the fact that they do not need to understand the situation during an arrest, to respond to gasps for oxygen or hands raised high in submission as signs of a shared relation.

Similarly, my fieldwork in Tanzania coincided with the start of an increasingly repressive era of the Tanzanian de facto single-party state, which has simply jailed, beaten, or murdered people who have spoken up too loudly, who have not offered up acceptable and efficacious interpretants of its signs (Paget Reference Paget2017). Against this background, what does it mean to observe that, despite their militaristic “patrols” and occasional alliances with security forces or police, inspectors were mostly in the business of cajoling, buffering, and above all communicating with households? Perhaps it simply means: do not arm social workers. It is only when the media of violence are not cheap and plentiful that actors can sustain the complex work of being in relation.

Whereas semiosis is often thought of as a kind of infrastructural bridge that relates sign to interpretant (or object to sign), I hope to have shown in this article that public infrastructures are themselves thoroughly semiotic. Critical scholars of infrastructure have arguably been fascinated by secondness—the fact that all infrastructures are hosts to parasitic breakdowns and lines of flight that resist pretensions to mastery. But there is a danger here. To reduce the grid to the record of its actual material flows of current and currency (in all their parasitic variation) is to reduce an object to its signs and interpretants (and reduce a third to a second). Like all semiotic processes, infrastructures are relations. Each possesses an animating logic that governs modal distinctions regarding which variations it will stand for, insofar as those variations stand for it. Grasping such distinctions is especially important under precarious or inauspicious conditions like those endemic to urban African life. In Dar es Salaam, consumer and utility continuously interpret each other’s signs, judging what sorts of prices they might pay, what sorts of debt they might accrue, and what sorts of modifications they might tolerate. It is this modest work of interpretation that allows them to be in relation and, if not without strain, to keep electricity circulating.

Footnotes

This article is based on fieldwork generously funded by the Wenner-Gren Foundation, the National Science Foundation, and the Social Science Research Council. I wish to thank Anna Weichselbraun, Hannah Chazin, Zoe Crossland, Paul Kockelman, and the rest of the participants of the “Materiality, Meaning, Agency & Value” panel at the 2019 Society of the Social Studies of Science meetings, where I presented an early version of this article and received incredibly useful feedback. I am grateful to two reviewers for their thoughtful reads and improvements, as well as to Asif Agha for his critical insights and deft editorship. Finally, thanks to “Thierry” for his grace, good humor, and friendship.

1. At the time of my fieldwork, the exchange rate was roughly $1/Tsh. 1,600.

2. An important exception would be Nancy Munn (Reference Munn1992); see also Lisa Björkman (Reference Björkman2014) and Marcel Hénaff (Reference Hénaff2013) for Peircean analyses of the triadic logic of Maussian gifts and Levi-Straussian reciprocity, respectively.

3. In 2010, a field survey (Azorom and AETS Reference Azorom2011) estimated that that TANESCO had suffered system losses of 24.4 percent. This combines “technical” losses occurred in the transmission and distribution (TD) process (12.8 percent) and commercial losses, i.e., billing or meter irregularities (11.6 percent). Starting around 2004, commercial losses have been hovering around 12 percent, a doubling of their previous 10-year average of 5–6 percent (Vagliasindi and Besant-Jones Reference Vagliasindi2013, 350).

