Tracing (in)visibilising practices: engaging with simulations for architecture and spatial planning

ABSTRACT

Using ethnographic vignettes from the development of simulations for architectural design and spatial planning in university contexts, this article discusses how the material, embodied, and tacit dimensions of developing and doing research with simulations can be opened up to analysis. We ask how (in)visibility comes to matter in simulations and how it is made and unmade in different situations. In particular, we explore how specific enactments of the things that are simulated and the practices involved in producing and handling them are rendered (in)visible through the simulation process. We argue that focusing on the politics of (in)visibility in making simulations can render power relations underpinning architectural and spatial planning practices more apparent.

KEYWORDS:

• STS
• simulation
• digital materialities
• data visualization
• (in)visibility

1. Introduction

The expansion of digital design methods in architecture and planning—especially of computer simulation—has introduced new possibilities for learning about and experimenting with future built environments. Such methods allow architects and planners to compile aspects of space, such as physics, human movement, temperature, energy flows, or light, into one interface and model processes through which they interrelate (Fritz, Hirschberg, and Hovestadt 2020). Digital simulations are supposed to uncover some of the potentialities in both speculative and existing built environments by allowing the targeted manipulation of isolated factors and parameters. Consequently, it is important to keep in mind that they are not capable of fully capturing all the complexities and contingencies of spaces (Blok and Farias 2016; Guggenheim 2016). The promise of such integrated models is that complex processes, for example, interrelations between the materiality of the built environment, the behaviour of inhabitants and environmental conditions, could be tested and better understood in order to address societal challenges and crises. Thus, in funding proposals, conference presentations and public debates, the focus is often on the visualized ‘output’ of simulations. In contrast, this contribution attends to the making of architecture and spatial planning simulations in academic contexts, where different knowledges, data, disciplines, and actors are brought together, and asks what is (in)visibilised in this process.

Architects and theorists of architecture have debated how implementing ‘non-visual information’ such as numeric data into the design process expands and transforms architecture’s response-abilities and capabilities to imagine and care for ‘other worlds, especially those which have reached a precarious condition’ (Sanchez 2019, 206). The emergence of new computational skills, practices and a growing sense of responsibility for managing the complex dynamics of built environments has led to new understandings of what being an architect and doing architecture and spatial planning mean (Ahrens and Sprecher 2019; Bernstein 2004; Loukissas 2008). At the same time, simulating specific process that are of interest while drawing on diverse data, practices and knowledges requires decisions about what to show and what not to show (Igelsböck 2016). Focusing on the ‘output’, i.e. integrated visualizations, renders practices that go into their production invisible (Hill 2020). This particularly concerns practices such as adapting, curating, assessing and integrating data sets.

Considering that (in)visibility is relational and situated and that ‘it is impossible to define anything inherently as visible or invisible’ (Star and Strauss 1999, 24), we ethnographically attend to different moments when simulations and their ‘objects’ are conceived and visualized, analyze the traces of such visualization practices, and consider what they render (in)visible in the simulation. We argue that tracing some of the practices, relations and things that remain implicit affords the opportunity to better understand what relations visualization practices enact in the process of developing and using spatial planning simulations. We therefore attend to how (in)visibility comes to matter, and how it is made and unmade in different situations. Doing so allows us to reflect on the politics of (in)visibility, and on the power relations underpinning visualization choices made in architectural and spatial planning. As Suchman notes, what separates explicit and tacit practices ‘is not only a difference between what we can see, talk, or even think about, but also between what our social milieu sanctions as legitimate to be seen, spoken or thought’ (2016, 130). We are thus interested in the configurative power (Knuuttila and Merz 2009) through which simulations for spatial planning—as necessarily selective and simplifying models—unfold, reassemble, and obscure design practices, and in the relations produced by them.

In exploring these dynamics, we draw on ethnographic research with two different university sites in which researchers use and develop computer simulations that visualize how built environments participate in urban ecologies. One of these sites involves simulating sound propagation in urban environments, and the other simulates the flows of materials implicated in urban development. In our analysis, we consider different kinds of (in)visibility, such as what becomes imperceptible from different positionalities, what is excluded because it is deemed insignificant or useless for specific audiences or purposes, or considered not do-able with the given resources, and what might seem impossible to visualize with specific media. We contribute to a rapidly growing body of literature critically examining computing practices. This literature addresses the implications of materializing processes in computing, examining how computing is constituted within academia or the cultural contexts, values and politics underlying allegedly neutral practices in programming (Britton, Klumbyte, and Draude 2019; Malazita and Resetar 2019).

