I. Introduction
The judicial evaluation of complex and evolving climate risks lies at the core of climate litigation, and existing research on local climate governance shows that public participation may lead to more ambitious local climate policies.Footnote 1 One could therefore expect that plural perspectives on risk, generated through community participation, may also lead to court decisions that are more responsive and better aligned with public needs. However, while previous studies on climate litigation have examined how scientific data (the Intergovernmental Panel on Climate Change – IPCC reports being the most widely known) can influence climate judgments,Footnote 2 less is known about whether and how courts incorporate other forms of extralegal knowledge in their decisions and the broader consequences of this incorporation, despite their considerable relevance to understanding impacts on affected communities.Footnote 3 As put by Sanz et al., while emissions projections are essential for assessing impacts also at the local scale, knowledge of vulnerability is crucial for enabling decisions that reflect the unique needs of local communities.Footnote 4
In this paper, we focus on how judges negotiate and use extralegal knowledges, purposefully used in the plural,Footnote 5 including, for instance, qualitative studies on climate risks and public knowledges, such as the risk perceptions of affected communities, in climate litigation across two sociolegal contexts with contrasting constitutional traditions and evidentiary procedures: Brazil and the Netherlands (more specifically, the Dutch Caribbean). One substantial difference in this regard concerns how extralegal knowledge is brought into legal procedures, which may shape the types of evidence considered and the judicial approach to incorporating societal voices in climate risk governance. While Brazilian constitutional litigation routinely admits civil society as amici curiae with the right to submit and present evidence orally, in the Dutch tradition, extralegal knowledge generally reaches the judge through the parties. This contrast speaks directly to our research question on how plural extralegal knowledges shape judicial risk work. One can hypothesise, for example, that ample societal access to courts may reflect judges’ access to a broader array of extralegal knowledge or that by considering evidence brought by amici curiae, rather than exclusively by petitioners and defendants or having to check evidence that is in the public domain, judges may feel more socially legitimised to translate it into judicial interventions.
In doing so, we aim to explore not only how different risk perspectives shape judicial interventions in each climate case, but also the interactional dynamics among the parties involved in litigation. Ultimately, the findings demonstrate how courts across contrasting sociolegal contexts are reframing climate risk by engaging with legal and extralegal knowledges. As we will see in the cases examined in this paper, in their assessments, judges deal with extralegal knowledges ranging from scientific evidence to social-scientific data on the impacts of climate change on vulnerable communities. These different sources of extralegal knowledge, stemming from different epistemic domains (e.g. techno-scientific, social-scientific), present judges with multiple risk perspectives – sociotechnical, cultural-symbolic – to be interpreted, negotiated and translated into legal decisions. In this interaction, extralegal and legal knowledges must be articulated as judges deliberate on whether risks to communities and ecosystems are being aggravated by governments’ greenhouse gas (GHG) emissions, and which interventions are necessary (or not) to remedy, adapt, or prevent harm.
In this scenario, in addition to interpreting one highly legitimised form of extra-legal knowledge (e.g., climate science data) and the risk relations it delineates, judges must also negotiate techno-scientific knowledge with other stakeholders’ perspectives of risk, including affected communities and what they understand to be at risk. Through the incorporation (or not) of these perspectives, judges’ risk work Footnote 6 becomes particularly visible in climate litigation decisions. This concept, inspired by sociocultural approaches to the sociology of risk, emphasises ‘working practices to assess or manage risk’Footnote 7 as dependent on multiple knowledges, aimed at interventions, and embedded in social relations. Such a perspective works as a helpful lens to illustrate that, rather than relying solely on techno-scientific notions of risk, courts play a role in configuring understandings of climate risk through epistemic plurality.
In this paper, we engage with risk work as a sensitising conceptFootnote 8 to elucidate the role of courts in managing climate risk, not to test or extend the framework itself, but to structure our analytical attention toward three features of judicial practice. We frame: (1) judicial decision-making as a social process, to analyse how social-scientific and techno-scientific evidence are reworked in legal contexts as society calls courts to distribute responsibility for managing climate change risks; and (2) judges as professionals who must interpret and negotiate different notions of risk in translating them into legal decisions, demystifying them as passive science applicators and their decision-making as mere applications of law. The framework thus offers a productive lens for making visible dimensions of climate litigation that purely doctrinal or science-focused accounts tend to obscure.
In this regard, we address the following research question:
How have courts in two high-profile climate cases decided in Brazil and the Netherlands incorporated a plurality of extralegal knowledges into their decisions, and what can the comparison tell us about the ways in which courts are (re)configuring climate-related risks?
II. Courts as epistemic sites, judges’ decision-making as risk work
Over the past decades, climate science has identified human influence as a main and “unequivocal” cause of global warming that affects “everyone, everywhere.”Footnote 9 But, like any other science, climate science is characterised by epistemic uncertainties, particularly regarding timing, magnitude and distribution of impacts.Footnote 10 Still, these uncertainties have not halted courts’ engagement with climate science knowledge.Footnote 11 Climate cases such as Urgenda v. the State of the Netherlands (2019), Milieudefensie v. Shell (2021) and Klimaseniorinnen v. Switzerland (2024) are landmark decisions in which climate science data has informed courts in attributing responsibilitiesFootnote 12 for managing climate change risks, highlighting how science, law and public participation become intertwined.
