Printer Friendly

Environmental research from here and there: numerical modelling labs as heterotopias.

Abstract. This paper examines the production of local and remote knowledge about the flows of a large Italian lake (Lake Como) on the basis of a dual ethnographic engagement with lake fishermen and computer modellers. The focus is the relationship between these two ways of knowing a lake, in relation to the lake itself and the places where the knowledge is produced. I first argue that despite being seemingly located at different edges of the environmental knowledge spectrum--the local and the remote and computer-mediated--the fishermen's and scientists' knowledges emerge from similar forms of skilled engagement in their respective environments. In other words, knowledge develops in place and fisherman and scientist do not know the same place in different ways; rather they know different places in similar ways. This raises the question of what kind of places these environmental modelling labs and offices are, and how they relate to the places they talk about, which are located elsewhere. I propose here the use of Michel Foucault's concept of heterotopia to model this complex articulation of the 'here' and 'there' in environmental modelling.

Keywords: environmental modelling, heterotopia, science and technology studies, lake, practice theory

"Knowing must be reconnected with being, epistemology with ontology, thought with life." Tim Ingold (2011, page 129)

Introduction

It was a cool autumn evening on a large lake in Northern Italy. I was on a small fishing boat, staring at a yellow buoy in the water a few hundred meters away. Alessandro, a local fisherman, was trying to hold the boat reasonably still near the rocky shore. He asked me: "So, what do you see?" Unsure, I asked whether the buoy was going down (meaning south). "No, it's going up [north], and it's quite fast", he replied. Only a couple of weeks after starting fieldwork at the lake, my eyes were not trained to look at it from its shores. I clearly remember catching myself thinking how much easier it would be to evaluate the buoy's displacement from the sky, almost wishing I were back in my Australian university office looking at a map-like representation of the lake on my computer screen.

The lake was the Lario, also known as Lake Como, which extends three long arms between the alpine foothills of Lombardy. I had studied its internal water currents from Australia, as part of a PhD research project in physical limnology. I had later come to the lake as a postgraduate student in anthropology,<u this time to learn about the lake from its professional fishermen, one of whom was Alessandro. I was interested in exploring different ways of knowing the lake, through the sociocultural and spatial contexts where these understandings had developed.

Processes of knowledge production, scientific or not, are always embedded in places. In the words of the philosopher Edward Casey: "To live is to live locally, and to know is first of all to know the places one is in" (1996, page 18). The places where scientific knowledge is produced have been exposed to the eye of science studies scholars for several decades (eg, Knorr-Cetina, 1999; Latour and Woolgar, 1986; Turnbull, 2000), but much remains to be explored about the relationships of scientific knowledge production with concepts of space and time (Law and Mol, 2001). This is especially the case as the practice of science has become increasingly computer-mediated and based on simulation models that are designed to produce knowledge claims about 'other' places and times than those they are embedded in.

Computer modelling activities, ubiquitous in environmental science and engineering, contribute in important ways to the shaping of discourses about environmental systems (Oreskes, 2003; Yearley, 1999). It is important to understand these practices in relation to the places they talk about. Environmental models come in various forms [see, eg, Wainwright and Mulligan (2004) for an overview]; the process-based models that combine scientific theory of natural processes in the form of mathematical descriptions with computational techniques are the most widely used in environmental fluid dynamics, and provide a particular focus for this paper.

In this paper I am addressing the conceptual and methodological question of the articulation of the 'here' and 'there' in computer-mediated environmental research, through the comparison of two ways of knowing Lake Como: one pertaining to local fishing practices and the other to the development and use of a computer simulation model of the lake. More specifically, I am interested in the following questions: (1) how does knowledge about a lake develop in the distinct contexts of remotely based numerical modelling and local fishing practice? (2) what are the similarities and differences between these ways of knowing? And (3) what kind of place is an environmental modelling lab and what is the relationship between the numerical model and the environmental system it is designed to simulate?

I use the term environmental system in relation to those ecosystems that environmental scientists study and model. In environmental science contexts, these are broadly understood as parts of the 'natural world' in general, which are affected by human activity--a disembodied definition of 'environment' that has become prevalent since the 20th century. Somewhat contrastingly, when I refer to environment in this paper, it is in line with the more phenomenological definition of an environment as relational to an organism--the space within which it exists and engages (Ingold, 2000). (2) For example, with this terminology, Lake Como was the environmental system that I studied as an environmental scientist, while my environment was the network of physical, social, and digital places within which my research practice unfolded, centred around an Australian university research institute.

Heterotopias

I ground my analysis in ethnographic data and the concept of heterotopia, proposed by Michel Foucault (1966a). The word heterotopia contains the notions of alterity (hetero) and place (topos); Foucault proposed it in reference to particular kinds of places, literary or physical, that problematise the grounds of knowledge by juxtaposing and combining multiple places that should normally be incompatible (1966a). In his 1967 lecture to a group of architects, he defined heterotopic places as real emplacements (and thus different from utopias) that simultaneously represent, contest, and reorder other places (2004).

Foucault did not develop the notion of heterotopia further than in these brief interventions of the mid-1960s, which has since led to a wide array of interpretations and uses of the term (Johnson, 2006). The scope of this paper is not to offer another interpretation of the term or a critique of its usages. (3) Rather, I simply borrow it to think through and articulate my ethnographic data. I argue that heterotopia constitutes a helpful frame to examine the spatiality of knowledge production in environmental modelling, because by enclosing and juxtaposing different places and times it exposes and problematises their connections, which is what I am interested in exploring here.

The first section below sets the scene for two ethnographic narratives of knowledge in the making that will constitute the basis for the analysis. Building on these intertwined stories of fishing and computer modelling of Lake Como, I use the concept of heterotopia to guide an inquiry into the spatial, temporal, and practical relationships between the numerical simulation model and the environmental system (the lake) it is said to represent.

