Introduction

I began to reconsider my opinion of humanitarian architecture on a wet November day in Venice, as rain lashed down on the city and darkness had just fallen. At the time, I was visiting the 2016 architecture Biennale, which had the theme ‘Reporting from the Front’: a phrase chosen by director Alejandro Aravena to encapsulate the way that architecture is ‘an endeavour that has to tackle many fronts, from guaranteeing very concrete, down-to-earth living standards to interpreting and fulfilling human desires’ (Aravena, 2016: 3–4). It was the last week before the Biennale closed for the season, and I had, over the previous summer, read a great deal of enthusiastic commentary on the event and its explicitly humanitarian intentions. I was keen to see the exhibits, especially given my long-running scepticism about the ability of architects to play a useful role in humanitarianism. However, after walking through the many rooms and halls of the Biennale on my first day there, I remained distinctly underwhelmed. It seemed to me that the event was dominated by unrealistic, top-down designs that were too resource intensive and far too politically unpalatable to ever work in practice. But then I stumbled upon the Austrian pavilion.

At the end of that wet afternoon my feet were tired from walking, my head was bursting from seeing too many unworkable ideas and I was looking for somewhere to shelter from the rain. I saw a white modernist building lit up in the dark, tucked away in a far corner of the Giardini. I ran to take cover. It featured an exhibit called Places for People: a sparse but simply furnished demonstration of real interventions rather than idealistic projections, describing three projects that had worked with refugees to make modest but important improvements to their emergency shelters. The ideas were a refreshing change from the rest of the Biennale because they were so practical and focused on everyday life, with thoughtful and humanistic ambitions. The projects were based on a simple idea: not to construct new shelters but to improve the empty office buildings that lay empty across Vienna after the financial crash. The walls of the bright white pavilion were illustrated with simple photographs, quotations and publications describing the approach, transforming dull grey offices into liveable accommodation by focusing on furniture and furnishings. Among other things, this entailed installing softer lighting, distributing simple materials to filter the harsh florescent bulbs, erecting divides to address the lack of privacy and adding splashes of colour and comfort throughout. It was, I immediately felt, an important if modest idea.

The three Viennese projects were simple but effective, cheap but transformative, fast but sensitive. They had been implemented with small amounts of money and a tight timeframe, yet they managed to respond creatively to the peculiar situation in Vienna, where huge empty office buildings had been allocated to shelter new asylum seekers during the ‘summer of migration’ in 2015. The architects had focused on adding simple furnishings that created a more homely environment, articulating a careful, human-centred approach that had interpreted shelter not as four walls and a roof but as a calming and secure internal space. The aim of these projects was to turn impersonal and soulless ‘non-places’ (Augé, 1995) into ‘places for people’ (Delugan-Meissl et al., 2016). The idea left a real impression, especially as I had spent much of that year examining mass refugee shelters in cities such as Berlin, where the vast Tempelhof airport had become a symbol of a bureaucratic and unhomely response. I was so impressed by the Austrian pavilion that, a few months later, I travelled to Vienna to visit the projects and speak to their designers in detail.

This article presents some reflections from the research I conducted on that trip, as well as conclusions from a much wider study of humanitarian shelter that I have run from the University of Oxford over the past three years. It argues that architects involved in humanitarianism could benefit from channelling the spirit of the Viennese projects and, rather than focusing on new buildings and structures, should think about the objects that populate them. This, I suggest, can offer a valuable corrective to the trend towards ‘innovative’ technologies that fetishise final products, ignore social life and narrow attention to basic biological needs. The projects in Vienna, in other words, imagined more creative ways to create ‘Places for People’ in situations of displacement – places that were practical, cheap and substantially improved daily life. The research presented in this paper emerges from interviews and fieldwork conducted between 2016 and 2019 as part the Architectures of Displacement project, which was funded by the Economic and Social Research Council of the UK and managed from the Refugee Studies Centre in Oxford.²

In the next section of this article, I set out a series of common criticisms of architecture by humanitarians, pointing to frequently unrealistic utopianism and a lack of practicality. In the second section, I set out the differences between innovation and architecture, summarising the key issues with humanitarian innovation and suggesting that these do not necessarily apply to architecture – and illustrating innovation with reference to the Ikea-funded Better Shelter. I describe this as an instance of ‘building without architecture’: the construction of a house or home but without the sensitivity that really marks out good-quality architectural thinking. In the third and final section, I turn back to the Viennese case study in order to describe the opposite position: ‘architecture without building’. This involves designing on the smaller scale, thinking about objects, furniture and the items that populate a space. This approach, I conclude, presents fertile new ground for architectural humanitarianism.

