Studying scale and scaling in the haptic design process of two architects who lost their sight
In this chapter, we explore how two architects who lost their sight found ways of continuing their practice.1 The first architect we study is Carlos Mourão Pereira, a Portuguese architect who lost his sight in 2006. However, he continues to run his small-scale office and even found his sight loss an opportunity to further his understandings of the multi-sensory nature of the built environment. In this context, he started designing a series of bathing facilities, and our study focuses on the design of one facility in particular: the Sea Bathing Facility in Lourinha on the Portuguese coast. These facilities are not yet built; for Pereira, they are in the first place a way to learn new design methods and explore his new understandings of space. The second architect, Christopher Downey, already had a more extensive career in architecture when he lost his sight. But, like Pereira, he chose to continue his design practice. He joined Smith Group, an architecture firm based in San Francisco, to assist them in the design of a new polytrauma and blind rehabilitation centre in Palo Alto. He is a design consultant for people with a visual impairment and assists in the communication with the client organisation, of which some people are blind as well.
In both Pereira's and Downey's practices, as in any architectural practice, scaled representations play an important role in the design process. Scale in architecture considers the different relations between a building and its elements, its environment or context, the human body, and the design process. Furthermore, these static notions of scale can be extended with a more dynamic notion of scaling to describe how architects develop a design by moving cyclically between different scaled models.
For Ralf Weber and Silke Vosskoeter (2008), scale is defined as ‘the perceived or apparent size of a building’. They analyse scale, then, in terms of human scale, outer scale and inner scale; in other words, how the building relates to the size of people, the context of the building and the building's individual architectural elements. This notion of scale only concerns the perception of an existing building. Phillipe Boudon (1971), on the other hand, defines scale as a relationship between the conceived architectural space and the perceived architectural space, linking the building to its conception; that is, the design process. Architectural space in this definition is ‘a whole consisting of two spaces, real space and mental space’ and scale is ‘the rule of transition of one space into the other’. Albena Yaneva (2005) goes even a step further in relating the notion of scale to the design process. She argues that it is insufficient to understand scale as a static notion of a proportional relation between a model and the real building, referring back to Boudon. Instead, she advances the more dynamic notion of scaling as ‘some frequently repeated moves, such as “scale up”, “jump”, “scale down”. Their successive repetition and redundancy compose a rhythmic conduit through which the building develops’.
However, the notions of scale discussed so far are interpreted implicitly or explicitly in a mainly visual way. Overall, architecture is characterised by a bias towards the visual. This features in how the built environment is analysed in architecture (Pallasmaa 2005), but also in the way designers in general think and work (Cross 1982). Martha Dischinger (2006) finds evidence for this overemphasis on the visual (and also intellectual and conceptual) dimensions of architecture in the traditional tools to represent architecture using mostly visual media. Both Pereira and Downey thus had to reinterpret, or even reinvent, the tools and representations they were used to in their design practice to overcome the consequences of this visual bias in architecture and its design process. They rely more on their other senses when interacting with their tools and representations. For instance, Pereira experiments with sound, and even smell, for the (re)presentation of his sea-bathing facility, a basin that offers a multi-sensory experience of the sea in a safe and more controlled environment (Vermeersch and Heylighen 2011). But when it comes to their day-to-day design practice, both architects rely to a large extent on their haptic sense to interact with the numerous artefacts involved in designing.
We will refer to the design tools Pereira and Downey use as haptic. They are developed to provide design information through the skin and muscles of the hands. Jack Loomis and Susan Lederman (1986) define haptic perception as a combination of tactile perception and kinaesthetic perception. Tactile perception is mediated solely by variations in cutaneous stimulation, which provides information by means of receptors in the skin. The cutaneous sense can inform you, for example, of the temperature, roughness or texture of a certain material. Likewise, kinaesthetic perception is mediated by variations in kinaesthetic stimulation, which provides information about dynamic and static body posture by the relative positioning of head, torso and limbs. For instance, perception of the length of a rod held between thumb and index finger is informed by kinaesthesis. So when Pereira and Downey evaluate a model by taking it in their hands, or letting their hands move over its surfaces, they are informed of its shape, size, temperature and texture by the haptic sense through their fingertips, and the relative positions of their fingers, hands and arms.