References

Althusser, Louis. 2006. “Ideology and Ideological State Apparatuses (Notes towards an Investigation).” In The Anthropology of the State: A Reader, edited by Aradhana Sharma and Akhil Gupta, 86–111. Malden, MA: Blackwell.Google Scholar
Azorom, AETS (Application Européenne de Technologies et de Services). 2011. “Electricity Loss Reduction Study—TANESCO and ZECO: Final Report Executive Summary.” Ministry of Finance Millennium Challenge Account Tanzania. AETS, Lons. https://www.aets-consultants.com/projects/energy/electricity-loss-reduction-study-tanesco-zeco.Google Scholar
Björkman, Lisa. 2014. “‘You Can’t Buy a Vote’: Meanings of Money in a Mumbai Election.American Ethnologist 41 (4): 617–34.10.1111/amet.12101CrossRefGoogle Scholar
Bloch, Maurice. 2015. “Durkheimian Anthropology and Religion: Going In and Out of Each Other’s Bodies.HAU: Journal of Ethnographic Theory 5 (3): 285–99.10.14318/hau5.3.019CrossRefGoogle Scholar
Bohannan, Paul. 1955. “Some Principles of Exchange and Investment among the Tiv.American Anthropologist 57 (1): 60–70.10.1525/aa.1955.57.1.02a00080CrossRefGoogle Scholar
Degani, Michael. 2017a. “La véranda, le climatiseur et la centrale électrique: Race et électricité en Tanzanie postsocialiste.” Afrique contemporaine 261–62 (1): 103–18.Google Scholar
Degani, Michael. 2017b. “Modal Reasoning in Dar es Salaam’s Power Network.American Ethnologist 44 (2): 300–314.10.1111/amet.12480CrossRefGoogle Scholar
Degani, Michael. 2018. “Shock Humor: Zaniness and the Freedom of Permanent Improvisation in Urban Tanzania.Cultural Anthropology 33 (3): 473–98.10.14506/ca33.3.08CrossRefGoogle Scholar
Elyachar, Julia. 2010. “Phatic Labor, Infrastructure, and the Question of Empowerment in Cairo.American Ethnologist 37 (3): 452–64.10.1111/j.1548-1425.2010.01265.xCrossRefGoogle Scholar
ESMAP (Energy Sector Management Assistance Programme). 1998. “Tanzania—Power Loss Reduction Study (Vol. 2): Transmission and Distribution System—Reduction of Non-Technical Losses.” Report no. 204B/98, World Bank, Washington, DC. http://documents1.worldbank.org/curated/en/240121468778464452/pdf/multi-page.pdf.Google Scholar
Foucault, Michel. 1983. “The Subject and Power.” In Michel Foucault: Beyond Structuralism and Hermeneutics, edited by Hubert L. Dreyfus and Paul Rabinow, 208–26. Chicago: University of Chicago Press.Google Scholar
Fredericks, Rosalind. 2018. Garbage Citizenship: Vital Infrastructures of Labor in Dakar, Senegal. Durham, NC: Duke University Press.10.1215/9781478002505CrossRefGoogle Scholar
Ghandan, Rebecca. 2009. “Connected Geographies and Struggles over Access: Electricity Commercialization in Tanzania.” In Electric Capitalism: Recolonizing Africa on the Power Grid, edited by David A. McDonald, 400–436. London: Earthscan.Google Scholar
Graeber, David. 2012. “Dead Zones of the Imagination: On Violence, Bureaucracy and Interpretive Labor.HAU: Journal of Ethnographic Theory 2 (2): 105–28.10.14318/hau2.2.007CrossRefGoogle Scholar
Haraway, Donna. 2008. When Species Meet. Minneapolis: University of Minnesota Press.Google Scholar
Hart, Keith. 2001. Money in an Unequal World: Keith Hart and His Memory Bank. London: Textere.10.1177/14634990122228755CrossRefGoogle Scholar
Hénaff, Marcel, and Robert Doran. 2013. “Living with Others: Reciprocity and Alterity in Lévi-Strauss.Yale French Studies 123:63–82.Google Scholar
Jackson, Steven J. 2014. “Rethinking Repair.” In Media Technologies, edited by Tarleton Gillespie, Pablo J. Boczkowski, and Kirsten A. Foot, 221–40. Cambridge, MA: MIT Press.10.7551/mitpress/9042.003.0015CrossRefGoogle Scholar
Keane, Webb. 2018. “On Semiotic Ideology.Signs and Society 6 (1): 64–87.10.1086/695387CrossRefGoogle Scholar
Kockelman, Paul. 2007. “From Status to Contract Revisited Value, Temporality, Circulation and Subjectivity.Anthropological Theory 7 (2): 151–76.10.1177/1463499607077296CrossRefGoogle Scholar