The Humanitarian–Architect divide

Why do so many designers associate shelter with building? Why is construction so often the default response? These questions are central to the way that humanitarians have historically engaged in the problem of emergency shelter (Burnell and Sanderson, 2011; Davis, 2011). For many years, the idea that ‘shelter is a process not a product’ has been one of the most central, foundational principles of the sector: the notion that humanitarians should help people shelter themselves, rather than providing a completed building to inhabit (Davis, 1978; Davis and Alexander, 2015; Sanderson and Burnell, 2013). Final structures and prefabricated units cost a great deal of money, after all, and rarely take into account the needs and desires of recipients. They can also result in monotonous urban environments, with uniform shelters placed in rows with little thought for public space (Agier, 2011; Hyndman, 2000). In some circumstances, there may be no alternative to constructing new shelters, such as in the wake of natural disasters or the immediate aftermath of a severe emergency. Construction in such circumstances, however, can still be approached in a more bottom-up manner. Many humanitarian agencies, for example, distribute building materials and train local communities to ‘build back better’, rather than constructing complete shelters themselves (Lyons et al., 2010).

When it comes to refugees, constructing shelter from scratch can generate even bigger problems. There are a great many political sensitivities and widespread opposition to camps, and local authorities may balk at any form of building that suggests permanent settlement (McConnachie, 2016; Turner, 2016; UNHCR, 2014; van der Borght and Philips, 1995). If we look at the response to the civil war in Syria, for example, only a minority of refugees ended up living in purpose-built accommodation or large planned encampments due to such sensitivities (Chatty, 2017; Miller, 2017), and, despite the enormous size and powerful imagery of Za’atari and Azraq camps in Jordan, the vast majority of refugees from Syria found shelter on their own terms: staying with friends and relatives, renting accommodation from landlords, occupying abandoned buildings or building their own makeshift shelters. The humanitarian response, as a result, often ended up assisting refugees in their own strategies and included providing cash for rent, distributing basic materials to improve self-built shelters and working with landlords to upgrade rented accommodation in a way that aimed to benefit refugees as well as improving the housing stock. This range of options, in other words, can be pursued without the need for prefabricated, flat-pack homes, and arguably without the need for architects at all.

The problem is compounded by the way that architects and other designers frequently respond by promoting brand-new products that look flashy and exciting but are completely impractical in emergencies. Indeed, a vast number of ‘refugee shelter’ proposals has emerged over the last few years, particularly in response to the widespread media coverage of migration during the summer of 2015: designs that have included villages made from shipping containers (Hill, 2015), new generations of tents in the desert made from ‘structural fabric’ (Nolen, 2015), hexagonal refugee shelters (Gander, 2016) and stackable homes (McEachran, 2014), as well as a series of still more fantastical ideas such as the ‘Wearable Habitation Coat’ (Hartley, 2016) and the proposal to reclaim a ‘refugee island’ from the Mediterranean Sea (Taylor, 2016). These designs have been widely circulated through social media and promoted by architectural newsletters, such as Dezeen and Arch-Daily, with large events such as the 2016 Venice Biennale adding a range of even more ambitious designs to the mix (see also Aquilino, 2011; Charlesworth, 2014; Meinhold, 2013; Sinclair and Stohr, 2006). Faced by this stream of ideas and proposals, many aid workers have concluded architects are moving in the opposite direction from humanitarians. Whereas humanitarians focus on process, architects focus on product. Whereas humanitarians aim to get existing shelters up to minimum standards, architects develop utopian visions in glorious isolation from reality.

Since I first wrote about this humanitarian–architect divide a few years ago (Scott-Smith, 2017), many people responded that the divide is often exaggerated, pointing to more sensitive ideas and experienced humanitarians with architectural training. Indeed, the world of humanitarianism includes many architects, and many more architects consider themselves humanitarians. Nevertheless, it is common to find misunderstandings, stereotypes and a substantial difference between the humanitarian and architectural approach to refugee shelter. As part of the Architectures of Displacement research project I interviewed many humanitarian shelter specialists who, off the record, suggested that architects are utopian dreamers, imposing top-down designs that are completely out of touch with the realities of the field. Many years ago, Stewart Brand subjected architecture to delightfully good-tempered criticism along these lines in his book and documentary series How Buildings Learn (Brand, 1994). The problem with architecture, Brand suggested, was it is so often driven by the grand visions of an expert designer, who focus on producing finished blueprints and plans that are not open to adaptation or truly responsive to the needs of inhabitants. A similar line of argument also emerged in the work of John Turner (1972), who advocated the importance of placing dwellers in control, rejecting the top-down tendency of designers with adaptability, affordability and ease of maintenance. Brand and Turner’s criticisms echo those of many humanitarian staff, who wonder why architects so often come up with impractical and top-down approaches even when faced with emergency conditions.

This despair, and arguably this misunderstanding, cuts both ways. From my experience interviewing architects for the project, I have found many professionals deeply interested in the topic of emergency shelter but dejected by the way that humanitarianism is organised. They look at the money spent on large-scale humanitarian responses and compare this with the conditions people are living in, seeing a narrow range of restrictive, unimaginative, mean-looking shelters in refugee camps and, outside of camps, an unambitious and piecemeal approach that involves a bit of cash, some extra plastic sheeting and the distribution of insulation in winter. Architects often conclude that this represents a lack of vision and creativity. At the very least, it represents a short-term mindset, where bare-minimum standards are addressed rather than a holistic understanding of what shelter can do. From an architectural perspective, many simple mistakes are made in the use of materials, no careful thinking goes into the layout of harsh, soulless settlements and there is generally a need to think about the intricacies of the landscape, the local culture and basic issues such as the quality of space and the direction of sunlight (see, e.g., Weinreich and Montgomery, 2016). Architects see their own training as precisely to consider such issues, which are unfortunately often ignored.

The humanitarian–architect divide, as I see it, can be crudely characterised by the tendency towards utopian, expensive and unworkable ideas from architects, on the one hand, and slow, limited and unimaginative responses from humanitarians, on the other. It is a divide that might be considered a stark caricature, yet it is certainly one that endures in the expressions of each side. My aim in the rest of this article is to nuance the distinction by pointing out that many shelter designs ascribed to architects are, in fact, the work of industrial designers, while also pointing out that many architects are involved in far more sensitive and thoughtful interventions – with the Viennese example being particularly useful in this regard. In the next section, I distinguish between architecture and innovation, pointing to the Ikea-funded Better Shelter as an example of innovation rather than architecture (or, as I will put it, ‘building without architecture’). In the third and final section, I return to Vienna, showing how architectural engagements cannot be reduced to simply building, and using this as an illustration of a more careful and sensitive approach that may have a great deal more potential for humanitarian action.

Building without Architecture

Over the past few years, there has been an expansion in literature on humanitarian innovation: a concept that first burst on to the humanitarian scene in 2009 and has since seen its currency rise and fall (Scott-Smith, 2016). The idea of this movement was to inject new life and creative thinking into the humanitarian sector, freeing humanitarian action from the perceived stultification caused by the unwieldy bureaucracies that dominated in the past (Betts and Bloom, 2014). The concept of innovation has sometimes been articulated in terms of the 4Ps – new products, new processes, new positioning and new paradigms – which flowed from the private sector to turn powerless beneficiaries into empowered consumers (Ramalingam et al., 2009). The central argument, for many, is that the humanitarian sector lacks the cut and thrust of a competitive market: it has become unwieldy and rigid and needs to absorb something of the ‘adapt or die’ ethos that prevails in the world of business.

Critics have for some years pointed to the problems with ‘innovation’. Some have argued that supposedly innovative ideas were nothing new, resembling the move to participation and accountability in the 1990s (Sandvik, 2014). Others have suggested that there was something particularly worrying in the focus on digital technologies, the accumulation of data and the application of biometric technologies, which generated insidious implications for privacy and power (Duffield, 2015, 2019). The whole focus on entrepreneurship has also been subject to critique, since it appears to tie humanitarianism into neoliberal governance and the drive to reduce dependency through a wider doctrine of ‘resilience’ (Evans and Reid, 2013; Pugh 2014; for an alternative view, see Scott-Smith, 2018a). New products and technologies, meanwhile, have been accused of becoming a fetish: objects with seemingly magical powers that emerge from nowhere, obscuring far more serious political and economic inequalities (Scott-Smith, 2013, 2016).

