Image Image Image Image Image Image Image Image Image Image

Bartlett School of Architecture, UCL

Scroll to top


No Comments

Democratizing the production of architecture

Democratizing the production of architecture

Architecture is a pervasive presence in our lives. In fact, buildings are so ubiquitous that the general terms for position in English, ‘inside’ and ‘outside’, place us in constant reference to which side of architecture we are located (Zizek 2010), and shelter is considered a universal and basic human need (UN 1954; Maslow 1954). However, the production of architecture has undergone radical change. Whereas vernacular, or ‘people-built’, architecture used to be the norm (Isaacs 1974; Parvin 2015), today there are strict regulations that determine who can legally build architecture. These norms for architectural design ‘harden’ the design process insofar as the production of architecture is complicated by building codes, laws, and professional status. If people cannot afford to legally obtain shelter, they often resort to building their own homes. Such buildings are called slums, labelled as informal and illegal, and face a threat of being destroyed (Mitlin and Satterthwaite 2004). If one can afford legally built architecture, it is often required that one hires a professional designer. In many cases, this involves architects.


Fig 1: An ‘informal' housing settlement in Rio de Janeiro, Brazil, shows how a large portion of the global population interacts with archtiecture (Davis 2007).

Fig 1: An ‘informal’ housing settlement in Rio de Janeiro, Brazil, shows how a large portion of the global population interacts with archtiecture (Davis 2007).


In this context, architectural services are often only financially possible for wealthy people, governments, and corporations.1 As a result, architectural services are disproportionately influenced by the agenda of the wealthy and powerful, despite the overwhelming need for housing for poor people. In response to this disjunction, this paper will explore how the production of architecture might be softened through democratization of the design process, especially through automated technology, in order to make housing more accessible. After defining key terms, it will examine open-source design, automation, and the social implications of softer architectural processes.

Key Terms

Before proceeding, it is necessary to establish a few key definitions. First, architecture will refer to inhabitable structures that are a ‘means, presupposing a purpose external to them’ (Hegel 1973). Although the use of the term has expanded to include a wide variety of concepts, from computer code structures to virtual spaces, this discussion is centered on housing and public buildings and will use ‘architecture’ to refer to physical structures that people inhabit. Consequently, ‘architectural’ will be used in its conventional definition to describe things which relate ‘to the art or practice of designing and constructing buildings’ (Oxford 2015). In contrast, ‘architectonic’ will refer to things which have architectural qualities without necessarily pertaining directly to architecture (Curl 2006). While each of the terms warrants consideration and debate, to do so is outside the scope of this work.2 Secondly, slums will be referred to as ‘informal housing’ not because they ‘have no form, but because they exist outside the legal and economic protocols that shape the formal city’ (McGuirk 2015, 25).

Finally, ‘democratization’ will be understood as the genuine involvement of people within processes that impact their lives, borrowing from Dahl’s fundamental understanding of political democracy as dependent on ‘universal suffrage’ (Dahl 1998). In this context, there is a clear distinction between democratizing access to design products and design processes, which this paper will explore in detail.

Open-source design

In order to narrow the gap between the demand for architecture and access to its production, architects and designers have made efforts to (re)democratize housing design and production. An early example is Isaac’s (1974) book, How to build your own living structures, which contained simple plans for affordable, eco-conscious architecture. His intent was to not only help people produce architecture, but to ‘increase sensitivity’ to tools so that people would eventually ‘surprise and amaze’ themselves by making their own designs and creations. His designs were intentionally modular, or on a ‘matrix’, so that people could expand according to their needs and intuition. The impact of his work is relatively unkown, perhaps because his teachings were linked with a radical, countercultural lifestyle change. Goals of architectural autonomy, however, continue to be explored by others.

Fig 2: Isaacs sitting in his ‘Superchair', a furniture design from his open-source initiative (Isaacs 1967).

Fig 2: Isaacs sitting in his ‘Superchair’, a furniture design from his open-source initiative (Isaacs 1967).


