Organized Session

Histories of the Less-Than-Known: Questions of Risk and Uncertainty in Global 20th Century Science

Organizer

Aaron Van Neste

Harvard University

Chair

Toshihiro Higuchi

Georgetown University

Metadata

Session Abstract

How have people and organizations lived and made decisions under conditions of risk, uncertainty or incomplete information? How do these conditions manifest and what happens when they persist? This panel aims to explore the origins and effects of uncertainty on individuals and communities, from the development of scientific practices to personal and communal coping strategies. We suggest that risk and uncertainty must be understood as socially embedded and emerge from existing structures of race, class, gender, capital, and geopolitics. Yet, at the same time, risk and uncertainty have the power to trouble and disrupt the solidity of these categories. They are linked on a psychosocial level to our fears and anxieties - of nuclear radiation, birth defects, unstable food systems, and the racial other. Attempts to quantify, control and mitigate them - to turn incalculable uncertainty to calculable risk, and to position risk within algorithms solvable with available data - are marked by both what Ted Porter has characterized as "trust in numbers," as well as the skepticism of numbers' utility in improving the welfare of individuals.
Taken together, our papers span three continents and the ocean, and encompass the majority of the twentieth century. They engage in conversation with a diverse theoretical literature, from Ulrich Beck's "Risk Society" and Charles Perrow's concept of "normal accidents" to the class and expertise politics of risk (e.g. the work of Brian Wynne) to tactics for creating uncertainty (e.g. the work of Allan Brandt and Naomi Oreskes). Our panel seeks to contextualize risk and uncertainty within scientific and other knowledge communities while problematizing these concepts in both definition and practice.

Presenter 1

Settling Uncertainties on Radiation Risk

Sumiko Hatekayama

University of Pennsylvania

Abstract

In assessing the health effects of radiation, a reliable system for estimating individual radiation dose-known as a dosimetry system-is indispensable to studying how radiation doses correlate to observed health outcomes. When evaluating the health impacts of the atomic bombings of Hiroshima and Nagasaki, the dosimetry systems developed by the Atomic Bomb Casualty Commission (ABCC) and its successor the Radiation Effects Research Foundation (RERF) have long been held as the "best available." Large epidemiological studies of hibakusha-atomic-bomb survivors-conducted by ABCC/RERF and other entities, have relied on them. The radiation protection community throughout the world depends on risk estimates defined by these epidemiological studies. However, archival materials highlight the inescapable uncertainties that have always underpinned dosimetry work. Often, consensus was as much about policy as about data: uncertainties were "settled" by a series of institutional decisions which reflect the intricate web of cultural, political, and economic interests surrounding ABCC/RERF. Working with hibakusha and other nuclear victims, I have also encountered situations where laws and regulations that leverage insights from ABCC/RERF exacerbate people's suffering rather than mitigate or alleviate it. In this paper, I show how, based on the actual symptoms experienced by hibakusha, some scientists came to contest the work done by ABCC/RERF. At ABCC/RERF, uncertainties were often dismissed on the basis that they would not impact risk estimates at a population level. In contrast, those who contested the work of ABCC/RERF-initially as a part of a lawsuit-focused instead on individuals' health and welfare. I argue that the contestation over radiation risk sheds light on how technical questions are the subject of long-term scientific negotiation, inflected by the interests of different parties, and bound up with questions of data, law, politics, and biology.

Metadata

Presenter 2

Many Anomalies Which I Cannot Explain: Early Twentieth-Century Actuarial Science and the Making of the Semi-Tropics

