The Emergence of Transdisciplinary Research and Disaster Science

Introduction: The Origins of Disaster Science as a Scientific Field

The use of disaster science (DS) as a concept to designate a scientific field of study is relatively new. It is becoming institutionalized very rapidly through the ongoing attempts to establish a transdisciplinary Research Institute for Disaster Science at Tohoku University in Sendai, Japan (Yonezawa et al., 2019). Other moves toward institutionalization include the founding of the journals Progress in Disaster Science in 2019 and the Journal of Disaster Science in 2016, and in 2014, the creation of the Disaster Science Research Initiative to Enhance Responder Safety and Health to encourage transdisciplinary research (TR) on health-related issues associated with disasters, as part of the U.S. National Institute for Occupational Safety and Health. DS is interdisciplinary in scope and has gained increasing recognition among scholars from Asia, North America, and Europe, as shown by the results of a search in Scopus (www.elsevier.com/solutions/scopus/how-scopus-works/content), the electronic reference tool in late March, 2020, which revealed that 92 publications had used the term. From 1993, when the first mention occurred to 2010, 14 documents utilized it, or .82 articles per year. Of the remaining 78 publications, 19 were published during 2011 to 2013 (6.3 articles per year) and 59 from 2014 to 2020 (8.4 articles per year). The upward trend in the average frequency of publications per year is clear-cut. With 19 mentions, social scientists are not the most frequent users of the label. Instead, scholars in three other disciplinary fields take the lead. These are in the disciplines of earth and planetary sciences (34), engineering (27), and environmental sciences (24).

The United States, with 17 publications to its credit, is third in the register of countries represented by the authors. China, with 31 and Japan with 20, occupy the first and second ranks, respectively. European countries (the United Kingdom, Germany, Slovakia, France, and Norway) come fourth, with 17 citations among them. Hui Liu et al. (2020) document Tsinghua University’s preeminence in the number of publications examining evacuation, including traffic evacuation modeling, group behavior and crowd evacuation, disaster emergency management, simulation, and emergency traffic evacuation; and exit choice in simulation studies.

The source of funding used by these authors also demonstrates the international aspect of DS. Not all 92 publications are funded, but 14 are by the National Natural Science Foundation of China, and 5 by Japanese sources such as Tohoku University and the Japan Society for the Promotion of Science. In contrast, the National Science Foundation in the United States had sponsored only two publications. The authors’ professional affiliations show similar patterns. Twenty-eight authors were associated with institutions like China’s Northeast Normal University, Ministry of Education, Academy of Science, and Institute of Water Resources and Hydropower Research.

The English language, by far, is the lingua franca used by the authors, which is revealed in the journals used for their publications. The four most popular outlets used are the Journal of Disaster Research (10 publications), Natural Hazards (9), Journal of Natural Disasters (4), and International Journal of Disaster Risk Reduction (3). The reasons for the emergence of this new field of DS is outside the limited scope of this article. It occurs at a time of increasing costs associated with damage caused by various hazards worldwide and corresponds to a more general interest in the use of interdisciplinary scientific collaboration in the amelioration of the vulnerabilities of societies to extreme events such as those produced by the rapid warming of the earth (on the shortcomings of interdisciplinarity, see Conrad, 2002). Reflected of this worldwide awareness is the emphasis placed by the 2015 Third UN World Conference on Disaster Risk Reduction (better known as the Tokyo Conference) on Integrated Research on Disaster Risk (Koike et al., 2018), which made a strong call for the need for multidisciplinary scientific research efforts to guide disaster remediation programs.

Definitional Issues

The relative recency of the emphasis on DS and the shift toward TR in the study of disasters has allowed a variety of interpretations of what the new emphases mean. There is a degree of confusion, for the methodological and theoretical bases for the new science of disasters is still an incomplete project. DS is a new field in which the several scholarly scientific disciplines participating in the study of disasters use their theories and methods primarily. This interdisciplinary arrangement creates inefficiencies and limited payoffs. How to remedy them is the topic of several publications. For example, Alexander (2002) appraised the articles published in the journal Disasters between 1976 and 1997 before the idea of DS was widely discussed. He concluded that during these 21 years, the field of disaster studies had advanced but also had suffered from overspecialization, as well as from a poor transfer of knowledge to potentially interested users. Alexander discussed several important changes during these years, such as that the accessibility to statistical information had improved over those years, even though it was difficult to make comparisons using these data sources because of their lack of agreement about the definition of key concepts. He lamented that disaster relief and mitigation had not been incorporated into economic development and noticed that the physical sciences and engineering had obtained the lion’s share of funding compared with the social sciences (including political sciences, sociology, geography, and social anthropology). He also noticed that divisions among the social sciences seemed to be due to their overemphasis on the primacy of their pet disciplinary discourses (e.g., for sociology, the analysis of complex organizational behavior.) In some ways, forestalling the emergence of DS, he mentioned that there was a need to treat hazard, risk, and disaster as integrated phenomena (Alexander, 2002) as well as study the links between culture, architecture, structural design, and social organization.

