ChapterTwo

Khulna Frustrations: Embankments, Siltation, and Dying Rivers

I returned to Bangladesh for fieldwork in August 2014. Three years had passed since my last visit to Khulna District with Sanvi. I stayed with relatives in Dhaka for the first month and carried out more than forty-five semistructured key-informant interviews with development professionals in the capital. In doing so, I gained insights into the official narratives of development projects and the many hats that Bangladeshi professionals wear as employees of aid-funded projects, while also obtaining practical information for the rural part of my fieldwork. I continued to meet former colleagues, and our repeated meetings turned into informal conversations. When I spoke in Bangla with interlocutors, I found that statements or opinions would shift depending on the language used. When speaking in English—the language of international development projects—they tended to regurgitate official aid narratives, whereas when speaking in Bangla, they revealed personal reflections and opinions that often were at odds with the official narrative of their respective development interventions. This linguistic code-switching illustrated the current reality of a Bangladeshi development professional working in English-dominated development projects. Conducting most of these interviews or conversations in Bangla was often rewarded with very honest and blunt statements about the reality in which the local people live and the “golden handcuffs” of their comfortable and well-paid jobs (see also Mosse 2011b).

In 2014, Bangladesh was frequently affected by national strikes (hartals) and my former colleagues recommended that I travel to Khulna on a Friday (by air) as the opposition party would not call for a general strike on the prayer day. When I arrived at the domestic terminal of Dhaka airport it was quiet and orderly, a stark contrast to the roads subject to national strikes characterized by burning cars and buses. The boarding process was quick, and it took only half an hour to fly to Jessore, followed by a two-hour drive on the airline’s private bus to Khulna. I arrived late at night and was met by the driver from my former sister-organization, who drove me to my designated guesthouse. At that point I felt that I was still working under the guise of a development consultant. The expensive flight, the private air-conditioned-bus, the private car, and the guesthouse were perceived as the most time-effective and safest means for development professionals from Dhaka to work in Khulna. Living only a short walk from my former interpreter Sanvi in Khulna, however, enabled me to immerse myself in the perspective of residents. I soon arranged to stay with local Bangladeshi families in Khulna city during the weekends away from Nodi.

I visited Sanvi frequently, both at her office and at home, often talking on her family’s rooftop, which was filled with various plants, fragrant flowers, and trees (banana, henna, papaya). On World Rivers Day, I joined Sanvi and her colleagues in a workshop organized by a grassroots environmental movement (GEMOB) in southwest coastal Bangladesh. We walked along the wide streets of Khulna, shaded by trees, and reached the venue located at the ground level of an NGO office. The front gate was open and the basement parking area was filled with several rows of plastic chairs facing a table. Above the table was a large poster with the words “Save our Rivers” written in Bangla. The main panelists, all Bangladeshi, sat at the table—a woman activist and development professional at a rights-based NGO, a lawyer at an environmental NGO, and a professor from the local university. There was no fan, no air-conditioning. The conditions were like many places in Bangladesh, and so fundamentally different from the air-conditioned, mosquito-proofed bubble of Dhaka’s development industry. For the first time working in Bangladesh, I attended a “development” meeting conducted entirely in Bangla with no presence of international “experts” or donors. It was fully independent and autonomous from development-project funding as the GEMOB network does not depend on, nor seeks to attract, donor funding. Thus, rather than an official performance of narratives emphasizing how project activities are needed and successful so as to legitimize them to donors and relevant stakeholders (Mosse 2005; Green 2003; Heaton Shrestha 2006), it proved to be a space for frank discussion among Khulna citizens. All participants worked on issues related to development and environment in this coastal region and voiced their personal views on pressing environmental problems.

The discussions ranged from embankments, siltation, and industrial pollution to transboundary water conflicts such as India’s Farakka and Tista barrages, as well as India’s new river-linking project that would worsen the already poor state of the Ganges. Rather than speaking about the virtues of a specific project, the speeches were filled with context: with history, environment, politics, and societal issues. An NGO worker named Amir spoke about the long history of the silted delta and the environmental degradation caused by the Coastal Embankment Project in the 1960s, of which Nodi is one:

Every year, more than a billion tons of sediment is not depositing on the land, but on the riverbed, so the fertility of agricultural land is decreasing. By depositing outside the polders [the Dutch word for embankments], the silt is raising the riverbed. During the monsoon, the water inside the embankment cannot drain out into the river. This increases the risks of jalabaddho floods [waterlogging and drainage congestion] and tidal surges. Since the creation of embankments, the river’s sediment has nowhere to go and our rivers and canals are silting up—they’re dying. Embankments are causing riverbed rise and waterlogging; these issues have no link to sea level rise. Even Bangladesh’s most famous “water specialist” never mentions sedimentation because of pressure from international donors. We [in Khulna] are not able to do anything. We may continue this movement for the environment, but we are all voiceless against those [who hold the real power] in Dhaka.

In his speech, Amir eloquently captured a key concern regarding how embankments exacerbate siltation and worsen damaging jalabaddho floods—a concern shared by several applied researchers, government officials, and Khulna activists with the rural people inside the embanked floodplains. He also pointed to the perceived hegemonic status of donors in “dictating” development in Bangladesh via the power-wielders in Dhaka (the donors, government agencies, and larger international NGOs run by fellow Bangladeshis of the upper-middle classes). In his view, the fact that they continue to promote the construction of embankments despite the devastating siltation they cause highlights how the elites in Dhaka do not understand the local environment.

