The December 26, 2004 Indian Ocean tsunami caused by the Sumatra-Andaman earthquake was an unprecedented natural disaster that severely affected fourteen countries, including the island nation of Sri Lanka. Across Sri Lanka, the death toll surpassed 31,000, with 516,150 people displaced by the initial tidal wave or ensuing flooding.^1,2 The financial need for housing reconstruction was estimated at US$487 million, which was primarily contributed by international agencies mobilized after the disaster and distributed through owner-driven and donor-driven housing reconstruction programs.^3,4

The 2004 tsunami occurred after two decades of civil war between the Sri Lankan military and ethnic Tamil separatists, known as the Liberation Tigers of Tamil Elam (LTTE).⁵ From 1983 to 2009, ethnic strife segregated the island nation, whose population is majority Sinhalese buddhist and approximately one quarter Tamil.⁶ On the East Coast, the population has historically been majority Tamil and prior to the tsunami, uneven power relations between the two groups provoked violence, destruction, and human displacement to these rural areas.^7,8 During wartime, derelict roadways and infrastructure inhibited supply chains from reaching the coast, public services were inadequately funded and administered, and an estimated one million internally displaced persons (IDPs) were left without homes.^9,10 The Sri Lankan Civil War created an outsized vulnerability in the country’s East, where marginalized civilian populations inhabiting a degraded landscape faced a diminished capacity to provide food, security and shelter. Against the sectarian violence wrought legible across the geography of the nation, the 2004 tsunami landed on the East coast and inordinately affected the impoverished region, destroying 50,450 homes.¹¹ 

The natural disaster immediately superseded the Civil War in determining spatial distribution and usage along the eastern seaboard; it temporarily halted the armed conflict as the LTTE ceded political and spatial control to the Government of Sri Lanka, foreign donors, and non-governmental organizations, before the war ended in May 2009.¹² However, the geographic isolation of the East Coast has endured, particularly with regards to a lack of infrastructural improvements to extend supply chains from the capital city of Colombo and connect the coastline of the East Coast with the rest of the country.

The resilience of a socio-ecological system was seminally defined by the ecologist Brian Walker and his colleagues as “the capacity of a system to absorb disturbance and reorganize while undergoing change so as to still retain essentially the same function, structure, identity and feedbacks.” ²⁰ In the event of ecological, social, or economic conditions that degrade a system to the point where it cannot recover, a fundamentally new system must be created, or transformed to sustain human settlement.²¹ Because humanitarian aid’s most accessible and longitudinal contributions are spatial in nature, housing built after natural disasters disproportionately informs the social, economic, and ecological landscapes of vulnerable zones. At best, aid housing enhances physical and psychological security and helps rebuild localisms and place-based affinity, but when poorly planned and executed it can compound inequality and inaccessibility. 

Across cultures, the house acts as both a social center and form of identity, simultaneously material and symbolic while articulating identity and power.^22, 23  Seminal works, including Clare Cooper Marcus’ House as a Mirror of Self (1995), Carole Despres’ “The Meaning of Home” (1991) and Amos Rapoport’s House Form and Culture (1969), expand on the house as a complex entity defined by cultural, social, demographic and political factors beyond spatial composition. Compared to all other forms of dwelling, the house has a greater impact on the social, economic and political realities of individuals as an indicator of resource ownership.^24,25 

As such, it requires a greater amount of contextual sensitivity than any other physical infrastructure, both in architectural characteristics and construction through distribution.^26,27 

When international donors, consultants and organizations are involved in the construction of the home, the highly personal politics of housing collide with geopolitics. This can create reconstruction patterns which reflect, sustain and reproduce patterns of inequality and exploitation when there is unequal agency between donor and recipients.²⁸ Sri Lanka utilized two models in housing reconstruction: a donor-driven and homeowner-driven, or sometimes called cash-based or community-based approach.^29,30 In the donor-driven approach, housing reconstruction is entirely conceptualized, built and given to recipients by the donor agency, while the owner-driven approach offers funding, technical assistance, troubleshooting, training and supervision by the donor agency, while the recipient maintains full control over the physical housing reconstruction process, including the design of the home.^31,32 This is an increasingly popular methodology among the donor community, and was used extensively in the reconstruction of Sri Lankan homes after the 2004 tsunami, in which the International Federation of the Red Cross constructed 68% of their pledged homes through this approach, though the donor-driven approach was used by a majority of donors across the country.³³ 

After the tsunami, the lack of a national construction industry with the appropriate number of contractors, equipment, skilled workforce and modern management practices and access to finance was debilitating to reconstruction work.³⁴ Community-based development which trains the workforce and stimulates the regional economy, spurring trade, local markets and supply lines, provides manual labor and enhanced livelihoods after the immediate rebuilding process has completed.

