The Lives of Data: Essays on Computational Cultures from India
Suvir Kaul · Critical AI · 2024
The internet age has brought with it a tsunami of numbers, or more accurately, of data. Slowly but certainly we are being transformed into citizens whose lives are dictated by their ratified—“authenticated”—relations with the state, all mediated by computational processes stored on giant servers. We are not just consumers whose purchasing power and consumption patterns are tracked but, increasingly, consumers whose choices are shaped by marketers who decide what we should see, hear, and desire. Our very aspirations are crafted for us by social media: the flags we mobilize under, the collectivities we claim as our own, the antipathies that fuel our political lives. In The Lives of Data: Essays on Computational Cultures from India, Sandeep Mertia curates a collection of essays that explore the “cultivation of relationalities of data” in India, for, as he writes, the stakes are high when it is data that “inscribes and mediates political power and socio-cultural difference” (9). The essays collected here discuss the infrastructure of data and the processes that sort information. As Mertia reminds us, the life “of any kind of data is shaped by actual and potential relations with other existing data, classifications, paper and digital infrastructure, statistical techniques, data collection and cleaning practices, and possibilities of circulation” (10).Mertia's introduction, “Relationalities Abound,” provides a useful example (the new Aadhar card regime in India) to explain to the uninitiated this new world of data collection and management: Perhaps the biggest irony about big data is that it has little to do with data per se. Rather, it has a lot to do with classifications, connections, and patterns that emerge or can be generated when large-scale, high-dimensional, real-time, and variably un/structured data are mashed up with other data. Individual Aadhaar card, credit card, internet history, or any other machine-readable data in itself does not mean much. It has to be positioned within the relationalities and infrastructures that demonstrate for example, the uniqueness of a fingerprint relative to the biometric data of 1.3 billion others, or the classification of one's online purchases in a cluster of other users who might be interested in buying a ‘related’ product. (14–15)And of course, this new world has its agents and their priorities: those who manage vast platforms for profit; data brokers who “market government-owned as well as private telecom and financial data”; policy makers who decide how data is to be sequestered, sold, or used in the surveillance of citizens, and others. Mertia is fully aware of the use of digital technologies to do harm or to exploit. The volume is an effort to “rethink the status quo, for a potentially decent, sustainable, and open-ended sociality, alterity, and generativity in computational cultures” (22).There are fourteen essays in this volume covering a wide range of topics germane to the life of data in India. Some of them revel in the theoretical and methodological questions raised by critical analysis of the infrastructures and data collection models at work today, others ground such analyses in examples from the field as platforms (supposedly developed to ease citizen-state relations) end up victimizing vulnerable populations. Not surprisingly, the Aadhar card, which aims to enroll all citizens of India, receives repeated attention. For example, Ranjit Singh, in “Study the Imbrication: A Methodological Maxim to Follow the Multiple Lives of Data,” quotes Annemarie Mol to warn us that “the point of asking what is being counted” by the Aadhar database “is not to argue that counting is doomed to do injustice to the complexity of life. This is certain. The point, instead, is to discover how and in what ways” (Mol 2002: 235; qtd. on 52). Singh shows how the identity anchored by the Aadhar card is imbricated with government social welfare schemes, which it enables and extends (ironically, this is true both when it works and when it does not).In “Reporting the World's Largest Biometric Project,” Anumeha Yadav, whose newspaper reports have illustrated the multiple challenges posed by the imbrication of Aadhar and welfare schemes, describes the problems faced by agents contracted to pay rural workers using their biometrics and a micro-ATM. In one instance, the machine did not recognize the fingertips of four of seven workers because their fingertips had been rubbed raw by manual work. They were asked to revive their fingerprints with skin-softening creams, flour, and lemon juice! Moreover, while enrollment in Aadhar was meant to be voluntary, those who did not enroll were denied essential services. Yadav offers a categoric conclusion: the government's “major claims on welfare delivery and efficiency had not come true across schemes. In Rajasthan, Jharkhand, Chhattisgarh, and Gujarat, there was little evidence that centralized