When I stayed in Khulna city during the weekends, I often visited Professor Hossain and his family. He is active in local academia as well as in development research for various donor-funded projects. Sitting in his living room, we often discussed the various questions arising from my fieldwork. While we discussed the ever-increasing number of climate change projects in the Khulna District and the reflections made at GEMOB, Professor Hossain revealed how development professionals feel compelled to reproduce donor narratives: “People don’t believe in science, but they believe in their own idea. Donors have a presumption: if ice caps are melting, Bangladesh will drown. For the rest of us, we [Bangladeshi development professionals] must communicate in a way that relates to the ideas of donors; it becomes difficult if we do not conform to their climate change story.”

Through Professor Hossain, I met Gaurav, who works at a well-connected Bangladeshi NGO. At the end of my stay in Khulna, I went to Gaurav’s office with a box of chocolates as symbolic thank you for all his help with my fieldwork. His organization was quite adept at attracting donor funding for development projects and, as always, he was very busy. He mentioned that he was helping another UK researcher who worked on a collaborative project on “climate resilience.”4 A quick glance at their survey revealed that the questions were phrased in a way that villagers would tailor their answers to ensure that they are targeted as project beneficiaries—as they initially tried to do with me. I found it strange that, although climate resilience toward floods was the key issue, none of the questions appeared concerned with embankments, the main flood-protection mechanism, or damaging jalabaddho floods. Though Gaurav agreed with this, he did not offer any feedback on the questionnaire: “the formal agreement is on ‘climate resilience’ so my hands are tied.” He was increasingly weary of conforming to donor narratives: “Apa [sister], I want to work with research. I’m dissatisfied with this [NGO] work. None of the projects are looking at the holistic picture. It’s all about ‘climate change and disaster.’ Few [donors] understand the ecology as a whole.” Gaurav showed me a cartoon (fig. 2.1) and stated that donors are like blind men thinking they are holding the answer, but they are missing the big picture. He continued: “A lot of problems in the coastal zone are natural phenomena that have existed for a long time. But now everything is about climate change and disaster risk reduction.”

Figure2.1. Donors like blind men examining an elephant. Adapted by Mark Robinowitz, base image ©Word Info. Line art modification courtesy of Ata Mojlish.

Arturo Escobar (1995, 7) argues that Western discourses of development have become hegemonic in “underdeveloped” countries. He and James Ferguson (1990) use Foucauldian discourse analysis to illustrate the inequality in knowledge production in international development. Their postdevelopment theory of a dominant development discourse imposing itself over actors in the Global South may help explain the frustration Amir and his colleagues express over the power donors hold. An underlying assumption of the hegemonic discourse argument is that donors intentionally exert their power over Southern partners. Based on my previous water management research in such a Western development organization, I thought that a Foucauldian lens of internalizing this discourse could be used to understand climate change as a development discourse, a development governmentality, a power that works to “render invisible,” with varying degrees of success, critiques of anthropogenic environmental degradation such as obstructed sedimentation caused by flood-protection embankments.

While Amir, Professor Hossain, and Gaurav criticize the pressure they feel from donors to neglect siltation, they and their Bangladeshi colleagues also knowingly participate in—and thereby reproduce—these narratives that ignore the multifaceted environmental challenges facing Khulna’s rivers. Over the course of my fieldwork, it became clear that there is not a single dominant hegemonic discourse exerting agency and power in a way that could render other issues invisible, nor have Bangladeshis uncritically internalized donor narratives. Instead, I found a dynamic landscape of development as a practice occurring through complex and interacting processes of funding, translation, and brokerage, which often produces unintended effects (Mosse 2014). Mr. Shahid made this poignantly clear at the GEMOB workshop through the following metaphor of climate change as a spice (masala): “Climate change shobche darun masala [is the most amazing spice]. Add climate change, poverty alleviation, and gender and you will have a recipe for success for your development project [funding application]. But will this recipe help save the river?”

Mr. Shahid’s critique of how he and his colleagues use development spices to secure project funding for their organizations and staff payroll, even though this may not necessarily help “save the river,” struck a chord with me. Like Shahid, I used climate change as a masala and obtained several funding offers for my research. However, at times I felt that my own strategic use of climate change to attract donor funding deflected attention away from the very environmental problems I sought to investigate.

The use of climate change as a spice highlights the importance of examining the relationship between the simplifying rhetoric used to mobilize support in international development and the world as experienced within the lives of development actors (Mosse and Lewis 2006), where one must distinguish between the intentions of those working in the aid industry and the effects of their work (Gardner and Lewis 2015). It thus involves performativity and diversity, enabling an analysis of the actors and networks—their incentives and motivations—required to form, shape, and interpret development actions and interventions, including how they engage with, and participate in, the production and reproduction of dominant donor discourses.

Development professionals (including international agency and donor staff, consultants, fieldworkers, NGO workers, applied anthropologists) have to secure their place within particular and complex institutional and social contexts. Development practice is thus embedded in social relations. “They [development professionals] work hard to maintain relationships, to negotiate their presence within foreign bureaucracies or NGOs for access and influence and manage interfaces within and between agencies. There is the messy, practical, emotion-laden work of dealing with contingency, compromise, improvisation, rule-bending, adjustment, producing viable data, making things work, and meeting delivery targets and spending budgets. In doing so, they must negotiate national identity, race, age or gender. They have to manage personal security, family relations, loneliness, stress, and anxiety.” To negotiate both access and influence, development professionals must make themselves bearers of context-free ideas with universal applicability (Mosse 2011c, 16).

In this case, Gaurav must reproduce climate change as a context-free notion that corresponds to donor ideas of rising sea levels as the main causative agent for environmental problems in the coastal zone. Like Ghanaian NGO workers, Gaurav is not embracing these institutional and political structures uncritically—as local development workers they are “aware that the institutional apparatus and discourses through which development interventions emerge can be as much part of the ‘problem’ as part of the ‘solution.’ At times they profess getting ‘caught’ in the process: that their own actions and thoughts are carried by ideological influences beyond their control and at tangent to their beliefs” (Yarrow 2011, 162).

The sense of feeling caught resonates with the fact that these development practitioners are also actors embedded in the same institutions and projects that reproduce such narratives (Bierschenk 2008; Arvidson 2008; Yarrow 2011). Despite their local knowledge and expertise, Bangladeshi development professionals like Gaurav and Professor Hossain participate in development interventions that they deem ineffective, but by doing so they also contribute to the incontestability of Western models (Rottenburg 2009). Thus, the issue is not only one where donors and policymakers in Dhaka simply exert a hegemonic power over development projects in Khulna. The picture that emerges is far more complicated.

Climate Reductionism: Embankments as Adaptation Infrastructure

A few weeks after the GEMOB workshop and my initial meeting with these Khulna civil society members, I returned to Dhaka to attend the closing conference from my prior research project and to present the project findings. My last project had fully funded this trip. It covered airfare and several nights at the conference hotel in one of Dhaka’s upscale neighborhoods. Entering the marble-clad lobby of this cosmopolitan hotel, I felt that I could be anywhere in the world. The air-conditioned room itself was furnished with a Western spring mattress, an impeccably clean Western-style tiled bathroom, and a flat-screen TV. It was a starkly different experience compared to the humble simplicity of an isolated embanked polder like Nodi where there were hardly any brick houses, let alone electricity or indoor plumbing. Participants flew in from all over the world to attend this workshop. I could hear Professor Hossain’s words in the back of my mind:

All these projects—all these research programs—have large overhead costs, perhaps up to 15 or 20 percent. They are flying in international “experts,” paying them per diems at Western rates. It’s not free. These experts are not CEOs of profit-making companies or working in high-income countries, they are coming to work here in Khulna or Dhaka! Is it then right that these consultants—Germans, Americans, Dutch—are getting such high salaries? There are so many consultants, hundreds of consultants, but no one has the right background to understand the context. When they hire people, they only look at the organizations and degrees, not what they did or what they know. What is the point of this? We have plenty of educated, knowledgeable Bangladeshis, but in this system we are not experts. Instead, a sizeable amount of money is spent on foreign consultants and on projects that replicate each other and spend vast amounts of money on overhead. If I can do a study on how much of the total amount of project funding is actually used for field-level training and providing food at these events, you would be astonished! The costs . . . disappear with all these middlemen: so little money reaches the field. With only a fraction of this [development] money you could do one issue at a time and resolve it once and for all!

Bangladeshi researchers are, despite their excellent qualifications, undermined and discredited among certain pockets of donors and consultancy firms, illustrating how they are not experts in this system (Hanlon 2020). This illustrates a wider tendency that the most important beneficiaries of development live in the North and not in the Global South (Goldman 2005, xi), where thanks to the large budgets of technical cooperation, expatriates profit more from aid than tribal villagers (Mosse 2005, 126–30).

Furthermore, Professor Hossain expressed frustration over the fact that those Bangladeshis who are able to profit and initiate themselves with large-scale development projects with sizeable budgets are often based in Dhaka or abroad. They, in turn, only subcontract smaller contracts to people working in regions like Khulna.

The morning after my arrival, an air-conditioned microbus transported my recently arrived international colleagues and me from the hotel through the congested roads of Dhaka to the venue—a majestic red-brick building characteristic of the Bangladesh government offices. The headquarters of all of Bangladesh’s ministries and their agencies are in Dhaka, along with all the main offices of donor embassies and private consulting firms. The great domed air-conditioned meeting hall was filled with hundreds of people from Dhaka’s development industry and foreign “experts” who had flown in from all over the world.

The conference targeted international donors to showcase both old and new projects. All the conference speeches were thus in English, the donor language. The performative aspects of development brokerage in maintaining such narratives of legitimacy and coherence was illustrated by the minister of water resources who opened the conference. Like the participants of the Khulna environmental meeting, he started off talking about how the flood-protection embankments built in the 1960s under the Coastal Embankment Project resulted in damaging siltation. He, like Amir, brought attention to how these embankments obstruct more than 1.5 billion tons of sediment deposition in the delta, causing many of Bangladesh’s rivers, tributaries and canals to silt up. The top tier of the Ministry of Water Resources in Bangladesh clearly recognized the consequences of embanking four thousand kilometers in the coastal zone and how it had resulted in a widespread problem that necessitates substantial maintenance to support agrarian livelihoods. Contrary to their Khulna colleagues’ beliefs, these policymakers in Dhaka do understand the environmental problems at stake.