Shelter, like all the humanitarian clusters, has been subject to the winds of innovation, and a good place to see this is in action is at AidEx: a large humanitarian trade fair where inventors and entrepreneurs introduce their ‘innovative’ new products to the marketplace. AidEx is a classic space of innovation: its stated aim is to improve the efficiency of aid and to offer a forum where the private sector and the humanitarian sector can get together. Its director has often remarked on the size of the aid industry, and when I interviewed him in 2017 he suggested that this market was worth well over a hundred billion dollars a year.³ The exhibition floor at AidEx is divided into sectors, showcasing technologies for the nutrition, logistics, water and sanitation, and the event was dominated by the central idea that the right technologies would solve deeply rooted humanitarian problems. It seems very much a male-dominated space, populated by, in the rather uncharitable words of a colleague, ‘boys with their toys’. The shelter section of the expo floor, for example, features a range of tents, drop-in modules, ‘plug and play’ prefabricated units, shelters that stack like Styrofoam cups and expand like an accordion, and a range of other unusual designs.

Very few of these exhibits, however, actually involve any architects. I have been to AidEx for several years now and have always noticed how the participants tend to be product designers, touting their objects as mass-produced, portable units to be shipped and deployed anywhere in the world. In this mixture of engineering, industrial design and entrepreneurship, innovation is very much the driving force, with its concern for profitability and universality. Innovation, however, is not the same as architecture. One might point out that certain generations of architectural modernism fall into the same trap of mechanistic and homogenised mass solutions, yet this is certainly not the central thrust of architectural training, which offers something very different to replicable product design. Architects are meant to design for a particular client, paying detailed attention to the specifics of a site and cultural context. Architects are meant to focus on the unique appropriateness of a single design, carefully tailored to a situation. Architects are meant to consider the ‘softer’ side of shelter, looking at the quality of the space and the sensitivity of the aesthetics. Architects are trained to think about homes as deeply contextual, rooted in iterative processes of design. The result may indeed by utopian and unworkable, but it is very different from the work of innovators and humanitarian product designers, who focus instead on universal shelters for global distribution, on meeting minimum standards at the lowest possible cost and who often fetishise the object, searching for the big and ‘game changing’ idea.

A good way to illustrate this distinction between innovation and architecture is through the Ikea-funded Better Shelter, which is one of the most prominent examples of humanitarian innovation in the shelter sector. This flat-pack shelter unit for refugees was designed by a group of industrial designers in Stockholm between 2009 and 2013, and it is easy to confuse with architecture because it has been widely praised by architectural correspondents in major newspapers and it won the architectural category of the worldwide Design of the Year competition (Scott-Smith, 2018b, 2019; Wainwright, 2017). No architects, however, were involved in its development. Even the main designer admitted, when I interviewed him in 2017, that it is something of a stretch to call this architecture. ‘If it were architecture,’ he told me, ‘it would be a brutalist architecture.’ Rather than thinking about specific locations, he explained, his aim was to produce something universal. ‘It’s a spatial design,’ he acknowledged, but ‘it’s very much a product.’⁴ When pressed on the difference between architecture and product design he stressed that architecture was more concerned with aesthetics, the specific needs of a client and the unique conditions of a site. The Ikea-funded shelter, however, had been devised to respond to the constraints of cost and transportability, while aiming to be applicable everywhere. ‘The end users are by definition unknown, so our considerations are very generalistic about what people need. What is the common denominator for all people on this planet in terms of shelter? We look at how to produce this in a rational way.’ ⁵

The Ikea-funded Better Shelter, therefore, is a prominent example of humanitarian innovation rather than an example of architecture. The central aim of this design, in the end, is not to think about how specific people live but to make a universal shelter that is cheap, practical and affordable. The aim is not to design something comprehensive for a particular site that brings together aesthetics and human utility but to design a product that can be flat-packed, put in a box, and shipped anywhere in the world. The Better Shelter can be considered an archetype of the innovation movement, not least because it was one of the most visible projects emerging from the UNHCR innovation unit, demonstrating many of the core features of an innovation mindset. The whole project was based on private-sector funding from Ikea; it was designed to ‘disrupt’ the established set of responses in emergency shelter (and in particular the prevalence of the tent), and it paid great attention to the use of new technology, such as innovative materials and a photovoltaic panel. Most significantly, the product was shaped by price, impact, scale of production and the desire to produce an affordable product for the humanitarian marketplace.⁶

In order to underline this distinction between innovation and architecture, I propose to describe products like the Better Shelter as instances of ‘building without architecture’. They approach shelters as structures with universal aspirations that respond primarily to concerns of price and mass production, rather than engaging with how people actually use space and improving the quality of their environment. Architects, of course, may be responsible for their own share of insensitive, top-down, mass-produced and monolithic residential blocks (see, e.g., the critique of modernism in Brand, 1994 and Scott, 1998), but they are nevertheless trained in what is fundamentally a humanistic discipline, driven by a particular kind of process. Products like the Better Shelter are quite different, producing structures without architectural thinking. Moreover, as we will see in the next section, architecture is not reducible to building and construction. In fact, some of the best architectural interventions may not actually involve constructing anything at all.

Architecture without Building

When I arrived in Vienna to look in more detail at the projects in the Austrian pavilion, my first task was to piece together the background. It was, I quickly learned, the Biennale itself that triggered the architectural interventions I had first seen on that rainy November day in 2016, because Alejandro Aravena, the curator of the event, had publicised his theme and each country pavilion had responded. In the Austrian pavilion, the scheme was passed to a firm called Liquid Frontiers, who reached out to three architectural practices to plan the various interventions that would become the subject of their exhibit: Caramel architects, EOOS and The Next Enterprise; and a non-governmental organisation, Caritas, was brought on board as a key partner who managed refugee shelters across the country – and in particular some of the large office buildings that were soon targeted for renovation. After meeting with the architects and the Biennale curators, they together decided to use the money and prestige of the event to spend on modest improvements to real shelters, with a view to finding something genuinely practical rather than the flashy and utopian schemes that so often dominated such exhibitions.

The interventions were all united by the idea of creating ‘places for people’ (Delugan-Meissl et al., 2016), and when I arrived in Vienna I visited all three sites.⁷ Each was situated in a different concrete office or administrative building, characteristically bureaucratic: their interiors were marked by the usual, uninspiring aesthetics of a workplace, such as thin carpets and lino, fluorescent strip lighting, a dropped ceiling and large open-plan spaces. Refugees had been placed in these empty buildings primarily because they were available at short notice for medium-term rent; and although they were warm and dry, they were not suitable for medium- or longer-term accommodation. Indeed, when the architects spoke to the residents after several months of use, they all heard a litany of complaints. The open spaces allowed for no comfort or privacy. The mealtimes were fixed and the food was monotonous. There was no way residents could cook for themselves, preventing them from gaining the most basic control over their lives. There was particular dislike of the poor-quality lighting, which seemed to capture all the problems of living in an office block: the bulbs overhead were harsh and they could not be controlled, being turned on and off at fixed times by centre managers. The crucial problem, therefore, was lack of control over social life. People were sheltered in the very bare, material sense, yet they felt completely at sea in an institutional environment in which cultural and social conditions were, by necessity, given secondary importance.

The idea of ‘Places for People’ showed that even basic shelters could become more sensitive, that even office buildings could be converted into pleasant living spaces without too much time and money. Of the three Viennese projects showcased in the resulting exhibition, this approach can be illustrated the first: a project led by Caramel architects, which intervened in a building that overlooked a railway line to the south of the city centre. On the upside, the location was relatively central as well as well connected. On the downside, the building was a drab office with large rooms, dingy internal staircases and poor-quality lighting. It was around 3,000 square metres in total, divided into open-plan spaces that housed up to ten or twenty people sleeping on mattresses in rows on the floor. There were familiar problems with this arrangement, including the monotony of regular routines, a complete lack of privacy and overhead lights that were turned off at 11pm and then on again at 5am. The solution proposed by the team at Caramel architects was very simple but turned out to be enormously effective: invest in parasols, blankets, pot plants and personal lamps. These were installed to create divisions throughout the offices: parasol ‘roofs’, fleece-blanket ‘walls’ and divisions that blocked light from outside and created small, cosy, private spaces with their own small reading lamp.