The idea that courts are institutional epistemic sites has been widely explored in Science and Technology Studies (STS),Footnote 13 particularly in examining legal processes in which courts, amid institutional constraints,Footnote 14 engage with scientific evidence (or refuse to) to produce legal decisions, where scientific knowledge is translated, negotiated, transformed, stabilised and sometimes contested in decision-making. This engagement requires courts to articulate legal actors’ understandings, scientific knowledge and normative judgments of climate risk in the face of epistemic uncertainties.
While a techno-scientific notion of risk, technical and quantifiable, tends to be framed by many as the main legitimate source of risk knowledge, “science does not have a monopoly on value judgments” and cannot respond alone to these questions in climate change governance.Footnote 15 Climate litigation involves other stakeholders’ perspectives on risk in making this value judgment, thereby making evident the interwovenness of science and social order in producing actionable meanings, such as responsibility for climate risk management, a notion that cannot be taken for granted.Footnote 16
Disputes between climate scientistsFootnote 17 and critiques of the hegemony of techno-scientific forms of climate knowledgeFootnote 18 for understanding and distributing climate risk highlight the relational Footnote 19 and dynamic nature of risk and its social construction. While all risk definitions tend to include “an object deemed to ‘pose’ the risk, a putative harm, and a linkage alleging some form of causation between the object and the harm,”Footnote 20 understandings of which objects are harmful (“risk object”) and which are being harmed (“objects at risk”) may vary according to the value attributed to them,Footnote 21 illuminating how notions of risk are embedded in sociocultural, political and historical contexts and may vary over time.
The conceptualisation of “risk work”Footnote 22 with which this paper engages has been proposed in relation to healthcare professionals’ work in everyday interactions, especially in risk interventions (e.g., diagnosing and prescribing treatment), where they must negotiate probabilistic and experiential risk knowledges (e.g., epidemiological data and patients’ risk perceptions). In seeking to explain “how risk gets translated into different contexts,”Footnote 23 it focuses on how risk interventions are situated within social relations and shaped by multiple knowledges, professionals’ identities, and the broader social context of their work. As such, risk assessment is never separated from the ways in which risks are constructed, communicated and managed, which makes the concept particularly relevant for understanding courts’ role in climate governance.
Similar to healthcare risk interventions, courts’ role in managing climate change risk governance is inevitably situated within interactional dynamics (e.g., claimants, defendants, judges) and informed by risk knowledges (e.g., techno-scientific, social-scientific, and normative). In our analysis, we focus on three interwoven core features of risk work – risk knowledges, risk interventions and social relations – and the tensions emerging from this interwovenness. Unlike healthcare workers, however, judges have double roles – as both (legal) experts and (climate science) lay people – in negotiating these different risk knowledges to make climate-related decisions.
In terms of risk interventions related to climate litigation, courts’ choices of injunctions for adaptation, mitigation or structural remedies reflect interpretations of different climate knowledges.Footnote 24 The types of interventions chosen, in turn, indicate the ways in which courts are distributing responsibility and, consequently, reorganising social relations with and among the parties, as well as with their imagined audience (e.g., society, scientists, the market). In the following section, we describe our operationalisation of this theoretical framework.
III. Case studies: methodological remarks
1. Selection of case studies
In our document-based research, we examined legal materials from two recent high-profile climate cases in the selected jurisdictions that attracted broad public attention for their judicial intervention in response to comprehensive governmental failures in climate protection. Namely, the cases of Greenpeace Netherlands and 8 citizens of Bonaire v. The Netherlands, decided on 28 January 2026 by the Hague District Court in the NetherlandsFootnote 25 and PSB et al. v. Brazil (on deforestation and human rights), ADPF 760,Footnote 26 decided on 14 March 2024 by the Brazilian Supreme Court and, as of the date of writing (February 2026), subject to judicial oversight.
The purposive sample is analytically justified by the social relevance of the cases, their potential to affect vulnerable communities (namely, the Amazonian Indigenous peoples and the Bonairian islanders), and, therefore, the opportunity to dialogue with socio-legal knowledge about risks to those communities. We acknowledge that the cases also differ along other dimensions, notably court level (apex constitutional court v. first-instance court), claim type (constitutional rights v. tort and ECHR), and scope (national deforestation and climate policy v. mitigation and adaptation in a post-colonial territory). Rather than treating these as confounders to be controlled out, we approach them as part of the sociolegal context that shapes judicial engagement with risk, and we remain explicit about their interpretive weight throughout the analysis. The Brazil–Netherlands pairing was also preferred over other landmark rulings (e.g., Germany’s Neubauer, Colombia’s Future Generations) for being more recent, reflecting the latest state of climate litigation.
a. Greenpeace Netherlands and 8 citizens of Bonaire v. The Netherlands
The Dutch case represents the most recent climate development in the Kingdom of the Netherlands, following the widely influential cases of Urgenda Foundation v. State of the Netherlands and Verein KlimaSeniorinnen Schweiz and Others v. Switzerland (ECtHR), and was described as “another milestone in the ongoing struggle to strengthen Dutch climate targets.”Footnote 27 The dispute was initiated by the NGO Greenpeace Netherlands and 8 citizens of Bonaire,Footnote 28 who argued that the Dutch State has failed to adequately protect the island of Bonaire, a municipality of the Netherlands located in the Caribbean, from the impacts of climate change. The petitioners sought both stronger national mitigation efforts and concrete adaptation measures tailored to the island’s particular vulnerability to climate risks.