More recent treatments of the meaning of DS portray it as an applied tool for finding practical solutions to the many difficulties international development organizations face when they try to solve the difficulties disasters generate. Illustrative of this emphasis is the activities of the United Nations Office for Disaster Risk Reduction’s Scientific and Technical Advisory Group (STAG; Aitsi-Selmi et al., 2015; see also Revi et al., 2015). According to STAG, disaster risk reduction (DRR) initiatives should involve practitioners, local stakeholders, and scientists, and it considers scientific know-how as a crucial component. Thus, a 2013 STAG report highlights 10 case studies where science, technology, and innovation were successfully used and communicated in the carrying out of disaster risk interventions. The framework of the UN Third World Conference on Disaster Risk Reduction acknowledged the importance of science and technology for DRR, sustainable development, and climate change mitigation and the need to “identify and enhance synergies and be aligned with the full cycle of prevention, mitigation, preparedness, recovery, rehabilitation, and response related to disasters and other potential global emergencies” (Aitsi-Selmi et al., 2015., p. 14). Undoubtedly, the UN perspective on DS has become very significant in advancing a clear and comprehensive view of disasters, and DS as an applied science has many important benefits that should be recognized. However, to the extent that it limits the future scope of this new science, the applied emphasis may relinquish its greater potential usefulness in helping eventually change the culture of modern societies that is necessary for the mitigation of the effects of disasters (on this point see Mileti, 1999). Relevantly, Rahman and Fang (2019) in their recent in-depth review of Priority 1 of the Sendai Framework for Disaster Risk Reduction, seem to imply that the Priority 1 stress on disaster risks veiled other needed emphases on vulnerability, coping capacity, exposure, the nature of hazards, and environmental factors associated with disasters.

Some authors emphasize that for DS to succeed, there is a need to develop a unified conceptual framework that would be responsive to the requirements of the different sciences interested in the study of disasters, and that would facilitate the identification of important matters worth pursuing. However, the development of a grand theory of disasters is problematic since there is not even agreement at present about how to define disasters (Perry, 2005) and whether the term natural disasters is misleading. There is also disagreement on the meaning of key concepts (see next), as shown by the presence of multiple models of vulnerability (Birkmann, 2013) and the varied conceptual underpinnings of emergency management (McEntire, 2004). And despite the multiple disciplines and methodological approaches of scholars in the field (see Lach, 2014; Peek et al., 2020), not much seems to have changed in more than 65 years to judge from I. L. Janis’ (1954, p. 15) “disaster-ology myth,” in which he documented the frustrated hope of scholars searching for a common conceptual framework. Janis (1954, p. 15) pointed that it was:

. . . valuable to take an occasional comprehensive look at the ramified effects of large-scale disasters for the special purpose of trying to discern hitherto unnoticed interrelationships. But . . . many of the theoretical issues and hypotheses considered under the general heading of disaster research cannot be expected to hang together (and will not be) illuminated by a common set of explanatory concepts and principles.

Unknown by Janis, of course, is the immediate goal of transforming the disaster field of studies through the creation of a transdisciplinary approach to research on disasters.

More than a decade after Alexander’s 2002 article, Wachtendorf (2019, p. 347) suggested the need for a comprehensive and inspiring grand question that would energize the field of DS:

Missing (is) immense, glorious questions—ambitious objectives, of the kind the space program demanded . . . our questions in disaster science have the potential to speak to the fundamental questions of human survival . . . without feeding into more ambitious aspirations, we risk those incremental advances getting lost in the disciplinary discourse, in pursuit of academic journal impact factor ratings . . .