In what follows, we lay out the conceptual basis for our discussion of studying (in)visibility in simulation and describe our empirical approach. We then present and analyze two vignettes from fieldwork with architecture researchers. Each of these engages with (in)visibilising practices, illustrating what such practices may generate and what we can learn through studying them. In closing we discuss at what costs these (in)visibilities may come.

2. Approaching simulation in architecture and spatial planning

2.1. Making sense of simulation practices: current literature

Computer simulations as ‘a sequence of states undergone by a digital computer, with that sequence representing the sequence of states that some real or imagined system did, will or might undergo’ (Parker 2009, 488) have been a longstanding concern of STS scholarship. Such scholarship has emphasized that simulations are performative: they shape socio-material relations and practices by complicating boundaries between representation and intervention, and act as trading zones or infrastructures connecting diverse professional actors, things and knowledges (Cardoso Llach 2019; Galison 1997; Hacking 1983; Houdart 2008; Knuuttila and Merz 2009; Loukissas 2012). While performing the realities they model, simulations become entangled with the emergence of epistemic practices and communities.

Situated between model and experiment, simulation combines representation and intervention; a trait that has been extensively theorized in STS and the philosophy of science (Ammon and Capdevila Werning 2017; Cardoso Llach 2015; Galison 2014; Lenhard 2007; Sismondo 1999). STS scholarship in particular has emphasized the performativity of simulations; they not only (re)construct and approximate (Mommersteeg 2022) the ‘objects’ and worlds that they model, but also reconfigure the scientific communities and identities that form around them. Studies show how relations among actors, identities, bodies and senses can become reconfigured with the process of modelling, whether through different sensing practices in the making of oceanographic models, reconfiguring professional identities in architecture, or becoming entangled with embodied performances of protein folds (Helmreich 2009; Loukissas 2009; Myers 2009). Igelsböck (2016) uses the term integration machine to critically interrogate the interplay between the performative dimensions of simulations—how they distribute heterogeneous knowledges, responsibilities, societal relations and identities—with the ideas, imaginations and promises that inform simulation, as well as those that are kept out of the process. Loukissas (2009) alludes to the ‘new kind of intimacy [of using digital simulations in architecture] that makes explicit what is shared [with others] and what is not’ (169), indicating that understanding simulations as trading zones might overemphasize shared vision. Cardoso Llach (2019) indicates that simulations are prone to conflict and pockets of resistance, despite promising to allow some actors to set standards for data and practice and communicate across disciplinary boundaries.

At the same time, digital simulations do things to data, which in the widest sense can be regarded as records of values, observations or measurements representing things and phenomena (Borgman 2015). In spatial planning and architecture, simulations spatialize data, assigning them to locations and relating them to other properties in order to create specific visual effects such as textures, colours or charts. The field of data visualization studies examines charts, maps and graphs to research how they are made, shared and understood while attending to their worlding capacities—understanding that ‘(t)hey embody and engender not only particular ways of seeing (…) but also ways of knowing and ways of organizing collective life in our digital age’ (Gray et al. 2016, 228/229; Kennedy and Engebretsen 2020). This research shows that data visualizations often fail to reveal data structures and formats, the sensory and affective aspects of those things represented by the data, and processes of data creation and visualization (Hill 2020; Kennedy and Engebretsen 2020; Rettberg 2020).

Relatedly, STS has produced a rich body of literature on the creative potentials and invisibility of data practices, particularly those pertaining to care and maintenance (Edwards et al. 2011; Plantin 2018; Schwennesen 2019). This work highlights the human labour ‘behind’ computation, its material and environmental costs, and the ways in which these are often erased from view (Crawford 2021; Irani 2019). While data practices are ‘performative in the sense that they help to enact—that is make up—the very realities they ostensibly only describe’ (Ruppert and Scheel 2021, 37), data practices are also performative in the sense that they imbue wider knowledge regimes with their materialities and with their ‘recurring patterns, regularities, logics, strategies, self-evidence, and rationalities’ (Blanchette 2011; Camus and Vinck 2019; Ruppert and Scheel 2021, 36). For these and many more reasons, spaces and processes where data is produced and handled, including selecting, formatting, sharing, storing and visualizing are of great interest to STS scholarship.

While much of the scholarship in STS highlights the empowering effects of attending to marginalized practices, Star and Strauss (1999) caution us to attend to the complex and situated relations of (in)visibility, power and autonomy. Even though making certain practices (such as care practices) visible can enhance their appreciation, it is important to bear in mind that increased awareness might impinge on actors’ discretion and autonomy. Invisibility thus ‘protects’ practices that have become obscured by virtue of routine, or those involving tacit knowledge from scrutiny, reductionism and standardization. In this article we therefore use the notion of (in)visibility as an orientation to trace the politics of visualization practices and processes, and consider what they render invisible from the perspective inscribed (Akrich 1992) in the simulation. Specifically, we are concerned with ‘behind the scenes’ practices that might at most leave traces in simulations—despite, as we will argue, imbuing their creation and content.

