Before outlining some challenges in DRR-related knowledge production and application, we need to explain our conceptual approach to knowledge. As visualized in Fig. 1, we distinguish between different qualitative levels of understanding: facts, data, information, knowledge, and wisdom. Although we adopted our organization of understanding from widely recognized models in the information and knowledge literatures (Cleveland 1982; Ackoff 1989; Rowley 2007), our intention is neither to promote a specific approach nor to contribute to the debate on the theoretical underpinnings of information science. However, we believe that a qualitative distinction of the different levels of understanding provides a sound basis from which researchers can better relate to policymakers and practitioners in the DRR domain. Moreover, a more differentiated use of the term “knowledge” is needed because in the floods of information of the information age the term is often confused and replaced with “providing information”. This is partly related to the advancement of information technology, which is increasingly producing and delivering facts and data, although much of that information remains unorganized, untapped, or unused.
From Facts to Wisdom
All of the elements of the continuum of understanding are abstract concepts and the distinctions between each stage are fuzzy. For instance, what constitutes information for one person may be just data to others because they may not have the context needed to make full use of that information. We agree with Cleveland (1982) that it is not important to search for universal agreement on the distinctions between the terms. For the purpose of this article, it is sufficient to consider data as a set of objective but meaningless facts that have not been processed and contextualized into usable information. In a scientific context, facts and data are generated through research and represented as structured records of measurements and observations. While facts and data do not have any inherent structure, information has context. As indicated by the original meaning of the verb inform, that is, “to give form to” something, to become information data need to be intentionally processed, organized, and structured in a useful way so that we can draw conclusions. Information is “data with meaning” that makes a difference.
Knowledge is created by accumulating and organizing information with respect to breadth, depth, and amount. Facts, data, and information are necessary mediums for eliciting and constructing knowledge. According to Davenport and Prusak (1998, p. 5) knowledge is “a fluid mix of framed experience, contextual information, values and expert insight that provides a framework for evaluating and incorporating new experiences and information”. While information is static, knowledge is dynamic, built through social interaction and experience, with the result that the “objective” facts, data, and information are considered and evaluated from different perspectives. One can neither count on one person’s knowledge transferring to another, nor assume that a knowledge transfer will have the desired impact. This is why awareness raising, training, and education are reasonable components of DRR policies, and why integrated, coproduced knowledge is so critical to implement. In the “linear” knowledge production model, academia’s best contribution to problem resolution lies in the adequate transfer of knowledge (through communication, education, patenting, or publication) to other actors charged with the implementation of such knowledge in the form of products, procedures, regulations, or problem solutions. A nonlinear understanding of knowledge production, however, assumes that relevant knowledge can be produced by any kind of actor—academic or lay—who must be acknowledged for his or her specific perspective on a given problem (Weichselgartner and Truffer 2015). Different actors can be the origin of new ideas, and knowledge flows can go in all directions, from practitioners to researchers, or from basic science to policy, and so on. In the DRR domain, however, the coproduction of knowledge is limited and implementation gaps between research and practice persist. A recent analysis of the characteristics of disaster risk research illustrates that most research on disaster risk is still discipline- or multidiscipline-centric, largely produced by North American and European scholars, and has limited success as an evidentiary basis for policy improvements (Gall et al. 2015).
Wisdom represents an even higher level of understanding: it is evaluated and reflected understanding, or integrated and applied knowledge. As with knowledge, wisdom operates within people. Experience that creates the building blocks for wisdom can be shared, but needs to be communicated with even more understanding of the personal contexts than in the case of knowledge sharing. Through the transitions from facts to wisdom, not only understanding increases but also the degree of participation and connectedness, resulting in a higher complexity. As Cleveland (1982) summarizes it: information is horizontal, knowledge is hierarchical, and wisdom is flexible. Furthermore, facts, data, and information deal with the past, whereas knowledge deals with the present. When we gain wisdom, we add more context and start dealing with the future as we are now able to envision the path ahead and design for what will be, rather than for what is or was. While the elements shown in Fig. 1 are abstract concepts and clear distinctions are difficult, a differentiation of the qualitative levels of understanding appears useful for identifying shortcomings in current DRR. Systematic research on DRR-related knowledge systems would not only advance our schematic understanding but also provide important insights into the roles of knowledge in DRR.