If housing aid is an in-kind instead of cash gift, and attempts to introduce or define specific new models or constructions of home, the local agency of vulnerable places may be challenged and eroded by poorly-adapted global construction practices, materials, and conventions.³⁵ This erosion is striking in post-tsunami Sri Lanka through a peculiar iconography of aid projects: every aid-sponsored house, school, road, and bridge carries a tile or sign attributing the source of its funding, many bearing prominently the flags of Japan, Canada, or the United States (Fig. 1). In isolated communities that did not have roads until post-tsunami construction, these symbols are tangible reminders, over a decade after the event, of the brief intersection of rural Sri Lanka with international aid. In the present day, there is no accountability, either from the organizations to beneficiaries or inhabitants to donors, regarding ownership or occupation of these dwellings. Furthermore, external reviews found the Government of Sri Lanka relaxed the stipulations and requirements of donor-driven housing as compared to owner-driven housing for international NGOs completing reconstruction work.³⁶

Traditionally, beneficiaries are formally accountable to the organizations who provide aid funds, while attempts to create standards of accountability from organizations to beneficiaries beyond roughly quantifiable statistics are overlooked.³⁷ The metrics of accountability - numbers of dwellings constructed, miles of road laid, fiscal commitment to projects - mask the nuances that could be provided through qualitative and longitudinal evaluation, particularly in the form of surveyed feedback from users. This is particularly crucial in housing reconstruction projects, in which buildings can become empty within months of their completion. Qualitative data from users is crucial in extending the lifespan of these projects, enacting cultural preferences in spatial layout and organization and generating knowledge of best practices to carry forward. 

The landscape of Sri Lanka’s East coast is largely homogenous. Long swaths of the coast are sandy beaches with steep, three-foot vertical drops in sand that denote the high and low-tide markers. Fishing vessels and mariculture activities are primarily launched from and return to sheltered harbors with sandy bottoms that offer protection from the fierce waves and near-shore coral reefs elsewhere along the eastern coast. Chains of shore-parallel, elongated lagoons form a barrier between the barren sandy beaches and inland rice paddies, and serve as biological hotspots for migrating birds, reptiles and larger mammals, including elephants.⁴⁴ Both the lagoons and surrounding rice paddies occasionally become 

hyper-saline, particularly along the southeastern coast, disrupting the fishermen and farmers whose livelihoods depend on the freshwater inflows.⁴⁵ The tsunami, which penetrated up to two kilometers inland across the east coast, extended to and salinated rice paddies, many of which lay fallow during the Sri Lankan Civil War. Low-lying scrub vegetation dominates the landscape, which lies in the country’s dry zone. Shrubs and trees are characterized by thick bark and interlacing canopies for water retention. The vertical landscape is dominated by coconut palms, particularly around the coastal freshwater lagoons. Additional fruit and vegetable species adapted for dry climates, including pineapple, banana, chilies, maize and mung bean can be successfully grown across the east coast, and have traditionally been grown in sustenance gardens adjacent to homes to supplement diets.⁴⁶

The interpretive fieldwork for this research was undertaken over the course of one month from May to June 2017 in Pottuvil, a town of roughly thirty-five thousand residents on Sri Lanka’s eastern coast (Fig. 2). Here we investigated three local projects emblematic of different delivery systems for architectural and urban design principles in disaster relief housing assistance. Pottuvil was severely afflicted by the tsunami and due to its size, it was a site in which NGOs constructed homes in close proximity to owner-driven and informal reconstruction efforts using drastically different models, and therefore proved a productive site for comparative study. 

