Science and Policies of Deforestation in the Amazon: Reflecting Ethnographically on Multidisciplinary Collaboration

  • Marko Monteiro


This chapter discusses results from an ethnography of a multidisciplinary scientific project focused on the Amazon. The ethnography shows that arriving at usable scientific results through interdisciplinary work can be a challenge because such sharing is conditioned by misunderstandings between disciplinary boundaries. These misunderstandings involve disparate views on the potentials of modelling to be applied to social phenomena, which affects how the knowledge production process was carried out in practice. By making these challenges visible and analysable, ethnography can be a powerful tool in such cooperative efforts, helping scientists to navigate issues which, although usually seen as problems, can be mobilized as an important part of the process of knowledge production.


Ethnography Anthropology of science and technology Science-policy interface Environmental infrastructures STS Brazil 


This chapter discusses results from an ethnography of a multidisciplinary and collaborative scientific project, with a focus on the Amazon. It argues that ethnographic engagement with scientific projects can help to illuminate some of the challenges of undertaking complex, large-scale environmental research. In the case analysed here, this collaboration involves multidisciplinary work between modelling, social science and environmental sciences. Such a broad scope of cooperation brings unexpected difficulties as well as unexpected gains to the science produced. As models become increasingly relevant to environmental policies in Brazil and other countries (Edwards 2010; Miguel and Monteiro 2014a, b; Shackley and Wynne 1996), and as large-scale scientific projects focused on climate and environment multiply, it becomes increasingly urgent to better understand how such science is produced in practice and the challenges such multidisciplinary efforts face.

I also argue that ethnography can be a powerful tool in both understanding how such projects operate (providing knowledge on how models are constructed and incorporated into policy) and in helping to make explicit some of the challenges of doing such cooperative work. Making these challenges visible can become a useful way to reorient how scientists deal with the difficulties of cooperating between disciplines. Instead of dismissing them as “problems with communication” which can be easily fixed (Monteiro and Keating 2009: 9), scientists could benefit from such ethnographic insight and start intentionally mobilizing such issues in the very process of knowledge production. This chapter does not propose to solve these problems or provide the relevant tools in any detail; my aim here is to begin the conversation by showing, through an ethnographic example, how cooperation faces challenges which are more than just lack of understanding, but relate to the specific ways in which interdisciplinary work happens in practice.

As current research has shown, interdisciplinary work in practice can be less marked by open-minded cooperation than by ambivalence, critique and even dishonesty (Fitzgerald et al. 2014), which permeate a politics of interdisciplinary work that is still not fully understood or described. Cooperation between social and natural sciences can be challenging: research priorities are not always shared, and ethnographies of such settings have shown how social scientific insights have been neglected or even met with hostility by natural scientists (Rabinow and Bennett 2012). As will be discussed in the ethnography of the Amazalert project, social scientists will often resist the kind of reduction necessary for data to become amenable to become model input (Jeffrey 2003), which creates difficulties in effectively materializing cooperative work, even as such cooperation is prioritized in current science. Suspicion and mistrust between disciplines is one feature of such cooperation that marks how many of these efforts are enacted in practice.

Yet it is in those places of noise and friction between different epistemic cultures and different social worlds attempting to produce answers together (i.e. scientists from different disciplines trying to come to grips with the drivers of deforestation in the Amazon) that a possible place for ethnography in such projects can be imagined as a productive part of enacting more open, participatory and humble science and technology. As an inherent part of interdisciplinary work, problems in communication cannot be simply “mitigated” or overcome by “more effective” ways of communicating: such misunderstandings should, on the contrary, be reflected on as part of the results of such endeavours, and recognized as part of the collaboration itself.

Ethnographers of science become deeply entwined in the practices they observe (Latour 1995; Latour and Woolgar 1986), sometimes participating in them directly (Forsythe 1998), as ethnographers of other social phenomena also do (Mosse 2006). However, by sharing relevant worldviews and becoming fluent in sciences other than their own, ethnographers become more than just participant observers: they are potentially also active knowledge producers, especially in projects related to climate change, in which there is interest in the “human and social” aspects of natural phenomena. These claims emerge from my own experience of doing ethnography of science, but these experiences are arguably generalizable to other instances where the ethnographer is embedded in collaborative scientific efforts in which both “natural” and “social” variables are being analysed (Forsythe 1998), and where expertise (scientific and otherwise) is an object of ethnographic interest in itself (Mosse 2007).

Ethnographies of science have since the mid-1980s consolidated a rich body of work (Franklin 1995; Hess 2001; Knorr-Cetina 1983), helping us delve into the intricacies of knowledge-making, offering new ways to perceive how science is constructed in practice (Lynch 1982) and enabling a richer understanding of how natural phenomena circulate as knowledge beyond their localized settings (Latour 1995). Latour and Woolgar’s now classic study (1986), alongside others, has established a powerful way of undertaking the study of science in the making which has become very influential in Science and Technology Studies (STS) and beyond. However, in the study by Latour and Woolgar the force of the argument is partly an effect of the rhetoric of detachment present in the narrative: the anthropologist analysing seemingly bizarre and exotic practices as an outsider.

This narrative of othering the scientific practices observed, albeit a powerful way of making what would be familiar to the reader very exotic, also built into the argument a rather positivistic way of looking at science through ethnography and does not represent the view presented in this chapter about the place of ethnography in collaborative scientific projects.1 Our point is to discuss entanglements between ethnography and the processes of knowledge production under ethnographic analysis, and not detachment, and to think how ethnography can participate in the processes of interdisciplinary knowledge production. Ethnography can help us to understand problems of communication as an integral part of how knowledge production happens, and not as something to be removed from the process.

I additionally argue that the project’s final results are more “humble” (Jasanoff 2003) when compared with the claims made at its outset about the potentials of modelling for public policy. Jasanoff argues that the traditional post-war social contract between science and society, whereby scientists, usually producing knowledge isolated from society, were expected to be able to provide reliable expert advice, is increasingly in crisis. This social contract is based on a great amount of freedom for science to govern itself, being thus protected from outside interference or participation. She states that we need “technologies of humility”, which are methods or institutionalized habits of thought to try to come to grips with those aspects of knowledge at the fringes of certainty and understanding, “the unknown, the uncertain, the ambiguous, and the uncontrollable” (Jasanoff 2003: 227).

