“Structural disaster”: beyond risk sociology

Since risk is usually uncovered by estimating directly unobservable parameters, such as safety coefficients or the breakdown rate of novel components, the estimation of risk accompanies a considerable range of uncertainty, which can be corrected by introducing assumptions that could, in turn, be influenced by the social situations of those in charge of estimation. In its extreme, for example, the introduction of ad hoc assumptions could give rise to mutually opposite estimations of risk (for example, estimation within safety versus estimation without). In addition, it is not unusual that different assumptions leading to such disparate estimations are due to heterogeneous social situations among those who made them (see Figure 1.2).

A divergence in the risk assessment of the “Cambrian sheep” case (Wynne 1987, 1996) during the aftermath of the Chernobyl accident, the heavily context-dependent safety assessment of the nuclear reactors of electrical companies in the context of restarting them after the Fukushima accident, as well as other nuclear-related estimations (Eden 2004), can effectively illustrate this. Ultimately, the identification of a particular risk could be dependent not only on scientific knowledge but also on heterogeneous social situations; therefore, a particular social decision made in a particular social situation could be responsible for generating a particular kind of risk. Apart from risk society arguments, previous sociological studies on risk have focused on the subtler relationships between the responsibility of social decision-making, the contents of risk, and scientific knowledge.

Conversely, in the context of the allocation of responsibility for risk incurred as a result of decision-making, the agents in question have tended to be reduced either to human beings as a whole or to dichotomous agents, such as victims who are exposed to risk versus perpetrators who caused it (and can often escape from it) (Worm 1996; Connelly 2008). To explore extreme events within the nature-artifacts-society interface, it is necessary to sidestep such a bipolar categorization of agents, as the patterns of behavior among agents involved in extreme events who exist in the gray area between the victims versus perpetrators dichotomy could provide an in-depth look at the structural factors involved.

Here, the new concept of “structural disaster” can provide a deeper understanding of the social backgrounds of extreme events, detect a previously unforeseen precursor of similar events, and devise appropriate measures against future events. The term “structural disaster” is defined as a state consisting of one or more of the following characteristics (Matsumoto 2012: 46, 2014: 191):⁶

• 1 Adherence to erroneous precedents, which can cause problems that are carried over and reproduced.
• 2 The complexity of a system under consideration and the interdependence of its units, which can aggravate problems.

Figure 1.2 The Structure of Diverging Estimations of Risks

Note: “Conservative” and “Daring” represent a mutually diverging risk assessments of an incident. Dotted lines indicate the possibility of external influences from heterogeneous social situations.

• 3 The invisible norms of informal groups, which essentially hollow out formal norms.
• 4 Quick fixes to the problems at hand, which lead to further such fixes as temporary counter measures.
• 5 Secrecy developing across different sectors, which blurs the locus of those agents responsible for the problems that need to be addressed.

The concept of structural disaster in the science-technology-society interface was developed to provide a sociological account of the repeated occurrence of similar type failures (Matsumoto 2002: 25-27). Structural disaster can clarify a situation in which novel and undesirable events have happened but where there is no single agent to blame and no way to prescribe remedies. The reason for denominating a disaster as a failure of the science-technology-society interface rather than that of science, technology, or society' independently is critical to understanding the development of this book’s primary argument. For example, if nuclear physics is completely successful in explaining a chain reaction, then how did technology' such as nuclear engineering fail in controlling the reaction, as in the case of Chernobyl and its aftermath (Wynne 1996)? Similarly, if nuclear engineering is nearly successful in containing radioactive materials within reactors, then how did social decision-making fail, as in the case of Three-Mile Island (Perrow 1984, 1999, 2007; Walker 2004)? If society is completely successful in setting goals for the development of renewable energy' technologies, then how could science and/or technology' fail, as in the case of Ocean Thermal Energy' Conversion (Matsumoto 2005)?

In short, the success or failure of science, technology, and society are intimately interwoven but cannot be overlapped (Latour [1993]1996; Hecht 2012). In particular, there seems to be something missing “in between” that has unique characteristics of its own; for example, institutional arrangements (Fricke and Moore 2006), organizational routines (Vaughan 1996; Eden 2004), tacit interpretations of a formal code of ethics or protocols (Jobin 2013), invisible customs and semantics behind real-time logics (Perin 2005), or the networks of different organizational interests. The failure to interface can be an indicator of one or more of these states of affairs.

This book focuses on, among other things, structural similarity in the patterns of behavior of heterogeneous agents that interact in the science-technology-society interface under a specific social condition. Of course, “structural” here does not privilege commensurability and/or pure formality without corrosive elements. What is described through the concept of structural disaster could be either commensurable or incommensurable; it could have more or fewer corrosive elements depending on the time, place, and social contexts of a disaster (Fortun 2001: 366-367). Rather, what is asserted is that if structural similarity is observed among heterogeneous agents within and without disasters, this could have its own clarifying power to observers apart from its meanings to the agents involved. By focusing on persistent secrecy before, during, and after the Fukushima accident, this chapter explores the power of the concept of structural disaster.