Kockelman, Paul. 2017. The Art of Interpretation in the Age of Computation. New York: Oxford University Press.10.1093/acprof:oso/9780190636531.001.0001CrossRefGoogle Scholar
Kohn, Eduardo. 2013. How Forests Think: Toward an Anthropology beyond the Human. Berkeley: University of California Press.Google Scholar
Lacan, Jacques. 2007. Écrits: The First Complete Edition in English, translated by Bruce Fink. New York: Norton.Google Scholar
Lévi-Strauss, Claude. 1969. The Elementary Structures of Kinship. Boston: Beacon.Google Scholar
Lewinson, Anne S. 2006. “Domestic Realms, Social Bonds, and Class: Ideologies and Indigenizing Modernity in Dar Es Salaam, Tanzania.Canadian Journal of African Studies/La Revue Canadienne des Études Africaines 40 (3): 462–95.10.1080/00083968.2006.10751402CrossRefGoogle Scholar
Malinowski, Bronislaw. (1922) 2002. Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea. London: Routledge.10.4324/9780203421260CrossRefGoogle Scholar
Mann, Michael. 2008. “Infrastructural Power Revisited.Studies in Comparative International Development (SCID) 43 (3): 355–65.10.1007/s12116-008-9027-7CrossRefGoogle Scholar
Mauss, Marcel. (1979) 2004. Seasonal Variations of the Eskimo: A Study in Social Morphology. Translated by Henri Beuchet. Boston: Routledge.Google Scholar
Mauss, Marcel. (1925) 2016. The Gift: Expanded Edition. Translated by Jane I. Guyer. Chicago: HAU Books.Google Scholar
Mbembe, Achille. 2001. On the Postcolony. Berkeley: University of California Press.Google Scholar
Munn, Nancy D. 1992. The Fame of Gawa: A Symbolic Study of Value Transformation in a Massim (Papua New Guinea) Society. Durham, NC: Duke University Press.Google Scholar
Paget, Dan. 2017. “Tanzania: Shrinking Space and Opposition Protest.Journal of Democracy 28 (3): 153–67.10.1353/jod.2017.0053CrossRefGoogle Scholar
Parry, Jonathan, and Maurice Bloch. 1989. Money and the Morality of Exchange. Cambridge: Cambridge University Press.10.1017/CBO9780511621659CrossRefGoogle Scholar
Sahlins, Marshall David. 1972. Stone Age Economics. Chicago: Aldine Atherton.Google Scholar
Serres, Michel, and Lawrence R. Schehr. 1982. The Parasite. Baltimore: Johns Hopkins University Press.Google Scholar
Simpson, Audra. 2014. Mohawk Interruptus: Political Life across the Borders of Settler States. Durham, NC: Duke University Press.Google Scholar
Singh, Bhrigupati. 2014. “How Concepts Make the World Look Different: Affirmative and Negative Genealogies of Thought.” In The Ground Between: Anthropologists Engage Philosophy, edited by Veena Das, Michael D. Jackson, Arthur Kleinman, and Bhrigupati Singh, 159–86. Durham, NC: Duke University Press.Google Scholar
Star, Susan Leigh. 1999. “The Ethnography of Infrastructure.American Behavioral Scientist 43 (3): 377–91.10.1177/00027649921955326CrossRefGoogle Scholar
Tripp, Aili Marie. 1997. Changing the Rules: The Politics of Liberalization and the Urban Informal Economy in Tanzania. Berkeley: University of California Press.10.1525/9780520327436CrossRefGoogle Scholar
Vagliasindi, Maria, and John Besant-Jones. 2013. Power Market Structure: Revisiting Policy Options. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/13115.10.1596/978-0-8213-9556-1CrossRefGoogle Scholar
Viveiros de Castro, Eduardo. 2014. Cannibal Metaphysics. Minneapolis: University of Minnesota Press.Google Scholar
Weichselbraun, Anna. 2019. “Of Broken Seals and Broken Promises: Attributing Intention at the IAEA.Cultural Anthropology 34 (4): 503–28.10.14506/ca34.4.02CrossRefGoogle Scholar
Winther, Tanja. 2008. The Impact of Electricity: Development, Desires and Dilemmas. New York: Berghahn Books.10.3167/9781845454951CrossRefGoogle Scholar
Figure 0

Figure 1. The author hands out electricity bills at a TANESCO branch office