Rachel Bunker

Rutgers University

Abstract

This paper will examine the advent of the "semi-tropical" as a salient category amongst early twentieth-century actuarial scientists as a means for understanding the influence imperial expansion had on broader ideas about risk and uncertainty. That insurers and their cohorts of actuarial scientists comprised some of the earliest inventors of risk has been documented. As historians Dan Bouk, Jon Levy, and others have chronicled, during the nineteenth century, the mortality calculations of actuarial scientists propelled the creation of a new type of financial instrument, the risk, which transformed the inevitable, yet seemingly unpredictable event of economic loss due to death into a measured and priced commodity. In doing so, actuarial science contributed to an increasing distinction made by early economists and social scientists, which differentiated knowable, measurable risk from true unknowable uncertainty. As this paper will show, however, practical efforts to harness mortality data and bring regions outside of the territorial United States and western Europe into the actuarial fold and the realm of commodification muddled neat separations between risk and uncertainty. In particular, white actuaries' anti-blackness framed circum-Caribbean populations, as well as culture and politics, as too unruly and idiosyncratic to conform to the law of statistical averages. This paper follows British and American actuaries as they traveled to the Caribbean and engaged in their own ethnographic research. The paper argues that faced with racialized and suspect data, actuaries adopted the liminal and malleable spatial category of the "semi-tropics," which was neither calculable nor wholly unknowable.

Metadata

Presenter 3

The Power to Become a Monster: Pregnancy as a Site of Risk in Early Twentieth-Century Biology and Medicine

Miriam Rich

Yale University

Abstract

This talk examines scientific and medical understandings of pregnancy as a site of risk in the
early twentieth-century United States. It explores how scientific and medical experts of this
period - including physicians, embryologists, and eugenicists - conceptualized, represented,
and mobilized pregnancy risk in new and sometimes competing ways. In this period, physicians
and healthcare providers sought to foreground and intervene on prenatal risk as part of a
broader shift in the medical management of pregnancy; biologists of this period likewise
worked to emphasize and characterize the nature of prenatal risk. In 1908, the embryologist
Franklin Paine Mall foregrounded gestation as a period of critical environmental risk to human
development, asserting that "the power to become a monster is present in every ovum,
provided the conditions surrounding the ovum be inimical to normal development." Eugenic
scientists such as Charles Davenport, while rejecting the environmentalist emphasis of
embryologists like Mall in favor of a strict hereditarian stance, likewise constructed and
mobilized discourses of pregnancy as a site of risk: specifically, these eugenicists portrayed
pregnancy as a time of danger and vulnerability to racial degeneration. This talk will explore
how, in interpreting and managing pregnancy as a site of risk, scientific and medical experts
engaged in particular types of social and ideological work. Competing discourse on pregnancy
and risk became a way to work out contested cultural meanings of human development,
difference, and descent.

Metadata

Presenter 4

They Rarely Remain Steady for Long: Fluctuations of Marine Species and the Question of Sustainable Fisheries

Aaron Van Neste

Harvard University

Abstract

While sustainability (using a resource such that the supply does not run out) has guided fishery science since at least the 1930s, the concept of sustainable fisheries is undermined by aspects of fish population biology that were widely known much earlier. In 1914, Norwegian biologist Johan Hjort demonstrated that the year-classes (the population of fish that survive their first year of life) of herring fluctuated by up to 30x from one year to the next. A single strong year-class could dominate the population for years until another strong year-class appears. However, recruitment fluctuations proved nearly impossible to predict. Scientists attempted to correlate fluctuations with potential environmental causes, but in each case, the correlation would eventually fail, leading to a perception of recruitment fluctuations as truly unknowable uncertainty.
By the 1950s, fisheries scientists had given up on correlating environmental conditions with recruitment fluctuations. Some models developed for predicting year-class linked recruitment to population size, while others ignored recruitment altogether. By removing a major source of uncertainty, these models could forecast future population sizes. Even though they were frequently inaccurate, this apparent precision appealed to managers and steady-state recruitment models were even enshrined in legislation. To achieve sustainability, the seemingly unsolvable problem of recruitment fluctuations had to fall away. But even as they vanished from management models, the spectre of recruitment fluctuations haunted fisheries science through a legacy of collapsed stocks, prompting ecologist Ramón Margalef to remark of fish populations in 1960, "They rarely remain steady for long... fluctuations lie in the very essence of the ecosystems." This paper argues that the absence of recruitment fluctuations from the fisheries science of the mid-late 20th century poses a fundamental challenge to that science's claims of sustainability.

Metadata