biometrics schemes served local needs” (87). Preeti Mudaliar, in “Broken Data: Repairs in the Production of Biometric Bodies,” confirms Yadav's findings as she examines the feelings of alienation and fear when individuals are denied access to biometric identification. Those “experiencing a sense of exclusion” feel orphaned not only from this new digital infrastructure but also alienated “from a larger scheme of collective belonging” (95).The volume also discusses the increasing evidence for problems with field surveys and data derived from them. Prerna Mukharya and Mahima Taneja, in “Outline India: Field Notes on Data Practices and Innovations,” describe their experiences in conducting data collection surveys, and of the importance of predicting and collating “potential and actual errors that arise because of comprehension and cognitive, regional, linguistic, and contextual variations.” Commenting on the “challenges of using technology such as tablet assisted personal interview platforms and UAVs,” they argue that “while such technologies aid in expediting and monitoring data collection, maintaining quality, and adding a layer of geospatial information to assist in evidence-based policy-making,” they face the risk of “deepening the subject–enumerator gap and losing qualitative insights” (106). In a salutary reminder of all that is at stake as the development sector bases its programs on data, Guneet Narula, in “Collecting Open Data: Data Practices, Tools, Limitations, and Politics,” reminds us that these data bases are far from open and models some of the practices that are required to make developmental sector data open to other researchers and practitioners. He leaves us with a crucial question: can donor organizations, that fund so much of the development sector, be “convinced that open data is worth their money” (112).Gaurav Godhwani, Noopur Raval, and Aakash Solanki provide telling instances of the very uneven penetration of data practices in both the private sector and government agencies. Paper and digital media jostle each other for supremacy, and the latter is still often subservient to the former for record-keeping. In “Making India's Budgets Machinable,” Godhwani argues for the importance of open data in enabling productive contact between citizens and state agencies, but rues the fact that there are “very few open budget data initiatives across the globe, and those that exist are yet in their nascent phase” (114). Anirudh Raghavan's “The Work of Waiting: Syndromic Surveillance and the Paradox of Immediacy,” is a particularly interesting exploration of the ways in which putatively data-driven governmental agencies cook data in order to legitimize their actions, or to justify inaction. His example is the Integrated Disease Surveillance Programme, which was created under “the larger rubric of global health security” (148). Data collected by health centers and supplied to the central agency is meant to drive the rapid response teams that respond to clusters of disease. However, the flood of signals sent to the center overwhelms the possibility of fieldwork by teams of microbiologists, epidemiologists, and pathologists, who take weeks to conduct their investigations. Thus, data is often treated as noise, not to be acted upon for one reason or the other. Typically, slum clusters are treated as repositories of ill health, so that their data is suspect—action must wait till the data is curated. This delay is not an aberration but is built into the structure of the system; as a unit chief puts it, “If you use data without waiting, you get panic, confusion. Data becomes counter-productive” (153).I will leave the last word to Lilly Irani, whose essay “Hackathons: Labor, Politics, and the Organization of Public Passions” is a reminder that the oceans of data that we seem to drown in can be reshaped into a public good, but with a great deal of civic collaboration and effort. The “passions provoked by ‘open data,’ ‘innovation,’ and ‘nation-building’ can prove potent resources for experiments in statecraft, private-sector research and development, or activist infrastructures. They can do so in ways that strengthen the adaptive capacities of investors and governments, or they can do so in ways that strengthen the reproduction of resistance and transformative efforts” (68). Irani points to hackathons, collaborative digital labor that designs “demos or prototypes of software-to-come,” often sponsored by private companies or the government to “expand accumulation.” However, there are also “hackathons in which people gather to care for infrastructures and data that sustain the publics and their politics in the face of environmental extraction” (68–69). Irani writes that as “a vehicle of care . . . hackathons . . . attract people to the often invisible labor of protecting data, expanding access, and sustaining resources that expand the field of political contestation” (76). She offers us a pithy, and hopeful, slogan: “Other hackathons are possible” (74).