With these debates in mind, and understanding that abstraction in both data and simulation renders things such as relations, practices, processes or identities (in)visible while being acutely aware of the ‘worlding’ that goes along with making data visualizations and simulations, we discuss how visibility and invisibility are made and unmade in spatial planning simulations. Following data visualization scholars’ invitation ‘to study the conditions under which such visual texts [visualisations of numeric data] are generated, disseminated and thought to benefit processes of sense-making, learning, and engaging’ (Kennedy and Engebretsen 2020, 19), our discussion contributes to literature that highlights the hybridity and co-becoming of researchers and their objects, and how materialities of research practices and relations shape simulated ‘objects’ in subtle and (in)visible ways (Loukissas 2009; Myers 2009).

2.2. Ethnographically engaging academic architecture researchers and simulation

The article draws on empirical material produced within the first author’s dissertation project, which explores simulation practices in spatial planning and architecture as developed and researched in academic settings. This involves attending virtual academic conferences, reading articles and reports authored by actors in the field, speaking to researchers working on simulation tools, examining simulations together with researchers, and mapping and drawing together the collected impressions as part of the analysis (Awan 2017; Clarke, Friese, and Washburn 2018). To access the multiplicity (Mol 2003) of different enactments of spatial planning and architecture in simulations, the dissertation project builds on digital, flow oriented ethnography in which the lead author produces vignettes (Markham and Gammelby 2018; Pink et al. 2016). To us, ‘[w]riting vignettes is a layered practice of re-interpretation’ (Bloom-Christen and Grunow 2022, 13) that is imbued with authors’ situated perspectives—thereby problematizing clear distinctions between ethnographic data and analysis.

In this article we discuss two ethnographic vignettes, drawing on visits to two different university architecture/planning departments, in which simulation researchers show their labs, talk through their simulations, and explain their everyday work. In both cases, the researchers’ work is publicly funded, and falls into the loose category of performance simulations (Fritz, Hirschberg, and Hovestadt 2020), which seek to test certain effects and relations within existing built environments. One site focuses on simulating sound propagation in built environments, and the other on modelling the (re)use of building materials in the city. The first vignette illustrates how the ‘object’ of the simulation, urban sound propagation, is multiplied through different visualization practices. The second shows how the multiplicity emerging from the different visualization practices and the ways they enact material flows related to urban development in the city, is reduced to create a coherent simulation. Ongoing informed consent ensures interlocutors’ anonymity. We therefore focus on key aspects and technicalities of their work without disclosing exact details about the applications they are developing.

Emphasizing that (in)visibility is relational and situated, the kinds of (in)visibility that take precedence in particular moments, contexts and for particular audiences and users becomes an important empirical question. Our analysis thus attends to (in)visibilising practices—practices that render particular things and aspects of simulated ‘objects’ and visualization processes visible or invisible in the simulation. We specifically focus on the performativity of processes of (in)visibilisation that are, as we argue, integral to simulation and achieving the necessary degree of simplification. We play on the mutual constitution of visibility and invisibility and discuss what emerges from and what imbues such processes. We ask: How are certain practices, things, and enactments made (in)visible in the simulation? And how does this matter?

To follow such concerns, we attend to moments where different ontological assumptions about a simulated ‘object’—emerging from different visualisations made in the development of simulations—confront each other. Such moments of ontological politics comprise instances where ‘the real, the conditions of possibility we live in’ (Mol 1999, 75) are actively shaped by mundane practices through which we interact with ‘the real’. Thus, reality does not precede the practices through which we interact with and make sense of it, but is shaped within such moments. As a result, when objects are enacted across contexts that entail different perspectives and practices, their ontologies are multiplied. Annemarie Mol (2003) describes this through the case of atherosclerosis which is diagnosed, hence enacted, in different ways across different medical fields.

Sensitized to this multiplicity (Mol 2003), we draw on the concept of epistemic dissonances (Farías 2015) which emphasizes the epistemic function of diverse, sometimes competing visualizations made in architectural practice. It expresses how alternative design options emerge from creating, juxtaposing and discussing the specific and sometimes contradictory ways that different visualizations such as models, maps, or renderings enact future buildings. Epistemic dissonances are moments when ‘[a]n actor holds expectations regarding what an object actually is, how it behaves or which effects it produces, which when suddenly disappointed are called into question, leading to a thorough revision of the knowledge’ (2015, 275). We attend to moments when epistemic dissonance is created and dealt with in the development of simulations, and how this relates to situated instances of (in)visibilisation.