The firm had just 10,000 euros and three months to realise this vision, which completely changed the feeling of the space. They were dealing with a set of financial and temporal constraints that would be familiar to many humanitarians, yet they demonstrated how architectural thinking did not need to be grand and unrealistic. With around 300 people living in the building and fewer than 50 euros to invest per person, their basic approach was to go to a garden centre and purchase large parasols: clusters of colourful and cheerful umbrellas that could be erected in the open-plan offices and that filtered the harsh strip lighting above. The architects then suspended large blankets around the edge of each parasol, clipped on with cable ties around simple plastic plumbing pipe from a builder’s merchant. This enclosed the space beneath in fabric dividers, and inside each parasol ‘room’ they added a small lamp, plant and a pair of earphones. When the office lights were switched off, people could now read without disturbing their neighbours; when the lights were on, they could create a sheltered and darker space, with an old plastic bottle and cable tie acting as a ‘doorbell’ to alert them to the presence of someone outside. It was a simple approach, creating a personal living space for everyone that was focused particularly on the sensory aspects that really improve everyday life in grim surroundings. The blankets dampened the communal noise outside. The parasol filtered the light above. The dividers created personal space. And the whole arrangement was created from cheap materials available from online outlets, builder’s merchants and garden stores.

The budget, of course, excluded the labour costs, but the whole process allowed refugees to get involved in production. Many had skills in sewing and construction; even more had time on their hands, and their involvement led to a huge amount of new ideas about how to create more secure divisions as well as variation in the different ‘rooms’ created. The cheap, synthetic blankets came in different colours that brightened up the atmosphere for everyone: they were hung like curtains by punching holes in the top and using cable ties as curtain rings; these in turn were suspended around the perimeter of the parasol fabric. The parasols were chosen for their heavy and stable base and the lamp for its flexible cord and clip-on feature. As the architects explained it to me, their role was to get things started – to give a little push, to make suggestions – but ultimately this was a process in which the inhabitants themselves created places and configurations that worked for them. Their aim was to create small spaces of comfort and control.

From the very beginning, the idea was to produce something deeply practical yet nevertheless humanistic and holistic. It was an approach shared in all three of the projects, which were brought together by the curator of the Austrian pavilion to illustrate how ‘architecture is a truly social discipline’. The second project, from EOOS, featured ‘social furniture’: a range of specially designed tables, chairs, desks and drawers that improved the homeliness of life in communal shelters. And the third project, from The Next Enterprise, created compact sleeping and studying units for office blocks. Each involved the careful consideration of how spaces are sensed – how they look and feel – combined with the imperative to act in emergencies. In the words of the curator, they showed that the moral act of sheltering people should not be divorced from a consideration about where stay and how it feels to live. They all stressed the importance of looking beyond bare biological needs, structural integrity and minimum standards. A humanitarian gesture, one of them said to me, has to provide truly human spaces to live.⁸

Conclusion

The Viennese projects illustrate how architecture might think beyond the provision of a complete house and the stereotypical ‘four walls and a roof’ in humanitarian emergencies. It shows how architects can work with humanitarians to improve shelters in an affordable manner, using their expertise to push beyond the standard set of approaches and often without building anything at all. Alejandro Aravena, the director of the 2016 Biennale, is famous for his idea of ‘half a house’, which promotes the idea that architecture can be incremental, allowing people to adapt and finish their own homes. Like the Caramel project in Vienna, this involves moving away from the fetishistic attachment to finished buildings, advocating ‘half houses’ instead, which are literally only half completed. They have a full gable roof, only half filled below, and if one looks at the house and draws an imaginary vertical line down from the apex of the roof, Aravena’s project is complete on only one side of this line. The ‘half house’ has been used for post-earthquake reconstruction in Chile, and it is based on a fundamentally humanitarian calculation: half a house costs half the money, which can help twice as many people. The central idea in this approach is that it is better to offer a longer-lasting, better-quality, incomplete house than a smaller, inferior complete one – not least because people will invest their own energies and fulfil their own desires when they get around to completing it.