The claim combined three core arguments. First, the plaintiffs contended that insufficient GHG reduction policies violate the State’s duty of care under Dutch tort law and its positive obligations under Articles 2 and 8 of the European Convention on Human Rights (ECHR). Second, they argued that the Netherlands has failed to adopt adequate climate adaptation measures for Bonaire, despite the island’s special vulnerability to sea-level rise, heat and water scarcity. Finally, but perhaps the most impactful in terms of human rights violations, the petitioners claimed that this lack of protection amounts to discriminatory treatment, as residents of Bonaire receive significantly less climate protection than those living in the European Netherlands. As noted by legal scholars, the case also exposes the colonial legacy in the climate adaptation debate, revealing that “climate harm is historically sedimented.”Footnote 29 The case represents, therefore, “not only a landmark decision on climate protection, but also a reminder that today’s climate vulnerability in the Caribbean can be traced, in part, to the colonial past.”Footnote 30
b. PSB et al. v. Brazil (on deforestation and human rights)
The Brazilian case is a constitutional climate lawsuit initiated by seven political parties and 10 civil society organisations introduced as specialist entities requesting admission as amici curiae before Brazil’s Supreme Court, challenging the federal government’s failure to control deforestation in the Amazon, a key driver of Brazil’s GHG emissions. The applicants claimed that the government deliberately weakened and failed to implement the Action Plan for the Prevention and Control of Deforestation in the Amazon (PPCDAm), which had previously been central to reducing deforestation and meeting Brazil’s climate commitments. Because deforestation releases vast amounts of carbon and undermines the Amazon’s role as a global climate regulator, the claim framed this policy failure as a violation of constitutional rights, including the right to an ecologically balanced environment and related rights to life and health. The petitioners, therefore, asked the Court to require the government to effectively implement existing anti-deforestation policies in order to curb emissions and address the climate emergency.
2. Document selection
The key document included for each case was the main judgment. We considered: for the Dutch case, a 64-page court decision; for the Brazilian case, an 867-page document containing the votes of 11 judges that together constitute the court decision. Whenever available, we also consulted submissions by the parties and civil society admitted to the process, as well as the accompanying documents, given the role this information played in the decisions.
For the Brazilian case, in addition to the decision, we also included all documents in which actors requested admission to participate in the process (the initial claim brought by the petitioners and petitions requesting admission as amici curiae), together with the decisions that assessed who could participate. These 7 documents together totalled 220 pages.
For the Dutch case, at the time of document selection, on February 2026, some additional documents from the case file were available at the Climate Litigation Database managed by the Sabin Center for Climate Change Law at Columbia Law School, such as petitioners’ summons, a decision on admissibility, and the plaintiff’s reply and change of plea,Footnote 31 without a translation to English offered by the court, totalling 375 pages. Unlike the Brazilian case, the full case record was not publicly available.
3. Analytical approach
We employed a theoretically informed reflexive thematic analysisFootnote 32 to identify three features of risk work, as defined by Brown and Gale (2018): risk knowledges, risk interventions, and social relations, supported by a two-cycle coding procedure, as described by Saldanã.Footnote 33 In the first cycle, we applied descriptive codes to instances of extralegal knowledge cited in the rulings, distinguishing between natural science, statistical, social science, technical and policy & governance knowledge. In the second cycle, we consolidated these initial codes into three higher-order analytic categories: techno-scientific, socio-scientific, and policy & governance, with an additional “multi-domain” category for instances that operated across more than one category. These categories then served as an analytic lens for reflecting on how multiple knowledges reconfigured risk relations in the rulings under analysis.
In our analysis, we started by identifying the stakeholders,Footnote 34 extralegal knowledgesFootnote 35 and their sources in each jurisdiction. Risk knowledges were categorised according to their underlying dominant epistemic logics of knowledge production: (1) techno-scientific (e.g., knowledge about climate processes and climate change effects in the forms of measurement assessment reports, climate attribution science research papers, technical and monitoring reports); (2) social-scientific (e.g., knowledge about the historical, social, cultural and economic causes and consequences of climate change, including qualitative and quantitative perspectives on the production of vulnerability and inequality) and (3) policy/governance (e.g., institutional and normative knowledge pertaining to climate policy such as policy documents, recommendations, and governmental communications). Whenever these sources contained knowledge from more than one domain, we classified them as multi-domain.