In what follows, we use two ethnographic vignettes to unpack two moments when visualization practices come to matter in simulations, discussing how simulations, simulated ‘objects’, researchers and research relations ‘become with’ the situated (in)visibilities produced in data visualization practices.

3. Encountering the (in)visible: two vignettes

3.1. Making sound visible: multiplying acoustic materiality

During my visit to a university architecture studio, researchers showed me how they used an open-source programme to simulate and research urban noise propagation. The programme represented sound as colourful particles bouncing off a line (emulating a façade from an eagle’s eye perspective). As the particles move away from a dot that represents the sound-output and bounce off the line, their colours change, indicating the decay of sound pressure, which is commonly perceived as volume. As soon as one of the researchers clicks a grey button, the dots are suspended in mid-air—their colours stick: the closest are red, representing high pressure, the furthest represent low pressure and are blue.

Having learnt about ‘acoustic waves’ in school, I ask why the simulation displays sound as particles. To answer my question, the researchers go back to an ‘analogue’ experiment they did some time ago. They play a video of tiny, sand-like particles evenly diffusing on a metal plate with the voice of an opera singer accompanied by an orchestra playing in the background. Then, as the voice is foregrounded and the singer holds a note for a couple of seconds, the particles form a vivid grid-like shape on the plate. As the orchestra chimes in, the particles scatter again.

One of the researchers adds,

the frequency of the opera singer’s voice—and the anti-nodes of its interfering acoustic waves—make some areas on the plate vibrate. This creates a shape by moving the sand away from the vibrating spots to areas that remain still. However, even referring to sound as a wave is an abstraction that helps us better understand how sound moves and interacts with things, for instance how interfering frequencies create these patterns. Sound is neither literally ‘a wave’ nor moving particles, but a wave-like pattern of disturbances created by energy transmissions when they displace particles of mediating substances such as air or water. These transmissions can be detected by our ears and be interpreted as sound by our brain.

Apparently, the matter-moving properties of sound could be better understood with the wave-pattern in mind, but its propagation and the pressure that is lost through these transmissions is easier to grasp when it is visualized as particles with changing colours. ‘The sand-experiment really makes the ways various frequencies impact matter visible. It shows that sound can make our cells oscillate, that it physically impacts our bodies’, they further explained. By showing the researchers how frequencies move sand, the experiment made the wave properties of sound visible. This alerted the researchers to the significance of frequency for understanding sound’s impact on matter.

The sand-experiment had thus sensitized the researchers to the ability of frequencies to not only evoke emotional but also somatic affect, making the cells in our bodies oscillate. A WHO study indicating negative health effects of urban noise supported this view, leading the researchers to consider focusing their project on finding ways to mitigate only ‘harmful’ sounds in the city. However, personal encounters with sound in their homes and workspaces pointed to the situated, context-dependent and experiential dimensions of sound, raising questions around how to define and categorize ‘harmful’ frequencies. Some sounds are soothing and important to the identity of spaces—an aspect that they thought should be considered and not erased by sound simulations.

The different enactments of sound that the researchers encounter while developing the simulation—as rays produced by a ready-made simulation, as matter-moving waves that become visible by conducting an experiment—multiplies sound as an ‘object’ of research and the ways researchers relate to it. It is no longer a simple thing that can be straightforwardly simulated, but something that can be visualized in diverse ways, each visualization revealing different potentialities that carry their own distinct significance. For instance, the open-source acoustic simulation is intended for understanding sound propagation only through ‘ray-tracing’, tracing the distribution of sound and the loss of pressure by representing it as particles moving along geometric rays. On the flipside, this means somewhat neglecting the wave properties of sound, thus also how its frequency matters for more complex, open and diverse acoustic systems like urban areas.

However, enacting (and subsequently juxtaposing) sound as a series of linear energy transmissions and as waves that move matter elicits epistemic dissonance (Farías 2015). It reveals that sound is at once a health- and environmental issue and an aesthetic component of built environments that can be experienced in different ways, and that a simulation integrating the distinctive benefits of both of the different, somewhat contradictory wave and particle representations of sound is yet to be designed. The affective dimensions of the dissonances also stimulate the researchers to relate their findings to embodied experiences and personal aesthetic sensibilities—making the entanglement of the physical and affective qualities of sound evident.

But the project duration and funding conditions of the project make it difficult to develop methods and tools to visualize sound in a way that represents its multiplicity. For the time being, the aspects of sound that do become visible to the researchers during the process of developing the simulation—despite being invisible in the simulation itself—inspire the researchers’ future plans to make different properties of sound visible, possibly by combining different software and experiments.

