The kind invitation to give a seminar in the series called Flooding the Field gave me the opportunity to reflect on and discuss some experiences and ideas about interdisciplinarity in flood-related research. I shared some of my experiences with the Sustainable Flood Memories project, where I had worked in 2011-2012, conducting research along the River Severn in Gloucestershire. I introduced the idea of “hydrosociality”, which I had come across and learned to appreciate during this project: I had learned that floods are never just about water levels. What makes high water into an emergency or even a disaster is the particular ways that political and material arrangements channel both where the water goes and who is most vulnerable. That the material and the political are not just closely related on the floodplains, but that they are very often two sides of the same coin is epitomised in the slogan that I came across repeatedly and which stated ‘One man’s flood defence is another man’s flood!’ This rendition of the popular cliché ‘One man’s terrorist is another man’s freedom fighter’ juxtaposes flood defences and terrorism, political oppression and flooding, and thereby illustrates how people at flood risk experience high water as a simultaneously hydrological and political threat. 

Of course, the hydrosocial dynamics of flooding characterize much of my current research in the Mackenzie Delta in Arctic Canada, too. For example, the roughly 600 inhabitants of Aklavik, a hamlet in the middle of the Mackenzie Delta live with high flood risk. In fact, it was this flood risk, along with erosion and general muddiness, that led the Canadian government to decide in the 1950s that the delta was not a safe place for Aklavik, and to relocate it onto dry and higher ground on the edge of the delta. This new town, with specially developed infrastructure to provide water and sewage disposal despite the Arctic conditions, and other modern, Canadian features, came to be known as Inuvik. However, not all Aklavik inhabitants were keen on moving to Inuvik, since it would have been too far from their hunting and fishing camps and their winter traplines. The main course of the Mackenzie River runs between Inuvik and Aklavik, and crossing this watercourse was tricky on windy days during the summer, and near impossible during early winter and late spring, when the ice conditions are a lot less reliable there than elsewhere in the delta. Moreover, many of those who did move, lured for example by the better job opportunities in Inuvik, soon returned to Aklavik, having found out that the well-paying jobs and the modern housing were reserved for white Canadians from the South of the country, while Indigenous people were limited to low-paying, menial labour and were housed in what came to be known as the ‘tent city’, far away from the fancy buildings and infrastructures. 

Meanwhile, back in Aklavik, large floods have become less frequent, which is caused, among other factors, by a reduction in the thickness of river ice. Nevertheless, they still happen occasionally. The last large flood happened in 2006, and involved airlifting a quarter of Aklavik’s population into Inuvik. One of the most interesting aspects of this flood was that the water didn’t actually enter any houses, since all currently inhabited dwellings are constructed on pilings above ground so as to not interfere with the permafrost below. Also, most people own boats, so getting around wouldn’t have been an issue either. What turned the flood into an emergency was that Aklavik, too, had gotten used to running water, flushed toilets and sewage disposal. With the floods, the trucks supplying fresh water and pumping sewage from each house’s private tank had trouble reaching some of the houses. Additionally, the road towards the place where the sewage truck was supposed to dump its load, affectionately called “Shit Lake” locally, was flooded, so that there was no safe place to dispose the wastewater. In short, it wasn’t the water in itself that made this flood problematic, but the infrastructure that emulated modern, high ground technology, including flushing toilets, roads, trucks, etc.

In the spring of last year, Aklavik people were anxious about another possible flood. The ice had been unusually thick in the winter, which made it liable to jam upon breaking up in late May and cause water levels to rise. The hamlet administration had improved the road to the sewage dumping point and felt fairly confident that they would be able to handle a flood, but it had a new source of worries: in case they would have to evacuate people to Inuvik, as had been the case in 2006, it was not clear at all where they could be lodged given the current Covid-19 pandemic and the respective physical distancing rules. In 2006, the evacuated families had stayed with relatives and friends in Inuvik, or were housed in the town’s recreational complex for the few days until the waters receded and it was safe to return home. This seemed not a good option in 2020. While the flooding issue had thus shifted from sewage infrastructure to epidemiological considerations, it turned out that there was no major ice jam around Aklavik and therefore no flooding or need for evacuations.