There is a growing need, I shall argue, for what we may call the ‘technologies of humility’. These are methods, or better yet institutionalized habits of thought, that try to come to grips with the ragged fringes of human understanding—the unknown, the uncertain, the ambiguous, and the uncontrollable. Acknowledging the limits of prediction and control, technologies of humility confront ‘head-on’ the normative implications of our lack of perfect foresight. They call for different expert capabilities and different forms of engagement between experts, decision-makers, and the public than were considered needful in the governance structures of high modernity. (Jasanoff 2003: 227)

As argued below, in incorporating ethnographies more explicitly in the production of knowledge in highly interdisciplinary scientific efforts, one can begin to build such technologies or methods for humility by addressing incommensurability in social vs natural science data and understandings of science; and in terms of deflating expectations on the role modelling can have on policy.

After discussing in section “Environmental Science and the Amazon” how large-scale collaborative projectshave become increasingly common ways of producing specific images of the Amazon and its relation to local and global climate, in section “Doing Ethnography of Science-Policy Interfaces” I will delve into the potentials of doing ethnography at the interface of science and policy, where specific types of expertise are seen as having a privileged capacity to aid decision-making. Much like other scientific endeavours in/about the Amazon, Amazalert aimed to produce policy-relevant science, helping policy actors to make better decisions about environmental problems and sustainability. In section “Following the Amazalert Project” where the ethnographic data from Amazalert is discussed in some detail, I point out how difficulties related to incorporating social variables into modelling practices are revealing of broader challenges of cooperating among disparate fields of expertise. In the conclusion, I argue that the ethnographic engagement with science can help to illuminate some of the challenges of working between disciplines, making it a valuable tool to produce knowledge about science in the making and the challenges of cooperation.

Environmental Science and the Amazon

Recent science has constructed the idea of the Amazon as a complex system of interactions and feedback mechanisms, which demands multidisciplinary research (Monteiro et al. 2014). Such work has mobilized an array of research groups, major scientific endeavours and growing funds from several countries, especially from the USA and Europe. These projects are usually implemented through large multidisciplinary teams and build large material infrastructures in order to conduct increasingly ambitious work. Noteworthy among these is the LBA project—Large Scale Biosphere Atmosphere Experiment in Amazonia,2 which is one of the largest collaborative research efforts in the region. It has had a significant impact in forming scientific capacity and in producing knowledge on the Amazon as a regional entity with a very specific role in the global climate system (Avissar et al. 2002; de Gonçalves et al. 2013).

In August 2015 the LBA inaugurated ATTO—the Amazon Tall Tower Observatory.3 This 325-metre tower will collect data on the biosphere-atmosphere interactions to help science to understand the impact of climate change, the role of different chemical compounds in the atmosphere and the role of urban pollution (Andreae et al. 2015). As in other projects of this size, ATTO seeks to produce knowledge that is policy-relevant, thought to be central to mitigate the destruction of the forest and the possible climate changes brought by deforestation. Another recent project, AmazonFACE (Lapola and Norby 2014), seeks to understand if the Amazon forest, understood as an “engine of photosynthesis” will consume enough atmospheric carbon dioxide (CO2) to slow global warming down (Grossman 2016). This project will spray pure CO2 into experimental lots to see how the forest reacts to CO2 levels expected be reached by 2050.

Several of these projects work with diverse technologies besides monitoring towers. Among these, remote sensing and computer modelling are becoming increasingly more present in environmental and climate sciences (Edwards 2010; Monteiro 2012; Rajão 2012). Remote sensing provides rich data on deforestation, on the dynamics of land use and has participated for quite some time in the formulation of environmental policies in Brazil, especially through the work of INPE, Brazil’s National Institute for Space Research (Miguel and Monteiro 2014a, b; Pereira 2008). Modelling has thus gained importance over the last decades, especially in climate science, and has had enormous impact on our perception of environmental processes and climate changes.

Furthermore, models are a central element of global climate politics and policies (Lahsen 2005, 2009; Miguel and Monteiro 2014a, b). Thus one can argue that the emergence of a scientific view of the Amazon as a complex system of interactions is co-produced with policies relating to the Amazon and its role in global climate dynamics (Miller 2004). There is a clear relationship between the growing scientific interest in the Amazon and the complex array of policies that have been created in Brazil regarding deforestation and preservation of the Amazon rainforest (Hochstetler and Keck 2007). Recently, models have taken centre stage in climate governance worldwide (Miller and Edwards 2001) as well as locally in Brazil (Miguel and Monteiro 2014a, b). Since the growing interest in the Amazon is also related to the perception of its relevance to understanding climate dynamics, as evidence in large-scale project framings, models have also become important in the imagination of how the Amazon should be governed (Monteiro 2015).

Doing Ethnography of Science-Policy Interfaces

The interface of science and policy has been a central topic in STS since its inception (Irwin 2008; Jasanoff 1987; Nelkin 1975, 1992), but the theme of ethnographic studies of science and technology has only come to the foreground recently (Hess 2001). Policy has gained greater attention from anthropologists (Mosse 2005; Shore and Wright 2003), but most of those studies do not tackle the central place of scientific knowledge in policy-making and policy implementation. This opens interesting and little charted territories for research at the intersection of anthropology and STS. Such studies have the potential to point to new areas of enquiry and for intervention in science and policy.

Beyond the experience of fieldwork itself, ethnography enables us to reconsider the place for STS in (re)creating technoscience. This includes discussions on governance (Guston and Sarewitz 2002) and “responsible innovation” (Stilgoe et al. 2013), discourses which call for a more active and constitutive role for social science and reflexive STS approaches in the production of new technologies (Macnaghten et al. 2005). The anthropological take on STS allows one, for example, to analyse the performative nature of scientific framings of environmental problems; beyond deconstructing traditional frames, this enables us to reconstruct practices and even attempt to participate in constructing different, more relevant (responsible, inclusive, democratic, sustainable) frames and policy approaches (Fortun and Fortun 2005). Ethnography can thus have a critical place at the nexus of scientific knowledge, technological systems and public policies.

As projects such as Amazalert and others build knowledge that relates to pressing environmental problems, they help to co-construct specific policies and transform local, national and sometimes global frames and practices related to issues such as climate change. Deforestation takes on new meaning, as it becomes part not only of destroying a rich and massively biodiverse biome, but of irreversibly changing local (and potentially global) environments and climates. As their frames of meaning change, the policies set up to monitor and mitigate deforestation also change, involving science policy (setting up programmes to produce more and “better” climate knowledge), rural and urban policies (as land use changes in the Amazon are major drivers for large scale deforestation) and development policies (as the environment is perceived as enabling services one can pay for in various ways or attempt to conserve, and economic activities are re-organized along those principles as well).