Lake Como: a waterscape, a water body

Lake Como is a deep, long, and narrow lake in northern Italy, surrounded by mountains that channel regular winds. Except in times of exceptional fog, one can always see the opposite side from any point on the shore, which inspired the poet Percy Bysshe Shelley to describe it as "a mighty river winding among the mountains and forests" (1882, page 25). The three arms of the lake extend north, southeast, and southwest, respectively, totalling approximately 170 km of winding shoreline. The southwestern or Como arm is the most intricate, with many places on the shoreline revealing only very small portions of the lake from the shore. The southeastern arm (Lecco) is edged by sharp mountains, while the northern arm (Alto Lago) is closer to the Alps, wider than the other two, more open, and exposed to the winds.

The closed and mountainous landscape of Lake Como is said to have shaped the mentalities of the people from the lakeshore villages, sometimes referred to as laghee or laghisti and described as down-to-earth and strong-minded people (Pensa, 1981). Most of the villages north of the urban centres of Como and Lecco count a few hundred inhabitants; they were built on the steep mountainsides, down to the lake where a few small boats are berthed in protected docks. When I conducted research by the lake, my main ethnographic concern was with the people from these villages for whom the lake is a daily life and work environment: a group of about thirty professional fishermen between nineteen and seventy-eight years old, operating on the lake all year round. (4)

When I started my PhD in Environmental Engineering at the Centre for Water Research at the University of Western Australia, my research interest in Lake Como was limited to its physical attributes. The lake was in many ways like the dozens of other lakes my research group had been studying: a bounded body of water exchanging water, heat, and momentum with the atmosphere above, resulting in internal water movements. As physical limnologists, we sought to understand these water currents in bodies of standing water in general and to predict them for specific lakes, estuaries, and seas. I began studying Lake Como through this lens late in 2007; the lake had become my study site mainly because relevant atmospheric and hydrological data were already available.

The first two years of my studies were focused on physical limnology. During that time I saw the lake only once, on a one-day boat trip with scientists from Milan. I had no awareness of the professional fishermen of the lake. I was working under the supervision of two Australian academics specialising in environmental fluid mechanics, and I was in close contact with other students focusing on similar topics but different sites. My studies were based on data from three buoys (which I named lake diagnosis systems, hereafter LDS) permanently installed in the lake at strategic locations (which I named 77, T2, and T3) with meteorological stations to record atmospheric data, and downward chains of temperature sensors. The sensors recorded water temperature every metre or so from the surface down to the depths at the buoy location (up to 150 metres). Once transmitted in near real-time via satellite to the Perth-based research centre, these data allowed me to see variables of the lake environment [such as temperature (eg, figure 1.1), wind speed and direction, radiation, humidity] and use them to produce a three-dimensional computer model of the lake.

In the following section I present two narratives of daily routines: one of lake fishing and another one of academic limnology. In both stories Lake Como has a central place. The two stories cover a 24-hour time period and are based on data collected at different times between 2008 and 2012, when I was alternately conducting research in Australia and Italy. The fishing trip described took place on 27 and 28 October 2010. It was one of several trips in which I took part during fieldwork that included participant observation, ethnographic interviews, and mapping of winds, fishing patterns, and lake currents with the lake's fishermen. The limnologist's work narrative is drawn from my own notebooks, e-mails, and memories relating to the time spent studying the lake from Australia in 2008 and 2009 and then again, more reflexively, between 2010 and 2012. My office was located in a research institute focused on environmental fluid dynamics, where I interacted daily with other researchers from this field.

There were countless nights in 2008 and 2009 when my numerical simulations of Lake Como's hydrodynamics and the fishermen's nets ran at the same time, along with the lake's flows. Writing up the two stories now allows me to present them as happening in parallel, one at the lake and the other at the Australian university office, in a way that supports a comparative analysis of these two ways of knowing the lake.

Crosscurrent practices of fishing and limnology

Getting ready

Alessandro fishes in the central area of the lake, facing the tiny village that has been his family's home for many generations. He is a son and a grandson of fishermen. He lives very close to the small self-managed fish-processing facility where he transforms the fish caught for retail with a colleague, the family-owned restaurant that specialises in lake fish, and the small harbour (molo) where he keeps his two fishing boats.

That day, in the late afternoon, Alessandro is preparing his gear to go fishing. It is a sunny autumn day with "a bit of vento [strong northerly wind] in the air", as he puts it. He is planning which nets to set and where, going by the weather conditions he expects for the night and the last few days' catches: besides a few whitefish and perch, he aims to catch some shad (with nets called pendenti) to start preparing the missultin, a local dried-fish specialty. He will set three sets of nets before sunset, let them work through the night, and retrieve them at dawn, which is the usual rhythm of the lake's fishermen.

At this time of year, the shad are found near the surface, while the whitefish are caught at a depth of about 30 m and go deeper as the season advances. Yesterday morning Alessandro found most of the whitefish caught in the lower part of the nets (called oltane), so he has increased the length of string connecting the floats to the top line of the nets. As for the sedentary fish (perch), his father and grandfather had shown him productive spots that he still exploits and maintains, alternating with other spots to give the fish schools enough time to recover between catches. Alessandro runs each series of nets through his fingers to check for tears and disentangle any major knot. As he goes along he coils the nets onto aluminium rods that will be placed onto purpose-made holders that he has had specially fitted on the boat to support his work routine. Most nights on his shore the shallow lake current goes north. He gives a quick, habitual look at the long weeds that grow underwater near the harbour: they are bending toward the north. He is anticipating that his pendenti will drift north through the night and preparing to set them south of the harbour, so he can track and retrieve them more easily as they move closer to his house. The afternoon light is starting to soften; Alessandro is almost ready.

[FIGURE 1 OMITTED]

In Australia, some 15 000 km away, I sit in a university office behind a desktop computer, with a map of Lake Como on the wall above the screen. To my left is a window with a view of a garden, to my right a PhD student from Venezuela who studies the coastal ocean of Western Australia. Behind me sits the third student sharing the office, from China and studying a hurricane in the Caribbean Sea. For two years, I have been working with numerical data of water temperature and meteorological variables automatically recorded every minute in Lake Como and transmitted via satellite. 1 process these matrices for computer-mediated analyses of the lake's water currents, including through the development of a three-dimensional hydrodynamic computer model of the lake. My numerical lake is based on a map of Lake Como's bed (bathymetry: figure 1.2) and made up of hundreds of thousands of three-dimensional grid cells. The cells contain 'digital water' whose momentum, temperature, and salinity are calculated at regular time intervals with a series of equations solved numerically in the model.