This vignette discusses how researchers working with acoustic simulations for urban spaces enact the ‘object’ of their simulation, sound, through different methods, apparatuses, and visualization practices, and thereby multiply it. Similar to how Mol (2003) describes how different diagnostic practices enact atherosclerosis in multiple ways, the different visualization practices portrayed in the vignette produce multiple enactments of sound. We also see that new concerns and commitments arise from these enactments and how subsequently the researchers grapple with visualizing sound in a way that captures the aspects that are important to them.

Crucially, sound had become so much more than the physical qualities emerging from the simulation and the experiment. The materialities and affects emerging from the researchers’ enactments of sound inspire them to trace how its particle and wave properties might become enmeshed with built environments and their inhabitants. Negotiating which of these realities would be visualized in the simulation, and which ones would come to fruition in future research, constitute moments of ontological politics (Mol 1999). Such negotiations present moments when reality is shaped within mundane practices, for instance by being decisive for the apparatuses and methods the researchers would use to make sense of sound. Concretely, different ontological assumptions about sound—materialized by the experiment and the simulation—confront each other, create dissonances, and expose new facets for the researchers to explore, recombine or drop when going forward with their project.

The dissonances arising from these visualizations have stimulated the researchers to find associations between simulating sound, individual health concerns and embodied experiences. Citing their own lived experiences in their cities, homes, and offices, sound emerged as an aesthetic and political issue, one that they consider neglected in architectural form finding and their formal education. Visualization practices have, in this case, made the cracks in geometric simplifications visible, making the researchers interrogate prior assumptions about which properties of sound matter and should be simulated.

In this case, epistemic dissonance doesn’t actually indicate possibilities to ‘resolve’ emergent incongruencies between the simulation and the experiment. It rather opens up new technical and moral questions that eventually reconfigure the aims and methods of the project and, crucially, how researchers enact being an architect by developing new response-abilities. The desire to simulate sound in a meaningful way, and the emergent technical difficulties and trade-offs of doing so, generated care in the sense that it encouraged the researchers to challenge what they described as the primacy of the visual and the boundaries of architectural practice. It inspired them to craft their own niches for sound to exist within architectural design—interstices to address sound in a way that would relate different situated accounts of sound stemming from experiment, simulation, and experience. Further, the potentialities emerging from adopting sound as a matter of care (de la Bellacasa 2011) were ‘not limited to their critique of power, but also to creating a relationship through that critique’ (97).

The questions beckoned by the dissonances allow the researchers to redefine their approach to the political issues that they feel are at stake in architectural practice, such as caring for the environment. In this way, despite not being visualized in the simulation, experimenting with and reflecting on the multiple materialities of sound through visualization practices might redefine how the researchers relate to their environments, society, and enact being an architect through their work, as much as it shapes the current simulation and how it enacts sound. This episode further emphasizes how ‘research and researched are mutually constituted with each other, and within artistic research there is space for questioning, refusing, cruising and fighting against itself’ (Lorenz 2017, p. 31). In this subtle way, the visualization of something invisible, such as sound, comes to shape built environments (and how they sound) and researchers’ subjectivities.

3.2. Visualizing the materiality of buildings: invisibilising the multiplicity of building data

A project leader at a different university lab explains to me how they arrived at the current state of their simulation. The simulation is supposed to visualize how hypothetical urban development scenarios such as construction, renewal, augmentation, or demolition relate to flows of building materials in different areas of a city. This ‘tool’ will support policy makers and planners in tapping the full building potential of areas by visualizing the material impacts of building policies or economic and demographic shifts. Visualizing how such aspects relate is understood to be necessary for achieving ‘sustainable’ urban development: more efficient land and material use in the city, preventing urban sprawl and recycling building materials.

The first step to model relations between development scenarios and material flows was building an understanding of what buildings are made of (their ‘material stock’) and how this material stock is distributed throughout the city. This meant combing through hundreds of construction files to trace buildings’ genealogies, their building periods, and whether they are under historic preservation or have recently been renovated. These hand-made data on building materials from the construction files were then ascribed to building periods and related to open government data (OGD) showing the distribution of building periods across areas, allowing the researcher to spatialize the so called ‘sample-data’ from the construction files. However, despite the immense efforts put toward producing and spatializing sample data that would give insight into the city’s material stock, the researcher tells me that the data is not reliable enough to determine individual buildings’ substances: OGD does not account for the building periods of 80% of the buildings in outer districts. Additionally, being well acquainted with the quirks of the building industry, the researcher knows that building periods and construction files are not always perfect indicators for the used materials, rendering the sample-data somewhat ‘unreliable’.