In terms of science and technology policies aimed at the region, the central place that the Amazon has taken in global climate science has enabled the implementation of ever larger projects to study all aspects of the biome, its relation to the atmosphere, the particles that help cause rain, land use changes, among many other topics. This involves setting up complex technological infrastructures, which drive scientific projects but also create problems for the STS-oriented ethnographer. How does one follow such diverse actors? How does one produce a coherent narrative that analyses, interprets and maps the issues at hand?4

One needs to follow not only scientific concepts (i.e. savannization, deforestation, climate change, tipping points) and their transformations, but also the technological infrastructures that allow such concepts to emerge and proliferate—such as the monitoring towers being built all over the Amazon—as well as the remote sensing infrastructure so crucial to deforestation research (Rajão and Vurdubakis 2013). As infrastructure studies have shown (Edwards 2010), climate science has participated in the construction of a global governance that goes beyond science itself. In order for this to unfold, one needs to understand such infrastructures as complex social arrangements (Edwards 2003). These involve distributed systems spread across many sites, institutions, dealing with a multitude of different disciplines and scholarly traditions. The ethnographic sensibility in STS helps us see such infrastructures as more than machines or objects, but as emergent systems enacted through the practical achievements of multiple actors (Morita 2013; Pickering 1993). Scientists, satellites, monitoring towers, supercomputers, laptops, smartphones and laws are all enmeshed in producing these associations and also in creating lock-ins that constrain future developments.

In the case of Amazalert, understanding the idea of a tipping point in Amazonian deforestation is crucial for reflecting on how the whole project was set up and legitimized. According to recent debates on deforestation, strong evidence suggests that if continued beyond a certain point irreversible changes will create a chain reaction that would alter the ecological aspects of the biome as well as the local and regional climates that interact with them (Nepstad et al. 2008). This is related to a possible savannization of the Amazon (Nobre et al. 1991), which is also a marked point of debate in current research on deforestation and climate in the region. Some authors suggest that one of the possible futures of the Amazon, as it is pushed beyond a tipping point after which a chain of irreversible changes occur, is to become more of a savannah-like biome, drier and more prone to fires (Malhi et al. 2009; Vergara and Scholz 2011).

This leads to the idea of constructing an early alert system: if there is such a tipping point, and if this is related to deforestation dynamics which include complex natural and social variables (Amazalert 2015), then understanding these dynamics and producing better models could enable us to enact policies now that will avoid irreversible changes in the future. Modelling is central to this effort as it is thought to be capable of producing reliable information about future trends, which in turn can help to lead to actions that avoid future damage. Such damage includes the loss of valuable “environmental services” related to carbon emissions, water and biodiversity, among many others.

While this idea is interesting and rich in research possibilities, constructing such a system carries a touch of optimism about technological potentials that helped create some tensions in the participatory workshops developed within Amazalert. Different conceptions of the possibilities of modelling socio-natural dynamics were one of the sources of friction and controversy that were observed in collective debates, and such controversies were very hard to mitigate and resolve. Such misunderstandings have been observed in research with similar groups developing multidisciplinary research (Monteiro and Keating 2009) and it remains an understudied aspect of such multidisciplinary collaborations.

Following the Amazalert Project

The Amazalert project,5 where the ethnographic research that informs this chapter was conducted, sought to produce knowledge about possible tipping points in the Amazon. It lasted from 2011 to 2014 and aimed to produce “early alerts” of possible points of no return in order to provide policy-makers with reliable information about what to do to avoid irreversible damage to the forest. Publicity materials lay out the project objectives in a way that suggests certainty (i.e. by stating that models will be improved by bringing in stakeholders to participate in creating scenarios for the future), but also that make explicit how the project leaders are keenly aware of the uncertainties involved.

Modelling was a central aspect of the work, as the project intended to develop a blueprint for an early alert system based on better predictions. However, this objective was premised on the admission that predictions vary according to what kind of model is used to make them. In interviews and conversations, project members were clear that they hoped to increase robustness in the model by incorporating “social” variables, which included the local stakeholders’ perspectives on scenarios for the Amazon.6

The strategy outlined by the scientists, as assessed by analysing both project materials and interviews, was to produce innovative computer models that included the “human elements” of deforestation to obtain more robust predictions. Robustness in this context means a model that can make better, more accurate predictions regarding the dynamics of deforestation. This involved organizing workshops with local stakeholders and policy-makers to produce the scenarios of deforestation and point out the more relevant variables, which would in turn be incorporated into the models. This led to interesting dynamics within the project, creating frictions that were not fully dealt with by the end of the activities, but which can become rich data for ethnographic reflection.

Amazalert held two workshops in order to achieve the task of incorporating socio-economic data into the model. The first workshop was held in Belém between 24 and 26 of June 2013, involving non-governmental organizations (NGOs) located in the Amazon region. This workshop developed two scenarios, a pessimistic and an optimistic about the Amazon, environmental degradation due to land use change and other economic activities. I attended the second workshop on 23 November 2013 in Brasília. The scenarios developed in the first workshop were discussed in Brasília by people from policy-related areas. These discussions occurred in separate groups, involving representatives from many government and policy institutions (Ministry of Defence, Ministry of Fisheries, Ministry of the Environment, Ministry of Agriculture, scientific institutions such as Embrapa7 and INPE, among others).

Groups were divided into broad topics: Economic Issues, Social Issues and Natural Resources. Interestingly, the Economic Issues group was the largest one, while the Social Issues group the smallest. This last group included two anthropologists, including myself, and was both spatially and thematically marginal in the whole discussion. While the reasons for separating “Economics” and “Social” issues remained unclear to me, it seemed that “social” in this workshop referred to problems related to marginalized populations, politics and identities, while economic referred to more recognizable indicators such as income, or production chains (in agriculture, cattle farming, etc.), which were more readily available for quantification.

The duality between natural phenomena and human/social processes remained problematic in different instances. The native representations of the scientists, while attempting to suggest an inclusion of social data into models, seemed at the same time to reinforce the ontological division between these realms, leading to irreducible differences that were seen as challenges to modelling in the terms set up by the project leaders.

Reducing Society into Models

In the interviews conducted before and after the workshop, the leading scientists expressed the complex and problematic nature of the undertaking. Their wish to incorporate socio-economic data into models, while theoretically a tool to increase the model’s predictive power, raised several dilemmas, including the question of how one “reduces” such social data into quantitative variables that would be amenable to become “input” in a model. This was not completely clear to them and it became less clear during the workshop as some participants, including myself, questioned this more directly.