That afternoon I am preparing to run another simulation of lake hydrodynamics on one of the centre's modelling computers (locked in a room downstairs) that 1 access from my desktop through the intranet. I have defined the grid size and the time step at which the model will run on the basis of elements such as the resolution of input data, the precision needed for my analyses (depending on the process of interest and its scale), the computational efficiency of the model, and feedback from previous simulations. For tonight's run, it will compute values for square cells with sides 200 m long, for a time period of three weeks with a calculation for every minute of (virtual) time, done by the model in a couple of (actual) seconds, resulting in a total simulation time of about 15 hours. If I start the run before I leave the office that night, 1 should get the results the next morning: three weeks of model-produced data of water temperature and velocity in my numerical lake, in three dimensions.

I am now checking the integrity of each of the datasets 1 will use (such as wind speed and direction, solar radiations, air humidity, and temperature). 1 do so by writing (coding) short scripts with the software MATLAB (5) and running the data files through these to check for gaps and inconsistencies. After addressing these, I plot the time series again to make sure the data 'look good' and 1 then export the clean data into new files, in a standard format to be read by the model. Practically, these tasks correspond to focused sequences of, for example, the following actions: typing code lines--running the script, which crashes--correcting the script--running--plotting--inspecting the data--deciding on an appropriate way to clean or filter the data--coding--running--plotting, and so on. This goes on until 1 feel that 1 have seen enough of the data, that they make sense, and that they are in the right format for the model.

Some of the scripts I use were adapted from those of more senior users who had shared them on the research centre's intranet, Internet forums, or informal e-mail conversations, often following trips to the coffee machine, a place where many ideas are shared. By going through these codes, running them, and trying to understand the flow of ideas and the logic behind them, I have learnt to write my own. After months of trial and error and navigation of the data in MATLAB I can barely imagine handling the large matrices of numbers constituting my data in any other way. As I go on preparing the simulation, I place the various data files in folders appropriately named and located for the modelling software (called ELCOM (6)) to find and process. I use a consistent folder structure to keep things organised.

Drifting nets and running code

When everything is sorted and the boat loaded, Alessandro skilfully manoeuvres it out of the small harbour. He likes to get on the lake after the other fishers of the area: this allows him to adjust his route to the location of nets that have already been set. Two drift nets carried by different currents can tangle and result in damage. In the event of an incident between pendenti and oltane belonging to different fishermen, the tacit responsibility lies with the owner of the pendenti; these are the nets that usually travel faster (as they are set at shallower depths) and become most damaged (as they get caught in the oltane's floats lines).

After setting oltane and smaller nets to catch perch, Alessandro starts sailing southward to set his pendenti, which will stretch between a depth of roughly 2 m and 10 m. A couple of kilometres away, he stops to check the velocity of the current, which seems particularly strong. He crafts an improvised drifter with an old jacket to catch the current, tied to a weight to keep it at the desired depth and to a buoy to follow the displacement at the surface, and throws it in the water (this is the scene that opened this paper). The buoy is drifting north rather fast, so after picking it up Alessandro takes the boat further south. He lets go of his nets progressively. As the nets run through his hand and into the water (figure 1.3), he subtly navigates the boat to compensate for the effect of the superficial current so the floats end up on an imaginary line across the basin (west-east), hundreds of meters long. All fishermen set their drift nets more or less perpendicular to the shores at the beginning of the night, across from the main north-south currents that keep them stretched (figure 1.4).

To differentiate between the nets, the fishermen attach lights of different colours and flashing frequencies to their extremities. There is a light indicating a set of nets further to the south that Alessandro judges of little concern because it is behind a peninsula that tends to block the current. There is also the noise of an engine several kilometres to the north; he knows which fisherman it is, and that he is likely to be setting oltane that will probably drift southward. Alessandro removes a float near one end of his series of nets to make that part of it hang in deeper waters, where he knows the current is slower. This is to slow down the northward progression of his series of nets and to make it rotate, so it can work for longer in the water without risking an encounter with the other fisherman's oltane. He is now thinking he may have to retrieve his nets earlier than usual anyway, around midnight, to avoid an incident.

Back on the lakeshore, Alessandro watches the movements of the nets for a while: the white light of his pendenti, the blue light of the other fisherman's oltane. When he is reassured that nothing abnormal is happening, he goes home. He will keep an eye on the lights through the night, checking their position several times from his house or the harbour, between short periods of rest. As he told me on another occasion: "It's better for me to sleep two hours in the afternoon than to sleep at night.... I just can't throw them [the drift nets] in and forget about them."

When everything is sorted for the numerical simulation, back in my Perth office I log on remotely to the modelling computer, mechanically typing my password and the usual Linux commands. The late-afternoon light filters through the eucalyptus trees by my office window. Before starting my simulation, I check whether there are other people's runs on the modelling computer: the centre's staff and students share access to a few of these machines, and too many simulations at once can lead one of them to crash. 1 type the commands to get the simulation started: "nohup ... bin/elcom.exe > log.txt &." Lines of the log appear every second or so in the command shell on my monitor, proof that my simulation is running (figure 1.5).

I watch these run down my screen for a few seconds. When I am reassured that nothing abnormal is happening, I go home. I have another look or two from there, connecting remotely from my home computer: the simulation is still running. I go to sleep hoping that it does not crash during the night and looking forward to pulling out the results the following morning.

The catch

At 11.30 pm by the small harbour the moon is big and the air is cold. Lights from the villages and villas on the lakeshore reflect on the lake's surface. The blue flashing (figure 1.6) attached to the other fisherman's nets appears to be drifting faster and faster towards Alessandro's nets. Alessandro arrives wearing a waterproof suit, and hops on the boat to retrieve his nets. The position of the two lights and the visible floats suggests that his series of pendenti has indeed rotated clockwise, which has prevented entanglement between the two sets of nets. This is not a technique Alessandro learnt from his father: back then, with handmade nets and rowing boats, they would not risk setting nets when the currents were that strong. He came up with this technique with experience, as he observed and experimented with the effect of currents on the motions of his nets.