Today, Alaistar Parvin’s company WikiHouse uses an online database of crowd-sourced architectural designs for homes that can be mostly fabricated Computer Numerical Control (CNC) router and plywood, with the exception of some off-the-shelf products like vapor barriers, windows, and doors (WikiHouse 2015). The designs can be constructed by a small group of people with limited construction experience. The guiding principle behind WikiHouse is that architecture should be designed by ‘the people’, for ‘the people’, and is motivated by a rejection of commercial, large-scale housing development strategies (Parvin 2015). Though it is a step toward democratizing the design process, it is still restricted by highly technical barriers. To own and operate a CNC router requires a large amount of capital and technical knowledge. Similarly, designing a WikiHouse requires design skills, proficiency with 3D modeling software, and, ultimately, an architect to ensure it meets legal standards. As a result, many of the available homes are designed by architects or aspiring designers. In other words, the average person is as involved in designing a WikiHouse as they are in designing Ikea furniture. Additionally, the ease of downloading a digital model could produce issues of architecture that is not appropriate for its context. For example, a person may choose a Scandinavian design based on aesthetic preferences, but end up with homes that are thermally inappropriate for rural Argentina. Still, this issue could be easily resolved by including educational materials about appropriate technical needs for designs.

Fig 3: Diagram showing the WikiHouse design and fabrication process (WikiHouse 2013).

Fig 3: Diagram showing the WikiHouse design and fabrication process (WikiHouse 2013).


Another organization which aimed to improve access to professionally-designed architecture was Architecture for Humanity (AH), which utilized volunteers, pro bono architectural services, and open competitions to produce building designs for people in developing countries. While the efforts of the organization likely helped their clients, AH expanded access to architectural products, not architectural processes. As such, the organizations impact was limited to the amount of projects they could fund and manage, and arguably made little contribution toward democratizing design. Unfortunately, mismanagement resulted in the organization declaring bankruptcy in January of 2015 (King 2015). The challenge of providing free or low-cost professional architectural services emphasizes a limitation of both wikiHouse and AH: namely, that they rely on professional services. However, that architects necessarily be professionals –or people– is not an uncontested assumption.

Given the constraints of what can be built legally, one solution to the need for housing is to expand access to professional design services. As illustrated by the previous two examples, the potential impact of this strategy is bottlenecked by access to low-cost professional services or the funds to purchase them. Another, more promising approach is to enable non-professionals to produce legally and structurally-sound architecture. One method for achieving this is to automate crucial moments of the design process with computers, such as structural requirements or mechanical components. Around the same time that Isaacs was producing his instructions for DIY eco-homes, Negroponte proposed that computer programs could act as autonomous and interactive participants in the design process (1975). Drawing on cyberneticist Pask’s conversation theory (see Pask and Boyd 1987; Negroponte 1975, 7—31), Negroponte suggested that architecture could be produced through person-computer dialogues. Negropante admits that his group worked ‘very sporadically and without much success’ on automated design, but the ideology that guided their research was explicit: to demystify the design process and empower citizens to configure their living spaces (1975, 102—123). Researchers at MIT advanced this concept, designing an automated program that generates architectural designs according to user-defined parameters (Merrell, Schkufza, and Koltun 2010). The significance of such a tool is global: millions of people that would otherwise never interact with architectural drawings could produce customized homes to US building standards by simply interacting with a computer program for a few minutes. This computer ‘architect’ could be duplicated at no cost.

Fig 5: An image from Negroponte's (1975) book showing a computer interpretation of a hand drawn sketch in an attempt to automate the design process.

Fig 5: An image from Negroponte’s (1975) book showing a computer interpretation of a hand drawn sketch in an attempt to automate the design process.

Fig 6: Examples of computer generated houses from the Merrell et al. (2010) program made at MIT.

Fig 6: Examples of computer generated houses from the Merrell et al. (2010) program made at MIT.


The slew of computer-generated homes is not an inspiring collection of design; all of the houses utilize conventional styles from American suburbs, with large garages and a devotion to pitched roofs that even results in triangular windows (see the grey house at the front left of Fig 6). The tool is effective if one is content with that particular style, but could be frustrating for people with different tastes or more creative ambitions. Still, to the average citizen, it offers an otherwise unattainable possibility: the ability to fully customize his or her home. This is not significant because it makes a luxury more accessible; it offers buyers an alternative to ‘[purchasing] a standard, generic house produced by a speculative developer’ and enriches urban fabrics by avoiding ‘impersonal mass-housing’ (Larson et al. 2004, 1). In doing so it addresses Parvin’s dissatisfaction with the housing industry without relying on donated professional services or technical fluency.

Technology as a softening agent

It is possible that technology will play a role in softening architectural processes. The power of technology, especially networked technology, is that it can record and distribute large amounts of data quickly. The ‘internet of things’ movement describes a world where every object in our life is hardwired into a network, enabling us to ‘track and count everything, and greatly reduce waste, loss and cost’ (Ashton 2009). As technology becomes smaller and cheaper, this agenda is progressing–billboards photograph and categorize us by demographics, thermostats keep track of when we feel cold, and cell phones constantly update our locations (Greenfield 2015). Greenfield (2013) argues, however, that this knowledge is primarily serves systems management and private corporate interests, despite the fact that the data is generated by the public in public spaces. He claims that the ‘smart city’ concentrates knowledge and power for a few ‘administrators’, limiting our freedom rather than expanding it (Greenfield 2015). His vision for technology in urban contexts is entails integrated data and communication systems that are distributed and controlled publically.

In line with Greenfield’s claims, many videos explaining the ‘Internet of Things’ are produced by major business corporations, such as McKinsey & Company, Forbes, or IBM, shown in the video above.


Greenfield’s concerns about privatization are illustrated by Usman Haque’s Pachube software. Pachube was originally built as an open-source internet of things platform that Haque envisioned being ‘a little like YouTube’ for sharing ‘real time environmental data’ (Haque 2011). Pachube was acquired by a private company in 2011, just 3 years after it was founded (ibid.) Tech bloggers that had followed the project were ‘surprised that Pachube sold itself so early’, especially because it ‘aimed to be the leading open development platform for the internet of things’ (MacManus 2011). Regardless of whether one believes the $15 million acquisition has sinister or benign implications, the fact remains that the software is no longer freely available or customizable, and is less accessible for those without capital.

Even in Greenfield’s ideal vision of publically-managed big data, there are issues with controlling large, dynamic, and real-time flows of data. The first is apathy: people already have limited time, and inviting people to manage raw data would simply not have an impact. Similar arguments are made against direct-democracy (general votes on all political issues, including the passing of laws), suggesting that it ‘hopelessly overburdens voters’ or effectively excludes them by rewarding bureaucrats that ‘spend the money…[to] achieve considerable “monopoly” power’ (Frey 1994, 338; Romer and Rosenthal 1979, 563). While making big data more accessible arguably increases access for citizens who have the time and interest to use it, it could run the risk of providing private corporate interests even more access to information about public behaviors.

The second issue faced by public data is technical; namely, that the data sets are so large that they are incomprehensible. Glanville (2000) explains, however, that the ‘unmanageable’ nature of this data can actually be a creative opportunity, allowing for the unexpected and unpredictable to emerge from uncontrollable entities. Still, manipulating large data sets requires a vetting process, where pertinent information is condensed so operators can make informed decisions (Glanville 2000). The nature of how information is condensed determines who controls and benefits from technology. Like Haque’s Pachube, architecture-generating programs could either drastically expand access to design in an open-source system, or allow companies to undercut market rates for architectural services while maintaining limited access if it is owned privately.


This review explored the aspirations and implications of open-source design, from the analogue eco-building project by Isaacs to highly technical, automated software by MIT. Several organizations that aim to provide open-source architecture were critiqued according to the level which they actually democratized the design process. AH failed in its emphasis on linking people with professionals, whereas WikiHouse still relies on a high level of technical skill. Next, this paper examined the general trend of technology toward collecting big data, as illustrated by the internet of things and Greenfield’s research on smart cities. This trend raises questions regarding ownership of public information, and relateds to the potential distribution, use, and capturing of architectural-automation technology.

As it stands, there is still gap between Isaacs, Negroponte, and Parvin’s visions for open-source architecture and actual building processes. Specifically, open-source architecture faces two barriers that prevent it from meeting the demand for housing: the technical barrier of physically building structures (such as the ability to access and operate tools like a CNC router or purchase land and materials), and the information barrier the prevents people from genuinely interacting with the design process (both in the legal requirements —which still ultimately require an architect to navigate–and the technical knowledge of operating design software). Organizations that try to bridge this gap by making professional services more accessible, such as WikiHouse and Architecture for Humanity, struggle with funding.

A video by The Architecture Foundation showing the self-built community in London inspired by architect Walter Segal. Link to video here.


A more promising solution is to include the public as designers instead of as pro bono clients. In order to truly ‘soften’ the architectural process, research must focus on the democratization of the design process and of the production of architecture. MIT’s automated design program is a step in this direction, but perhaps relies too much on stylistic conventions that hamper creativity. Using design as a methodology, this paper will explore the possibility of creating an open-source architectural design tool that genuinely involves the public as designers while maintaining the creative dimension of the process. Research will focus on how democratization, itself a complex term, translates into physical design. It will also examine role that technology and automation will play in softening design processes, as well as strategies for controlling ‘unmanageable’ data.

List of Figures:

Fig. 1 Davis, Crystal. 2007. Slum in Rio de Janeiro, Brazil. Image. Accessed 3 Dec 2015.

Fig. 2 Isaacs, Ken. 1967. Image of ‘Superchair’. Accessed 3 Dec 2015.

Fig. 3: WikiHouse. 2013. Diagram of WikiHouse Process. Image. Accessed 3 Dec 2015.

Fig. 4: Carron, Margaux. 2014. Image of WikiHouse prototype. Accessed 3 Dec 2015.

Fig. 5: Negroponte, Nicholas. 1975. Soft Architecture Machines. Boston: MIT Press. Page 120.

Fig. 6: Merrell et al. 2010. Computer-Generated Residential Building Layouts. Accessed 3 Dec 2015. Stott.

Key References:

Greenfield, Adam. 2013. Against the Smart City. New York City: Do projects.

–––. 2015. “Adam Greenfield.” The Bartlett International Lectures. London: The Bartlett School of the Built Environment.

Negroponte, Nicholas. 1975. Soft Architecture Machines. Boston: MIT Press.

WikiHouse. 2015. “WikiHouse.” WikiHouse Foundation.

Works Cited:

Ashton, Kevin. 2009. “That ‘Internet of Things’ Thing.” RFID Journal, June.

Curl, James Stevens. 2006. “A Dictionary of Architecture and Landscape Architecture.” Oxford. Oxford University Press.

Dahl, Robert. 1998. On Democracy. New Haven: Yale University Press.

Frey, Bruno S. 1994. “Direct Democracy: Politico-Economic Lessons from Swiss Experience.” American Economic Review 84 (2): 338—42.

Glanville, Ranulph. 2000. “THE VALUE OF BEING UNMANAGEABLE-VARIETY AND CREATIVITY IN CYBERSPACE.” Netzwerke. Vienna: Falter Verlag.

Haque, Usman. 2011. “Pachube.” Haque Design+Research.

Hegel. 1973. Hege’s Lectures on Fine Art. Edited by T M Knox.

Isaacs, Ken. 1974. How to Build Your Own Living Structures. New York: Harmony Books.

King, John. 2015. “Architecture for Humanity Shut; Nonprofit Helped Disaster Victims.” SF Gate, January 17.

Larson, K, S Intille, T J McLeish, J Beaudin, and R E Williams. 2004. “Open Source Building: Reinventing Places of Living.” BT Technology Journal 22 (4): 187—200. doi:10.1023/B:BTTJ.0000047598.66712.e0.

MacManus, Richard. 2011. “Pachube Acquired: Why Did It Sell so Early?” Readwrite, July 20.

Maslow, Abraham. 1954. Motivation and Personality. New York: Harper.

McGuirk, Justin. 2015. Radical Cities. London: Verso.

Merrell, Paul, Eric Schkufza, and Vladlen Koltun. 2010. “Computer-Generated Residential Building Layouts.” ACM SIGGRAPH Asia 2010 Papers on – SIGGRAPH ASIA ’10 29 (6): 1. doi:10.1145/1866158.1866203.

Mitlin, D., and D. Satterthwaite. 2004. Empowering Squatter Citizen. London: Earthscan.

Oxford. 2015. “Oxford English Dictionary Online.” Edited by Angus Stevenson, Julia Elliott, and Richard Jones. Oxford English Dictionary. OED Online. Oxford University Press.

Parvin, Alastair. 2015. “An Open Manifesto for Housing.”

Pask, Gordon, and Gary Boyd. 1987. “CONVERSATION THEORY AS BASIS FOR INSTRUCTIONAL DESIGN.” In Interactive Media: Working Methods and Practical Applications, edited by D Laurillard, 97—115. Chichester, England: Ellis Horwood.

Romer, Thomas, and Howard Rosenthal. 1979. “Bureaucrats Versus Voters: ON the Politifcal Economy of Resource Allocation by Direct Democracy.” The QUarterly Journal of Economics 93 (4): 563—67. doi:10.1080/02724980343000242.

UN. 1954. The Universal Declaration of Human Rights.

WHO. 2014. “Global Health Observatory Data.”

Zizek, Slojav. 2010. “Architecture and Aesthetics.” Navarra: Fundacion Arquitectura y Sociedad.

Submit a Comment