Stakeholders were identified from the decision in the Dutch case and, in the Brazilian case, from the petitioners’ initial claim, amicus curiae admission petitions, and the admission decisions. The decisions and accompanying documents in each jurisdiction grounded our analysis of extralegal knowledges, judicial interventions and the social relations among parties, judges and admitted amici curiae. Finally, we explored the tensions among the three features of risk work regarding matters of truth and legitimacy pertaining to knowledges, interventions and relationships between the parties.
IV. Presentation of findings
1. Extralegal climate risk knowledges in the Dutch and Brazilian cases
a. The Dutch case
The Dutch case drew on a wide range of extralegal sources, including natural and social science research, technical consultancy reports, policy assessments and institutional communications (see Table 1).
Extralegal knowledge in Dutch Bonaire and Brazilian climate cases, organised by epistemic domain
Table 1. Long description
The table categorizes extralegal knowledge sources in Dutch Bonaire and Brazilian climate cases by epistemic domain. It has 10 columns: Epistemic domain, Epistemic domain (Social-scientific), Source type, Source name, Source description, University and independent research institutes, Academic literature, Commissioned technical consultancy reports, Government ministries and agencies, International public and civil society bodies, National public bodies, Epistemic domain (Techno-scientific), and Epistemic domain (Polygovernance). The table contains 10 rows with specific entries under each column. Each row provides detailed information about the source type, source name, and source description for various epistemic domains.
This information included not only techno-scientific understandings of climate processes or environmental impacts on the BES islands, but also qualitative or mixed-methods research on the local socio-cultural impacts of climate change. The latter highlighted the particular vulnerability of island and coastal regions’ populations, such as in studies conducted by Wageningen University, the Institute for Environmental Studies (IVM) at the Vrije Universiteit Amsterdam, and the private consultancy firm Witteween+Bos commissioned by the Ministry of Infrastructure and Water Management. By using expert interviews, participatory mapping and social media analysis on the BES islands,Footnote 36 these reports assessed the impacts of climate change not only on Bonairians’ close environment (e.g., agriculture and fisheries, energy and telecom infrastructures, water management) but also on their dietary vulnerabilities, health, recreation and tourism, and tangible and intangible cultural heritage. The court also referred to the Terramar report detailing exchanges on climate risks with the population of Bonaire, which indicated that climate risks were communicated to the public on the island.Footnote 37 The decision framed, therefore, all these elements (close environment, health, culture, etc.) as objects at risk.
These findings represent a shift in how courts engage with extralegal knowledge in climate litigation in the Netherlands, compared to the first Dutch climate case, Urgenda, decided by the Hague District Court in 2015 and ultimately confirmed by the Dutch Supreme Court in 2019. While the three judgments in Urgenda relied exclusively on the IPCC assessment reports and its techno-scientific approaches to climate science in the decision-making process, the Bonaire judgment by the Hague District Court, in 2026, relied on a broader array of knowledges, engaging more explicitly with social-scientific climate risk perspectives, as exemplified by the mentions to socio-cultural impacts of climate change on the livelihood of vulnerable communities in the Dutch island of Bonaire.
While the change may connect with the centrality of claims about inadequate adaptation in the Bonaire case, which assessment is facilitated by concrete data on specific climate impacts on vulnerable communities and their most urgent needs, it also seems to reflect the consolidation of judicial understandings in domestic and international litigation about the urgent need for climate action in order to respect fundamental rights. In other words, once jurisprudence in landmark cases like Urgenda and KlimaSeniorinnen has established evidence-based climate duties from international law, new climate decisions can more easily expand the scope of evidence considered beyond the mitigation standards “accepted” and “deemed necessary” by science and the international climate policy, often referred to in the Urgenda judgment.
b. The Brazilian case
The Brazilian decision contained extensive empirical information on deforestation dynamics, including detailed statistical data on forest loss rates (often disaggregated to show impacts within Indigenous territories) and on the State of deforestation policy. This evidentiary basis allowed the Court to establish both the scale of environmental degradation and its disproportionate incidence on protected and traditional territories. However, as shown in Table 1, institutionally produced techno-scientific knowledge, such as that resulting from satellite monitoring, statistical data analysis and policy performance indicators, was privileged in the decision-making process.
In this regard, information on annual deforestation rates from the National Institute for Space Research (INPE), the Amazon Deforestation Calculation Program (PRODES), and the real-time deforestation alert system (DETER) was the most cited, mentioned by 9 of the 11 Supreme Court justices. Together with institutional and statistical data about the anti-deforestation governance in Brazil, it formed the core factual cluster shared by most justices, regardless of how extensively they engaged with other knowledge. Other sources, including scientific literature, social science studies and international assessments, were eventually introduced by some of the Justices. These included, for example, studies on the fiscal and economic drivers of deforestation; links between illegal deforestation, organised crime, land grabbing and fraudulent authorisations; weak land rights fuelling deforestation increasing inequality and reducing economic opportunities for local communities; psychological impacts of climate change on young people, including anxiety and depression; and qualitative analysis of indigenous land management practices and their contribution to environmental conservation. Some sources also appeared in only one opinion, suggesting that, beyond the shared evidentiary foundation, each justice constructed their own epistemic narrative across the case, with some opinions relying primarily on legal sources, while the rapporteur and the second vote contain the broadest range of epistemic sources and domains.