Well that’s difficult. It’s very easily said, we do modelling and we do policy advising, but to actually really couple the two of them… Very often it happens in these projects is that there is a modelling component and a policy component, and the modellers do the modelling, and the policy people, they do the policy studies, and then at some point they start saying to each other ‘we should really integrate’, but then, it’s very difficult to talk to each other.

(Interview with one of the PIs for the project during the second workshop)

The dilemma can be roughly glossed as a duality between two resistances: that of the social scientists, who resisted reductionist framings of the social drivers of deforestation; and that of modellers, who resisted the concept of “irreducible complexity” while seeking to actively build models that incorporate those data. Irreducibility would be a challenge to the task of modelling, as it would mean a problem impossible to solve: if social variables are irreducible to numbers or values they could never be modelled, even in approximate terms, and that would make models irremediably limited. But it would also mean a challenge to policies, which tend to favour clear objectives that are minimally measurable and trackable, even though the potential of indicators to actually drive the science–policy interface is problematic (Sébastien et al. 2014).

This dilemma became more explicit in relation to other areas that might be considered “social”, including policy. The difficulty in communication between areas was implicit during the Brasília workshop, as the whole dynamics was geared towards producing usable data for modelling. But other issues emerged during the process, including questions of violence in the region, the absence of the state, and conflicts around land (including but not exclusively native Amerindian populations being driven out of their traditional lands). These were suggested in the “social issues” group by people with extensive experience on the ground, but were very hard to capture in terms of “drivers”, or incorporate into the scenarios built in the previous workshop in Belém that were structuring the debate in Brasília.

Several discussions (which were filmed and transcribed) related to how technology could enable better, cleaner and more sustainable economic activities in the Amazon, which were preconditions for the more optimistic scenarios to come into being–that is, less deforestation while the region would still achieve relative development.

But my comment is this, that this is exactly the natural tendency. We want the small, the medium producer, also the family farm, to make use of a better technology, because by doing this, they will be saving space and making that production better in relation to the natural resources that exist in that region.

(intervention of the representative of a ministry in the economic issues group)

In this example, we see a kind of perception that the use of more intensive technologies would save land and thus prevent deforestation. As cattle in Brazil is still raised extensively, relying on open grasslands, the tendency as production grows is for deforestation to expand correspondingly. A more intensive use of technology, according to many, would curb that expansion while still enabling growing outputs. This same view was not restricted to cattle or farming, but also appeared during a debate around natural resources in general:

I think another important aspect of the long-term mining would be the incorporation of technology in the minerals, so that we don’t just keep selling ore without any work or processing in Brazil. I think this adds value, because otherwise we will always be a colony, right!?

(intervention of the representative of a public university during discussions in the natural resources group)

The incorporation of technology was seen as a positive aspect in all sectors: it would enable more productive agriculture and potentially result in more productive uses of Brazil’s natural resources. For example, more technology would, according to some, enable the country to reduce exports of raw minerals and increase exports of industrialized goods such as steel or even machinery and capital goods. More technology in farms would also mean broader use of confined cattle, which in theory would need less land and promote less deforestation. These debates went on as the members of the economic issues group debated different scenarios. But the pervasiveness of “technology” as a value, almost in itself, was prominent. The idea that some generic form of technology would necessarily promote more sustainability was not problematized or reflected upon. Which technologies? Embedded in/impacting which social relations? Controlled by which groups?

I was the one elected by my group (social issues) to present our debates to the other groups. The collective sharing happened at the end of the workshop, and my presentation aimed to sum up the topics we covered collectively:

We did a pre-discussion on the issues that were perhaps missing [from the other groups], and we talked about the social responsibility of companies, being debatable that this is on the rise. And also on royalties [from petroleum, mining, etc.]. Where do these royalties go? How are they used? There are city governments that will build a palace or a fancy plaza, but won’t invest in sewage. Also, large infrastructure works, we discussed compensations and how to mitigate social stress and if, possibly, impact studies were focusing too much on the physical environment and not on the broader social issues. We talked about political culture, which relates to royalties since, for example, corruption and ‘clientelismo’8 have a large impact … if you have a context of corruption and impunity, the money will come in and will not be well invested.

(excerpt from my presentation summarizing the topics covered in the social issues group)

The discussion went much further and deeper than this, especially in the social issues group, in which I participated. We discussed inequality, citizenship, responsibility and topics that are apparently unrelated to environment or economy. However, they help us to understand problems related to violence and local politics, and also explain why environmental laws are not followed, why land keeps being cleared illegally, and why traditional groups are driven out of their lands by large-scale farming, among many other variables that can help to explain deforestation dynamics. We discussed how science and technology could interact with and create a dialogue with local knowledges, while also debating how large infrastructure projects can help create new production chains in the Amazon region that would rely less on deforestation.

A central concern throughout was finding ways of transforming data on socio-economic processes into input that could be incorporated into the model. This meant, in the discussions, developing indicators that would quantify some relevant information about key aspects of interest (e.g. deforestation rates, rainfall, forest fires, etc.). Natural indicators would be mostly related to climate such as sea surface temperatures, precipitation and droughts. But social indicators remained an unanswered question, which the project leaders hoped would be better defined in the next stages of the project. My interviews with the leaders revealed this to be a central concern and a central blind spot of modelling for environmental research, or for using modelling as a source of data for policy: which social indicators should be measured and monitored? How could the measurements and the monitoring become feasible? Which institutions should get involved and why? How can such knowledge inform policies that enable positive outcomes in terms of preservation of the Amazon? The dilemma was addressed by an anthropologist present at the discussions:

[…] the social sciences, although they also work with statistics, […], they don’t test, and there are no laws that describe society in the same way that you find regularities and hypotheses, through the way natural systems function. So, social tipping points, that’s a complicated thing.

(intervention from an anthropologist during open debates in the Brasilia workshop)

He suggested that indicators for violence could be a starting point, as they are connected to governance and citizenship, among other social issues that are deeply interrelated to environmental governance and possibly climate change mitigation efforts (or their chances of being implemented). However, the issue of social and political “tipping points” remains fascinating: as he put it, there are no limits to social problems as historically the human capacity to create violence and exclusion is very high. Others present argued that tipping points may not be an adequate language to talk about social issues, but that thresholds could be imagined that would delimit situations that we would not want to reach involving violence, poverty and income, among other indicators available.

In the end, there was no solution to this conundrum, even though it revealed very relevant blind spots in the project strategy. This controversy is an example of the kinds of complexities that can emerge in such highly diverse debates, which are very hard to close or resolve in the time allotted for them in the one-day workshop. Results had to be written up at the end of the day, when exhausted participants tried to condense the very rich discussions into central points that were to be taken up by modellers afterwards. These controversies were not discussed in the closing debates and were not explicitly addressed in the final results presented. This indicates that such controversies were not seen as “results” to be dealt with, which makes us reflect on why an ethnographer saw them as interesting and how that could be possibly reflected upon as part of such a scientific undertaking.

Potential Contributions from Ethnography

Some days after the workshop, I interviewed one of the leading modellers involved in the project and asked them about the integration of human factors into the model:

Anthropologist: Have you dealt with the human dimension? Is this a recent trend in climate science?

P1: This is a very recent thing from what I can see, we don’t know how to quantify this yet, it is very hard to translate into numbers, so it is a relatively recent thing.

A: I’m asking because I have heard at INPE, when I was interviewing people working at [PRODES and DETER9], the remote sensing guys, this [came up] and I was unsure if it came from INPE or from the [project leaders].

P1: It is something that has been emerging in the last 5 or 10 years… In the 90s we learned a lot about the climate, I mean this perception in science that human beings alter the climate, many projects were developed in order to understand how this was happening and we learned that in order to understand we had to understand all processes at the same time, so little by little people were seeing that we needed to study the whole process in an integrated way.

(Skype interview with senior scientist working with climatology and responsible for most of the modelling in Amazalert)

This answer helps to illustrate the growing relevance of human and social elements in climate science, which also contextualizes the current problems scientists face in effectively promoting dialogue between social and “hard” sciences on deforestation and other environmental problems. The workshops and other activities that were intended as moments to produce social inputs for the model in Amazalert were very short and did not, in my view, allow enough time for those questions to even be adequately posed, let alone dealt with. This is an instance where an ethnographer, attentive to this dilemma from the start, could proactively suggest ways through which non-quantifiable knowledge might be integrated in some way into the models. What this could mean, practically, is that these issues of incommensurability should be dealt with explicitly and as problems in themselves, which did not happen in the project. They were not addressed explicitly (although in corridor talk they came up often); and they were not perceived as part of the problems to be solved by the scientific effort at hand.

This also indicates that the mere fact of having embedded ethnographers or social scientists as active members of such projects is not enough to make these collaborations more open to uncertainty or to their own limitations. As these large-scale projects grow in number and in relevance, many social scientists are brought into their teams. But are the projects attentive to how these different epistemic cultures are interacting? In the case of Amazalert, social scientists were expected to bring data and insights about the social variables present in deforestation dynamics in order to make the model’s performance more robust. This is important as an innovative way of producing models, but the project did not foresee an institutional place for the discussion of incommensurability as a problem in itself that needed to be dealt with; and participants did not take advantage of discussions of these problems that came up in workshops when finalizing the project’s mains results.

What could an ethnographer do here in order to address those problems and make them more productive in scientific terms? The first thing would be to explicitly address them in the group setting, with adequate time to process what they could mean to the modelling initiative, for example. The non-quantifiable nature of some crucial data cannot be solved by producing (numerical) indicators, or by just ignoring the issue. On the other hand, innovative modelling could be achieved if the multidisciplinary team, by bringing those problems to the fore and accepting that they were relevant to the whole enterprise and not “problems of communication” or mere lack of shared expertise. The attempt to find solutions to this problem could move scientists to better understand the limits of the model and its applicability, among other unforeseen results.

Making Results Humbler

A noteworthy element in the project’s final results as compared with its initial goals was the greater awareness among its participants of the complexities and uncertainties involved in setting up the technology of an early alert system. The results presented at the end of the project were significantly ‘humbler’ in relation to what the proposal laid out in the beginning. While the project’s initial language seemed relatively certain that such a system could be set up and that it could be a useful tool for policy-makers, the language in the final report shows a much more cautious approach:

Despite its relative resilience if deforestations stays low, AMAZALERT has shown that severe degradation of Amazonia is possible when severe climate change and deforestation progress simultaneously. However, the type of change can vary strongly and can be difficult to predict, because signals of change may come only after a biophysical threshold has been crossed and decline will already be rapid or irreversible (…). Early warning for such change will therefore also have to be approached from a broad perspective. (…) Thresholds should be defined that account for society’s coping capacity as well as with the uncertainty in prediction of natural ecosystem degradation or instability. In this envisioned early warning system (EWS), new scientific insight and technical capability should be constantly adopted and tested for effectiveness. (Amazalert 2015: 9, emphasis mine)

This excerpt is taken from the final summary for policy-makers and thus is written in a language directed to decision-makers. It is not a scientific paper, where uncertainties are often openly discussed (Shackley and Wynne 1996). As exposed above, the early warning system proposed is an open system, which would have to be constantly adapted as novel scientific insights and technical capabilities become available. Although the use of models as sources of information for policy is increasing, studies point to the lack of confidence that policy-makers have in using them (Brugnach et al. 2007). This lack of confidence comes from the presence of perceived uncertainty, indicating that policy-makers tend to have high expectations with regard to models output.

Such an open system, “constantly adopted and tested for effectiveness”, appears to be open to adjustments due to new knowledge and thus aware of its own limitations. This explicitly partial and incomplete apparatus is humbler than what was proposed at the start of the project in the sense suggested by Jasanoff (2003), being therefore a type of technology more open to participation and questioning than a closed system, locked into a model thought to correctly predict environmental or climate phenomena. It is hard to say why the final summary for policy-makers incorporated the uncertainty in more explicit terms than the initial proposal. Yet it is significant that the results made those limitations explicit and clear, which signals an openness to uncertainty that can be an opportunity for broader collaboration with ethnographers and other social scientists interested in addressing the inherent limitations of modelling as a tool for policy.


Current discussions around science have proposed new forms of producing knowledge that are closely tied to its context of application (Gibbons 2000), and that are open to participation from stakeholders and other actors outside science (Guston and Sarewitz 2002; Owen et al. 2012). Amazalert incorporated many of these elements, internalizing within the project’s activities discussions between scientists and groups of stakeholders and policy-makers. And yet it produced results that are more open to uncertainty and complexity than those initially proposed, which makes it an interesting case for debate.

Contemporary societies live in a context of being constantly at risk (Beck 1996), and in such an uncertain environment science has been called to do more than speak “truth to power”. Jasanoff’s concept of technologies of humility becomes thus an intellectual answer to the challenges of enacting science and technologies that are more participatory, accountable and aware of their own limitations.

While it would perhaps be too ambitious to call Amazalert a fully realized example of a “humble” way of producing science and technology for environmental problems, it is a good example to reflect on in terms of the hopes and challenges of doing such types of “other” sciences. The project effectively worked with scientists and experts from a myriad of disciplinary backgrounds, which is a dynamic of scientific work that increases uncertainty and noise in communication. The project did not, however, make explicit the problems that arose in this kind of interdisciplinary work, problems which were here made visible through ethnographic analysis. Perhaps this was so because analysing these problems was not in the scope of the project; but I argue here that to do so, possibly through the incorporation of ethnography in some explicit measure in the dynamics, can be a useful way to explore the potential gains that arise from misunderstanding and noise between disciplines.

One of the more salient misunderstandings is the problem of incommensurability between different ways of perceiving data and its quantifiability, as observed in the participatory workshops. By bringing in professionals from different areas of expertise and disciplinary backgrounds, these workshops made controversies about data harder to close. This raises important issues related to how such incorporation can happen productively in such projects and how the limitations detected can be addressed. The problems raised related to different perceptions about the quantifiable nature of inputs from different sciences, a feature that is central to the possibility of modelling certain variables. While modellers wanted to include social variables to make models more robust and reliable, social scientists were sceptical about this possibility. This scepticism related to the perception of the great difficulty (perhaps impossibility) of quantifying social phenomena.

The project’s results are, as discussed above, humbler than initially proposed in the beginning of the activities. But the project did not fully process or take formally into account the complexities and challenges of the participatory workshops. While these topics were being discussed informally, they were not dealt with as part of the scientific agenda of the project, which may have impeded the scientists from fully appreciating the richness of the experience they made possible. It is proposed here, based on the experience of doing this ethnography, that this methodology can be productively incorporated into such projects as a formal and explicit aspect of the science being developed, and that this can promote significant gains in how environmental (and other complex) scientific collaborations are developed.

How this can be achieved is still unclear, but it is a topic that will probably gain growing relevance as large interdisciplinary projects multiply and as the human dimensions of climate change are increasingly believed to be central do climate change research. Moreover, the demands for social scientists to participate in climate research and governance (Victor 2015) will continue to push this issue to the fore of climate and science policies.

The incorporation of “human dimensions” into scientific and political concerns about climate change opens an interesting window of opportunity for such interactions and collaborations among ethnographers and environmental scientists, and should be the object of more reflection and action on the part of STS scholars. Large-scale scientific projects are, as all social environments, a complex and dynamic arena of conflict and interaction. But they are particularly central in our societies as they produce powerful effects on our perceptions of the world and on our political actions in relation to climate and on the environment. Therefore, the effective participation of ethnographers can enable relevant engagement with such transformative practices and can enable more interesting interfaces between science and policy.


  1. 1.

    Earlier critiques of Latour and the actor–network approach (Bloor 1999; Collins and Yearley 1992) have engaged with different aspects of this debate: Bloor’s work attacks Latour’s critique of the Strong Programme, SSK and the sociology of knowledge. Collins and Yearley, on the other hand, argue that actor–network approaches, although presented and perceived as radical, end up reinforcing a conservative position through a relativistic view of science and technology.

  2. 2.

    “The LBA Program is managed by the Ministry of Science, Technology and Innovation and coordinated by the National Institute of Amazon Researches/INPA. In the 17 years of research, the program was important in forming human resources, including over 600 masters and PhD researchers in Brazil. Over 150 cutting edge projects, partnering with around 280 national and international institutions, carried out by 1400 Brazilian scientists and 900 scientists from other Amazonian countries, from 8 European nations and North American institutions, sought to study and understand climate and environmental changes underway in order to favor sustainable development in the Amazon” (Source: All translations done by the author.

  3. 3.

    “The long-term objective of the ATTO is to measure the impact of global climate changes in the Amazon forests through measuring the interactions between the forest and the atmosphere, besides enabling novel research on the chemistry of the atmosphere (gas exchanges, chemical reactions and aerosols), mass and energy transportation processes in the limit of the atmosphere and processes of cloud formation and development” (Source:

  4. 4.

    Fischer (2001), Hine (2007).

  5. 5.
  6. 6.

    The incorporation of local perspectives was explained as a way of mitigating inherent uncertainties in the modelling as well as a way to make models more relevant for policy. In interviews, the principal investigators of the project discussed how this was also seen as a new frontier in modelling work, and the incorporation of social variables in such environmental models continues to be seen as a challenge. Because they attempt to model processes that have a lot to do with human action in the environment, such initiatives are becoming more common.

  7. 7.

    The Brazilian Agricultural Research Corporation (Embrapa), founded in 1973, is a research corporation, owned by the state and focused on developing technologies and technical scientific knowledge for Brazilian agriculture.

  8. 8.

    This can be roughly translated as the preponderance of personal relationships and bargaining to the detriment of the public interest in dealings with/through the state apparatus.

  9. 9.

    “Brazil has two systems for tracking deforestation: PRODES (Program to Calculate Deforestation in the Amazon) and DETER (Real-time Detection of Deforestation), which allow to rapidly identify where deforestation is occurring. PRODES, which has a sensitivity of 6.5 hectares, provides Brazil’s annual deforestation estimates (measured each August) while DETER, which has a coarser resolution of 25 ha, is a year-round alert system that updates IBAMA, Brazil’s environmental protection agency, every two weeks. This gives authorities the technical capacity—although not necessarily the political will—to combat deforestation as it occurs” (source:



  1. Amazalert. 2015. Final Report Summary – AMAZALERT (Raising the Alert About Critical Feedbacks Between Climate and Long-term Land Use Change in the Amazon). Amazalert Project. Accessed 14 Dec 2016.
  2. Andreae, M., O. Acevedo, A. Araújo, P. Artaxo, et al. 2015. The Amazon Tall Tower Observatory (ATTO) in the Remote Amazon Basin: Overview of First Results from Ecosystem Ecology, Meteorology, Trace Gas, and Aerosol Measurements. Atmospheric Chemistry and Physics 15: 10723–10776.CrossRefGoogle Scholar
  3. Avissar, R., P. Silva Dias, M. Silva Dias, and C. Nobre. 2002. The Large-Scale Biosphere-Atmosphere Experiment in Amazonia (LBA): Insights and Future Research Needs. Journal of Geophysical Research: Atmospheres 107(D20): LBA 54-1–LBA 54-6.Google Scholar
  4. Beck, U. 1996. World Risk Society as Cosmopolitan Society? Ecological Questions in a Framework of Manufactured Uncertainties. Theory, Culture and Society 13(4): 1–33.CrossRefGoogle Scholar
  5. Bloor, D. 1999. Anti-Latour. Studies in History and Philosophy of Science 30(1): 81–112.CrossRefGoogle Scholar
  6. Brugnach, M., A. Tagg, F. Keil, and W. de Lange. 2007. Uncertainty Matters: Computer Models at the Science–Policy Interface. Water Resources Management 21(7): 1075–1090.CrossRefGoogle Scholar
  7. Collins, H., and S. Yearley. 1992. Epistemological Chicken. In Science as Practice and Culture, ed. A. Pickering, 301–326. Chicago: The University of Chicago Press.Google Scholar
  8. de Gonçalves, L., J. Borak, M. Costa, S. Saleska, et al. 2013. Overview of the Large-scale Biosphere–Atmosphere Experiment in Amazonia Data Model Intercomparison Project (LBA-DMIP). Agricultural and Forest Meteorology 182: 111–127.CrossRefGoogle Scholar
  9. Edwards, P. 2003. Infrastructure and Modernity: Force, Time, and Social Organization in the History of Sociotechnical Systems. In Modernity and Technology, ed. T. Misa, P. Brey, and A. Feenberg, 185–225. Cambridge, MA: MIT Press.Google Scholar
  10. ———. 2010. A Vast Machine: Computer Models, Climate Data, and the Politics of Global Warming. Cambridge, MA: MIT Press.Google Scholar
  11. Fischer, M. 2001. Ethnographic Critique and Technoscientific Narratives: The Old Mole, Ethical Plateaux, and the Governance of Emergent Biosocial Polities. Culture, Medicine and Psychiatry 25: 355–393.CrossRefGoogle Scholar
  12. Fitzgerald, D., M. Littlefield, K. Knudsen, J. Tonks, and M. Dietz. 2014. Ambivalence, Equivocation and the Politics of Experimental Knowledge: A Transdisciplinary Neuroscience Encounter. Social Studies of Science 44(5): 701–721.CrossRefGoogle Scholar
  13. Forsythe, D. 1998. Using Ethnography to Investigate Life Scientists’ Information Needs. Bulletin of the Medical Library Association 86(3): 402–409.Google Scholar
  14. Fortun, K., and M. Fortun. 2005. Scientific Imaginaries and Ethical Plateaus in Contemporary US Toxicology. American Anthropologist 107(1): 43–54.CrossRefGoogle Scholar
  15. Franklin, S. 1995. Science as Culture, Cultures of Science. Annual Review of Anthropology 24: 163–184.CrossRefGoogle Scholar
  16. Gibbons, M. 2000. Mode 2 Society and the Emergence of Context-sensitive Science. Science and Public Policy 27(3): 159–163.CrossRefGoogle Scholar
  17. Grossman, D. 2016. Amazon Rainforest to get a Growth Check. Science 352(6286): 635–636.CrossRefGoogle Scholar
  18. Guston, D., and D. Sarewitz. 2002. Real-time Technology Assessment. Technology in Society 24: 93–109.CrossRefGoogle Scholar
  19. Hess, D. 2001. Ethnography and the Development of Science and Technology Studies. In Handbook of Ethnography, ed. P. Atkinson, A. Coffey, S. Delamont, J. Lofland, and L. Lofland, 234–246. London: Sage.CrossRefGoogle Scholar
  20. Hine, C. 2007. Multi-sited Ethnography as a Middle Range Methodology for Contemporary STS. Science, Technology and Human Values 32(6): 652–671.CrossRefGoogle Scholar
  21. Hochstetler, K., and M. Keck. 2007. Greening Brazil: Environmental Activism in State and Society. Durham: Duke University Press.CrossRefGoogle Scholar
  22. Irwin, A. 2008. STS Perspectives on Scientific Governance. In The Handbook of Science and Technology Studies, ed. E. Hackett, O. Amsterdamska, M. Lynch, and J. Wajcman, 583–607. Cambridge, MA: MIT Press.Google Scholar
  23. Jasanoff, S. 1987. Contested Boundaries in Policy-Relevant Science. Social Studies of Science 17(2): 195–230.CrossRefGoogle Scholar
  24. ———. 2003. Technologies of Humility: Citizen Participation in Governing Science. Minerva 41(3): 223–244.CrossRefGoogle Scholar
  25. Jeffrey, P. 2003. Smoothing the Waters: Observations on the Process of Cross-Disciplinary Research Collaboration. Social Studies of Science 33(4): 539–562.CrossRefGoogle Scholar
  26. Knorr-Cetina, K. 1983. New Developments in Science Studies: The Ethnographic Challenge. Canadian Journal of Sociology 8(2): 153–177.CrossRefGoogle Scholar
  27. Lahsen, M. 2005. Seductive Simulations? Uncertainty Distribution Around Climate Models. Social Studies of Science 35: 895–922.CrossRefGoogle Scholar
  28. ———. 2009. A Science–Policy Interface in the Global South: The Politics of Carbon Sinks and Science in Brazil. Climatic Change 97: 339–372.CrossRefGoogle Scholar
  29. Lapola, D., and R. Norby 2014. Amazon-FACE: Assessing the Effects of Increased Atmospheric CO2 on the Ecology and Resilience of the Amazon Forest–Science plan and Implementation Strategy. Accessed 16 Dec 2016.
  30. Latour, B. 1995. The ‘Pedofil’ of Boa Vista: A Photo-Philosophical Montage. Common Knowledge 4(1): 144–187.Google Scholar
  31. Latour, B., and S. Woolgar. 1986. Laboratory Life: The Construction of Scientific Facts. Princeton: Princeton University Press.Google Scholar
  32. Lynch, M. 1982. Technical Work and Critical Inquiry: Investigations in a Scientific Laboratory. Social Studies of Science 12(4): 499–533.CrossRefGoogle Scholar