To make matters worse, the data are also considered ‘unclean’, because in practice, there is often a notable disconnect between the construction files and a building’s actual material stock. Deviations from the original construction files are rarely amended throughout the building process and continuously changing bureaucratic standards for filling them in often leaves the files incomplete and ‘polluted’ by obsolete or simply faulty information. ‘You cannot even be certain about what contemporary buildings are made of’, they tell me. According to the researcher, anyone who ‘knows’ the city can easily recognize the approximate building periods of areas and spot whether buildings have been renovated or augmented. Therefore, being rendered in the simulation, the flaws and distortions in the data are easy to spot. The researcher stresses that ‘on the level of visualizing these data relations, this version of the simulation is a real miss’.

Using visualizations to figure out how to relate and spatialize the data ‘behind’ the rendering in a way that is meaningful to intended users presents a more feasible approach to simulating urban development than counting on incomplete public datasets and ‘polluted’ sample data. One instance when visualization proved to be pivotal to the project is when the researcher conferred with waste-management researchers about the practicality of the simulation for stimulating recycling: The simulation was designed to fragment the city map into tiny cells, consisting of only a few buildings, with each cell displaying the predominant material using a specific colour. Initially, it seemed perfectly sensible to extrapolate the ‘material stock’ from the sample-data for each of these tiny cells. Later, when the researcher showed the visualization to waste-management researchers, it turned out to be a misstep: The ‘waste-managers’ find the degree of detail totally unnecessary, as the local waste-industry treats the city’s geography as only a few large blocks corresponding to the geographic distribution of waste management facilities.

Thus, in the subsequent phase of the project, the researcher is now approaching things differently:

This time we will rely less on data. Before simulating, I show the collaborators mock-ups of how I want the interface to look and what the simulation should be able to do—taken directly from the scribbled notes and put into Miro.’

The researcher opens an online whiteboard with crude images of interfaces and ‘handwritten’ notes. ‘I have learnt from the last phase of the project that it is necessary to be very graphic when explaining things. Now, I know that I need to guide others through my plans.’ One issue here seems to be the differences among the ‘visual languages’ of project partners from different disciplines.

The engineers and waste-managers use Excel-lists and flowcharts to communicate their ideas of how to relate data. As spatial planners, maps and 3D-visualisations are our ‘native language’, well … it was a mistake to just explain things verbally and to not sit next to each other and process what is happening by sketching out your thoughts on a piece of paper. Visualising your ideas really makes you re-evaluate your conceptions of the city and what matters.

This encounter presents simulation as a material and collective process, with producing visualizations to make sense of and visualize data (relations) in a meaningful way at its core. Simulation is a material process in the sense that (among many other practical aspects) it encompasses making different digital and pen-and-paper visualizations. These visualizations afford different ways of seeing and making sense of urban development. As discussed earlier, simulation involves tinkering—assessing, representing, materializing, and relating data through various visual iterations such as Excel-lists, charts, 3D maps and scribbles on paper (e.g. Farías 2015; Houdart 2008). Here too, simulation turns out not to be a straightforward translation of data into a visual representation, but an iterative process, mediated by the researchers’ and their partners’ situated, professional, and common sense knowledge of the city, and the affordances of different visualizations.

Urban development and its materiality emerge as barely graspable through data alone, which turned out to be unreliable, polluted and inconsistent. This led to changes in strategy: instead of taking data as a starting point, the researchers started the project’s next phase by sketching charts and mock-up interfaces to make sense of how the relations between datasets and spaces could be assembled in order to simulate what they were interested in. To make the data and the researcher’s vision of how to implement it in the simulation ‘readable’ to project partners, it needed to be visualized in ways that would accommodate actors’ diverse disciplinary backgrounds and visual habits. The researcher mentions the necessity to ‘translate between their own language’, 3D models, maps, code, and scribbles, and partners’ visual languages by drawing. The researcher appreciates drawing as it encourages them to repeatedly rethink and reconstruct their endeavour and bring it into harmony with research partners’ interpretations. Thus, datasets, local interests, knowledges, institutions such as the waste-management industry, and the different material affordances of visualizations collectively constitute the simulation.

The encounter with waste-management illustrates that making visualizations can reveal that not even seemingly trivial things such as selecting the ‘right’ cell size to spatialize data on the city map are self-evident. Here too, enacting the ‘object’ of the simulation, material flows in the city, through visualizations such as scribbles, charts or renderings, has a multiplying effect. The epistemic dissonances emerging from such visualizations make actors’ different relationships to the city visible. If irreconcilable, such dissonances might even stand in the way of a consolidated, ‘useful’ simulation.

