Machine-assisted big data (MABD) research is enabling quantitative studies of large-scale social phenomena, including societal responses to climate change. The rise of MABD science is causing both enthusiasm and concerns. Reviewing prominent criticisms of MABD and their relevance for MABD explorations of macro-structural factors shaping media coverage of climate change, this article finds that the quality and contributions of such studies depend on avoiding common pitfalls. The review focuses specifically on MABD studies attempts to identify and make sense of correlations-or lack thereof-between climate vulnerability and climate coverage in different countries. The review draws on insights from a single, nationally focused, context-attentive, and relatively more qualitative "small data" study in the Global South (Brazil) to shed critical light on assumptions, claims, and policy recommendations made based on the computer-assisted macro-studies. The review illustrates why more narrowly focused and qualitative small data studies are complementary and indispensable. Besides providing vital understanding of causal relationships that elude MABD studies, more narrowly focused and context-sensitive qualitative studies can foster understanding of the consequential mediating roles of place-specific meaning-making and political strategizing in how climate and weather phenomena are framed by social actors and mass media in particular places. These are dimensions that escape the Big Data quantitative methods, but that are vital to sound policy advice, as illustrated by the Small Data research from Brazil. This article is categorized under: Social Status of Climate Change Knowledge > Knowledge and Practice

Climate change certainly shapes weather events. However, describing climate and weather as the cause of disasters can be misleading, since disasters are caused by pre-existing fragilities and inequalities on the ground. Analytic frames that attribute disaster to climate can divert attention from these place-based vulnerabilities and their socio-political causes. Thus, while politicians may want to blame crises on climate change, members of the public may prefer to hold government accountable for inadequate investments in flood or drought prevention and precarious living conditions. To be both strategic and moral, framing choices must therefore be sensitive to context-dependent political meanings and particularities, and to how the values implicit within analytic frames about the causes of disasters shape policy responses. Such sensitivity requires multicausal analysis of weather-linked disasters to illuminate a broader range of means to reduce the damages associated with climate change and weather extremes. Through examples from around the world, and especially Brazil, we discuss how and why climate-centric disaster framing can erase from view-and, thus, from policy agendas-the very socio-economic and political factors that most centrally cause vulnerability and suffering in weather extremes and disasters. We also offer a theoretical discussion of why attribution is not neutral. Analytic frameworks always embed choices about factors that matter, and thus are inherently normative and consequential for understandings of responsibility and action. This article is categorized under: Social Status of Climate Change Knowledge > Climate Science and Decision Making Highlight Attributing crises only to climate is inadequate from a mechanical, moral, and strategic policy points of view.

Collaboration between public administration organizations and various stakeholders is often prescribed as a potential solution to the current complex problems of governance, such as climate change. According to the Advocacy Coalition Framework, shared beliefs are one of the most important drivers of collaboration. However, studies investigating the role of beliefs in collaboration show mixed results. Some argue that similarity of general normative and empirical policy beliefs elicits collaboration, while others focus on beliefs concerning policy instruments. Proposing a new divisive beliefs hypothesis, we suggest that agreeing on those beliefs over which there is substantial disagreement in the policy subsystem is what matters for collaboration. Testing our hypotheses using policy network analysis and data on climate policy subsystems in 11 countries (Australia, Brazil, the Czech Republic, Germany, Finland, Ireland, Japan, Korea, Portugal, Sweden, and Taiwan), we find belief similarity to be a stronger predictor of collaboration when the focus is divisive beliefs rather than normative and empirical policy beliefs or beliefs concerning policy instruments. This knowledge can be useful for managing collaborative governance networks because it helps to identify potential competing coalitions and to broker compromises between them.

According to a recent article in this journal, the failure of policy action on climate change despite scientific consensus points to a broken science-society contract. To avoid this tragedy of climate science, the authors call for a moratorium on its production. As scholars of, and participants in, global science-policy interfaces, we recognize the authors assumptions and reasonings but also see an urgent need for a deeper understanding of the current limitations of environmental research, and the challenges of connecting knowledge to policy and society. Rather than a blanket moratorium, we argue that what is needed is a profound transformation of environmental research. This entails a shift in research priorities towards currently marginalized approaches in social sciences, humanities and participatory research, to generate a much-needed understanding of obstacles to action and just and equitable strategies for overcoming them with due consideration of issues of justice and equity. We also propose a new science-society contract that recognizes the politics of environmental knowledge. This is necessary to enable critical reflection on what interests environmental research serves whose knowledge needs are excluded, and with what consequences. We recognize that our proposal can be uncomfortable and that it challenges deeply held beliefs in the neutrality of science. However, deep reprioritization in environmental science and science policy are urgently needed to strengthen the contribution of environmental research to the transformative changes that it calls for.