How, then, are design artefacts used in a haptic way? And how does this relate to certain haptic qualities that Pereira and Downey want to investigate through these scaled representations?2 One aspect that makes visual representation so powerful is optical consistency while scaling, so that different scales (for example, city and building, building and building detail) can be reshuffled and recombined (Latour 1990). However, when trying to find these qualities in haptic representations, this seems problematic at first sight, because of the proximal nature of the haptic sense. While scaling in visual terms maintains internal proportions – the elements of a pattern look smaller when taking a few steps back, but the pattern itself is not perceived differently – this is not so evident for a haptic pattern. Taking a few steps back is not possible when dealing with haptic perception. Scaling the texture of a surface will maintain internal proportions, but the haptic perception of that texture will alter much more distinctively than when perceiving that texture mainly visually.
When Christopher Downey joined the Smith Group, he had to find a way of designing that could fit in with the day-to-day practice of the architects working there. As they rely heavily on BIM (building information modelling) based on an integrated CAD model, Downey had to find a way to access the CAD drawings taken from this model and to formulate and communicate design ideas based on his interpretation. Therefore, he plots these drawings on an embossing printer (a type of matrix printer that prints Braille dots and patterns on a thick sheet of paper), which he can then read with his fingertips and hands. To design, he combines these drawings with Wikki Stix®, wax sticks he can cut, bend and stick to each other and to the paper (Figure 1). This allows him to test design ideas by shaping some sticks and temporarily fixing them to the original drawing. After reading the whole with his hands again, he can re-adjust the sticks and manipulate the sketch he has just made, analogous to sketching on tracing paper laid over a print-out or another drawing.
In Pereira's case, we observed a similar dynamic process of manipulating and (re-)interpreting a design situation, but in a way that is somewhat further removed from more day-to-day architectural design tools. Pereira too uses his hands to explore a model or a raised line drawing, and to manipulate that model or drawing, but his hands sometimes even become the model. In a conversation we observed, Pereira used his hands to explain a certain shape to one of his collaborators. His collaborator took that hand and started pointing on it to address specific points in the design, and even started gently manipulating the shape of Pereira's hands. This became a very fluent process of fixing and manipulating to further develop this shape. As Pereira explained, ‘the hands can become anything’, so they can change scale level almost instantly within a design conversation. Using their hands, Pereira and his collaborator can go back and forth between building part, building, site and so on.
Figure 1 Christopher Downey combining Braille plots with wax sticks to design.
Both Pereira and Downey are seeking in their more haptic design practice to incorporate artefacts that allow them to dynamically develop their design ideas in interaction with spatial representations and other collaborators. A design is not just an outcome of a process going on in the designer's mind, but grows in the interaction between the designer's thoughts, actions (like sketching) and the objects being created (like drawings). Donald Schön (1983) describes this process as a reflection-in-action, where ‘[the designer] shapes the situation, in accordance with this initial appreciation of it, the situation ‘talks back’, and he responds to the situation's back-talk’. Yaneva (2005) relates these interpretative actions of designers to not just one drawing, but to a particular rhythm of scaling up and scaling down, as moving between different scaled models of a design. In the example of Pereira, we observed this process of scaling as it occurred within a single flexible ‘model’ being the hands of Pereira and his collaborator. This way of designing is almost like a conversation in that it allows for a flexible interpretation and a quick change in configuration, but it has the advantage of a model as it constitutes a spatial representation. On the other hand, it is more fleeting in nature as there are no physical traces of this haptic communication.
Of course, Pereira also makes more lasting representations of his designs, and again we observed this concern with different scales of models that together form the sea-bathing facility. What the right scale is for one of these models depends on the aspect of the design that needs to be investigated, the haptic nature of handling the model and the model itself. For instance, Pereira makes a series of models of the sea-bathing facility that all concern certain aspects of the design, from the larger context of the site to a more detailed spatial configuration of the bath itself. These models are all at different scales, but of approximately the same size. Bryan Lawson (1994) has found a similar phenomenon in a more traditional visual design process. Several architects he interviewed liked to sketch on a sheet of A4 or A3 paper because this allows for a quick overview. The size of the models Pereira makes also allows for a quick ‘overview’, not in a visual sense, but haptically. The size of Pereira's models is thus partly characterised by how he can encompass these models by his hands allowing for an exploration of a specific detail without losing contact, literally, with the broader context of the whole model (Figure 2).
Next to these smaller models, Pereira makes (or asks a collaborator to make) larger models where his hands can go inside. These give him more information about the spatial qualities. We perceive spaces from within, so besides models he can encompass with his hands, Pereira also needs models he can enter with his hands to let his hands be encompassed by them. These models are of course larger than the former ones, but together they allow Pereira to go from a representation of the whole to a more detailed representation of the space itself. In a workshop with students, we observed how, at one point, he felt both types of models at the same time. A student had to explain a house using a number of 3D models she made. One smaller model gave an impression of the outer volumetric of the house, and could be held by Pereira in one hand. The other model was larger and represented the different volumes of the separate spaces of the house. The latter gave him more detailed information about the spatial qualities of the inside of the house, while the former allowed him to put these spaces in the context of the larger whole.
Figure 2 Carlos Pereira using multiple models at different scales encompassed by his hands.
Like Pereira, Downey too had to overcome some particularities of haptic perception when producing his Braille prints. For him, it was not possible to just print out the CAD drawings that his colleagues had produced for an ink and paper plot. As he remarks: ‘you need to get the size that is appropriate, and you need to shed some information. […] Visually you can easily separate things, but if all that becomes tactile, it's overwhelming.’ He had to find the appropriate scale to be able to read the different plans, sections, elevations and so on with his hands. But also the limitations of the technology to create these drawings have their impact on the chosen scale. The maximum paper size is A3 and the resolution (the distance between the raised dots) is limited. Therefore, Downey needs to discuss with a colleague how to adapt the amount of information in a single drawing and its appropriate output scale. Unlike Pereira, who makes models to the size of his hands, Downey chooses to print out the drawings as large as possible to incorporate as much detail as possible. He then reads such drawings part by part to get a gradual understanding of the whole.
In the previous section, we explored how Pereira and Downey had to alter their design practices to keep designing, and how they did so by relying mostly on their haptic sense. In the next section, we will further analyse how this haptic design process relates to haptic qualities of their designs. Their specific interaction with the haptic models they make allows them to investigate certain qualities that are more in line with their ambitions to develop a more multi-sensory architecture, learning from their bodily experience of the built environment.
Since Pereira lost his sight, he started developing a new understanding about aesthetics in architecture. Before, he was searching for a strong visual image, much in line with contemporary Portuguese architecture: clean surfaces meeting in distinct sharp edges, making for a clear visual composition and interplay of lines. Now he rather explores shapes that are more comfortable to touch, no sharp edges that cut the hand, but shapes that are better adapted to the hand and invite to touch (Vermeersch and Heylighen 2010). Very important for him in this sense is how he can gain an adequate understanding of the haptic qualities he is trying to pursue in his design. As described above, haptically representing architectural concepts about, for example, location, approaching a site, general layout and so on, is one thing. However, actually representing the haptic qualities of a building is something different. One point of departure for both Downey and Pereira is that certain places in a building are more likely to be touched than others, and when designing for the haptic sense you should first and foremost pay attention to those spots.
To begin their exploration of a building's haptic qualities, Pereira and Downey concentrated first on discrete points in space where people are already touching (parts of) the building. Downey mentions doorknobs, handrails, table tops and so on; Pereira pays considerable attention to the design of handrails and also mentions doorknobs, window sills and so on. The small size of these parts allows them to explore these on a scale of one to one, which is a very appropriate scale to get haptic feedback. For instance, Pereira makes clay models to explore the haptic aspects of a particular shape, such as the ending of a handrail (Figure 3). In such models, the material is of less importance for him, and is dictated by the ease of working and the shape he wants to test, rather than by the tactile aspects of, for instance, the texture. Even on a scale of one to one, they still work with models and not with prototypes.
In this same sense, Downey reinterpreted a tool already in use in the design firm he joined. The interior designer there used to work with material palettes to get a feeling of the visual composition of materials in future spaces. At one meeting they discussed the difference in texture between two flooring materials, which were chosen by the interior designer to contrast with each other. But this choice was mostly informed by the visual sense. In the discussion, which usually proceeded verbally and visually, Downey put the materials on the floor and tried to distinguish between these materials by moving his cane over them, and found no such contrast. In this example, the shapes of the material samples are less important; rather, it is all about the difference in textures of flooring materials and the differences that are perceivable through a cane.
When Boudon (1971) studied the question of scale in architecture, he was still discussing the design process in architecture as a mental matter. However, architects have distributed their design and building processes over multiple artefacts from as early as 2100 BC, increasingly so as history progressed (Porter 1997). So when Boudon defines scale as the rule of transition between real space and mental space, we should regard the mental space as only partly playing out inside our minds and for the other part being distributed over the different externalized representations produced in a design process. How the conceived architectural space relates to the perceived architectural space then becomes apparent in the design process of Pereira and Downey. We observe a close similarity between their haptic perception of the models and artefacts they use in their design process and the aspired haptic qualities in the designed building or part of the building.
Figure 3 Carlos Pereira exploring haptic qualities while shaping a clay model.
When Pereira is shaping the end of a handrail using a block of clay (Figure 3), the movements he makes with his hands during that process are analogous to the movements his hands would make using that handrail. During this process he manipulates and interprets the model, both in a haptic way. His haptic reading then informs him of the haptic qualities of the handrail he is designing. Tim Ingold (2000), in discussing the theories of James Gibson and Maurice Merleau-Ponty, reminds us of the importance and the role of movement in the act of perception. We are not motionless observers of our environment, but we are actively seeking out information in that environment. If we perceive a space, we are mostly moving through it and, even if we take a pause in our walk, we still move our heads, scanning the environment with our eyes, while sounds and smells come to us informing us of our next steps. Therefore, the relation between the movements Pereira and Downey make during their reading of their respective haptic models and the movements they make in the built environment allows them to anticipate haptic qualities while designing.
Although Pereira and Downey interact in a similar way with their models as with the environment these models represent, this is not a literal translation. When Downey puts the material samples on the floor, he is still limited in his movement by the size of the samples, and focuses specifically on the question of transition and contrast. Also, when Pereira shapes his clay model of a handrail, he still sits at his desk and is not moving down the stairs. Peter Paul Verbeek (2005) uses the concepts of ‘reduction’ and ‘amplification’ to discuss how objects, and thus also design tools, mediate our perception of the environment. These design tools allow Pereira and Downey to focus on a certain aspect of a material or a distinct part of a building, but these models are still representations and not the building itself. The interaction with these samples then is also partial in comparison with the eventual interaction with the built environment, but nonetheless brings Pereira and Downey closer to a perception of what it could be.
The models and samples used in the above examples are still on a scale of one to one and thus allow for a direct comparison between the perceptions of the models and of the built environment. They can give an equivalent impression of the haptic qualities in these design solutions. But we also found some more immediate analogues between the haptic reading of a scaled drawing or model and how Pereira and Downey move through a building. Due to Downey's choice of printing his Braille drawings on a larger format of paper, he has to explore that drawing sequentially. The way he then moves his hands over a plan, for instance, follows the same lines and routes as if he walked through that building, and he searches with his fingers for elements he would search for in that space with his cane. In the interview, he refers to the position of columns in a corridor. One idea was to place them in front of the wall, but as Downey went through the plans, he found that this would hinder his movement through that corridor as he would be unable to just follow the wall with his cane (or follow the line representing the wall on the Braille plan with his finger).
We found a similar handling of scaled models in the case of Pereira. He also moves through his larger models, those he intended to go into with his hands, by following the walls with his fingers as if he were exploring that space when built. In a sense, his finger becomes his body, mimicking bodily movements through space, and thus informing him about the perceptual qualities of the design. For Boudon (1971), scale had to do with the relation between conception and perception, but as these examples clearly illustrate, the process of conception cannot take place without perception. For instance, perspective drawings are widely used in architecture to give an impression of the visually perceivable qualities of a design. However, as Porter (1997) remarks, ‘perspective was a contradiction to the very nature of visual perception as it caused the viewer to freeze in time and space’. On the other hand, in the design processes of Pereira and Downey, this aspect of motion is crucial to their haptic interpretation of the representations they make in such a way that it relates to their own haptic perception of the environment. Therefore, the decisions they make concerning these haptic qualities, be it through a model on a scale of one to one or on a smaller scale, are informed by a process of perception itself. In that sense, we could state that scale is also a relation between perception in the design process and our perception of the built form.
If movement is crucial in the perception of the (built) environment, as Ingold (2000) pointed out, it is equally important in the perception of tools used in the design process. Although these movements are not entirely the same, as we have argued, in a way they are linked through the architecture of models and building, allowing the architects to use movement within the model to get an impression of what the movement in the built form might be like, and vice versa. Perception of the built environment might colour the way Pereira and Downey perceive their models, and how they relate one with the other. Or, as Boudon (1971) formulates it: ‘just like the conception of architectural space makes perception intervene, the perception of the architectural space cannot take place without letting conception intervene’.
In addressing the notion of scale, we studied the design tools of two architects who lost their sight at a certain point in their careers, and who then decided to continue their design practices. Our study focused on the haptic aspects of the tools they use, and how these tools help them to design for haptic perception. Most of these tools in themselves are not so distant from the known tools used in the design process in architecture, but Pereira and Downey use them in a haptic rather than visual way. The power that resides in scale and scaling is explained, in part, by optical consistency. This aspect, however, is not so readily transferrable to further an understanding of how scaled representations can help a design process in architecture based (for the most part) on haptic perception – within this design process, but also in relation to the built environment. Nevertheless, both Pereira and Downey still use scaled models and drawings to design. They scale whole buildings or specific elements to the size appropriate, not for a quick overview but for encompassing or exploring with their hands. Yet also what they want to investigate with these models and drawings is also important in choosing the right scale. First focusing on specific building elements to explore haptic qualities of a design, they work at a scale of one to one, relating their perception of the model closely to the perception of, say, the handrail to be built. But also in smaller-scale models or drawings, we observed how the way in which they move through a raised line plan or a model is not so far off from how they would move through that environment when built.
The study of scale in the haptic tools Pereira and Downey use teaches us that scale has to do not only with the relation between conception and perception, or with moving between representations. Scale also relates perception of models and drawings to the perception of the environment to be built.
This study has received funding from the European Research Council under the European Community's Seventh Framework Programme (FP7/2007–13)/ERC grant agreement no. 201673. Peter-Willem Vermeersch received support from the Research Fund K.U. Leuven. We would like to thank Megan Strickfaden and Greg Nijs for their share in the data collection. Last, but not least, we would like to thank Carlos Mourão Pereira and Christopher Downey for their time, enthusiasm, patience and honesty.
1 The material we base our studies on comes from multiple sources. The study of Chris Downey relies on published materials and a semi-structured in-depth interview we conducted in the office where he works. For the study of Carlos Mourão Pereira, the material comes from published documents (writings, drawings and photographs) and a semi-structured in-depth interview, complemented with observations in his office and observations of a workshop with students. The interviews were audio-recorded and transcribed for further analysis. The observations were recorded in field notes, and photographs and video taken by a collaborator of Pereira.
2 Pioneering work on haptic qualities in the built environment has been conducted by Jasmien Herssens. She studies haptic perception in architecture by working together with blind people and learning from their daily interaction with the built environment (Herssens and Heylighen 2009, 2010, 2011). Her work inspired us to further study aspects of haptic perception in the design process.