While the researchers in the first vignette embrace multiplicity as enriching their understanding of sound as a matter of care, in this vignette the dissonances end up informing the researcher’s decisions about what to make (in)visible. For instance, increasing cell-size to accommodate the waste-management industry’s vision obscures the city’s diverse material make-up. By making sense of urban development through different visualization practices, researchers could localize dissonances, get a sense of how data could be aligned and spatialized otherwise, decide which inconsistencies to make invisible, and omit elements such as ‘polluted’ data, that ‘disrupt’ what they perceive as a useful simulation. In this way, the researchers’ aspirations to produce ‘useful’ visualizations that faithfully represent specific aspects of the built environment, resonate with Igelsböck’s (2016) observation that computer simulations include ‘black boxing, delegating, integrating and sorting out’ (152) data and knowledges. In this sense the process of simulating urban development and visualizing invisible things entails invisibilising the multiplicity emerging from data-visualizations that look ‘wrong’ or ‘unnecessarily’ granular city maps.

What also becomes visible here is the porosity of simulations and the ways in which materialities, data, embodied practices of seeing and showing, and local perspectives are entangled with the design space. In attempts to align perspectives and create a single coherent simulation, various (data) practices, perspectives, different enactments of urban development, and moments in which the politics of simulation are negotiated become obscured. But in the design process, these diverse data and visualization practices are prominently visible and shape the code. This implies not only that ‘the linguistic practices of [humans and software] influence and interpenetrate each other’ (Hayles 2005, 59), but also that the many interstices in which urban development is being enacted are folded into the singular appearance of the simulation, rendering them part of these ‘digital’ spaces.

4. Discussion

In presenting and discussing two vignettes, we have outlined some aspects of the politics of (in)visibility in the development of urban simulations. By examining the involvement of experiments, data and visualizations within two simulation projects, we were able to analyze how their ‘objects’, urban sound and development, are materialized and what becomes visible and invisible in the process. This concerns both (in)visibilising emergent enactments and affordances of simulated ‘objects’ and the perspectives involved in developing the simulations. We now briefly reflect on these findings, and on how the lived experiences and structural constraints of researchers are implicated in such politics.

In the first vignette, we saw how enacting sound through different practices such as ray tracing or the sand experiment displays multiple acoustic materialities, accentuating either propagation or frequency by enacting sound as a particle or a wave. The acoustic materialities emerging from these enactments, researchers’ affective responses, their relations to the city, their aspirations to care for urban (sound) ecologies, and their subjectivities were entangled with the unequal possibilities of simulating the multiplicity of acoustic materiality. While the experiment opened up to the researchers what physical and emotional affect sound, enacted as waves, might unfold, these newfound potentialities needed to be reconciled with what is do-able in the project. Although invisible in the simulation, the experiment and the dissonances it reveals leave behind new perspectives on caring for environments through architecture and a plethora of technical and moral questions to inspire future practice.

The second vignette highlights how urban development is enacted through different visualization practices that reveal its multiplicity, such as creating sample-data and relating it to the city map with public data, thus spatializing it, and drawing sketches and mock-ups. Thereby, dissonances between the different data sets, the researchers’ common sense of the city, and the different visual languages of the professions involved in the project became apparent, emphasizing the necessity to reassess the project’s strategy. We showed how these dissonances are concealed to produce a ‘useful’ simulation, for instance by withdrawing from data as a point of departure or adapting the scale. Our argument has been that close examination of such processes can show how knowledges and practices are coordinated, and how emergent ambiguities are dealt with and hidden in later stages of the simulation. The residues of such ambiguities and invisibilised practices are however more obvious in this case: The departure from the idea of a complete and reliable dataset shifts the emphasis to visual communication strategies that accommodate actors from other fields and brings forth various charts, graphs and sketches. Crucially, the process leaves behind an interface that inscribes the perspective of the local (waste)industry and obscures the material diversity of the building stock.

Looking closely at these (in)visibilising processes has made visible the values, tacit knowledge, materialities, affect and multiplicity that underpin simulations and complicate their ostensibly unambiguous reductionism. We consider moments when ‘complications’ are dealt with as moments of ontological politics because they give shape to how simulations become. Such moments unlock opportunities for researchers to explore and (re)define their ‘objects’ and how to do architecture within the spaces they inhabit. Simulating is thus a process of paying attention, of ‘thinking with eyes and hands’ (Suchman and Trigg 1993, 153), and of carefully (un)folding the multiplicity of ‘objects’.