Hoping for a good catch, he starts pulling the nets out at a steady pace, skilfully removing the fish from the net with bare hands as he goes along. Some fishers use gloves when it gets cold, but bare hands are more precise and prevent breaking too much of the fine netting when disentangling the fish. The live shad are silver with green and purple shades as the scales on their bellies refract the light. They emit a short gurgling noise as Alessandro takes them out of the net with dexterity and throws them in a plastic box. In the morning he will clean and prepare the fish with his wife's help, and in the afternoon he will be out on the lake again if the vento does not pick up too much. His young son might join him on the boat.

At about 9 am the following morning I am back in my Perth office. I copy the output files to my computer, open MATLAB, and start coding to process and visualise the model results. I am hoping for interesting insights into the variability of the main north-south currents. In particular, I have defined a series of 'curtains' through the lake that are roughly perpendicular to the shores (west-east): the model should have output values of water temperature, velocity, and direction in the cells of these curtains at every calculation iteration.

I have also programmed the model to extract calculated values of water temperature in my numerical lake at the same locations and depths where, in the 'real' lake, sensors are installed that have been measuring water temperature. Comparing these two sets of temperature data (the measured and the calculated) helps me 'validate' the model. I look for sense in the patterns on the graphs: what matches the field data and what does not, what I may need to correct for the next simulation, what is worth exploring some more, and possibly what is worth publishing. I try a few different graphs to better highlight the results that I feel are meaningful. A more senior researcher directs my attention to features that catch his eye, and suggests ways to improve the fit of the model. We decide that I should change the grid size in one area of my lake and adjust the wind drag coefficient, a parameter that influences the calculated transfer of momentum from air to water. This idea leads me back to the technical literature, the ELCOM science manual, and the source code itself. Between runs, readings, and adjustments my modelling practice will go on until my graphs and tables can support an article about the lake's hydrodynamics (Laborde et al, 2010), where I describe the 'new knowledge' that emerged from my work, making the fluid and messy circumstances of this emergence look crystalline and polished to the reader--another skill in itself.

What kind of environmental knowledge?

Extracting elements from the ethnographies, I first show that knowledge about the lake emerged in very similar ways in both cases, albeit in very different contexts. This first part of the analysis draws on recent developments in cognitive science (eg, Mercier and Heintz, 2013) and the philosophy and anthropology of knowledge (eg, Ingold, 2000; November et al, 2010; Turnbull, 2000).

Through water or pixels: navigating the task at hand

The narratives suggest that Alessandro's lake and mine had little in common, despite being connected through what we each called 'Lake Como'. In the next section I address how these two environments "hang together", to use Annemarie Mol's expression (2002, page 55). What I would like to start with here is an analysis of the development and use of knowledge in the two contexts described in the ethnographies.

The ways Alessandro organises his gear and adjusts the setting of his nets to the dynamic conditions in his environment are embedded in his skilled fishing practice [see Lauer and Aswani (2009) and Palsson (1994) for similar observations in different fishing contexts]. Meanwhile in Australia I was focused on my computer screen coding MATLAB scripts to navigate the data, guided by a familiarity with my digital workspace and the program's toolboxes. I am arguing that the knowledge embedded in these tasks of fishing and scientific research is of the same kind. Typing one's password is like turning the key to start the boat, coding a script like negotiating a fishing route. It is the dynamic interaction between person (fisher or limnologist) and environment (Italian lake or Australian workspace) that affords the conditions for the emergence of knowledge, within the space of the task at hand. At the same time, in both cases knowledge guides adjustments of a course of action to the information sensed in these interactions; it is at once enacted and transformed, in a process of navigation (Ingold, 2010; November et al, 2010).

The body is of course central to this process of navigation. For example, consider the way Alessandro's hands and my own worked through their daily tasks: every day his hands feel and assess the nets (passare le reti), push buttons and manoeuvre levers with precise timing on the boat, hold and twist slippery fish with just enough pressure to get them quickly out of the nylon meshes. Later they chop, scale, and fillet the fish, quickly and precisely. Mine cannot do any of that anywhere near as efficiently, of course--I tried. Instead, every day my hands typed Linux commands and MATLAB codes, drew sketches in a notebook to navigate my Lake Como that stretched from my desk to the online library system and the modelling machines with their familiar names from popular culture ('Vader'). These are cognitive processes, but quietly carried out, and deeply embedded in our interactions with our respective environments (Wilson, 1998).

The need to reassociate body, mind, and environment in theories of knowledge has been recognised by a broad range of scholars interested in processes of human cognition (eg, Clark, 1997; Marchand, 2010; Varela et al, 1991; Wilson, 1998). Perhaps one of the most influential attempts to grapple with the embodied dispositions that tacitly infuse practice by underlining and informing patterns of acting, thinking, and feeling is Pierre Bourdieu's concept of habitus (1972). Embedded in the habitus is a "sense of the game" (1997, page 23) that guides in our case the fisherman's and limnologist's actions without the need, most of the time, to consciously aggregate their perceptions and to reason through appropriate responses.

This is not to say that all knowledge is enacted unconsciously, as some critics of Bourdieu emphasised (eg Farnell, 2000). Conscious reasoning and skilful hunches are intertwined threads of knowledge production. Reasoning is a competence that is contingent on a social environment and field of practice, which is of course as true for scientists as for anybody else (Mercier and Heintz, 2013). When reasons and explicit understandings are produced, they do not emerge from a vacuum but from a history of engagement, and they will be brought to life again in practice, whether that practice is catching fish or drafting a scientific article. For instance, the arrival of new fishing gear like an echo-sounder might at first lead the fishermen to a reflection on their fishing habits and strategy, but soon enough it will become "just as much a 'natural sign', directly sensed, as birds in the air or natural landmarks" (Palsson, 1994, page 918). The same can be said for concepts of physical limnology such as the thermocline, or tools of computational fluid dynamics like a specific data format or a new computer: one might spend time thinking conceptually or mathematically about them, but soon they will just be a familiar entity to be seen on graphs, fed to the model, or discussed by the coffee machine.