Wachtendorf’s bounteous approach ignores Janis’ forewarnings and is opposite of the narrower UN emphasis on application. Her perspective on DS has merit, for it resonates with some of the central values of the scientific communities, such as enhancing their relevance and prestige.

Whether the promotion of such ambitious goals is the most effective way to proceed is open to debate. The experiences of development economics come to mind. William Easterly (2001), at the time chief economist of the World Bank, documented the professionalism and excellent values of many of the economists working for the World Bank’s poverty reduction and other economic development goals. The worldwide elimination of poverty was for many years after the formation of the United Nations in 1945, its most important goal. Billions of dollars were invested over decades in multiple programs to achieve it. Nevertheless, most of these programs failed. None of them were able to obtain the treasured goal of reducing poverty, which keeps growing. Easterly (2001) concluded that the failures of the poverty programs were due to their very goals, which were so extensive that it was difficult to identify the appropriate means to measure their success and whether their employees were doing what they were supposed to do. The officers overseeing the projects could not show that they had achieved the goals of reducing poverty in developing countries. Considering these painful experiences, Easterly (2001) argued that economic development projects needed to be more delimited than they had been in the past and have well-specified, attainable goals to make it possible to judge whether these goals had been achieved, thus allowing for the elimination of those projects that failed. Perhaps DS could derive some useful lessons from this experience. Rather than a grand unified theory, what is now necessary, as stated previously, is the multidisciplinary coordination of research amalgamated by established shared rules and centered on the answers to theoretical middle range questions (Merton, 1949). Needed are limited questions regarding disasters, or to put it differently, the study of measurable phenomena and social organization with limited conceptual ranges rather than the entire conceptual structure used to derive them. There is a need for disciplined patience that, over time, may allow for the cumulation of scientific knowledge that would provide the empirical foundation to the new science.

Authors such as McEntire et al. (2002), advance disaster management as the future paradigm for the field despite its narrow emphases. Similarly Ismail-Zadeh et al. (2017) offered another road map. These authors make the valuable distinction between disciplinary, multidisciplinary, interdisciplinary, and TR. They argue for the value of the transdisciplinary approach to scientific research on disasters (see also Bernstein, 2015). In their terms, TR involves scientists from different disciplines who agree to study an issue or problem relevant to the diminution of the effects of disasters comprehensively (Rhoten and Parker, 2004; see also Peek et al., 2020). They “exchange data and information, share resources, create conceptual, phenomenological, theoretical and methodological innovations, integrate disciplines, and move beyond discipline-specific approaches.” (Ismail-Zadeh, 2018, p. 233). TR first requires the abandonment of traditional disciplinary silos and then their eventual reconstitution into an integrated theory and methods, as well as a shared view of disaster demands that may include other interested parties and community actors. TR encourages scientists to train on how best to establish coordination and collaborations across the different disciplines. Ismail-Zadeh et al. (2017, p. 972) discussed three of TR’s dominant features. The first is that there should be agreement among the researchers about the research problem they will pursue and how members representing different disciplines are going to communicate and collaborate. The second is the need to create a consensus that the goal of the study is the amelioration of the effects of well-specified problems through “solution-oriented and transferable knowledge.” The third is the need to integrate the knowledge generated by the research in ways that are of value to multiple disciplines from a practical as well as a scientific standpoint. It is hard to overemphasize the usefulness of this statement. The Ismail-Zadeh (2018) model dovetails to a considerable extent with the earlier observations made by the National Research Council’s Committee on Disaster Research in the Social Sciences (Kreps, 2006), which reviewed several features common to successful projects and emphasized four of them. Two are having the encouragement of senior leaders and the monetary backing of granting agencies. The other two are near duplicates of Ismail-Zadeh’s (2018) thinking, namely, the need for a setting for continuous dialogue and an emphasis on an applied problem.

At its core, DS is a transdisciplinary field of study (Van Niekerk, 2012; Lach, 2014). Research projects in DS need to have project scientists from different disciplines willing to show mutual respect in their dealings with each other and reach a sincere collective agreement among themselves about what they will do and how they will do it. Also critical in TR is the importance of agreeing on the practical value of the study. These features increase the chances that the participating scientists will be able to share in the development of the design of the overall study, the selection of the methods to be used, and coordinate their efforts to address the research questions posed.

