top of page


  • Dr Sechaba Maape

The Shelter


For the past 20 or so years I have been in search of what a contemporary African architecture could be. This interest was first sparked in my second year of architecture studies at the Wits School of Architecture and Planning where I was taught by architect Peter Rich. He made me appreciate that there is indeed value in the architecture of African people. His inspiration came from Pancho Guedes and Ndebele architecture. I then decided to pursue this further in a masters’ thesis in 2011 where I started looking at my home in the Northern Cape as an inspiration for an African approach. This was the first time I took a serious look at the spaces my ancestors lived in.

This then led to a PhD study which was not specifically searching for an African architecture, but nevertheless exposed me to ideas and spaces that gave me invaluable insights. These especially came from studying the caves in Kuruman that were inhabited as early as the time our modern human ancestors came to exist about 105 000 years ago, who it turns out occupied these caves and shelters. After completing the PhD study in 2017 I had amassed an enormous amount of information regarding the habitation life of South Africans long before any colonial settlers arrived there. At the same time, I was designing and building two buildings in Kuruman, both of which being off grid buildings made of waste material. I began to develop an architectural language; however, it was only in 2019 that I began to put the pieces together, practice and theory, to develop a framework for an architecture that is inspired by Africa, and more specifically my hometown Kuruman.

Inspired by the caves, rock shelters and homesteads of Africa

During my research, particularly when I began the Phd study, I started hanging out with Dr Heinrich Kammayer. I had a discussion with Heinrich about the Wonderwerk Cave near Kuruman, and how people would have possibly inhabited this space. Of course, this was a speculation, but Heinrich mentioned that as far as he is concerned, people particularly the late Holocene occupants of the shelter would have likely lived at the mouth of the cave because the back was too dark and not conducive for living. This is of course what is reflected in the archaeology. He said that he suspects people would have set up their small grass dwellings in the front portion of the cave and possibly outside, treating the cave more like a rock shelter. This again has been argued by archaeologists elsewhere. It was the first time I recognised the way in which the shelter, as in the rock shelter and cave created a place for human habitation. I later applied this idea in relation to the Ga-Mohana rock shelter also near Kuruman. This shelter is very large and would have possibly been occupied by large families, or at least had the potential to do so.

The evolution of the African shelter: Drawings by Dr Sechaba Maape

I began looking at the way Khoe and San people in the area, particularly the Korana, built their dwellings. These were mainly beehive style grass domes. These domes can easily survive out in the open, but what occurred to me is that when under a shelter like Ga-Mohana, the hut would do far less work, particularly in relation to wind breaking as well as rain. In a painting done by colonial explorer Samuel Daniell, one can see the grass huts or arches of the Korana who lived in the Kuruman area. In some cases, you can see wind breaks at the back of the hut.

One can imagine that when staying under the shelter of Ga-Mohana or Wonderwerk Cave, one would no longer need the windbreak, or even very tight weaving for waterproofing. Another important factor about both these shelters is that at some point in time, and even so today during rainy seasons, the shelters themselves would have streams of water running down the brow. Almost in a kind of paradise fashion, water would flow from what my community call the breast (tufas) of the shelter. In addition, water would be available at nearby water holes. Both the shelter and the cave are near waterholes, the Kuruman eye in the case of

Painting of Korana by Orange River: Source

Ga-Mohana, and Bushman’s hole in the case of Wonderwerk. In this sense, these two locations provided the ideal human habitat.

Later in my research I began looking at the architecture of my culture, Tswana architecture, which was also a dominant architectural feature of the Kuruman area and has been studied by people like Franco Frescura. A powerful insight came for me when reading one of Frescura’s papers about Tswana architecture where he also refers to a painting by Daniell. In the painting one sees the dome hut similar to that of the Korana, the two having had contact and even intermarrying and sharing some elements of language, but rather than the dome being under the shelter of a cave or a rock shelter, my people placed their dome under a human built shelter. This was a cone on posts, similar to the cone on cylinder that we all know. The image that Frescura refers to is a painting of the Batlhaping community of the 1800’s who are still currently the ruling tribal authority in Kuruman. Moreover, the image shows what could be a homestead, within which is a granary for food storage.

Daniell's painting of a Tswana Homestead. Source:

It is not clear from the image if more than one family stayed in this homestead, but we know that this is a common practice in Tswana culture and society. The homestead could be seen as an extension of the idea of the shelter, and rather than having beehive huts being under one canopy as in Ga-Mohana, in the case of my people they would be in one enclosure or container. In Daniell’s painting the homestead is enclosed by a grass wall, but we also know that at Dithakong this was stone instead of grass, perhaps a newer innovation. And like Ga-Mohana and Wonderwerk cave, Tswana settlements would be near water, either near rivers or waterholes.

(As a side note I should say that it is a pity that the only visual representation we have of these dwellings are from colonial representations, which obviously have their own issues, however they are generally understood to represent factual information particularly when looked at next to contemporary dwellings made by similar communities.)

Key to these two examples is that the shelter allows or is allowed by an outdoor living. The climatic conditions in South Africa for the most part, make outdoor living possible. The shelter also makes it possible to have a much more ephemeral and lightweight, ‘under insulated’ architecture, meaning that the membrane between outside and in, the cell wall, is very thin. In the case of those living in huts under the shelter of a rock canopy, not much fuss would be made regarding insulating against wind and rain. Similarly, the shelter would create a place where we are all under one roof. Rather than being isolated and separated in individual dwellings that require large amounts of insulation (as in individual houses), the shelter itself is the ‘house’, and the huts are the rooms. The spaces in between would be the living spaces, the in-between spaces where the family, and the extended family meet. This one could say is also evident in the Tswana homestead, and here we have even more evidence of the social manifestation of the shelter. In fact in both of Daniells paintings above, the dwelling in both the case of the Korana and the Tswana are under the shelter of trees. The latter has an additional physical shelter seen in the cone protecting the beehive. However, the canopy that covers the whole family is now the enclosure of the homestead. The same principle of the outdoor living applies here, and the hut becomes the place for sleeping, although one could argue that the veranda remains the indoor/outdoor space of shelter. This is evident in Daniell’s painting where one can see a low wall between the posts providing a place to sit perhaps when the rain does not allow one to be in the open. This kind of outdoor communal living was mediated by many social and cultural norms but was nevertheless a place where extended family members lived together. Perhaps places that had need for more intense insulation may have led to the development of less communal spaces and perhaps the emergence of ever more individualised family clusters as in the western nucleus family.

Exploring the shelter

Therefore, one of the key findings, but by no means the only, that I came to after all the years of studying Kuruman was the notion of the shelter. By shelter I am referring to a simple structure, a roof supported by posts with no walls. Perhaps another word to describe this is a canopy. My first architectural encounter with the idea of the shelter came when I was doing my explorations in my master’s dissertation. This at the time was not a conscious attempt at exploring the shelter, but regardless this is precisely what I was doing. The design I proposed was a centre for oral history and cultural development, and it was situated on a piece of land in Kuruman where my community first lived and were subsequently forcibly removed in the 60’s. The land was a buffer zone between a coloured township and the main CBD which was in those days part of South Africa and predominantly white. My family and others were moved to what later became part of Bophuthatswana, a native reserve and labour camp.

Kuruman center for oral history and cultural development: Dr Sechaba Maape

In the dissertation I did a series of sections where I explored a shelter or canopy structure which housed outdoor activities. These were meant to be what I called informal story telling spaces under the shelter, partly indoor partly outdoor. My supervisor at the time advised me to turn this canopy into a shape that was reminiscent of an actual tree like structure which provided shade, a very poetic move, although retrospectively I see would have not served as a shelter from the rain. After my completion of the dissertation, I worked on another project with architect Mpheti Morojele where the shelter featured as part of the design. This was a cultural centre in Limpopo, and again in this design we explored the shelter. The aim here was again to create sheltered outdoor spaces under which various activities would occur. In this design some spaces were sheltered from the sun, some both the sun and the rain.

Explorations of the shelter: Dr Sechaba Maape

Tree canopy as a derivative of the shelter: Dr Sechaba Maape

In 2016-2017 I was approached by a client in my hometown of Kuruman to design a building that would help her generate income. She had virtually no money other than the little pension money she cashed in to follow her dreams. Due to the very low budget we needed to develop a strategy that would be both cost effective as well as simple to construct. This led to a strategy of building columns made from stock brick that would hold the roof up, and these columns would sit on a slab and foundation.

The Pallet House: Dr Sechaba Maape

Between the columns was where the design became interesting, we procured timber pallets from a nearby solar plant who was giving away these pallets for free as part of their way of giving to the local community. Therefore, the columns and the roof remained conventional construction, and the walls were novel. In a follow up building, we applied the same method but this time around the infill was tyres and locally manufactured bricks.

The Tyre House: Dr Sechaba Maape

Both these buildings were off grid, they were on the same property which has a borehole, and both used solar PV for their electricity needs. Although these buildings had walls to infill between the columns, they brought in a key element to the notion of the shelter, namely being the notion of self-sustaining.

The evolution of the idea

I have seen other architects use the shelter as a concept, and perhaps this has something to do with the possibility of shelter living due to our climate in South Africa. I have seen similar ideas in the work of practices such as MMA architects, Earth Works, but probably best known in the work of Kerearchitecture by world famous architect Francis Kere.

School designed by Andre Eksteen with Joubert, Kammeyer and De Villiers architects. Source

Kigutu International Academy in Burundi by MMA architects. Source:

National Park of Mali by Francis Kere. Source:

These are spaces sheltered by a single roof under which a variety of open and enclosed spaces make up an indoor-outdoor environment. In a recent project where Afreetekture collaborated with Lemon Pebble architects, we revisited the shelter, and in this case, we drew inspiration from the Ga-Mohana shelter, as well as the beehive under cone on post. The brief was to create a traditional medicine facility for the late Gogo Mbai Mbai. Her house in Burgersfort was where she was practicing her healing and traditional medicine production. She had built back yard rooms and had a central courtyard reminiscent of a homestead. After visiting the Gogo and seeing her house and the way she set up her healing facility, we found inspiration for the new design.

Concept Sketch of Traditional Medicine Facility: Drawing by Dr Sechaba Maape

The building was located near a river, and the idea was to use the river water to irrigate the new medicinal plants Gogo was growing. The site was in an area that has no bulk infrastructure, therefore we needed to make provision for the collection of water and possible solar PV installation. As a result, we proposed a large flat roof on concrete posts, with a series of earth pods under the roof. There were four pods under the roof made of clay. The idea here was to design relatively in expensive simple structures, which could be ephemeral and not needing large amounts of bonding agents. The roof would protect the pods underneath from rain and weathering. Further away was a fifth pod. This pod is meant for overnight healing, the idea here being that the pod would disintegrate and weather, and part of the healing process would be that the afflicted person, depending on their ailment, will see the pod as themselves, and overtime they will repair the pod as a process of mending their self.

Axo of Traditional medicine facility

3D visualization of Traditional Medicine Facility

Plan of Traditional Medicine Facility

The main facility would sit on a plinth to protect from flooding and lift the pods from moisture. Therefore, the pods, like the Pallet House and Tyre House, would be sandwiched between two protective layers, the roof, and the floor. This would allow for the soft parts of the building to be soft, including the bodies of the occupants, protected by an exoskeleton. These are the outdoor spaces of the Tswana and other culture’s homestead enclosures, including the low walls for seating and social interaction. The solar PV and water collection would mean that the building would be plugged into the system of the cosmos in general, like the dwellings of the Korana and the Tswana, linked to the rivers and the sun. The solar PV would mean that the building is in direct contact with its power source, and the water collection, along with the river meant that the building is in right relation, or perhaps original relation in the sense that it aligns with human environment relations conducive for human habitation.


Le-Corbusier’s Dom-Ino House or Marc-Antoine Laugier’s rustic hut are ideas that have influenced Western Architecture in a profound way, but perhaps these ideas are echoes of the shelter; the basic structure above one’s head, the protective canopy that allows for human social life to occur. I often tell my students that we need to go back to our own versions of the acropolis, like what these architects did when they found inspiration for some of the ideas that continue to haunt us today. I believe that the idea of the shelter is still in its infancy, and that if we look carefully at it we may generate new profound insights about architecture that have the potential to decentralize the West as the original generator of architectural ideas.

Copyrights: The images in this blog produced by the author are subject to copyright laws and are not to be used without permission from the author.

633 views0 comments
bottom of page