As this illustrates, a hydrosocial approach can show how water flows are implicated in a wide gamut of other dynamics, including political economy, livelihoods, infrastructure, colonial exploitation, and epidemics, to name but a few. Bringing together hydrological and social dynamics is not only a concern for anthropologists and human geographers, however, but also of growing interest among hydrologists and engineers. Just as anthropologists have realised that society and culture are intricately related to water flows, so have hydrologists understood that it makes little sense to analyse water flows without taking account of human agency and infrastructure. The latter call this approach “socio-hydrology," where “humans and their actions are considered part and parcel of water cycle dynamics, and the aim is to predict the dynamics of both." What at face value appears like two complementary approaches – one taking social research into the water, the other extending hydrology into society – are, however, two sets of radically different epistemologies and trajectories, the one interested in the fine-tuning of hydrological models, the other in the critique of hegemonic and unsustainable environmental ideas and practices. Just like the term ‘memory’ meant rather different things to the interdisciplinary research team of the flood memory project, the conjunction of water and society means something else to hydrologists than to anthropologists.

In recent years, a number of commentaries have juxtaposed socio-hydrology and hydrosociality in order to gauge what combines and unites them. Anna Wesselink and colleagues, for example, note how socio-hydrology aims to solve practical problems in society-water relations, notably flooding, by attempting to formulate an integrated, quantitative model with both social and hydrological inputs akin to Earth System Science. They characterise hydrosocial research, conversely, as a child of political ecology, focused on critique rather than providing solutions, and interested in the mutual constitution of society and water, rather than their interaction. Although I find that Wesselink and colleagues are unnecessarily stereotyping the two approaches in contrasting them, for instance by characterising hydrosocial approaches as “constructivist” and thereby ignoring the burgeoning field of new materialism, their discussion identifies shortcomings to both approaches and points into a useful direction. It also addresses the fact that hydrosocial research is no match for socio-hydrology and other Earth Systems Science approaches, which command a far greater amount of prestige and funding in the current academic and policy world, so that there is little hope in trying to integrate the two perspectives. Instead, the authors suggest that the two research traditions keep working in parallel, with each approach "recognising respective strengths without either party setting out to change the other’s paradigms."

Maria Rusca and Giuliano Di Baldassarre have also considered the relation between these two approaches to the mutual shaping of water and society. Invoking the ideal of interdisciplinary research as producing more nuanced analyses for better informing policy, they concentrate on how these approaches are compatible and complementary. While they do identify some superficial divisions, which can be – and often are – crossed in actual research, like the qualitative-quantitative divide, their compassionate argument for a convergence of water and society research systematically circumvents the question on whose terms and for what purpose this integration is meant to happen. Both authors being based at an Earth Science department, it comes as no surprise that their argument reads like an advertisement geared at critical geographers to engage with socio-hydrology, with little if any movement in the other direction. It remains unclear who or what is supposed to benefit from an integration of hydrosocial research into socio-hydrology, other than the same governance and management bodies, and with that the assumptions and powers that be, which are routinely criticised by critical geographers.

To me, these two commentaries – the contrasting one and the integrative one – indicate that different approaches to the social in flooding are justified, and that attempting to integrate them would do more harm than good. Casper Bruun Jensen and Atsuro Morita make a similar point arguing that Earth Systems Science’s modelling of river delta crises systematically performs “the social” as various different things, only some of which can be integrated in the model. The latter include particular quantifiable reductions of social life, perhaps such things as population density, residence preference and land use. But alongside as an object of their models, Earth Systems Sciences understand the social also as the context of their models, for instance in the guise of policies that should take the models to heart. Moreover, Jensen and Morita identify a third variant of the social, which the modellers of social-ecological systems ignore altogether. That is the positionality and agency of the researchers themselves, who – through their very modelling work – contribute to configuring water-society relations. As an alternative strategy to this awkward attempt at integrating different scales and purposes of analysis, Jensen and Morita propose what they call “sophisticated conjunctions” of various approaches that bring different, but often incommensurable, insights to understanding a given problem, thereby making up for each other’s disciplinary blind spots and other specific limitations. Key in this approach is acknowledging, and then working across, difference, rather than trying to integrate these differences away into the smallest common denominator. This chimes with Wesselink and colleagues’ suggestion to take “narrative” as a boundary object between hydrosocial research and socio-hydrology, as both fields create particular narratives to make sense of their study material. 