c. Differences in the consideration of extralegal knowledge
While both jurisdictions articulated techno- and social-scientific knowledges in their decision-making processes, each court emphasised different kinds of knowledge. The Bonaire decision, for example, showed a pronounced engagement with social-scientific data on climate risks to the affected population’s ways of life and to their immaterial heritage, such as cultural practices. It also engaged with the population’s perceptions of climate threats through the Terramar report, which included output from a workshop organised with Bonaire inhabitants. In contrast, the engagement with social-scientific knowledge in the Brazilian decision seemed to focus on academic studies about socio-economic drivers of deforestation and the role of indigenous peoples in protecting the Amazon forest, rather than articulating the vulnerability of specific communities through specific data about local material and immaterial harms, or through the experiences of affected populations.
At first sight, the limited engagement with studies or testimonies on specific impacts to local communities can appear counterintuitive in the Brazilian case, given the process’s broad participatory nature. That involved multiple civil society organisations acting as amici curiae and introducing documents that portrayed how vulnerable groups experience the impacts of climate change, such as children’s letters and drawings. At the same time, it is not possible to ignore that the differences in epistemic pluralism can relate to the scope of each case. Accordingly, the Dutch case concerned mitigation and adaptation policies in a geographically and historically specific part of the Dutch territory, i.e., an island and former colony, which very directly connects to the effects suffered by the local populations. The Brazilian case, on the other hand, primarily concerned the broader national issue of the climate consequences of deforestation. The references to indigenous communities were made, therefore, not to show their vulnerability, but to demonstrate that their practices and ways of life can contribute to forest preservation.
2. Climate risk interventions in the Dutch and Brazilian casesFootnote 38
a. The Dutch case
The Dutch Bonaire case illustrates a relatively concise remedial structure combining rights-based declarations with forward-looking injunctions. The court first declared that the State had violated Article 8 of the ECHR, and the non-discrimination guarantees of Article 14 of the ECHR and Article 1 of Protocol No. 12, recognising that residents of Bonaire were exposed to climate risks without equivalent protection. It then ordered two forms of governmental action. The first was an adaptation injunction, requiring the State to prepare and implement a national climate adaptation plan that explicitly includes Bonaire and aligns with the United Arab Emirates Framework for Global Climate Resilience, with targets to be met by 2030. The second was a mitigation injunction, directing the State to incorporate economy-wide emission reduction targets into national legislation, consistent with Article 4(1) of the Paris Agreement and subsequent international commitments. Therefore, the interventions established legal obligations, while largely leaving the design and implementation of climate policy to the political branches.
In deciding on these interventions, judges relied on a combination of legal and extralegal knowledge to interpret existing domestic and international norms and articulate them in their assessment.Footnote 39 On the legal side, it relied heavily on existing international and domestic case law in the judgment’s assessment sections, notably the ECtHR KlimaSeniorinnen case. The latter was mentioned 46 times in the footnotes and 43 times in the footnotes of the assessment part. The ICJ Advisory Opinion was mentioned 21 times in the footnotes and 18 times in the assessment footnotes. The Advisory Opinion of the IACHR was mentioned 3 times in total, and 2 times in the assessment. The Urgenda ruling was mentioned 10 times in the footnotes and 7 times in the assessment footnotes.
One interesting aspect of the translation of extralegal knowledge into legal terms in this case concerns the use of the KlimaSeniorinnen decision as an assessment framework for mitigation and adaptation claims,Footnote 40 as well as its articulation with local evidence. By clarifying for the first time that the application of Articles 2 and 8 of the ECHR in climate cases includes a positive obligation to adopt and implement a credible pathway to climate neutrality with evidence-based, binding emissions targets, monitoring and updates, international case law played a key role in legitimising the consideration by a domestic court of extralegal knowledge in the judicial interpretation of risk. In this case, the judicial consideration of a study on climate-related health risks in the Caribbean sought to establish the sufficient plausibility of the claim that climate change already poses an increased risk of illness and mortality for certain groups in the Caribbean region (which the State also did not dispute).Footnote 41 This evidence-based reasoning supported the court’s conclusion that the violation of Article 8 of the ECHR was possible.Footnote 42
Other examples of the articulation of extralegal and legal risk knowledge in deciding on which interventions would be suitable in this case were references to: the IVM studies showing cultural-heritage risks due to the concentration of a large part of the buildings and Bonairian cultural heritage in low-lying areas of the island, which are likely to be flooded, to support the conclusion that the State had failed to fulfil its duty of care to adopt timely climate-adaptation measures for Bonaire to reduce the vulnerability of its inhabitants to climate change,Footnote 43 and to studies and knowledge exchange platforms commissioned by the Dutch governmentFootnote 44 to support the conclusion that further court-ordered research into the historical, current and future consequences of climate change on Bonaire was unnecessary.
b. The Brazilian case
The Brazilian case, PSB et al. v. Brazil, concerning deforestation in the Amazon, followed a different approach. Rather than a single judgment, the litigation unfolded through a sequence of judicial votes and decisions that progressively shaped a structural remedial framework. An early opinion by Justice Cármen Lúcia proposed declaring an “unconstitutional state of affairs” in Amazon protection and ordering the government to produce a detailed implementation plan for the federal anti-deforestation policy, including quantified deforestation limits, measurable indicators, institutional reforms, budget allocations, and regular reporting. The Supreme Federal Court, by a majority, ultimately declined to adopt the unconstitutional-state-of-affairs declaration, reflecting tensions in the votes over how to translate the wide array of evidence into a legal decision.
While the Justices did not contest the evidence demonstrating structural and comprehensive failures in the national forest conservation policy, a disagreement arose over whether the court should translate the factual scenario into a legal declaration of an unconstitutional state of affairs. In this regard, Justice André Mendonça considered the potentially difficult-to-predict adverse international and political macro-consequences of such a declaration, including diplomatic pressure, comparative disadvantages or obstacles to multilateral negotiations, and political backlash regarding judicial overreach. According to the Justice, such unintentional effects could, in turn, be addressed by adopting a more dialogical, experimental and managerial approach through a structural injunction, inspired by the South African doctrine of “meaningful engagement.”Footnote 45
In this regard, the majority of opinions adopted the core idea of a structural injunction requiring the federal government to present and implement a comprehensive policy plan. Therefore, the final decision acknowledged structural failures in Amazon governance and ordered the State to engage meaningfully with the policy framework, supported by reporting duties, fiscal safeguards, and institutional coordination. What followed was not the end of the case but the beginning of an ongoing judicial monitoring procedural stage. The court issued a series of follow-up orders extending deadlines, convening contextualisation hearings, requesting technical support from the Supreme Court’s Complex Structural Litigation Unit (NUPEC), partially approving the government’s plan, and assessing its implementation over time.
c. Differences in judicial interventions
While both cases articulated legal and extralegal knowledges in ordering their interventions, the judicial roles in each intervention differed. Compared with the Dutch case, the Brazilian litigation reflects a far more iterative and supervisory judicial role, in which the court remains actively involved in overseeing compliance. While the Bonaire judgment relied on declaratory findings and framework injunctions that guide legislative and executive action, PSB v. Brazil illustrates a form of structural climate governance through litigation, where judicial intervention unfolds through continuous dialogue with governmental institutions and sustained monitoring of policy implementation.
A distinctive feature of the judicial intervention in the Brazilian case relative to the Dutch case is its stronger dialogical element, in which the Supreme Court Justices asked the government to elaborate a protection plan and are monitoring its implementation together with social actors in the process. Such an iterative, flexible and participatory approach may bring several advantages from a socio-legal perspective, such as making decisions more responsive to human needs and ESCR realisation and resulting in gains in judicial legitimacy, both from a procedural (judgments more open, inclusive, and transparent) and societal perspective (higher acceptance of judicial decisions). More collaborative decisions may also mitigate separation-of-powers judicial concerns or criticism in this respect, as courts will not impose climate policies on governments but will enable the creation of policies that respect fundamental rights. In this vein, Justice André Mendonça’s opinion expressed concern about political backlash, proposing the use of dialogical structural injunctions, as well as the South African doctrine of “meaningful engagement” to address criticism regarding the separation of powers and the judicial technical capacity to formulate public policies.
In contrast, the remedies in the Bonaire judgment did not include an oversight judicial phase to monitor the implementation of the requested orders. It is noteworthy, however, how the Dutch decision also shows a more dialogical feature than previous climate decisions in the pioneer climate Dutch case of Urgenda, by ordering the State to come up with a plan under certain requirements, rather than unilaterally ordering a specific injunction that, according to some critiques, turned a broad duty of care into international human rights norms into a specific standard of care. It was, therefore, praised as an example of remedial innovation in the Dutch context.Footnote 46
3. Social relations in the Dutch and Brazilian casesFootnote 47
a. The Dutch case
Unlike the Brazilian case, in the Netherlands, third parties generally cannot request admission as amici curiae. Courts may invite or allow third parties to submit observations only in specific cases, particularly before the highest administrative courts or the Supreme Court, and participation is limited to written submissions.Footnote 48 Despite the procedural limitation, the Bonaire judgment noted that three residents of Bonaire were allowed to make statements during oral proceedings on 7 and 8 October 2025, which were added to the case file. This process suggests an indirect, institutionally mediated judicial engagement with civil society by privileging evidence sources such as reports from universities, consultancy firms, and (inter)governmental organisations.