Conscious intentions and ideas confer adaptive and creative qualities to practical engagement, even when skill allows attention to reach out in the field of practice instead of being reflectively focused on the thoughts themselves (Farnell, 2000; Ingold, 2011). This is evidenced in the narratives: consider the way Alessandro innovated to adapt to the presence of another net that he anticipated would drift towards his own, and the hunch-driven process of trial and error that characterised my computer coding practice. Creative improvisation arose out of skilled practice in these cases, shaped by prior episodes of engagement with an environment and combined with the commitment and intentionality that are part of the practice (catching fish and preserving nets on the one hand, producing publishable results and getting a PhD on the other). This line of thought, developed in a collection of essays edited by Elizabeth Hallam and Tim Ingold (2007) also elucidates how tradition and creativity are far from mutually exclusive but rather two sides of the same coin (see also Ingold and Kurttila, 2001; Knudsen, 2008).

Figli d'arte

The environments where practice takes place, or 'taskscapes' as Ingold once called them (2000), are of course also social: other people in the community of practice are central to the emergence of knowledge (Lave and Wenger, 1991). One of the fishermen described himself and his peers as figli d'arte, literally 'children of the art'. This expression highlights the importance of growing up by the lake, exposed to day-to-day fishing practice (arte from the Latin ars: 'practical skill') and under the influence of older fishermen through months or years of apprenticeship. Listening to fishermen's stories and seeing fishing places and techniques while helping by rowing or holding nets supported the coupled assimilation and development of practical fishing skill and of an unspoken 'code of conduct' of the lake fishermen (Pensa, 1981). Ingold, after James Gibson (1979), described this process as an education of attention, a central concept in his theory of relational knowledge acquisition (Ingold, 2000, pages 132-151). (7)

As illustrated in the narrative, my own development of scientific knowledge in graduate school largely occurred by following the conceptual and experimental paths of more experienced faculty and fellow graduate students, and by integrating aspects of these into my own practice. I was immersed at the research centre in a social environment of increasingly 'like-practised' people as their ways of approaching, conducting, and writing up research permeated my own. Decisions about what 'looks good', 'makes sense', and is 'the right format' were largely automatic and based on my experience with the modelling environment and the tacit expectations of my community of practice, shaping what could be described as a form of context-specific common sense. The tightness of this research network manifested itself through the production of recognisably similar articles from within the research centre (see also Knorr-Cetina, 1999).

Similar ways of knowing different places

Echoing Casey's call for knowledge to be "reconstrued as specifically placial, as a matter of acquaintance with places, knowing them by means of our knowing bodies" (1996, page 45), I hope to have shown in this section that Alessandro and I did not have different ways of knowing one lake (Lake Como); instead, we developed knowledge in similar ways, but in and about very different places. For each of us, knowledge about the lake-as-waterscape and the lake-in-modelling-lab emerged from a progressive and fluid education of attention, contingent on an engagement with the environment through a process of navigation, at times guided by more skilled others.

However, if our respective knowledges were about such different lakes, we were still able to talk about one 'Lake Como', a kind of boundary object (Star and Griesemer, 1989). How was this entity constructed across the different places of our day-to-day practices? To explore the connections between the multiple facets of 'Lake Como' as an object of knowledge and the articulation of the 'here' and 'there' in environmental research, I propose that it is useful to think of the environmental modelling lab as a heterotopia, in the sense proposed by Foucault (1966a; 2004). The next section applies this framework to the ethnographic case studies and develops its conceptual and methodological interests.

Heterotopias of environmental modelling

A key characteristic of Foucault's heterotopias is that they articulate imperceptible transitions between different places and times. Heterotopias are 'absolutely real' (my desk, the computer, the keyboard, the ELCOM code) and 'absolutely unreal' (the digital lake, its shoreline, the water filling its tenths of thousands of rectangular prisms), as well as "absolutely different from all the places that they reflect and speak about" (that is, Lake Como) (Foucault and Miskowiec, 1986, page 24). (8) A heterotopic numerical (spatial) model acts like a mirror reflecting another place, while also highlighting the differences between the reflected and the reflection, the 'here' and the 'there' (Hetherington, 2011). Foucault associated other properties with heterotopias. First, heterotopias break with the traditional notion of time, for example by engaging in forms of temporal accumulation (such as a museum or library). They are connected to but also somehow isolated from the places around them; kinds of sanctuaries whose access is reserved for those who know the code of entry or conduct. Heterotopias interrogate the places they talk about, either by exposing their illusory quality or by creating a space "as perfect, as meticulous, as well arranged as ours is messy, ill constructed, and jumbled" (Foucault and Miskowiec, 1986, page 27).

These propositions about heterotopic spaces can help to highlight characteristics and functions of the spaces of environmental modelling. Numerical models facilitate contextual shifts between on the one hand the data to be processed here and now, and on the other scientific formal knowledge embedded in the code and derived from many places and times. This also reflects the will to enclose that characterises some heterotopic spaces such as the museum or library (Foucault, 1966a; 2004). The ELCOM code was produced over thirty years ago as an evolved version of an earlier code, with inputs from different researchers that included theory, computational techniques, and results from empirical studies. In other words, my lake contained memories of Isaac Newton, Claude-Louis Navier, George Stokes, Joseph-Louis Lagrange, and Joseph Boussinesq, and countless researchers since them who have contributed to the field of computational fluid dynamics. It also contained traces of lakes, laboratory tanks, and seas from the USA, Australia, Kenya, Argentina, Switzerland, and other places where somebody had once run an experiment or made an observation that had been formalised into fluid mechanics theory. My model was supposed to speak about Lake Como, and it arguably contained more of Lake Como than any of these other places because of the empirical data I fed it, measured in Lake Como. These data recorded from the sensors installed in the lake were the most obvious connection between my heterotopic lake and the place it talked about, the bridge on which shifts in space and time could operate. The combination of sensors, data logger, satellite, computer, and its variants and articulations operated as a form of 'blind's man's cane' in the sense of Gregory Bateson's allegory (1972), allowing me to get sensory information about the lake, then transform it, integrate it, and let it become part of my heterotopic taskscape.