What I take from this is that, first, there may be a point to pause and consider for what purposes and on whose terms research paradigms are to be integrated, and whether this is something we want to participate in, especially given the current political economy of academia, where critical water studies are extremely marginal compared to mainstream hydrological sciences. We must also get rid of the illusion that different disciplines have singular access to different aspects of the world. We can approach the social in flooding from the perspectives of different disciplines, each of which asks a different question of this relationship and uses different methods and paradigms to answer them. Water is not the sole purview of hydrology, and social practice not owned by anthropology. And third, we might want to take a cue from what people in floodplains do and how they navigate the multiple knowledges and their hierarchies produced by the disciplinary division of labour in academia and policy-making.

The people of the lower River Severn in Gloucestershire, for instance, did not seem to make a fundamental difference between state-sanctioned hydrological facts and their own experiences and stories about flooding. Of course, they were acutely aware of the hierarchies in these knowledges, for instance when it came to Environment Agency support or insurance policies, but they also knew that everyone was feeling their way through the dark, including the flood modellers and urban planners. One most amazing floodplain inhabitant from the city of Gloucester, whom I had the pleasure to meet, was Ray Armishaw. He was passionate and very knowledgeable about the River Severn, for instance about the tidal bore that travelled up the river under certain circumstances. But he was equally passionate and knowledgeable about flooding, and very critical of the expert opinions and local council policies that kept allowing more obstructions like causeways, buildings and a solid waste disposal site to be placed in the floodplain, which would redistribute flood water and aggravate flood risks. In this video, he explains some shortcomings in a hydrological model, using a knitting needle and a camera filming footage on a television and an old photograph. Along with other clips, he gave this to me so that I could have it digitized at my university and put it online for others to access. Ray Armishaw has no use for discussions about knowledge integration, since he is able to himself make “sophisticated conjunctions” between different domains.

Similarly, in the Mackenzie Delta, people look into a variety of sources to judge how high or risky the coming flood may be. Some of the factors are local, like the thickness of the river ice. This is something people know best from ice fishing, for which they drill or chisel holes through the ice. Depending on how much they have to drill or how long it takes to chisel, they know what to expect. But other factors cannot be directly experienced, like the amount of snowpack in the catchment or the development of temperature and wind. Some Aklavik Inhabitants therefore follow updates on social media or subscribe to an email newsletter with the so-called “Mackenzie-Beaufort Breakup Report” through which the Geological Survey of Canada provides satellite images and hydrological data every few days during the flood season. People then assemble the heterogeneous knowledges from break-up reports, local ice thickness, and a host of other sources of information in order to decide, for example, how long it is still safe to travel along particular routes, or when to prepare their camps for the floods. Importantly, they do not need to “integrate” these sources of information, but they are useful for their decisions anyway.

Perhaps there is a lesson for us as researchers in this. Perhaps it’s fine – even beneficial – if socio-hydrology and hydrosocial analysis do things differently. As an anthropologist, for instance, I am regularly very grateful for solid hydrological research on the watercourses where I work. For example, it would have been extremely difficult for me to find out that the annual spring flood peak in the Mackenzie Delta has shifted forward by eight days since the 1960s, were it not for hydrologists to have measured and analysed this. But this is useful for me not because we have an integrated framework or feed into the same model, but because it provides me with another way of knowing a transforming delta. 

When we explore the relationship between floods and society, we find fundamentally different knowledge practices, with very different approaches and purposes. Some aim at predicting and planning, others at facilitating cohabitation, some focus on the catchment scale, others on the neighbourhood, the household or the global climate. Such diversity may indeed be key in addressing this problem. Asking different questions of the same issue will yield different answers, but these answers are likely to complement each other and cover each other’s blind spots. “Integrating” these questions and perspectives into a common framework, based on the lowest common denominator or – worse – on the paradigms of the best-funded disciplines, would be a loss. For tackling real-world challenges, pursuing incommensurable questions remains as crucial as collaborating. And in collaboration, we must pay attention to existing power differentials between knowers and knowledges, for instance between the unquestioned authority of large-dataset, quantitative approaches – as useful and convincing as they may be – and small-scale, qualitative studies that are too easily discounted as anecdotal, unrepresentative opinion pieces with little scientific relevance. In short, the relation between floods and society are too multifaceted to leave them to any one approach or integrated framework. Disciplinary multiplicity is helpful in developing different questions and perspectives. If anything needs to change, it’s the willingness to listen to each other and support “sophisticated conjunctions” between approaches.