Despite not directly engaging with the public in the case, the court acknowledged the need for participatory governance procedures. The judgment mentioned, for instance, that “When taking adaptation measures, member states should follow a country-driven, participatory and fully transparent approach, taking into account vulnerable groups, communities and ecosystems. Adaptation measures should be based on the best available scientific knowledge, indigenous peoples’ knowledge and local knowledge systems to ensure that adaptation is properly integrated into relevant socio-economic and environmental policies and measures.” Footnote 49
Therefore, although the Dutch court did not directly engage with Bonaire residents in public hearings, it considered whether the government had organised societal exchanges to promote knowledge and awareness of climate change challenges. The Terramar workshop report explains, in this regard, that:
A workshop on 20 October 2023 was jointly organized with Islanders at the Helm, TERRAMAR, stichting Climate Adaptation Services (CAS), KNMI, and WUR to discuss and prioritize climate impacts. This workshop was part of the Trans-Atlantic Platform (TAP) developed within Islanders at the Helm to promote knowledge and awareness of climate challenges through courses, lectures and workshops. For the workshop, 63 contacts were invited from government, nature organizations, social-cultural and economic sectors. Presentations were given about climate scenarios, the climate impact atlas, and adaptation approaches.
In this context, the court assessed whether community participation was present in climate governance, checking whether it included the exchange of knowledge with society and was based on scientific, indigenous peoples’ and local knowledge to ensure the development of appropriate adaptation policies. The decision considered, for example, the actual implementation of a participation mechanism in climate governance informed by extralegal knowledge. In particular, the establishment of a climate table group to draft a climate adaptation plan for the BES islands, including representatives from various sectors and governments and led by an independent chair, was considered to reach the conclusion that, from 2023, the Dutch State would be catching up on overdue climate measures and making room for participation and knowledge sharing by residents and local organisations.Footnote 50 The acknowledgement of this duty was relevant to reordering social relations between the State and the local population, ensuring the voice and a place for the latter in climate governance.
b. The Brazilian Case
While a plurality of information was added to the process via civil society organisations admitted as amici curiae,Footnote 51 the analysis of extralegal knowledge in the Justices’ opinions revealed that this openness to public participation did not reflect in a particularly notable consideration of knowledge about the vulnerability of communities affected by deforestation and climate change or the lived experiences of affected communities in the judgment. In this sense, the Brazilian decision exemplifies a model of adjudication in which participation operated primarily at the level of institutional inclusion. The Court acknowledged that certain communities were more affected, but it did not engage with how those communities interpret environmental harm. This reinforces a framing of the dispute that foregrounds structural policy failure, while experiential forms of knowledge remain marginal to the legal analysis. In terms of social relations, the Brazilian decision also put in evidence the court’s legitimacy concerns with its position relative to other branches of government, in a context marked by increasing political tensions and polarisation.
c. Differences in social relations
Despite the court’s indirect engagement with civil society organisations in the Bonaire case, except for the petitioners, by highlighting the need for societal participation and input in governance processes, it acknowledged a new form of relationship between the State and the population. One in which local communities’ perspectives become an unavoidable part of climate governance.
The Brazilian experience points to a related but distinct dynamic: broader participation did not translate into the incorporation of local experiences. Despite the proceedings’ remarkable procedural openness, the justices’ opinions remained vastly anchored in quantitative forms of institutionally produced techno-scientific knowledge. This apparent contradiction may be explained, at least in part, by judges feeling safer addressing legal or institutional sources in a politically sensitive and highly polarised context, where the court faced criticism for opposing policies implemented during Bolsonaro’s government, and justices showed evident concern about political backlash and accusations of overreach.
In the Bonaire case, the incorporation of social-scientific knowledge positioned culture as an object at risk, while the State’s inaction and failure to prevent a discriminatory situation was framed as a risk object. This reconfiguration is notable because it implicated the State in a relational framing of climate risk, and it was not without consequence for social relations. The Bonaire judgment directly produced societal reaction, including public criticism and personal criticism of judges involved in the case.Footnote 52 This suggests that judicially reframing risk in ways that some actors contest can generate significant tensions and conflicts, which are likely to be aggravated in politically polarised contexts.
In the Brazilian case, the object at risk remained, at first glance, more familiar (the Amazon forest), but the risk object that the Court identified was a structural failure of governance, one that demanded a correspondingly structural response. The Court’s engagement with evidence of institutional dismantlement, budget cuts and the perverse incentive structure created by retrospective regularisation mechanisms, such as the amnesty provisions of the 2012 Forest Code, which the Court itself identified as corroborating “the logic of waiting for the State to regularise the illicit,”Footnote 53 reflected, for example, a diagnosis that goes beyond episodic wrongdoing to implicate systemic state conduct.
V. Discussion
Turning to the first part of our research question on how courts in Brazil and the Netherlands have incorporated a plurality of extralegal knowledges into their decisions, the analysis shows that both courts articulated techno-scientific and social-scientific knowledge. Courts engaged with the latter either by identifying the vulnerability of local communities to climate change, as in the Bonaire case, or the relevance of indigenous practices in forest conservation, as in the Brazilian case. Importantly, the findings reveal that in both cases, techno- and social-scientific knowledges were treated as complementary rather than opposites in the assessment part of the judgment or in the justices’ opinions. The Dutch case illustrates this particularly well, as accounting for the vulnerability of the population of Dutch Caribbean islands to climate impacts helped the court to address the unequal treatment resulting from the colonial legacy. This openness to social-scientific perspectives may be related both to the nature of the case, which centred on adaptation measures in a former Dutch colony, and to the advancement of international case law, which acknowledges that climate policies should be evidence-based.
The second part of the research question, concerning what the comparison can tell us about the role of courts in (re)configuring climate-related risks, yields several interconnected findings. Our comparative analysis reveals that the articulation of techno- and social-scientific extralegal knowledge by the courts, beyond informing judicial reasoning, reconfigured relations of risk, identifying new “objects at risk” and new “risk objects,” and redistributing responsibility accordingly. In both cases, the way courts mobilised extralegal evidence, whether more locally grounded, as in Bonaire, or operating at a broader policy scale, as in Brazil, served to identify new risk relations: either new objects, or new connections between existing ones. This reconfiguration translated into the redistribution of responsibility to governments.
The findings invite a broader theoretical reflection on the role of courts as epistemic institutions. Existing scholarship in science and technology (STS) studies has long recognised courts as sites where scientific knowledge is transformed, stabilised and contested in the process of legal decision-making. The risk work framework used in this paper develops this insight by attending to the processual and relational dimensions of judicial knowledge negotiation. This foregrounds not only which epistemic inputs are admitted, but how judges, as professionals navigating plural and sometimes competing risk knowledges, actively shape the meanings that those inputs acquire in legal contexts.
In this regard, the cases examined here contribute to the STS literature on courts and knowledge production in two directions. First, they show that the epistemic work performed by courts in climate litigation is not currently confined to the evaluation of techno-scientific evidence, but it increasingly encompasses social-scientific forms of knowledge whose admission and weight remain contingent on procedural architecture, constitutional tradition, and political context. Beyond illuminating a recent turn towards more (explicit) epistemic pluralism in courts,Footnote 54 this contingency means the epistemic work courts do is not just about truth-finding but about legitimising particular ways of knowing climate risk over others. Second, the cases illustrate that judicial choices about knowledges are not epistemically neutral: in selecting and legitimising particular risk framings over others, courts co-produce the very understanding of climate risk that grounds the distribution of legal responsibility.
Finally, our findings highlight how this articulation of extralegal and legal knowledges contributed to the consolidation of climate litigation as a legal tool for climate governance.Footnote 55 Like in other recent climate cases,Footnote 56 judges drew on previous climate cases as argumentative precedents by mobilising them as a source of legal authority and reproducing their underlying narratives on harms and duties. Risk work, in this sense, is epistemic legitimation work and institutional layering,Footnote 57 a process through which courts not only negotiate extralegal knowledge but actively articulate it with legal knowledge and authorise particular understandings of climate risk as the basis for distributing legal responsibility.
The study is not without limitations. The geographical scope and number of cases examined make the analysis necessarily contextual. Two cases in two jurisdictions cannot support generalisations. What the comparison can do, and what we hope it has done, is illustrate with granularity how risk configuration operates in judicial practice across contrasting socio-legal settings. A further limitation concerns the research design’s exclusive documentary nature. While document analysis allowed us to reconstruct the epistemic choices made in each case, it did not grant access to the personal experiences and institutional pressures that may shape how individual judges interpreted and negotiated risk. Triangulating document analysis with phenomenologically oriented interviews with judges and other legal actors would be a promising avenue for future research, particularly for identifying the practical reasoning that underlies risk work in climate litigation settings.
These limitations also inform a wider research agenda. The dynamics identified here, such as the tension between procedural openness and epistemic narrowness, the judicial co-production of understandings of risk and the influence of political polarisation on epistemic choices, are likely to operate differently in litigation directed against corporations rather than states, where different evidentiary demands, liability frameworks and stakeholder relations come into play. Extending the comparative framework to other jurisdictions, particularly those in the Global South where climate vulnerability intersects acutely with colonial legacies and institutional fragility, would further enrich understanding of how courts navigate the plurality of risk knowledges in climate governance. A related open question concerns how to define the scope of “local” and “local populations” in climate litigation, and how courts can engage local communities’ lived experiences directly, beyond organisational intermediaries, in ways to illuminate how environmental changes are perceived, experienced, and managed on the ground.
As climate litigation continues to expand in scope and ambition, so does the need for socio-legal research attentive to the epistemic and legitimising dimensions of judicial practice.
VII. Conclusion
This study has positioned risk work as a useful theoretical framework for exploring court decisions through a sociological lens. By reframing judicial decision-making in climate litigation as a form of professional risk practice, the analysis illuminates how courts do not merely apply law to scientific facts but actively reframe climate risks at the intersection of legal and extralegal knowledges, historical context, and politically polarised environments. The comparison between the Bonaire and Brazilian cases demonstrates that this reframing is neither uniform nor inevitable: it is shaped by procedural architecture, constitutional tradition, the nature of the claims brought before the court, and the social and political pressures that surround high-profile litigation. In this sense, risk work offers socio-legal scholarship a conceptual vocabulary for accounting for the situated and relational character of judicial engagement with climate knowledge, and, by extension, with climate governance itself.
Supplementary material
The supplementary material for this article can be found https://doi.org/10.1017/err.2026.10122
Funding Statement
Open access funding provided by Tilburg University.
Competing interests
The authors declare none.
Open access