Foucault's heterotopias are somewhat protected spaces, not easily accessible from the outside. This is also true for spaces where environmental modelling is carried out, which generally require some sort of affiliation for access. As computational capabilities expand, numerical environmental models also tend to be increasingly complex and, by the same token, more and more opaque to outsiders. Accessing them requires an authorisation, some prior theoretical background, and a great deal of time to become familiar with the codes and the ways to run them. This relative separation from the outside world plays a part in conferring to these places the qualities of a refuge and a place for play (through simulation). In these spaces the modeller can run scenarii about environmental systems, with safeguards such as the 'clear all; close all; clc' command that resets the MATLAB workspace to a fresh start.

Benjamin Genocchio (1995) warned about the temptation to ask whether a kind of place can be considered heterotopic simply because all places have some internal heterogeneity. This is true of the lake as waterscape too: it encompasses various times and places on a limnological level, with waters of different ages containing traces of the mountains around, the winds above, and fish species imported from other places; and it encompasses times and places on a sociocultural level, with, for example, artefacts and bodies from the pre-Roman times, still lying in the darkness of the lakebed, along with their multispatial and millennial stories of commerce, contraband, politics, and other dimensions of human life. It would not be productive to locate the heterotopic space of environmental modelling in resistance or opposition to all other social spaces (Saldanha, 2008), or even, in this case, to the lake as waterscape. Rather, the question is: how useful is it to think of the lake-in-modelling-lab as a heterotopia? Below I propose a number of conceptual and methodological gains from the use of a heterotopic lens to look at places of environmental modelling.

Conceptual implications

In heterotopias the real and the virtual cohabit, and so do the present and the absent, the tangible physical grounds of the representation and its distortions, biases, and imperfections that distinguish it from a utopia. This is important because it contradicts the common idea that numerical models are simply abstract representations of natural systems and that modellers live in virtual worlds, only hanging on to reality by the thin thread of data. This is not the case: although digital, my lake and those of my fellow limnologists supported engaged practices and were real places, in the sense proposed by John Agnew (2011). Let me illustrate this point with a very short story: one of my fellow PhD students in environmental engineering had spent over four years studying and writing a thesis about Lake Kinneret, also known as the Sea of Galilee in Israel. After finishing her PhD, she went to Israel to see the lake. When she got there, her Facebook profile read: "Jenny 'sees her lake for the first time'." Her lake, for the first time. This short sentence illustrates how heterotopic and heterochronic her lake was. Yet, she had developed a strong sense of attachment to the place. Indeed, like me, she had spent a lot of time at her lake, living moments of bored routine, shared excitement, and computational frustration: the practice of environmental science. Our numerical lakes too were the "knots of stories" (Ingold, 2011, page 234), connected and located within vernacular geographies (Knuutila, 2005). Drawing on an expression first introduced by (Stewart) Brand in relation to buildings, I argue that the idea of a model is crystalline, while its day-to-day reality is fluid (Brand, 1995, page 2; see Ingold, 2011, page 314; Law and Mol, 2001).

Another conceptual implication of heterotopia in this case is that there is no essential Lake Como to be known through any one lens, but the ways it is conceptualised and bounded as a place are always arbitrary (Genocchio, 1995). This is an interesting ontological point, especially in the context of environmental fluid dynamics studies, as Newtonian mechanics embodies the search for universal descriptions of an absolute physical 'whole'. Yet, heterotopias of environmental modelling, whether they contain mechanistic models or not, are places of intense knowledge production precisely because of the shifts and overlaps they afford between different places and times (Topinka, 2010), even as those seemingly merge. This resonates with Law and Mol's attempt to 'situate technoscience', in which they propose a heterotopic conceptualisation of a technoscience object as "a stable pattern of conjoined alterity in which continuity depends upon discontinuity, or presence upon absence, the movement or displacement between here and there.... Here, then, and paradoxically, the global is already included in the local" (2001, page 618).

Methodological considerations

It follows that heterotopia may be useful as a methodological tool in science and technology studies to somewhat frame and 'locate' multisited research through its articulation of several places in one (Candea, 2007). Knowledge production is inherently spatial, and heterotopia as a framework exposes the complexities of this spatiality, providing a potentially useful approach for the study of multilayered spaces through a focus on what makes them 'hang together'. The framework of heterotopia is also open: approaches similar to juxtaposing fishermen's stories of the way a lake feels with the very visual stories of hydrodynamics told through computer graphics are productive, especially if these stories are judged problematic, because they put a spotlight on the assumptions about place and knowledge that make these unexpected combinations an issue [in the manner of Jorge Luis Borges's Encyclopaedia, cited in Foucault (1966b)].

When the usual grounds of knowledge are unsettled, there is space in the interstices and overlaps for insight and critique. After my time with the fishermen, another layer of complexity had come to merge with my heterotopic Lake Como. I realised as I spent time by the lake that the fishermen's way of knowing the lake engaged the body and the senses in very different ways to my prior digital navigations, which emphasised visual and vertical dimensions. This is illustrated in the opening paragraph by my failed attempt at reading the lake from a horizontal, water-level perspective (see also Ingold, 2011, pages 233-236). As I learnt about the fishermen's sophisticated understandings of the lake's currents, I also realised that I had had access, by probing into the lake's depth with my technological 'blind man's cane', to fragmentary information about the lake that was inaccessible to them. The exchanges about our respective ways of seeing and grasping the lake and its flows led to moments of intense insight and reflexivity, a process that I address elsewhere (Laborde, 2013). This leads to a corollary that is useful for interdisciplinary research and other work across communities of practice: a connection between knowledge groups should be found by focusing on the environments and stories of knowledge in the making, instead of on the products of this knowledge.

Finally, as places of intense knowledge production, the heterotopias of science and technology should be studied because they are also places where narratives are shaped and standards set. This is sometimes problematic when, for example, simulation model results are taken to be representations of reality without openly addressing the connections--and their messiness--between a heterotopic space of environmental modelling and the environmental system it talks about (Barnes, 2014).