Nevertheless, the minister later dismissed the role of embankments in causing siltation and instead stressed that the sediment in the delta requires regular excavation of canals and rivers: “The fault is not with the polders [embankments], but with us. We should have excavated the rivers and canals, which we did not do. The problem is that we lack funding.” This is problematic since the CEP’s extensive embanking of floodplains in coastal Bangladesh was, and still is, unsuitable to the local conditions of the Bangladesh delta with its eroding rivers, heavy monsoon rains, and complex relations between siltation and flooding. Indeed, this unsuitability is the result of modeling the embankments after Dutch dikes (polders) in the Rhine delta, which sees only 1 percent of the sediment in the Ganges-Brahmaputra delta (Hanlon 2020, 30). This illustrates how the historical counterparts of the World Bank, USAID, and the Netherlands implemented the project according to the technical blueprints designed by international consultants. This foreign solution did not translate well into a heavily sedimented, tropical monsoon delta and has resulted in longstanding—and difficult to overcome—problems of siltation and waterlogging.

The existence of polders now restricts the types of solutions available. Bangladesh cannot simply remove embankments as they are used as roads and people have built their homes atop or outside of them. These embankments replaced the old practice of temporary earthen embankments while efforts to remove parts of the embankment to enable flooding and silt deposits through Tidal River Management failed because of the institutional inability to provide monetary compensation in a complex landscape of tenure and under-tenure holdings. Bangladesh must therefore deal with the fact that the extensive embanking of now 139 coastal floodplains requires large sums of money to mitigate the negative ecological effects of donor-funded infrastructure: repairing (and relocating) continuously eroding embankments, excavating canals, and dredging rivers. Even though such maintenance arose in the first place to address negative ecological effects of donor-funded infrastructure, donors like the World Bank refuse to fund periodic maintenance—seeing it as the responsibility of Bangladesh through the Ministry of Water Resources (Dewan, Mukherji, and Buisson 2015). By doing so, the World Bank avoids taking responsibility for the negative effects of projects they themselves supported.

Furthermore, donor-prompted structural adjustment policies, including the downsizing of the Bangladesh Water Development Board (BWDB), exacerbated the state’s inability to meet the regular maintenance demand of the silting and eroding delta (MoWR 2005). They lack both staff and a regular budget for maintenance. The Ministry of Water Resources is therefore highly dependent on donor projects in order to excavate canals and repair embankments (Dewan, Buisson, and Mukherji 2014).

In light of this, it is perhaps not odd that the minister—after having spoken about embankments causing the problems of siltation and stressing how this necessitates regular excavation of canals and rivers in the delta—concluded the speech by casting these very same embankments as a form of climate adaptation: “Polders are necessary, especially in the light of climate change. We must thoroughly redesign our polders; we need to address the threat of rising sea levels. The southwest coastal zone is the most backward region of this country because of salinity, tidal surges, cyclones, and sea level rise. I don’t believe climate change will destroy Bangladesh. We can fight our way. Bangladesh will survive and it will survive very well.”

Not only does this remark illustrate that not all adaptation projects capitalize on ideas of Bangladesh sinking, it also highlights how climate change can be used as a spice even here. In this case, like many official speeches I have observed in Bangladesh’s development industry, it reiterated the project rationale and project goals in the conclusion (polders as climate adaptation), irrespective of the preceding content (polders cause siltation). This highlights the importance of seeing development as a practice that requires joint coherence-making and the rearranging of causal events to legitimize the success of projects (Mosse 2005)—in this case, the introduction of the World Bank’s newest technical intervention, the Coastal Embankment Improvement Project (CEIP).

Yet not all stakeholders were concerned with maintaining this translation, and a joint interpretation must be successful in enrolling supporters (Mosse and Lewis 2006). The following speaker, Dr. Samir, the minister’s colleague at a policy department tasked to coordinate all water issues in Bangladesh, spoke about the problem of heavy siltation outside the embankments. He argued: “We are currently not allowing sedimentation inside the polder, we need to correct this error.” In contrast to the minister, Dr. Samir, whom I later interviewed, did not attempt to make any link to climate change. His organization is rarely involved in donor-funded or capital-intensive infrastructure projects, like embankments.

These remarks about siltation, however, were notably ignored by the next speaker, part of the World Bank’s CEIP, which is implemented by the BWDB. In his speech, this Bangladeshi water professional did not mention siltation. Instead, he focused on the link between the rationale of climate change and his own project: “Due to this region’s vulnerability, the World Bank is spending US$400 million [in] loans to improve the coastal embankments.”

The CEIP casts higher and wider embankments as a technical solution to Bangladesh’s vulnerability to climate change (Government of Bangladesh 2013). It proposes that existing embankments are ill-equipped to help Bangladesh adapt to climatic change and the project would therefore “improve the coastal embankments to increase climate resilience toward natural disasters and rising sea levels” (World Bank 2012). Such a translation of the metacode of climate change connects and legitimizes embankments as a way to deal with rising sea levels. However, since this translation also casts floods as caused solely by rising sea levels, this arguably constitutes a climate reductive translation.

As Dr. Samir emphasized, and as detailed in the preceding chapter, the sedimentation outside the polder and the consequent heightened riverbeds are a much greater threat to Bangladesh’s coastal zone than the predicted rise of sea levels (Auerbach et al. 2015; Brammer 2014). So why was it that the problem of embankments—so well-known to Bangladeshi development professionals in the wake of the 1960s polders as well as documented since colonial times—came to be repackaged as climate change adaptation?

Colonial watertight embankments and railways supported the state and capitalists in many ways: for the state it reduced annual repair costs and served a military function that outweighed environmental concerns and expanded the business and profits of railway interests. Coastal embankments promoted by the World Bank since the mid-1950s allowed a similar deflection from the environmental problems that they caused. How could this similar deflection continue today under the guise of climate adaptation?

Code-Switching: Climate Change as a Metacode

After the conference, I wanted to understand how development brokers who were connected to the new World Bank embankment project legitimized it, considering the problems of an embanked and silting delta. I met Mr. Kazi, one such development professional, in an air-conditioned office in an affluent part of Dhaka. This was our first meeting and I introduced myself as a researcher from a Western university. He pitched the project to me in English—as if I were a donor—and stressed that embankments can protect against rising sea levels. I then asked, this time in Bangla, whether the project is a way of obtaining maintenance funds for the silting delta to excavate canals and repair eroding parts of the embankment. Mr. Kazi switched the conversation back to English: “The current polder system is fully functioning and is in no need of maintenance. However, if we consider climate change: sea levels will rise, cyclones will increase, and floods will turn into permanent waterlogging with increased and erratic rainfall.”

I persisted in Bangla, asking whether there are other problems with embankments beyond global warming. This time he replied in Bangla, providing an in-depth, historical account of the negative ecological effects of embankments, highlighting his long work experience on water issues in Bangladesh.5 I questioned the benefits of the project in light of this, and Mr. Kazi reverted back to English: “The World Bank is funding this project to improve coastal embankments. They are funding this project due to climate change; this is a climate change adaptation project.” This explanation reiterated and emphasized the legitimacy of this internationally funded project and its links to global warming, conforming to the paradigm-maintenance and the policing of knowledge production characteristic of the World Bank (Broad 2007).

Not only did Mr. Katz switch between a public and hidden transcript of the project (Mosse 2005), he also alternated between languages. His switch from Bangla to English when changing narratives is arguably an example of code-switching (Bullock and Toribio 2009), a strategy used by Bangladeshis to negotiate their multivocal identities to maintain a “strategy of neutrality” (Wilce 1998), in this case to balance between identities as an internationally funded project employee and a locally knowledgeable Bangladesh.

Mr. Kazi’s code-switching demonstrates the high degree of agency and strategic maneuvering that Bangladeshi development professionals do to maintain official narratives that can attract funding and make the project appear coherent, and thereby successful. They compartmentalize knowledge to maintain the organizational need for ignorance about what is going on locally (van Ufford 1993). Here, climate change acts as a metacode—an official script (Rottenburg 2009) or a public transcript (Scott 1998) that strips out context to appear neutral and universal. In this case, it glosses over the context of how embankments worsen siltation. Such an acknowledgment of contextual knowledge would require engaging with issues beyond the scope of the project. To discuss long-term solutions often enters into political ground, where donors will not pay for maintenance, such as regularly excavating canals or solving the siltation problem caused by embankments (Dewan, Mukherji, and Buisson 2015). Thus, the metacode is a tool to make development projects appear free from politics.

Knowledge or technologies claimed as universal—such as climate adaptation, as the conversation with Mr. Kazi illustrates—are not context free at all: they are “embedded, albeit in unacknowledged ways, in the particular prejudices and structures of the originating policy-making institutions, and has to lose (or hide) its context and history to become relevant as international development policy” (Mosse 2014, 518). The avoidance of context through the metacode helps resolve differences and enables cooperation between heterogeneous development actors, with differing—and sometimes conflicting—agendas (Rottenburg 2009). The thoughts and actions of Bangladeshi development actors are thus not automatically shaped by an external discourse that they have internalized (cf. Nijbroek 2012). Competing environmental knowledge contained within the same person illustrates that epistemic politics of differing knowledges are not necessarily divided between different groups of “experts” (Vaughn 2017): for example, “scientific” versus “indigenous knowledge” (West 2005), “early adapters” versus “local activists” (Paprocki 2016b), or even different segments of the population (coastal community, project employee, government official). Rather, such a strategy to broker environmental knowledge in a way that helps perform ignorance of local context, by separating public transcripts from hidden transcripts, is particularly useful for ensuring the continued coherence of a project linked to one’s professional survival.

In Bangladesh’s development industry, English is the main language of donor-facing communication and “development performances,” while hidden transcripts are more safely spoken (and written) about in Bangla.6 By taking the opportunity to speak frankly in Bangla about the complex environmental problems and then switching to English to reiterate the formal project narrative, these brokers are able to sustain “the need for ignorance” to stabilize representations of events so the project and its activities appear coherent. They must balance their identities as internationally funded project employees who must provide for their families, and that of locally knowledgeable Bangladeshis.

The Fear of Speaking Out:Paradigm Maintenance and Blacklisting

The active avoidance of context, as clearly illustrated in the code-switching between Bangla and English, arguably forms part of paradigm maintenance, the policing of knowledge and ideology, within and between interlinked organizations (Mosse 2011a; Broad 2006, 2007; Uchiyamada 2004; Woods 2006; St. Clair 2006). This is particularly the case for the World Bank, which has consistently funded embankments following Bengal’s independence from the British Raj. Bretton Woods institutions like the World Bank and the International Monetary Fund are embedded in transnational structures of power, knowledge, and capital that—through their links to US–British geopolitics—generate “the project of development” that fuels highly uneven and unstable relations of capitalist production (Goldman 2005, xvi).

Yet, as ethnographers show, the World Bank is not a monolithic global actor with a singular agenda, but is made up of elite networks that depend on collaboration with a diversity of different actors and institutions both in the Global North and South (Goldman 2005, 11). The different hierarchies of power are also visible within the institutional structures of the World Bank. Anthropologists working with social issues for the World Bank highlight how the organization is dominated by economists and how its publications in development economics are cited more often than those in academic journals (Goldman 2005, 102). The dominance of the economics paradigms, even when bundled with social and environmental sustainability, provide the ontological backdrop for framing goals, definitions, and measurement of development “success” and “progress” (Broad 2006; St. Clair 2006; Mosse 2011c). These form part of a grander “assembly line of knowledge production” inside bank headquarters that is carefully controlled (Goldman 2005). For example, its research department plays a central role in “policing” certain paradigms through incentives in hiring, promoting, and publishing, as well as selectively enforcing rules, discouraging dissonant views, and manipulating data (Broad 2007; Goldman 2005, 148–49) with career incentives, or “golden handcuffs,” that reward “right thinking” (Mosse 2011b).

The careful control and self-policing of World Bank knowledge is evident in Climate Change Risks and Food Security in Bangladesh—a World Bank–funded study that casts embankments as critical “protective infrastructure” (Yu et al. 2010, 105). The report itself is an interesting compilation of contradictions. In the first chapter, the authors acknowledge the distinction between annual monsoon (borsha) floods essential for agriculture and low-frequency, high-magnitude floods (bonna) that are highly damaging to agriculture and rural livelihoods (Yu et al. 2010, xvi, 10). However, in the remainder of the report, they do not specify which type of floods they are referring to. This is particularly problematic when they state that embankments have played a major role in reducing flood risks, but without indicating which type of flood (Yu et al. 2010, 82). Moreover, in certain passages the authors explicitly recognize that (1) substantial information on the local context of Bangladesh is missing; (2) predictions are difficult; (3) the competing processes related to sedimentation and accretion are “largely unknown”; and (4) this is an area for future research. The complex matters of siltation, active river morphology, and transboundary water sharing are omitted from an otherwise sophisticated analysis owing to “resource limitations,” or because they have been slated for “future research.”

Nevertheless, the report creates a climate reductive translation linked to rising sea levels in order to recommend adaptation measures in the form of economic access and support, combined with technical solutions such as irrigation efficiency, which are directed toward individual farmers (Yu et al. 2010, 105). Such solutions are easily incorporated into the current paradigm and institution in which both the authors and the report are embedded. Arguably, this illustrates how development professionals and applied researchers act as brokers between their own organizations, donors, and particular development interventions. This also highlights that the reproducing silences or ignorance of complex environmental issues is not simply a matter of “corruption,” but of negotiating institutional politics of knowledge production.

The political economy of development captures how much of the capital lent by the World Bank passes through the hands of Southern governments and travels directly to firms in the Global North, the main actors who carry out development projects, and who supply the capital goods and services. These dynamics of politics and capital may help explain why the latest embankment project—similar to both the 1960s CEP and 1990s FAP—is passed off as a climate project, with its original critics seemingly jumping on the bandwagon.7 Mr. Balam is a research consultant at a private research organization specializing in GIS modeling in Dhaka. Technical studies are awarded large budgets in the age of climate modeling and assessing climate risk in Bangladesh. His organization has conducted research on water problems in the coastal regions for several decades. When Mr. Balam expressed a critical opinion regarding increased Chinese involvement in Bangladesh, I asked why they are working with them on projects. He then replied: “We [Bangladeshis] only think about our own shorir [health, bodies]; if there’s money we’ll pursue it.”

During our conversation, which was entirely in Bangla, he criticized flood-protection embankments for silting up Bengal’s rivers and lamented how experiments to mitigate the problems have been nearly impossible to implement. I was therefore surprised when he revealed that his organization was in the process of bidding for a research component in a climate change project intended to expand polders and thus further obstruct sedimentation processes. He mentioned their collaboration with various Western companies and top-tier Western universities for a project worth more than US$150 million.

In Africa, Asia, and Latin America, a World Bank consultant earns more than thirty times that of an equally qualified economist, making it much harder to say no to the World Bank (Goldman 2005). At first glance, one could write off Mr. Balam as “greedy” and participating in development to “extract development revenue” (Bierschenk, Chauveau, and de Sardan 2000). However, when I asked him what the project will achieve, he replied: “More studies, and very little implementation. Implementation is difficult.” He then began to tell me about his involvement in tidal river management to deal with the problem of siltation in the 2000s:

We tried a project where we purposefully broke the embankment during the monsoon to get rid of jalabaddho. The problem with this is the issue of compensation: does it go to the absentee landlord or cultivator? This has not been resolved in terms of how we compensate those that are affected. The main thing is that water needs to be allowed inside the polder. The riverbed is rising, and it is rising rapidly. This was highlighted in previous studies as well. In a large study a few years ago, they had initially only looked at polder management but realized that they also must look at drainage which is connected to siltation: the sluice gates stop working due to siltation. But building higher embankments as the World Bank suggests is not sustainable. It will only increase the risk by imbibing a false sense of security. Our embankments are not built like Dutch dikes. The construction itself is vulnerable.8

This is at the crux of paradigm maintenance and technical solutions: it is too difficult to implement what is required. By participating in the technical game with climate change as a metacode, Mr. Balam can earn considerably more than if he had openly criticized donor projects. Yet, perhaps even more importantly, considering his previous experience of attempts to resolve siltation, there was also a self-realization of the difficulty of implementation and that short-term projects cannot resolve the vast problems of the coastal zone. He ended our conversation with the following words: “There is no real democracy in Bangladesh. It is limited, an imported model from the UK. The right to speak exists formally in Bangladesh, but people are afraid to speak.”

The decision to accept donor projects, regardless of misconceived ideas, is not only about financial and career incentives. Nor is it simply a matter of “corruption.” As Mr. Balam points out, there is a fear of openly criticizing projects. A focus on development as profit gained through collusion could analytically miss the importance of understanding that to not participate in this game may entail blacklisting and losing one’s livelihood that sustains not only oneself, but also one’s wider family network. Like one of my interlocutors pointed out: the development industry is an important employer for middle-class, educated Bangladeshis and for aspiring rural youth. To maintain their professional careers, to enable the survival of their organizations and the paychecks for their staff, to support their families, they choose to reproduce these donor narratives in their translations of climate change. This reflects the wider issue of structural constraints in the development industry, as several Bangladeshi development professionals, consultants, and researchers that I interviewed stated that there was no point in discussing how another embankment project might be futile or make things worse. They stressed the fear of being “blacklisted,” that their work would be dismissed and discarded for mentioning local realities that jeopardize state politics or project legitimacy.

During our discussion of riverbank erosion and sedimentation, I asked Mr. Manzur, who once held a leading position at BWDB, why new and higher embankments are being built when all parties involved know it is ineffective and does not resolve the siltation of canals. He replied: “Rajar kono dosh nei [the king is never at fault]. The donors are the ones with money; thus, they are above criticism. A few years ago, I published an article in an English-language newspaper publicly criticizing the World Bank. I asked why the World Bank will not bear the yearly maintenance costs of the Jamuna [Brahmaputra River] as the embankment is damaged in several places and it was part of their project. I was promptly blacklisted; this was over a decade ago.”

Manzur draws attention to how his open critique in an English-language paper entailed publicly violating the technical game and the resulting consequences. He lost his current position and was no longer able to find equivalent positions, as others feared that association with the blacklisted individual would reduce their own chances of participating in similar projects. In this sense, Manzur was no longer able to participate in the technical game as his capacity to sustain his networks was heavily reduced.

The efforts to maintain certain project narratives (paradigm maintenance) could be interpreted as a top-down exertion of World Bank power, silencing those opposing the dominant narratives. Large-scale coastal embankments have been controversial since the Krug Mission Report of 1957, long before discussions of embankments as climate change adaptation. The Coastal Embankment Improvement Project, like the Flood Action Plan (FAP) of the 1990s and the colonial railways of the 1880s, is a capital-intensive infrastructure project. The CEIP, in addition to the costly infrastructure component, also entails the execution of several well-paid studies. Powerful financial interests who stood to profit lobbied for both the construction of railways and the FAP—development interventions that saw strong opposition from civil society and those with local knowledge, like Sir Arthur Cotton (chap. 1).

In the Flood Action Plan, several (though not all) Bangladeshi and international construction firms, research consultants, and development organizations sought to take advantage of large-scale infrastructure projects funded by international donors that would “maximize the flow of funds” (Hossain, Islam, and Saha 1987, ii; Shaw 1992; 2014, 232). Those not agreeing with the formal narrative of the FAP were also blacklisted. For example, an “expatriate” team for a USAID study for the FAP suggested that Bangladesh ought to live with floods rather than attempt to control the unruly and dynamic delta through costly embankments (Rogers, Lydon, and Seckler 1989). Their nuanced analysis and clearly engaged understanding of the complex environmental processes of the Bengal delta explicitly identified the problems of low-frequency and high-magnitude floods (bonna) and the necessity of regular monsoon floods (borsha). The study infuriated then Prime Minister Ershad for its “defeatist tone” and led the US ambassador in Dhaka and other US officials to distance themselves from the report (Boyce 1990). The authors and their report were dismissed and were not even invited to attend the final workshop meetings of the FAP (Boyce 1990).

The team’s dismissal highlights the link between blacklisting and paradigm-maintenance—that is, the policing of knowledge production—as these dissidents and their research unsupportive of (profitable) large-scale embankments were in effect boycotted. The debates and concerns on flood-protection infrastructure in the 1990s contain important parallels to today’s discussions on climate change adaptation. Government officials, NGOs, and donors work together to adopt climate change as a metacode for their projects. If one does not participate in this technical game, the development professional and his or her organization may become blacklisted and end up completely outside the development industry. Nevertheless, by participating as brokers in this aid game, Bangladesh’s development professionals run the risk of reproducing silences over important problems such as obstructed sedimentation, which in turn results in deteriorating rivers, jalabaddho, and increased vulnerability to climatic risk.

Translating Climate Change Adaptation in Nodi

The World Bank’s CEIP project actively avoided the issue of siltation and erosion in the Bengal delta, but by no means are climate change projects in Nodi part of an all-powerful development discourse that seeks to render siltation invisible. Rather, power is dispersed and reinforced through the strategic tactics of development actors that belong to widely different development assemblages. Climate adaptation projects shaped by development assemblages more aligned with local interests, such as Nodi’s poorer constituents, can be bottom-up and locally relevant.