What does this add to critical studies of simulation and (in)visibility? First, we add to the research on invisible (data) practices. We present how ‘disembedding background work’ (Star and Strauss 1999, 15)—such as the visualization practices we have described—can make the multiplicity and locality of architecture and spatial planning simulations, like the situated knowledges and institutional structures permeating them, invisible. Yet wider ethnographic experience with academic architecture research suggests that some of the associations made by local visualization practices—the ways they shape projects, articulations of architecture-society relations, researchers, and their ‘objects’—stay invisible in project proposals, funding applications and conference presentations, but leave residues in the simulations themselves. Is this lack of attention a rejection of complexity, echoing what is broadly considered do-able or a core objective in architecture (Sanchez 2019)? Or is it holding on to Computer Aided Design’s promise of a ‘utopia of a total representation’ (Cardoso Llach 2015, 101)—a finite space of possibilities that can be fully explored?

We cannot answer these questions here, but have observed the nuanced views of researchers concerning what their simulation can and cannot represent. Being less concerned with ‘total representation’ when enacting sound or urban development, researchers discover potentialities of ‘objects’, find ways to relate to them, and crucially, further explore and develop their own skills and politics. Within our vignettes, researchers, simulations and simulated ‘objects’ become together in the process of creating simulations, providing insights into the intra-action (Barad 2007) of data, people and matter. Thus, our second contribution adds to research exploring how the scientific self, affect and embodiment are entangled with performing and making sense of ‘objects’ (cf. Myers 2009, 2015).

Third, we would like to stimulate discussions that challenge the ostensibly given and stable parameters and perspectives inscribed in simulations; how they come about, at what cost, what traces of obscured enactments are discernible, and how such inscriptions could have, and still can, be done otherwise.

We see the invisibilised dissonances, visualization practices and the ways they help researchers explore hidden aspects of built environments discussed in this article as part of growing critique of traditionalist conceptions of architecture that reject expanding response-abilities and fail to recognize the importance of invisible things (e.g. Sanchez 2019; Schumacher 2011). We have learnt from our interlocutors that transgressing perceived disciplinary boundaries can be challenging and time consuming. For example, going beyond the use of ready-made software to explore how sound affects urban environments and matter complicates a seemingly straightforward task. However, it also causes excitement and promises to give architecture research new avenues to care for other worlds and to explore a ‘natural world […] cohesive beyond these narrow territories’ (Cupkova 2019, 48) delineated by perceived disciplinary boundaries (Sanchez 2019). We believe that such discussions could make room for new, more explicit normative commitments and make architects’ and planners’ politics more visible. Such ambitions should be pursued in a reflexive way that is wary of latent encroaching/colonizing inflections in the vocabularies and metaphors mobilized in theory texts, and the hierarchies that may emerge from working together with others using digital design platforms (Cardoso Llach 2015).

We find that the ecologies of practice and associations that become visible upon paying attention to mundane visualization practices, rupture the boundaries of a supposedly discrete digital design space (Cardoso Llach 2019). Attending to visualization practices has made the embodied, material practices and local contexts that lie beyond what we can see on the screen, visible. It allows us to observe how taken for granted things such as the materialities of the ‘objects’ being simulated and the stability and reliability of the data representing them come undone. Crucially, tracing visualization practices unfolds pivotal moments when things can be done and thought otherwise and opens them up to reconsideration. Our account here has displayed some of the work and ingenuity that enable visualization practices, and indicated that they are careful, situated explorations of built environments and spatial politics. Crucially, they allow researchers to integrate and develop what they care about, and who they become as researchers.

Acknowledgements

We would like to thank the architecture and spatial planning researchers who have, through sharing their work, inspired this article. We are very grateful for their trust and the insights that they have so generously shared with the first author. We also thank our colleagues Kathleen Gregory and Bao-Chau Pham for reviewing this article and supplying literature resources that have substantially contributed to our argument.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Notes on contributors

Esther Dessewffy

Esther Dessewffy is a PhD candidate at the department of Science and Technology Studies at the University of Vienna. Ethnographically researching practices and materialities engaged in making computer simulations for architecture and spatial planning in academic contexts, Esther maps the politics of the more than human relations that imbue simulations.

Andrea Schikowitz

Andrea Schikowitz is a postdoctoral researcher at the Department of Science and Technology Studies, University of Vienna. Her current research focuses on knowledge practices and -infrastructures in distributed urban planning and controversies.

Sarah R. Davies

Sarah R Davies is Professor of Technosciences, Materiality, and Digital Cultures at the Department of Science and Technology Studies, University of Vienna. Her current work explores the inter-section between digital and epistemic practices and forms of life.