Several examples show the type of studies that would profit from a TR approach (Kreps, 2006). There are studies to measure the accurateness of loss estimation tools, to shed light on the cost-effectiveness of different loss mitigation efforts, to help understand the efficacy of various financial tools to transfer risks, to evaluate the success of different techniques for response and recovery for specific categories of persons such as remote sensing and communication networking systems, and to measure vulnerability and community resilience. A satisfactory answer to questions such as these presumes the presence of specialists in the various disciplines interested in the study of disasters working together on a problem of common interest and adapting their approaches to be able to incorporate and appreciate the insights of team participants from different specialty areas.

A Way to Understand the Differences Among Disaster-Related Scientific Fields

Skeptics of the viewpoint that TR will become the predominant practice characterizing DS will point to the chronic inability until now of most disaster studies to develop fruitful collaboration among the different specialists participating in studies of the various aspects of disasters. This is an important and valid characterization of the past and anticipated future of DS. Nevertheless, there is a reason for guarded optimism that progress is possible here as well. The next paragraphs make the argument that there is a need to appreciate the extent to which the discrete specialties differ. Doing so will help design collaborations among different specialties and develop appropriate elements of the scientific design which would minimize the effects of their differences or help exploit these differences on behalf of their shared goals.

Fortunately, recent advances in the philosophy of science allow us to assume that the differences in the methods and theories of the approaches interested in the study of disaster are of degree rather than kind. By its logic, until now seemingly insurmountable differences do not constitute ipso facto an impenetrable obstacle to the eventual fruitful synchronization of these professions. An important contribution to this literature is Hoyningen-Huene’s (2013) systematicity methodology, which offers a reasonable procedure to begin to understand in comparative terms the characteristics of the disciplines active in the disaster field (but for criticism, see Oreskes, 2017). Bschir et al. (2019) point to three key features of Hoyningen-Huene’s (2013) approach to the nature of science. In paraphrase, first, he does not exclude from science the humanities, history, the social sciences, or even certain aspects of the arts. He also does not emphasize a set of conditions necessary to label a practice scientific and thus does not try to answer the question “What is science?” by listing a set of criteria and conditions necessary for a profession to be scientific. Instead, his interest is in contrasting science as a form of knowledge to other forms of knowledge, particularly common knowledge. In this manner, this author examines the complexity of science in all its many manifestations. The principal goal of Hoyningen-Huene’s (2013) philosophy of science is to offer an appropriate conceptualization of science that would involve, contrast, and differentiate common sense knowledge from scientific knowledge in relative rather than absolute terms.

Instead of telling what science is, Hoyningen-Huene (2013) offers multiple dimensions that vary by their degrees of systematicity or by the extent of the systematic character of their knowledge, and that can accommodate both common sense and science. Math and physics are disciplines that score very high in all the nine components of systematicity, while the social sciences are highly systematic only in some of these dimensions. The nine components identified by Hoyningen-Huene (2013) are description, explanation, prediction, the defense of knowledge claims, critical discourse, epistemic connectedness, an ideal of completeness, knowledge generation, and the representation of knowledge. In a very abbreviated form and paraphrasing the original text in Hoyningen-Huene’s (2008, 2013), they are as follows:

A Computer Simulation Tool for the Integration of Different Fields in DS

A common observation in DS is that computational research is becoming widespread in all its subfields. Transcending the disciplinary boundaries to link computational models from different subfields and thus simulate overarching disaster scenarios is not a trivial task given the great differences in the types of models used and programming languages employed. One of the promising ways by which to address this problem is to use distributed simulation, which can link independent simulators that execute on the same or separate processors. An example of this is the computational tool developed by Lin et al. (in press). The tool is the Simple Real-Time Infrastructure (SRTI; GitHub, 2019). SRTI has a low barrier to use by researchers with limited programming or computer science background. SRTI allows researchers from different backgrounds and fields to link their computational models together to study the effects of natural hazards on community resilience through a client–server structure. Data communication between the distinct simulators in different fields is done through sockets and uses a publish–subscribe pattern. Each simulator is a black box that interacts through the SRTI distributed computational platform. Figure 1 shows how SRTI links various simulators.

OPEN IN VIEWER

Figure 1.

Distributed simulation using SRTI.