Some critics have dismissed Aravena’s idea as providing substandard products to the poorest people in the world, accepting that ‘less is enough’ (Boano and Perucich, 2016). Yet the ‘half house’ is meant to prove something very different. It is not driven by the idea that ‘less is enough’ but rather by the conviction that it is possible to do far more with the same amount of money. It also draws attention to the importance of bottom-up rather than top-down visions, working with inhabitants to shape the final form of a living space, an important and long-standing idea in developing world housing. It is an idea neatly captured by John Turner’s maxim that ‘when dwellers are in control, their homes are better and cheaper than those built through government programs or large corporations’ (Turner and Fichter, 1972: 1).⁹

One could say that Aravena responded by halves and the Caramel project went all the way. Aravena stopped construction after half the building, and Caramel stopped construction completely. The circumstances, of course, were very different. One project was dealing with displacement after natural disaster, the other with refugee flows. One project responded to the need for new and more permanent homes, the other for temporary, emergency accommodation. The projects, however, reflect a similar idea: that shelter is not a finished house but a process that needs to involve beneficiaries and allow them to shape the final outcome. This in itself is not new, but the Caramel scheme also reminds us that shelter is more than just a physical space. It is a network of people and objects (Latour, 2005). It includes the things that populate a building as well as the external structure, and much can be done to design new objects and creatively adapt existing spaces. Architectural engagement in humanitarian circumstances, in other words, can involve deliberately avoiding construction of an actual shelter and intervening in a much subtler, small-scale manner instead. This focus on the objects that populate existing buildings has much to contribute when intervening in emergency conditions.

It is worth noting that the Caramel architects were responding to an emergency situation: a mass arrival of asylum seekers who were sleeping on the streets. Emergencies are often used to justify measure that would ordinarily be far from ideal. Homogenous mass-accommodation blocks or shelters, for example, promise a quick way to meet immediate needs. Yet these magic bullet solutions offer just an illusion of speed and success, and very often more modest solutions can work just as effectively. Most importantly, the Caramel intervention indicates that there is, in fact, time for process-driven forms of ‘architecture without building’ even in emergency conditions. One might argue that these are not scalable models for all circumstances, but they certainly indicate a kind of architectural engagement that might have lasting impact in humanitarian action, especially when there is a mass influx of refugees to urban environments – a situation that is increasingly common.

The distinction between architecture and building goes back to the work of John Ruskin (1871: 183), who wrote in the nineteenth century how ‘it is very necessary, at the outset of all inquiry, to distinguish carefully between Architecture and Building’. Ruskin meant, in a rather elitist manner, that architecture is florid and highly developed: more important than simple, serviceable building. This is not the argument I am advancing in this article. Pevsner (1945: 15) made a similar point: ‘A bicycle shed is a building,’ he wrote, ‘and Lincoln cathedral is a piece of architecture.’ Again, the idea was that architecture does more than simple building, which on its own is basic and functional. This is an elitist and narrow view of architecture that I absolutely do not endorse. In the 1960s, Bernard Rudofsky wrote the book Architecture without Architects (1964), which turned Ruskin and Pevsner’s judgement on its head. It argued that ‘there is much to learn from architecture before it became an expert’s art’. Rudofsky drew attention to vernacular buildings that were constructed by their own inhabitants rather than by architects (see also May, 2010). The idea of promoting architecture without building is, in many ways, consistent with Rudofsky’s philosophy, which was to see design as much less fixated on final structures and much more concerned with process. The real insight of the Viennese projects, however, is that this process can focus on furniture rather than large structures.

Rudofsky was an explicit inspiration for the Austrian pavilion and its title Places for People, which was a reference to an earlier Rudofsky book, Streets for People: A Primer for Americans. In that text, Rudofsky (1969) argued that urban design had been too focused on the needs of the automobile, rather than the needs of people. Streets, he argued, should have crucial social purposes, facilitating pleasure, enjoyment and conviviality, which are too often lost. Places for People makes a similar call. It asks that we move beyond the focus on buildings and structures to embrace the way that existing spaces can be made more amenable to social life. Another architectural reference point might be that earlier Venice Biennale directed by Aaron Betsky. This took place in 2008 and had the subtitle Architecture beyond Building. It stressed that the architect did not just construct buildings but invested in concepts, created artistic interventions and made theoretical and ideological statements, leading to what we might call, for want of a better phrase, the more impractical side of the profession. This is the side that generates such irritation among humanitarians, but as I hope to have indicated in this article, it can also be deeply practical not to build. In humanitarian situations, sometimes not building is exactly what is required, and architecture can succeed not by being more permanent and more developed, nor by being more conceptual and artistic, but by designing on a smaller, more modest scale.