The research was conducted through informal, semi-structured interviews and three site visits to the developments investigated in this paper, as well as sketching, photography and observations of the three housing typologies. Interviews were conducted based on accessibility to target populations, namely five fishermen within each housing development, each of whom acted as the head of household within their family. The sites for case studies were based on the accessibility of information regarding the schemes and the knowledge of an NGO worker who facilitated translation and had familiarity with the projects. Interviews were used to collect qualitative information on living arrangement, habitation and improvements within each housing scheme. In addition, expert interviews were conducted with NGO leaders by phone, primarily regarding the lessons learned through post-tsunami housing reconstruction and applied to current building projects. 

The interviews were conducted with a Tamil translator working with a local NGO, who also facilitated photography of homes and other sites constructed after the tsunami. Thus, the methods used were impacted by the language barrier and necessity to translate questions from English to Tamil. It should be noted that inhabitants of the East coast, and in some cases within single housing schemes, sometimes speak only either Tamil or Sinhalese, leading to situations in which neighbors and community members cannot communicate in a common language. These barriers to communication are heightened by the aforementioned history of strong cultural conflicts between Tamils and Sinhalese which prevented interviews from being undertaken with both populations over the course of individual site visits. 

The Ecoshell system, developed by the Monolithic Dome Institute, a commercial organization based in Texas, has been deployed by various organizations as aid housing across the world. Monolithic’s domed houses - reinforced concrete erected using inflatable formwork - are emblematic of a type of formally and technologically distinctive bespoke architectural prototype given as donor-driven in-kind aid. The town of Inspector Eatham, on an elevated plain on the northeastern edge of Pottuvil, is the site of one such relief housing development in Sri Lanka constructed by the Solid House Foundation (SHF), a Dutch NGO. Following the tsunami, some 60 domed ecoshell homes were built by SHF within an existing fabric of agricultural property ownership and small homes and developments (Fig. 3).

Construction of the concrete shells begins with the excavation of a circular foundation, into which a reinforced concrete slab or ring beam is poured. When initial curing is complete, inflatable formwork using air pumps is inflated on top, and a cage of rebar is erected around it. Finishing layers of concrete are either troweled onto the formwork or propelled as shotcrete. Community meetings were held and a common prototype, 9 m [29.5 ft.] in diameter, was drawn by SHF to guide the construction process (Fig. 4).⁴⁷

Palmyra palm tree leaves - which have a rigid central spine from which long, lancelet leaves extend from either side - woven into mats, termed cadjan in the local vernacular, have been used as a building material across Sri Lanka and greater Asia since antiquity, particularly along the poorer east coast, where geographic isolation and poverty demands the use of easily obtainable, repairable building materials.⁵⁰ Cadjan huts near the coastline in Pottuvil are generally occupied by fishermen, whose boats can be pulled up the beach next to their homes, providing immediate access to the sea and secure storage for vessels. The rectangular, single-story structures are generally entered only from the side opposite seaward breezes, and the space inside is partitioned by divisions made of additional woven cadjan mats (Fig. 6). Their structural integrity decays over one to three years, after which they must be replaced with additional leaves, which can be sustainably sourced from the area. There is generally no flooring constructed, particularly in fishermen’s dwellings built directly onto the sandy beach.

Through interviews and observation, it became clear that the erection of cadjan buildings is generally undertaken by communal groups rather than solely the inhabitants of a house. The roof of a traditional home can be assembled using cadjan by three to four able-bodied men in two days. Generally, the main structure is erected by community members, while secondary considerations not required for shelter, including fencing, are added at the discretion of the family. In times of heavy winds, the buildings must be weighted down unless a framework of timber is secured to a concrete foundation, which has not been common practice along the beach in Pottuvil. Cadjan buildings are especially vulnerable to flooding and require barriers of sandbags, vegetation, or walls to keep water from reaching the structures. These houses offer limited physical security, and rely on the social fabric of the community to ensure privacy, which has historically been tenuous along the east coast.

Previously in Sri Lanka, the Tsunami Evaluation Coalition found serious problems in housing relief stemming from supply-driven approaches, inappropriate housing designs and an over-reliance on the donor-driven process.⁵¹ Introduced by international NGOs, the housing typologies we investigated were pre-established generic prototypes that have specific advantages in durability and standardization under idealized circumstances. However, localized site-specific concerns were largely ignored in favor of schemes that position houses as isolated single-family assets, in contrast with the strong history of flexible family living in Sri Lanka.⁵² Furthermore, these aid projects stand rooted in foreign urban models that neglect the realities of programmatic integration familiar to Sri Lankans, specifically the integration of residential and commercial space. In areas surrounding these housing projects residences commonly have a permeable first floor with retail space, while families live in the back of the building. This space is generally used to sell cold water, petrol and pre-packaged foods.

Construction methods and materials were imported and not readily sourced from the eastern coast, where undeveloped infrastructure was degraded during the war, and which remains difficult to reach. The relative isolation of the east coast made the establishment of supply lines more difficult than in other areas of the island, leading to increasing costs for materials as demand increased with consistently limited supply. Demanding technical oversight or capital influx for maintenance and repairs by international donors or NGOs, these housing typologies indulge in the rhetoric of participatory and inclusive community projects while making complete handover to the community untenable. Ultimately, physically well-constructed housing that does not anticipate cultural, economic, and environmental needs can become obsolete artificially quickly, long before weathering or future natural disasters might pose problems.

It is critical that occupants are able to alter their dwellings to adapt to their changing needs. In post-disaster situations, it is paramount that housing constructed immediately in the aftermath of a natural disaster can be altered and reconfigured to the requirements of families after they have recovered and re-established their livelihoods. We suggest specific reasons why inhabitants do not easily physically adapt to the models we investigated: site planning and architectural characteristics. The site strategy of the JICA project (see Figs. 5, 8) creates small, fenced lots of approximately 230 m² [2,476 sq. ft.]. While the house model built from a UN prototype uses standardized construction methods familiar to local builders, the small lots and serial arrangement are reminiscent of conservative western single-family residential developments, on which it would be difficult to significantly extend or modify a home, and impossible to construct a second home for non-nuclear family.⁵⁶ 

Local authorities who would normally approve sites, infrastructure and accompanying public services were largely backlogged and bypassed in the aftermath of the tsunami, leading to poorly planned, paved streets for vehicular access that restrict possibilities to convert and repurpose outdoor spaces. While SHF’s domes fit more readily into larger existing plots, with ample space for extension, their unfamiliar geometry may have proved a barrier to meaningful modification (see Fig. 7). The beneficiaries themselves were critical of the one-size-fits-all approach that ignored household size, age and disabilities, therefore requiring owner-built modifications and extensions after donor funding had been spent by international organizations, not the inhabitants themselves.⁵⁷ Without channels for input during the design process, any modifications or alterations that happen after construction are completed when families have the economic means to alter their homes. Yet the economy of the East Coast is isolated from the rest of the nation, making economic recovery challenging beyond the initial influx of aid in the direct aftermath of the tsunami. 

Questions of cultural and livelihood appropriateness were mostly specifically articulated in reference to kitchen design. Throughout the tsunami-affected area, many homes reconstructed by foreign donors were equipped with ventilation appropriate for gas-burning stoves, not the traditional biomass-burning models. This design was undertaken in consideration of the deforestation of the East coast and depletion of biomass sources, as well as the negative health impacts of burning biomass for cooking. However, gas is unaffordable for the majority of families along the East coast, 75% of whom live beneath the poverty line.⁶⁴ Recipients of these homes added ventilated attachments to their home built of cadjan mats, or continue to cook using biomass stoves in improperly ventilated areas. In addition, in homes constructed with only two rooms and without a chimney, there is no adequate place to designate as a kitchen (Fig. 9).

Stranded assets lose economic value well ahead of their anticipated useful life due to changes in market forces, societal norms, or environmental shocks.⁷⁰ Aid housing projects risk becoming this kind of liability when useful social, environmental, and economic context is ignored in favor of universal prototypes. In the case of Sri Lanka’s eastern coast, residents elected to vacate their allocated homes in settlements superficially modeled on international urban or architectural schemes. Locals identified many plots and homes within both the SHF and JICA developments that stood abandoned, and noted that many families owned multiple vacant properties but inhabited only one. Although expensive, structurally robust aid housing built with foreign design and construction techniques satisfies an immediate demand for safety and security, housing in the form of prescriptive gifts-in-kind undermines the ability of local families to respond, adapt, or migrate when facing contingencies in their environment and economy, such as flooding that may inundate or salinate nearby farmland, or urban reconfiguration that may result from shifting shoreline topography.