Conclusion

To conclude, let us return to where this paper started: the boat by the rocky outcrop, the drifting buoy, Alessandro, and myself. It should now be clear to the reader why in this instant 1 felt like a fish out of water: this lake I was looking at had very little to do with the numerical lake I had lived with for years. I have argued here, on the basis of ethnographic data and recent developments in the scholarship of cognition, that a local practitioner and a remote scientist do not have fundamentally different ways of knowing the same place, as is often assumed; rather, they have the same kind of knowledge of different places. This premise frames knowledge as being integrally contingent on an individual's navigation of his or her physical and social environment, and therefore it reconnects knowing with being, to start answering the call from Ingold that opened this paper.

I then showed how environmental modelling labs could be productively modelled as heterotopias, to emphasise the inner workings of their complex and multilayered spatialities. Heterotopia as a methodological tool helped highlight mechanisms of intense knowledge production in these taskscapes of science and technology, operating through the juxtapositions and manipulations of physical and digital places and times. It also provided a framework to integrate multisited research 'in place'. Finally, this approach encourages a conceptual shift away from cognitive and epistemological differences between communities of practice and instead towards differences between the physical and social environments that support these practices (here, the lake as waterscape and the lake as numerical model). In doing so, it reconnects epistemology with ontology, by reconnecting the knowledges (here, the fishing knowledge and the scientific knowledge) with the worlds within which they emerge. I disagree with Ingold when he suggests that a "sense of astonishment ... is so conspicuous by its absence in contemporary scientific work" (2011, page 129). It is admittedly absent from its written products because of the tacit rules of scientific publishing, but the interstices and overlaps in the heterotopias of 'science in the making' are full of astonishment. My heterotopic Lake Como provided an inspiring way to think about and look at a lake, the sky, and clouds not as bounded entities but as fluid space that stratifies, mixes, and flows along particular patterns relevant across fishing and fluid dynamics--reconnecting, finally, thought with life.

doi: 10.1068/d14128p

Sarah Laborde

Department of Anthropology, The Ohio State University, Smith Laboratory, 174 W 18th Avenue, Columbus, OH 43210, USA; e-mail: laborde.7@osu.edu

Received 5 June 2014; in revised form 19 October 2014; published online 17 March 2015

Acknowledgements. The research for this paper was financially supported by a postgraduate scholarship from the Australian Government and a postgraduate award from the University of Western Australia (UWA). The School of Social and Cultural Studies of UWA partly funded the fieldwork in Italy, while the limnology studies were conducted at the Centre for Water Research. In Australia I wish to especially thank Sandy Toussaint; in Italy, Massimo Pirovano and the fishermen of the Lario, especially Alessandro Sala and his family. This paper went through several versions--I am grateful to the editors of Environment and Planning D: Society and Space and two anonymous reviewers, as well as Nick Harney, Jessie Barnes, Marijn Nieuwenhuis, and Mark Moritz for helpful feedback. I also thank Tom Boellstorff, the then editor-in-chief of American Anthropologist, and anonymous reviewers for their thorough reviews of an early version in 2011. Finally, I thank the participants of the UQ-UWA Anthropology 2012 workshop, as well as Patrick Gallagher and the participants in the Mundane Mapping and Grounded Truths panel at the 2014 meeting of the American Association of Geographers, for critically engaging with the ideas underpinning this paper.

References

Agnew J, 2011, "Space and place", in The SAGE Handbook of Geographical Knowledge Eds J Agnew, D N Livingstone (Sage, London) pp 316-330

Bateson G, 1972 Steps to an Ecology of Mind: Collected Essays in Anthropology, Psychiatry, Evolution, and Epistemology (Chandler, San Francisco, CA)

Barnes J, 2014, "The politics of prediction: modelling the Nile's water futures", paper presented at the 2014 AAG annual meeting, Tampa, FL, 8-12 April

Bourdieu P, 1972 Esquisse d'une theorie de la pratique (Droz, Geneve)

Bourdieu P, 1997 Meditations Pascaliennes (Seuil, Paris)

Brand S, 1995 How Buildings Learn: What Happens After They 're Built (Viking, New York)

Candea M, 2007, "Arbitrary locations: in defence of the bounded-site" Journal of the Royal Anthropological Institute 13 167-184

Casey E S, 1996, "How to get from space to place in a fairly short stretch of time: phenomenological prolegomena" Senses of Place Eds S Feld, K H Basso (School of American Research Press, Santa Fe, NM) pp 13-52

Clark A, 1997 Being There: Putting Brain, Body, and World Together Again (MIT Press, Cambridge, MA)

Farnell B, 2000, "Getting out of the habitus: an alternative model of dynamically embodied social action" The Journal of the Royal Anthropological Institute 6 397-418

Foucault M, 1966a, "Les Heterotopies" France Culture 1 December, Institut National de l'Audiovisual, Paris, http://www.discogs.com/Michel-Foucault-Utopies-Et-Heterotpies/release/1974541

Foucault M, 1966b Les mots et les choses: une archeologie des sciences humaines (Gallimard, Paris)

Foucault M, Miskowiec J, 1986, "Of other spaces" Diacritics 16 22-27

Foucault M, 2004, "Des espaces autres" Empan 2(54) 12-19

Genocchio B, 1995, "Discourse, discontinuity, difference: the question of other spaces", in Postmodern Cities and Spaces Eds S Watson, K Gibson (Blackwell, London) pp 35-46

Gibson J J, 1979, "The Ecological Approach to Visual Perception" (Houghton Mifflin, Boston, MA) Hallam E, Ingold T, 2007 Creativity and Cultural Improvisation (Berg, Oxford)

Hetherington K, 2011, "Foucault, the museum and the diagram" The Sociological Review 59 457-475

Hodges B, Dallimore C, 2014 Estuary, Lake and Coastal Ocean Model: ELCOM v2.2 Science Manual (Centre for Water Research, University of Western Australia)