A point that is significant for DRR efforts and that is addressed in the new SFDRR is the circumstance that knowledge is embedded in a physical object (person or organization) and shaped by perception, experience, and culture. The same applies to knowledge products such as knowledge management databases, information platforms, or lessons learned documents. Therefore, one needs to distinguish between two types of knowledge: explicit and tacit (Polanyi 1967; Nonaka and Takeuchi 1995). Explicit knowledge can be easily processed by a computer, transmitted to others in formal language and electronically, or stored in databases. It is this type of knowledge that current knowledge management practices try to capture and most of the knowledge issues addressed in the SFDRR deal with explicit knowledge. In contrast, tacit knowledge is personal knowledge embedded in individual experience and involves intangible factors, such as personal beliefs, insights, perspectives, and value systems. It is knowledge that is hard to encode, formalize, and articulate with formal language. It is ephemeral, transitory, personal, context-specific, and cannot be resolved into information or itemized in the manner characteristic of information. Before it can be communicated, it must be converted into a form—words, models, or numbers—that can be understood. Moreover, tacit knowledge has two dimensions: a technical (procedural) one, encompassing the kind of informal experiences and skills often captured in the term know-how, and a cognitive one, encompassing beliefs, perceptions, ideals, values, emotions, and mental models. As we will illustrate in the next sections, it is particularly these dimensions of tacit knowledge that need to be better addressed and captured by DRR research and policy.
The Sendai Framework for Disaster Risk Reduction (SFDRR) and Knowledge
The SFDDR takes into account that informed decision making and coordinated action require reliable knowledge. The Sendai Framework’s implementation is guided by several principles, and Paragraph 19 directly refers to knowledge: “Disaster risk reduction requires a multi-hazard approach and inclusive risk-informed decision-making based on the open exchange and dissemination of disaggregated data, including by sex, age and disability, as well as on the easily accessible, up-to-date, comprehensible, science-based, non-sensitive risk information, complemented by traditional knowledge” (UNISDR 2015b, p. 9). While this statement seems straightforward at first glance, the problem of who should collect, disaggregate, and disseminate the data is less so. Additionally, there are some details on the “how” that require further explanation and discussion, a process that Glantz (2015) accurately referred to as “lessons learned about lessons learned”.
The SFDRR sets four priorities for action: (1) understanding disaster risk; (2) strengthening disaster risk governance to manage disaster risk; (3) investing in disaster risk reduction for resilience; and (4) enhancing disaster preparedness for effective response, and to “Build Back Better” in recovery, rehabilitation, and reconstruction. Particularly Priority 1 relates to issues of knowledge, listing 23 requirements that are directly or indirectly linked to information and knowledge (UNISDR 2015b). For instance, point (h) advises to “promote and improve dialogue and cooperation among scientific and technological communities, other relevant stakeholders and policymakers in order to facilitate a science-policy interface for effective decision-making in disaster risk management” (UNISDR 2015b, p. 11). An effective implementation of this advice, however, requires a certain understanding of knowledge production processes, of the existence of different types of knowledge, and of the causes hindering the transfer and use of information. Therefore, it would be useful to present potential means and provide actual opportunities for bridging gaps between bottom-up and top-down actions, between local and scientific knowledge, and between issue domains such as DRR and climate change adaptation. Research in this direction exists and propositions have been made (Kasperson and Berberian 2011; Gaillard and Mercer 2013; Kelman et al. 2015).
The SFDRR (UNISDR 2015b, p. 9) points to the importance of promoting “the collection, analysis, management, and use of relevant data and practical information” at national and local levels, as well as to “ensure its dissemination, taking into account the needs of different categories of users”. This is reasonable since many countries do not systematically collect disaster-related facts, data, and information. Depending on the agency or institution, the collection ranges from hazard type to risk exposure and disaster damage. Thus, knowledge is scattered among various actors and arenas with limited coherence, coordination, and sharing. The existence of a national web site that displays disaster-related data is not evidence for the existence of a national disaster information system. Little information is available on the extent to which households, businesses, and government institutions from outside the sector visit these web sites or whether the information available is actionable (UNISDR 2015a).
More importantly, there is hardly any reassessment and evaluation of collected and used data and information. Learning includes the processes of generating, acquiring, and sharing knowledge, as well as incorporating the newly acquired knowledge into future activities. Especially after disaster occurrence, it would be appropriate to reconsider existing data, information, and knowledge, preferably within larger spatial and temporal scales to capture feedback loops (López-Peláez and Pigeon 2011).