The occupants of peripheral rural areas and humanitarian benefactors must make many difficult choices in the face of increasing environmental volatility. In the future, there will likely be less aid assistance available to those living on the coast outside major metropolitan areas, particularly in impoverished nations. This necessitates a difficult shift in the nature of aid from the typological towards the systemic - from the long-term expensive prototype to a low-cost, high-turnover, resilient culture of building. 

Solutions are found at the nexus of the spatial and social. Frameworks that provide technical specifications in a common language are useful only when paired with technical training that allows builders and carpenters to mobilize locally, offering agency to communities upended by natural disaster. Critically, during the reconstruction process the National Housing Development Authority provided a manual based on their previous successful projects in coastal Sri Lanka, Guidelines for Housing Development in Sri Lanka (2005), which was largely ignored by both the Government of Sri Lanka and foreign organizations. The best practices guide provides a framework for community engagement, technical specifications, and climatically appropriate design.⁷² To bridge the gap between local, tested expertise and foreign skills and knowledge bases, “train the trainer” seminars offered continuously to at-risk communities, sponsored and overseen by foreign NGOs but taught by local designers and organizations might promote experimentation with traditional and imported forms, and increase both physical and social resilience. 

A part of owner-driven housing schemes elaborated on by the United Nations Habitat Program, training seminars that teach community members building and technical skills work by harnessing the capabilities of the community before a disaster, creating a network of community members capable of proactively sharing their skill set. The Indian Housing Program sponsored by UN Habitat in the northern districts of the island capitalized on this knowledge transfer, and offered specialized training from architects and designers to homeowners, including troubleshooting during the construction process.⁷³ On the ground training can work in concert with technical documentation provided by international donors and NGOs, allowing for housing prototypes to be offered while homeowners possess the information to modify them to be culturally and familiarly appropriate. This kind of training can appropriate foreign knowledge and integrate it within a community, as a longer timescale for knowledge transfer ensures that trainers can modify knowledge into best practices that integrate with existing ones. 

A return to primitivism is neither required nor suggested. Instead, support for localized knowledge that builds agency and the transfer of knowledge and hard skills, not rigid typologies, offer remote societies the opportunity to propel forward even when they return to geographic isolation. Best practices that treat the community as an asset-base for development and that do not rely on a constant influx of foreign aid help to promote longevity and sustainability of projects, but treating the construction of dwellings as an objective with a clear end ignores the realities of social life and civil and municipality works that are critical for a functioning, robust, healthy society. Specifically included within this framework is proper construction of sewage treatment, roadways, and common places, which are often neglected by international NGOs or thought of as secondary. On Sri Lanka’s eastern coast, mitigating ethnic tensions and the sources of decades of civil war is a critical priority, and should be seen as such, not as an afterthought to physical reconstruction in the wake of a natural disaster. Unresolved ethnic tensions and inequity are still legible across the country, particularly in legislation on the East Coast, and psychological barriers within communities segregate physical assets, such as the JICA housing compound. In light of the disproportionate poverty of the coast compared to predominantly Sinhalese areas, conflicts that arise from unequal distribution (or aid perceived as such) might be managed better by frameworks for distributed, bottom-up knowledge.

Limitations of this study included the language barrier between English and Tamil, requiring a translator who expressed a clear bias for the Sinhalese population, and restricted access to Tamil inhabitants of the typologies investigations. Data from the Government of Sri Lanka was accessed online, and there are clear gaps in information accessible relevant to the East Coast, which were supplemented, when available, with information from the LTTE, which must be considered biased due to the history of Civil War within the country. 

There is a clear need for continuing research on disaster relief housing in the context of global climate change and projected increases in resource scarcity and growing economic inequality. Future research that studies differences in vernacular architecture, between the Tamil and Sinhalese tradition could greatly contribute to best practices in culturally appropriate design, and more importantly create a set of best practices and construction standards to follow in the production of housing. Research into “Vaastu shasthra,” a traditional Tamil set of beliefs regarding land use and planning, is very limited, and must be used to further best practices for housing that will not be abandoned due to ethnic or religious reasons. Further research that quantifies the number of abandoned houses provided by in-kind, donor-driven aid across tsunami-affected areas in Sri Lanka will be useful in determining the role of ethnic conflict in housing allocation, and suggest where design interventions can be most impactful.