Ingold T, 2000 The Perception of the Environment: Essays on Livelihood, Dwelling and Skill (Routledge, London)

Ingold T, 2009, "Point, line and counterpoint: from environment to fluid space" Neurobiology of "Umwelt": How Living Beings Perceive The World Eds A Berthoz, Y Christen (Springer, Berlin) pp 141-155

Ingold T, 2010, "Footprints through the weather-world: walking, breathing, knowing" Journal of the Royal Anthropological Institute 16 S121-S139

Ingold T, 2011 Being Alive (Routledge, London)

Ingold T, Kurttila T, 2000, "Perceiving the environment in Finnish Lapland" Body and Society 6(3-4) 183-196

Johnson P, 2006, "Unravelling Foucault's 'different spaces'" History of the Human Sciences 19(4) 75-90

Knorr-Cetina K, 1999 Epistemic Cultures: How the Sciences Make Knowledge (Harvard University Press, Cambridge, MA)

Knudsen S, 2008, "Ethical know-how and traditional ecological knowledge in small scale fisheries on the eastern Black Sea coast of Turkey" Human Ecology 36 29-41

Knuutila T, 2005 Models as Epistemic Artefacts: Towards a Non-representationalist Account of Scientific Representation PhD thesis, Department of Philosophy, University of Helsinki

Laborde S, 2013, "A research practice like Escher's drawing hands: reflections on crossing disciplinary boundaries on Lake Como, Italy" Collaborative Anthropologies 6 290-306

Laborde S, Antenucci J P, Copetti D, Imberger J, 2010, "Inflow intrusions at multiple scales in a large temperate lake" Limnology and Oceanography 55(3) 1301-1312

Latour B, Woolgar S, 1986 Laboratory Life: The Social Construction of Scientific Facts (Princeton University Press, Princeton, NJ)

Lauer M, Aswani S, 2009, "Indigenous ecological knowledge as situated practices: understanding fishers' knowledge in the Western Solomon Islands" American Anthropologist 111 317-329

Lave J, Wenger E, 1991 Situated Learning: Legitimate Peripheral Participation (Cambridge University Press, Cambridge)

Law J, Mol A, 2001, "Situating technoscience: an inquiry into spatialities" Environment and Planning D: Society and Space 19 609-621

Marchand T H J, 2010, "Making knowledge: explorations of the indissoluble relation between minds, bodies, and environment" Journal of the Royal Anthropological Institute 16 1-21

Mercier H, Heintz C, 2013, "Scientists' argumentative reasoning" Topoi 1-12 doi: 10.1007/S11245-013-9217-4

Mol A, 2002 The Body Multiple: Ontology in Medical Practice (Duke University Press, Durham, NC)

November V, Camacho-Hubner E, Latour B, 2010, "Entering a risky territory: space in the age of digital navigation" Environment and Planning D: Society and Space 28 581-599

Oreskes N, 2003, "The role of quantitative models in science" Models in Ecosystem Science Eds C D Canham, J J Cole, W K. Lauenroth (Princeton University Press, Princeton, NJ) pp 13-31

Palsson G, 1994, "Enskilment at sea" Man 29 901-927

Pensa P, 1981 Noi gente del Lario (Pietro Cairoli, Como, Italy)

Saldanha, A, 2008, "Heterotopia and structuralism" Environment and Planning A 40 2080-2096

Shelley P B, 1882 Select Letters of Percy Bysshe Shelley Ed. R Garnett (Kegan Paul, Trench and Trubner, London)

Star S L, Griesemer J R, 1989. "Institutional ecology, 'translations' and boundary objects: amateurs and professionals in Berkeley's Museum of Vertebrate Zoology, 1907-39" Social Studies of Science 19 387-420

Topinka R, 2010, "Foucault, Borges, heterotopia: producing knowledge in other spaces" Foucault Studies 9 54-70

Turnbull D, 2000 Masons, Tricksters and Cartographers: Comparative Studies in the Sociology of Scientific and Indigenous Knowledge (Routledge, London)

Varela F, Thompson E, Rosch E, 1991 The Embodied Mind (MIT Press, Cambridge, MA)

von Uexkull J, 1992, "A stroll through the worlds of animals and men: a picture book of invisible worlds" Semiotica 89 319-391

Wainwright J, Mulligan M, 2004 Environmental Modelling: Finding Simplicity in Complexity (John Wiley, Chichester, Sussex)

Wilson F R, 1998 The Hand: How its Use Shapes the Brain, Language and Human Culture (Pantheon, New York)

Yearley S, 1999, "Computer models and the public's understanding of science: a case-study analysis" Social Studies of Science 29 845-866

(1) In 2012 I was awarded a PhD in environmental engineering and anthropology. The ethnographic research that led to this paper was conducted between 2008 and 2011.

(2) See Ingold (2009) for a later and more integrating discussion of "organisms in their environments' in relation to earlier theories, especially von Uexkull's (1992) concept of the Umwelt.

(3) For a discussion of the concept of heterotopia and its various uses, see http://www.heterotopiastudies.com

(4) In 2010 sixty-nine people were holders of a professional licence granted by the Department of Fisheries of the Como and Lecco provinces, permitting professional fishing in Lake Como. Many of them were, however, retired, seasonal, operating elsewhere, or family members of fishermen helping with the activity on the land.

(5) The Math Works, Inc., Natick, Massachusetts, United States.

(6) Estuary lake and Coastal Ocean Model (Hodges and Dallimore, 2014).

(7) See also Bourdieu's (1972) description of learning as a development of habitus through socialisation (for example, parenting) and Jean Lave and Etienne Wenger's (1991) descriptions of relational (as opposed to genealogical) learning in communities of practice.

(8) For the original text in French, see Foucault (2004).
COPYRIGHT 2015 Sage Publications, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Laborde, Sarah
Publication:Environment and Planning D: Society and Space
Geographic Code:8AUST
Date:Apr 1, 2015
Words:9510
Previous Article:Wet ontologies, fluid spaces: giving depth to volume through oceanic thinking.
Next Article:Neoliberal Niagara? Examining the political history of fish consumption advisories in New York State.
Topics:

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |