Financial companies are increasingly leveraging financial technology (fintech) to monopolize financial data, ostensibly to maximize profit for their clients and enhance their power in society. This trend both exposes the serious limitations of individual consent models of data protection and undermines the nature of financial data as a shared resource by excluding data contributors from its governance. This chapter posits that consumer associations, acting as data trusts, can play a crucial role in overseeing financial data-opolies while fostering the development of the community governance of data. Beyond addressing privacy harms, these associations can promote the attainment of important social goods, including the prevention of predatory and discriminatory lending and the expansion of access to financial capital. By leveraging and shaping both formal and informal rules-in-use that may facilitate these efforts, consumer data trusts can ultimately enhance the legitimacy of financial data commons. A discussion of BlackRock’s Aladdin platform and a European consumer association’s lawsuit against Meta illustrates the argument.

To conclude, I turn to Tessa, a test veteran widow, and her uneasy reflections on the claim-making process. Recognizing the double-edged nature of seeking justice, its enlivening and injurious potential, she asks that we take seriously the enduring challenges of injury claims, that we continue to interrogate how and why contemporary institutions turn justice making into pathways for ongoing harm. An exploration of the injurious nature of law is, I argue, crucial for understanding both the political origins of injury, and possible meaningful political responses to injury. This in turn reveals the necessity of reimagining injury in ways that can account for the historical, political, social, and institutional harms that accrue from our very systems of redress. That nuclear injuries have proven so difficult to respond to without enacting further harm must inform our current approaches to nuclear risk and disarmament, and necessitates a reimagining of injury law. Such a reckoning offers us possibilities to prioritize new modes of collective responsibility, care, and justice in the aftermath of radiation exposure and other invisible environmental harms.

In this chapter, we explore an unsupervised learning problem: estimating a distribution function from two-dimensional data. Although there is no response variable, the workflow mirrors that of supervised learning. We select the best-fitting function within a family by maximising the sum of the log of the distribution's values at the observed data points. As in supervised learning, excessive flexibility leads to overfitting, while insufficient flexibility leads to underfitting. We use cross-validation to identify a function family that achieves a happy medium.

Chapter 1 attends to the historical context of British bomb testing through an exploration of contrasting accounts of nuclear risk. It juxtaposes archival and official narratives of military risk management with test veterans’ own memories of hazardous events. This reveals the jarring mismatch that test veterans experienced when encountering official records detailing moments their younger selves experienced first-hand. I explore how radiation measurement technologies, material infrastructures, and a military culture of hierarchical trust negated the perception of risk, yet still left room for servicemen to experience doubt and fear. This chapter also places the harms to service personnel in the wider colonial context of racial hierarchies and the theft of lands, revealing how stories of Indigenous injury became uncertain undercurrents within test veterans’ own narratives.

This chapter introduces simple and multiple linear regression models – core tools in predictive modelling due to their simplicity and interpretability. These models assume the response variable is a linear function of the predictor(s), plus a noise term. The regression function gives the expected response given the predictors. The coefficient of determination, R2, measures how much of the variance in the response is explained by the model. In simple linear regression, R2 equals the square of the Pearson correlation between response and predictor; in multiple regression, it equals the square of the correlation between response and predicted values. Each coefficient in multiple regression reflects the expected change in the response for a one-unit increase in that predictor, holding others fixed. Standardising predictors lets us compare coefficient sizes. Strong collinearity between predictors increases uncertainty in the fitted coefficients. Models using only a subset of predictors may generalise better than those using all and overfitting. The squared error risk of a modelling procedure – its expected test error – can be broken down into bias, variance and irreducible noise.

The effectiveness of comparative studies resides in the breadth and suitability of the cases used in pursuit of a research question. We have selected thirteen societies to develop a comparative understanding of how premodern economies were organized and operated. These span a broad range of societies in terms of organization, complexity, and their place in time and space. They include societies from around the world: six from the Americas and seven from Eurasia and Africa (Figure 2.1). They are diverse in adaptation and scale, and include horticultural, foraging, pastoral, mixed economy, and sedentary agricultural groups. Examples include tribal, chiefdom, and ancient state-level societies. Despite this diversity and the historical independence of the Americas and Eurasian/African examples, commonalities exist in economic structures because of the cumulative and shared nature of economic behaviors that we hope to capture.

In this chapter, we examine our first supervised learning problem, focusing on how to construct prediction functions and assess their performance. Given data consisting of predictor–response pairs, we can learn the parameters of a prediction function by minimising a loss, such as the residual sum of squares, which measures the discrepancy between actual and predicted responses. Using more flexible families of prediction functions typically reduces loss on the training data, but excessive flexibility can lead to overfitting: fitting to noise rather than the systematic component of the relationship. Overfitting results in poor prediction performance on new, unseen data. To estimate how a prediction method will perform on unseen data, we use cross-validation. However, when we compare many prediction methods using cross-validation, the best-performing method often appears better than it truly is; its apparent performance is an unreliable guide to its future accuracy. Prior knowledge is crucial for selecting plausible prediction methods to compare. Finally, we can use bootstrapping to quantify uncertainty in prediction functions and their predictions.

Moral Philosophy, Mental Philosophy, Reformism, Brahmo

On an undisclosed date in the 1880s, the gods came to College Square. There, under the impressive colonnade of the recently completed university building, Varuna informed Brahma, first among the gods, that the recent appointment of Prasanna Kumar Ray (1849–1932) to a professorship in “mental philosophy” (manobijnan śastra) inaugurated a new era in the history of higher education in Bengal. That the gods would descend from Kailaśa to Presidency to celebrate the academic achievements of mortals might appear eccentric, even idiosyncratic, but Durgacarana Ray's puranic novel Debagaṇera marttye āgamana (“The Gods’ Visit on Earth”) was registering a pervasive worldly concern: the enduring interest among the Bengali bhadralok in the “philosophy of mind.”

This chapter investigates how the fortunes of “mental philosophy” rose and fell in College Square. In the early nineteenth century, a set of distinctively Scottish arguments soldering together epistemological realism, neo-Baconian anti-scholasticism, and theories of moral intuition into a single tradition of “common sense” philosophy came to acquire a disciplinary life as an increasingly global subject—“mental philosophy.”Though bemoaned in Britain by “philosophical radicals” as a conservative project unable to make a case for moral change, the common-sense arguments of “mental philosophy” found an unforeseeably radical reception in Calcutta, where Henry Louis Viven Derozio (1809–31) turned to the Scottish idea of self-evident morality to mount an explosively original challenge to extant śastric norms.

As a contribution to a volume devoted to the history of one iconic educational institution, readers might well expect its object of analysis to be fairly clearly defined. Yet, as Michel Foucault reminds us, “concepts of discontinuity, rupture, threshold, limit, series, and transformation present all historical analysis not only with questions of procedure, but with theoretical problems.”1 There are two kinds of discontinuities that I want to interrogate. First, the coherence and specificity of disciplinary backgrounds, rather than a general sense of belonging to the Presidency College as a whole. Did disciplinary affiliation, to let us say English literature or history or chemistry, remain confined to the object of analysis alone or did it contribute in some way a student's or teacher's more general social identity? Did it, for instance, in any way shape their political and social choices?

By raising this question, I also want to signal my distance from histories that invoke ahistorical and generalized categories such as “Science” or “Humanities.” Particularly in the case of “Science,” there is now enough scholarship that establishes that whatever might be the political utility of invoking the sign of science, a singular, unified notion of science as a category of historical analysis makes little sense.