Treating simulators as black boxes is one of the key advantages of SRTI for use in DS simulations. Users from a specific discipline (e.g., civil engineering) do not need to know (and do not have the training to know) the details of how simulators from another part of a simulation (e.g., disaster health) work. Each group just needs the output from a simulator if they provide a given input. For example, a disaster health system simulator that addresses injuries in collapsed buildings only needs to know the extent of damage suffered by a structure due to an extreme event, which comes from a structural simulator (a civil engineering part of the project). The SRTI framework is extensible. New simulators can be easily plugged to it, enabling more complex and realistic simulations of disaster scenarios. SRTI supports different programming languages and different computer systems, and it achieves extensibility and scalability for complex systems in hazard simulation. Some key features of the SRTI include the following: (a) it is open-source and free to use (GitHub, 2019; Lin et al., in press); (b) it permits distributed simulations across multiple machines; (c) its precompiled components can be downloaded and directly used without installation, and (d) data messages are parsed as “strings” using JSON format. ¹

As shown in Figure 2, the structure of the SRTI consists of three key components: the RTI Server, the RTI Lib, and the users’ simulators. In the SRTI, data communication is through sockets. The RTI Server is a server-side application that acts as the shared connection point for simulators in a simulation system. It can be run either on the same machine or a separate computer from the individual simulators. The RTI Lib is a client-side API (Application Programming Interface) library that allows a simulator to connect to an RTI Server. The precompiled RTI Lib API must be locally stored and referenced on the same machine as the simulator. The RTI Lib gives the details of socket communication, message creation, and parsing. Using the RTI Lib API, the connection to the RTI Server and the transmission of messages is straightforward.

OPEN IN VIEWER

Figure 2.

High-level architecture of the SRTI v1.00.00.

The simulators here refer to the computational models provided by a user that are connected to the SRTI Server. The user needs to modify their simulators to reference the RTI Lib and its API as required. The source code, precompiled libraries, documentation, and other information associated with the SRTI are available at The RTI Server. Both it and the RTI Lib is precompiled so that the user only has to download the information. The RTI Server was written in Java and executes on systems with the free Java Runtime Environment installed. At this time, the RTI Lib API is available in native Java and C++ so that the simulators written in Java, C++, and Matlab are supported. Both the RTI Server and RTI Lib API are also available as source code. Other compatible languages can also do JSON parsing and socket communication.

SRTI has an optional component, that is, the RTI Server GUI. GUI stands for graphical user interface and it automatically opens when launching the RTI Server. It provides a visual window that describes the “hostname” and “port number” of the RTI Server as well as a list of the connected simulators. Users can also inspect a live feed and history of data messages received by the RTI Server through the RTI Server GUI. Additional details of SRTI are in Lin et al. (in press), and its use for modeling community response to natural hazards is in Lin and El-Tawil (2020).

Conclusion

This article has documented the relative recency of the effort to establish DS. It has shown the ongoing multiplicity of attempts to define DS and its links to transdisciplinary scientific designs to advance it in response to the methodological and theoretical difficulties identified in interdisciplinary disaster studies. Together with the value of TR, we counsel against the attempt to develop a grand unified theory for the field of disaster studies, instead suggesting the value of more circumscribed empirical studies guided by theories of the middle range and a set of rules guiding the style of professional relations necessary to make transdisciplinary collaboration successful. We then point to the real uncertainties of the outcomes of this effort to reinvigorate the study of disasters by considering the continued absence of agreement about crucial concepts among social scientists studying disasters and the need to understand why there are so many different specialties failing to carry out successful collaborative research. As a partial answer to this question, we mention the new philosophy of science approach to the understanding of science known as systematicity methodology. Such an approach has the potential value of assisting in the development of a more nuanced, holistic understanding of the degree and type of systematicity of the various sciences involved in the study of disasters, which in turn could provide an discernment of the relative ability of scientists representing different disciplines in the study of disasters to work together as team members. It could also guide interventions to make such outcomes more likely.

We then describe critical new tools available in the study of disasters and their explicit assumption of the presence of such interdisciplinary teams, and we close with a brief introduction to a new simulation integrator, a tool that allows computer simulation programs from different disciplinary specialties interested in disasters to interact and produce more sophisticated answers to the questions posed by the original simulation. Witnessing such advances, it may be that despite appearance to the contrary, things often change for the better. Perhaps this is so as well for the study of disasters. The most important message of this article is that it calls attention to the critical need for specialists in the study of disasters to move beyond acknowledging the problems inherent in interdisciplinary studies and consider how best to bring about a successful approach to TR efforts.