While closer cooperation between academics and practitioners in making data available for research purposes is desirable, the common practice is that datasets are not shared but guarded by secrecy and nondisclosure agreements (Milton 2014). Even when datasets are freely accessible, they often remain empirical, unstructured, and meaningless facts. As a result, although risk information is being generated and disseminated on a large scale, we do not know how far it reaches and whether it changes risk perceptions and awareness levels (UNISDR 2015a). In the DRR domain, a drawback is the lack of agreed standards and clearly defined responsibilities and accountabilities in knowledge management.
According to Senge (1990, p. 19), learning organizations are “organizations in which people continually expand their capacity to create the results they truly desire, where new and expansive patterns of thinking are nurtured, where collective aspiration is set free, and where people are continually learning to see the whole together”. The basic rationale for such organizations is that in situations of rapid change only those organizations that are flexible, adaptive, and productive will excel. For this to happen, Senge argues, organizations need to discover how to tap people’s commitment and capacity to learn at all levels. Such a view is in contrast to many of the recommendations outlined in the SFDRR. Top-down activities that lack incentives and possibilities for integrating diverse societal actors will inevitably disregard valuable experience and expertise.
The 2002 Elbe River floods in Germany and Hurricane Katrina in 2005 in the United States are prominent examples illustrating that for DRR-related organizations effective policy implementation is as important as making the right policy. The river floods and the hurricane not only exposed the vulnerability of highly developed countries to natural hazards but also disclosed deficiencies in social learning and applying knowledge. In both countries, the investigation of DRR and response processes characterized the responses to the two events as “a failure of initiative” rather than “a failure of knowledge” (Weichselgartner and Brévière 2011). But there is, of course, a nexus between the two. Both knowledge and initiative require information and a coordinated process for sharing it. The disasters highlighted shortcomings in effectively transferring organized information into applied knowledge, that is, knowledge into wisdom. Katrina was known to be headed to New Orleans several days before the hurricane made landfall, and the potential damage had been known and understood years ahead of time (Select Bipartisan Committee to Investigate the Preparation for and Response to Hurricane Katrina 2006). Yet the socioeconomic and political impacts of the hurricane were enormous. In April 2006, the Elbe River again caused severe damage—despite a comprehensive lessons-learned process initiated shortly after the Elbe River floods in 2002 (DKKV 2004). The failure to mitigate and respond more effectively to natural hazards—which in both cases had been predicted in theory for many years, and forecast with startling accuracy for days ahead of time—underlines the importance of addressing the distinct domains of disaster risk reduction, knowledge management, and social learning together.
While the SFDRR addresses both the creation and dissemination of knowledge through various recommended activities, the analysis of available and used knowledge is hardly mentioned. However, an evaluated and reflected understanding based on a lessons-identified approach is critical to further improving DRR towards applied knowledge. Organizations in the DRR domain must increase their efforts to identify lessons learned and move from single-loop to double- and triple-loop learning (Argyris and Schön 1978). Single-loop learning refers to an incremental improvement of action strategies without questioning the underlying assumptions. Given or chosen goals, plans, strategies, and rules are operationalized rather than questioned. Double-loop learning refers to a revisiting of assumptions—cause-effect relationships are a good example from the DRR domain—and to questioning the governing variables themselves by means of a critical reexamination. In triple-loop learning, one begins to reconsider underlying values, beliefs, and worldviews. Such learning may then lead to an alteration in the governing variables and a structural change or a shift in the way in which strategies are framed. Researchers have developed frameworks that can guide analysis of how multilevel and multiloop learning processes influence the dynamics of factors underlying the adaptive capacity of governance and management systems, for instance regarding natural resources and floods (Pahl-Wostl 2009; Pahl-Wostl et al. 2013).
The SFDRR points at important shortcomings in the DRR domain, above all how little integration of knowledge systems occurs at community, regional, and national levels. The same is true of the institutional level. Closer collaboration between the different organizations working in DRR would improve the quality and utilization of their “knowledge products”. Closer cooperative efforts with organizations from related knowledge domains such as climate change would harness additional expertise (Kelman 2015). More incentives and political backing for knowledge sharing are needed. Hardly any resources are committed to specific efforts to improve knowledge management in DRR. Collected data and information are usually not organized for different audiences and translated into different languages. Issues of power and competition at institutional and administrative levels can severely hinder the sharing of data and information. In the following sections, we portray the impacts of storm Xynthia on the Atlantic coast of France to illustrate some of the challenges the implementation of the SFDRR faces, as well as the practical efforts of the French National Observatory for National Risks that are aimed at reducing some of the existing shortcomings.