Caring for Dying Canals

My interlocutors in Nodi consisted of female and male landless agricultural day laborers and farmers. They viewed the canals as public goods for irrigation and for catching wild fish; most expressed concern over the many dead canals (mora khal) where there had once been deep waterbodies connecting the river to the fields. For example, Nitesh spoke about the once great Bhadra River where steamboats could pass, but which is now merely a silted canal: “The Bhadra River is now the Dead Bhadra [Mora Bhadra]. Siltation caused the canals to die. Riverbank erosion damages the embankment and might wash away our homes any time. We can no longer catch fish from the canals or irrigate our crops.”

Figure2.2. Canal excavation in Nodi, May 2015. Photo by author.

Several people in Nodi emphasized that if these public canals were excavated every five years (even excavated canals silt up over time) they could grow more crops locally while retaining more water from the monsoon for the water-scarce dry season. This was in response to a highly praised climate change adaptation project entitled Enhancing Resilience to Disasters and the Effects of Climate Change, which was funded by the World Food Program (WFP) together with “new donors” such as Brazil and Japan. This cash-and-food-for-work scheme provided rural training and work to women and men in the coastal communities in Bangladesh.

Nodi experienced the first phase of the project, when it hired around two to three hundred people to excavate large canals that had silted up across the polder (see fig. 2.2). It thus provided income opportunities to local families, including landless day-laboring single mothers without husbands. By widening and deepening the heavily silted canals, the project envisaged a greater absorption capacity of monsoon rains—expected to increase with climatic variability—while allowing for water storage from the monsoon, thus providing a space for fishing and irrigation as a public common. This labor-intensive project saw the main cost of wages deposited to the private bank accounts of locally hired labor and also distributed food grants through a local NGO. Unlike the (inter)national research bids for embankment research in Dhaka, the project was designed to be bottom-up, where local government officials, community members, and NGO staff discussed and identified which of these activities to implement (World Food Program 2012). Whether climate change adaptation projects can benefit local populations or not depends on how climate change adaptation is translated in the shaping of a development intervention, which itself depends on the constitutions of donors and their implementing partners in specific development assemblages.

Badrul’s project used climate change adaptation funds to install wells and latrines in rural areas and was locally relevant and valuable, as was the World Food Project’s interventions that both excavated canals and provided local employment opportunities. The different activities of these two projects—safe drinking water and canal excavation as a remediation of siltation—complicate ideas of climate change as an adaptation regime causing agrarian dispossession (Paprocki 2018).

Conclusion: Why Critique Climate Reductionism?

In his 2004 article “Why Has Critique Run out of Steam? From Matters of Fact to Matters of Concern,” Bruno Latour—a key figure in the social construction of science and the lack of scientific certainty—reflects on skeptics appropriating critical studies of science for their own agenda (2004, 227). He laments: “Why does it burn my tongue to say that global warming is a fact whether you like it or not? Why can’t I simply say that the argument is closed for good?” For Latour, to understand the social construction of climate change science is to renew empiricism and to get closer to the facts. Even something as real as global warming can also be socially constructed. This is why the distinction between material climatic change and discursive climate change is useful. Focusing on a better understanding of climate reductive translations illustrates how distinct assemblages of development brokers come together to make a climate-related development project exist and how they actively work to maintain its existence. Global warming is a serious cause for concern, but the way in which it is translated into everyday politics and economic activities is one that is embedded in social context.

Climate change as a spice, or metacode, not only attracts donor funding, but provides a space for resolving tensions of actors and agencies with different agendas. Bangladeshi development brokers compartmentalize their context-specific environmental knowledge by code-switching between the public transcript in English and the unofficial, hidden transcript in Bangla. This enables them to participate in the technical game and maintain ignorance as they recognize that a long-term strategy is unachievable within a development professional’s short-term project tenure. In the World Bank’s Coastal Embankment Improvement project, climate change adaptation was translated into a seemingly neutral and technical metacode that removed local context and rearranged the causal reasoning that legitimized the project. By doing so, it neglected the problem of how embankments exacerbate siltation and damaging jalabaddho floods that may worsen the livelihoods of coastal people.

The translation of climate change depends on specific assemblages of brokers and multiple heterogeneous development actors to highlight how the translation of embankments as adaptation does not necessarily form an antipolitics machine or an adaptation regime. When actors with different agendas, priorities, and knowledge backgrounds come together, they form a distinct assemblage of socially embedded relations that will jointly create a shared translation—linking their project activities to the metacode of climate change—that will be different from assemblages consisting of other brokers with a different set of diverse agendas. It is the composition of brokers and donors in these assemblages, their various institutional agendas and mandates, that shape whether the final interventions fit the needs of the populations they are seeking to assist. Thus, the lack of benefit to local people in the World Bank project may illustrate that this specific assemblage of brokers is too far removed from rural perspectives to partake in the act of translating climate change to the latter’s benefit. It may also illustrate that the officials involved in World Bank projects are better at harnessing dominant narratives for their own particular agenda. Furthermore, the assemblages of capitalist interests and state capitalists that benefit from capital-intensive projects and studies may not be interested in a translation of climate change that creates benefits for the rural poor.

Climate change is one of the most pressing issues of our time, and it is important to critically scrutinize the official scripts of development projects using climate change as a metacode. In this way, those who pursue old agendas under the guise of climate adaptation and resilience may exacerbate Bangladesh’s climatic vulnerability now or in the long-term—as embankments inevitably do.