  33. Macnaghten, P., M. Kearnes, and B. Wynne. 2005. Nanotechnology, Governance, and Public Deliberation: What Role for the Social Sciences? Science Communication 27(2): 268–291.CrossRefGoogle Scholar
  34. Malhi, Y., L. Aragão, D. Galbraith, C. Huntingford, et al. 2009. Exploring the Likelihood and Mechanism of a Climate-Change-Induced Dieback of the Amazon Rainforest. Proceedings of the National Academy of Sciences 106(49): 20610–20615.CrossRefGoogle Scholar
  35. Miguel, J., and M. Monteiro. 2014a. Por que devemos nos interessar por modelos climáticos? Climacom 1(1): 6–17.Google Scholar
  36. ——— 2014b. Modelos do Clima, Modelos Políticos: Uma breve historiografia da previsão numérica e modelagem climática no CPTEC/INPE. Anais da REACT-Reunião de Antropologia da Ciência e Tecnologia, Campinas, 2014.Google Scholar
  37. Miller, C. 2004. Climate Science and the Making of a Global Political Order. In States of Knowledge: The Co-production of Science and the Social Order, ed. S. Jasanoff, 46–66. London: Routledge.Google Scholar
  38. Miller, C., and P. Edwards. 2001. Changing the Atmosphere: Expert Knowledge and Environmental Governance. Cambridge, MA: MIT Press.Google Scholar
  39. Monteiro, M. 2012. Imagens de satélite como sítio etnográfico? Interpretando práticas de sensoriamento remoto no Brasil. In Ciência, Tecnologia e Sociedade no Brasil, ed. M. Kerbauy, T. Andrade, and C. Hayashi, 251–278. Campinas: Alínea.Google Scholar
  40. ———. 2015. Construindo imagens e territórios: Pensando a visualidade e a materialidade do sensoriamento remoto. História, Ciências, saúde – Manguinhos 22(2): 577–591.CrossRefGoogle Scholar
  41. Monteiro, M., and E. Keating. 2009. Managing Misunderstandings: The Role of Language in Interdisciplinary Scientific Collaboration. Science Communication 31(1): 6–28.CrossRefGoogle Scholar
  42. Monteiro, M., S. Seixas, and S. Vieira. 2014. The Politics of Amazonian Deforestation: Environmental Policy and Climate Change Knowledge. WIREs Climatic Change 5: 689–701.CrossRefGoogle Scholar
  43. Morita, A. 2013. Traveling Engineers, Machines, and Comparisons: Intersecting Imaginations and Journeys in the Thai Local Engineering Industry. East Asian Science Technology and Society 7(2): 221–241.CrossRefGoogle Scholar
  44. Mosse, D. 2005. Cultivating Development: An Ethnography of Aid Policy and Practice. London: Pluto Press.Google Scholar
  45. ———. 2006. Anti-social Anthropology? Objectivity, Objection, and the Ethnography of Public Policy and Professional Communities. Journal of the Royal Anthropological Institute 12(4): 935–956.CrossRefGoogle Scholar
  46. ———. 2007. Notes on the Ethnography of Expertise and Professionals in International Development. Ethnografeast III ‘Ethnography and the Public Sphere’, Lisbon.Google Scholar
  47. Nelkin, D. 1975. The Political Impact of Technical Expertise. Social Studies of Science 5(1): 35–54.CrossRefGoogle Scholar
  48. ———. 1992. Controversy: Politics of Technical Decisions. Thousand Oaks, CA: Sage.Google Scholar
  49. Nepstad, D., C. Stickler, B. Soares-Filho, and F. Merry. 2008. Interactions among Amazon Land Use, Forests and Climate: Prospects for a Near-term Forest Tipping Point. Philosophical Transactions of the Royal Society 363: 1737–1746.CrossRefGoogle Scholar
  50. Nobre, C., P. Sellers, and J. Shukla. 1991. Amazonian Deforestation and Regional Climate Change. Journal of Climate 4: 957–988.CrossRefGoogle Scholar
  51. Owen, R., P. Macnaghten, and J. Stilgoe. 2012. Responsible Research and Innovation: From Science in Society to Science for Society, with Society. Science and Public Policy 39(6): 751–760.CrossRefGoogle Scholar
  52. Pereira, G. 2008. Política Espacial Brasileira e a Trajetória do INPE (1961–2007). PhD dissertation, State University of Campinas.Google Scholar
  53. Pickering, A. 1993. The Mangle of Practice: Agency and Emergence in the Sociology of Science. The American Journal of Sociology 99(3): 559–589.CrossRefGoogle Scholar
  54. Rabinow, P., and G. Bennett. 2012. Designing Human Practices: An Experiment with Synthetic Biology. Chicago: University of Chicago Press.CrossRefGoogle Scholar
  55. Rajão, R. 2012. ICTBased Monitoring of Climate Change Related Deforestation: The Case of INPE in the Brazilian Amazon. NICCD. Accessed 16 Dec 2016.
  56. Rajão, R., and T. Vurdubakis. 2013. On the Pragmatics of Inscription: Detecting Deforestation in the Brazilian Amazon. Theory, Culture and Society 30(4): 151–177.CrossRefGoogle Scholar
  57. Sébastien, L., T. Bauler, and M. Lehtonen. 2014. Can Indicators Bridge the Gap between Science and Policy? An Exploration into the (Non) Use and (Non) Influence of Indicators in EU and UK Policy Making. Nature and Culture 9(3): 316–343.CrossRefGoogle Scholar
  58. Shackley, S., and B. Wynne. 1996. Representing Uncertainty in Global Climate Change Science and Policy: Boundary-Ordering Devices and Authority. Science, Technology and Human Values 21(3): 275–302.CrossRefGoogle Scholar
  59. Shore, C., and S. Wright. 2003. Anthropology of Policy: Perspectives on Governance and Power. London: Routledge.Google Scholar
  60. Stilgoe, J., R. Owen, and P. Macnaghten. 2013. Developing a Framework for Responsible Innovation. Research Policy 42: 1568–1580.CrossRefGoogle Scholar
  61. Vergara, W., and S. Scholz 2011 Assessment of the Risk of Amazon Dieback. The World Bank. Accessed 5 Nov 2015.
  62. Victor, D. 2015. Embed the Social Sciences in Climate Policy. Nature 520: 27–29.CrossRefGoogle Scholar

Copyright information

© The Author(s) 2017

Authors and Affiliations

  • Marko Monteiro
    • 1
  1. 1.Department of Science and Technology PolicyInstitute of Geosciences (IG), State University of Campinas, UNICAMPCampinasBrazil

Personalised recommendations